Académique Documents
Professionnel Documents
Culture Documents
BP EXPLORATION
1995 British Petroleum Company PLC Text originated by BP Drilling Department Manual produced by ODL Publications, Aberdeen, Tel (01224) 637171
WELCOME
Click here to zoom in on text, then click on text to scroll through
Ladies and Gentlemen: Following is the Second Edition of the BP Well Control Manual first issued in 1987. When issued it was expected to be a living document, accounting for changes in technology and experience, it still is. Now, eight years later, horizontal and extended reach wells, coil tubing drilling and under balance drilling have or will become part of our kit for improved profitability. Our objective with this Second Edition is to bring three changes to the operating groups: 1) 2) 3) Issue the manual in an electronic version as a pilot which may lead to collecting all of the manuals on a server or CD-ROM. Make available Excel based well control worksheets which have been incorporated into the manual. Modify parts of Volume I Chapters 1 and 6 for high angle and horizontal well operations.
In a separate file we have issued the HTHP Well Control Manual. Future updates will tie this manual with the BP Well Control Manual. Publication of the manual in electronic format should make the abundance of information in it more accessible to you. A powerful search capability and hot button references are part of the software package we have selected. Software used is compatible with Macintosh, MS-DOS and DEC hardware platforms making it accessible to BP and our contractors when needed. Electronic publishing makes modifications easier and we solicit your suggestions for correction, clarification, change or addition to the manual. If we have not managed to make the resource more useful and clear to you we have failed our objective. Your views on how well we have done are important. To open and use the manual please read the section below. While use of the electronic version of the manual is encouraged there is still the option of printing a hard copy of the manual. Hard copies can still be obtained from ODL in Aberdeen at a cost for printing and shipping. Originally this manual was not issued as policy. In the October 1994 Drilling Managers Meeting this and two other documents, the Drilling Policy Manual and Casing Design Manual, were designated as the three core policy documents covering our operations. Every effort has been made in this edition to tie to the other two documents.
March 1995
HOW TO USE
This manual has been converted into Adobe Acrobat software and is a read only version, ie you cannot make any changes to text or figures, you can copy the text and figures and paste them in to another application.
To go back or forward to a previous move you have made, use the Acrobat arrows in the Menu Bar.
The header at the top of each page has been hot spotted, to return you to the Main Contents page of the Volume you have selected.
Once you have reached the section you require (e.g. 1.1 General), the hand cursor will appear with an arrow inside it. Press the mouse button on the section you require to read, and you will be zoomed into the section, press it again and it will scroll through that section, at the end of the section it will reset to the beginning of the section.
Excel Worksheets
Each example of a Worksheet in the manual is linked to a blank Excel Template for you to use for your own calculations, just click on the example Worksheet and Excel will automatically open. To return to the manual, simply Quit out of Excel.
Printing
When printing to a US Letter size printer please click on the Shrink to Fit box in the Print dialogue box. Printing of Excel Worksheets is through Excel.
Manual Contents
March 1995
Volume 1 Contents
Nomenclature Abbreviations 1 PREPARATION
Section 1.1 1.2 1.3 1.4 1.5 INSTRUMENTATION AND CONTROL MANPOWER ORGANISATION DRILLS AND SLOW CIRCULATING RATES USE OF THE MUD SYSTEM KICK TOLERANCE Page 1-1 1-9 1-15 1-27 1-35
March 1995
6.2
6.3
March 1995
NOMENCLATURE
SYMBOL A a An b c C Cp Ca CL CR D Dshoe Dwp dbit dh dhc do di dcut dc F Fsh FPG g G DESCRIPTION Cross sectional area Constant Total nozzle area Constant Constant Annular capacity Pipe capacity Cuttings concentration Clinging constant Closing ratio Depth Shoe depth Depth of openhole weak point Bit diameter Hole diameter Hole/casing ID Pipe OD Pipe ID Average cuttings diameter Drilling exponent (corrected) Force Shale formation factor Formation Pressure Gradient Gravity acceleration Pressure gradient UNIT in.2 in.2 bbl/m bbl/m % m m m in. in. in. in. in. in. lb SG psi/ft psi/m SG psi/ft m m m sec/m m m/hr psi lb SG
Gi H Hi Hp ITT K L MR M m MW
Influx gradient Height Height of influx Height of plug Interval Transit Time Bulk modulus of elasticity Length Rotary exponent Migration rate Matrix stress Threshold bit weight Mud weight
March 1995
SYMBOL N OPG P P Pa Pbit Pcl Pdp Pf Pfrac Pfc Pi Pic Plo Pmax Poc Pp Pscr PV Q Qmud Qgas Re R Rw ROP S Sg Sw t TR T TD TVD V
UNIT rpm SG psi/SG (The units of subsurface pressure may be either psi or SG) psi psi psi psi psi psi/SG psi/SG psi psi psi psi/SG psi/SG psi psi/SG psi cP gal/min gal/min gal/min ohm-m ohm-m m/hr meq/100g psi Fractional Fractional seconds min degrees C, F, R m m bbl
Adjustment pressure Annulus pressure Bit pressure drop Choke line pressure loss Drillpipe pressure Formation pressure Fracture pressure Final circulating pressure Hydrostatic pressure of influx Initial circulating pressure Leak off pressure Maximum allowable pressure at the openhole weak point Wide open choke pressure Pore pressure Slow circulating rate pressure Plastic Viscosity Flowrate Mud flowrate Gas flowrate Reynolds number Resistivity Resistivity of water Rate of Penetration Shale factor Overburden pressure Gas saturation Water saturation Time Transport Ratio Temperature Total Depth True Vertical Depth Kick tolerance
March 1995
SYMBOL V
DESCRIPTION Volume
UNIT
v vmud vp vs W
bbl cc ml l Velocity m/min m/s Mud velocity m/min Average pipe running speed m/min Slip velocity m/min Weight gm kg lb Weight lb/ft lb/bbl SG Weight of pipe lb/ft Baryte required for weighting up lb/bbl Average cuttings weight SG Weight on Bit lb Offset () Yield Point lb/100ft2 Compressibility factor Viscosity cP Poissonss Ratio Maximum effective principle stress psi/SG Tectonic stress psi/SG Porosity Fractional Fann reading lb/100ft2 Tectonic stress coefficient Density SG Bulk density SG
March 1995
ABBREVIATIONS
API RP BHA BOP BRT DWT ECD EMW H2S IADC ID KTOL LCM LMRP LO MAASP OBM OD PMS PV ROP SCR SG SPM YP American Petroleum Institute Recommended Practice Bottomhole Assembly Blowout Preventer Below Rotary Table Dead Weight Tester Equivalent Circulating Density Equivalent Mud Weight Hydrogen Sulphide International Association of Drilling Contractors Internal Diameter Kick Tolerance Lost Circulation Material Lower Marine Riser Package Leak off Maximum Allowable Annular Surface Pressure Oil Base Mud Outside Diameter Preventive Maintenance System Plastic Viscosity Rate of Penetration Slow Circulating Rate Specific Gravity Strokes per Minute Yield Point
March 1995
1 PREPARATION
Section 1.1 INSTRUMENTATION AND CONTROL 1.2 MANPOWER ORGANISATION 1.3 DRILLS AND SLOW CIRCULATING RATES 1.4 USE OF THE MUD SYSTEM 1.5 KICK TOLERANCE Page 1-1 1-9 1-15 1-27 1-35
March 1995
Page
1-2 1-2 1-4 1-6
Illustrations
1.1 1.2 1.3 Suggested Instrumentation for a Floating Rig Suggested Instrumentation for a Fixed Installation Suggested Fluid Measurement System
1-1
March 1995
1 General
It is essential that an appropriate level of control equipment is provided on every rig in order that a well that is under pressure can be accurately monitored. In general, during a well control incident, there is a necessity for more accurate instrumentation than under conditions encountered during routine drilling. The level of instrumentation on every rig therefore must be evaluated in order to assess its s uitability for well control purposes. This evaluation should ideally be carried out inconjunction with the pre contract rig audit and any deficiencies made good prior to contractaward. The purpose of this section is to highlight the important aspects of instrumentation and control and to recommend a standard level of equipment for all rig types. The level of instrumentation that is recommended will ensure that a suitable level of control is afforded during unusually critical operations, and that adequate back-up is provided. Therefore, much of this equipment would not be necessary in routine circumstances. However equipment failure is most likely when the equipment is highly stressed. It is in these situations that serious incidents can develop if a suitable level of back-up instrumentation and control equipment is not to hand.
2 Pressure Gauges
When a well is under pressure it is important that accurate pressure measurements can be made. Each rig will normally be equipped with gauges to read standpipe pressure and annulus pressure. The gauges that are fitted to the choke panel and at the drillers console are often the only gauges available for well control purposes. Although the standpipe and choke manifold will generally be fitted with Cameron gauges, these are considered to be so inaccurate as to have little application to well control. All of these gauges will have a fullscale deflection that is at least equal to the working pressure rating of the equipment. In all cases, this means that it will be necessary to install gauges of lower rating in order that relatively low pressures can be accurately recorded. This will be especially important with high pressure equipment. It is also important that suitable pressure gauges are installed at the choke manifold in case the well has to be controlled from this position. This will apply to land rigs which may be equipped only with manual chokes and the majority of rigs that are equipped with both manual and remote operated chokes. Accurate readout of pump pressure and choke pressure is, in the majority of cases, all that is required. However an extra pressure reading is required on a floating rig in order that the wellhead pressure can be monitored through the kill line. In order to be able to install additional pressure gauges it may be necessary to fabricate manifolds and install high pressure instrument hose between the choke panel and the standpipe/choke manifold. All this equipment must be rated to the working pressure of theequipment.
1-2
March 1995
STANDPIPE 1
CAMERON GAUGE
STANDPIPE 2
K C
CHECK VALVE
STANDPIPE MANIFOLD
HYDRAULIC FLUID INLET
CAMERON GAUGE
D K C
SW AC O
CHOKE PANEL
KILL LINE
CHOKE MANIFOLD
MANUAL CHOKES
BUFFER TANK DRAIN FROM BOP OVERBOARD LINE POORBOY DEGASSER FLOWLINE
CHOKE LINE
D DRILL PIPE K KILL LINE C CHOKE LINE 1/4in NEEDLE VALVES CHECK VALVE/HYDRAULIC FLUID INLET
WEOX02.001
1-3
March 1995
So in general: There must be gauges available to read choke pressure, standpipe pressure and kill line static pressure in the case of a floating rig. The above gauges must be readable from the manifold if manual chokes are fitted to the manifold. It must be possible to easily install and remove low range pressure gauges at the choke panel and at the choke manifold.
Suggested pressure recording systems for a floating rig and a fixed installation are shown in Figures 1.1 and 1.2. The proposed systems can also be used for measuring slow circulating rate pressures (SCRs). The following points should be noted from the proposed systems: A good selection of gauges should be available. Gauges should be calibrated on a regular basis with a Dead Weight Tester. It is suggested that the gauges are checked at each BOP Test and at this stage the pressure monitors in the mud logging unit should be checked against the rig equipment. It must be easy to change the gauges. A hydraulic fluid hand pump should be available to purge the lines at suitable points as shown. Consideration should be given to completely isolating the supplementary pressure monitoring system from that originally fitted to the rig. This would ensure that the original system was closed and hence in no way susceptible to leaking needle valves or misuse of the supplementary system. Sensitive low pressure rated gauges should be removed from the system unless required. The piping and manifolding should be permanently installed. It would be a good idea to fabricate a cover for the manifolding at the choke manifold and choke panel. The gauges that are used to measure the slow circulating rate pressures should be used to monitor well pressures in the event a kick is taken. A stroke counter, similar to the battery operated Swaco unit, is recommended for remote installation at the choke manifold. It should be removed when not required. A suitably isolated terminal should be located at a convenient point at the choke manifold, in order that the signal from the limit switches on the pumps can be transmitted to the counter.
3 Pump Control
It is desirable that the remote control of the pump used to kill a well that is under pressure is located reasonably close to the choke operator. In most cases the rig pumps will be used. Generally, the Driller will control these pumps from a position that is close to the choke panel. Most choke panels contain a meter that displays the cumulative output of the pump. Therefore, in the majority of cases, if the well is controlled with a remote operated choke, the man on the pump will be able to co-ordinate with the choke operator.
1-4
March 1995
TO STANDPIPE
TO STANDPIPE
STANDPIPE MANIFOLD
D C
CHOKE PANEL
D
SW
AC
TO BURN PIT
CHOKE MANIFOLD
CAMERON GAUGE
C CHOKE LINE D DRILL PIPE 1/4in NEEDLE VALVES CHECK VALVE/HYDRAULIC FLUID INLET
WEOX02.002
1-5
March 1995
However, if the choke manifold contains manual chokes, the choke operator may be some considerable distance from the man on the pump and a monitor of the pump output. In such cases, it is recommended that a remote pump output meter is positioned at the choke manifold. This will be especially important on land rigs which may be equipped only with manual chokes and where often the choke manifold is located at some distance from the rig floor. A further complication may arise if a kill pump or cement pump is used during a well control operation. It may become necessary to use these pumps on any rig, but the use of a relatively small displacement pump will be standard well control procedure on a floating rig that is drilling in deep water. Therefore, on a floating rig, it is desirable that it is possible to control and monitor the kill/cement pump from the rig floor.
4 Fluid Measurement
During stripping operations, as well as during a volumetric kill, it is important to beable to accurately measure small volumes of fluid bled from, or pumped into thewell. API RP 53 recommends that a trip tank or other method of accurately measuring the drilling fluid bled off, leaked from, or pumped into a well within an accuracy of half a barrel isrequired. Most rigs will not have suitable equipment to do this. It is usually assumed that the choke manifold lined up across a manual choke to the trip tankis a suitable fluid measurement system. However , in most cases this will not be a satisfactory arrangement because of the relatively large volume in the line between the choke and the tank. In general, there is a requirement for a line from the well, terminating at a manual choke positioned directly above a measuring cylinder, such as the trip tank (hydraulically activated chokes are not suitable for this application). However a bleed line from the well to the mixing tanks on the cement/kill pump may be sufficient. The most satisfactory arrangement is to use a strip tank as shown in Figure 1.3. This tank would typically have a 3 to 4 bbl capacity so that very small volumes of fluid can be measured. After bleeding into the strip tank, the tank contents can be emptied into the trip tank where the total volume of mud bled from the well, together with the mud leaked past the preventers, can be measured. Although it is not ideal, it may be sufficient to use a Lo-Torq valve instead of amanual choke to bleed fluid to the tank. However, during a long operation this is likely to wash out and so provision should be made to easily and quickly replace the valve. It is not recommended to bleed mud into a measuring tank that is situated in a confined area when there is a possibility that gas is entrained in the mud.
1-6
March 1995
PRESSURE GAUGE
MANUAL CHOKE
LEVEL INDICATOR
FLOWLINE RETURNS
TRIP TANK
WEOX02.003
1-7/8 1-7
March 1995
Page
1-10 1-10 1-12
Illustrations
1.4 An Example Communication System
1-13
1-9
March 1995
1 General
This section is intended to provide a guideline for the allocation of individual responsibilities during a well control incident. It is Company policy that a well control contingency plan should include the allocation of individual responsibilities. The contingency plan should be drawn up in conjunction with the drilling contractor and should be regularly reassessed. Well control drills provide an opportunity to assess the effectiveness of the contingency plan and to identify and make good any inadequacies.
2 Individual Responsibilities
The well control contingency plan must allocate the responsibilities of all those concerned in the operation. Circumstances at the rigsite may dictate that these responsibilities be modified in the event of an incident; however, the following can be used as guidelines for the allocation of responsibilities in the event of a well control incident:
1-10
March 1995
However, in the event that the well gets out of control, the Company Representative has the right to assume complete control and supervise the work required to regain full control of the well. (This entitlement is a standard condition of Company drilling contracts.)
It may be prudent to send an extra Mud Engineer to the rig in the event of a well control incident to ensure constant supervision of the mud system.
1-11
March 1995
3 Communication
One of the Company Representatives responsibilities is to organise a pre-kill meeting once the well has been shut-in. The purpose of this meeting is to ensure that all those involved in the supervision and implementation of the well control operation are familiar with the procedures that will be used to kill the well. This meeting is also the first stage in the process of communication during the well control operation. Experience has shown that even the most well conceived well control procedures can go badly wrong if communication before and during the operation is not properly organised and effective. It is therefore most important that the well control contingency plan details the method and line of communication for each individual involved in the operation. The objectives of a suitable system of communication are: To ensure that all information relevant to the well control operation is communicated to the Company Representative. To ensure that those involved in the supervision of the operation are at all times in communication with the Company Representative. To ensure that all those involved in the operation are aware of the line and method of communication that they should use. To ensure that communication equipment on the rig is adequate, and is used during the well control operation in the most effective manner possible.
Figure 1.4 shows an example of a possible communication system on a semi-submersiblerig for use during standard well control operations. The following can be noted from this example: After the kick is taken, the well is shut-in and closely monitored. The Company Representative calls a pre-kill meeting of those involved in the supervision of the operation. Responsibilities are allocated to those involved in the operation by the supervisors who attended the meeting. Each line and method of communication is defined. It should be noted that: The rig telephone system is not overloaded. The most important lines of communication to and from the Company Representative (denoted by those inside the broken line) are best maintained with the use of hand held radios. The use of intrinsically safe hand held radios ensures that all those inside the broken line can listen in on each others communication. Depending on the type of operation it may be necessary to include others within the broken line.
1-12
March 1995
(1) KICK TAKEN WELL SHUT-IN WELL BEING MONITORED (2) PREKILL MEETING
COMPANY REPRESENTATIVE COMPANY DRILLING ENGINEER SENIOR CONTRACTOR REPRESENTATIVE TOOLPUSHER MUD ENGINEER MUD LOGGING ENGINEER
TOOLPUSHER
SUBSEA ENGINEER
MATES
CONTRACTOR STAFF
DRILLER
CONTRACTOR SHOREBASE
DRILL CREW
PUMPMAN/ DERRICKMAN
DRILLER
CONTRACTOR SHOREBASE
MARINE STAFF
PUMPMAN/ DERRICKMAN
RT
S/S
MUD ENGINEER
RT
SUBSEA ENGINEER
COMPANY SHOREBASE
WEOX02.004
1-13/14 1-13
March 1995
Page
1-16 1-16 1-17 1-17 1-19 1-19 1-21 1-22 1-23
Illustrations
1.5 1.6 1.7 SCR Pressure Plot Choke Line Pressure Loss Data Sheet An example Determination of Choke Line Losses
1-15
March 1995
1 General
Both BOP Drills and the recording of slow circulating rate pressures will be carried out on a routine basis on all rigs. This section covers the reasons why it is necessary to carry out BOP Drills, to regularly record SCRs, as well as recommended procedures.
2 BOP Drills
The purpose of BOP Drills is to familiarise the drillcrews with techniques that will be implemented in the event of a kick. One of the major factors that influences the wellbore pressures after a kick is taken is the volume of the influx. The smaller the influx, the less severe will be the pressures during the well kill operation. In this respect, it is important that the drillcrew react quickly to any sign that an influx may have occurred and promptly execute the prescribed control procedure. Drills should be designed to reduce the time that the crew take to implement these procedures. The relevant Drills should be carried out as often as is necessary, and as hole conditions permit, until the Company Representative and the Contractor Toolpusher are satisfied that every member of the drillcrew is familiar with the entire operation. Every effort must be made to ensure that the Drill is carried out in the most realistic manner possible. Where practical, there should be no difference between the Drill and actual control procedures. Once satisfactory standards have been achieved, the Drills (D1, D2 and D3, as appropriate) should be held at least once per week. If standards fall unacceptably, the Company Representative should stipulate that the Drills are conducted more frequently. It is important that returning drillcrews have frequent Drills. The following Drills should be practised where applicable: D1 Tripping D2 Drilling D3 Diverter D4 Accumulator D5 Well Kill (Suffix R to be included if the remote panel was used) These codes should be used to record the results of the Drill on the BOP Drill Record Proforma. This form should be sent to the Drilling Superintendent fortnightly. The results of each Drill must also be recorded on the IADC Drilling Report.
1-16
March 1995
1-17
March 1995
When the pipe is on bottom, the following procedure can be used as a guideline for the drill: Without prior notice, the Company Representative gradually increases the apparent pit level by manually raising the float. The Driller is expected to detect the pit gain and take the following steps: 1. Pick up the kelly (or topdrive) until the tool joint clears the BOPs and the kelly cock is just above the rotary table. 2. Shut down the pumps. 3. Check the well for flow. 4. Report to the Company Representative. 5. Record the time required for the crew to react and conduct the Drill on the IADC drilling report. When the bit has been tripped to the previous casing shoe, a further Drill may be conducted that will result in the well being shut-in. Therefore after tripping the bit to the shoe, the following procedure may be used as a guideline for this Drill: Stop tripping operations and install the kelly (or topdrive) and start circulating. Having been instructed to do so by the Company Representative, the Driller is expected to take the following steps to shut-in the well: 1. Pull up until the tool joint clears the BOPs. 2. Shut down the pumps. 3. Open the choke line valve. 4. Close the annular preventer. 5. Record the casing and drillpipe pressure. * 6. Double check spaceout, close and lock hang-off rams and hang-off pipe and check that the kelly cock is just above the rotary table. 7. Notify the Company Representative that the well has been shut-in. 8. Record the time taken for the crew to shut-in the well on the IADC drilling report. * If on a floating rig The procedures adopted during these Drills should be in line with the shut-in procedures as outlined in the Standing Orders. These procedures are outlined in Chapter 4.
1-18
March 1995
1-19
March 1995
Accumulator Working Pressure Rating 1500 psi 2000 psi 3000 psi
Min. Acceptable Precharge Pressure 750 psi 950 psi 950 psi
Max. Acceptable Precharge Pressure 850 psi 1100 psi 1100 psi
1-20
March 1995
3. If the accumulator pumps are powered by air, isolate the rig air system from the pumps. A separate closing unit air storage tank should be used to power the pumps during this test. When a dual power (air and electric) source system is used, both power supplies should be tested separately. 4. Close the annular preventer and open one choke line failsafe valve (orHCR valve). Record the time (in seconds) required for the closing unit pumps to close the annular preventer plus open the choke line valve and obtain 200 psi above the accumulator precharge pressure on the closing unit manifold. It is recommended that the time required for the closing unit pumps to accomplish these operations does not exceed two minutes. 5. Close the choke line failsafe (or HCR valve) and open the annular preventer. Open the accumulator system to the closing unit and charge the accumulator system to its designed operating pressure using the pumps.
1-21
March 1995
All these factors must be taken into account when deciding at what rate to displace the kick. However the absolute upper limit for the displacement rate may be restricted by the pressure rating of the surface equipment, in particular the setting of the pump relief valve. It should be noted that it is potentially hazardous to displace a kick from the hole when the surface pressure is close to the relief valve setting. In order to estimate the circulating pressures during the displacement of a kick, it is necessary to know the friction pressure in the circulating system at low rates. For this reason, it is useful to have determined the SCR pressure before a kick is taken. At a given rate of circulation, the initial circulating pressure can be estimated from the sum of the shut-in drillpipe pressure and the SCR pressure. Company policy states that SCRs should be conducted regularly and at least: Once per tour (or at 300m intervals during the tour). When the bit is changed. When the BHA is changed. When the mud weight or properties are changed.
The range of circulation rates used will be dependent upon many factors, but should fall within the limits of 1/2 and 4 barrels per minute. If oil base mud is in the hole, when back on bottom after a trip, circulate bottoms up before measuring SCRs. At these relatively low pump speeds the volumetric efficiency of the rig pumps may be significantly less than at normal speeds used during drilling. It is therefore recommended that the volumetric efficiency of the rig pumps is checked at low pump speed, such as when pumping a slug prior to a trip. It is useful to plot the SCRs on a graph as shown in Figure 1.5. The drillstring internal friction should be calculated at the SCRs and used to determine the annulus frictional pressure as shown. The annulus frictional pressure is a major factor that will influence the rate at which the kick will be displaced from the hole (using standard well control procedure the annulus frictional pressure will be added to wellbore pressure as the pump is brought up to speed to kill the well).
1-22
March 1995
PSCR3
PSCR2
PSCR1
SCR1
SCR2
SCR3
WEOX02.005
1-23
March 1995
It is important that the choke line frictional pressure is accurately known at a wide range of circulating rates. From this information the additional load on the wellbore can be assessed at a range of displacement rates and subsequently the most suitable rate can be selected. The following procedure should be implemented in order to properly assess the choke line frictional pressures at slow circulating rates. This procedure should be carried out initially when the BOP and riser are installed and before drilling out of each subsequent casing shoe. 1. Install suitable pressure gauges to record standpipe and choke pressures during circulation. 2. Record SCR pressure at a range of rates from 1/2 to 4 bbl/min down drillpipe and up the riser. 3. Open choke line valves. 4. Line up choke manifold to route flow across a fully opened remote operated choke. Route returned flow through the poorboy gas separator to the shakers. 5. Space out to ensure no tool joint is opposite annular preventer. 6. Close annular preventer. 7. Circulate down the drillpipe and up through the choke line until returns are uniform. 8. Record SCR pressure at same rates as before. Record the choke pressure at each rate. 9. Calculate the choke line frictional pressure at each rate. Figure 1.6 shows a form that can be used to record the data. The form also shows how to determine the choke line friction pressure from the recorded data. Figure 1.7 shows an example determination of choke line losses. The choke line losses should be adjusted for changes in mud weight as shown on the form. The accuracy of this adjustment is however questionable over a wide range of mud weights. In order to verify choke line losses after drilling out of the casing shoe, it is acceptable to isolate the well and pump down the choke line at the range of slow circulating rates.
1-24
March 1995
WELL No
25
RIG
RIG 19
DATE
25/7/87 J. P.
CORRECTED CHOKE LINE LOSS AT MUD WEIGHT (psi) CORRECTED CHOKE LINE LOSS AT MUD WEIGHT (psi)
133/8in CASING RUN AND TESTED / 135/8in STACK INSTALLED AND TESTED 1.4SG OBM/PV24CP/YP100 lb/100ft2
SCR PRESSURE UP CHOKE LINE (psi) CHOKE PRESSURE AT SCR (psi) MEASURED CHOKE LINE LOSS RECORDED BY CORRECTED CHOKE LINE LOSS AT MUD WEIGHT (psi)
4.78
40 30 20
RIG PUMPS: NATIONAL 12 - P - 160 985 1435 80 680 400 985 590 55 40
1-25
March 1995
3.58 2.39
CEMENT PUMP - HT - 400 (4in PLUNGER) 1.00 0.5 0.25 120 50 0 190 65 0 25 0 0 45 10 0
(1)
(2)
(3)
(2)-(1)-(3)
WEOX02.006
400
600
50
POC = 50psi PCL = PSCR (up choke line) PSCR (up riser) POC = 600 400 50 PCL = 150psi
where
PSCR = Slow Circulating Rate Pressure (psi) PCL = Choke Line Pressure Loss at SCR (psi) POC = Choke Pressure recorded at SCR with choke wide open (psi)
WEOX02.007
1-26
March 1995
Page
1-28 1-28 1-29 1-31 1-34
Illustrations
1.8 An example Mud Gas Separator operating at maximum capacity 1-32
1-27
March 1995
1 General
Well control contingency plans should outline the manner in which the mud system will be utilised during standard well control operations. This section is intended to highlight the major factors that will determine the most satisfactory arrangement of the mud system in such circumstances.
2 Pit Management
The following guidelines should be considered when specifying pit arrangements:
1-28
March 1995
The kick can be displaced from the hole using either the Wait and Weight Method or the Drillers Method. The most satisfactory arrangement of the pits will be different for each technique and clearly will be rig-specific. There are three different stages at which the mud can be weighted up for these two techniques: The Wait and Weight Method In a typical situation when it is impractical to weight up a complete hole volume prior to displacement of the kick. This will therefore entail that some mud is weighted while the kick is displaced from the hole. The volume that is weighted prior to displacement of the kick will depend, for a given hole capacity, on the rate at which baryte can be added into the system in relation to the desired rate of displacement. In the unusual situation when there is adequate surface volume, a complete hole volume of kill mud can be prepared before displacement of the kick.
The Drillers Method In this case the mud is weighted either while the kick is displaced with original weight mud or after the first circulation depending on the availability of baryte and tank space.
The maximum rate at which the mud can be weighted can be determined for a given required mud weight increase from the following formula: Maximum possible rate = at which the mud can be weighted (bbl/min) Baryte delivery rate (lb/min) Baryte required to weight up (lb/bbl)
1-29
March 1995
Therefore for the following example: Required mud weight increase = 0.2 SG (from 1.5 SG to 1.7 SG) Baryte required = 1490 X (1.7 - 1.5) = 117 lb/bbl 4.25 - 1.7 If the maximum barytes delivery rate for the rig = 350 lb/min Then: Maximum rate at which the = 350 = 3 bbl/min mud can be weighted 117 This figure therefore gives an indication of the maximum displacement rate if the mud is weighted as the kick is displaced from the hole.
The limiting factor for an oil base mud may be the rate at which viscosity can be built into the base oil. Building viscosity is usually a less important factor when water base muds are used. Shear equipment is required for building viscosity using clay viscosifiers in new base oil. Some offshore rigs have jet line mixers to help build viscosity. In circumstances in which large volumes of new oil mud must be built, it would be useful to know the rate at which new mud can be sheared to a level at which barytes can be suspended. This rate is determined by shearing a known volume of new mud until the minimum viscosity is reached. As a guideline, the minimum viscosity would be represented by a yield point of 10, and a 10 second gel reading of 3. In emergency situations, viscosity can be built quickly using an oil mud polymer (Baroids LFR 2000 as an example) at 4 lb/bbl in conjunction with organophilic clays. However, it is recognised that these polymers can cause high temperature gelation of the mud, and as such, they are not recommended for use in high temperature wells.
1-30
March 1995
1-31
March 1995
The back pressure due to the flow of gas should be monitored with a pressure gauge as shown in Figure 1.8. Some warning of the possibility of a gas blow-through will be given when the registered pressure approaches the hydrostatic head of the fluid in the discharge line. It should be noted that the maximum hydrostatic head available may not be that of the mud in the event that large volumes of oil or condensate are displaced tosurface. If the safe operating limit of the separator is approached, the choke can be closed in (while ensuring that the well is not overpressured) or the flow switched to the overboard line or the burn pit.
GAS OUTLET 8in ID MINIMUM GAS BACK PRESSURE REGISTERED AT THIS GAUGE (Typically 0 to 20psi) STEEL TARGET PLATE INLET INSPECTION COVER SECTION A-A TANGENTIAL INLET
30in OD
INSPECTION COVER
TO SHAKER HEADER TANK MAXIMUM HEAD AVAILABLE DEVELOPED BY THIS HEIGHT OF FLUID eg: 10ft HEAD AT 1.75 SG GIVES 7.6psi MAXIMUM CAPACITY 10ft APPROX
WEOX02.008
1-32
March 1995
Lines that are required to handle high velocity gas must be as straight as possible to minimise erosion. Significant erosion is likely to occur in the path of high velocity gas and solids, therefore the redundancy in flowlines and manifolds downstream of the choke must be analysed on all rigs.
1-33
March 1995
5 Chemical Stocks
(a) Baryte and mud chemical stocks
Company policy details the minimum stocks of baryte and mud chemicals that should be held at the rigsite. The policy states that: Sufficient weighting material stocks must be maintained on site such that the entire mud circulating volume can be raised by a minimum of 0.25 SG (See formula in Paragraph 3 ). Reserve stocks of bentonite or viscosifier must also be on site to enable this increase in mud weight to be effected. Where transport and logistics are not assured (offshore and remote locations) the minimum onsite weighting material stock must be 100 tonnes. This is a minimum standard, and as such, the Company Representative may wish to stock a greater quantity of baryte and chemicals.
1-34
March 1995
Page
1-36 1-36 1-37 1-40 1-41 1-42
Illustrations
1.9 Kick Tolerance Values Through a Zone of Increasing Pore Pressure 1-43 1-44
1-35
Rev 1 March 1995
1 General
Many definitions of kick tolerance have been used in the drilling industry. Within BP, Kick Tolerance is defined as the maximum volume of kick influx that can be safely shut-in and circulated out of the well without breaking down the formation at the openhole weak point. It is now an accepted part of the Company Casing Design policy to determine the casing setting depth by the Limited Kick Method. It is therefore particularly important that the kick tolerance in critical hole sections be accurately monitored. This section explains how to calculate kick tolerance and when to calculate kick tolerance. In critical hole sections, it is important to calculate kick tolerance on a regular basis. This is because kick tolerance changes as a function of hole depth, BHA geometry, mud weight, formation pressure and influx type, etc.
Although these assumptions may seem unrealistic, the simple methods have gained wide acceptance in the drilling industry because they are simple and generally yield conservative (safer) kick tolerance. However these methods have an inherent shortcoming: they do not measure how quickly an influx will grow. This is to say that in some cases formation deliverability may be such that the well could not be shut in before the kick tolerance volume was exceeded. Therefore the same kick tolerance between two wells may not mean that they share the same level of risk ! 2 Computer Kick Simulators In the recent years many sophisticated computer simulators have been developed which can provide a good approximation of kick conditions from the stage when it flows into the wellbore to that when it is circulated out. In the simulations, assumptions used in the simple methods are replaced by mathematical models. Among many other applications, the kick simulators can be used for kick tolerance calculations. They can predict the maximum pressures at any point of the annulus and the results are more accurate and less conservative than using the simple methods. In addition, as simulators can simulate how quickly an influx will flow into the wellbore, they can predict how much time the rig crew have to shut in the well before the influx exceeds the kick tolerance limit. Therefore simulators can be used to provide direct indications in the level of risk involved under various scenarios.
1-36
March Rev 1 1995 March 1995
However, due to complexity, kick simulators are recommended only in the situations where kick tolerance is considered critical based on the simple methods. Some computer kick simulators are available from the Drilling & Completions Branch, BP Exploration, Sunbury.
The following procedure can be used to calculate the kick tolerance: 1 Estimate the safety factor to be applied to the Maximum Allowable Annular Surface Pressure (MAASP) When the influx is displaced from the hole, there will be additional pressures acting in the wellbore. The following are some of the possible causes of such additional pressures during circulation: Choke operator error (depending upon the chokes condition, operator s experience,etc.) Annular friction pressure (depending on the hole size, mud properties, etc.) Choke line losses (in particular on floating rigs)
The safety factor (SF) to be applied to the MAASP will be the sum of these additional pressures. The drilling engineer must use his/her judgement to determine the most appropriate safety factor. 2 Calculate the Maximum Allowable Annular Surface Pressure (MAASP) Without Breaking Down the Weak Point Formation: MAASP = Pleak 1.421 x MW x TVD wp SF (psi) where: MAASP MW Pleak SF TVDwp Maximum allowable annular surface pressure (psi) Mud weight in hole (SG) Leak-off pressure at the openhole weak point (psi) Safety factor (psi) Vertical depth at the openhole weak point (m)
It should be seen that MAASP is determined based on the consideration of the formation fracturing pressure at the openhole weak point. So it is considered only when there is a full mud column from the weak point to the surface (i.e. the influx is still below the weak point). If lighter fluids (such as a gas influx) occupy the annulus above the weak point, the surface pressure in excess of MAASP may not cause downhole failure. Therefore from the moment the top of an influx has been displaced past the openhole
1-37
Rev 1 March 1995
weak point, MAASP is no longer a consideration and may be exceeded by a margin which should be determined based on the casing burst strength and the pressure ratings of BOP stack and choke manifold. The method for estimating the position of the influx top is described in Vol.I, Chapter 6, Section 6.1. 3 Calculate the maximum allowable height of the influx in the openhole section:
H max =
where: Hmax Gi Pf TVD h 4
(m)
(MW Gi)
Maximum allowable height of the influx (m) Influx gradient (SG) Formation pore pressure (psi) Vertical depth of openhole (bit) (m)
Calculate the maximum allowable influx volume that Hmax corresponds to at the initial shut-in conditions Vbh = H max where: Vbh C1 bh
x
C1 / cos( bh)
(bbl)
Maximum allowable influx volume at initial shut-in condition (bbl) Annular capacity around BHA (bbl/m) Hole angle in the bottom hole section (degree)
If the bottom hole section is horizontal (or above 90 degree), the hole angle used in the calculation should be the openhole angle immediately above the horizontal section. The kick tolerance should be the sum of the calculated volume (Vbh ) plus the annular volume of the horizontal section. In cases where Hmax /cos( bh ) is greater than the length of BHA, the maximum allowable volume (Vbh) should be calculated partly based on the annular capacity around BHA and partly around drillpipe. 5 Calculate the maximum allowable influx volume that Hmax corresponds to when the top of the influx is at the openhole weak point Vwp = Hmax where: Vwp C2 wp
x
C2 / cos( wp)
(bbl)
Maximum allowable influx volume when top of the influx is at the openhole weak point (bbl) Annular openhole capacity around drillpipe (bbl/m) Hole angle in the openhole section below the weak point (degree)
In cases where Hmax /cos( wp ) is greater than the openhole drillpipe length below the weak point, the maximum allowable influx volume (Vwp) should be calculated partly based on the annular openhole capacity around drillpipe and partly around BHA.
1-38
March 1995 1995 Rev 1 March
Convert the maximum allowable influx volume at the weak point (Vwp ) to what would be at the initial shut in condition Based on Boyles law, the maximum allowable influx volume at initial shut-in corresponding to Vwp will be:
V bh' = V wp X Pleak Pf
7
(bbl)
The actual kick tolerance should be the smaller of Vbh (Step 4) and Vbh' (Step 6) Example: Bit depth: Current hole size: Hole angle: Mud weight in hole: BHA length / OD: Drillpipe OD: Estimated pore pressure at 4000 m: Last casing shoe: Leak-off test EMW: Annular back pressure at SCR: Safety margin for choke operator error: 4000 m 12-1/4" Vertical 1.60 SG 182 m / 8" 5" 1.58 SG 2695 m 1.72 SG 70 psi 150 psi
1. Estimate the safety margin to be applied to MAASP: SF = 70 + 150 = 220 psi 2. Calculate MAASP: Leak-off pressure, Pleak = 1.421 x 1.72 MAASP = 6587 - 1.421
x x
3. Calculate the maximum allowable influx height in the openhole section: Pore pressure gradient, P f = 1.421 x 1.58
x
H max =
= 178m
4. Calculate the maximum allowable influx volume at the initial shut-in condition: Annular capacity around BHA, C1= (12.252 - 8 2) / 313.8 = 0.2743 (bbl/m) As the BHA length (182 m) is longer than Hmax (178 m), so the influx is around BHA only when it is at the bottom of the hole. Therefore: Vbh = 178 x 0.2743 = 49 bbl 5. Calculate the maximum allowable influx volume when the top of influx is at the casing shoe: Annular capacity around openhole DP, C2= (12.252 - 5 2) / 313.8 = 0.3985 (bbl/m) Openhole DP length = 4000 - 2695 - 182 = 1123 m ( > H max of 178 m)
1-39
Rev 1 March 1995
Vwp = 178 x 0.3985 = 71 bbl 6. Convert Vwp to the initial shut-in condition: Vbh' = 71 x 6587 / 8981 = 52 bbl 7. Therefore the actual kick tolerance is 49 bbl.
H max =
where: Gi Pf SF TVD h
Influx gradient (SG) Formation pore pressure (psi) Safety factor mainly determined by the choke operator error margin (psi) Vertical depth of openhole (bit) (m)
1-40
March 1995 1995 Rev 1 March
d Calculate the influx volume that Hmax corresponds to when the gas influx top has reached the surface: Vsurf = Hmax x Ccsn where: Vsurf Ccsn Maximum allowable influx volume when the influx top reaches surface (bbl) Annular capacity in the casing near surface (bbl/m) (bbl)
V 2 = V surf X
f
Psurf Pf
(bbl)
The actual kick tolerance volume is the smaller of V2 (step e) and V 1 (stepa).
1-41
Rev 1 March 1995
Figure 1.9 shows an example of the type of calculations that should be worked. The kick tolerance figures shown are those that would typically be calculated before a transition zone. As shown, the current bit depth is 3500 m and the kick tolerance has been calculated at various intervals across the zone of increasing pore pressure. The kick tolerance has been calculated for the mud weight currently in use, for the maximum mud weight anticipated for the section, and intermediate weight. From these figures, it is clear that a serious situation would develop if a kick was taken from the high pressure zone with the mud weight currently in the hole. This might occur if either the pore pressure developed more rapidly than predicted, or if the steady increase in pore pressure was undetected at the surface. The kick tolerance figures for the intermediate mud weight show that even at this weight, the kick tolerance would be small if the high pressure zone was unexpectedly encountered. The kick tolerance is finally calculated at the maximum mud weight. These figures show a final minimum kick tolerance of 50 bbl at that mud weight. The table also shows the kick tolerance if the pore pressure developed higher than predicted of 1.6 SG. In general these figures indicate that drilling should proceed cautiously through the zone of increasing pore pressure. On the basis of these figures, it may be decided to weight up the mud a certain amount before the predicted increase in pressure occurs. The decisions that are made on the basis of kick tolerance figures such as these will be largely dependent upon the particulars of each situation, including the level of confidence placed in the pore pressure prediction.
1-42
Rev 1 1995 March 1995 March
Figure 1.9 Kick Tolerance Values though a Zone of increasing Pore Pressure
PORE PRESSURE (psi) 3000 4000 5000 6000 7000 8000 9000
9000
9.2ppg
11,000
13,000
12 12 12 12 12 12 12 12
WEOX02.009
1-43
Rev 1 March 1995
Well:
Example Calculation
Units: (UK/US):
UK
Input Messages:
Non-Horizontal
Other Parameters:
Bottom Hole Assembly OD ? Bottom Hole Assembly Length ? Drillpipe OD ? Gas Hydrostatic Pres Gradient ? Pressure Safety Factor ? Mud Weight in Hole ? Annular Capacity Around BHA: Annular Capacity Around DP: Fracturing Pres at Weak Point: Max Allowable Shut-in Csg Pres: Min Pore Pressure at Kick Zone: Maximum Allowable Gas Height: Kick Tolerance at Min Pore Pres: Max Pore Pressure at Kick Zone: Maximum Allowable Gas Height: Kick Tolerance at Max Pore Pres: 1.60
Pore Pressure Gradient
Comments:
8981 178 48.7 9094 120 33.0
1.60 1.59 1.59 1.58 1.58 1.57 33 41 Kick Tolerance (bbl) 49 33 41 49 1.60 1.59 1.58
For more infor or help, please contact YUEJIN LUO, BP Exploration, Sunbury, Tel: 853-2424, Fax: 853-4183
1-44
March 1995 1995 Rev 1 March
Figure 1.10 Example Calculations using Excel Kick Tolerance Calculator (cont'd)
APPENDIX:
Maximum Allowable Gas Influx Volume Based on Surface Equipment Rating & Casing Burst
At Minimum Pore Pressure Gradient: Maximum Allowable Gas Height When Gas Arrives at Surface: Max Allowable Gas Vol. on Shut-in: At Maximum Pore Pressure Gradient: Maximum Allowable Gas Height When Gas Arrives at Surface: Max Allowable Gas Vol. on Shut-in:
1.60
Comments:
(m) (bbl) 2460 613
(m) (bbl)
2403 547
1.60 1.59 1.59 1.58 540 550 560 570 580 590 600 610 620 547 580 613 1.60 1.59 1.58
1-45
Rev 1 March 1995
Formation pressures are contained by the hydrostatic pressure of a column of drilling fluid this is primary well control. If primary control is lost the blowout preventers are closed and secondary well control techniques are used to kill the well. Primary control is maintained by ensuring that a full column of drilling fluid of an appropriate weight is allowed to exert its full hydrostatic pressure in the hole.
Industry wide experience has shown that the most common causes of loss of primary control and hance the well kicks are: Swabbing during trips. Not adequately filling the hole during a trip. Insufficient mud weight. Lost circulation.
The evidence also shows that the majority of kicks have occurred during trips.
This chapter outlines the measures that are required to eliminate or minimise the risk of a kick due to the above causes, and to minimise influx volumes if a kick occurs.
March 1995
2.1
1 2 3 4
Paragraph
Illustrations
2.1 2.2 2.3 Typical Trip Tank Hook-up on a floating rig BP Trip Sheet example of a completed sheet Example of Standing Orders for Driller
2-1
March 1995
1 General
Industry wide experience has shown that the majority of well control problems have occurred during trips. It is therefore particularly important that special attention is paid to ensuring correct tripping procedure. During tripping the potential exists for a significant reduction in bottomhole pressure due to the following effects: Reduction in ECD as the pumps are stopped. Swab pressures due to pipe motion. Reduction in height of the mud column as pipe is removed from the well.
The procedures required to deal with an influx when the pipe is off bottom are not so straightforward as when the pipe is on bottom. Every effort must therefore be made to ensure both that the well is stable prior to initiating a trip out of the hole, and that correct tripping procedure is strictly adhered to.
2 Prior to Tripping
Considerable preparation is required before the trip is commenced. The following are among the most important actions that should be carried out prior to tripping: Circulate the hole The mud should be conditioned to ensure that tripping will not cause excessive swab/ surge pressures. Any entrained gas or cuttings should be circulated out. The mud weight should be such as to ensure an adequate overbalance will exist at all times during the trip. Determine the maximum pipe speed Swab/surge pressures should be calculated at various tripping speeds using the appropriate formulae. (See Chapter 3, Volume 2. ) The maximum average pipe speed should be selected bearing in mind the estimated overbalance or trip margin. Line up the trip tank Company policy states that: A trip tank must be available on every rig and be complete with a mechanically operated indicator of the trip tank level, visible from the Drillers position. The trip tank level must also be monitored from the Mud Loggers cabin.
2-2
March 1995
RIG FLOOR
ROTARY TABLE
DIVERTER
RETURNS TO SHAKERS
FLOWLINE
2-3
March 1995
TRIP SHEET
WELL No
26 3250m 5 5
in in
RIG
RIG 20
CHANGE BIT No 20
INITIAL BIT DEPTH
DRILLER
SHEET No
3250m
: : : : :
: : : : :
91/2 in
in in
Doubles
Stands
53 STANDS
STAND STAND
No Increment
increment (bbl)
accum (bbl)
1 2 3 5 7 10 15 20 25
1 1 1 2 2 3 5 5 5
30.5 30.0 29.4 28.6 27.2 25.9 23.8 20.1 16.6 13.2
(1)
(2)
(3)
(4)
(1)-(3)
(2)-(4)
WEOX02.011
2-4
March 1995
Figure 2.2 shows a completed example of the BP trip sheet. This trip sheet should be used if the contractor cannot provide a similar sheet. The basic requirement for a trip sheet is that a clear method of comparing calculated with actual hole fill volumes is provided. The cumulative discrepancy between the two values should also be recorded. The trip sheet for the last trip out of the hole should be available for comparison. Provide the Driller with the necessary information The Driller should be told the reason for the trip. He should be told of any indicators of increasing pore pressure or near balance that were identified during drilling before, or since, he came on shift. He should be fully aware of the procedures to be adopted in the event of a kick while tripping. An example of the standing orders that should be provided to the Driller is shown in Figure 2.3. Drill floor preparation Crossovers should be available on the rig floor to allow a full opening drillpipe safety valve to be made up to each tubular connection that is in the hole. A drillpipe safety valve (kelly valve) should be available on the rig floor. It should be kept in the open position. A back-up safety valve, such as a Gray valve, should be available close to the rig floor. This valve should only be used in the event that the drillpipe safety valve does not hold pressure, or if stripping in the hole is required and no dart sub is fitted. The rig crew should be completely familiar with, and practiced in, their responsibilities in the event of a kick.
3 Tripping Procedure
Having completed the preparations as outlined in the previous section, the trip out of the hole can be started. The following procedure is proposed as a guideline: 1. Flow check the well with the pumps off to ensure that the well is stable with the ECD (equivalent circulating density) effect removed. 2. Pump a slug. This enables the pipe to be pulled dry and the hole to be accurately monitored during a trip.
2-5
March 1995
WELL NO
15 ON ALL TRIPS
COMPANY REP
RIG
RIG 12 SMB
TOOLPUSHER
15/6/87
K.D.
IF ANY OF THE FOLLOWING OCCUR: 1. HOLE NOT TAKING CORRECT VOLUME DURING THE TRIP 2. THE WELL IS FLOWING 3. 4. 5. 6. 7. 8. Or if there is any other possible indication of a kick.
YES
NO
1. SET THE SLIPS 2. INSTALL OPEN DP SAFETY VALVE 3. CLOSE DP SAFETY VALVE 4. OPEN CHOKE LINE VALVE (S) 5. CLOSE ANNULAR PREVENTER 6. CHECK THAT WELL IS SHUT IN 7. NOTIFY COMPANY REPRESENTATIVE 8. INSTALL KELLY 9. LINE UP STANDPIPE MANIFOLD 10. OPEN DP SAFETY VALVE 11. RECORD DP AND CSG PRESSURE 12. IF IN OPENHOLE: ENGAGE BUSHINGS, ROTATE THE PIPE 13. PROCEED AS DIRECTED
WEOX02.012
2-6
March 1995
The following formula can be used to calculate the volume of slug to ensure a length, L, of dry pipe: Vsl = MW X L X Cp (bbl) (MWsl MW) where Vsl L Cp MWsl MW = = = = = volume of slug (bbl) length of dry pipe (m) internal capacity of the pipe (bbl/m) slug weight (SG) mud weight in the hole (SG)
As a general rule, the slug should be mixed to maintain a minimum of 2 stands of dry pipe. It is important to accurately displace the slug to the pipe. In this manner, the Driller will know the weight, depth and height of the slug at all times during the trip. 3. For the first 5 10 stands off bottom, monitor the hole through the rotary. This is to check that the annulus is falling as pipe is removed from the hole. The pipe wiper should therefore be installed only after the first stands have been pulled. The trip tank should not be overfilled at this stage to ensure that swabbing is clearly indicated, should it occur. The circulating pump should be switched off at this stage and the hole filled from the trip tank, after each stand. 4. Circulate the hole across the trip tank and continue to trip out, monitoring hole volumes with the aid of the trip sheet. 5. Conduct a flowcheck when the BHA is into the casing shoe. 6. Conduct a flowcheck prior to pulling the BHA through the stack. Be aware that the required hole fill volume per stand of heavy weight and drill collars will be greater than for drillpipe as the BHA is being removed from the hole. If unsure of the overbalance, consideration should be given to conducting a short round trip. Once back on bottom, the overbalance can be assessed from the level of the trip gas at bottoms up. If the hole does not take the correct amount of fluid at any stage in the trip, a flowcheck should be carried out. If the flowcheck indicates no flow and the cause of the discrepancy cannot be accounted for at surface, the string should be returned to bottom while paying particular attention to displacement volumes. After circulating bottoms up, it may be necessary to increase the mud weight before restarting the trip out of the hole. If the flowcheck is positive, the well should be shut-in according to the procedure indicated in the standing orders. Subsequent action will be dependent upon the conditions at the rigsite (See Chapter 5).
2-7
March 1995
In these circumstances the following procedure is recommended prior to pulling out of the hole: 1. Flow check the well. 2. Circulate bottoms up. 3. Check trip to the shoe monitoring hole volumes. 4. Flow check at the shoe and run back to bottom. 5. Circulate bottoms up. Close in the BOP and circulate through the choke when the potential influx is at 500m below the stack, watching for any pit gain. 6. If necessary increase the mud weight and perform a further check trip. This procedure can be relaxed if, after several trips under the same conditions, the well remains stable. The following procedure is recommended in these circumstances after a round trip. 1. When back on bottom prior to any further drilling or coring, circulate bottoms up to check for trip gas. 2. Circulate until potential influx is at 500m below the stack, watching for any pit gain. 3. Close in the well and circulate the potential influx through the choke.
2-8
March 1995
2.2
Paragraph
Illustrations
2.4 Bottomhole Pressure Reduction due to gas cutting 2-12
2-9
March 1995
1 General
Primary well control is achieved by controlling formation pressures with the hydrostatic pressure of the drilling fluid. The drilling fluid may be contaminated with cuttings and formation fluids during drilling. These contaminants can significantly alter the effective hydrostatic pressure exerted by the drilling fluid, and in certain circumstances, this can cause loss of primary control. Hydrostatic pressure will be reduced once drilling stops as a result of the loss of annulus frictional pressure and the removal of cuttings from the annulus. The settling of cuttings to the bottom of the hole may significantly reduce the hydrostatic pressure further up the hole. This section outlines the techniques that can be used to predict the effect of drilling fluid contamination on the hydrostatic pressure.
2 Gas Cutting
When drilling through a formation that contains gas, it is inevitable that the mud will become contaminated with gas from the drilled formation even if the formation is penetrated overbalance. Drilled gas will enter the mud system at a rate determined by the following factors: Rate of penetration, ROP (m/hr) Hole diameter, d h (in.) Formation porosity, (fractional) Gas saturation, Sg (fractional)
The rate of gas entering the mud at bottomhole conditions, Qgas (gal/min), is given by the following formula:
Qgas =
dh 24
1.285 X ROP X X Sg
(gal/min)
Therefore as an example in the following conditions: ROP dh Sg Bottomhole pressure Hole depth and depth at which gas enters the mud, D = = = = = 25 m/hr 12 1/4 in. 0.2 0.75 6000psi
= 3020m
Qgas = 12.25 24
2-10
March 1995
Therefore at atmospheric pressure the gas flowrate is given by: Qgas = 1.26 X 6000 = 514 gal/min at atmospheric pressure 14.7 This simplified calculation treats the gas as ideal and does not consider the effects of temperature. In this hole section the flowrate of mud is 700 gal/min; the actual mud weight at surface can be calculated using the following formula:
MW act = MW X Qmud Qmud + Qgas
= = = =
actual mud weight at surface (SG) uncut mud weight (SG) flowrate of mud (gal/min) flowrate of gas (gal/min)
Therefore in this case the actual (or gas cut) mud weight at surface is given by:
MW act = 1.4
It should be stressed that this figure is an estimation of the actual mud weight at the flowline and as such will not reflect the actual density of the mud in the hole. The percentage gas cutting is given by: Percentage cut = MW MWact X 100 MW Which in this case gives a figure of: Percentage cut = 1.4 0.81 1.4
X
The following formula can be used to estimate the bottomhole pressure reduction due to gas cut mud: P = 14.7 (MW MW act) ln (96.46 X MW X D) MW act 1000 (psi)
where P = bottomhole pressure reduction due to gas cutting (psi) D = depth at which gas enters the mud (m) Figure 2.4 shows the effect of various levels of gas cutting for two different mud weights using the above formula. It should be noted that these curves represent an ideal gas; temperature and solubility effects are not considered. In this case: P = 14.7 (1.4 0.81) ln (96.46 X 1.4 X 3020) 0.81 1000 P = 64psi (psi)
2-11
March 1995
0 0
20
40
60
80
100
-1000
-2000
-4000
2.1 SG 1.05 SG
WEOX02.013
2-12
March 1995
Therefore the average mud weight in the hole is equal to: MW = (6000 64) = 1.38 SG 3020 X 1.421 It can be seen that what appeared to be significant gas cutting, at 42%, caused a very small reduction in the bottomhole pressure and actually only reduced the effective mud weight by 0.02 SG, or by a factor of 1.4%. The actual reduction in bottomhole pressure is caused by the gas when it has considerably expanded. This expansion does not occur until the gas has been circulated to near the surface. As can be seen from the previous example, this near surface expansion has a small effect on the bottomhole pressure in a deep well for moderate levels of gas cutting. However the effect of near surface expansion may be critical in relatively shallow hole sections. The effect of gas cutting in a relatively shallow hole is demonstrated with the following example: dh Instantaneous ROP D MW = = = = 24 in. 80 m/hr 300m 1.13 SG Sg Pump output Formation pressure = = = = 0.3 0.7 750 gal/min 1.03 SG
1.285
80
0.3
0.7
= 21.6 gal/min at bottomhole conditions Gas flowrate at surface is given by: 21.6
X
1.03
The actual mud weight at surface is given by: 750 750 + 645
X
1.13 = 0.61 SG
The average mud weight in the hole is given by: (1.13 X 1.421 X 300) 44 = 1.02 SG 300 X 1.421 Quite clearly the potential exists for the well to kick in this situation, given that the pore pressure at this depth is assumed to be normal at 1.03 SG.
2-13
March 1995
Industry experience has shown that excessive gas cutting in shallow hole has in many cases been the cause of shallow gas blowouts in offshore environments. The previous example shows the possible effect of gas cutting in shallow hole. However it should also be noted that in shallow hole the annulus pressure loss during circulation will be negligible, and the expansion of the gas may cause mud to be unloaded at surface, thereby further reducing the hydrostatic head of the mud column. It is therefore important that the ROP is restricted in shallow hole. High pump output should also be maintained to disperse the gas within the mud to minimise variations in SG.
3 Cuttings Contamination
One of the most important functions of the drilling fluid is to transport cuttings from the bit to the surface. The presence of cuttings in the annulus will increase the effective hydrostatic pressure of the fluid column. If this increase is excessive, it can cause losses which may possibly lead to the loss of primary control. It is therefore useful to be able to estimate the additional pressure caused by the cuttings in the annulus. In order to be able to estimate this additional pressure, a measure of the ability of the drilling fluid to remove the cuttings from the well is required. The cuttings slip velocity is defined as the velocity of the cuttings relative to the velocity of the mud. There are many factors that influence the cuttings slip velocity, however the following relationship can be used to estimate its value: Slip Velocity, v s = 108 where vs MW wcut dcut = = = = =
X
0.667
slip velocity (m/min) average viscosity (cP) mud weight (SG) average cuttings weight (SG) cutting average diameter (in.)
However, if the particle Reynolds number is greater than 2000, the following formula should be used to calculate the slip velocity:
vs = 34.56
1 2
MW
d cut
2-14
March 1995
The transport ratio is defined as the ratio of the actual cuttings velocity to the mud velocity; it is therefore determined as follows: Transport ratio, TR = v m vs vm where vm = Q 0.134(d hc 2 do 2) (m/min)
and vm = mud velocity (m/min) Q = pump output (gal/min) dhc = hole/casing ID (in.) do = pipe OD (in.) The concentration of cuttings in the annulus can be calculated from the following formula: Ca = ROP X d bit2 X (1 ) 448.4 X Q X TR where ROP = rate of penetration (m/hr) dbit = diameter of the bit (in.) = porosity The extra pressure caused by the cuttings in the annulus is given by the formula: P = (w cut MW) X 1.421
X
sum (L
Ca)
where L = the length of each section (m) The cuttings concentration must therefore be determined for each section of hole. Consider the following example for a 17 1/2 in. hole section drilled from a floating rig. Casing shoe at 900m Casing ID = 22 in. Riser ID = 22 in. Bit size = 17.5 in. Drillpipe OD = 5 in. Collar OD/length = 8 in./180m Mud weight = 1.5 SG The slip velocity Average viscosity Pump output ROP Openhole length Cuttings density Cuttings diameter = 50 cP = 700 gal/min = 50 m/hr = 180 m = 2.5 SG = 0.3 in.
The velocity of the mud in 17 1/2 in. hole is given by: Velocity = 0.134
X
= 21.6 m/min
= 11.4 m/min
2-15
March 1995
This gives a transport ratio of 64% in 17 1/2 in. hole and of 32% in 22 in. hole. The cuttings concentration, Ca, in the 17 1/2 in. hole is given by: Ca = 50 X 17.5 2 448.4 X 700 X 0.64 = 0.076 (= 7.6%)
In the 22 in. hole section: Ca = 50 X 17.5 2 448.4 X 700 X 0.32 = 0.152 (= 15.2%)
The porosity is not considered. The additional hydrostatic pressure due to the cuttings is determined as follows: P = (2.5 - 1.5) X 1.421 X [(0.076 X 180) + (0.152 X 900)] P = 214 psi This additional pressure therefore increases the effective mud weight to a figure givenby: MW = (1.5 X 1.421 X 1080) + 214 1080 X 1.421 = 1.64 SG
2-16
March 1995
2.3
1 2 3 4 5 6 7
Paragraph
Illustrations
2.5 2.6 Balanced Plug Technique Lost Circulation Remedies
2-22 2-24
2-17
March 1995
1 General
Lost circulation can occur as a result of the following: Unconsolidated or highly permeable low pressure formations (including depleted reservoirs and at the base of long permeable reservoirs). Natural fractures. Induced fractures. Cavernous formations.
Lost circulation is undesirable primarily for three reasons. Firstly, that a loss of hydrostatic head may lead to the well kicking and secondly, that the cost of the replacement mud required may be considerable. Thirdly, it precludes accurate monitoring of the hole. This section is intended to outline how to identify the different types of loss zone and, in each case, to determine the most appropriate remedy.
2-18
March 1995
2-19
March 1995
Natural fractures Can occur in many rock types. May cause a gradual loss of mud to the hole, however if drilling proceeds and more fractures are exposed, complete losses may be experienced.
Induced fractures Horizontal fractures may be induced at relatively shallow depths after the formation of mud rings and by overloading the annulus. The formation of a mud ring will be indicated by an increase in pump pressure and the drillstring becoming tight. Vertical fractures may occur at greater depth and may be caused by any pressure surge on the formation. Usually indicated by sudden and complete losses.
Cavernous formations Normally only experienced in limestone formations. Loss of returns may be sudden and complete. May be accompanied by the bit dropping up to several feet depending on the height of the cavern.
Underground blowout Condition where the act of shutting in on a kick induces a fracture in the openhole above the point of influx. Kick fluids flow, usually from the lower active zone to the zone which has been fractured. Generally indicated by unstable pressure readings at surface. The depth of the loss zone must be established in order to calculate the hydrostatics involved and to determine the remedial action required. The loss zone can be located using a Temperature Survey, which operates by identifying a discontinuity in the temperature gradient within the wellbore. A noise log may also be used. Correlation with the known lithology at the confirmed loss zone is very important to identify the type of formation that has been fractured.
2-20
March 1995
The best displacement method for placing plugs is to use the balanced plug technique. This is however not always possible to achieve or desirable, depending on the rate of loss or the type of slurry to be pumped. In general, placement techniques will be as follows (refer to Paragraph 6 for recipes):
2-21
March 1995
TUBING
MUD
SPACER
where h = height of spacer (m) H = height of plug (m) L = drillpipe/tubing length (m)
PLUG WEOX02.014
2-22
March 1995
Calculate the height of the cement plug before the pipe is pulled out H(m) = Volume of slurry(bbl) annulus cap (bbl/m) + pipe cap (bbl/m)
where H = height of the plug (m) Calculate the plug displacement volume Displacement volume (bbl) = (L H h) where L = Drillpipe/tubing length (m)
X
2-23
March 1995
LOST CIRCULATION REMEDIES SEVERITY OF LOSS, bbl/hr 1 10 EFFECTIVE IN LOST CIRCULATION REMEDIAL TECHNIQUE Technique 2A Plug of fine bridging agents in mud Technique 3A High-filter-loss slurry squeeze with fine bridging agents Technique 1 Pull up and wait (primarily for induced vertical fracture) Technique 2B Plug of medium bridging agents in mud Technique 3A High-filter-loss slurry squeeze with coarse bridging agents Technique 3B or 3C High-filter-loss slurry squeeze with coarse bridging agents Technique 4B Thixotropic cement or other cements (4A, 4C, 4D) Technique 5B Mud + diesel-oil-bentonite plus cement Technique 5A Downhole-mixed soft plug (mud-diesel oil-bentonite) Technique 7B Downhole-mixed hard plug (sodium silicate, calcium chloride, cement squeeze Flo-Check) Technique 3A, 3B or 3C High-filter-loss slurry squeeze with 25 35 lb/bbl or coarse bridging agents Technique 5B Downhole-mixed soft/hard plug continuously mixed in large amounts
LOSS ZONE GEOMETRY to horizontal loss zones** to induced vert fractures to horizontal loss zones** to induced vert fractures
WBM yes yes yes yes yes yes yes yes yes yes
Partial
10 500
Complete
500 complete
Complete
yes
no
Complete
Technique 1 Pull up and wait Technique 5B Downhole-mixed soft/hard plug Technique 5A Downhole-mixed soft plug Technique 7B Downhole-mixed hard plug (sodium, silicate, calcium chloride, cement squeeze, Flo-Check) Technique 3B or 3C High-filter-loss slurry squeeze with coarse bridging agents Technique 4A Neat portland cement Technique 7B Downhole-mixed plug (sodium, silicate, calcium chloride, cement squeeze, Flo-Check)
partial no no yes
Usually not in use where loss zones are horizontal. They consist of porous sands and gravels, natural fractures, and honeycomb and caverns.
2-24
March 1995
Technique 3B High filter loss slurry squeeze As Technique 3A but include the following: 15 30 lb/bbl medium and coarse LCM
Technique 3C High filter loss slurry squeeze As Technique 3A but include the following: Reduce Diearth concentration to 10 25 lb/bbl Use barytes as inert filler at 300 lb/bbl Add cement at 70 lb/bbl Place in 30 bbl slugs into loss zone with 200 psi squeeze pressure. Note: Wherever possible, slurry formulations should be tested prior to spotting to eliminate possible premature setting. When this is the case, always be aware of the thickening time and avoid leaving cement in or opposite the pipe beyond this time.
Technique 4A Neat cement slurry Neat cement slurries give high compressive strength plugs. Mix Class G cement at 1.90 SG in water
Technique 4B Extended cement slurry (using bentonite) Prehydrated bentonite slurry gives a degree of fluid loss control and plating effect to help stop lost circulation. Coupled with this, a lightweight slurry can be formulated (1.58 SG) which helps in areas of serious lost circulation. A further benefit is that reasonable compressive strength characteristics are found with slurries of this type. Add 10 lb/bbl bentonite to pre-treated fresh water (with 0.25 lb/bbl caustic and 0.25/lb/ bbl soda ash). Mix cement up to 1.58 SG.
Technique 4C Aggregated cement slurry (with sand or ground coal) Add aggregrate to the neat cement slurry at 1.90 SG up to a maximum weight of 20 35 lb/sack of cement in the mix.
2-25
March 1995
Technique 4D Thixotropic cements Cement of this type exhibits good flow characteristics when being pumped and a quickly developing gel strength when stationary. This thixotropic behaviour is beneficial for the following reasons: A plug of cement displaced past the loss zone is self supporting and does not fall back under its own weight. The cement will tend to remain next to the wellbore when squeezed into fractures due to their rapidly developing gel strength. Due to the temperature and chemical formulation sensitivity of this type of slurry, it is not recommended to use this cement without rigorous quality control and testing prior to each job. Halliburton Thixset 1 or 2 are examples of this type of cement.
Technique 5A Downhole mixed soft plug This type of lost circulation pill is designed to mix with a water base mud or formation water in the downhole environment and subsequently be squeezed into the formation. Mix 10.5 gal of diesel or base oil to 100 lb of bentonite. Granular or fibrous LCM may be added to this mix if required, ie mica at 10 ppb plus walnut at 10 ppb. This mixture must be kept away from contact with water until it is placed out of the drillpipe. To do this, a 10 bbl oil spacer is pumped ahead of a plug, followed by 10 bbl after the plug. The principle of this plug is to form a rubbery plug whenever the mixture contacts the water base mud. Formation water will assist the hydration of the bentonite.
Technique 5B Downhole mixed soft/hard plug This type of lost circulation pill is designed to mix with a water base mud or formation water in the downhole environment. It can be designed to form an initially fluid mixture of a soft or semi-hard nature depending on its composition, and can be squeezed into the formation where it will harden and develop compressive strength. The proportion of mud to the pill in the final mix downhole will determine the strength of the plug. For example, a 1:1 mix with fresh water will result in a soft plug, whereas a 1:3 (water/mix ratio) mix will result in a hard plug. In every case however, pilot tests should be carried out at surface for various mixes, prior to spotting the pill. Mix on surface 300 lb of G neat cement and 158 lb of bentonite to 1 bbl of diesel or base oil. All water should be excluded from the mix on surface.
2-26
March 1995
Technique 6 Downhole mixed soft plug Oleophilic clay and water This type of plug formulation is designed for use in an oil base mud. It works by the same principle as 5A, except that the clay disperses in water and hydrates in oil (the opposite of a bentonite squeeze). Mix on surface 280 lb of oleophilic clay to 1 bbl of water. Add lignosulphonate at 4/lb/ bbl water. An example of oleophilic clay is Baroid Geltone. The spacers ahead and behind this plug must be water based.
Technique 7A Surface mixed soft plug (polymer type) These formulations are mixed on surface, where polymers are blended with activators and extenders, to give a delayed thickening reaction. This allows enough time to place the plug in the loss zone before the chemical reaction takes place. Haliburton Temblok is an example of this type of material. This treatment is only temporary and the yield strength breaks down fairly quickly. It should be followed by a cement slurry to effect a permanent seal.
Technique 7B Downhole mixed hard plug Haliburton Flocheck can be used for this. This is a Sodium Silicate material which on contact with calcium ions forms insoluble Calcium Silicate. By pumping a CaCl 2 brine to the formation, followed by the Flocheck material, plugging of the formation occurs when the two chemicals mix in the formation matrix. Placement as follows: Pump 50 bbl 10% (by weight) CaC12 followed by 10 bbl fresh water. Then pump 35/bblof Flocheck followed by a further 10 bbl fresh water. Care must be taken to ensure that CaC1 2 does not come into contact with Flocheck on surface as it will go hard in the pits. This treatment, whilst permanent, may be reinforced by a cement slurry.
7 Drilling Blind
In certain circumstances it may become necessary to drill ahead without any returns at surface, ie drilling ahead blind. This may be required if all attempts as laid out in Paragraph 6 have failed. Once the decision to drill blind has been made, the main objective will be to set casing in the first competent formation penetrated.
2-27
March 1995
Although no cuttings will be obtained while drilling blind, casing seat can be located by logging and by keeping up a penetration log whilst drilling ahead. The hole has to be logged frequently, for example every 100m or whenever the penetration rate suggests a formation change. Once a competent formation has been identified, the new formation has to be penetrated by at least 20m to successfully set and cement the next casing string. Whilst drilling blind the following precautions must be taken: Use one pump for drilling as normal with the other continuously filling the annulus with water. Assign personnel to monitor the flowline for returns at all times. Pick the drillstring up off bottom every 2m drilled to ensure that the hole is not packing off above the bit. Keep one pit full of viscous mud at all times ready to pump to the hole. If one pump requires repair, use the cement unit to fill the annulus continuously. After drilling each single, wipe the hole over a full single and kelly length prior to drilling ahead. Wipe the hole over the length of a stand if using a topdrive. If overpull is experienced wipe the hole 3 or 4 times. Spot a viscous pill around the bit prior to making each connection. This pill should be balanced in and outside the pipe. If, during drilling, the fluid in the annulus reaches surface, stop drilling immediately. Pick up the drillstring so that the BOPs can be closed if required. Stop the pump on the drillpipe and the annulus. Close in and observe for any pressure build up. If there is no pressure on the annulus, start up the pump on the drillpipe and circulate bottoms up through a fully opened choke (if this is possible). The loss zone may be plugged with drill cuttings. Drill ahead if everything is normal to a predetermined depth, if the area is well known. Stop and log if the area is not well known to determine if a suitable casing seat has been found and has been sufficiently penetrated. If there is pressure on the annulus be prepared to adopt procedures for an undergroundblowout.
2-28
March 1995
Stop the pumps and monitor the well whenever the bit is pulled into the previous casing shoe. Be prepared to shut in at all times during the trip.
If wireline logging is required when complete loss of returns exists then the following precautions must be taken. When logging, the pump should be kept continuously on the hole. The only exception is when static fluid level has to be established. Logging is best conducted using through drillpipe logging tools, with open ended drillpipe run to the casing shoe.
2-29/30 2-29
March 1995
Page
3-2 3-2 3-2 3-3 3-4 3-4 3-6 3-6
3-1
March 1995
1 General
When drilling with returns to surface, a kick cannot occur without any warning sign. This Chapter outlines and explains the signs that indicate either that a kick has occurred or that a kick may soon develop.
2 Drilling Break
One of the first indications that a kick may occur is an increase in penetration rate, or a drilling break. Many factors influence the rate of penetration, but an increase in penetration rate can be caused by an increase in formation porosity, permeability or pore pressure. A change in all or one of these formation parameters may create the conditions in which a kick could occur. For this reason any drilling break should be checked for flow. Even if the flowcheck indicates no flow, the reason for each drilling break should be determined. As an example, a drilling break could be caused by drilling into an impermeable transition zone above a permeable reservoir. Because the formation is impermeable, it is unlikely that any significant flow would be noticed during a flowcheck. However, the formation may be considerably underbalanced by the mud column. If drilling continued and the reservoir was penetrated, a kick would be taken. Consideration must therefore be given to circulating bottoms up before drilling ahead after a negative flowcheck, especially in critical sections of the well.
3-2
March 1995
4 Pit Gain
(a) While Drilling
A gain in pit volume, that was not caused by the movement of mud stocks at surface, is confirmation that a kick is occurring or has occurred. This is the most reliable indicator of a kick. Consequently, every effort must be made to ensure that pit levels are accurately monitored at all times. Very small influx volumes may not be detected at surface as they occur. This may be due to the fact that, either the initial influx was particularly small, or the influx flowrate was very slow. This could be the case if the formation has low permeability or if a more permeable formation was only very slightly underbalanced. In such cases, the influx may be detected before it is circulated to the surface if it expands significantly as it rises up the hole. In general, the greater the amount of gas that is contained in the influx, the greater the expansion of the influx will be as it rises up the hole. As a result, the greater the proportion of gas in the influx, the more likely it is that the influx will be detected as it is circulated up the hole. Consequently, a low volume influx heavy oil or brine that does not contain any appreciable quantity of gas, will be relatively difficult to detect at surface. However, if the active system is accurately monitored, pit gains of less than 10 bbl should be detected reliably, even on floating rigs.
3-3
March 1995
3-4
March 1995
Gas can enter the mud for one or more of the following reasons: As a result of drilling a formation that contains gas even with a suitable overbalance. As a result of a temporary reduction in hydrostatic pressure caused by swabbing as pipe is moved in the hole. Due to the pore pressure in a formation being greater than the hydrostatic pressure of the mud column.
3-5
March 1995
A permeable formation that is only slightly underbalanced may only cause a small flow into the well. The first evidence of this at surface is likely to be gas cut mud, accompanied by a small pit gain. The initial pit gain may be so small that it is only detected as it expands as it is circulated up the hole. In the case a tight formation is underbalanced, there may be little or no actual flow of gas into the wellbore. Therefore, drilling such a formation may show only gas cut mud, even if the underbalance is relatively high. This is a relatively difficult situation to detect and is also potentially dangerous.
7 Increase in Hookload
If an influx occurs while drilling, an increase in hookload may be noticed at surface. Influx fluids will generally be lighter than the drilling fluid, especially so if the influx is gas. Displacement of the drilling fluid by the influx will reduce the buoyancy of the bottomhole assembly. This will increase the effective weight of the drillstring, a change that is likely to be registered as an increase in hookload. An increase in hookload may only be noticed after a considerable volume of influx has occurred. It is not therefore a reliable method of detecting a kick at an early stage.
3-6
March 1995
March 1995
4.1
1 2 3 4 5
Paragraph
4-1
March 1995
1 General
Offshore shallow gas accumulations are normally associated with recently laid down sand l enses that are totally enveloped by mudstones. When encountered at shallow depths,lenses tend to be highly porous, permeable and relatively unconsolidated. They arecommonly thin, flat and normally pressured. However , overpressured lenses have been encountered. Overpressure at this depth is generally caused by inclination of the lens which has the effect of increasing the height of the lens and hence the pore pressure gradient at the top of the lens. In some areas, shallow gas has been associated with buried reefs or vuggy limestone which can be extremely porous and almost infinitely permeable. Shallow gas kicks are generally caused by loss of hydrostatic head due to one or a combination of the following: Overloading the annulus with cuttings and hence causing losses. Drilled gas expanding and unloading the annulus. Improper hole fill while tripping.
Consequently it is strongly recommended to take the following general precautions to minimise the possibility of inducing a shallow gas flow: Drill pilot hole Drill riserless Restrict ROPs Accurately monitor the hole
Shallow gas flows are often extremely prolific, producing very high flow rates of gas and considerable quantities of rock from the formation; particularly so when a long section of sand has been exposed. In the event of a shallow gas flow, the Company Representative must immediately liaise with the Senior Contractor Representative to make preparations to evacuate initially non-essential personnel from the rig. The eventuality of having to completely evacuate the rig must also be addressed (the contractors emergency evacuation procedures will be implemented). A well should not be drilled through a shallow seismic anomally (bright spot), which may indicate the presence of shallow gas. If a bright spot is present at the proposed drilling location it is good practice to move the rig to avoid the hazard. The new drilling location should, if possible, be located on a shallow seismic shot point. It should be noted that the absence of bright spots does not rule out the possibility of the existence of shallow gas. Further to this, the absence of shallow gas in one well of a series drilled from a surface location does not guarantee the absence of shallow gas in subsequent directional wells drilled from the same surface location.
4-2
March 1995
The contingency plan must be discussed in detail at the pre-spud meeting. A gas blowout in open water produces a 10 degree cone of low density water and a discharge of highly flammable gas. The intensity of the blowout depends to a large extent on the water depth and current. The plume is likely to become more dispersed with greater water depth, whilst the effect of a current would be to displace the plume away from the rig. Within a plume of expanding gas, a floating vessel will suffer some loss of buoyancy; however, this diminishes rapidly with water depth such that the effect on a semi-submersible at operating draft would be negligible. The eruption of the gas would tend to displace a vessel, and if constrained by its moorings, might cause a drillship to keel towards the plume, thereby reducing its freeboard further. Under calm conditions, the gas cloud would disperse slowly and would constitute a fire hazard if the gas became entrapped in a confined area. The severity of the hazard can only be assessed at the time, and whilst there is unlikely to be an immediate danger to crew or vessel, the following precautions or considerations should be addressed before and whilst the surface hole is open: The rig should be moored with length of moorings remaining in the locker to allow the rig to be winched 400 ft away from the plume. If practical, the windlasses should be held on their brakes and the chain stoppers only applied after surface casing is set. All hatches should be secured to prevent invasion of voids by inflammable gas or downflooding if the freeboard is reduced by loss of buoyancy or heel. This is critical for a drillship. Facilities and personnel should be continuously available at short notice to slack off the moorings closest to the plume and heave in those up current (but not down wind). Before spudding, a contingency plan should be prepared detailing individual responsibilities and duties. Drill pilot hole, limiting the ROP and circulate at a high rate to distribute the cuttings and drilled gas. A float valve should always be run in the drillstring. Sufficient mud should be kept on site to fill the hole volume twice. (Typically at 1.15/SG.)
4-3
March 1995
Weather conditions and current should be continuously monitored and the sea surface should be checked for evidence of gas.
If a shallow gas flow is detected: If there is no immediate danger to personnel or the rig: 1. Attempt to control the well by pumping mud/seawater at a maximum rate. If the gas flow is endangering personnel or the rig: 2. Drop the drillstring or shear the pipe (See Section 6.2). 3. Winch the rig to a safe position outside the gas plume.
3 Gas encountered whilst drilling for surface casing from a Floating Rig with a riser
In relatively shallow offshore environments, the conductor is usually set in a formation that is too weak to contain the pressure of a gas kick. If a kick is detected in such circumstances, the well should be diverted in order to avoid an underground blowout and the possibility of the gas broaching around the conductor shoe. It is Company policy that where the situation demands that a riser is to be used when drilling for the surface casing, an annular preventer and subsea dump valves are installed at the mudline, in addition to the normal diverter system at surface. Industry experience has shown that current diverter systems cannot be relied upon to safely control shallow gas blowouts. As a result, shallow gas flows should be controlled at the seabed, using the subsea dump valves at the mudline and annular preventer. Immediate preparations should then be made to unlatch the pin connector or LMRP and winch off location, up current but not down wind. A contingency plan must be developed, prior to spud, in conjunction with the Drilling Contractor to cover the following situations: The procedures to be adopted in the event of a shallow gas flow. The procedure for winching the rig off location. The procedure to be adopted in the event of failure of any of the major components of the BOP/riser/diverter system.
The contingency plan must be discussed in detail at the pre-spud meeting. The surface diverter system ensures that there is a back-up system available in the event of a failure of the subsea system. It can also be used to divert gas which may be in the riser above the stack.
4-4
March 1995
The following precautions, in line with those listed in Paragraph 2, should be taken routinely whilst the surface hole is open: The rig should be moored with length of moorings remaining in the locker to allow the rig to be winched 400 ft away from the plume. If practical, the windlasses should be held on their brakes and the chain stoppers only applied after surface casing is set. All hatches should be secured to prevent invasion of voids by inflammable gas or downflooding if the freeboard is reduced by loss of buoyancy or heel. This is critical for a drillship. Facilities and personnel should be continuously available at short notice to slack off the moorings closest to the plume and heave in those up current (but not down wind). Before spudding, a contingency plan should be prepared detailing individual responsibilities and duties. Care should be taken to ensure that the annulus does not become overloaded with cuttings, causing losses or cuttings liberated gas, and hence the possibility of unloading the annulus. This is achieved by drilling a pilot hole, limiting the ROP and circulating at a high rate to distribute the cuttings and drilled gas. Facilities should be continuously available to fill the annulus rapidly from surface in the event of sudden losses. Care should be taken to monitor the hole and ensure that it remains full whilst tripping. A float valve should always be run in the drillstring. Sufficient mud should be kept onsite to fill the hole volume twice. (Typically 1.15/SG.)
Should the well start to flow, the following procedure can be used as a guideline: 1. Open the subsea dump valves. 2. Close the annular preventer and allow the gas to vent at the seabed. If there is no immediate danger to personnel or the rig: 3. Attempt to control the well by pumping sea water/mud at a maximum rate. If the gas flow is endangering personnel or the rig: 4. Consider dropping the drillstring or shearing prior to (5) (See Section 6.2). 5. Unlatch the LMRP or pin connector and winch the rig to a safe position outside the gas plume. In the event of failure of the subsea diverter system there remains the option to divert at surface or to unlatch the LMRP or pin connector, thereby venting the gas at the wellhead. Diverting at surface is not recommended, however if it becomes absolutely necessary to divert at surface, proceed as follows: 1. Maintain maximum pump rate. 2. Space out so that the lower kelly cock is just above the rotary table. 3. Open the diverter lines, close the shaker valve and diverter element thereby diverting returns overboard.
4-5
March 1995
4. Shut down all non-essential equipment and machinery to minimise potential sources of ignition. Deploy fire hoses beneath the rig floor . 5. Prepare to unlatch the pin connector or LMRP and winch to a safe position. If the situation is deteriorating and loss of control is imminent: 6. Consider dropping the drillstring or shearing the pipe prior to (7) (See Section 4.3). 7. Release the pin connector or LMRP and winch the rig to a safe position outside the gas plume.
4 Gas encountered whilst drilling for surface casing from a Bottom Supported Rig
Shallow gas reservoirs are potentially much more hazardous when penetrated from a jack-up or platform. Because the conductor extends almost to the rig floor, the products of a kick are discharged directly into a hazardous zone. In the event of a shallow gas flow, the diverter will immediately be closed in order to direct the flow overboard. The reliability of the diverter system while subject to the stress of a shallow gas flow is uncertain and so the possibility of equipment failure at this stage must be considered. On a bottom supported rig, a hazardous situation is created if a restriction forms in the diverter line. The subsequent pressure build up may cause gas to broach around the casing to the seabed. In this event there is a real risk that the seabed becomes fluidized, thus inducing a sudden reduction in spudcan resistance. The following precautions should be taken routinely whilst the surface hole is open: Care should be taken to ensure the annulus does not become overloaded with cuttings, thus causing losses or gas to be liberated from the cuttings to such an extent that the annulus unloads. This is achieved by drilling pilot hole, limiting the ROP, and circulating at a high rate to distribute the cuttings and drilled gas. Facilities should be continuously available to rapidly fill the annulus from surface in the event of sudden losses. Facilities should be available and care taken to monitor the hole and ensure that it remains full whilst tripping. A float valve should always be run in the drillstring. A means of diverting the flow away from hazardous zones, without restricting flow or imposing backpressure on the well, should be available for immediate activation. Sufficient mud should be kept onsite to fill the hole volume twice.
4-6
March 1995
Should the well start to flow, the following procedure may be used as a guideline: 1. Maintain maximum pump rate. 2. Space out such that the lower kelly cock is just above the rotary table. 3. Ensure that diverter lines are open, close shaker valve and diverter element thereby diverting returns overboard. 4. Shut down all non-essential equipment and machinery to minimise potential sources of ignition. Deploy fire hoses beneath the rig floor. 5. Evacuate all non-essential personnel. 6. Monitor the sea for evidence of gas breaking through outside the conductor. (Evacuate all personnel if any evidence is detected.)
4-7/8 4-7
March 1995
4.2
1 2 3
SHUT-IN PROCEDURES
Page
4-10 4-10 4-11
General Fast Shut-in Shut-in Procedure
Paragraph
Illustrations
4.1 4.2 4.3 Kick while Drilling, Floating Rig, Fast Shut-in Kick while Drilling, Fixed Rig, Fast Shut-in Kick while Tripping, Fast Shut-in
4-9
March 1995
1 General
It is Company policy that a well kick will be shut in and controlled at the BOP stack on hole sections below the surface casing. The procedures to be adopted in the event of a kick while drilling ahead from the surface casing shoe are drawn up at the discretion of the Company Representative and the Company Drilling Superintendent. There are various methods of shutting in a well that is flowing. In general, the best method is that which ensures that the well is safely shut in and the influx volume is minimised. The smaller the volume of influx, the lower will be the pressures in the wellbore and at surface throughout the kick control process. It is the responsibility of the Company Representative to ensure that the Contractor is made aware of the procedures that should be initiated in the event of a well kick.
The speed with which the Drillcrew carry out these procedures is a critical factor. In this respect, if a primary indicator of a kick, such as either a pit gain or an increase in returns flowrate is detected, no time should be spent flowchecking the well. In such circumstances, the kelly (or topdrive) should be picked up, the pumps stopped and the BOP closed immediately.
Speed and proficiency are achieved by regular drills. It is a further responsibility of the Company Representative that he ensures these drills are carried out at suitable intervals to ensure the drillcrews are proficient at implementing the shut-in procedures. The forms illustrated in Figures 4.1 to 4.3 should be used to make absolutely clear the shut-in procedures that will be used on each rig. These forms are intended primarily for the Driller, however copies should be distributed to other relevant personnel including the Toolpusher and, where appropriate, the Subsea Engineer. When a standard shut-in procedure is finalised, this procedure should be written on a large notice board that will be positioned prominently on the rig floor.
2 Fast Shut-in
Drilling management have issued the following guideline:
In the event that a kick is detected, or suspected, the choke line valve(s) are opened and the BOP closed.
4-10
March 1995
On a floating rig, the annular BOP will be used to initially shut-in the well. On a fixed rig, the pipe rams may be used to initially shut-in the well, in order to speed up the procedure, if the position of the tooljoint in relation to the pipe ram is known with confidence. The advantage of this method is quite clear, namely that the operation is relatively simple in comparison with the soft shut-in. Consequently, mistakes are unlikely and the time taken to close in the well will be minimised. At all times, be aware that the pressure rating of the standpipe equipment is generally less than that of the BOP stack and the choke manifold.
3 Shut-in Procedure
It is the responsibility of the Company Representative and the Company Drilling Superintendent to define the shut-in procedure that will be implemented in the event of akick. The following forms are examples of the information that should be provided to the Driller: Figure 4.1: Kick while Drilling, Floating Rig, Fast Shut-in. Figure 4.2: Kick while Drilling, Fixed Rig, Fast Shut-in. Figure 4.3: Kick while Tripping, Fast Shut-in.
4-11
March 1995
WELL NO
24
COMPANY REP
RIG
RIG 20 K.D.
TOOLPUSHER
10/3/87
IF ANY OF THE FOLLOWING OCCUR: 1. DRILLING BREAK *2. INCREASED RETURNS FLOWRATE *3. PIT GAIN 4. CHANGE IN PUMP SPEED OR PRESSURE 5. SUDDEN CHANGE IN PROPERTIES OF RETURNED MUD 6. 7. 8. 9. 10.
LOWER KELLY COCK IS 2.5m 1. PICK UP UNTIL ABOVE ROTARY UPPER PIPE (Space out to ensure that a tool joint is clear of rams) 2. SHUT DOWN THE PUMPS 3. FLOWCHECK THE WELL IF NECESSARY (Do not flowcheck if 2* or 3* as above have been detected.)
YES
NO
1. OPEN UPPER CHOKE LINE FAILSAFE (S) 2. CLOSE UPPER ANNULAR 3. CHECK WELL IS SHUT IN 4. NOTIFY COMPANY REPRESENTATIVE 5. CHECK SPACEOUT 6. CLOSE UPPER PIPE RAMS 7. ADJUST ANNULAR CLOSING PRESSURE 8. HANG OFF ON UPPER PIPE RAMS 9. CLOSE RAMLOCKS 10. PROCEED AS DIRECTED
WEOX02.015
4-12
March 1995
WELL NO
28 FOR WELL No 28
COMPANY REP
RIG
RIG 15 J.P.
TOOLPUSHER
15/9/87
J.B.H.
IF ANY OF THE FOLLOWING OCCUR: 1. *2. *3. 4. 5. 6. 7. 8. 9. 10. DRILLING BREAK INCREASED RETURNS FLOWRATE PIT GAIN CHANGE IN PUMP SPEED OR PRESSURE SUDDEN CHANGE IN PROPERTIES OF RETURNED MUD
LOWER KELLY COCK IS 2m 1. PICK UP UNTIL ABOVE ROTARY 5in PIPE (Space out to ensure that a tool joint is clear of rams) 2. SHUT DOWN THE PUMPS 3. FLOWCHECK THE WELL IF NECESSARY (Do not flowcheck if 2* or 3* as above have been detected.)
YES
NO
1. OPEN CHOKE LINE VALVE (S) 2. CLOSE ANNULAR PREVENTER 3. CHECK THAT WELL IS SHUT IN 4. RECORD DP AND CSG PRESSURE 5. NOTIFY COMPANY REPRESENTATIVE 6. PROCEED AS DIRECTED
WEOX02.016
4-13
March 1995
WELL NO
28 ON ALL TRIPS
COMPANY REP
RIG
RIG 10 H.H.
TOOLPUSHER
23/7/87
A.J.N.
IF ANY OF THE FOLLOWING OCCUR: HOLE NOT TAKING CORRECT VOLUME DURING THE TRIP THE WELL IS FLOWING 5. 6. 7. 8. Or if there is any other possible indication of a kick. 1. 2. 3. 4.
YES
NO
1. SET THE SLIPS 2. INSTALL OPEN DP SAFETY VALVE 3. CLOSE DP SAFETY VALVE 4. OPEN CHOKE LINE VALVE (S) 5. CLOSE ANNULAR PREVENTER 6. CHECK THAT WELL IS SHUT IN 7. NOTIFY COMPANY REPRESENTATIVE 8. INSTALL KELLY 9. LINE UP STANDPIPE MANIFOLD 10. OPEN DP SAFETY VALVE 11. RECORD DP AND CSG PRESSURE 12. IF IN OPENHOLE: ENGAGE BUSHINGS, ROTATE THE PIPE 13. PROCEED AS DIRECTED
WEOX02.017
4-14
March 1995
4.3
1 2 3 4 5 6 7
Paragraph
Illustrations
4.4 4.5 4.6 4.7 Shut-in Pressure Build-up Curve showing the effect of influx migration Well Control Operations Log An Example Calculation showing how to evaluate the type of influx fluid An Example of the possible increase in wellbore pressure due to influx migration 4-17 4-18 4-22 4-23
4-15
March 1995
1 General
When a flowing well is shut in by closing the BOPs, the flow will continue until shut-in pressures have built up to balance the static reservoir pressure. In most cases, this will mean that the flow will stop almost immediately the BOPs are closed and that the shut-in pressure will stabilise within a few minutes. In general, only if the well has been flowing for some time will the kick zone pressure take time to build up to a maximum after the well has been shut in. In most cases, when a kick is taken, the inflow into the wellbore occurs for only a short time and the drawdown is relatively small. As a result, pressure in the wellbore will stabilise quickly after the well is shut in. However, there have been many cases of surface pressures taking several hours to stabilise. The reasons for this can be one, or all, of the following: The influx originated from a low permeability zone. The influx created instability in the wellbore, leading to the hole sloughing and packingof f. The influx is migrating up the hole. The surface lines or subsea choke line is partially packed off.
This section covers the procedures that may be required during the time the well is shut in prior to circulation.
4-16
March 1995
STABILISED PERIOD
ANNULUS PRESSURE
DRILLPIPE PRESSURE
Figure 4.4 Shut-in Pressure Build-up Curve showing the effect of influx migration
Figure 4.5 shows a form that can be used to record the build-up of drillpipe and casing pressure. This form should also be used to keep a complete record of events during the well control operation.
4-17
March 1995
WELL NO
28 /
RIG
RIG 9
DRILLPIPE CHOKE PIT LEVEL/ PRESSURE PRESSURE VOLUME (psi) (psi) (bbl )
REMARKS
03.00 03.01 03.02 03.03 03.04 03.05 03.06 03.07 03.08 03.09 03.10 03.11 03.12 03.13 03.14 03.15 03.16 03.17 03.20 03.25 03.30 03.45 04.00 04.30 05.00 05.10 05.15 05.20 05.24 05.30 05.40 05.50 06.00 06.10
300 360 420 460 520 590 630 700 720 740 760 770 775 775 780 780 780 780 780 780 780 780 780 782 782 782 782 782 1085 1010 880 755 630 450
450 500 560 600 660 730 770 840 860 880 900 910 920 920 925 925 925 925 925 925 925 925 925 925 925 925 925 925 925 910 903 930 945 950
120 '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' '' 120
100bbl 1.75 SG MUD MIXED IN TANK No 1 VERIFY EQUIPMENT CORRECTLY LINED UP CO REP AND TOOLPUSHER TO RIG FLOOR START CIRCULATION BRING PUMP UP TO 25 SPM PUMP UP TO SPEED RETURNS THROUGH DEGASSER
WEOX02.019
4-18
March 1995
2. Carefully monitoring both the pump and casing pressure, pump to the hole at a controlled rate (very slow). 3. Record the increase in pump pressure and the volume of mud pumped. The relationship between the pump pressure and the volume of mud pumped will be linear as the mud in the drillpipe is compressed. If pumping is continued after the pressure equalises across the float valve, the valve will open. As the valve opens, the pump pressure will increase slower than before; this change should be easily recognisable at slow pump rates. Stop the pump when this change is noticed. The casing pressure is also likely to show an indication of the valve opening. 4. Isolate the pump at the standpipe. 5. Record the shut-in drillpipe pressure as the pump pressure recorded immediately before the float valve opened. 6. If the casing pressure rises at any stage, immediately stop the pump. Isolate the pump. Bleed off the excess pressure from the casing. As an example, if the casing pressure rose 50 psi and this extra pressure was considered undesirable, bleed 50psi from the casing and record the shut-in drillpipe pressure as 50 psi less than the final pump pressure. The utmost care must be taken in carrying out this procedure. As outlined, this procedure involves pumping into a closed well. The well is pressurised at the start of the operation, and so any excessive additional pressurisation caused by pumping into the well may overpressure the openhole section.
4 Trapped Pressure
In some circumstances it is possible that pressure, in excess of that caused by the kick zone, can be trapped in the well. There are three possible causes of this phenomenon: The pumps were left running after the well was shut-in. The influx is migrating up the hole. Pipe has been stripped into the well without bleeding the correct volume of mud.
Trapped pressure of this kind will result in surface pressures that do not reflect the actual kick zone pressure. However if the surface pressure built up at any point after the well wasshut-in, this is confirmation that there is no trapped pressure in the well. Pressure may be trapped in the well if the surface pressure appears constant and no pressure build has been seen. The drillpipe pressure is used to determine the kick zone pressure and hence the mud weight used to kill the well. An artificially high drillpipe pressure reading, used to determine the kill mud weight, will result in overkilling the well.
4-19
March 1995
The following procedure can be used to check for trapped pressure: 1. Ensure that accurate pressure gauges are fitted to the drillpipe and annulus. Carefully monitor the drillpipe and casing pressure. 2. Using a manual choke, bleed a small volume of mud from the annulus to a suitable measuring tank. (1/2 barrel is a suitable amount.) 3. Shut in the well. Allow pressure to stabilise. If pressure has been trapped in the well, the drillpipe pressure and casing pressure will have fallen. If the drillpipe pressure does not drop after bleeding mud from the annulus, no pressure is trapped in the well. Be aware that, if there is no trapped pressure in the well, each increment of mud bled from the well will cause a further influx into the well. Therefore, if no reduction in drillpipe pressure is detected after bleeding 2 3/bbl from the well, no more mud should be bled off. An increase in casing pressure is a sure sign that additional influx has entered the well. Therefore, if this occurs, no more mud should be bled from the well. 4. If both the drillpipe pressure and casing pressure have decreased, continue to bleed mud from the well in 1/2 bbl increments. 5. When the drillpipe pressure no longer decreases as mud is bled from the well, record the drillpipe pressure as the shut-in drillpipe pressure. Stop bleeding mud from the well. It should be stressed that bleeding mud from a well that has kicked is an operation that must be carefully implemented. Whilst it is undesirable to overkill the well, it is potentially hazardous to increase the size of the influx, which is clearly a possibility if this procedure is not properly carried out. A firm recommendation is that the volumes bled from the well at this stage are kept to a minimum, unless influx migration is obviously occurring. If there is some doubt as to the true shut-in drillpipe pressure, even after bleeding mud from the annulus, it may be prudent to use the Drillers Method to circulate out the kick, rather than continue bleeding mud. This procedure is not recommended if the kick zone is suspected to have low permeability. Bleeding even very small quantities of mud from the annulus may reduce the pressure of a tight kick zone below its final shut-in pressure. The drillpipe pressure will continue to decrease, giving the false impression at surface that the bottomhole pressure is still greater than the actual kick zone pressure. A possible consequence is that the operator may inadvertantly reduce the bottomhole pressure significantly below the kick zone pressure and cause a further influx into the wellbore.
4-20
March 1995
The pit gain at surface provides a guide to the volume of the kick. With this information, together with the annular geometry and the surface pressures, it is possible to estimate theinflux density . The type of influx fluid can then be evaluated, using the following as aguide: Influx fluid Gas Oil Water Calculated Influx gradient (psi/ft) 0.05 0.2 0.3 0.4 > 0.4
Figure 4.6 shows an example of how to determine the influx type. This calculation is only an approximation, for the following reasons. Firstly, it is assumed that the influx is a discrete bubble, whereas it is more likely to be eccentric to the hole and contaminated with mud. Secondly, the effective mud weight in the annulus is not likely to be the same as in the drillpipe, due to cuttings loading the annulus, and possibly, contamination of the mud with formation fluid. Thirdly, the hole may be out of gauge. It is important, however, that this calculation is carried out for the additional reason that it provides a check of the validity of the kick data. It is useful to know the type of influx before circulation is initiated. Although most formation fluids, including formation water, contain some gas, the calculated influx gradient provides a guide to the proportion of gas in the fluid. The proportion of gas in the influx determines two important factors, firstly, the well bore pressures during displacement, and secondly, the pit gain during displacement. If the gas contains sufficient heavy hydrocarbon molecules at reservoir conditions, condensate fluids may be formed as the gas is displaced from the hole. This will not occur for a dry gas that does not contain a sufficient proportion of heavy molecules. Gas will come out of solution from an oil influx when the influx pressure reduces below the bubble point pressure during displacement. For light oils, a significant quantity of gas will be produced. It is recommended that all kicks are assumed to contain a certain proportion of gas. Prior to circulation therefore, an estimation should be made of the maximum pressures that will be encountered during circulation, and provision should be made for a pit gain during this period. (See Chapter 5, Volume 2 for hand calculation techniques.)
6 Influx Migration
After a kick is taken, there is usually a tendency for the influx to migrate up the hole. Thistendency is caused by the dif ference in density between the influx fluid and the mud. Influx migration up a closed-in well can cause excessive pressures within the wellbore if suitable control procedures are not implemented. Figure 4.7 shows an example of the potential increase in bottomhole pressure caused by gas migration.
4-21
March 1995
Figure 4.6 An Example Calculation showing how to evaluate the type of influx fluid
HOLE DRILLSTRING DIMENSIONS PRESSURE BALANCE
1. Determine the bottomhole pressure DRILLPIPE SURFACE PRESSURE 1.7SG MUD 2. Determine the hydrostatic pressure of the influx ANNULUS SURFACE PRESSURE
855
500
+
HYDROSTATIC PRESSURE OF MUD IN THE DRILLPIPE
MUD MUD
+
HYDROSTATIC PRESSURE OF MUD IN ANNULUS
81/2in HOLE
61/4in COLLARS
+ =
INFLUX HYDROSTATIC PRESSURE
INFLUX
=
BOTTOMHOLE PRESSURE
Identify the influx fluid as follows: 1. Determine the bottomhole pressure Bottomhole pressure
= = = =
Drillpipe pressure + mud hydrostatic pressure 500 + (1.7 x 1.421 x 4000) 500 + 9663 10,163psi
2. Calculate the height of the influx in the annulus Influx volume = Recorded pit gain = 20bbl Annular capacity at collars = 0.1058bblm Height of influx = 20/0.1058 = 189m 3. From pressure balance Annulus surface pressure + Hydrostatic pressure of the mud + Hydrostatic pressure of the influx = Bottomhole pressure 855 + 1.7 x 1.421 x (4000 189) + Pi Pi, hydrostatic pressure of the influx = 10,163psi = 10,163 - 855 - 9206 = 102psi = Pi/height of the influx = 102/(189 x 3.2808) = 0.16psi/ft
Influx gradient
Therefore the influx is mainly gas 4. The following formula can also be used routinely to calculate the influx density Density of the influx (SG) = MW Pa Pdp h x 1.421 = 1.7 855 500 189 x 1.421 = 0.378 SG 0.16psi/ft
WEOX02.020
4-22
March 1995
SURFACE PRESSURE 0
250psi
3195psi
2975m 3000m BOTTOMHOLE PRESSURE BOTTOMHOLE EMW (SG) 6180psi 1.45 GAS @ 6180psi 9160psi 2.15 (12140psi) (2.85)
WEOX02.021
Figure 4.7 An Example of the possible increase in wellbore pressure due to influx migration
Influx migration does not always occur, but when it does, the rate at which the influx rises up the hole is dependent on several variables. Experiment has shown that a gas bubble will migrate up one side of the annulus as mud falls down the opposite side. Bearing this process in mind, it is clear that the factors that predominantly affect the rate of rise of the influx will be the following: The viscosity of the drilling fluid. The more viscous the mud, the more difficult it is for the mud to fall down the annulus to allow the influx to migrate. The difference in density between the mud and the influx. The buoyancy force causes the influx to migrate. Any interaction between the mud and the influx fluid. Migration will be slowed if the viscosity of the mud is increased as a result of contamination with the influx fluid. In severe cases, migration may be completely prevented.
4-23
March 1995
4-24
March 1995
Page
5-2 5-2 5-2 5-5 5-5 5-7 5-9
Illustrations
5.1 5.2 5.3 5.4 5.5 5.6 Preparations for the Well Kill Decision Analysis Pipe off Bottom (Drillpipe in the Stack) Decision Analysis Pipe off Bottom (Drillcollar in the Stack) Decision Analysis No Pipe in the Hole Decision Analysis Flow to a Fracture above a High Pressure Zone Decision Analysis Flow to a Fracture or Loss Zone below a High Pressure Zone
5-1
March 1995
1 General
This Chapter is intended to provide guidelines to the decision making process in the event that a kick is taken in a variety of different situations. In reality, the specific conditions prevailing at the rigsite at the time that the kick is taken will determine the best course of action to take in order to kill the well. This Chapter should therefore not be used as a guide at the moment that a kick is taken. However, it is anticipated that general familiarity with the analysis presented in this Chapter will enable rigsite personnel to be better prepared to deal with a situation in which the well has kicked. The techniques referred to in this section are covered in detail in Chapter 6, Well Kill Techniques.
2 Pipe on Bottom
If a kick is taken with the pipe on bottom, the well will be shut-in immediately unless the decision has previously been made to divert. Having established that the well is safely closed in, it will be necessary to decide on the most appropriate method of killing the well. This decision is the responsibility of the Company Representative. Having decided on the most appropriate course of action, the Company Representative is responsible for ensuring that contractor personnel are made aware of the procedures that will be used to kill the well. The general procedure that is presented in Figure 5.1 represents the steps that should be taken in preparation to kill the well. These steps are applicable to any situation in which a kick is taken.
5-2
March 1995
PREKILL MEETING
DECISION MADE AS TO MOST APPROPRIATE METHOD OF KILLING THE WELL DRILLING SUPERINTENDENT IN TOWN SHOULD BE MADE AWARE OF THE SITUATION
COMPLETE PREPARATIONS
WEOX02.022
5-3
March 1995
Figure 5.2 Decision Analysis Pipe off Bottom (Drillpipe in the Stack)
WELL KICKS PIPE OFF BOTTOM (Drillpipe in stack)
IS IT POSSIBLE TO STAB A SAFETY VALVE? YES STAB AND CLOSE FULL OPENING SAFETY VALVE
NO
HANG OFF
SHEAR PIPE
CLOSE ANNULAR
INSTALL DP DART OR INSIDE BOP THE SEVERITY OF THE SITUATION DICTATES THAT STRIPPING WITH RIG EQUIPMENT IS IMPRACTICAL
CONSIDER SNUBBING
YES
NO
ATTEMPT TO REDUCE SURFACE PRESSURE CONSIDER: VOLUMETRIC LUBRICATION BULLHEADING CIRCULATE OUT INFLUX
YES
ATTEMPT TO LOWER PIPE THROUGH STACK ATTEMPT TO REDUCE SURFACE PRESSURE CONSIDER: VOLUMETRIC LUBRICATION BULLHEADING CIRCULATE OUT INFLUX
NO
POSSIBLE TO LOWER PIPE THROUGH ANNULAR? YES ATTEMPT TO REDUCE SURFACE PRESSURE CONSIDER: VOLUMETRIC LUBRICATION BULLHEADING CIRCULATE OUT INFLUX YES
YES
NO
YES
CONSIDER SNUBBING
5-4
March 1995
Further complications that may arise in this situation are numerous, but include the following: There is not sufficient weight of collars to strip through the annular BOP. Well pressures force the collars out of the hole. An internal blowout through the drillstring.
The appropriate course of action required in these situations will depend to a large extent on the particular conditions and equipment at the rigsite. However Figure 5.3 is intended as a guide to dealing with such situations.
5-5
March 1995
Figure 5.3 Decision Analysis Pipe off Bottom (Drillcollar in the Stack)
IS IT POSSIBLE TO STAB A SAFETY VALVE? YES STAB AND CLOSE A FULL OPENING SAFETY VALVE
NO
CLOSE ANNULAR
YES
MINOR LEAK
LEAK STOPS
YES
NO
IS IT POSSIBLE TO LOWER PIPE INTO THE HOLE? YES STRIP IN UNTIL DRILLPIPE IN THE STACK
NO
YES
CONSIDER SNUBBING
WEOX02.024
5-6
March 1995
However, if the influx is someway down the hole, it may not be possible to reduce the surface pressure significantly. If the influx is migrating up the hole, it may be possible to kill the well by implementing the Volumetric Control Method. On fixed offshore and land rigs, the only practical method of controlling the well may be with the use of a snubbing unit. Snubbing units have been used in exceptional circumstances on floating rigs. Figure 5.4 represents a full analysis of the decision making process in the event that a kick is taken with no pipe in the hole.
Particular attention should therefore be paid to these aspects. In critical well sections, consideration should be given to installing casing rams in the BOP stack prior to running casing; this is only practical in surface stacks. Specialist shear rams are available that can shear up to 13 3/8 in. casing; these may be considered applicable in certain situations. It is impractical to detail the procedure required in the event that a kick is taken while running casing or a liner. The immediate priority however will be to close in the well, but the most suitable control technique can only be determined bearing in mind the particular conditions at the rigsite. The subsequent options available can be summarised as follows: Cross over to drillpipe (unless current string weight is too great) and strip to bottom to kill the well. Cross over to drillpipe, strip in until drillpipe is in the stack and kill the well at current shoe depth. Kill the well with the casing across the stack. Drop the casing. Shear the casing.
5-7
March 1995
NO
NO
YES
NO
YES
NO
DO SURFACE PRESSURES INDICATE THAT INTRODUCING PIPE INTO THE HOLE IS POSSIBLE?
YES
YES
NO
YES
NO
BULLHEAD KILL MUD INTO THE WELL PREPARE CONTINGENCY TO DEAL WITH THE FRACTURED ZONE KILL WELL
NO
YES
WEOX02.025
5-8
March 1995
The major factors that will determine the most appropriate course of action will include the following: The length and type of casing run. The possibility and consequences of the casing becoming stuck. The possibility and consequences of collapsing the casing. The feasibility of circulating out a kick by conventional means. (The relatively small annular clearance may cause excessive pressures in the annulus, or may possibly completely restrict circulation.) The feasibility of killing the well by other means such as bullheading or by volumetric control. The BOP stack configuration and ram types. The likelihood of the casing being forced out of the hole by the well pressure.
7 Underground Blowout
(a) Flow to a Fracture above a High Pressure Zone
The majority of underground blowouts in the past have been as a result of a fracture to a weak zone up the hole as high pressure zone is penetrated. Figure 5.5 shows a decision analysis for identifying and dealing with an underground blowout of this type. If an underground blowout is suspected, on no account should attempts be made to control the well using standard techniques. If the annulus is opened, reservoir fluids will be allowed to flow up the wellbore to surface, thereby increasing surface pressures. The first action, after shutting in the well, will be to perform a positive test. The purpose of this test is to determine whether or not the hole is a closed system. A small displacement pump is lined up to the drillpipe and a small amount of fluid is pumped. If the drillpipe and casing pressure increase, there is no indication of fracture in the openhole. If the drillpipe pressure does not increase, or if any increase is not evident on the casing, then a fracture in the openhole is indicated. In order to halt an underground flow, it is necessary to pump fluid at a high rate down the drillpipe and up the annulus; thus effecting a dynamic kill. The fluid will eventually have to be at kill weight in order to balance the kick zone EMW. However, it will also have to be as thin as possible to ensure that it can be pumped at high rate without excessive surface circulating pressures. Generally the kill mud must flow at least as fast as the underground flow if it is not to be dispersed by the flow as it passes out of the bit. The kick zone EMW can at best be estimated because reliable drillpipe pressure will not be available. The mud weight required to kill the well will depend on the position of the fracture in the wellbore and the average weight of the fluid occupying the annulus between the fracture and surface.
5-9
March 1995
Figure 5.5a Decision Analysis Flow to a Fracture above a High Pressure Zone
SHUT IN THE WELL
SUSPECT UNDERGROUND BLOWOUT BECAUSE: 1. DRILLPIPE ON VACUUM 2. PRESSURE BUILDUP CLEARLY INDICATES FORMATION HAS FRACTURED 3. ANNULUS PRESSURE FLUCTUATING
NO
YES
1. DO NOT BLEED FLUID FROM ANNULUS 2. LINE UP ONE PUMP TO THE ANNULUS. LINE UP MUD AND IF NECESSARY WATER SUCTION
IF ANNULUS PRESSURE IS BUILDING, PUMP MUD AT SLOW RATE DOWN ANNULUS. IF ANNULUS CANNOT SUPPORT MUD, PUMP WATER
5-10
March 1995
Figure 5.5b Decision Analysis Flow to a Fracture above a High Pressure Zone (continued)
PREPARE 2 x ANNULUS VOLUME OF KILL WEIGHT MUD (AT MIN PV AND YP USE FRICTION REDUCER IF AVAILABLE). REMOVE KELLY INSTALL HP CIRCULATING LINE
PUMP KILL WEIGHT MUD AT MAXIMUM RATE KEEP PUMPING UNTIL ALL THE MUD IS USED STOP ONLY IF SURFACE PRESSURES BECOME EXCESSIVE
1. CHECK MUD IS AT KILL WEIGHT 2. REDUCE MUD VISCOSITY 3. REDUCE DRILLSTRING INTERNAL FRICTION 4. PUMP LARGER PLUG
TRY AGAIN
DRILLPIPE AND ANNULUS PRESSURES INDICATE THAT UNDERGROUND FLOW HAS CEASED?
YES
TAKE STEPS TO SECURE WELL OPTIONS: 1. CEMENT BHA IN PLACE 2. POOH TO PLUG FRACTURE 3. POOH TO RUN CASING
NO
1. MIX LCM PILL (100bbl MIN FOR LARGE ANNULUS) 2. MIX 2 x ANNULUS VOLUME OF KILL WEIGHT MUD 3. PUMP LCM PILL DOWN ANNULUS UNTIL JUST ABOVE FRACTURED ZONE
PUMP KILL MUD AT MAXIMUM RATE DOWN DRILLPIPE PUMP LCM PILL DOWN ANNULUS AND INTO FRACTURE KEEP PUMPING UNLESS SURFACE PRESSURE LIMITS ARE REACHED
1. CHECK MUD IS AT KILL WEIGHT 2. REDUCE MUD VISCOSITY 3. REDUCE DRILLSTRING INTERNAL FRICTION 4. PUMP LARGER PLUG
TRY AGAIN
DRILLPIPE AND ANNULUS PRESSURES INDICATE THAT UNDERGROUND FLOW HAS CEASED?
YES
TAKE STEPS TO SECURE WELL OPTIONS: 1. CEMENT BHA IN PLACE 2. POOH TO PLUG FRACTURE 3. POOH TO RUN CASING
NO OPTIONS: 1. BACK OFF, STRIP UP INTO CASING, SQUEEZE HIGH FILTER LOSS CEMENT SLURRY TO PLUG WELL 2. IF CIRCULATION IS POSSIBLE ON BOTTOM, PUMP FRESHWATER AT MAXIMUM RATE TO SLOUGH HOLE OPTIONS: 1. STRIP UP INTO CASING. HAVING INSTALLED DART SQUEEZE HIGH FILTER LOSS CEMENT SLURRY TO PLUG WELL 2. PUMP FRESHWATER AT MAXIMUM RATE TO SLOUGH HOLE WEOX02.027
YES
NO
5-11
March 1995
Figure 5.6 Decision Analysis Flow to a Fracture or Loss Zone below a High Pressure Zone
NO
YES
PUMP LCM PILL SET CEMENT PLUG ON BOTTOM CIRCULATE THE HOLE
NO
5-12
March 1995
The fracture may only support a column of water, in which case it will be necessary to balance the kick zone pressure with the sum of the hydrostatic pressure of the kill weight mud from the kick zone to the fracture and the hydrostatic pressure of the water above the fracture. If the first attempt to control the flow is unsuccessful, the most likely causes will beeither that the volume or the velocity of kill mud was insufficient. Subsequent options therefore include increasing the volume of the kill mud pumped and pumping at a greater rate. If the rig pumps have been operating at maximum output there remains the options to bring more pumps to the rigsite or to reduce the frictional resistance of the drillstring by such measures as: Removing the nozzles of the bit with a charge run on wireline. Perforating the BHA close to the bit. Pumping a lighter, less viscous mud ahead of the kill weight mud in order to reduce the velocity of the inflow.
As indicated in Figure 5.5, if these measures do not bring the well under control, there remains the option to mix an LCM pill or soft plug (See Chapter 2, Section 2.3 ) and displace it down the annulus and into the fracture as the kill weight mud is pumped down the drillpipe. The pump rates on the drillpipe and the annulus should be such as to ensure that the LCM pill is completely displaced into the fracture over the period of time that will be required to pump the prepared volume of kill weight mud. Past experience has shown that in many cases, having halted the underground flow, a further flow has been initiated by attempts to pull off bottom. If the decision is made to pull off bottom having halted an underground flow, extreme care should be taken. The industry has given the term Baryte plug to the heavy weight pills required to deal with underground blowouts. The recommended procedure for mixing and spotting a baryte plug, to deal with an underground blowout, is covered in Chapter 6.
5-13
March 1995
Drilling under pressure will however only be used in circumstances in which lost circulation of this type has been anticipated, the high pressure zone has low permeability and the correct equipment, including a rotating head, is available onsite. See Chapter 2, Section 2.3 for LCM and cement plug recipes.
5-14
March 1995
March 1995
Page
6-2 6-2 6-3 6-5 6-8 6-11 6-14 6-23
Illustrations
6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 An Example Completed Kick Sheet The Kill Line Monitor Subsea BOP Gas prior to Removing Gas from Below the Preventers Removing Gas from a Subsea BOP Stack Lower pipe rams closed hang off rams opened Removing Gas from a Subsea BOP Stack Kill and choke lines displaced to kill weight mud Removing Gas from a Subsea BOP Stack Kill and choke lines displaced to water Removing Gas from a Subsea BOP Stack Gas pressure bled down, gas occupies choke line Removing Gas from a Subsea BOP Stack Diverter is closed, the annular is opened and the gas is displaced from the stack The Effect of Choke Line Losses Casing pressure greater than choke line pressure 6-14 6-21 6-24 6-25 6-26 6-27 6-28
6-29 6-30
6.9
6.10 The Effect of Choke Line Losses Casing pressure after initial circulation is less than choke line loss
6-31
6-1
Rev 1 March March 1995 1995
1 General
This section covers the basic steps that are required to implement the Drillers Method, the Wait & Weight Method on both a fixed installation as well as a floating rig. Further discussions on the theories behind the methods are covered in Vol.2, Chapter 5. Company policy is that a contingency plan must be developed regarding the implementation of the well control methods for both Company operated rigs and rigs that are under a Company contract. This section is intended to assist in drawing up these contingency plans. All the well control techniques are designed to ensure that: Bottom hole pressure is maintained constant and equal to, or slightly greater than, the formation pressure. This is the key to well control practice. These techniques use the principle that: The drillpipe pressure is used to monitor bottom hole pressure. In the event of any well control incident it is important that a diary of events is kept. The Well Control operations log can be used initially for this (See Figure 4.5). A full report should eventually be issued and submitted to Line Management.
6-2
Rev March 1 March 1995 1995
Under the above circumstances, the Drillers method should be considered. The Drillers Method requires that two complete hole circulations are carried out before the well can be killed. After a kick is taken and the well shut-in, the kick is displaced from the hole by the first circulation with the original mud. In the mean time the mud is weighted up to kill weight, and the second circulation carried out to kill the well. The advantages of the Drillers Method over the Wait & Weight Method are: The kick can be displaced from the hole soon after the well is shut-in. The earlier circulation may reduce the risks of stuck pipe and other hole problems. Influx fluids can be displaced from the well, even if suitable mud weighting material is not available. It avoids the need to initiate a volumetric control during the waiting period.
3 Kick Sheet
The kick sheet should be used to record all the relevant well and kick data. Figures 6.1a, 6.1b and 6.1c show an example kick sheet. The procedures for completing the kick sheet are shown in Figure 6.1d. The general well data, drillstring/annulus contents, circulating times and the mud pump data should be recorded routinely and available at all times in the kick sheet. In case a kick is taken, the relevant kick data should be recorded in the kick sheet. The shut-in procedure and the interpretation of the pressure data are covered in Chapter 4. Based on the kick data, a decision should be made regarding what method be used to kill the well. In addition to the standard methods which have been described in the previous paragraphs, some special techniques should be also considered. These special techniques are discussed in Section 6.2. If the decision is made to displace the kick from the hole by using one of the standard methods, the relevant parameters should be calculated and recorded in the kick sheet.
6-3
Rev 1 March March 1995 1995
Pdp = Stabilised shut-in drillpipe pressure (psi) MW1 = Original mud weight (SG) TVD = True vertical depth of kick zone (m)
It is not recommended practice to weight the mud any higher than the kill weight during the well killing operation. After the well has been killed however, the mud weight should be raised to provide suitable overbalance.
Total quantity of baryte required (lb) = Wb x Total Active Mud Volume Total active mud volume = Drillstring Vol + Annulus Vol + Surface Active Vol
The stocks of baryte at the rigsite must be at least 10% greater than the calculated quantity of baryte required.
6-4
Rev 1 1995 March 1995 March
A limiting factor, particularly in the case of oil mud, may be the rate at which viscosity can be built in the mud. This, and associated problems of building mud weight are discussed in Chapter 1 in Use of the Mud System. Having established the maximum possible circulation rate, the actual circulation rate will be determined on the basis of several factors. These factors are detailed in Chapter1 in Drills and SCRs. The chosen SCR and the relevant pumping data should be recorded in the kick sheet.
6-5
Rev 1 March March 1995 1995
The final circulating pressure can be estimated as follows: Pfc = P scr (at MW1) x MW2 MW1
(d) Determine the displacement times and the cumulative pump strokes
At all times during circulation, it is important to know the position of the influx in the wellbore, as well as the volume of hole that has been circulated to kill weight mud. The key points during the circulation are as follows: When the kill weight mud reaches the bit. When the top of the influx is circulated to the casing shoe or openhole weak point. When the influx is circulated to the choke.
Before circulation is started, the estimated circulating time and the corresponding total pump strokes to each point should be calculated. Pumping time to reach point = of interest Total strokes to reach point = of interest Volume to be displaced (bbl) (min) Pump rate (bbl/min) Volume to be displaced (bbl) (stk) Pump output per stroke (bbl/stk)
6-6
March Rev 1 1995 March 1995
Maintain the choke pressure equal to the original shut-in casing pressure as the pump is slowly brought up to speed. This may take 1/2 to 1 minute. Once the pump is up to speed record the initial circulating pressure.
If the actual initial circulating pressure is considerably different from the calculated value, stop the pump, shut in the well and investigate the cause. If the actual initial circulating pressure is equal to, or reasonably close to the calculated value, continue the displacement and adjust the standpipe pressure schedule accordingly. Any marginal difference between the actual and calculated initial circulating pressure is most likely to be due to the fact that the SCR pressure used to calculate the initial circulating pressure was inaccurate. The actual SCR pressure, and hence the corrected final circulating pressure, Pfc, can be determined from the initial circulating pressure as follows:
As the drillpipe is displaced with kill weight mud, the standpipe pressure should be stepped down according to the standpipe pressure schedule. (The standpipe pressure will have a natural tendency to drop as the kill weight mud is displaced down the drillpipe.) Once the drillpipe has been displaced to kill weight mud, the drillpipe pressure should be maintained at the final circulating pressure for the rest of the circulation. The pit gain, drillpipe pressure, choke pressure and all other relevant information should b e recorded during displacement using the Well Control Operations Log (SeeFigure 4.5). These will help to determine the down hole condition during all stages of the kill operation. As the influx is displaced up the hole, the drillpipe pressure will tend to drop as the influx expands. (This expansion will not occur if the influx is water or oil.) This effect will be especially marked if the influx contains a significant quantity of gas. The choke should therefore be adjusted to compensate for this. For example, if the drillpipe pressure drops by 70 psi below that required, the choke pressure should be increased by approximately 70 psi. The pressure on the drillpipe will increase after a lag time which can typically be 2 seconds per 300m of drillstring depth. This technique will be most effective at the early stages of displacement; and less so at later stages of the displacement, if the well contains a significant proportion of gas. When the influx reaches the choke, the choke pressure will start to decrease due to the differences in density and viscosity between the influx and the mud. If the influx contains significant quantities of gas, the drop in choke pressure may be quite substantial, and the choke will have to be closed down quickly. As the influx is circulated from the well and mud is circulated to the choke, the choke pressure will begin to rise rapidly. The choke should therefore be opened to allow the choke pressure to drop sufficiently to re-establish the final circulating pressure on the drillpipe, and hence maintain constant bottom hole pressure.
6-7
Rev 1 March 1995
Once the hole has been circulated to kill weight mud, the pump should be stopped, the well shut-in, and the casing and drillpipe checked for pressure. There should be no pressure on either the casing or the drillpipe. However, if there is still some pressure on the casing, circulation should be restarted to clear the contaminated mud from the annulus. Once the well has been completely killed, a flowcheck on the choke line return should be carried out before the rams are opened. If this flowcheck indicates no flow, the rams should be opened and a further flowcheck on the annulus carried out. In line with Company policy, a further complete hole circulation should be carried out prior to continuing operations. A suitable overbalance can be added to the mud at thisstage.
6-8
March Rev 1 1995 March 1995
The following steps can be used as a guide for the procedure for the displacement of thekick: 1 Bring the pump up to speed for the first complete circulation Line up the pump to the drillpipe and route returns through the choke manifold to the mud gas separator. Set the stroke counter on the remote choke panel to zero. Open the remote operated choke at the same time as the pump is slowly brought up to speed. Consider stroking the drillstring up at this point. Maintain the choke pressure equal to the original shut-in casing pressure as the pump is slowly brought up to speed. This may take 1/2 to 1 minute. Once the pump is up to speed record the initial circulating pressure. If the actual initial circulating pressure is considerably different from the calculated value, stop the pump, shut-in the well and investigate the cause.
If the actual initial circulating pressure is equal to, or reasonably close to the calculated value, continue the displacement, holding the standpipe pressure at the value recorded when the pump was first brought up to speed. Any marginal difference between the actual and calculated initial circulating pressure is most likely to be due to the fact that the SCR pressure used to calculate the initial circulating pressure was inaccurate. The actual SCR pressure can be determined from the initial circulating pressure as follows:
Pscr = Pic Pdp
This adjusted value for the SCR pressure should be used for estimating the circulating pressures for the second complete circulation. 2 Circulate the influx from the well maintaining constant bottom hole pressure
Influx behaviour during circulation will be similar to the Wait and Weight Method requiring similar choke manipulation. Choke pressures will inevitably be higher than if the Wait and Weight Method had beenused. These higher pressures will be reflected downhole, causing greater stress in the openhole. Once the influx has been displaced from the hole, the shut-in drillpipe and shut-in casing pressure should be equal. If the casing pressure is higher than the drillpipe pressure, this is evidence that there is still some kick fluid in the annulus, or the mud weights are out of balance. Prior to circulating kill weight mud into the hole, the calculations as outlined in Paragraph3 Kick Sheet should be carried out. The following further calculations are then worked:
6-9
Rev 1 March 1995
Pfc = Second circulation final circulating pressure (psi) MW1 = Original mud weight (SG) MW2 = Kill mud weight used for second circulation (SG)
(d) Determine the displacement times and corresponding cumulative pump strokes
These figures will be the same as for the first circulation.
The initial circulating pressure should be the same with the standpipe pressure during the first complete circulation. If this is the case, continue the displacement and adjust the standpipe pressure schedule accordingly. If the initial circulating pressure has changed considerably, stop the pump, shut in the well, and investigate the cause.
6-10
Rev 1 1995 March 1995 March
Circulate the hole to kill weight mud maintaining constant bottom hole pressure
As the drillpipe is displaced with kill weight mud, the standpipe circulating pressure should be stepped down according to the standpipe pressure schedule. Once the drillpipe has been displaced to kill weight mud, the final drillpipe circulating pressure is held constant by manipulating the choke. As kill weight mud is circulated up the annulus, the drillpipe pressure will tend to increase. The choke should be adjusted to ensure that the drillpipe pressure is maintained at the final circulating pressure; thereby ensuring constant bottom hole pressure. When the returned mud is at kill weight, the pump should be stopped and the well shut-in. The well should be checked for pressure. Once the well has been killed, a flowcheck on the choke line return should be carried out before the rams are opened. If this flowcheck indicates no flow, the rams should be opened and a further flowcheck on the annulus carried out. In line with Company policy, a further complete hole circulation should be carried outprior to continuing operations. A suitable overbalance can be added to the mud at this stage.
6-11
Rev 1 March March 1995 1995
If there is more than one drillpipe section (tapered string), then the size factor should be calculated for each of the sections. BHA can be treated as part of the drillpipe section. = where: Pfc - P scr 1 + 2 1 2 Pfc Pscr = = = = Drillpipe friction constant, (psi.in 5/m) Drillpipe size factors for section 1 and 2, (m/in5) Final circulating pressure (psi) Slow circulating pressure with original mud MW1 , (psi)
ii. Calculate the friction pressure increase when the kill mud reaches each of the critical depths in the drillpipe (kick-off, end-build, end-tangent, etc.). If the critical depth is above/at the drillpipe section cross-over point, then: P friction = x MD ID 1 5
If the critical depth is below the drillpipe section cross-over point, then: P friction = x [ 1 + (MD - L1 ) ] ID 25
where: P friction = Friction pressure increase due to kill weight mud, (psi)
6-12
March Rev 1 1995 March 1995
MD = Measured depth at the critical depth, (m) iii. Calculate the static drillpipe pressure when the kill weight mud reaches each of the critical depths: Pstatic = Pdp x (1.0 where: Pstatic Pdp TVD TVD h TVD ) TVD h
= Static drillpipe pressure, (psi) = Drillpipe pressure before the kill weight mud is circulated, (psi) = Vertical depth at the critical depth, (m) = Vertical depth at the open hole kick zone, (m)
iv. Calculate the standpipe pressure when the kill weight mud reaches each of the criticaldepths. Pstand = P scr + Pfriction + Pstatic where: Pstand = Standpipe pressure, (psi) The results of the above calculations should be recorded in the Kick Sheet. These calculations should be carried out if the hole has a maximum angle greater than 30de grees. Figures 6.1a shows an example of a completed kick sheet for a high angle well. Figure6.1b sho ws the standpipe pressure schedule for pumping down the kill weight mud. It shows that the standpipe pressures required to maintain a constant bottom hole pressure are lower for a high angle well (with build-hold profile) than if the well was vertical. So if the standpipe pressure schedule for a vertical well was used (dotted straight line in Figure 6.1b), excessive high well pressures would result, which would increase the risk of fracturing the formation at the casing shoe or openhole weak point.
However the kick influx density/type (gas, water or oil) can not be determined based on the above data (as using the method described in Section 4.3). A gas kick is recognised when it is being circulated through the low angle or vertical hole section, where gas expansion causes a continuous increase in the casing pressure. So the first attempt to kill the well should be using one of the standard techniques.
6-13
Rev 1 March March 1995 1995
If the kick can not be circulated to surface using the standard techniques, it indicates that the kick influx is free gas which has been trapped in the inverted or horizontal hole section. To remove the trapped gas, the mud must be circulated with an annular velocity above a critical value. This critical annular velocity is about 100 ft/min when the hole angle is between 90~95 degrees. In an 8-1/2" hole, this corresponds to a critical flow rate of 4.6 bbl/min, which is higher than the normal range of SCR during a well control operation. So prior to drilling an inverted or horizontal hole section, the pump pressure at a SCR corresponding to 100~150 ft/min should be recorded in the kick sheet. The following procedures may be attempted to remove the trapped gas from the inverted or horizontal hole section: Start circulation with the original mud at a flow rate corresponding to an annular mud velocity of 100~150 ft/min until the entire horizontal hole section has been displaced; Reduce the flow rate to a normal SCR and proceed using one of the standard well killing techniques. After one complete circulation, stop the pump and shut in the well to check the pitgain. If there is still a positive pit gain, that indicates that some gas is still trapped downhole. Repeat the previous procedures.
In cases where the high flow rate can not be achieved to remove the trapped gas, consider bullheading the gas back into the formation. As the trapped gas should stay near the kick formation, bullheading is more likely to succeed in an inverted hole section. The bullheading technique is described in Section 6.2.
These problems are particularly acute in deep water. However, well control procedures should be modified in line with those described here, even in relatively shallow water, to take account of these problems. The drillpipe pressure is still used to monitor bottomhole pressure.
6-14
March Rev 1 1995 March 1995
Rigname 9-5/8"
On Order (MT):
Date:
27-Oct-94
Casing Burst (psi):
Time:
TD= 4291
1.44 40
Length (m)
DRILLSTRING CONTENTS
DP/DC Section ID Capacity (bbl/m) Vol (bbl) Cumulative Volume (bbl)
4.276 3 2.25
ID
5180 60 60
Length (m)
303.5 304.5
Cumulative Volume (bbl)
ANNULUS CONTENTS
4291 949 60
Length (m):
831.5 837.9
Vol (bbl):
2.9
1151 1501
Max Pres (psi): Max Pres (psi): PUMP 2 SPM bbi/min Pscr
5-1/2" 5-1/2"
4723 4723
Stroke Vol (bbl/stk): Stroke Vol (bbl/stk): TRAVEL TIMES (MIN / STROKES)
Surface to Bit
Bit to Shoe
30 40 50
Time of Kick:
30 40 50
268 / 8055 445 / 13368 201 / 8055 334 / 13368 161 / 8055 267 / 13368 0.98 20 55
350 psi
Shut-in DP Pres (psi), Pdp= Kill Mud Weight (sg), MW2= Chosen Pump SPM: Time Started:
460 81.1
2.565
0.086
SCR (bbl/min):
750
410
5" DP + BHA
4.276
5300
3.708
16.27
0 350 1328
350 1000
20 77
239 905
316 160
4 15
5300
1667
309
3612
60
Pfc=
410
6-15
March 1995
March 1995
800 750 700 650 600 550 500 450 400 Kick off
410
6-16
End Build
Bit 350 0 500 1000 1500 2000 Pump Strokes 2500 3000 3500 4000
SUMMARY OF FORMULAE
1. MAASP = P wp - 1.421 x MW 1 x D wp 2. Pipe Internal Capacity (bbl/m) = ID 2 / 313.8 3. Annular Capacity (bbl/m) = (DH 2 - DP 2) / 313.8 Pdp 4. Kill Mud Weight (sg), MW2 = MW 1 + 1.421 x TVDh (MW2 - MW1 ) 5. Baryte Required (lb/bbl) = 1490 x (4.25 - MW 2) 6. Initial Circulating Pressure (psi), Pic = P dp + P scr 7. Final circulating Pressure (psi), P fc = P scr MW 2 x MW 1 13. 10.
= L 1 / ID15 = L 2 / ID25
11. 12.
Pfc - Pscr Drillpipe Friction Constant, = 1+2 Static Pressure When Kill Mud at TVD (psi): Pstatic = P dp x (1.0 - TVD / TVDh ) Friction Pressure Increase When Kill Mud at MD (psi): a. When MD above/at DP cross-over point: MD Pfriction = x ID 15 b. When MD below DP cross-over point: (MD - L1) Pfriction = x [ 1 + ] ID 25
8. Pumping Time to Reach Depth of Interest (min): Volume to be displaced (bbl) = Pump output (bbl/min) 9. Pump Strokes to Reach Depth of Interest (stk): Volume to be displaced (bbl) = Pump stroke volume (bbl/stk) 14.
Pfriction
6-17
March 1995
/ 313.8
2
Annulus capacity [bbl/m] = [ (Hole size, inch) (Pipe OD, inch) 2 ] / 313.8 Volume [bbl] = (Capacity, bbl/m) x (Length, m) Total active mud volume [bbl] = (Total circulating system vol, bbl) + (Surface active mud vol, bbl)
/ (Pump output, bbl/min) / (Pump stroke volume, bbl/stk) / (Pump output, bbl/min) / (Pump stroke volume, bbl/stk) / (Pump output, bbl/min) / (Pump stroke volume, bbl/stk)
Bit Shoe [min] = (Total open hole annular volume, bbl) [stk] = (Total open hole annular volume, bbl) Shoe Choke [min] = (Total casing annular volume, bbl) [stk] = (Total casing annular volume, bbl)
KICK DATA
Record all the relevant kick data (time, hole depths, mud weight, shut-in DP & casing pressures, pit gain). All the kill parameters should be calculated.
Kill mud weight [sg] = (Mud weight in hole, sg) + [(SIDPP Pdp, psi)
(Kill mud weight, sg) (Mud weight in hole, sg) Barite required [lb/bbl] = 1490 x 4.25 (Kill mud weight, sg) Total quantity of barite required [MT] = (Total active mud volume, bbl) x (Barite required, lb/bbl) / 2205 Initial circulating pres P ic [psi] = (SIDPP Pdp, psi) + (SCR pres P scr , psi) Final circulating pres P fc [psi] = (SCR pres P scr , psi) x (Kill mud weight, sg)
6-18
March 1995
Calculate the pump stroke and the corresponding standpipe pressure when the kill mud has reached the depth at MD/TVD (the point for calculation such as kick off, end-build, etc.):
Volume [bbl] = (Drillstring capacity, bbl/m) x (Measured depth MD, m) Pump stroke [stk] = (Volume, bbl)
(TVD, m) Static (shut-in) pressure Pstatic [psi] = (SIDPP Pdp, psi) x [ 1.0 ] (Hole TVD, m) Friction pressure increase (due to kill mud) Pfriction : - If MD (point for calculation) is above or at DP1/DP2 cross-over point: P friction [psi] = (DP Friction Const, ) x (MD, m) - If MD is below DP1/DP2 cross-over point: [ (MD, m) (Top DP1 length L 1 , m) ] P friction [psi] = x [ (DP1 size factor, 1 ) + (DP2 ID, inch) 5 Standpipe Pressure Pstand [psi] = (SCR pres P scr , psi) + (Pfriction , psi) + (Pstatic , psi)
6-19
March 1995
Choke line losses are generally not significant at slow circulating rates in shallow water and so the calculations required during the implementation of both the Drillers Method and the Wait and Weight Method on a floating rig, drilling in shallow water, do not account for choke line losses. The calculations as covered in Paragraphs 4 to 6 (which cover the normal implementation of the Wait and Weight Method and the Drillers Method) are therefore still applicable. In deep water, when choke line losses can be significant, it is necessary to assess the effect of choke line losses on wellbore pressures during circulation. In which case further calculations, as covered in Paragraph 8, Accounting for Choke Line Losses in Deep Water, are required to account for choke line losses. Standard procedure (as detailed in Paragraphs 4 to 6) should be modified along the following lines when using either the Wait and Weight Method or the Drillers Method on a floatingr ig: 1 Bring the pump up to speed Line up to monitor wellhead pressure through the kill line. See Figure 6.2 for a schematicof the kill line monitor . (Bear in mind that the kill line may not contain mud at this stage.) Line up the pump to circulate down the drillpipe and route returns through the choke manifold to the mud gas separator. Set the stroke counter on the choke panel to zero. Record the pressure registered on the kill line monitor. Open the remote operated choke at the same time as the pump is started on the hole. Hold the kill line monitor pressure constant as the pump is brought up to speed. Once the pump is up to speed the initial circulating pressure should be checked.
Circulate the kick to the wellhead maintaining constant bottomhole pressure In the case of the Wait and Weight Method the standpipe pressure will be reduced in line with the standpipe pressure schedule. In the case of the Drillers Method the standpipe pressure is maintained at initial circulating pressure as the kick is displaced from the hole. When the total strokes pumped indicates that the influx is approaching the wellhead the kill line monitor should be carefully checked for any sudden drops in pressure. A drop in pressure registered on this gauge indicates that the influx has entered the choke line, however this drop may not always be detected.
Circulate the influx out of the well maintaining constant bottomhole pressure It is recommended that the influx is displaced up the choke line at a considerably reduced rate in order that the choke does not have to be adjusted at an unrealistic rate. This may involve shutting in the well at this point and restarting the displacement at the minimum pump speed. A considerable increase in choke pressure will generally be required as gas or lightweight influx displaces mud from the choke line.
6-20
March 1995
Figure 6.2 Use of Kill Line Monitor for Wellhead Pressure on Floating Rig
DRILLPIPE PRESSURE GAUGE
PUMP
VALVE CLOSED
VALVE OPEN
RETURNS
SEA
CHOKE LINE
SEABED
6-21
Rev 1 March March 1995 1995
An increase in the pressure recorded at the kill line monitor may indicate that the original mud behind the influx has started up the choke line. In the case of the Wait and Weight Method, once the returns are at kill weight, the pump should be stopped and the well checked for pressure. In the case of the Drillers Method, the well will be circulated to kill weight mud prior to step (4). 4 Remove BOP gas It is quite possible that some gas will have accumulated under the closed BOP duringdisplacement of the kic k. This gas must be removed from the stack before the BOP is opened. The recommended technique is to isolate the well, displace the kill and choke lines to water (maintaining the BOP gas at original pressure), bleed gas up choke line, open the annular and allow riser to U-tube, displacing the gas up the choke line. Diesel may be used instead of water if low mud weights have been used to kill the well. (Adequate facilities should be available to deal with the returned diesel.) For the example stack shown in Figure 6.4, for which trapped gas has the potential to be a serious problem, this technique is implemented as follows: Isolate the well from the BOP stack by closing the lower pipe rams. (See Figure6.4.) Circulate kill mud down the kill line, across the stack and up the choke line. Route r eturns through the degasser. Record the kill line circulating pressure. (SeeFigure 6.5.) Shut the well in. Line up to circulate water down the kill line and up the choke line. Slowly displace the kill line to water. As the kill line is displaced to water increase the kill line circulating pressure by an amount equal to the difference in hydrostatic pressure between the kill mud and water at the depth of the stack. (This will ensure that the gas pressure is unchanged.) Keep pumping water across the stack and maintain the final circulating pressure.W hen the returns are clear water, stop the pump and shut in at the choke. (See Figure6.6.) Close the subsea kill line valve(s). Bleed pressure from the choke line. (See Figure 6.7.) (The pressure that has been trapped in the gas bubble is used to ensure that the gas bubble expands as the choke is opened to displace all the water from the choke line. Having bled all the pressure from the choke line the gas bubble should be almost at atmospheric pressure.) Close the diverter. Line up the trip tank/pump to circulate the riser under the diverter. Slowly bleed back the upper annular closing pressure. Open the annular. Allow the riser to U-tube. Take returns up the choke line. Fill the hole as required. (See Figure 6.8.) Be prepared to deal with gas in the riser. Displace the riser and kill and choke lines to kill weight mud.
6-22
March 1995
Open the lower pipe rams. Open the diverter and flowcheck the well.
6-23
March 1995
Figure 6.3 Subsea BOP Stack prior to Removing Gas from Below the Preventers
KILL LINE
CHOKE LINE
MUD
UPPER ANNULAR
GAS
VALVE OPEN
VALVE CLOSED
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.031
6-24
March 1995
Figure 6.4 Removing Gas from a Subsea BOP Stack Lower pipe rams closed hang off rams opened
KILL LINE
CHOKE LINE
MUD
UPPER ANNULAR
GAS
VALVE OPEN
VALVE CLOSED
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.032
6-25
March 1995
Figure 6.5 Removing Gas from a Subsea BOP Stack Kill and choke lines displaced to kill weight mud
KILL LINE
CHOKE LINE
MUD
UPPER ANNULAR
GAS
VALVE OPEN
VALVE CLOSED
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.033
6-26
March 1995
Figure 6.6 Removing Gas from a Subsea BOP Stack Kill and choke lines displaced to water
KILL LINE
CHOKE LINE
MUD
GAS
UPPER ANNULAR
VALVE OPEN
VALVE CLOSED
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.034
6-27
March 1995
Figure 6.7 Removing Gas from a Subsea BOP Stack Gas pressure bled down, gas occupies choke line
KILL LINE
CHOKE LINE
MUD
GAS
UPPER ANNULAR
VALVE OPEN
VALVE CLOSED
GAS PRESSURE BLEEDS DOWN TO DISPLACE WATER FROM CHOKE LINE RESULTANT GAS PRESSURE IS CLOSE TO ATMOSPHERIC
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.035
6-28
March 1995
Figure 6.8 Removing Gas from a Subsea BOP Stack Diverter is closed, the annular is opened and the gas is displaced from the stack
KILL LINE
CHOKE LINE
MUD
GAS
UPPER ANNULAR
VALVE OPEN
VALVE CLOSED
LOWER ANNULAR
BLIND/SHEAR
PIPE RAM
PIPE RAM
PIPE RAM
PIPE RAM
WEOX02.036
6-29
March 1995
Figure 6.9 The Effect of Choke Line Losses Casing pressure greater than choke line pressure
800
400
800
800
1385
430
SCRS AND CHOKE LINE LOSSES SPM PSCR PCL 20 400 150 30 680 250 40 985 370
WEOX02.037
6-30
March 1995
Figure 6.10 The Effect of Choke Line Losses Casing pressure after initial circulation is less than choke line loss
INITIAL SHUT-IN CONDITIONS CIRCULATION STARTED AT 30SPM
400
100
400
400
780
150
DRILLPIPE PRESSURE EQUALS THE SUM OF THE ORIGINAL SHUT-IN DRILLPIPE PRESSURE PLUS THE SCR PRESSURE PLUS THE CHOKE LINE LOSS PLUS THE WIDE OPEN CHOKE PRESSURE MINUS THE SHUT-IN CASING PRESSURE
100
100
100
200
600
50
UNABLE TO KEEP THE KILL LINE PRESSURE CONSTANT. EVEN WITH THE CHOKE WIDE OPEN THE KILL LINE PRESSURE INCREASES BY THE SUM OF CHOKE LINE LOSS AND WIDE OPEN CHOKE PRESSURE MINUS THE ORIGINAL SHUT-IN PRESSURE
SCRS AND CHOKE LINE LOSSES SPM PSCR PCL 20 400 150 30 680 250 40 985 370 KEY MUD GAS
WEOX02.038
6-31
March 1995
These pressures as well as the annulus frictional pressure will act at all points in the wellbore and circulating system. The effect of these additional pressures must therefore be analysed at all points in the system and in particular at the openhole weak point. 2 Calculate the initial circulating pressure The initial circulating pressure is calculated to estimate the standpipe pressure once the pump is up to speed. For Case A: For Case B: where: Pscr Pdp Pcl Pa Poc the initial circulating pressure = Pdp + P scr the initial circulating pressure = Pdp + P scr + P cl + P oc P a = show circulating rate pressure (psi) = shut-in drillpipe pressure that reflects the kick zone pressure (psi) = choke line frictional pressure at SCR (psi) = annulus shut-in pressure (psi) = choke pressure recorded while circulating at SCR with the choke wide open (psi)
Calculate the final circulating pressure The final circulating pressure, when kill weight mud reaches the bit, for each case is calculated as follows: For Case A: Final circulating pressure = Pscr
X
MW2 = weight of the kill mud (SG) MW1 = weight of the original mud (SG)
4 Monitor pressure at the kill line monitor as the pump is brought up to speed For Case A, the pressure at the kill line monitor is held constant as the pump is brought up to speed. The choke pressure will decrease by an amount equivalent to the choke line friction pressure once the pump is up to speed. For Case B, the pressure at the kill line monitor will be constant as the pump is brought up to speed. However at some point before the pump is up to the SCR the kill line monitor pressure will start to increase. Once the pump is up to speed the choke will be wide open and the pressure at the kill line monitor will have risen by the proportion of the choke line friction pressure that is not compensated for. (The increase will be equivalent to Pcl + P oc P a.) 5 Check the initial circulating pressure once the pump is up to speed If the initial circulating pressure is significantly different from the calculated value, the pump should be stopped, the well shut in and the cause for the discrepancy determined. If the initial circulating pressure is equal to or reasonably close to the calculated value, the displacement should be continued. Any marginal difference is likely to be due to the fact that the actual SCR pressure is different from the value used to calculate the initial circulating pressure. The actual SCR pressure can be established from the initial circulating pressure recorded when the pump is up to speed.
6-32
March 1995
For Case A, the actual SCR pressure can be determined from the initial circulating pressure as follows: P scr = Pic P dp For Case B, the actual SCR pressure can be determined as follows: P scr = Pic P dp P cl P oc + P a For the Wait and Weight Method the final circulating pressure must be recalculated as follows: For Case A, the final circulating pressure can be determined as follows: Pfc = Pscr
X
MW2 MW1
For Case B, the final circulating pressure is determined as follows: Pfc = (Pscr
X
The standpipe pressure should therefore be redrawn to take into account these adjusted figures. 6 Assess the effect of choke line losses at the latter stages of kick displacement For Case A: In the latter stages of the displacement the choke pressure required to maintain constant bottomhole pressure will drop. This drop will be most significant once the original mud behind the influx is at the choke. If the required choke pressure drops below the sum of the choke line loss and the wide open choke pressure, it will no longer be possible to completely compensate for the choke line losses. The resultant increase in wellbore pressure at this stage will be given by: Increase in pressure = Pcl + P oc P a In practice, the choke will be wide open at this stage and the standpipe pressure will rise above final circulating pressure. When the hole has been circulated to kill weight mud, the circulating pressure will have increased by the sum of the choke line losses and the wide open choke pressure. For Case B: As the influx expands the choke pressure required at surface will increase. As the required choke pressure increases it will be possible to compensate for a greater proportion of the choke line losses. If the required choke pressure increases to a value equal to the sum of the choke line loss and the wide open choke pressure it will be possible to compensate for the complete amount of the choke line losses. It should be noted that the most critical period in terms of downhole pressures is likely to occur at early stages in the displacement. In this respect the change in choke line loss compensation at latter stages in the displacement is unlikely to be a critical factor.
6-33/34 6-33
March 1995
6.2
SPECIAL TECHNIQUES
Page
Subsection 2.1 VOLUMETRIC METHOD 2.2 STRIPPING 2.3 BULLHEADING 2.4 SNUBBING 2.5 BARYTE PLUGS 2.6 EMERGENCY PROCEDURE
6-35/36 6-35
March 1995
6.2
Paragraph
1 2 General Static Volumetric Method (Drillpipe pressure used to monitor bottomhole pressure) Static Volumetric Method (Choke pressure used to monitor bottomhole pressure) Lubrication Dynamic Volumetric Control
6-38
4 5
Illustrations
6.12 Static Volumetric Method an example of control of bottomhole pressure at the choke 6.13 Static Volumetric Control illustrating the consequences of improper procedure 6.14 Volumetric Control Worksheet an example for a land rig 6.15 Static Volumetric Method choke pressure used to monitor bottomhole pressure 6.16 Dynamic Volumetric Method used to remove gas from below a stack
6-37
March 1995
1 General
The Volumetric Method can be used to control the expansion of an influx that is migrating during shut-in periods. It can therefore only be used if significant migration is occurring. This may occur only in the case of gas kicks. This method can be used during shut-in periods prior to displacement, or as a means of safely venting an influx from a well in which circumstances prevent the implementation of normal well control techniques. Situations in which the Volumetric Method may be applicable therefore include: During any shut-in period after the well has kicked. If the pumps are inoperable. If there is a washout in the drillstring that prevents displacement of the kick. If the pipe is a considerable distance off bottom, out of the hole or stuck off bottom. If the bit is plugged. If the pipe has been dropped.
There are four techniques that may be required to deal with an influx that is migrating up the hole. These are as follows: Static Volumetric Control: When the drillpipe is on or near bottom and can be used to measure bottomhole pressure. Static Volumetric Control: When the drillpipe cannot be used to measure bottomhole pressure. Lubrication: When the influx has migrated to the stack this technique is used to replace the influx with mud as the influx is bled at the choke. Dynamic Volumetric Control: This technique may be used as an alternative to the above but is most applicable as an alternative to lubrication on a floating rig.
The following Paragraphs can be used as guidelines for the implementation of the above mentioned procedures.
6-38
March 1995
The following guidelines can be used: 1 Record the shut-in drillpipe and choke pressures After the well has been shut-in the surface pressures can be used to identify the influx type. These calculations are covered in Chapter 4. If the influx contains a significant proportion of gas, it will be necessary to allow the influx to expand considerably as it migrates up the hole. 2 Develop annulus pressure profile The annular pressures during migration of the influx will be similar to those resulting from circulation with the Drillers Method. In this respect, a PC or programmable calculator can be used to develop the annulus pressure profile as for the Drillers Method. The maximum wellbore pressures can therefore be estimated along with the anticipated pit gain. 3 Determine migration rate After the surface pressures have built up to values which reflect the kick zone pressure, further increases will be due to migration. The rate of migration can be estimated from two pressure readings, recorded either both on the drillpipe or both on the casing, taken at a known time interval apart. The distance D (m), migrated up the annulus of constant cross section in the time interval T (min) is given by: D= P2 P1 MW X 1.421 = = = = = (m)
where P1 P2 MW T MR
surface pressure at start of interval (psi) surface pressure after interval T (psi) mud weight in the hole (SG) time interval (min) migration rate (m/hr)
60
(m/hr)
T Allow drillpipe pressure to build by overbalance margin The drillpipe pressure should be allowed to build by a suitable overbalance margin. This margin will be registered on the drillpipe as an increase in pressure over and above the final shut-in pressure. The overbalance margin may typically be in the range 50 to 200 psi. 5 Allow drillpipe pressure to build up by operating margin The drillpipe pressure should be allowed to build by a further margin to ensure that the overbalance is maintained as mud is bled from the well. The operating margin may also typically be in the range 50 200 psi depending on the resultant wellbore pressures at each stage in the operation.
6-39
March 1995
Bleed increment of mud from the annulus to reduce drillpipe pressure After the drillpipe pressure has built by the sum of the overbalance margin and the operating margin, the kick zone will be overbalanced by the sum of these two values. Mud should then be bled from the annulus to reduce the drillpipe pressure to a value representing the final shut-in pressure plus the overbalance margin. A manual choke should be used for this operation to ensure adequate control. It is strongly recommended that small volumes of mud are bled off at a time to allow time for the drillpipe pressure to respond. There will be a considerable delay time between choke and drillpipe pressure in a deep well and especially if the influx contains gas.
Continue process until influx migrates to the stack This process should be repeated until the influx migrates to the stack. Arrival of the influx at the stack may be preceded by bleeding gas cut mud from the well. However, if gas is observed at the choke, the well should be shut-in and mud lubricated into the well. If gas is bled from the well the bottomhole pressure will drop and eventually cause a further influx. When the influx has migrated to the stack, surface pressures should no longer rise as migration will cease to occur. This may not be the case on a floating rig when some migration may occur up the choke line. Use the Volumetric Control Worksheet to record all the relevant data (See Figure 6.14.)
Lubricate mud into the hole or implement the Dynamic Volumetric Method See Paragraphs 4 and 5 as follows.
6-40
March 1995
For example, if the choke pressure increases by 100 psi, a volume of mud equivalent to a hydrostatic pressure in the annulus of 100 psi is bled at the choke at constant choke pressure. In this manner, control over the bottomhole pressure is achieved. It should be noted that this method is only applicable if the influx is migrating as the mud is bled from the well . The rate of influx migration determines the time required to bleed each increment of mud from the well. Figure 6.12 illustrates this technique. In this example, the following conditions apply: Operating margin = 150 psi Annulus = 8 1/2 in. Mud weight = 1.85 SG
X
Hydrostatic equivalent of mud = 445.7 1.85 (72.25 25) Therefore bleed 150 = 8.5 bbl of mud 17.5
As can be seen from Figure 6.12, the influx must migrate (1824 133 =) 1691m while the 8.5 bbl of mud is bled from the well. It is clear that this operation will take several hours. If the operating margin was quickly bled from the well, the original influx would expand by approximately 0.4 bbl before the bottomhole pressure drops to the original kick zone pressure. If the remaining 8.1 bbl were bled from the well, this would cause a further influx of 8.1 bbl, as shown in Figure 6.13. As the influx migrates further up the hole, the time required to bleed the 8.5 bbl increment from the well will decrease significantly. In this example, the influx must migrate 570m (approximately 2 hours) as the next increment is bled from the well. If the rate of influx migration is maintained, this time will continually reduce until the influx is at surface. Volumetric control is similar to the Drillers Method although the influx moves up the hole under the influence of migration. The resultant wellbore pressures as well as the required pit gain will be similar for the two techniques. The following guidelines can be used: 1 2 3 Record shut-in choke pressure Develop annulus pressure profile Determine migration rate The first three steps are carried out in the same manner as for the previous technique. 4 Calculate hydrostatic pressure of mud per barrel The hydrostatic pressure of the mud per barrel should be calculated at the point in the annulus directly above the influx. It can be calculated as follows: Hydrostatic pressure per barrel = 445.7 X MW (d hc 2 d o2 ) (psi/barrel)
where MW = mud weight in the hole (SG) dhc = hole/casing ID (in.) do = drillstring OD (in.)
6-41
March 1995
Figure 6.12 Static Volumetric Method an example of control of bottomhole pressure at the choke
1. AT INITIAL SHUT-IN 2. INCREASE IN SURFACE PRESSURE FOR OVERBALANCE MARGIN 850psi Pa
650psi Pa
PRESSURE IN BUBBLE 10,000psi HEIGHT OF INFLUX 66m VOLUME OF INFLUX 10bbl DEPTH 3615m
76m
PRESSURE IN BUBBLE NOW 5405psi PRESSURE IN BUBBLE 10,000psi VOLUME OF INFLUX 10bbl KEY MUD 133m GAS
1824m
6-42
March 1995
Pa > 1000psi
WEOX02.040
Figure 6.13 Static Volumetric Control illustrating the consequences of improper procedure
5 Allow choke pressure to build by overbalance margin The choke pressure should be allowed to build by an overbalance margin that may typically be in the range 50 200 psi. 6 Allow choke pressure to build by operating margin The choke pressure should be allowed to continue building a further similar amount to provide an operating margin. The total margin will depend on the resultant wellbore pressures at each stage in the operation. 7 Bleed increment of mud from the well at constant choke pressure A suitable volume of mud should be bled from the well to reduce the bottomhole pressure by an amount equivalent to the operating margin.
6-43
March 1995
UK
DATE AND TIME 15:30 1.85 ANNULUS: ANNULUS: psi/bbl psi/bbl 150 Hydrostatic of Mud Bled/ Lubricated (psi) 0 0 0 -150 0 -150 0 -150 0 -150 psi Overbalance P1
WELL NO
26
RIG
20/08/95
SHEET NO
MUD WEIGHT IN THE HOLE, sg HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF OVERBALANCE MARGIN: TIME 200 psi
LUBRICATING MUD WEIGHT, sg sg MUD in sg MUD in sg MUD in sg MUD in 5 X X HOLE: HOLE: 8.5
17.46
psi/bbl psi/bbl
Migration Rate 2
Time (min) P2
Rate, (mpm)
20 5 11.2311
OPERATING MARGIN: Choke or DP Choke Monitor Pressure (psi) 650 Change in Monitor Pressure (psi) 0 200 150 0 150 0 150 0 150 0
OPERATION If DP pressure can't be read see page 6-36 of Vol. 1 of BP Well Control Manual
( hr
min)
(psi) 0 200 350 200 350 200 350 200 350 200
Total Volume of Mud (bbl) 100 100 100 108.5 108.5 117 117 125.5 125.5 134
19:00 19:15 19:25 19:25 / 01.25 1:35 1:35 / 3:15 3:30 3:30 / 4:45 4.:55 4:55 / 5:30 Influx Migrating Influx Migrating Bleed Mud at Choke Influx Migrating Bleed Mud at Choke Influx Migrating Bleed Mud at Choke Influx Migrating Bleed Mud at Choke
+ ve increase
- ve bled
+ ve overbalance - ve underbalance
+ ve bled
- ve decrease
+ ve lubricated
- ve lubricated
WEOX02.199
6-44
March 1995
The choke pressure must be held constant as the mud is bled from the well. As an example (refer to Figures 6.12 and 6.13): Operating margin = 150 psi Annulus = 8 1/2 in. X 5 in. Mud weight = 1.85 SG Hydrostatic equivalent of mud = Therefore bleed 150 17.5 445.7 1.85 = 17.5 psi/bbl (72.25 25)
As can be seen from the example in Figure 6.12 the bottom of the influx has had to migrate from 133m off bottom, to 1824m off bottom, whilst bleeding off 8.5 bbl of mud. This could take considerable time. If the operating margin, in this case 150 psi (8.5 bbl), had been quickly bled off and assuming no migration during this period, the bubble would have expanded by only about 0.36/bbl before bottomhole pressure (BHP) dropped to kick zone pressure. This would result in a further influx of 8.14 bbl. Subsequent volumes bled from the well will require less migration distance, ie for anincrease of bubble size to 27 bbl (after next bleed off), the distance from bottom will be 2395m.
2200 PRESSURE BUILDUP GAS MIGRATING TO SURFACE
1750
1600
MUD BLED AT CHOKE (at constant choke pressure until volume bled off corresponds to Operating Margin) OPERATING MARGIN
1450
1300
1150 OPERATING MARGIN 1000 OPERATING MARGIN 850 OVERBALANCE MARGIN 650 FINAL SHUT-IN ANNULUS PRESSURE
8.5
17
25.5
34
42.5
51
59.5
68
76.5
WEOX02.041
Figure 6.15 Static Volumetric Method choke pressure used to monitor bottomhole pressure
6-45
March 1995
Continue the process until the influx migrates to the stack This process should be repeated until the influx migrates to the stack. When the influx has migrated to the stack surface pressures should no longer rise as migration will cease to occur. This may not be the case on a floating rig when some migration may occur up the choke line. Use the Volumetric Control Worksheet to record all the relevant data. Figure 6.14 shows a completed example.
Lubricate mud into the hole or implement the Dynamic Volumetric Method See Paragraphs 4 and 5. If this process has been implemented because the pipe was off bottom, it may be feasible to circulate the influx out of the hole when the influx has migrated to the bit. See Figure 6.15 for a typical choke pressure schedule for the Static Volumetric Method.
4 Lubrication
This technique may be used to vent the influx from below the stack while maintaining constant bottomhole pressure. Lubrication is most suited to fixed offshore and land rigs. It can be used to vent gas from the stack after implementing the Static Volumetric Method, as well as to reduce surface pressures prior to an operation such as stripping or bullheading. Lubrication is likely to involve a considerable margin of error when implemented on a floating rig because of the complication of monitoring the bottomhole pressure through the choke line. When the influx has migrated to the stack it is quite possible that the choke line will become full of gas cut mud. In this situation it is impractical to attempt to maintain control of the bottomhole pressure with the choke. However lubrication is simpler to implement than the Dynamic Volumetric Method. For this reason alone, it may be considered for use on a floating rig. The following guidelines can be used to lubricate mud into a well: 1 Calculate the hydrostatic pressure per barrel of the lubricating mud This is done in the same manner as for the Volumetric Method. 2 Slowly lubricate a measured quantity of mud into the hole Line up the pump to the kill line. Having determined the safe upper limit for the surface pressure, the pump should be started slowly on the hole. Mud should be lubricated into the well until pump pressure reaches a predetermined limit. At this point the pump should be stopped and the well shut in. The well should be left static for a period while the gas migrates through the mud that has been lubricated into the well.
6-46
March 1995
The exact amount of mud lubricated into the well should be closely monitored. 3 Bleed gas from the well Gas should be bled from the well to reduce the surface pressure by an amount equivalent to the hydrostatic pressure of the mud lubricated into the well. If the surface pressure increased as the mud was lubricated into the well, the amount that the pressure increased should be bled back in addition. Ensure that no significant quantity of mud is bled from the well during this operation. If mud appears at the choke before the surface pressure has been reduced to its desired level, shut the well in and let the gas percolate through the mud. Returns should be lined up through the mud gas separator to the trip tank to ensure that any volume of mud bled back with the gas is recorded and accounted for. 4 Repeat this procedure until all the influx has been vented from the well This procedure should be repeated until all the gas has been vented from the well. It is likely that it will be necessary to reduce the volume of mud lubricated into the well at each stage during this procedure. This is due to the reduction in volume of gas in the well. If the influx was swabbed into the well and the mud weight is sufficient to balance formation pressures, the choke pressure should eventually reduce to zero. However, if the mud weight in the hole is insufficient, the final choke pressure will reflect the degree of underbalance. It will then be necessary to kill the well.
6-47
March 1995
Circulate down the kill line and up the choke line Ensure that it is possible to monitor the active pit level accurately. Route returns through the mud gas separator.
Bring the pump up to speed As the pump is brought up to speed, the kill line (or pump pressure) must increase by an amount equal to the kill line pressure loss. However if it is not possible to compensate for the choke line pressure loss, the kill line pressure will inevitably increase by more than the kill line pressure loss. The kill line circulating pressure will be monitored during the operation to remove gas from the well.
Reduce kill line pressure in line with drop in pit level As gas is bled from the well, the pit level will drop while the choke operator adjusts the choke to maintain a constant kill line circulating pressure. This will result in mud being lubricated into the well. If the kill line circulating pressure is held constant as mud is lubricated into the well (as gas is removed), the bottomhole pressure will increase. Therefore, as the pit level decreases, the kill line pressure should be reduced to account for the greater hydrostatic pressure in the annulus. As an example: Drop in pit level = 10 bbl Annulus = 8 1/2 in. X 5 in. Mud weight = 1.85 SG Hydrostatic equivalent of mud = 445.7 X 1.85 (72.25 25) = 17.5 psi/bbl
Therefore reduce kill line circulating pressure by 17.5 X 10 = 175 psi This procedure should be continued until all the influx has been vented from below the stack. This will be indicated by a constant pit level. If the well has been completely killed by removing gas from the stack, the final circulating kill line pressure will be equal to the sum of the kill line pressure loss, the choke line pressure loss and the wide open choke pressure. If the well is not yet completely killed at this point, the final circulating kill line pressure will be greater than this value. See Figure 6.16 for an example kill line pressure schedule for this technique.
6-48
March 1995
Figure 6.16 Dynamic Volumetric Method used to remove gas from below a stack
6-49/50 6-49
March 1995
6.2
Paragraph
1 2 3 4 5 6 7 General Monitoring Well Pressures and Fluid Volumes Annular Stripping Annular Stripping Procedure Ram Combination Stripping Ram Combination Stripping Procedure Dynamic Stripping Procedure
Illustrations
6.17 A Guide to Interpretation of Surface Pressure Changes during Stripping 6.18 The Effect of the Pipe/BHA Entering the Influx 6.19 Surge Dampener Fitted to the Closing Line of an Annular BOP 6.20 Example Stripping Worksheet showing effect of migration and BHA entering the influx 6.21 Surface BOP Stack Suitable for Ram Combination Stripping 6.22 to 6.25 Annular to Ram Stripping 6.26 Equipment Rig-up for Dynamic Stripping
6-51
March 1995
1 General
Stripping is a technique that can be used to move the drillstring through the BOP stack when the well is under pressure. Stripping places high levels of stress on the BOPs and the closing unit, and requires a particularly high level of co-ordination within the drillcrew. Company policy is that a contingency plan must be developed regarding stripping procedure for both Company operated rigs and rigs that are under a Company contract. This Section is intended to aid in the drawing up of this contingency plan and as such the following are proposed as the most important considerations: How to move the tool joint through the BOP. Wear on BOP elements and the control unit. The level of redundancy in the BOP and the control system. Wellbore pressures in relation to the maximum allowable pressure for equipment and the formation. The monitoring of pressure and fluid volumes. The organisation and supervision of the drillcrew. Controlling increases in wellbore pressure due to surge pressure. The condition of the drillpipe. (Drillpipe rubbers should be removed and any burrs smoothed out.) The possibility of sticking the pipe. The control of influx migration. Manufacturers information regarding minimum closing pressures for annular preventers. (This information should be available at the rig site.) The procedure to be adopted in the event that the surface pressure approaches the maximum allowable as the pipe is stripped into the influx.
See Figure 5.2 in Chapter 5 for a decision analysis related to stripping operations.
6-52
March 1995
(a) To compensate for the volume of pipe introduced into the hole
To avoid over pressuring the well, a volume of mud equal to the volume of pipe and tool joints (the volume of metal plus the capacity) introduced into the well, must be bled off. Where possible, mud should not be bled from the well while the pipe is stripped in. It is recommended that mud is bled from the well during each connection. This ensures that there is a clear indication at surface of the BHA entering the influx. However it is recognised that there may be situations when it is impractical to bleed mud from the well at connections. Such situations include: If the surface pressures are close to maximum allowable prior to the stripping operation. If a high pressure water kick is taken. In these circumstances the effective compressibility of the fluid in the hole will be low and as such there may be a very large pressure rise as pipe is stripped into the well. If the pipe has to be stripped out of the hole. In this case, there will be a tendency for the volume of metal removed from the well to be replaced by influx fluid.
(c) To allow an increase in surface pressure as the BHA enters the influx.
When the BHA is run into the influx, the height of the influx will be considerably increased. This can cause a significant decrease in hydrostatic pressure in the annulus, requiring a greater surface pressure to maintain a constant bottomhole pressure (See Figure 6.18). A potential problem arises if this condition is undetected. The choke operator may continue to bleed mud from the well to maintain a constant surface pressure and inadvertantly cause further influx into the wellbore. It is therefore important to accurately monitor the total volume of mud bled from the well. It is recommended that the potential increase in surface pressure resulting from entering the influx should be estimated before stripping into the hole.
6-53
March 1995
START STRIPPING IN
CONTINUE STRIPPING
NO
YES NO
CONTINUE STRIPPING
CONTINUE STRIPPING
NO
HAS THE CORRECT VOLUME OF MUD BEEN BLED FROM THE WELL?
YES
YES
INFLUX IS MIGRATING
NO
NO
YES
YES
WEOX02.043
6-54
March 1995
1. Start stripping
2.
Height of influx in annulus has increased Overall hydrostatic in annulus decreases Surface pressure required to balance formation pressure increases
GAS INFLUX
MUD
WEOX02.044
6-55
March 1995
3 Annular Stripping
There are two stripping techniques, Annular and Ram combination stripping. The decision analysis presented in Chapter 5 , Pipe off Bottom Drillpipe in the Stack outlines the basis upon which the most suitable stripping technique is selected. Annular stripping is considered to be the most satisfactory technique. It involves less risk than ram combination stripping for the following reasons: Annular stripping is a relatively simple technique. During annular stripping the only item of well control equipment that is subject to high levels of stress is the annular element. The control system is not highly stressed during the operation (as is the case during ram combination stripping). The annular element can be changed out on a surface stack when pipe is in the hole by inserting a split element. The upper annular preventer, on a floating rig, is the only stack component that is subject to wear and this can be changed without pulling the complete BOP stack.
Ram combination stripping is possible on all types of rig but involves significantly more risk when implemented on a floating rig. The surface pressure is the overriding factor which determines whether or not it will be possible to implement annular stripping. However, it is also necessary to consider that the operating life of an annular element is severely reduced by increased wellbore pressure. Field tests* carried out on Hydril and Shaffer 5K Annulars, show good performance at 800 psi wellbore pressure, but at 1500 psi and above the performance was severely reduced and unpredictable. If surface pressures indicate that annular stripping is not possible, attempts should be made to reduce the pressures in order to enable annular stripping to be used. The most appropriate technique will depend on the position of the influx in the hole. The options are; to circulate out the influx, to lubricate the influx from the well or to bullhead. To ensure that the annular is not subjected to excessive pressures as the tool joint is stripped through the element, a surge dampener must be placed in the closing line (See Figure 6.19). This may not be necessary on a surface stack if the pressure regulator can respond fast enough to maintain a constant closing pressure as a tool joint is stripped through the annular. As a word of caution, some drilling contractors have installed check valves in the control lines to the BOPs; the purpose being to ensure that the BOP stays closed if the hydraulic supply is lost. However, if a check valve is installed in the closing line to an annular BOP, it will not be possible to reduce the closing pressure once the annular has been closed. In order to reduce the annular closing pressure, in this case, it will be necessary to open the annular having closed another ram to secure the well. * Tests carried out by Exxon Prod. Research 1977.
6-56
March 1995
OPENING LINE
CLOSING LINE
WEOX02.045
Figure 6.19 Surge Dampener Fitted to the Closing Line of an Annular BOP
6-57
March 1995
2 Monitor surface pressures Surface pressures should be monitored after the well has been shut-in to check for influx migration. If the influx is migrating it will be necessary to implement volumetric control during the stripping operation. If the pipe is off bottom, it will not be possible to identify the type of influx in the usual manner. However, a high surface pressure caused by a relatively small underbalance usually indicates that the influx contains a significant quantity of gas. 3 Determine the capacity and displacement of the drillpipe It will be necessary to bleed mud from the well to compensate for the volume of pipe introduced into the hole. This volume is equal to the sum of the capacity and the displacement of the pipe. There are various tables which outline these quantities, but a reasonable estimation can be made as follows: Displacement and capacity = do2 X 0.003187 (bbl/m) where do = outer diameter of the pipe (in.) Allowance should also be made for the extra volume of metal in the tool joints. 4 Calculate hydrostatic pressure per barrel of mud Should migration occur, it will be necessary to bleed from the well at constant choke pressure to allow the influx to expand. The hydrostatic pressure equivalent of the mud in the hole is calculated as follows: Hydrostatic pressure equivalent = 445.7 X MW (d hc2 d o2) where MW = mud weight in the hole (SG) dhc = hole/casing ID (in.) do = drillstring OD (in.) or if the pipe is above the influx: Hydrostatic pressure equivalent = 445.7 X MW dhc 2 (psi/bbl) (psi/bbl)
For more details on this technique, See Sub-section 2.1 Volumetric Method in this chapter. 5 Estimate increase in surface pressure due to BHA entering the influx It is possible to estimate the maximum possible pressure increase due to the BHA entering the influx as follows: Max possible surface = 445.7 pressure increase where MW Gi V dhc do = = = = =
X
(MW G i)
1 (d hc2 d o2)
1 dhc 2
(psi)
mud weight in the hole (SG) influx gradient, converted to SG (water = 1 SG) volume of influx (bbl) hole/casing ID (in.) BHA OD (in.)
6-58
March 1995
Allow surface pressure to increase by overbalance margin An overbalance of 50 to 200 psi should be maintained throughout the stripping operation. If the influx is not migrating, the overbalance margin can be applied by bleeding a volume of mud that is less than the volume of pipe introduced into the hole, at the start of the operation.
Reduce annular closing pressure The BOP manufacturers recommend that the closing pressure is reduced, prior to stripping, until a slight leakage occurs through the BOP. This reduces the wear on the annular by lubricating the element during stripping.
Strip in the hole The pipe should be slowly lowered through the annular while the surface pressure is accurately monitored. The running speed should be reduced when a tool joint passes through the annular. Mud should be bled from the well at each connection, unless surface pressure limitations dictate that this should be carried out more frequently. The pipe should be filled with mud at suitable intervals, typically every 5 stands. Use original mud weight. A person should be posted at the Drillers BOP Control Panel at all times to be ready to shut-in the well in the event of failure of the annular preventer.
Monitor surface pressure Surface pressures and all relevant data should be recorded on the Stripping Worksheet. (See Figure 6.20.) Use Figure 6.17 as an aid to the interpretation of changes in surface pressure.
10 Strip to bottom. Kill the well The only sure method of killing the well will be to return the string to bottom and implement standard well kill techniques.
6-59
March 1995
Figure 6.20 Example Stripping Worksheet showing effect of migration and BHA entering the influx
STRIPPING WORKSHEET
For worksheet calculation enter information in shaded cells.
WELL NO 3 1.75 2000 RIG Rig 10
Units (US/UK)
uk
10/7/87 10:30 1.75
MUD WEIGHT IN HOLE INITIAL BIT DEPTH STRIPPING DATA VOLUME OF MUD DISPLACED BY OVERBALANCE MARGIN VOLUMETRIC CONTROL DATA
5 120 psi
Inch Pipe
Drillpipe
bbl/m 150
: psi (Max)
2.15
bbl/stand
HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF HYDROSTATIC PRESSURE PER BARREL OF TIME OPERATION
5 6.5 8.5
x x HOLE
8.5 8.5
ANNULUS ANNULUS 10.80 10.80 psi/bbl psi/bbl Hydrostatic of Mud Bled/ Lubricated Overbalance
( hr 10:05
(psi) 550
(psi)
bbl
(psi) N/A
(psi)
(bbl)
(bbl)
2000
Drillpipe Dart Installed Strip in Stand No 1 Strip in Stand No 2 Bleed Mud at Connection Strip in Stand No 3 Bleed Mud at Connection Strip in Stand No 4 Bleed Mud at Connection Strip in Stand No 5 (Assume BHA has entered flux) 770 890 770 890 770 890 770 950 120 120 -120 120 -120 120 -120 180
2000 2027 2054 2054 2081 2081 2108 2108 2135 27 54 54 81 81 108 108 135 2.2 4.4 4.4 6.6 6.6 8.8 8.8 11.0 N/A N/A N/A N/A N/A N/A N/A N/A 120 240 120 240 120 240 120 240 0 0 2.2 0 2.2 0 2.2 0 0.0 0.0 2.2 2.2 4.4 4.4 6.6 6.6
Bleed Mud at Connection Strip in Stand No6 Bleed Mud at Connection Strip in Stand No 7 Bleed Mud at Connection Strip in Stand No 8 Bleed Mud at Connection Strip in Stand No 9 Bleed Mud at Connection
+ve overbalance
M
-ve decrease
M NA if bled to
compensate for pipe - ve underbalance
WEOX02.197
6-60
March 1995
There is an inlet at the stack between the two BOPs used for stripping. There is a suitable level of redundancy in the stack to ensure the lowest BOP is not used during the stripping operation. API RP 53 (issued 1984) states: The lowermost ram should not be employed in the stripping operation. This ram should be reserved as a means of shutting in the well if other stack components of the blowout preventer fail. It should not be subjected to the wear and stress of the stripping operation.
In a critical situation, it may be possible to modify a surface stack to suit these conditions after a kick has been taken. An example surface stack that is suitable for ram combination stripping is shown in Figure 6.21. The risks involved in ram combination stripping can be assessed by considering the following points: The high level of drillcrew co-ordination required. The level of stress placed on the BOP elements. The level of stress placed on the BOP control system. (During ram combination stripping, the accumulators are charged to maximum operating pressure and isolated from the BOP. The pumps are used for operational functions.) The possibility of replacing the worn BOP elements during operation. On a floating rig, the reduction in level of redundancy within the subsea BOP stack as the ram preventer is used.
6-61
March 1995
Figure 6.21 Surface BOP Stack Suitable for Ram Combination Stripping
ANNULAR
BLIND RAM
FLANGED ACCESS POINT TO STACK FOR USE DURING RAM COMBINATION STRIPPING
PIPE RAM
PIPE RAM
WEOX02.046
6-62
March 1995
Figure 6.22 Annular to Ram Stripping stop stripping in when tool joint is above the annular
MUD
VALVE OPEN
ANNULAR
VALVE CLOSED
PIPE RAM
PIPE RAM
WEOX02.047
6-63
March 1995
Figure 6.23 Annular to Ram Stripping close pipe ram bleed ram cavity pressure
MUD
VALVE OPEN
ANNULAR
VALVE CLOSED
PIPE RAM
PIPE RAM
WEOX02.048
6-64
March 1995
Figure 6.24 Annular to Ram Stripping strip in until tool joint is just below annular
MUD
VALVE OPEN
ANNULAR
VALVE CLOSED
BLIND RAM
PIPE RAM
PIPE RAM
WEOX02.049
6-65
March 1995
Figure 6.25 Annular to Ram Stripping use rig pump or cement pump to equalize across pipe ram
MUD
VALVE OPEN
ANNULAR
VALVE CLOSED
BLIND RAM
FROM PUMP
PIPE RAM
PIPE RAM
WEOX02.050
6-66
March 1995
Isolate the accumulator bottles at full operating pressure The accumulators should be kept as back-up in the event of pump failure. Allow the surface pressure to increase by the overbalance margin Reduce annular closing pressure and strip in
8 9
10 Stop when tool joint is above annular (See Figure 6.22.) 11 Close pipe ram at normal regulated manifold pressure 12 Bleed ram cavity pressure Before the annular is opened it will be necessary to bleed down the pressure below it. (See Figure 6.23). 13 Reduce ram operating pressure 14 Open annular. Lower pipe 15 Stop when tool joint is just below annular (See Figure 6.24.) 16 Close annular at maximum operating pressure 17 Pressurise ram cavity to equalise across ram (See Figure 6.25.) Do not use wellbore pressure to equalise across the ram. 18 Reduce annular closing pressure 19 Open pipe ram 20 Continue to strip in according to the above procedure. Kill the well Fill the pipe as required.
6-67
March 1995
MUD
VALVE OPEN
ANNULAR
VALVE CLOSED
BLIND RAM
MANUAL CHOKE
PUMP
WEOX02.051
6-68
March 1995
The main problem associated with this technique is that migration and entrance into the gas bubble may not easily be detected at surface. If no allowance is made for these complications, further influx may be allowed to occur. To avoid this, the mud tank levels should be closely monitored to ensure that the levels rise, or drop, in direct relation to the volume of pipe that has been stripped into, or out of, the well. If any discrepancy is noticed, the well should be shut-in and the surface pressures verified. Influx migration should be dealt with using the Volumetric Method. The Dynamic Stripping technique can be used during either annular or ram combination stripping. For annular stripping it is implemented along the following lines: (For details of Steps 1 to 6, See Paragraph 4 Annular Stripping Procedure) 1 2 3 4 5 6 7 8 9 Install drillpipe dart Monitor surface pressures Determine the capacity and displacement of the drillpipe Calculate hydrostatic pressure per barrel of the mud Estimate the increase in surface pressure due to the BHA entering the influx Allow the surface pressure to increase by the overbalance margin Line up the pump to the choke line (See Figure 6.26.) Ensure that the manual choke is fully closed. Open choke line valve(s) Open the manual choke at the same time as the pump is brought up to speed
10 Maintain final shut-in pressure on the choke 11 Reduce annular closing pressure 12 Strip in the hole 13 Monitor surface pressures and pit level If the choke pressure increases significantly as the pipe is stripped into the hole, either reduce the pipe running speed or increase the circulation rate. Use the Stripping Worksheet to record all the relevant data. It is very important to accurately record pressures and mud volumes while stripping. 14 Strip to bottom. Kill the well Fill the pipe as required.
1-69/70 6-69
March 1995
6.2
Paragraph
1 2 3 4 General When to Bullhead The Important Factors Procedure
Page
6-72 6-72 6-72 6-73
Illustrations
6.27 Well Shut-in after Production tubing full of gas prior to bullheading 6.28 Example Guide to Surface Pressures during a Bullheading Operation 6.29 Well during Bullheading Operations 6.30 Well after Bullheading Operations tubing displaced to kill weight brine
6-71
March 1995
1 General
Bullheading is a technique that may be used in certain circumstances during drilling operations to pump an influx back into the formation. This technique may or may not result in fracturing the formation. Bullheading is however a relatively common method of killing a well during workover operations. This technique is generally used only during workover operations when there is adequate reservoir permeability.
2 When to Bullhead
During operations, bullheading may be considered in the following situations: When a very large influx has been taken. When displacement of the influx by conventional methods may cause excessive surface pressures. When displacement of the influx by conventional methods would result in an excessive volume of gas at surface conditions. If the influx in suspected to contain an unacceptable level of H2 S. When a kick is taken with the pipe off bottom and it is not considered feasible to strip back to bottom. When an influx is taken with no pipe in the hole. To reduce surface pressures prior to implementing further well control operations.
6-72
March 1995
4 Procedure
In general bullheading procedures can only be drawn up bearing in mind the particular circumstances at the rigsite. For example there may be situations in which it is considered necessary to cause a fracture downhole to bullhead away an influx containing H2S. In another situation with shallow casing set, it may be considered totally unacceptable to cause a fracture in the openhole. During a workover operation a procedure for bullheading will be drawn up along the following lines: 1 Calculate surface pressures that will cause formation fracture during bullheading Calculate also the tubing burst pressures as well as casing burst (to cover the possibility of tubing failure during the operation). 2 3 Calculate static tubing head pressure during bullheading Slowly pump kill fluid down the tubing. Monitor pump and casing pressure during the operation
As an example consider the following well (See Figure 6.27). Well information: Depth of formation/perforations at 3100 m Formation pressure Formation fracture pressure Tubing 4 1/2 in. N80 Vam Internal capacity Internal yield Shut-in tubing pressure Gas density = = = = = = 1.06 SG 1.66 SG 0.0499 bbl/m 8430 psi 3650 psi 0.1 psi/ft
0.0499
(bbl)
= 155 bbl Maximum allowable pressure at pump start up = (1.66 X 3100 = 6300 psi Maximum allowable pressure when the tubing has been displaced to brine at 1.06 SG = (1.66 1.06) X 3100 = 2640 psi
X X
1.421) (0.1
3.2808 X 3100)
(psi)
1.421 (psi)
6-73
March 1995
Figure 6.27 Well Shut-in after Production tubing full of gas prior to bullheading
3650 psi
PACKER
WEOX02.052
6-74
March 1995
Static tubing head pressure when tubing has been displaced to brine = 0 psi (ie the tubing should be killed)
The above values can be represented graphically (as shown in Figure 6.28). This plot can be used as a guide during the bullheading operation. Figures 6.29 and 6.30 show a schematic of the well at two stages of the operation.
10000 TUBING BURST 9000 8430 8000 WORKING PRESSURE RANGE DURING BULLHEADING OPERATION STATIC TUBING PRESSURE THAT WOULD FRACTURE FORMATION INCLUDING 500psi SAFETY FACTOR (if fracturing is a consideration)
2640 2000 1000 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 155 STATIC TUBING PRESSURE TO BALANCE FORMATION PRESSURE 2140 1000 0
WEOX02.053
6-75
March 1995
4000psi 60bbl OF THE TUBING DISPLACED (FROM FIG 6.28, TUBING PRESSURE WITHIN ACCEPTABLE LIMITS) BULLHEAD BRINE
PACKER
PERFORATIONS
WEOX02.054
6-76
March 1995
Figure 6.30 Well after Bullheading Operations tubing displaced to kill weight brine
0psi
PERFORATIONS
WEOX02.055
1-77/78 6-77
March 1995
6.2
Paragraph
1 2 3 General Snubbing Units Selection of a Snubbing Unit
Illustrations
6.31 Rig Assisted Snubbing Unit 6.32 Concentric Cylinder Snubbing Unit 6.33 Multicylinder Snubbing Unit 6.34 Force Diagram for Snubbing Pipe
6-79
March 1995
1 General
Snubbing is a technique used to force pipe into a shut-in well when the upthrust due to well pressure makes it impossible to strip the pipe through the BOP under its own weight. Snubbing is relatively common in some areas in workover operations, when the well may be allowed to continue flowing as remedial work is carried out. Snubbing may be considered during drilling operations for well control purposes, either when it is impossible to introduce pipe into a well that is under pressure, or if the rig BOP system is not considered adequate to provide reliable pressure containment during a prolonged stripping operation. A snubbing unit can be used to introduce a range of sizes of pipe into the well. It can be used to snub tubing, drillpipe and even casing in exceptional circumstances. The lowermost components of the snubbing unit are the snubbing BOPs, which are made up to the top flange of the annular preventer on the rigs stack. This flange is often poorly maintained because it is normally made up to the bell nipple and does not generally need to form a pressure seal. It must therefore be inspected and, if necessary, repaired before the snubbing BOPs are nippled up. The snubbing BOPs are likely to be too tall to fit underneath the rotary table and too wide to go through it. To overcome this problem, the snubbing company can provide suitable spacer riser sections to bring the assembly above the rig floor. The weight of the snubbing unit is supported by the wellhead. Guy lines from the work platform prevent lateral movement. Snubbing units can therefore be rigged up on land rigs and fixed offshore installations in a relatively straightforward manner. Snubbing units are not commonly used on floating rigs, however they have been used successfully in the past for well control operations. In order to use a snubbing unit on a floating rig, pressure containment must be established between the rig BOP and the unit on the rig floor. Drillpipe or tubing may provide this pressure containment, in which case small diameter tubing may be run into the well through the drillpipe or tubing. An operation of this type can only be carried out in relatively calm seas so that the rig heave does not cause excessive movement of the snubbing unit.
2 Snubbing Units
(a) The Rig Assisted Type
The rig assisted unit uses the travelling blocks to generate the snubbing force through a series of pulleys and cables. (See Figure 6.31.) The rig assisted unit can handle larger diameter pipes such as casing up to 13 3/8 in. and have snubbing capacities of 80,000 lb to 400,000 lb. These were the first snubbing units used and the few that are currently available are operated by Otis and Cudd Pressure Control.
6-80
March 1995
TRAVELLING BLOCK
BALANCE WEIGHT
TRAVELLING SNUBBERS
SNUB LINE
STATIONARY SNUBBERS
PLATFORM
PUMP INLET
SAFETY PREVENTERS
WELL PRESSURE
WEOX02.056
6-81
March 1995
The unit consists of a set of travelling snubbers which are connected to the travelling block. The travelling snubbers grip the pipe and force it into the well as the blocks are raised. A set of stationary snubbers grip the pipe while the travelling snubbers are being raised (by the counter balance weights) for a new bite on the pipe. Once sufficient pipe has been run to reach the balance point, the travelling snubbers will be removed and the pipe will be run in by conventional stripping.
The units are operated from the work platform which is on top of the hydraulic jack assembly. From this position the speed of the pipe and the slips are controlled as can be the rotary table, if required. Stationary and travelling slips are operated in sequence to grip the pipe as it is snubbed into the well. One operator will control the BOPs and equalising valves. Another operator will co-ordinate the pipe handling, using the counter balance system.
6-82
March 1995
PISTON
PISTON EXTENDED AND TRAVELLING SLIPS CLOSED PRIOR TO FORCING PIPE INTO WELL
PISTON RETRACTED AND TRAVELLING SLIPS OPEN BEFORE PISTON IS AGAIN EXTENDED
WEOX02.057
6-83
March 1995
POWER TONGS BOP CONTROL PANEL CONTROL PANEL WORK PLATFORM TRAVELLING SLIPS COUNTERBALANCE WINCH
TELESCOPING MAST
STATIONARY SLIPS
WEOX02.058
6-84
March 1995
The snubbing force is calculated as follows: Snubbing force, Fs = F p + Ff (w a Ly X 3.281) (wb X L z X 3.281) where Fp = Pw Ao (See Figure 6.34) where F s Fp Ff wa wb Ly Lz Ao = = = = = = = = required snubbing force (lb) force due to well pressure (lb) frictional force (lb) weight of pipe (lb/ft) buoyant weight of pipe (lb/ft) length of pipe above BOP to the travelling snubber (m) length of pipe in the hole (m) outside cross sectional area of pipe (in.2)
COMPRESSION FORCE Fs
POINT OF APPLICATION OF TRAVELLING SNUBBERS
Fs
wa Ff
PIPE
Ly
SNUBBING BOP
Ff (wa)(Ly) wb Pw Lz
WELLBORE
Fp Fp (wb)(Lz)
Equilibrium Equation (from Forces = 0) Therefore: Fs = Fp + Ff (wa) (Ly) (wb) (Lz) Where
Fs Fp Ff wa wb Ly Lz
= required snubbing force (lb) = force due to well pressure (lb) = frictional force (lb) = weight of pipe (lb ft) = bouyant weight of pipe (lb ft) = length of pipe above BOP to the travelling snubber (m) = length of pipe in the hole (m)
WEOX02.059
6-85
March 1995
Snubbing force for the first joint of pipe. In this case, the length of pipe in the hole (Lz) is zero, and the length of pipe above the BOP is considered insignificant. Consider the following example: The well is shut in with a wellhead pressure of 5000 psi. 2 7/8 in. tubing produces a frictional force of 3000 lb at the stripping rams. The area of pipe exposed to the wellbore pressure therefore equals 6.492 in. Snubbing force, F s = Fp + F f = (6.492 X 5000) + 3000 = 35,460 lb The snubbing force, Fs, if there is already some pipe in the hole. In this case the length of the pipe above the BOP is again considered insignificant. As an example: 2 7/8 in. tubing of 6.5 lb/ft is run empty to 1000 metres in 1.2 SG mud. The wellhead pressure is 5000 psi. Drag in the hole is 2000 lb, friction at the BOPs is 5000 lb. Ai Ao wi wo wa wb D = = = = = = = internal cross sectional area area of pipe (in.2 ) outside cross sectional area area of pipe (in.2 ) weight of fluid inside the pipe (SG) weight of fluid in annulus (SG) weight of pipe in air (lb/ft) buoyant weight of pipe (lb/ft) depth of tubing (m) (lb)
wb = wa + (w i X Ai) (wo X Ao) wb = 6.5 + (O X Ai) (1.2 wb = 3.12 lb/ft Therefore the snubbing force is given by: Fs = Fp + Ff (wa X Ly) (w b X L z) Fs = (6.492 X 5000) + 2000 + 5000 (3.12 = 29,200 lb Size of the Unit The dimensions of the unit must be checked against the internal dimensions of the derrick, if the unit is to be used with a rig on location. Lifting Force The unit must be able to provide a reasonable overpull, over and above the weight of the maximum string weight.
X X
(lb/ft)
1000 X 3.281)
(lb)
6-86
March 1995
Tubular Selection If there is already pipe in the hole, this will determine the most suitable type of pipe to be used. Drillpipe can be used, however the following points should be considered: Drillpipe will require a relatively high snubbing force because of its large crosssectional area at the tool joints. Drillpipe does not have gas-tight connections. The drillpipe must be in good condition and inspected thoroughly before running in. Tubing is more commonly used for snubbing for the following reasons: The force required to snub it in is very much less, and the unit required corresponding smaller. External flush tubing can be run through the stripper rubbers without the need for sequencing the rams. The following points must also be considered: The limitations imposed by the ID of the tubing on the maximum pump rate. External upset tubing will be slower to run, but will be easier to control, if it starts to be forced out of the well. Premium connections are desirable because they are gas tight. The collapse strength of the tubing. The susceptibility of the tubing to failure due to buckling.
6-87 1-87/88
March 1995
6.2
Paragraph
1 2 3 4 5 6 7 8 Characteristics of Baryte Plugs Deflocculation Pilot Tests Slurry Volume Pumping and Displacement Rate Preparation of a Baryte Plug After Pumping a Baryte Plug Baryte Plug Procedure
Page
6-90 6-92 6-92 6-92 6-93 6-93 6-93 6-94
Illustrations
6.35 Field Mixing of Baryte Plugs
6-91
6-89
March 1995
(c) Settling/Non-settling
Since baryte settling is of little value downhole and troublesome on the surface, it should be an optional feature of the slurry recipe. Figure 6.35 shows two recipes for baryte slurries. The recipes are identical except that one contains XC polymer to eliminate baryte settling. It would seem reasonable to use the settling recipe for small jobs or where the settling baryte might really be helpful downhole. For large kill operations, the non-settling recipe would be preferred. Bentonite or some polymer other than XC could be used to suspend the baryte in a slurry. The particular recipe in Figure 6.35 was selected because it is prepared easily in both fresh and seawater and because XC solutions are shear-thinning enough to allow good pumpability while adequately suspending the baryte in the pits.
6-90
March 1995
6-91
March 1995
Baryte-plug-type slurries can be prepared with all of the baryte substitutes which are now on the market. In general the recipes in Figure 6.35 do not require change except that, in some cases, the higher density of the substitue allows higher slurry weights than were possible with baryte. For example, hematite slurries can be prepared to 3.00 SG using the non-settling recipe in Figure 6.35. Replace the baryte with 870 lb hematite per final bbl of slurry. The non-settling recipe is strongly recommended for hematite slurries because of the relatively coarse grind of oil-field hematite.
2 Deflocculation
For years it has been standard practice to add a thinner to baryte slurries used for well control. Both lignosulphonates and phosphates have been used, with the phosphate SAPP having the widest acceptance. Chemicals of either type can deflocculate a baryte slurry to improve pumpability and allow settling into a firm cake. The choice of deflocculant will influence the baryte slurry properties as follows: Use of SAPP gives a slurry with fairly high fluid loss (50cc). SAPP will not deflocculate in sea water or in the presence of some contaminants which occur in natural baryte. Use of lignosulphonate gives a slurry with low fluid loss (5cc). Lignosulphonate is effective in sea water and tolerates both contamination and elevated temperatures.
Use of a high fluid loss baryte slurry is advantageous, possibly because it might dehydrate and plug the wellbore, or promote, perhaps, hole instability. On the other hand, a low fluid loss slurry would reduce the chances of differential sticking. Faced with this choice, prudence suggests using the more reliable lignosulphonate rather than the somewhat unpredictable SAPP. The recipes in Figure 6.35 contain lignosulphonate.
3 Pilot Tests
Because of variation and possible contamination of ingredients throughout the world, it is always advisable to pilot test a baryte slurry. Prepare a sample of the slurry using the recipe chosen and the ingredients at the wellsite. After being stirred well, the sample should have the expected density and be easily pumpable. If the baryte needs to settle in the wellbore, this should also be checked ahead of time. Reasonable settling is 2 in. in a mud cup after a 15 minute wait. The settled cake should be hard and somewhat sticky rather than soft and slippery. The settling test is not a guarantee that the baryte pill will form an effective plug under downhole conditions, but will certainly give an indication of the settling characteristics.
4 Slurry Volume
Slurry volumes depend upon the amount of openhole and the severity of the kick. These volumes normally range from 40 bbl to 400 bbl. The slurry volume should be 125 to 150 percent of the annular capacity necessary to give the height of plug desired, but should not be less than 40 bbl. If a second baryte plug is required, the slurry volume should be greater than the first.
6-92
March 1995
6-93
March 1995
The decision after placing a baryte plug is whether to pull pipe or not. The goal of pumping a high-density slurry is to achieve a hydrostatic kill; the decision whether to pull pipe depends on an assessment of the success of this kill. If a hydrostatic kill was probably achieved then it is usually best to pull up above the slurry and try circulating mud. If there is doubt about the hydrostatic kill it may be better to stay on bottom to be ready to pump a larger baryte plug if needed. The risk in pulling out is that the pipe may become stuck off bottom or may have to be stripped back to bottom if the baryte plug fails. The risk of staying on bottom is that the pipe may become stuck or plugged. It is possible to keep the pipe free by moving it (especially in a non-settling plug) but there is no way to circulate (to avoid plugging) unless the pipe is pulled above the top of the baryte slurry.
6-94
March 1995
Perforate the drillstring near the top of the baryte plug. Attempt to circulate It may be difficult to tell whether the well is circulating or flowing from charged formations. Pressure communication between the drillpipe and annulus is one clue; a pressure increase should have appeared on the drillpipe from annulus pressure or on the casing from hydrostatic pressure in the drillpipe when the perforation wasmade. Consideration should be given to circulating with lighter mud because of the known lost returns zone. Well will circulate: Use drillpipe pressure method to circulate annulus clear of formation fluid. Run a free-point log. Begin fishing operations. Well will not circulate: Squeeze cement slurry through perforation. Cut displacement short on final stage to provide an interior plug or set wireline bridge plug. WOC and pressure test plug. Run free-point log. Perforate the pipe near the indicated free point. Circulate using drillpipe pressure method until annulus is clear. If well will not circulate, squeeze perforations with cement or set a wireline bridge plug above perforations and perforate up the hole.
6-95
March 1995
Monitor the annulus If no pressure is on the annulus, continue working the pipe, and observe annulus mud level. If the annulus is full, begin circulating at a low rate keeping constant watch on pit levels. If the annulus is not full, fill annulus with water and observe. If annulus will stand full, begin circulating at a slow rate. Consider cutting mud weight, if feasible. If pressure is on the annulus, circulate the annulus using normal well control techniques. Continue working the pipe. If returns become gas-free, the baryte plug was successful and the well is dead. If returns do not become essentially gas-free after circulating two or three annular volumes, the baryte plug was not effective. A second plug will be necessary.
Trip out of the hole after verifying that the well is dead If the bottom part of the hole is being abandoned, then a cement plug should be placed on top of the baryte.
6-96
March 1995
6.2
Paragraph
1 2 Use of Shear Rams Dropping the Pipe
Page
6-98 6-99
6-97
March 1995
Most shear rams are designed to shear effectively only on the body of the drillpipe. Procedures for the use of shear rams must therefore ensure that there is no tool joint opposite the ram prior to shearing. Be aware that many subsea stacks have insufficient clearance between the top pipe rams and the shear rams to hang off on the top rams and shear the pipe. Specialist shear rams, such as the Cameron Super Shear Rams, are available that are designed to shear 7 in. drillcollars and casing up to 13 3/8 in. OD. It is clearly important however, that rigsite personnel are aware of the capabilities and operating parameters of the shear rams installed in the rigs BOP stack. Optimum shearing characteristics are obtained when the pipe is stationary and under tension. It is therefore recommended practice that the pipe weight is partially hung off prior to shearing. Hanging the pipe off also ensures that there is no tool joint opposite the shear rams. Maximum operating pressure should be used to shear the pipe. The following procedure can be used as a guideline for shearing the pipe in the case of an internal blowout: 1 2 3 Space out to ensure that there is no tool joint opposite the shear rams Close the hang-off ram Hang off on the rams Ensure that the pipe above the hang-off rams remains in tension. 4 5 6 Prepare to operate the shear rams Close the shear rams at maximum accumulator pressure Monitor the well. Implement appropriate control procedures
6-98
March 1995
Once the pipe has been dropped the well is shut-in with the blind/shear rams. However, re-establishing control of the well in this situation will be time consuming and costly. It is clearly important to be sure that the pipe will clear the stack once it has been dropped (especially on a floating rig in deep water). The possibility of damaging the ram packings must also be considered. There are two techniques that can be used to drop the string:
6-99
March 1995
10 Close the blind/shear ram 11 Close the choke Both of these techniques involve a certain amount of risk. The most suitable method in each case will depend on the particular conditions at the rigsite.
6-100
March 1995
6.3 COMPLICATIONS
Paragraph
1 2 3 4 5 6 7 8 9 10 11 12 13 Plugged Bit Nozzle Plugged Choke Cut Out Choke Pump Failure Pressure Gauge Failure String Washout Stuck Pipe Well Control Considerations in Horizontal Wellbores Hydrates Surface Pressures Approach the MAASP Impending Bad Weather Loss of Control Well Control Considerations in Slim Hole Well
Page
6-102 6-102 6-102 6-103 6-103 6-103 6-104 6-104 6-105 6-109 6-110 6-111 6-111
Illustrations
6.36 Temperature at which Gas Hydrates will Freeze (Katz) 6.37 Natural gas expansion Temperature reduction curve (NATCO) 6.38 Height of 10 bbl Gas Influx in Annulus 6.39 Reduction in Bottom Hole Pressure Due to 10 bbl Gas Influx 6.40 Annular Friction Pressure Drop 6.41 Swab Pressure in a 1000 m Hole 6-106 6-107 6-113 6-114 6-115 6-116
6-101
March 1995
2 Plugged Choke
A plugged choke is indicated by an unexpected increase in choke pressure accompanied by an equal increase in drillpipe pressure. Some plugging of the choke is to be expected if the annulus is loaded with cuttings. Clearly the first course of action is to open the choke in an attempt to both clear the restriction in the choke and to avoid overpressuring the well. If this action is not successful the pump should be stopped immediately. After switching to an alternate choke the excess pressure in the well should be bled at the choke and the displacement restarted in the usual manner. One of the reasons for displacing a kick at slow circulation rates is to avoid overpressuring the well if cuttings plug the choke. In this respect, circulation rates should be minimised in critical conditions if the annulus is likely to contain a substantial volume of cuttings.
6-102
March 1995
Having established that there is no loss of circulation a worn out choke should be suspected. There may come a stage when it is no longer possible to maintain a suitable circulating pressure even with the choke apparently fully closed. At, or preferably before this stage, the flow should be switched to another choke and repairs effected to the worn choke.
4 Pump Failure
The most obvious indicator of failure at the fluid end is likely to be erratic standpipe pressure together with irregular rotary hose movement. This may be preceded by an unexplained drop in circulating pressure. If pump failure is suspected, the pump should be stopped and the well shut-in. The displacement should be continued with the second rig pump, or if necessary, the cement pump. The faulty pump should be repaired immediately.
6 String Washout
A washout in the drillstring may be indicated by an unexpected drop in standpipe pressure, while the choke pressure remains unchanged. The recommended procedure in the event of a drillstring washout is to stop the pump and shut the well in. Every effort must be made to ensure that the washout is not enlarged by extended circulation and drillstring manipulation. The most critical situation would be in the event of a washout close to the surface. Should this occur, it is unlikely that it will be possible to displace the influx from the hole effectively, unless the influx is above the washout. If the washout is identified as being near the bottom of the well, it may be possible to displace the kick from the well effectively. In this case, there will of course be the risk of parting the drillstring with continued circulation. Regardless of the depth of the washout, it will be necessary to re-establish the correct circulating pressure if the pump is restarted. Excessive downhole pressures may be caused if the original circulating pressure is maintained at the standpipe. It is advisable to periodically re-establish the circulating pressure, if the circulation is contained for prolonged periods through a washout.
6-103
March 1995
7 Stuck Pipe
The complication of stuck pipe during a well control operation can cause serious problems, most especially if the pipe is stuck off bottom. Unfortunately, the likelihood of the pipe becoming stuck during a well control operation is increased if the pipe is off bottom. The pipe should be rotated, to minimise the risk of sticking the pipe, if the well is shut-in with the pipe off bottom and the BHA in openhole. Due to the relatively high wellbore pressures during a well control operation, the most likely cause of stuck pipe is differential sticking. However, mechanical sticking may result if the hole sloughs and packs-off as a result of the contact with the influx fluids. If the pipe is differentially stuck with the bit on bottom, continue the operation because it is most likely that circulation can still be carried out in order to kill the well. Efforts to free the pipe can be made once the well has been killed. Should the pipe be differentially stuck with the bit off bottom, the situation is complicated in that it will generally not be possible to reduce the wellbore pressure at that depth by circulation. It may be possible to free the pipe by spotting a freeing agent. However, if the influx was swabbed in, it may be possible to regain control of the well by volumetric control. If the pipe is mechanically stuck, a combination of working the pipe and spotting a freeing agent can be used in attempting to free the pipe.
6-104
March 1995
There is a greater potential for swabbing when a large surface area of reservoir is exposed. Correct tripping procedure must be rigorously adhered to. It is quite feasible, in a horizontal well, that the horizontal section is full of reservoir fluid and yet the well be dead. It is therefore recommended that extreme caution be paid when tripping back into such a reservoir after a round trip. When back on bottom it is recommended to circulate bottoms up through the choke manifold. In the event of a kick whilst tripping it may not be possible to drop or pump down the dart. This will depend on the hole angle at the dart sub position. If it is not possible to install the dart into the dart sub, the Gray valve can be used.
9 Hydrates
Natural gas hydrates have the appearance of hard snow and consist of chemical compounds of light hydrocarbons and liquid water. They are formed at temperatures above the normal freezing point of water at certain conditions of temperature and pressure (See Figure 6.36). This formation process is accelerated when there are high gas velocities, pressure pulsations or other agitations, such as downstream of a choke and at elbows, which cause the mixing of hydrocarbon components. During well control operations, gas hydrates may cause the following serious problems: Plugging of subsea choke/kill lines, preventing opening and closing of subsea BOPs, sealing off wellbore annuli and immobilising the drillstring. There have been recorded incidences of such occurrences with subsea stacks in water depths of 350m and deeper. Plugging of surface lines at and downstream of the choke or restriction. This is particularly hazardous when high gas flowrates are experienced through low pressure equipment (such as the poorboy separator and gas vent line). The formation of hydrate plugs under these conditions can rapidly overpressure low pressure well control equipment.
The major factors which determine the potential for hydrate formation are gas composition, liquid content and pressure and temperature. The formation of hydrates can be predicted using Figure 6.36. It should be noted that the conditions for hydrate formation can be created at a subsea stack operating in a cold water environment. Figure 6.37 can be used to predict the temperature drop associated with a pressure drop (across a choke, for example). As an example, if gas at 3000 psi and 90 F was choked to 1800 psi, the temperature would be expected to drop to 55 F, in which case, hydrate formation could be expected.
6-105
March 1995
The purpose of this chart is to determine the temperature below which hydrates will form, when sufficient liquid water is present.
4000
3000
2000
ET
HA
NE
400
0.
300
AV
0.
200
100 90 80 70 60
35
40
45
1.
50
0.
0. 9
55
60
65
70
75
80
85
TEMPERATURE (F)
Example: With 0.7 specific gravity gas at 1000psia, hydrates may be expected at 64F. At 200psia this would be 44F.
WEOX02.061
6-106
March 1995
NATURAL GAS EXPANSION TEMP REDUCTION CURVE BASED ON 7 SP GR GAS (From NATCO)
160
70
00
160 150
60 55 50 00 00 00
150
0 50
140
SU
R
140
R ED
OP
130
00 40
35
00 30
120
45 00
TE
00
INITIAL TEMP RISE
110
MP
P RO
DU
ET
R OP
ES
100
80
0 20 0
0 50
105 - 80 = 25
6-107
15
10 50 0 00
90
80 70
BASE LINE
70
00
60
60 50
0
S VE
50
40
40 30 20 10 0
30
20
CO
R CU S PY GA L A T TH U F EN 0 C T 0 N 0 TA U/1 NS BT
10
EXAMPLE REQUIRED: REDUCE GAS PRESSURE FROM 2400 # PSI AT 80F TO 1500 # PSI DETERMINE INITIAL TEMPERATURE RISE NECESSARY SO THAT AFTER EXPANSION TO 1500 # PSI THE FINAL TEMPERATURE WILL BE 75F
0 1000 2000
WEOX02.062
4000
5000
March 1995
Hydrates can be combated by one or a combination of the following: Injecting antifreeze agents such as methanol into the gas flow; this has the effect of dissolving liquid water deposits, and thus lowering the temperature at which hydrates would be expected to form. Methanol is often injected at the subsea test tree during well testing operations from a floating rig. The most appropriate place to inject methanol at surface is at the choke manifold. The point of injection should be upstream of the choke. High pressure chemical injection pumps (as manufactured by Texsteam) are suitable for this application. Heating the gas above the temperature at which hydrate will form. During gas well testing operations, a steam exchanger will usually be provided for this purpose. Experience has shown that this is the most effective and reliable method of preventing the formation of hydrates. The combination of heating and antifreeze injection is ideal. Reducing line pressure in order to allow the hydrates to melt. This is a temporary measure and not always practical. Once hydrates have formed, it often takes a considerable length of time to clear the line.
It is important that adequate contingency is provided, along the above lines, to deal with hydrates, if it is suspected that the potential exists for hydrate formation. Subsea water temperatures and pressures should be checked as well as the potential for hydrate formation at surface in the event of a gas kick.
6-108
March 1995
The consequences of overpressuring the openhole weak point as in the first option can be assessed, bearing in mind the following factors: The depth of the casing shoe. The quality of the cement job. By how much the openhole weak point will be overpressured. The length of time that the openhole weak point will be overpressured. The characteristics of the openhole weak point. Any safety factor included in the calculation of the MAASP. The possibility of broaching around the casing.
The consequences of underbalancing the formation as in the second option can be assessed, bearing in mind the following factors: The type of kick zone fluid. The permeability of the kick zone. The degree of underbalance. The length of time that the kick zone will be underbalanced.
The appropriate course of action should therefore be selected on the basis of these factors. However, in general, a kick zone should only be underbalanced in exceptional circumstances such as when the zone is known to have low permeability. This can often be assessed from the rate of pressure build after shutting in a well that has kicked.
6-109
March 1995
If additional time is available, consideration should be given to spotting a heavy pill or plug on bottom to either kill the well hydrostatically or provide a barrier to migration. Bad weather may cause problems regarding the supply of chemicals and barytes to all types of rigs. In this respect, it may be necessary to implement the Drillers Method, should there not be adequate chemical stocks at the rigsite. In certain areas of the world, severe cold may cause operational problems. Points of particular concern are, BOP operating fluid, manifolds and flowlines.
12 Loss of Control
Loss of control during a well control operation may result from excessive loading of pressure control equipment or exposed formations. However there are recorded incidents of equipment failure at pressures significantly below rated values. These failures have been attributed to faulty manufacture, lack of proper maintenance, or corrosion. High pressure equipment is considered particularly susceptible to failure when exposed to corrosive fluids such as H2 S. It is not possible to detail specific procedures in the event of loss of control during a well control operation. However, in critical situations, action should be taken bearing in mind that the absolute priority is the safety of rigsite personnel.
6-110
Rev 1 March March 1995 1995
Greater Bottom Hole Pressure Reduction As the result of the greater influx length, the same volume of formation influx will result in a greater reduction in the bottom hole pressure in a slim hole well. Asshown in Fig.6.39, a 10 bbl gas influx will reduce the bottom hole pressure by about 743 psi in a 3.5"x2.5" slim hole well and only 94 psi in a conventional 8.5"x5" well.
Higher Annular Friction Pressure Also due to the reduced annular size, the annular friction pressure drop can be many times higher in a slim hole well than in a conventional well, as shown in Fig.6.40. Therefore the friction pressure drop can become significant during well control operations in a slim hole well whereas it is all but ignored in the case of a conventional well.
Higher Swab and Surge Pressures Fig.6.41 compares the swabbing pressure in both slim hole and conventional wells. It can be seen that the swabbing pressure is much higher in a slim hole well than in a conventional well. Also the swabbing pressure increases more rapidly in a slim hole well with increasing the trip speed.
Effect of High Drillpipe Rotational Speed During a slim hole drilling operation, the drillpipe is often rotated at a much higher rate than that during a conventional drilling operation. Due to the high rotational speed together with the small annular size, the drillpipe rotation can result in a significant increase in the annular friction pressure and the ECD. This effect must be taken into account in the well control procedures. Otherwise, the weak formation may be broken down when the drillpipe starts to rotate, or a kick influx be induced when rotation stops (whilst still maintaining circulation).
6-111
Rev 1 March March 1995 1995
6-112
Rev 1 March March 1995 1995
Friction Pressure (psi) Brine: 4.0 cPMud: PV=15/YP=10 13.8 31.2 41 68.9 128.6 254.3
1 0 400
8.5 x 5 6 x 4.5
66 199 523
8.5 x 5 6 x 4.5
6-113
Rev 1 March March 1995 1995
10 10 200
523 m
3.5 x 2.5
3.5 x 2.5
128.6 254.3
Figure 6.39 Reduction in Bottom Hole Pressure Due to 10 bbl Gas Influx
Figure 6.38b: Reduction in Bottom Hole Pressure Due to 10 bbl Gas Influx
(1.0 SG Density Difference Between Mud and Gas) 800 Mud: PV=15/YP=10 31.2 600
6-114
400 200 283 psi 0 94 psi 8.5 x 5 6 x 4.5 Size of Annulus (inch)
743 psi
3.5 x 2.5
Height of Ga (m)
250 200
Brine: 4.0 cP
8.5/5
Mud: PV=15/YP=10
31.2
41
6/4.5 40.9 41 41.3 68.9 41.7 42.5 44.8 46.2 3.5 x 2.5 48.3 55 58.8 64.8
3.5/2.5 45.8 46 46.3 48.8 47.8 50.9 52.5 55.1 66.4 92.2 147 90 90.8 92.1 94.2 98.5 110 117.7 128.4 168 238 368
swab pressures
6-115
Rev 1 March March 1995 1995
Reduction in (psi)
March 1995 1995 Rev 1 March
3.5/2.5
6-116
90 90.8 92.1 94.2 98.5 110 117.7 128.4 168 238 368
Volume 2 Contents
Nomenclature Abbreviations 1 THE ORIGINS OF FORMATION PRESSURE
Section 1.1 1.2 1.3 1.4 1.5 INTRODUCTION NORMAL FORMATION PRESSURE SUBNORMAL FORMATION PRESSURE ABNORMALLY HIGH FORMATION PRESSURE SHALLOW GAS Page 1-1 1-9 1-11 1-17 1-33
March 1995
4 FRACTURE GRADIENT
Paragraph 1 GENERAL 2 STRESSES IN THE EARTH 3 FRACTURE ORIENTATION 4 FRACTURE GRADIENT PREDICTION 5 DAINES METHOD OF FRACTURE GRADIENT PREDICTION 6 AN EXAMPLE PRESSURE EVALUATION LOG 7 LEAK OFF TESTS 8 LEAK OFF TEST PROCEDURE 9 INTERPRETATION OF RESULTS 4-2 4-2 4-3 4-4 4-4 4-7 4-9 4-10 4-11
March 1995
March 1995
NOMENCLATURE
SYMBOL A a An b c C Cp Ca CL CR D Dshoe Dwp dbit dh dhc do di dcut dc F Fsh FPG g G DESCRIPTION Cross sectional area Constant Total nozzle area Constant Constant Annular capacity Pipe capacity Cuttings concentration Clinging constant Closing ratio Depth Shoe depth Depth of openhole weak point Bit diameter Hole diameter Hole/casing ID Pipe OD Pipe ID Average cuttings diameter Drilling exponent (corrected) Force Shale formation factor Formation Pressure Gradient Gravity acceleration Pressure gradient UNIT in.2 in.2 bbl/m bbl/m % m m m in. in. in. in. in. in. lb SG psi/ft psi/m SG psi/ft m m m sec/m m m/hr psi lb SG
Gi H Hi Hp ITT K L MR M m MW
Influx gradient Height Height of influx Height of plug Interval Transit Time Bulk modulus of elasticity Length Rotary exponent Migration rate Matrix stress Threshold bit weight Mud weight
March 1995
SYMBOL N OPG P P Pa Pbit Pcl Pdp Pf Pfrac Pfc Pi Pic Plo Pmax Poc Pp Pscr PV Q Qmud Qgas Re R Rw ROP S Sg Sw t TR T TD TVD V
UNIT rpm SG psi/SG (The units of subsurface pressure may be either psi or SG) psi psi psi psi psi psi/SG psi/SG psi psi psi psi/SG psi/SG psi psi/SG psi cP gal/min gal/min gal/min ohm-m ohm-m m/hr meq/100g psi Fractional Fractional seconds min degrees C, F, R m m bbl
Adjustment pressure Annulus pressure Bit pressure drop Choke line pressure loss Drillpipe pressure Formation pressure Fracture pressure Final circulating pressure Hydrostatic pressure of influx Initial circulating pressure Leak off pressure Maximum allowable pressure at the openhole weak point Wide open choke pressure Pore pressure Slow circulating rate pressure Plastic Viscosity Flowrate Mud flowrate Gas flowrate Reynolds number Resistivity Resistivity of water Rate of Penetration Shale factor Overburden pressure Gas saturation Water saturation Time Transport Ratio Temperature Total Depth True Vertical Depth Kick tolerance
March 1995
SYMBOL V
DESCRIPTION Volume
UNIT
v vmud vp vs W
bbl cc ml l Velocity m/min m/s Mud velocity m/min Average pipe running speed m/min Slip velocity m/min Weight gm kg lb Weight lb/ft lb/bbl SG Weight of pipe lb/ft Baryte required for weighting up lb/bbl Average cuttings weight SG Weight on Bit lb Offset () Yield Point lb/100ft2 Compressibility factor Viscosity cP Poissonss Ratio Maximum effective principle stress psi/SG Tectonic stress psi/SG Porosity Fractional Fann reading lb/100ft2 Tectonic stress coefficient Density SG Bulk density SG
March 1995
ABBREVIATIONS
ASN BHA BHC BHT BGG BRT CDP CEG CG DE DIL DRG DST ECD EMW ES FDC FIT HCR ID ITT LMRP MWD OD PV RFT RMS ROP SLS TD TG UV WOB YP Amplified Short Normal Bottomhole Assembly Borehole Compensated Tool Bottomhole Temperature Background Gas Below Rotary Table Common Depth Plot Cation Exchange Capacity Connection Gas Drilling Engineer Dual Induction Laterolog Designated Resident Geologist Drillstem Test Equivalent Circulating Density Equivalent Mud Weight Electrical Survey Formation Density Compensated Tool Formation Interval Tester High Closing Ratio Internal Diameter Interval Transit Time Lower Marine Riser Package Measurement while Drilling Outside Diameter Plastic Viscosity Repeat Formation Tester Root Mean Squared Rate of Penetration Long Spacing Sonic Tool Total Depth Trip Gas Ultra Violet Weight of Bit Yield Point
March 1995
March 1995
1.1
1 2 3 4
INTRODUCTION
Page
1-2 1-2 1-6 1-7
General Subsurface Pressures Pressure Seals Pressure Gradients
Paragraph
Illustrations
1.1 1.2 Composite Overburden Load for Normally Compacted Formations Schematic Diagram of Subsurface Pressure Concepts
1-4 1-5
Tables
1.1 Types of Formation Pressure Seals
1-6
1-1
March 1995
1 General
All formations penetrated whilst drilling a well exert pressures of varying magnitudes. To gain an understanding of the origins of these pressures, it is neccesary to define and explain certain subsurface pressure concepts. These are explained in this Section.
2 Subsurface Pressures
(a) Hydrostatic Pressure
Hydrostatic pressure is defined as the pressure due to the unit weight and vertical height of a fluid column. The size and shape of the fluid column do not affect the magnitude of this pressure. Mathematically: P=rXgXD where P g D = = = = hydrostatic pressure average fluid density gravitational acceleration vertical height of fluid column (1-1)
Relating this to drilling operations and commonly used oilfield units gives: P = C X MW X D where P MW D C = = = = hydrostatic pressure (psi) fluid density or mud weight (lb/gal or ppg) vertical depth (ft) conversion constant (psi/ft per lb/gal) (1-2)
The constant, C, is necessary to allow the use of oilfield imperial units (psi, ft, lb/gal). It has a value of 0.052 psi/ft per lb/gal and is derived as follows: Using consistent units (pressure in lb/sq.ft, length in ft, density in lb/cu.ft) and rearranging equation 1-2, C would be numerically equal to 1: C= P D X MW = 1 lb/sq.ft/ft per lb/cu.ft
Substituting the standard conversion constants of 144 sq.in/sq.ft and 7.48/gal/cu.ft gives: C=1
X
7.48 144
lb/sq.ft ft X lb/cu.ft
sq.ft/sq.in cu.ft/gal
C = 0.052
lb/sq.in ft X lb/gal
1-2
March 1995
So in imperial oilfield units (psi, ft, lb/gal), equation 1-2 becomes: P = 0.052 X MW D For the Companys system of units (psi, SG, m): P = C'
X
(1-3)
SG
(1-4)
where SG = specific gravity of the fluid (no units) D = vertical depth (metres) C' = conversion constant (psi/m) NOTE: Specific gravity (SG) is not a unit of density. It is the ratio of the density of afluid to the density of fresh water at a specified temperature, and hence has no units. The constant, C', has a value of 1.421 psi/m and is derived as follows: To express equation 1-2 in terms of SG (as in equation 1-4), the constant C' must be related to the density of fresh water, which is 8.33 lb/gal. Hence for fresh water: C' = C
X
C' = 0.052
C' = 0.433 psi/ft Expressing this in terms of metres using 3.2808 ft/m gives: C' = 0.433
X
SG X D
(1-6)
SG X D
(1-7)
where D = vertical depth (ft). The overburden pressure gradient (OPG) is given by: OPG = S D = 0.433
X X
SG
OPG = 0.433
1-3
March 1995
1. 2. 3. 4.
Constant gradient 1.0psi/ft Gulf of Mexico, Texas and Louisiana, USA Santa Barbara Channel, California, USA North Sea area
1 4 2 3 1
DEPTH 1000m
WEOX02.063
1-4
March 1995
However, because the degree of compaction of sediments varies with depth, the overburden pressure gradient is not constant. Worldwide experience indicates that the probable maximum overburden gradient in clastic rocks (fragmental sedimentary rocks ie sandstone, shale) may be as high as 1.35 psi/ft. Onshore, with more compact sediments, the overburden pressure gradient may be assumed to be close to 1 psi/ft. Offshore however, overburden gradients at shallow depths will be much less than 1 psi/ft due to the effect of the depth of sea water and large thickness of unconsolidated sediment. Figure 1.1 shows average overburden gradient for various areas.
PRESSURE
NO RM AL HY DR OS
DEPTH
VE
Formation Pressure, Pf
TA TIC GR AD IEN T
EN
IE
Matrix Stress, M
Overburden Pressure, S = Pf + M
WEOX02.064
1-5
March 1995
3 Pressure Seals
For abnormal pressures to exist, there must be a permeability barrier which acts as a pressure seal. This seal restricts or prevents the movement of pore fluids and thus separates normally pressured formations from abnormally pressured formations. The origins of a pressure seal may be physical, chemical or a combination of the two. The types of formation pressure seals are listed below in Table 1.1.
Nature of Seal Massive siltstones Shales Massive salts Anhydrite Gypsum Limestone, marl, chalk Dolomite Faults Salt and shale diapirs
Examples Gulf Coast, USA, Zechstein in North Germany, North Sea, Middle East, USA, USSR.
Transverse Combination
Worldwide Worldwide
Table 1.1
1-6
March 1995
4 Pressure Gradients
As indicated previously in Paragraph 2(b) under Overburden Pressure, it is common practice to express subsurface pressures in terms of pressure gradients, or pressure per unit depth, psi/ft or psi/m. It should be realised that densities such as mud weights in lb/gal orSG, also express pressure gradients. These units can easily be converted to psi/ft or psi/m using the conversion constants derived earlier in Paragraph 2(a). Rearranging equation 1-3 gives: PG = P D = 0.052 X MW (1-9)
where PG = pressure gradient (psi/ft) at depth D (ft), and rearranging equation 1-6 gives: PG = P D = 1.421 X SG (1-10)
where PG = pressure gradient (psi/ft) at depth D (ft). By converting subsurface pressures to gradients relative to a fixed datum, it is possible to directly compare formation pressures, fracture pressures, overburden pressures, mud weights and equivalent circulating densities (ECDs) on the same basis (See Chapter 3). The datum chosen is usually sea/ground level for initial planning purposes. Once a rig has been allocated for the well, then the datum chosen for final well planning and whilst drilling is the rotary table level (since mud hydrostatic pressure starts from just below this level). During drilling operations, it is standard practice to express all pressure gradients in terms of equivalent mud weight (EMW) either in lb/gal or SG. This allows direct comparison of downhole pressures to the weight (density) of the mud in use. EMWs can be calculated from rearrangements of equations 1-9 to 1-11: EMW (lb/gal) = EMW (SG) EMW (SG) = = P (psi) 0.052 X D (ft) P (psi) 1.421 X D (m) P (psi) 0.433 X D (ft) (1-12) (1-13) (1-14)
NOTE: From this point on ppg will be used instead of lb/gal as the abbreviated version of pounds per gallon. Example: For a formation pressure of 5970 psi at 3500m BRT, what is the formationpressure gradient in psi/ft? What is the equivalent mud weight in ppg and SG?
1-7
March 1995
pressure depth
= 0.52 psi/ft
Equivalent mud weight from equation 1-12 EMW = 5970 (ppg) 0.052 X 3500 X 3.2808
EMW = 10.0 ppg From equation 1-13 EMW = 5970 1.421 X 3500 = 1.20 SG
1-8
March 1995
1.2
1 2
Paragraph
Tables
1.2 Average Normal Formation Pressure Gradients 1-10
1-9
March 1995
1 General
Normal formation pressure is equal to the hydrostatic pressure of water extending from the surface to the subsurface formation. Thus, the normal formation pressure gradient in any area will be equal to the hydrostatic pressure gradient of the water occupying the pore spaces of the subsurface formations in that area.
Table 1.2
1-10
March 1995
1.3
1 2 3 4
Paragraph
Illustrations
1.3 1.4 Relationship between Piezometric Surface and Ground Level for an Aquifer System Temperature-pressure-density diagram for Water illustrating Subnormal Pressures caused by Cooling an Isolated Fluid Formation Foreshortening
1-13
1-14 1-15
1.5
1-11
March 1995
1 General
Subnormal formation pressure is defined as any formation pressure that is less than the corresponding pore fluid hydrostatic pressure. A subnormal formation pressure gradient is thus any gradient less than the pore fluid hydrostatic gradient. Subnormal formation pressures are often termed subpressures.
1-12
March 1995
INTAKE AREA
SUBNORMAL PRESSURES
PIEZOMETRIC SURFACE
GROUND LEVEL
HYDROSTATIC HEAD
RESERVOIRS WEOX02.065
Figure 1.3 Relationship between Piezometric Surface and Ground Level for an Aquifer System
The temperature-pressure-density diagram for water shown in Figure 1.4 illustrates this concept. Both temperature and pressure are dependent on depth. For a normal fluid (non-isolated) which is allowed to expand and contract freely, a temperature reduction associated with a depth change would follow the path indicated (which in this example corresponds to a temperature gradient of 2.5 C/100m). A lower pressure would result but it would still be equal to the normal hydrostatic pressure. In an isolated fluid system (ie/completely sealed shales), cooling must take place along a constant density path as shown. The pressure corresponding to the lower temperature is thus subnormal. If gas is present in the pores, the effects of temperature reduction will be greater as gas pressure is much more sensitive to temperature changes than water. Mechanisms which may create a reduction in subsurface temperature include uplift, erosion or a combination of uplift and erosion.
1-13
March 1995
Figure 1.4 Temperature-pressure-density diagram for Water illustrating Subnormal Pressures caused by Cooling an Isolated Fluid
1 2 = Initial conditions at depth 1 = Conditions at depth 2
1.0gm/cc
11
0.962
0.933
0.909
10
DENSITY
PRESSURE 1000psi
PRESSURE AT DEPTH 1 5
T2 0 50 100
TEMPERATURE C WEOX02.066
1-14
March 1995
2.5
C/
100
0.877
0.98
BED B
BED C C
WEOX02.067
(f) Osmosis
Osmosis is the spontaneous flow of water from a more dilute to a more concentrated solution when the two are separated by a semi-permeable membrane. In the subsurface environment, clays and clayey siltstones can act as semi-permeable membranes. If salinity differences exist between the fluids in the sediments on either side of clay beds, then osmotic flow can occur. If the flow is from a closed volume, the pressure will decrease and may become subnormal. Likewise, if the flow is into a closed volume, abnormally high pressures may result. Osmosis is discussed in more detail in Section 1.4 of this Chapter.
1-15
March 1995
4 Summary
The various suggested causes of subnormal formation pressures can be classed as artifically caused or naturally caused. Depleted reservoirs and piezometric surface (where pressure regime depends on the surface location of the well) may be classed as artificial causes, since these subnormal pressures do not originate in the subsurface formation, but are externally influenced. Conversely, the other causes of subnormal pressure discussed have origins in the formations themselves and can be thought of as being naturally caused. It is unlikely that any one of these processes may be the sole cause of subnormal pressures in any particular area. It is probable that a number of processes have contributed to produce the subnormal pressures, particularly in the light of the dependency of the processes on depth and temperature.
1-16
March 1995
1.4
Paragraph
Illustrations
1.6 1.7 1.8 1.9 Typical Formation Pressures caused by Compaction Disequilibrium Interlayer Water and Cations between Clay Platelets Schematic of Reaction of Montmorillonite to Illite Water Distribution Curves for Shale Dehydration
1-19 1-20 1-21 1-23 1-23 1-24 1-26 1-26 1-28 1-30
1.10 Diagenetic Stages in the alteration of Montmorillonite to Illite 1.11 Abnormal Formation Pressures caused by Tectonic Compressional Folding 1.12 Abnormal Pressure Distribution around a Piercement Salt Dome 1.13 Schematic Diagram of a Mud Volcano 1.14 Abnormally High Pressure due to Reservoir Structure 1.15 Schematic Diagram illustrating Osmotic Flow
1-17
March 1995
1 General
Abnormally high formation pressure is defined as any formation pressure that is greater than the hydrostatic pressure of the water occupying the formation pore spaces. Abnormally high formation pressure gradients are thus any formation pressure gradient higher than the pore fluid hydrostatic pressure gradient. Abnormally high formation pressures are also termed surpressures, overpressures and sometimes geopressures. More often, they are simply called abnormal pressures.
1-18
March 1995
The excess pore fluids help support the increasing overburden load, thereby retarding compaction further, and resulting in abnormally high pressured formations. Abnormal pressures resulting from this process will have a gradient no greater than the overburden gradient. This is because these pressures are produced by the excess overburden load being supported by the pore fluids. If beds of permeable sandstone that are hydraulically connected to zones of lower fluid pressure are present within an overpressured zone, adjacent clays will dewater to the sand bed. The adjacent clays will compact and decrease in permeability and porosity thus restricting further dewatering of the clay beds. The local pressure gradient across these clay/sand boundaries will be significantly higher than the overall pressure gradient, but are caused purely by leakage from the clays to the sand. Figure 1.6 illustrates typical overpressures caused by compaction disequilibrium. Areas in which abnormal formation pressures associated with high sedimentation rates have been encountered include the North Sea, the Gulf of Mexico, and the Gulf of Papua.
Very high local pressure gradient adjacent to permeable zones due to low permeability of the clays
DEPTH
CLAY
Overall formation pressure parallels the overburden pressure gradient, but may not reach extrapolated pressure gradient due to leakage from the clays Extrapolated initial formation pressure (parallel to overburden pressure gradient)
CLAY
SAND
PRESSURE
WEOX02.068
1-19
March 1995
Rock Salt Deposition Continuous rock salt deposition over large areas can cause abnormal pressures that may approach overburden pressure. Salt is totally impermeable to fluids and behaves plastically at relatively low temperatures and pressures, thereby exerting pressures equal to the overburden load in all directions. The fluids in the underlying formations can not escape as there is no communication to the surface and thus the formations become overpressured. Massive rock salt deposits are commonly found in the southern North Sea with abnormally high formation pressures sometimes being encountered in formations below or within these massive salts. For instance, one BP southern North Sea well required mud weights up to 1.94 SG (0.84 psi/ft) to control a saturated salt water flow from an anhydrite formation at the boundary between the Z2 and Z3 Units of the Zechstein halite formation.
(b) Diagenesis
Diagenesis is the alteration of sediments and their constituent minerals during burial after deposition. Diagenetic processes include the formation of new minerals, the redistribution and recrystallisation of the substances within the sediments, and lithification (sediments turning into rocks).
CLAY SHEET
H O
H O
H O H H O Na + O
Na + O H H
H O
H O
H O
Ca + + H H O O H H H O O H H H H H O Ca + +
H O H H O H H H O H Na + O H
Ca + + H
CLAY SHEET
WEOX02.069
1-20
March 1995
Clay Diagenesis The major constituents of marine shales are bentonitic clays of which montmorillonite is by far the most common. Montmorillonite has a swelling (expanded) lattice structure and contains approximately 70 to 85% water at initial burial in sea floor sediments. This water is present in the form of at least four layers of molecules adsorbed between clay platelets and up to ten layers held on the outside of the platelets. The clay platelets have a negative charge imbalance due to their structure. This causes the adsorption of interlayer water together with various cations (positively charged ions), principally sodium (Na +) and calcium (Ca++). The interlayer water is shown schematically in Figure 1.7. The environment at this initial burial stage would be alkaline, rich in calcium and magnesium (and of course sodium from salt water), but poor in potassium. After further burial, compaction expels most of the free pore water (non-adsorbed) and the water content of the sediment is reduced to about 30%. As burial progresses and the temperature increases, eventually all but the last layer of structured (adsorbed) water will be desorbed to the pore spaces. This causes the clay lattice to collapse and with the availability of potassium, montmorillonite diagenesis to illite occurs. This reaction is shown schematically in Figure 1.8. It involves adsorption of potassium at the interlayer and surface sites as well as the release of a small amount of silica.
O A O A K A O A A O A ILLITE Ky AL4 (Si8-y, Aly) O20 (OH)4 = Oxygen M W K + = Magnesium = Water = Potassium = Cation eg Ca ++, Na+ 3 LAYER SHEET INTERLAYER SITES 3 LAYER SHEET
O A O
MONTMORILLONITE (Al4-x Mgx)(Si8-y, Aly) O20 (OH)4 Negatively charged plates satisfied by interlayer water and cation adsorption
WEOX02.070
1-21
March 1995
The reaction is temperature (and hence depth) dependent. Initial dehydration may occur at temperatures as low as 6 C. Most of the interlayer water is liberated between 100 C and 250 C, but some of the more structured water remains to about 300 C. Water distribution curves showing the various shale dehydration stages are shown in Figure 1.9. At the second dehydration stage (See Figure 1.9), the water that is released expands due to a density reduction from the highly structured phase to the pore phase. Thus abnormally high pressures may result, particularly if the rate of expulsion of free pore water from the clay body is less than the rate of water release from the clay interlayers. Figure 1.10 is a schematic diagram showing the stages of alteration of montmorillonite to illite. If water escape from the clay body is restricted, the silica released in the diagenetic process will precipitate in the pore spaces. This may further reduce permeability and so assist in developing abnormal pressures. Sulphate Diagenesis Diagenesis in sulphate formations (gypsum, anhydrite) may cause abnormal pressures by creating permeability barriers, a fluid source and/or a rock volume change. Carbonate reservoirs are commonly overlain by evaporite sections (salt, anhydrite). Anhydrite (calcium sulphate, CaSO4) is formed by the dehydration of gypsum (CaSO4.2 H 2O) which liberates large amounts of water. There is a 30% to 40% shrinkage in formation thickness associated with this process. If this occurs at depth and in the presence of a permeability barrier, abnormal formation pressures may result. (The anhydrite itself is totally impermeable and may act as a vertical permeability barrier.) This process may have been the cause of the high pressure salt water flow discussed under Rock Salt Deposition in (a) Depositional Causes. Here, a mud weight of 1.94 SG (0.84 psi/ft) was required to control a saturated salt water flow from an anhydrite section sandwiched between massive salt sections. The process is, however, reversible. Anhydrite can take on water to form gypsum. There is an intermediate semi-hydrate stage (CaSO4.1/2 H2O) in which the rock volume would increase by 15 to 25%. If such rehydration was to occur at depth in a closed system, very high abnormal pressures could be developed. Diagenesis of Volcanic Ash Diagenesis of volcanic ash results in three main products: clay minerals, methane and carbon dioxide. Thus formations that originally contained large amounts of volcanic ash may become overpressured due to the production of gases from the volcanic ash. Areas in which this process has occurred include the NW coast of the USA and areas of the South China Sea region (Java, etc).
1-22
March 1995
% WATER 10 20 30 40 50 60 70 80 PORE WATER PORE AND INTERLAYER WATER EXPULSION SEDIMENT SURFACE
1st DEHYDRATION AND LATTICE WATER STABILITY ZONE LATTICE WATER STABILITY ZONE
INTERLAYER WATER
STAGE 1 Before diagenesis (about 3000 6000ft, below 60C) porosity = 20 to 35% clay is 70% montmorillonite 10 mixed layer 20% other MOST WATER IS BOUND WATER LOW POROSITY
STAGE 2 During alteration to illite (100 200C) high porosity = 30 to 40% clay is 20% montmorillonite 60% illite 20% other NOTE PARTICLE COLLAPSE STAGE 3 After diagenesis and compaction (over 200C) porosity = 10 to 20% clay is 70% illite 10% montmorillonite 20% other LOW POROSITY VERY LITTLE BOUND WATER FREE PORE WATER FROM DESORBED INTERLAYER WATER CLAY RELEASES SILICA, ADSORBS POTASSIUM
VOLUME LOST
WEOX02.072
1-23
March 1995
EXTENSION EXTENSION
COMPRESSION
COMPRESSION
1-24
March 1995
Faulting Faults may cause abnormally high formation pressures in the following ways: Slippage of formations along a fault may bring a permeable formation, eg a sand bed, laterally against an impermeable formation such as a clay. Thus, the flow of pore fluids through the permeable zone will be inhibited and abnormally high formation pressures may result. Non-sealing faults may transmit fluids from a deeper permeable formation to a s hallower formation. If this shallower formation is sealed, then it will be p ressured up by the deeper formation. (See Char ged Formations in (d) Structural Courses).
Uplift If a normally pressured formation is suddently uplifted, abnormally high pressures may be generated. Uplift is not a unique cause of abnormal pressure as the process that uplifts a buried formation will also uplift the overburden. For abnormal pressures to occur, there must be a concurrent geological process that reduces the relief between the buried formation and the surface. Such processes may be piercement salt domes, shale diapirs, faulting or erosion. Note that uplift and erosion may also cause subnormal formation pressures, depending on the type of formation and the amount of cooling that the formation undergoes. (See Temperature Reduction and Decompressional Expansion in Section 1.3 of this Chapter.)
Salt Diapirism Diapirism is the piercement of a formation by a less dense underlying formation. Salt will behave plastically at elevated temperatures and pressures and due to its lower density, will move upwards to form piercement salt domes in overlying formations. This upward movement disturbs the normal layering of sediments and overpressures can often occur due to the associated faulting and folding action. Additionally, the salt may act as an impermeable seal and inhibit lateral dewatering of clays thereby further contributing to the generation of abnormal pressures. The typical distribution of abnormal pressure zones around a piercement salt dome is shown in Figure 1.12. Abnormally high formation pressures associated with salt domes have been encountered worldwide, both onshore and offshore.
Shale Diapirism As with salt diapirism, this mechanism refers to the upward movement of a less dense plastic formation. In this case, high porosity (high water content) shales behaveplastically causing the formation of shale diapirs called mud volcanoes (See Figure 1.13). In practice, wherever mud volcanoes occur, there has been rapid Tertiary and/or late Cretaceous sedimentation. This rapidly loads underlying shales of low shear strength causing the formation of mud volcanoes. Formation pressures are abnormally high. For example, pressure gradients of 0.9 psi/ft have been measured around mud volcanoes on Aspsheron Peninsula in Azerbaidzhan, USSR.
1-25
March 1995
SAND
BASIN WARD
HORIZON
HORIZON
WEOX02.074
MUD VOLCANO
UPPER MIOCENE
SEA LEVEL
MIDDLE MIOCENE
LOWER MIOCENE
5000ft
Mile
WEOX02.075
1-26
March 1995
Earthquakes Earthquakes may cause compression in subsurface formations which causes a sudden increase in pore fluid pressures. For example, the 1953 earthquake in California caused production in the nearby Mountain View oil field to double over a period of several weeks after the earthquake.
D T)]
(1-15)
formation pressure at top of zone (psi) formation pressure at bottom of zone (psi) pressure gradient of fluid in zone (psi/ft or psi/m) vertical depth to top of zone (ft or m) vertical depth to bottom of zone (ft or m)
In the example shown in Figure 1.14, the formation pressure at the oil/water contact is normal hydrostatic pressure with a gradient of 0.452 psi/ft. Using equation 1-15, the pressure at the gas/oil contact is 4850 psi which gives an abnormally high formation pressure gradient of 0.462 psi/ft. Similarly, the pressure at the top of the reservoir is 4784 psi giving an abnormal gradient of 0.486 psi/ft. Obviously, in very large structures, especially in gas/water systems with long gas columns, the overpressures developed at the top of the gas column may be very high. Indeed one documented example in Iran quotes a pressure gradient of 0.9 psi/ft (approaching overburden gradient) at a depth of only 640 ft (195m).
1-27
March 1995
DEPTH At top of reservoir: Pf = 4850 0.1 x (10500 9842) Pf = 4784psi . . . FPG = 4784 = 0.486psi/ft 9842 At gas/oil contact: Pf = 5116 0.325 x (11319 10500) Pf = 4850psi . . . FPG = 4850 = 0.462psi/ft 10500 At oil/water contact: NORMAL HYDROSTATIC PRESSURE GRADIENT OF 0.452psi/ft Pf = 11319 x 0.452 Pf = 5116psi
CAP ROCK TOP OF GAS CAP D = 3000m (D = 9842ft) GAS/OIL CONTACT D = 3200m (D = 10500ft) OIL/WATER CONTACT D = 3450m (D = 11319ft)
WEOX02.076
Charged Formations Normally pressured, or low pressured porous and permeable formations at shallow depths, may be pressured up by communication with deeper higher pressured formations. This charging of the shallower formations may take place by fluid communication along non-sealing faults behind casing in old wells, or wells with faulty cement jobs, and whilst drilling a sequence of permeable formations with very large differences in pore fluid pressures (causing recharge salt water flows). Abnormal pressures caused by recharge can be very high, especially if gas is the medium that transmits the pressure (same mechanics as gas reservoir in Reservoir Structure, but over greater depth differences). Mud weights as high as 19 ppg (2.28SG, 0.988 psi/ft) have been quoted as sometimes required for drilling through shallow charged zones.
1-28
March 1995
However, shales are not totally impermeable and the time taken to heat the shales during burial should be sufficient to allow most of the excess pressures developed to leak away. The main effect of heating during burial is to retard compaction, and aquathermal pressuring is not thought to be a major cause of abnormally high formation pressures. Thermal Cracking At high temperatures and pressures caused by deep burial, complex hydrocarbon molecules will break down into simpler compounds. Thermal cracking of hydrocarbons will increase the volume of the hydrocarbons in the order of two to three times the original volume. If contained in an isolated system, this would result in high overpressures being developed. However, there is no conclusive evidence that thermal cracking is a significant cause of abnormal formation pressures. Permafrost In arctic regions, drilling and production operations may cause extensive thawing of the permafrost. If this thawed permafrost refreezes later in the life of the well, freezeback pressures, high enough to damage the casing, may result. Obviously, this situation may be avoided by proper well planning and casing design. Freezeback pressure gradients ranging from 0.66 psi/ft to as high as 1.44 psi/ft have been recorded in Alaska. Osmosis As defined in Section 1.3 , osmosis is the spontaneous flow of water from a more dilute to a more concentrated solution when the two are separated by a suitable semi-permeable membrane. This action is represented schematically in Figure 1.15. For a given solution, the osmotic pressure (differential pressure across the membrane) is almost directly proportional to the concentration differential; and for agivenconcentration dif ferential the osmotic pressure increases with temperature. Theoretically, osmotic pressures of up to 4500 psi can be produced across a semi-permeable membrane with solutions of 1.02 gm/cc NaCl in water and saturated NaClbrine. Clay and clayey siltstone beds can act as semi-permeable membranes. If salinity differences exist in the sediments above and below such beds, then osmotic flow can occur. If the flow is into an isolated system, then a pressure increase will occur in this system. Alternatively, the osmotic pressure developed across these beds may inhibit the vertical flow of water from compacting shales, thereby contributing to the development of abnormal pressures. However, the efficiency of clay beds as semi-permeable membranes in the sub-surface environment is unknown. It is thus currently believed that osmosis is a minor cause of abnormal formation pressures.
1-29
March 1995
0 3 2 1 3
0 1 2
H2O
H2O CI-
CLAY MEMBRANE
H2O
H2O
1-30
March 1995
SALINE WATER
FRESH WATER
Na+
In the North Sea abnormal pressures occur with widely varying magnitudes in many geological formations. The Tertiary sediments are mainly clays and may be overpressured for much of their thickness. Pressure gradients of 0.52 psi/ft are common with locally occurring gradients of 0.8 psi/ft being encountered. An expandable clay (gumbo) also occurs which is of volcanic origin and is still undergoing compaction. The consequent decrease in clay density would normally indicate an abnormal pressure zone but this is not the case. However, in some areas, mud weights of the order of 0.62 psi/ft (1.43 SG) or higher are required to keep the wellbore open because of the swelling nature of these clays. This is almost equal to the low overburden gradients in these areas. In the Mesozoic clays of the Central Graben, overpressures of 0.9 psi/ft have been recorded. One BP well encountered a formation pressure gradient of 0.91 psi/ft in the Jurassic section. In the Jurassic of the Viking Graben, abnormal formation pressure gradients up to 0.69psi/ft have been recorded. In Triassic sediments, abnormally high formation pressures have been found in gas bearing zones of the Bunter Sandstone in the southern North Sea. Also in the southern North Sea, overpressures are often found in Permian carbonates, evaporates and sandstones sandwiched between massive Zechstein salts.
4 Summary
Of all the processes that may be responsible for causing abnormally high formation pressures, it is unlikely that any one will be the sole cause in any particular area. The effects of several processes will probably combine to cause the observed abnormal pressure. Certain processes are thought to be either ineffective or uncommon as causes of abnormal pressures. These include uplift (as a sole mechanism), osmosis, thermal cracking, permafrost and earthquakes. A recent report(6) has found that the most significant cause of abnormally high formation pressures in depositional basins is compaction disequilibrium, with aquathermal pressuring contributing to a small extent. Clay dewatering (diagenesis) was found to have little effect. However conditions within clays during dewatering are very similar to these developed during undercompaction; and the two processes probably occur concurrently, while undercompaction is recognised as the primary mechanism. The significance of aquathermal pressuring as a cause of abnormal pressure is temperature and hence depth dependent. This is also true of the diagenetic process. With increasing depth aquathermal pressuring is thought to be a contributory factor in all cases of abnormal pressure generation.
1-31 1-31/32
March 1995
1.5
1 2 3 4
SHALLOW GAS
Page
1-34 1-34 1-34 1-35
General Definition Origins of Shallow Gas Characteristics of Shallow Gas
Paragraph
1-33
March 1995
1 General
Shallow gas accumulations present a major hazard to drilling operations. Gas influxes taken at shallow depths cannot generally be shut-in for fear that the pressures involved will fracture the formation at the previous casing shoe, thereby causing an underground blowout, or flow around the casing to the seabed.
2 Definition
For the purposes of drilling operations, shallow gas can be defined as any gas accumulation encountered at any depth before the first pressure containment casing string is set. For well planning purposes, possible gas bearing zones at shallow depths may be identified from shallow seismic sections (bright spot technique See Section 2.2 of Chapter 2 ). These are normally used down to a depth of about 1000m below surface or mudline.
1-34
March 1995
1-35
March 1995
Suggestions for further reading: 1. EXLOG, 1981. Theory and Evaluation of Formation Pressures. Exploration Logging Inc., USA. 2. EXXON, 1975 Abnormal Pressure Technology. Exxon Company, USA. 3. FERTL, W.H., 1976. Abnormal Formation Pressures. Elsevier Scientific Publishing Company, Amsterdam. 4. FERTL, W.H. and CHILINGARIAN, G.V., 1976. Importance of Abnormal Pressures to the Oil Industry. Soc. Petrol. Eng., Paper 5946. 5. GEARHART, 1986. Overpressure. Gearhart Geodata Services Ltd., Aberdeen. 6. MANN, D.M., 1985. The Generation of Overpressures During Sedimentation and Their Effects on the Primary Migration of Petroleum. Report GCB/156/85. BP Research Centre, Sunbury. 7. SELLEY, R.C., 1985. Elements of Petroleum Geology. W.H. Freeman and Company, New York. 8. SHEPHERD, M., 1984. Forties Field: Shallow Gas Hazards in the Main Field Area. Report GL/AB/1880. BPPD Aberdeen.
1-36
March 1995
2-5
2-23
2-65
March 1995
2.1
1 2
INTRODUCTION
Page
2-2 2-2 General The Transition Zone
Paragraph
Table
2.1 Techniques used to Predict, Detect and Evaluate Formation Pore Pressures 2-3
2-1
March 1995
1 General
Knowledge of formation pore pressure is of prime importance in the planning, drilling and evaluation of a well. Good estimates of formation pore pressures and fracture pressures are required to optimise casing and mud weight programmes, and to minimise the risk of well kicks, stuck pipe, lost circulation and other costly drilling problems. The following sections describe the techniques used to predict, detect and evaluate formation pore pressures at the various stages of drilling a well. Table 2.1 summarises these techniques. Methods for predicting and evaluating fracture pressure are covered separately in a later section of this Manual. Abnormally high pressured zones are by far the most common encountered, and the most important, in drilling operations. This Chapter is therefore mainly concerned with methods of predicting, detecting and evaluating abnormally high formation pressures.
2-2
March 1995
Table 2.1
Seismic (interval velocity) Drilling rate Drilling exponents Other drilling rate methods Torque Drag MWD logs Gas levels Flowline mud weight Flowline temperature Resistivity, salinity and other mud properties Well kicks Pit levels Hole fill-up Mud flow rate Bulk density Shale factor Volume, shape and size Miscellaneous methods Sonic (int. transit time) Resistivity log Density log Other logs Wireline tests (RFT/FIT) Drillstem tests
Cuttings parameters
Wireline logs
After drilling
2-3/4 2-3
March 1995
2.2
Paragraph
2-20 2-21
Illustrations
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 Pressure/Depth Plot Schematic Diagram illustrating Seismic Reflection System and Seismic Traces Schematic Diagram showing Common Depth Point (CDP) Seismic Ray Paths Schematic Plot of Offset versus Two Way Travel Time for Common Depth Point System Example Seismic Velocity Analysis Plot Example of Stacking Velocity Data on a Seismic Section Seismic and Sonic ITT versus Depth Plots for Abnormally Pressured Well Log-log Plot of Seismic Interval Transit Time ITT Departure versus Formation Pressure Gradient
2-7 2-9 2-10 2-11 2-13 2-14 2-17 2-18 2-19 2-20 2-22
2.10 ITT Ratio versus Formation Pressure Gradient 2.11 Example of Drilling Hazard Log over Shallow Section
Table
2.2 Calculation of Depths and Interval Transit Times
2-16
2-5
March 1995
1 General
At the planning stage of a well, several early decisions are made that are directly influenced by the predicted formation pore pressure profile for the well. The magnitude of the expected formation pressure influences the pressure rating of the casing and wellhead/BOP equipment to be used, and can ultimately influence drilling rig selection. Casing programme design and mud weight programmes should be tailored to the predicted formation pressures for thewell. Other related aspects of well planning that are influenced include, cement programmes, completion equipment, contingency stocks of casing, and mud chemicals/baryte stocks to be held. Thus, accurate formation pressure predictions are required in order to optimise well planning. Good well planning will, in turn, help to minimise the risk of costly problems whilst drilling. There are normally (but not always) two sources of formation pressure information for the well location being considered. The first and most widely used is offset well data. However, in areas where there are no offset wells or they are considered to be too far away to give reasonable data, then seismic data may be used to predict formation pore pressures. Seismic analysis may also be useful in validating offset well data for the location being considered.
2-6
March 1995
Leak Off Test 30/41 Leak Off Test 30/42 500 Predicted Formation Pressure 3/10b-a
Holocene to Eocene TERTIARY TO RECENT
1000
1500
Palaeocene
2000
2500
DEPTH (m)
Upper
CRETACEOUS
3000
SG EQUIVALENT
3500
psi ft
Lower
4000
4500
2.61 1.128 2.51 1.085 2.41 1.042 2.31 0.998 2.21 0.955 2.1 0.911
Upper JURASSIC
5000
Middle Lower
SG EQUIVALENT 1.0 1.1 1.2 1.3 1.4 1.5 psi ft 5500 0 0.477 0.434 0.564 0.521 0.607
0.651
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000 PRESSURE (psi) WEOX02.078
2-7
March 1995
Exploration wells In well explored regions, such as certain areas of the North Sea, the offset well data should be reliable enough for a good estimate to be made of the pore pressure profile for the proposed exploration well. However, if the nearest offset wells are far away, then the pressure data should be treated with caution when applying it to the proposed well. If there are insufficient pressure data available for any one profile to be predicted, then the alternatives should be considered and the worst case evaluated for each particular aspect of well planning. Analysis of seismic data may be required to back-up the pressure profile predicted from offset wells. In areas where there is no offset well information or they are too far away to be of any use, then seismic data analysis may be the only method available to predict the pore pressure profile (See Paragraph 3, Seismic Data).
In exploration areas where there is a well established Company office, the predicted pressure profile is usually compiled by the Designated Resident Geologist (DRG) for the well. The pressure depth plot should be obtained as soon as possible and the data must be checked immediately by the Drilling Engineer responsible for planning the well. The DE must ensure that the pressure data is the best available, whilst also accepting that the accuracy of the data will vary depending on the number and proximity of nearby wells. In areas where there is no established exploration office, or where the pressure profile is required prior to compilation by the DRG, then the well planning DE will have to predict the formation pressure profile. The DRG or Area Geology Group must be consulted. The DRG or Area Geology Group will determine which offset wells are most geologically similar to the proposed well and hence the best source of formation pressure data. Also, geological features such as faults and unconformities in the area will be identified. These may affect the way in which the pressure data are applied to the proposed well. Notes on formation pressures from offset wells are often given in the Drilling Proposal document, together with the lithological prognosis and other pertinent data (well location, target depths, total depth etc). Petroleum Engineers should also be consulted, as they may have additional pressure information, especially regarding expected reservoir pressures.
3 Seismic Data
In hydrocarbon exploration, seismic data are mainly used to identify and map prospective reservoir traps and to estimate the depths of formation tops in the lithological column. Seismic data can also be used to predict formation pressures quantitatively, or at least to give an indication of the entrance into abnormally high pressures. In new or relatively unexplored areas, seismic data are often the only information available from which pore pressure data can be derived. Seismic data can also be used to indicate the possible presence of shallow gas bearing sands. This is done using data from high resolution shallow seismic surveys which are normally used down to a maximum depth of about 1000m below surface or the mudline.
2-8
March 1995
3.1
Refl C
Time
Refl B Refl A
Up hole time
First breaks 8 9 10 11 12
1 Shot Moment
Geophones
Shot Point
Geophones
V1
WEOX02.079
Figure 2.2 Schematic Diagram illustrating Seismic Reflection System and Seismic Traces
2-9
March 1995
With increasing depth and compaction, the density and elasticity of shales increases which results in increasing seismic velocity with depth. Overpressured shales are undercompacted. This results in lower density and elasticity for that depth. The seismic velocity in overpressured shales is thus lower than in normally pressured shales at similar depths. Thus we need formation interval seismic velocity data to detect and evaluate overpressured shales. These data are readily available from seismic surveys. Seismic data are acquired by creating acoustic waves, by some form of explosion (orimplosion), and measuring the time taken for the wave to travel down to subsurface reflecting beds and back to the surface. The surface point of origin of the wave is calledthe shot point and the reflected waves are detected at surface by a series of geophone (or hydrophones if offshore) placed at known distances from the shot point. The system is shown schematically in Figure 2.2, together with the seismic traces recorded by the geophones. The whole system is moved across the surface and the measurements are repeated from a new shot point. The process is continued along a pre-determined seismic line. By using the geometric relationships between the shot points and geophone positions, it is possible to identify a series of seismic traces that have approximately the same reflection point on a reflecting bed. This point is known as a common depth point (CDP), and the seismic paths associated with this point are shown in Figure 2.3. For clarity, only the first reflecting bed is shown, but obviously the deeper reflecting beds will also have corresponding CDPs vertically below, the reflections from which will appear on the series of seismic traces. The distance between the shot point and any particular geophone is termed the offset.
Reflecting bed A
Figure 2.3 Schematic Diagram showing Common Depth Point (CDP) Seismic Ray Paths
2-10
March 1995
Figure 2.4 Schematic Plot of Offset versus Two Way Travel Time for Common Depth Point System
Offset, x to
Time, t
The equations of the dashed lines through the seismic reflections are of the form: x = V t2 - to2 where to = vertical two way reflection time to reflecting beds (ie offset, x = o) V = stacking velocity (average velocity) Thus the stacking velocity, V, is the variable defining the hyperbolae which best fit the seismic reflections.
WEOX02.081
2-11
March 1995
Reflecting Beds
In practice, the seismic traces from the same CDP are collected together to form a gather in which seismic traces at the various offsets are plotted against the reflection time. A simplified schematic plot of offset versus reflection time is shown in Figure 2.4. With greater offset, the path length of the wave is longer (See Figure 2.3) and the reflection time for the same reflecting bed increases. Curves can be drawn through the peaks on the seismic traces, corresponding to the same reflecting beds, as shown by the dashed lines in Figure 2.4. The geometry of the CDP seismic system is such that the equation of the curve through the seismic peaks (known as a seismic event) from a horizontal reflector should be a hyperbola. The variable defining the shape of the hyperbola is called the stacking velocity or the normal moveout velocity. In practice, the peaks on the seismic traces do not lie exactly on a hyperbola. Velocity analyses are performed to determine the velocity value that gives a best fit hyperbola to the data. This is done by investigating the hyperbolic function with a range of stacking velocities at increasing time increments, and comparing the result to the actual data from the seismic traces on the gather. The results from the velocity analysis are output in the form of a plot of stacking velocity versus reflection times. A typical example plot from an actual analysis is shown in Figure 2.5. The plot appears as a series of contours defining a number of peaks. Due to the mathematical computations involved in the analysis, the peaks represent the best fit stacking velocity values and the corresponding vertical two-way reflection times for each reflecting bed. The stacking velocities obtained are not the true average velocities from the surface to the reflecting bed. However, the stacking velocity is usually considered to approximate to the root mean square (RMS) velocity (as indicated on the horizontal axis in Figure2.5). The RMS velocity is the average velocity along the actual path of the seismic wave. In many cases, this is also considered to be equal to the vertical average velocity from the surface to the subsurface reflecting bed. Thus, the velocity-time pairs (as they are called) from the velocity analysis can be used to calculate the depths of the reflecting beds. The stacking velocities are used to compute the vertical two-way reflection times for each of the seismic traces on the seismic gather. The seismic gather can then be stacked to form one complete seismic trace for that particular CDP. A seismic section is then produced by displaying the stacked traces for each CDP along a seismic line. The stacking or RMS velocities are also used to calculate the interval velocities between reflecting beds, which is the property that we require to detect and evaluate abnormal pressure.
2-12
March 1995
10000
11000
12000
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
13000
WEOX02.082
4000
5000
6000
7000
8000
2-13
March 1995
9000
SP 561 TIME 0 200 300 650 1150 1450 1700 1850 2050 2200 2350 3100 5000 VRMS 1470 1470 1527 1685 1986 2218 2368 2668 2750 2850 2851 2930 3150 VINT 1470 1635 1809 2320 2942 3098 4923 3416 3972 2866 3165 3479
140 550
146 600
LINE CB-82-41
WEOX02.083
2-14
March 1995
The first step in the method is to obtain the stacking velocity data for a range of CDPs near to the proposed well location. The stacking velocities used for these CDPs should be given in panels displayed above the surface line on the seismic section. An example is shown in Figure 2.6. At this point, it is worth checking the stacking velocities given in the panels against the velocities obtained from the CDP velocity analyses. This is because stacking velocities are chosen to produce a good CDP stack (clean appearance) and may not be equal to the values that would be obtained from a velocity analysis such as that in Figure 2.5. A geophysicist should be consulted to decide which stacking velocities should be used, although more often than not, the velocities given in the panel on the seismic section will suffice. The interval velocities are then calculated from the two-way time and stacking velocity (average velocity) using Dixs formula: (V i12)2 = where Vi12 t1 t2 V1 V2 t 2(V 2)2 t 1(V 1)2 t 2 t1 = = = = = interval velocity between reflecting beds 1 and 2 (m/s) two-way travel time for reflecting bed 1 (s) two-way travel time for reflecting bed 2 (s) average velocity to reflecting bed 1 (m/s) average velocity to reflecting bed 2 (m/s) (2-1)
In the example shown in Figure 2.6, the interval velocities have already been computed using Dixs formula. The depths to the reflecting beds are calculated from: D = t.V 2 where D = depth of the reflecting bed (m) t = two-way travel time for the reflecting bed (s) V = average velocity to reflecting bed (m/s) Note that the two-way time in the panel in Figure 2.6 is given in milliseconds (ms). This needs to be converted to seconds for use in equation (2-2) (1ms = 10 3 sec). A table should be drawn up as shown in Table 2.2. The final step in the calculations is to convert interval velocities, a term used by geophysicists, into interval transit times which is a term more familiar to drilling engineers. This is done by simply taking the reciprocal of the interval velocity. Note that interval transit times are expressed in micro-seconds per metre ( sec/m) (1 sec = 10-6 sec). A plot of interval transit time (ITT) versus depth can then be constructed. The interval transit time is plotted as a vertical line over the depth interval, for which it was calculated. This results in a plot similar to a sonic log plot but in which the data are averaged over long sections and not, as with the wireline sonic log, over a few feet only. A plot of the data from Table 2.2 is shown in Figure 2.7. The corresponding wireline sonic log plot is also shown for comparison. Note that ITT is plotted on a logarithmic scale and depth on a linear scale. The types of scales that are used are discussed further in (c) Interpretation. (2-2)
2-15
March 1995
Two-way time t (millisecs) 0 200 300 650 1150 1450 1700 1850 2050 2200 2350 3100 5000
Average (stacking) velocity V (m/s) 1470 1470 1527 1685 1986 2218 2368 2668 2750 2850 2851 2930 3150
Depth
Int. velocity (Dixs formula) Vi (m/s) 1470 1635 1809 2320 2942 3098 4923 3416 3972 2866 3165 3479
Int. transit time ti ( sec/m) 680 612 553 431 340 324 203 293 252 349 316 287
D (m) 0 147 229 548 1142 1608 2013 2468 2819 3135 3350 4542 7875
Table 2.2
(c) Interpretation
As stated, overpressured shales have lower interval velocities, and therefore higher interval transit times than normally pressured shales at the same depth. The normal shale compaction trend line on the ITT depth plot decreases with depth. Thus an increase in interval transit time away from the normal trend line indicates the presence of abnormal pressures. This is shown by the shaded section in Figure 2.7. From the seismic ITT plot (stepped profile), the top of the abnormal pressures would probably be estimated to be at 2300m to 2500m. When the well was drilled the top of the abnormal pressures was found to be at about 2000m. There is a certain amount of conflict surrounding the types of scale that should be used for plotting ITT data. The format used in Figure 2.7 assumes that the normal compaction trend is a straight line on semi-logarithmic scales. This method is recommended by Fertl(17), as it enables ITT data to be directly compared with other pressure indicators that are plotted using the same linear depth scale (composite plots). Alternatively, Pennebaker(25) suggested that the normal compaction trend should be a straight line on log-log scales. An example plot of this format is shown in Figure 2.8.
2-16
March 1995
Figure 2.7 Seismic and Sonic ITT versus Depth Plots for Abnormally Pressured Well
LITHOLOGY
500
SEISMIC DATA
1000
1500
DEPTH (metres)
2500
3000
4000
mal
com
pac
tion
tren
Nor
sandstone
4500
100
200
300
400
500
600
800
WEOX02.084
2-17
March 1995
NORMAL TREND
DEPTH
TOP OF OVERPRESSURE
2-18
March 1995
Another major problem in interpreting seismic ITT plots is the placing of the normal compaction trend line. Referring to Figure 2.7, it would be most difficult to determine the exact position and gradient of the normal compaction trend line from the seismic data alone. The various non-shale lithologies affect the data quite considerably and even with the actual sonic log from the well overplotted, the correct position of the normal compaction trend line is still open to debate. One possible solution to this problem is to make numerous seismic ITT (and sonic log ITT, if available) plots for the region being investigated. It may then be possible to determine the position and gradient of an average normal compaction trend line for the region. A full discussion of other problems associated with the interpretation of seismic ITTplots is given by Barr (2) and are further discussed in relation to sonic log plots in Section 2.4 of this Chapter.
1.0
GULF COA ST
2.25
0.9
W
ILC
OX
WE XAS ST TE DELA WARE
BASIN
0.8
S CK BU RG
FR IO
1.75
S EA
AS
0.7
LF GU
VI
SO
UT
I CH
CO
NA
0.6
NO
H RT
SE
1.50 (FRIO, VICKSBURG, AND WILCOX SOUTH TEXAS GULF COAST AREAS) 1.25
0.5 1.00 0.4 0 10 20 30 40 50 60 70 80 abnormal t pressured shale normal t pressured , u sec/ft shale
WEOX02.086
2-19
March 1995
2.00
0.4
0.5
1.25 0.6
1.50 0.7
1.75 0.8
2.00 0.9
2.25
1.0
1.2 T/ Tn
1.4
1.6
WEOX02.087
Detailed high resolution seismic surveys as well as conventional seismic data are used to identify potential gas bearing zones at shallow depths by using a technique known as bright spot analysis. The high resolution seismic data are acquired over a survey grid with perhaps only 150m between seismic lines, the grid covering an area of only a few square kilometres around a proposed well location. The data are processed to produce detailed seismic sections usually down to a maximum depth of about 1000m.
2-20
March 1995
Gas bearing formations may produce high amplitude anomalies on the seismic reflection traces of the seismic section. These high amplitudes (relative to the other seismic reflections) are caused by strong seismic reflections due to the velocity impedance contrast between the gas bearing formation and the overlying formations. These amplitude anomalies appear visually on the seismic section as bright areas. The lateral extent of the bright spots can be mapped on a horizontal section, or sections, and the area of the proposed well location examined in detail. It may be necessary to move the well location to avoid drilling into a possible shallow gas zone as indicated by a bright spot. It must be noted that the high resolution seismic technique cannot usually detect a gas sand that is less than 2 to 3 metres thick, although such a thickness of gas accumulation may be enough to cause a shallow gas blowout. Also, the absence of bright spots does not mean that there will be no shallow gas and conversely, bright spots do not always contain gas. However, it is wise to avoid drilling through any bright spots if possible. Ideally, the Geophysicists must be responsible for analysing the shallow seismic data at the proposed well location. Once the well location has been finalised, the Drilling Engineer should liaise closely with the Geophysicists and Geologists to produce a drilling engineering hazard log over the depths covered by the shallow seismic survey. An example hazard log is shown in Figure 2.11. It will not be possible to predict formation pressures for shallow gas formations from the seismic data. However, drilling personnel should always be aware that shallow gas bearing formations may be overpressured, though this is not normally the case.
4 Summary
The importance of reliable formation pressure data must be stressed. It is the responsibility of the well planning DE to ensure that the pressure data used are the most accurate available. Whenever possible, pressure data from offset wells should be used to predict the pore pressure profile for well planning. Direct pressure measurements such as those from RFTs, drillstem tests and well kicks should give more accurate data than pressures derived from well logs. Seismic methods of pressure prediction should only be used in the absence of offset well data. Occasionally, seismic analysis may be necessary to endorse the data from offset wells, although there is no guarantee that this will be successful. A recent development by Geochemistry Branch at Company Research Centre, Sunbury is worthy of note. A compaction model has been developed that may have an application for predicting formation pressures. This model may be useful for pressure prediction in areas with very few or no offset wells, especially if used in conjunction with seismic data. At present, the model is being validated against actual well data.
2-21
March 1995
RTE
SEABED
100
18 5/8in (580m)
600 620
800
850
FAULT
1000
WEOX02.088
2-22
March 1995
2.3
Paragraph
4 5 6 7
Illustrations
2.12 Example showing Increase in Penetration Rate on Entering an Abnormally High Pressure Zone 2.13 Effect of Lithology Variation on Penetration Rate 2.14 Effect of Bit Condition on Penetration Rate when Drilling into an Overpressured Zone 2.15 Schematic Diagram showing Typical response of Corrected d-exponent in Transition and Overpressured Zones 2.16 Schematic Diagrams showing Various Typical d c-exponent Responses 2.17 Schematic Diagram showing dc-exponent Response to Overcompaction caused by Ice Sheet Loading 2.18 Example of Formation Pressure Determination from the d c-exponent plot using the Ratio Method 2-26 2-27 2-28
2-23
March 1995
Illustrations
2.19 Example showing the Equivalent Depth Method for Formation Pressure Determination from dc-exponent Plots 2.20 Example showing Formation Pressure Determination from the dc-exponent Plot using Lines Constructed from the Eaton Equation 2.21 Example showing the Normalized Penetration Rate Method for Determination of Formation Pressures 2.22 Schematic Diagram showing Mud Gas Levels as an Indicator of Formation Pressures 2.23 Example of Mud Gas Levels showing Trip Gas, Kelly Gas (Kelly Cut), and Recycled Trip Gas 2.24 Schematic Diagram showing Theoretical Geothermal Gradients and Temperature Profile through an Overpressured Zone 2.25 Schematic Diagram showing Expected Flowline Temperature Response on Drilling through an Overpressured Zone 2.26 Example Flowline Temperature Plots showing Raw Data Plot, End-to-end Plot and Trend-to-trend Plot 2.27 Example Horner Temperature Plot for Estimation of True Bottomhole Temperature (BHT) 2.28 Example of Typical Response of Differential Mud Conductivity/Delta Chlorides 2.29 Schematic Shale Bulk Density/Depth Plot 2.30 Variable Density Column for Measuring Shale Bulk Density 2.31 Response of Shale Bulk Density/Depth Plots in Overpressures caused by Various Mechanisms 2.32 Shale Factor/Depth Response to Overpressure caused by Compaction Disequilibrium and Clay Diagenesis 2.33 Characterisation of Shale Cavings Caused by Underbalanced Conditions and Stress Relief 2.34 Mud Logging Unit Functions and Information Flow Diagram 2-36
2-49
2-49 2-50 2-51 2-53 2-54 2-55 2-56 2-58 2-59 2-62
Table
2.3 General Mud Logging Sensor Specifications 2-63
2-24
March 1995
1 General
The aim of formation pressure evaluation whilst drilling is to determine the optimum mud weight to contain any formation pore pressures encountered, whilst maximising rates of penetration and minimising the hazards of lost circulation and drillstring differential sticking. To achieve this, formation properties have to be closely monitored in order to detect any changes that may indicate the transition from a normally pressured zone to an abnormally pressured zone or vice versa. Abnormally pressured zones may exhibit several of the following properties when compared to normally pressured zones at the same depths. Higher porosities Higher temperatures Lower formation water salinity Lower bulk densities Lower shale resistivities Higher interval velocities Hydrocarbon saturations may be different (ie higher saturation)
Any measureable parameter which reflects the changes in these properties may be used as a means of evaluating formation pressures. The parameters commonly used to evaluate formation pressures while drilling are listed in Table 2.1. It should be remembered however, that the above properties also vary with differing lithologies. Lithological variations should always be taken into account when interpreting changes in drilling and mud parameters. As the aim of formation pressure evaluation whilst drilling is to reduce the risk of taking well kicks, this section concentrates on the techniques used to achieve this. The pressure evaluation techniques in Table 2.1 that are associated with kicks are not discussed here.
2 Drilling Parameters
2.1 Rate of Penetration
Rate of penetration varies with the weight on the bit, rotary speed, bit type and size, hydraulics, drilling fluid properties and formation characteristics. If the weight on bit, rotary speed, bit type, mud density and hydraulics are held constant, then the rate of penetration (ROP) in shales will decrease uniformly with depth. This is due to the normal compaction increase in shales with depth. However, the undercompaction present in transition and abnormally pressured zones, together with the reduction in differential pressures across the bottom of the hole, result in an increase in penetration rate. It should also be noted that slower penetration rates have often been observed in the cap rock (pressure seal) overlying transition zones.
2-25
March 1995
The increase in ROP on drilling into a transition zone can be best seen on a plot of ROP versus depth. The average ROP over 0.5 to 2m depth increments (depending on whether thedrilling is slow or fast) is plotted as shown in Figure 2.12. A normal compaction trend can be established in shales as shown. A new trendline must be established for each new bit r un. An increase in penetration rate away from the normal compaction trend may indicateabnormal pressures provided that the drilling and mud parameters, and lithology , remain constant.
ROP
DEPTH
NEW BIT
TOP OF OVERPRESSURES
WEOX02.089
Figure 2.12 Example showing Increase in Penetration Rate on Entering an Abnormally High Pressure Zone
Complications arise due to lithology changes and bit dulling. Sandstone usually drills much faster than shales. This is normally shown by a sharp increase in ROP as the sandstone is penetrated. This effect, known as a drilling break is shown schematically in Figure 2.13. The normal compaction trend must be established over the shale sections only. Drilling breaks must always be flow checked regardless of whether the current estimated pore pressure gradient is less that the mud weight. Occasionally, the transition zone may be only a few metres thick if there is a very good pressure seal. This may make it very difficult to identify an increase in ROP as being one due to increased pore pressure, because it may be masked by a drilling break.
2-26
March 1995
ROP
DEPTH
WEOX02.090
2.2
Drilling Exponents
From the preceding discussion on ROP curves, it is clear that a method of accounting for the effect of drilling parameters is desirable in order to make ROP a better indicator of abnormal pressures. The d-exponent attempts to achieve this.
2-27
March 1995
sand shale
DEPTH
WEOX02.091
Figure 2.14 Effect of Bit Condition on Penetration Rate when Drilling into an Overpressured Zone
(a) d-Exponent
In 1965, Bingham(4) proposed a generalised drilling rate equation to relate all the relevant drilling parameters:
ROP = a WOB N B
where ROP N B WOB a d = = = = = =
(2-3)
penetration rate (ft/min) rotary speed (rpm) bit diameter (ft) weight on bit (lb) rock matrix strength constant (dimensionless) formation drillability exponent (dimensionless)
2-28
March 1995
Jorden and Shirley (21) rewrote equation 2-3 for d, the drillability exponent. They inserted constants to allow the use of more common oilfield units and let the matrix strength constant, a, equal 1. This removed the need to derive values for the matrix strength constant, but made d-exponent lithology dependent:
log ROP 60N d= 12WOB log 106 B
= = = = =
drillability exponent (d-exponent) (dimensionless) penetration rate (ft/hr) rotary speed (rpm) bit diameter (in.) weight on bit (lb)
NOTE: The constant 106 is simply a scaling factor inserted in the equation in order to give values of d in a convenient workable range, normally about 1.0 to 3.0. In constant lithology, d-exponent will increase with depth as the ROP decreases due to the increased compaction and differential pressures across the bottom of the hole. However, when an overpressured zone is penetrated, compaction and differential pressure will decrease and will result in a decrease in d-exponent. Hence d-exponent is, in general, related to the differential pressure at the bottom of the hole which in turn is dependent on pore pressure.
FPG N ECD
(2-5)
where dc = corrected or modified d-exponent (dimensionless) FPGN = normal formation pressure gradient (ppg, SG) ECD = equivalent circulating density (ppg, SG) This correction has no theoretical basis but has been successfully used worldwide. ECD should be used whenever possible but use of the actual mud density has been found to be acceptable. The response of d-exponent in overpressure is shown schematically in Figure 2.15. The dc-exponent may be plotted with either semi-log or linear co-ordinate axes. Either system will produce an approximately linear, normal compaction trendline, as indicated in Figure 2.15. In practice, the semi-log co-ordinate system gives a more efficient data display (values of dc are normally in the range 0.5 to 2.0) and is a more suitable format for making formation pressure estimates from dc-exponent. A d c-exponent plot should be commenced as soon as drilling begins. Values should be calculated at 0.5 to 2m intervals, depending on penetration rate. This is normally done automatically by the Mud Loggers computer and displayed as required. The values may also be plotted up automatically to enable trends to be spotted as early as possible.
2-29
March 1995
NORMAL PRESSURE
DEPTH
TRANSITION ZONE
OVERPRESSURED ZONE
dc
WEOX02.092
Figure 2.15 Schematic Diagram showing Typical response of Corrected d-exponent in Transition and Overpressured Zones
The normal dc trendline should be established as soon as possible in order that transition zones to abnormal pressures can be recognised as they are being drilled. However, it is often difficult to precisely establish the normal dc trendline due to scatter in the dc values calculated. This variation in d c values is mainly caused by: Lithology As previously stated, d-exponent increases with depth and compaction in constant lithology. This implies that d-exponent is mainly applicable to shales. Changes inlithology will thus cause changes in the value of d c. If the lithology change is relatively minor, such as silty shales, then a slight decrease in dc values may be observed which may not affect the overall trend significantly. Cuttings analysis should help to identify true shale points for use in establishing the normal trend if the dc values show a largescatter .
2-30
March 1995
SAND
MUDSTONE SAND
DEPTH
MUDSTONE
CALCITIC MUDSTONE
MUDSTONE
OVERPRESSURE
OVERPRESSURE
MUDSTONE SAND
MUDSTONE
CALCITIC MUDSTONE
MUDSTONE
dc
dc
(c)
ROCK BIT 12 1/4in / 25 000 lb W/B = 2040 lb/in SMOOTHED CURVE RAW DATA
(d)
SMOOTHED CURVE
DEPTH
INSERT BIT
DEPTH
ROCK BIT
RAW DATA
dc
dc
(e)
(f)
NEW BIT
DEPTH
NEW BIT
DEPTH
NEW BIT
OVERPRESSURE
NORMAL PRESSURE
FRESH BIT
DULL BIT
dc
dc
WEOX02.093
2-31
March 1995
For major lithological variations, such as interbedded sandstone/shale, the normal trend must be developed through the shale sections only. The increased ROP in sand sections will give sharply decreased dc values. (It may be possible to develop normal trendlines for the various other lithologies but these are of little use in overpressure evaluation and may only serve to confuse matters.) The important message here is that lithology variations must be taken into account when interpreting dc-exponent plots. The response of d c in various lithologies is shown schematically in Figure2.16 (a) and (b). Hydraulics Changes in drilling hydraulics may produce changes in dc-exponent. This also applies to formations that are susceptible to jetting. Therefore, it is often impossible to establish a normal dc trend in soft, unconsolidated sediments, such as those commonly drilled in offshore top hole sections. Bits The different drilling actions of different types of bits, ie mill tooth or insert, can cause variations and trend shifts in dc. It is sometimes necessary to plot a smoothed curve to account for trend shifts as shown schematically in Figures 2.16 (c) and (d). Changes in hole size will also produce a trend shift in dc. The effect of bit wear is to produce an increase in dc values towards the end of the bit run, as indicated in Figure 2.16(e). The new bit should give a new dc trend that continues along the previous trend provided that it is the same type of bit and none of the other parameters have varied significantly. The effect of drilling into an overpressured zone as the bit dulls is shown schematically in Figure 2.16 (f). A dull bit may mask the decrease in dc which would be expected if the bit was fresh. In extreme cases, bit dulling may totally mask or even produce an increase in dc values even though an overpressured zone has been penetrated. Thus it can be seen that the position of normal trends should be established with great care, as should the practice of shifting trends from raw data to produce smoothed curves. Two further noteworthy phenomena that may cause difficulty in interpreting the plotsare: Unconformities/Disconformities The presence of an unconformity/disconformity in the geological age of formations being drilled will often change the character of the normal trendline. The different compaction histories and sedimentary conditions of the formations above and below an unconformity/disconformity may result in not only a shifted normal dc trendline, but also a change in slope. A new trendline should be established after drilling through an unconformity/disconformity.
2-32
March 1995
1.2
1.3 SG
G 1S 1.
SG
OVERCOMPACTED
DEPTH
OVERPRESSURED WEOX02.094
Figure 2.17 Schematic Diagram showing dc-exponent Response to Overcompaction caused by Ice Sheet Loading
Ice Sheet Compaction Ice sheet compaction can often cause a good normal compaction trend to be established at shallow depths in top hole sections. This is due to the increased compaction of the near surface sediments caused by the weight of a once present overlying ice sheet. This may lead to a normal d c trend being developed through dc values that are too high. The compacting influence of the ice sheet is often dissipated after the first few hundred metres and the d c-exponent then appears to decrease to a new normal trend, falsely indicating an increase in pore pressure. This effect is shown schematically in Figure 2.17.
2-33
March 1995
The ratio method The magnitude of the formation pressure can be related to the dc deviation on the semi-log plot using the ratio method: FPG O = FPGN X dcN dcO (2-6)
where FPG O = actual formation pressure gradient at depth of interest (psi/ft, SG or ppg) FPG N = normal formation pressure gradient (psi/ft, SG or ppg) dcO = observed corrected d-exponent at depth of interest dcN = expected corrected d-exponent on normal trendline at depth of interest
Normal shale trend line Normal formation pressure Gradient is 1.08 SG SANDS
2.04
DEPTH
SG
WEOX02.095
Figure 2.18 Example of Formation Pressure Determination from the dc-exponent plot using the Ratio Method
Equation 2-6 is only valid for the semi-log dc plots as it is assumed that dc is an exponential function of depth. By rearranging the above equation into: dcO = d cN X FPG N FPG O (2-7)
2-34
March 1995
and substituting known values of FPG N and dc at various depths, it is possible to calculate a series of values of dcO , equivalent to various values of formation pressure gradient, FPGO. These series of values of dcO can be plotted on the semi-log dc plot as lines parallel to the normal d c trendline. The formation pressure gradient at any desired depth can then be estimated directly from the dc plot. Figure 2.18 shows an example d c plot with equivalent formation pressure gradient lines drawn in. NOTE: Transparent overlays ready marked with equivalent formation pressure gradient lines are sometimes available for reading formation pressures directly off the dc plot. As it is never certain exactly what depth and dc scales were used to construct these overlays, their use should be avoided in making formation pressure gradient estimates. The ratio method is a very simple method of making formation pressure estimates from dc-exponent. However, it ignores the effect of the variable overburden gradient (See Overburden Pressure in Chapter 1, Section 1.1), which controls compaction trends. This effect is reflected in the d c-exponent trend, but is considered not accurately defined by it. An alternative method of calculating formation pressures from the dc plot is the equivalent depth method. Equivalent Depth Method Due to the increase in compaction with depth, the formation matrix stress also increases, and the formation becomes harder to drill. In overpressured formations the compaction and matrix stresses are less than would be normally expected at that depth. The equivalent depth method attempts to relate these values to the depth at which they would be normal. The method assumes that the matrix stress (grain to grain contact pressure) is equal at all depths having the same value of dc. Matrix stress (M) is related to pore pressure (P f) and the overburden pressure (S) as shown by equation 1-8 (See Chapter 1, Section 1.1). This equation can be rearranged to give: Pf = S M (2-8)
This equation holds at any depth. Therefore, referring to the example dc plot in Figure 2.19, the actual formation pressure gradient (FPG O) at the depth of interest (D O) is given by: FPGO = PfO = DO SO M O D O DO (2-9)
FPGO = OPGO MO DO where OPG O MO = overburden pressure gradient at depth of interest (psi/ft) = matrix stress at depth of interest (psi)
The overburden pressure gradient is known because it is continually estimated by the Mud Loggers and updated from wireline formation density or sonic logs. (Theoverburden gradient is required for estimating fracture pressures as well as for making pore pressure estimates.) However, the value of the matrix stress at the depth of interest is unknown.
2-35
March 1995
A line is then constructed vertically upwards from the value of dc at the depth of interest until it crosses the normal dc trendline at the equivalent depth (DE), as shown in Figure 2.19. At this equivalent depth, both the pore pressure and the overburden pressure are known. Thus, equation 2-8 can be solved for the matrix stress (ME) at the equivalent depth (DE): ME = SE PfE In terms of gradients: ME = SE = PfE = OPG E FPG E DE DE DE ME = DE (OPGE FPG E) (2-11) (2-10)
where OPGE = overburden gradient at equivalent depth (psi/ft) FPGE = formation pressure gradient at equivalent depth (psi/ft) which also equals the normal formation pressure gradient at the equivalent depth FPGNE (psi/ft)
DE
DEPTH
DO
WEOX02.096
Figure 2.19 Example showing the Equivalent Depth Method for Formation Pressure Determination from dc-exponent Plots
2-36
March 1995
Since the matrix stress at the depth of interest and equivalent depth are considered equal (equal d c values), then substituting equation 2-11 into equation 2-9 gives: FPGO = OPGO D E (OPGE FPG NE) DO where FPG O OPG O OPG E FPG NE DO DE = = = = = = (2-12)
formation pressure gradient at depth of interest (psi/ft) overburden pressure gradient at depth of interest (psi/ft) overburden pressure gradient at equivalent depth (psi/ft) normal formation pressure gradient at equivalent depth (psi/ft) depth of interest (ft) equivalent depth (depth at which dc is equal to value at DO) (ft)
NOTE: Equation 2-12 can be used directly with gradients in SG, lb/gal or psi/ft and depths in metres or ft. The equivalent depth method has been successfully used to estimate formation pressures from both semi-log and linear scale d c plots. However a major flaw in the theory occurs when the equivalent depth of a particular overpressured formation is found to be above the rig floor. This will be the case if high overpressures are developed at relatively shallow depths. Also, the method relies on determining the intersection point of a vertical line with the normal compaction trendline. It therefore becomes inaccurate when the normal compaction trendline is very steep, as is usually the case on the semi-log dc plot. The Eaton Method The most accurate estimates of formation pressure from dc-exponent are considered to be obtained from the Eaton equation. This empirical equation was again developed from the basic relationship between pore pressure, overburden pressure, and matrix stress (equation 2-8). For normal pressure conditions: MN = S O PfN (2-13)
Eaton then introduced a term to relate the dc-exponent (drilling rate) response in overpressures to the reduction in matrix stress: MO = MN DcO d cN
1.20
(2-14)
(2-15)
Rewriting equation 2-13 for an abnormally pressured situation gives: MO = S O PfO Substituting equation 2-16 into equation 2-15 then gives the Eaton equation: PfO = SO (SO PfN) DcO d cN
1.20
(2-16)
(2-17)
2-37
March 1995
Dividing through by the depth (D), gives the equation in terms of gradients:
PfO = SO SO PfN DO DO DO DO dcO d cN
1.20
1.20
where FPGO, FPGN, OPGO, dcO and dcN are the same terms as explained for equations 2-6 and 2-12. By rearranging equation 2-18 and substituting known values of FPGN, d cN and OPG, it is possible to plot a series of d cO lines equivalent to various values of FPG O (in a similar manner to that previously explained for the Ratio method). An example of this construction is shown schematically in Figure 2.20. Formation pressure gradients can then be read directly from the dc plot. Eaton originally developed the equation for use in estimating formation pressures from shale resistivity plots (See Section 2.4 ), but found that it applied equally tocorrected d-exponent. The value of the exponent, 1.20, was derived from actual well data. All the methods for estimating formation pressures from dc-exponent plots rely on correct placement of the normal compaction trend. The difficulties in achieving this have previously been discussed and highlight the fact that identification of overpressured zones should not be based on dc-exponent calculations alone. Other abnormal pressure indicators, which are often more basic in nature than dc-exponent calculations, should always be checked. These indicators must support, as far as possible, any formation pressure conclusions drawn from the dc plot. Drilling factors that are not accounted for by dc-exponent are drilling hydraulics, bit tooth efficiency (bit wear) and matrix strength (lithology dependent). Also, the relationship between ROP and the various drilling parameters is not so simple as is implied by the dc-exponent equation. These factors have led to the development of more refined drilling exponents in which attempts have been made to model the various drilling/formation interactions more closely. In particular, mud logging companies have developed their own drilling exponents from which they make formation pressure estimates. Exlogs Nx (normalised exponent) and Nxb, and Anadrills A exponent are examples of these more refined drilling exponents. The theory of these drilling exponent methods will not be discussed in detail here as their formulae are of a proprietary nature and are not generally available. Suffice it to say that the methods still rely on estimating a normal compaction trend and spotting deviations from it caused by pore pressure changes and not by lithology or drilling changes.
2.3
There are a number of other drilling rate methods for estimating formation pressures that are worthy of note. As these methods are generally more complex than d-exponent methods, they have not gained wide acceptance and thus tend only to be used by their originators.
2-38
March 1995
Figure 2.20 Example showing Formation Pressure Determination from the dc-exponent Plot using Lines Constructed from the Eaton Equation
1.80
1.68
1.56
1.44
SG
NORMAL TREND
DEPTH
TOP OF OVERPRESSURE
WEOX02.097
2-39
March 1995
2990
8.53m/hr
4.11m/hr 3050 1.28 SG CIRCULATED 1.38 ECD ALL AROUND NEW BIT 1.28 SG
ECD = 1.38 SG
3080
8.23m/hr
12
Figure 2.21 Example showing Normalized Penetration Rate Method for Determination of Formation Pressures
2-40
March 1995
DEPTH (metres)
3020
As drilling proceeds, a plot of normalised penetration rate against depth is constructed. The observed penetration rate is mathematically corrected to the normalised penetration rate by applying arbitrarily chosen normal parameters according to the equation:
ROPN = ROPO
where ROP W N Pbit Q m = = = = = = =
WN m WO m
NN NO
PbitN QN PbitO QO
(2-19)
penetration rate (ft/hr or m/hr) weight on bit (lb) rotary speed (rpm) bit pressure drop (psi) mud flow rate (gpm) threshold bit weight (weight necessary to initiate formation failure) (lb) rotary exponent
Values of and m are given by Prentice(26). If the normal conditions are chosen so that most of a bit run can be drilled at these conditions, then no corrections will be necessary and ROPN will equal ROPO. Each bit run is treated as an individual unit and is plotted up as shown in the example in Figure 2.21. Changes in mud weight are also plotted separately. Drilling trends are fitted to each bit run, or part bit run, at constant ECD, as shown in the example. Provided that the ECD and formation pressure remain constant, the bit will dull and the ROPN will follow the dulling trend. If a deviation from the dulling trend is noted at constant ECD, this then indicates either a lithology change or a change in formation pressure. Lithology changes are generally abrupt, and easily identified. Formation pressure changes show a more gradual deviation from the dulling trend, as shown in the example plot at about 9950 ft and 10,100 ft. Vidrine and Benit (32) developed a graphical relationship between differential pressure across the bottom of the hole and the percentage decrease in ROP caused by this overbalance. Using this relationship, the extrapolated dulling trend ROP N and the observed ROPN at a particular depth are used to estimate the actual formation pressure at that depth. The method is detailed in full by Prentice(26) together with worked examples and a comprehensive discussion of the theory behind the method. The method is quoted as being the most responsive of all methods used to indicate the changes in formation pressure, but no data are presented to support this claim.
(b) Sigmalog
This method was developed by AGIP and Geoservices(3). Basically, it is a plot of a calculated rock strength parameter versus depth. The method is based on the following drilling rate equation (developed by AGIP):
t = WOB . N B . ROP0.25
0.5
0.25
(2-20)
where t = raw rock strength parameter and WOB, N, B and ROP are as previously defined. The raw rock strength is then corrected to the rock strength parameter, o, using experimentally derived relationships to account for depth and bottomhole differential pressure (assuming a normal formation pressure gradient). The Sigmalog is then constructed by plotting o versus depth.
2-41
March 1995
In normally pressured formations, o will increase with increasing depth and compaction. A normal compaction trend can be established and a decrease in o away from the normal trend will indicate an increase in formation pressure. When this occurs, the relationship used to correct t to o is reworked to determine the true bottomhole differential pressure (not the assumed one). The formation pressure can then be calculated from the differential pressure and the ECD for the mud weight in use. Various factors such as faults, unconformities/disconformities, poor bit efficiency, coring etc, cause shifts in the normal trend. However all the normal trends have the same slope, and the shifts of the trendlines are proportional to the shifts in the values of o. Correct shifting of the normal trendlines is thus of prime importance in calculating formation pressures from the Sigmalog. Despite this problem, it is claimed that the Sigmalog is an excellent formation pressure evaluation tool and can be applied both in shale and non-shale lithologies. The Sigmalog is commonly used by Geoservices to estimate formation pressures.
2.4
Hole Characteristics
2-42
March 1995
Torque can be useful in detecting large increases in pore pressures, for example when crossing a fault line into overpressured formations. However, sudden large increases in torque can also be caused by a locked cone on the bit, a sudden change in formation type, and by stabilisers hanging up on hard stringers. Both torque and drag are not considered to be valid overpressure indicators when drilling high angle deviated holes. Also, increases in torque due to abnormal pressures are difficult to distinguish from the normal torque increase with depth. When drilling from a floating rig the vessel motion and varying offset from the wellhead tend to produce significant torque fluctuations that make interpretation very difficult.
Hydrocarbon gases enter the mud system from various sources during the drilling of a well. The gases in the return mud stream are extracted from the mud for analysis in the mud logging unit. There is no quantitative correlation between measured gas levels and formation pressure. However, changes in gas levels can be accounted for by relating them to the actual drilling operation in progress (drilling, tripping etc) and the mud weight in use. Tentative pore pressure estimates may then be made. The main sources of gas in the mud system are: Gas liberated from drilled cuttings. Gas flowing into the wellbore due to underbalanced conditions.
The gas levels from these sources are dependent upon the formation gas saturations, the mud weight and the particular drilling operation.
2-43
March 1995
2-44
March 1995
MUD WEIGHT C
PRESSURE PROFILES
GAS LEVELS
Connection C C
DEPTH
Background Gas
C C C
C C C C C C Indicates connection
WEOX02.099
Figure 2.22 Schematic Diagram showing Mud Gas Levels as an Indicator of Formation Pressures
One major problem with this type of interpretation is to distinguish connection gas peaks caused by effective mud weight reduction due to stopping circulating, from gas swabbed into the wellbore when the drillstring is picked up. Swabbing effects are much more difficult to quantify than simple reductions from the ECD to static mud weight. This may result in higher than actual pore pressure estimates being made, especially if the connection gases observed are entirely due to swabbing. Clearly, it is good practice to use connection procedures that minimise swabbing. If used consistently, this will aid in the interpretation of connection gas levels.
2-45
March 1995
Due to the complex causes of trip gas, it may only be used qualitatively in estimating formation pressures. The early onset of trip gas after circulation is resumed may indicate that much of the openhole is slightly underbalanced. Other abnormal pressure indicators must be consulted to confirm this.
GAS LEVEL
TOTAL GAS
MUD WEIGHT
10
20
30
40
50
60
70
10
TIME
20
30
40
50
60
70
CIRCULATION STARTED
WEOX02.100
Figure 2.23 Example of Mud Gas Levels showing Trip Gas, Kelly Gas (Kelly Cut), and Recycled Trip Gas
2-46
March 1995
Kelly gas (also known as kelly cut) is caused by air being circulated around the system from a partly empty drillstring or kelly after a trip or connection. The air is pumped into the borehole as a slug of mud aerated with compressed air. This enhances any gas diffusion effects from formations to the borehole and may result in enrichment of the aerated mud with the hydrocarbon gases. A gas peak will thus be recorded when this mud is circulated back to the surface. Kelly gas due to connections is rarely seen as the kelly is usually kept full of mud during connections by closing the lower kelly cock. Kelly gas after a trip is sometimes observed (as shown in Figure 2.23) but should be easily distinguishable from other gas peaks by experienced Mud Loggers. Although indicating the presence of hydrocarbon gases, kelly gas is of no value for formation pressure evaluation. Recirculated trip gas (or any other recirculated gas) behaves in a similar way to kelly gas, and should be anticipated by the Mud Loggers from knowledge of the mud system total circulation time. An example is shown in Figure 2.23. Carbide gas is used to check the calculated total circulation time and is caused by the Mud Loggers putting calcium carbide down the drillpipe at a connection. The carbide reacts with the water in the mud to produce acetylene, a hydrocarbon gas that is detected as a large sharp gas peak when circulated round to surface. The circulation time can then be used to back calculate the openhole volume and thus to check for hole enlargement. It must be noted that evaluation of formation pressures from gas levels relies entirely on hydrocarbon gases being present to some extent in the well being drilled. Occasionally, very dry holes are drilled which may be overpressured, but show very low background gas levels. In these wells, it is very difficult to use gas levels as a reliable formation pressure indicator.
3.2
Temperature
Due to the radial flow of heat from the earths core to the surface, the subsurface temperature increases with increasing depth. The geothermal gradient is the rate at which the temperature increases with depth and is usually assumed to be constant for any given area. However, it has been found that the temperature gradient across abnormally pressured formations is generally higher than that found across normally pressured formations in the same area. This phenomenon can be explained by considering the thermal conductivity of the formations. Since water has a thermal conductivity of about one-third to one-sixth that of most formation matrix materials, then formations with a higher water content (higher porosity) will have a lower thermal conductivity. These formations will thus have a higher geothermal gradient across them. Overpressured shales usually have a higher water content than normal and will thus have higher than normal geothermal gradients across them. The top of an overpressured shale should therefore be marked by a sharp increase in geothermal gradient. This may often be reflected by an increase in the temperature of the return mud in the flowline. Also, the caprock immediately above a pressure transition zone often shows a reduced geothermal gradient due to increased compaction (higher thermal conductivity) and a lower than normal temperature at the top of the transition zone. This effect is shown schematically in Figure 2.24. Again, this may be reflected in the flowline mud temperature by a reduced flowline temperature gradient. In some cases, the flowline temperature may even fall (negative gradient) and be then followed by a large increase as the overpressured zone is penetrated, as shown schematically in the plot of flowline temperature versus depth in Figure 2.25.
2-47
March 1995
The example in Figure 2.25 is, of course, an idealised case. The flowline temperature very clearly reflects the changes in formation temperature and there are no other influences on the mud temperature. In practice, there are many other factors that affect the flowline temperature and make the interpretation of flowline temperature plots very difficult, especially offshore. Such factors include: Circulation rate. Rate of penetration. Time elapsed since the last trip (the mud in the hole heats up during a trip). Volume of the mud system. Surface treatments such as adding water, mud chemicals or weighting material. Ambient temperature (diurnal temperature changes, such as those encountered in desert regions, may cause large fluctuations in flowline temperatures). Lithology effects (sandstones and limestones generally have higher thermal conductivities than shales). Cooling effect of the sea around long marine risers.
Various methods are used to improve the interpretation of temperature-depth plots. Surface effects can be minimised by measuring the temperature of the mud in both the flowline and the suction pit (mud temperature into the hole), and then plotting lagged differential temperature. A sharp increase in differential pressures may then indicate entry into a pressure transition zone. However, the temperature trends (flowline and differential) are still found to be obscured by discontinuities at bit trips, wiper trips and other periods with no circulation. These discontinuities split the temperature depth plot into a series of unconnected depth segments, as shown in the left hand curve in Figure 2.26. Since overpressure indications are based on temperature gradient changes rather than on the magnitude of the flowline temperature, each depth segment on the temperature-depth plot can be investigated separately for gradient changes. It may, however, be helpful to plot the segments end to end, disregarding the absolute temperatures, to produce a smoothed curve. Also, end to end plotting of the individual segment trendlines may be of value, but care is required to ensure that this technique does not smooth out obvious gradient changes within an individual segment. The three techniques for plotting flowline temperature are shown in Figure 2.26.
2-48
March 1995
DEPTH
OVERPRESSURE
WEOX02.101
Figure 2.24 Schematic Diagram showing Theoretical Geothermal Gradients and Temperature Profile through an Overpressured Zone
DEPTH
Figure 2.25 Schematic Diagram showing Expected Flowline Temperature Response on Drilling through an Overpressured Zone
2-49
March 1995
NEW BITS
NB
NB NB
DEPTH
NB NB NB NB NOTE TEMPERATURE REDUCTION NB NB NB NB NB RAW DATA END-TO-END PLOT TREND-TO-TREND PLOT TOP OF OVERPRESSURE GRADIENT
FLOWLINE TEMPERATURE
WEOX02.103
Figure 2.26 Example Flowline Temperature Plots showing Raw Data Plot, End-to-end Plot and Trend-to-trend Plot
Due to the many factors affecting the flowline mud temperature, it is very difficult to interpret temperature-depth plots to evaluate formation pressures. At least, changes in the gradient of the plots may suggest that an overpressured zone has been penetrated. It is unlikely that flowline temperature will be the primary indication of abnormal pressures, though it may well be useful to support other pressure indicators.
2-50
March 1995
tL
280
270
260
250
240
WEOX02.104
Figure 2.27 Example Horner Temperature Plot for Estimation of True Bottomhole Temperature (BHT)
When drilling, the formations in the lower section of the hole are cooled by the mud in circulation. When circulation stops, the mud temperature begins to rise and gradually approaches the formation temperature. It is estimated that about four days are required for the mud temperature to reach equilibrium with the formation temperature. A modified Horner expression is used to model the temperature increase with time. By extrapolating
2-51
March 1995
the temperature increases to infinite time, it is possible to estimate the formation temperature. The Horner temperature expression is:
T = Tf c.log
where T Tf c tC tL
tC + t L tL
(2-21)
= measured temperature ( F or C) (from each wireline logging run) = actual formation temperature (F or C) = constant = circulation time at TD = time since circulation stopped
A plot of T versus log ((tC + tL)/tL) should thus give a straight line, as shown in Figure2.27. At infinite time after circulation was stopped (i.e. tL = infinity), the value of log (t C +t L)/tL) equals zero. Hence, extrapolating the plot to intercept the temperature axis gives the estimated actual formation temperature, as shown in Figure 2.27. The geothermal gradients between the logging run end points can then be calculated. Increases in the geothermal gradient may indicate the presence of abnormal pressures. Unfortunately the actual formation temperature can only be estimated at logging points. Thus, only three or four formation temperatures can be estimated from which geothermal gradients can be established. These gradients are thus average gradients over significant depth intervals and they can only be established after each hole section has been drilled. Hence, they are generally of little use in pressure evaluation while drilling, but may confirm any flowline temperature trends that were noticed earlier.
3.3
Resistivity/Conductivity/Chlorides
The resistivity of a formation depends on the porosity and the dissolved salts concentration in the formation pore water. Due to their higher pore water content, overpressured shales generally have lower resistivities than normally pressured shales at the same depths. When using water base muds, an attempt can be made to monitor this formation property by measuring the mud conductivity (conductivity is simply the inverse of resistivity). The mud conductivity at the flowline and suction pit can be measured and a conversion made to chlorides. An increase in the differential chlorides, known as delta-chlorides, may then indicate abnormal pressures. It is doubtful whether an increase in mud conductivity due to the release of pore water from drilled cuttings would be measurable. This is due to the volume of pore water released being minute compared to the volume of mud. However, pore water influxes from more permeable formations may be seen as changes in mud conductivity or delta-cholrides. Hence, a warning of underbalanced conditions may be given. The system is best suited to situations where there is a large difference between pore water and mud salinity. In these situations, the response of differential mud conductivity is similar to that of mud gas levels showing influx peaks at connections or a gradual increase due to underbalanced conditions. This is shown schematically in Figure 2.28. Obviously, mud conductivity as an abnormal pressure indicator has many limitations. A large salinity contrast between mud filtrate and formation fluids is required. Thus, the method is of little use in saline mud systems, unless of course, the mud filtrate salinity is much greater than the formation water salinity. This could be the case with saturated salt and potassium chloride (KCl) mud systems, and may well result in a mirror image plot to that shown in Figure 2.28.
2-52
March 1995
DEPTH
INFLUX AT CONNECTION
CONTINUOUS INFLUX
3.4
Continuous recording of the flowline mud weight will show mud density changes due to gas cutting or formation influxes. Some influxes are not always picked up by an increase in return mud flow or by an increase in mud pit level, especially if the influx occurs gradually due to a very low permeability formation. Thus, an underbalanced situation due to abnormal pressures may be indicated by a slight reduction in the flowline mud weight.
4 Cuttings Parameters
(a) Shale Bulk Density
The bulk density of normally compacted shales increases with depth. Overpressured shales are generally undercompacted and thus have higher porosities and lower bulk densities than would be expected. If shale bulk density is plotted against depth as drilling progresses, then a normal compaction trendline can be established. A decrease in shale bulk density away from the normal compaction trendline may then indicate the presence of an overpressured zone. A schematic shale bulk density plot is shown in Figure 2.29.
2-53
March 1995
The magnitude of abnormal pressures can be calculated from shale bulk density plots using the equivalent depth method (as described previously for d-exponent plots).
DEPTH
TOP OF OVERPRESSURES
2.4
2.6
WEOX02.106
Alternatively empirical curves, relating observed bulk density deviation from the normal trend to formation pressure gradient, can be used. However, such curves are area dependent, so can only be used if the appropriate area curve is available. Hence it will usually be necessary to use the equivalent depth method if formation pressure magnitudes are required from shale bulk density plots.
2-54
March 1995
The most common methods of measuring shale bulk density at the rigsite are: Mud Balance Shale cuttings are added to the mud balance cup until the balance reads 1.0 SG (8.33ppg) with the cap on. The cup is then topped up with fresh water and re-weighed (W). The shale bulk density is then given by: Bulk density (SG) = Density Column A graduated column of fluid is prepared from a mixture of two fluids of different densities such that the density of the mixture varies with column height. The column is calibrated using beads of known density which settle at different heights in the column. Selected shale cuttings are then dropped into the column and the height at which they settle is converted to shale density using the calibration curve. The method is illustrated in Figure 2.30. 1 2W (2-22)
250
SG 2.2
200
150
2.3 Shale 2.38
FLUID LEVEL cc
100
Shale Density
2.48
50
2.65
Figure 2.30 Variable Density Column for Measuring Shale Bulk Density
The mud balance method has the advantage of being fast and simple and uses a good quantity of cuttings to obtain a good average bulk density. The density column, however, requires selection of individual cuttings and multiple determinations to obtain an average density value. The mud balance method is probably the more representative method.
2-55
March 1995
Use of shale bulk densities for the detection and evaluation of formation pressures frequently has the following limitations: Presence of shale gas in the cuttings decreases the bulk density values determined. Cavings from higher up the hole may be part of the sample. The reliability of the data depends on the consistency and care taken by personnel, when carrying out the density determinations. Formation age boundaries and unconformities may cause shifts in the normal compaction trendline. It may be necessary to determine individual normal compaction trends for each geological age unit. Variations in the lithology, such as high carbonate content, silty/sandy shales etc, may cause significant variations in the bulk density determinations. Only good clean shales should be plotted. The presence of high density minerals, such as pyrite, will increase bulk density values and may mask the onset of abnormal pressures. Density measurements on cuttings from water base muds are usually low due to the absorption of water by the cuttings. Less reactive muds, such as oil base muds and highly inhibited water base muds, will give more accurate cuttings densities.
SHALE DENSITY
NORMAL PRESSURE
DEPTH
OVERPRESSURE
COMPACTION DISEQUILIBRIUM
CLAY DIAGENESIS
AQUATHERMAL PRESSURING
TECTONIC PRESSURING
WEOX02.108
Figure 2.31 Response of Shale Bulk Density/Depth Plots in Overpressures caused by Various Mechanisms
2-56
March 1995
The response of shale bulk density values in abnormal pressured zones will vary with the type of mechanism that caused the overpressure. This is illustrated by the idealised plots shown in Figure 2.31. However, as most overpressures in shales are caused by compaction disequilibrium and aquathermal pressuring, the most common response will be a decrease in shale bulk density at the top of an overpressured zone. (See Chapter 1 Section 1.4 for explanations of the various causes of abnormal formation pressures.)
Despite the above limitations, shale bulk density plots can be a very valuable indicator of abnormal pressures. They should be constructed during the drilling of all exploration and appraisal wells, and are most useful when long shale sections are encountered.
titrant normality
(2-23)
Pure montmorillonite clays have a high shale factor of about 100 meq/100gm. This is due to the presence of many loosely bound cations (Na+ , Ca++) between the clay platelets. However, pure illite clays, due to their tightly bound cation (K+ ) between clay patelets, have low shale factors of 10 to 40 meq/100gm. Thus, shale factor can be used to identify the montmorillonite/illite content of shale samples. For abnormal pressure evaluation, however, the use of shale factor is limited as it is dependent on the various mechanisms that may cause overpressures. Generally, shale factor decreases with depth as montmorillonite is converted to illite. In o verpressured intervals caused by compaction disequilibrium (see Chapter 1 Section1.4 ) clay dewatering has been restricted, which in turn restricts montmorillonite diagenesis to illite. Thus a larger proportion of montmorillonite will be present in the overpressured zone, resulting in an increase in shale factor. This is shown schematically in Figure 2.32(a). However, overpressures caused by clay diagenesis (montmorillonite dehydration) will show a decrease in shale factor on entering the overpressured zone. The proportion of montmorillonite has been reduced by conversion to illite, with the release of large amounts of water. This causes increased pore pressure if water escape is restricted. This shale factor response is shown schematically in Figure 2.32 (b).
2-57
March 1995
Since compaction disequilibrium is thought to be the major contributing mechanism to overpressure development in shales, the shale factor response of Figure 2.32 (a) will probably be the most dominant. However, the contribution of other overpressure mechanisms will complicate the interpretation of shale factor plots. This often results in shale factor being of little use in the detection of abnormal pressures.
SHALE FACTOR
SHALE FACTOR
DEPTH
DEPTH
OVER PRESSURES
OVERPRESSURES
Figure 2.32 Shale Factor/Depth Response to Overpressure caused by Compaction Disequilibrium and Clay Diagenesis
(c) Cuttings Character
The presence of cavings in drilled cuttings samples is an indication that the borehole wall is unstable. Cavings are much larger than normal drilled cuttings and are readily seen at the shale shakers. They are thought to be produced by two different mechanisms which result in cavings of different shapes and sizes, these two mechanisms are: Underbalanced drilling Borehole stress relief
In underbalanced drilling conditions, the pore pressure in the formation adjacent to the borehole is greater than the pressure in the borehole. In impermeable formations, such as shales, the pressure differential due to an underbalance may be high enough to exceed the tensile strength of the shales. The shale will thus fail in tension and form cavings which fall into the borehole. These cavings are usually long, splintery, concave and delicate, as illustrated in Figure 2.33 (a).
2-58
March 1995
The natural stresses that are present in the earths crust vary regionally and with depth, lithology etc. Drilling a hole through formations will relieve some of these stresses depending on the hole angle and direction in relation to the principal formation stresses. The result may be that the formation stress at the borehole wall is greater than the stress (pressure) due to the mud column. The borehole wall may then fail either in compression from vertical stresses or in tension due to horizontal stresses, or a combination of both. Cavings produced in this manner tend to be blocky and rectangular in shape, as shown in Figure 2.33 (b). Thus, the presence of cavings in cuttings samples will not necessarily mean that the hole is underbalanced. However, other overpressure indicators should always be examined in detail to confirm whether abnormal pressures are being encountered. Even if it can not be confirmed that the hole is underbalanced, it may still be necessary to increase the mud weight to regain hole stability, and avoid the problems caused by excessive amounts of cuttings/cavings being present in the hole.
SIDE SCALE
FRONT
SIDE
0.5in to 1.5in
TYPICALLY CRACKED
PLAN
WEOX02.110
Figure 2.33 Characterisation of Shale Cavings Caused by Underbalanced Conditions and Stress Relief
(d) Other Methods
Several other methods of formation pressure evaluation based on measurements on shale cuttings have been developed. These include shale cuttings resistivity, filtration rate of shale cuttings slurry, filtrate (shale water) colour index, shale cuttings moisture index, redox and pH potential of cuttings slurry and slurry filtrate. These methods are fairly complex and time consuming and thus have not gained wide acceptance as rigsite techniques. A more detailed discussion of these techniques is given by Fertl(17).
2-59
March 1995
2-60
March 1995
The combination of MWD logging techniques and downhole/surface measured drilling parameter techniques should enhance the ability to detect and evaluate formation pressures whilst drilling is in progress. Developing MWD technology is continually assessed by Drilling Division, and reports periodically issued.
These functions, and their relation to information flow through a typical mud logging unit,are illustrated in Figure 2.34. The level to which the latter function is required depends on the t ype of well being drilled. Usually exploration and appraisal wells require mud logging s ervices capable of a higher level of formation pressure evaluation than for development wells.
2-61
March 1995
Figure 2.34 Mud Logging Unit Functions and Information Flow Diagram
KELLY POSITION
DEPTH
PENETRATION RATE
CARBON DIOXIDE
HYDROCARBONS
MUD FLOW
TOTAL GAS
UV BOX
CHROMATOGRAPH
COMPUTATION DISPLAY
DATA STORAGE MUD pH/PHS REMOTE DATA DISPLAY MUD RESISTIVITY EVALUATION MUD WEIGHT FORMATION CUTTINGS DENSITY
MUD TEMPERATURE
PIT LEVEL/PVT
DRILLING PARAMETERS KELLY HEIGHT DRILL RATE TOTAL DEPTH HOOK LOAD WEIGHT ON BIT STANDPIPE PRESSURE BIT REVOLUTIONS ROTARY SPEED TORQUE CASING PRESSURE
GEOCHEMICAL ANALYSIS
CEC
MUD PRESS
FORMATION LOG
PRESSURE LOG
GEOCHEMICAL LOG
BASIC ADDITIONAL
WEOX02.111
2-62
March 1995
Table 2.3
2-63
March 1995
7 Summary
The majority of the whilst drilling formation pressure indicators discussed are only applicable to massive shale sections interbedded with sandstone/siltstones. However, as most of our drilling occurs in sedimentary basins containing such sections, then the techniques discussed are of direct relevance to our drilling operations. The most reliable abnormal pressure indicators in shales are probably d-exponent (or other drilling rate method) in combination with gas levels and cuttings character (cavings). Occasionally, one indicator may be particularly effective in showing the onset of abnormal pressures, but this will probably not be apparent until drilling has progressed well into the overpressured zone. It is stressed that all formation pressure indicators must be carefully examined to confirm the possible abnormal pressures that may be implied by a particular overpressure indicator. Also, the possibility of lithological changes should always be borne in mind when sharp changes in abnormal pressure indicators are observed.
2-64
March 1995
2.4
Paragraph
Illustrations
2.35 Schematic diagram showing the Operating Principle of the Sonic (BHC) Logging Tool 2.36 Schematic diagram showing Shale Sonic Interval Travel Time Response in Overpressures 2.37 Schematic Shale Resistivity/Depth Plot showing Response in Overpressures 2.38 Shale Resistivity/Depth Plot illustrating the Problems Associated with Formation Pressure Interpretation 2.39 Empirical Correlations for Estimation of Formation Pressures from Shale Resistivity Ratio 2-67 2-68 2-71 2-73 2-74
2.40 Log-derived Shale Bulk Density Plot on Semi-logarithmic Scales 2-76 2.41 Schematic diagram showing the RFT Pre-test and Sampling Principle 2.42 Diagram showing the Operation of the RFT Sample Probe 2.43 Example of an RFT Analogue Pressure Recording 2.44 Example of a Typical Drillstem Test String (for high pressure gas well) showing Position of Gauges 2.45 Example of a Typical Pressure Chart from a Mechanical Gauge placed below the Tester Valve in the DST String 2-78 2-79 2-79 2-81 2-83
2-65
March 1995
1 General
After each intermediate and reservoir hole section has been drilled, the formations are electrically logged to evaluate their physical characteristics and hydrocarbon potential. Some of these logs can be used to estimate formation pressures to confirm (or otherwise) the estimates made whilst the hole sections were being drilled. Formation pressures calculated from wireline logs are estimates only. Direct formation pressure measurements are normally taken in the reservoir hole section(s) using a wireline repeat formation test (RFT) tool. Also, formation pressures are directly measured in the shut-in (pressure build-up) periods during drillstem testing (DST) of potential reservoir formations.
The sonic logging tool measures the time, t, required for a compressional sonic wave to travel through one foot (or metre) of formation. This is known as the interval transit time (ITT) and is the reciprocal of formation interval velocity. The principle of operation of the sonic tool (borehole compensated (BHC) tool) is shown in Figure 2.35. Sonic pulses from two transmitters travel through the formation, and are picked up by two pairs of receivers. The time difference between sonic arrivals at each pair of receivers is measured. The average time difference is then recorded to compensate for borehole geometry and tool tilt. As discussed in Section 2.2 of this Chapter, overpressured shales show a higher sonic ITT than normally pressured shales at the same depth. Thus, a plot of sonic ITT in shales versus depth on semi-logarithmic axes should show a straight line compaction trend in normally pressured shales. Departures from this line towards higher shale ITT values indicates abnormal pressures. The normal compaction trend and sonic log departure in overpressures are shown in the schematic sonic log plot in Figure 2.36. A discussion of the problems associated with the interpretation of ITT depth plots, is given in relation to seismic ITT data in Section 2.2 of this Chapter. The main problem areas are: Scales Two types of formats have been proposed for plotting ITT-depth data. These are log-log plots (as suggested by Pennebaker(25)), and semi-log plots, as suggested above. The semi-log format is recommended as the linear depth scale enables direct comparison of sonic ITT data with other overpressure indicator plots. Normal Trend Line It is sometimes very difficult to confidently establish the position of the normal shalecompaction trend line. The depth interval over which the sonic log data are obtained in normally pressured upper hole sections is often too small to reliably establish the normal compaction trend. This is because logs are normally only obtained from below surface casing.
2-66
March 1995
UPPER TRANSMITTER
t1 t = t2 t1
t2
MUD CAKE
LOWER TRANSMITTER
WEOX02.112
Figure 2.35 Schematic diagram showing the Operating Principle of the Sonic (BHC) Logging Tool
Different lithologies frequently have vastly different sonic ITTs. Care should be taken to ensure that the normal compaction trend line is established through ITT values in good clean shale sections only. It may be necessary to make sonic log plots from several wells (if data are available) in the area of interest. These may then be used to determine the position and gradient of an average regional normal compaction trend line. The BHC sonic tool has a depth of investigation of only a few inches into the borehole wall. Hence, reactive shales that absorb water from the drilling mud, may exhibit higher ITT values (higher porosity) than would be recorded if the shales were non-reactive. These higher ITT values may falsely indicate the presence of abnormal formation pressures. A deeper reading long spacing sonic (SLS) tool is sometimes run. When available, the sonic log data from this deeper reading tool should be used in preference to those from the BHC sonic tool.
2-67
March 1995
Figure 2.36 Schematic diagram showing Shale Sonic Interval Travel Time Response in Overpressures
DEPTH
TOP OF OVERPRESSURES
t
WEOX02.113
2-68
March 1995
Unconformities/disconformities may produce a marked sudden shift in sonic ITT values and may require a second separate normal compaction trend line to be established.
Once the position of the normal compaction trend lines has been firmly established on the semi-log sonic ITT-depth plot, then the depths and magnitudes of suspected abnormal pressures may be calculated. Several methods are available for estimating the magnitude of abnormal pressures from sonic log plots:
(2-12)
= formation pressure gradient at depth of interest (psi/ft) = overburden pressure gradient at depth of interest (psi/ft) = overburden pressure gradient at equivalent depth (psi/ft) = normal formation pressure gradient at equivalent depth (psi/ft) = depth of interest (ft) = equivalent depth (depth at which sonic ITT is equal to value at DO) (ft)
NOTE: Equation 2-12 can be used directly with gradients in SG, ppg or psi/ft and depths in metres or feet. It is necessary to obtain overburden pressure gradient data for the well being investigated in order to use the equivalent depth method. These data should be available in the form of an overburden gradient-depth plot in the Mud Loggers report for the well. The advantages and disadvantages of this method are discussed in Section 2.3 of thisChapter .
2-69
March 1995
(c)
Eaton Equation The following equation was presented by Eaton(12) for calculation of formation pressures from sonic ITT plots, the derivation of which is exactly analogous to equation 2-18, which was developed for dc-exponent plots:
tN tO
3.0
(2-24)
where FPGO and OPGO are as defined above and, FPG N = normal formation pressure gradient (psi/ft) t N = extrapolated normal trend sonic ITT at depth of interest ( sec/ft) t O = observed sonic ITT at depth of interest (sec/ft) The value of the ITT ratio exponent, 3.0, was derived from actual well data. Despite the problems outlined earlier, it is considered that the use of sonic ITT data provides the most reliable method of formation pressure evaluation from well logs. The use of an empirical correlation provides the quickest method of estimating the magnitude of abnormal pressures from sonic ITT plots. However, if a correlation is not available for the area of interest, it will be necessary to use either the equivalent depth method or the Eaton equation (or both). These latter methods require overburden pressure gradient data which should be readily available in Mud Loggers reports for the well(s) under investigation.
2.2
Resistivity Log
The resistivity of shales depends on the following factors: Porosity Salinity of pore water Temperature
Temperature varies approximately linearly with depth and hence formation resistivities can be corrected for temperature. Also, the salinity of the pore water should not vary significantly with depth. Porosity is thus the major factor controlling shale resistivity. Under normal compaction (i.e. in normal pressure environments), shale resistivity increases with depth since porosity decreases. A plot of shale resistivity versus depth will thus show an increasing trend with depth. In clean shale sections, any departure from this normal trend towards lower shale resistivities may indicate an increase in porosity and hence overpressures. Shale resistivity (Rsh) is plotted on a log scale versus depth on a linear scale. The shape and slope of the normal trend line will vary with the age and type of shales present. This will lead to individual normal compaction trends being developed for each area investigated. It is unlikely that any two areas will have identical normal compaction trends. A schematic shale resistivity-depth plot is shown in Figure 2.37. The normal compaction trend line may be a curve or may approximate to a straight line over certain depth intervals, depending on the area under investigation.
2-70
March 1995
DEPTH
CAP ROCK
TOP OF OVERPRESSURE
0.4
0.6
0.8
1.0
1.5
2.0
3.0
2-71
March 1995
Originally, shale resistivities were plotted from the amplified short normal (ASN) curve of the now absolute ES (electrical survey) logging suite. Today, a variety of resistivity logging tools are run, from which shale resistivity plots may be made. The tools are designed for various depths of investigation from shallow to very deep. The deep reading tools record the true resistivity of virgin formation and thus near borehole effects (shale hydration, mud filtrate invasion in permeable zones) do not affect the resistivity values recorded. The deep reading logs that should be used for resistivity plots are the ILd curve from the dual induction laterolog (DIL) tool and the LLd curve from the dual laterolog (DLL) tool. The dual laterolog tool requires a conductive mud, so it will not work in oil base muds. The dual induction laterolog will work in oil base or water base muds and tends to be the resistivity log that is normally run. Possible problems that may be encountered with shale resistivity plots are: Only shale resistivities in thick clean shales must be plotted. It may be necessary to consult a geologist in order to pick good clean shales from the well logs. Use the deepest reading resistivity curve available to plot true shale resistivities. It may be very difficult to firmly establish the shape and position of the normal compaction trend line from the resistivity plot for just one well. An average regional trend may have to be established from the resistivity plots of many wells in the area of interest. Unconformities/disconformities and variations in geological age may show sudden changes in shale resistivities which will affect the position of the normal trend line. Changes in formation water salinity may give false pressure indications. For example, shales in the proximity of large salt masses (e.g. salt domes) have very low resistivities due to increased pore water salinity. This may indicate higher-than-actual formation pressures. Also, shales at depths less than 1000m below surface or the mudline, usually contain formation water fresher than sea water. This results in high resistivity values that may indicate lower-than-actual formation pressures.
The problems associated with interpreting shale resistivity plots are illustrated in Figure2.38. Once the normal compaction trend has been firmly established, it is possible to estimate the magnitude of any abnormal formation pressures indicated by the shale resistivity plot. Again, there are several methods available:
2-72
March 1995
DEPTH
ne
Pressure top
ou s tr en d
nd l tre rma No
Region 'B' Lithology, not pressure, change 0.1 0.5 1.0 5.0
WEOX02.115
Rsh (ohm-m)
Figure 2.38 Shale Resistivity/Depth Plot illustrating the Problems Associated with Formation Pressure Interpretation
(b) Equivalent Depth Method
This method is identical to that previously discussed for dc-exponent plots (Section2.3 ) and sonic log plots (earlier this Section). Again, equation 2-12 is valid for use with shale resistivity plots:
(2-12)
where DE = equivalent depth (depth at which shale resistivity is equal to the value at the depth of interest, DO) (ft) and FPG O, OPG O, DO, OPGE, and FPGNE are as previously defined in connection with dc-exponent plots and sonic ITT plots. As explained previously, overburden gradient data must be obtained (from Mud Loggers report) in order to use this method.
2-73
March 1995
0.4
0.5
0.6
North Sea (limited data) (Timko, 1972) 1.50 HottmanJohnson, 1965 South China Sea (Limited data) Timko, 1972 1.75
0.7
0.8 East Cameron Timko-Fertl, 1970 0.9 Eaton, 1972 (Range) 1.0 10 15 20 30 Normal R(sh)/observed R(sh) 40 50 2.25 2.00
1.25
Figure 2.39 Empirical Correlations for Estimation of Formation Pressures from Shale Resistivity Ratio
(c) Eaton equation
Equation 2-25 was proposed by Eaton (12) for calculating formation pressures from s hale resistivity plots (derivation analogous to equation 2-18, developed for dc-exponent plots):
Rsh(N) R sh(O)
1.20
where FPGO, OPGO and FPGN are as defined for equation 2-24 (sonic log plots), and Rsh(N) = extrapolated normal trend shale resistivity at depth of interest (ohm-m) Rsh(O) = observed shale resistivity at depth of interest (ohm-m) Again, the value of the shale resistivity ratio exponent, 1.20, was derived from actual well data. Overburden pressure gradients for the well are also required (from Mud Loggers well report) in order to use equation 2-25.
2-74
March 1995
where F sh = shale formation factor (dimensionless) Rsh = shale resistivity (ohm-m) Rw = formation water resistivity (ohm-m) Basically, the method involves computing a formation water resistivity (Rw) depth profile from the SP (spontaneous potential) curve in clean, shale free water sands. Values of Rsh are then obtained from thick, clean shales from whichever resistivity log is available (ILd or LLd curve). Values of Fsh at depths corresponding to the Rsh values are then calculated from equation 2-26. A plot of Fsh versus depth on semi-log scales (linear depth scale) then shows a straight line trend in normally pressured formations, F sh increasing with depth. Departure from the normal trend towards decreasing Fsh values then indicates abnormal pressures. The magnitude of any abnormal pressures can then be calculated using the equivalent depth method (as discussed in (b) above). The major drawback with this method is the calculation of R w values from the SP curve. The method is subject to inaccuracies, is difficult and is very time consuming. The advantage of this method is that it takes into account changes in formation water resistivity, R w. Other methods rely on the assumption that formation water resistivity remains relatively constant with depth. The method is detailed in full by Foster and Whalen(18) and Fertl(17). All the pressure evaluation methods using resistivity logs were developed for the US Gulf Coast and would appear to work quite well for this region. However, they have been found to be of limited use in the North Sea. Formation water salinity variations cause erratic tool responses which make it virtually impossible to construct a normal compaction trend.
2.3
Density Log
The formation density logging tool consists of a radioactive source which bombards the formations with medium-energy gamma rays. The gamma rays collide with electrons in the formation which cause the gamma rays to scatter. The degree of scattering is directly related to the electron density and therefore the bulk density of the formation. The scattered gamma rays that return to the borehole are picked up by detectors in the logging tool. In the FDC (formation density compensated) logging tool, the gamma ray source and two detectors are mounted on a skid that is pushed against the borehole wall by an eccentering arm. The skid has a plough shaped leading edge to cut through any mud cake present on the borehall wall. Any mud cake that is not removed will effect the tool reading. The dual detectors of the FDC tool automatically compensate for mud cake effects. The corrected bulk density (Pb) and the correction made () are recorded on the FDC log.
2-75
March 1995
DEPTH
CAP ROCK
TOP OF OVERPRESSURES
2.0
2.1
2.2
2.3
2.4
2.5
2.7
WEOX02.117
2-76
March 1995
A plot of shale bulk density versus depth on either linear or semi-log scales will show a straight line normal compaction trend. Since the bulk density of shales is inversely proportional to porosity, and an increase in shale porosity indicates abnormal pressures, then a decrease in shale bulk density from the normal compaction trend line will indicate abnormal pressures. The semi-log type plot is shown schematically in Figure 2.40. The densities from non-washed-out pure shale sections should be plotted. After the normal compaction trend line has been established, the equivalent depth method (See Sonic and Resistivity Logs) may be used to estimate the magnitude of formation pressures. The use of shale bulk density trends from the formation density log should be a fairly reliable overpressure indicator. However, it has been found that unless borehole conditions are ideal (uniform gauge hole), the formation density log will not be as accurate or reliable for pressure evaluation as other techniques based on sonic or resistivity logs.
2.4
Other Logs
Other wireline logs that have been used to evaluate formation pressures include the spontaneous potential (SP) log, the neutron porosity log (CNL), the thermal neutron decay time log (TDT), and also downhole gravity and nuclear magnetic resonance (NMR) logs. These techniques are discussed further by Fertl(17). Also, the use of an MWD gamma ray log for formation pressure evaluation of US Gulf Coast shales, has been discussed by Zoeller(34).
The repeat formation tester (RFT) is an electric wireline tool designed to measure formation pressures and to obtain fluid samples from permeable formations. After it has been run in the hole, the tool can be set any number of times. This enables a series of pressure readings to be taken and permits the Logging Engineer to pre-test, or probe the formation for permeable zones before attempting to take a fluid sample or a pressure recording. The RFT was developed from the formation interval tester (FIT) which is only able to take one, less accurate, pressure measurement whilst taking a sample. However, the FIT is able to take a pressure measurement/sample in cased hole by using a shaped charge to perforate the casing. A schematic diagram of the RFT pre-test and sampling principle is shown in Figure 2.41.
2-77
March 1995
CHAMBER No 1
Figure 2.41 Schematic diagram showing the RFT Pre-test and Sampling Principle
When the tool is set, a packer moves out on one side and back-up pistons move out on the opposite side. This forces the packer against the borehole wall and holds the body of the tool away from the wall to reduce the chances of differential sticking. The probe is then forced into the formation and opened by retracting the filter probe piston. This operation is shown in Figure 2.42. The two pre-test chambers are then operated sequentially, each sampling a small volume (10cc) of the formation fluid at different rates (assuming that the formation is permeable). A filter in the flowline probe prevents sand entry into the tool and the piston cleans the filter when the tool is retracted. A strain gauge pressure transducer monitors the pressure during the pre-test. The pressure is continuously recorded at surface in both analogue and digital form. An analogue pressure recording from a typical pre-test is shown in Figure 2.43.
2-78
March 1995
PISTON
FLOWLINE FILTER
UNCONSOLIDATED SAND
PROBE
WEOX02.119
Figure 2.42 Diagram showing the Operation of the RFT Sample Probe
FLOWRATE, Q
PRESSURE, P
HYDROSTATIC PRESSURE P1 P2
FORMATION PRESSURE
TIME, t
WEOX02.120
2-79
March 1995
The initial pressure (See Figure 2.43) before the tool is set is the hydrostatic pressure of the mud column. When the tool is set, the pressure rises slightly due to the compression of the mud cake by the packer. The probe piston then retracts giving a drop in pressure due to the flowline volume expansion and communication with the formation. When the piston stops retracting, there is a slight pressure rise because the packer continues to compress the mud cake until the tool is fully set. The pressure then drops again as the first 10cc pre-test piston starts to retract (at time tO). After about 15 seconds, the first pre-test chamber is full (at time t1) and the second piston begins moving at a rate 2.5 times faster than the first piston. The pressure thus drops further until the second pre-test chamber is full (at time t2 ). The pressure then builds up towards a final pressure, which is usually that of the original formation pressure(30). Finally, the probe and packer are retracted and the mud hydrostatic pressure is again measured. Thus, the RFT provides three distinct pieces of pressure data: The mud column hydrostatic pressure (two readings). The formation pressure. The pressure transient induced by the withdrawal of a small sample of formation fluid (2 x 10cc).
The two mud hydrostatic pressure readings are compared to verify the stability of the tools recording system. The two values should be within a few psi of each other. The formation pressure is used to verify estimates made whilst drilling the well and to construct a reservoir pressure profile. This will yield data on the pressure gradients and nature of the reservoir fluids. The pressure/flowrate/time data from the pre-test sample withdrawal can be used to calculate reservoir characteristics, such as permeability. Hence, the RFT provides accurate data on formation pressures. However, formation pressure data can only be obtained from permeable formations such as reservoir sandstones. These formations may or may not be at the same pressure as adjacent shales. RFTs are normally run at the request of the Geologists/Petroleum Engineers to seek information on potential reservoir formations. However, in deep high pressure wells, the RFT is being increasingly run to obtain accurate formation pressures before potentially troublesome drilling operations (such as coring) are commenced. Accurate knowledge of formation pressures in such wells allows fine mud weight adjustments to be made to minimise the risk of swab/surge pressure problems.
2-80
March 1995
Figure 2.44 Example of a Typical Drillstem Test String (for a high pressure gas well) showing Position of Gauges
DESCRIPTION
Flowhead
Tubing
Lubricator Valve Tubing 5in PIPE RAMS 5in Slick Joint Tubing 5in Slick Joint MUD LINE Tubing Downhole Safety Valve (surface controlled) Tubing Annulus pressure operated Downhole Shut-in Tool (including tubing reverse-out facilities) Tubing Nipple Tubing (2 joints) Crossover Pressure Gauge Carrier + 2 Gauges Drill Collar (1 joint) Pressure Gauge Carrier + 2 Gauges Drill Collar (1 joint) Pressure Gauge Carrier + 2 Gauges No-Go Shoulder of Seal Assembly
Permanent Packer Millout Extension Seal Assembly Seal Bore Extension Liner
WEOX02.121
2-81
March 1995
3.2
Whenever drillstem tests are carried out on potential reservoir formations, various pressure gauges are run in the hole with the test string. The purpose of these pressure gauges is to record the downhole pressure during the sequence of flow and shut-in periods that comprise the drillstem test (DST). The pressures recorded during the test are used to calculate reservoir characteristics such as formation pressure, permeability, skin damage and productivity index. Various types of pressure gauges are available. These are run in conjunction with clocks and recorders, and include: Mechanical gauges normally bourdon tube (BT) type pressure gauges with mechanical clocks and recorders. Electronic gauges strain gauge, quartz crystal or bourdon tube type pressure gauges with electronic clocks. Data are recorded on various types of electronic memories and read from the gauge on surface after the test by a special reader. Electronic surface read out (SRO) gauges strain gauge or quartz crystal type pressure gauges linked by cable to the surface where downhole pressures are continuously monitored and recorded.
The mechanical and electronic gauges can be run in various ways/positions in the test string: Set in a wireline nipple (hence retrievable during or after a test). Hung off in the tailpipe (below the packer) using a DST hanging kit. Placed in a bundle carrier or gauge carrier in various positions in the string.
The SRO gauges are always placed above the tester valve (above the packer) as they are connected to surface equipment by a cable. A typical DST string is shown in Figure 2.44 (for a gas well test). This illustrates the various positions of the pressure gauges in the DSTstring. After a DST has been successfully completed, the test string is pulled and the pressure gauges are retrieved for the pressure charts to be read. A typical valid pressure chart from a mechanical gauge placed below the tester valve is shown in Figure 2.45. Note that a linear plot of the pressures recorded by an electronic gauge should have the same general form, without the baseline. The significant events during the test (marked by capital letters) on Figure 2.45 are as follows: A: Atmospheric pressure at surface.
A-B: The gauge is run in the hole with the test string and records increasing hydrostatic pressure. The early steps effect is the result of pauses to pump the water cushion into the test string. B: C: At test interval depth, the gauge records the hydrostatic pressure of the mud column. The packer is set, squeezing the sump below the packer and causing an increase in pressure.
D-E: The tester valve is opened and the gauge is suddenly subjected to the reduced hydrostatic pressure of the water cushion alone.
2-82
March 1995
C D B G M L N
PRESSURE
I E F H J K BASE LINE
TIME
WEOX02.122
Figure 2.45 Example of a Typical Pressure Chart from a Mechanical Gauge placed below the Tester Valve in the DST String
E-F: F: The influx of reservoir fluid into the test string adds to the pressure of the partial water cushion. The tester valve is shut after an initial 5 to 10 minute short flow period.
F-G: The reservoir pressure slowly builds up. After 30 minutes, no more build up is seen. The gauge now gives an estimate of the virgin reservoir pressure (G). G-H: The tester valve is now opened again and the reservoir is exposed to hydrostatic pressure of the fluids in the test string. H-I: I-J: J-K: The reservoir flows again and the gauge pressure increases until the water cushion reaches the surface. As the reservoir fluid replaces the water cushion in the test string, the gauge pressure decreases until all the water cushion has been unloaded (J). The pressure continues to fall due to wellbore effects before steadying out as the flow into the wellbore becomes radial.
K: The tester valve is closed at the end of the second flow period. K-L: The reservoir pressure starts to build up again as it returns to equilibrium. L-M: The packer is unset at the end of the second build up period and the pressure gauge again reads the pressure of the annulus mud column. N-O: The test string is pulled out of the hole and the gauge pressures reduces. O: Finally, the gauge is back on surface and reads atmospheric pressure.
2-83
March 1995
Analysis of the pressure build up data from the shut-in periods can then give accurate estimates of the reservoir formation pressure. An example of this analysis is given in the BP Guide to Testing Operations. Thus, data from drillstem tests can give accurate estimates of formation pressures. However, the pressure data can only be obtained from permeable reservoir formations that are considered to have sufficient hydrocarbon potential to warrant the expense of a drillstem test. As with RFT pressure data, the reservoir pressure calculated from DST data may or may not be the same as the pressures in adjacent shales.
4 Summary
The most accurate estimates of formation pressures are obtained from wireline RFT measurements and drillstem test pressure data. However, these direct measurements are only possible in permeable formations such as sandstones and limestones. These methods are clearly not applicable to impermeable shale sections (where the majority of overpressures are developed). Conversely, estimates of formation pressures from wireline logs are restricted to shalesections, with assumptions made as to the pressures in any adjacent permeable sections. The recognition of a normal shale compaction trend line is of vital importance whenestimating formation pressures from log-derived shale properties. Of the various logs available, the sonic log is usually the best log for quantitative pressure evaluation as it is r elatively unaf fected by changes in hole size, formation temperature, and formation watersalinity .
Section 2 References
1. ANSTEY, N.A., 1976. the New Seismic Interpreter Videotape Manual, International Human Resources Development Corporation, Boston, Massachusetts, USA. 2. BARR, M.V., 1983. An Appraisal of Seismic Reflection Techniques for the Recognition and Prediction of Abnormal Formation Pressures. Report PEB/55/83. BP Research Centre, Sunbury. 3. BELLOTTI, P. and GERARD, R.E., 1976. Instantaneous Log Indicates Porosity and Pore Pressure. World Oil, Oct. 1976. 4. BINGHAM, M.G., 1965. A New Approach to Interpreting Rock Drillability. Oil and Gas Journal, Nov. 2 1964?Apr. 5 1965. 5. BOURGOYNE, A.T., 1971. A Graphic Approach to Overpressure Detection While Drilling. Pet. Eng. 43(9): 76?78. 6. BP, 1985. A Guide to Testing Operations. BP Exploration Co. Ltd., Operations Support Division, London. June 1985. 7. BP, 1986. A Wellsite Guide to Logging Operations. BP Exploration Co.Ltd., Logging Operations Branch, London. January 1986. 8. BP, 1985. Resident Geologists Manual. BPPD Aberdeen. 2nd Edition, Sept. 1985. 9. COCHRANE, D.F. and HARDMAN, P., 1986. Shallow Gas Hazards in Drilling Operations. Report DTG/L/1/1986. BPPD London.
2-84
March 1995
10. COMBS, G.D., 1968. Prediction of Pore Pressure from Penetration Rate. SPE Paper2162. 11. DIX, C.H., 1955. Seismic Velocities from Surface Measurements. Geophysics, 20: 68?86. 12. EATON, B.A., 1975. The Equation of Geopressure Prediction from Well Logs. SPEPaper5544. 13. EXLOG, 1980. Field Geologists Training Guide. Exploration Logging Inc., USA. 14. EXLOG, 1979. Mud Logging: Principles and Interpretations. Exploration Logging Inc., USA. 15. EXLOG, 1981. Theory and Evaluation of Formation Pressures. Exploration Logging Inc., USA. 16. EXXON, 1975. Abnormal Pressure Technology, Exxon Company, USA. 17. FERTL, W.H., 1976. Abnormal Formation Pressures. Elsevier Scientific Publishing Company, Amsterdam. 18. FOSTER, J.B., amd WHALEN, H.E., 1966. Estimation of Formation Pressures from Electrical Surveys Offshore Louisiana. SPE Paper 1200. 19. GEARHART, 1986. Overpressure. Gearhart Geodata Services Ltd., Aberdeen. 20. HOTTMAN, C.E., and JOHNSON, R.K., 1965. Estimation of Formation Pressures from Log-derived Shale Properties. Journal of Petroleum Technology, 17: 717-723. 21. JORDEN, J.R., and SHIRLEY, O.J., 1966. Application of Drilling Performance Data to Overpressure Detection. Journal of Petroleum Technology, 18: 1387-1394. 22. LESSO, W.G. and BURGESS, T.M., 1986. Pore Pressure and Porosity from MWD Measurements. IADC/SPE Paper 14801. 23. MANN, D.M., 1985. The Generation of Overpressures During Sedimentation and theirEf fects on the Primary Migration of Petroleum. Report GCB/156/85. BP Research Centre, Sunbury. 24. MINTON, R.C., 1986. Technical Specification for Drilling Mud Logging Service. ReportDTG/D/4/86. BPPD Aberdeen. 25. PENNEBAKER, E.S., 1968. An Engineering Interpretation of Seismic Data. SPE Paper2165. 26. PRENTICE, C.M., 1980. Formation Pressures from Normalized Penetration Rate Plots. Prentice and Records Enterprises, Inc., Lafayette, Louisiana, USA. 27. REHM, W.A., and McCLENDON, R., 1971. Measurement of Formation Pressure from Drilling Data. SPE Paper 3601. 28. ROESLER, R.F., BARNETT, W.C., and PASKE, W.C., 1986. Theory and Applications of an MWD Neutron Porosity Sensor. SPE/IADC Paper 16057. 29. SCHLUMBERGER, 1972. Log Interpretation Volume 1 Principles. Schlumberger Ltd., New York, USA. 30. SCHLUMBERGER, 1981. RFT Essentials of Pressure Test Interpretation. Schlumberger Ltd.,
2-85
March 1995
31. SINGH, J., 1987. A Review of Measurement-While-Drilling Systems. Report DTG/L3. BPPD London. 32. VIDRINE, D.J., and BENIT, E.J., 1967. Field Verification of the Effect of Differential Pressure on Drilling Rate. SPE Paper 1859. 33. ZOELLER, W.A., 1970. The Drilling Porosity Log DPL. SPE Paper 3066. 34. ZOELLER, W.A., 1983. Pore Pressure Detection from the MWD Gamma Ray. SPEPaper12166.
2-86
March 1995
Page
3-2 3-2 3-2 3-4 3-7 3-10 3-12
Illustrations
3.1 3.2 3.3 3.4 3.5 Hydrostatic Pressure The Effect of Flowline Elevation shown in relation to calculation of formation pressure Example Calculation of the Equivalent Circulating Density (ECD) Theoretical Variation in Swab/Surge Pressure when tripping pipe at constant speed Pressure Surges associated with Lowering Pipe into a Borehole 3-3 3-5 3-6 3-11 3-12
3-1
March 1995
1 General
Primary well control is maintained by controlling formation pore pressures with the hydrostatic pressure of the drilling fluid. Primary well control is exercised between two distinct limits; these being the maximum formation pore pressure gradient and the minimum fracture pressure gradient in a section of openhole. This Chapter is intended to outline the various factors that can influence the actual pressure exerted by the drilling fluid in the wellbore during routine drilling operations. The effect of the following is considered: Flowline elevation. Circulation. Tripping pipe.
Easy to use formulae are presented to predict the effects of these factors.
2 Hydrostatic Pressure
The hydrostatic pressure of a column of drilling fluid is determined, in theory, by the density, and vertical height of the fluid above a point of interest. The density of the drilling fluid and the height of the fluid column are related to the hydrostatic pressure as follows: Hydrostatic pressure (psi) = MW (SG) X D (m) Figure 3.1 shows a sample calculation.
X
1.421
3-2
March 1995
MUD @ 1.5 SG The hydrostatic pressure at total depth in well A and well B = Density of the (SG) x vertical depth (m) x 1.421 = 1.5 x 1000 x 1.421 = 2130 psi WEOX02.123
3-3
March 1995
It is important therefore that the effect of flowline elevation be considered when describing formation pressures in terms of an equivalent mud weight. This is because formation pressures are originally referred to sea-level, or the surface elevation, depending on whether the well is offshore or on land. Figure 3.2 shows an example of the calculation of the EMW of a normally pressured formation referenced to the flowline of a semi-submersible drilling rig.
It is clearly important to be able to estimate circulating pressure losses in order to be able to predict both the pump pressure and downhole ECD at specified circulating rates. The next paragraph details the formulae that can be used to estimate circulating pressurelosses.
3-4
March 1995
Figure 3.2 The Effect of Flowline Elevation shown in relation to calculation of formation pressure
SAND EQUIVALENT MUD WEIGHT REFERENCED TO THE FLOWLINE OF A SEMISUBMERSIBLE DRILLING RIG
SEA LEVEL
100m
SEA BED
200m
Formation pressure at 325m BRT = 1.03 x 1.421 x 300 = 439psi Normal pore pressure gradient = 1.03 SG Formation pressure at this point referenced to the flowline, in EMW = 439 = 1.421 x 325 = 0.95 SG
WEOX02.124
3-5
March 1995
HOLE STATIC
PUMP
2000m
500m
Total pressure at shoe = (1.5 x 2000 x 1.421) + 100 = 4363psi ECD at shoe = 4363 2000 x 1.421 = 1.54 SG Pressure drop = 150psi
Total pressure at TD = (1.5 x 2500 x 1.421) + 250 = 5579psi ECD at TD = 5577 2500 x 1.421 = 1.57 SG
WEOX02.125
3-6
March 1995
3. Calculate the pressure loss for the pipe section, assuming laminar flow. P= L X PV X v 8361.5 X d i2 + L X YP 68.6 X d i (psi)
4. Calculate the effective viscosity. = 8361.5 X P X d i2 (centipoise) LXv 5. Calculate the Reynolds number. Re = 422.8 X MW
X
d i2
The critical Reynolds number is assumed to be 2000 for Bingham fluids. If Re isless than 2000, the flow is assumed to be laminar and the pressure loss is calculatedusing the formula in step 3. If Re is greater than 2000, the flow is assumed to be nonlaminar and the pressure loss must be re-calculated using the formulae in steps 6 and 7:
3-7
March 1995
7. Calculate the pressure loss for the pipe section in non laminar flow. P= f
X
L X MW X v 2 315.8 X d i
(psi)
8. Calculate the critical velocity. (ie the velocity above which the flow will be non laminar)
vc = 7.76
X
PV + [7.76
(PV2 + (102.79 MW X di
YP
MW X di2 ) )2 ]
(m/min)
2. Calculate the pressure loss for the section of annulus assuming laminarflow . P= L X PV X v + L X YP 5574.32 (dhc d o)2 60.96 (dhc do ) (psi)
(centipoise)
(dhc d o)
The critical Reynolds number is assumed to be 3000 for Bingham fluids. If Re is less than 3000, the flow in this section of the annulus is assumed to be laminar and the pressure loss is calculated using the formula in step 2. If Re is greater than 3000, the flow is assumed to be non laminar and the pressure loss must be re-calculated using the formulae in steps 5 and 6: 5. Calculate the Fanning friction factor. f = 0.079 Re 0.25 6. Calculate the pressure loss for the section of the annulus in non laminar flow. P= f X L X MW X v 2 315.8 X (dhc d o) (psi)
3-8
March 1995
7. Calculate the critical velocity. (The velocity above which the flow will be non laminar)
vc =
11.63
PV + [11.63
YP X (dhc do)2))2 ]
(m/min)
2. Calculate the bit pressure loss. Pbit = where v vn Q di dhc do L PV YP MW 600 300 An P Pbit vn2 X MW (psi) 12.49 = = = = = = = = = = = = = = = = mud velocity (m/min) nozzle velocity (m/min) pump output (gal/min) ID of pipe (in.) ID of hole/casing (in.) OD of pipe (in.) length of section of pipe/annulus (m) plastic viscosity (centipiose) yield point (lb/100ft2) mud weight (SG) effective viscosity (centipiose) Fann viscometer reading at 600 rpm (lb/100ft2) Fann viscometer reading at 300 rpm (lb/100ft2) total nozzle area (in.2) section pressure loss (psi) bit pressure loss (psi)
These formulae can be used to estimate the pressure drop in each section of pipe and annulus. The standpipe circulating pressure can be estimated from the sum of the pressure losses across the bit and in all sections of the pipe and the annulus. The ECD at the bottom of the hole can be estimated from the total annulus pressure loss. The annulus pressure losses may also be estimated when circulating by subtracting the calculated pressure drop in the drillstring and the bit from the actual standpipe pressure (accounting also for surface pressure losses). This technique is likely to yield a more accurate estimate of the annulus pressure losses for the following reasons: The inside measurements of the drillstring are more accurate than the openhole internal diameter. The pressure drop through the bit is accurately modelled by the formula presented. The effect of loading the annulus with cuttings is measured directly.
3-9
March 1995
The main disadvantage of this technique stems from the fact that the majority of the pressure loss in the system is in the drillstring and across the bit. Therefore, a small error in the calculated pressure drop will cause a relatively large error in the estimate of the annulus pressure loss.
Therefore the factors that determine the magnitude of swab and surge pressures are assumed to be: The annular clearance. The viscosity of the mud. The gel strength of the mud. The speed of the pipe. The length of low clearance pipe in the hole. The position of the low clearance pipe in the hole in relation to the point of interest. The acceleration or deceleration of the pipe.
On the basis of these assumptions, typical variations in wellbore pressure due to swab and surge pressures whilst tripping pipe are shown in Figures 3.4 and 3.5. Recent studies however, have shown that steady state models are not adequate to model the behaviour of the mud while the pipe is tripped. It has been shown that swab and surge pressures are best modelled as a transient, rather than a steady state phenomenon. The transient model assumes that a pressure wave is propogated at the instant that the pipe begins to move; the wave then travels down the well at the speed of sound and is reflected back up the hole. As a result of this effect, the pressure at a point in the well oscillates. The oscillations will continue until either the pipe reaches a steady speed, or the pipe has stopped and the reflected pressure waves have diminished.
3-10
March 1995
Figure 3.4 Theoretical Variation in Swab/Surge Pressure when tripping pipe at constant speed
BIT DEPTH
BIT DEPTH
2 Influence of BHA
SURGE PRESSURE AT A
SWAB PRESSURE AT A
WEOX02.126
3-11
March 1995
A Negative Surge Pipe Lifted from Slips B Positive Pressure to Break Mud Gel C Minimum Pipe Velocity D Maximum Pipe Velocity E Negative Surge Sudden Pipe Stoppage
PRESSURE
0 C
A TIME
WEOX02.127
Figure 3.5 Pressure Surges associated with Lowering Pipe into a Borehole
The latest swab/surge software models the behaviour of the mud as a transient phenomenon and also accounts for the following factors: The compressibility of the mud. The elasticity of the wellbore. The change in rheological properties of the mud with pressure and temperature. The temperature profile in the wellbore. The elasticity of the pipe.
3-12
March 1995
The formulae used for the steady state model are relatively easy to use and, as such, may be used in the field to approximate swab/surge pressures. The following procedure should be used to calculate swab/surge pressure for either open or closed pipe: 1. Estimate the velocity of the mud for a given pipe running speed.
For closed pipe: v = CL + do 2 2 dhc do 2
X
vp
do 2 di 2 2 dhc do 2 di
vp
where v = velocity of the mud (m/min) CL = clinging constant vp = average running speed of the pipe (m/min) The clinging constant, K, is assumed to equal 0.45 in the absence of detailed formula that are used to predict this quantity. 2. Determine the maximum mud velocity. The maximum mud velocity is generally taken to be 1.5 calculated in (1)).
X
3. Determine the swab/surge pressures due to the pipe movement. The swab/surge pressure resulting from the pipe movement can be estimated by substituting the maximum annular mud velocity as calculated in (2) into the formulae for annular pressure loss (Bingham or Power Law). The swab/surge pressure is added to the hydrostatic pressure of the mud if the pipe is being run into the hole and subtracted if the pipe is being pulled. Therefore:
EMW at point of interest = MW sumP (SG) D X 1.421
where sumP = total swab/surge pressure (psi) D = vertical depth to point of interest (m) Preston Moores method can be used to approximate swab/surge pressures due to the movement of a drillstring that contains a bit with nozzles. The range of values for the resultant swab/surge pressure that are predicted by this technique should be treated with some caution, as it is generally assumed that it will predict low values of swab/surge pressures. The upper limit for swab/surge pressures for a drillstring with a bit and nozzles will be represented by the value calculated for closed pipe.
3-13
March 1995
The procedure for calculating swab/surge pressures for a drillstring that contains a bit and nozzles is as follows: 1. Calculate the velocity of the mud around the drillpipe for open pipe. Use the formulae as shown for the previous technique. 2. Calculate the swab/surge pressure generated by the drillpipe due to the pipe movement. The swab/surge pressure can be calculated by substituting the annular mud velocity in the formulae for annular pressure loss (Bingham or Power Law). 3. Calculate the velocity of the mud around the collars. Use the following formulae: v(drillcollar) = v(drillpipe) X Adp Adc where Adp = cross-sectional area of drillpipe annulus (in.2) Adc = cross-sectional area of drillcollar annulus (in.2 ) 4. Calculate the swab/surge pressure generated at the collars due to pipemovement. Use the formulae for annular pressure loss (Bingham or Power Law) and v(drillcollar) as calculated in (3). 5. Calculate the total annular swab/surge pressure. This is equal to the sum of the swab/surge pressures at the drillpipe and the collars, or the sum of (2) and (3). 6. Calculate the swab/surge pressure inside the drillstring. Using Preston Moores assumption that the fluid level outside the pipe equals the level inside the pipe, the velocity of the mud inside the pipe equals the velocity outside. 7. Calculate the swab/surge pressure generated inside the drillpipe. Assuming that the mud velocity outside the pipe equals that inside the pipe, use the formulae for internal pressure loss (Bingham or Power Law). 8. Calculate the swab/surge pressure generated inside the drillcollar. Assuming that the mud velocity outside the drillcollar equals that inside the collar, use the formulae for internal pressure loss (Bingham or Power Law). 9. Calculate the swab/surge pressure generated at the bit. Using the formulae: vn = Q An X 10.23 (m/sec) (psi)
where in this case the mud flowrate, Q, is equal to the mud flowrate through the collars.
3-14
March 1995
10. Calculate the total internal swab/surge pressure due to the pipe movement. This is equal to the sum of the swab/surge pressures inside the drillstring, (6) plus (8), plus the bit swab/surge pressure as calculated in (9). 11. Estimate the actual swab/surge pressure due to the pipe movement. It is assumed that the actual swab/surge pressure will be between the values calculated in (5) and (10). The resultant swab/surge pressure is added to the hydrostatic pressure of the mud if the pipe is being run into the hole and subtracted if the pipe is being pulled. Therefore: EMW at the point of interest = MW D sumP X 1.421 (SG)
where sum P = total swab/surge pressure (psi) D = vertical depth to point of interest (m)
3-15/16 3-15
March 1995
4 FRACTURE GRADIENT
Paragraph
1 2 3 4 5 6 7 8 9 General Stresses in the Earth Fracture Orientation Fracture Gradient Prediction Daines Method of Fracture Gradient Prediction An Example Pressure Evaluation Log Leak Off Tests Leak Off Test Procedure Interpretation of Results
Page
4-2 4-2 4-3 4-4 4-4 4-7 4-9 4-10 4-11
Illustrations
4.1 4.2 4.3 4.4 Principal Stress Orientation Poissons Ratio for Different Lithologies An Example Pressure Evaluation Log A Typical Fracture Test
4-3 4-5
4-12
4-1
March 1995
1 General
The absolute upper limit of primary well control is the point at which the wellbore pressure equals the fracture pressure of the exposed formation. At this point a fracture is initiated and the wellbore can no longer be considered to be a closed system. This will lead to loss of mud from the hole and the possibility of the loss of primary control. In order to drill a well safely therefore, it is useful for the Drilling Engineer to be able to predict and measure fracture pressures. At the well planning stage, the fracture gradient can be estimated from offset well data. Ifthis information is not available then Daines Method can be used to predict the fracture gradient. As the well is drilled, Leak Off Tests are carried out to assess the mud holding capability of the openhole. It is Company policy that these tests be carried out to leak off point, which in most cases will represent a pressure that is less than the actual fracture initiation pressure. The leak off pressure is converted to an equivalent mud weight which determines the upper limit of primary control for the next hole section. LO tests are generally carried out once in each openhole section after drilling out of the shoe. However the test should be repeated when weaker zones are drilled into. It is not practical to conduct a leak off test at every change in formation and consequently it is useful to be able to predict the fracture gradient of new formations without conducting further leak off tests. Before covering the techniques that are used to predict fracture gradient, it is appropriate to explain the origins of the stresses that occur naturally below the surface of the earth.
In most cases, the maximum stress will be vertical, due to the pressure of the overlying rock and pore fluid. This is defined as the overburden pressure. In a tectonically relaxed area the maximum stress will, in most cases, be vertical and the stresses in the horizontal plane will be equal. At shallow depths however, the horizontal stress may be greater than the vertical stress, even in a tectonically relaxed area. Figure 4.1 shows the effect of tectonic forces on the principal stresses. A small tectonic force ensures that the two principal stresses in the horizontal plane are no longer equal. This has the effect of creating an actual intermediate stress.
4-2
March 1995
'1 = maximum principal stress '2 = intermediate principal stress '3 = minimum principal stress
1 Tectonically relaxed area '1 is vertical '2 = '3 induced fractures will be vertical
2 under the influence of a small tectonic stress '1 is vertical '2 = 3 an actual intermediate stress is created
3 in an area that is significantly affected by tectonic stress '1 is horizontal induced fractures will be horizontal
WEOX02.128
3 Fracture Orientation
A fracture will be created if wellbore pressures exceed the minimum principal stress at any point in the openhole. The fracture will propogate along the path of minimum resistance, which will be at right angles to the direction of the minimum principal stress. Fractures will therefore be vertical when the minimum principal stress is horizontal, and horizontal if the minimum principal stress is vertical. (See Figure 4.1). Consequently induced fractures will be vertical in areas where tectonic forces are negligible, except possibly at very shallow depths. However horizontal fractures may be formed in areas where tectonic forces are significant. In effect, it is necessary for the applied pressure to lift the weight of the overburden for horizontal fractures to be formed. This is unlikely to occur at depth when overburden pressure will, in most cases, be greater than pressures due to tectonic forces.
4-3
March 1995
t = Pfrac 'l
where t = Pfrac = '1 = = Pf =
Pf l
tectonic stress (psi) fracture pressure (psi) maximum effective principle stress (psi) Poissons ratio for the rock formation pore pressure (psi)
4-4
March 1995
Clay, very wet Clay Conglomerate Dolomite Greywacke: coarse fine medium Limestone: fine, medium medium, calcarenitic porous stylolitic fossiliferous bedded fossils shaley Sandstone: coarse coarse, cemented fine very fine medium poorly sorted, clayey fossiliferous Shale: Calcereous (<50% CaC03) dolomitic siliceous silty (<70% silt) sandy (<70% sand) kerogenaceous Siltstone Slate Tuff: Glass From Weurker H.G: 0.21
Annotated Tables of Strength and Elastic Properties of Rocks, Drilling, reprint Series SPE Dallas (1963).
4-5
March 1995
The overburden pressure is determined from density logs, or from bulk densities determined from the cuttings. 2. Calculate the tectonic stress coefficient. The tectonic stress coefficient can be calculated as follows: = t / '1 where = tectonic stress coefficient This value is used to predict the magnitude of the tectonic stress throughout the nexthole section until the next LO test can be used to recalculate the figure. It is however generally the case that '1 remains directly proportional to t throughout the well, if the rock strata are horizontal and the basin structure does not change significantly with depth. Having calculated the above figures at the first LO test, the fracture pressure can be calculated as drilling proceeds in the following manner: 3. Calculate the maximum principal stress at the point of interest. The magnitude of the maximum principal stress is calculated from the pore pressure and the overburden pressure as follows: '1 = S Pf where S = overburden pressure (psi) Pf = pore pressure (psi) The overburden pressure can be calculated from density logs, or from the bulk density values determined from the cuttings. 4. Calculate the tectonic stress at the point of interest. The magnitude of the tectonic stress is calculated from the maximum principal stress and the tectonic stress coefficient as follows: t = '1
X
5. Calculate the fracture pressure at the point of interest. Using Figure 4.2 to determine a value for the Poissons ratio for the rock, the fracture pressure can be calculated from the following formula:
Pfrac = t + 'l
+ Pf l
(psi)
where P frac = fracture pressure at the point of interest (psi) This procedure can be repeated as the well is drilled in order to map the trend in fracture gradient with depth.
4-6
March 1995
Pfrac = t + 'l
+ Pf l
(SG)
An interesting case would be to estimate the fracture gradient of a sand at these conditions and at this depth. Using the same formula, but substituting a Poissons ratio of 0.01 for a typical shallow marine sand, the fracture gradient is calculated as follows:
The possible variation in fracture gradients at these depths is therefore quite significant. After 1450m, the clays have sufficiently dewatered due to compaction to support a horizontal stress. As a result, the fracture gradient is reduced to a value that is less than the overburden gradient. This means that vertical fractures may be formed at pressures lower than the overburden pressure. The tectonic stress coefficient is calculated from the result of the LO test carried out at the 18 5/8 in. casing shoe. This is the first point at which the clays are assumed to be adequately compacted so as to predict a reasonable figure for the tectonic stress coefficient as follows:
t = Pfrac
'l
Pf l
(SG)
4-7
March 1995
therefore the tectonic stress coefficient is given by: = t '1 = 0.165 = 0.39 (1.95 1.53)
and this value of is used to calculate the tectonic stress in subsequent rock strata. From the 18 5/8 in. shoe to 2880m, the fracture gradient increases in line with the overburden gradient. At 2880m, the pore pressure gradient begins to decrease, causing a reduction in the calculated fracture gradient to 1.88 SG at 3100m. At 3120m, the formation changes to a sandstone interbedded with siltstone. A Poissons ratio of 0.06 is chosen for these loose fine grained sands which results in a reduction in the calculated value of the fracture gradient to approximately 1.72 SG. At 3220m, the formation changes to limestone, for which a Poissons ratio of 0.28 is used. Therefore at 3400m, the fracture gradient is calculated as follows using Daines formula:
Pfrac = t +
'l
+ Pf l
(SG)
The LO test at the 13 3/8 in. shoe shows a fracture gradient of 2.13 SG, which is slightly higher than the predicted figure. The fracture pressure then increases with depth and pore pressure throughout the 12 1/4 in. section to a calculated maximum of 2.23 SG at the 9 5/8 in. casing point. The LO test at this point confirms a 2.21 SG fracture gradient. Mud was lost to the sandstone stringers at the base of the limestone (4200m) at an ECD of 2.06 SG. This figure is therefore taken to be the minimum fracture gradient in the 8 1/2 in. hole. However, the actual fracture gradient of the mudstone increases with depth and in line with the pore pressure, to 2.205 SG at 4429m. A Poissons ratio of 0.06 is used to calculate the fracture gradient in the sandstone section after 4429m. The fracture pressure in the sand remains constant at 2.16 SG until the formation becomes interbedded with mudstone, at which point, the calculated fracture pressure increases to 2.22 SG. The underlying mudstone has a calculated fracture gradient of 2.22 SG.
4-8
March 1995
Company policy is that: Leak off tests or competency tests will be performed prior to drilling each new hole section (except for conductors). The following guidelines are offered: Leak off tests should be performed after drilling 3 to 5m of new hole below any casingshoe. Leak off tests should be taken to leak off unless: The pressure exceeds that to which the casing was tested. (On a development well) where the pressure may be limited to that required to drill safely the next section of hole (competency test).
When drilling through sands, or permeable rock, at any point below the casing shoe, consideration should be given to carrying out a further LO test to ascertain the new rock strength, and thus, the ability of the hole to contain a kick. Leak off tests should not be conducted in brittle formations (eg fractured limestone). Company policy is therefore to restrict applied pressures to a maximum represented by the LO point. The reason for this is that, in many cases, it is not certain that an induced fracture will heal completely to withstand the pressure that originally caused it to fracture. Field evidence, however, suggests that in most cases induced fractures will heal completely. However it is difficult to predict the circumstances in which fractures will not heal completely and hence permanently weaken the formation. It has been suggested that the drilling process locks additional stresses into the rock around the wellbore, thereby increasing the pressure required to cause a fracture. If a fracture is created, these additional stresses are released and consequently the pressure required to reopen the fracture may be less than that originally required. It is accepted, however, that particularly brittle rocks, such as limestone, will show very little inelastic behaviour prior to fracture. As a result, there may be no clear leak off before a fracture occurs. A brittle formation may be permanently weakened by an induced fracture and consequently it is not recommended to conduct LO tests in such formations.
4-9
March 1995
The bulk modulus of elasticity of a drilling fluid is determined by the characteristics ofthe base fluid as well as the solids content of the fluid. The following figures can beused: K, water = 290,000 335,000 psi K, BP H3HF Base Oil = 160,000 260,000 psi The bulk modulus of actual drilling fluids will be greater than these figures by an amount related to solids content. Plot a graph of pressure versus mud pumped to establish linearity prior to the LO test. 5. Drill out of the shoe and 3 to 5 m of new hole. 6. Circulate and condition the mud. 7. Pull up into the casing. 8. Line up the pump to the annulus and displace all lines to the well to mud. 9. Close the BOP.
4-10
March 1995
10. Run the pump at a constant 0.3 to 0.5 bbl/min. Monitor the pressure build up, and accurately record the volume of mud pumped. Plot pressure versus volume of mud pumped. 11. Stop the pump when any deviation from linearity is noticed between pump pressure and volume pumped. Record the final pump pressure and calculate LO EMW. 12. Bleed back mud from the well and compare with the volume pumped.
9 Interpretation of Results
Figure 4.4 shows the result of a typical fracture test carried out in a consolidated low permeability formation in a tectonically relaxed area. NOTE: It is Company policy that the test is stopped at leak off point. From points 1 to 2, the exposed rock is deforming elastically as the relationship between pressure and volume pumped is linear. At point 2, the pressure in the wellbore at the exposed formation is equal to the sum of the pore pressure and the minimum horizontal effective stress. In other words, any cracks that exist at the wellbore and in the vertical plane will be in a state of equilibrium, the applied pressure exactly counteracting the naturally occurring compressive forces. At point 3, which represents the leak off point (because it is the first noted deviation from the linear relationship), the pump would normally be stopped and the pressure bled down in line with Company policy. If the pump was left running, the pressure would eventually build to fracture pressure as shown. From points 2 to 4, the formation is deforming plastically, in that for the same increment of applied stress (pressure), a greater level of strain (volume) is produced. The difference between the pressure at point 2 and the pressure at point 4 represents the pressure required to initiate the fracture. If the pump was stopped at point 4, as is shown on the diagram, the fracture would not propogate further into the formation and the pressure will drop to point 5. The pressure at point 5 should be equal to the pressure at point 2. If the pressure is then bled down, the returned volume should be equal to the volume pumped into the hole; if it is significantly less, then the fracture may be still be open. If the pump was kept running after point 4, the fracture would propogate into the formation at a pressure slightly lower than point 4, or the fracture propogation pressure.
4-11
March 1995
4 3 2 LEAK OFF PRESSURE (the pump would normally be stopped at this point)
1 1 1 2 bbl PUMPED 3 2 3 4
TIME, MINUTES
WEOX02.130
4-12
March 1995
Page
5-3 5-3 5-8 5-19 5-24 5-27
Illustrations
5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 Choke and Standpipe Pressure during the first circulation of the Drillers Method Pit Gain during the first circulation of the Drillers Method
5-5 5-5
Shoe Pressure during the first circulation of the Drillers Method 5-6 Choke and Standpipe Pressure during the second circulation of the Drillers Method Choke and Standpipe Pressure during the Wait and Weight Method Shoe Pressure during the Wait and Weight Method Choke Pressure during the Drillers Method for various influx volumes Choke Pressure during the Wait and Weight Method for various influx volumes Shoe Pressure during the Drillers Method for various influx volumes
5.10 Shoe Pressure during the Wait and Weight Method for various influx volumes 5.11 Choke Pressure during the Wait and Weight Method and the Drillers Method for two different influx volumes 5.12 Choke Pressure during displacement of a gas kick using the Drillers Method for various kick intensities 5.13 Choke Pressure during displacement of a gas kick using the Wait and Weight Method for various kick intensities
5-1
March 1995
5.14 Choke Pressure during displacement of a gas kick using the Wait and Weight Method for various kick intensities 5.15 A Comparison of the Shoe Pressure during displacement using the Drillers and Wait and Weight Method for two gas kicks of different intensities 5.16 A Comparison of Shoe Pressures during displacement of a 20 barrel gas kick for various shoe depths 5.17 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 3000m 5.18 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 2500m 5.19 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 2000m 5.20 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 1500m 5.21 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 1000m 5.22 Choke Pressure during displacement of a water kick using the Wait and Weight Method 5.23 Comparison of Choke Pressures during displacement of a gas kick on a fixed rig and a floating rig 5.24 Choke Pressure for various Water Depths during displacement of a gas kick 5.25 Determination of the Required Rate of Choke Manipulation for a Deep Water Subsea Well 5.26 Estimated Choke Line Losses (psi) for Various Choke Line Lengths (3in. ID) 5.27 Annulus Pressure Loss for various Well Configurations 5.28 Choke Pressure during displacement of a gas kick with overbalanced mud 5.29 Shoe Pressure during displacement of a gas kick with overbalanced mud 5.30 Annulus Pressure Worksheet 5.31 Graph of Pseudo-critical Temperature and Pressure for Hydrocarbons 5.32 Compressibility Factors for Natural Gas
5-13
5-14 5-15 5-15 5-16 5-16 5-17 5-17 5-19 5-20 5-21 5-22 5-24 5-25 5-26 5-26 5-31 5-33 5-34
5-2
March 1995
1 General
When a kick is taken with the pipe on bottom, the well can be killed using either the Wait and Weight Method or the Drillers Method. The Wait and Weight is the preferred method. The procedures used to implement these techniques on either a floating or a fixed rig are detailed in Volume 1. Both these methods ensure that the bottomhole pressure is maintained constant and equal to, or slightly greater than, the kick zone pressure. In order to fully understand the implementation of these methods, it is important to understand the surface and downhole pressures that are caused by displacing a kick from the hole using either the Drillers Method or the Wait and Weight Method. This chapter is intended to cover the variations in surface and subsurface pressures during these methods, and to explain the most important factors that affect the magnitude of these pressures. All the pressure plots shown in this chapter are developed by computer programme. The pressures are determined by simulating the displacement of a gas kick from a well with the model of a discrete bubble of gas. The actual pressures seen when a kick is taken may be different from those predicted by the programme; however the plots can demonstrate the influence of the major factors that affect the wellbore pressures during circulation. The pressure plots contained in this chapter are generated on the basis that the bottomhole pressure is constant and exactly equal to the kick zone pressure.
5-3
March 1995
Figure 5.1 shows the choke pressure during the displacement of a kick with the Drillers Method for a surface BOP. Point A represents the shut-in casing pressure. From point A to point B, the casing pressure drops as the influx is displaced past the BHA. This drop is caused by a reduction in height of the influx as the influx is displaced from the BHA annulus to the drillpipe annulus. The choke operator will open the choke to maintain the appropriate standpipe pressure. From point B to point D, the influx is expanding as it is circulated up the hole and hence the choke pressure required to balance the kick zone pressure is increasing. The choke operator will therefore close in on the choke to maintain the correct standpipe pressure. At point C, the gas has expanded to occupy its original height in the annulus when opposite the BHA. At point D, the gas arrives at the choke; the choke operator will have to close in on the choke to ensure that the choke pressure does not drop significantly as the low density gas passes across the choke. From point D to point E, the gas is passing the choke; the choke operator will have to open the choke to reduce the choke pressure to maintain the correct standpipe pressure. The choke pressure required to balance the kick zone pressure reduces as the gas passes the choke because the column of gas in the annulus is continually decreasing in height. At point E, the gas has been displaced from the well and the choke pressure will stabilise at a value determined by the degree of underbalance. Figure 5.2 shows the pit gain, or the volume of the kick, as it is displaced to the choke. Figure 5.3 shows the pressure at the casing shoe as the kick is displaced from the hole. From point P to point Q, the pressure drops as the influx is displaced past the BHA. From point Q to point R, the pressure increases as the influx expands as it is circulated up to the casing shoe. At point R, the top of the influx has arrived at the casing shoe and from point R to point S the influx is circulated past the casing shoe. Once the influx has been circulated past the shoe, the pressure at the shoe will remain constant as the influx is circulated to the choke, as long as the choke is correctly manipulated. It can be seen from Figure 5.3 that, in this case, the shoe pressure was at maximum when the well was shut-in. In other words, the influx did not expand to its original height in the annulus before it arrived at the choke. However, if the shoe was shallower, the maximum shoe pressure might have been when the influx was circulated to the shoe. Figure 5.4 shows the standpipe and choke pressure during the second circulation during which the well is circulated to kill weight mud. Having established the initial circulating pressure, the standpipe pressure must be reduced as the drillpipe is displaced to kill weight mud. In practice, very little choke manipulation will be required at this stage because the standpipe pressure will drop automatically as the kill weight mud is pumped down the drillpipe. Once the kill weight mud starts up the annulus, the choke size will have to be increased so that the correct final circulating pressure is maintained. Once the hole has been displaced to kill weight mud, the choke pressure required to maintain the final circulating pressure will be zero. In practice therefore, the choke will be wide open at this point and it may not be possible to keep the standpipe pressure down to the final circulating pressure.
5-4
March 1995
Figure 5.1 Choke and Standpipe Pressure during the first circulation of the Drillers Method
Figure 5.2 Pit Gain during the first circulation of the Drillers Method
5-5
March 1995
6000 5800 5600 5400 5200 5000 4800 0 200 400 600 800 WEOX02.133
P Q
Figure 5.3 Shoe Pressure during the first circulation of the Drillers Method
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
2000 1800 1600
STANDPIPE PRESSURE
SCR2
CHOKE PRESSURE
DRILLPIPE VOLUME
200
400
ANNULUS VOLUME
600
800 WEOX02.134
Figure 5.4 Choke and Standpipe Pressure during the second circulation of the Drillers Method
5-6
March 1995
1400
C
1200 1000 SCR1 800 600
SCR2
DRILLPIPE VOLUME
200
400
Figure 5.5 Choke and Standpipe Pressure during the Wait and Weight Method
Figure 5.6 illustrates the pressure at the casing shoe for both the Wait and Weight Method and in comparison with the Drillers Method. Between point P and point Q, the shoe pressure decreases as the influx is displaced past the BHA. The influx expands as it is circulated to the shoe at point R, after which, the pressure at the shoe decreases. At point S, the kill weight mud starts up the annulus and hence reduces the choke pressure below that for the Drillers Method. Between point T and point U, the original weight mud is displaced past the shoe until point U, when the kill weight mud arrives at the shoe. The pressure at point U is equal to the kick zone equivalent mud weight, and thus represents the minimum pressure that the shoe will see once the well has been killed.
5-7
March 1995
In this case therefore, the maximum shoe pressure is unaffected by the technique used to kill the well, however, the shoe will be under pressure significantly longer if the Drillers Method is used.
6000 5800 5600 5400 5200 5000 4800 0 DRILLPIPE VOLUME 200 400 600 800 WEOX02.136
P Q
Figure 5.6 Shoe Pressure during the Wait and Weight Method
5-8
March 1995
200
400
600
800 WEOX02.137
Figure 5.7 Choke Pressure during the Drillers Method for various influx volumes
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
2000 1800 1600
200
400
600
800 WEOX02.138
Figure 5.8 Choke Pressure during the Wait and Weight Method for various influx volumes
5-9
March 1995
Figure 5.9 shows the shoe pressures as various influx volumes are displaced from the well using the Drillers Method. Figure 5.10 shows the shoe pressures as the same influx volumes are displaced using the Wait and Weight Method. Figure 5.11 shows a comparison of choke pressure during the Wait and Weight Method against the Drillers Method for influx volumes of 20 bbl and 50 bbl. Influx volume is therefore a variable that has significant influence on wellbore pressure during the displacement of a kick. However, it is the only variable that the rig crew have some control over for a given kick situation; it is therefore particularly important that shut-in procedures are implemented as quickly as possible, even if there is some doubt as to whether the well is flowing.
5-10
March 1995
20bbl
200
400
600
800 WEOX02.139
Figure 5.9 Shoe Pressure during the Drillers Method for various influx volumes
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
6600 6400 6200
20bbl
200
400
600
800 WEOX02.140
Figure 5.10 Shoe Pressure during the Wait and Weight Method for various influx volumes
5-11
March 1995
50bbl
200
400
600
800 WEOX02.141
Figure 5.11 Choke Pressure during the Wait and Weight Method and the Drillers Method for two different influx volumes
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
2000 1800 1600
1400 1200
1.91SG
1000
1.87SG
800
1.83SG
600
1.79SG
400
1.75SG
200 0 0 200 400 600 800 WEOX02.142
Figure 5.12 Choke Pressure during displacement of a gas kick using the Drillers Method for various kick intensities
5-12
March 1995
200
400
600
800 WEOX02.143
Figure 5.13 Choke Pressure during displacement of a gas kick using the Wait and Weight Method for various kick intensities
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
2000 1800 1600
1.91SG
1.75SG
DRILLER'S METHOD
Figure 5.14 Choke Pressure during displacement of a gas kick using the Wait and Weight Method for various kick intensities
5-13
March 1995
WAIT AND WEIGHT METHOD 1.75SG DRILLER'S METHOD WAIT AND WEIGHT METHOD
600 800 WEOX02.145
Figure 5.15 A Comparison of the Shoe Pressure during displacement using the Drillers and Wait and Weight Method for two gas kicks of different intensities
Once the kill weight mud starts up the annulus, shoe pressures will be lower than if the Drillers Method is used. As discussed in (b), the higher the kick intensity, the more marked the difference between wellbore pressures during the Wait and Weight Method and the Drillers Method. However, for a given kick intensity the significance of the difference between the two techniques is also influenced by the depth of the shoe. The shallower the shoe, the more significant is the effect of the kill weight mud on pressure reduction at the shoe. The hole configuration therefore can influence the pressures seen at the shoe during displacement. Figure 5.16 shows a comparison of the shoe pressures for the same kick, for various lengths of openhole. A 20 barrel kick is taken at 3500m and is then displaced from the hole; Figure 5.16 shows the pressure variations at 1000m, 2000m and 3000m. As can be seen from Figure 5.16, if the shoe had been at 3000m, the maximum pressure at the shoe is clearly at initial shut-in. However, if the shoe was at 1000m, the shoe pressure is actually greater than at initial shut-in when the influx is displaced to the shoe. This situation is brought about when the influx expands to occupy a greater height in the annulus than it did at initial shut-in before it is displaced to the shoe. This generally requires a considerable length of openhole. Figures 5.17 to 5.21 compare the shoe pressures during displacement of a gas kick for a range of shoe depths, using both the Drillers and the Wait and Weight Methods.
5-14
March 1995
3500m 1000m, 2000m, 3000m MW1: 1.7SG KICK ZONE EMW: 1.83SG
+100 INITIAL (SHUT IN PRESSURE) 0
-100
-200
SHOE AT 3000m
SHOE AT 2000m
SHOE AT 1000m
-300
Figure 5.16 A Comparison of Shoe Pressures during displacement of a 20 barrel gas kick for various shoe depths
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
8300
8200
Q P
8100
8000
7900
DRILLER'S METHOD
Figure 5.17 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 3000m
5-15
March 1995
7000
6900
6800
DRILLER'S METHOD
6700
6600
6500
Figure 5.18 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 2500m
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
5900 5800
5700 5600
5500
DRILLER'S METHOD R
5400 5300
Figure 5.19 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 2000m
5-16
March 1995
4400 4300 4200 4100 4000 3900 3800 0 200 400 600 800 WEOX02.150
DRILLER'S METHOD
Figure 5.20 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 1500m
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW:
3500 3400 3300
3200 3100 3000 2900 2800 2700 2600 2500 0 200 400 600 800 WEOX02.151
DRILLER'S METHOD
Figure 5.21 A Comparison of Shoe Pressures during displacement of a gas kick shoe at 1000m
5-17
March 1995
Figure 5.17 shows the shoe pressure for a 20 barrel kick taken at 3500m for the shoe at 3000m. From initial shut-in to point P, the pressure decreases as the influx is displaced past the BHA; from point P to point Q the pressure increases as the influx expands as it is displaced up towards the shoe. At point Q, the top of the influx arrives at the shoe and from point Q to point R the pressure at the shoe drops as the influx is displaced past it. From point R to point S, the pressure at the shoe remains constant as the original mud occupies the annulus from the bottom of the hole to the shoe. However, the pressure at the shoe is further reduced at point S when, in the case of the Wait and Weight Method, the kill weight mud starts up the annulus. Figure 5.18 shows the shoe pressure for the shoe at 2500m. A similar pressure profile is shown to that in Figure 5.17; however in this case the influx expands more before it arrives at the shoe due to the greater length of openhole. Figure 5.19 shows the shoe pressure profile for the shoe at 2000m. In this case, the kill weight mud starts up the annulus at point R, when the tail of the influx is passing the shoe. Figure 5.20 shows the shoe pressure profile for the shoe at 1500m. In the case of the Drillers Method, the shoe pressure almost reaches its original shut-in value. In the case of the Wait and Weight Method however, the kill weight mud starts up the annulus at point P, before the influx arrives at the shoe. The shoe pressure is reduced by the kill weight mud from this point on. Figure 5.21 shows the shoe pressure profile for the shoe at 1000m. In the case of the Drillers Method, the shoe pressure now increases past the shut-in value as the influx is circulated to the shoe. However, in the case of the Wait and Weight Method, the kill weight mud starts up the annulus at point P, and this has the effect of reducing the maximum pressure that the shoe experiences. Figures 5.17 to 5.21 show that the Wait and Weight Method has only a small influence on the maximum shoe pressure for wells of this type, even when the shoe is relatively shallow. The most important point however is that the time that the shoe is subject to high pressure is substantially reduced when the Wait and Weight Method is used. The reduction in shoe pressure due to the kill weight mud is most significant when there is a long section of openhole (as is seen in Figures 5.17 to 5.21).
5-18
March 1995
Figure 5.22 shows a typical choke pressure profile for a salt water kick displaced from the hole by the Wait and Weight Method. From point P to point Q, the choke pressure remains relatively constant as the drillpipe is displaced to kill weight mud. From point Q to point R, the choke pressure drops as the kill weight mud starts up the annulus. This is in marked contrast to the gas kick where the expansion of the kick at this stage tends to increase the choke pressure. From point R to point S, the influx passes the choke with a corresponding drop in choke pressures. From point S to point T, the choke pressure drops as the original mud behind the influx passes the choke. At point T, the kill weight mud arrives at surface.
500
400
300
200
R
100
S
0 0 200 400 600
T
800 WEOX02.152
Figure 5.22 Choke Pressure during displacement of a water kick using the Wait and Weight Method
An oil kick is likely to behave in a similar manner to the gas kick when displaced from the well. The term oil covers a large variety of fluids, ranging from viscous black oil that contains very little gas, to very light oils that have very high gas oil ratios. Most oil will contain gas at reservoir conditions which will come out of solution and expand as it is displaced up the hole. Essentially, therefore, the majority of oil kicks will behave similarly to a gas kick.
4 Subsea Considerations
If a kick is taken from a floating rig, the influx will be displaced to the surface through a small diamater choke line that is attached to the drilling riser. The fundamental difference between well control procedures on a fixed and a floating rig originate from the necessity of having to circulate through this choke line.
5-19
March 1995
The potential problems caused by circulating through the choke line can be summarised asfollows:
(a) Choke pressures will be significantly higher than for an equivalent well drilled from a fixed rig.
This is due to the fact that the height of the influx is considerably increased as it is displaced from the annulus to the choke line. Figure 5.23 shows a comparison between the choke pressure during displacement of a gas kick from a well drilled in 1000m of water and a similar well drilled from a fixed rig. The influence of the choke line is apparent in that the maximum choke pressure is increased from 1200 psi to approximately 2600 psi.
FIXED RIG: WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW, MW2: FLOATING RIG: WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW, MW2:
2800
6 1/4in/180m 5in DP WAIT AND WEIGHT 20bbl GAS 1000m/3in ID 6 1/4in/180m 5in DP WAIT AND WEIGHT 20bbl GAS
2400
FLOATING RIG
2000
1600
1200
FIXED RIG
800
400
Figure 5.23 Comparison of Choke Pressures during displacement of a gas kick on a fixed rig and a floating rig
5-20
March 1995
Figure 5.24 shows the choke pressures during displacement of the same 20 bbl influx for a variety of water depths. It can be seen that the choke pressure is not, in this case, significantly affected by a water depth of 100m.
FIXED RIG: WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW, MW2: FLOATING RIG: WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW, MW2:
2800
6 1/4in/180m 5in DP WAIT AND WEIGHT 20bbl GAS 1000m, 500m, 100m/3in ID 6 1/4in/180m 5in DP WAIT AND WEIGHT 20bbl GAS
2400
1000m WATER
2000
500m WATER
1600
1200
100m WATER
800
FIXED RIG
400
Figure 5.24 Choke Pressure for various Water Depths during displacement of a gas kick
(b) The rate of increase of choke pressure required as the gas enters the choke line may be unrealistically high at normal displacement rates.
As can be seen from Figure 5.25 the increase in choke pressure required as the influx is displaced up the choke line is equivalent to 64 psi/bbl. This can be converted to a rate of choke manipulation for various displacement rates as follows: At 4 bbl/min At 3 bbl/min At 1 bbl/min At 0.1 bbl/min = = = = 64 64 64 64 x x x x 4 3 1 0.1 = = = = 256 psi/min 192 psi/min 64 psi/min 6.4 psi/min
5-21
March 1995
It can therefore be seen that normal displacement rates have the potential to require an unrealistic rate of manipulation of the choke. In this case, the most satisfactory rate of displacement would be of the order of 1 bbl/min.
As gas invades the choke line: p = bbl/pumped As the mud behind the gas enters the choke line p = bbl/pumped 2000 28 = 71psi/bbl 1800 28 = 64psi/bbl
2400
2000
800
400
Figure 5.25 Determination of the Required Rate of Choke Manipulation for a Deep Water Subsea Well
It should be noted however that these calculations are based on the assumption that the gas influx enters the choke line as a discrete bubble without mixing with the mud ahead of it. This may not always be the case, however the figures quoted above certainly indicate that the normal kick displacement rates have the potential to cause such complications. A further problem exists in that when the gas enters the choke line, the drillpipe pressure will only register the drop in bottomhole pressure after the lag time, which can be substantial in deep wells.
5-22
March 1995
On a fixed rig, the lag time will not be so problematical because the required rate of choke manipulation is generally lower. In other words, the bottomhole pressure will drop only very slightly before the drillpipe pressures registers that drop and the choke operator closes in the choke (to increase the choke pressure and hence the bottomhole pressure). The lag time between the choke and the drillpipe pressure gauges is generally considered to be of the order of 2 seconds per 300m of drillstring length. This lag time, however, will be significantly affected by the type and size of the influx in the hole. It can be seen therefore that there may be a lag time of approximately 20 seconds in deep wells. If the required rate of choke manipulation is 420 psi/min as the influx is displaced up the choke line, the bottomhole pressure may have dropped 130 psi before the drillpipe pressure gauge registers this drop. Clearly this is an additional reason for displacing the influx through the choke line at a rate that is substantially slower than normal slow circulation rates.
(c) The rate of decrease of choke pressure required as the mud behind the gas reaches the base of the choke line may be unrealistically high.
In a similar manner, the required rate of choke manipulation as the mud behind the influx enters the choke line may be unrealistically high at normal slow circulating rates. In this case, the potential problem is that the well may be overpressured, leading to the possibility of fracturing the exposed formation. Figure 5.25 shows that the choke pressure would theoretically have to be reduced at 71psi/bbl which corresponds to the following rates for various displacement rates: At 4 bbl/min At 3 bbl/min At 1 bbl/min At 0.1 bbl/min = = = = 71 71 71 71 x x x x 4 3 1 0.1 = = = = 284 psi/min 213 psi/min 71 psi/min 7.1 psi/min
This is clear indication that normal displacement rates are unsuitable when displacing a gas influx through a long choke line.
(d) The frictional pressure as a result of circulating through the choke line may be significant at slow circulating rates.
Choke line frictional pressure may be significant, when added to the wellbore pressures resulting from the displacement of a kick. In certain circumstances, it may be of a magnitude such as to cause formation breakdown. There are special techniques that can be used to eliminate the effect of choke line losses during displacement of a kick. One such technique, namely the use of the kill line monitor, is described in Chapter 6 of Volume 1. Choke line losses are generally insignificant in relatively shallow waters, but can be significant in waters of 500m or greater. Figure 5.26 shows a table of estimated choke line losses for various choke line lengths. When very slow displacement rates are used, (such as 1 bbl/min) choke line losses are generally insignificant, even in deep water.
5-23
March 1995
3 bbl/min
1.5
1.7
1.9
2.1
CHOKE LINE LENGTH (m) 1000 m 500 m 100 m 4 bbl/min MUD WEIGHT (SG) 120 90 17 1.5 200 100 19 1.7 220 110 22 1.9 245 125 25 2.1
CHOKE LINE LENGTH (m) 1000 m 500 m 100 m 260 130 26 295 145 30 325 165 33 360 180 36
Figure 5.26 Estimated Choke Line Losses (psi) for Various Choke Line Lengths (3 in. ID)
5 Safety Factors
During well control operations, it is clearly necessary to maintain the bottomhole pressure slightly greater than the kick zone pressure. This will provide a margin of error for the choke operation that will prevent a second influx occurring. However, excessive additional pressure may needlessly overpressure the wellbore and possibly cause the formation to fracture. In general, every effort should be made to ensure that no additional pressures are applied to the openhole at early stages in the displacement of the kick when downhole pressures will generally be at a maximum. The following are possible causes of additional pressures during the displacement of a kick:
5-24
March 1995
Annulus pressure loss at 3 to 4bbl/min for mud weight range 1.3 SG to 1.7 SG = 20 to 25psi
7in @ 3200m 9 5/8in shoe @ 3500m 8 1/2in hole 270m of 6 1/2in collars TD 4000m 3 1/2in drillpipe Annulus pressure loss at 3 to 4bbl/min for mud weight range 1.7 SG to 2.1 SG = 100 to 125psi 7in @ 4200m 200m 4 3/4in collars 6in hole TD 4500m Annulus pressure loss at 3 to 4bbl/min for 1.5 SG mud = 150 to 180psi 2.1 SG mud = 190 to 240psi 9 5/8in @ 3200m
WEOX02.157
5-25
March 1995
6 1/4in/180m 5in DP OVERBALANCED MUD INFLUX: 20bbl GAS BHA: PIPE: TECH:
1400
1.9SG
CHOKE PRESSURE (psi)
1200
1000
800
600
400
200
Figure 5.28 Choke Pressure during displacement of a gas kick with overbalanced mud
WELL DEPTH: SHOE DEPTH: MW1: KICK ZONE EMW: MW2:
5000
6 1/4in/180m 5in DP OVERBALANCED MUD INFLUX: 20bbl GAS BHA: PIPE: TECH:
4800
4600
1.85SG MUD
4200
4000
Figure 5.29 Shoe Pressure during displacement of a gas kick with overbalanced mud
5-26
March 1995
Heavier than kill weight mud is often considered in order to either add a small overbalance after the kick has been displaced from well or to kill an underground blowout. From the examples in Figures 5.28 and 5.29 it can be seen that even a relatively small overbalance will increase the wellbore pressures during kick displacement. Overbalance should be added to the mud after the well has been killed.
5-27
March 1995
The formulae that are presented as follows are recommended for use at the rigsite for quick estimations of annulus pressures during the displacement of a kick, if a computer is not available for this purpose:
S 2
(psi)
(psi)
PO X VO X Z D ZO X TO
TD
VO (if temperature and compressibility effects are ignored) = = = = = = = = = = = = = = = depth to the top of the influx (m) pressure at the top of the influx (psi) influx temperature for influx at depth, D ( R) influx compressiblity factor for influx at depth, D original influx pressure (psi) original influx temperature ( R) original influx compressibility factor original inlfux volume (bbl) kill mud weight (SG) original mud weight (SG) annular capacity immediately below the influx (bbl/m) well total depth (m) height of original mud below influx (m) hydrostatic pressure influx (psi) kick zone pressure (psi)
Therefore to determine the choke pressure at gas to surface, use D = O. To determine the pressure at the openhole weak point when the top of the influx is at the openhole weak point, use D = depth of openhole weak point. The hydrostatic pressure of the influx is assumed to remain constant during displacement. However it may be adjusted for substantial changes in annular capacity using the following formula: Pi1 = P i2 X C2 C1 where Pi1 Pi2 C1 C2 = = = = hydrostatic pressure of influx at original conditions (psi) hydrostatic pressure of influx at point of interest (psi) original annular capacity (bbl/m) annular capacity at point of interest (bbl/m)
5-28
March 1995
S 2
(psi)
1.421 + Pi Pf
(psi)
S 2
(psi)
where: S = H1 X 1.421
X
1.421 + Pi Pf (psi)
S 2
(psi)
where: S = (TD Dwhd +V cl ) MW1 X 1.421 + Pi Pf (psi) C and Pchoke Dwhd Vcl C = = = = choke pressure at gas to surface (psi) wellhead depth below rotary table (m) internal volume of choke line (bbl) annular capacity immediately below the influx (bbl/m)
5-29
March 1995
(psi)
= (1253 X 1.421 X 1.7) + (3500 1253 1000 + 28.5 ) 1.83 0.1604 = -2310 psi Substituting into:
2 Pchoke = S + K X MW2 X 1.421 C 4 1 2
1.421 + 59 + 9101
S 2
(psi)
1 2
2 Pchoke = -2310 + 20 4
-2310 2
(psi)
= 3225 psi Therefore the maximum anticipated pressure during displacement is 3225 psi. It would however be anticipated that this figure represents the maximum possible pressure at surface and, as such, the actual maximum pressure would be expected to be lower than this value. (For derivation of these formulae ref: Blowout Prevention, Theory and Application by Peter Mills, 1984, D. Reidel Publishing Company.)
5-30
March 1995
Figure 5.30a
Rig Name: bbl length, ft length, ft length, ft length, ft length, ft bbl/ft bbl/ft bbl/ft bbl/ft bbl/ft Volume Volume Volume Volume Volume bbl bbl bbl bbl bbl bbl
Well No: Kick Zone Depth: Casing Shoe Depth: Kick Zone Pressure: Casing Pressure: Pit Gain Influx Height Influx Hydrostatic = = = = = TVD,ft TVD,ft psi psi bbl ft psi
Date:
Drillstring Internal Volume Annulus, in. Annulus, in. Annulus, in. Annulus, in. Annulus, in. ID X OD ID X OD ID X OD ID X OD ID X OD
Total Annulus Volume Hydro Pressure of Mud Below Influx ( psig ) (9) Influx Hydrostatic (psi) (10)
Original Mud Volume of Mud Pumped (bbl) (1) Below Influx Vol (bbl) (2) Height ( ) (3) Pressure (psi) (4) Vol (bbl) (5)
Kill Mud Below Original Mud Height ( ) (6) Pressure (psi) (7)
Hydrostatic of Annulus Fluids (psi) (19) Req'd Back Pressure (psi) (20) Pressure at the Shoe (psi) (21)
F (12)
R (13)
Factor (14)
5-31
ft ft ft ft ft
WEOX02.196
March 1995
Figure 5.30b
The worksheet provided can be used to estimate annulus pressures during the displacement of a kick. The worksheet can be used as follows for the Wait and Weight method: (if the Drillers method is used (5), (6) and (7) are left out of the calculation) (1) Barrels (bbl) of kill weight mud Estimate the volume of kill mud pumped for the gas to arrive at the point of interest. (2-4) Original mud below gas This volume is equal to the volume of kill weight mud pumped until the drillpipe is displaced. At this point and subsequently this volume will remain constant at the drillpipe internal volume. Convert this volume to height and hydrostatic pressure equivalent, in the annulus. (5-7) Kill mud below original mud This volume is zero until the internal volume of the drillstring has been displaced. Once the kill mud starts into the annulus, its height and hydrostatic pressure should be calculated. (8) Metres of mud below gas The total height of mud below the influx. (9) Pressure of mud below the gas Equal to (4) + (7). (10) Gas hydrostatic pressure In a constant annulus size it is assumed that the gas hydrostatic pressure remains constant as the influx expands. The gas hydrostatic must however be corrected for substantial changes in annular dimensions using the following relationship:
This is equal to the kick zone pressure minus the total hydrostatic pressure of the mud below the influx and half of the gas hydrostatic pressure. (12-13) The gas temperature This is estimated from the surface pressure and the temperature gradient in the well unless more detailed information is available. The temperature in F can be converted to R by adding 460. Use Figure 5.31 (BP Well Control Manual, Volume 2) to determine the pseudo critical temperature and pressure of the gas (assume gravity is 0.7 unless logging unit has detected presence of CO2 or H 2S or unusually heavy hydrocarbon components). The pseudo reduced values are then calculated as follows:
T( R) T pseudo critical
V2 = T2 X Z2 X P1 X V1 P2 X T1
V T P Z = = = = Influx volume (bbl) Influx temperature ( R) Influx pressure (psia) Compressibility factor
Influx height =
(19)
Total hydrostatic pressure of annulus fluids This equal to (9) + (10) + (18).
(20)
Required back pressure This is the difference between the kick zone pressure and the total hydrostatic pressure of the fluid in the annulus (19).
(21)
Pressure at the shoe The pressure at the shoe is determined by either: Subtracting the hydrostatic pressure of the annulus fluids from the bottomhole to the shoe from the bottomholepressure Adding the hydrostatic pressure of the fluids from the shoe to the surface to the required back pressure (20)
This procedure will be repeated until the influx is positioned at the appropriate point in the well. For example if the first calculation shows that the top of the influx is above the shoe (assuming that the point of interest is when the top of the influx arrives at the shoe), the calculation should be reworked for a smaller volume of mud pumped. For the first approximations it is a good idea to neglect the effect of temperature and compressibility in order to speed the calculation.
5-32
March 1995
700
Miscellaneou
Pseudo Critical Pressure psia
s gases
650
Cond
ensat
e well
fluids
600
550
500
es
450
M
e isc
lla
ne
g us
as
Con
den
ew sat
ell
flui
ds
400
350
WEOX02.162
5-33
March 1995
1.05
1.0
2.6 2.2
2.8 2.4
2.0
1.0
0.9
1.9 1.8
1.7
1.5
0.95
1.4
1.6
0.8
1.45 1.4
COMPRESSIBILITY FACTOR Z
1.
1.
05
1.5
1.7
0.7
1.6
1.35
1.3
1.
1.
1.
3
1.5
0.6
1.25
1.2
0.5
.15
1.8 2.0
1.4
0.4
1.
1
2.6 3.0
2.4
1.3
0.3 0.25
1.2
3.0
1.1
2.8
1. 05
1.1
2.6 2.2
2.4
1.9
1.0
2.0 1.8
1.1 1.05
1.2
1.7 1.6
1.4 1.3
1.0
January 1, 1941
0.9 7 8 9 10 11 12 13 14 15
0.9
WEOX02.163
5-34
March 1995
COMPRESSIBILITY FACTOR Z
Blowout preventers and associated equipment provide the means of controlling a well after primary control has been lost. The basic requirements for effective BOP equipment include: A properly designed and cemented casing string that can contain pressures encountered whilst drilling. A properly designed and installed wellhead assembly that can support, and seal between, the casing string and the BOP stack. BOPs securely anchored to the wellhead and capable of closing off the annulus against openhole or any tool that is run into the hole. A control system to operate the BOPs which features adequate redundancy and acceptable closing times. A choke system that can maintain a variable back pressure on a well whilst it is circulated. A kill system which gives flexibility to pump to the hole via the annulus or drillstring. Instrumentation that allows control of the well killing operation.
March 1995
6.1
1 2
WELLHEADS
Page
6-2 6-2
Surface Wellheads Subsea Wellheads
Paragraph
6-1
March 1995
1 Surface Wellheads
A conventional wellhead for use on land, platform and jack-up rigs, comprises a series of spools and is based on a starting head. The starting head is anchored to the surface string of casing, with the weight of the casing transferred to the cement, or mud line hanger. Additional support may be provided by the conductor. The major components of the surface wellhead are as follows:
(a) Spools
Conventional wellhead spools generally incorporate the following features: API standard flanges, or hub to suit a clamp, top and bottom. The flange face has a machined groove, often inlaid with stainless steel, to suit an API gasket. A set of seals, sometimes energised with plastic packing, in the base of the spool, to pack off around the preceding casing stub. A bowl to accommodate the next casing string slips. Studded or flanged side-arms below the bowl,which provide communication to the casing annulus. Often the side-arm is threaded to accept a plug to facilitate valve removal. Lock down screws are provided in the top flange of most spools, to retain, and sometimes to energise the pack-off, and also to retain the bore protector. Ports are provided to allow the pressure testing of the flange seals (i.e. the void between the slip and seal assembly, the upper spool seals, and the ring gasket).
2 Subsea Wellheads
A subsea wellhead as used on floating rigs, consists of one or two wellhead housings, casing hangers/pack-offs and a guide base. It is positioned just above the seabed. The wellhead housings are normally made up onto the conductor and 13 3/8 in. casing in the case of a 2 stack system, or onto the conductor and surface casing in the case of a single stack system. They perform four functions: Support of casing strings by means of an internal upset on which the first casing hanger lands. Subsequent casing hangers land off on the previous seal assembly.
6-2
March 1995
Pressure isolation of the casing annulus from the wellbore by providing a polished bore on which the seal assembly packs off. Pressure containment between the wellhead housing and the BOP, by provision of a polished stainless steel inlaid profile for a gasket in the hub bore. Support of the stack which lands on the hub and latches onto a profile on the outside of the hub.
Commonly 21 1/4 in. housings are rated to 2000 or 5000 psi, 16 3/4 in. housings to 5000 or 10,000 psi, 18 3/4 in. and 13 5/8 in. housings to 10,000 psi or 15,000 psi. The following are the major items of equipment associated with the subsea wellhead:
6-3/4 6-3
March 1995
6.2
1 2 3 4 5 6 7
Paragraph
Illustrations
6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 Annular Preventer Sealing Elements Hydril Annular Preventer Type MSP Hydril Annular Preventer Type GK Hydril Annular Preventer Type GL Shaffer Annular Preventer Cameron Annular Preventer Type D Packing Unit Selection (from Hydril) Secondary Rod Seal Cameron Type U Ram Preventer Opening and Closing Ratios
6-8 6-9 6-10 6-11 6-12 6-13 6-14 6-17 6-18 6-21 6-22 6-25 6-26 6-27 6-28 6-34
6.10 Approved BOPs for Company Operations 6.11 Availability and Bore of Blowout Preventers by Major Manufacturers 6.12 5M Surface BOP Stack 6.13 10M/15M Surface BOP Stack 6.14 Four Inlet/Outlet 10M/15M Subsea BOP Stack 6.15 Three Inlet/Outlet 10M/15M Subsea BOP Stack 6.16 Specifications for BOP Flanges, Ring Gaskets, Flange Bolts and Nuts
6-5
March 1995
6.17 Hydril Drilling Spool Data 6.18 Choke Manifold, 10M/15M 6.19 Standpipe Manifold 6.20 Subsea Diverter Stack
6-6
March 1995
1 Annular Preventers
Annular preventers have a doughnut shaped elastic element with bonded steel internal reinforcing. Extrusion of the element into the wellbore is effected by upwards movement of a hydraulically actuated piston. The element is designed to seal around any shape or size of pipe and to close on openhole. (See Figure 6.1.) An important function of annular preventers is to facilitate the stripping of the drillpipe in or out of the well, with pressure on the wellhead. Undue wear of the element is avoided by the use of pilot-operated hydraulic regulator, which controls closing pressure. The majority of annular preventers currently in use are manufactured by Hydril (Types MSP, GK, GL, GX), Shaffer (Spherical) and Cameron (Type D), these are illustrated below (See Figures 6.2 to 6.6) together with a summary of major operating features. The following are the most important aspects of the operation of annular preventers: To obtain maximum sealing element life, hydraulic closing pressures should conform to the manufacturers recommendations for pressure testing and operational use of the preventers. Excessive closing pressures, when coupled with wellbore pressure sealing effects, cause high internal stresses in the element and reduce element life. Cavities should be flushed out and the element inspected following each well. Preventers should be stripped and inspected annually. Seals should be replaced and all sealing surfaces inspected. Cap seals should be replaced when changing elements. Drilling tools, especially rock bits, should be run cautiously through BOPs to minimise element damage. Elements of annular preventers do not, on occasions, retract fully. The type of elastomer (natural rubber, synthetic rubber, neoprene) used in the packingelement should be the most suitable for a particular wellhead environment. See Figure 6.7. Although most models and sizes of annular preventer will seal an openhole in an emergency operation, it is not recommended as such gross deformation of the elastomer causes cracking and accelerated wear. Closing pressures should be regulated to the pressures specified by the manufacturers. This information should be available at the rigsite. When stripping, the closing pressure should be regulated to the minimum required for a slight weeping of mud past the element. Closing pressures higher than this will increase element wear. The pipe should be moved slowly, particularly as tool joints pass through the element. The manufacturers also provide information regarding recommended closing pressures during stripping operations. Surge vessels on the closing ports will help to smooth-out surge pressures as tool joints pass through the element.
6-7
March 1995
CUTAWAY DRAWING SHOWING HOW RUBBER IS MOULDED AROUND STEEL SEGMENTS (HYDRIL)
(CAMERON)
OPEN
CLOSED ON PIPE
CLOSED ON PIPE
WEOX02.164
6-8
March 1995
PACKING UNIT
PISTON
OPENING CHAMBER
CLOSING CHAMBER
Operating Features: 1. 2. 3. 4. 5. Will close on open hole and hold 2000psi (but not recommended). Primary usage is in diverter systems. Automatically returns to the open position when closing pressure is released. Sealing assistance is gained from the well pressure. Good stripping capability of the packing unit since (fatigue) wear occurs on the outside of the packing unit.
Most annular preventers are designed to use wellbore pressure to assist in maintaining closure. In some circumstances and depending on the preventer size, the well pressure can maintain closure without any closing hydraulic pressure being applied. An annular preventer should never be operated without some closing hydraulic pressure applied. The reason is that with only well pressure maintaining closure, the packing unit may suddenly open with only a small surge or reduction in well pressure. Also, the pressure seal may be lost around the body of the drillpipe after a tool joint passes through the element during stripping operations.
6-9
March 1995
PISTON
CLOSING CHAMBER
Operating Features: 1. 2. 3. 4. 5. Will close on open hole (but not recommended). Sealing assistance is gained from the well pressure. Requires high closing pressures when used in subsea installations. Has provision to measure piston travel to gauge element wear. Available with a bolted top.
WEOX02.166
6-10
March 1995
PACKING UNIT
OPENING CHAMBER
PISTON
SECONDARY CHAMBER
Operating Features: 1. 2. 3. 4. 5. 6. Will close on open hole (but not recommended). Some sealing assistance is gained from well pressure. Bolted cover for easier element charge. Primarily designed for subsea operations. Has provision to measure piston travel to gauge element wear. Has a secondary chamber which can be connected four ways to achieve different effects: a. Minimise closing/opening fluid volumes; b. Reduce closing pressure; c. Automatically compensate (counterbalance) for marine riser hydrostatic pressure effects in deep water; and d. Operate as a secondary closing chamber.
WEOX02.167
6-11
March 1995
OPENING CHAMBER
PISTON
CLOSING CHAMBER
Operating Features: 1. 2. 3. 4. 5. 6. Will close on open hole (but not recommended). Requires higher closing pressure in subsea applications. Some sealing assistance is gained from the well pressure. No provision for measuring piston travel. Currently the most common annular preventer for subsea use. Important to check seals in upper adapter head when changing an element and replace if necessary.
WEOX02.168
6-12
March 1995
A 1 in. valve can be installed on both the opening and closing lines next to the annular preventer. These valves must be in the open position at all times except when testing hydraulic lines and hydraulic chamber seals. These valves can be used to verify seal leaks between the opening and closing chambers of an annular preventer.
PACKING UNIT
Operating Features: 1. 2. 3. 4. 5. 6. Quick-release top latch for easy element change. Most sizes use less closing fluid than Shaffer and Hydril annular preventers. Overall height is less than Hydril and Shaffer annular preventers. Two piece packing unit. Operational problems have been experienced with this preventer. Requires 3000psi closing pressure to close an openhole.
WEOX02.169
6-13
March 1995
DRILLING FLUID COMPATIBILITY Waterbase fluid Oil base/oil additive fluid Oil base fluid
Closing pressures must be adjusted when annular preventers are operated subsea. The manufacturers recommendations for the required adjustment pressure are summarised below: For Hydril GK and MSP, the adjustment pressure is related to the mud weight, the water depth, and the water density as follows:
P = (MW
X
1.421 X D) (w CR
D X 1.421)
(psi)
where P MW w D CR and
CR =
= = = = =
adjustment pressure (psi) mud weight in the riser (SG) sea water density (SG) water depth (m) annular closing ratio
For example CR for Hydril 13 5/8 in. 5M GK = 2.56 CR for Hydril 21 1/4 in. 2M MSP = 4.74 and so: Subsea closing pressure = Surface closing pressure + Adjustment pressure For Hydril GL operated subsea (with the secondary chamber connected to the openingline): Adjustment pressure as for Hydril GK in subsea operation.
6-14
March 1995
For Hydril GL operated subsea (with the secondary chamber connected to the closingline): Adjustment pressure = K X Adjustment pressure (as determined for Hydril GK) where
K = Closing chamber area Closing chamber area Secondary chamber area
For Hydril GL operated subsea (with secondary chamber in hydraulic communication with the riser): No adjustment required.
For the NL Shaffer annular operated subsea, tests carried out by Exxon indicated that the required adjustment to the closing pressure is given by the following: For the 16 3/4 in. 5M: For the 18 3/4 in. 5M: P = (0.335MW 0.335)D P = (0.339MW 0.318)D
6-15
March 1995
Cameron offer the following sizes of variable bore rams: BOP Bore 11 in. 11 in. 13 5/8 13 5/8 16 3/4 16 3/4 18 3/4 18 3/4 Pipe Size Range 5 5 7 5 7 5 7 5 in. 2 7/8 in. 1/2 in. 3 1/2 in. in. 4 1/2 in. in. 2 7/8 in. in. 3 1/2 in. in. 2 7/8 in. 5/8 in. 3 1/2 in. in 2 7/8 in.
6-16
March 1995
closed position after they are shut hydraulically. In an emergency, the screws can be used to close the rams if the hydraulic system fails. If the locking screws are used to close the rams, the hydraulic closing unit valve handle should be turned to the closed position. This will eliminate the possibility of hydraulic oil being trapped on the opening side of the actuating pistons. An optional hydraulic lock mechanism (Camerons Wedge Lock, Shaffers Poslock and Hydrils MPL) can be used in place of locking screws to lock the rams in the closed position. The hydraulic lock holds the rams closed until unlocking pressure is applied even though the primary control pressure is released. The hydraulic ram lock was developed for subsea BOP stacks and can be used on land rigs in place of the manually operated locking screws.
CHECK VALVE ENERGISING RING HYCAR LIP SEAL RETAINER RING AND LOCKING RING PLASTIC PACKING RING
'O' RING VENT TO ATMOSPHERE PREVENTER BODY OPERATING CYLINDER PREVENTER BONNET
WEOX02.171
6-17
March 1995
SIZE
WP (psi)
Hydril Ram Open 1.5 1.5 1.7 6.6 2.6 2.6 Close 5.4 5.4 8.2 7.6 5.3 5.3
7 1/16 in.
3,000 5,000 10,000 15,000 2,000 3,000 5,000 10,000 2,000 3,000 5,000 10,000 15,000 3,000 5,000 10,000 15,000 2,000 3,000 5,000 10,000 10,000 15,000 2,000 3,000 5,000 10,000 2,000 3,000
3.37
7.11
9 in.
11 in.
2.5 2.5 2.5 2.5 2.2 2.3 2.3 2.3 5.6 2.3 2.3 2.3 3.6 4.1 1.3 1.3 5.1 4.1 1.0
7.3 7.3 7.3 7.3 9.9 7.0 7.0 7.0 8.4 6.8 6.8 6.8 7.4 9.7 7.0 7.0 6.2 7.2 7.0
13 5/8 in.
16 3/4 in.
1.63
7.11
26 3/4 in.
6-18
March 1995
It should also be noted that, for high wellbore pressures, pressures greater than 3000 psi may be required to open some ram preventers. Bonnet Seals Bonnet (or door) seals are exposed to wellbore pressures and fluids. Since they can be subjected to high pressures and temperatures without being backed-up by another seal, bonnet seals are critical to the integrity of the BOP system. The seals are generally of fibrous/rubber construction and require careful handling and installation. Manufacturers recommendations should be observed meticulously. Bonnet seals should be replaced each time bonnets are opened. Bonnet seals should be handled carefully, particularly on installation, and be stored at controlled temperatures in darkness. They should be discarded after storage for one year. Bonnet bolts should be made up to manufacturers recommended torques, which can be extremely high with some compression-type seals. Due regard should be paid to the type of lubricant used, eg make-up torque is reduced by approximately 50% if a molybdenum disulphide lubricant, rather than an API5A lubricant, is used. Bonnet faces, preventer faces and seal grooves should be clean and dry before seal installation and make-up. Bonnet seals should be tested after installation. The following are the most important aspects of the care and maintenance of ram preventers: Pipe rams should not be closed on openhole or on mis-matched pipe. This would induce excessive extrusion of the elastomer and can cause cracking or bonding failures. Ram recesses should be washed out and the ram element inspected following each well. Preventers should be stripped, inspected (particularly all sealing surfaces) and seals replaced annually. When in good operating condition, ram preventers should close with 300 psi or less hydraulic pressure without wellbore pressure. If high closing pressure is required during test operations, the preventer should be checked first for debris in the ram cavity and then inspected for piston rod misalignment or other mechanical problems. Wellbore pressure helps close ram preventers. They are designed to hold pressure from the lower side and will not seal properly if installed upside down. Also, ram preventers are not designed to be pressure tested from the top side and this can damage the preventer. Field experience has proven that ram preventers are more likely to leak with a lowwellbore pressure than a high pressure. For this reason, they should be tested at 200/300psi prior to the rated working pressure test. Ram preventers will close faster than annular preventers, especially in the larger sizes. Usually, ram preventers require only one-third or less of the hydraulic fluid volume to close compared to an annular. In instances where mechanical problems prevent rapid closure of the annular preventer, a ram preventer should be closed immediately to minimise additional well flow. The main closing unit control handle for operating blind or blind/shear rams should always be protected against accidental closure with pipe in the hole. Numerous costly incidents have resulted from accidentally closing the blind rams and flattening or cutting
6-19
March 1995
the drillpipe during well control or drilling operations. A flip-up cover without locking device should be used. If the handle is locked in the open position, it prevents closing the preventer from a remote station. Shear rams are not recommended for land rig operations. When aluminium drillpipe is used, special consideration must be given to ram size selection. For example, 5 in. aluminium drillpipe has an outside body diameter of 5.150 in., versus a 5.000 in. body diameter for 5 in. steel pipe. Thus, regular 5 in. ram blocks must be slightly modified to seal and not damage the main tube section of aluminium pipe. In addition, 5 in. aluminium pipe has a tapered transition zone for a length of 41 in. to 46 in. on both the box and pin ends from 5.150 in. OD up to 5.688 in. OD. Standard rams will not seal on the tapered end sections. Variable bore rams can be used to seal on the body and end sections of aluminium drillpipe. Ram preventers can be used to strip drillpipe in or out of the hole under pressure, but it is necessary to use two preventers which have sufficient distance between rams to isolate a tool joint box. The drilling spool provides this space in a five preventer stack. The upper and lower rams of a double ram preventer are too close together for this purpose. Excessive hydraulic pressure should not be applied on the rams when stripping pipe under pressure because it tends to wear the resilient material of the ram. The lowest ram in the BOP stack should never be used for stripping since it is always considered the master valve.
Multiple stack systems Bore 21 1/4 in. 13 5/8 in. 11 in. Working Pressure 2M or 5M 5M, 10M or 15M 5M, 10M or 15M
Figure 6.10 shows a summary of approved BOPs. Figure 6.11 shows the availability and bore of BOPs from the major manufacturers.
6-20
March 1995
OFFSHORE DIVERTER SYSTEMS Hughes Offshore Hughes Offshore Hydril C Type KFDS (Floating) Type KFDJ (Platform) Type FSP (Floating and Platform) (with Flow Selector) ACCEPTABLE RUBBERS* Nitrile Water and Oil Muds The following types of Hydril rubbers are available: 1. Natural rubber (black) Water base muds 2. Synthetic rubber (red) All muds 3. Neoprene rubber (green) Low temperature service and oil muds The following types of Shaffer rubbers are available: 1. Natural rubber (black) Water base muds 2. Nitrile (blue) Oil and water base muds ACCEPTABLE RUBBERS* Super Wear Water and oil muds Super Wear Water and oil muds Nitrile Water and oil muds The following types of Shaffer rubbers are available: 1. Natural rubber (black) Water base muds 2. Nitrile (blue) Oil and water base muds
ANNULAR PREVENTERS Cameron Type D Hydril MSP, GK, GKS, GS, GL, GX
Shaffer Spherical
RAM PREVENTERS Cameron Type QRC Cameron Type U Hydril Type Ram Shaffer Type LWP Shaffer Type LWS Shaffer Type SL Koomey Type PL PB *
All BOP manufacturers specify their rubber elements and rams as H2S resistant; however, H 2S exposure reduces the service life of rubber products. The performance of these products can vary significantly, depending on the extent of exposure and H2S content. Shaffer Type 70 ram blocks are not acceptable because of metallurgical and rubber packer failures.
**
6-21
March 1995
March 1995
Blowout Preventer Nominal Size 7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 9 in. 9 in. 9 in. 9 in. 11 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 16 3/4 in. 16 3/4 in. 16 3/4 in. 18 3/4 in. 18 3/4 in. 18 3/4 in. 21 1/4 in. 21 1/4 in. 21 1/4 in. 21 1/4 in. 26 3/4 in. 26 3/4 in. 29 1/2 in. 30 Working Pressure 3,000 5,000 10,000 15,000 2,000 3,000 5,000 10,000 2,000 3,000 5,000 10,000 15,000 3,000 5,000 10,000 15,000 2,000 3,000 5,000 10,000 5,000 10,000 15,000 2,000 3,000 5,000 10,000 2,000 3,000 500 1,000 Cameron U Cameron QRC Cameron D Hydril GK Hydril GL Hydril GX Hydril HSP Hydril Ram Shaffer LWS Shaffer SL Shaffer Spherical Koomey
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 16 3/4 in. 16 3/4 in. 18 3/4 in. 18 3/4 in. 21 1/4 in. 20 3/4 in. 21 1/4 in. 21 1/4 in. 26 3/4 in.
7 1/16 in. 7 1/16 in. 9 in. 9 in. 11 in. 11 in. 13 5/8 in. 16 3/4 in.
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 16 3/4 in. 18 3/4 in. 18 3/4 in. 21 1/4 in.
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 9 in. 9 in. 9 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 16 3/4 in. 16 3/4 in.
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 9 in. 9 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 18 3/4 in. 18 3/4 in. 21 1/4 in. 20 3/4 in. 21 1/4 in.
7 1/16 in. 7 1/16 in. 9 in. 11 in. 11 in. 21 1/4 in. 20 3/4 in.
7 1/16 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 16 3/4 in. 18 3/4 in. 18 3/4 in. 21 1/4 in.
7 1/16 in. 7 1/16 in. 7 1/16 in. 9 in. 9 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 16 3/4 in. 18 3/4 in. 21 1/4 in. 21 1/4 in.
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 18 3/4 in. 18 3/4 in. 21 1/4 in. 20 3/4 in. 21 1/4 in. 21 1/4 in.
6-22
The test pressure rating of BOP equipment is a one off test, conducted on the BOP (or valve) body at the time of manufacture to a pressure 50% greater than the working pressure. In service, working pressure ratings should not be exceeded. It is acceptable to use annular preventers rated at 5000 psi less than the rams for certain 10M and 15M applications.
4 Stack Configurations
Company policy regarding minimum stack configurations for all categories of land and offshore operations is detailed in the Drilling Policy and Guidelines Manual. Figures 6.12 to 6.15 show examples of acceptable stacks for various applications. The particular details of each well will however dictate the most suitable stack for each application.
6-23
March 1995
Well can be circulated either under the annular preventer or under the upper pipe rams. Lowermost rams not used for stripping operations and only used when no other ram available for this purpose. If casing rams are required they should be positioned in the upper pipe ram preventer cavity. The rams should be changed out on the trip out of the hole prior to running casing, before pulling the BHA through the stack. The bonnet seals are tested against the test plug and the annular prior to running casing.
6-24
March 1995
ANNULAR BOP
BLIND RAMS
PIPE RAMS
SECTION A 1 2 1. Flanged gate valves 2in minimum ID same working pressure as 'A' section. The outside valve is the working valve during drilling operation. This valve is removed and reused after completion. 2. Tee with tapped bullplug, needle valve, and pressure gauge. 3. Flanged gate valve 2in minimum ID same working pressure as BOP stack. 4. As 3. or flanged spring-loaded type check valve 2in minimum ID same working pressure as BOP stack. 5. Drilling spool two flanged side outlets 3in choke and 2in kill line minimum IDs. 6. Flanged hydraulically controlled gate valve 3in minimum ID same working pressure as BOP stack. 7. Flanged gate valve 3in minimum ID same working pressure as BOP stack. Top of annular preventer must be equipped with API flange ring gasket. All flange studs must be in place or 8. holes filled in with screw type plugs. NOTES: Unless specified otherwise in the Bid Letter and/or Contract, the contractor will furnish and maintain all components shown except the 'A' section and items 1 and 2, which will be furnished by the Company. The choke line between the drilling spool and choke manifold should not contain any bend or turn in the pipe body. Any bend or turn required should be made with a running tee with a blind flange or welded bullplug. All connections should be flanged or welded. All fabrications requiring welding must be done by a certified welder. Welds should be stress relieved.
WEOX02.175
CASING SPOOL
6-25
March 1995
9
FLOWLINE FILL UP LINE
ANNULAR BOP
BLIND RAM OUTLET FLANGE (USED ONLY FOR RAM COMBINATION STRIPPING) UPPER PIPE RAM
3 2
SECTION B
1 2
SECTION A
1. Flanged gate valves 2in minimum ID same working pressure as 'A' section. The outside valve is the working valve during drilling operation. This valve is removed and reused after completion. 2. Tee with tapped bullplug, needle valve, and pressure gauge. 3. Flanged gate valve 2in minimum ID same working pressure as 'B' section. 4. Flanged gate valve 2in minimum ID same working pressure as BOP stack. 5. Drilling spool two flanged side outlets 3in choke and 2in kill line minimum IDs. 6. Flanged hydraulically controlled gate valve 3in minimum ID same working pressure as BOP stack. 7. Flanged gate valve 3in minimum ID same working pressure as BOP stack. 8. Top of annular preventer must be equipped with API flange ring gasket. All flange studs must be in place or holes filled in with screw type plugs. 9. The ID of the bell nipple must be less than the minimum ID of the BOP stack. NOTES: Unless specified otherwise in the Bid Letter and/or Contract, the contractor will furnish and maintain all components shown except the 'A' and 'B' sections and items 1 and 2, which will be furnished by the Company. The choke line between the drilling spool and choke manifold should not contain any bend or turn in the pipe body. Any bend or turn required should be made with a running tee with a blind flange or welded bullplug. All connections should be flanged or welded. All fabrications requiring welding must be done by a certified welder. Welds should be stress relieved.
WEOX02.176
6-26
March 1995
KILL LINE
CHOKE LINE
RISER CONNECTOR
BLIND/SHEAR RAMS
PIPE RAM No 2
PIPE RAM No 3
PIPE RAM No 4
WELLHEAD CONNECTOR
WEOX02.177
6-27
March 1995
KILL LINE
CHOKE LINE
RISER CONNECTOR
BLIND/SHEAR RAMS
PIPE RAM No 2
PIPE RAM No 3
PIPE RAM No 4
WELLHEAD CONNECTOR
WEOX02.178
6-28
March 1995
Annular and ram combination stripping is possible with this stack. Ram combination stripping is however considered impractical from a floating rig. The lowermost ram is not used for stripping operations.
Disadvantages: Reliance is placed on the annular preventers when running casing. Unless variable or suitable sized pipe rams are installed initially, the stack must be pulled and redressed before using tapered strings. Annular preventers are not to be considered an adequate substitute for pipe rams when using tapered strings. The requirement to have the facility to hang off on pipe ram no. 2 and shear on the pipe body generally means that it is not possible to use a double preventer for these two rams. If it is not possible to shear on the pipe body when the drillstring is hung off on pipe ram no. 2, the drillstring should be hung off on pipe ram during well control operations. This is undesirable and reduces the flexibility of the stack.
6-29
March 1995
Choke and Kill Outlets To comply with Company policy, the choke line must have a minimum ID of 3 in., the kill line may be as small as 2 in. (but this might restrict operational flexibility should it be required to substitute for a washed out choke line). During normal operations, the inner (manual) choke and kill line valves should be open and the outer (HCR) valves closed.
Remote Kill Line On a land rig, a remote kill line can be tied in to the kill line so that it may be used whichever preventer is closed. The remote kill line should be rated at the pressure rating of the BOP stack and should terminate at a similarly rated flanged valve, at least 100 ft from the well. The purpose of this line is to enable a pump truck to be tied into the well in an emergency situation.
Wellhead Outlets It is recommended that wellhead spool outlets are not used for a choke and kill line tie-ins. Each wellhead spool should have dual valve isolation on one side and valve removal plugs (VRP) should be installed on the non-active side.
Check Valves Traditionally, a check valve has been installed outboard of the stack valves on the kill line. Now many rigs, particularly jack-ups, have the facility to use the kill line to augment, or replace, the choke line. In such a hook-up, check valves are omitted. Company policy is that check valves are not mandatory on the kill line.
Choke and kill lines are generally fabricated in line with the following specifications: All connections should be flanged, clamped or welded. Screwed fittings, unions and chicksans should not be used on the choke lines, although minimal use is acceptable on kill lines. All welding should be carried out under shop conditions with machine cut weld preparations. All welding should be conducted by certified welders to approved weld procedures and all welds should be suitably non destructively tested and pressure tested prior to use. Lines, particularly the primary choke line should be installed with the minimum number of bends. Where bends are required, targeted tees, or block tees should be used. Swept bends are undesirable. Choke lines should be well braced, to withstand severe vibration. Supports should be fitted as required, but these should not be welded to the choke line.
6-30
March 1995
Choke and kill lines are tied into BOP outlets, not to drilling spools or the wellhead. Generally BOP stacks for exploration wells should have 4 ram, and 2 annular preventers. This provides some flexibility in case a ram or element fails during a well killing operation. On some rigs, a hydraulically actuated cement dump valve is provided on the kill line. This valve may be used to dump cement returns, thereby avoiding long circulation times up the riser in deep water. It can also be used to flood the riser if it becomes evacuated and in danger of collapsing. However, the dump valve should be treated with caution. Misuse, or inadvertent opening, could cause considerable loss of hydrostatic head in the well. Often dump valves are considered to be unnecessary and are omitted on most rigs. The following points should be noted regarding the major choke and kill line components: All valves should be failsafe. Two valves are required per outlet. Valves should be installed as close to the BOP outlets as possible, and preferably in line with the outlets. Side-arms and valves should be well protected by the framework around the stack. Targeted tees should be used for all 90 degree bends. Choke and kill connections at the lower riser disconnect should be rigidly supported by the framework, so that they will not part when full working pressure is applied simultaneously to both lines. The choke and kill line across the ball/flex joint should be flexible and not restrict movement of the joint up to its maximum designed deflection. Riser couplings and the LMRP stab plates should be designed to withstand induced loadings when full working pressure is applied simultaneously to both lines. The choke and kill lines across the telescopic joint should be able to accommodate the maximum designed travel of the joint. All surface connections should be flanged, clamped or welded. Screwed fittings, chicksans and unions should not be used. Lines should be installed with the minimum number of bends. Where bends are required, targeted tees or block tees should be used. Swept bends are not desirable. Choke lines should be anchored to withstand vibration. Supports should be fitted as required, but these should not be welded to the choke line. Both the choke and kill line should be tied into the choke manifold to allow one to replace or augment the other.
6-31
March 1995
pressures are required for compatible operations with standard BOP closing units. The choke line valve must be operable from both the main and remote closing units. As an optional feature, hydraulically operated valves are available with stem and handle for manual operation (to close but not open) in case of hydraulic system failure. Although numerous companies manufacture HCVs, Cameron Iron Works and NL Shaffer supply the majority of remotely operated choke line valves since they are initally ordered as a component of the BOP stack. On most rigs, the hand operated gate valves used for the choke manifold and kill line are usually the same type as the HCV. Cameron introduced the HCR (High Closing Ratio) as the first remotely controlled valve for choke line service. This valve has the same basic design and operational features as a Cameron QRC preventer. The HCR valve has been used so extensively throughout the industry that most oil field personnel refer to any make of remotely controlled valve as the HCR. Because the HCR is limited to 5000 psi working pressure, the advent of 10,000 psi and higher working pressure BOP required additional valve development. Currently, Camerons type F hydraulically operated gate valve is probable the most widely used and is available with rated working pressures from 3000 psi to 15,000 psi. NL Shaffers choke line valve is the type DB which is rated for 5000, 10,000 and 15,000 psi working pressures. Other reputable valve manufacturers equipment may be acceptable for choke line service; however, prior well control reliability and experience should be verified.
6-32
March 1995
Prior to a modification made in June, 1972, the valve was marginally failsafe in 440m of water; but now a modified version, Model EDU, is available that is independent of water depth. WKM Valve (Model M with D-2-C Operator) This valve will fail safe closed in 570m water depth. It has no balancing stem; therefore, the body volume decreases when the valve opens. WKM claims that the body cavity is so large compared with this change that pressure locking is no problem. The problem of body lock is solved in the same way as fluid lock when the valves are in tandem. Basically, flow paths within the body allow redistribution of the small volume change resulting from valve stem movement. Cameron Valves Cameron has three subsea valve designs: 1) the Type A valve has a solid gate for uni-directional sealing, 2) The Type AF valve has bi-directional sealing capability with a ported outlet in the lower body cavity to prevent liquid locking, and 3) the type DF valve is bi-directional with a balancing stem ported to the sea and a vertical fluid passageway on the outer surface of the gate to prevent pressure locking. Cameron rates the Type A and the AF valves for service in water depths to 300m. The Type DF valve is rated by Cameron for service to a water depth of 1800m. All valves were originally designed with a dog attached to the gate to rotate the seats a fraction of a turn when the valve was opened, which reportedly would provide uniform seat wear. Cameron now recommends removal of the dog since its action can induce stresses which may cause the seat to fracture if settled baryte and/or drilled solids prevent seat rotation. Vetco Type VS Valve The Vetco Type VS subsea gate valve is a full-bore, through-conduit gate valve. A metal-to-metal seat is provided between the moving sealing member (gate) and the stationary seats. Since the valve lacks a balancing stem, the manufacturer limits the valve to water depths of 900m to ensure failsafe closure.
6-33
March 1995
March 1995
Figure 6.16 Specifications for BOP Flanges, Ring Gaskets, Flange Bolts and Nuts
All blowout preventers, drilling spools, adapter flanges will be furnished with the specific API ring joint flange equipment listed below: RATING OF BOP STACK 2000 psi wp and 3000 psi wp Installations 5000 psi wp Installations
6-34
APPROVED FLANGES
API Type 6B with Type R Flat Bottom Groove API Type 6B with Type R Flat Bottom Groove or API Type 6BX w/Type BX Groove API Type 6BX with Type BX Groove
API Type BX
* Acceptable material for flange ring gaskets as per API Spec 6A, Wellhead Equipment. Sweet Oil Sour Oil or Gas Low Carbon Steel Type 316 stainless steel preferred but Type 304 stainless steel acceptable except for high risk H2S wells.
** In some H2S applications, ASTM A-193 Gr B M a maximum Rockwell hardness of 22 may be acceptable. If used, flanges should be derated per Table 1.4B of API 6A.
WEOX02.180
BORE
CONNECTIONS
SIDE OUTLETS
(B) SPOOL CENTRE LINE TO FLANGE OR HUB FACE (in.) 13.25 13.25 13.50 13.50 13.50 15.18 15.18 16.44 15.00 15.00 15.25 15.25 16.50 16.50 17.25 17.25 18.62 18.62 18.62 18.62 22.50 17.25 17.25 19.00 19.00 19.00 19.00 20.88 20.88 20.88 20.88 22.52 22.52 22.52 22.52 21.75 21.75 21.75 21.75 21.75 25.25 25.00 27.52 27.52
7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 7 1/16 in. 9 in. 9 in. 9 in. 9 in. 11 in. 11 in. 11 in. 11 in. 11 in. 11 in. 11 in. 11 in. 11 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 13 5/8 in. 20 3/4 in. 20 3/4 in. 20 3/4 in. 20 3/4 in. 21 1/4 in. 21 1/4 in. 21 1/4 in. 21 1/4 in. 21 1/4 in. *29 1/2 in. *29 1/2 in. *30 in. *30 in.
7 1/16 in. 3,000 Flange 7 1/16 in. 3,000 Flange 7 1/16 in. 5,000 Flange 7 1/16 in. 5,000 Flange 7 1/16 in. 5,000 Flange 7 1/16 in. 10,000 Flange 7 1/16 in. 10,000 Flange 7 1/16 in. 15,000 Flange 9 in. 3,000 Flange 9 in. 3,000 Flange 9 in. 5,000 Flange 9 in. 5,000 Flange 11 in. 3,000 Flange 11 in. 3,000 Flange 11 in. 5,000 Flange 11 in. 5,000 Flange 11 in. 10,000 Flange 11 in. 10,000 Flange 11 in. 10,000 Hub 11 in. 10,000 Hub 11 in. 15,000 Flange 13 5/8 in. 3,000 Flange 13 5/8 in. 3,000 Flange 13 5/8 in. 5,000 Flange 13 5/8 in. 5,000 Flange 13 5/8 in. 5,000 Hub 13 5/8 in. 5,000 Hub 13 5/8 in. 10,000 Flange 13 5/8 in. 10,000 Flange 13 5/8 in. 10,000 Hub 13 5/8 in. 10,000 Hub 20 3/4 in. 3,000 Flange 20 3/4 in. 3,000 Flange 20 3/4 in. 3,000 Flange 20 3/4 in. 3,000 Hub 21 1/4 in. 2,000 Flange 21 1/4 in. 2,000 Flange 21 1/4 in. 2,000 Flange 21 1/4 in. 2,000 Hub 21 1/4 in. 2,000 Hub 29 1/2 in. 500 Flange 29 1/2 in. 500 Flange 30 in. 1,000 Flange 30 in. 1,000 Flange
3 1/16 in. 3,000 Flange 3 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Flange 4 1/16 in. 5,000 Flange 3 1/16 in. 10,000 Flange 4 1/16 in. 10,000 Flange 4 1/16 in. 15,000 Flange 3 1/16 in. 3,000 Flange 3 1/16 in. 5,000 Flange 2 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Flange 3 1/16 in. 3,000 Flange 3 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Flange 4 1/16 in. 5,000 Flange 3 1/16 in. 10,000 Flange 4 1/16 in. 10,000 Flange 3 1/16 in. 10,000 Hub 4 1/16 in. 10,000 Hub 4 1/16 in. 15,000 Flange 3 1/16 in. 3,000 Flange 3 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Flange 4 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Hub 4 1/16 in. 5,000 Hub 3 1/16 in. 10,000 Flange 4 1/16 in. 10,000 Flange 3 1/16 in. 10,000 Hub 4 1/16 in. 10,000 Hub 3 1/16 in. 3,000 Flange 3 1/16 in. 5,000 Flange 4 1/16 in. 5,000 Flange 3 1/16 in. 3,000 Hub 7 1/16 in. 2,000 Flange 3 1/16 in. 5,000 Flange 4 1/16 in. 5,000 Flange 3 1/16 in. 5,000 Hub 4 1/16 in. 5,000 Hub 7 1/16 in. 500 Flange 12 in. 500 Flange 7 1/16 in. 5,000 Flange 12 in. 1,000 Flange
510 525 510 500 525 1025 1075 1400 700 725 710 725 950 975 1065 1290 2190 2215 1285 1310 1710 1055 1080 1755 1780 1050 1075 3325 3355 1925 1950 2590 2615 2540 2565 1850 1800 1850 1850 1825 2380 2320 2500 2450
16.50 16.50 16.50 19.75 19.75 21.12 21.12 22.38 18.12 18.12 18.12 18.12 18.62 18.62 23.38 22.38 25.12 25.12 25.12 25.12 29.75 19.38 19.38 22.38 22.38 22.38 22.38 27.75 27.75 27.75 27.75 27.12 27.12 27.12 27.12 23.38 23.38 23.38 23.38 23.38 31.75 31.75 40.00 40.00
6-35
March 1995
API high-pressure connections are pressure sealed by means of ring-joint gaskets made of soft iron, low-carbon steel or stainless steel. API Type RX and Type BX ring-joint gaskets are pressure-energised seals but are not interchangeable. Rings that have been coated with Teflon, rubber or other resilient materials are not acceptable. All flanges in the stack and side-outlets should be fitted with new ring-joint gaskets each time they are assembled. It is important that the ring groove in the flange be clean and dry prior to flanging up. API Standard 6A, Wellhead Equipment, provides specifications for flanged wellhead fittings. API Type 6B flanges are available in the following pressure ratings: 2000 psi to 5000 psi range. API Type 6BX flanges are available for the 5000 psi to 30,000 psi range. Figure 6.16 lists specifications for BOP flanges, ring gaskets and bolts. Bolts must always be the right size not larger and not smaller than required for the specific bolt holes. Hub and clamp connectors are principally used on subsea BOP stacks to reduce the weight and height. The bolts are designed for easier make-up, especially in cramped quarters, because the wrench movement is downward instead of horizontal. When clamp connectors were first used there were numerous problems with the clamp loosening during drilling operations and creating a hazard in well control situations. This problem has been greatly reduced by the manufacturer furnishing recommended bolt torque make-up values and the avialability of power torque wrenches on the rigs. Cameron Iron Works clamp connections are installed on most major manufacturers hub and clamp preventers. When a clamp connected BOP stack is used, recommended torque requirements should be obtained from the manufacturer and all bolts should be made up to the required torque with power wrenches.
6-36
March 1995
A bypass line to the poorboy degasser is provided in order to be able to deal with returns in the event of failure of the buffer tank. It is recognised that the majority of choke manifolds installed on drilling rigs comprise a buffer tank into which all the lines downstream of the chokes are tied. Field personnel should be aware that this design compromise seriously reduces the flexibility/redundancy of the manifold. If the buffer tank cuts out, the manifold is in effect rendered useless. Consideration should therefore be given to installing split buffer tanks and separate flare lines or, as previously mentioned, a bypass line upstream of the buffer tank. All connections should be flanged, welded, or clamped. Field welding is not acceptable. Company policy specifies that choke manifolds should incorporate at least two variable chokes on offshore rigs, one of which must be remotely adjustable. On some manifolds, mandatory in some areas, an additional outlet from the buffer chamber is provided, so that hydrocarbons can be directed via a production separator to a flare. An inlet to facilitate the tying-in of a specialised choke manifold during formation testing is also provided. On wells where there is a possibility of encountering hydrogen sulphide, all equipment and material should be suitable for sour service. The control panel for the chokes should be near the Drillers station, and should have read-outs for standpipe manifold pressure, choke manifold pressure and pump stroke counters. A pressure gauge reading standpipe pressure should be located at the choke manifold if manual chokes are used during a well kill operation. The MAASP function, where fitted, should not be used. A recording chart for standby pressure and choke manifold pressure, may also be considered. This chart can be used when testing BOPs, or when handling kicks. Under normal drilling conditions, valves on the choke line and manifold should be left open up to the valve immediately upstream of the remotely operated choke that will be used in the event of a kick. The valves downstream should be open to the poorboy degasser and mud tanks. The remote adjustable choke(s) should be left closed. The outer choke (HCR or failsafe) valve on the BOP stack should be closed during drilling. It must be possible to record choke pressure when the well is shut-in with the choke manifold lined up in this manner.
6-37
March 1995
TO POORBOY DEGASSER
TO MUD PITS
2 4
1. 2. 3. 4.
10000psi gate valves. 5000psi gate valves. Remote controlled chokes. Manually adjusted chokes.
6-38
March 1995
SECONDARY STANDPIPE
PRIMARY STANDPIPE
AUXILIARY TIE-IN POINT TO CHOKE MANIFOLD OR PRESSURE GAUGE 1. 5000psi gate valves. 2. Gate valve to suit pressure rating of standpipe manifold.
MUD PUMP
MUD PUMP
WEOX02.182
7 Diverters
If a kick is taken when conductor is set in incompetent formation, the well will not be shut-in, but instead, will be diverted. A surface diverter system, consisting of an annular preventer and vent lines, allows the flow to be directed to a safe area, away from the rig and personnel. Vent lines should be as large (12 in. minimum on offshore rigs) and as straight as practical, so as to minimise back pressure, erosion and the risk of plugging by well debris. The lines should be sufficiently braced to absorb severe shock loadings; sections likely to suffer erosion e.g. bends, should be reinforced. There should be no restriction to the bore, any valves in the lines should be full opening ball valves. Periodically, the lines should be flushed through to ensure that they remain unobstructed. To prevent the well being inadvertently shut in, any valves in the vent line should be designed to automatically open when the diverter is closed. An acceptable alternative is to elevate the vent line above the flowline, so that no valves are necessary. If the BOP stack is installed, the control panels should be clearly marked that the well is not to be closed in, but that the diverter is to be actuated.
6-39
March 1995
21in HST RISER COUPLING PIN MUD BOOST LINE CONNECTION 211/4in 2000 MSP ANNU-FLEX
FLEX JOINT
ANNULAR BOP
C/K VALVE
30in LATCH
WEOX02.183
6-40
March 1995
The working pressures of the diverter and vent lines is not of prime importance (particularly on floating rigs where the slip joint packing may be the limiting factor); 500 psi is a typical rating. Company policy states that subsea wells should be drilled riserless until a pressure containment string is set. This is to avoid allowing shallow gas flow to the rig. If however it becomes necessary to drill for surface casing with a riser, Company policy states that the well will be diverted subsea in the event of a shallow gas flow. The most likely stack up that will be used to divert subsea will comprise the following: Pin connector with subsea dump valves (minimum ID 10 in.). LMRP with annular preventer.
This will be a relatively inexpensive stack that will in most cases be made up mainly from existing rig equipment. In the event of a shallow gas flow the dump valves will be opened and the annular closed to divert subsea. In order to move the rig the LMRP can be disconnected and the well allowed to flow at the seabed. Various stacks have been custom made for diverting subsea in areas of high incidence of shallow gas. An example is shown in Figure 6.20; the diverter stack comprising: Flex joint Annular preventer Hydraulic connector Blind/shear ram Spool piece with two outlets with dump valves Choke/kill line Hydraulic connector
In the event of a shallow gas flow, the dump valves will be opened and the annular closed. In order to move the rig off location the blind/shear rams can be closed and the connector released.
6-41/42 6-41
March 1995
6.3
1 2 3 4
CONTROL SYSTEMS
Page
6-44 6-44 6-46 6-47
General Power Source Control Manifolds Accumulators
Paragraph
Illustrations
6.21 Subsea Stack Function Schematic 6.22 Annular Preventers fluid required to operate 6.23 Ram Preventers fluid required to operate 6.24 Compressibility Factor Nitrogen
6-43
March 1995
1 General
The Control System provides the means to individually close and open each BOP and valve, conveniently, rapidly, repeatedly and at the correct operating pressure. The equipment should be designed to operate when, in emergencies, primary rig power may not be available. The essential elements of a control system are: Power Source(s) Control Manifolds Accumulators Connecting Pipework/Hose Bundle and Wiring
Detailed specifications for a particular application will be governed by the number, size and pressure rating of BOPs. Water depth considerations will also influence the design of subsea BOP control systems. An example arrangement for subsea BOP systems is shown in Figure 6.21.
2 Power Source
(a) Primary Power Source
The primary power source should be an electrically driven pump (or pumps) located at the main control manifold. For 3000 psi accumulator systems, the pump(s) should incorporate a pressure switch set to cut in and out at 2800 psi and 3000 psi respectively. Diesel driven pumps may be substituted for land rig applications. The electric pump output should be twice that of the secondary air pumps. The combined electric and air pumps should be sufficient to charge the accumulator system from pre-charge to operating pressure in less than 15 minutes, also to close an annular preventer (without accumulator assistance) in less than 2 minutes.
6-44
March 1995
DRILLERS PANEL
DRILLERS PANEL
DRILLERS PANEL
RESERVOIR
RESERVOIR
RESERVOIR
SOLENOIDS RIG AIR COMPRESSOR 3000psi PUMP RIG AIR COMPRESSOR 3000psi PUMP RIG AIR COMPRESSOR POD SELECTOR AIR VALVE YOUR RIG PILOT REGULATOR POD SELECTOR POD MOUNTED REGULATOR SPM VALVE
KR KR KR
3000psi PUMP
KR
KR
KR
YOUR RIG
YOUR RIG
REDUNDANT POD
REDUNDANT POD
REDUNDANT POD
RAM PREVENTER CLOSING LINE BOP RAM PREVENTER OPENING LINE WELLHEAD
RAMS CLOSED
BLOCK POSITION (RAMS CLOSED, NO PRESSURE) 3000psi Accumulator Fluid Pressure 125psi Rig Air Pressure Vent/or No Pressure Regulated KR Fluid Pressure
RAMS OPENED
WEOX02.184
3 Control Manifolds
The BOP control systems should ideally be equipped with 3 control manifolds or panels.
6-46
March 1995
Covers, or interlocks, for critical functions, eg shear rams, wellhead disconnect. Visual and/or audible warning devices for low accumulator pressure, air pressure, or fluid levels. Where applicable, controls for diverter functions.
4 Accumulators
The hydraulic fluid required to operate the BOP functions is stored in accumulators, pressurised against a nitrogen inflated bladder. The accumulators should be located near the main control manifold location. The purpose of the accumulators is to provide a store of hydraulic energy and a high rate supply of hydraulic fluid to the BOP functions. The response time of the BOP functions is therefore independent of the output of the pumps. For subsea installations, at least two accumulators should be isolated from the main bank to provide pilot line pressure. Also, to ensure acceptable response times, additional accumulators should be mounted on the BOP stack. Accumulator bottles should be used as surge dampeners on annular preventers for stripping operations on both surface and subsea BOP stacks.
(a) Accumulator/Precharge
Operating pressure of accumulators is generally 3000 psi. The optimum bladder inflation, or precharge pressure, is governed by the minimum acceptable pressure remaining in the accumulators after operation of the preventers. About 1200 psi is required to hold some annular preventers closed. A precharge of 1000 psi will retain a small liquid reserve in the accumulator when pressure in the system falls to 1200 psi.
6-47
March 1995
The following is an example of the technique that can be used to size accumulators for a surface stack (comprising one Hydril GL 18 3/4 in. 5M annular and 3 Hydril 18 3/4 in. 10M ram preventers): Volume to close: 1 Annular 1 Ram 1 HCR valve = 44 gal = 17.1 gal = 0.6 gal
Total fluid required = 61.7 gal X 1.5 = 92.55 gal Precharge to 1000 psi, maximum operating pressure = 3000 psi, minimum operating pressure = 1200 psi
ANNULAR PREVENTERS GALLONS OF FLUID REQUIRED TO OPERATE AN OPEN HOLE Size and Working Pressure Inches 6 6 7 1/16 8 8 10 10 11 11 12 13 5/8 13 5/8 13 5/8 16 16 16 3/4 16 3/4 16 3/4 18 18 3/4 20 20 20 21 1/4 30 30 psi 3,000 5,000 10,000 3,000 5,000 3,000 5,000 5,000 10,000 3,000 3,000 5,000 10,000 2,000 3,000 3,000 5,000 10,000 2,000 5,000 2,000 3,000 5,000 5,000 1,000 2,000 Close 2.9 3.9 9.4 4.4 6.8 7.5 9.8 25.1 11.4 18.0 34.5 17.5 21.0 28.7 21.1 Hydril GK Open 2.2 3.3 3.0 5.8 5.6 8.0 Close GL Open Balancing NL Shaffer Spherical Close 4.6 4.6 7.2 11.1 11.0 18.7 Open 3.2 3.2 5.0 8.7 6.8 14.6
9.8 14.2 24.3 12.6 14.8 19.9 14.4 44.0 44.0 20.0 19.8 19.8 8.2
33.8
33.8
17.3
33.0
25.6
48.2 32.6
37.6 17.0
58.0
58.0
29.5
61.4
47.8
6-48
March 1995
RAM PREVENTERS GALLONS OF FLUID REQUIRED TO OPERATE ONE SET Size and Working Pressure Inches 4 1/16 6 6 7 1/16 7 1/16 8 8 10 10 10 11 11 11 12 13 5/8 13 5/8 13 5/8 13 5/8 13 5/8 13 5/8 13 5/8 16 16 3/4 16 3/4 16 3/4 16 3/4 18 18 3/4 20 20 20 20 21 1/4 21 1/4 21 1/4 21 1/4 21 1/4 21 1/4 26 3/4 26 3/4 psi 10,000 3,000 5,000 10,000 15,000 3,000 5,000 3,000 5,000 5,000 10,000 10,000 15,000 3,000 3,000 3,000 5,000 5,000 10,000 10,000 15,000 2,000 3,000 5,000 5,000 10,000 2,000 10,000 2,000 2,000 3,000 3,000 2,000 2,000 7,000 7,500 10,000 10,000 2,000 3,000 Close Cameron U Open Close QRC Open Close 0.59 1.22 1.22 1.22 1.22 1.17 1.17 1.17 1.17 0.81 0.81 0.95 0.95 1.19 6.35 6.35 2.58 2.58 1.74 2.98 LWS Open 0.52 0.99 5.89 5.89 2.27 2.27 1.45 2.62 Cylinder Size Inches 6 6.5 6.5 14 14 8.5 8.5 8.5 8.5 10 14 14 8.5 10 10 14 14 14 8.5 10 14 14 6.00 24.88 8.11 8.11 9.35 8.11 9.35 20.41 23.19 26.54 30.15 10.50 10.50 23.00 7.61 7.61 8.77(S) 7.61 8.77(S) 17.78 20.20(S) 21.14 27.42(S) 9.84 9.84 7.05 5.07 7.80 5.07 16.88 4.46 6.68 4.46 15.35 14 8.5 10 8.5 14 14.55 13.21 17.1 15.6 Close NL Shaffer SL Open Close E Open Manual (a) Close Open Hydril Auto (a) Close Open
2.75 2.75
7.00 4.50
7.00 8.10 3.55 4.46 4.46 10.52 7.00 10.52 3.65 3.0 3.55 2.9 2.9 5.4 11.5 5.4 11.5 11.8 11.8 4.9 11.2(S) 4.9 11.2(S) 11.8 11.8(S) 12.0 5.9 12.0 12.9 12.9 5.9 4.9 11.2(S) 4.9 11.2(S) 11.8 11.8(S)
6-49
March 1995
5.42 6.36(S) 5.42 6.36(S) 11.29 6.00 9.45 9.45 11.19(S) 11.19 7.05
15.6
14.1
14.42 16.05
12.65 13.86(P)
14 14
Therefore: P1 = 1000 + 15 = 1015 psi P2 = 1200 + 15 = 1215 psi P3 = 3000 + 15 = 3015 psi (See Figure 6.24) where: P1 P2 P3 V1 V2 V3 Z = = = = = = = precharge pressure (psi) minimum operating pressure (psi) maximum operating pressure (psi) bladder internal volume at precharge pressure (gal) bladder internal volume at P2 (gal) bladder internal volume at P3 (gal) compressibility factor for nitrogen Z1 = 1.00 Z3 = 1.06 Z3 = 1.06 T = 80 F V1 = 10 gal (11 gal bottle minus 1 gal bladder replacement)
Using the gas law: P X V = constant TXZ So in this case: 1015 X 10 1.00 V2 = 8.52 gal V3 = 3.57 gal The useable volume per bottle is given by: V2 V3 = 8.52 3.57 = 4.95 gal/bottle Therefore there is a requirement for: 92.55 4.95 = 19 bottles = 1215 X V2 1.02 = 3015 X V3 1.06
6-50
March 1995
2.2 2.1 2.0 100F 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 2000 4000 6000 8000 10000 12000 14000 16000 18000
WEOX02.187
0F
COMPRESSIBILITY FACTOR
6-51
March 1995
Surge Dampening Surge vessels should be provided for subsea annular preventers to facilitate stripping, according to manufacturers recommendations. Some preventers require surge vessels on the opening as well as closing sides. Nominal 10 gal capacity accumulators should be used.
Total fluid required = 63.5 gal Precharge to 1000 psi plus the hydrostatic of the control fluid. Therefore: P1 P2 P3 where: = = = = 1000 1747 1200 3000 + 15 + (500 X 1.03 X 1.421) psi Z1 = 1.01 + 15 + 732 = 1947 psi Z2 = 1.00 + 15 + 732 = 3747 psi Z3 = 1.09 T1 = 80 F T2 = 40 F T3 = 40 F
P1 = precharge pressure (psi) P2 = minimum operating pressure (psi) P3 = maximum operating pressure (psi) V1 = bladder internal volume at P1 (gal) V2 = bladder internal volume at P2 (gal) V3 = bladder internal volume at P3 (gal) Z = compressibility factor for nitrogen
6-52
March 1995
Using the gas law: PXV TXZ So in this case: 1747 X 10 1.01 X 540 V2 = 8.23 gal V3 = 4.66 gal The useable volume per bottle is given by: V2 V3 = 8.23 4.66 = 3.57 gal/bottle Therefore there is a requirement for: 63.5 3.57 = 18 bottles = 1947 X V2 1.02 X 500 = 3747 X V3 1.06 X 500 = constant (T in R)
A simple hook-up is impractical for subsea applications too many individual lines to be handled easily and the pressure drop through the length of line would be too great for acceptable reaction times. Instead, hose bundles are employed, which contain one high capacity (1 in.) conduit (to transfer the hydraulic fluid required to operate all functions and recharge the subsea accumulators) and up to 64 pilot (3/16 in.) lines (to direct and control the flow of fluid to a particular function). The bulk line is teed with the subsea accumulators and terminates at a regulator which reduces the accumulator pressure to operating pressure. The output of the regulator is manifolded to the pilot valves. The pilot lines terminate in function dedicated pilot (SPM) valves which respond to accumulator pressure when a function is selected. Each then allows regulated fluid to flow, via a shuttle valve, to a particular function. The displaced fluid from the opposite function is vented at its pilot valve.
6-53
March 1995
The pilot valves and regulators are housed in a wireline retrievable pod, which is duplicated to provide complete redundancy. A shuttle valve located at each function allows control by either pod. When considering a subsea system, the following should be noted: Company policy (after API RP53) recommends that the systems ensure ram preventers close within 45 seconds and annular preventers within 60 seconds of surface actuation. Electro-hydraulic systems will be required where water depths preclude satisfactory closing times with all hydraulic systems. Systems should be duplicated in all hydraulic and electric lines from the main control panel to the BOP stack functions, i.e. there should be 100% redundancy. The Drillers panel and the remote panel should be designed to select and operate either system. Dynamically positioned vessels and rigs operating in hazardous areas should have an acoustic back-up system to secure the well and release the riser. Any unused functions (such as when the low pressure stack in a two stack system is run) should be blanked off to ensure that fluid is not vented by inadvertent operation of that function.
6-54
March 1995
6.4
1 2 3 4
ASSOCIATED EQUIPMENT
Page
6-56 6-57 6-60 6-62
Mud Control and Monitoring Equipment Mud Gas Separator Drillstring Valves Rotating Heads
Paragraph
Illustrations
6.25 Typical Trip Tank Hook-up on a floating rig 6.26 An example Mud Gas Separator 6.27 Grant Rotating Head
6-55
March 1995
6-56
March 1995
RIG FLOOR
ROTARY TABLE
DIVERTER
RETURNS TO SHAKERS
FLOWLINE
6-57
March 1995
Basically, there are two types of mud gas separators: Atmospheric and Pressurised. The atmospheric type separator is standard equipment on virtually all rigs and is referred to in the field as a gas buster or poorboy separator. The main advantage of this type of separator is its operational simplicity which does not require control valves on either the gas or mud discharge lines. A pressurised mud gas separator is designed to operate with moderate back pressure, generally 50 psi or less. Pressurised separators are utilised to overcome line pressure losses when an excessive length of vent line is required to safely flare and burn the hazardous gas an extended distance from the rig. The pressurised separator is considered special rig equipment and is not usually provided by the contractor. This type of separator is installed on rigs drilling in high risk H2S areas and for drilling underbalanced in areas where high pressure, low volume gas continually feeds into the circulating fluid. During well control operations, the main purpose of a mud gas separator is to vent the gas and save the drilling fluid. This is important not only for economic reasons, but also to minimise the risk of circulating out a gas kick without having to shut down to mix additional mud volume. In some situations the amount of mud lost can be critical when surface volume is marginal and on-site mud supplies are limited. When a gas kick is properly shut in and circulated out, the mud gas separator should be capable of salvaging most of the mud. There are a number of design features which affect the volume of gas and fluid that the separator can safely handle. For production operations, gas oil separators can be sized and internally designed to efficiently separate gas from the fluid. This is possible because the fluid and gas characteristics are known and design flow rates can be readily established. It is apparent that gas busters for drilling rigs cannot be designed on the same basis since the properties of circulated fluids from gas kicks are unpredictable and a wide range of mixing conditions occur downhole. In addition, mud rheological properties vary widely and have a strong effect on gas environment. For both practical and cost reasons, rig mud gas separators are not designed for maximum possible gas release rates which might be needed; however, they should handle most kicks when recommended shut-in procedures and well control practices are followed. When gas flow rates exceed the separator capacity, the flow must be bypassed around the separator directly to the flare line. This will prevent the hazardous situation of blowing the liquid from the bottom of the separator and discharging gas into the mud system. Figure 6.26 illustrates the basic design features for atmospheric mud gas separators. Since most drilling contractors have their own separator design, the Drilling Foreman must analyse and compare the contractors equipment with the recommended design to ensure the essential requirements are met. The atmospheric type separator operates on the gravity or hydrostatic pressure principle. The essential design features are: Height and diameter of separator. Internal baffle arrangement to assist in additional gas breakout. Diameter and length of gas outlet. A target plate to minimise erosion where inlet mud gas mixture contacts the internal wall of the separator, which provides a method of inspecting plate wear. A U-tube arrangement properly sized to maintain a fluid seal in the separator.
6-58
March 1995
GAS OUTLET 8in ID MINIMUM GAS BACK PRESSURE REGISTERED AT THIS GAUGE (Typically 0 to 20psi) STEEL TARGET PLATE INLET INSPECTION COVER SECTION A-A TANGENTIAL INLET
30in OD
INSPECTION COVER
TO SHAKER HEADER TANK MAXIMUM HEAD AVAILABLE DEVELOPED BY THIS HEIGHT OF FLUID eg: 10ft HEAD AT 1.5 SG GIVES 6.5psi MAXIMUM CAPACITY 10ft APPROX
WEOX02.189
Figure 6.26
The height and diameter of an atmospheric separator are critical dimensions which affect the volume of gas and fluid the separator can efficiently handle. As the mud and gas mixture enters the separator, the operating pressure is atmospheric plus pressure due to friction in the gas vent line. The vertical distance from the inlet to the static fluid level allows time for additional gas breakout and provides an allowance for the fluid to rise somewhat during operation to overcome friction loss in the mud outlet lines. As shown on Figure 6.26, the gas-fluid inlet should be located approximately at the midpoint of the vertical height. This provides the top half for a gas chamber and the bottom half for gas separation and fluid retention. The 30 in. diameter and 16 ft minimum vessel height requirements have proven adequate to handle the majority of gas kicks. The separator inlet should have at least the same ID as the largest line from the choke manifold, which is usually 4 in. Some separators use tangential inlet, which creates a small centrifugal effect on the gas-fluid mixture and causes faster gas breakout.
6-59
March 1995
The baffle system causes the mud to flow in thin sheets which assists the separation process. There are numerous arrangements and shapes of baffles used. It is important that each plate be securely welded to the body of the separator with angle braces. A 6 in. minimum ID gas outlet is recommended to allow a large volume of low pressure gas to be released from the separator with minimum restriction. Care should be taken to ensure minimum back pressure in the vent line. On most offshore rigs, the vent line is extended straight up and supported to a derrick leg. The ideal line would be restricted to 30 ft in length and the top of the line should be bent outward about 30 degrees to direct gas flow away from the rig floor. If it is intended that the gas be flared, flame arresters should be installed at the discharge end of the vent line. As previously mentioned, when the gas pressures in the separator exceeds the hydrostatic head of the mud in the U-tube, the fluid seal in the bottom is lost and gas starts flowing into the mud system. The mud outlet downstream of the U-tube should be designed to maintain a minimum vessel fluid level of approximately 3 1/2 ft in a 16ft high separator. Assuming a 1.44 SG mud and total U-tube height of 6 ft, the fluid seal would have a hydrostatic pressure equal to 3.7 psi. This points out the importance for providing a large diameter gas vent line with the fewest possible turns to minimise line frictional losses. The mud outlet line must be designed to handle viscous, contaminated mud returns. As shown in Figure 6.26, an 8 in. line is recommended to minimise frictional losses. This line usually discharges into the mud ditch in order that good mud can be directed over the shakers and untreatable mud routed to the waste pit. In recent years, there have been a number of serious accidents caused by the failure of mud gas separators during well control situations. Primarily these have resulted from drilling contractors not updating their separator design and personnel training standards to handle high pressure gas kicks for deeper drilling operations. It is important that drilling personnel understand the limitations of all well control equipment and are trained to take remedial action before pressure or capacity limitations occur. The key initial decision that must be made is the pump rate at which the kick will be circulated out. Large influx, high pressure gas kicks should always be pumped out at low rates (generally 1 bbl/minute or less) to minimise the gas release rate at the surface where rapid gas expansion occurs. Circulating out at a slow rate reduces the risk of exceeding pressure limitations for the well control equipment and provides additional decision reaction time.
3 Drillstring Valves
Drillpipe valves are used to close in the well on the drillpipe bore and to protect surface equipment. The valves may be permanently in place, or installed at surface when required and may be of a manual shut-off or automatic check valve type. Drillstring valves should be rated to the same pressure as the BOP and tested at the same frequency. Some of the drillstring valves impose restrictions on future operations when installed. For example, both the inside BOP and drop-in valve, when in place, prevent access below them to the drillstring bore.
6-60
March 1995
6-61
March 1995
All the items in the drillstring above the landing sub must have sufficient ID to allow the check valve to pass. This includes kelly cocks, mud savers etc.
4 Rotating Heads
When used, rotating heads are installed above the BOP stack. They provide a seal on the kelly or drillpipe. A drive unit, attached to the kelly, locates in a bearing assembly above the stripper rubber. Some applications for rotating heads are: Drilling with air or gas, to divert the returns through a Blooey line. To permit drilling with underbalanced mud, by maintaining a back pressure on the wellbore. As a diverter for surface hole. To keep gas away from the rotary table. This is especially important where Hydrogen Sulphide can be expected.
Realistic working pressures for rotating heads are 500 to 700 psi. It is recommended that they are not installed for routine gas cap drilling (unless sour gas is expected) since their use precludes observation from the rig floor of annulus fluid level. Figure 6.27 shows a schedule of the Grant Rotating Head.
6-62
March 1995
KELLY BUSHING
BOWL
STRIPPER RUBBER
WEOX02.190
6-63/64 6-63
March 1995
6.5
1 2 3 4 5
EQUIPMENT TESTING
Page
6-66 6-66 6-67 6-71 6-71
General BOP Equipment and Wellheads An Example Test Procedure Test Frequency Pressure Tests of Casing
Paragraph
Illustrations
6.28 Choke Manifold Schematic 6.29 An example BOP/Choke Manifold Test Procedure 6.30 Schematic of BOP Pressure Tests 6.31 An example BOP Equipment Test Report
6-65
March 1995
1 General
The consequences of a failure of BOP equipment under operating conditions can be far reaching. Rigorous BOP testing procedures are required in order that problems may be identified under test conditions, and rectified before an emergency arises. Equipment should be tested at the time of installation on the wellhead and at regular intervals thereafter, in accordance with Company standard policies and guidelines (unless contradicted by local policies). Common causes of failure include: Casing wear. Plugging of lines with baryte. Wellhead or BOP connections working loose through vibration. Deterioration of seals in valves and BOPs. Leaks and faults occurring in control systems.
The recommended procedures in this section cover BOP stack installations at surface andsubsea.
6-66
March 1995
After latching a subsea stack, a tensile test should be applied to ensure connectors are properly latched before any pressure testing. A pressure test should be carried out to the pressure rating of the wellhead or connector, on initial installation of the stack; thereafter, the connector will be tested during BOP tests to the pressure that the BOPs will be tested to. Also, the control system should be function tested on both pods.
6-67
March 1995
Gauge Transmitter 31 32 30 26
Auto Choke 29 35 34
Manual Choke 25
Manual Choke 16
Kill Line
36
22
17
12
To Drain
To Diverter Overboard
6-68
March 1995
All tests will be carried out using a suitable test plug with only the specified drillcollarweight below; i.e. test plugs will not be run on top of a bottomhole assembly except when testing the blind/shear rams agains a backed-off test plug, if tested on a separate run. When pressure testing blind/shear rams against casing consideration should be given to pressure differential that already exists due to any difference in the weight of the mud inside and outside of the casing. All tests should be recorded on a chart. When testing blind/shear rams against a backed-off test plug, monitor volumes pumpedclosely .
TEST 1
CHOKE MANIFOLD VALVES CLOSED 3, 7, 9, 15, 18, 20, 24, 26, 30, 32, 33 2, 30
UPPER ANNULAR, UPPER INNER KILL, LOWER INNER KILL LOWER ANNULAR, UPPER OUTER KILL, LOWER OUTER KILL UPPER PIPE RAMS, UPPER OUTER KILL, UPPER INNER CHOKE, LOWER INNER CHOKE UPPER PIPE RAMS, UPPER INNER KILL, UPPER OUTER CHOKE, LOWER OUTER CHOKE MIDDLE PIPE RAMS, LOWER OUTER KILL, UPPER OUTER CHOKE
2, 30
3, 6, 11, 13
7 8
3, 7, 12, 17, 19, 22, 33, 35, 36 4, 7, 8, 10, 16, 19, 21, 25, 29 LOWER PIPE RAMS, LOWER INNER KILL, UPPER INNER CHOKE
1, 7, 8, 26, 28
6-69
March 1995
KILL
CHOKE
CHOKE
CHOKE
TEST VALVE
LOWER ANNULAR
TEST VALVE
TEST VALVE
TEST VALVE
LOWER ANNULAR
TEST VALVE
LMRP CONNECTOR
LMRP CONNECTOR
UPPER PIPE
UPPER PIPE
MIDDLE PIPE
MIDDLE PIPE
MIDDLE PIPE
LOWER PIPE
LOWER PIPE
LOWER PIPE
CONNECTOR
CONNECTOR
CONNECTOR
CHOKE
CHOKE
CHOKE
TEST VALVE
LOWER ANNULAR
TEST VALVE
TEST VALVE
LOWER ANNULAR
TEST VALVE
TEST VALVE
LOWER ANNULAR
TEST VALVE
LMRP CONNECTOR
LMRP CONNECTOR
LMRP CONNECTOR
UPPER PIPE
UPPER PIPE
MIDDLE PIPE
MIDDLE PIPE
LOWER PIPE
LOWER PIPE
CONNECTOR
CONNECTOR
CONNECTOR
WEOX02.193
6-70
March 1995
4 Test Frequency
Pressure testing of BOP equipment should be carried out according to Company policy, however in general: After installation of the wellhead component and BOP stack and prior to drilling out each casing string. At intervals not exceeding 14 days. At any time requested by the Company Drilling Representative.
Results of pressure tests should be recorded on IADC reports, and on the BOP test form. An example of a typical BOP test form is presented as Figure 6.31. The following additional points should be considered: Annular and ram (pipe) preventers should be operated on each trip into the hole with the bit at the shoe (perhaps as part of a kick drill). Blind (but not blind/shear) rams should be operated each time the bit is out of the hole. Choke line pressure should be monitored before re-opening the rams. Kelly cocks should be operated daily. Choke and kill valves should be operated daily, and lines pumped through. Choke manifold line-up should be checked each tour.
Where significant casing wear is possible, a ditch magnet should be installed to monitor metal returns. If severe casing wear is suspected, actual wear should be measured by wireline calliper tools and then the casing tested to the minimum acceptable pressure.
6-71
March 1995
Mention here leaks experienced in testing parts used, faulty or missing equipment and remedial action
SIGNATURES
Driller:
Toolpusher:
6-72
March 1995