Académique Documents
Professionnel Documents
Culture Documents
3.1.4 Tutorial 03: Modelling an oil well with black oil PVT matching
File: ~/samples/PROSPER/T03_OilWellPVTMatching.OUT
It is required to:
- Build a PROSPER well model for an oil producing well
- Calibrate the black oil correlations against lab data
- Use the model to estimate the initial flow rate against a well head flowing pressure of
250 psig
- Generate a well performance curve (Rate vs. well head flowing pressure)
- Generate lift curves to be utilised by a simulator (Reveal, GAP, MBAL, Eclipse, VIP,
etc)
3.1.4.2 PVT Input Data
The basic available PVT data are:
Parameter Value
Solution GOR: 400 scf/stb
Oil Gravity: 30 API
Gas gravity: 0.75 (Air =1)
Water salinity: 80000 ppm
Impurities (CO2, N2, H2S): None
Bubble point pressure: 2500 psig at 200 degF
Further PVT data are given in tabular form below at 200 degF:
Pressure in psig Gas Oil Ratio in scf/ Oil FVF in rb/stb Oil Viscosity in
stb centipoises
1500 237 1.138 1.34
2000 324 1.178 1.15
2500 400 1.214 1.01
3000 400 1.207 1.05
4000 400 1.198 1.11
Note: The overall heat transfer (also called U value) is a composite number that
captures the different heat transfer mechanisms taking place: conduction, convection
and radiation. It can be thought of as the inverse of the thermal resistance in the
surroundings of the wellbore.
Fluid: Oil and Water. All other settings can be left to their defaults options.
Parameter Value
Solution GOR: 400 scf/stb
Oil Gravity: 30 API
Gas gravity: 0.75 (Air =1)
Water salinity: 80000 ppm
Impurities (CO2, N2, H2S): None
In order to enter the lab measurements at different pressures, select the | Match Data
button and populate the table as follow:
PVT regression
Now select |Regression | Match All to kick the non-linear regression algorithm. The
following screen will confirm when the regression is completed.
In order to quality-check the regression, select | OK | Parameters and this is what can be
seen:
Please consult the online help to get the details about the regression algorithm.
Ideally, Parameter 1, which is a multiplier, should be close to unity whereas the shift
(Parameter 2) is expected to be around nil.
On the basis of the results obtained here, we will proceed with the combination Glaso /
Petrosky.
For this, select | Done | Done and choose the appropriate black oil correlations as per
screen-shot below:
Now by selecting | Edit, the software will take the user through all the screens necessary
to input the equipment data starting with the deviation survey
3.1.4.5.4.1 Deviation Survey
Here are the data:
Note: It is possible to input or import a large deviation survey table into PROSPER. A
filtering algorithm is implemented that will select up to 20 of the most relevant data
points describing the well trajectory.
For this, the | Filter button located on the top right of this screen is to be used.
Select | Done and the screen labelled Surface Equipment will follow.
3.1.4.5.4.2 Surface Equipment
This model will not include any surface equipment.
Selecting the "Cancel" button located on the top left of the screen will take the user to
the next screen: the down hole equipment description screen.
3.1.4.5.4.3 Down Hole Equipment
The down hole equipment includes the tubings, casings, nipples, Sub-surface Safety
valves and so on.
The deepest entry in the down hole equipment section is the datum depth for the
static reservoir pressure that will be entered in the IPR section.
It is recommended to take the top perforation as the reference depth for the static
reservoir pressure. This implies that the down hole equipment description must stop at
the top of the perforation. This is also true for multiple-zone completion because the
pressure drop between the zones will be taken into account by the appropriate Inflow
model (multi-layer IPR model with dP or multilateral IPR model).
With | Done the entry screen of the Equipment data section will be accessed:
If one then selects | Draw Down hole this is the sketch of the well bore that appears:
To proceed, simply select the Input Data button (red if invalid) in the top right corner of
the screen above:
From the screen above, select the the tab "Mech/Geom Skin" at the bottom left corner
of the screen to enter the mechanical skin:
Mechanical skin: +2
Once the IPR data input is completed, it is a good practice to create an IPR plot to verify
that everything makes sense. For this, simply select | Calculate and the software will
create and display the following plot:
Please select | Main and then save the PROSPER file with | File | Save to secure all
changes made so far.
3.1.4.5.6 Estimation of the well flow rate
The response of a well bore combines fluid properties (PVT), reservoir data (IPR) and
tubing response (VLP).
The fluid properties are used to compute the reservoir response (IPR) and the tubing
response (VLP).
For a given set of boundary conditions (given reservoir pressure and well head flowing
pressure), the well flow rate is the intersection between the IPR curve and the VLP
curve.
In order to compute this intersection point with PROSPER, simply select | Calculation |
System | 3 Variables as per screen-shot below:
After selecting the system calculation, one needs to enter the well head flowing pressure
as per screen-shot below:
The top node pressure is the pressure downstream of the system. Since here there is
no pipeline in this PROSPER model, the top node pressure is the well head flow
pressure.
The parameters Water Cut and GOR were entered earlier on the IPR screen. For the
fluid description, there was also a need to enter the GOR on the PVT screen.
However, during the life of a well either parameter may change. Therefore, this screen
allow the user to sensitize both.
When sensitizing on the water cut and / or the GOR, PROSPER will use the new
sensitivity values for the calculations It will no longer use the GOR and the water cut
values entered on the IPR screen or the GOR from the PVT screen. The same principle
applies to all sensitivity variables specified elsewhere.
Surface Equipment Correlation: This applies to any pipeline in the model. Since this
model does not include any pipeline, the choice of a pipeline correlation is irrelevant.
Vertical Lift Correlation: Petroleum Experts 2.
Solution Node: Bottom Node
Rate Method: Automatic - Linear
Left-Hand Intersection: DisAllow
One can proceed with | Continue | Continue and Calculate with the following results:
One can read an oil rate of 10 500 stb/d as solution rate. The corresponding BHFP
reads: 2584 psig.
If one selects | Plot | System Plot, the system plot (VLP + IPR plot) is displayed as
follow:
Select | Main to get back to the main PROSPER screen and save the file to secure all
changes made so far.
In this example, we will vary the well head flowing pressure from 50 psig to 1200 psig.
For this, simply select | Calculation | System (IPR VLP) | 3 Variables | Continue and here
select the sensitivity variable "First Node Pressure":
One can manually populate the cells with the desired values or generate them
automatically with | Generate. Afterwards, one needs to enter the first value, the last
value, the number of values and the sequencing algorithm:
To complete the process, simply select | Generate and then the input cells are
populated accordingly:
One can see that on this screen up to 3 variables can be sensitized on simultaneously. It
is generally possibly to sensitize on any number of parameters.
With | Continue | Calculate, ten (10) sets of VLP and IPR are computed.
To visualise a plot of the solution rate versus first node pressure, select | Sensitivity.
This is the plot of the Bottom Hole Flowing Pressure versus First Node pressure, the
first node being the well head.
In order to plot the oil flow rate versus the first node pressure, one simply needs to select
| Variables and then | Gas Rate for the Y axis variable:
3.1.4.5.8.1 Introduction
Lift curves express the pressure drop across a tubing for a given set of variables.
Different software providers use different formats and acronyms for lift curves.
For naturally flowing oil wells, the variables involved in lift curve generation are:
- liquid rate
- Well Head flowing pressure (= first node pressure)
- Water Cut and
- GOR
Even though in this case, four (4) variables will change, the lift curves are referred to as
3-Variables VLP. The rate variable is taken for granted as one cannot calculate
pressure drop without flow rate.
The 3 VLP variables for a naturally flowing oil well are: WHFP, GOR & Water Cut.
Recommendation:
In this example, we will demonstrate how to generate lift curves with PROSPER since lift
curve generation is a key task that PROSPER is used to perform. However, if a project
involves the generation of lift curves for multiple wells, a more efficient approach would
involve the use of GAP, the network modelling and optimisation tool developed by
Petroleum Experts.
Here are the reasons why:
- GAP will instruct PROSPER to generate the lift curves using the very same PROSPER
well model created by the user and linked to the well icon in GAP
- GAP uses PROSPER to generate and save lift curve for any number of wells in batch-
mode. This can be done overnight for instance to save time.
- GAP will ask for the right VLP variables input depending upon the well type. The
variable selection for lift curves can sometimes be delicate because different simulators
use different names for the same quantity (example gas lift injection rate versus. GLR
injected).
3.1.4.5.8.2 Steps
In order to generate wells' VLP, one first needs to work out the variables involved. Here
these variables are liquid rate, WHFP, GOR and Water Cut.
The second step is to choose values for the variables in such a manner that the host
application (GAP, MBAL, REVEAL, Eclipse, etc) can reasonably interpolate between
them to find the solution needed. Extrapolation and very large variable spacing are to be
avoided.
In order to enter the data for the lift curve generation , select | Calculation | VLP (Tubing
Curves) | 3 Variables
The next screen shot can be used to enter the range of rates to use:
A few notes:
The values for top node pressure, water cut and GOR will be overwritten through the
sensitivity variables.
One should make sure that the right VLP correlation is selected. Here, the Petroleum
Experts 2 correlation will be used.
The "User Selected" rate method gives control to the user to enter the rates he /she
think will best describe the expected operating ranges of the well.
One should also ensure that the first and last node are correctly selected.
In order to proceed, simply select | Continue, select the relevant variables and populate
the entries as per screen-shot below:
At the end, a message will confirm that the calculation has completed.
It is always a good practice to plot and visualise the lift curves before exporting them.
One can see that the lift curves do all have the right shape (J -curve) and look smooth.
The following screen will appear where one has to select the format for the host
application (here Eclipse):
The process is completed with | Continue | Save and the file is saved and edited
automatically for viewing:
Select | File | Exit | Main and this completes the exercise.
It is required to:
- Estimate the initial oil rate against a well head flowing pressure of 250 psig
- Perform sensitivity runs on the horizontal well length for 500 ft, 1000 ft, 1500 ft, 2000 ft,
2500 ft and 3000 ft.
- Sensitize simultaneously on water cut of 0 %. 10%, 20%, 30%, 40% and 50%.
3.1.5.2 PVT Input Data
Parameter Value
Solution GOR: 400 scf/stb
Oil Gravity: 30 API
Gas gravity: 0.75 (Air =1)
Water salinity: 80000 ppm
Impurities (CO2, N2, H2S): None
Bubble point pressure: 2500 psig at 200 degF
Further PVT data are given in tabular form below at 200 degF:
Pressure in psig Gas Oil Ratio in scf/ Oil FVF in rb/stb Oil Viscosity in
stb centipoises
1500 237 1.138 1.34
2000 324 1.178 1.15
2500 400 1.214 1.01
3000 400 1.207 1.05
4000 400 1.198 1.11