Vous êtes sur la page 1sur 14

SPE/IADC 119746

Case History: Automated Drilling Performance Measurement of Crews and


Drilling Equipment
Ketil Andersen and Per Arild Sjøwall, StatoilHydro; Eric Maidla and Buddy King, Nexen Data Solutions Inc.; and
Gerhard Thonhauser and Philipp Zöllner, TDE Thonhauser Data Engineering GmbH

Copyright 2009, SPE/IADC Drilling Conference and Exhibition

This paper was prepared for presentation at the SPE/IADC Drilling Conference and Exhibition held in Amsterdam, The Netherlands, 17–19 March 2009.

This paper was selected for presentation by an SPE/IADC program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have
not been reviewed by the Society of Petroleum Engineers or the International Association of Drilling Contractors and are subject to correction by the author(s). The material does not
necessarily reflect any position of the Society of Petroleum Engineers or the International Association of Drilling Contractors, its officers, or members. Electronic reproduction, distribution, or
storage of any part of this paper without the written consent of the Society of Petroleum Engineers or the International Association of Drilling Contractors is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE/IADC copyright.

Abstract
This paper describes the development and implementation of a unique new capability to automatically measure the
performance of drilling crews, drilling equipment and downtime. The methodology and supporting technology make it
possible to carry out a detailed comparison of the performance of equipment and crews across a large number of installations.
The system was tested on 8 drilling units (fixed and mobile) in the North Sea.

When the data is analysed and plotted as a histogram (frequency against time taken for each task) the information provided
gives:
• The management a process to select the best supplier based on the performance of its equipment when compared to
the technical limit of that equipment.
• The crews the most consistent and efficient way of working (best practice).
• The management an automated measuring tool to set and follow up the targets for crew performance.
• A tool that makes ‘hidden down time’ visible, so appropriate actions can be taken to improve performance.
• A process to see if operations can be eliminated or considerably reduced in time (reaming, circulating, etc).
When it comes to crew handling of the equipment the results show a performance improvement of 30%. There are potential
time savings of between 40 to 60% in some individual drilling tasks such as slip to slip connection times and weight to
weight connection times. This is achieved by performing each operation in a more efficient and consistent way. Optimizing a
number of identified KPIs for drilling, including tripping and casing running times, shows a potential saving of between 8 to
15% of the total well construction time. Automated Drilling Performance Measurement (ADPM) has been proven in over 30
wells on 8 rigs in the North Sea.

This paper also describes the implementation strategy and approach including the elements of the change management and
training that are required to implement this new drilling performance process.

Introduction
One of StatoilHydro’s objectives is to increase the number of wells delivered into production each year. It is the company’s
aim to reverse the dropping trend, shown in Figure 1, where drilling performance measured in meters per day for European
wells, as reported by Rushmore, has dropped in the last 3-4 years. Our goal for the future must be to exceed the best
performance delivered in any previous year.
2 SPE/IADC 119746

Figure 1: European trend of average meters drilled per day based on Rushmore data.

One element in achieving this target is to find the hidden down time. Use of an Automated Drilling Performance
Measurement (ADPM) system to measure the detailed performance of the drilling process is a step change in our ability to
efficiently identify hidden down time. Figure 2 is showing the time efficiency break down definitions used in this paper.

Figure 2: An Example of a Weight to Weight Time per Connection for a sample well showing hidden down time. This KPI can be
used to target improvement opportunities.

Reduction in non-productive time has been studied in great detail by Bond, et al.[1], van Oort, et al. [2], Thonhauser, et al, [3]
and Thonhauser [4]. The difference between those studies and this paper being that an ADPM system has been used. New
visual aids have also been developed that represent the performance of equipment and crews. This system also provides base
line data, in a structured manner that allows rig crew and management to improve the overall team efficiency.
SPE/IADC 119746 3

The method presented here involves all personnel in the well construction process including the drilling contractor, service
companies and operator personnel. The process uses an objective measurement tool to measure operations in an unbiased
manner.

Automation is necessary to measure detailed performance for ‘routine’ drilling operations because it would be very difficult,
if not impossible, to manually measure these individual operations in a consistent way (e.g. slip to slip connections during
tripping operations, etc). In ‘Non-routine’ operations, in particular the ones leading to downtime time, data and information
are generally well described in the morning report (e.g. different equipment breakdown by contractors and service companies,
waiting on weather, etc). This is further illustrated in Figure 3 and Figure 4 below.

The following definitions have been used throughout this paper.


• A “Slip to Slip Connection” represents the time in slips during tripping operations limited to drill pipe connections
only (i.e. tripping of BHA components etc. is not considered). The KPI starts when the drill string is put in slips and
ends when it comes out of slips.
• A “Weight to Weight Connection” represents the time between two drilled stands. This KPI starts when the drill
string is lifted off bottom and lasts until the string is on bottom again. This operation includes all wellbore
conditioning as well as the drilling slip to slip connection conducted during this interval.

Figure 3: Automated Drilling Performance Measurement (ADPM) detection example for Well A.
4 SPE/IADC 119746

Figure 4: Comparison of a 24 hour period between the ADPM system and the Daily Morning Report. There is a large discrepancy
between the two reports (duration values in hours rounded).

It can be seen in Figure 4 that the morning report shows a 13h30min drilling with no detail given for any wellbore
conditioning time. By using the ADPM system it is possible to identify that only 6h18mins were actually used for making
hole and that 13h41mins were used on wellbore treatment time (i.e. reaming and washing upwards/downwards, and
circulating.)

Implementation Approach

Figure 5 shows our approach to the implementation of the Automated Drilling Performance Measurement System from data
measurement through to decision support and the implementation of any actions that may result. Not only is it necessary to
ensure that the technology is robust and accurate, it is just as important to ensure that all the personnel involved have been
adequately trained and have trust in the system and the results that it gives. Relationships between the teams involved must
be open and honest to allow this system to be successfully implemented.

Figure 5: Some of the prerequisites for improvement.

As outlined in Figure 5 the following pre-requisites have to be in place:

• The appropriate levels of instrumentation and sensors on the rig equipment


• IT infrastructure, databases and bandwidth to store and transmit the data and information
SPE/IADC 119746 5

• Real-time processing and analysis combined with visualisation and/or pattern recognition
• Clearly defined work tasks
• A process to continuously improve and innovate
• The appropriate level of training and competency for the teams involved
• A high level of collaboration and communication both between rigs and between the rigs and the support offices
• Visibility of performance data to all relevant parties

In order to overcome existing limitations, as outlined in Figure 6, a new way of working has to be developed as part of the
implementation strategy.
New Way of Working

Current Situation Goal


Difficult to measure efficiency
between man and machine Automated measurements

Major time variations in


Minor time variations
identical tasks

Major variations in execution


Standardized task execution
of identical tasks

No detailed reference Detailed reference measure-


measurements (once daily) ments with other teams and eq.

Figure 6: The desired new way of working

The ADPM system makes visible information and data in areas of the operation where it was previously not possible to get
the level of detail needed for the performance to be improved. Once it is possible to break down the operation into component
parts, it is possible to focus on the objective of reducing the non-productive time.

Figure 2 illustrated the opportunity to improve the “Weight to Weight Connection” key performance indicator (drilling
connection KPI). Figure 7 shows an example of a process to achieve the desired improvement in this KPI. What is important
here is that the ADPM is able to identify the hidden down time that was previously invisible. The ADPM makes it possible to
target and remove the hidden down time and significantly improve performance.

Figure 7: Process followed to enhance the drilling performance.


6 SPE/IADC 119746

Automated performance measurement is vital to identifying improvement areas in drilling operational tasks. These automatic
measurements can also be used to indentify best practices leading to the development of standardized procedures. A standard
is defined with the involvement of the whole team and needs to compensate for the differences in equipment that individual
rigs may have. The standard must be realistic, achievable and have the buy in from all teams involved. The objective here is
to agree on a safe and consistent procedure that can be followed given the appropriate level of training and competency. The
objective is not only to reduce time taken but to improve the operational efficiency and safety performance by the consistent
use of standardized tasks. The identification and removal of unnecessary tasks is also carried out.

Implementation Strategy

The ADPM system was implemented in StatoilHydro over a period of a few years. The first part, which was started in
February 2004, was to implement a database system capable of transferring data utilizing the WITSML protocol. This took
place over 2 ½ years.
The second part was a trial phase to check for the algorithm accuracy, data frequency issues, and implementation issues.
Three wells on three rigs were selected to be analyzed between December 2006 and March 2007. Among the main findings
were:

1. Accuracy of the automatic operations detection:


1.1. Data Quality Check
Morning Reports and the digital log plots were checked against the ADPM outputs and the accuracy and
precision were confirmed.
2. Data Issues:
2.1. Time Synchronization
Comparing the real-time data to the daily reports from the DBR system (StatoilHydro’s Daily Drilling Report)
showed a time-shift of one hour. This didn’t pose a problem to the ADPM system made it hard to compare
both reports against each other – this was corrected.
2.2. Bit Depth Oscillations
One well contained very strange bit depth oscillations – this was flagged to the data provider and then fixed.
2.3. Data Channel Spikes
Two wells contained very strange spikes on some of their data channels. This was also flagged to the data
provider who solved the problem.
2.4. Data Frequency Issues
Some data points from two wells were closer to 30 second time frequency rather than 10 second or less. This
made it hard to detect certain KPI’s e.g. Slip to slip connections. This was corrected by working with the data
provider who then started providing a consistent 5 second data sampling rate.
2.5. WITSML Server Up-Time
This was a continuous problem that was fixed during the trial.
2.6. WITSML Server Data Scramble
Some drilling parameters in the database contained data from two different providers that were mixed into the
same channel. It was impossible to unscramble since the data frequency from each channel was not constant.
The solution was to go straight to the original data from one single provider.
3. Usability of the Information:
The ADPM daily tracking report was shown to complement the daily morning reports produced by the
company’s representatives on the rig. Key Performance Indicators were then established against which the
crew’s performance was measured.

Following on from this trial StatoilHydro defined a number of KPI’s that could be measured using the ADPM system. In
addition to this, a set of visualizations techniques were developed to analyze the large amounts of data collected; these
included the histogram plots for different time intervals, daily, weekly etc. An example of these KPI’s is given in the
summary table (Fig 15).

A second trial phase was carried out to check upon the scalability of the system. Four new rigs were chosen for the trial
which took place between April and June 2007. The scalability issues were addressed, however, some of the data issues
from the first trial showed up again. This lead the group to believe that there was a need for continuity between trials,
without any interruption, otherwise the same systematic recurring issues would continue to arise.

A third phase was to implement the Automated Drilling Performance Measurement System on a continuous basis on 5 rigs.
This started in January 2008 and required a greater amount of involvement from all the operating groups. A significant
SPE/IADC 119746 7

amount of effort was required to ensure the appropriate amount of team engagement to make the implementation of the
system a success.

Improvement Process
The key principal underlying the ADPM process is that ‘if we can measure an operation, we can improve it’. However, if we
cannot measure it, we don’t know if we can improve it or not.

The ADPM process creates a significant amount of digital data (eight drilling functions measured every 1 to 5 seconds). This
is an enormous amount of data to be processed. It also needs to be presented in an easy and straightforward manner and to be
able to convey a quick and precise message to those managing the operations.

Histogram plots were selected as the preferred visual aid as they can summarize large quantities of data and are easy to
interpret. These plots were used to analyze any KPI at any level (rig performance, individual crew performance, aggregated
group performance by hole size, etc.).

An example of how the histograms are used is illustrated in Figure 8. In this particular example a process for selecting a best
practice is outlined that takes safety and efficiency into consideration. Very short task times can be attributed to bad
measurement, data imperfection, or unsafe practices which a simple investigation should clarify. The idea here is to define a
best practice that is safe, achievable and measurable. It is essential that any best practice has the buy in from all the crews
involved if it is to be implemented successfully.

• Step 1: Define Current


Best Practice State
• Step 2: Identify
Exceptions
– Exceptional Short –
Maybe taking Shortcuts
– Exceptional Long –
Operational Problems
• Step 3: Define Best
Practice as Target for
Future Operations

Figure 8: Using histograms to select best practice.

Once agreement has been reached on the best practice target value, the histograms will help to show whether the best practice
is actually met, and if not, what the amount of deviation from the target value is. This is further illustrated in Figure 9.
8 SPE/IADC 119746

Figure 9: Using histograms to verify the consistency around the best practice target.

The best practice should be a realistically achievable target by properly trained crews performing routine operations under
good conditions. It should also be noted that the target should not be a fast moving target. The idea here is to take small steps
towards a sound and safe practice and only revise the best practice after sufficient experience has been gained at the current
performance level (a year or so). The primary aim is to have a consistent delivery of the best practice rather than pushing the
technical limit.

As will be demonstrated later in the case history, the fact that routine operations can be done consistently and close to best
practices will yield significant savings in safety, time and money.

• Further Improvements by
implementing new
technology to lower
Average Operations
Duration
• Team Effort to define
new Target

Figure 10: Using histograms to select a new best practice target.

The final step in this process it to make sure that the measurement and performance gains are sustainable into the future even
after all targets are met.

Implementation Example
Figure 11 and Figure 12 are examples of the starting data sets that can be used to establish a best practice for a drilling
“weight to weight” connection. All the drilling connection task times were measured and the best times for each of the tasks
were used to define a ‘best composite time’ for the entire operation. This is shown in Figure 12.

From the analysis it was determined that a certain amount of reaming and washing was necessary but that each crew was
making different decisions on how much reaming and washing was appropriate. How much time is necessary for this
SPE/IADC 119746 9

operation is a judgment call which should come from torque and drag plots supplied by engineering. The driller should
follow the best practice and only make changes from it if he observers deviating trends from established engineering plots.
Kucs, et al. [5] and Freithofnig, et al. [6].

A standard procedure for the connection task can be defined with the goal to have only small variations around the
implementation of standard procedure in operation. Figure 11 shows the comparison of a “fast” and a “slow” weight to
weight connection. As can be seen the ‘fast’ connection took just over half the time of the ‘slow’ connection.

14.9 min

26.4 min

Figure 11: Overlay of “Fast” and “Slow” Weight to Weight Connections as part of the analysis.

In order to analyze weight to weight connection the entire task was broken down into smaller steps. These are:

• Reaming up
• Reaming down
• Waiting on connection, which represents the hidden down time from stopping reaming and washing until starting the
slip to slip connection
• Slip to slip connection
• Waiting on drilling, which is also hidden down time and should be eliminated

Figure 12 shows this operation broken down into the smaller steps as a standardized best practice. This process of breaking a
task down into smaller steps is key to the overall performance improvement approach and allows the comparison of different
work steps to identify potential areas for improvement.
10 SPE/IADC 119746

9.2 min

Drilling Connection
(Weight to Weight Connection)

Figure 12: Example of a best practice Weight to Weight Connection

If there is little or no drag one should not ream the stand as is shown in Figure 13. Additional time of 0.5min is invested to
check whether up and down weight is consistent with the theoretical drag values. By analyzing the plot it can also be seen
that time can be saved by pumping up the MWD survey while drilling. In this example, it is possible to save 5.3 minutes for
each connection using this standardized procedure:

Saved reaming time -3.2min


Saved time waiting on survey -2.6min
Additional time to take weight readings +0.5min
Total savings -5.3min

Table 1: “Weight to Weight Connection” time saving potential

3.9 min

Figure 13: Theoretical best composite time ([7] A.W. Iyoho etal.) for a “Weight to Weight Connection”

Reducing the hidden non-productive time has lead to significant benefits in safer task handling and a reduction in wellbore
aging (formation exposure to drilling fluids chemicals and pressure). This is expected to lead to less formation related
problems helping achieve earlier production at lower costs. All this leads to the main target of drilling more wells in a
calendar year.
SPE/IADC 119746 11

Case Study
It is not only the measurement of the data that is important but the manner in which this information is provided to the rig
crews, service people and operator personnel. All efforts must be made to ensure that the people involved in the
implementation of this process see it as a value adding approach to achieve the optimal performance targets for the drilling of
the well. Having an open and honest culture between the teams and ensuring that all the parties are involved is a key success
factor. All teams involved must feel that they can not only challenge the process but are recognised for their part
implementing the continuous improvement process. With this in mind, the case studies will cover a variety of KPI
measurements and define a summary of the potential savings.

The first example, Figure 14, comes from four rigs performing the same operation under similar conditions. The KPI
measured is the “Slip to Slip” time while tripping pipe. What can be seen in this plot is:
• Both Rigs A and B manage to make tripping connections consistently around 2 minutes as is shown by their 50%
value.
• The Rigs C and D perform the same operation consistently at 1.5 and 1.3 minutes respectively, this is shown by their
corresponding 50% values.
• When looking at the high performance end of the spectrum it is possible to see that Rig C and D manage to perform
this operation below 1 minute for more than 2600 repetitions.

Rig A Rig B

Rig C Rig D

Figure 14: Example of comparison of 4 offshore rigs in the North Sea Area

The observations in Figure allow the following conclusions to be drawn:


• The lower end of the data spread reveals the technical limit for the rig. Based on this observation it is possible to
start a detailed operational analysis breaking down the slip to slip connection into its component parts to investigate
the equipment performance:
12 SPE/IADC 119746

• Iron roughneck performance


• Hosting system performance
• Racking system performance, etc.
• This type of equipment performance capability analysis is then performed with the drilling contractor and the crews
to find the improvement potential.
• Comparison can also be made between different equipment suppliers. Ultimately this information could be used as
part of an equipment procurement strategy.

Figure 15 illustrates another example of making tripping connections with 4 crews on the same rig. Each crew has performed
this task a large amount of times. In order to further analyse this data we have also included a table that summarizes some of
the data, for instance, Crew B has performed this operation 879 times compared to Crew A that has carried it out 1886 times.
The average values contained in the table are important as these allow us to see if there is a learning curve (or improvement)
over time. The best practice time was set at 1.17 minutes based on Crew D’s performance.

The cumulative time to the right of the target value is called the ‘savings potential’. This value is an indication of the savings
that can be realized once the crew starts improving and moves towards the best practice time for an operation. In addition to
the savings potential the histogram plots also show the scatter of the data. This gives an indication of the consistency with
which the crew is performing an operation. This may also highlight the need for further training.

Figure 15: Comparison of four crews on the same rig carrying out a slip to slip connection.

Many other KPI’s may be indentified and analysed in this way. By adding these performance opportunities together an
overall savings potential can be calculated. A summary of these findings for one well is shown in Figure 16. The table is self
SPE/IADC 119746 13

explanatory and points out that just by focusing on routine operations there is a potential saving of 7.6 days for the 61 day
well. This further confirms the value added of ADPM as a means to achieve the company’s goal of increasing the number of
producing wells delivered per year.

Slip to Slip Time Moving Time


KPI Weight to Weight Drilling Stand*

Tripping Casing/Liner Tripping Casing/Liner

Phase All 16“ 12 ¼“ 8 ½“ All 16“ 12 ¼“ 8 ½“ 16“ 12 ¼“ 8 ½“ 16“ 12 ¼“ 8 ½“

Lower Cut-Off [min] 0.5 0.5 0.5 0.5 0.25 0.5 0.5 0.5 10 10 10 0 0 0

Upper Cut-Off [min] 10 30 30 30 10 30 30 30 90 90 90 300 300 300

Benchmark [min] 1.15 3.86 1.82 5.12 0.55 1.57 1.04 1.03 25.32 12.78 19.92 42.07 31.10 91.71

Savings Potential [h] 60.1 5.7 14.9 1.4 36.7 5.6 6.3 0.9 23.0 10.1 9.1 55.2* 12.2* 21.8*

Savings Potential [%] 60.73 37.06 71.91 37.68 66.98 59.49 63.20 68.73 37.66 42.25 35.05 47.94* 25.83* 22.26*

* only 10% are included for


Time analysed by proNova 61 days total savings potential
Total Savings Potential (incl. 10% of Drilling Stand Time) 7.6 days

Figure 16: Example of savings potential for a North Sea Area offshore well.

Figure 17 shows the performance improvement for the weight to weight connection time KPI for 8 StatoilHydro wells. There
has been an improvement of 28% for this KPI as a result of the use of the Automated Drilling Performance Measurement
System.

Figure 17: Performance improvement of 28% for the weight to weight connection time has been seen following analysis on 27 wells
on 8 rigs.
14 SPE/IADC 119746

Conclusions
Based on the findings of this work the following conclusions can be made:

• A new approach to improving the drilling process based on rigorous measurement of drilling performance has been
developed and implemented.
• Breaking down an operational task into its component parts and measuring the performance of each part is key to the
process. Automation is necessary to measure detailed performance for ‘routine’ drilling operations and to identify
the ‘hidden downtime’.
• Simple visual aids that turn data into information and enhance the decision making process are vital to getting the
system used in the field.
• An approach the looks at the entire process from data measurement through to decision support and includes team
engagement is essential to having a successful implementation.
• ADPM can also be used to indentify best practices leading to the development of standardized procedures.
• This approach can track the performance of different crews and normalise their performance for the different
equipment they are using.
• ADPM has been proven to work in over 30 wells across 8 different rigs in the North Sea showing that the system is
scalable.
• This approach can identify tasks that can either be eliminated or considerably reduced in time.
• This study has shown a performance improvement of 30% for the crew handling of the equipment in specific tasks.
• This approach has identified a potential of 40-60% in the time taken for some specific drilling tasks.
• Optimizing a number of identified KPIs for the drilling operation has shown that there is a potential saving of
between 8 to 15% of the total well construction time by using this approach.
• ADPM provides management with visibility of the performance of crews and equipment across the entire rig fleet
and allows targets to be set for crew and equipment performance.
• Performance data and information provided by ADPM can be used to support a drilling equipment procurement
strategy.
• The use of Automated Drilling Performance Measurement can make a step change in is drilling performance and
reverse the current downward performance trend for meters drilled per day.
• The capability that ADPM provides has the potential to form the basis for closed loop automated control for drilling
operations.

Acknowledgements
The authors would like to thank StatoilHydro, Nexen Data Solutions Inc. and TDE Thonhauser Data Engineering GmbH for
the permission to publish this paper. Special thanks go to the proNova team and all who have been involved defining and
implementing the vision of a new way of working to improve drilling performance.

References

1. Bond, D.F., Scott, P.W., Page P.E., and Windham, T.M., “Applying Technical Limit Methodology for Step Change in
Understanding and Performance”, SPE/IADC 35077 presented at the 1996 SPE/IADC Drilling Conference, New Orleans (USA),
March 1996
2. vanOort, E., Taylor, E., Thonhauser, G., Maidla, E., “Real-time rig-activity detection allows to indentify and minimize invisible
lost time”, World Oil April 2008, Page 39ff
3. Thonhauser, G., Wallnoefer, G., Mathis, W., “Use of Real-Time Rig Sensor Data to Improve Daily Drilling Reporting,
Benchmarking and Planning - A Case Study”, SPE 99880 presented at the 2006 SPE Intelligent Energy Conference, Amsterdam
(The Netherlands), April 2006
4. Thonhauser G., ”Using Real-time Data for Automated Drilling Performance Analysis”, European Oil and Gas Magazine 4 2004,
Page 170ff
5. Kucs R., Hermann F. Spörker H.F., Thonhauser G. “Automated Real Time Hookload and Torque Monitoring”, IADC/SPE
112565 presented at the 2008 IADC/SPE Drilling Conference held in Orlando, Florida, U.S.A., 4–6 March 2008
6. Freithofnig, H.J., Spoerker, H.F., Thonhauser, G., “Analysis of Hook Load Data to Optimize Ream and Wash Operations”, SPE
85308 presented the SPE/IADC Middle East Drilling Technology Conference & Exhibition held in Abu Dhabi, UAE, 20-22
October 2003
7. A.W. Iyoho, SPE, K.K. Millheim, SPE, B.K. Virginillo, “Methodology and Benefits of a Drilling Analysis Paradigm”,
IADC/SPE 87121, presented at the IADC/SPE Drilling Conference held in Dallas, Texas, U.S.A., 2-4 March 2004.

Vous aimerez peut-être aussi