SPE 116519

Intelligent Fields Management at Woodside: A Low-Cost Step Improvement

in Field Management Using Off the Shelf Technology
Bruce James, Kevin A. Kerr, Stephanie Lim, Ed Lewandowski, Craig Knight, and Richard Bell,
Woodside Energy Limited

Copyright 2008, Society of Petroleum Engineers

This paper was prepared for presentation at the 2008 SPE Asia Pacific Oil & Gas Conference and Exhibition held in Perth, Australia, 20–22 October 2008.

This paper was selected for presentation by an SPE 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 and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, 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 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 copyright.

Abstract

Using improved integration, data efficiency and more informed decision making, Woodside’s Intelligent Fields Management
(IFM) has created a step change improvement in the way the Cossack-Wanaea-Lambert-Hermes (CWLH) fields are managed.
There is tremendous value in IFM in numerous and often unexpected places. However, the greatest value likely lies within the
improved integration of data and work practices and longer-term reservoir management. The benefit-cost ratio of IFM on the
CWLH fields is likely in the order of ten to one. This paper demonstrates what a mid-sized oil and gas company can
accomplish within 12 months and at relatively low-cost using off-the-shelf technology and starting with little pre-existing
knowledge.

CWLH is four oil fields with 11 subsea wells tied back to a floating production storage offtake (FPSO) vessel located offshore
on the North West Shelf of Australia. The initial focus of Woodside’s Intelligent Field Management program is on near-term
field management issues, including reducing well testing, daily optimisation, production allocation, and managing sand
erosion. Woodside’s IFM program is currently being implemented on two other oil assets, with an expansion to gas and
longer-term reservoir management planned for later in 2008.

The Petroleum Experts’ Integrated Field Manager™ is the calculation engine used in the IFM program, allowing for custom
workflows to be automated and scheduled. For example, workflows have been created that can run every 30 minutes and
automatically identify elements of an integrated asset model, where measured behaviour significantly diverges from calculated
behaviour. Powerful visualisation for interpreting data, particularly across disciplines, was critical for success and was
realised through ISS Group’s Babelfish™ web based viewer.

Successful implementation relies on a highly motivated interdisciplinary team, dedicated IT and management support and
close interaction with asset teams. Long-term success will require ongoing IT and super-user support.

Introduction

This paper presents Woodside Energy Ltd’s initial experiences in implementing its Intelligent Fields Management (IFM)
program. The program’s overall goal is to improve the management of oil and gas fields through better integration of
workflows and data, with an emphasis on real time production data and downhole flow control. This paper focuses on
improving the integration of work processes and information.

IFM is part of an industry-wide revolution in work practices using real time data, similar to efforts at other companies such as
Shell, Chevron, and BP. IFM type ideas have been applied in many ways. A common area of application is in production
optimisation (Stenhouse: 2008, Madray et al: 2008 and Oberwinkler et al: 2005), and allocation and surveillance using data
driven neural networks (Poulisse et al., 2006). Other applications include 4-D seismic (Calvert. and Bakulin, 2007), closed
loop management (van Dijk and Goh, 2008) and real time collaboration centres (Hickman et al., 2008).
2 SPE 116519

This case study describes how Woodside’s IFM program was implemented and demonstrates what a mid-sized oil and gas
company can accomplish in 12 months at relatively low cost starting with little pre-existing knowledge.

The time is ripe for implementing IFM-like-programs throughout the Exploration and Production industry for three reasons:

1. field management is becoming increasingly difficult, as new field sizes are shrinking and complexity and water
depths are increasing;
2. high quality production data is more readily available with technological improvements, particularly real-time
production data from permanent downhole pressure gauges. This data is highly valuable, but is often poorly used or
managed.
3. IFM-like-programs are easily technically feasible at a reasonable cost. Technology includes integrated asset
modelling packages, powerful data visualisation packages and software such as the Petroleum Experts’ Integrated
Field Manger™ that allow automation and customisation of workflows.

The current IFM program was initially focused on near-term management issues, such as well test deferral and daily
optimisation on the CWLH field. It is currently proceeding on two other oil fields. Initial efforts are also underway to expand
the program to gas fields and to other parts of the life cycle, such as reservoir management.

The remainder of the paper will provide an overview of the CWLH fields and discuss how the IFM program was implemented,
including developing custom workflows. Finally, results and learnings of the program will be presented.

The Cossack-Wanaea-Lambert-Hermes Fields

The Cossack-Wanaea-Lambert-Hermes Fields (CWLH) oil fields are located offshore Western Australia, 125-145km north-
west of Karratha (Figure 1). The Lambert and Hermes fields are situated 15km to the north of the Wanaea and Cossack fields.
The fields lie on the inner continental shelf in water depths of 75-135m.

Figure 1 - Location map and layout of the Cossack-Wanaea-Lambert-Hermes fields

The fields are being produced through the Cossack Pioneer Floating Production Storage Offtake (FPSO) facility using a
subsea system of 11 wells tied back via flow-lines. Production from the Wanaea and Cossack fields began in November 1995.
The Hermes field came on stream in October 1997, with production from Lambert-3 tied back to the Cossack Pioneer via a 16
km flow-line. The Lambert field began production in September 1999.

The CWLH fields are operated by Woodside Energy Ltd on behalf of the CWLH Joint Venture comprising Woodside Energy
Ltd (33.3%): BHP Petroleum (16.7%), BP (16.7%), Chevron (16.7%), and Japan Australia LNG (16.7%).

Well testing is conducted by diverting individual wells into the dedicated test separator. The test separator measures oil,
water, and gas flow rates.
SPE 116519 3

The wells have been excellent producers, with high recovery due to high reservoir permeabilities, excellent reservoir
continuity and strong aquifer drive. The fluid quality is also excellent with API gravities in the range of 47 to 48 degrees, a
viscosity ranging from 0.17 to 0.5 cP, and gas-oil-ratios ranges from 140 to1300 scf/stb. All four fields are under saturated.

Current Challenges
With increasing water cut and currently limited gas lift capacity, the wells are becoming more difficult to operate. At
present, three wells receive continuous gas lift to remain on line, whilst Lambert-4 and Lambert-7 remain shut-in awaiting
sufficient gas lift capacity. Gas lift capacity is currently at only 10 MMscf/d; however, this is planned to be increased in the
near future.

The pressure gauges within the CWLH development are of variable quality. All tubing head pressure gauges are
operational, and considered to be accurate. In contrast, there are only four wells with operational downhole pressure gauges.
There are known inaccuracies in total facility gas metering which is currently being resolved.

Despite issues surrounding data acquisition, CWLH represented an excellent candidate for implementing IFM. The
significant production deferral associated with well testing provided a strong business case, whilst increasing water cuts and
gas lift is making the field more difficult to optimise. Ultimately, the absence of quality downhole data was problematic, but
not critical to the success of IFM.

WELL RATE ESTIMATION AND WELL TESTING

Accurate and reliable well rate estimation forms the foundation of many of IFM’s processes, including well allocation,
daily production optimisation, flow assurance management, and reservoir management.

A common well rate estimation method is to use a well performance curve that estimates the well rate for a given measured
tubing head pressures. Another method is the wellhead method. It relies on finding the intersection point between the vertical
lift performance (VLP) curve and the inflow performance relationship (IPR) (Figure 2). The VLP governs the well rate up the
tubing while the IPR governs the rate within the reservoir. The wellhead rate is where the VLP and IPR cross, or where the
VLP and IPR rates are equal.

The well rate depends upon a number of factors including fluid PVT, well geometry and tubing roughness. However, for
an oil well, the estimated rate is typically dependent on four key well parameters that can change with time:

1. the average pressure in the reservoir around the well (Pres);

2. water cut or percentage of the liquid stream that is water;
3. gas-oil-ratio (GOR) or volume of gas produced per unit volume of oil produced; and
4. productivity index (PI), which represents the change of flow into a well from the reservoir for a given change in
pressure.

The typical approach for estimating these well parameters is well testing. However, there are a number of problems with
well testing. Firstly, a single rate test will measure GOR and water cut, but may only provide information on the combination
of Pres or PI.

Secondly, obtaining accurate well test information is problematic on the CWLH field. The physical layout of production
flowlines, manifolds and risers of the subsea infrastructure introduces a number of practical issues such as variations between
operating and testing conditions, and potential cross contamination within flowlines.
4 SPE 116519

Bottom Hole Flowing Pressure

VLP

IPR

Flow Rate
Figure 2 - Well performance plot showing inflow performance relationship and vertical lift performance curves

Well testing on the CWLH field causes production deferral and is a resource intensive activity. A multi-rate test may be
conducted to estimate Pres and PI. However, due to a substantially greater production deferment, multi-rate tests are rarely
conducted.

One of the goals of IFM is to quickly and accurately estimate these four parameters and use them in well models with
measured tubing head and down hole gauge pressures to estimate well rates throughout the month.

Methodology

The technological backbone of IFM is the Petroleum Expert’s (Petex) Integrated Field Manager™ and ISS Group’s
Babelfish™. Real time field data flows from the field to a Honeywell Process Historian Database™ (PHD) production
historian. Data then flows from the PHD database to the Petex Integrated Field Manager™, which acts as the calculation
engine. Currently, calculated results are written back to the PHD database, but other options are currently being considered.
Visualisation of information is carried out by ISS’s Babelfish™. Babelfish™ is a web based visualisation package that allows
easy viewing of a wide range of data by employees across Woodside. It can automatically be linked to any data in the PHD
database, enabling easily accessible, powerful visualisation across multi-disciplinary teams - a critical success factor for the
project.

Petroleum Experts Integrated Field Manager™

The Petroleum Expert’s (Petex) Integrated Field Manager™ is an umbrella program for controlling Gap (network), Prosper
(wells) and MBal (material balance) integrated asset models. It is powerful because:

1. Workflows can be written in C# to automate and schedule tasks. This allows many repetitive or complicated
tasks to be automated and scheduled. Ordinarily these tasks would be poor utilisation of engineering resources
and would not be done.
2. Managing data is automated. Results from all calculations may be saved to a database, such as PHD, for safe
storage and visualisation.

Any program linked in an integrated asset model through Petex Resolve™, such as a HYSYS™ plant model or an Eclipse
reservoir simulation, can be automatically linked to Petex Integrated Field Manager™. In this way, Petex Integrated Field
Manager™ can be used to control very large and complex integrated models of gas and oil fields.

Another very useful feature of the Petex Integrated Field Manager™ is Model Catalogue, which is used to archive versions
of well and network models. As a model is updated, reasons for changes are recorded and the most current model is always
identified. This is particularly useful for managing the large number and versions of well, network, and subsurface models
that are typically developed for a large oil or gas field. Without effective management, these models are quickly scattered
among a myriad of network drives.
SPE 116519 5

Well Rate Methods

Petex Integrated Field Manager™ also has built in automated workflows. One of the most valuable is well rate estimation
that uses a number of different methods to estimate well rates, including: the wellhead, the VLP, the IPR, and the choke rates
(Figure 3). The wellhead method (discussed above) will calculate rates using the tubing head pressures and reservoir pressure
as boundary conditions. The VLP method calculates the rate using the VLP curve and the tubing head and downhole gauge
pressures. The IPR method uses the IPR curve and downhole gauge pressure and estimated reservoir pressure. The choke
method uses the pressure drop across a choke.

P&T

VLP
WellHead
P&T
IPR

Reservoir
Pressure

Figure 3 - Well rate estimation methods

These methods can be used to calculate different well rates, at regular periods. Currently well rates are being estimated
every 30 minutes on CWLH. Comparing the different well rates methods is powerful for detecting subtle changes in well
conditions. If all well parameters are tuned properly then the wellhead, VLP and IPR methods should calculate the same rate.
However, as conditions change, the different rate methods diverge.

As the different calculation methods react differently to changes in well conditions, clues may be provided by analysing the
nature, and magnitude of rate divergences. For example, the choke rate is sensitive to changes in friction, compared to the
VLP, which is sensitive to changes in density. The IPR is sensitive to changes in PI or Pres. While the choke method may not
necessarily give accurate estimates, analysing its trend over time may be useful.

As an example, should the GOR of a well increase, the density in the tubing will decrease, causing the downhole gauge
pressure to decrease and the pressure drop across the tubing to decrease. The VLP rate will decrease because of the measured
pressure drop (Figure 4). The decrease in the downhole gauge pressure will cause an increase in the IPR rate from the
increased pressure drop assumed across the reservoir.

In contrast, an increase in water cut will increase the density in the tubing which will increase the downhole gauge
pressures and the pressure drop across the tubing. This will cause the VLP rate to increase and the IPR rate to decrease
(Figure 4).

An increase in GOR will increase the friction through the choke causing an increased pressure drop. The choke method
will register an increase in rates from the increased choke pressure drop. The VLP rate as discussed above will move in the
opposite direction to the choke rate showing a decreased rate.

While these changes give clues to the changes that are occurring, they do not necessary indicate what the changes are.

Visualisation
A key goal of IFM was to represent data visually, making it easily accessible and able to be analysed across all disciplines
in the Company. ISS Group’s Babelfish™ was used to simultaneously visualise both calculated data from Petex Integrated
Field Manager™ and measured facility data. It is a web based visualisation tool enabling creation of customised web pages
and data trends.
6 SPE 116519

Increase in GOR Increase in Water-cut

VLP
IPR
Well Head
Oil Rate

Oil Rate
Well Head
VLP IPR

Figure 4 - Impact of increasing GOR and water cut on different well rate

A number of different customised pages were created to allow data to be viewed by interdisciplinary teams. These
included system dashboards to easily monitor how the system is operating. A well summary page is used to monitor estimates
by all well rate methods for each well.

Other visual elements included:

• Well page - summarising various well parameters and historical trends. This also shows current valve positions (Figure
5). The page has been used several times by engineers to rapidly troubleshoot problems.
• Subsea summary page - monitors the subsea system by comparing measured pressures and temperatures with
calculated pressures and temperatures (Figure 6). It is also used to monitor virtual temperatures at a location like the
riser base where there is no temperature gauge. Knowledge of the temperature history is valuable to manage the riser
liner life, which to replace is a significant commercial decision.
• Reservoir engineering page (Figure 7) - allows users to easily access well parameter estimates made when matching
well models. The page also allows potential changes in Pres, water cut, GOR and PI to be made in context of other
wells within the reservoir picture, rather than being ‘isolated’ to matching within the well only.

Figure 5 - Well monitoring page with valve symbols and related historical trend data
SPE 116519 7

Each one of these visual pages also provides a one-stop-shop for critical reservoir-related documentation, with live links to
documents within the corporate data base (i.e. PVT Reports). This ensures that documents are readily available to all
disciplines, and that the latest and correct document is available.

Figure 6 - Subsea summary page for comparing measured and calculated pressures and temperatures
as well as virtual temperature monitoring

WANAEA-8

CO

WANAEA-3

WANAEA-5 WANAEA-6

WANAEA-1

WANAEA-10st1WANAEA-7ST1

WANAEA-2

WANAEA-9HST1

WANAEA-4

Figure 7 - Reservoir summary page (edited)

8 SPE 116519

Custom Workflows
A number of custom workflows have been created that allow routine and/or complex tasks to be automated. These are
workflows that are written in C# and can be scheduled on a regular basis or executed on demand. These workflows are
extremely powerful for improving monitoring, surveillance, and optimization.

Gas Lift
The gas lift workflow calculates gas lift rates every 30 minutes using the measured pressure drop across the gas lift chokes.
This work flow was extremely valuable as it allowed individual well gas lift rates to be estimated, which, due to minimal or
inaccurate metering, would otherwise be unknown. Due to a restricted gas lift capacity, the estimated gas lift rates allowed gas
lift optimisation to be confirmed.

Automatic System Quality Control (SQC)

One of the most valuable workflows is an automated system quality control, which is used to detect parts of the system that
are not in tune. A GAP network model is run every half hour using only the separator pressure and individual reservoir
pressures as boundary conditions. All flow rates, pressures, and temperatures within the network are then calculated. Large
deviations between estimated and measured values flag parts of the system that are not matched. Results are summarised with
other tuning data in a SQC Dashboard. Production engineers and production technologists can quickly scan this page to
determine how well the overall system fits and to locate problems (Figure 8).

The fit between the VLP, wellhead, and IPR rates for wells with downhole gauges is immediately visible. Another
column is used to compare the fit between measured and estimated tubing head pressures. Finally, misfits between measured
total oil, water and gas and estimates from the well models and the gap network models are available.

Figure 8 - System quality control dashboard used for monitoring system fit (indicative example only).

Allocation
IFM will be used to improve the daily well allocation process by enabling a more precise and transparent production
allocation process. Allocation will be based on GAP model results using the latest best estimate of well parameters and gas lift
rates.

Traditional production allocation utilises well performance curves generated monthly. This potentially leads to production
allocation being conducted using a well performance curve (WPC) up to 1 month out-of date, resulting in infrequent,
potentially major, changes in WPC. Using IFM is a quality assurance step, ensuring that the WPCs used for production
allocation are not more than 4 days out-of-date (potentially over a long weekend).

In addition, production allocation through IFM is achieved using secure transparent databases, with changes logged,
recorded, and auditable – a significant step away from traditional MS Excel based production allocation tools. MS Excel based
production allocation tools have tended to be uncontrolled, and difficult to audit.
SPE 116519 9

A large number of additional workflows are currently being developed including optimisation, and sand erosion
management.

Well and Network Tuning: Improvements in Integrated Ways of Working

One of the most fundamental uses of IFM is to keep well and network models continuously tuned. This forms the
foundation of well testing requirements as well as improving near-term optimisation, flow assurance management, reservoir
surveillance, and long-term reservoir management. Tuning needs to be repeated on a regular basis depending on how fast well
or reservoir conditions change. In order to be useful, tuning must be fast and easy to achieve. The steps include:

1. Quickly examining the system to easily find misfits and to get clues for why any parts of the system are not in
tune. Misfits can be effectively spotted using a combination of the automatic SQC workflow and Babelfish™
Dashboards, such as the SQC Dashboard (Figure 8).
2. Use existing data to predict values for uncertain parameters that need tuning. A key way of reducing the
number of uncertainties is to use all the relevant data and information on changing reservoir and well conditions,
including historical trends and forecasts, as well as theoretical parametric relationships such as PI vs. water cut.
3. Tuning misfit wells with downhole pressure gauges first.
4. Tuning wells with broken or no downhole pressures gauges next. Fits to tubing head pressure can be used for
these wells.

Once the well models are tuned up, then the network is checked. If the entire system is tuned within tolerance of 5% and
historical trends honoured, then well testing is not required. A well would be tested if the degree of uncertainty is too large.

The entire process is conducted with active collaboration between the Production Engineer and Production Technologist,
with close communication with the Reservoir Engineer.

Results

Implementing IFM on CLWH has had a significant and positive impact, including increasing production, improving work
efficiency, and increasing integration and decision making. Many of these improvements were unexpected and hence are
difficult to quantify a priori.

Less Well Tests

In a retrospective test case using historical data, IFM was capable of accurately estimating well rates for the entire field
over an 11 week period without well tests. Based on this positive outcome, it is anticipated that the number of well tests
needed to tune the well and network models will be much less than previously required. However, the full benefits from well
test deferral will become evident with experience.

Reducing the number of well tests also has the additional benefit of reducing the number of process disrupting activities,
thereby maintaining stability. This is especially relevant given a production system with increasing water cuts and reservoir
decline, which has resulted in numerous well kick-off and well clean-up issues directly attributable to process disruptions.

Better Efficiency and Troubleshooting

IFM has improved efficiency in a wide range of areas. First model catalogue has reduced the time spent looking for the
latest well and network models. The most up-to-date models are always known.

In another area IFM has also resulted in improved system monitoring and troubleshooting efficiency. For example, the
automatic system quality control allows system misfits to be detected at the click of a mouse. With the automatic SQC
workflow and the Babelfish™ dashboards, engineers can easily monitor the well and network models easily and in a way not
practically possible before. Lower quality work has been replaced with higher end work such as looking for optimisation
opportunities.

Integrated Data and Analysis

Perhaps the greatest immediate benefit of IFM has been a step change improvement in integrating both data analysis and
work practices. Tuning well and network models requires using and honouring accurate data. The automatic SQC for tuning
the well and network models integrates all measured tubing head pressures, downhole gauge pressures, total water, oil, and gas
rates in a way that has never been done at Woodside before.
10 SPE 116519

As well as using all measured production data, historical trends are now combined with forecasts to predict how well
parameters are changing with time. Woodside is also in the process of developing parametric relationships between
parameters, such as water cut, GOR, and PI. These will be used to provide reasonable boundaries to potential PI and water cut
changes. Updated well parameters are saved to PHD for storage and easy visualisation and historical trends are archived and
will not get lost in spreadsheets.

Using this trend information is particularly needed when fitting parameters, such as water cut in wells without downhole
gauges. For example, in a well that does not have a downhole gauge, trend data was used very well to predict a change in
water cut over 5 months (Figure 9). To ensure continued calibration of the model, when a predicted WC is 10% above the last
Well Test calibration point on a particular well, a well test will be conducted. On the subject well this point was reached after
5 months. Confidence was increased when after 5 months without a well test, the well test WC was 2% above RE Forecast
and 3% below the short term forecast. Predicting short-term WC change this way is proving successful in maintaining a match
at both well and system levels.

The tuning process requires regular interaction between the production engineer, production technologist and reservoir
engineer to develop a holistic understanding of system parameters, and has led to improved collaboration both formally and
informally.

70

60
WC (%)/PI (stb/d/psi)

50

40

30

20

10

0
13-Feb-02 28-Jun-03 09-Nov-04 24-Mar-06 06-Aug-07 18-Dec-08

Figure 9 - Water cut trend data was successfully used to predict future water cut at a well without a downhole pressure gauge

Improved Model Characterisation

A by-product of increased integration has been improved model characterisation, as well as greater correlation between the
well and network models and the reservoir forecast. For example, production forecasts are now compared weekly with
measured and calculated trends in well parameters, resulting in potentially problematic well forecasts being quickly identified.

In an interesting example, a comparison of reservoir pressures for wells within a well connected reservoir identified two
inconsistent wells (A and B) that were off trend from the other wells (Figure 10). One of the wells was surrounded by wells
that were on trend. A subsequent analysis of pressure build ups verified that the reservoir pressures of the two wells should, in
fact, have been on trend. Using IFM quickly identified inconsistencies between adjacent wells. Finding a well off trend would
be a clue of baffling or compartmentalization.
SPE 116519 11

Reservoir Pressure (@top Perf - psig)

3300 3350 3400 3450 3500 3550 3600
2780
WA6
2790

2800 A WA1
TVDSS Top Perf (m)

2810 WA3

WA8
2820
WA7
2830

2840

2850

2860
B WA9

Figure 10 - Two wells, A and B, had reservoir pressures that were off trend

IFM demonstrates the value of having permanent downhole pressure gauges. For example, it was found that IFM system
matching on wells with downhole gauges can be a better predictor of water cut than well testing because sometimes well test
results are inaccurate for a number of reasons. In one example, tuning with IFM indicated that a higher water cut was needed
to match the well than was indicated by the most recent well test, which was suspect (Figure 11). A subsequent well test of
improved rigour was then conducted that was consistent with that estimated by IFM and historical trends.

60

50

40
WC (%)

30

20

10
Historical Invalid IFM Matched

Time (Months)

Figure 11 - IFM estimate of water cut was more reliable than a estimate from a well test that turned out to be poor

Downhole gauge pressures may also be used to detect extremely subtle changes in well conditions. For example, changes
of 0.2% water cut were detected from the divergence of different rate methods. Unfortunately, however, the cause for the
change may not always be known for certain if many parameters are changing simultaneously.

In another example, the magnitude and duration of well clean up times (caused by cross flow during shut-in) may be
calculated by analysing the various well rate estimation methods (Figure 12). These calculations have significant implications
for the timing of well tests, and may also provide a case for a water shut-off or similar intervention.
12 SPE 116519

12,000
Well shut-
Well normal flow Well clean-up Well normal flow
in
10,000
Daily Oil Rate (bopd)
8,000

6,000

4,000
IPR Method
2,000 VLP Method
Wellhead Method

0
0 1 2 3 4 5 6 7
Days

Figure 12 - Example of monitoring cross flow in a well

Improved Prioritisation
The improved visualisation achieved by IFM has enabled incorrectly set valve positions to be identified quickly.
Previously, this was a relatively time consuming activity. For example, while troubleshooting an issue with the gas-lift
compressor, the cause of the problem was found to be a closed Xmas tree valve. Although this problem had gone unnoticed
offshore it was detected in 15 minutes using the Babelfish™ web pages.

IFM provides a much better link between data and value. This improves how broken or malfunctioning gauges are
prioritised for maintenance. For example on CWLH, there was a malfunctioning water meter. In the past a broken water
meter would be low on the priority list for being fixed, compared with urgent daily tasks. However, with IFM water cuts in
each well are important to predict. An accurate total water meter is now critical for tuning the well models. As a result,
maintaining the water meter was put at a high priority.

In another example, early in the implementation it was found that the pressures from the downhole gauges oscillated
between correct and incorrect values. It had been an ongoing problem for a couple of years. Downhole gauges provide critical
information to IFM for monitoring well behaviour. As a result, this problem was quickly fixed by developing a simple filter to
filter out bad pressure data.

All of the work in IFM in the past year was accomplished by a small team of up to six to eight people, most of whom were
part-time. Their backgrounds included production engineering, production technology, reservoir engineering, data
management, programming, and computer support. The team started with little IFM type experience, but did have the ability
to act quickly. Once the technology and workflows are developed for one asset they can be replicated much more easily,
cheaply and quickly on other assets. As such, IFM type implementations can be done at a relatively low cost.

Conclusion

Intelligent Fields Management and similar concepts are extremely valuable in improving field management, primarily through
improved integration and work practices. The benefits are pervasive throughout the company and are often unexpected. The
benefit-cost ratio for CWLH is likely on the order of ten to one.

Crucial technical elements of the IFM program include:

• The Petex Integrated Field Manager™ - allows custom workflows to be automated and scheduled, so that valuable,
but complex work that is often tedious, is easy to carry out. It also automatically archives data to data bases for
security and storage.
SPE 116519 13

• Babelfish™ visualisation - allows complex data to be analysed at the click of a mouse. It also provides easy access to
a wide range of data across multi-disciplines and increases transparency so that staff and management can see what the
IFM team is doing.

IFM is most valuable in offshore oil fields that have excellent instrumentation, particularly downhole pressure gauges working
in each well. Downhole pressure gauges are particularly valuable because they can detect subtle changes in well conditions.
Even though not all downhole pressure gauges were working within the CWLH field, IFM still proved valuable. Given this
result, IFM can be cost-effective in many other types of fields including offshore oil fields with few if any working downhole
pressure gauges, offshore gas fields, and large onshore oil and gas fields.

The minimum field requirements for an oil asset with relatively stable conditions, such as a strong aquifer drive, to
successfully implement IFM include:

1. a calibrated integrated asset model,

2. operational, accurate tubing head gauges,
3. facility meters for total oil, water and gas,
4. test facilities for calibrating well models, and
5. a means of estimating gas lift rates if gas lift is present.

For fields with rapidly changing conditions, more instrumentation will be needed, such as working down hole gauges.

It is likely that the greatest value in IFM and similar programs is in longer-term reservoir management and not near-term
management, which has been the focus of our current efforts. However, the immediate benefits indicate that the time is ripe
for implementing IFM concepts. As new fields become more complicated and oil prices are high: every barrel counts, the
value in using IFM is significant. At the same time, technological advances mean the concept can be implemented quickly and
at a reasonable cost. High frequency data from downhole and tubing head pressure gauges is becoming a standard and
software for the technological backbone, such as Petex Integrated Field Manager™ and data visualisation packages are
commercially available.

The greatest challenge in implementing IFM is dealing with the soft issues, such as changing the way work is done. Key
success criteria for implementing IFM include:

• having a small inter-disciplinary team that can focus, act quickly and work across disciplines,
• close interaction with asset teams,
• management support, and
• strong IT support, particularly in the area of database management.

Perseverance towards a common goal also helps because implementing new technology takes time. Ongoing support from a
central technical team of IFM super-users and IT experts will be required.

Real-time communication of offshore data to the office and back again is fundamental to improving production, reducing
downtime and making better use of staff resources. Intelligent Fields Management is a way of integrating hardware, software
and people, resulting in better decisions and more efficient operations.

References

Calvert, R. and A. Bakulin, 2007 New 4-D seismic monitoring techniques as enablers for smart fields. SPE 108207, SPE
Digital energy conference and Exhibition held in Houston TX April.

Hickman, A, A. Guidry and S. Seaton, 2008. Real-time collaboration-efficient problem solving and extending resources, SPE
112141 SPE Intelligent Energy Conference and Exhibition, Amsterdam, the Netherlands, 25-27 Feb.

Madray, R., C. Coll, G. Veitch, C. Chiboub, M. Butter, S Azouzi, S Bahri, B. Yaich, T Saada, 2008. Integrated Field
Modelling of the Miskar Field, SPE 113873, SPE Europec/EAGE Annual Conference and Exhibition held in Rome, Italy, 9–12
June 2008.
14 SPE 116519

Oberwinkler, C and Stundner, M. 2005 From Real-Time Data to Production Optimization, SPE Production and Facilities (SPE
97008).

Poulisse, H., P. van Overschee J Briers C. Moncur and C-D Goh, 2006 Continuous well production flow monitoring and
surveillance, SP# 99963 SPE Intelligent Energy Conference and Exhibition, Amsterdam, The Netherlands, 11-13 April 2006.

Stenhouse, B, 2008 Modelling and Optimisation in BP Exploration and Production Case Studies and Learnings, 2008, SPE
Intelligent Energy Conference and Exhibition, The Netherlands, 25-27 February, 2008.

van Dijk F. and K.C. Goh 2008 Closing the Loop for Improved Oil and Gas Production Management SPE 111997 Intelligent
Energy Conference Amsterdam, The Netherlands, 25-27 February 2008.

Acknowledgements

We would like to acknowledge the many other valuable contributors to the IFM program including: Mark Adams, Mubariz
Aliyev, Mike Buchanan, Nenad Burmez, Betsy Donaghey, Anthony Hernandez, Mike Krzus, Chris Lihou, Greg Mogan,
Martin Rice, James Stewart, Keith Spence and Bill Wallwork. We also greatly appreciate the support from the Woodside
Technology Fund that provided the seed money for this program and our executive sponsor Kevin Gallagher. We also thank
the CWLH joint venture partners for permission to publish this paper. Integrated Field Manager and Babelfish are trademarks
of Petroleum Experts and the ISS Group, respectively.

Biography

Bruce James
Bruce James is the IFM Development Leader at Woodside and has worked for Woodside for almost 3 years. Before that he
worked for Unocal and Shell. Bruce Holds BASc and PhD degrees in geological engineering from the University of Waterloo
and University of British Columbia, respectively as well as an MS in Engineering Economic Systems and Operations Research
from Stanford University.

Kevin A. Kerr
Kevin A. Kerr was the Cossack Pioneer Production Engineer during the initial phase of IFM implementation, and has worked
at Woodside for 5 years. Prior to working at Woodside, Kevin worked as a Process Engineer at Hatch Associates. Kevin holds
a BE (Petroleum) and BSc (Chemistry) from The University of Western Australia.

Stephanie Lim
Stephanie Lim is a Graduate Production Engineer and is the current Production Engineer for the Cossack Pioneer. She has
worked at Woodside for 12 months and prior to working at Woodside, Stephanie completed a BE (Petroleum) and BCom
(Finance) at The University of Western Australia.

Ed Lewandowski
Ed Lewandowski is a Senior Production Engineer and has worked at Woodside for almost 20 years. He has previous roles at
Woodside include software, planning, and reliability engineering. He as also worked previously for Shell and EDL
Engineering. Ed has a BE (Mechanical) from the University of Newcastle.

Craig Knight
Craig Knight is a Graduate Production Technologist and has worked at Woodside for 18 months. Prior to working at
Woodside Craig completed a BE (Petroleum) at the University of New South Wales.

Richard Bell
Richard Bell is E&P data analyst and is the IT project manager for IFM. Richard was worked for Woodside for 1.5 years.
Prior to working for Woodside, Richard was a geophysicist for 6 years, and spent 5 years in technical IT support roles for
SAIC and Geotech. He has a BSc (Applied Geology) from Sunderland University and an MSc (Petroleum Geology).