SPE 163312

Use of Integrated Modeling for Production Optimization in KOC North

Kuwait Jurassic Gas Field
Satinder Malik, Shell Kuwait; Noura Al-Mai, KOC; Qasem Dashti, KOC

Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Kuwait International Petroleum Conference and Exhibition held in Kuwait City, Kuwait, 10–12 December 2012.

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
The North Kuwait Jurassic Gas (NKJG) reservoirs are currently under development by KOC. The fractured carbonate
reservoirs contain gas condensate and volatile oil at pressures up to 11,500 psi with 2.9% H2S and 1.5% CO2. Currently
around 20 active wells are producing to an Early Production Facility (EPF-50) that falls short of achieving the desired
capacity and capability to handle production efficiently.
To understand wells and field performance, an integrated system model comprising of wells, flow line and gathering system
separator network was created. The setting up a model and its use is an integral subset of WRFM (Wells, Reservoir and
Facilities Management) process that is essential for effectively managing the current asset and for further field development.
The application of the model is to be an enabler for wider implementation of the WRFM process in KOC and a tool to meet
the following objectives:
 It should closely match the EPF-50 throughput
 It should identify problem wells
 It should identify well intervention and data gathering opportunities
 It should be a tool for training engineers on nodal analysis and WRFM
 It can be used for back allocation
 It could be used to benchmark against any existing models

The model has shown close approximation with field metered production and is already achieving many of its desired
objectives.
This paper describes the use of integrated nodal analysis model to generate data gathering and well intervention opportunities
not only to operate the facilities efficiently but understand well and reservoir behavior for input to full field development
plan.

Background

Exploration activity during the last ten years, targeting Jurassic carbonate reservoirs in North Kuwait (Fig 1), has culminated
in the discovery of six major tight gas condensate fields, encompassing an area of about 1,800 sq km with a reservoir gross
thickness of about 2,200 ft. These fields are the first free-gas fields in Kuwait, which were put on early production during
2008. The reservoirs are characterized with dual porosity matrix system, dominated by low porosity and permeability, in deep
HP/HT conditions, with wide variety of hydrocarbon fluids ranging from volatile oil to gas condensate with sour gas. Typical
per well production rates are up to 5,000 BOPD/BCPD and 10 MMSCFPD, making them an excellent commercial success.

The Jurassic of Kuwait unconformably overlies the Triassic clastic and evaporative Minjur formation; consists of
transgressive 2000ft of Marrat section. Based on sequence stratigraphic studies a depositional model has been developed to
provide a framework for the reservoir model. The depositional model incorporates Sabkha, tidal flats, lagoon, back shoal,
shoal, shore face, inner shelf, outer shelf, and slope and basin depositional environments . Hyper saline brines were generated
in the lagoon, seeped downward, selectively dolomitizing the underlying strata, creating secondary porosity and permeability
and significantly improving the reservoir characteristics in some of these fields. Significance of Natural Fracture in Jurassic
2 SPE 163312

carbonate reservoirs has been established trough production in few of the key areas. Fig 2 highlights the stratigraphic column
of the NKJG reservoirs. Marrat and Najmah /Sargelu are the two significant producing reservoirs. Marrat formation is the
main producing reservoir from fracture and matrix reservoir. Unit C (Middle Marrat) porous limestone & dolomites have the
best reservoir properties. Najmah & Sargelu formation are secondary reservoirs producing gas and condensate from
fractures.
Reservoir facies are dolomites, dolomitic limestones and limestones.Matrix properties are low with porosities ranging from
3% to 24% and permeability ranging from 0.001md to 100md. Good production is generally attributed to fracture dominated
areas. Reservoir fluids are understaurated, near critical fluids and could be either gas condensate or volatile oil.

Fig 1: North Kuwait Jurassic gas fields

SPE 163312 3

Fig 2: Jurassic type section based on a typical well. The main Jurassic reservoir is unit “C” of the Marrat formation.
Abbreviations are Gamma Ray (GR), Litho –Density Log (LDL in gm.cc) and Compensated Neutron Log (CNL)-(Ref. 6)

Introduction
The IPSM (Integrated Production System Modelling) model integrates nodal analysis software (Prosper) and system model
(GAP) to generate pressure and rate profile from reservoir to surface facilities intake (Separator) to export or sales point. It’s
a unified process that links the simulation model at the reservoir levels to nodal analysis at the well and surface network,
incorporating the basic physics at each node (SPE 140632)
The use of the IPSM model falls under the following categories:
• Surveillance- Prosper – dated output. For individual well performance analysis
• Optimization- network model. For system performance analysis (subsurface to surface sink)
• Forecasting – integrated model. If reservoir model coupled.
• Development planning- integrated model with a potential to link to a planner and economic model
4 SPE 163312

The benefits of the IPSM are

1. Cross discipline integration within and outside a given asset
2. Linking various functions (subsurface and surface) as one unit
3. Understanding system bottlenecks
4. Identifying & testing opportunities before committing resources
5. Identifying flow assurance issues
6. Scenario analysis and concept selection
7. Decision making and communication tool
8. Generate business value
The application of the model is to be an enabler for wider implementation of the WRFM process in KOC and a tool to meet
the following objectives:
 It should closely match the EPF-50 throughput
 It should identify problem wells
 It should identify well intervention and data gathering opportunities
 It should be a tool for training engineers on nodal analysis and WRFM
 It can be used for back allocation
 It could be use to benchmark against any existing models

And finally to use the model for short term forecasting against given constraints and as an activity planning tool.

Process- work flow

WRFM (Wells, Reservoir and Facilities Management) is a process that integrates a well defined path from identifying to
realizing an opportunity in an operating asset. Fundamentally, WRFM is about maximizing production and reserves and then
applying efficiencies to reduce costs to maximize asset value by understanding and optimizing the performance of reservoirs,
wells, and facilities all the way from reservoir to the wells, through the wells to the flow lines and facilities and then on to the
sales point. (Ref. Shell internal documentation)
Fig 3 depicts a WRFM process path also known as WRFM value loop that is recommended to be followed for optimizing a
given asset performance.

Fig 3: WRFM value loop (Courtesy Shell International)

As observed ‘Interpretation & modeling’ is an integral part of the WRFM value loop. Actually using the very exercise to
build a model and run it is being used to introduce the concept of WRFM value loop in KOC.
Making a model is not a standalone exercise to be done in isolation. It involves all functions of value chain (chasing a
molecule from reservoir to export). Fig 4 highlights different functions that are essential to be engaged in making a robust
and reliable model and were engaged and involved in making the current EPF-50 & EPF-150 model (Early Production
Facilities model).
SPE 163312 5

Prod.
Tech.
Facilities
Engrs

Well
Engrs

Management
oversight INTERFACE MANAGEMENT

Reservoir
Engrs

Well
Interventon/
Surveillance

Geo/PP
Process
Engrs

Fig 4: Interfaces between different functions – essential for making a robust IPSM

The highlight of the process that was and is being followed to make and keep the current model updated is depicted in Fig 5.

Fig 5: The process in making IPSM

6 SPE 163312

Model Build
Fig 6 highlights the bird’s eye view of the EPF-50 process path. Currently around 24 wells are connected and flowing to
EPF-50 with majority of wells with their own dedicated flow line to the EPF-50. Only few wells share their flow line with
other wells. The wells are either routed through ‘dry’ or ‘wet’ header train and fluid goes through 3 stages of separation.
The current EPF-50 & EPF-150 model does not have a coupled reservoir model. KOC is working on building MBAL model
and once they are ready, the reservoir model will be coupled with IPSM and the model could be then be used for prediction
runs and short term forecasting. The current model is being used for understanding well performance, identifying problem
wells, well and system performance analysis and back allocation.

Fig 6: Level 0 (high level) EPF-50 process schematic

Fig 7 depicts the GAP model that is constructed to integrate EPF-50 multi well model, generated individually in prosper with
surface facilities. The model can be run to generate following scenarios:
 Optimized case
 Non optimized case
 Constraint case
 Unconstraint case
 Prediction /Short term forecasting
The model currently is being used to generate well specific intervention opportunities. Note that the reservoir model (MBAL)
is not connected with wells and hence the current model cannot be used for prediction runs and short term forecasting.

Fig 7: EPF-50 GAP model without reservoir model connection

SPE 163312 7

As a general rule, the NKJG reservoir fluids are near critical fluids with reservoir temperature close to critical temperatures.
Since the reservoir fluids are near critical and behave either as ‘gas condensate’ or ‘volatile oil’, EOS (Equation of State)
model is used for generating PVT properties and phase envelope.
EOS is not a predictive tool but a relationship between P, V, &T. It needs detailed lab data to calibrate the model to honor the
mass balance. Those wells that had PVT model based on BHS (Bottom Hole Samples) generated more robust PVT definition
than those based on surface recombined samples.
A validated and calibrated PVT model is used to generate individual well model in prosper. Prosper well models are
matched, validated and calibrated using pressure and well test data.
As well model is central to system analysis, it is essential to have a robust well model and a calibrated PVT model is essential
for building a robust well model. For near critical fluids, PVT data is a must and it is recommended to get as many fluid
samples as technically and commercially feasible. A good PVT data base also helps in better delineation of the reservoir.

Model Application
The model is currently being used to generate data gathering and identifying well intervention opportunities, back allocation
and to optimize production by indentifying optimum choke setting for individual wells. For the newly drilled wells that are
not hooked to current facilities, the model is being used to prioritize well hook up candidates under a constraint environment
of limited header space.
Identify problem wells
Fig 8 shows an example of P-Q curve generated for a typical well. If well tests points consistently do not fall on the curve it
highlights either the well is damaged or stimulated. Skin values are used to get a match on P-Q curve only after well reservoir
parameters like pressure and Kh values are assured. This triggers data acquisition requirements to confirm either the model
parameters or acquire data such as well test/ PBU etc to design an appropriate well intervention activity.

Fig 8: A typical well P-Q curve

Fig 9 highlights the process that is currently being followed to indentify problem wells and generate data gathering and well
intervention opportunities. As the reservoir is very dynamic with strong aquifer support and is not very homogeneous, the
reservoir parameters for e.g. reservoir pressure, Kh are not constant and hence regular data gathering is essential to keep the
model current and validated.
8 SPE 163312

Fig 9: Flow path for generating data gathering and well intervention opportunities
Optimizing choke size
There is a self imposed constraint on the choke size for wells to mitigate against the risk of sudden decline in well
productivity. Coupled with the choke size, another constraint is the company policy of not producing the wells below bubble
point.
But recently, few currently producing wells have shown a sudden & significant drop in their production performance and that
has negatively impacted daily field production throughput. In addition current wells are required for reservoir studies to get
data for future field development planning and reservoir modelling. This may necessitate closing few wells for extended
duration to gather reservoir data.
To reduce the impact of production deferment arising out of extended shut in of wells and compensate for declining
production from some wells, IPSM model was used to shortlist wells that have a potential for further optimizing choke size to
make up for the loss in production for the shut in wells.
IPSM/GAP model is run with following constraints with inline choke either bypassed or beaned up to understand well
response:
 FBHP (Flowing Bottom Hole Pressure) above Pb (Bubble point)- Initially constraint is set to be 2000-3000 psi
above Pb
 Flow line pressure must not exceed 1400 psi

High safety cushion is given for bubble point constraint because the well will need to be monitored very closely to prevent
pressure declining precipitately after bean up and have enough time to take preventive action if required.
A list of prioritized wells with potential to further optimize choke size is continuously updated to its current.
The recommendation given to field is to open up the well in stages and continuously monitor well performance at every
incremental choke opening. Other subsurface parameters like proximity to fault, standoff from free water level, fracture
network etc are also considered.
SPE 163312 9

Prioritization of Hook up candidate wells

The current production facility can only cater to limited number of wells. The field is being appraised more and new zones
are being drilled and tested along with increasing number of wells being drilled in existing proven zones to cater to future
expansion to meet 1 BCF target. The newly drilled wells should be tested for longer duration to get an assurance on
production potential and to obtain well and reservoir data for field development planning and facilities design.
The model has been used to gain an insight on indentifying those wells that can be hooked temporarily to existing production
facility to facilitate their clean up and conduct long term testing.
Following objectives were considered for short listing the wells that could be connected to existing facility for long term
testing:

 Wells should meet some appraisal value for e.g. testing of new zones
 Wells should have short jump over flow line length to nearest producing well (save cost & time for hook up)
 The wells should not adversely impact currently producing wells performance and production
 The wells should have sufficient energy i.e. pressure to justify and give confidence in their deliverability besides
being conducive for long term testing to give reservoir properties values to update wells model
 The results should influence appraisal drilling program.
 Confirm the pressure is still holding in Virgin areas (compartmentalization, etc), - will influence drilling program

Understanding facilities constraint & production back allocation to wells

The current facilities are designed to cater to certain composition of fluid stream especially with respect to H2S & CO2. The
model is being used to get an insight on how the percentage of H2S & CO2 changes in the gas stream when the fluid goes
through multiple stages of separation. Since the current facility can only handle certain volume of Oil, Gas and Water, the
model is also being used to generate scenarios on fluid composition with different wells on stream. This is to prioritize wells
brought online.
The model also has been used to back allocate production as reported by fiscal meter to respective wells. This is made
possible by keeping the wells model updated by doing frequent well tests. As the reservoir is very dynamic a regular data
gathering campaign has been initiated based on model results. The wells that show significant deviation from the simulated
result are earmarked for well intervention to gather data for updating the model. This will become a part of WRFM process.

Challenges
There are numerous challenges involved in setting up a model. Few of the significant challenges encountered during setting,
calibrating and validating the model were
• PVT characterization
• Generation of robust phase envelope for all types of fluids
• Ever changing well performance due to the very nature of reservoir performance
• Well tests coupled with PTA/pressure surveys
• Assurance on well tests and PTA results
• Reliability/Availability of production facility

Conclusion
The IPSM has proven to be an effective, robust and reliable tool for understanding the performance of wells, identifying
problem wells, facilities constraints and generating well intervention and data gathering opportunities. It is the backbone for
Wells, Reservoir and Facilities management initiative by the asset and has proven to be a valuable tool for production back
allocation to wells. It has also given a training platform for new engineers to understand nodal analysis and system
optimization methodology.
GAP model is also being used to get an insight on phase behavior of near critical fluids and develop an understanding on the
gas composition, especially H2S & CO2 at different stages of separation. This information is critical for asset integrity and
flow assurance. A robust, validated PVT model for near critical fluid is most important parameter for generating a reliable
and robust Prosper/ GAP model.
The greatest challenge in keeping the model updated is the availability of reliable, validated data be it well test, pressure build
up analysis or a good PVT analysis.
Once reservoir model is connected, the IPSM could also be used for short term forecasting and as a predictive tool.
The IPSM approach has proven to be an effective multifunctional and multi discipline integrator.

Acknowledgements
The authors wish to thank the Ministry of Oil, Kuwait and KOC management to publish this paper. This project of
developing Prosper & GAP model for current and future production facilities has been a joint KOC-Shell initiative. The
10 SPE 163312

successful culmination of this project has established ground work for further work to develop predictive and forecasting
model that will serve the backbone for the KOC Wells, Reservoir and Facilities management initiative.
The project success would not have been possible without support from KOC management.

References
1. E.A.Ageh, A.Adegoke, O.J.Uzoh, “Using Integrated Production Modeling (IPM) as an Optimization Tool for Field
Development Planning and Management’, SPE 140625
2. Ehimhen Okoh, Shekar Sathymoorthy, Elozino Olaniyan, Ozioma Ezeokeke, “ Application of Integrated Production
System Modelling in Effective Well and Reservoir Management of the Bonga Field”, SPE 140632
3. L.K.Harms, “Better Results Using Integrated Production Models for Gas Wells”, SPE 93648
4. Tejaswi Shrestha, Suzanne Hunt, Paul Lyford, Hemanta Sarma. “Workflow for Integrated Production Modelling of
Gas Wells in the Northern Cooper Basin’ SPE 116936.
5. C. Correa Feria, “ Integrated Production Modeling- Advanced but, not Always Better”, SPE 138888
6. Samir Yousif and Ghalib Nouman , MOO, Kuwait “Jurassic Geology of Kuwait”, GeoArabia Vol.2, No. 1, 1997
7. ‘Shell International’ internal WRFM process