spe distinguished lecturer seriesno cap on field productivity (1 simulation runs)no cap on field...

SPE DISTINGUISHED LECTURER SERIESis funded principally

through a grant of the

SPE FOUNDATIONThe Society gratefully acknowledges

those companies that support the programby allowing their professionalsby allowing their professionals

to participate as Lecturers.

Shahab D. Mohaghegh, Ph.D. – WVU & ISI SPE Distinguished Lecture Series, 2007 - 2008

And special thanks to The American Institute of Mining, Metallurgical,

and Petroleum Engineers (AIME) for their contribution to the program.

Smart Completions, Smart Wells and SPE Distinguished Lecture 2007-2008

Now Smart Fields; Challenges & Potential SolutionsChallenges & Potential Solutions

Shahab D. Mohaghegh, Ph.D.West Virginia University &g yIntelligent Solutions, Inc.

2

Smart Oil Field Technology

Smart Completion:Smart Completion:Remotely monitor & control downhole fluid production or injection.Downhole control to adjust flow distributionsadjust flow distributions along the wellbore to correct undesirable fl id f t tfluid front movement.


Smart Oil Field Technology

Smart Well:Using permanent downhole gauges for continuous monitoring of pressure, flow rates, … and automatic fl t lflow controls.Capability of automatic interaction using extensive downhole communication.


Smart Oil Field

The Missing link


Characteristics of Smart Fields

Availability of high frequency data.Making

reservoir t

The Missing link

management decisions based

on real time data from the

field.

Possibility of intervention, control and management from a distance.



Availability of high frequency data.Making

reservoir t

The Missing link

management decisions based

on real time data from the

field.

Possibility of intervention, control and management from a distance.



Making reservoir management decisions based on real-time data from the field.Considerations:

Reservoir management tools.Uncertainties associated with the geological model.Predicting the consequences of the decision.Real-time optimization.


Hardware / Software

Intelligence requires a combination of hardware and software.We have made strong advances in hardware.gSoftware development is lagging.Intelligent Systems will play a pivotal role:Intelligent Systems will play a pivotal role:

Artificial Neural NetworksFuzzy Set TheoryFuzzy Set TheoryGenetic Optimization


Hardware / Software

Surrogate Reservoir Models (SRM) are developed to address the software need of smart fields.SRM are reservoir management tools for smart fields:

Real-time full field reservoir simulation & modeling

Predictive modelingUncertainty analysisR l ti ti i ti


Real-time optimization

Removing The Bottle-NeckFull Field Flow Models for

Reservoir Simulation & Modeling. One of the major tools

Real-Time, High Frequency Data Stream

Modeling. One of the major tools for integrated Reservoir

Management

Ti S l Ti S lTime Scale:

Seconds, Minutes, Hours

Time Scale:

Days, Months, ….

How can the bottle-neck be removed?

Perform analysis at the same time scale as the High Frequency Data Streams; in seconds, or better yet, in

REAL TIME


REAL-TIME

SURROGATE RESERVOIR MODEL Definition

Surrogate Reservoir Models are replicas ofSurrogate Reservoir Models are replicas of the numerical simulation models (full field flow models) that run in real-timeflow models) that run in real time.REPLICA.

A d ti f k f t i llA copy or reproduction of a work of art, especially one made by the original artist.A copy or reproduction especially one on a scaleA copy or reproduction, especially one on a scale smaller than the original.Something closely resembling another.


g y g

Characteristics of SRM

SRMs are notSRMs are notresponse surfaces.statistical representations of simulation modelsstatistical representations of simulation models.

SRMs areengineering toolsengineering tools honor the physics of the problem in hand.adhere to the definition of “System Theory”adhere to the definition of System Theory .

INPUT OUTPUTSYSTEM


Case Study

Lets see an example of a SurrogateLets see an example of a Surrogate Reservoir Model in action.This case study demonstrates developmentThis case study demonstrates development of a surrogate reservoir model (SRM) that will run in real-time in order to accomplish therun in real time in order to accomplish the objectives of the project.


Background

A giant oil field in the Middle EastA giant oil field in the Middle East.Complex carbonate formation.165 horizontal wells165 horizontal wells.Total field production capped at 250,000 BOPD.Each well is capped at 1,500 BOPD.Water injection for pressure maintenance.


Background

Management Concerns:Management Concerns:Water production is becoming a problem.Cap well production to avoid bypass oilCap well production to avoid bypass oil.Uncertainties associated with models.

Technical Team’s Concerns:Technical Team s Concerns:May be able to produce more oil from some wells (which ones? How much increase?) without ( )significant increase in water cut.Increasing well rate may actually help recovery.


Objective

Increase oil production from the field by identifyingIncrease oil production from the field by identifying wells that:

will not suffer from high water cut.will not leave bypassed oil behind.


Objective

Accomplishing this objective requires:Accomplishing this objective requires:Exhaustive search of the solution space, examining all possible production scenarios, while considering uncertainties associated with the geological modeluncertainties associated with the geological model.Hundreds of thousands of simulation runs; thus development of a Surrogate Reservoir Model (SRM) based on the Full Field Model (FFM) became a requirement.


Flow Model Characteristics

Full Field Flow Model Characteristics:Full Field Flow Model Characteristics:Underlying Complex Geological Model.Industry Standard Commercial Reservoir Si l tSimulator165 Horizontal Wells.Approximately 1,000,000 grid blocks.Approximately 1,000,000 grid blocks.Single Run = 10 Hours on 12 CPUs.


Very Complex GeologyNaturally Fractured

Carbonate Reservoir.

Reservoirs represented in th Fl M d l


the Flow Model.

Steps Involved in SRM Development

Identify Clear ObjectivesIdentify Clear ObjectivesDesign SRM’s input and outputGenerate DataGenerate DataBuild SRMValidateAnalyzeResults & Conclusions


SRM’s Objective

Accurately Reproduce the following for theAccurately Reproduce the following for the next 25 to 40 years.

Cumulative Oil ProductionCumulative Oil ProductionCumulative Water ProductionInstantaneous Water Cut


SRM’s Input & Output

OUTPUT was identified by the ObjectiveOUTPUT was identified by the ObjectiveCumulative Oil ProductionCumulative Water ProductionCumulative Water ProductionInstantaneous Water Cut

INPUT must be designed in a way to captureINPUT must be designed in a way to capture the complexity of the reservoir.

Well-based SRMWell based SRMWell-based SRM gridCurse of dimensionality


Curse of dimensionality

Curse of Dimensionality

Complexity of a system increases with itsComplexity of a system increases with its dimensionality.Tracking system behavior becomesTracking system behavior becomes increasingly difficult as the number of dimensions increases.dimensions increases.Systems do not behave in the same manner in all dimensionsin all dimensions.

Some are more detrimental than others.


Curse of Dimensionality

Sources of dimensionality:Sources of dimensionality:STATIC: Representation of reservoir properties associated with each well.DYNAMIC: Simulation runs to demonstrate well productivity.


Well-Based Surrogate Reservoir Model

Surrogate Model Elemental Volume.Surrogate Model Elemental Volume.1 2 3 4 5 6 7 8

1

22

3

44

5


The elemental volume includes 40 SRM blocks.

Curse of Dimensionality, StaticPotential list of parameters that can be collected on a “per-well” basis.p

Latitude Longitude

Parameters Used on a per well basisLatitude Longitude

Deviation Azimuth

Horizontal Well Length Productivity Index

Distance to Free Water Level Water Cut @ Reference PointDistance to Free Water Level Water Cut @ Reference Point

Flowing BHP @ Reference Point Oil Prod. Rate @ Reference Point

Cum. Oil Prod. @ Reference Point Cum. Water Prod. @ Reference Point

Distance to Nearest Producer Distance to Nearest Injectorj

Distance to Major Fault Distance to Minor Fault16 Parameters


Curse of Dimensionality, Static

Potential list of parameters that can be collected onPotential list of parameters that can be collected on a “per-SRM block” basis.

Mid Depth Thickness

Relative Rock Ttype Porosity

Initial Water Saturations Stylolite Intensity

Parameters Used on a per segment basis

Initial Water Saturations Stylolite Intensity

Horizontal Permeabil ity Vertical Permeabil ity

Sw @ Reference Point So @ Reference Point

Capil lary Pressure/Saturation Function Pressure @ Reference Point

12 Parameters


Curse of Dimensionality, Static

Total number of parameters that need representation during the modeling process:

12 t 40 id bl k/ ll 48012 parameters x 40 grid block/well = 480

16 parameter per well

Total of 496 parameter per well

Building a model with 496 parameters per well is notBuilding a model with 496 parameters per well is not realistic, THE CURSE OF DIMENSIONALITY

Dimensionality Reduction becomes a vital task.


Dimensionality Reduction becomes a vital task.

Curse of Dimensionality, Dynamic

W ll d ti it i id tifi d th hWell productivity is identified through following simulation runs:

All ll d i t 1500 2500 3500 & 4500All wells producing at 1500, 2500, 3500, & 4500 bpd (nominal rates)

No cap on field productivity (4 simulation runs)No cap on field productivity (4 simulation runs)Cap the field productivity (4 simulation runs)


Need to understand reservoir’s response to changes in imposed constraints.

Curse of Dimensionality, Dynamic

Well productivity through following i l tisimulation runs:

Step up the rates for all wellsNo cap on field productivity (1 simulation runs)No cap on field productivity (1 simulation runs)Cap the field productivity (1 simulation runs)


Need to understand reservoir’s response to changes in imposed constraints.

Data Generation

Total of 10 simulation runs were made to t th i d t t f th SRMgenerate the required output for the SRM

development (training, calibration & validation)validation)Using Fuzzy Pattern Recognitiontechnology input to the SRM was compiledtechnology input to the SRM was compiled.


Fuzzy Pattern Recognition

I d t dd th “C fIn order to address the “Curse of Dimensionality” one must understand the behavior and contribution of each of thebehavior and contribution of each of the parameters to the process being modeled.Not a simple and straight forward task !!!Not a simple and straight forward task. !!!



To address this issue, we use Fuzzy PatternTo address this issue, we use Fuzzy Pattern Recognition technology.



Parameter: Pressure @ Reference


Parameter: Pressure @ Reference

Key Performance Indicators


Validation of the SRMat

e M

odel

)t %

(Sur

roga

Wat

er C

ut

Water Cut % (Reservoir Simulator)


Water Cut % (Reservoir Simulator)

Validation of the SRMod

el)

urro

gate

Mo

rodu

ctio

n (S

lativ

e O

il Pr

Cumulative Oil Production (Reservoir Simulator)

Cum

u


( )

Validation of the SRM


Using SRM for Analysis

Identify wells that benefit from a rate increaseIdentify wells that benefit from a rate increase and those that would not.Address the uncertainties associated with theAddress the uncertainties associated with the simulation model.Generate Type curves for each wellGenerate Type curves for each well.

Design production strategy.Use as assisted history matching toolUse as assisted history matching tool.

To perform the above analyses millions of simulation runs were required.


p y qUsing the SRM all such analyses were performed quite quickly.

Optimal Production Strategy

Well Ranked No. 1

IMPORTANT NOTE: This is NOT a Response Surface – SRM was run hundreds of


ptimes to generate these figures.


ll k d 100

IMPORTANT NOTE: This is NOT a Response Surface – SRM was run hundreds of i h fi

Well Ranked No. 100


times to generate these figures.


Wells were divided into 5 clustersWells were divided into 5 clusters.Production in wells in cluster 1 can be increased significantly without substantialincreased significantly without substantial increase in water production.

Cl 1Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5

12 Wells 14 Wells 22 Wells 37 Wells 80 Wells

Best Performance

12 Wells


Best Performance

Analysis of Uncertaintyy y

Objective:Objective:To address and analyze the uncertainties associated with the Full Field Model using Monte gCarlo simulation method.


Analysis of Uncertainty

Motivation:Motivation:The Full Field Model is a reservoir simulator that is based on a geologic model. g gThe geologic model is developed based on a set of measurements (logs, core analysis, seismic, …) and corresponding geological and geophysical interpretations.



Motivation:Motivation:Therefore, like any other reservoir simulation and modeling effort, it includes certain obvious g ,uncertainties.One of the outcomes of this project has been the identification of a small set of reservoir parameters that essentially control the production behavior in the horizontal wells in this field (KPIs)behavior in the horizontal wells in this field (KPIs).



Following are the steps involved:Following are the steps involved:1. Identify a set of key performance indicators that

are most vulnerable to uncertainty.y2. Define probability distribution function for each of

the performance indicators.a. Uniform distributionb. Normal (Gaussian) distributionc Triangular distributionc. Triangular distributiond. Discrete distribution



Following are steps involved:Following are steps involved:3. Run the neural network model hundreds or

thousands of times using the defined probability g p ydistribution functions for the identified reservoir parameters. Performing this analysis using the act al F ll Field Model is impracticalactual Full Field Model is impractical.

4. Produce a probability distribution function for cumulative oil production and the water cut atcumulative oil production and the water cut at different time and liquid rate cap.



Following are steps involved:Following are steps involved:5. Such results bounds to be much more reliable

and therefore, more acceptable to the , pmanagement or skeptics of the reservoir modeling studies.



Average Sw @ Reference point in Top g w @ p pLayer II

Value in the model = 8%Lets use a minimum of 4% and a maximum of 15% with a triangular distribution

4 8 15



Average Capillary Pressure @ ReferenceAverage Capillary Pressure @ Reference point in Top Layer III

Value in the model = 79 psiValue in the model 79 psiLets use a minimum of 60 psi and a maximum of 100 psi with a triangular distribution

60 80 100


00


PDF for HB001 Cumulative Oil and Cumulative Water production at the rate of 3,000 blpd cap after 20 years.

Actual Models are available & can be demonstrated after the presentation.


p

Type Curves

Type curves can be generated in secondsType curves can be generated in seconds to address sensitivity of oil and water production to all involved parameters.p pType curves can be generated for:

Individual wellsIndividual wellsEach cluster of wellsEntire fieldEntire field


Type Curves

Cum. Oil Production as a function of

Average Horizontal Permeabilityin one of the top layers.


Type Curves

Water Cut as a function of

Average Horizontal Permeabilityin the well layers.


Type Curves


Average Vertical Permeabilityin one of the top layers.


Type Curves


Average Vertical Permeabilityin the Well layers.


Results & Conclusions

Upon completion of the project managementUpon completion of the project management allowed production increase in six cluster one wells.After 8 months of successful production rest of the cluster one wells were also put onof the cluster one wells were also put on higher production.It has been more than 15 months since theIt has been more than 15 months since the results were implemented with success.


Results & Conclusions

A successful surrogate reservoir model wasA successful surrogate reservoir model was developed for a giant oil field in the Middle East.The surrogate model was able to accurately mimic the behavior of the actual full field flowmimic the behavior of the actual full field flow model in real-time.


CONCLUSIONS

Development of successful surrogateDevelopment of successful surrogate reservoir model is an important and essential step toward development of next generation p p gof reservoir management tools that would address the needs of smart fields.


spe distinguished lecturer seriesno cap on field productivity (1 simulation runs)no cap on field...

Documents