Control Optimization of Oil Production under Geological Uncertainty ¹´²´³Agus Hasan, ²´³Bjarne Foss, ¹´³Jon Kleppe 03/25/22 NPCW 2009 NTNU ¹Department of Petroleum Engineering, NTNU ²Department of Cybernetics Engineering, NTNU ³Center for Integrated Operations in Petroleum Industry Nordic Process Control Workshop 2009 Porsgrunn, Norway 29-30 January 2009
23
Embed
Control Optimization of Oil Production under Geological Uncertainty ¹´²´³Agus Hasan, ²´³Bjarne Foss, ¹´³Jon Kleppe 9/6/2015 NPCW 2009 NTNU ¹Department.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Control Optimization of Oil Production under Geological Uncertainty
¹´²´³Agus Hasan, ²´³Bjarne Foss,¹´³Jon Kleppe
04/19/23 NPCW 2009NTNU
¹Department of Petroleum Engineering, NTNU²Department of Cybernetics Engineering, NTNU
³Center for Integrated Operations in Petroleum Industry
Nordic Process Control Workshop 2009Porsgrunn, Norway29-30 January 2009
Outline
04/19/23 NPCW 2009NTNU
Objectives and Motivations Closed-loop Reservoir Management Case Study Part 1 Optimization
Optimization MethodsReservoir Control StructureBinary Integer ProgrammingOptimization Results
Part 2 UncertaintyGeological UncertaintyHistory MatchingResults
Conclusions and Recomendations
Objectives and Motivations
• Efficient: Fast enough• Accurate• Robust• Applicable: can be used in practical way
Which optimization method should we choose in our problem?
Objective function: Net Present Value (NPV)
injProd, ,
, ,1 1 1
, , , ,
1n
n n NNNo o j w o j n
w inj inj itn j i
r q x u m r q x u mNPV r q
b
Objectives:
Find operating combination conditions of down-hole valve settings that optimize the water flood. Investigate potential for improvement as function of reservoir properties and operating constraints.
04/19/23 NPCW 2009NTNU
Closed-Loop Reservoir Management
Production System
(Reservoir, Well)
Reservoir Simulator
Optimization
Optimization
Calc.NPV
Data
Identification and Updating
Identification and Updating
Control andOptimization
GeologicalUncertainty
04/19/23 NPCW 2009NTNU
Case StudyGrid cells : 45 x 45 x 1 = 2025
2-phases : Oil-Water
Assumptions:
1 Injector and 1 Producer well
Each well was divide into 45 segments
Each segments was modeled as a separated “smart well”
No flow boundaries
Incompressible and Immiscible fluids flow
No capillary pressure
No gravity effect
04/19/23 NPCW 2009NTNU
(Brouwer 2004)
Initial Data• Porosity : 0.2 (uniformly distributed)• IOIP : 324000 sm3 = 2041200 bbl• Injection rate : 405 sm3/day• Water Injection price : $ 0 / bbl• Oil produced price : $ 60 / bbl• Water produced price : $ 10 /bbl• Discount rate : 0• Three different permeability cases:
04/19/23 NPCW 2009NTNU
Reservoir Simulator
04/19/23 NPCW 2009NTNU
Mass balance
Darcy’s Law
Saturation Equation
Pressure Equation
Non-optimized Results
04/19/23 NPCW 2009NTNU
PART 1 Optimization
04/19/23 NPCW 2009NTNU
Optimization Methods
Reactive Control Shut-in well with water cut above some threshold
Results (Cont’d)Saturation profile without Uncertainty (800 days)
Saturation profile with Uncertainty (800 days)
Conclusions A new production optimization technique has been presented.Optimization proces based on Binary Integer Programming has beensuccesfuly applied and gives improvement in Net Present Value. BinaryInteger Programming gives more benets in the sense of NPV improvementthen regular Reactive or Proactive Control. Binary Integer Programming is a robust optimization technique undergealogical uncertainty such as permeability distribution. The optimizationprocess also showed that water saturation at breakthrough was observed tobe more uniformly distributed across the reservoir after the optimizationprocess as compared with the unoptimized case. The scope for improvement depends on the type of heterogeneity in thepermeability field. Because the NPV performance of the optimal waterflood depends less on geological features than that of a conventional waterflood, the scope for improvement partly depends on the performance ofthe conventional water flood. The scope for improvement depends on the relative magnitudes of the oilprice and the water cost, and on the length of the optimization window.
04/19/23 NPCW 2009NTNU
Recommendations
The effects of capillary pressure, compressibility, and gravity were notinvestigated in this study.
Results obtained in this study may therefore only be representativefor situations were gravity and capillary effects are relatively small. Gravity maypositively or negatively affect the sweep efficiency. The scope for improvement andthe shape of the optimal control functions may thus change if capillary or gravityforces are signicant. Therefore, their exact effects should be investigated.