production optimization an industry perspective - … optimization – an industry perspective alex...
TRANSCRIPT
Production Optimization –
An Industry Perspective
Alex Furtado Teixeira – PETROBRAS/CENPES
Mário César Mello Massa de Campos – PETROBRAS/CENPES
Trondheim, 24 September 2013
9TH International Conference on Integrated Operations in the
Petroleum Industry
Agenda
Decision Making Process
Motivation
Expected Benefits
Adopted Strategy
BR-SiOP (Production Optimization System)
Quantifying Benefits
Main Challenges
Conclusions
Trondheim, 24 September 2013
Decision Making Process
Data Decisions
Experience
Data and Simulator:
Better understanding of the process;
Indentify opportunities;
Model the process;
Experience:
Knowledge or skills that the engineer gained over time working in the asset;
Decision Support Tools:
Search for an optimal operating point;
Evaluate a huge number of alternatives when operating under severe process constraints;
Enhance process understanding.
Decision
Support
Engineer
Simulator
Trondheim, 24 September 2013
Mathematical Optimization - Structure
Objective:
Maximize Oil Production;
Maximize NPV (Net Present Value);
Minimize Operational Cost.
Decision Variables:
Production Choke Opening;
Gas Lift Flow rate;
Well Routing;
Water/Gas Injection Flow rate.
Equality and Inequality Constraints:
Water Handling Capacity;
Liquid Handling Capacity;
Compression Capacity;
Amount of Gas Available for Gas Lift;
Pressure Constraints.
Feasible Region!
Degrees of Freedom!
Single or Multi-Objective?
Motivation
The characteristics of the wells are changing over time, which implies in a new optimal operating point.
30.0
40.0
50.0
60.0
70.0
14/1/2004 1/8/2004 17/2/2005 5/9/2005 24/3/2006 10/10/2006 28/4/2007
Test Data
Wate
rcu
t [%
]
50
70
90
110
14/1/2004 1/8/2004 17/2/2005 5/9/2005 24/3/2006 10/10/2006 28/4/2007
Test Data
GO
R [
m3/m
3]
GOR
Watercut
Trondheim, 24 September 2013
Motivation
The availability of critical process equipments is changing over time, which also implies in a new optimal operating point.
0
500
1000
1500
2000
2500
3000
3500
30/6/2012 5/7/2012 10/7/2012 15/7/2012
Date
Ga
s F
low
rate
[N
m3
/d]
Total Gas
Gas Lift
Produced Gas
Flare
5 days
Unexpected shutdown of a compressor!!!
Trondheim, 24 September 2013
Expected Benefits
Increases production (1 – 3%);
Reduces operational cost;
Fast response to abnormal situations;
Reduces time of the decision making process;
Increases robustness and reliability of the decision making process;
Increases interaction among disciplines.
Trondheim, 24 September 2013
Adopted Strategy
OPTIMIZER
New Production
Test Data
Simulator
Recommendations
Decisions
Update
Models!
Run model automatically
covering the entire operational
envelope
Tables
Actions
Valid
Models
Validate
recommendation!
Trondheim, 24 September 2013
Sec. – Min.
Production Optimization Loop:
Adopted Strategy
Simulators
Optimizer
Decoupled Strategy:
The engineers can use different simulators with the same optimizer.
Trondheim, 24 September 2013
A B C D
Adopted Strategy
Objective: Maximize Oil Production;
Decision Variables:
Gas Lift Flow rate;
Pressure upstream choke;
Gas Flaring;
Status of the wells (Opened or Closed);
Well routing (subsea manifold).
Constraints:
Compression Capacity;
Liquid Handling Capacity;
Water Treatment Capacity;
Limit allowed to the gas flaring;
Gas lift flow rate and wellhead pressure
lower and upper bound per well.
Characteristics of the scenario: Offshore
platforms with satellite oil wells.
MINLP
MILP
Piecewise
Linearization
(SOS2)
(Gunnerud, V. and Foss, B., 2009)
(Codas, A., Campos, S. R. V., Camponogara, E.,
Gunnerud, V. and Sunjerga, S., 2012)
Adopted Strategy
Production Engineers
Petrobras Network
72 servers HP Dual Xeon quad 3.0 GHz with Microsoft
Windows Server 2003 x64 SP 2.
Asset A Asset B Asset C
Computer Cluster located at Cenpes
Web User Interface
Benefits:
Improved optimization performance with parallel computing;
Easier to maintain and continuously improve the system;
Optimized number of solver licenses.
Trondheim, 24 September 2013
BR-SiOP – Production Optimization System
Trondheim, 24 September 2013
Web Graphical Interface:
BR-SiOP – Production Optimization System
Trondheim, 24 September 2013
Web Graphical Interface:
BR-SiOP – Production Optimization System
Graphs available per well:
Oil Flow Rate versus Gas Lift Downhole Pressure versus Gas Lift
Trondheim, 24 September 2013
BR-SiOP – Production Optimization System
Sensitivity Analysis – Process Constraints:
Trondheim, 24 September 2013
BR-SiOP – Production Optimization System
Sensitivity Analysis – Process Constraints:
Trondheim, 24 September 2013
BR-SiOP – Production Optimization System
Sensitivity Analysis – Process Constraints:
Trondheim, 24 September 2013
BR-SiOP – Production Optimization System
Sensitivity Analysis – Process Constraints:
BR-SiOP is an useful tool to assist engineers in the quantification of losses due to unavailability of
critical process equipments.
Trondheim, 24 September 2013
Quantifying Benefits
Case Study 1 – FPSO with 13 wells:
Total Produced Oil Flow Rate [m3/d]
17,500.00
17,600.00
17,700.00
17,800.00
17,900.00
18,000.00
18,100.00
1 2 3 4
Scenarios
With Optimizer
Without Optimizer
Total Produced Oil Flow Rate [m3/d]
Scenarios - Days
Comparison based in production historical data – normal operating scenario.
Gains ranging from 0.37 to 1.55%
Trondheim, 24 September 2013
Total Produced Gas Flow Rate [Sm3/d]
2,000,000.00
2,005,000.00
2,010,000.00
2,015,000.00
2,020,000.00
2,025,000.00
2,030,000.00
2,035,000.00
2,040,000.00
2,045,000.00
2,050,000.00
1 2 3 4
Scenarios
With Optimizer
Without Optimizer
Quantifying Benefits
Case Study 1 – FPSO with 13 wells:
Total Produced Gas Flow Rate [Sm3/d]
Scenarios - Days
Comparison based in production historical data – normal operating scenario.
Gains ranging from 0.24 to 1%
Trondheim, 24 September 2013
Total Gas Lift [Sm3/d]
1,350,000.00
1,400,000.00
1,450,000.00
1,500,000.00
1,550,000.00
1,600,000.00
1,650,000.00
1,700,000.00
1,750,000.00
1,800,000.00
1 2 3 4
Scenarios
With Optimizer
Without Optimizer
Quantifying Benefits
Case Study 1 – FPSO with 13 wells:
Total Gas Lift [Sm3/d]
Scenarios - Days
Comparison based in production historical data – normal operating scenario.
Trondheim, 24 September 2013
Quantifying Benefits
Case Study 2 – FPSO with 13 wells:
Action of the Platform
Operators
Action Based in the
Recommendation of the Optimizer
Comparison based in a real scenario – limited compression capacity.
Trondheim, 24 September 2013
Quantifying Benefits
Case Study 2 – FPSO with 13 wells:
0 600 1200 1800 2400 3000
Time (sample number)
1,450,000
1,500,000
1,550,000
1,600,000
1,650,000
1,700,000
1,750,000
1,800,000
To
tal G
as L
ift
[Sm
3/d
]
1,737,561
1,603,805
1,564,978
Time (sample number)
To
tal
Gas
Lif
t [S
m3/d
] Comparison based in a real scenario – limited compression capacity.
Trondheim, 24 September 2013
Quantifying Benefits
Total Oil Production [m3/d]
16950
17000
17050
17100
17150
17200
17250
17300
17350
With Optimizer Without Optimizer
Case Study 2 – FPSO with 13 wells:
Total Oil Produced [m3/d]
With Optimizer Without Optimizer
Comparison based in a real scenario – limited compression capacity.
Production Increased 1.18%
Trondheim, 24 September 2013
Main Challenges
Instrumentation
Regulatory Control
Advanced Control
Optimization
Planning
S.P.
The production optimization strategy presented assumes that the process is “stable”. It means that dynamic instabilities like severe slugging must be addressed by the regulatory and advanced control layers.
Production Optimization Layer
Trondheim, 24 September 2013
(OTC-24286)
Reduces variability.
Main Challenges
Simulator
(Model)
Parameters
Variables Results ? Optimizer
Uncertainty
Uncertainty Propagation:
Results
Uncertainty Uncertainty
Watercut
GOR
Reservoir Pressure
Productivity Index
Trondheim, 24 September 2013
Main Challenges
Histograma
0
5
10
15
20
25
4 7 10 12 15 17 20 23 25M
ais
Bloco
Fre
qü
ên
cia
Histograma
0
5
10
15
20
25
30
3842
4030
4218
4406
4594
4782
4971
5159
5347
Mais
BlocoF
req
üê
nc
ia
Input Parameters
Produced Oil
Histograma
0
5
10
15
20
25
5101
89
5255
73
5409
56
5563
40
5717
24
5871
07
6024
91
6178
75
6332
58M
ais
Bloco
Fre
qü
ên
cia
Gas Lift
Uncertainty Propagation:
Due to the uncertainty the optimum will be a region instead of a point.
This region can be explored by the field operator or by the well control system (closed loop).
Opportunity for the use of Robust Optimization Techniques!
Trondheim, 24 September 2013
Objective: Maximize Oil and Gas
Production;
Decision Variables:
Gas lift flow rate;
Pressure upstream choke;
Well routing in the subsea manifold;
Status of the wells (opened or closed).
MEG flowrate;
Constraints:
Gas, oil and water treatment capacities;
Pressure in the pipeline network;
CO2 content in the gas;
UTGCA gas and condensate treatment
capacities.
Collaboration:
Integrated Production Optimization System:
Main Challenges
Mathematical
Optimization
Main Challenges
Human Aspects:
Proactive x Reactive approach (continuously search for the opportunities for improvements);
Training (good understanding of what is inside of the optimizer);
Good integration between onshore and offshore team (operator is part of the optimization loop);
Without these items the gain can be 0%.
Trondheim, 24 September 2013
Main Challenges
Human Aspects: Training and knowledge dissemination.
Production Optimization:
Training courses;
Workshops;
Conferences.
Conference – Rio, May 2013.
Training Course – Rio, January 2012.
Workshop – Rio
January 2012.
Conclusion
A decision support system to assist engineers and operators in the production optimization process was developed and made available for three different assets from Campos Basin (P-54, P-50 and P-53).
The preliminary results obtained with the use of the decision support system are promising.
Two case studies with a quantification of the benefits of the decision support tool for production optimization were presented.
One of these case studies was a real example where the use of the decision support system resulted in an increase of about 1.18% in the total oil flow rate produced by the platform.
The main technological and human challenges related with production optimization were presented.
Trondheim, 24 September 2013
THANK YOU!
Contact:
Alex Furtado Teixeira
Email: [email protected]
Mobile: +55 21 8162-0774
Phone: +55 21 2162-4397