Upload
doandieu
View
259
Download
4
Embed Size (px)
Citation preview
May 16 – 20, 20162015 Gas-Lift Workshop 1
Oilfield Production Modeling & Optimization
• Kevin Wade, Lead - Oil & Gas Production System
Process Systems Enterprise (PSE), UK
39th Gas-Lift WorkshopHouston, Texas, USAMay 16 – 20, 2016
Agenda
• Overview of PSE
• Overview of PSE’s Production modeling Technology
– Focus on Oilfield Optimization Technology
• Discrete optimisation
• Case example
Private, independent company
incorporated in UK
1997
Company ‘spun out’
Acquires technology
PSE HISTORY: FROM RESEARCH TO INDUSTRY
1989 – 1997 Now
Advanced Process modeling platform
100s of person-years of
R&D with industry
Simulation & modeling,
Optimization, numerical
solutions techniques,
supply chain
London HQ Korea JapanUS NJUS TX Switzerland
Thailand Malaysia ChinaTaiwan
140+ people~60 PhDs
Background
PSE Business Sectors
Oil & Gas Chemicals &Petrochemicals
Life Sciences, Consumer & Fine Chemicals
Power & CCS
Overview of PSE’s Production modeling Technology
Overview of PSE’s Production Optimization Technology
Scope
Oilfield Production Optimization
Maximise production value from an oilfield by adjusting process and well behaviour
Decide
which wells to use
which routing to take to the surface
how much gas-lift to apply to each well
……
…within operational envelope
Scope
Well and well-network systems:
From: Sandface
To: Topside Separator
Overview of PSE’s Production Optimization Technology
Technological basis
Key elements of gPROMS Oilfield Optimizer
1. gPROMS Platform
2. gPROMS Oilfield Model Library
3. gPROMS Oilfield Optimizer
4. Interface to existing systems and data
Process flowsheeting technologies
The Sequential Modular (SM) approach
?
WATER
ETHANOL
MIXED REF-1
REF-0WGS-IN
PRODUCTS
REFORMER
HEATER COOLER
MIXER
WGS
1 2
3
4
5
67
8
9
10
f(x) = 0One large set of equations (103-106 variables)
Solved simultaneously
Process flowsheeting technologies
The Equation Oriented (EO) approach
1 2
3
4
5
67
8
9
10
Conceptually much simpler …
Complete Field Optimization
Field Optimization
• Existing tools can optimise production on
– Well level
– Manifold level and
– Field level
• Given a number of field constraints
• These focus on continuous variables
– Gas lift injection rate (0-x) MMscf/d
– Choke Delta P (0-x) psi
• We also optimise on discrete variables
– Well status (on/off)
– Well routing (pipeline A or pipeline or ….)
– Pipeline routing (riser A or riser B or …)
– Riser routing to (separator A or separator B or …)
Feb. 4 – 8. 20132013 Gas-Lift Workshop 13
Discrete Optimisation
• Traditional tools can only solve continuous optimisation
• They can be made to optimise discrete cases by:
– Brute force approach
– using GA
• Brute Force: Ok if options are small
• For I = 1 to number of wells on/off options
– For j = 1 to number of well routes
• For k = 1 to number of pipe routes
– For l = 1 to number of separators
» Call continuous solver
• Next l, k, j, i
• GA: has limitations
Feb. 4 – 8. 20132013 Gas-Lift Workshop 14
Innovate UK project
Case study performance
Solver parameter settings
RunObj Func[$m/day]
CPU [hr] Param #1 Param #2 Param #3
1 3.3467 136 1.E-04 1.E-08 1.E+062 3.2705 91 1.E-02 1.E-06 1.E+063 3.2015 66 1.E-06 1.E-09 1.E+034 3.1789 137 1.E-06 1.E-09 1.E+035 3.1742 125 1.E-03 1.E-06 1.E+066 3.1354 29 1.E-06 1.E-09 1.E+03
Best solution significantly better than
that from industry-standard tool
Quality of solution varies significantly
with solver parameter settings
Long execution times 5½ days to get best solution strongly dependent on solver settings
IS IT POSSIBLE TO DO BETTER THAN THIS ?
Best-in-Class MINLP solver
Solver parameter settings
RunObj Func[$m/day]
CPU [hr] Param #1 Param #2 Param #3
gOO 3.5071 1.6 N / A
1 3.3467 136 1.E-04 1.E-08 1.E+06
2 3.2705 91 1.E-02 1.E-06 1.E+06
3 3.2015 66 1.E-06 1.E-09 1.E+03
4 3.1789 137 1.E-06 1.E-09 1.E+03
5 3.1742 125 1.E-03 1.E-06 1.E+06
6 3.1354 29 1.E-06 1.E-09 1.E+03
Innovate UK project
Case study performance
Significantly better solution:
$58.5m/year
Significantly faster solution:
O(100) speed advantage
Performancenot dependent on
solver parameter settings
Conclusions now confirmed across a range of case studies
Best-in-Class
MINLP solver
Innovate UK project
Project outline
• Project start: 1st November 2015
• Project finish: 30th September 2016
• WP1: System scoping & specification
– Stakeholder interviews
• WP2 – Technology assessment, gap analysis & project planning
• WP3 – Development of the framework and further development of numerical solvers
• WP4 – Construction of the validation case study model
• WP5 – Industry feedback and exploitation plan
PSE is now seeking input and potential test applications from stakeholders.
Case examples
gPROMS Oilfield Optimization
Case Study 1: Optimization Onshore
• Challenge
– Optimize revenue (from both oil & gas)
– Separator: maximum gas rate constraints
– Well: Maximum liquid rate & maximum drawdown
Client: Confidential
Field type: Gas Condensate
Well count: 120
Pipeline count: 200
Separator count: 3
Routing combinations: >1 Million
Case Study 1: Optimization Onshore
Current Technology vs gPROMS results
Matched simulation results of current world-leading Optimization tool to within 0.3%.
Constraints violated
Gas rate [MMScf/day] / separatorOil rate [bbl/day]
Revenue [MM$/day]A B C
(max 1030) (max 571) (max 161)
Simulation
Current Technology
973 610 155 221,615 $ 24.08
gPROMS 974 608 161 222,235 $ 24.15
Optimization(Continuous)
Current Technology
974 555 157 215,380 $ 23.4
gPROMS 991 571 1601 224,924 $ 24.37
Optimization(Field Configuration)
Current Technology
974 555 157.2 215,380 $ 23.4
gPROMS 1030 571 161 239,570 $ 25.83
Case Study 1: Optimization Onshore
Current Technology vs gPROMS results
Increase in production of 4.4% with gPROMS ~ $1.0M/d!
Gas rate [MMScf/day] / separatorOil rate [bbl/day]
Revenue [MM$/day]A B C
(max 1030) (max 571) (max 161)
Simulation
Current Technology
973 610 155 221,615 $ 24.08
gPROMS 974 608 161 222,235 $ 24.15
Optimization(Continuous + Well
Status)
Current Technology
974 555 157 215,380 $ 23.40
gPROMS 991 571 161 224,924 $ 24.37
Optimization(Field Configuration)
Current Technology
974 555 157.2 215,380 $ 23.4
gPROMS 1030 571 161 239,570 $ 25.83
gPROMS solution into current technology
991 567 154 223,880 24.25
Case Study 1: Optimization Onshore
Current Technology vs gPROMS results
Increase in production of 11.0% with gPROMS ~ $2.5M/d!
Better utilisation of separators
Gas rate [MMScf/day] / separatorOil rate [bbl/day]
Revenue [MM$/day]A B C
(max 1030) (max 571) (max 161)
Simulation
Current Technology
973 610 155 221,615 $ 24.08
gPROMS 974 608 161 222,235 $ 24.15
Optimization(Continuous + Well
Status)
Current Technology
974 555 157 215,380 $ 23.40
gPROMS 991 571 1601 224,924 $ 24.37
Optimization(Field Configuration)
Current Technology
974 555 157.2 215,380 $ 23.40
gPROMS 1030 571 161 239,570 $ 25.83
Case Study 1: Optimization Onshore
Current Technology vs gPROMS results
Increase in production of 11.0% with gPROMS ~ $2.5M/d!
Better utilisation of separators
Gas rate [MMScf/day] / separatorOil rate [bbl/day]
Revenue [MM$/day]A B C
(max 1030) (max 571) (max 161)
Simulation
Current Technology
973 610 155 221,615 $ 24.08
gPROMS 974 608 161 222,235 $ 24.15
Optimization(Continuous)
Current Technology
974 555 157 215,380 $ 23.4
gPROMS 991 571 1601 224,924 $ 24.37
Optimization(Field Configuration)
Current Technology
974 555 157.2 215,380 $ 23.4
gPROMS 1030 571 161 239,570 $ 25.83
221,615
215,380 215,380
222,235224,924
239,570
SIMULATION OPTIMISATION (CONTINUOUS) OPTIMISATION (FIELD CONFIGURATION)
Production Rate b/d
Other gPROMS
gPROMS Oilfield Optimization
Case Study 2: Optimization Offshore
• Challenge
– Optimize Oil Production
– Riser: maximum fluid velocity speed & limited
gas lift injection gas
– Well: Maximum liquid rate & maximum drawdown
Client: Confidential
Field type: Gas Lifted Oil Field
Well count: 13
Riser count: 5
Separator count: 3
Routing combinations: > 300,000
Matched simulation results of current world-leading Optimization tool to within 1.1%.
Oil rate [bbl/day] Revenue
[MM$/day]
Simulation
Current Technology
85,254 $ 9.26
gPROMS 86,107 $ 9.36
Optimization(Continuous + Well Status)
Current Technology
90,404 $ 9.82
gPROMS 95,464 $ 10.34
Optimization(Field Configuration)
Current Technology
90,404 $ 9.82
gPROMS 105,432 $ 11.37
Case Study 2: Optimization Onshore
Current Technology vs gPROMS results
Case Study 2: Optimization Offshore
Current Technology vs gPROMS results
Increase in production of 5.6% with gPROMS ~ $0.5M/d!
Oil rate [bbl/day] Revenue
[MM$/day]
Simulation
Current Technology
85,254 $ 9.26
gPROMS 86,107 $ 9.36
Optimization(Continuous + Well Status)
Current Technology
90,404 $ 9.82
gPROMS 95,464 $ 10.34
Optimization(Field Configuration)
Current Technology
90,404 $ 9.82
gPROMS 105,432 $ 11.37
Case Study 2: Optimization Offshore
Current Technology vs gPROMS results
Increase in production of 16% with gPROMS ~ $1.5M/d!
Oil rate [bbl/day] Revenue
[MM$/day]
Simulation
Current Technology
85,254 $ 9.26
gPROMS 86,107 $ 9.36
Optimization(Continuous + Well Status)
Current Technology
90,404 $ 9.82
gPROMS 95,464 $ 10.34
Optimization(Field Configuration)
Current Technology
90,404 $ 9.82
gPROMS 105,432 $ 11.37
Case Study 2: Optimization Offshore
Current Technology vs gPROMS results
Increase in production of 16% with gPROMS ~ $1.5M/d!
Oil rate [bbl/day] Revenue
[MM$/day]
Simulation
Current Technology
85,254 $ 9.26
gPROMS 86,107 $ 9.36
Optimization(Continuous)
Current Technology
90,404 $ 9.82
gPROMS 95,464 $ 10.34
Optimization(Field Configuration)
Current Technology
90,404 $ 9.82
gPROMS 105,432 $ 11.37
Conclusions & Discussion
• Modeling
– gPROMS technology can demonstrably model the field as accurately as established production modeling tools
– Existing production system models can easily be imported
– The process and production system can be modelled in the same environment
• Optimization
– Standard gas lift / choked well (continuous Optimization)
• gPROMS Oilfield Optimization => Reliably better solutions that established production modeling tools
– Discrete Optimization (well status and routing)
• gPROMS Oilfield Optimization => Significant increase in production and/or revenue
gPROMS Oilfield Optimization
Conclusions
Equation-based modeling and Optimization
Best practice multiphase flow approaches+ validation
Thank you