novel digital technology and methodology enables...
TRANSCRIPT
JOGMEC TECHNO-FORUM 2018
Novel Digital technology and methodology enables subsurface & surface integration
for the Oilfield Optimization
Toyo Engineering Corporation November 27 , 2018
- 2 -TOYO ENGINEERING CORPORATION ©2018
Market Forecast : How technology unlocks oil and gas resources
⇒ Innovation technology is being welcomed to unlock oil & gas.⇒ Optimization of Operation will become viable option (eg: Middle East)..
Reference : 2018 BP Technology Outlook
Technically Recoverable Oil & Gas Resources Technology advances to 2050 could
increase recoverable oil reserves by around 50%, compared to around 25% for gas.
- 3 -TOYO ENGINEERING CORPORATION ©2018
Market Forecast : Potential volume increases and cost reductions through Technology
Reference : 2018 BP Technology Outlook
Technology can play a major role in improving access to oil & gas and in reducing the costs of production.
The potential reduction in average lifecycle costs will be ~30% for oil and gas resulting from technology advances to 2050.
The lifecycle costs of different oil types and gas resources vary considerably.
- 4 -TOYO ENGINEERING CORPORATION ©2018
Collaboration with Baker Hughes
Upstream Midstream Downstream
Subsurface & Surface Integration for Oilfield Development Explore Digital solutions utilizing GE Predix, a unique cloud-based platform
System Integrator EPC Upstream Business
Technology Provider Subsurface Digital
- 5 -TOYO ENGINEERING CORPORATION ©2018
Digital PlantPlant Owner
User Interface
Actual Plant
Predictive Maintenance
Real Time Mirror Plant
Vertical Integration with ERP
Engineering Digital Twin
Virtual Plant
E/O/M/B Service
Data/Record
Virtual Plant
DX-PLANT Digital Solution Scheme
- 6 -TOYO ENGINEERING CORPORATION ©2018
EPC : Digital Twin Portfolio
- 7 -TOYO ENGINEERING CORPORATION ©2018
Project Management and Field Implementation
Reservoir Characterization and Modeling
Development Planning for Field and Facilities. Mitigate the risks.
Reservoir Management and Optimization
Start of Field Development
Application Study of IOR, or 2ndary
, or 3rdary Recovery Application Flooding Study with Feasibility Check
Model Development with Production Model & Facility
(Study & FEED)Production Mode Monitoring
Abandon of Field
A continuous process to assist oil and gas companies optimizethe economic performance of oil and gas fields
Upstream Business : Total Support during Project Cycle
- 8 -TOYO ENGINEERING CORPORATION ©2018
Big Data Approach for Field Management
Applications
Data Driven Model
Sparse events and no perfect fittings. Correlation with statistics can be
made, but it sometimes shows poor. Forecast can be done without much
causality.
Machine Learning is used to • extract functional relationship• estimate missing data
Blue = Collected data. Red / Green = Estimated data. Poor correlation with statistical models
Data + estimated data + physical model = prediction
Time
Machine learning
Time
Para
met
er
Physics Based Model
Physics-based models capture variation with causality.
Not all data available to build model Time consuming process
Para
met
er
Time
Casing (Pipe) ModelReservoir Model Pump Model Tubing Model Pressure Gradient
@ Surface
Flow Rate, Q
Head
Efficiency
Horse Power
Best Efficiency Point
⇒ Hybrid Model with Machine Learning can speed up prediction with causality.
Reference : BHGE
- 9 -TOYO ENGINEERING CORPORATION ©2018
Production Optimization Injection Rate CO2 Concentration WAG Schedule Draw Down (ESP)
Surface Facility Operation Optimization Daily Production Short Term Production Mid Term Production
Subsurface/Surface Monitoring Pressure & Temperature Oil/Gas/Water/CO2 Rate CO2/Water Breakthrough Formation Stress
Update Control
Calibrate Model
Analytics
Validate Model
Physics-based ModelData Driven Model
Reservoir Simulation with Hybrid Model
Prediction
Probabilistic learning
AI (Deep learning)
Deep domain models
Integrated Dynamic Production Optimization
Background image reference : NETL