Performance Optimization
Stochastic Model for Well Production Optimization
- Client
- Major Oil Operator
- Country
- Italy
- Year
- 2006-2015
- Industry
- Oil & Gas
- Service
- Operations & Asset Performance
For a major oil operator, kwantis added a stochastic stage to production optimization, with the input uncertainties described by statistical distributions fitted on field data and production tests.
The challenge
Reservoir behavior and complexity bring many uncertainties into production optimization modeling. The modeling is used to assess the confidence level of the optimized plant configuration, and the impact of uncertain inputs on the expected production increase.
Our approach
kwantis added a two-stage analysis to the optimization cycle:
- Generate an “optimized field configuration” from “Most Likely” data for all uncertain parameters (mostly pressure/flow rate curves).
- With that plant configuration locked, combine the probability distributions of each uncertain parameter in a Monte Carlo simulator.
The input distributions were fitted on past field data and production tests, and correlation coefficients were calculated to establish the relationships between the variables.
The outcome
The statistical definition of the inputs significantly improved the quality of the analysis. The optimized configuration now comes with a confidence level based on the uncertainty of its inputs.
