Forecast Optimisation Under Uncertainty

Understanding model uncertainty is paramount to the success of our operations. Understanding model uncertainty is paramount to the success of our operations. This is why we spend a large amount of time performing extensive analysis, which in most cases leaves us with a number of potential cases of varying probability.

It is then very difficult to assemble these results and generate forecasts from these potential cases. How can you tackle the challenges in hand without compromising, when bound by tight deadlines? tNavigator has many unique, advanced tools for handling uncertainty, meaning you do not have to settle on a single scenario, or reduce the number of parameters:

• Forecasting functions allow you to easily and automatically generate forecasts from an ensemble of cases.
• In-built optimization tools enable you to go beyond simple forecasts, applying an optimization algorithm over any number of potential cases and automatically generating a curve of best cases – known as a pareto front.
• Forecast and optimization under uncertainty gives you confidence about uncertainties, to decide the best possible course of action.

Integrated Uncertainty and History Matching Workflows

Classic uncertainty and history matching practices have limitations as to what is modifiable, and fails to take into consideration uncertainty of the model building process itself.

tNavigator’s integrated workflows address these issues by defining uncertainty and matching variables within the model building process. This allows you to take into consideration every possible source of uncertainty and generate models with different structural, property, fluid and dynamic properties.