Prediction of Oil Production With Confidence Intervals*

 Application of Prediction Theory

Basic Idea: I

Basic Idea: II

Basic Idea: III

Basic Idea: IV

New Idea:  Arrival Time Error Models

Arrival Time Model is Simple, Robust

Illustration:
Posterior reduces choices and uncertainty

New Result:

Problem Formulation

Simple Reservoir Description

Ensemble

 Upscaling

Slide 15

 Examples of Upscaled,
 Exact Oil Cut Curves

Slide 17

 Bayes Theorem

Errors and Discrepancies

   Example

Slide 21

Slide 22

Slide 23

Model Reduction:

Slide 25

Slide 26

Slide 27

Slide 28

Confidence Intervals

Slide 30

Slide 31

Arrival Time Error Analysis

Covariance Matrix: 10x10 Scaleup

Slide 34

Slide 35

Slide 36

Slide 37

Slide 38

Slide 39

Confidence intervals for arrival time error model (%)

Arrival time error model vs. solution value model: confidence intervals (%) for s = 0.6 and 10x10 scaleup

Summary and Conclusions

References