How to Make a Good Reservoir Model: It’s Not the Software, It’s the Design (G036)

Mark Bentley: TRACS International Consultancy and Langdale Geoscience.

How can you tell the difference between a ‘good’ reservoir model and a ‘bad’ one? This short, focused class is designed to draw out the common reasons for ‘good’ and ‘bad’ outcomes, under the premise that models add value only when they add clear value to business decisions. The theme throughout the event will be the overriding issue of model design and the five areas of common error: model purpose; selection of elements; use of determinism and probability; model scale; and uncertainty handling. Advice will be given on how to review models, what questions to ask the model builders, and how to determine whether the output from models can be relied upon and used to support decisions. The course will close with a set of questions to ask yourself and others, suitable for reference in peer reviews or assists.

Classroom version: A 1-day course comprising a mix of lectures, case studies and exercises. The manual will be provided in digital format and participants will be required to bring a laptop or tablet computer to follow the lectures and exercises.

Virtual version: Two 4-hour interactive online sessions presented over 2 days. A digital manual and exercise materials will be distributed to participants before the course. Some reading and several exercises are to be completed by participants off-line.

Fundamental. Designed for people who want an update or refresh on working with reservoir models without having to spend a week out of the office. The class will provide an opportunity for learning, inspiration and discussion with other modelers.

You will learn to:

1. Explain the common causative factors for modelling ‘disappointments’.
2. Define model purpose and explain the use of framing.
3. Understand the fluid sensitivity to selection of model elements.
4. Describe techniques for handling small-scale detail in large models.
5. Be able to select between techniques for quantifying uncertainty.
6. Implement QC tips to evaluate your (and other people’s) models.