This paper will explore a new statistical approach to EUR estimation and show how quantile regression can greatly simplify EUR estimation, reserves estimation, completion evaluation, and production efficiency. Two key benefits include the speed with which information may be analyzed and an increase in its accuracy. This approach supplies geoscientists and engineers with statistically valid and consistent estimates of production to then compare with geological controls in an area. An additional benefit of this method is an index of production and completion efficiency.

Today EURs are commonly computed based on one well at a time, resulting in estimates that are inconsistent and difficult to use in a broader geologic context. When using a large numbers of wells at a time, however, statistical guidance can provide new insights. These insights include:

  1. ensemble estimates of EUR,

  2. estimates for individual wells contributing to the ensemble,

  3. estimates of P90, P50, and P10 EURs and

  4. estimates of production and completion efficiency.

When quantile regression is used, EUR estimation is much faster. This allows engineers and geoscientists more time to understand the results and to integrate them with geologic models. It also enables more informed completion decisions and can help identify production problems in plenty of time to solve them. Production efficiency can be quantified by looking at the rugosity of individual production data for a well and then quickly compared to the entire group of wells in an area to identify problems. A strong correlation is demonstrated between results that are based on statistical compilation calculations and results generated from an individual well's cumulative production. Individual wells are very noisy, so fitting them is slow and results can easily vary from engineer to engineer. Statistically generated curves use much more data, so results are more robust. Thus P10, P50, and P90 EURs can be computed with significantly more confidence from a statistical group of wells treated as an aggregate, giving a quick and accurate picture of an area to be purchased or developed.

Quick and easily accessible estimates of EUR, reserves, and completion and production efficiency are invaluable for rapid assessment of acquisitions and proposed drilling programs, as well as comparisons to reservoir model results. The use of quantile regression in this approach allows consideration of all the wells in an aggregate yet still retains proper P10/P50/P90 curves as well count changes with time, resulting in a much better and longer-term estimate.

This content is only available via PDF.
You can access this article if you purchase or spend a download.