SPE Members


This study analyzes the Devonian shale gas reservoirs of the Appalachian Basin and the applicability of using initial test data to assign reserves in this reservoir system. Data has been collected on over 500 such wells and recovery has been related to initial test data to assess the reliability of reserve estimation. This study concludes that there is no true linear relationship between recovery and initial test data.


When a gas well is evaluated for completion, one of the first thoughts with regard to its future is on well payout time. Once the well has paid for itself, net revenue and profitability assist in enabling further exploration and production. The payout time depends upon reserve forecasting, and reserve forecasting depends upon estimating the recoverable reserves.

Gas reserves are usually more difficult to estimate than oil reserves. On a volumetric basis there are many factors whose effect on an estimate is multiplicative. Drainage area, volume factor, and recovery factor are particularly difficult to obtain on an individual well basis and usually area averages will be substituted for actual values. Information from well test analysis is also dependent upon data derived from other tests. Analogy is often a popular choice, but with only the most preliminary reservoir data it is difficult to assess reserve potential compared with other wells. The options available to estimate reserves are limited in the early life of the well.

Additional complication is introduced in gas reserve estimation when the reservoir is tight and naturally fractured. Such a reservoir system is exemplified in the Devonian shales of the Appalachian Basin. The gas is released in a multiple-step fashion, first as free gas flushed from the open fractures, then as adsorbed gas streaming off the fracture surfaces, and finally as absorbed gas released slowly from the shale matrix. Of course, there will be overlap of the three release mechanisms, but the overall effect on the production rate is to produce an unusual hyperbolic decline curve on a production rate versus time plot. Forecasting such a curve to estimate recoverable reserves would be questionable for a new well, since it has no production history upon which to extrapolate a curve.

At this stage, the evaluation engineer reviews the available data — usually the initial shut-in pressure and open flow rate — and attempts to relate these to other wells in an effort to draw analogy. The purpose of this paper is to study the relationship of initial reservoir data to cumulative recovery in order to assess the reliability of using analogy to estimate reserves for the Devonian shales.


The area studied for this paper is depicted in Figure 1 and represents a twenty county area along the common borders of West Virginia, Kentucky, and Ohio. The area selected has similar geological, geochemical, and reservoir engineering characteristics in light of the shales' attributes throughout the extensive Appalachian Basin. Additionally, there is a wealth of information in this area, and this area is considered the most promising for future development.

P. 283^

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