There are two primary ways to determine reserve quantities: the volumetric method and decline curve analysis.
The volumetric method
This calculation is derived from the formulae –
Gas-in-Place = 43,560 * A * H * f * Sg * Bg
Reserves = (Recovery Factor) * (Gas-in-Place)
This method may appear to be fairly ironclad until we consider the uncertainties that are inherent in each of the measurements required as inputs to the equations. It relies on the accuracy of the maps made by the geologist and geophysicist, which are derived from interpretations of complex data that, for the most part, are derived from physical parameters beneath the surface. These parameters are the inputs for the areal extent of the hydrocarbon accumulation (A), the thickness of the pay zone (H), the porosity of the rock (f), the gas or oil saturation (Sg or So) and fluid composition (Bg). Interestingly, but also a mathematical reality, is that if each variable is off by only 10%, then the gas-in-place will be off by approximately 40% and the reserves could be off by approximately 50%.
Decline curve analysis
Decline curve analysis is widely accepted as the more accurate method for reserve determination, presuming ample production history is available to make the predictions. This is a little bit like trying to determine what a child will look like once he grows to maturity. When he is an infant, we might look at the size of his feet or hands to estimate his mature weight or height, or by analogy consider the height and weight of his mother and father. As he reaches the teenage years, we can be more exact in our prediction, but ultimately we only really know once he is fully grown.
An example using a tight gas sand producing well may help to understand the hazards in this method:
Based on decline curve analysis, after the first 100 days of actual production with output measured in thousand cubic feet per day (Mcfpd), this well would only show an expected ultimate recovery of about 230 million cubic feet (MMcf).
As time goes by, more production data is obtained, and decline curve analysis would predict that the well is capable of recovering more gas.
After about five years, a much clearer picture appears of what the well will actually produce.
This example helps explain the uncertainties in extrapolating decline curves to a well’s ultimate recovery or reserve quantity. In this case, the reserve prediction continued to increase with time. It is not uncommon, however, to have the opposite occur in instances such as a well prematurely ‘watering out,’ having mechanical failures resulting in early abandonment or any number of unexpected problems.