A drilling prospect with several potential pay zones is modeled. Analytical arguments involving conditional probabilities reveal that geological dependence has no bearing on either the expected number of successful layers or the expected combined reserves. Monte Carlo simulation sheds light on the dispersion of the reserve distribution. A case study using an offshore Louisiana prospect with three layers illustrates the principles described. The general approach outlined in this model of layered pay zones can be applied to other types of prospects having multiple components that may be related.
When engineers and geoscientists examine a new drilling prospect, two important questions arise:
I. What is the chance of success, P(S)? That is, if we had 100 identical prospects like this one, how many would yield reservoirs capable of generating hydrocarbons? The fraction of successes is regarded as the probability of success. This number can be 5% or less for rank wildcats and 90% or greater for exploitation or development prospects. Strictly speaking, we are concerned here with geologic success or discovery as opposed to commercial success, where the amount of producible hydrocarbons justifies further exploration and development.
2. What is the size of the reserves associated with this prospect? How many barrels of oil equivalent (BOE) or Bcf of gas will this prospect generate if it is successful? Incidentally, we make no distinction here between reserves and estimated ultimate recovery (EUR). Readers sensitive to this issues may replace the word reserves throughout the paper with EUR.
In its simplest form, the proposed wellbore will penetrate a single layer that represents a potential reservoir. Often, the first question is approached by considering several key geological components, called chance factors, estimating the probability of success (i.e., adequacy or existence) for each and multiplying these probabilities together to get the desired probability. Thus we might assert
Different geoscientists posit anywhere from three to seven factors. We will use five factors in our model, but any number would do so far as the logic is concerned.
Regardless of the number of factors used, one assumption is paramount: the chance factors are independent of one another. That is, the existence of a Source rock has no bearing on the existence of Seal.