This reservoir engineering presentation will show the production results from over 1,300 horizontal wells drilled throughout the world, and will compare these results to predictions. Because of the impact of heterogeneity and the difficulty in predicting permeability fields, it is shown that production rates are a difficult thing for reservoir engineers to predict, with an average error bar of ±50% for any single well's prediction. These are similar odds one might get at a blackjack or craps table!
It is commonly accepted that petroleum reservoirs have properties that are “heterogeneous”; however, the “h-word” can mean different things to different geoscientists. The explorationist may think about heterogeneous depositional environments that span scales of tens or hundreds of miles, while a production geologist may be more concerned about heterogeneities that occur within a field on the scale of a mile or two. Log analysts help determine heterogeneity within a few inches or feet of the wellbore, while core measurements and thin sections literally show us detail at the microscopic or pore level. Reservoir engineers are most interested in heterogeneity at the inter-well scale, where pertinent data sources may be (transient) well tests and long-term production information.
What is “heterogeneity”? Loosely defined, it might be considered as variations in natural or induced reservoir properties. The most common heterogeneities are variations in porosity, permeability, and pay thickness. But are you sure your well has low-permeability, or is it damaged? Is the rapid early pressure decline indicating limited reservoir, or is it simply transient fluid flow in a tight reservoir? Is your 3,000' horizontal well contributing from the entire section or just from a few hundred feet? Is the variation in GOR or WOR due to different compartments or different connectivities to the gas-cap or aquifer? Does any of this really matter?
Yes, it matters, and what ties everything together is the concept of “scale-dependent heterogeneity”. Heterogeneity exists at all scales of measurement, and each scale of measurement impacts performance in a different way. Is it reasonable that core analyses indicate 100 md average permeability, whereas pressure transient tests indicate 30 md and long-term production tests indicate 10 md? In many cases, an example of which will be discussed, the answer is “yes, all these measurements are reasonable and correct, they simply apply at different scales”.