Tests from compartmentalized and layered reservoirs can be difficult to interpret for connected volume and connectivity. The problem comes from the timing of the response from "primary" and "secondary" drainage volumes, either as well-compartment and distant compartments or as dominating and dominated layers, with reduced communication between volumes posing more challenges. Although simple material balance considerations can be used to estimate the total volume, the difficulty lies in determining individual volumes and connectivity. The term "connectivity" is used here in a broad sense to cover effects of differential depletion, a critical element of recovery from compartmentalized reservoirs.
Results from the paper are applied to both pressure transient tests and to production data (rate data) analyses based on decline curves and type curve methods. For pressure transient tests the distinction between extended buildup behavior and single-rate extended drawdown behavior, e.g., based on deconvolution data, is of particular importance, with different boundary characteristics picked up from the two data types. For production data analyses the main challenges are associated with effects of differential depletion and the concept of boundary dominated flow both for closed systems and for systems with pressure support or influx. Similar challenges apply to compartmentalized systems produced under simple depletion and pseudosteady state conditions.
In order to separate facts from fiction, all examples are based on analytical and numerical models. This is important since critical parameters such as pressure communication and flow between compartments and between layers must be known in order to determine how these effects can be properly investigated and understood based on pressure transient and well performance data. The basic analytical model used in this paper is fully transient with flow based on local gradients rather than on differences in average pressures between compartments which tend to yield data with delayed and exaggerated response.