Estimation of original hydrocarbons in place is a fundamental responsibility of a practicing reservoir engineer and is essential for constraining production forecasts and reserves estimates within physically reasonable bounds. In traditional reservoir engineering, most non-empirical reserves estimation techniques are based on single-phase pseudo-steady-state flow theory. More recently, many researchers have tried to extend these single-phase techniques to multiphase flow through the introduction of pseudo-pressure and pseudo-time transformations. For multiphase flow, definition of these transformations is not unique, and different researchers have proposed alternative methods of computing pseudo-functions with varying degrees of success. A major difficulty with the pseudo-function approach is that a relative permeability-saturation model must be selected for the system of interest; in our experience, system relative permeability curves are seldom known. This work proposes a flowing material balance method that does not depend on generation of pseudo-functions, nor upon knowledge of system relative permeability relationships. The method is based on mass material balance, and only utilizes production and PVT data. We present complete details of its derivation along with its application to both synthetically generated and actual field data sets. We also compare results of our analysis to results obtained from interpreting multiphase production data using single-phase techniques. We show that application of single-phase analysis methods to multiphase production data can result in large errors in estimated original hydrocarbons in place.
The primary job of any practicing reservoir engineer is to create production forecasts that are used for booking reserves, calculating economics, etc. The most widely used tool for generating these forecasts is empirical decline curve analysis (DCA) based on the work of Arps (1945). The practice of using empirical decline curves to forecast production is as much an art as it is a science; while Fetkovich (1980) showed that certain flow regimes and reservoir types are associated with specific decline curve parameters, the explicit identification of flow regimes and reservoir types is difficult and often uncertain. Furthermore, in practice, many combinations of decline curve parameters can yield "good" historical production data matches while resulting in large discrepancies in long-term production forecasts. This necessitates a wealth of experience on the part of the investigator to select appropriate decline curve parameters; the result is a production forecast that is, at best, only valid under current operating conditions, and at worst, highly misleading with respect to future production. To better constrain production forecasts and EUR estimates within physically reasonable bounds, the practicing reservoir engineer needs an independent estimate of original hydrocarbons in place (OHIP).
