A generalized analysis of partitioning interwell tracer tests

A partitioning interwell tracer test (PITT) is a method for estimating oil volume and/or oil saturation in the swept zone between a set of injectors and producers in a reservoir. One of the methods for analyzing PITTs is the method of moments, which is based upon calculating the first temporal moment of the tracer concentrations in the produced fluids. PITTs have many advantages over other methods for estimating oil saturation by measuring over a much larger volume than a single well tracer test and a well log. It is especially important to know the remaining oil saturation as accurately as possible before applying enhanced oil recovery methods. PITTs also provide valuable information on swept volumes between wells, flow paths, and breakthrough times. A very general derivation of the method of moments applied to PITT data is presented in this dissertation. This derivation shows that the method of moments can be used for three-dimensional, heterogeneous reservoirs under very general conditions. The general derivation and its verification with numerical simulations shows that the method is not limited to residual oil saturation as generally assumed, but can be extended to mobile oil saturation (or any multiphase flow problem). PITTs in naturally fractured reservoirs are an extreme example of heterogeneous reservoirs that can be analyzed by the method of moments, although the time to conduct such tests can be generally very long. For this reason, the concept of natural tracers was investigated and analyzed. The technique of using natural tracers is based on the idea of measuring a naturally residing petroleum organic component such as organic alcohols and acids. Since natural tracers originate in the oil itself, its use can be less expensive and more environmentally friendly than the use of injected chemical or radioactive tracers, and can take less time to produce a useful signal. The synthetic tracer data in naturally fractured reservoirs as well as to singleporosity heterogeneous reservoirs are generated using numerical simulators. These data were analyzed under a wide range of reservoir conditions using both the method of moments and inverse modeling using a program developed at TAMU.