A new flood monitoring technique is presented that models both the capture cross section (sigma) and ratio/porosity responses of a pulsed neutron logging tool using known reservoir, hydrocarbon and flood agent properties. Open hole logs and pulsed neutron logs acquired before, during and alter flooding are combined to determine individual hydrocarbon, saltwater and flood agent saturations. This new model uses total neutron migration length and direct ratio computations to compute the flood agent volume fraction. Improved accuracy of the computed fluid saturations and the ability to handle specific flood agents are two enhancements to flood monitoring provided by this new approach. Accuracy is especially important because often the results from a limited number of monitor wells are used to project the sweep efficiency and economics of the flood process for an entire reservoir. This new method is applied to a CO2 flood and the resulting fluid saturations for oil, saltwater and CO2 are analyzed and discussed. These results are compared with those based on the older excavation effect method.
The progress and efficiency of flood projects have been estimated for years by various methods. Uncertainties in the mixing formulae for log responses have often limited flood monitoring to a qualitative nature. We have developed a flood analysis model to quantify the changing fluid saturations of an active flood project using open hole and pulsed neutron logs. This new model is called TMDFLOOD. In order to quantitatively monitor a flood project, information about the chemical and physical properties of the specific reservoir, flood agent and hydrocarbons must be obtained. Production and core data are important sources of this information. The monitoring process then consists of open hole logs and pulsed neutron logs recorded prior to flooding to define log responses at initial (base) conditions and pulsed neutron logs run during the life of the flood project (monitor). Suggested procedures are given in Table 1.
Most floods are long term projects that last from several months to several years. During the life of a flood, downhole mechanical conditions can change drastically. The two exponential model utilized in the TMD system minimizes many borehole related log anomalies. However, there can be unforeseen changes in the downhole tubular goods, addition of gravel packs, or logistical problems that make it impossible to obtain logs that are consistent with the original base conditions. Particularly troublesome problems include gas entry into the wellbore or other changes in the wellbore fluid. These problems can render neutron porosity measurements ineffective and should be avoided when possible. Use of unperforated monitor wells is highly recommended. Another problem is that it may not be possible to obtain base pulsed neutron logs. In these cases, synthetic sigma and ratio/porosity curves must be computed from high quality open hole volumetric analyses and accurate reservoir parameters. If the borehole properties change enough to significantly effect the log responses, renormalization to base conditions is sometimes needed. Due to the long term nature of enhanced recovery projects, sufficient log data often becomes available to provide a good quality check on any necessary data renormalizations or environmental corrections.