The setting of pretest parameters for the acquisition of pressure and mobility data with wireline formation testers has largely been left to field engineers operating the tool and their knowledge of local conditions. If the test sequence is not successful the pretest parameters, such as the number of drawdowns, pretest volume, rate and duration, are either adjusted to achieve a better result or the tool is moved to a nearby location where the chance of success is deemed to be better. This strategy "works" if the information necessary to make an assessment is readily available. During drilling operations, where telemetry rates are usually very low and test times are necessarily short, this approach is not likely to be acceptable.
This paper describes pretest sequences that optimize the process of acquiring pressure and mobility data during formation tester operations without compromising the quality of the data. All sequences employed consist of three phases: identification when the mud cake has been breached and the sandface pressure is below formation pressure; a short "investigation" phase in which the formation is interrogated to obtain an initial estimate of its properties; and one or more "measurement" phases in which the best estimates of the formation pressure and mobility are obtained. The measurement phases use the information acquired during the investigation and/or previous measurement phases to design an optimal test sequence such that a stabilized sandface pressure is achieved at the end of each measurement phase subject to an overall time constraint. This is done in the downhole tool by solving a small- scale optimization problem. Associated with the sequences are quality indices derived from the results of the individual phases that indicate how well the procedure has succeeded and the validity of the data acquired.
The test sequences have been implemented on standard wireline formation tester tools and a formation pressure while drilling tool. Field trials with the new sequences are discussed and the results are compared with standard operating sequences.