Summary Extensive pressure buildup testing was carried out in the gas-lift wells of the Falah field offshore Dubai. These tests were to gather reservoir information for use in a three-dimensional (3D) reservoir simulation model. Pressure testing was carried out with conventional Amerada bombs, downhole memory recorders, and continuous monitoring by means of an electric line unit with a Geophysical Research Co. (GRC) gauge. We also attempted to use a downhole shut-in tool with surface pressure readout. pressure readout. Log-log plots of pressure change vs. time displayed two different periods of wellbore storage in some wells as a result of a mechanism that occurred following shut-in. This type of behavior has not been reported in the literature and leads to a very complicated interpretation of the test data. The data were analyzed by type-curve matching techniques to identify the interpretation model. Radial flow was observed in only two wells. Other wells showed dominance of wellbore storage in spite of long shut-in periods-as long as 3 weeks. This was caused in part by expansion of annulus gas into the tubing part by expansion of annulus gas into the tubing immediately after shut-in. The gas expansion caused wellbore fluids to be injected in the reservoir and appeared as the first period of wellbore storage. The second period of wellbore storage occurred when the flow was period of wellbore storage occurred when the flow was reversed. This mechanism was modeled analytically by writing the superposition equation for a single well with wellbore storage and skin. Recognizing this mechanism, we designed a special test with reduced production rate that proved successful in providing interpretable data. This paper presents the results of buildup testing in three wells. The buildup shapes show marked differences caused by changes in wellbore conditions. Recognizing this type of buildup behavior is very useful indesigning the tests,cautioning against data that may appear as a straight line to Homer plot, andshowing inadequacy of short-term buildups to gather pressure data for history matching purposes in field-scale pressure data for history matching purposes in field-scale simulation models. Introduction A pressure buildup survey program was undertaken in the gas-lift oil wells of Falah field to supply data for a detailed simulation study. An electric line unit with GRC gauge was used to record the buildup in Well FA-5 for about 3 weeks. One week after the survey had started, the remaining wells were shut in conjunction with scheduled maintenance for the gas-lift compressor. Before the wells were shut in, downhole memory recorders were lowered in four additional wells, which were shut in for 7 to 10 days. Data from all Falah wells except Well FA-6 showed a slope that increased after the generally expected period of wellbore storage when a log-log plot of deltap vs. deltat was prepared. This increase in four out of five wells surveyed prepared. This increase in four out of five wells surveyed showed that wellbore storage was a real phenomenon rather than malfunctions in the pressure recording equipment. With the exception of Well FA-5, the subject wells had been surveyed upon initial completion and showed perfectly normal buildups. Thus, the recent buildup perfectly normal buildups. Thus, the recent buildup shapes were caused by reservoir pressure decline resulting from depletion and current wellbore pressures. It is theorized that as a result of low flowing bottomhole pressures (BHFP's), gas from the casing pushes the pressures (BHFP's), gas from the casing pushes the liquid into the formation upon shut-in, which results in two different periods of wellbore storage. Previous authors have hinted at this type of mechanism but did not show data to demonstrate it. Well FA-5 provides an excellent example of it. To overcome the flow-reversal mechanism, we decided to repeat the buildup survey in Well FA-5 with a downhole shut-in tool. Unfortunately, this approach had to be abandoned because of mechanical failures. We then decided to select another well, FA-4, and to stabilize it at a reduced rate in an effort to raise the BHFP and to reduce, if not eliminate, the reversal of flow after shut-in. The data obtained in this manner, although still dominated by wellbore storage, provided interpretable results. Three examples of buildup tests are included in this paper. paper. The first example discusses Well FA-5 data and demonstrates the flow-reversal mechanism, which was analyzed as such and as a changing-wellbore-storage phenomenon by type-curve matching. phenomenon by type-curve matching. The second example illustrates tests on Well FA-4 upon initial completion (Oct. 1978), Nov. 23, 1981, and Dec. 23, 1981. The third test on this well was done after the well was stabilized at a reduced flow rate. The third example includes data from Well FA-6, which showed normal buildup in Aug. 1979 and during the July 1981 survey. JPT P. 466