No abstract

Interference testing although primitive in terms of its introduction and idea to the petroleum industry, still stands to this day as one of the most cost effective and efficient ways of establishing communication between wells and determining the reservoir transmissibility in the region. This paper discusses the methodology and results obtained from a four month pressure data acquisition campaign for a transient interference test performed in a carbonate reservoir known as Marrat, in the Giant Burgan field of Kuwait. The Marrat long term interference test was conducted around a water injector pilot with distances as far as 0.9 km at the subsurface locations between the injector and producer wells. Therefore, the interference test was used to evaluate the transmissibility between the injector and the nearby observation wells. The producer wells were shut-in for the entire length of the test so as not to create any disturbances that could hinder the interpretation processes. After conducting this test, a better understanding of the subsurface uncertainty as well as communication between the wells was highlighted. Other objectives were added to the tests which were to determine the water bank distance from the injector, as well as to describe the least resistive path that the water prefers to travel in. The tests showed that not all the wells responded to the pressure pulse, and therefore the assumption that a fault was isolating one of the wells. One of the main conclusions was a strong directional transmissibility that was at first associated with a high permeability corridor corresponding to the depositional environment. The other conclusion was the orientation of the fracture plane which could cause this high directional transmissibility. A comparison and integration of the acquired pressure data with a separate geologic stochastic model was constructed and discussed in this paper. Based on the integration work of the interference test and the geologic study it was therefore concluded that a secondary recovery using water flooding would be beneficial and necessary for sustaining Marrat reservoir production in the long term based on the location of both producer and injector wells.

The Wara reservoir has been producing for over 60 years and its pressure has slowly decreased over the years, now below saturation pressure in some structurally-high areas where gas cap has increased in size compared to very small initial gas caps in these areas. A peripheral, water injection project is being considered to maintain the pressure above bubble point and improve oil recovery from the flank areas. However, limited information is available concerning Wara reservoir heterogeneity. Shut-in of all Wara producers provided an "once-in-a-lifetime" opportunity to carry out a fieldwide pressure data acquisition campaign. Over a period of six months (Nov. 2006-May 2007, 127 static bottom-hole pressure (SBHP) surveys, 26 pressure buildup (PBU) tests (including buildup tests for 2 active wells of the interference test program) and 3 interference tests were conducted. Each interference test involved one active well and 3-6 observation wells. This paper describes a systematic methodology to select wells to test on a fieldwide basis, test design exercise, data acquisition program execution, and observations and conclusions reached from this data set.Over the duration of the campaign, an increasing pressure trend was observed in almost all wells slated for PBU and interference tests. This observation was made possible due to multiple SBHP recordings in these wells. The linearlyincreasing pressure trend had to be taken into account in the analyses of the buildup tests to avoid interpreting wrong boundary conditions. Wellbore dynamics also initially affected some pressure buildup tests and were later circumvented by the use of a downhole shut-in tool (DHST) to limit the phase segregation effect. The transient analyses revealed different flow regimes, from infinite-acting radial flow, to dual-layer and radial composite with or without sealing faults. From the SBHP data, maps have been made to assist in identifying compartments. One interference test showed anisotropy which could be critical in the implementation of a water injection project. Overall, the interpretation of the data from this extensive data acquisition campaign has shed light on reservoir heterogeneities that have been integrated with changes made in a fullfield, "Wara only", history-matched simulation model.

Interference testing, although primitive in terms of its introduction and idea to the petroleum industry, still stands to this day as one of the most cost effective and efficient ways of confirming communication and evaluating reservoir properties between wells. Similarly, a pressure build-up is one of the most accurate ways of estimating dynamic reservoir parameters surrounding the well, providing that the shut-in of the well is allowable. On the other hand, a drawdown test is not usually recommended due to the instability of the flow rate, and hence, the uncertainty in the parameter estimation when analyzing the transient of the pressure drawdown. In this project, due to production constraints a drawdown test was run for the active horizontal well as a substitute to the pressure build-up. It was therefore decided to couple the drawdown test with an interference test so as highlight the subsurface uncertainties. In order to achieve these objectives, careful design and operational coordination between the different asset teams and contractors is crucial to obtain interpretable and useful data. Water production was observed in some of the nearby wells, and therefore communication between the horizontal well and the surrounding wells needed to be verified. The main objective of this project was to evaluate the reservoir parameters and connectivity surrounding the important horizontal well. In this test, the horizontal well was the active well in a five well interference test. The results of the test indicated different pressure behaviors seen from the observation wells corresponding to the pulse created by the horizontal well. Communication was established in some of the wells, whereas, faults were also verified in the surrounding regions. In addition, the drawdown analysis of the horizontal well showed all the flow regimes that relate to a horizontal wells' signature as well as boundary behavior which coincide with the interference test results. The results of the drawdown analysis indicate the possibility and accuracy of conducting a pressure transient analysis using this method without being constrained with production objectives, and hence not shutting the well in.

Interference testing is the oldest but still the most effective way of establishing communication between wells and determining the reservoir transmissibility. However, data can be difficult to interpret and the results can be misleading. Fortunately, simple steps can be performed to validate the data and obtain first estimates of the formation parameters. We demonstrate this methodology for an interference test performed in the Greater Burgan field in Kuwait.A pilot project was started to understand how to successfully inject water in the Wara reservoir. Seven wells were drilled in an area away from the existing wells: one injector at the center of a 250 m-radius hexagon formed by six producers. An interference test was performed between the injector and the producers. The main objective of the study was to evaluate the transmissibility between wells and the permeability anisotropy in the formation. In five of the producers, the target sands were oil bearing, whereas surprisingly, the same sands were water bearing in the sixth well. Consequently, a second objective was added to the study: to check whether the sixth well was in communication with the other wells and to determine the origin of the water.The tests showed that all wells responded to the pressure pulse, including the sixth well, thus refuting the assumption that a fault was isolating it. The fall-off analysis of all the wells highlighted the presence of boundaries, a finding that was consistent with the fluvial depositional environment. Moreover, the analysis showed that the channel was narrowing near the sixth well. Therefore, we could hypothesize that the sixth well had been drilled in a zone with perched water trapped by the channel boundaries. A few weeks after the test, the oil cut started to increase in that well, confirming our hypothesis.The findings from this pilot project proved the efficiency of waterflooding as secondary recovery method and were used to design the pressure maintenance program.