Pressure Falloff Testing of Enhanced Coalbed Methane Pilot Injection Wells

Abstract

As part of the effort to interpret Amoco's enhanced coalbed methane recovery pilot, each of the four nitrogen injectors was tested with a pressure falloff test (PFOT) after approximately five months of injection. An additional test was done on the final injector at the termination of the pilot. The tests were analyzed to obtain flow capacity, wellbore condition, and average reservoir pressure for use in subsequent simulation studies. Stress effects on reservoir permeability and wellbore skin were investigated. Results from the PFOTs yielded an injectivity index which can be used to estimate compression requirements for enhanced coalbed methane recovery projects.

Discussion

Amoco's enhanced coalbed methane recovery pilot in the northern San Juan Basin consisted of four nitrogen injectors surrounding a central production well. Although four coals are present in this area, only one of them was used for the pilot. Prior to the pilot, the only pressure transient test run on any of the five pilot wells was a postfrac pressure falloff test (PFOT) on the central producer. This test yielded an absolute permeability of 32 md, a wellbore skin of +2.6, and an initial reservoir pressure of 1,313 psig. To aid in pilot interpretation, a total of five pressure falloff tests were run on the injectors. Objectives of these PFOTs were to determine effective permeability to gas, wellbore skin, and average reservoir pressure. Details of each PFOT are discussed below and are summarized in Table 1.

Downhole gauges and surface shut-in were employed in all tests. The injectors were shut in sequentially and each shut-in nominally lasted two weeks. Four of the five tests reached radial flow, were analyzed with traditional semilog methods, and were subsequently matched with either of two simulators. Both of these simulators are single-phase, analytic codes which are not strictly applicable to a coalbed which produces both gas and water. However, the high gas-water ratio of the center producer, approximately 31 mcf/stb, was indicative of a nearly dewatered coalbed and use of such simulators appears warranted.

Average reservoir pressures were calculated using the methods Mathews-Brons-Hazebroek (MBH) and Reynolds, Bratvold and Ding (RBD). Both used real gas pseudopressures based upon p* values obtained from semilog analysis.

The PFOT in the northern injector, Simon 15-3, yielded an effective permeability to gas of 4.4 md and a wellbore skin of +1.7, indicative of a slightly damaged wellbore This well had 22 ft of perforations and the upper 16 ft of the perforations were broken down. The calculated skin factor for this particular partial penetration geometry was 1.61, in good agreement with that obtained from the PFOT. Average reservoir pressure was 1,001 psig. This test showed the effect of offset wells at late time. The simulator match employed closed boundaries at 680 and 960 ft and open boundaries at 250 and 800 ft. While this combination of boundary conditions is certainly not unique, it does provide an excellent match of the test data, as seen in Figure 1.

The test in the eastern injector, Simon 15-2R, initially showed decreasing pressure, then at late time, increasing pressure which is attributed to interference from offset wells. Early time data indicated radial flow was reached and traditional semilog methods were used to obtain starting values for the history matching exercise. The history match effort included all data and resulted in an effective permeability to gas of 3 md and a fracture half-length of 21 ft (equivalent to a skin of -3.94). The match used four open boundaries at 240 ft. Figure 2 is a log-log plot showing both the pressure data and the simulated match. As seen from the figure, the match was good at early and intermediate times but poor at late time.

P. 307