Portal de Periódicos da CAPES

The Multiwell Experiment - A Field Laboratory in Tight Gas Sandstone Reservoirs

1990; Society of Petroleum Engineers; Volume: 42; Issue: 06 Linguagem: Inglês

10.2118/18286-pa

1944-978X

D.A. Northrop, K.H. Frohne,

Oil and Gas Production Techniques

Summary The U.S. DOE's Multiwell Experiment (MWX) was a field laboratory aimed atimproved characterization and gas production from low-permeability reservoirstypified by the Mesaverde Group in western Colorado. A broad spectrum ofactivities was conducted over 8 years at a site containing three closely spaced(less than 225 ft [less than 68 m]), deep (7,550 to 8,350 ft [2300 to 2550 m])wells. The results yielded insights and contributions into the technology ofgas production from this resource. Introduction New and improved technology is required to enhance natural gas productionfrom the extensive low-permeability sandstone reservoirs of the U.S. This largepotential resource has more than 600 Tcf [17 Gm3] of technically recoverablegas. More specific to this study, a resource analysis of the Mesaverde Group'stight sandstones in the Piceance basin estimated 420 Tcf [11.9 Piceance basinestimated 420 Tcf [11.9 Gm3] gas in place, with 68 Tcf [1.9 Gm3] identified astechnically recoverable. Government-sponsored tight gas research efforts to stimulate production fromthese reservoirs began in the mid-1960's. The initial focus was on thedevelopment of stimulation technology with the use of nuclear explosives toinduce fracturing. Results showed that substantial further development andpublic support were necessary. Efforts then focused onmassive-hydraulic-fracturing (MHF) tests, cost-shared with industry in severalwestern basins. Field tests were conducted in the Piceance (Colorado), Uinta(Utah), and Greater Green River (Wyoming) basins, and consisted of coring, special logging, and productivity testing followed by large hydraulicstimulations. Results from tests conducted through 1979 were disappointing anddid not improve technology or consistently enhance production. Review and analysis of this pool of reservoir and fracturing informationindicate that tight gas reservoir parameters varied widely within the threebasins. Well production performance following the cost-shared productionperformance following the cost-shared MHF's was unpredictable. The basic shortcomings were that these tests failed to separate reservoir behavior fromstimulation effectiveness and did not provide sufficient data to define thecritical factors affecting gas production. As a result, the U.S. DOE developed and operated the MWX, aresearch-oriented field laboratory, during 1981–88. The principal objective wasto obtain sufficient principal objective was to obtain sufficient informationon the geologic and technical aspects of gas production from the widespreadMesaverde Group to unlock its tight gas resource. A key feature of the MWX was three wells spaced 110 to 215 ft [34 to 66m]apart (Fig. 1). Drilling and configuration for the three wells are described inRefs. 6 through 8, and general project overviews detailing objectives, plans, and activities are in Refs. 9 through 13. Detailed core, log, and welltest datafrom such close well spacings and direct geologic study of nearby surfaceoutcrops of the Mesaverde provided detailed reservoir characterizations. Interference and tracer tests, as well as the use of fracture diagnostics inoffset wells, provided additional uncommon information on stimulation andproduction behavior. Another key feature was the synergism resulting from abroad spectrum of supporting activities: geo-physical surveys, sedimentologicalstudies, core and log analyses, well testing, in-situ stress determination, stimulation experiments with fracture diagnostics, and reservoir performanceanalyses. This paper highlights some of the insights on natural gas production fromlowpermeability sandstones and the technological contributions gained from theMWX. Employees at Sandia Natl. Laboratories and CER Corp. were the principalinvestigators at the MWX, with the DOE providing research management. Manyother indepenlent researchers added to the information pool with studies of MWXcore, logs, and pool with studies of MWX core, logs, and ether experimentaldata. This paper summarizes the results from this unique project. Description The MWX's focus was the Mesaverde formation of northwest Colorado. Thisthick sequence was deposited during the Late Cretaceous over a broad region ofthe western U.S., and contemporaneous formations are found in the Green River, Wind River, Uinta, and San Juan basins. The MWX field laboratory is in theRulison field in the east-central portion of the Piceance basin. The site is inSection 34, T6S, R94W in Garfield County and is 7 miles [11 km] southwest ofRifle, just south of the Colorado River (Fig. 1). Here, the Mesaverde formationlies at 4,000 to 8,250 ft [1220 to 2520 m] between the overlying Wasatchformation and the underlying Mancos shale. Three wells were drilled: MWX-1 to 8,350 ft [2550 m], MWX-2 to 8,300 ft[2530 m], and MWX-3 to 7,565 ft [2305 m]. More than 4,100 ft [1250 m] of 4-in.[10-cm] core-approximately 30% of it oriented-was cut with greater than 99%recovery. Numerous logging programs containing both standard and experimentallogs were conducted. The three wells are exceptionally straight, with relativeseparations of 110 to 215 ft [34 to 66 m] at depth (Fig. 1). Significant gasshows were encountered throughout the section in all three wells and mudweights as high as 15 lbm/gal [1800 kg/m3] were required to maintain wellcontrol. Fig. 2 shows gamma ray logs from the three wells over the entireMesaverde at the MWX site; the various tested zones are identified. Activities at the MWX site were conducted from Aug. 1981 to Dec. 1987. Natural productivity tests were conducted productivity tests were conducted inthe Cor-coran/Cozzette marine sandstones (Aug. 1982 to July 1983). Completestimulation experiments were conducted in the lenticular sands of the overlying(nonmarine) paludal (July 1983 to Aug. 1984), coastal (through June 1986), andfluvial (through Dec. 1988) intervals. Each experiment consisted of detailedinterval characterization, prefracture well tests, stress tests, prefracturewell tests, stress tests, stimulationrelated tests (e.g., step-rate, flowback, and minifracture), a propped stimulation with fracture diagnostics, andpostfracture well tests. postfracture well tests.