Gas production from the Frontier formation at 18,300-ft depth in the Frewen Deep #4 well, eastern Green River basin (Wyoming), was uneconomic despite the presence of three sets of numerous, partially open, vertical natural fractures. Production dropped from 360 Mcf/D to 140 Mcf/D during a 10-day production test, and the well was abandoned. Examination of the fractures in the core suggests several possible reasons for this poor production. One factor is the presence of mineralization in the fractures. Another more important factor is that the remnant porosity left in the fractures by partial mineralization is commonly plugged with an overmature hydrocarbon residue (pyrobitumen). Reorientation of the in-situ horizontal compressive stress to a trend normal to the main fractures, which now acts to close fracture apertures during reservoir drawdown, is also an important factor.
The Frewen Deep #4 well is located in Sweetwater county, southwestern Wyoming (Section 13 of Township 19 North, Range 95 West). The target of the well was natural gas from sandstones of the Frontier formation (Fig. 1) at a depth of approximately 18,300 ft. The Frontier formation consists of Cretaceous-age sandstones and shales. The main reservoir sandstone is about 40-ft thick at this location, with thick over- and underlying shales.
Amoco Production Co. formed the Frewen Deep Unit in 1988. Its purpose was to evaluate the hydrocarbon potential of the Cretaceous sedimentary section in a 16 sq miles area on the south flank of the Wamsutter Arch. This arch trends WNW-ESE and divides the eastern Green River basin into two subbasins, the Great Divide basin to the north and the Washakie basin to the south (Fig. 2A).
The Cretaceous sedimentary section is commonly productive in stratigraphic traps along the crestal portion of the Wamsutter Arch, as in the Echo Springs-Standard Draw and Wamsutter fields. The Frewen Deep Unit was formed to explore for deeper production in the Lakota formation. The initial unit well, the Frewen Deep #1, was drilled to a total depth of 19,299 ft on a southward-plunging, fault-related anticline. It was completed in the Lakota formation, but extended production tests from this zone indicated noncommercial rates. Shows had been observed while drilling through the Frontier formation to the deeper horizon, and this zone was targeted for testing. Unfortunately, the wellbore became mechanically unusable during the course of moving uphole to test the Frontier. Mechanical problems associated with the great depth, problems with the completion fluids, as well as problems with the casing integrity in this well were grounds for the decision to evaluate the formation in a completely new well. The Frewen Deep #4 well was drilled as a replacement, offset 600 ft from the #1 well (Fig. 2B).
Much of the Frontier formation in the #4 well was cored with good recovery (86 ft), even though the core contains numerous partially mineralized vertical natural fractures. The fractures have obvious open porosity at depth (Fig. 3), with bridgings of mineralization holding open apertures locally up to 5 mm wide. Four fracture sets, based on character and strike, were differentiated in the core. These included three sets of irregular but numerous natural fractures, designated F1, F2, and F3 in order of their formation (based on observed cross-cutting relationships). The 86 ft of core had been slabbed and extensively sampled before our study, and the fractures themselves are commonly multistranded. Both of these factors make exact fracture counts difficult to obtain. Pervasive fracturing of the core suggests that the reservoir must be highly fractured, although the actual data set consists of approximately 10 F1 fractures, eight F2 fractures, and two F3 fractures. Fracture heights along the vertical axis of the core range from a maximum of about 4 ft for the F1 fractures down to several inches for F3 fractures. A fourth set of fractures consists of 30 regularly spaced, coring-induced1 petal fractures striking parallel to each other and to the F3 fractures.
Gas in the drilling mud and the presence of open fractures seemed to promise significant gas production, but the initial production rate was not high and declined precipitously to an uneconomic level. We analyzed the natural and coring-induced fractures in the Frewen core during this study to assess the possible reasons for the low and declining production despite the presence of significant natural fracturing in the reservoir. This paper documents the conclusions from the core study and also offers an interpretation for the origin of these unusual fractures.
Well History and Reservoir Properties.
The Frewen Deep #4 well was spudded on 18 October 1990 and reached a total depth of 18,600 ft on 3 March 1991. Three separate conventional cores (totaling 86 ft recovered) were taken through the Frontier formation. Horizontal Dean Stark air permeabilities were measured at each foot in the sandstone core; 61 measurements yielded an average permeability of 0.007 md (range 0 to 1.23 md), an average porosity of 3.7% (range 0.8 to 7.1%), and a flow capacity of 1.7 md-ft. Geophysical logs were collected over the objective interval, including induction and neutron/density suites.
Mud weight at total depth was in excess of 15 ppg, indicating a pressure of approximately 14,489 psi (minimum) at the reservoir level. Shows of gas requiring the use of a gas buster to de-gas the mud began at 18,225 ft and continued during coring operations. Shows periodically supported 10- to 20-ft (estimated) flares. Below 18,380 ft, the mud did not require de-gassing to remain manageable and control the well.
Multiple sets of casing were set in anticipation of high pressures: we set 13 3/8-in. surface casing at 2,358 ft, 9 5/8-in. intermediate casing at 10,835 ft, and 51/2-in. casing at 18,114 ft before initiating coring operations. A 5-in. liner set from 18,114 to 18,593 ft completed the casing of the well. Each of the casing and liner strings was cemented in place and an acceptable bond was achieved.
Completion operations began on 23 April 1991 when the well was perforated from 18,316 to 18,344 ft with 6 shots per foot, 6,000 psi underbalanced. The well did not flow. Swabbing was required to achieve a 15 to 20 Mcf/D flow rate for 7 days. Subsequently, we performed a CO2 breakdown, with 110 tons CO2 pumped at 8.5 bbl/min into 14,400 psi tubing pressure. The well flowed back CO2 and gas at a rate of 500 Mcf/D (>25% CO2) and was shut in preparatory to flow testing and bottomhole pressure buildup.