Kern River Steam-Foam Pilots

Abstract

Summary Shell conducted two steam-foam pilots in the Kern River field, one on theMecca lease (1980–85) and the other on the Bishop fee (1982–86). The pilotsconsisted of four contiguous inverted five-spots covering 12 and 14 acres [4.8and 5.7 ha], respectively. The Mecca pilot began after 9 years of steam soaksand 10 years of unconfined steam-drive; the Bishop pilot was preceded by 19years of steam soaks and 1 year of steamdrive. Steam foam in both pilots wasgenerated by continuous injection of 250 B/D [39.7 m3/d] per pattern [coldwater equivalent (CWE)] of 50% quality steam, with 4 wt% NaCl and 0.5 wt%AOS-1618 surfactant in the aqueous phase and 0.06 mol% N2 in the vapor. Introduction The Kern River oil field is in the eastern San Joaquin Valley about 4 miles[6.4 km] north of Bakersfield, CA. The productive sand intervals (Sand J, M, and Q) were deposited in the Pliocene and Pleistocene as a high-energyalluvial-channel complex. Both pilots (Fig. 1) are near the northeast andeastern (updip) boundaries of the Kern River field (Fig. 2). Compared to theheart of the field, the pilot areas are characterized by a higher content ofpore-filling clays, lower oil gravity and higher viscosity, and lower net/grosspay ratios. Tables 1 and 2 give concise, chronological pilot summaries. Geology The steamdriven Sand M in the Mecca pilot (Fig. 3 and Table 3) is dividedinto the Upper, Main, and Poor M sands. The Upper M is about 15 ft [4.6 m]thick and is separated from the Main M by a 3- to 5-ft [0.91- to 1.52-m] clayor clayey-silt layer that is seemingly continuous across the pilot area, withthe exception of the northeastern producer, Mecca Well 27. Below this claybreak, there is a 40-ft [12.2-m] -thick sand package, the Main M, which has nosignificant silt or clay breaks. The Poor M comprises the remaining 25 ft [7.62m] between the Main M and the Sand M base. No definite breaks separate the PoorM from the Main M. In some wells, the Poor M grades from a good-quality sand toa poor-quality pebbly sand mixed with silt and clay. In other wells, the Poor Mconsists of pebbly sand, silts, and clays randomly mixed throughout theinterval. These silt and clay layers do not appear to be continuous from wellto well. Average gross thickness for Sand M in the Mecca pilot is 83 ft [25.3m], and net pay thickness is 67 ft [20.4 m]. The remaining 16 ft [4.9 m] can besplit between poor sand (7 ft [2.1 m]) and clay or clayey silts (9 ft [2.7 m]). The geometrical pilot area is 11.6 acres [4.7 ha]. Because steam foam invadedthe Poor M, the net pay thickness was adjusted to 74 ft [22.6 m], increasingthe original oil in place (OOIP) in the geometrical pilot area from 1,190,000to 1,400,000 STB [189 × 10–3 to 223 × 10–3 stock-tank m3]. The latter value isthe upper bound on the volume of oil originally in the Mecca pilot. Sand Q inthe Bishop pilot (Fig. 4) is a 100-ft [30.48-m] -thick sand package dividedinto several layers by 0 to 3 clay or clayey-silt barriers that are continuousacross at least a single pilot pattern and by 0 to 3 discontinuous barriersthat are randomly distributed throughout the pilot area. Average pay thicknessof good sands in the Bishop pilot is 65 ft [19.8 m] and total thickness ofbarriers is 15 ft [4.6 m], and there is 20 ft [6.1 m] of poor-quality sand. TheOOIP in the geometrical pilot area (14 acres 5.7 ha]) is 1,500,000 STB (238 x10-3) stock-tank m3]. Cation exchange capacity (CEC) is 8 meq/100 g in MeccaSand M and approaches 9 meq/100 g in Bishop Sand Q. These large CEC's arecaused by a high content of montmorillonite and illite (85% in calcium andmagnesium form initially). Oil in the Tank The Kern River pilots were designed to evaluate the performance of steamfoam when applied either very late in a steamdrive (10 years at Mecca) or veryearly (1 year at Bishop). Production and injection histories of the pilots aredetailed in Table 4. In evaluating the pilots performance, we are interestedonly in the incremental oil production resulting from improved volumetric sweepby steam foam (observed in both pilots), and possibly from reduction ofresidual oil saturation (ROS) in the foam-swept zones (none with AOS-1618). Weare disregarding any acceleration of oil production, such as the early responseof the Mecca pilot immediately after the start of foam injection. An unconfinedsteamdrive was run during 1970–80 at Mecca. Such a drive in a dipping reservoirshould produce more oil than a corresponding confined drive, and possibly evenmore than the OOIP because of gravity drainage. JPT P. 496⁁