This case history will discuss how MARCITsm polymer gels were used at injection wells to correct channeling problems in a waterflood, so that additional oil could be swept toward producing wells and recovery improved. As a result of the polymer gel project, oil is up, water is down, and 36,000 to 50,000 barrels of incremental oil have been recovered to date. Project payout was achieved in less than ten months using an oil price of $10 per barrel.
The 160-acre Hardwicke-University Unit is in the McElroy Field of the Permian Basin in Upton County, Texas. The unit consists of eight active injection wells and ten active producing wells (#1, #3, #4, #6, #7, #10, #11, #15, #16, and #22) arranged in irregular patterns on 10-acre spacing (Fig. 1.). The unit produces from the San Andres/Grayburg dolomite formation at depths ranging from 3,500 to 4,000 feet. Net reservoir thickness is approximately 250 feet, and average porosity is 10%. The permeability contrast in the reservoir is high, since vugs and fractures lie adjacent to much tighter matrix rock with average permeability of 3.0 md. Oil gravity is 31.2° API with a bottomhole temperature of 110° F.
Original-oil-in-place (OOIP) is estimated at 18–23 million barrels. Approximately 13% (2.9 million barrels) of OOIP was recovered during primary operations, with an additional 150,000 barrels (less than 1% OOIP) produced from secondary recovery, for a secondary-to-primary-recovery ratio of 0.05.
This waterflood has not produced much secondary oil, and it is believed that reservoir heterogeneity has been the primary reason for the low recovery. Sweep efficiency has been limited by water cycling through high permeability rock between injection and producing wells; movable oil, trapped in matrix rock, has been by-passed. Since the highly permeable rock in this reservoir has probably been swept of secondary oil, continued cycling of water through these areas would not be effective in sweeping the tighter rock that contains the bulk of the remaining secondary reserves.
Waterfloods that exhibit strong channeling between injectors and producers can benefit the most from injection-side polymer gel treatment. A good treatment candidate exhibits the following:
low secondary oil recovery,
high injection rates and low pressures,
initial, short-lived oil response at producers followed by rapid water increase, and
significant volume of mobile oil still in place. This waterflood satisfies these requirements.
Water channeling is most common in reservoirs that have a high permeability variation (Kv) and/or an adverse mobility ratio. This paper will focus on the issue of heterogeneity since it is the primary problem with this waterflood. The degree of negative impact on sweep caused by water channeling can be predicted by calculating the Dykstra-Parsons Kv from cores. For example, a Kv approaching zero indicates a homogeneous reservoir, and sweep problems would not be expected. Conversely, a Kv of one suggests that the reservoir is extremely heterogeneous. As a rule of thumb, if the Dykstra-Parsons Kv exceeds 0.55, low oil recovery due to sweep problems may be expected.
So, how much oil is by-passed when an adverse Kv exists? Figures 2 and 3 show that a Kv of 0.69 and 0.80 can be translated to mean that 50% of the injection water will flow through only 16% and 9%, respectively, of the "pay." Cores taken from this reservoir indicate a Dykstra-Parson Kv of 0.75, which is considered to be adverse and in need of sweep improvement technology.