Economic method improves conformance in CO2 flood

As COEOR becomes a more mature technology, operators are experiencing excessive CO₂ production, similar to mature waterfloods that produced excessive amounts of water. This conformance issue of injected fluids bypassing oil-rich zones can be the result of several contributing factors ranging from poor cement to fracturing and even vuggy features caused by rock dissolution. As with any remediation method, properly defining the problem prior to making a recommendation is of paramount importance.

Kinder Morgan is one such operator experiencing an aging CO₂ flood in its SACROC unit in Scurry County, Texas. This unit encompasses more than 50,000 acres and was the first CO2 flood when it was initiated in 1972. However, the facilities in the field have limited handling capacity, and as injection wells begin to channel CO₂ to offset producers rather than process oil, the injection efficiency is reduced resulting in significant handling cost and lost recovery opportunity for the injected CO₂.

The conformance issue at SACROC was studied by Kinder Morgan engineers and service company engineers. Ultimately, it was decided that the majority of the CO₂ cycling through the reservoir was moving through the fractures and areas of vuggy porosity. The reservoir is made of a Middle Pennsylvanian through Early Permian carbonate that was deposited on the Horseshoe Atoll. The marine to near-shore depositional environment led to stratigraphic complexity and high-permeability fractures that exist in some parts of the reservoir. Furthermore, because of the high vertical permeability exhibited as both fractures and the anisotropy of any vertical barriers, correcting the conformance problems with near-wellbore solutions like patches and cement squeezes led to short-term solutions as the highly mobile CO₂ cross-flowed back into the offending zones.

FIGURE 1. This map shows producing wells in the SACROC unit, Scurry County, Texas. (Images provided by TIORCO LLC and Kinder Morgan)

The complexity of this issue led the operator and the service providers to develop a method of identifying and attempting to quantify the volume of large channels. This method included evaluating anomalously high injectivity indices in injection wells and high-rate CO₂ producing wells but most importantly the gas-oil ratio (GOR) and water-oil ratio (WOR) as a function of the volume of fluid injected. Because the company incorporates a water-alternating-gas injection, this GOR and WOR vs. injection volume calculation can be made several times over the history of the pattern. By calculating the volume of injected fluid prior to the injected fluid’s breakthrough, a reasonable estimate of the channel volume could be made.

In injection pattern 141-3, as shown in Figure 1, offset wells were making extremely high volumes of CO₂ and water while seeing very little oil response. After using the methods described previously, it was determined that the channel volume of the thief zone in the pattern was somewhere between 40,000 bbl and 80,000 bbl. This volume was much too high to address with traditional void space-filling materials like cement, so the operator opted for injecting a large volume of crosslinked polyacrylamide.

Crosslinked polyacrylamide gels have been used throughout the world to reduce water flow through fractures in producing wells and injection wells in waterfloods. One system using chromium acetate, MARCIT, has been used in fractured CO₂ floods in the Rocky Mountains in fields such as Rangely and Wertz.

However, a reservoir with highly fractured carbonate and possible solution caverns under CO₂ injection presents an extreme situation in terms of differential pressure for the gels to withstand.

FIGURE 2. The increase in wellhead pressure (green) as polymer concentrations (blue) are increased is displayed in this graph. The two pressure anomalies present are a result of equipment malfunction.

FIGURE 3. The results of gel treatment are presented on this time-rate plot for all producing wells in the 141-3 injection pattern.

To combat the extreme differential pressure in the near-wellbore region, large volumes of gel consisting of high molecular weight polymer and a strong near-wellbore cap were recommended. The large volume was recommended to prevent cross-flow of subsequently injected fluids back into the fractures, and the high molecular weight polymer was used to increase injection viscosity in an attempt to overcome gravity segregation to the bottom of the zone. The cap used, CAPIT, was a combination of different molecular weight polyacrylamide products (medium and low) that, when blended together, form a stronger gel than can otherwise be generated while maintaining injectivity with conventional pumps. Both systems are injected as a viscous fluid that gels in situ and are capable of reducing fractures from multi-Darcy permeability to micro-Darcy permeability.

For the 141-3 pattern, the initial design was to inject 20,000 bbl of MARCIT gel ranging from 4,000 ppm to 10,000 ppm followed by 50 bbl of CAPIT gel with a concentration of 30,000 ppm. The goal of the treatment was to inject the entire volume under the fracture pressure to avoid the gel penetrating out of zone while also observing a steady increase in injection pressure indicating sufficient fill-up of the high-permeability features and ensuring that subsequent injection will be diverted to lower permeability features. However, pressure increases slowed significantly, and the volume of the 7,000 ppm and 8,000 ppm gel injected was greater than designed. In total, 28,025 bbl of gel were injected.

The response to this treatment was significant. Where prior to the treatment CO₂ injection led to very little oil production, after the treatment, CO₂ injection cycles resulted in rates in excess of 300 b/d of oil.

When plotted on a GOR vs. cumulative oil plot, an incremental 60,000 bbl of oil reserves are added to the pattern production. With an overall treatment cost of US $400,000, these reserves were added at a cost of $6.67/bbl. Subsequent to this treatment, Kinder Morgan has repeated this strategy several times and achieved similar results.

Treating conformance problems in-depth with large volumes of polymer gels is expensive, but when appropriately applied, it can be highly successful. In fractured CO₂ floods where offset wells have the potential to produce at very high oil rates when CO₂ processes highly oil-saturated rock, blocking offending fractures with correctly formulated gels can lead to economic projects and extend the life of the flood.

References available.