Response to Comment on "Comparison of Water Use for Hydraulic Fracturing for Unconventional Oil and Gas versus Conventional Oil".

Fracturing for Unconventional Oil and Gas versus Conventional Oil” T general perception is that hydraulic fracturing (HF) for oil production in unconventional reservoirs uses excessive amounts of water to produce energy. Our analysis of water demands over the production life of conventional oil wells as compared to unconventional oil wells contradicts the “general perception”. Results of our analysis based on oil production to date and estimated ultimate recovery in the two main oil producing plays in the U.S., Eagle Ford and Bakken, suggest that HF water use is within the lower range of water used for oil production from conventional reservoirs. However, Lampert, in this comment, suggests that one must account for potential water use during the mature production phase from unconventional reservoirs, and, in particular, enhanced recovery, which could significantly increase the currently estimated water use. Therefore, it would be inappropriate to compare HF water use at well completion in unconventional plays with water use over the life of wells in conventional plays, which include water use for primary (drilling), secondary (water flooding), and tertiary (steam, chemical, or CO2 injection) recovery. Results of previous published studies support our analysis. Studies of the Bakken reservoir indicated that the oil-wet nature of the play should “reduce the effectiveness of water flooding”. Recent simulation studies in the Bakken indicate that large water volumes could not be injected at sufficient pressure to increase oil production because of low water injectivity, limiting the ability or the value of waterflooding. Modeling analysis by Iwere et al. (2013), cited in Lampert’s comment, showed little variation in recovery with waterflooding (6.7% recovery of original oil in place) and natural depletion (6.4% recovery) for the Middle Bakken formation. This study indicated that the injectivity of the Middle Bakken and underlying Three Forks formations is very poor. Results of lab core experiments indicate that increased recovery in the Eagle Ford play using water with 2% NaOH could be attributed to dissolution; however, cores from the Barnett play were damaged when subjected to the same fluids. Similar experiments on Marcellus shale cores showed that they were too tight and did not allow any fluid imbibition. Lampert suggests that CO2 injection may work particularly well for enhanced production from unconventional reservoirs but indicates that CO2 injection often results in higher water use to oil production ratios than waterflooding alone based on the review by Wu and Chiu, 2011. However, data from Wu and Chiu (2011) refer to oil production from conventional reservoirs only and may reflect the currently common water alternating gas (WAG) approach to enhance oil recovery in which slugs of water and CO2 are successively injected. In contrast, modeling analyses in the Bakken indicate that CO2 injectivity is sufficient to enhance oil production, increasing the oil recovery factor by 13−15% after 18 years of continuous CO2 injection 3 (that is, no WAG and consequent water use). Modeling by Iwere et al. (2013) also suggests a slight increase in recovery after continuous CO2 injection. In conclusion, these studies indicate that CO2 injection, with little or no additional water, may be favorable in some plays. Therefore, high water use associated with waterflooding, CO2 injection, and steam injection related to production from conventional reservoirs, depicted in Figure 1 of Lampert, may not be applicable to enhanced recovery techniques in unconventional reservoirs. Although not explicitly mentioned by Lampert, and with an argument congruent to his, some authors have suggested that water would also be used in the mature phase of unconventional wells because of the need for refracturing wells. Refracturing water use would significantly increase the currently estimated water use. However, published literature does not seem to support the need for comprehensive refracturing of all wells, as documented in several field studies. The studies generally show an increase in estimated ultimate recovery after refracturing but also that most wells are not good candidates for refracturing. Examples include studies of ∼20 wells each in the Eagle Ford and Bakken plays, five wells in the Woodford play, 13 wells in the Barnett, and one well in the Eagle Ford. Two studies evaluated different screening tools to assess the potential for refracturing, suggesting that ∼15−20% of wells in the Barnett would be suitable for more in-depth analysis, and another study suggesting that 6% of wells in the Barnett (11 out of 188 wells) would benefit from refracturing. In summary, evaluation of refracturing is in the early phase of analysis with most studies converging to the conclusion that only a subset of wells will ever undergo the operation. Analysis of the existing data indicates that the percentage of wells that would be suitable for refracturing would be ≤20%. Without conflicting data or new research suggesting otherwise, it seems reasonable to compare HF water use for oil production from unconventional reservoirs with water use over the different stages of maturity for oil production from conventional reservoirs. In addition, if refracturing is limited to 10−20% of wells in unconventional plays, it would not change the results of our analysis as water use for oil production in unconventional reservoirs would still be in the lower range of that used for oil production in conventional reservoirs. Bridget R. Scanlon* Robert C. Reedy J.-P. Nicot Bureau of Economic Geology, Jackson School of Geosciences, University of Texas at Austin, Austin, Texas 78712, United States