Mini-coreflood experiments validate in vitro community control
To evaluate the effect of our top-down engineering approach on oil recovery, we applied the same treatments to a miniature coreflood system. The mini-coreflood (MCF) design (Figure 5A-C ) was used to simulate microbial enhanced secondary oil recovery from representative core material to evaluate the effects of the carbon and micronutrient supplements in parallel. In addition to pressure generated by the microbes in the MCF, we evaluated the composition of the microbial community and the produced oil. After initial oil recovery from brine flooding was completed, 9:1 mixtures of effluent and molasses-nutrient supplements were injected into the MCFs and allowed to sit for 10 days in an anaerobic environment to seed the core and incubate the growing microbes. Similar to our in vitromicrocosms, we found that molasses was able to induce microbial growth and activity, generating up to 26 PSI. Nitrate-treated floods generated the most pressure while the molybdate and combination treatment generated less pressure (Figure 5D ). We suspect that the suppression of H2S and the limitation of sulfate-reducing bacteria account for the differences in pressure generation between the molybdate-treated cultures and the untreated or nitrate-treated cultures. Because increased pressure applies more force on the trapped oil deposits, the amounts of oil recovered trended with the pressure produced (Figure 5E/F ). We also found that the microbial community in the MCF was distinct from our in vitroscreening, reflecting the contribution of the high petroleum hydrocarbon concentrations in these samples. MCF microbial communities drastically shifted toward Bacteroides under the molybdate and combination treatment whereas the untreated and nitrate-treated MCFs were dominated by Lachnoclostridium (Figure 5G ). As at the end of thein vitro microcosms, both Bacteroides andLachnoclostridium were prominent members of the final MCF microbiome (Figure 4, Figure 5G ), but these MCF trials contain significant amounts of the oil hydrocarbons that can be used as an additional carbon source. While the V4-V5 16S rRNA gene region amplified here is not specific enough to identify the species of these genera, microbes belonging to the Lachnoclostridium group have been found to degrade some hydrocarbons and some reduce sulfate to H2S.47–49 Similarly, we observed increases in the abundance of organisms belonging to the genusPseudomonas which is not a prominent producer of gases and has been linked to naphthalene,50 normal paraffin,51 and polycyclic hydrocarbon degradation.52,53 As shown in (Figure 5E/G ) the proportion of n-paraffin dropped from ~35 %w/w to ~25 %w/w of the relative abundance whilePseudomonas sp. increased in abundance in the molybdate and combination MCFs. This condition suggests that Pseudomonas sp.degraded around 30% of the long chain n-paraffins (C14 – C29) in a period of 14 days (Supplemental Figure 7A, Supplemental Table 1 ). Although isolates were not identified here, Pseudomonas sp., such as P. proteolytica, P. xanthomarina, and P. aeruginosa , are expected to induce n-paraffin degradation starting from n-tetradecane (C14) by alkane hydroxylases systems.53–55 Certainly, biodegradation of long chain n-paraffins is recognized to have a positive influence in oil recovery processes by means of reducing the viscosity of the crude oil while increasing its fluidity.56 The relative abundance of monocyclo-paraffins increased from ~30 %w/w to ~40 %w/w in total corresponding to the increment of the C9 – C17 fraction for all treatments (Supplemental Figure 7B, Supplemental Table 1 ), likely due to degradation of the n-paraffins. However, no specific trend was observed for the abundance of iso-paraffins. Aromatics such as naphthalene also decreased in abundance with treatment and the presence of Pseudomonas sp. (Figure 5F/G ). Species such as P. mendocina, P. putida, P. fluorescens, P. paucimobilis, P. vesicularis, P. cepacia, P. testosteronei, P. aeruginosa, and P. stutzeri have been reported to induce naphthalene biodegradation.52,57,58 In contrast, the triaromatic fraction (C17) increased as a response of the supplements here applied. The relative abundance of dicyclo- and monocyclo-paraffins increased when these treatments were applied, likely because of n-paraffin degradation, and thus, changed the overall composition of the produced oil (Figure 5E/G, Supplemental Table 1 ). While the results presented in Figure 5 represent one trial of these treatments, an additional replicate treatment at a different time (Supplemental Figure 8A-C ) showed n-paraffin degradation for all treatments. However, the same trend for the aromatic polycyclic hydrocarbon fraction is not seen in the replicate likely because the microbiomes of the produced water used to seed these wells were different (Supplemental Figure 8D ). Regardless of these specific differences, we observed reproducible control of the oil well microbiome as a function of nutrient supplementation to alter produced oil paraffin profiles while limiting H2S production and pressurizing the reserve.
Our results demonstrate a clear correlation between pressure generation and oil recovery in MEOR processes. While the H2S-limiting treatments did not substantially enhance pressure generation in the presence of hydrocarbons and confined geometry, we were able to validate the role that molybdate plays in controlling souring and corrosion by oil well microbes. However, field trials and economic analysis are needed to determine the viability of our process relative to current surfactant-based oil recovery processes. The modest improvements in oil recovery may be offset by the increased operating costs needed for corrosion maintenance, surfactant production, and microbial control via more expensive ammonium quaternary disinfectants currently used.19 Moreover, we demonstrated that the specific oil compositions were a strong function of microbial activity, which could be controlled via top-down engineering. Further optimization of the nutrient formulation, such as increased amounts of molasses or other nutrients, may enhance pressure generation or alter the microbial dynamics so that more oil can be recovered and specific hydrocarbon compounds can be enriched or depleted. In agreement with the in vitro findings, these results suggested that we can control the microbial population in small scale oil recovery settings and can use that to concurrently modulate oil composition.