Numerical simulation of reservoir souring in chalk reservoirs

The fourth law of ecology states “there is no such a thing as a free lunch”, or in other words, every gain comes with a price. Seawater flooding of hydrocarbon reservoirs is not an exception. Though seawater injection increases oil recovery, it transforms the natural anaerobic ecosystem of hydrocarbon reservoirs in many ways, for examples by introducing sulfate as a new electron acceptor and reducing temperature. A consequence is the so-called souring process, the dominating activity of sulfate-reducing microorganisms (SRM) and production of toxic, corrosive hydrogen sulfide (H2S). Various competing electron acceptors such as nitrate and perchlorate have been proposed and used to inhibit or alleviate SRM activity. These electron acceptors can be reduced microbially through many different metabolic pathways, some of which can compromise the efficiency of treatment measures. In order to reduce the environmental footprint of seawater injection, non-ishothermal bio-reactive transport models have been developed to predict microbial reservoir souring and to design mitigation plans. Former numerical models (i) have employed various implicit assumptions, (ii) are not able to detect which metabolic pathways prevail under a given physiochemical condition, and (iii) for a given metabolism, merge the metabolism rate of all species into a cumulative rate. As such, aiming to better predict and to remediate environmental impacts of seawater flooding and subsequently reduce the long-term risks, this thesis reveals some of the unknowns in reservoir souring modeling.
This thesis includes seven manuscripts. The first two manuscripts employ batch and flow experiments to understand the underlying reactions of microbial reservoir souring. In the 1^st manuscript, a workflow is established to determine the fate of nitrate in nitrate-based reservoir souring mitigations. Dissimilatory nitrate reduction to ammonium is found to be the main nitrate reduction pathway. Additionally, it is shown that nitrite inhibition is the major mechanism of reservoir souring prevention by nitrate. In the 2^nd manuscript, various mitigation measures including nitrate, nitrite and perchlorate injection are assessed based on their efficiency in remediating souring. Nitrite injection is shown to be the most effective method to suppress reservoir souring. The remaining five manuscripts utilize numerical models and focus on the rate of metabolic reactions, considering the fact that microbial metabolisms in petroleum reservoirs are catalyzed by multiple microorganisms and not a single microbe. The 3^rd manuscript  investigates the reliability of production water samples for souring studies. It shows that there is a significant difference between the sulfate reducing microbial community of the produced water and that of the reservoir. Therefore, utilizing only the produced water samples may be misleading in reservoir souring studies. The 4^th manuscript suggests a methodology to find kinetic parameters of an equivalent strain such that it characterizes a microbial guild (a group of microorganisms with a common metabolism). It demonstrates that cardinal models may not characterize the temperature dependency of a guild. The 5^th manuscript suggests a novel relationship that can characterize the effect of guild shifts on the maximum specific growth rate in batch experiments accurately. The 6^th manuscript illustrates that using the proposed relationship reduces the absolute normalized difference by one order of magnitude. The 7^th manuscript evaluates the error associated with assuming a constant growth yield and ignoring the minimum catabolic Gibbs energy yield that microorganisms require to maintain transmembrane electric potentials. 
The results of this thesis significantly improve the state-of-the-art of reservoir souring models and clearly highlight the currently applied errors associated with implicit assumptions used in reservoir souring studies. The most important learnings are: (i) souring rates should not be derived only from produced water samples, (ii) rate equations that describe a single microbe does not represent reservoir souring, (iii) growth yields are not constant, and (iv) there is a thermodynamic limit on souring rates. The obtained results from this thesis can be used to adopt laboratory procedures and numerical equations to prevent flawed souring estimates and subsequently improve microbial souring predictions.