Impact of Biodegradation - Consideration in the Polymer Flood Model
A. Behr, S. Mukherjee and D. Prasad
Event name: IOR 2017 - 19th European Symposium on Improved Oil Recovery
Session: Poster Introductions 2
Publication date: 24 April 2017
Info: Extended abstract, PDF ( 2.45Mb )
Price: € 20
Polymers are susceptible to degrade under reservoir conditions due to impact of various factors. It is well known that synthetic polymers such as partially hydrolyzed polyacrylamide (HPAM) are more prone to mechanical and chemical degradation. However, biological degradation by microorganisms is deemed to be the primary property changing mechanism in case of biopolymers. Thus, for a sound technical and economic viability using polymer flooding, it is important to include this possible phenomenon to complete the analysis of the biopolymer behavior in the reservoir. Mathematical modelling of oil recovery by polymer injection involving biodegradation has so far not been investigated and documented extensively in the literature. In this paper, two numerical approaches as well as a distinct analytical solution are presented which allow quantitative analysis of the biodegradation impact on the polymer flooding. The analytical model assumes radial flow with the polymer degradation rate proportional to its concentration (decay model). For the first numerical approach, this kinetic model is generalized for realization on a real reservoir model. The other numerical approach is more sophisticated and considers explicitly biopolymer together with bacterial population in-situ. In this case, the model distinguishes clearly between the planktonic and the sessile bacterial communities. Biodegradation process, the mass exchange between polymer molecules and bacteria, is described by the chemical reaction equations. We adapted a CMG STARSTM for numerical solution of the model which was applied to interpret a pilot test with biopolymer. A reasonable history match was achieved for the polymer injection well using bottomhole pressure as the primary control parameter, although this was not considered as unique. Our primary focus lies on the characteristic degradation time, which is an explicit input parameter in the first model. We estimated the characteristic degradation time in the model from the outcomes of the multiwell and multiple huff and puff tests. However, the challenge lies in the estimation of this value, due to the complex kinetic of the model in the second approach.