Model upsizing for integrated evaluation of polymer degradation and polymer flooding lab data

Abstract

In oil recovery operations through polymer flooding, it is essential that the polymer maintains its stability until near the producer well, avoiding degradation phenomena. The present paper seeks to assess the representation of the degradation effects associated with polymer flooding for heavy oil recovery through one-dimension simulation models with different sizes.Base models were developed using data from the literature and previous lab work. The models consider adsorption, accessible pore volume, residual resistance factor, polymer concentration and non-Newtonian effects. Aiming at including degradation effects, a methodology of model size increase was developed, allowing to combine the evaluation of long-term dependent effects with displacement mechanisms. A sensitivity analysis was performed for the degradation, considering four levels of polymer half-life time. The lab-scale models were submitted to an upsizing procedure, generating models 100 times and 10.000 times bigger.The outcomes showed that the higher the level of the reaction frequency factor, the faster the degradation process. In the presence of original and degraded polymers in the porous media, we observed two advance fronts. The front of the degraded polymer advanced faster for a higher reaction frequency factor, while the advance front of the original polymer was delayed. We analyzed the results in terms of cumulative produced water, cumulative produced oil, average pressure, oil recovery factor, cumulative mass of the injected polymer, cumulative mass of the degraded polymer, water phase saturation, and water phase viscosity.The procedure for model upsizing was matched using a linear scaling factor. The method permits parameters found in laboratory experiments to be simulated in field-scale. Normalized results allowed comparing the impact of the main physical phenomena related to polymer flooding in the mentioned results, independent of the model size.