This paper describes results from a series of comparative core floods and static compatibility tests examining the differences in laboratory test procedure, scale inhibitor (SI) returns and modeling approaches for non-aqueous and aqueous scale inhibitor treatments. Two types of non-aqueous system, one ethylene glycol based and two oil soluble products each containing penta-phosphonate scale inhibitors, were investigated. Detailed compatibility and injectivity tests were carried out before core flooding and special consideration was required in the treatment process due to their hydrophobic property. To understand the scale inhibitor transport and retention mechanisms for these non-aqueous products, comparisons were made with the relevant aqueous applications in terms of scale inhibitor return performance, flow back permeability, possible formation damage and wettability conditions which might account for any post-treatment changes. In addition, possible approaches to mathematically modeling these core floods were studied. This paper will focus on the application of a partitioning model in a standard reservoir simulator. An alternative two-phase mathematical model for such systems, which includes both inter-phase partitioning and adsorption has been described in detail in another recent publication.1
All core floods were performed using outcrop Clashach sandstone material rather than reservoir cores, and hence the advantages of deploying non-aqueous treatments over the conventional aqueous treatments would not be significant. However, the experimental and modeling results help to capture the features of these non-aqueous systems in an understandable way. Results confirmed the existence of complicated phase behavior during the core floods using two oil soluble products. Examination of the core after flooding using an environmental scanning electron microscope (ESEM) indicated decreased water wetness following the two oil soluble scale inhibitor treatments compared with the aqueous treatment. Numerical modeling results show that the behaviour of this system is most consistent with the assumption that the SI in the non-aqueous system is only slightly soluble in the oil phase during the oil post flush.