Application of Quantitative Structure-Property Relationship (QSPR) Method for Chemical EOR Surfactant Selection

摘要

The selection of high performance surfactants for chemical EOR is a challenging and time consuming task. A surfactant formulation, typically a blend of at least two surfactants must be developed for each case study. A tool to pre-select suitable surfactants would thus be highly valuable. In this paper, we describe the development of a quantitative structure-property relationship applied to the selection of surfactants for chemical enhanced oil recovery. A correlation is drawn between surfactant structures and optimal salinities, i.e. the salinity which corresponds to a minimum in interfacial tension. A comprehensive and coherent database has been generated using a high-throughput screening robotic platform and industrial products belonging to different families of surfactants: olefin sulfonates, alkyl ether sulfates and alkyl glyceryl ether sulfonates. This database has been built for specific reference conditions (temperature, oil, brine hardness). Industrial surfactants, most often constituted of a variety of molecules, have been carefully analyzed in order to identify predominant species. The structures of these compounds have then been drawn using molecular design tools, and molecular descriptors were generated for the whole set of amphiphiles. Finally, various statistical approaches have been used to develop multi-linear regressions correlating combinations of the most relevant molecular descriptors with the experimentally determined optimal salinity of surfactant mixtures. Our results indicate that a strong correlation exists between the surfactant structure and its optimal salinity. A limited set of descriptors can be used to predict this critical property with predictive models. These models can then be used to select promising existing products as well as to identify candidate raw materials or products for industrial surfactants development. We also demonstrate the ability of our models to predict optimal salinity of surfactant blends typically used in chemical EOR. Future developments will be focused on extrapolation of these models to the prediction of other application properties for chemical EOR (e.g. interfacial tension value) and to broaden the application domain to a wide range of conditions (temperature, brine composition, type of oil).