The effect of brine composition, concentration, temperature and rock texture on zeta potential and streaming potential coupling coefficient measured in sandstones and sandpacks

Measurements of the streaming potential component of the self-potential (SP) have been proposed to monitor subsurface flow in a number of settings. Numerous studies report laboratory measurements of the streaming potential coupling coefficient at laboratory temperatures in rock samples saturated with simple NaCl and KCl brines at low ionic strength. However, temperatures are considerably higher in many subsurface settings, such as deep saline aquifers, geothermal fields, and hydrocarbon reservoirs; moreover, natural brines are often significantly more saline and contain a wide variety of ionic species. We report measurements of the streaming potential on sandpacks of controlled mineralogy, and intact sandstone core samples, saturated with high salinity natural and artificial brines at elevated temperatures. We measure streaming potential using an experimental set-up that incorporates in-situ measurements of saturated rock conductivity, brine temperature, brine pH, brine electrical conductivity, pressure difference and voltage at temperatures up to 150oC. We find that the streaming potential coupling coefficient and interpreted zeta potential are negative and decrease in magnitude with increasing temperature. Measurements of brine pH show significant decrease with increasing temperature and we suggest this reduces the mineral/brine surface charge and, hence, the zeta potential. Such changes in pH must be accounted for when matching theoretical models to experimental data and when interpreting the streaming potential component of subsurface SP measurements in field settings. The temperature dependence of the zeta potential is consistent between four different natural sandstone samples, and sandpacks composed of two different sands. Increasing the content of multivalent cations in the brine causes a decrease in interpreted zeta potential, consistent with published data at low salinity.