Abstract Pumping energy is a key component of the groundwater governance challenge, yet it is largely missing in the discourse on agricultural use of groundwater. A sub-category of literature studying groundwater-energy nexus tends to focus on groundwater depletion hotspots where entrenched interests and long-standing histories restrict the range of feasible energy pricing options. Using an agent based model, we estimate expected impacts of expanding groundwater irrigation under five different energy provision models in Odisha, an Indian state with among the lowest irrigation coverage, and therefore, free of path dependent policies. We find aquifer properties play a crucial role in mediating the groundwater-energy nexus. For this study region, on average, the maximum volume of water that can be pumped from a well of a specific depth in an alluvial aquifer is approximately 150 times the volume that can be pumped from a well in a hard-rock aquifer. Therefore, the risk of over-consumption and aquifer depletion is a far greater challenge in alluvial than hard-rock aquifers. Risks of groundwater consumption and depletion can be limited in hard-rock aquifers provided the number and depths of wells can be controlled. Capital subsidies for well construction could be an effective policy to increase irrigated area so long as economic incentives for digging deeper are not distorted. Our results imply that solar pumps are a relatively safe option for hard-rock regions where deep drawdowns naturally limit the extent of over-extraction. Solar pumps are also estimated to be among the most economical for expanding irrigation. Using a novel dataset comprising of biophysical and socioeconomic data, we find hard-rock regions to have limited irrigation coverage, high availability of annually replenishable groundwater, and high concentrations of marginalized farmers. Therefore, groundwater irrigation expansion in hard-rock areas could have dual benefits of ensuring future food security and targeting poverty reduction.