Reducing global carbon emissions will require diverse industrial sectors to use energy more efficiently, electrify, and operate intermittently. The water sector is a transformation target, but we lack energy quantification tools to guide operational, infrastructure, and policy interventions in complex water sourcing, treatment, and distribution networks. The marginal energy intensity (MEI) of water supply quantifies the location-specific, instantaneous embedded energy in water delivered to consumers. We describe the first MEI algorithm and elucidate the sensitivity of MEI to generalizable water system features. When incorporated in multi-objective operational and planning models, MEI will dramatically increase the energy co-benefits of water efficiency, conservation, and retrofit programs; maximize energy flexibility services that water systems can deliver to the grid; and facilitate full cost recovery in distribution system operation.