The increased availability of water end use measurement studies allows for more mechanistic and detailed approaches to estimating household water demand and conservation potential. Here, probability distributions for parameters affecting water use are estimated from end use studies and randomly sampled in Monte Carlo iterations to simulate water use in a single-family residential neighborhood. This model represents the existing conditions and is calibrated to metered data. A two-stage mixed integer programming optimization model is then developed to estimate the least-cost combination of long- and short-term conservation actions for each household. This least-cost conservation model provides an estimate of the upper bound of reasonable conservation potential for varying pricing and rebate conditions. The models were adapted from previous work in Jordan and are applied to a neighborhood in San Ramon, CA in the EBMUD service area. The existing conditions model produces seasonal use results very close to the metered data, and the least-cost conservation model suggests water demand is relatively inelastic with respect to price, "cash for grass" programs will not be effective, and clothes washer rebates are the most cost-effective rebate programs. |