Demand-side energy management systems (EMSs) are expected to be used as a method for further balancing the supply and demand of power systems with high levels of renewable energy generation, such as photovoltaic (PV) power. Energy demand and solar radiation must be predicted to realize the optimal operation scheduling of demand-side appliances by a home energy management system (HEMS), including heat pump water heaters, PV systems, and solar-powered water heaters. This paper presents a HEMS model that controls a residential battery system connected to a rooftop PV system taking into account of energy load and PV generation forecast errors. Forecast accuracy is verified by real HEMS data from 160 households; the impact of forecast errors on household economics is examined. Furthermore, the contribution to the entire power system is examined using the model under the dynamic pricing system. (C) 2015 Elsevier Ltd. All rights reserved.