A transition to a decentralised, decarbonised energy system for the domestic sector is constrained by the difficulty of obtaining energy balance between fluctuating demand and the intermittent, non-dispatchable power supply delivered by most renewables. A microgrid system including a mix of renewable generation technologies, energy storage and demand response (DR) systems has been modelled using a linear programming approach, based on real world data of residential energy consumption and weather variables. This model allows the exploration of the effects of fluctuations in demand and supply, microgrid scale and configuration, energy management options and alternative optimisation criteria. The model demonstrates quantitatively that a mixed-renewables microgrid system can reduce demand fluctuations and improve energy balance. Peak demand hour fluctuations were reduced by up to 19% for a simulated microgrid containing 144 households with one renewable unit and four batteries per household, with a renewables mix of 83% photovoltaic (PV) panels and 17% wind turbines. With this system, the demand on macrogrid energy supply was reduced by 16%, CO2 emissions associated with energy use were reduced by 10% for all hours of operation, and by 74% during the hours of renewable supply. These findings suggest that microgrids using contemporary technologies can contribute significantly to CO2 mitigation targets. (C) 2012 Elsevier Ltd. All rights reserved.