Detailed temporal consideration has been a major challenge for energy systems models with typical time horizons of years and decades. This presents particular issues in investigating electricity generation, capacity and storage, whilst retaining broader trade-offs sectors, technology pathways and timing of investments. This paper reports on a methodology for temporal disaggregation in the widely applied energy service driven, technology rich, cost optimizing, linear programming MARKAL energy system model. A flexible time slicing feature is developed to enhance representation of diurnal and seasonal electricity demand curves through disaggregation of resource availability and energy service demands. In a first application of a temporal UK MARKAL model, a range of runs investigate the role of electricity storage at supply and demand sides. The results display considerably enhanced insights, notably on the role and preference of demand-side electricity storage over supply-side storage. On average, the system chooses about 7-10% of electricity demand as storage. On the supply side, hydrogen-based electricity storage is greatly preferred but stored-hydrogen is used in the transport sector rather than for power system balancing mechanism. (c) 2010 Elsevier Ltd. All rights reserved.