Countries are pushing for the use of local, renewable energy sources in order to reduce the dependence on fossil fuels for energy supply. One of the main problems with several renewable energy sources is their variability, which can be solved with energy storage. With buildings representing an important share of energy consumption, and given the growing capacity of distributed generation, distributed energy storage in buildings is expected to become increasingly present. In this context, the optimal dimensioning of home installations of photovoltaics and lithium-ion batteries, and the impact of such installations on the grid, is of the utmost importance. While there have been developments on this field, some important handicaps remain, notably the independent treatment of installation optimisation and grid impact and the substantial result differences between studies. In this paper, photovoltaics and lithium-ion storage installations are optimised through the use of real, high-resolution data from several individual households, based on realistic cost figures, and through well-defined metrics that correctly grasp the problem at hand. The impact on the grid as well as possible mitigation measures are also analysed. Results show that up to about 30% of electricity self-sufficiency can be obtained using only PV and close to grid parity. Above 40% self-sufficiency, energy storage must be used, strongly increasing the cost of such installations. Economies of scale play an important role suggesting a preferential implementation for larger users or at a community-scale. Feed-in limits seem to be a good solution to attenuate grid impact. On the other hand, a higher share of a capacity-based component on the grid tariff strongly affects the economic viability of such installations for the average household. These results are important for studies on distributed photovoltaics and energy storage as well as for energy policy. Also, the large range of results made available, calculated on a free market perspective using a simple control mechanism, provide a much-needed benchmark for further comparable studies. (C) 2017 Elsevier Ltd. All rights reserved.