Purpose: This study aims at accounting for the variation in electricity production, processes and related impacts depending on season (heating, cooling), day of the week (tertiary building) and hour of the day. In this context, this paper suggests two alternative methods to integrate grid-building interaction in life cycle assessment of buildings and districts. Methods: An attributional dynamic method (AD) and a marginal dynamic method (MD) are compared with an annual average method (AA), representative of standard practice, using electric space heating as an illustrative case. The different methods are based on a dispatch model simulating electricity supply on an hourly basis, averaging historically observed climatic and economic variability. The meteorological inputs of the model are identical to those of the building energy simulation. Therefore, the environmental benefits from smart buildings and onsite renewable energy production are more accurately evaluated. Results and discussion: Using electricity production (or supply) data for a specific past year is a common practice in building LCA. This practice is sensitive to economic and meteorological hazards. The suggested methodology is based on a proposed reference year mitigating these hazards and thus could be seen as more representative of average impacts. Depending on the chosen approach (average or marginal) to evaluate electricity supply related impacts, the carbon footprint of the electric space heating option for the studied low-energy house in France is evaluated to 61.4 to 84.9 g CO2eq kWh−1 (AA), 78.8 to 110.2 g CO2eq kWh−1 (AD) and 765.1 to 928.7 g CO2eq kWh−1 (MD). Compared to wood and gas boiler, 22–107 and 218–284 g CO2eq kWh−1 respectively, the ranking between the different technical options depends on the chosen approach. Uncertainty analysis does not undermine the interpretation of the results. Conclusions: The proposed electricity system model allows a more precise and representative evaluation of electricity supply related impacts in LCA compared to standard practices. Two alternative methods are suggested corresponding to attributional and consequential LCA. The approach has to be chosen in line with the assessment objectives (e.g. certification, ecodesign). Prospective assessment integrating long-term evolution of the electric system and influence of global warming on buildings behaviour are identified as relevant future research subjects. © 2016, Springer-Verlag Berlin Heidelberg.