Calculating carbon dioxide (CO2) emissions associated with electricity is a key component in the field of Life Cycle Assessment (LCA), but is often cited as challenging due to the complex nature of electricity systems despite its importance to the outcome. While calculating the operational CO2 emissions associated with electricity generation is an active research field, the embodied CO2 emissions, typically referred to as embodied carbon, of network assets has far less representation in the literature. This paper focuses on the CO2 emissions aspect of LCA to calculate the embodied CO2 of network assets in relation to the operational grid CO2 over time. Several functional units are defined: CO2 per operational year, CO2 per asset cost, CO2 per functional unit of electricity (kW h) and the relationship between embodied emissions and operational emissions in an electricity system over time. Hybrid functional units are then applied in order to better attribute the embodied carbon to the network functions. The hybrid functional units involve network asset lifetime and the issue of temporal horizons. Several suitable horizons are suggested and the comparison of results highlight the importance of the timeframe on results. The relationship between temporal horizons and environmental discounting is discussed and recommendations are made on the appropriate level of discounting depending on the temporal horizon and the purpose of the LCA. The paper uses data from the Great Britain electricity system where planned investment in network assets is (sic)12bn at distribution level (Dx) and (sic)16.4bn at transmission level (Tx) over the next eight years. By using GB network data for embodied carbon, demand and asset data, as well as data from the decarbonisation of electricity generation, indicative results are provided into the way in which embodied carbon impacts could change over time, showing that by 2035, the embodied carbon of the transmission network could contribute almost 25% of total emissions associated with electricity. On a regional basis, DNO level network assets could reach anywhere between 40% and 130%. This network data is also used to show that new network investment could account for up to 6.5% of DNO level network embodied carbon when front loaded during the RIIO-ED1 period. (C) 2016 Published by Elsevier Ltd.