Life-cycle Assessment and Energy Systems Analysis of Second-life Li-ion Batteries
Examensarbete för masterexamen
With a large increase in electric vehicles, a subsequent increase in the retirement of Li-ion batteries (LIBs) will follow. The environmental performance of second-life LIBs in Swedish context has so far only been assessed in a LCA by Janssen et al. (2019), which this study intends to build upon. Energy system modelling is used as a mean to further understand the second-life use case of LIBs in Swedish context. Hence, what flows are needed for energy systems analysis to be successfully integrated with LCA and how it can enrich the assessment on the environmental performance of second-life LIBs is also explored. Using energy systems modeling, more than 2000 Swedish households with energy storages are modeled for economic efficiency for the homeowners maximizing the value of PV self-consumption. How the utilization of the storage differs between households, installation sizes and relative remaining storage capacities of the LIBs (as they initiate the second life) is also investigated. The results include estimated degradation, which is used to estimate the duration of the LIB’s second life and consequently the second lifetime performance. The environmental burden from manufacturing, can be allocated between the LIB’s first and second life by their respective energy throughput. Hence, the utilization of the storage in the second-life use case will directly effect this allocation. Furthermore, this study investigates if there is any environmental benefit in extending a LIB’s lifetime in Sweden from a marginal and average perspective and will identify the main processes contributing to the impact. The specific utilization of the storages was found to be similar between households and the different installation sizes modeled, resulting in similar environmental performances. This is likely because the utilization of the storages are mostly limited by storage volume rather than availability of PV. The environmental performance of a second-life LIB was found mainly to depend on the energy throughput during the second life, which is directly linked to the allocated burden of manufacturing, emissions from charging and avoided electricity consumption. With similar utilization between storages, and thus degradation rates, the state of health (SoH) of the LIB as it enters the second life was determining for the second-life duration and total energy throughput. A residential second-life LIB was found to charge 0.5 to 1 MWhel/kWhNSC (nominal storage capacity) over a second lifetime of 4 to 9 years, depending on the initial SoH. Thus, a second life was estimated to relieve the first life from 15 to 29% of the burden of manufacturing by energy allocation. Under the assumption that the LIBs give rise to 250kg CO2eq/kWh battery storage (Janssen et al., 2019), then 17 to 31g CO2eq/kWhel supplied from the second-life LIB will originate from allocated burden of manufacturing. A high energy throughput results in less impact per kWh electricity supplied. It is considered that emissions are related to charging PV electricity to the storage. For the chosen average perspective, the emission intensity of the PV electricity charged was assumed to be an attributional LCA value of 41g CO2eq/kWhel (Schlömer et al., 2014). From the marginal perspective a consequential LCA value was assumed of 76.7g CO2eq/kWhel (Jones & Gilbert, 2018). By allocated manufacturing impacts and emissions related to charging the storage, the second-life LIB investigated in this thesis was found to be able to supply electricity at 62 to 76g CO2eq/kWhel from the average perspective and 102 to 116g CO2eq/kWhel from the marginal. By both perspectives, emissions related to charging the storage with PV electricity was the main contribution to the environmental impact while impacts allocated from processes prior to the second-life were mainly constituted by battery pack manufacturing. As a storage is introduced to the system, some electricity and its related emissions were considered avoided by the storage. by average accounting, the avoided emissions were considered to come from the energy mix while by marginal accounting, from the marginal technology (waste incineration). With positive and negative impacts, a net impact of the storage can be calculated. From the average perspective, extending the performing lifetime of a LIB with a second-life in Sweden resulted in an environmental burden with a net impact of 22 to 37kg CO2eq/kWh storage. From the marginal perspective, an environmental benefit was found at net impact -88 to -33kg CO2eq/kWh storage. The environmental burden is caused by the storage replacing grid electricity with lower emission intensity than it is able to supply electricity at. Similarly, the environmental benefit comes from replacing electricity, on the margin, with higher emission intensity than that supplied by the storage. In a static system, typically depicted in LCA, avoided emissions would have been calculated with the total electricity consumption avoided and the average emission intensity of this electricity. By using average accounting for avoided emissions and hourly resolution in the energy systems model, the resulting emissions were 7 to 8% lower than when calculated for a static system, using the same data. This indicates that incorporating energy systems modeling into the LCA added value to the assessment.