The transportation sector is a significant source of greenhouse gases and air pollution, contributing to climate change and affecting human health. The transition from traditional Internal Combustion Engine (ICE) vehicles to Battery Electric Vehicles (BEVs) is a critical feature of efforts to meet climate targets in the UK and around the world. In this blog, Dr Iain Staffell and Dr Daniel Mehlig from Imperial’s Centre for Environmental Policy draw on their recent research to explore the impacts of BEV charging. Iain leads the market design work of the IDLES Programme, investigating the regulations, policies and market structures needed to steer us towards the optimal integrated energy system of the future.
The growth in sales of passenger BEVs offers a huge opportunity to replicate the success of electricity sector decarbonisation in the transport sector. At the end of December 2021, there were over 395,000 BEVs on UK roads and that figure continues to rise. While BEVs do not produce greenhouse gas emissions on the road, their environmental credentials relative to conventional ICE vehicles are still the subject of some debate, partly due to the source of electricity needed to charge them. Both sectors must function successfully together to avoid shifting the pollution problem elsewhere, i.e. displacing exhaust emissions with electricity generation emissions from power plants. Understanding the link between BEV charging and the electricity system is essential if we’re to integrate the two.
Although claims that BEVs can be responsible for greater emissions over their lifetime than their ICE equivalents are widely refuted, studies into the impacts of BEVs typically only focus on a single pollutant – CO2 – and employ simple methodologies using annual average electricity generation mixes. Calculating the true emissions produced by BEV charging is complicated by the time-varying nature of electricity generation and a lack of clarity over the generating technologies needed to meet the additional load. We have recently completed a study that – for the first time – pairs nationally representative charging data with high resolution electricity generation data, enabling us to take account of variation in the electricity grid’s carbon intensity with time and look beyond CO2 by calculating air pollutant emissions of PM2.5, NOx, and SO2.
Different methods are used to allocate emissions from BEV charging and it’s worth spending some time thinking about the aspects that influence the choice of method, since these relate to the research questions you want to explore. Firstly, assessments can either be retrospective (examining present or past BEV activities) or prospective (studying future BEV activities). Retrospective studies have typically been hampered by a lack of data; however, this has improved in recent years. Secondly, the emissions calculated can be categorised as average or marginal. Average emissions are the real-time emissions of electricity generation, calculated from the sum of all emissions produced by the power plants supplying electricity to the grid in a given period, divided by the net output of these power plants. On the other hand, marginal emissions are calculated by looking at the change in load on the system (in this case demand for BEV charging) and the emissions that would result from meeting this demand. Average and marginal emissions can provide very different results and deciding which approach to go for depends on whether the load is viewed as static and unchangeable, or additional and variable.
In the short-run, demand from new BEVs presents an additional and variable load on the system, requiring an increase in output from the marginally operating plants, and so constitutes a marginal emissions approach. Over time, each individual vehicle contributes to the overall charging demand of the UK’s fleet of BEVs. This fleet-wide pattern of charging demand is accommodated by the electricity system through the building and operating of new infrastructure. There is currently no methodology that can estimate how emissions change from the short-run to the long-run from historical data alone. Therefore, we can only approximate the long run emissions from charging BEVs using the average emissions approach.
The third and final aspect of the modelling approach relates to the temporal characteristic of the method – are you considering daily and seasonal variation in your electricity generation or are you using a generation mix averaged over minutes, days, months, or years? The most common approach uses yearly averages since the data required to undertake a time-varying study is typically difficult to obtain.
These three aspects helped define our approach, where we chose to undertake a retrospective analysis of the decade from 2010 to 2020, using data at a 30-minute time resolution allowing for daily, seasonal, and yearly affects to be captured. We wanted to observe the impact of introducing new BEVs into the fleet throughout the decade and so used a marginal emissions approach. To compliment this, we also employed an average emissions approach to allocate emissions to the existing BEV fleet.
In our study, we used two real-world charging profiles (showing when BEVs are charged throughout the day) and two theoretical charging profiles to mimic smart charging technologies that exist in the real world. These charging profiles were paired with electricity generation data from Electric Insights. Uniting the charging profiles and electricity generation data and then using the different methods to calculate average and marginal emissions has enabled us to make a number of interesting observations that we dig into in detail in the paper. The research provides valuable insights for those working on, or just interested in, BEV life cycle emissions.
The study shows that emissions (average and marginal) of CO2, NOx, and SO2 steadily reduced from 2012 to 2019, although PM2.5 emissions have largely remained constant. The phase out of coal generation was a major reason for this decline, being displaced by gas, imports and renewables since 2012. Tracking the data over the past 10 years is interesting because you can observe via the marginal and average emissions the fundamental changes in the energy system over this period. For example, the change from coal being both the major baseload and marginal contributor to being overtaken by gas in the marginal mix from around 2012.
For the typical BEV entering the fleet in 2019, where the new charging demand is considered marginal in the short-run, we found a 23% increase in CO2 emissions for the short-run marginal emissions when compared to average emissions. We investigated what this increase may mean over the lifetime of the BEV if electricity generating infrastructure did not respond to the change in pattern of charging demand. We referred to results of a recent life cycle assessment (LCA) and found that a much greater increase would be required to bring the lifetime emissions of the BEV anywhere close to that of an ICE vehicle.
Our findings suggest that smart charging in the real world may risk increasing emissions in the short run. This is due to how current smart charging strategies are designed to minimise the real-time emission intensity of electricity but neglect the impact this additional demand may have on marginal emissions. To improve the efficacy of smart charging in the real world, the focus should be switched from minimising average emissions to minimising marginal emissions and will likely require opportunities to charge at any point in the day.
Finally, the methods used to calculate emissions in this work and the simpler methods used in LCA studies were scrutinised. We found that neglecting the time-varying nature of BEV charging and electricity generation and instead using a yearly average grid mix and the resulting grid CO2 emission factors resulted in an underestimation of CO2 emissions by 4%. This change is small compared to other areas of uncertainty in typical LCAs, and so is a fair approximation for studies employing this simpler methodology.
We’re excited that we were able to show the value of coupling UK-scale information describing both the electricity generation mix and how vehicles were charged and compare across different methods for calculating emissions from EV charging. Ultimately, we hope this work helps develop understanding around the link between charging BEVs and the impact on the electricity system so that we can set in place the best strategies – for the electricity system and consumers – to enable efficient decarbonisation of our transport sector.
Read more about the IDLES Programme here.