Before EDF, much of my professional career has been spent working in the more traditional areas of efficiency and renewables in the energy sustainability industry. The past few years working on ‘beyond supply’ initiatives at EDF has broadened horizons and introduced me to the fascinating world of storage, EVs, flexibility and wholesale energy market trading. Ultimately, all the pieces of the puzzle will have to fit together if we are to realise our Net Zero ambitions.
One aspect I’ve always found mildly dissatisfying is how storage and flexibility are sometimes considered as just supporting and enabling technologies for the low-carbon transition vs. providing direct and independent decarbonization benefits.
Carbon impact of efficiency vs. storage
With an efficiency project, the decarbonization equation is straightforward:
Carbon Reduction [kg.CO2] = [X] MWh/yr saved x [Y] kg.CO2/MWh x [Z] years
where [X] is the annual project energy savings, [Y] is a grid carbon intensity factor and [Z] is the project lifetime. Figuring out the equivalent number of trees planted or houses taken off the grid is just one more arithmetic step away.
With a battery storage project, there is typically no such convenient figure to calculate. Instead, the discussion on the wider societal benefits tends towards improving grid resilience, helping to manage the intermittent nature of renewables and providing essential grid services like frequency response. In addition, recent progress in overcoming ethical concerns with precious metals in the battery supply chain, as well as the long list of second-life battery applications are also cited positively.
However, I think there's more to the carbon story and we may be ‘selling batteries short’, especially when considering batteries trading in the wholesale energy markets. The reasoning is outlined below, and while I’ve focused on batteries, the same concept will apply to other flexible loads. This includes electric vehicles employing smart charging to avoid expensive peak periods as well as households making decisions about when to run their washing machines.
Carbon and price are correlated
Firstly, below is a graph of UK hourly grid carbon intensity vs. UK hourly day-ahead energy prices in 2019. Every circle represents an hour in the year and the colour of the circle represents the fraction of low-carbon power (wind, solar, hydro and nuclear) making up the generation mix for the hour.
Very simply, there is a trend between the carbon intensity of energy and its price. The next graph below is the same data shown differently, with each square of the grid representing a day in the 2019 year, and the carbon vs. price trend for the day shown within. Mostly positive slope lines in the grid show that this trend holds well throughout the year.
What drives this trend? Like all commodities, the price of electricity is driven by the dynamics of demand and supply. On the supply-side, increasing quantities of minimal marginal cost (and low-carbon) renewables in the UK grid tend to depress short-term power prices, as unlike fossil-fuelled coal or gas power stations, renewables do not need to factor in the cost of fuel per unit of electricity generated when offering a price to the market.
On a windy sunny day in the summer with low demand, we can expect low energy prices with a relatively high fraction of the generation mix made up of renewables. On the other hand, during a still and dark winter evening, more inefficient gas and coal plants will need to run to keep up with high electricity demand. This will drive up the price as well as the average carbon content of energy.
A day in the life of a battery
Now, let’s consider the carbon impact of a battery operating in the wholesale energy market with a very simple trading strategy of running one single cycle per day, charging and discharging at the trough and peak of the day-ahead price respectively. In reality, our trading team maximize value for our growing portfolio of storage assets by constantly trading across multiple markets via our PowerShift platform, as explained by my colleague Chris Regan here. However, the general concept of buying low and selling high still holds true, so we can persist with this simplification.
Now, storage by itself is intrinsically neither low-carbon or high-carbon. But after making allowances for charging efficiency, if the carbon intensity of energy used to charge a battery is lower than the carbon intensity of energy the battery displaces when discharging, this results in a net carbon reduction.
Consequently, for a merchant battery operating in wholesale market arbitrage mode, the act of buying low and selling high automatically results in decarbonisation benefits. This is because, as illustrated in the example graph below, it is typically displacing ‘high carbon’ energy when discharging during a high-price period with ‘low carbon’ energy used to charge the battery during a low-price period.
In reality, the marginal grid carbon intensity should be considered instead of the average intensity, as it’s the marginal generation unit supplying power to the grid that will be impacted by the battery operation. But data on marginal intensity is not readily available and nor is there is a single method for objectively determining marginal rates. As such, using average figures is a reasonable first-order approximation.
For illustration, with all the simplifying assumptions noted, a 1 MWh battery with a 90% roundtrip efficiency would have saved 19,000 kg of carbon (equivalent to about 20 houses off the grid) with the above trading strategy in 2019. For comparison, the equivalent efficiency savings required to generate the same carbon benefits across a similar load would have been 1-2%.
A couple of qualifications
Co-location: The above analysis has used the grid carbon intensity. Stationary batteries that form part of microgrids or are co-located with other generation assets will need to be considered differently. In particular, co-location of batteries tied directly with solar may improve the carbon benefits.
UK Specific: The premise that trading storage assets carries a carbon benefit is only based on the present grid and generation mix characteristics in GB and may not necessarily apply elsewhere.
So, to summarise...
Carbon is impacted by when we use electricity along with how much. For the “when”, we are fortunate that given the correlation between price and carbon, what’s good for our wallets is also good for the environment and our Net Zero goals.
In the domestic space, perhaps the carbon benefits of charging up our electric vehicles and running the dishwasher during off-peak hours will be a greater driver for influencing behaviour vs. just cost benefits.
In the business space, perhaps carbon reporting requirements should evolve beyond employing a static GHG conversion factor to annual energy consumption and instead provide a truer picture by accounting for time of consumption.
Lastly, for companies investing in storage assets that are traded in the markets, perhaps the direct carbon impacts should be reflected in the benefits case as well.