Support for policy to expand electricity storage appears to be increasing. The Smart Energy Council and Clean Energy Council have both backed such plans. Attention is increasingly focused on the details of such policies and their costs. This note looks at the economy of “grid oriented” batteries behind the meter.
In November 2021, Green Energy Markets proposed to extend the small-scale solar subsidy scheme to home batteries. In February of this year, Senator Helen Haines submitted a bill to parliament based on their analysis and ideas.
In May, Peter Harris, Peter Sheehan, Ted Woodley and I released a report suggesting that electricity storage policies should be at the heart of the Australian government’s energy policy. We proposed a broad scheme for storage large and small behind and in front of the meter and including EVs, broadly modeled on the renewable energy target.
The electricity storage market in Australia is growing rapidly. Technology and commercial development in this part of the market is rapid. Much remains to be done to better understand this for policy-level assessment so that customers can extract value from storage and benefit the industry. This note responds to that need.
We report here on preliminary studies trying to understand how much income a battery behind the meter (BTM) would earn if it were focused purely on the 5-minute wholesale market and wanted to maximize its profits in that market.
We further wonder how this would change if such a BTM battery were used to exploit arbitrage opportunities in both wholesale and retail markets (i.e. arbitrating prices in the customer’s retail offering and assuming they have BTM solar.)
A BTM battery so completely grid-oriented is not the norm. There are more and more cases of ‘virtual power plant’ batteries that can be shipped centrally, but not in the regular (daily) way we assume here.
The commercial construct we use for this review is where a household (or a utility company) invests in a BTM battery as a high-frequency trading device. Would this be profitable? If so, this kind of commercial arrangement may well find a market.
We assume a battery of 5 kW/10 kWh (typical for residential batteries behind the meter) and we also assume that the household has solar panels and that the battery is used in such a way that it is fully charged during the period from 08.00 a.m. to 5 p.m. and fully discharged in the period from 5:30 p.m. to 9 p.m., every day.
For the wholesale market prices, we have taken these 5-minute prices for the period October 1, 2021 (when 5-minute prices were introduced in the NEM) to August 31, 2022.
For Case A (the BTM battery that is purely aimed at the wholesale market), we assume that the battery will be charged at 5 kW during the 24 cheapest intervals of five minutes (i.e. 2 hours) in the charging period and then discharged at 5 kW. kW for the 24 most expensive five-minute intervals in the discharge period.
For Case B (the BTM battery that takes full advantage of wholesale and trade opportunities behind the meter), we set a charge cap of 5 cents per kWh (this is the assumed feed-in rate for excess solar on the roof). ). And we impose a floor price of 40 cents per kWh on the discharge (this is the assumed retail price of grid-supplied electricity in the peak period).
This means that we assume that the customer always has 5 kW of solar energy available to charge the battery (if necessary) during the charging period, and that the customer has sufficient demand to consume 5 kW in each case that he the battery pulls during the charging period. discharge period.
For most households this is optimistic (there will be many instances where 5 kW will not be available and households rarely have 5 kW demand at a time).
A more realistic analysis would use a sample of actual data from 5 minutes of demand and rooftop solar power generation. Case B should therefore be seen as an upper bound of the income that a 5kW/10kWh battery could receive if it had been programmed to take full advantage of the wholesale and BTM trading opportunities.
The first chart below shows the number of 5 minute cases during our study period in which the battery charges and the number it discharges for Case A (BTM battery aimed at the wholesale market) in Queensland.
From the chart you can see that it will most likely charge from 10:30am to 1:30pm when rooftop solar peaks and thus wholesale prices in Queensland are at their lowest.
Conversely, we can see that the battery is very likely to discharge from 5:30 PM to 7:30 PM when prices are usually highest (the blue line in the chart is the median spot price in Queensland in each interval).
The second graph (below) shows the results for Case B. You can now see that a $50/MWh cap on the charge price and a $400/MWh cap on the discharge price results in a much more even spread of charging and discharging .
How do these business results translate into revenue for our two cases? The chart below shows the annual income that our 5kW/10 kWh battery arbitrage battery would earn in the wholesale market in each of the five NEM regions (we assumed income in the 12th month equal to the average of the 11 modeled months).
As we see in this chart, income is lowest in Tasmania, where hydro storage limits prices in peak periods and increases prices in off-peak periods compared to other NEM regions.
South Australia and Queensland have the highest incomes. This follows the observation that their abundant solar power drives prices down during the charging periods; and since very expensive gas generation usually determines wholesale market prices during unloading hours, the combination of the high prices during those hours and the low prices during the loading hours leads to nice arbitrage income.
The table below shows in the first column the annual income for the wholesale market-oriented battery (case A). The second column shows the annual income for the wholesale and retail battery (case B).
|Case A: Annual income (wholesale oriented)||Case B: Annual income (wholesale + retail oriented)|
The first (data) column shows a healthy income for a wholesale-only battery, particularly in Queensland. Tasmania, as expected, has the lowest income and not enough to justify installing a BTM battery in the wholesale market.
The second column (Case B) shows healthy income across all NEM regions and also that home charge/discharge capability is most/least valuable in those states with the least/most attractive wholesale arbitrage opportunities, namely Tasmania/Queensland.
Assuming an installation cost of approximately $10k for our assumed battery, these numbers suggest an attractive net return, particularly for batteries capable of serving wholesale and retail markets, across all regions.
Our analysis shows that growing storage in the NEM is unlikely to require much subsidy. It reveals quite significant arbitrage earnings in the wholesale market and potentially lucrative arbitrage earnings in the retail market.
While NEM prices were quite extreme during our study period, this may be becoming a new normal – gas is now extremely expensive and the continued expansion of solar energy will provide bountiful cheap deliveries in the middle of the day.
Nectr pioneered VPP-enabled solar battery bundles and Energy Australia has now introduced this model in their main retail operations (having started it in their retail venture) and compensation for participating in a “virtual power plant” is now common among those retailers that offer VPP batteries.
What can be learned about how these work in the real world?
Given the many factors that influence the economics of storage, this analysis suggests that it is important not to pick winners. The real market is far too complex and dynamic for any of us to fully understand.
The best we can hope for is to encourage savvy risk takers to try it and then analyze who succeeds and who doesn’t. It would be a bad idea to consider behind-the-meter instead of front-of-meter batteries or large batteries versus small batteries more worthy than the other for policy support.
Bruce Mountain and Ben Willey, Victoria Energy Policy Center