Smart Energy Council – Solar & Storage 2018 Article

Energy Storage Economics – choosing the right sized battery to reduce peak demand

There is a real need to change our energy mix to cleaner sources to minimise its impact on climate and energy storage is often proposed as a way to assist in the  facilitation of higher levels of renewable energy generation.  I feel fortunate that much of my work of late has been  in this area where the economic analysis of energy storage is crucial to ensure the appropriate system is employed, especially in cases where the technology is still expensive.

When looking at the economics of energy storage, the area that causes most initial confusion is the distinction between energy (in kWh or MWh) and power (kW or MW).   Depending upon the application, there may be a focus on power or energy i.e. for applications such as some off-grid systems, frequency control, or energy market arbitrage the focus may be on power, whilst in others such grid connected batteries or battery back-up the need for low cost per unit of energy storage may be important.  Usually, on a cost per kWh it is more expensive to have a more powerful battery i.e. if we had a 1MWh battery and there were 3 power ratings available in the market of 2MW, 1MW and 500kW, we would expect the 2MW to be most expensive, then the 1MW and finally the 500kW solution.

It might seem simple then just to buy the cheapest per kWh energy storage system but this is not always the case as it is best to work out the most economic choice i.e. one needs to compare the value of the potential savings from different configurations and then examine this against the purchase price.  As an example, we have taken the 2017 historic load in South Australia and compared different power ratings, for where the battery has 30 minutes, 1 hour or two hours of storage with a round trip efficiency of 90% to see how different systems perform in minimising the annual peak demand.

In this analysis, there are two measures; the demand reduction in MW for each storage solution depending upon the MW power rating, and the effectiveness of the chosen solution in MW per MWh of storage.  Whilst increasing the size of the battery increases the reduction in peak demand, the system becomes less effective per unit of energy in reducing this figure.  If in this example, the battery faced a fixed annual peak demand charge, then a smaller battery size may be a better investment unless the battery’s economics improve faster with scale. Also, the batteries with higher power (or less energy capacity for a given power rating) have a better financial return which needs to be assessed against their likely higher cost.  As an example, at 50MW, there is no benefit in having a 1 hour battery as against a 30 minute battery, so only after the installed costs are known, can an assessment be made of the optimal economic solution.

This example only looks at demand reduction but it shows that a thorough analysis is required to understand the economics.  Similar effort needs to be taken to analyse system size for wholesale market or retail price arbitrage, ancillary service markets and for network support, which is too much for this short article.  Once these other income streams are identified, the potential conflicts between these opportunities must be assessed.   Ideally, further analysis will overlay other factors such as battery degradation, different technology characteristics such as useful life, roundtrip efficiency and depth of discharge limitations, each of which will usually result in a reduced economic return.  Only when all of this is performed can a proper assessment of a battery ‘s investment be done.

Warwick Forster- Apogee Energy (www.apogeeenergy.com.au)

As published in Spring Edition of Solar & Storage (https://www.smartenergy.org.au/sites/default/files/uploaded-content/field_f_magazine_upload/ss_2018_issue_3_sept_web_1a.pdf)


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.