Scoping Australia’s virtual power plant potential
Australia now has over 2 million small-scale solar PV systems installed, according to the Clean Energy Regulator (CER). The milestone came just at the end of a record-breaking year for the Australian solar sector - a year which also saw the introduction of a number of battery incentive programs. The CER numbers show that over 20% of Australia’s stand-alone homes now have solar, with 6 gigawatts (GW) of solar capacity installed on homes around the country.
While this is impressive - and indeed, Australia remains a world leader in terms of household solar penetration - there’s still plenty more to go. According to a report published late last year by UNSW in conjunction with the APVI and Solar Citizens, there is enough residential roof space to accommodate anywhere between 43GW to 61GW of solar capacity in Australia - 7 to 10 times more than what is currently installed.
This stunning number shows that Australia is just getting started as a distributed energy superpower. It also highlights the enormous role that virtual power plants (VPPs) could play in the future energy system.
With an already large and rapidly growing number of distributed solar & battery systems, the role for virtual power plants (VPPs) on Australia’s National Electricity Market (NEM) could be enormous. The table provides an overview of the potential for battery-based VPPs on the NEM. In a high growth scenario, VPPs could constitute up to 11 gigawatts (GW) of dispatchable household battery capacity - meeting as much as 25% of forecast maximum operational demand on the NEM.
But even a business as usual (BAU) growth scenario shows Australia hitting 3.4GW of battery-based VPP capacity by 2040 - enough to meet nearly 8% of maximum demand on the NEM. The opportunity will be larger with demand response management (e.g. curtailing use of air conditioners, pool pumps and other equipment) also incorporated into the VPP envelope.
We note, however, that VPPs can - and will - do more than ‘just’ wholesale energy trading; they will also be deployed to deliver crucial network support services to stabilise the grid on location and time-specific bases. While outside the scope of this article, this aspect of VPP function will have comparable importance to energy trading in the years and decades to come.
(The tool we created to make these estimates is publicly available here.)
SwitchDin enables vendor-neutral VPPs using batteries and demand response
The impact of Distributed energy on the NEM will only grow
Already, Australia’s high level of solar uptake is having an impact on the electricity system. Its most obvious manifestation is the depression of daytime electricity demand and wholesale electricity prices on the National Electricity Market (NEM), followed by sharper spikes as the sun begins to set. This phenomenon is referred to as the ‘duck curve’ because of the shape it makes when graphed out (see animation).
The emergence of the duck curve is attributable not only to the 6GW of residential rooftop solar, but also to larger commercial & industrial solar projects and (to a lesser degree) another 1GW of utility-scale solar. Small-scale solar systems, however, bring their own suite of challenges and opportunities which go beyond wholesale energy market dynamics, having implications for network stability as well.
Like solar, growing uptake of home battery storage will alter demand patterns on the NEM and have an impact on network function. By their nature, batteries (which store and discharge electricity) offer a greater degree of flexibility than solar PV (which only generates electricity). In order to maximise their utility, they need to be coordinated in conjunction with other batteries through VPP programs; this is where the greatest potential lies for distributed energy resources (DERs) on the grid.
Scoping Australia’s potential virtual power plant capacity
So how big a role could there be for battery-based VPPs to play on the grid?
The answer depends on a number of questions:
How many NEM-connected homes will have solar PV?
What proportion of these homes will have batteries?
What is the average battery capacity for each home?
What proportion of the battery-equipped homes will be signed up for a VPP program (i.e. ‘VPP-ready’)?
What proportion of the VPP-ready homes will be available for dispatch at any given time?
We look at each of these in detail below.
1. How many NEM-connected homes will have solar PV?
43GW is the conservative ‘maximum potential’ figure floated by UNSW/APVI/Solar Citizens for stand-alone homes (not even counting businesses), translating into about 12 million solar homes nationally or roughly 10 million solar homes on the NEM if the NEM / non-NEM ratio remains the same as what it is now.
But a rollout to this full number will take time. BNEF’s forecasts give an idea of the pace, estimating a cumulative total of about 15GW of residential solar capacity installed by 2030, translating into (very roughly) just over 4 million solar homes on the NEM. AEMO’s estimates are not far behind, with about 13GW of rooftop solar on the grid by 2030 under their ‘neutral’ scenario. Both anticipate about 20GW by 2040.
BAU estimates of solar homes on the NEM:
4 million solar homes by 2030
7 million solar homes by 2040
2. What proportion of the solar homes will have batteries?
Australia’s residential battery storage market is newer than the home solar market. Because there are more unknowns (including the impact of incoming battery incentives), forecasts for battery uptake vary more than with solar. Case in point: AEMO’s ESOO 2018 ‘neutral’ scenario modelling anticipates that by 2038 there will be 2.6GW of behind-the-meter battery capacity installed, while BNEF’s estimate is nearly six times greater at 15GW.
A more useful reference point for BAU battery uptake is the 15% solar/battery integration estimate that AEMO anticipates by the end of 2028. If 15% of 2038’s estimated 7 million solar homes have batteries, there would be roughly 900,000 battery-equipped homes.
But in comparison to other projections, 15% looks conservative. For example, if BNEF’s 15GW figure is correct the uptake percentage would be close to 50% - translating into roughly 3 million battery-equipped homes on the NEM.
Meanwhile, the federal Labor Party has pitched a policy platform that would put Australia on a path to 1 million battery-equipped homes by 2025 - should they win the election. This would work out to be about 30% of solar homes with batteries.
Even if this federal program never comes into effect, however, the suite of state-based programs already on the cards represent a commitment of incentives for about 150,000 household. Both BNEF’s New Energy Outlook and AEMO’s ESOO 2018 reports were published before this wave of incentives came about - so their 2019 updates are likely to be more bullish.
BAU estimates of number of solar homes with batteries on the NEM (at 15% uptake):
500,000 battery-equipped solar homes by 2030
900,000 battery-equipped solar homes by 2040
3. What is the average battery capacity for each home?
The federal Labor Party’s proposed battery incentive would support systems up to 4 kilowatt-hours (kWh) in energy storage capacity. The details may change, but batteries of this size are on the small side, with some stand-alone products on the market being as large as 13-15kWh.
While energy storage capacity is relevant for VPP operation (to be discussed in a separate piece), the capacity figure we’ll focus on for the time being is power output, in kilowatts (kW). Although batteries themselves have have a rated power output capacity of their own, this is almost always modulated by the inverter to which they are connected (true even for ‘all-in-one’ battery/inverter solutions); residential inverter output ranges are generally between 3.5kW and 7kW, with 5kW being typical.
Conservative estimate of total distributed battery capacity on NEM (assuming 5kW capacity per home):
2.4GW distributed battery capacity by 2030
3.8GW distributed battery capacity by 2040
4. What proportion of the battery-equipped homes will join a VPP?
AEMO’s ESOO 2018 estimates that the percentage of signed up for a VPP program will be about 28% in 2038, with lower rates in the preceding years. (For simplicity’s sake, we use 28% for 2028 as well).
Interestingly, the number of homes participating in a VPP program will be a subset of the total number of VPP-ready homes (who have the technical capacity to join a VPP but may choose not to).
We note, however, that actual uptake rates are likely to be higher than those below, as the battery incentive & VPP program commitments from South Australia (90,000), New South Wales (40,000) and Victoria (10,000) alone are set to usher in nearly 150,000 battery-equipped homes by the early 2020s. Many of these will enter the market with VPP participation on their minds, as VPPs become an integral part of the ‘normal’ value proposition for installing battery storage.
Conservative estimate of VPP-ready battery capacity on NEM:
0.7GW by 2030 (approx 140,000 homes)
1.1GW by 2040 (approx 250,000 homes)
5. What proportion of the VPP-ready homes will be available for dispatch at any given time?
VPPs differ from conventional power plants in that they are built from fleets of DER assets not owned by the operator; indeed, the batteries ‘in’ a VPP are first and foremost the property of households/businesses, who mostly use them to reduce reliance on grid electricity. Because of variations in household energy consumption patterns, only a portion of VPP-ready homes will be available to dispatch energy into the NEM at a given time.
There are no estimates available indicating what the availability percentage might be as VPP programs become more common. However, most VPP operators and tech platforms are not passive in their approach to market opportunities (including anticipated demand/price spikes) - using smart software to ‘quarantine’ and pre-charge batteries in preparation for possible dispatch. We therefore assume a 70% availability figure in our modelling.
We highlight again that they only take into account the possibility of energy trading on the NEM using batteries, and not demand response capability, which VPPs can also enable - and which already play a role in some network-based VPP trials in which SwitchDin is involved.
We also emphasise that this analysis does not focus on the arguably more enduring prospect of network services, which VPPs can also supply. Payment to households for provision of network services will only grow more important with time, as such services become a more widely accepted and utilised approach to addressing network challenges.
Final total - BAU / conservative estimate of VPP capacity available for instantaneous dispatch on NEM:
0.5GW by 2030
0.75GW by 2040
Examining ‘growth’ and ‘high growth’ scenarios
So far we’ve used conservative inputs to get a baseline idea of what VPP capacity in Australia could be in a decade and in two decades’ time - arriving at somewhere between 1/2 - 3/4 GW by 2040 if each battery contributes 5kW.
However, these numbers could easily be much greater given that there are many additional sources of growth which are unaccounted in the baseline/BAU scenario.
Q2. Solar homes install batteries at a faster rate thanks to incentives and falling battery prices:
30% of solar homes with batteries (up from 15%)
Q3. Average inverter size remains at 5kW
Q4. More VPP programs available, more households accept VPP benefits as part of battery ownership
50% of battery-equipped homes in a VPP (up from 28%)
Q5. VPP technology is capable of effectively and efficiently quarantining battery capacity for market response
90% dispatch availability among VPP-ready homes (up from 70%)
2.1GW by 2030
3.4GW by 2040
High growth scenario
Q2. Home battery prices drop dramatically, resulting in widespread uptake
50% of solar homes with battery storage
Q3. More households install larger PV and/or battery systems, increasing average inverter size
Average of 7kW inverter (battery output) capacity per home
Q4. VPP programs abound and participation becomes a key benefit of battery ownership
70% of battery-equipped homes in a VPP
Q5. 90% dispatch availability among VPP-ready homes
About 7GW by 2030
About 11GW by 2040
Putting it into context
To get a feel for what these numbers would mean, it’s helpful to understand that:
AEMO forecast maximum operational demand on the NEM is:
~40GW by 2030
Growth scenario: 2.1GW VPP capacity would be equal to ~5.4% maximum operational demand
High growth scenario: 7GW VPP capacity would be equal to about 18% of max operational demand
~45GW by 2040
Growth scenario: 2.5GW VPP capacity would be equal to ~7.7% maximum operational demand
High growth scenario 11GW VPP capacity would be equal to about 25% of max operational demand
Comparing these figures to the capacity to a conventional ‘peaker’ power plant is also helpful - even if the comparison is imperfect. For example, Australia’s largest gas-fired power plant, Colongra Power Station in NSW, has an output capacity of about 0.7GW.
While both gas plants and VPPs can be deployed to meet spikes in electricity demand on the NEM, there are some important differences between the two. For example:
VPPs are comprised of consumer-owned assets; the primary purpose of these assets is to enable households to bolster their energy self-reliance and reduce energy costs while also reducing everyday demand on the NEM - they’re not sitting idle when they’re not actively ‘VPP-ing’
VPPs can incorporate on-site demand part of their decision-making process/algorithms - meaning that in addition to helping produce power during times of high demand, they can also reduce loads, which potential multiplies their capacity (see section below)
VPPs are scalable and can be built incrementally - making them a potentially lower-risk investment than a conventional large-scale power plant
VPPs are more versatile, capable of supplying geo-specific network services as well as energy market participation
There’s More to a VPP than energy trading with battery storage
The focus of this article has been VPPs built from distributed battery storage systems for the purpose of energy trading on the NEM. The reality is, however, that VPPs of the future will use much more than just batteries to do much more than just energy trading.
Households with air conditioners, pool pumps, and other demand response enabled devices (DREDs) will be paid to reduce their consumption, providing a similar service to a battery-based VPP - but without the battery
Battery discharge for participation in the frequency control ancillary services (FCAS) market with as part of a VPP program
Payments to households for localised network services, including demand response, solar curtailment and battery energy export
Combined, these additional capabilities will help increase the potential for VPPs in Australia beyond the BAU estimates in this article.
Explore the numbers yourself
We’ve put together a calculator that provides rough estimates of future potential VPP capacity based on a number of variables related to the five questions in this article. (This tool will be refined as time goes on.)
Click here or on the image below to access it.
SwitchDin is enabling the distributed energy future
Big changes are afoot in Australia’s energy system, and SwitchDin is here to help tie it all together. We enable VPPs, microgrids and other types of distributed energy resource management, with a number of projects up and running.
Learn more about what we do - or get in touch below to talk about how we can work together.