How much solar can the grid expect to absorb overt the next 30 years? And how much storage will be installed to manage it? These are important questions to the Australian Energy Market Operator, which knows a lot about supply from generators bigger than 30MW but a lot less about the massive and fast growing number of PV systems on houses, businesses, factories and paddocks around the country.
Solar is destined to play a major role in the transition to clean energy as coal stations are slowly retired over the next 20 years, but the energy source is flighty, to say the least. The electrons start to flow at sun-up and peak at noon but by sundown it’s all over. Solar energy is reliable, sure, and cheap (and getting cheaper) – but AEMO knows the sub-30MW end of the market is a bucking bull that it needs to understand and attempt to tame.
In June, AEMO took receipt of a report by Green Energy Markets that it had commissioned to help it get a handle on what to expect over the next 30 years from small- and mid-scale solar and batteries.
It’s a cloudy outlook the solar sector, with the number of installations predicted to fall sharply in a couple of years and never recover the 2020 peak. As the incentives to own PV are crimped by falling feed-in tariffs, network export limits and dwindling government subsidies, the industry is destined to change or adapt.
Not quite sunset for solar
For the report, PV systems were divided into different buckets for size: up to 15kW, 15kW-100kW, 100kW-1MW and 1MW-30MW. This is to account for the different revenue characteristics, says Green Energy Markets director analysis and advisory and report co-author Tristan Edis. “If you’re in front of the meter you’re effectively a power station and only see the wholesale price,” he says. “Those electricity customers behind the meter face different types of tariffs.” Some owners of commercial systems are not be allowed to export, whereas owners of residential systems see about 80% of their solar energy exported in return for whatever the local retailer is offering.
Edis recalls from about nine or so years ago debates about a sensible system size for residential solar. Is there any point going bigger than 1.5-2kW when any more would be surplus to requirements? The counterpoint, he remembers, was: the fixed costs of installing solar are high, so we may as well “deck the roof out with as much solar as we possibly can”. On that basis, he argues the residential solar boom has been driven by suppliers, not customers. With the cost of borrowing at record lows, homeowners will agree to a sales pitch that pushes the biggest system possible: 6.6kW. The vast majority of generation, however, is exported.
The report predicts total installed PV capacity below 30MW of between 75GW and 33GW by 2050, under five scenarios. The growth of solar will not be steady going, however. Edis predicts a dramatic fall in sales of new systems and replacements of old systems by 2021 and 2022. This year, 2020, may go down in history as the peak for solar. A drop off around 2028 reflects the end of rebates in Victoria.
A growing portion of the market from around 2022 will come from replacement of old systems, he says. “We think all people will replace their systems, and that is going to be a major contributor to keeping the market afloat.”
Edis expects feed-in tariffs will “plummet” over the next two to three years to staunch the effects of rising levels of solar on the grid. “The lowest point of the day on the wholesale market is now midday,” he says. In effect, the success of solar has screwed things up for solar. It’s not great news for the solar industry. “Revenue [from exports] for household solar systems is going to plunge, and that’s going to make it much harder to sell a solar system into the future.”
Edis expects feed-in tariffs in South Australia, for example, to fall from 16c/kWh to 4c.
The second constraint to sales of large home systems is already coming over the horizon as networks set limits on inverter export capacity. And thirdly, the networks are going to move from a set price per kWh, irrespective of time of day, to time-of-day tariffs where the daytime tariff is lowest. In South Australia, for instance, SAPN’s approved network charge during the peak solar output period is 25% of the 24-hour charge, he says. The combination of those things – a fall in feed-in tariffs, emerging network constraints, a fall in network charges – will be a major blow to solar businesses.
But it’s not all doom and gloom because it will be a great time to sell batteries – “especially from the revenue side”. It won’t be a natural transition. Solar retailers will need to sharpen their battery sales skills, instead of looking at it as simply too hard when there is other work to do. “The only reason a battery looks remotely OK in a payback is because its piggybacking off the solar payback looking so good,” Edis says. That will change as feed-in tariffs drop.
Based on an assumption of substantial falls in the cost of storage, Edis expects batteries will be mandatory inclusions new system purchases for homes and businesses from 2029. If the prediction for 33-75GW of solar under 30MW by 2050 looks excessive, bear in mind most of it will be absorbed by batteries for later use. Also, an aggregate of output for all systems in the NEM will likely never exceed 60% of rated capacity at any time.
The authors didn’t evaluate the impact electric vehicles will have on the market, but some are predicting they will evolve into powerful grid assets, whether as fleets or privately owned “batteries on wheels” capable of soaking up surplus solar. Community batteries were also left out, relying as they would do on the network sector showing rare lightning initiative.
The slow removal of state incentives and the rundown of the CER’s Small-scale Technology Certificates by 2030 will see revenues from solar fall between now and 2030 and fail to keep up with inflation until 2050. “Post-subsidy, we’re anticipating the cost of a solar system will rise over time,” Edis says, as the phenomenon of system cost reductions outpacing government subsidies peters out.
If the report is right and 2020 is peak year for solar, you’ve reached the end of the story now so you’d better get back to work.