When talking about the rollout of distributed energy sources it’s really important to consider different technologies from the perspective of different parties. There are some big trends in the energy market now; solar is booming *still* in Australia, stationary batteries are coming and EV charging is starting have an impact, and one’s view on each of these trends will depend a lot on whether or not one benefits from them! This post is going to discuss how thinking about these technologies from the owner’s view can give us a fair idea of where these trends are headed. I’ll illustrate with some examples.
If you have a poke around the excellent Australian PV Institute website you will see the amount of installed solar just keeps increasing. This is what we expected. Folks like Tony Seba, with strong “from the customer’s perspective focus” have been forecasting this since about 2010 and earlier. The theory goes, that from the first solar installation this snowball has been growing. The first solar applications, probably space or military, looked really good from the users’ perspective. They could get power in a remote location cheaper than alternative sources and so built a factory to make solar panels. Then it turned out that there were other places that needed electricity and those users bought solar panels to solve their problems. This made the first solar factory busy enough to make someone build another one, but this one makes slightly better value panels. Now new consumers are looking at the panels because they’re a bit better value and the market of consumers opens up. Every time a new factory makes better value panels the number of places where solar is a good idea increases. Looks like Tony was right.
This has been quite a disruption to energy markets. In Australia, consumers who previously didn’t have much option apart from buying grid power or not, could suddenly make their own electricity and give a lot less money to electricity generators and networks. Solar brings technical and financial challenges to the utilities and over the years they have changed the rules to try and protect their income. Again, this is as predicted, sometimes doomily called the “Solar Death Spiral”. We know solar keeps getting cheaper, and increasing the number of places it’s a good idea to install. So it gets installed. Utilities earn less money so they change the way electricity is priced to try and earn the same money from less electricity sales. Solar gets cheaper and more households buy solar, this time targeting the new rules. Utilities earn less, change the rules and consumers respond. Electricity metering changed from gross to net, as is happening in California at the moment, and households responded by installing more solar. Tariff structures were changed, feed in tariffs reduced, export banned in some locations and now networks want to be able to turn off solar remotely. As we can see from the APVI data, each time the rules were changed Australians responded by installing more solar.
When the rules change, the solar impacted by the change is already installed, so it might as well stay installed making electricity. There’s no point removing it, even if it’s not generating as much value as previously. The impact is that the utilities just get to sell less and less electricity and try to recover the same costs from fewer kilowatt hours. The Death Spiral theory goes that in about 2040 the utilities will be competing to sell one kWh each year for about a trillion dollars. That forecast seems about right.
So my first forecast outcome, which isn’t particularly bold, is that solar will keep getting installed, even if there are some ways in which it doesn’t look like it makes economic sense. Folks have been fretting about price cannibalisation, where the prevalence of rooftop solar depresses daytime electricity demand and reduces the revenue of grid connected solar and wondering when households will stop. But they won’t, because from the household’s or end user’s perspective, I realise I’m using them interchangeably, that doesn’t impact their decision. Solar keeps getting cheaper, it is cost effective against reduced grid power use, grid power keeps getting more expensive, so there will be more places where it’s cost effective, and the value of grid power during the day will keep falling. This will probably lead to periods where grid solar installations don’t seem like a good idea, like now, but I suspect that will just be part of the ebb and flow. Solar overall keeps growing and I don’t think it will stop.
Extending this idea to batteries, and I talked about this at length in the battery post in the Capacity Problems section, taking the end-user’s perspective I think we will end up with a massive overbuild of batteries.
Batteries are deployed to solve problems, and those problems won’t necessarily take all of their time. For example, the Goulburn EV Charging battery in the Battery Post, which I worked on, is only needed when all the EV chargers are running and the sun isn’t shining, maybe 10 hours a week. This requires about 10 hours of charging each week, leaving about 150 hours for the battery to do other things, like be available for frequency correction or playing in the wholesale market during the day. The cost of the battery is paid back by the work it does for that 20 hours a week, so from the battery owner’s perspective any further revenue is a bonus. So when the battery bids into the frequency market against gas turbines and hydro plants, the battery doesn’t need to make much money because it was paid for elsewhere, and probably got charged for free from excess solar somewhere, and is competing against technologies that have fuel costs. Batteries will beat hydro and gas in both frequency and wholesale markets because of this.
This happens at large and small scales. Households have started installing batteries, roughly 200,000 at this point. Assume these are 5kW each, which is a standard Tesla Powerwall 2, and that’s about 1GW of available power. From the user’s perspective, the economics of residential batteries aren’t great, most analyses I’ve done have the payback at about 10 years, when solar is closer to 5 in most places. That doesn’t mean they’re worthless though, the economic value might not be great but buyers are getting something they value, and presumably that value doesn’t require the battery to continuously charge and discharge, allowing these batteries to be deployed in other uses.
At large scale the Waratah Super Battery demonstrates the same idea. This is being deployed as a “virtual transmission solution that unlocks latent capacity in the existing transmission system”. In other words, the existing transmission wires have times when they want to flow more power than they can. These events are short lived and infrequent, maybe one or two hundred hours a year, leaving about 8000 hours a year for these batteries to provide other services in the market. So batteries are getting installed and paid for by savings that look good from the user’s perspective and don’t use all of the battery’s capacity. Batteries are getting cheaper so the number of places where there are good installations from the user’s perspective will grow. These batteries too will be underutilised and have spare capacity for market services at very low cost and so I predict we will end up with a lot more batteries than people think we need for a clean grid. They will be high-power, short duration batteries, solving capacity problems, not long duration batteries mimicking incumbent generators. As an aside, this is why I think Snowy 2 is not the best use of funds. It is effectively a 2-week battery and I don’t see many 2-week long problems in the grid.
Is this happening? Are a lot of batteries being installed and do we have evidence that they’re ahead of schedule? It’s a hard question to answer at the moment, we’re early in the battery transition and the data on batteries isn’t great, but I’ll have a big crack at it when I revisit the Forecast looking back on 2023. Sunwiz haven’t updated their market snapshot for 2023 yet, but at the end of 2022 had Australia at about 180,000 battery installations for a total of almost 2,000 MWh of storage. At grid scale OpenNEM has almost 2GW of registered storage connected to the grid. For future batteries RenewEconomy’s Battery Map has 6.2GW under construction (across Australia) and about double that announced. So roughly 1GW household, 2GW grid scale and 6GW in the pipeline. On track for about 10GW/20GWh by the end of 2024. The Integrated System Plan hopes for about 10GW storage power in 2025-26 and about 20GWh, so maybe we’re a tiny bit ahead of schedule.
My second forecast then is that we will end up with enough storage to meet the ISP forecast with battery projects that were cost effective from the owner’s perspective. As with solar, consumers will keep installing batteries to reduce the amount of electricity they buy and will at times have more battery than they need. This spare capacity will be used as grid storage and I can easily imagine having enough storage installed to meet or beat the ISP timeline above.
For this reason I think Snowy 2 will look redundant when it’s finished. 2GW is the same power as all the grid scale batteries installed now, which can cycle about 2GWh/day. By 2035 we could well have 350 separate 1GW batteries that can cycle daily rather than every two weeks.
How do we apply these lessons to EV charging? I think the scale of EV charging helps. An “average” Australian household uses about 20-30kWh of electricity per day. If they drive an average EV the average daily distance it adds about 7kWh to their electricity consumption. It’s hard to verify, but I’ve heard that about 80% of EVs in Australia are charging on a little charger that plugs into a normal socket. It makes sense. They only need about 7kWh each day, the wall socket can deliver about 2.3kWh per hour, in 3 or 4 hours the car is completely recharged. This is a cheap and low fuss solution. We charged two cars for about 6 months in our house with that system.
What got us to change, and what might provoke others to move to a smart charger or V2G system in future? We went to a smart charger because we kept tripping the circuit the charger was on. Most power point circuits are rated to about 16 amps and if 10 amps is devoted to the car charger if I put the washing machine and dishwasher on at the same time the internet went off.
I’ve argued elsewhere that dynamic power management or smart charging solves most EV charging problems. Capacity problems are short lived and cars are usually plugged in for much longer than they need to be, so it’s trivial to slow down charging briefly to avoid upgrading the power supply, which is very expensive. From the household’s perspective smart charging fixes my over-current problem by putting the charger on a dedicated circuit, and including a power meter in the system that measures demand in the switchboard and moderates charger power accordingly to stay within limits. So now our cars charge as fast as they can, with the capacity is available. This step in control cost us about $2,000, charger and installation.
Having a smart charger allows a household to manage the charge rate of their car, which from the grid’s perspective is the same as charging a battery. We have a grid problem of high voltage from solar output and having a load of EVs on smart chargers would be one way to manage this. Have the cars ready to charge and ramp up the chargers as the sun comes up.
If every household got to this point the grid would have a good flexible load to damp voltage rise from solar during the day. That’s worth something, but even better would be the ability to take power back from the battery and meet the evening peak. Looking at the way our solar and battery works we need power from about 5pm to 9pm to meet the cooking dinner and showering peak. After that power use drops to almost nothing until 5am when I launch out of bed and turn on the coffee machine.
Thinking of the problem at this scale helps too. The average car is using about 7kWh/day, a decent commuter might be closer to 20kWh/day. Car batteries are mostly over 60kWh. Our old Leaf is 24kWh but the modern Teslas, Hyundais and BYDs are 60kWh and higher. There should be spare capacity in the car batteries, and the value to the grid is in charging them during sunlight hours and discharging them from 5-9pm. That means cars plugged in when the sun is shining at work or home, then discharging at home during the evening. What is required to access this?
Taking power back out of the car requires a different type of charger, that can take DC from the car and turn it into AC electricity at the right frequency and voltage for the grid it’s connected to. We call this function V2G, vehicle to grid. This is pretty much a bidirectional solar inverter with a DC car charging plug on the end. Early examples of these chargers have been too expensive in Australia, but they look to be obsolete now. The first plug standard approved for DC charging was CHAdeMO from Japan, rather than the more common (in Australia) CCS2 which EU, Korean, Chinese and US manufacturers all use. That left Nissan Leafs as the only viable V2G vehicle, which is a small market, so I think the chargers were servicing a small sector to recover their costs. I mention this mostly as a reason why we have never taken the step, despite being well set up for it. Our Leaf is stationary most of the time and we have loads of excess solar during the day. But the charger was about $10,000 installed, and might have saved us 10kWh per day, maybe $2 electricity per day at best, about a 14 year payback and there are better ways to spend the money. I could add 10kWh to the spare Solax inverter I have for about a third of that price.
Things are about to change though as CCS2 enables V2G in new cars and different chargers become available. I’m optimistic that we’ll see a 5kW, bi-directional charger for CCS vehicles in Australia by the end of 2025 for about $3,000 installed. I’ve got a SolarEdge SE5000 inverter which is a 5kW bidirectional solar inverter/battery charger that we bought in 2016 for $2,500. If that happened the payback drops to about 4 years and we would probably take that step.
How many EV owners will take this step? Given the assumption previously that about 20% of EVs are charging on smart chargers at the moment, I’d be surprised if take-up of V2G in the home got above 30% by 2035. I’ve oscillated on V2G over the years. Early thought from Tesla and JB Straubel was that it was an expensive use of a car battery, and for a Tesla Model S that cost $250,000 that was probably true. Cars are getting cheaper now though and there are 40kWh EVs available for about $40,000. If we can make that V2G for another $3,000 that sounds interesting. From an Australian policy perspective there is a lot of potential value in car batteries. 200,000 V2G enabled vehicles delivering 5kW would deliver 1GW of power during the evening peak, about half the power of Snowy 2 which will be 2GW in 2029. There were 100,000 EVs delivered to Australia during 2023, how many V2G vehicles can we have enabled by 2029? I think V2G will go close to providing more power than Snowy when it turns on. CSIRO forecasts we’ll have 2.5m vehicles by then and at least 11% of vehicles able to contribute some V2G which seems reasonable.
I don’t have a firm forecast on V2G. I think uptake will really be driven by the hardware availability first, then the electricity policy environment. If there are cars in Australia, approved by their manufacturer for use in V2G, and reasonably priced CCS2 chargers become available then I can see loads of households taking that step. Without these ingredients V2G could easily be the sort of technology only nerds like me are interested in, which is probably not the best outcome for Australia. There’s a lot of potential emissions reductions in that spare battery capacity if we can find the policy and technology drivers to access it.
Summing up, and I’m aware this will read as “the market will save us!” to some, but: I think the Australian grid will get as much solar and battery storage as we need to meet the ISP plan and ultimately decarbonise the grid, just through households and businesses making decisions that are rational from their perspective. And the broader point is that if you don’t consider the transition from the perspective of the end user you can forget that people who already have solar installed don’t care if grid solar operators make less money.
Evcricket's Energy 