Virtual power plants (VPP) get quite a lot of press from time to time. These consist of a battery of batteries under software control that can be used to supply a grid shortfall.
Is calling them a “power plant” just PR spin for a way of gaming the rules of the electrity grid to make money, or are they actually a power plant?
The concept of people making money from selling the extra electricity they get with a roof full of panels is pretty attractive to people with the money to cover their roof with panels and spend thousands on batteries. Meredith Anguin's book "Shorting the Grid" took the above image (from the US Government) to task for being singularly misleading. Will it be black women with bucket loads of money installing private wind turbines and battery banks? No. It will generally be rich people, meaning old white males with plenty of spare capital.
UAE Barakah Vs Tesla Powerwalls
Can we compare a massive virtual power plant with a massive real live power plant?
On the one side we’ll have the UAE Barakah nuclear reactors, a set of 4xAPR1400 South Korean nuclear plants. These were built over the past decade and are coming on line as we speak.
On the other side we’ll consider a virtual power plant made from rooftop PV panels and batteries.
How many Australian households would Barakah power? The Darlington Point) solar plant reckons it can power 115,000 homes from its estimated annual energy production of 685 GWh. Which is around 6MWh per annum per household.
As I write (Feb 2021), AGL is offering a rooftop PV and battery bundle for $15,244. The PV system is a 6.6kW, which will produce about 9MWh per annum in Australia, so it’s clear that the Darlington Point website authors were only talking about modest electricity users and not people signing up the to AGL deal. But if a modest user was to sign up, they’d clearly have plenty of extra electricity to sell from that battery.
First, how many households can the Barakah plant power? At 6MWh per year … about 7.3 million. What about high-end users of electricity, those using 9MWh/year? Barakah would power about 4.9 million of these.
So let’s compare Barakah with a virtual power plant of 4.9 million modest users with a 6.6kW rooftop system selling off their excess to anybody who wants it. Both systems need a grid, so we can ignore it in our comparison.
Which is more expensive, Barakah or the AGL virtual power plant?
The Barakah plant cost about AU$30 billion to build, and may well cost as much again to operate over its 60 year lifetime. The lifetime costs are just a guess; based on an estimate I saw on the World Nuclear News website. The costs of the AGL VPP are simple to calculate, just multiply 4.9 million by the price of the units; which comes to about AU$74 billion dollars. The batteries have a 10 year warranty, so being generous, let’s allow them 15 years. So we need to multiply the cost by at least 3 because we’ll need to replace all these systems. So the cost is now looking like about $200 billion.
But that’s just one of the costs. There are ecological costs also.
The fuel costs of nuclear reactors are tiny. You need to mine about 1,120 tonnes of uranium per annum to keep Barakah operating each year, but the amount of energy produced is so large that the cost per kwh is tiny; fractions of a cent.
But the distinction between fuel and infrastructure is rather arbitrary. It used to be that the consumables in running a laser printer were the paper and toner. Then they became the paper and the cartridges containing the toner. And now some cartridges even contain contain drums. Printers are now so cheap and shoddy that they are almost throw away consumables in their own right.
Batteries act like consumables in a VPP. As an aside, this is particularly obvious if you think about lead-acid batteries. They are cheap but don’t last long, particularly if you cycle them in ways they weren’t designed for. But they are recyclable. But recycling lead-acid batteries is very carbon intensive and involves several highly dangerous toxic processes. In rich countries the industry is highly regulated, but globally, things are different. Lead acid battery recycling is the most toxic industrial process on the planet, as measured by loss of healthy years of life to those involved. End of aside.
Lithium-ion batteries are a better fit for a VPP than lead acide, but still have a relatively short life span and are expensive and difficult to recycle.
The weight of the batteries in our AGL VPP, assuming we are using Tesla Powerwalls, is 559,238 tonnes; and they need to be replaced every 10-15 years.
When I tell people that uranium mining is a major environmental plus for nuclear power, they are incredulous, but the numbers make it clear. Over the course of 60 years, Barakah will need somebody to mine about 67,200 tonnes of uranium. That’s tiny compared to tthe mining required for the batteries. Much of the battery mining will, for the foreseeable future, come from sweat pit mining and processing in Africa and China. In the US, most uranium mining is done without a mine. You drill a couple of holes and pump in some slightly saline water; you can them pump out the water and it will contain uranium. It's called in-situ leaching. But where, as in Australia, the uranium is is a sideline for copper mining, a more traditional hole in the ground is used. Olympic Dam is a copper mine and BHP is busily talking up their green credentials by saying their copper will power the green future. And it's true. Olympic dam produces about 230,000 tonnes of copper and 4,000 tonnes of uranium annually ... enough for about 20 reactors.
But aren’t I ignoring the material used to build the actual reactor? Isn’t it huge? The concrete building is certainly fairly large, but the bit that does the work is about 13m high and 4m wide; the reactor pressure vessel. There’s not much difference in the size of a research reactor and a power reactor. It’s just that most of the research reactor is empty and usually visible as an open pool of water. Whereas the power reactor is packed with rods of uranium fuel.
The reactor pressure vessel for an APR1400 (a very large reactor in output) is about 780 tonnes of very sophisticated metal fabrication. An 8MW wind turbine motor is nearly 500 tonnes; so 780 is actually tiny relative to the 1400MW output.
But a reactor has a considerable amount of reinforced concrete; and pipes, lots of pipe.
Data on exactly how much reinforced concrete is hard to come by, but back in 2009, there was a document on the Westinghouse website that gives us a good estimate: perhaps 100,000 m3 of concrete and 12,000 tonnes of steel reinforcing. For our VPP, each participating family has about 500 kg of panels on their roof, plus the electronics. In total our VPP needs 107 million panels weighing 2.4 million tonnes; not including the electronics.
Energy and material costs these days are much tougher to grasp than they used to be. The tiny size of a mobile phone gives no hint as to its energy and material cost. An article by energy analyst Vaclav Smil calculated the following:
“New cars weighed more than 180 times as much as all portable electronics [produced each year], but required only seven times as much energy to make.”
So it is with distributed energy like rooftop solar. The massive ecological costs underpinning renewable energy aren’t visible when the final products are spread out; but they are just as real.
Summary
I began by saying I wanted to structure a VPP that would make for a reasonable comparison with Barakah. If anybody can think of more simple features to improve the comparison, then please let me know.
But there are still problems with my assumptions. Barakah provides true 24x7 reliable power. The VPP does not. There will be many times during a year when the batteries aren’t able to get you through the night because of a run of dull days. That’s not a problem if you are on a reliable grid, but it does indicate that we are still not comparing apples with apples.