With VIC & SA in a heatwave this week I remembered to re-post my Conversation article about solar and peak electricity demand, five months later. I’m not very good at blogging.
I’ve added a bonus graph to the post.
When the sun don’t shine, the power don’t flow … or does it?
By Kerry Burke, University of Newcastle
Renewable energy seems to be on a roll. One million Australian homes have rooftop solar cells. There’s so much renewable energy it’s reducing wholesale electricity prices. But then, that old chestnut pops up: reliability. How do we make energy when there’s no sun?
When the heat is on, the sun is shining
The electricity market is effectively two separate markets. There’s the market for energy: the coal, gas, operations and maintenance component of running the network.
Then there’s the market for capacity. To serve our electricity needs, we build enough power stations to supply our highest imagined electricity needs, then let some or most of them sit idle for all the parts of the year when we’re not using quite as much electricity.
We tend to think about the cost of power as being related to things such as coal and gas prices, or operation expenses; but a large part of the cost of electricity is simply the cost of money involved in having standby generators ready for that one moment when we all want to use electricity at the same time.
This creates a problem for the owner of a solar or wind farm. Because they can’t guarantee capacity due to the vagaries of weather, in effect, they can only sell energy
But distributed solar is different. The time when we all decide to use a lot of electricity at the same time is when we all turn our air conditioners on mid-afternoon on a scorching hot summer day. At that time, the sun is clearly shining. For distributed solar to be interrupted, it would have to be overcast over an entire capital city, in which case the temperature wouldn’t be that high and we wouldn’t actually need spare capacity.
Alongside the growth of solar installation, there has been a growing enthusiasm for sharing data on solar output. One such site is PVOutput.org. This large sample of real-time solar system behaviour under real world conditions is combined with data from the renewable energy regulator on installation rates. It allows real-time solar generation in the National Electricity Market to be calculated.
Using this data we can have a look at how renewable energy actually handles peak demand.
During the demand peak on top five highest demand days in each state, solar reduced demand by at least 21% of its installed capacity in each state. Surprisingly, Victoria and South Australia are actually better at supplying peak demand with solar. The low latitude creates long days and daylight saving time shifts consumption earlier into the day when the sun is higher in the sky. Less surprising is that north-west and west facing solar are better at meeting peak demand.
Moving from gas to solar
It’s even possible to put some rough numbers on this ability to meet peak demand. A lot of the recent work on the value of distributed solar has been conducted by regulatory bodies as they try to determine a fair price for solar feed-ins.
The regulators and their consultants all fall into the error of assessing distributed solar as only an energy source, without regard for solar’s value as capacity.
The easiest way to think about peak capacity is to value it the same as a gas turbine. In fact distributed solar has several advantages over gas turbines. Solar can be installed incrementally, as needed. New transmission lines are not needed. More importantly, a large portfolio of small solar systems will always work. Like a car engine, gas turbines sometimes fail to start.
Nevertheless, gas turbines are the electricity market’s standard response to peak demand because they are the cheapest type of conventional generator to build. Comparing distributed solar to the cheapest generator should understate its value as a provider of peak capacity.
It is of course necessary to correct for a few things; solar doesn’t produce at its full rated capacity in the afternoon with the sun striking at an angle and gas turbines suffer transmission line losses and output reductions in hot weather.
After crunching all the numbers, the ability of distributed solar to supply peaks has capacity value equivalent to 10-20% of the unsubsidised installation cost. That’s on top of the energy value as calculated by all the regulators. It’s worth more in the south, because of the longer days. It’s worth another 3-6% of the system cost if the system faces north-west or west.
Importantly, the value is a capital value, not cents per kWh. Capacity isn’t about how much power is produced. It’s about how much is there when you need it most.
A working paper with more detail is available here
Kerry Burke is affiliated with Westpac and the University of Newcastle.