Yesterday I wrote about how the NSW utility regulator, IPART, had decided that households should only be paid 5-8 cents for the electricity they export to the grid from their solar PV system, whereas they’ll be charged somewhere between 30 and 44 cents for any electricity they consume from the grid.
That article explained that the difference between what IPART felt households should get paid for electricity they export versus what customers pay for electricity they import, is related to charges for network infrastructure (poles, wires, transformers etc.). These network charges represent nearly half of the average household electricity bill, so it’s a big deal. According to IPART, “PV exports are unlikely to provide system-wide benefits that will materially reduce either distribution network or transmission network costs in NSW.”
IPART came to this conclusion because the vast majority of solar PV is installed on residential households, which impose their greatest demand on the network between 6-8pm. This is a time when solar system output is low. Networks infrastructure capacity is governed entirely by the maximum peak in demand, not average demand over the day. So even though solar PV reduces the load on networks over much of the day, this makes relatively little difference to infrastructure requirements.
Of course, as I said in yesterday’s article, this completely misses the fact that household demand and solar output could be shifted in time so that loads on the network were reduced. And also misses the fact that government has so far refused to implement proper time of use pricing that would encourage such a shift.
But it also prompts the question, what if we installed solar PV in distribution networks which primarily serve businesses rather than households?
Well Western Power is one of the few network businesses in Australia that has actually bothered to examine this and publicly reveal the data. They simulated the likely effect of solar PV on peak demand for their network areas serving commercial businesses (offices, retail); industrial loads (manufacturing) and a mix of commercial and residential loads. This involved simulating a scenario where solar installed generating capacity was equal to 10 per cent of the peak demand of a local network area.
For network feeders serving commercial businesses, solar PV output is a pretty close match for taking out the peak. If you installed PV equal to 10 per cent of maximum demand, you’d take 6 per cent to 7 per cent off peak demand, which is illustrated below. The blue line is demand without PV, and the other lines represent the effect of 10 per cent solar PV under varying levels of solar radiation.
In terms of an industrial area you get a slightly better result of about 7 to 8 per cent reduction.
And mixed commercial and residential provide a reduction between 5 and 7 per cent.
But guess what?
If you install a solar PV system in these network areas, or any other type of power generation system, there’s no transparent, standardised pricing system for power you export to the grid. Instead, you’ll have to negotiate a payment with a 900 pound gorilla called the monopoly network business. And this gorilla has little incentive to negotiate nicely.
These charts also make you wonder why on earth government support policies for solar PV focus on either:
-- Small household systems (The REC multiplier and state government feed-in tariffs have system size restrictions) or;
-- Extremely large systems (Solar Flagships supports installations of 50MW )
These both provide relatively little reduction in network peak demand, meanwhile government support for PV systems in commercial and industrial areas is almost non-existent.
Surely it wouldn’t have anything to do with political mathematics?
(households = voters; large fields of solar panels = photo opportunity).