How much is too much PV?

The BBC reported today that the owner of a caravan park in Northern Ireland was refused permission to install a 50kWp solar panel array because the local substation could not handle the load - he was only allowed 20 kWp. This comes just a couple of weeks after another item reporting that solar PV panel applications were booming as developers tried to get in before the grid was swamped. The latest figures from DECC show that Solar PV supplied only 2.9% of renewable electricity in 2012 - that is 1,200 GWh from PV compared to 20,000 GWh from wind and wave power, and 41 GWh in total [3]. Renewable sources as a whole contributed only 11.3% in 2012, so PV provided only 0.3% of supply in 2012 [4]. On the other hand, a briefing note from the National Grid to DECC reported the grid could handle 10 GW of PV capacity - in 2012 we had 1.7 GW [5]. So what is the problem?


The National Grid paper was looking at the impact of Solar PV installations on total supply, and how this could be matched to demand. National Grid said that above 10 GW of capacity there would be problems on sunny summer days because PV supply is at a peak and demand is fairly low. We would need to constrain wind farms more often and there would be increasing difficulty with voltage regulation. However, there are also very severe constraints at the local level, of various kinds.

In general, cables and transformers don't mind which way power flows, they can handle the same power going in or coming out, so it should be OK to generate as much as we ever use. The biggest demand is on winter evenings and the worse case average domestic peak demand (coldest day) is very roughly 1.6 kW/household. This means if more than a third of households have a 4 kWp array the cables may not handle it. Well that is still a lot of capacity. However, that is not the only issue. Here are some of the problems that can occur when there are power generators distributed through the grid.

Where you have long cabling lengths you get a variable voltage drop going down it, depending on the power being drawn off by the houses and businesses conntected to it. This is corrected by equipment installed at suitable intervals and controlled by monitoring devices at rather longer intervals. However, if there are generators along the line as well as loads, then the voltage may increase rather than decrease and the monitoring equipment can't tell what is going on everywhere, so it is possible some of the line may go over voltage. Going over voltage is worse than going under voltage because it can burn things out.

There is another more serious safety issue due to generators being the 'wrong side' of the monitoring points. If there is a short circuit somewhere, perhaps caused by a squirrel in a substation then this should be very quickly spotted up the line because of a drop in voltage and the fault will be isolated. But given sufficient power generators on the line they can 'hide' this drop so the section is not isolated, power continues to flow and could cause a fire, or worse.

Another problem can arise due to safety equipment in the PV generating equipment. The invertors that connect your panels to the grid are designed to shut off if they detect a fault on the grid, usually because the voltage drops - this is important so that the engineers can fix the grid without PV generators electrocuting them. Also, as well as monitoring the voltage they also check the frequency, and shut off if this goes out of normal range. Unfortunately a drop in frequency could be due to a sudden extra load, in which case taking generators off makes the situation worse. Of course a few small generators won't make much difference but a lot could turn a minor glitch into a brown out.

The National Grid has to operate safely and there would quite rightly be huge outcry if anyone were hurt by their equipment. They are also expected to meet high standards for reliability, voltage and frequency control and quality. So we can understand, even applaud, reasonable caution in allowing connections which could conceivably compromise safety or reliability. Around the world different grid systems have different guidelines. For example in Ontario, Canada, the limit for any branch of the network is 7% of peak winter demand whereas in California they allow 15% as a matter of course, and are considering up to 50%[8].

The problems I have mentioned, and others which I have not, can all be mitigated with additional monitoring and/or smarter equipment - but there is a cost to installing this. A PV installer based in Cambridge tells me that a couple of projects he has been involved in were restricted in size, to around 30 kWp. They would have been allowed more if they paid to upgrade the line, but that would have been too expensive. So, it is quite likely that local constraints will be a more severe limitation to renewable micro-generation than the national supply level problem in many places, especially in rural areas where equipment is old.




[1] Parts of Northern Ireland's electricity grid overloaded BBC 13/Nov/13
[2] Grid capacity worries spark UK solar farm boom BBC 29/Oct/13
[3] Regional Renewable Statistics (DECC)
[4] Digest UK Energy Statistics 2013 (DECC)
[5] Solar PV (electricity systems) and the National Grid: a briefing note for DECC (Dec 2012)
[6] How Much Should Self-Generators Pay For The Grid? (EU Energy Policy Blog 1/Nov/2013)
[7] PV Generation and its effect on utilities SolarPro
[8] Islanding (Wikipedia)