You’ve heard the one about how the growth of grid-connected solar power is causing voltage rise on low voltage (LV) networks?
I sure have. I’ve written plenty about the phenomenon, often with an eye to what networks are doing to try and counter the problem.
However, courtesy of an e-mail from Green Energy Markets’ Tristan Edis, I have learned to my surprise that the “blame solar power” narrative simply isn’t true. Yes, voltages on the LV network are routinely – almost always – out of compliance, but it barely matters whether solar is on the network or not.
Edis’s e-mail alerted us to a huge report that landed in May, at the height of the COVID-19 lockdowns. The COAG Energy Council’s 175-page LV Voltage Report was prepared for the Energy Security Board by the University of New South Wales, and is part of the Energy Council’s DER Integration Plan.
In particular, Edis drew our attention to a surprising finding: voltage rise is rampant in National Electricity Market (NEM) networks, even when those networks have only a low penetration of DER.
This is the finding reported in the LV Voltage Report’s cover note [PDF]:
“The key finding of the paper is that, even in the absence of solar PV, there is a significant level of high voltage across all DNSPs in all NEM states as highlighted in the following chart. The nominal voltage standard in the NEM is 230V – more than 95% of readings were found to be higher than this.”
Further:
“The increasing level of output from solar PV is, in turn, slightly increasing the voltages on most electricity networks. However, the report also finds that many sites experience higher voltages during the night when solar PV is not operational.”
“Higher voltages at night” is a far cry from industry (and media) hype that solar power is causing meltdown in our LV networks. As Edis remarked to us in the e-mail, it looks a lot like networks want solar to carry the blame for their own over-voltage non-compliance problems.
How much non-compliance? In UNSW work by Baran Yildiz presented in the report, in just one DNSP:
- At least 22% of sites in the dataset experienced voltage over 253V at least 1% of the time;
- At least 10% experienced voltage over 255V, 2% of the time;
- 3% of sites experienced voltage over 253V at least 20% of the time; and
- 2% of sites experience voltage over 255V, at least 10% of the time.
Suburban voltages in the NEM. Source: COAG Energy Council LV Voltage Report
If Not Solar, Then What?
The report’s cover note says network operators have, in part, decided to let LV network voltages rise “to support additional air conditioning loads installed over the last three decades”, and in part, it’s because 20 years ago Australia adopted a nominal 230V standard, but some DNSPs are still to transition from the older 240V standard. And, of course, voltage rises and falls with variation in network loads.
In spite of apocalyptic warnings that too much PV is making the grid unstable, solar power’s impact on the grid is so far mild – in most postcodes, solar has contributed around 2V rise per kilowatt exported from PV*. Across both NEM states and DNSPs, the median impact on LV networks ranges from 2V to 5V.
(*I’ve been corrected by study author Simon Heslop in the comments. I’ve left the original text in place, but will quote from the report’s cover note for clarity: “The UNSW analysis showed that although sites in some postcodes experience significant
voltage rises per kW of site PV export, most are lower than 2 V/kW export. At both the state/territory and DNSP level, the median voltage impact at ranges from 2V to 5V. There is an impact on voltage from solar PV, and particularly the distribution of higher voltages. In turn, voltage impacts solar PV exports.”)
Sure, we can expect the impact to grow as more rooftops get more solar panels – but only if DNSPs do nothing to manage the transition.
Curtailment? Mostly Not Much, Costing Less Than $12 Annually
Of course, high grid voltages lead to curtailment either on a dynamic basis (a solar inverter’s output is scaled back on instruction from the network) or by limiting new connections to the network.
UNSW PhD researcher Naomi Stringer spoke to the Webinar launching the report about her group’s work estimating the financial impact of curtailment on solar owners, and it turned out to be mostly minor.
Stringer’s team analysed 1,300 systems in South Australia, using data collected every minute on 24 clear-sky days, and assuming a 25c/kWh retail price and 6c/kWh feed in tariff for its financial modelling.
Curtailment is happening – Stringer found 53% of sites experienced some curtailment, but the average curtailment across all sites was just 1%, and the average across curtailed sites only 6%.
The low levels of curtailment are reflected in the minimal financial impact it’s currently having on customers. Stringer’s modelling found most experienced so little curtailment that the financial impact is between $3 and $12 per annum.
“However, there is a small number of prosumers which are significantly impacted and can experience considerable financial penalty. The most impacted prosumer is estimated to lose approximately of the order of 30-90 per cent per year.”
For most of us, though, buying a battery to avoid curtailment is not a good investment.
About The Study
The dataset the report used is huge: Solar Analytics’ roughly 12,000 sites cover all the NEM states and DNSPs. Voltage at those sites was sampled every five seconds for as long as a year, and exports every five minutes.
About Richard Chirgwin
RSS - Posts
I think this revelation warrants an inquiry.
It has been well know in energy circles for years. Everyone knew it was the case, we just didn’t have the data or analysis to quantify it. Now we do. And to their credit the DNSPs, ESB, AER and AEMO (surely I can squeeze in another energy TLA) are all taking appropriate actions and working to enable more rooftop solar
Watched 101 Solar today, comment was made about the best system would be a 6.6Kw, I had a 5Kw good quality system installed 6 years ago by SunWise, Bunbury WA. Recently was told that to add 6 panels (1.6Kw) facing west, would cost another $3000,-.
Just wondering if that seems like a fair quote?
Thanks Robert
Hi Robert
That is a significant amount to pay for an extra 1.6 kilowatts of solar for an older system. In WA it’s likely to save you under $200 a year so it will have a long simple payback period. You may be better off investing in another energy efficiency measure that will have a faster payback. Alternatively, you could save your money and in the future if you decide to get a battery or electric car you could upgrade your whole solar system then and replace it with a new larger one.
As for whether or not it is a fair quote, it’s not a small job as it will involve the rewiring of existing panels. There is also a good chance the extra panels will require optimizers which adds to the cost. Without knowing the details of the hardware used it is difficult to say how fair the price is. But if it is being done by Sunwise I know they do good quality work.
Yes, I believe in Santa Clause too!!
Follow the money trail and find the companies gaming the system to their advantage. They have form in gaming the system.
The small solar generator gets screwed as usual. My bills show this since December last year.
“it looks a lot like networks want solar to carry the blame for their own over-voltage non-compliance problems.”
OMG! Who’d have thunk it?!~
I had quite a bit of trouble with overvoltage (postcode 2576). It caused my solar to disconnect on many occasions and we did complain. They did wind it down a bit but eventually we moved the solar to another phase, which was more stable.
The voltage was frequently over 255v and I have noted 270v on a number of occasions. How much extra strain is this putting on our lights and other electrical equipment? I reckon it’s costing us quite a bit!
That’s not great, but most electronics won’t be particularly affected by a higher voltage. The power supply may have more strain on it though.
Resistive heating (toaster, kettle, hair dryer, electric hot water, etc) and incandescent lights use power in proportion to the voltage, so these will definitely get hotter and last a shorter time when running on a higher voltage. And your toast will also cook your toast quicker.
If I recall my electrical theory correctly the power is not proportional to the voltage but increases with the square of the voltage. So your power consumption goes up significantly with a small over voltage.
P= V x I = I^2 x R = V^2 / R
i.e. power is proportional to the square of the voltage.
1) ‘power is proportional to the square of the voltage.’ only assuming the resistance stays constant with power. For heating elements, the resistance will increase as it gets hotter.
2) For resistive heaters, at a higher voltage, the appliance will simply get to temperature quicker and switch off (e.g. kettle)- essentially no difference in energy consumed (will be a tiny difference due to slightly different losses).
3) For most appliances with switch mode power supplies, as the voltage rises the current will reduce to keep the power constant. I’m told most SMPS’s are designed to work most efficiently at about 230V, and they convert less efficiently at higher voltages – which is where high voltages cause a very slight amount of extra energy to be consumed. But it is nothing like a square relationship, or even in 1:1 proportion to the voltage rise.
With the design constraints on modern design of solar inverters it is not true that they will cause over voltage. They are required to self limit their output to a maximum voltage in line with local network requirements.
In my case I have a system with micro-inverters and due to small variations in their sensing of voltage it means that they progressively shut down, one after the other as voltage increases. This means a soft, progressive reduction in total system output as voltage increases.
Why are the electricity network managers so reluctant to turn down the voltage? Is it because there are less losses at the slightly higher voltage?
They aren’t reluctant, they do it, but its time and labour intensive as the distribution transformers need their taps changed manually.
I found the distribution company very reluctant to come out .. and when they did come out they turned it down a minuscule amount so that the voltage is still bumping on the top end all the time. ..
I don’t think this would have much effect on the L/V network but Ohms law says we will pay more for any resistive load we are using when the voltage is higher. Only a small amount per consumer but I guess a significant amount of money over the network.
There’s a number of reasons. Cost is one, load at the end of a loop another and load at different times of the day/season a third. In most cases the industry wants to look the other way but it also refuses to update its out of date system aided and abetted by the current coal controlled government which dishonestly uses the renewable industry as a scapegoat to try and suppress an industry its donors do not want to thrive.
I got my 8.7 kW system to work correctly only after I pushed the installer to swap out the single phase inverter with a 3 phase inverter and got Ausgrid to lower its ridiculously high street voltage. Have not had a problem since.
The study takes measurements at the home switchboard, and also uses devices which don’t meet voltage compliance measurement requirements. See AS 60038:2012 and AS 61000.3.100. The high voltage findings therefore can’t be taken to confirm non-compliance. This is stated in the report.
“in most postcodes, solar has contributed around 2 V rise per kilowatt exported from PV”
This is false, only 34% of postcodes have a V/kW > 1, approximately 5% have a V/kW > 2, and many of these postcodes don’t have enough sites for this value to considered representative of the postcode as a whole
Simon – thanks for this correction.
Richard Chirgwin
They aren’t reluctant, they do it, but its time and labour intensive as the distribution transformers need their taps changed manually.
Over 20 years ago when we bought our house, well before solar systems were the norm, I was continually replacing lights, light dimmers, PC power supplies etc. I wasn’t sure why it was happening, but suspected power surges, and when I heard another person on our street had similar issues I contacted the power supplier who checked and said all was “within spec”. Ten years ago we demolished and built a new house, and installed solar. The inverter kept switching itself off – I found it was reporting well over 255V almost constantly, and frequently hitting 260V+ before shutting down to protect itself. Got the power company out again, and again they said it was within spec but this time I showed them my proof (a log showing the grid power over a number of weeks). They put a monitor on the street and agreed that there was an issue and tuned the power down at the pole on the end of the street. Seems that at set locations there are regulators that boost or drop the power as required – the closer you are to these, the higher your voltage can be and it drops the further away you are. We (and our neighbour also having issues) were only a few houses away from the nearest regulator….
Basically it makes sense that with a power grid that is supposed to be 240V, that the power would vary along the grid and some users would encounter under or over voltage – and for me at least the cause was not my solar, but rather having solar enabled me to both identify and have fixed the issue! Just a shame it took so long…..
My Fronius inverter turns off now & then due to “over voltage”. So far it has always happened about half an hour prior to any solar production.
Household voltage is normally around 255v.
Postcode 5033
Report your high street voltage to SAPN – it is outside the Australian Standard.
The standard practice in the USA is for the utility to run the LV network as close to the upper limit allowed by ANSI as possible. The reason is economics. Increased revenue from higher consumption ( ohms law ) many but not all domestic devices will draw more power at higher voltage. Additionally by turning down the voltage in the lower income areas of town the utility can show the regulators that they have lowered the bills of those customers versus the rest of the city.
I have just last week put in a new 6.6 kw solar system. I had to pay an additional $2000 as well as a new meter box to upgrade my cables from the power pole..10mm to 16mm cable. Apparently this will reduce the voltage output
The larger cable will reduce the voltage DROP.
So your solar system will be less likely to drop out due to over voltage. Depending on the distance between the pole and your meter box, 16mm^2 sounds reasonable. I’m surprised it was originally only 10mm^2.
There could be effects due to reactive power too.