Good solar monitoring has a long list of benefits. Now we can add ‘solving high voltage problems’ to that list.
The phrase “the power of data” is already a cliche, but only because it’s true. This week, we recount the story of how Solar Analytics’ R&D head Jonathon Dore used his company’s monitoring data to dramatically reduce curtailment of output from the solar power system at his Blue Mountains home.
Dore’s experience came to our attention when he Tweeted his comment on an ABC story about curtailment:
I had the same problem with voltage limiting my #solarenergy output, but my @SolarAnalytics monitoring gave me the evidence to get it fixed. #abc730 pic.twitter.com/1AWtJM63uL
— Jonathon Dore (@jonathondore) August 17, 2020
The ABC was discussing a UNSW research project for the Energy Security Board we covered in July, which found that the impact of solar PV on low voltage network voltages is far less than the networks’ operational decision to tolerate higher voltages.
In the ABC story, Andrew Dillon of Energy Networks Australia emphasised the need for more data – and that’s where Dore’s experience comes in.
Dore installed his solar power system in October 2018 and given his role at Solar Analytics, he made sure he had good solar monitoring from the start.
“I could see fairly quickly I was getting curtailed,” Dore said.
Being installed to post-2015 standards, his solar system supports curtailment rather than simple cut-off, which made the impact less obvious than if his exports dropped to zero. But his monitoring system presents a daily summary of what it would expect to export versus what was measured and he could also see the impact of his own consumption on exports.
“If I turned on the air conditioner, the power output from the PV went up, because the aircon took the voltage down”, Dore explained.
“And with Solar Analytics data, we could see the grid voltage rising to over 250V.”
In spite of his contacts within distributed network service providers (DNSPs), Dore decided to follow up the issue the same way any other Endeavour Energy customer would.
“I wanted to go through the customer experience, so I went on the on the website, hit ‘general inquiries’, and filed my query, saying I could see the voltage was high on all three phases.”
To his surprise, Endeavour got back to him the same day, and in the afternoon a technician arrived to tap down the transformer on the two phases with the highest voltage.
The technician explained that the provider was taking the voltage down not to support Dore’s solar system, but for the safety of customers’ appliances.
With the distribution network 5V lower than it had been, Dore said:
“the system is back to its full output, and the expected-to-measured export ratio on the Solar Analytics dashboard was back up to 100 per cent”.
With his R&D hat on, Dore believes this application for Solar Analytics data will become increasingly common. He said at the moment solar monitoring data is used mostly by installers, because they have the skills to use the data.
The export expected-versus-measured report lets the system owner or retailer take the data to their DNSP and get them to fix issues.
“Mostly, DNSPs re pretty good when they find out about overvoltage – the problem is, they don’t usually find out about it.”
He said the networks are caught in a Catch-22 situation: they have standards, and if someone comes to them with data about overvoltage, they’re obliged to fix it. However, they don’t have enough people to fix every distribution transformer all at once.
“Back in the old days, the DNSPs were the enemy – now, they have good people, but they’re trying to do a big job.”
The other problem is simply money: networks don’t believe customers would be willing to pay for ubiquitous collection of and access to LV network data.
“We have been working on an automated detection algorithm to detect this issue more cleanly, and let people know what’s going on,” Dore said.
What Solar Analytics can see is most systems are fine – it’s just a few that are taking a “massive chunk” out of homeowners’ production.
And the system that’s curtailed might just be yours.
About Richard Chirgwin
RSS - Posts
Whereabouts in Solar Analytics do you find that voltage graph? ie. The one at the very bottom of Jonathon’s tweet? (I can find the other one, for out system, but not the voltage. It looks very useful!)
Thank you 🙂
Hi Cupcakes,
I made that one myself using data downloaded from my site. To do this, you need fleet manager access (usually given to the solar analytics reseller who provided the system and monitoring to the customer). So if you’re a Solar Analytics reseller, you’re all good – go to ‘Live Data’ to eyeball it or go to ‘Data Download’ -> ‘Max Site Data’ and make all the funky graphs you like.
We don’t usually provide this access to the solar owner though because in it’s current form, it’s not very user friendly for the lay person. As the voltage issues get worse though, I’m sure we’ll find a way to present this in a way that’s more accessible to the system owner.
Hey Jonathon!
Thanks so much for your reply and for explaining all this. If you find a way to present this voltage info directly to solar owners in future, I’d be overjoyed :).
We love our Solar Analytics; this would make it even better.
Thanks so much!
Does monitoring data from SolarEdge provide a similar output to SolarAnyltics ?
Would i just call SAPN?
I have seen my voltages spike to 258 and sometimes 260.
Richard,
I have a system with 2 of SolarEdge SE5000 inverters on a rural split phase system. When installed, the Essential inspector wrongly restricted me to 8.4Kw total output (His limit was based on the Total current flowing through the Neutral wire, which is incorrect in a reasonably balanced system).
The reality of these limits is not as large as people may think. Most systems only generate near the limit at peak sunshine. Even on my optimised, Northerly & Westerly system, I only generate near the peak from about 10am to 2pm.
This might make a difference of a few Kw´s generated. The bigger problem is that when the inverter is set to Kw limit, IF the voltage gets to the upper limit, the inverter will fault & reset, which stops generation for up to 2 minutes.
On my system, the output is Voltage controlled. This means the output winds back as the voltage rises, so I am outputting the maximum the system will generate at the voltage limit.
This also helps my neighbors, who will suffer less voltage imposed dropouts because my system controls the voltage (I have historical approval, but my neighbors are limited to 3Kw export, There are 5 customers on our 25Kw transformer). We had the transformer voltage adjusted when this system was installed.
To my mind, the important thing is to try to design your system to maximise early morning & late afternoon output because that will make more difference to the daily total power generated (providing of course the panel wattage is the 33% allowed over the Inverter rating)
My system a couple of days ago, on a sunny day generated 65Kw (in winter!). It has generated over 80Kw in the summer.
So, my comment is yes, the monitoring will find over-voltage problems. But so will looking at the data from the inverter (obviously on an internet connected inverter) The cost of a Solar-Analytics is not always justified. (But I have a Solar-Analytics on a location that does not have internet access, so uses 4G connection: justified for that instance)
At home here, I have just installed an IoTaWatt energy monitor, feeding to an EmonCMS database to optimise my local power usage: charging my EV, Hot water, pool etc. Still in process but very promising. This was necessary because my inverters are difficult to read locally, unless you are happy with the limitations of the Inverter manufacturer websites. (Solar Edge & Victron).
The Iotawatt is an IOT wifi device. It can run single, split or 3 phase voltage references, & works with many common current transformers. (Up to 14 inputs)
The Iotawatt is an Open-Source design made in the US. $100US+ post + 10% + locally procured voltage transformers & Current transformers.
Like Doug, I am using the Open Energy Monitor system – in my case an EmonPi and two EmonTX (for three phase) – to monitor usage, generation and grid voltage.
It was very useful being able to supply a comprehensive set of data to my local provider (EvoEnergy) to demonstrate consistent and ongoing over-voltage issues. Two of the three phases entering my property were peaking well over the limit, and one was even averaging above the limit. It also showed that the peaks were twice daily – once at about 2am (from memory) and then again from about 10am on during the day.
EvoEnergy were quick to fit some monitoring kit of their own to confirm my readings, and then to make sub-tap adjustments to lower the feed voltage. It took them a couple of goes to find the sub-station that was causing the issue, but since then, I haven’t seen any issues.
The monitoring kit requires a bit of fiddling, but anyone interested in getting good data would find it fascinating. One thing I like is that I can collect data down to a 5 or 10 second interval, so get a much more granular view of what is happening, than the 15 minute view from the Enphase interface.
Call SAPN Chris, They will put a logger on the street, and in about a week if their records on the logger show over voltage they will attend and turn it all down,
I get this happening at least once a month to one of my customers.
It took me nearly 6 months to get ACTEW AGL in ACT to adjust the transformer. I was getting frequent over-voltage dropouts (not curtailment). Despite it being fairly obvious where the problem was, they installed a monitoring device during wet/cloudy days and failed to record any dropouts or significant over voltage. With the assistance of Solar Hub contacts I was finally able to convince them to act and tap down our local transformer and the system has performed perfectly ever since.
What suburb/area was that Brian?
Yup, same problem Brian … we’re just outside Canberra, and I experienced drop-outs due to high grid voltage which was evident just by looking at the smart meter (>253v indicated). Not much point in going to ACTEWAGL though, but with SolarHub’s help we got Essential Energy out to check and they agreed after a good few weeks of monitoring incoming grid voltage. Some further months elapsed before the techs came out and moved the local transformer down 2 taps (6v each) with the result the system works perfectly now … grid voltage now a consistent 238-242 volts and no trips at all.
My DNSP blew me off when it came to reporting over voltage. They flat out refused to do anything. There is no consistency over how this issue is handled around the NEM.
https://www.abc.net.au/news/2020-08-27/authorities-power-to-switch-off-south-australia-solar-panels/12602684
Once Tesla incorporates MMB into Powerwall3, we should probably all flee-the-grid.
This aberrant behaviour by power companies will quickly backfire on them once storage systems increase in reliability, longevity and warranties double.
Roll-on Battery Day!~
Hi Lessor
I’d guess that if you “flee-the-grid”, your excess energy can’t be used to reduce CO2 emissions.
Then there’s economics – while you save the connection fee, you lose FiT earnings – although I agree, this is a balancing act (our FiT here in the West is so small, you have to export a lot to cover the connection fee).
Our DNSP (Powercor) have been fantastic. Had a 9.45kW solar sytem with a fro jus 8.1kW inverter installed in May 2019. Only then did we become aware on voltage fluctuations from 190-258 VAC at the smart meter. Powercor adjusted the transformer several times, changes phases, checked connections, replaced the service wire, and eventually replaced the 1920’s wiring in the street with ABC in March this year. So far so good, no voltages below 224, or above 253 at the Samer meter. Let’s see what December brings….
I’m experiencing high voltage issues too, and used monitoring to provide evidence to my installer who in turn notified the electricity distributor who are currently monitoring the grid.
I use graphing software called Grafana to graph data that comes from the inverters Modbus TCP interface. If anyone’s interested here’s my setup which should work with SolarEdge inverters, and maybe SunSpec compatible inverters – https://github.com/jsloane/solar-monitoring-stack
I had this problem also – could see it clearly on the Solar Analytics Live Monitor. Reported to the DNSP (Energex in SE QLD). They put on a data logger and eventually fixed the problem (we have single phase so I guess they moved us to a different phase).
Change “ration” to “ratio”? (expected-to-measured export ratio)
Yeah… I imagine “ratio” is what was meant instead of “ration”. Well caught. I’ll go out on a limb and correct the typo without checking with Richard first. But if I’m wrong, Finn will probably put me on half rations. As if I can somehow live on only 12 kilograms of chocolate a day…
Hi all
Synergy here in WA have apparently announced that due to grid-stability issues arising from high DER implementation, that they need to take steps to discourage excessive rooftop PV generation.
Apparently?, because the ABC report seemed to be confused – they spoke of reducing the subsidy (which I had thought meant the STC rebate). But no, they then went on to talk about the FiT – which to me could be considered a simple financial transaction (we sell our excess to them at 7c/kWh, they on-sell it for considerably more – 26c/kWh + gst) – I would not have considered this a form of “subsidy”.
Apparently, our FiT is to drop to 3c/kWh, but will be raised to 10c/kWh during “peak” times. I was told the peak time is from 3 pm to 9 pm – however without batteries, I kind of doubt many owners of PV are going to be exporting much after 5-6pm, even in summer.
Let’s wait to hear from the shonky installers capitalising on this announcement (get a battery – benefit from the higher FiT….).
I, to, have experienced high voltage curtailment – but after several attempts our DNSP finally dropped the transformer taps (at the sub-station, not at our local grid transformers – apparently this was too hard, or maybe they had no more taps to go as we started out with a 250V main).
I’m wondering WHY some still think it has nothing to do with solar PV? You’d think that without DER, the increasing load on the grid would simply LOWER voltage, not raise it. Perhaps the transformer gods have been increasing the tap voltages to keep the people at the end-of-the-line from hitting the the lower end of the regulated minimum voltage limit?
If so, how come the DNSP can still now re-LOWER the transformer taps when requested to deal with an over-voltage situation (if they can be bothered)?
I can’t help but think that adding PV could only make things worse.
The solution to both problems, of course, is to “strengthen” the grid – that is, to increase wire sizes, double up on feeder cables, transmission lines, etc. A difficult thing, I would have thought, to avoid cutting off power from time to time to allow the work to be done.
It is not as if Europe hasn’t established a precedent – maybe we should LEARN from Germany’s experience?
Yes Ian – the reason power companies want to keep the voltage high is because our Australian distribution system started at 250 V, then went to 240 V and now has a nominal voltage of 230 V. But a lot of the infrastructure was designed for the higher voltage. And the higher voltage they run the less current they have in their cables (for a given power delivered). So losses are lower and the need for more copper is averted. It is therefore not surprising that some power distribution companies are reluctant to drop their supply voltage. Annoying, but not surprising.
In some cases the problem of inverters dropping out due to high voltage is made worse by having too much voltage drop between the domestic switchboard and inverter. If the maximum allowable voltage drop is 2% that is 5 V @ 253 V. ie. the inverter will need to turn off at 253 V when the mains voltage at the switchboard is 247 V. Oversizing the cable from switchboard to inverter can therefore reduce the likelihood of overvoltage problems and will have the additional benefit of reducing energy losses from solar generation.
And yes – I did mean to say 248 V at the switchboard!
Things electrical are complex for sure and that’s why we need experts to clarify issues – thanks. The over-voltage issue seems to be concerning. In principle, it must be possible to have what can be called a ‘voltage controller’ at the ‘boundary’ of each ‘electrical consumer/generator’ (property that consumes or generates electricity) and that accepts or provides only a ‘safe’ voltage. Would that not solve the problem? Of course such a thingy may have many disadvantages, and they could well be prohibitive in a practical or realistic sense. However that should not deter us from exploring alternatives to the current practice of tinkering and futzing at the edges. What a shame it is that now that we have a viable clean energy generation system being deployed as a distributed resource we seem to need to curtail its output. That to me sounds like a system that is far too complex or in quintessential Aussie terminology is stuffed!
We have such a “thingy” Heinz – it’s called a transformer! We also have on-load automatic tap-changing transformers, if this becomes necessary.
But Heinz, this really would be “futzing at the edges” – as most houses would need one, and they are EXPENSIVE.
The “general” solution, is to upgrade all the wiring in our system, to heavier diameters/greater area of cross-section. And, upgrade the capacity of existing transformers.
An “externality” cost of more DER, but solvable.