Skip to main content
Observer Ethical awards Winners 2011

YouGen Blog

Voltage spikes cause drops in solar PV performance

Posted by Rob Palgrave on 12 November 2012 at 10:36 am

A members of my local Friends of the Earth group installed solar PV on her roof at the  start of the year. This is in Surrey. The generated energy is much lower than  expected, and the installer has not been able to resolve the issue so far.  Suspicion falls on the compatibility of the inverter to match the (apparently) very spiky local grid connection. I'd be interested to hear from anyone who has had a similar experience, or can shed light on the issue.

The installation comprises: 18 Schott Perform Mono 180—190 wp (9  front roof and 9 back roof), 3.42kWp net capacity; east/west facing roof 30 deg pitch; Sunny Boy inverter

After the feed-in tariff claim at the end of first 6 months (30 June 2012) British Gas commented - the  reading (680kWh) was only 48% of what would have been expected over this  timeframe.

The installer was asked to attend – they checked everything out on site, said they could see nothing wrong but would contact  manufacturers etc. Which lead nowhere.

About 2 weeks later an alert appeared on Bluetooth display re the inverter saying “voltage spikes”.  The daily graphs had started going flat between 11 and 4 approx. (About 22  July by this time and hot weather).  

Installer came back in – explained grid normal voltage setting  230v (standard G38/1) but made a call to Sunny Boy  and was instructed to change setting to 250v – apparently legal supply should  be about 230 but with parameters of operation 250v should be adequate – and  the owners were told to contact electricity distribution company SSE to have a voltage monitor installed. There have since been flat periods in graphs (but no error messages).

A voltage monitor was duly installed for a week and SSE has written to the solar PV owner  confirming that there are voltage spikes above and beyond their legal parameters. Apparently this will require serious work in the area with electricity switch off throughout. That was in  August and nothing has happened by end of  October.

Have voltage  spikes have been affecting the system all along (including first 6 months)?  SSE were non-committal.

The explanation given by SSE is that they get spikes in electricity supply in hot weather. If no spikes in first 6 months what can  be done to establish the problem? Installer seems unable to envisage this  possibility although there were no evident outages in this period per graphs. Is the problem truly solved – the installer says yes.

About the author: Rob Palgrave is a guest blogger

If you have a question about anything in the above blog, please ask it in the comments section below.

Like this blog? Keep up to date with our free monthly newsletter

Comments

15 comments - read them below or add one

James Lund

James LundComment left on: 12 August 2016 at 10:07 am

this is a bit late i suppose but maybe of interest to someone, I just installed an Eco Box grid voltage monitor, result! no more voltage drop outs! couldnt be happier! invertors are designed to cut out if the incoiming voltage gets too high. i cant recommend this product enough, I even had my inverter replaced trying to solve this problem. found mine re a seller in Southport www.thevascogroup.co.uk. 

report abuse

playford1651

playford1651Comment left on: 22 December 2012 at 12:32 am

In reply to @msundog: the manufacturer says it uses HF transformer technology. As to whether it needs to  be connected to an rcd I have no idea. The reputable company that installed it, & subsequently checked the electrics of the PV system, must have thought so!!

report abuse

Sims Solar Ltd

Sims Solar LtdComment left on: 10 December 2012 at 11:50 pm

Anne not sure if your problem is solved but thought i'd share the findings of a call i had with SMA today. The type 20 TL inverter you have is set with a ref voltage of 230v which with +10% gives the 253v referred to in other comments. The type 21 TL (current model) has a reference voltage of 240 so greater tolerance. SMA recommend 6mm AC cable because of cable resistance requiring the inverter to output at higher voltage to overcome resistance. They suggested a quick test, turn off inverter at the inverter located AC isolator. Turn back on and note grid voltage immediately displayed, when inverter starts to generate make a note of this AC voltage, if the difference is greater than 5v this could be contributory. Ideally you want this difference to be as low as possible. Might be worth noting whether the screen is scrolling G83 or G59 as the latter will be potentially more sensitive to elevated supply.

In summary i suggesting some tests to clarify whether an elevated supply voltage is being exacerbated locally.

report abuse

Anne Mathieson

Anne MathiesonComment left on: 5 December 2012 at 4:24 pm

As owner of the solar panels referred to in Robert Palgrave blog - the comment on 19 November asked for details of our inverter: it is Sunny Boy SB 4000TL-20 with nominal and max  AC power of 4000W.

The installer has given us a G83/1 certificate from SMA Solar Technology AG (the manufacturer) giving results of G83/1 tests and a declaration that all such devices shipped to UK com ply with G83/1 - these settings can only be changed by use of a password protected tool apparently and I understand that my installer changed this setting back in July when the 'Voltage Spikes' error message came up.

We have now had a telephone message from SSE (following confirmation from them of voltage spikes back in August) that instead of tapping down at the substation (involving switch off of local electricity supply) they were going to 'back feed the supply' and 'spread it across the network' thus avoiding switching off the electric supply locally. This is supposed to have happened in last week of November. 

Having followed the blog, there is no mention of this being an alternative way to deal with voltage spikes - if feels as if we are being fobbed off by SSE - and without the electric supply switch off there is no evidence that anything has been done. Has anybody heard of this? Was this a valid way for them to deal with the problem? 

 

report abuse

Rudge Energy

Rudge EnergyComment left on: 2 December 2012 at 6:45 pm

There have been a large amount of cases such as this over the summer, especially this year in (the sparse) sunny periods.

The grid voltage is nominally quoted as being 230v, but can be between 220v and 245v, sometimes more. The grid voltage will vary during the day at any one location with the amount of load on your local substation.

Loading the local substation will make your local grid voltage slump, depending on the amount of load. Now with people having solar PV systems installed being commonplace, a sunny day will cause the grid voltage to rise, causing the opposite of loading the grid.

As indicated on other posts, most inverters come programmed to operate at 230v, but will have a tolerance factor to rise to a higher voltage before switching off due to overvoltage. We find that Sunny Boy inverters now come set to 253v max, but can be programmed to be a little higher in extreme situations. 

If your house is in an area where there is commercial activity (offices or factories), you will find a far lower grid voltage during the week, which can rise considerably over the weekend when the offices, etc are not operating. If in addition to this, there are a number of PV systems installed and the weather is sunny, some of these PV systems may start to drop out due to high voltage.

Though this is a common story in rural areas where the grid fed by overhead lines, town and city centres should less susceptable.

We've been out to quite a number of systems (not just ours) this year to carry out adjustments on PV systems that until Spring 2012 have been working fine. The big expansion of PV installations has had this knock on effect, they push the grid voltage up.

report abuse

Cathy Debenham

Cathy DebenhamComment left on: 30 November 2012 at 11:34 am

From @msundog on Twitter:

@YouGenUK .. Solving the RCD tripping problem .. does the system have a transformer-less inverter & does the circuit actually need an RCD?

report abuse

playford1651

playford1651Comment left on: 29 November 2012 at 1:11 pm

A problem that I have encountered is that since our PV system was installed in October 2011 we have had a problem with our RCD tripping. It does not happen all the time; we can go several weeks without it tripping. We have had the installer out to check the the system but can find no fault. We have had our electricians check our wiring (which was rewired in 2008) but they found no problem. The following week the RCD tripped 3 times in a week!!!. The electricians then switched the PV system to the non-RCD side of the consumer unit. This was in the 2nd week in August & since then we have had no problems.

The interesting thing to note is that the trips seem to happen between 10 & 11pm. Thinking back to other occasions the tripping seem to happen when the inverter went into sleep mode.

The system is a 2.9 kW consisting of 12 x Sanyo 240W panels with a Fronius IG Plus 30 inverter.

report abuse

RobertPalgrave

RobertPalgraveComment left on: 19 November 2012 at 3:16 pm

@ H2ecO Ltd and @SunGift Solar

- many thanks, the owner of the panels is going to pursue these points shortly and I'll report back

 

Rob

 

report abuse

SunGift Energy

SunGift EnergyComment left on: 19 November 2012 at 2:18 pm

I think another important question to ask in this case, which I must admit I over looked initially, is what model Sunny Boy has been installed? Unless it is a 4000TL or 3600TL it will only have one MPPT ( maximum power point tracker ) which will not be able to handle having two arrays on different roof aspects connected to it. If this is the case it would be a case of poor design or product knowledge by the installer. 

report abuse

H2ecO Ltd

H2ecO LtdComment left on: 13 November 2012 at 5:14 pm

SMA inverters have a set upper voltage limit - part of the G83 regulations above which they are not allowed to feed power into the grid. The normal solution is to get the National Grid operator to tap the transformer down to lower the voltage to your property. Your installer should have a good relationship with your DNO and should be able to help you with this.

I have seen another instance recently where a customer of another company was only producing about 40% of the predicted output. We were asked to go and have a look and when we got there, we found that his SMA 3000HF had not been set correctly to G83 settings. The inverter was an import fitted in the rush at the end of last year and the installers didn't know that it needed to be reset during commissioning. German Grid settings are much tighter than ours so his inverter was regularly cutting out in the middle of the day when it was supposed to be generating the maximum. This could easily be misaken for a grid voltage problem when it was in fact an inverter problem. Have a look at the readout on the front of your sunny boy inverter. If you see G83/1 flashing up as one of the bits of information, this isn't the problem but if there's no G83/1 mentioned after a restart, your inverter could be set to non-UK settings. This is easily rectified by an installer with a grid guard code and a laptop - but some installers don't know this.

 

report abuse

RobertPalgrave

RobertPalgraveComment left on: 13 November 2012 at 4:12 pm

@"MG Renewables LtdComment left on: 13 November 2012 at 12:14 pm

Try fitting a 100 milliamp breaker, this may prevent the inverter tripping out...just a thought. We do this as standard due to this issue in rural areas."

many thanks - could you provide a bit more information please? Cannot immediately see how a 100mA breaker helps ( the breaker rating will be something like 16A won't it?).

@ SunGift, Sims Solar and Alternative

Many thanks also, we are following up on these points

report abuse

MG Renewables Ltd

MG Renewables LtdComment left on: 13 November 2012 at 12:14 pm

Try fitting a 100 milliamp breaker, this may prevent the inverter tripping out...just a thought. We do this as standard due to this issue in rural areas.

report abuse

SunGift Energy

SunGift EnergyComment left on: 12 November 2012 at 4:07 pm

This is quite a common problem with the increasing number PV installations being installed. As part of our surveying process we record the onsite voltage so we can predict if we will run into any problems with the installation.

Luckily we have a great relationship with the inverter manufacturers and have the service software to enable us to adjust certain parameters to ensure the inverter will not shut down from slightly increased voltages. Our Engineers can make these adjustments via their laptops ,if necessary, although sometimes the voltage may be to high to work with.

In these cases we have supported our customers and helped them get the power company to adjust the voltage on the local transformer. I suggest you speak to your installer and ask for their advice/support to get this resolved.

report abuse

Sims Solar Ltd

Sims Solar LtdComment left on: 12 November 2012 at 1:11 pm

I attended a seminar hosted by SSE earlier in the year and this was an issue discussed. With hot weather driving demand and giving rise to higher grid voltages tripping is going to be nuisanc issue. The problem is also made worse by non notified installations, SSE reconned 50% of installs were not being notified to them.

Its not specified what inverter model is being used but if a sunny beam is being used that data logs. Case for discussion with SSE to see if Grid Guard settings can be changed.

We are located close to a main transformer and our voltage is typically close to 250V, fortunately we have had no tripping.

report abuse

Alternative

AlternativeComment left on: 12 November 2012 at 11:58 am

I had a similar problem, but picked it up quite quickly as I regularly monitor PV output. My voltage was well outside spec at certain times of the day, so I kept hassling the electricity company after they fitted the monitor. They reluctantly agreed to change the tapping on the local transformer, which dropped our voltage by 10 volts overall. This was enough to prevent the inverter from going offline during a surge.

report abuse

Leave a comment

You must log in to make a comment. If you haven't already registered, please sign up as a company or an individual, then come back and have your say.