Larger 6kWp+ domestic install before the FIT cut

[wtc]

I'm not an electrician, just a householder (though do work in an area given a kick in the family jewels by government decision last year). I am technically minded so want to understand that proposed systems are worth the spend. I have a decent south facing roof space that can take around 25-27 standard sized panels, with some horizon level shading from trees and a bit more early evening due to roof shape.

Since I have the roof space and can go for 6kWp+, I was hoping for a potentially better return due to the economies of scale (e.g. one inverter, one lot of scaffolding, etc), but the more competitive 4kWp arena means this is probably not the case. Furthermore, it is unfortunate that now I'm in a position to actually go ahead with solar PV, there seems to be a mad scramble to get an installation by the end of the year and/or secure the components needed.

I have had a few quotes so far but not entirely happy with varying aspects of them and have listed some thoughts below.

1. G59 response time
Given the time to the potential FIT cut, I have explicitly suggested a G83/2 install with capped inverter, but installers seem be wary of this and want to go with G59/3. My DNO is SSE, but what is unusual is that it is a former hospital site in Epsom, Surrey. The remaining parts of the hospital and all the surrounding area use UK Power Networks, but the new build houses and flats use SSE. The electricity infrastructure is therefore all new. Some of the early houses had a token 1kWp added by the developer, a few have retrofitted 4kWp, but most homes are apartments in the conservation area (I’m not) without solar PV.

I have enquired with SSE and they can only guarantee the OFGEM 45 working day turnaround, which now just takes me into next year. They advise G83/2 with 3.68kW limit is safer, but admitted they are working extra hours to process G59 applications. Am I right to think there is far too much risk in a G59/3 application, where risk means missing the current FITs or installer having put up poorer value cut down system as a G83/2 fallback.

Installers seem very confident of G59 success given the development, but can they really be sure?

Would it be worth the extra cost of using a G59 agency?

Does anyone have first hand experience of current SSE G59 responses?

2. SolarEdge and string length
One installer has suggested a four-string SolarEdge, with either 6 or 7 250w panels on each string. I think this should be two strings, to exceed the minimum of 8 on each string, but also to effectively deal with the shade and maximise the time with at least six panels alive on each string. He thinks I might be right, but what do those on the forum think?

3. SolarEdge limitation under G83/2.
Am I right in assuming that the SE4000-16A version that is an explicitly G83/2 approved model is a 100% safe bet to use under G83/2 (surely that's the point)? Is that even if fed with 5kWp of panels?

What about an SE5000 or SE6000, but with the generation capped at 3.68kW in the installer settings? These are both G83/2 type tested along with the smaller inverters, but do not directly meet G83/2 due to generating more than 16A. If installer capped, should a DNO accept this?

The much better option would be an export cap at 3.68kW using the SolarEdge meter, but on reading further, I’m pretty sure this would be G59 only.

My gut feel is that if a generation capped SE6000 is acceptable under G83/2, then to have it this year, followed by a G59 to apply for uncapped or export capped later (if worth the bother).


Thoughts from the experts on this forum are welcome.

Thanks,
James

 

[egg67]

Looks like you've looked into it, sorry do not know any thing about it but some one will be along at some point

 

[TedM]

If generation is capped at 3.68 kW and the inverter is certified G83 by the manufacturer then the installer should have no problem signing and issuing G83 paperwork to the DNO to say the system they have installed does indeed comply with G83 - under the normal post-notification basis.

You should be asking potential installers to quote and model returns based on 4/5/6 kWp of panels - all under G83.

It is also worth noting that one of the current proposals under the recent DECC FiTs consultation is to deny FiTs to extensions. So if this change is actually implemented and you add panels next year then the extension would not be eligible for FiT payments at all.

 

[whinmoor]

Quick answers:

1) Can't comment on the DNO in your area.
2) SE can take up to 25 panels on a single string. Not sure why you'd want to have two or more.
3) The SE4000-16A has been developed by SE just for the UK market to comply with G83 requirements. The most we've fed into it is 5.56 kWp (21 x 265w panels). That was just with a single string and it's working fine.

You need to consider how severe the shading actually is. Is the roof easily accessible if an optimiser fails or are you looking at a £400 scaffold? Would a regular, decent quality string inverter from SMA or ABB (Power One) be more appropriate?

 

[wtc]

Thanks for the replies so far.

The unusual thing about the SE4000-16A is that in the G83/2 type test certificate (and data sheet), it is shown as AC rated at 4000w, so not 3.68kW, potentially giving a gap. However, the 'G83 certificate' and device itself will show 3.68KW as the rating though.

I've had a look at the type test documents and the G83/2 and G59/3 are virtually the same, the only difference being that the harmonics are tested under BS EN 61000-3-2 for G83 and BS EN 61000-3-12 for G59.

Good point about the proposed zero FIT on system extensions (although it will be almost zero on new installs compared to where it is now). I gather this means if I had 4kWp on the roof this year, then 2kWp added next year, the practical implementation is that I would only get FIT payments on 2/3 of the metered generation after the extension. So I really should max out what I can get on the roof now.

 

[wtc]

3) The SE4000-16A has been developed by SE just for the UK market to comply with G83 requirements. The most we've fed into it is 5.56 kWp (21 x 265w panels). That was just with a single string and it's working fine.

Another oddity is the latest SolarEdge datasheet for the SE4000-16A (dated Sep2014) states the max DC input is 5000w, a slightly older one (Feb 13) has 4400w while a datasheet for the SE4000-GBR (dated Jul 11) has 4100w. the max has been getting bumped up over the years.

 

[BruceB]

Oversizing note:
http://www.solaredge.com/files/pdfs/inverter_dc_oversizing_guide.pdf
looks like there is some flexibility to operate outside recommendations but might blow warranty.

 

[whinmoor]

Our job was on four roof surfaces with steep pitches, different orientations and some shading. I think it was 8 panels facing east, 8 west, 3 south and 2 on a very steep pediment. We could have added two more panels on another pediment but I thought that was over-egging it.

According to the monitoring, it rarely clips.

We've also done a 5.3 kWp E/W split and that similarly hardly ever hits 3.68 kW. The inverter is probably working in its most efficient zone more often due to the over-sized array and the system easily beats its monthly targets every time.

I'd say that a 5.3 kWp E/W split array maybe acts like a 4 kWp south-facing array in terms of maximum production but the curve is flatter without the midday peak.

These jobs are in Yorkshire. You might want to be more cautious in Surrey.

 

[The Solar King]

Watch out for SE Warranty invalidation. You need to run configuration through their tool. If it doesn't show compatibility you need to specifically ask their tech team.

We ended up getting a G59 for a 5.44kw system using SE with Solar Frontier thin film panels. Configuration was confirmed as invalidating the warranty on a SE4000. Had to use a 5000 instead. Due to E/W and only a small amount South, think times this system will exceed 16A and clip are minimal.

 

[wtc]

Oversizing note:
http://www.solaredge.com/files/pdfs/inverter_dc_oversizing_guide.pdf
looks like there is some flexibility to operate outside recommendations but might blow warranty.

Watch out for SE Warranty invalidation. You need to run configuration through their tool. If it doesn't show compatibility you need to specifically ask their tech team.

BruceB - I'm guessing you're the same person responding on the other forum I where posted similar question (and I'm getting really helpful responses from both)

The oversizing note states that the maximum oversizing ratio allowed is 135%. However, it also states the AC output for this calculation is the lower of the rated/nominal power of the inverter or any software limit set.

If the inverter is an SE4000-16A, or even SE4000/SE5000/SE6000 set to max at 3.68kW or 16A, then the max DC input at 135% would be 4.968kWp so just a tiny margin below 5kWp. 5kWp itself is nearly 136% so, following to the letter, might invalidate an SE warranty. The SolarEdge software wouldn't allow an SE4000-16A unless the ratio is at least 136%. But then, the latest datasheet for this model states 5000w max DC.

It might appear to rule out the ~6kWp with an SE5000 or SE6000 software capped to 3.68kW, but in my (increasingly looking theoretical) case that would be temporary. Using StorEdge and/or feed in limitation at some point in the future would mean lifting that cap.

 

[BruceB]

Rushng out now, but at first sight i do not see that limiting output of a SE 4000/5000/6000 would affect the maximum on the input. The right person on the phone or email at SE could clarify.

Yes it is me over there.

 

[wtc]

Following up my original post, I did get a system installed in December on the old FITs, but not in excess of 6kWp though! It comprises 20x 250w panels to give 5kWp fed into a SolarEdge SE4000-16A inverter, so done under G83/2. This is within the warranty limits of the SE4000-16A.

It was a different installer to those who had given quotes around the time of my original post and a firm who inspired more confidence at the quote stage but I have also been happy with the end result.

There just didn’t seem enough time to get a G59 application in, and looking back now, the chance of it getting approved installed by mid-January would have been tight. As it is, under G83/2, it was installed before the rate cut.

The 20 panels have been installed on a single string. No need for four!

During December and January there was significant shading from trees. It was only into mid-February that the whole array has been getting into full sun. However, Tuesday it performed very well. It was only just clipping around midday, at just a fraction over 4kW according to the SolarEdge monitoring.

A few other observations from the quote process:
- No-one mentioned voltage optimisation, which I’m happy about.

- There was a fair bit of misleading in terms of self-consumption around the quote stage. My chosen installer was the most honest saying it may only be a third. I did have quotes allowing for 75% and 90% self-consumption. While these two did include iBoosts making a higher proportion possible, this was put into the financial projection as imported electricity saving to over-inflate the payback. A bit of a wheeze to make costly systems seem better value. Including the 7.7kWp system forecast to produce 7171kWh per year with self-consumption at 5487kWh, even though our household only uses around 4000kWh per year.

- There were different attitudes to the tree shading, ranging from stating there is no issue to carrying out detailed measurements and putting them into PVSOL (though I would only get the full results if I went ahead). But it made very little difference to the proposed panel layout between different quotes.

- My installer was happy to either connect to a spare way in the existing CU or put in a 2-way CU and I asked for the latter. Others seemed to be reluctant to offer the small CU when I had a spare way.

Going back to my original brief, I was hoping to benefit from economies of scale with the bigger system, since there was still only one inverter, one design, one set of scaffold etc. What I seemed to find was that there were competitive prices around the 4kWp mark, but going above this would cut the firms willing to quote and bring in G59 costs, so not necessarily giving a better return.

Thanks to those who posted.

 

[Worcester]

@wtc Thanks for posting back, it's always good to hear the final result.