Hybrid inverter for small in roof 2kw array - help please!

[Timmyk]

Hi! 

 

I’m new to the forum and PV and I’m just looking for some advice for a new in-roof system.

 

We’re currently having a dormer loft conversion and that has mandated a new roof at the front of the house (35 pitch) and over the bathroom (15 pitch) at the rear. They face west and east respectively. 

 

Instead of replacing the slates, we’ve decided to have in-roof solar panels put in. GSE mounts, black JA Solar 370w panels. There’s space for 3 panels on the west facing front and 2 on the east facing back. The total cost of the mounts/panels was £1500 - which isn’t much more than the new slates were going to be. 

 

Whilst I fully appreciate that it is a small systems with a less than ideal arrangement, given the small difference in cost to re-slating the roof we thought we’d give it a go. There is scope to later - with planning permission - put south facing panels on the flat roof of the dormer.

 

Ideally I’d like to use the system to charge batteries via a hybrid inverter and then use any power in the evening/overnight as the house is mostly empty during the day. Current draw even in the evening is low. We use about 3000kWh/year total.

 

I’m not experienced enough in this to know which inverter/batteries to go for. The Solis 3kW hybrid looks reasonable but 3kW is quite a lot higher capacity than the PV will be able to produce - my understanding is this will significantly reduc efficiency.

 

Can anyone suggest an inverter / batteries for this array? Or any other opinions/advice on this setup?

 

Thanks!

Tim

[Timmyk]

Photo above is the front roof, below is the small rear bathroom roof

 

Thanks!

[Conor]

I don't think you'll generate enough to effectively charge the batteries, except for peak days in summer. Once you bear in mind that you won't generate the full 2kw due to the orientation, and that your day to day background load will each up another couple hundred Watts at any one time, then you're left with very little.

 

Better option is to change your electric use habits (washing machine and dishwasher on time delays, solar diverter to heat hot water etc) to make the most of the daytime generation. And if you end up with an electric car in the future, then that will eat up the lot.

 

Don't forget that you'll need oprimisers on the panels to maximise generation. Can reccomend the solaredge optimisers and inverter setup.

 

That first roof looks big enough for 4 panels in landscape 2+2 configuration. Have you considered that? Remember, you can run the GSE trays right to the gutter and half a slate from the verge.

Edited April 8, 2022 by Conor

[Timmyk]

20 minutes ago, Conor said:

I don't think you'll generate enough to effectively charge the batteries, except for peak days in summer. Once you bear in mind that you won't generate the full 2kw due to the orientation, and that your day to day background load will each up another couple hundred Watts at any one time, then you're left with very little.

 

Better option is to change your electric use habits (washing machine and dishwasher on time delays, solar diverter to heat hot water etc) to make the most of the daytime generation. And if you end up with an electric car in the future, then that will eat up the lot.

 

Don't forget that you'll need oprimisers on the panels to maximise generation. Can reccomend the solaredge optimisers and inverter setup.

 

That first roof looks big enough for 4 panels in landscape 2+2 configuration. Have you considered that? Remember, you can run the GSE trays right to the gutter and half a slate from the verge.

Thanks for the fast reply.

 

About size - 4 portrait panels side to side including the side flashing would take up 4.7m. The entire width of the roof is 4.9m. So just about too small I thought but maybe I’m wrong on that. If I’m quick I can probably add to the order!

 

Landscape I had a really hard time finding a compromise between panel size and GSE trays that would work for them; I did try to spend a while looking. One ray of hope is putting extra panels on the flat roof further down the line.

 

That’s sad but fair enough about the battery. There’s nothing that could work?

 

Optimisers I can sort. One per panel, right?

 

Thanks again.

Tim

Edited April 8, 2022 by Timmyk

[Timmyk]

Got to be this one by the looks of it…

 

https://midsummerwholesale.co.uk/buy/SolarEdge/solaredge-compact-1500m-ext

 

[Dillsue]

If youre looking to expand the system onto your flat roof you probably want to get a bigger inverter or check the inverter linked to can handle the power input you may end up with. My personal preference would be an optimiser per panel so that a single optimser failing doesnt take out to much generation.

[Timmyk]

16 minutes ago, Dillsue said:

If youre looking to expand the system onto your flat roof you probably want to get a bigger inverter or check the inverter linked to can handle the power input you may end up with. My personal preference would be an optimiser per panel so that a single optimser failing doesnt take out to much generation.

 

Okay, thanks.

 

I had a chat with a supplier and it seems the panels I have bought produce too much current for the SolarEdge Compact anyway (apparently the panels make 11amps, the inverter can take 10.5.) I'm not 100% convinced on this though.

 

An optimiser per panel is doable for sure. Probably the Tigo ones I guess so there is flexibility with regards to inverter choice?

 

That still leaves the question of which inverter - particularly what size to go for given the array size and direction.

 

Any thoughts?

 

[markocosic]

Good thinking - it's great seeing "will do what I can do" 🙂

 

 

You need to consider:

 

- Max current

- Max open circuit voltage of the string (not going to be a problem for a small string)

- Minimum startup voltage and MPPT tracking range for the string (can be tricky with small strings)

- Shading (with conventional inverters you'll want one MPPT input for east and one for west)

- Maximum DC power / AC power (it's usually ok to attach more panels to the inverter than it is able to convert to DC; output will only be clipped occasionally)

 

Don't go too big. The efficiency of inverters is poor at low load.

 

If the shading isn't horrible I'd stick in a pair of these:in the appropriate (overall power) rating. One east. One west. Does 15A. Has WiFi. Stick them somewhere that you can access them to replace them later when the fail.

https://midsummerwholesale.co.uk/buy/sofar-inverters/sofar-1100-tl-g3

 

Else keep it stupid simple. Panels that work with any old inverter. Put the inverter somewhere easily accessible. They're the only bits that really fail.

 

 

No solaredge stuff. They company appear to be grade one a-holes and it's pushed hard by sales because once it's on the roof you're stuck with it due to the cost of getting up there to rip it out.

 

https://www.mcelectrical.com.au/solaredge-inverter-optimiser-review/

 

 

Batteries probably don't pay for themselves over their lifetime. They don't last forever. Do the math on the how much each charge/discharge effectively costs you and you'll probably decide that the batteries aren't worth it. PV on the other hand 30+ years. Inverters 15+ years.

 

 

I think you push panels down the front roof to avoid the chimney shading.

 

Push the panels on the extension as far away from the house as you can to minimise shading. The panels are made in strings. If you shade the bottom of a portrait panel then you shade all three strings and you knock out the panel. If you shade the bottom of a landscape panel then 2 of the three strings still work, the bypass diode will bypass the third, and the panel will produce. Depending on your shading landscape may (often will) perform better. On-roof fit installers hate landscape because it needs more rails and you're faffing more finding rafters etc. With GSE trays it matters not one jot.

 

https://www.sciencedirect.com/science/article/pii/S0038092X21003868

 

 

 

If you do need optimisers go with Tigo. They'll work with any inverter.

Edited April 8, 2022 by markocosic

[JohnMo]

I have tigo optimiser and Growatt inverter.  But mines 10 panels in a row as a single string. All seems to work ok.

 

With panels on different elevations you would ideally do two strings.

 

But I think I would just go down the route of micro inverter, no optimisers required keeps it all simple.

[markocosic]

Depends on shading - two inverters is simple (and they're cheap / not mounted on the effing roof under the effing panels, for when they fail before the panels do) - but optimisers will make better use of the PV in shading conditions

[Sandybay]

Hi, 

I have a 9 panel, 3.68kw array on a ground mount frame, panels are in a row, just about to run wire for grid connection, trying to do what I can before calling a professional for wiring it up, is it just a single cable from the panels to grid connection or do I run two separate lengths of cable? 

[Timmyk]

16 hours ago, markocosic said:

Good thinking - it's great seeing "will do what I can do" 🙂

 

 

You need to consider:

 

- Max current

- Max open circuit voltage of the string (not going to be a problem for a small string)

- Minimum startup voltage and MPPT tracking range for the string (can be tricky with small strings)

- Shading (with conventional inverters you'll want one MPPT input for east and one for west)

- Maximum DC power / AC power (it's usually ok to attach more panels to the inverter than it is able to convert to DC; output will only be clipped occasionally)

 

Don't go too big. The efficiency of inverters is poor at low load.

 

If the shading isn't horrible I'd stick in a pair of these:in the appropriate (overall power) rating. One east. One west. Does 15A. Has WiFi. Stick them somewhere that you can access them to replace them later when the fail.

https://midsummerwholesale.co.uk/buy/sofar-inverters/sofar-1100-tl-g3

 

Else keep it stupid simple. Panels that work with any old inverter. Put the inverter somewhere easily accessible. They're the only bits that really fail.

 

 

No solaredge stuff. They company appear to be grade one a-holes and it's pushed hard by sales because once it's on the roof you're stuck with it due to the cost of getting up there to rip it out.

 

https://www.mcelectrical.com.au/solaredge-inverter-optimiser-review/

 

 

Batteries probably don't pay for themselves over their lifetime. They don't last forever. Do the math on the how much each charge/discharge effectively costs you and you'll probably decide that the batteries aren't worth it. PV on the other hand 30+ years. Inverters 15+ years.

 

 

I think you push panels down the front roof to avoid the chimney shading.

 

Push the panels on the extension as far away from the house as you can to minimise shading. The panels are made in strings. If you shade the bottom of a portrait panel then you shade all three strings and you knock out the panel. If you shade the bottom of a landscape panel then 2 of the three strings still work, the bypass diode will bypass the third, and the panel will produce. Depending on your shading landscape may (often will) perform better. On-roof fit installers hate landscape because it needs more rails and you're faffing more finding rafters etc. With GSE trays it matters not one jot.

 

https://www.sciencedirect.com/science/article/pii/S0038092X21003868

 

 

 

If you do need optimisers go with Tigo. They'll work with any inverter.

Thanks for the fantastic response. 

 

Okay, sounds like dodging SolarEdge is better. Thanks for that advice. I knew the optimisers were only compatible with their inverter but hadn’t thought forward enough to think about what I’d need to do if/when one failed.

 

Shading shouldn’t be a problem, except for the different orientation they face. Ie - all of the west facing ones will get sun without shading, both the east facing ones will get sun without shading. Obviously that will be in the evening/morning respectively with a period in the middle of the day they *should* both get sunlight.

 

Re: landscape at the front. I probably could change my order to accommodate that. But I would need to work out which panels would go with landscape GSE frames - which I was finding hard before. I also need them relatively quickly as the builder wants to make the roof watertight this week coming. Any idea which panels would work with which GSE frames? Would probably need to be minimum 350w panels to make it worth changing from the 3 x 370w I have ordered.

 

Then the inverter. My feeling is that two small ones are probably more hassle than one slightly bigger dual MPPT. Are optimisers still needed if all the panels of each string will get sun equally, and they’re on separate inputs into the inverter?

Any recommendations for such an inverter?

 

15 hours ago, JohnMo said:

I have tigo optimiser and Growatt inverter.  But mines 10 panels in a row as a single string. All seems to work ok.

 

With panels on different elevations you would ideally do two strings.

 

But I think I would just go down the route of micro inverter, no optimisers required keeps it all simple.

 

Thanks - assume you mean a single small inverter with input for 2 strings, rather than a single input? I think wiring up two separate inverters to the house might create a bit much hassle. Any suggestion on which inverter?

 

12 hours ago, markocosic said:

Depends on shading - two inverters is simple (and they're cheap / not mounted on the effing roof under the effing panels, for when they fail before the panels do) - but optimisers will make better use of the PV in shading conditions

 

By the sounds of it I can get away without optimisers, if the panels are on separate strings to the inverter. thanks for the insight!

[markocosic]

East West need 1 mppt each. Optimisers not useful if not shaded.

 

Two small £250 inverters require you to mount with 8 screws and connect 2 AC cables.

 

One big inverter with dual input requires 4 screws and 1 AC cable.

 

They can share the cable running back to the consumer unit.

 

The DC side is identical. These even have the isolation switches and WiFi built in.

 

It's all of 30 minutes work to screw both units on the wall and wire them. Most of the work is getting the DC cables to them and the AC cables to them.

 

I don't know of a low power dual input inverter with dual DC isolation switches built in that's compatible with your current ratings and costs £500.

 

 

Search by all means but this is what I decided was best for a small south + west arrays on my own house.

 

When my hideously oversized and inefficient but £50 3.6 kw dual input Aurora inverter goes pop that is. Can't justify £450 to replace what already works given that I'm spilling much of the generation as is.

 

(1x "1.6 kW" array, 1x "1.2 kW" array, but built from £140 worth of secondhand 20yr old panels with bad orientation, because they were there on eBay and nicely offset the daytime hotel load)

[Timmyk]

Bit more research this morning. I think I can get 4 x Q cells G9+ 345w at the front and 2 at the back. So total system will be 2070w; 1380w west and 690w east.

 

5 minutes ago, markocosic said:

East West need 1 mppt each. Optimisers not useful if not shaded.

 

Two small £250 inverters require you to mount with 8 screws and connect 2 AC cables.

 

One big inverter with dual input requires 4 screws and 1 AC cable.

 

They can share the cable running back to the consumer unit.

 

The DC side is identical. These even have the isolation switches and WiFi built in.

 

It's all of 30 minutes work to screw both units on the wall and wire them. Most of the work is getting the DC cables to them and the AC cables to them.

 

I don't know of a low power dual input inverter with dual DC isolation switches built in that's compatible with your current ratings and costs £500.

 

 

Search by all means but this is what I decided was best for a small south + west arrays on my own house.

 

When my hideously oversized and inefficient but £50 3.6 kw dual input Aurora inverter goes pop that is. Can't justify £450 to replace what already works given that I'm spilling much of the generation as is.

 

(1x "1.6 kW" array, 1x "1.2 kW" array, but built from £140 worth of secondhand 20yr old panels with bad orientation, because they were there on eBay and nicely offset the daytime hotel load)

 

This is very useful - thank you. Sounds like two small inverters with a shared AC cable back to the consumer unit is the way to go. Still recommend the Sofar inverter you mentioned above?

 

Thanks again

Tim

Edited April 9, 2022 by Timmyk

[Timmyk]

Having had a look, I think 1 x Growatt 1500 TL-X for the west array and 1 x Growatt 750 TL-X for the East array would work well.

 

Any thoughts on those?

[markocosic]

No internal DC isolator. So you need more gubbins to fit them which will look ugly take longer and end up costing more. No monitoring via WiFi. Unless you buy more widgets.

 

Mppt range and startup range ok.

 

If you're sorting by price then do remember to include all the extra bits. You might want to ask your installer what they're prepared to fit too.

[Timmyk]

Ah okay. Understood.

 

By the looks of it, if I go with the ones you are suggesting then I'll need the Sofar 1100 inverter for the east roof (690w) and the 1600 for the west roof (1380w). Maybe even the larger 2200 for the west facing roof as there may be scope to expand onto the flat roof with south facing panels in the future. Correct?

 

My builder is going to install the GSE frames and panels. I'm going to connect them up to the inverters myself and get an electrician to hook those onto the consumer unit.

 

Hope that's not too mad a proposition.

[Timmyk]

Does anyone know if it is acceptable to run a 6m or so length of DC cable under the floorboards without a conduit? I think max voltage will be only 70 (690w/10a). Going to need to get the cable run from the back to the front of the house.

[Dillsue]

On 09/04/2022 at 11:46, Timmyk said:

Okay, sounds like dodging SolarEdge is better. Thanks for that advice. I knew the optimisers were only compatible with their inverter but hadn’t thought forward enough to think about what I’d need to do if/when one failed.

 

Re: landscape at the front. I probably could change my order to accommodate that. But I would need to work out which panels would go with landscape GSE frames - which I was finding hard before. I also need them relatively quickly as the builder wants to make the roof watertight this week coming. Any idea which panels would work with which GSE frames? Would probably need to be minimum 350w panels to make it worth changing from the 3 x 370w I have ordered.

As I understand it Solaredge inverters require a minimum of 8 optimisers so probably not an option anyway unless you go for one of the smaller optimiser integrated systems you mentioned previously.

 

I beleive Solaredge produce a tool that allows you to disable the safe 1volt output feature when comms to the inverter is lost. This allows you to use their optimisers with any inverter so not locked to a Solaredge inverter.

 

Im not suggesting you go for a Solaredge system but my experience of the 2 SE systems we have is nothing but positive....so much so a 3rd SE system is going on my daughters roof in the next few weeks.

 

If youre thinking of changing to landscape orientation on account of the shading from snow, consider how much snow you get in Cambridge, how long the snow persists for and how much generation youll loose for a couple of days of snow cover when your panels are generating next to nothing in the winter anyway

[Timmyk]

21 minutes ago, Dillsue said:

As I understand it Solaredge inverters require a minimum of 8 optimisers so probably not an option anyway unless you go for one of the smaller optimiser integrated systems you mentioned previously.

 

I beleive Solaredge produce a tool that allows you to disable the safe 1volt output feature when comms to the inverter is lost. This allows you to use their optimisers with any inverter so not locked to a Solaredge inverter.

 

Im not suggesting you go for a Solaredge system but my experience of the 2 SE systems we have is nothing but positive....so much so a 3rd SE system is going on my daughters roof in the next few weeks.

 

If youre thinking of changing to landscape orientation on account of the shading from snow, consider how much snow you get in Cambridge, how long the snow persists for and how much generation youll loose for a couple of days of snow cover when your panels are generating next to nothing in the winter anyway

 

Thanks. 

 

The orientation thing was actually about making better use of the roof space and having 4 x 345w panels instead of 3 x 370w. 

 

Seems SolarEdge divides opinions!

 

Because of the orientation of my arrays (one east, one west), it seems I need 2 small inverters to make it work. Just starting to wonder if it's even worth the bother having the 2 panels on the east side really. Only 690w created at the best of times 🤷‍♂️ And a reasonable cable run to get it to an inverter. Argh, this is more complicated than I thought it would be!

 

Edit: Just looking again at SolarEdge compact. The (singe) 6-input optimiser that comes with this inverter seems perfect for my needs re: 2 azimuths - it even states this as a use example in the product brochure. I could locate  the optimiser inside the loft so if I ever need to change inverter then it's easily removed. Then all the bases seem covered 👍

Edited April 10, 2022 by Timmyk

[Dillsue]

Not sure how much space youve got on the east side but you can get 450+ watt panels so not far short of 1kw if you can get two x 2100 x 1050 panels in. Only you know if you can get them in and if its worth the effort. Maybe worth running the 2 panel set up through PVGIS to see how much theyll produce

 

If I was in your place, Id speak to the local planning office and sound them out about panels on the flat roof. If they sound positive I design what you do now assuming that youll be adding to the system in the near(ish) future

[Timmyk]

19 minutes ago, Dillsue said:

Not sure how much space youve got on the east side but you can get 450+ watt panels so not far short of 1kw if you can get two x 2100 x 1050 panels in. Only you know if you can get them in and if its worth the effort. Maybe worth running the 2 panel set up through PVGIS to see how much theyll produce

 

If I was in your place, Id speak to the local planning office and sound them out about panels on the flat roof. If they sound positive I design what you do now assuming that youll be adding to the system in the near(ish) future

So with the current plan, it estimates 600kWh East and 1128kWh West / year.

 

That SolarEdge compact 2.0kW takes up to 8 panels, so I've got the potential to pop 2 panels on the flat roof in the future...or change inverter!

[Timmyk]

Managed to change/cancel my order for 5 panels and am now going for 4 x Qcells G9 345w to West, 2 x Qcells G9 345w to East, all in portrait with GSE frames.

 

Inverter will be the SolarEdge Compact 2kW extended, with the optimiser placed indoors in an accessible place if I ever need/want to change it. This inverter looks perfect for my needs as it accommodates different azimuths and has a low startup voltage, built in wifi and (I think) a built in DC isolator.

 

Whole thing should cost around £2400, which probably isn't terrible.

 

How does this setup sound? Is there anything obvious I've missed in this?

 

Thanks again!

Edited April 11, 2022 by Timmyk

[JohnMo]

You may want to add an external DC isolation switch also.

 

Will make future maintenance easier, such as inverter replacement, otherwise how do you disconnect the DC wires from the inverter etc.  The wires will always be live.

[joth]

7 minutes ago, JohnMo said:

You may want to add an external DC isolation switch also.

 

Will make future maintenance easier, such as inverter replacement, otherwise how do you disconnect the DC wires from the inverter etc.  The wires will always be live.

If you use SE optimizes they automatically disconnect their output (in fact, reduce it to 1V) when the inverter is shutdown.

Still, I'd want the DC isolator anyway.

 

The other option is only do maintenance at night or very cloudy days. And MC10 are fairly safe from accidental contact.

Still, I'd want the DC isolator anyway.