My first PV set up - advice please !

[bmj1]

Hi all,

 

Just finished my build, my typical daily usage looks like this:

 

 

(28kwh total in 24h)

 

I'd like to install solar PV (and perhaps a battery), objective is to maximise ROI.

 

I've got a flat roof with room for:

 

1) LONGi Solar Hi-MO X6 Explorer 535w Black Frame Solar PV Module - LR5-66HTH-535M

- x8

=> 4,280Kw system

 

Or alternatively:

 

2) Sunpower SPR-P6-400-BLK (400w)

- x10

=> 4,000Kw system

 

Or alternatively:

 

3) Aiko Neostar 2S 510W N-Type ABC, 60 Cell, All Black Solar Panel (510W)

- x10

=> 5,100Kw system

 

Or alternatively:

 

4) JA Solar JAM66D42-580/MB

- x10

=> 5,800Kw system

 

I'm in London, so I understand a 5.1Kw system can be expected to generated ~10Kwh of electricity per day (so roughly 36% of my daily usage). Spread over 8 hours of sunlight, that's 1.25Kwh of energy per hour.

 

With this in mind - I'm tempted to skip the battery altogether.

 

So I just need:

• PV panels (and mounts)
• 3-phase inverter (would a "Huawei SUN2000-8KTL 8KW 3 Phase Solar PV Inverter dual MPPT" be okay ?)

 

If I understand correctly, if I'm generating 10Kwh of electricity per day, then that is 3,650 per annum, so I'm saving ~£900 per annum at current prices. If I spend £2.5k on the panels and mounting kit, another £500 on the inverter, £250-£500 on a sparky for 1-2 days to wire it up, I'm looking at pay back in 4 years.

 

The other benefit of avoiding the battery is it's in a small plant area, and I've got my networking kit in there, and I don't really want to heat it up and need to put in A/C or some more active ventillation.

 

Key Questions:

1) How do I know if the panels are compatible with the inverter I've proposed above ?

2) Is an 8KW inverter sufficient for 5.1Kw of panels

3) Any other thoughts/suggestions ?

 

Edited Friday at 16:52 by bmj1

[JohnMo]

14 minutes ago, bmj1 said:

so I understand a 5.1Kw system can be expected to generated ~10Kwh of electricity per day (so roughly 36% of my daily usage). Spread over 8 hours of sunlight, that's 1.25Kwh of energy per hour.

 

Be careful with assumptions and PV. Output varies hugely, based on shading, direction, angle, and time of year. Expect almost nothing mid winter. The figures you quote are average figures so a bit meaningless.

 

You size inverter mostly to match it's output, you can overclock also, so move PV kW peak panels than kW rating of the inverter.

 

Various inputs into inverter need to be taken in to account, start up voltage, max volts, max amps etc.

[bmj1]

Ok - I guess I can start with the PV - see what I'm generating - and then size the battery and add that in - if it makes sense to do so ?

I'm not clear how I size the inverter though ? E.g. for the 5100Kwh system above, what size inverter would be appropriate ?

[SteamyTea]

Right, apart from my normal nag about units, don't use mean averages.

The important point to keep in mind is that PV is variable.

It can swing between 10% and 100% maximum rated output in a few seconds.

What this implies, in practice, is that you cannot use the output to run household appliances in real time.

This is why storage is used.

The cheapest type of storage is thermal. The most useful form being hot water. This is just a case using a standard cylinder, with an immersion heater in it, but controlled by a PV diverter.  All the diverter does is sense when there is excess PV, which would be exported to the grid, and diverts that to the immersion heater.  This is usually done at the 1 Wh level (usually one flash from the LED on your meter), though some diverters use a minimum current e.g. 5 amps as the switch (usually older systems).

Diverting to resistance heating can be done in a similar fashion.

Battery can be charged in a similar fashion, but they are more expensive to install.  The economic case is probably not realistic when looked at in isolation, but may be positive when variable import tariffs are taken into account.

 

As you have some usage data, and PVGIS can show you hourly expectations from a PV array in any orientation, you may want to spend a few hours doing some modelling and work out the cheapest system that gives you the most useful saving, this may involve you looking at your usage patterns and seeing if they can be changed e.g. your 8AM peak (is that electric showers, storing PV charged DHW can reduce that).

Batteries can also reduce that, just a case of working out where the biggest savings, over a year, are to be made.

 

One of the problems with PV installation is historical. When the FiT system started, the emphasis was in maximum yield per year, as this maximised payments, things are very different now. You may find that an optimally angled system i.e. south facing 35° pitch produced more than you can use for the 4 summer months, and almost nothing in the 4 winter months.

It really is a case of modelling it and not totally relying on storage, which may have to be topped up from the grid.

 

Start by visiting PVGIS and getting some data, I always use a 1kW_p system as a default then scale up once I have found the most useful orientation.

[bmj1]

Thanks @SteamyTea - here is my first cut with the 5.8Kwh array.

 

As you have pointed out, there is huge variability. In the month of July I'd basically generate my daily usage... (739kwh) - or 24kwh per day.

 

So with this in mind - how big a battery would be optimal ?

 

 

 

[JohnMo]

Start by using the correct units then it will all make sense.

 

Panels are sized in Watts. So 10 500W panels is 5000W or 5kW. That is the size of the inverter you need. So a 5kW inverter or smaller.

 

A PV array outputting 5kW for one hour is 5 kWh, for 2 hours is 10kWh. So using kWh in this discussion is a bit meaningless. Except if discussing daily output. 

34 minutes ago, bmj1 said:

28kwh total in 24h

 

PV in summer with no shading can generate for about 10 to 12 hrs, in winter 4 or 5 hrs. Summer output in your location is going to be closer 30+kWh, most of which is generated while you are at work. In winter 2kWh is more likely.

 

I see your lunch time usage is about 2.5kWh, a good summer day a 5kW PV array could be generating 5kW or between 12am and 1pm 5kWh. Great if paid for export, but not so if you are giving it away.

 

Plus what @SteamyTea says.

 

 

[jack]

45 minutes ago, bmj1 said:

• 3-phase inverter (would a "Huawei SUN2000-8KTL 8KW 3 Phase Solar PV Inverter dual MPPT" be okay ?)

...

1) How do I know if the panels are compatible with the inverter I've proposed above ?

 

Is your electrical connection 3-phase or single-phase? There's no point getting a 3-phase inverter if you aren't on 3-phase power.

 

21 minutes ago, bmj1 said:

Ok - I guess I can start with the PV - see what I'm generating - and then size the battery and add that in - if it makes sense to do so ?

 

Sure, although there are some good online PV generation estimation tools available. What might be harder to predict is your usage. Worse, even if you can estimate your usage patterns now, that will change signficantly if you get a battery. 

 

21 minutes ago, bmj1 said:

I'm not clear how I size the inverter though ? E.g. for the 5100Kwh system above, what size inverter would be appropriate ?

 

Assuming you don't expect to add more PV in the future, an inverter that can handle the peak power output is fine. Some people slightly under-rate the inverter, because in the UK, it's unlikely your panels will ever deliver their rated power output.  A 5 kW (or more) inverter would be fine for this installation, but even a 4 kW inverter would rarely top out. You might also need an inverter with an output limit if you don't get approval to install more than the standard ~4 kWp in solar panels.

If you've considering adding batteries in the future, you should consider getting an inverter (e.g., a hybrid inverter) that will work with a battery system. 

[bmj1]

1) I've got 3-phase.

 

2) Corrected chart for a 5.8Kw array:

 

 

3) I now understand that it's ideal to undersize the inverter by ~20% in the UK, so if sizing a 5.8Kw array, then a 5Kw inverter is about perfect.

 

4) Right now a bunch of energy companies are paying 15p/Kwh for SEG in London. So I'm think it makes sense to skip the battery altogether for now (but have a hybrid inverter). Anything I don't use, I'll export, and the net/net is it reduces my energy cost from 22p/Kwh -> 7p/Kwh (22p-15p).

 

As/when battery prices come down, or SEG prices drop, then it makes sense to bite the bullet and get a battery.

 

Thoughts ?

[SteamyTea]

Right, here is a better chart of what really happens.

 

It shows the mean power, based on 18 years of hourly data, plotted for a whole year (I shall do some monthly charts once I have finished my supper).

It also shows the minimum power, if that power is greater than 0 watts, and maximum power.  The Standard Deviation, in watts, is the middle 68% of the mean power.  This allows you to accurately estimate the usable power.

 

 

You will quickly notice that there is a huge difference between maximum power and mean power.

[JohnMo]

1 hour ago, bmj1 said:

Right now a bunch of energy companies are paying 15p/Kwh for SEG in London

But generally only for a MCS certified install, not a self installed system or one where you get your non-MCS certified electrician. So make sure you are fully informed before making any spend decisions

 

[Dillsue]

SEG is a government backed scheme that typically pays 4-6p/unit. The 15p/unit rate is a Power Purchase Agreement ie a contract between you and an eleccy company. There's no government involvement/backing and in theory could end at anytime. Lots of people including me sell under a PPA but best that you understand it's different to SEG

[Dillsue]

Hopefully you're aware that if you want to connect an inverter greater than 3.68kw output you need prior permission from your DNO so a 5kw inverter is a no no without approval

[JohnMo]

3 minutes ago, Dillsue said:

Hopefully you're aware that if you want to connect an inverter greater than 3.68kw output you need prior permission from your DNO so a 5kw inverter is a no no without approval

But not if a 3 phase, as per phase will be below 3.68kW, so should be G98, tell them after, I believe?

[Dillsue]

3 minutes ago, JohnMo said:

But not if a 3 phase, as per phase will be below 3.68kW, so should be G98, tell them after, I believe?

Yes...I missed the presence of 3 phase!

 

If the 3 phase inverter distributes the PV equally over all 3 phases but the house loads aren't evenly balanced across the 3 phases then it may be more beneficial to have a single phase inverter hooked up to the phase with the bulk of the load on it??

[Beelbeebub]

37 minutes ago, Dillsue said:

Hopefully you're aware that if you want to connect an inverter greater than 3.68kw output you need prior permission from your DNO so a 5kw inverter is a no no without approval

Most newer inverters have an export limiter so you can run an 8kw inverter with a 4kw export limit. 

 

I believe you still need to fill in a form but there isn't an approval issue. The worst case is they limit you to the 3.7Kw "default". 

[SteamyTea]

It is 16 Amps per phase.

Local voltage may vary so:

 

P = A x V

 

Taking the extreme allowed voltages of 230 V + 10% and 230 V - 6% you get

 

P_max = 253 x 16 = 4.048 kW

 

P_min = 201.2 x 16 = 3.4592 kW

 

 

[Dillsue]

10 hours ago, Beelbeebub said:

Most newer inverters have an export limiter so you can run an 8kw inverter with a 4kw export limit. 

 

I believe you still need to fill in a form but there isn't an approval issue. The worst case is they limit you to the 3.7Kw "default". 

That setup needs prior approval via a G99 plus G100 to cover the setup of the export limitation. I think there's an awful lot of inverters that support export limitation but unless the limit is factory set to 3.68kw and G98 documented as such, then you need consent before connecting.

Edited yesterday at 06:53 by Dillsue

[SteamyTea]

Here are the monthly charts, based on my location, which is a bit better than most people.

 

 

[JamesPa]

On 06/06/2025 at 17:24, bmj1 said:

I'd like to install solar PV (and perhaps a battery), objective is to maximise ROI.

 

I installed solar panels over 14 years ago.  If I were installing panels now I would

• Fit as much panel capacity as the roof will take - as others have said you can have greater capacity than the max output capacity of the inverter.  My only regret about what I did 14 years ago is I could have had more panels, and of course I cant now match the panel design
• Limit export  to the G98 value (16A/3.6kW) per phase for simplicity.  You arent going to exceed this very often (never if you have 3 phase)
• Get a decent export tariff (eg I get 15p/kWh for export with EON, Octopus do similar)
• Use DHW (ie a solar diverter) as the only on-site storage...
• ...unless you have an EV which is frequently at home during the day, in which case then rigging up a simple 'divert to EV if export > (say) 1.5kW' is worth considering - I did this for a while with a granny charger, Home assistant and a Shelley relay, but now don't bother as I have a night time import tariff that is less than my export tariff, so its easier just to charge on import at night.

 

I would probably fit a hybrid inverter however, ie one that enabled (DC coupled) batteries to be added later, but I wouldn't bother with the batteries until they get a lot cheaper.

 

With a decent export tariff you have to worry less about self consumption, so managing the use becomes easier, and the business case* for batteries more or less vanishes (at least for my usage pattern).  Of course export tariffs may reduce with time, but then the case for a battery will be stronger (and batteries anyway cheaper) hence why I would consider fitting a hybrid inverter now.

 

Modelling accurately in the real world house is rather difficult, so simplifying the question may help.

 

 

-----

 

*Unlike solar or an ASHP, the environmental case for batteries is at best arguable, possibly non existent, so my current personal take is that they need to justify themselves financially.

 

 

Edited 9 hours ago by JamesPa

[SteamyTea]

1 hour ago, JamesPa said:

Modelling accurately in the real world house is rather difficult, so simplifying the question may help.

Been trying to do that for my place.

I agree that fitting the maximum amount of panels is the best option because of their cheapness is best.

Storing thermal energy is also cheap, even a storage heater in a room will help.  Water is best.

 

The only problem with 'tariff switching' is that it is not a long term solution, so hard to make investment decisions around it.  This is part of the reason that the FiT was introduced, it took the risk out of the equation. And who foresaw the price rises and volutilty we have had over the last few years.

 

The cheapest energy is the kWh you don't need, working to reduce that is generally the biggest saving.

 

[JamesPa]

1 hour ago, SteamyTea said:

 

The only problem with 'tariff switching' is that it is not a long term solution, so hard to make investment decisions around it.  This is part of the reason that the FiT was introduced, it took the risk out of the equation. And who foresaw the price rises and volutilty we have had over the last few years.

I agree entirely.  That's one reason I haven't bothered with a battery, with current tarrifs it makes no sense for me and I can't predict future tariffs.  Only if either batteries come down significantly in price or the gap between export and import prices widens significantly (and looks to remain wide) can I justify this particular investment. 

 

Furthermore I wouldn't be surprised if, by the time that happens (battery prices don't seem to be falling that fast!), getting a cheap leaf with bidirectional charger turns out to be a great second car solution. 

 

Obviously the numbers vary from person to person.

[JohnMo]

1 hour ago, SteamyTea said:

working to reduce that is generally the biggest saving.

I just tweaked my cooling temperature up 0.2 degs today, so it doesn't needlessly fire up too often.

 

Added a heat pump dryer recently also, to replace a normal one. Big drop in consumption.

 

2 hours ago, JamesPa said:

but then the case for a battery will be stronger

We get zero paid for export, so the financial case for a battery was pretty easy. Any charging is done at night, when grid is pretty green normally, then via PV, so very green. The battery basically offsets nearly all usage to cheap rate and with PV allows you use it later and into the next day.

 

We generally import around 5kW, some days lots more (heating season) others less, at 13p/kWh, but consume 20 to 22kWh (with cooling) and more in the heating season, about 0.5kWh escapes to the grid each day. So the cost of consumed electricity is about 3-4p per kWh, for at least 6 months of the year. Since the 24 Dec we have only used 160kWh of expensive rate electric (30p), and since 24 March to 24 May after making a few changes only 20kWh, so about 0.3kWh a day (the battery seems to use this balancing loads).

 

With just PV on it own, we used the immersion diverter, did washing etc (but only so much you can do) and the rest escaped without being paid for it. Now almost all PV is used by us and it has a value, if surplus it goes to the battery or when full, 3kW is fed to immersion and heat pump is run a elevated (heating) or reduced (cooling) flow temps.

 

 

[SteamyTea]

16 minutes ago, JamesPa said:

Only if either batteries come down significantly in price

In reality, they need to be able to deliver at around 10p/kWh, then they can compete with natural gas.

 

Assuming a battery can do a 20% to 80% charge/discharge 3000 times, then:

 

1 kWh become 0.6 kWh

 

If the price is £500/kWh that becomes £500/0.6 kWh or £830/kWh.

 

830 [£] / 3000 [cycles] = £0.28/kWh.

 

So realistically you need to install at <£166/kWh.

 

Now if natural gas is not available, then the price difference is not so bad, but then it does rely on not paying to import energy, so at best they need to be £250/kWh.

I am not sure how close were are to that at the moment.

According to this:

https://heatable.co.uk/solar/advice/battery-storage-costs

The price in 2024 is between £265 and £415/kWh.

 

[SteamyTea]

12 minutes ago, JohnMo said:

Added a heat pump dryer recently

I used to have one on an old dryer.  While it worked well, I did find that the extra drying time was a bit annoying, though I am told they are better now.

When I moved down here full time, 20 years ago, I bought a washing line.  Cost £2 with pegs (from Poundland).

Best bargain ever.

[SteamyTea]

15 minutes ago, JohnMo said:

We generally import around 5kW

For how long?