Wiki: Domestic Solar Energy Pros and Cons

I realise this is just a dump of an email, but it's a bit broad brush and could do with being more specific and citing details. I think it's useful as a kind of 'things to ask' checklist, perhaps organised as bullet points?

Some comments:

You *can* export without MCS. Octopus at least will accept a standard electrical installation certificate issued by any sparky who is registered with one of the bodies like NICEIC, Elecsa, etc (basically any sparky who can do notifiable work, rather than handyman types). Octopus charge £250 for doing the paperwork to enable you.

There's also the 'plug in solar' route, which is becoming available. We don't know what will happen about export there (probably no export, but if a smart meter is in place maybe some supplier will allow it?)

You can of course import without restriction, just like every house in the country imports from the grid.

The 'I've known it take years' comment could really do with some evidence. Was it a bloke down the pub who told you this story, was it written up in a newspaper, etc?

Every installation is compatible with every energy company - it's just kWh exported. What you're talking about is 'smart' features where your import/export is set by an hourly tariff - this is only relevant if you have a battery (or an EV). For those it needs to support the API that gets them the hourly tariff data.

Inverters can also be small, as in microinverters the size of a VHS cassette which often live on the back of panels.

Loads are in kW not kWh.

Your DNO will say how much export you are allowed. The paperwork (G98) is easier if it's less than 16A per phase (3.68kW single-phase), as above this (G99) the DNO can impose restrictions and needs to pre-approve the design. You can have a larger inverter to eg charge batteries as long as the export is capped to the allowed G98/G99 limit.

Battery size also depends on the loads you have. Do you run lots of stuff like computers 24/7? Do you have electric heat like a heat pump, or is heating gas?

I don't think 'pricing of batteries' is unstable: pricing of batteries is coming down. Perhaps that trend may reverse at some point, but predicting the future is hard. Increasingly it's the cost of the packaging that's dominating rather than the cost of the cells themselves.

New tech like sodium ion may help to reduce packaging costs - eg as it'll work in sub-zero conditions it can be happily mounted outside. Sodium ion is also cheaper as the materials are abundant.

Another thing is distributed v centralised battery. Some battery systems are standalone in that they just connect to 230v AC, while others are integrated with the solar inverter. Integrated systems are cheaper, but mean you have to put solar inverter and battery in the same place. With a

The higher wattage panels are bifacial, meaning they have cells both sides. You might get that if using them as a fence, or above a light-coloured roof surface where a lot of light is reflected to the back side. If you are putting them right up against a dark coloured roof you won't get much benefit from being bifacial (I think the bifacial bit gives a ~15% boost to output, which you won't get).

What's mixed, ooi? It's just some spikes to stop birds getting behind the panels - eg if they don't nest they may sit and crap on the panels which is not great for efficiency.

ah, I hadn't thought about that. I could imagine all sorts of problems with ice dams and whatnot.

According to:

So north facing still 'works', but it doesn't work as well. Either way you need to do the sums - if you only have a N facing roof you can maybe still make it work.

In general, I think I might be tempted to reframe the article as:

1. Gather some basic information about your site:

- potential locations for panels

- roof data (dimensions, angle of pitch, compass bearing, any sources of shading like chimneys or dormers)

- what your energy consumption profile looks through the day in spring, summer, autumn, winter (eg smart meter data)

- what your electricity supply looks like (pictures of consumer unit, supply fuse, meter, etc)

- could you move more consumption to electric (EV, heat pump, induction hob, electric hot water)?

1. Calculate:

- how much roof area you have to play with

- how many panels you might reasonably fit (roughly speaking 1x2m panels at 500W each)

- roughly how much energy you might expect at different times of day and times of year at your location (eg via globalsolaratlas)

- what you might do with that energy (eg run baseload, store in a battery for evening)

- can you practically timeshift appliances, eg dishwasher, tumble dryer, EV charging) to peak generation time

1. Start talking to installers being more informed about what you actually want.

I think there's a danger with a lot of the solar quoting surveys which just ask 'how many panels do you want'; unless the answer is going to be 'as many as will fit' you probably want to be somewhat informed about what sort of number fits what you're going to do with it.

Theo

Yes, I can set it out a bit better, with context.

Good to know, thanks.

Yes.

A friend/former colleague/neighbour. Pretty sure he's not lying. Mine took days. Maybe stress it's anecdote.

Thanks - I'll delete that bit. From memory some energy companies ask for details of installed kit - assumed that was because it might matter, but evidently not.

Ah didn't know that - the ones I've seen are a lot larger.

Yep :-)

Yes, I meant that in terms of the inverter -shoudl have been clearer. Mine comes in several versions - 3kW up to 16kW (I think).

OK, thanks, I'll change that to stable, decreasing.

I'll copy/paste that if I may.

Yes, I was aware of bifacial - didn't realise it affected output on a piched roof/15cm from surface, live/learn, I'll mention that.

Just something my installer told me - sometimes they don't work, stay fixed, damaged over time. Some installs no problems reported. Just be reasonably sure you need them was the point I was trying to get across.

My neighbour's was in her gutter after the only dusting of snow we had last winter.

Interesting. Again, didn't know that. 50% is a lot more than I expected. In fact I was on a course and an installer made clear they wouldn't install N-facing, and the room appeared to agree (well, no-one disagreed).

Many thanks. I'll leave it a bit and come back to it when I have more time. Roof to get on :-)

Perhaps a section on locations for battery installs. Quit often installers seem to go for places that are "easy" but not necessarily ideal, e.g. lofts (high temperature swings bade for battery longevity), under stairs cupboard (fire risk right under your primary means of escape etc).

There is now a whole lot of recent regulation (April 2026), previously just recommendations. Batteries should not be installed in lofts, bedrooms, under stairs or or less than 1m from doors and windows. and if fitted inside a house consideration must be given for a fire proof enclosure and ventilation. Garages seem to be OK. Limited to 40kWh if fitted inside a house.

The battery on my installation weights 110kg. How could one get it into a loft> or even upstairs?

With difficulty ;-)

Is this a good place to ask what batteries people are using? I'm not convinced that solar makes sense for us, but a battery (or batteries) to load shift from peak rate to off-peak might, especially if it can live in a shed.

I think they're modular or disaassemble easily - maybe 4 x 25kg battery packs clip together inside the box sort of thing?

Interesting, thanks. Quite hard to stay ahead of the curve . . .

As ever with these stories, it helps to get to the bottom of why there was a delay. Did they just not reply to attempts to contact them? Was some bit of paper missing, that stalled the process because it was hard to generate? Did one party go bust before the process was completed?

ie is this a 'they take forever to do paperwork' problem, which could apply to anyone, or a 'unusual circumstances happened' which caused the delay, and if those circumstances don't happen to you (something you can maybe know in advance) then you're fine.

eg if you're having surgery for condition X, those people who also have condition Y make it much more risky and drag down the average. But if you don't have condition Y you know the odds are much better.

Newspaper reports tend to give a bit more detail which explain the delay, whereas fourth hand whispers down the pub don't.

Sometimes it's because they'll only offer the smart tariff if you have a compatible EV/charger/heatpump/etc that they can talk to. If you don't, they'll only offer their regular tariffs. eg 'Intelligent Octopus Go' needs a compatible EV or charger that allows them to control the charging period, but 'Octopus Go' will work with any.

Microinverters mean each panel (pair) is a separate AC generator. ie you're just distributing 230V AC around the place rather than higher current DC wiring.

The downside of putting them on the back of panels is access can be a problem if you need to replace them later - you might need to replace the inverter after 10 years but the panels after 20+. You could of course put them somewhere else.

Feel free.

Microinverters are the same idea on the generation side. They're particularly useful if you have several roof aspects, eg a complicated roof with only space for a panel or two on each section. Rather than trying to cable those back to a central inverter, and with quite different shading properties, you can just scatter panels and their microinverters about wherever you can find spare roof.

I think you could add a section about shading. Panel output is reduced by shading, and when you put panels in series then shading of one panel can reduce the current flowing through the whole string. Ideally you want the whole string to have similar shading, ie all face the same way, have nothing like chimneys or trees that might cast a shadow on one panel not all. But another thing you can do is have a power optimiser. This is a DC/DC converter that runs each panel/group of panels at its optimum via Maximum Power Point Tracking (MPPT), and then you combine those into a string feeding into the big inverter. That makes the best out of each panel without shading limiting the output.

Optimisers cost extra and are a bit similar to microinverters but without the AC side, so they're cheaper but still a substantial fraction of the cost of the panels.

It's a little bit of marketing inflating the numbers compared with what you might get in practice. OTOH if they can make bifacial for a small cost delta and it gets you a bit more generation in some cases, then why not?

It would double your payback time, which maybe means the economics doesn't work for a pro install. But if you were DIYing and could keep costs down (eg no scaffolding) then maybe doubling payback from say 3 years to 6 is acceptable.

Installers just want an easy life and maybe aren't interested in awkward jobs - eg this guy fitted vertical panels but had a hard time getting installers to do it because MCS didn't have an approved system:

Basically any cost calculation is no longer about the panels any more, it's all about the ancillaries (labour, scaffolding, mounting system). Those are where you can make the difference in payback times. Availability of the 'plug in' systems may help cut those costs down further - eg no need for MCS certification cutting labour costs.

Did you consider solar panels on that roof? :-)

Theo

Eat your spinach...

Many of them use 19" server-rack cases, so they come in lumps of 4U or 6U that are installed in a 19" rack inside the outer chassis, and then wired up. Typically everything operates using 48V. There may be a fancy cosmetic exterior so it looks less like a wiring closet from the outside.

Theo

This is something I followed from the beginning - not really 'a story'. The company was pretty useless by all accounts and is now dissolved.

He never had an explanation.

OK thanks, noted.

OK, thanks, didn't realise they were a thing.

Yes, thanks, I do mention shading more than once, but more detail could be included. As it happens the person I wrote in for is unlikely to be affected by shading.

Quite - especially for the 15%. My installer didn't even mention them. And I didn't think to ask, given my application (standard pitched roof). Live and learn :-)

Yes, I do mention that, but could add more detail.

Good point! It's actually got a half-decent aspect . . . but it's only 2.5m x

OK, so that's 'the company can't be bothered' rather than any systemic problem. Your job as client is to choose a company who can be bothered, not a bunch of cowboys.

(I think there is an escalation path via MCS, but not sure how that works)

If the panels are bifacial, be aware that the power output numbers for the panels include the bifacial aspect. So butted up against a dark roof you might expect a ~15% drop in output because there's nothing to reflect light to the back side.

It sounds perfect for a 'plug in solar' install with a panel and a microinverter. Just needs a mains connection, could be via a FCU. Wait for them to show up in Lidl...

Theo

Many are modular - installed in more modest units that can be ganged together.

There seems to be a trend for exterior wall mounted batteries so the form factor is different from rack mounted stuff like pylontech.

The advantage of pylontech is that they are fairly catholic to paralleling batteries of different ages.

It is apparently possible to parallel different makes of battery as long as they contain the same number of, same chemistry, cells in series by intercepting the CANbus signals from the inverter and spoofing them so the battery sees what it expects but that has proved beyond me.I still have a redundant 6.5kWh ac coupled battery and inverter I would like to make use of,(the battery at least).

My set up has been a decent cost saving investment in 10 years of retirement, my total gas and electric bills last year was £62.

I only use gas for DHW in winter and the only reason not to have it cut off is I think it would make a house sale difficult plus I have a cunning plan.

If the battery comes with its own inverter, it's just a 230v device I'd have thought? You can hook that up to the mains and it'll charge and discharge separately - it doesn't need to be paralleled with the other battery you have.

Now, deciding when to charge and discharge is another question. eg you probably don't want battery A to be discharging when battery B is discharging, otherwise you're just moving energy around (with losses) rather than storing or consuming. But that's something external control systems (eg Home Assistant) can tell them both what to do.

If your don't have comms between the battery and the inverter, that's where protocols like pylontech come in - there are widgets to translate from one protocol to another if they don't speak the same. There are also widgets to translate between EV batteries and domestic inverters[1].

Theo

Typo: "while battery B is *charging*"

On 15/05/2026 12:28, Theo wrote: of,(the battery at least).

Yes but apart from the extra ~80W overhead of the extra inverter there's the problem of conflicts between the solar PV hybrid inverter and the AC one. so under normal circumstances when no solar PV electricity is being exported one can charge the AC coupled battery. . I'm concerned what happens when there is no solar power and the house is running off battery. What conflicts might occur? I think the AC coupled battery will discharge first until it shuts down, then the hybrid battery would take over. As I am not sure I hesitate to do it.

Had the people that installed the hybrid just left the AC coupled unit alone I might have experimented but as they insisted on decommissioning it I cannot decide.

Yes as above.

I don't have home assistant as it is too technical and involved for me to understand.

Nice find Theo but probably beyond my capabilities.

I had a friend come round and try and sniff the protocols using a Teensy

I expect that each inverter has a current transformer around the meter tails to sense how much current the house is taking and how much is coming from/going to the grid. To run the house from battery they want to output just enough to match the loads so that the grid draw is zero, but also nothing is being exported.

So you can move these CTs around to set how much current the batteries are generating. eg if you put one on the live to the upstairs sockets circuit and one on the live to the downstairs sockets circuit, each battery would output enough to supply loads on those circuits. (they have a directionality of current flow, typically marked with an arrow, so make sure they aren't put on backwards, in which case they'll think import is actually export). There is no problem if both inverters are outputting at the same time - worst case you'll export when you didn't intend to.

You could also have relays to switch between different CTs, eg connect inverter A to the main house CT and inverter B to a dummy CT with no wires through it, with the result that inverter A will power the house and B will output (roughly) nothing. Then switch them around and inverter B will power the house. That's doable with no messing about with CANbus, and no software at all if you wish (although a bit of logic in software is always possible, eg 'if inverter A is no longer outputting [sensed by CT or by the red light on the front] then switch to inverter B')

Assuming there's nothing unsafe about it, I suspect they just weren't interested in trying to make them work together.

Theo

I think it was more to do with complying with the DNO, With two inverters and G100 approval it may have been the extra theoretical export potential.

I think your logic is probably right though. I shall see about mounting the inverter and think about where the CTs should go.

Theo I still have your email address from when we met in Haslemere, is it alright if I contact you ?

Andrew