Solar Panels UK - is it worth it?

[PITA]

So I'm 2 months in (or so) and thought it worth writing up my experience for people. Its been quite a trip with many on here helping massively along the way.

TL/DR:
We got panels, 19 of them.
Install was more involved than expected, but OK
Fully solar powered in June & July, saving about £180/month so far in the summer
So.energy - reasonably happy with.
Pure drive 10kw battery - probably wouldn't recommend but working fine

Full story

The Plan
Reminder of the plan since around March time. We happened to be talking to so.energy about panels and batteries when the Ukraine thing kicked off and the price of power started going mad(er), realising this likely would lead to additional supply demand on solar we fairly rapidly firmed up a commitment to using them and got on with some simulating. We already had full electric in the house - Heatpump for water and heating, M3, induction cooking and so on, so were using a lot of power, although roughly equivalent to the value of gas we were using pre-conversion.

On the advice of @PITA (I think) I investigated populating our split roof with some panels facing the NE, in addition to what installers would usually recommend on the SW roof. Frustratingly the NE roof is overall a better shape for panels, but that's life. We eventually settled on 8x375w panels on the NE, 14x375w on the SW and a 10kw Puredrive battery. They initially quoted for an 8kw Solus inverter, and that is what is currently installed.

So.energy were ...decent... through out. They always got back to questions, and answered some pretty techie stuff well. Only downside was that replies to emails generally took ~5 working days. But they were answered and always within that time, and always well, so I was actually reasonably happy. I stuck with them, as I said due to having a place in the queue booked, all the equipment looked decent although not cutting edge and a key thing for me - its a big company with interests beyond solar. So everything comes backed with a 10 year guarantee and there is a reasonable chance they will be around that long.

Total price was £12400, with payback at today's prices in 5.5 years. I'm hearing horrible rumours of 75p/unit in the future, which would make this pay off in 2.5 years, hopefully it doesn't come to that however!

The install
Install days approaches and a scaffolder turns up to measure. His plans call for blocking a neighbours drive for a week. I tell him that's not going to fly, so he does a re-design. He also inserts a bunch of 'safety measures' which I say sound like I good idea, and he takes that as authorisation to spend more than planned. Not sure why so.energy let him off with that, but it wasn't reflected in my price so *shrug*.

About this time I start getting calls from a lead at so.energy. Apparently my SW facing array should have Tigo optimisers as 4 of the panels are at slight different angles (possibly 15 degrees of a difference) and they wanted to install them on all the panels, but can't even get 4 to put on the ones with the variation. He insists that the inverter just won't start up at all with 4 panels on a different plane. I (still) think that's BS - they might limit the total output to the lowest common denominator, but with no shading and not that much variation I am unconvinced it would have ben worse than adding that extra kw of installed panel. 2 of the panels I wasn't that bothered about (they had been moved from the NE side to aid the aesthetics on the less productive side), but I really would have liked the other 2. Still no go - option of wait for the Tigos to be back in stock (August sometime), with the scaffolding up, or drop those 4 panels. I used these conversations to persuade them that with a split array I really didn't need the 8kw inverter as we would be maxing out in a practical sense at ~4.6kw. Given we should be aiming to clip a little, 8KW is massively oversized and the more senior designer agreed. We settled on a 5kw as it was the smallest their software would allow given the size of the array, even tho they were facing different directions.

All agreed, deliveries start turning up. But 22 panels and an 8kw inverter and the battery all arrive. This is going to be interesting.

After many discussions we agree that actually we can squeeze 11 panels into the space that they had planned 10, stick with the 8 on the NE, and they will ship a 5kw to me ASAP and do a swap. This ends up with 8x375 = 3kw installed NE array, and a 4.125kw SW facing array for a total install of 7.125kw installed panels.

The guys doing the install were great, but it was good I was very familiar with the house and there to solve problems. As we wanted all the cabling etc routed internally it was pretty much up to me to spec exactly where to route things. At one point I had to get under the house and rescue one of the guys as he got lost, and had planned to leave ~100m of armoured cable winding all over the place for the battery. I sorted that out and cut it down to the 7m it needed to be. Well worth taking a day off if you want to be particular about the install I would say!

Anyway, all in and installed in 2 days, stressful but largely happy result all around.

Performance
Since installation we have generated (and used almost all of) 1077kwh, saving around £300. Until last week we hadn't used any grid power at all, other than a couple of errors I made and charging the car. With a couple of duller days this week I have converted over to charging the battery at night on our E7 rate rather than running it down over night, trying to get a feel for what level it has to be at to make it through the day and evening on sunlight.

The split array is working pretty well, a typical bright day looks like this for the 2 arrays:
View attachment 842095
That NE array (in grey) really supporting and getting that early boost in the morning. On a cloudy day the 2 curves almost match, and thats where just the sheer number of panels helps.

I have started using solcast.com.au to predict the energy which is working well (which is where the data for the above pic comes from). However doing that has highlighted that the 8kw inverter is really hamstringing us on dull days. On a good day, the forecast will be within ~5% of what we get, but consistently if the forecast is 10kwh or lower, our actual production drops off a cliff. The 15th of this month was predicted to be ~8.6kwh, but we only produced 3, 16th predicted to be 10 and we got 6.3. Today (17th), 23.7 predicted, 22 delivered. It could just be a forecasting error, but there is enough other data to support that there is a problem at the bottom of the scale. This is because we are operating in the cliff part of the inverter efficiency curve in low output - 800w is 10% of the rated power, and destinctly in to sub 90% efficiencies. At 500w we could be down to ~60% (can't find the specific curve for our inverter, so guessing a bit). Swap to the smaller inverter is booked so hopefully these numbers will improve.

Battery
Now for the weird bit. The battery. These look pretty funky looking at the datasheet and pictures:

https://puredrive-energy.co.uk/wp-content/uploads/2021/05/Home-AC-10kWh_Data_Sheet.pdf

Reviews on line are largely positive (except one youtuber who I decided was just a dick and insisted on sending it back because the installer hadn't put any warning stickers on it. Hopefully no one here) so I went for it. A generous interpretation of the data sheet gives you a 5kw discharge rate and either an unspecified or 3.7kw charge rate if you do the sums. Both are unfortunately wrong. To be clear, the max discharge is 3kw (which is actually reasonable enough to do the job) but the max charge rate is only 1.8kw. Which is pretty meh. It also only discharges day to day to between 10 and 15%, after which it starts cycling between that and 20% if you are out of juice. This just means that you start paying extra for your power as you have to pay a conversion fee each time you put power in or out of the battery. The charge rate also limits you somewhat. Assuming the last kw is reserved, [email protected] needs 5 hours to charge, not the 4 octopus would generally give you.

The UI is also, rubbish. There is a nice app from Pure Drive that lets you monitor stuff:

View attachment 842089View attachment 842090

but you can't change anything there. Setting charge times or levels all comes through the HTTP interface for the battery its self. Quite a nice front end:
View attachment 842049
but setting anything is deep in a horrible menu system. 20 clicks (I just counted) through stuff like this:
View attachment 842050

The battery charge schedule is just off the bottom of that screen - I have to scroll past the option to enable 3 phase every time I want to change the charge level. Or stop that bouncing by putting a fake charge level to 15% until the cheap times at midnight. And again when I have to remember to turn that off before the next day (no one time option in the UI).

I love having access to all this, but day to day its a complete pain and TBH one day I'm going to break something either by fiddling, or by mis clicking on the phone.

What it does expose however is that this the puredrive battery may be a pretty neat battery, but the overall package is actually co-ordinated by a Victron Energy MultiPlus-II. I've not delved in too far yet, but there is a pretty large online community setup to support these. They mostly run boat battery systems, but they are also used for hospital backup power and so on. So pretty good stuff, but I still need a way around using the HTML interface.

Finally, it does do anti-islanding, but only for 1 down stream circuit, not the house. I'm still debating if I should attempt to use this and move the servers, lights and kitchen sockets onto it. It would be quite a pain as we already have 3 CUs now - a 4th is going to be really tight!

Definitely slightly grumpy about the battery, but it is working and I do plan to have an entertaining discussion with them at some point soon about the charge, discharge rates and a UI for controlling overnight charging and applicability of consumer protection laws around clarity of advertising.

Future plans:

• More automation of everything. I'm running fully manual just now to get a feel for the decision tree needed to minimise our grid intake. Plan however is to move my forecasting from excel into some python which will then configure the battery over night charge appropriately, set the heatpump to absorb any extra power and email me telling me to plug in the car (or not).
• I also have the Andersen A2 charger set up so it should be able to absorb extra power, but the CT clamp its using is reading inaccurately, and when it tries to grab power the house battery starts discharging into the car which I don't want (again, that round trip)
• There is 1 panel that could do with a Tigo optimiser (it gets shaded by a dormer later in the day), for £30 I'm willing to give it a shot and seeing if it helps)
• I need to get an eddie to divert excess power through the day into the HW, rather than triggering a heat cycle on the HP. Although this works, there are often a few 100w's being exported through the middle of the day as we exceed that 1.8kw absorption and capturing them to put into the water, even in an inefficient immersion way would be better

Hopefully this helps a few people and provides some interest about the Puredrive battery. Oh yea, and finally, don't let the sparky terminate any Cat5. This attempt took out my whole home network somehow until I tracked it down at the battery, snipped it off and re-did it for him. Imbecile.

View attachment 842087

So are you happy that you made the initial investment for the NE panel set now?

We've got 10 x 400w panels on NE and they really help support the system.

 

[Avendit]

So are you happy that you made the initial investment for the NE panel set now?

We've got 10 x 400w panels on NE and they really help support the system.

Absolutely. They are predicted to really help in the shoulder seasons to keep us off prime time power and they are adding meaningful additional power over the summer too.

With solar I've decided it's really simple - more is just more. There isn't any optimal array, or right size other than 'what ever the hell the max is you can bolt to the roof'. Elec use will go up as more things switch so you will never have too much. While someone is installing, you may as well throw up as much as you can. Incremental costs are negligible.

 

[PITA]

Absolutely. They are predicted to really help in the shoulder seasons to keep us off prime time power and they are adding meaningful additional power over the summer too.

With solar I've decided it's really simple - more is just more. There isn't any optimal array, or right size other than 'what ever the hell the max is you can bolt to the roof'. Elec use will go up as more things switch so you will never have too much. While someone is installing, you may as well throw up as much as you can. Incremental costs are negligible.

Indeed... and Solar Panels act as a thermal barrier, so keep your sun-facing roof space cooler.

Glad you're happy

 

[JaseW]

We have a system being installed next week: 9 x 400w panels in the SE roof. We don’t use a lot of energy (average around 4-5kWh/day when not charging). I never thought about doing NW as I didn’t think it was worth it but reading this thread and @Avendit’s experience maybe I should have

I will see how this goes and look into it at a future time, our north facing roof also has some obstacles so it may be for the best. I do agree though you should get as many up there as you can.

 

[PITA]

We have a system being installed next week: 9 x 400w panels in the SE roof. We don’t use a lot of energy (average around 4-5kWh/day when not charging). I never thought about doing NW as I didn’t think it was worth it but reading this thread and @Avendit’s experience maybe I should have

I will see how this goes and look into it at a future time, our north facing roof also has some obstacles so it may be for the best. I do agree though you should get as many up there as you can.

North walls are good for installing the electric hardware outside.

Like batteries, Inverters, Distribution Boards etc...

Mostly in the shade, gets decent airflow and if at ground level good for easy access maintenance.

Providing it's out of sight and in a decent area, should still all be there in the morning

 

[init6]

Thanks for the detailed review. I look forward to an update in the winter if you'd be so kind.

 

[ACarneiro]

Thanks for the detailed review. I look forward to an update in the winter if you'd be so kind.

Winter is grim for solar production, it’s to be expected.
I produced 1326 kWh in June and 105 kWh in December. Next to negligible for a month when I use about 1600 kWh.
It’s the time of the year, though, when load shifting with the batteries really comes to its own and whereas before it would make no financial sense, if the delta between cheap and expensive rates remains this high then it might be economically more viable.

 

[init6]

Winter is grim for solar production, it’s to be expected.
I produced 1326 kWh in June and 105 kWh in December. Next to negligible for a month when I use about 1600 kWh.
It’s the time of the year, though, when load shifting with the batteries really comes to its own and whereas before it would make no financial sense, if the delta between cheap and expensive rates remains this high then it might be economically more viable.

Batteries certainly seem to be a game changer in terms of shifting low cost energy around, but I'm interested in when the grimness starts this far North. @Avendit has built up a wealth of pre-install data, so it will be interesting to see how that compares with real life.

 

[Avendit]

I had a second post planned to gather as much of the process I went through to decide that Solar + Battery makes sense, and I think that these 2 questions will hopefully be answered. At least around expectations - we will wait and see for actuals!

Thanks for the detailed review. I look forward to an update in the winter if you'd be so kind.

Batteries certainly seem to be a game changer in terms of shifting low cost energy around, but I'm interested in when the grimness starts this far North. @Avendit has built up a wealth of pre-install data, so it will be interesting to see how that compares with real life.

Much of this post will be repeats of things I've said back around page 25-30, but I wanted to gather it all into one place as I've started referring non-tesla owners to the thread for the mine of information in here between what I have discovered, and others have described. So hopefully this is useful to people thinking about 'should I get solar' and lets people build their own decision process.

I started, back at the start of 2022 by recording my power usage for a few months. I mostly used cheap wifi plugs off amazon (Meross branded, but I'm sure there are others) and some z-wave controllers we installed for a good % of lighting we have, and built in metering from some big items like car charging and the Heat Pump. I managed to account for around 90% of my usage eventually:

Trust me, words were had about that heated airer (only 6p/hour to run! they said ). I can't remember exactly which months the data was gathered, but Q1 sometime. Most are measured, some like the washing machine were worked out from their spec sheets and the numbers and types of loads we did - some devices are frankly too much of a pain to get to the plug to measure.

I then split those into 'baseload' - stuff that is basic usage and just needs to happen, and that + kitchen. These were important to me as I was aiming to try and cover our usage from solar for as much of the year as possible.

I then grabbed a standard generation curve and did some calcs to work out which months we would be above all load, base load and failing. But that was before I knew about PVWatts. This is like the abetterrouteplanner.com of PV generation.

I used the so.energy solar app (So Energy Solar) to find how many panels I could fit on each face of the roof, then ran that KWp through the nrel.gov site. This gives an hour by hour breakdown of likely output using real weather data. Stick that into a pivot table for fun and profit:

Rows are months, columns hours. From this you can see that my base+kitchen load of 406KWh is covered April through Aug with my eventual solution. First time doing this it was somewhat rougher, and I did lots of variations based on different panel & battery configurations. It was about here that 'more is more' started to become really obvious. For example, remove that 11th panel to the SW I squeezed in during the install and August becomes a scrape for being self sufficient instead of easy.

I also did some calculations on what could be load shifted to the nights, and what things look like being able to shift 10kw of power into the battery in the shoulders and winter. Its all a bit messy and using too many estimates, so I won't show it, but its a worth while process. This finally arrives at how much the battery + solar might save each year:

Pretty boring graph, but shows (with the 30p/kwh setting I have just now) that payback is sitting at 5 years. If I change that to 50p/unit, payback is 3.5 years. This is using a very simple flat rate saving from the battery as that has to account for 10kw (actually 9 I now know) cheap power some portion of the days, complete grid reliance some days and increased self consumption other days. I put a flat figure of £800/ year - no idea if thats right tbh. Again, this is all fairly rough, but was plenty to convince me that panels + a battery were the right thing to do, and that throwing as much on the roof as possible was a key point.

Then it all gets real. Its time to drop down to some kind of analysis of which out of hours rate we should be on.

We start by analysing that data from pvwatts again. Looking across the year I selected an average December day (2.5kw generated), a completely average day across the whole year (12.5kw on average), a good day (~20kw generated), and the max day for giggles (34kw generated). A bit of analysis tells me I get 140 Dec-like days, 120 average days and 100 good days. For each type of day I then looked at the actual day:

This is the 'good day', although a spring one one with the heating on, rather than an autumn one without heating.

Again, that 1 extra panel shifted an extra 20 days over to 'good days' from the average bucket - well worth the £140 incremental cost.

The base load in this case is the general background always on stuff, with some averaging for lighting and similar. I don't carefully put 6kwh into the car every evening, but it simplifies the calcs - as long as that is in the off peak period, its all fine. Cooking and heat pump accounted for. Usage is the total and the last 3 columns show the battery levels through the day for 3 different tariff settings. Economy 7 that so.energy can offer me (and we are on now - 30/21p@7hr), the standard Octopus night option (35p/0.75p@4 hours), and Octopus go faster assuming I can get a 5hr slot late in the night (40p/8.25p@5 hours). The late slot is important as in the colder months there is a lump of Heatpump work to be done between 6 and 7am. Slotting that into the cheap rate is pretty important.

By tracking the battery levels through the day we can see when it is likely to run out, and therefore how much we have to pay to cover the difference. The winter looks much nastier and you can see teh different strategies needed. In the summer you want to start the day empty, charge to 100% then run down over night. In the winter below, we start full, last as long as we can then fill up on cheap power overnight:

We still end up with a pretty negative balance at the end of the day, but that is real heat pump data and is just the way our house is just now. More insulation coming, but yep, 38kwh of power going into heating and HW at some points last year - about £10/day. But you still win on the 9kw of battery (it bottoms out at 10% under normal usage) and the 2.5kw of sun power is offsetting some of the rest a little.

Anyway. Take the highest number in each column, x the day rate, + the over night usage (inc the car) + the cost of filling the battery (a pitiful 7.2kw for this battery in 4 hours) and stick it all together. Multiply by the number of days aligned to each category to get this:

These savings are against ~£5k of pre-panel yearly costs.

Due to that key time of having a cheap rate in the lead up to 7am, and standard octopus only doing 4 hours, economy 7 is actually slightly cheaper! I do need to talk to them after the next price cap update and see if I can get onto the 1:30am, 5hr go faster, but I'm a bit nervous of doing that before everything is settled with so.energy, given they did the install etc.

Anyway, hopefully this gives others thinking about it some ideas for their own spreadsheets, but I feel it really does need to go to this level of detail to be really sure you fully understand how much money panels will make you. Even my calcs are still not right - as I get used to the system I realise a 'good day' is actually much closer to managing without charging the battery over night. If I can get to 100% at sun down, then it can usually last through the night without paying for anything other than a car charge.

Quickly back to @init6 's question... December I expect to generate aprox 80kwh of power, worth a whopping £24 (at current prices). But the battery will save me £27 on econo 7 meter, or £80 on octopus. I'll update with actual numbers, eventually!

 

[init6]

Thanks for doing that. It will be a useful reference.

but I feel it really does need to go to this level of detail to be really sure you fully understand how much money panels will make you.

I think this is key. Lots of folk say the get XkW per year and then tell you in summer they have to run around dumping excess solar into water heaters, cars etc which IMHO is not very useful.

I've got a couple of more questions, but will leave that until tomorrow - late and all that.

thanks again.

 

[Mrklaw]

@Avendit fabulous stuff. I’ve done simlar but less granular - using octopus half hourly readings for whole house rather than splitting for baseload (you can estimate some baseload on certain days when everyone is out of the house).

Have you done any estimates for an additoinal battery for the HP? finger in the air for me feels like for an ’average‘ house, an 8-10kwh battery is value for money for baseload with solar (or a high differential ToU tariff like go), and if you add in heavy usage with a HP possibly consider a second battery although maybe incremental savings at that point?

 

[Simavon]

Re "stick as many panels as you can up" that's what we did and I don't regret spending the extra to get 8kW (9.9kWp) - which was quite a long way short of double the cost of a 3.6kW set-up. This results in over 60kWh output on the best days, and annual output of over 8MWh. Our configuration is E/W. Although you probably get more output from the same number of panels facing south, E/W allows good output for longer in the day and potentially saves more on bills - with the FIT having not kept up with the price of energy, importing less at 40p is probably saving a lot more than being paid 6-7p for what we generate.

Back in 2019 we didn't go for a battery due to the cost and the payback being uncertain. Now it looks much more viable, but it's proving hard to find anyone to even quote.

 

[Avendit]

@Avendit fabulous stuff. I’ve done simlar but less granular - using octopus half hourly readings for whole house rather than splitting for baseload (you can estimate some baseload on certain days when everyone is out of the house).

Have you done any estimates for an additoinal battery for the HP? finger in the air for me feels like for an ’average‘ house, an 8-10kwh battery is value for money for baseload with solar (or a high differential ToU tariff like go), and if you add in heavy usage with a HP possibly consider a second battery although maybe incremental savings at that point?

It depends what the aim is really. If you are wanting to be as completely grid independent as possible, then you would need more battery. I did this graph which helped me a bit:

Bit complicated, but green line above orange = good. Adding another 10kw of storage would barely grab you another 2 months, although going with 2x powerwalls to give 26kw of storage would get me another ~3.5 months covered by them I think (eyeballing on that graph)? Of the single battery options a powerwall would have been lovely - that extra 3kw would bring my 'average day' into the realms of being covered by the battery. But really this is a phycological win rather than a cost one I think, as its only saving a couple of KWh. On Octopus thats a 60p saving £120 a year. Nice extra, but given the PW wasn't available I didn't go that path. (one thing to consider is that the PW would have possibly come with 400w or 420w panels giving an extra 500w of generation which may have translated into another marginal gain that could flip the whole install preference)

I'm also fighting from the other side. External Wall insulation is going on in a couple of weeks, that should bring our U value of most of the house down to ~0.14. I have a thermally weak bathroom dormer to rebuild before the winter which will treat the biggest hole left in the house. I've optimised the HP since these stats were gathered and gained about 0.5 COP by targeting HW generation to warmer times of day which will lower costs a bit (more to gain on that still). Some of the always on devices like servers and media centres were due rebuilds so have been replaced with ultra low power equivalents. My server stack is now idling at under 80w while being waaay more powerful than it was, almost halfing the cost of running it. The lights that my wife insists are on most of the day are on motion sensors and dim to 1% when no one is in the room.

One frustrating thing I found was my Sonos system. 7 speakers/connect around the house, each pulling around 7w 24/7. I looked at putting them on timers, but the wifi plugs all use ~1w too. So the payback is really poor and we just settle for manually turning them off if we are away for an extended period.

Nibbling away so hopefully a better picture next year!

 

[Mrklaw]

one more question - how is a heat pump for HW? It seems like the worst thing they can do as they want to run cooler for efficiency. But assume its still more efficient than resistive heating an immersion?

 

[btc1k]

one more question - how is a heat pump for HW? It seems like the worst thing they can do as they want to run cooler for efficiency. But assume its still more efficient than resistive heating an immersion?

2-3x more efficient than a resistive heater.

 

[Avendit]

one more question - how is a heat pump for HW? It seems like the worst thing they can do as they want to run cooler for efficiency. But assume its still more efficient than resistive heating an immersion?

2.5 to almost 4x better than an immersion heater across the year (cop = coefficient of performance, ratio of power in Vs power/heat moved). For our tank that is taking water up to 48 degc. The HP can push on to 55, which I do if there is excess power, but according to the installer 48 is about the sweet spot. 350 litres at that temp will shower 4 people unless anyone really spends too long in there.

 

[Avendit]

PS, when we decided to get the HP, the break even point Vs Gas was a COP of 3. I don't track Gas prices any more, so no idea where this lies any more.

 

[Andywil]

View attachment 843020
2.5 to almost 4x better than an immersion heater across the year (cop = coefficient of performance, ratio of power in Vs power/heat moved). For our tank that is taking water up to 48 degc. The HP can push on to 55, which I do if there is excess power, but according to the installer 48 is about the sweet spot. 350 litres at that temp will shower 4 people unless anyone really spends too long in there.

Hot water should be heated to 60 at least once a week to avoid the risk of Legionnaire's disease which can thrive in cooler water water.

 

[Avendit]

Hot water should be heated to 60 at least once a week to avoid the risk of Legionnaire's disease which can thrive in cooler water water.

The HP defaults to doing that, although we have it turned off ATM.

 

[73henny]

My set up, installed yesterday. 5kW array & 6.5kW battery.

So far so good with one or two teething problems. One of the strings isn't performing quite as well as the other. The installer came back today within an hour of me calling them and tested both strings and they're the same and fine, so it might be a set up parameter in the inverter. He's going to investigate further tomorrow. What I'm seeing is one string has a voltage that fluctuates around 130v - 140v (the 'best' string), whilst the other is always at 150v. It sometimes has an output a few hundred watts lower than the 'best' string.

Anyone on here have a clue as to what the issue might be?