Everyone writes one rave post when they get a Solar Power System. Here's mine.
I've done my research and decided that where we live, and the way our house is oriented, means we were extremely unlikely to get any economic benefit from having a Solar system installed. We could marginally recoup value on a set of panels, but adding a storage system would never cost-in. So what would be the point?
tl;dr: Toys and Holier Than Thou Smugness!
In November 2019, I got made an offer....
So Tesla teamed up with Octopus to offer a special deal. Get a full solar system added, and one's rates would drop (considerably) as well as having parity on power exported. The big bonus here is when one is charging one's car a lot, this cuts the costs by nearly 50% just on the imported power rate. So I was interested.
We signed up to at least get a quote. And had a very professional survey done, and got the system on order in early January 2020. We were warned that the necessary regulatory approvals would take 6-8 weeks, which was fine - all being well the system should be in place for Spring and the return of sunny skies.
And then COVID happened.
The system was finally installed in late-June 2020
If the world hadn't locked down, it looks like we'd have been ready to install based on approvals etc in early April. Instead it happened over a couple of days in late June. I want to add that it was a very impressive experience, our installers - Energize UK - were fantastic, very professional, very well organised and with a great eye for detail. And the support from the Tesla side was great too - with a good inspection/follow-up process to give lots of reassurance that the system's gone in successfully.
The guys turned up at about 08:30 on the day of the install (the scaffolding having been erected the previous day), and by 14:00 I'd had my first entirely solar-powered cup of tea. Wonderful.
What we got: Hardware
We've got a 10 panel system, with nominal 3.15kW output, a single SolarEdge Inverter with each panel connected single-string via an Optimiser. We've also got a Tesla Powerwall 2 (13.5kWh storage) and the whole system's held together with a Tesla Gateway 2 to manage it.
As indicated above, we have a modest 4 bedroom house in the South of England. Our South-facing aspect is complicated by a gable over the main bedroom which limits the number of panels we can deploy. With the benefit of Energize's experience we determined that the west-facing aspect of the gable was unsuitable due to shadowing from next door and the chimney stack, leaving the south-facing main roof and east-facing gable as suitable. There's room for 3 panels on the East-facing aspect, and they managed to get 7 panels on the South-facing aspect, giving 10 in total:
 |
| Fig 1. The Solar array fitted to our roof. 7 panels face roughly South; 3 panels roughly East. |
There's nothing in the Attic apart from cables from the panels being routed; it's possible to fit the Inverter up there but our insulation is on the attic floor, meaning it's not a temperature controlled space; I can verify from experience it gets flippin' cold in Winter and stinkin' hot in Summer, and while the Inverter's rated for those conditions it's not ideal for longevity or performance. So that and the rest of the system are all located together in the Garage instead:
 |
| Fig 2. (top-left) SolarEdge Inverter (bottom-left) Tesla Gateway 2 (bottom-right) Tesla Powerwall 2 during commissioning. It's a lot neater and tidier now! |
There's a decent argument for keeping the Powerwall itself outside - mainly on safety grounds, as it's an energy store. In our case the passage between us and the neighbour is fairly narrow - sticking it there would get in the way of getting the bins in and out, and what I hadn't appreciated is that it's got a great big glowing green light on it which would draw attention. I've done some checking and the fire tests conducted show the overall risk to the house is fairly low so all in all, indoors in the (fire-isolated) garage is our best option.
I should note that so far we've not had an upgrade to our electricity meter. We're on the list for a SMETS-2 smart meter but so far no news on when that might happen. As a result, nice hot sunny days have this effect once the battery's full and the immersion heater's warmed the water tank:
What we got: Software
Tesla App
Out of the box we have accounts with Tesla, so the App works great and shows what the system's doing and allows for customisation. Haven't felt much of a need to do much with it except set a 10% power buffer reserve for powercuts; they're rare and usually shortlived here so at normal loads that'll give me plenty of time to either wait out a short cut or (after say 15 minutes) if it's not back start doing an orderly shutdown of our household electrics. The nice thing about the Tesla app is the live view of status and where power's coming from and going to.
 |
| Fig 3. Typical power routes during the day (the kettle's on for a cuppa) |
 |
| Fig 4. Charging the Car still requires some Grid power. |
The Powerwall can supply 5kW (I've never seen it go over this although it's supposed to be able to do a 7kW surge), and that's enough for most uses. Although boiling the kettle while the oven is on and the washing machine's operating is over this limit, in practice the loads aren't continuous and it's rare to find household "spikes" of power that go above what the battery + solar (during the day) can provide, so our Grid usage during daylight hours has been minimal.
The exception to this is shown in the 2nd screenshot, and that's when the car is charging. It's got a 32Amp (7kW) connection and it usually fires up outside peak solar hours (it's not been used much recently so it's mostly idle, but it does like to top-up by 1-2kWh every couple of days at least). So we do draw from the grid but the battery provides the bulk of the power for those small charges. One limitation at the minute is that there's no way of setting a priority for powerdraw - if I'm doing a big car charge I'd rather it all came from the grid and the Powerwall's energy was reserved for home use, but there's no way of telling the Gateway that. I suppose economically it makes no difference in our case but, hey, it'd be a nice to have.
SolarEdge Monitoring
An unexpected but welcome bonus was that the SolarEdge inverter has it's own monitoring software, and our installer hooked us up with an account there. Whilst it only tells you about the generation side (up to and including the Inverter), it does have a degree more detail than the Tesla app about that. I quickly became addicted to the diagrammatic view which shows the performance of each panel over time. Here's what they've generated since install:
 |
| Fig 5. SolarEdge Energy Production in diagram form. Green box is the Inverter (and total), each panel's production is shown in it's relative position on the roof. 1.1.9 currently in the lead but it's the last to be shadowed as the sun drops in the West. |
Ultimately, the SolarEdge stuff is less important for us - it'd be great if we were actually generating or exporting power, and it's valuable for giving feedback on how the system is performing (and I guess would be a good warning indicator of any panel issues, if their relative production started to drop off). But who doesn't love a good detailed statistic, eh?
Connectivity
Worth pointing out that both the SolarEdge and Tesla Gateway 2 need internet connectivity to show the data I've put above. The SolarEdge gateway has WiFi, Ethernet and (apparently) ZigBee connectivity options. The Tesla system is so dependent on connectivity it not only comes with WiFi and Ethernet but is also configured with a 4G SIM card for independent connectivity and monitoring. Shame we live in an area of poor signal, but it does work (intermittently).
The Energize guys configured both to connect via our home WiFi network using a Powerline extender to the Garage. Subsequently I found that to be both redundant (I have a mesh WiFi system with good coverage there) and unreliable, so I reconfigured both to directly connect to the home network. That's worked great ever since although I noted the Tesla Gateway can throw a hissy fit when things change near it; hasn't happened a while but I did lose a couple of days of data early on before I discovered this propensity.
The APIs
For me, the most interesting thing about the install is that there's both a remote and a local Gateway API. For the remote API one can query Tesla directly using one's account details and get configuration, health and performance data from the whole system. Of more immediate utility, though, the Gateway itself has a local REST endpoint that exposes much of the same data to local network sources...
So clearly I've been heavily using that and have a full time monitoring solution built using it to cover some of the (perceived) gaps in the Tesla App. To be fair, there's no way Tesla themselves could be expected to encapsulate the vast array of power billing mechanisms in use across the world - some of them so fiendishly complex it seems even the utility themselves don't understand them, so this isn't particularly a criticism just an area I wanted to address for my own needs. Having said that, it does somewhat erk me that there's no easy way to see what the battery is doing over time or to cross-reference that with the state of power use and consumption. So I wrote my own dashboard to provide all of that info instead:
 |
| Fig 6. Self-penned Dynamic Power Dashboard, bringing all the data into one page |
(If you look closely at the "Power Today" chart you can see the car's had itself a little top-off already). I'm particularly pleased with the little status update top-left giving an estimate of when the battery will be full or empty. It's not clever - it just plots a straight line over the last 10 minutes - but it's surprisingly useful as an indicator.
At some point in the proceedings I realised that the rig up on the roof also effectively acts as it's own weather station - the amount of power being generated giving an indication of how cloudy it is, and the times of day when generating starts and finishes providing a proxy for the day length. Clearly, this is also graphable and dashboardable:
 |
| Fig 7. Rapidly disappearing up my own fundament, I present to you "Solar Array as Weather Station, The Dashboard" |
The Economics
First major point to make is that we're about 6 weeks in to this setup, and that's over the hottest and sunniest days of the year. This is clearly not a good time to be assessing the economic worth of the system; I need at least a full year's worth of data before I can begin to do this sensibly.
Second point is that I'm not going to give absolute costs. One, to preserve a fig-leaf of privacy and Two, your costs will differ, possibly greatly, in hardware, fitting and (most importantly) utility pricing. Do your own research before you make a decision.
With that out of the way, I'll make the following observations:
- We get billed monthly for electricity. The last-but-one period was almost perfectly half before and half after the install, and as a result was almost exactly 50% of previous bills for the months before that
- I've just submitted a reading for this month's bill. We drew a total of 3 kWh from the grid in that period. This is going to make our bill approximately 0.5% of it's normal monthly value.
- (It's worth noting that I've been deliberately charging the car and running the immersion heater to avoid having to report negative numbers on the meter; I was aiming for zero but overshot yesterday as we did loads of washing and used the tumble-drier too)
- The Battery is working well for us in the job it's supposed to do - run off the grid as much as possible. To a first approximation, the Solar Panels are generating enough during the day to both supply the house and charge the battery sufficiently to last overnight.
- Actually the battery's proven it's big enough that, fully charged, it'll get us through an entire day of grotty weather. We only start drawing from the grid on the 2nd consecutive day of bad weather.
- Having the Battery act as a "soak" for our solar power means we're not exporting much excess. We seem to need 2 good days in a row to both charge the battery (...and top-off after overnight draws) and have enough left over to export. And we only export in the afternoon despite generating most of our power in the morning.
- If we were getting paid more to export, this would be a problem. At the moment, in this weather, it indicates we're well balanced between solar and battery.
- Conversely, the rate we're now on we get the same for exporting power as we pay to import. Strictly speaking, this makes the entire Powerwall system redundant and valueless (as every kWh it "saves" from the grid we could export and import at net zero cost anyway)
- According to the Tesla App, we've had two short "backup" events in the 6 weeks the system's been alive. These are times when the Gateway's detected the mains power has gone (or is about to go) out and has switched to entirely self-supported power. The first time we didn't notice anything at all. The second time some of the most sensitive electronics in the house rebooted which was noticeable.
- It's unlikely these were actual grid power outages - the Gateway's monitoring capabilities are necessarily paranoid to avoid brown-out scenarios, so it's more likely to go isolated at the slightest under-performance of the grid which nothing in the house would have noticed anyway.
- That worthless pointless weather-station dashboard above tells me we're already past peak power days for the year, with a long decline to winter ahead of us. We've had 4 continuous days of hot weather, yesterday was virtually cloudless, and we're down 20% on early-July's peak generation.
With all of the above noted, the most important figure economically is the time to break even on the install costs. Noting that we're a short time into ownership and that this is the best time of year for such a system, a simple calculation on monthly bill savings vs cost to install gives me a return on investment in a shade over 20 years. I expect once I can average this across a whole year it'll go up probably to 30 or so years.
Over that sort of period, there's too much variability in all factors - system lifespan, improvements in panel efficiencies, battery degradation and the changes in the utility market place - to have any confidence that the number is even meaningful let alone accurate. But I think it's clear that this is not an exercise that the thrifty and value-conscious should prioritise.
I'll note that the battery system roughly doubled the cost of the installation, and as noted under the deal we've got it's actually adding zero economic value. So you could eliminate it and reduce time to break even by a half too. Having said that, we only got the utility deal because we've got the battery and current starting-from-scratch tariffs (post FITS-era) aren't normally anywhere near that generous as balanced import/export, so whether that would be practical in the real world is anyone's guess.
Conclusions
Realistically: Too soon to say, but the omens aren't looking good for this being a great economic decision at the current time.
Having said that, there are human and other non-economic factors in play that aren't accounted for by the account given so far:
- I am far, far more aware of how much power we use around the house doing different things, and (by shouted osmosis) so is the rest of the family. This has already had the desired effect of making people think whether it's a good idea to leave lights on, or the TV, or whether that half-load of washing can wait. So we're slowly becoming a little bit more environmentally aware
- The shear Smug sense of satisfaction one gets when we go an entire day without drawing a joule of energy from the grid without having to make any sacrifices in lifestyle is impossible to put a monetary value on, but should not be discounted.
- I am desperately trying to source a high-quality sticker I feel good about attaching to the car that says "Solar Powered - I'm Driving on Sunlight, not Dead Dinosaurs" or similar. I have achieved what proponents of electric vehicles have pointed out as one of the hidden benefits: My car emits less CO2 than when I got it, not more.
- This system can only get smarter as we go. At the minute I'm manually switching the immersion heater on and off to "soak up" excess power and hopefully make an impact on our gas bill too; I'm well aware there are automated systems available to do this which I'm looking at, and the potential to make better use of smart switches throughout the house based on draw and capacity surely has some benefit.
So: I'm very pleased and satisfied with the system we've got. Ideally I'd like more panels because MOAR POWER but unless and until we move house (....which we're not) that's not practical. Economically it's probably a white elephant but that's really not why we got the system in the first place. Insert your own Green Manifesto statements here if you like, I'm going to stick with the basics. We did this because we could, and because doing it feels like it's better for the wider environment than not doing it, over the lifetime of the system. You will come to different conclusions.
