PV on the roof - What did yours cost?

[Joe.B]

Calling all those with PV on the roof!

I’ve been quoted £11K for a PV installation and I’m now trying to gauge whether that is a fair price.  I’d be interested in hearing from other YACF’s with PV about how much their systems cost.

A bit of background:

My roof is in 3 sections, facing East, West & North. Obviously the North is out of the picture so we’re looking at an E-W array, there is also a bit of shading from two chimney stacks.

Before receiving the proposal from the installer I made a few stipulations about the system:
·         To minimise shading losses I stipulated that I wanted panels with embedded bypass diodes.
·         Panels to be wired in parallel rather than series to minimise shading losses, (installers usually default to series as it uses significantly less wiring).
·         I asked for a Multi Power Point Tracking (MPPT) inverter/controller rather than a cheaper PWM device.
·         I want LiFePO battery storage rather than Li-Ion.

Our household consumption for March this year was rather typical for us:
•             Car Charging = 225KWh
•             Hot Tub = 275KWh
•             Rest of House = 310KWh

Given that we’ll be using the PV to charge the home battery, heat the immersion and run the hot tub, I don’t envisage having any surplus to sell back on the grid. (I’m currently building a diy solar thermal system for the hot tub which I’ll post about once complete).

They have come back to me with a proposal for a 10 x 380W (split 50-50 E/W), installation with a 5.8KWh battery. Cost £11K.

I’ve looked up the specified panels, inverter & battery on sparky trade sites and these items can be bought for £4500. Allowing another £1000 for other required components, we’ll assume £5.5K on parts.

That leaves £5.5K for survey’s, scaffolding, installation and commissioning.  This is the bit that I’m not sure about so I’d be grateful to hear what others here have paid.

 

 

[FifeingEejit]

I've been quoted just over £5.4k for a 5kw array and £4.8k for a 4kw array
And £4400 for a 5.2kw Givenergy battery system

This was before the VAT change

I really need to get back on the case.

[Diver300]

·         To minimise shading losses I stipulated that I wanted panels with embedded bypass diodes.
·         Panels to be wired in parallel rather than series to minimise shading losses, (installers usually default to series as it uses significantly less wiring).
·         I asked for a Multi Power Point Tracking (MPPT) inverter/controller rather than a cheaper PWM device.

Our household consumption for March this year was rather typical for us:
•             Car Charging = 225KWh
•             Hot Tub = 275KWh
•             Rest of House = 310KWh

With panels in parallel, the panels will be kept at the same voltage. If a panel  is shaded, that will cause its voltage to drop, and it won't supply any current, which makes the bypass diodes pointless. The bypass diodes are useful when a panel will still be delivering current with some parts of it shaded.

I think that dealing with shading is best done with separate inverters for each panel. Dave from the EEVblog (https://www.eevblog.com/) had some issues with shading and his latest installation has micro inverters on each panel IIRC.

MPPT stands for Maximum Power Point Tracking. It continually adjusts the panel voltage to get the maximum power.

Have you got a heat pump for the hot tub?

[Canardly]

Just accepted a local authority auction quote of £6.4 k for a 14 panel 4.7kw array. Additional costs will include a diverter for the immersion heater of £450. Battery is unnafordable atm.
Edit. Also extra for the pidgeon netting.

[fd3]

Just accepted a local authority auction quote of £6.4 k ...

How does that work?

[Joe.B]

Thanks for the reply's folks.  On the face of it my quote seems a bit on the expensive side, although given that I have specified particular preferences regarding panels/inverter/battery and the fact that I will require scaffolding on both sides of the house perhaps I should expect it to be a bit on the high side.

I did think about specifiing micro-inverters on each panel but in the end decided that they presented a bit of a challenge in the event of any repair/replacement that might be required.  Of course it could be argued that a single inverter results in a single point of failure.

When I finally gave in to Mrs B and ordered a hot tub I talked to the supplier about a heat pump but decided that it wasn't worth the extra cost. That was back when we were paying 15p/KWh and I do rather wish we'd paid extra for the heat pump now but I should soon have a solar thermal panel plumbed into the tub which should help reduce costs.

[Wombat]

When considering the cost of the parts versus total installed cost, don't underestimate the cost of the panel mounting hardware.  In discussion with my installer, it became evident that this is a large part of the cost, if a decent system is used.  The design of this also needs to be calculated and verified.

My own 4kWp system was £5700 in 2018, including the solar iBoost unit, to power the immersion heater from surplus PV generation.  It uses Canadian Solar 335w panels, and a Solax X1 twin mppt inverter.

[Canardly]

Just accepted a local authority auction quote of £6.4 k ...

How does that work?

Combine of several local authorities runs an auction with pre vetted installers. Households register interest and receive an individual proposal which they then have a month to think about. A small deposit then follows and all is subject to detailed survey. Price actually seems to be about current average, although a 30% discount is  being claimed. YMMV.

https://www.cambridgeshire.gov.uk/residents/climate-change-energy-and-environment/how-you-can-take-action/home-energy/solar-together-cambridgeshire

Well this is all beginning to turn into a bit of a disaster. This company does not communicate, return emails, use the telephone call back facility and is not following the procedures set out in its marketing bumpf, presumably agreed as part of the Local Authority procurement process.
Scaffolding is supposed to be erected up to seven days before installation and written confirmation provided that the agreed install date is still good up to four days before due installation date.
My installation was due on the 5th August, agreed when the major financial deposit was taken. No communications has been received since then.  No scaffold has been erected. Materials arrived  on the early evg of 3rd of Aug with no notice.  The 'call' centre is extremely difficult to contact (other than sales, surprise surprise) spent 20 mins on their arrange installation option yesterday during which time I went from 4th in the Queue to 3rd then gave up.
This bunch also seem to have been awarded another contract in Devon.

The final installation price with extras for what is basically a 4kw array is £5400 ish including some extras such as pidgeon netting and a diverter but no battery. I would sooner pay a bit more for a decent service. Does not bode well at all for after sales. Screams lack of resources or possible financial difficulty to me, having worked in the construction industry for 40 years.

Oh dear I have just read these, wish I had done so beforehand. May be time to seek an alternative. So much for having Local Authority sponsors.  Now would I have good title to the delivered kit I wonder?

https://uk.trustpilot.com/review/www.solartogether.co.uk?page=3


Edit.

In accordance with the the terms and cons have cancelled this. But as expected no acnowledgement. I have received an invoice for an installation I have not had. You could not make this up! There is slapdash and there is incompetent. ISO 9001 oh really!.  When chased for the payment I actually received a direct line to a human being. Told her that installation has not taken place. That was last monday. Nothing since.


A quick google produces a report from the London Mayors office highlighting 385 complaints being made about this company, very much in line with my own and other's experience. And yet, they are still being awarded further concessions. Something is very not right here.

https://www.london.gov.uk/questions/2022/2589

[fruitcake]

It's legitimate to ask the company why it will cost that much and to ask whether their price is negotiable. 

Something that you might consider as a preliminary stage is dropping any chimney stacks no longer in use and making good. The chimney brickwork would terminate below roof tiles, with the rafters/tiles then augmented to cover the gap. I've had quotes of £600 for that job.

The rationale for dropping the chimneys is that future maintenance on a chimney stack would become more expensive where the contractor needs to work around a PV installation. And your PV cells would receive more light.

[AllyCat]

Hi,

Rather a new interloper here as I'm not a serious cyclist, but I came across this thread (via another excellent one) and perhaps can contribute.  Sorry if after nearly 4 months I'm too late, so I'll keep this moderately short (for me).  First my credentials are that we had an 8 panel system installed last November, also via a Solar Together "Auction" (Surrey), which cost around £4k for the installed panels, etc. and +£2k for a LiFePO4 battery (3kwh) with upgrade to a Hybrid Inverter, etc.  That seemed a reasonable price but I'm not sure it would have been at +30%.  We had planned for 10 panels (which didn't change the cost much) but the shape of our otherwise "optimum" roof (South-facing at 37 degrees slope) couldn't accept more than 8.

Therefore, the quote of £11k seems high for what appears to have been offered.  However, IMHO your specification/stipulations seem "wrong".  Embedded diodes won't do anything in a Parallel-connected panel arrangement, but that seems inappropriate for an on-grid/domestic system.  Parallel-wired would give around 100 Amps at 35 volts (compared with 10 Amps at 350 volts for Series), which will be much less efficient (and probably unreliable) in generating the Inverter's internal ~400 volt bus.  An inverter of over 3 kW will probably have dual MPPT inputs, ideal for your separate E/W panels.

Whilst I appreciate your concern about locating "optimisers" directly at the panels (I felt the same), a diode effectively "writes off" any power from an affected panel, whilst the optimiser converts it to a useful voltage.  We were quoted £50 for individual panel optimisers or £30 each for the whole roof; since we have 1-2 panels occasionally shaded by a chimney stack, I specified just one optimiser for the more affected panel as an "experiment" and the overall power does seem significantly reduced when the shadow strikes the other panel.  I believe most domestic Inverters are MPPT now and LiFePO4 are (or were) actually cheaper (and better) than Li-Ion, but their lower energy density makes them less ideal for EVs.  Sadly the price of LiFePO4 battery packs seems to have almost doubled in the last few months, and personally I have a rather reactionary view to consider Lead Acid batteries if the Inverter can be located in a garage, but I won't expound on this at the moment.

My main reason for contributing is that your electricity consumption at 810 kw/month is unusually high; to generate that in March on a South-facing roof is likely to need more than 20 panels!  I haven't calculated how much worse would be E+W facing panels, but remember that if the elevation of the sun is lower than the slope of the roof (e.g. throughout the winter in UK), then the sun will be "behind" one of the sets of panels.  PV panels need direct sunshine and not at too much of a "glancing" angle; it's true that they do deliver "some" output under cloudy skies, but in the same way that you may make some progress to your destination if you carry your punctured bike on your back, compared with riding it? Typically, in a UK winter, the panels will deliver about 25% of their daily summer output and last December our system generated only 10% of their summer output (just 5 kWh per panel in the whole month).  Personally, I've found the following PV System Calculator very comprehensive and apparently quite accurate (seems to include localised Cloud Cover data):

https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html

There's more I could say, but this must be enough for now. 

Cheers,  Alan.

[Canardly]

Just cancelled the order for the PV panels (Statutory cooling period of only 14 days which commences when materials are delivered but before installation) and will go with one of the main energy providers instead I think. If this involves a significant wait so be it. Have been in touch with the LA and it seems I am not alone in experiencing problems with this particular firm. Shame, its such a good idea.

[orraloon]

Can anyone point me towards a PV panels on the roof basics introduction?

Recently moved in to my (final?) home, work to do, fine with that, south facing bungalow at top of a slope so sun shining bright on me just now.

Thinking about solar panels.  But have no knowledge on subject.  Would like to understand the basic principles before contacting potential providers, so I at least understand what they're talking about.

[SoreTween]

Post 1 of er, many.
I can help you here but it's best to explain things with how technology has evolved.

Back in the dark ages of solar PV, perhaps as much as 20 years ago  solar PV was wired like this:


This is a bit of an exaggeration as under load a PV panel produces 34 ish or 39 ish volts.  So the 'string' 6 panels under load would produce 207 to 234 volts.  But.  If your One Big Ass Inverter is not consuming what the panels are potentially producing the panels get grumpy & frustrated and their output voltage rises.  Think of it as pressure in a pipe.  Each panel is trying to pump water, if each one can shuffle their water load on to the next panel they are all happy.  But if there is a blockage the panels get angry and try to force their water through, each panel pushing hard as it can against the next.  If the inverter doesn't take the load the back pressure becomes interesting.

You could have one hell of a big pile of voles at the input terminals to the OBAI. 600V DC as the picture shows will ruin your day if you come into contact with it.

Despite the big hairy ass voltages in your roof this was an ok system.  Simple to wire as the panels very quickly gained a simple, standardised,  good quality connector.  Plug one to the next to the next & the installers were happy & the consumers ignorant.
There was a down side though.  Shade.  If a fat bar steward pigeon sat on one panel shading some of those little squares (more than one but less than half a panel) the output of the entire string dropped.  If one panel fell to 50% so did all the other panels in the string.  Imagine the fat pigeon has it's knarly, deformed talons crimped around the cable of that one panel crushing the cable for all it is worth.  Because the vole flow for that one cable is restricted so is the vole flow for all the other panels in the string.  In Canada, where the bears live, a big maple leaf landing on one panel could knock an entire string down to 50%.  So few chestnut leaves or the shade of a chimney on a UK panel.

Something had to be done.  So it was.

To be continued...

[Mrs Pingu]

This is interesting, I await the next instalment

[SoreTween]

Part deux of many.

Pretty much every component in the string inverter system above is a single point of failure.  Whatever breaks or fails, the whole system output is hit.  So what if instead one one big ass system you had lots of independent small ones? 
Enter the micro-inverter:


Each micro-inverter converts the DC from one panel into 240VAC.  Wiring is dead simple, the same standardised connectors are used on the DC side and the multi-drop cable on the AC side comes pre-made. 

The micro-inverters live on your roof under the panels:


If a leaf falls on one panel only that panel is affected versus losing output from your whole string:


The down side is that there is more to go wrong.  But really, who cares?  If a panel fails in a string setup your whole system output is zero until you can get the panel replaced.  If your string inverter fails zero output again.  With microinverters no one component can bring down the whole system, you might be down one panels worth of output for a while but the rest keep doing their thing.  No single point of failure - almost, bah :-(  I'll get to that in part 3.

[Diver300]

This is a bit of an exaggeration as under load a PV panel produces 60 ish or 72 ish volts.  So the 'string' 6 panels under load would produce 360 to 432 volts.  But.  If your One Big Ass Inverter is not consuming what the panels are potentially producing the panels get grumpy & frustrated and their output voltage rises.

Interesting thread.

Minor point of order here. PV panels will self-limit their voltage, and that voltage is only 20 - 30 % higher than the voltage at which they produce the maximum power. So if the inverter isn't taking current, there is a larger voltage, but it's not massively larger. 6 panels in series will produce enough voltage to kill you, whether they are running normally or no load is being taken.

The shading issue is a real problem.

[orraloon]

Thanks ST.  When's the next lesson? 😊

[rogerzilla]

PV cells sound a bit like bike dynamos, also trying to push a constant current.

[SoreTween]

Part C.

There is one more component required in a micro-inverter system not shown above, a controller.  The controller does two things, first it tells the micro-inverters what to do, what voltage and frequency to output and so on.  Second it interrogates the micro-inverters to see how much power they are producing and report that on to the cloud or a local PC.  A controller is required by law but it wasn't always so.  Once you could have a system installed & the installer would use a temporary controller to program up the micro-inverters and then be on his way.  The only reporting your owner got was from a meter measuring the system output.  Now all systems must have an on site controller fitted at all times.  This is good for the customer because they can get panel level monitoring.  It also means the profile of the micro-inverters can be adjusted quickly, the electricity company can ask the manufacturer to push out a minor adjustment to better match the grid profile.  When I installed my small system I downloaded the list of profiles and there were hundreds of them, far more than just the variants of 220, 230, 240 & 110VAC in combination with 50 or 60 Hz.  The bad news is that micro-inverters must shut down if they do not hear from a controller, I have no idea how long they will continue to produce if your controller goes silent.



Some controllers communicate wirelessly with the microinverters, others piggy back the control over the top of the mains.  Whatever, the controller is a single point of failure in an otherwise very resilient arrangement.  The good news is it will be fitted somewhere easily accessible, near your meter probably.  10 minute job to replace if that single point of failure does so in the middle of winter!

Aside:
I've just realised I've mixed up one thing.  I incorrectly recalled the significance of the numbers 60 or 72 as being the output voltage, that is wrong.  Panels usually have 60 or 72 cells in them, other variants exist too such as 54 or 96 but 60 cell & 72 cell are the most common.  It is not a simple one cell to one volt relationship.  I'll look it up the correct typical output voltages and make appropriate corrections.  Sorry about that.

[Kim]

Doesn't an individual solar cell give you (when open circuit) the usual P-N junction voltage of 0.6V or so?

Panels are presumably made up of series-parallel strings, akin to large batteries.

[SoreTween]

Figures updated.  Refresh your browser if you still see 72V in the part deux image.

Doesn't an individual solar cell give you (when open circuit) the usual P-N junction voltage of 0.6V or so?

Yes (for engineering values of yes which is more accurately yes-ish).  A quick check of a few panel datasheets gives about .55 - .57 per cell under optimal load and .65 - .68 open circuit.

Panels are presumably made up of series-parallel strings, akin to large batteries.

Solid yes.  The most common panels are actually 120 or 144 cells made up of two parallel strings so 2x60 or 2x72.  That is why output drops to half in the one leaf scenario, one of the parallel strings is not producing.

[SoreTween]

Part iv.

In the two systems above microinverters have a huge advantage which meant string inverter manufacturers were in danger of losing their lunch.  Something needed to be done about one leaf or a bit of shade halving or even knocking out all production from a string.  So optimisers were invented.  These are basically bypass valves that allow power from unshaded panels to bypass a shaded or faulty panel.  They sit in the system like this:


The optimisers live on the roof under the panels just the same as the micro-inverters do.  They are generally more than just a bypass valve, they also provide panel level monitoring.  The sting inverter communicates with the optimisers via a signal piggy backed onto the high voltage DC loop.  So now string systems have most of the advantages of a micro-inverter system.  Shading only affects one panel, a faulty panel doesn't knock out the string, and you have panel level monitoring.  However, the optimisers are now single points of failure for the string instead.

So that's the three main types of system.  Dumb string systems still have their place as they are obviously the cheapest, solar parks & industrial roofs where shading isn't an issue.  Commercial installation don't need panel level monitoring, if a string is under performing they'll send a person to investigate.  Most domestic installations ought to be micro-inverter or string with optimiser, unless it really does have zero deciduous trees in the vicinity and zero shading year round.

Next I'll have explain a bit more about the life cycle of panels.

[Feanor]

Im reading this with interest.

Can you also touch on the issue where the inverters need an existing supply to sync to, and whether they can run standalone in the case of a power cut (Island mode?), thx..

[AllyCat]

Hi,

Another feature of the PV/Forward Silicon Diode voltage issue is that the ~600 mV has a negative temperature coefficient of ~2 mV/degree C.  That's why all Panels lose about 0.3% of their output voltage/power for each degree C rise in temperature.

No, the "surprising" answer is that most PV systems must "shut down" if a power cut occurs.  It's primarily a "safety" requirement to prevent any locally generated power being sent out to the "National Grid" where people may be attempting to repair the cables!  I don't know how or if the MicroInverters can handle this situation, but our "Hybrid" (combined PV and Battery) Inverter does have a separate "Un-Interruptable" output facility up to 3 kW (for one 13A socket).  Sadly, our installer didn't even offer to connect this up as an option, which I probably would have accepted, if the price had been reasonable.  However, it is only primarily intended to run a PC or Router, etc., it's not going to keep the Lights on, or the Pump and Gas valve running in a typical Central Heating system.     Incidentally, we did have a (very rare) power cut recently and not only did the PV system "stop", but because the Router and all the communications also failed, the (one hour) event is still almost "invisible" on our data logs.

Perhaps it's also worth mentioning that there are separate "ac" and "dc" parts of the system.  The PV panels themselves generate dc (typically 200 - 350 volts) and so does an (optional) battery (typically 48 volts) whilst the (National) "Grid" is ac.  ALL other ac power, e.g. from the Inverter(s), must be perfectly synchronised to the 50 Hz Grid, or there will be a very big bang.    Thus a fundamental difference between MicroInverters and Optimisers (or basic PV panels) is that the "downlead" from the roof will carry ac instead of dc.  I guess that optimisers are far cheaper than MicroInverters since they only need to handle ~50v dc compared with ~230v ac (rms, so ~700v peak-peak).  Similarly, I believe that some battery packs interface with ac rather than dc.

Even relatively small batteries can be a large added cost to the installation, so it's worth considering their benefits.  At best, the most you will be paid for any energy you "Export" (via the "Smart Export Guarantee") will be about 25% of what you have to pay to buy the same energy, and a recent quote I got from BG was nearer 5%.  Thus it makes sense to store energy to use overnight (or alternatively charge the battery from lower cost overnight energy after "dull" days).  Also, in winter, early morning, late evening or "overcast" weather, etc., the PV system may be only "ticking over" at perhaps 0.2 - 1 kW, so the battery can handle any "surge" load such as boiling a kettle (3kW), Microwave Oven, etc., without needing to pull in any expensive power from the Grid.  Sadly, there's nothing that a battery can do about the considerable over-generation in summer (e.g. the last 10 days) and severe under-generation through most of the winter, in UK.

Cheers,  Alan.

[MikeFromLFE]

On my relatively small panel installation (6 panels) the calculated payback time of a cheap battery storage system was around 20 years.
I've decided against,  but I suspect that the technology isn't quite mature for battery backup at prices households can afford.

Sent from my moto g(30) using Tapatalk