Solar panels and battery - any real world reccomendations?

SoliD said:
Yeah 28 kwh here yesterday. Obviously not at the peaks of the summer but still hitting 4.7kwh at midday peak. One for those who've had it a few years what's the drop off on a clear day in winter? What sort of potential max would you see at the middle of the day in winter with no cloud cover?
Click to expand...

If the middle of the summer temps mean the panels lose some efficiency, I wonder if there is a cooler time of year for them when they operate better? Seen some material that mentions 25c so for example a day in April when its cooler but cloudless might produce more energy.
 
20.4 with my 4kw of panels yesterday.

May, June and July are my best performing months, although the last couple of years April's been good to. The other thing you need to consider is the 21 of June is the longest day, so by the time you get to August the days are quite a bit shorter.
 
Not bad for west of Scotland, 23kw for me yesterday so cant complain (got a pretty large array). Ive found Sep to be not too bad in the past, just a downhill till literally nothing in Dec/Jan
 
Last edited:
gpuerrilla said:
If the middle of the summer temps mean the panels lose some efficiency, I wonder if there is a cooler time of year for them when they operate better? Seen some material that mentions 25c so for example a day in April when its cooler but cloudless might produce more energy.
Click to expand...
I think it was back end of August when it was a bit cooler with northern winds but clear skies I had my best peak generation.
 
We had a cold but very sunny April last year, best solar month ever 570kWh from a 3.5kw system.

April/May, usually very good solar months as they are cooler.
 
Last edited:
Journey said:
Hoping for 1.21 jigga Watts next April xD
Click to expand...

I wish :p

I'm hoping to get the **** inverter working on the other house.... 2 months lost
Having to import one that will cover the high single string voltage, earliest in the country 21st Sept. (another Fronius)

On my house have run ducting and 2 strings of 6mm solar cable up the garden for future use. (4X20meters)

8 new EVE LF280K are 28 days into their trip to me, not long now, I hope.
 
Last edited:
Just worked out monthly averages, using January 2016 through to August 2022 for my 4kW array.

January​
February​
March​
April​
May​
June​
July​
August​
September​
October​
November​
December​
128.65​
206.87​
364.07​
523.55​
589.41​
579.52​
586.82​
513.86​
404.33​
249.45​
149.89​
103.07​
 
Journey said:
As I have said previously I do my best to predict forward based on knowledge of my use, but I only have a single 8.2kWh battery installed as of now. Usage is much lower for me usually sub 10-12kWh a day, so even on a terrible solar day I can almost guarantee that the battery will cover all my needs unless I am charging the car, and that is done in the Go Faster window of 21:30 to 02:30.

Have you worked out your usage patterns yet, using your graphs, what time of day are you likely to use lots of power and when are you not? Even though I don't need to I still do heavy draw stuff like washing and tumble drying/immersion use when after 21:30 when I know it will be a terrible day the following day, and make sure the battery is fully charged to run the house for the 02:30 to 21:30 the next day. I know you can't do that due to your power use, but you can work it out, but it is all about data gathering, and you are unique so can't be compared to others.



Seasonal really, as if you check sunrise/sunset time, and predict generation you'll get a good idea of what state the batteries will be in once you've used the data gathered above. Sadly a couple of weeks of solar data isn't enough, but I know you are itching trying to get it all setup perfect. I suppose to some extent I don't have automation really, as I have a habit of checking at least once a day the weather, and then making sure I'll have the battery where I want it for the following day. Some people can't be bothered, but the system will pay back faster (it's being sold anyway) and I feel better knowing I did my best to lessen the load on the grid at peak times. :)
Click to expand...

Great tips kind sir as always - yes, I'm getting there with adjusting.

Solcast was anticipating a 30kw day today, but looking much more cloudy, so may end up paying a bit for 'grid electricity ' as I set the battery to only charge to 50% last night. Lots of learning to 'tune' at the moment!
 
Just opportunity cost, its what risk you take to save a few quid over the month compared to paying full whack and working out if the risk is worth it for you.
 
Now if suppliers offered some fresh tariffs to encourage off-peak use and heavily discounted (rather than the archaic economy 7) then riding through the next couple of years would be straight forward. We can only hope the export price gets looked at too which is another archaic price structure out of touch with everything else.
 
katie279 said:
Great tips kind sir as always - yes, I'm getting there with adjusting.

Solcast was anticipating a 30kw day today, but looking much more cloudy, so may end up paying a bit for 'grid electricity ' as I set the battery to only charge to 50% last night. Lots of learning to 'tune' at the moment!
Click to expand...

The game within the game, I was talking to MKW about this very same thing recently, especially as I'm not getting paid anything at all for export right now.

Currently I am just checking the Solcast forecast data manually each night, for the next day, and adjusting the charge value myself which is working OK.

Longer term it would be nice to do something like https://community.hubitat.com/t/beta-givenergy-forecast-charge/90532 or https://www.speaktothegeek.co.uk/20...home-assistant-and-solar-forecast-automation/. Query the API for Solcast for the next day, adjust the battery SOC accordingly.

One point to bear in mind, that if you get it wrong, you wind up paying a lot more for daytime units than over-charge on night ones, a factor of about 5x the cost per unit. It's better to over-charge by 4 units than to draw 1 from the grid at peak.

Once SEG payments kick in, you'll be getting 4.1p/kwh roughly, if they don't change. New units on the cheaper tariff cost 7.5 or 8.25p/kwh depending on Go or Go Faster, so at worst, if you pay 8.25p/kwh overnight, you lose 4p/kwh for over-charging and export.

I'd err on the side of caution even if automating a lot of this, your own usage/drain stats will come into play as well, how much charge you need minimum to reach the next solar day-shift for example comfortably. I expect you'll want a minimum value, which if current SOC is higher won't get reached anyway.
 
Last edited:
Great tips HH - yes, agree I have Solcast set up with home assistant - so was thinking the same thing (ie automate the % based on predictions). Currently it under predicts Kwh by about 20% which is good news in a way, but this is only based on a couple of weeks data, so don't know if this'll shift through seasons.

I'm not aware, but it'd be good if these weather forecasts had some sort of 'confidence' rating - ie how certain are they that it's going to be a sunny day etc as that's the missing piece for an automation (ie I can be more aggressive with charge % the more sure they are)

As you say, we pay far higher if we undercharge, so I'm trying to work out the right balance of target %.

It's a fun game though - I think I'll closely monitor through winter as that'll be the most 'risky' and that'll give me a good sense for the rest of the year.

The other dimension is the capacity of batteries increases the margin of error - ie adding more batteries would allow an occasional 'error', so I could look at an extra battery or two.... ;)
 
Last edited:
katie279 said:
Great tips HH - yes, agree I have Solcast set up with home assistant - so was thinking the same thing (ie automate the % based on predictions). Currently it under predicts Kwh by about 20% which is good news in a way, but this is only based on a couple of weeks data, so don't know if this'll shift through seasons.

I'm not aware, but it'd be good if these weather forecasts had some sort of 'confidence' rating - ie how certain are they that it's going to be a sunny day etc as that's the missing piece for an automation (ie I can be more aggressive with charge % the more sure they are)

As you say, we pay far higher if we undercharge, so I'm trying to work out the right balance of target %.

It's a fun game though - I think I'll closely monitor through winter as that'll be the most 'risky' and that'll give me a good sense for the rest of the year.

The other dimension is the capacity of batteries increases the margin of error - ie adding more batteries would allow an occasional 'error', so I could look at an extra battery or two.... ;)
Click to expand...

The 1st link was talking about confidence ratings in one of the posts, at least from what it seemed to be saying, https://community.hubitat.com/t/beta-givenergy-forecast-charge/90532/3.

The problem with additional batteries is the benefit drops off the more you add. All depending on your figures but I know if I added a 2nd one it would be nowhere near as beneficial as my 1st one was.

If you have got enough battery to almost never charge from grid in the summer months, then that additional battery only really gets any use in the darker 3-4 months of the year, so your cost for that is set against the cost of either spamming import to 100% charge on existing at cheaper rates (honestly in December/Jan at least I'd likely just leave the battery to charge 100% always) or eat a few units at the regular price when needed.
 
My annual usage is 9500kWh if I get an electric car in the future it will be more.
I want to fit as many PV modules as possible. So i want 14 x 400w panels on my south roof and 14 x400w on the west roof. I can see only solar edge do a 10kw 1ph inverter. I see Goodwe also do a 10kw 1ph inverter but dont see its marketed in the UK yet.
Question is, is it possible to put a 5kw inverter on each array? can they share a battery system (10 to 20kw) or do they need separate batteries? or do they need some sort of synchronising on their outputs at the AC end if they each have their own battery, when they combine AC outputs for House/Grid? Of course I understand that its possible at some time at peak production there may be export power to grid when the batteries are full, and understand that this will be limited to a cap of 3.68kw. How is that achieved with 2 inverters?
I want to use all of my solar and also will fit a power diverter to immersion etc.
Anyone been in this position or has any input?
 
@Kot You won't necessarily be limited to 3.68kw, I recently got approved for 13kw, it all depends on your local grid.

You might want to look into Victron, there are other options of course, but that's just what I'm familiar with. I already have 4kw Solaredge system and am adding a Victron inverter, batteries and another 5kw array. The Victron will use the excess from the SE system to charge the batteries, as well as the new array.

You can use more than one inverter but it gets a lot more complicated especially when adding in batteries. With Victron you'd only need one inverter, and use separate MPPT solar charge controllers which connect to the batteries, but Victron is not cheap (mind you nothing is these days!).
 
Kot said:
Question is, is it possible to put a 5kw inverter on each array?
Click to expand...
Yes.
Kot said:
can they share a battery system (10 to 20kw)
Click to expand...
Yes again, but it would require an AC coupled battery.

Fronius do both a 10KW and 15kW single phase inverter, they are pretty hard to get a hold of right now though.
Kot said:
at some time at peak production there may be export power to grid when the batteries are full, and understand that this will be limited to a cap of 3.68kw
Click to expand...
This is incorrect, unless you only opt for G98 notification, instead of G99 via the DNO which is determined by the DNO based on your local circuit/capacity.
Kot said:
I want to use all of my solar and also will fit a power diverter to immersion etc.
Click to expand...
What is your budget?
As Ron-ski said, I'd be looking at a 48v Victron setup, with dual MPPT's (4 PV strings) into a larger inverter (Multiplus II 10000VA) , with DC coupled 48v batteries and a Cerbo GX device to control it all.
 
Has anyone set up a seperate or spare tablet/ipad to monitor the solar generation and export were their washing machine/kitchen is etc etc….so you can easily see whats being generated before switching things on???
 
Journey said:
As Ron-ski said, I'd be looking at a 48v Victron setup, with dual MPPT's (4 PV strings) into a larger inverter (Multiplus II 10000VA) , with DC coupled 48v batteries and a Cerbo GX device to control it all.
Click to expand...

@Kot You can't use the Multiplus II 10000 grid tied yet - rather ennoyinly its not yet been G99 certified, would need to be the Quattro 48/10000. Victron will get it tested along with the Quattro II, but no time scale when I asked recently.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom