On-grid domestic battery storage

EVandPV

chamelion said:

Martyn1981 said:

chamelion said:

Exiled_Tyke said:

chamelion said:

Folks, would this be a 'fair' calculation to use to work out the per kwh unit cost of battery storage over it's lifetime? This completely ignores the cost of electricity and 'savings' in this regard (which would just add to the below cost, not subtract from it).
Price per kwh = install price / (Usable capacity * lifetime (days) * cycles / day) 

So, if I'm installing a 7.2kwh £3500 system that lasts 10 years where I cycle the battery once a day, my per unit price could be:
price per kwh = £3500/(6.48kwh * 3650 * 1) = 14.8p/kwh. 

This effectively tells me that each kwh of energy I'm storing and using from my battery costs me 14.8p based on the install price?
So, for battery storage to be effective, the price above needs to be BELOW the unit price you currently pay for electricity - unless you think you can stretch out the lifetime to >10 years or do multiple cycles per day? (although even then, there's a lifetime number of cycles - so 6000 would mean 16 years with one cycle a day, or 8 years with 2 - presumably with some degredation). 

I prefer the calculation performed against the number of cycles rather than number of days. Hopefully this is more reliable and realistic?  It also brings the per unit cost down although some of the return is so far into the future it may not mean too much in reality. 

I guess there are a million different ways to do the math - and none give me pretty figures where I'm thinking "ah f* it i'll just get it given i've been looking at this for years!"

With cycles:
Price per kwh = install price / (Usable capacity * cycles)
= £3500 / (6000 * 6.48) = 9p?

That's better I guess. Basically saying for the lifetime of the system, however many years that'll be, if your electricity is more expensive than 9p then you'll have a reachable ROI. 

Personally I need the time window to work out my ROI. Using SolarChaser's input that he's done 200 cycles in 16 months, that's 200 cycles in 487 days. So 6000 cycles would take 40 years. That tells me more over the 9p calculation...

You could also add in an average battery capacity based on the warranty. So, say the 6,000 cycles warranty promises an 80% capacity at the end, then it might be £3,500 / (6,000 * (6.48 * 90%)) = 10p.

But, whenever I did these calcs going back, I stressed over and over that I was calculating the cost based entirely on the warranty, which is unfair, but I was really scared that I might mislead anyone if I used a 'better' number of cycles. However I suspect(?) that most folk expect to get more than the warranted number of cycles, much more?

Another point I realised and made was that the economics improve significantly (with this method) if the warranted cycles are increased. So all factors remaining the same, if the manufacturer doubles the cycle warranty, which might happen over time if batts perform well, then the cost of using the batts halves.

I also (going back to the start) kinda assumed that batts would halve in price over 5yrs, which doesn't seem to be happening, perhaps due to the massive demand from the expanding automotive side. Damn!

Lol we can make this formula as complicated as need be
Do the LUX systems themselves need replacement? I know solar PV ones they expect you need to replace your inverter after 10 years - is that the case with the battery systems too? I'm sure batteries last longer than the hardware managing them?

Our pv inverter is 8 years old and still working as well as it did on day one.
It's is in the loft where it stays cool all year round which probably helps.
If/when it packs in, there are direct replacements for it on fleabay for as little as £150, albeit refurbished units.

Zarch

mickyduck55 said:

Zarch said:

chamelion said:

Zarch said:

Here's a graph from my first week of using the 4.8kWh of batteries alongside my Solar.

Through targeted cheap-slot overnight imports on the Agile tariff alongside filling via springtime PV I've managed to keep the average import price per unit down to 5.3p

I think i'd need at least a third battery to get me to being closer to self sufficient.

What is satisfying though, is that i've shifted my usage % of generation waaaay the other way.  So rather than 70% of going back to the grid, i'm storing and later using most of what I currently generate.  This will obviously dip in % terms as more sun comes into play as we head into summer.

You've seen a lot of my posts in this thread where i'm trying to justify the expense to myself and it always works out against me. But just to see graphs like your above, even if not financially the right choice for me, makes me want to just drop the 3 grand today and pick up a battery system

PMSL......... More graphs, more data, more geekery and to hell with the finances!! 

Good morning Zarch... how do you collect the data?

That above all comes from my emonPi / emonTX set up.

Clamp based monitoring, one on live, one on PV, one on battery. (then scope to add one more when we get an EV charger)

Then their emonCMS software has the above app that does all the graphing.
https://guide.openenergymonitor.org/applications/solar-pv/

Not a cheap solution by any means, but if you want data....... I don't think it can be beat!!

Exiled_Tyke

mickyduck55 said:

A little premature but at the moment my batteries stop supplying my home when they reach 10% remaining charge.  Will there be any benefit to the longevity of the battery condition to change this say 20 % in the summer when I will almost certainly NOT need the 9kW available at full charge overnight?

I wouldn't expect to be doing complete cycles in Summer because 1: generation period is longer so will start drawing from the battery later in the evening. 2 Less cooking! and of course 3. batter will get full almost every day.  So I doubt it will dip below 20% very often. And to answer you question 'Yes' this will add to the batter life.   

My battery is 6.3kW.  The sunny days are at the moment filling the battery to capacity and if I don't cook or iron or do anything else which draws a heavy load, then even now I have more than 20% left by the time it starts charging up again in the morning. 

joefizz

Thrugelmir said:

Solarchaser said:

Generally not i would say.
The batteries are expected to survive 6000 cycles, so 20 years anyway

I didn't realise that they would last that long. Presumably at some point the capacity of the battery to store starts to degenerate though. 

Pylontech warranty is 80% at 5 years 70% at 7 years based on normal usage with solar only (not charging from mains).

Whilst technically it might be possible to say the batteries would last 20 years, its highly likely they would be pretty unusable for normal home operation past 10-12 years.

Anyone thinking it will get cheaper in the future really isnt paying attention to the current economic crisis.

Of course my plan has always been to replace the cells in the batteries, not the batteries themselves. That will be a much more cost effective option and by the time it comes round to it will be reasonable with lots of online videos or small local businesses offering return of post replacements.

I would reasonably estimate paying the same again as the current prices in 10 years time as a rough ballpark, but as mentioned the inverter and all the cabling etc will probably be good to go for the lifetime or with small replacement/repair costs.

It is worthwhile following the fallout from the covid outbreak and the upcoming financial crisis to see what happens with offshoring and sourcing things at overseas or below cost of production etc.

Saw a mention of a solar panel tax being talked about during the week..... Interesting times!

joefizz

Started hitting 99% regularly this last week, doesnt look like its going to go to 100% so will see if it drops off any further this year.

On the other hand, changed electric suppliers with a 55 quid sign up bonus this week, over half my yearly electric bill! Result! (well if they dont refuse me ;-))

EVandPV

Was having a look at the Github page on controlling a Lux inverter and it got me wondering if it might be possible somehow to get it to control the amount of overnight grid charging based on the weather forecast for the following day ?? I believe the Powerwall already does this.
Been a long time since I did any coding so I'm pretty out of touch and wouldn't know where to start.
Could see it being useful particularly during the months when pv generation is not likely to completely charge the batteries.

Zarch

EVandPV said:

Was having a look at the Github page on controlling a Lux inverter and it got me wondering if it might be possible somehow to get it to control the amount of overnight grid charging based on the weather forecast for the following day ?? I believe the Powerwall already does this.
Been a long time since I did any coding so I'm pretty out of touch and wouldn't know where to start.
Could see it being using useful particularly during the months when pv generation is not likely to completely charge the batteries.

This is something i've already been talking about with Chris (the man behind the Lux code).  He's been trialling using Solcast to forecast day-ahead solar generation and using that as a guide how to much charge overnight.

All tricky logic though.  Similarly, through my Twitter account (@energystatsuk) i've been mulling the same subject with followers.  There is already work happening, but all very disparate.   People with Powerwalls doing, Chris with his Lux etc.

Needs everyone to come together and make something open-source that will plug-in to various energy tariffs, weather/solar predictions and battery APIs.

EVandPV

Zarch said:

EVandPV said:

Was having a look at the Github page on controlling a Lux inverter and it got me wondering if it might be possible somehow to get it to control the amount of overnight grid charging based on the weather forecast for the following day ?? I believe the Powerwall already does this.
Been a long time since I did any coding so I'm pretty out of touch and wouldn't know where to start.
Could see it being using useful particularly during the months when pv generation is not likely to completely charge the batteries.

This is something i've already been talking about with Chris (the man behind the Lux code).  He's been trialling using Solcast to forecast day-ahead solar generation and using that as a guide how to much charge overnight.

All tricky logic though.  Similarly, through my Twitter account (@energystatsuk) i've been mulling the same subject with followers.  There is already work happening, but all very disparate.   People with Powerwalls doing, Chris with his Lux etc.

Needs everyone to come together and make something open-source that will plug-in to various energy tariffs, weather/solar predictions and battery APIs.

Great stuff Zarch !
Could you keep us updated here please. 

mickyduck55

Zarch said:

EVandPV said:

Was having a look at the Github page on controlling a Lux inverter and it got me wondering if it might be possible somehow to get it to control the amount of overnight grid charging based on the weather forecast for the following day ?? I believe the Powerwall already does this.
Been a long time since I did any coding so I'm pretty out of touch and wouldn't know where to start.
Could see it being using useful particularly during the months when pv generation is not likely to completely charge the batteries.

This is something i've already been talking about with Chris (the man behind the Lux code).  He's been trialling using Solcast to forecast day-ahead solar generation and using that as a guide how to much charge overnight.

All tricky logic though.  Similarly, through my Twitter account (@energystatsuk) i've been mulling the same subject with followers.  There is already work happening, but all very disparate.   People with Powerwalls doing, Chris with his Lux etc.

Needs everyone to come together and make something open-source that will plug-in to various energy tariffs, weather/solar predictions and battery APIs.

If you need any early adopter or testers I would be very interested in taking part

EVandPV

mickyduck55 said:

Zarch said:

EVandPV said:

Was having a look at the Github page on controlling a Lux inverter and it got me wondering if it might be possible somehow to get it to control the amount of overnight grid charging based on the weather forecast for the following day ?? I believe the Powerwall already does this.
Been a long time since I did any coding so I'm pretty out of touch and wouldn't know where to start.
Could see it being using useful particularly during the months when pv generation is not likely to completely charge the batteries.

This is something i've already been talking about with Chris (the man behind the Lux code).  He's been trialling using Solcast to forecast day-ahead solar generation and using that as a guide how to much charge overnight.

All tricky logic though.  Similarly, through my Twitter account (@energystatsuk) i've been mulling the same subject with followers.  There is already work happening, but all very disparate.   People with Powerwalls doing, Chris with his Lux etc.

Needs everyone to come together and make something open-source that will plug-in to various energy tariffs, weather/solar predictions and battery APIs.

If you need any early adopter or testers I would be very interested in taking part

Likewise. 👍