How do you manage your battery charge?

Boffinboy24

I haven’t got a battery - and had ruled out one for now due to uncertainty on value, but the “fun factor” keeps me curious about it.

I was wondering how do people manage overnight charging of their battery at different times of the year so as to not run dry, but also not overfill from the grid? I know the Tesla Powerwall can work off the solar forecast and is largely automatic “set and forget”, but not sure about others. Do you just set to charge a certain amount overnight in winter months, and a different amount (or not at all) in summer?

It must be a bit of a challenge to get right as weather is not consistent - yesterday I got 7 kWh from my solar array and the day before it was 33. Do you change battery behaviour daily if the forecast is bad? Or do you just get it “good enough”? Do other battery platforms have some forecasting / automation features based on weather? Given time-shifting and maximising self solar use are the main gains from the battery, optimising must be part of the game!

Exiled_Tyke

I simply charge up every night regardless. At this time of year, if it's been a sunny day then it typically takes about 3kWh to fill up. I'm guessing that second guessing the forecast and trying to judge whether or not to take from the grid, is likely to cost me far more on the days I get it wrong.  Yesterday was a good example.  The forecast here said sunny spells all day but the sun never came out.   Had I not started the day with a full battery then it would have been empty by the time I'd cooked dinner and drawing from the grid at 4 1/2 times the cost.   Or to put it another way my usual top up at this time of year may be sometimes costing me 22.5p per night (although could be more admittedly) .  BUT I only need to be short by 670Wh at the end of the day to exceed this 22.5p charge.  Without a full battery to start the day yesterday would have probably cost me about 50p in peak time import.  Give it a few weeks and the sightly shorter days make this risk far more pronounced. 

Also for me,  having a mainly west facing system, if I didn't start the day with a kWh or so in the  battery I'd be importing peak electricity as soon as I got up and put the kettle on.   

So yes it does seem a shame that most of the days at this time of year I'm giving a heck of a lot to the grid but I reckon it's the best way for me.  I could spend a lot of time looking at forecasts and still get it wrong. 

QrizB

If you've got a single-rate electricity tariff, best to leave your battery alone in "auto" mode to do its own thing.

If you've got dual-rate electricity (and don't have a Powerwall or similar) you have more options. Here's what I do:

• Winter: set battery to charge fully overnight. My battery & array are both small enough that the battery will always be empty before the next charging period comes around.
  
• Spring / autumn: set battery to charge partially overnight. Taper the charge back a bit every week or two.. The idea is to grab enough charge to get to 10am or so on battery, by which time the sun (if there is any) is on my array and I can start charging. The battery will usually be empty before the next charging period comes around.
• Summer: don't charge overnight other than exceptional circumstances (last night was one, see below) as the battery will usually last through the night until charging starts at 8am.

I'm not saying this is the ideal way to manage a battery, and it does require manual intervention in the battery inverter settings every so often. Yesterday, for example, I didn't generate enough electricity to fill the battery and it was at 30% SOC when I checked before bed so I set it to charge at 100W during the 4 hrs of my off-peak period. That ensured it got me through the night and far enough into the morning that solar could take over.

Edit: I was writing this post when E_T submitted his. I agree with him that if you've got an off-peak tariff that's much cheaper than the peak one (and, as it happens, we both do) it can make sense to buy extra power off-peak rather than risk running out and having to pay peak prices.

Second edit: There's also hot water; you can think of your HW tank as a 5-10kWh heat battery where you can deposit more electricity. For many (not all) solar PV owners, it's more expensive to buy gas than it is to divert surplus electricity to heat water. I'm fortunate to be on an old enough electricity tariff that it's cheaper to buy off-peak electricity than it is to buy gas, so (based on the mediocre weather forecast for today) last nght I ran the immersion and heated the top half of my HW tank from the grid.

Mikeyboy01443

I won't charge from the grid until its closer to winter, at which point I will change to Octopus Go so i can charge up overnight...
I will probably aim to to charge to maybe 50% overnight to see me through til the next day, i usually check the forecast for the next day to see if its a decent solar day, but as I'm just on the standard tariff there is no benefit to me charging up overnight at the moment...

Boffinboy24

Thanks all. Really helpful. I saw a Twitter post that suggests GivEnergy are adding some kind of forecasting mode to their battery. We are currently on an old E7 fix, which runs to September, and works out better than single-rate with our solar. No EV, but maybe in future. I would need to work out if Go would work out better or worse for us in the winter due to higher daytime rate. Not sure if it would be better unless we had a large battery, because we are quite high users - 20kWh a day or so. 

Our array is SSE so tends to generate slightly more in the morning, which maybe means we wouldn’t need to put as much charge in for sunnier months. ET - like your way of thinking of just filling up though if on a tariff like Go. 

Longer term from an ethical perspective I do feel like I should move to low rate electric/solar divert for hot water. Seems like it can actually be comparable to gas given the losses with gas efficiency and pipe work. I was intrigued that Giv also seem to be planning to offer a compatible diverter in future. I know they don’t make financial sense post FiT though. We can use our A2A heat pumps in shoulder months for heating which also helps reduce gas use a bit. 

orbit500

I use Go Faster to overnight charge a coupled 30kwh battery and daytime top up from my 10kw array. Currently I just fill up to 100% with no forecast while I gather data. I will move to a forecast system shortly.
There is a company called Solcast which, given your PV array parameters, can accurately predict the day ahead production. You can then use this to match the required battery % to finish the day with ready for the next overnight charge. Obviously if you're on Go or similar the aim is to completely avoid any peak import and secondly not export any excess PV as the SEG/TOU differential is very high.
This is all done via Home Assistant.

QrizB

orbit500 said:

There is a company called Solcast which, given your PV array parameters, can accurately predict the day ahead production.

I've just gone to their site https://solcast.com

They also do a hindcast - you give them your solar PV paramaters and they tell you how much you should have generated over the past week.

For me they hindcast 12.4 / 16.6 / 14.9 / 13.7 / 14.9 / 5.9 kWh/day vs. my actuals of 11.5 / 16.7 / 14.5 / 13.6 / 15.5 / 4.0 which is pretty close to the mark. (That was with the default 0.9 efficiency, which might be a bit optimistic for a decade-old system.)

EcoScruples

QrizB said:

orbit500 said:

There is a company called Solcast which, given your PV array parameters, can accurately predict the day ahead production.

I've just gone to their site https://solcast.com

They also do a hindcast - you give them your solar PV paramaters and they tell you how much you should have generated over the past week.

For me they hindcast 12.4 / 16.6 / 14.9 / 13.7 / 14.9 / 5.9 kWh/day vs. my actuals of 11.5 / 16.7 / 14.5 / 13.6 / 15.5 / 4.0 which is pretty close to the mark. (That was with the default 0.9 efficiency, which might be a bit optimistic for a decade-old system.)

I use this also and have found the Hindcast to pretty pretty spot on so far, the forecast not so, I seem to produce up to a Kw more peak power on most days than that one suggests

So far I have left my battery in maximum self consumption mode and will stick to that for the summer, I'll only look at topping up from the grid in the 3 winter months if I need to, I don't really have the patience to keep a daily çheck but who knows, maybe I'll change.

Magnitio

My instal was at the beginning of Feb, so no experience of the depths of Winter, but during the early months I was looking at the forecast and battery status each evening to determine how much extra to charge the batteries. It only takes a few minutes and is quite an interesting balancing act.

Last month I switched to Agile Octopus, which means the rates for both outgoing and incoming change every half hour. This adds an extra dimension to the balancing act e.g. delaying the charging of the battery from solar until the early afternoon when outgoing rates are lower. This of course depends on the forecast. In addition, on particularly bright cool days, the peak output from the panels exceeds the limit of the inverter, but if some power is diverted to the battey, all of the generation capacity can be used. Some days I get it wrong: for example on Friday the outgoing rates were high, so I exported several kW in the evening only to find that generation yesterday was extremely poor. I therefore used the gas boiler to heat the hot water tank rather than the immersion.

Quite often the difference in income/expenditure is minimal and really not worth it, apart from the "fun". At other times, it can be quite rewarding. Income over the past week was nearly £34 from the exporting of excess solar - part of that was through scheduling battery charge/discharge to maximise rates.

Mikeyboy01443

I tried looking at the solcast tools, i dont really get what I'm looking at tbh although i only gave it a quick 5 minute browse... 
How do i find out the azimuth / tilt etc, i think I had some of this info in quotes? But i wouldn't know how to measure it myself

EcoScruples

I was the same with Solcast, it looked really confusing but once you know which bits to look at it becomes quite simple. Stick to the API (live and forecast) PV output page and you'll be good.
For the Azimuth bear in mind it's an Australian site so that'll be back to front, I'm 5 degrees east of south facing so mine is 175 azimuth.