On-grid domestic battery storage

1961Nick

zeupater wrote: »

Hi

Might be a long post, but it's considered & concise .... I'd rather read 100 at that length & quality than even one 5 liner of something that doesn't stand up to a basic sanity test ....

As mentioned numerous times, there's a sweet spot on batteries where the charge rate can absorb the PV output, the capacity is high enough to compensate for 'normal' periods of poor generation whilst not being too large to justify the additional benefit (returns eventually flatline!) and the discharge rate covers the majority of well managed high loads (ie sequential appliance usage where possible), without expecting to cover the highest demand conditions such as cooking Sunday lunch (hob+oven+ ... etc) ... which you seem to have recognised.

The issue that most miss is that the sweet spot changes in line with overall electricity consumption & usage patterns ... this is why it's not unreasonable to suggest or recommend that efficiency improvement measures are considered before configurations are settled & purchase decisions are made ... it doesn't make sense to have to justify decisions after the fact by unnecessarily 'burning' energy as an accounting exercise! ...

I find the information on lags in demand following interesting as it confirms something that was discussed on here some time back regarding 100% of energy being supplied by a battery and achieving zero exports .... also the on the <200W 'decision making', that's a useful observation that seems to confirm something that I pondered some time back related to the management system working on a clamp based sensor as they tend to display this effect at low currents which is probably why most setups I've come across (/seen) use something similar to the SMA energy meter which should be more reliable (/accurate) & react faster ...

Regarding the clamp based export/import monitoring, do you know whether the system measures & compensates for daily grid voltage variance, PF etc, or does it simply take the voltage from a variable within the management system? .... there's lots of potential for import/export performance issues just on this simple point alone!


HTH
Z

It's great to have this thread back on track.:cool:

The induction hob/ME3000SP lag is something I've also observed. There appears to be 2 scenarios. The first is where the inverter is already discharging - the ramp up time is around 5 or so seconds every time the hob cycles 'on'. The second is when the inverter is in standby or charge mode - the switch over to discharge can take 20 - 30 seconds. The amount of actual import during the second scenario is noticeably higher & could be largely eliminated by increasing the base load to a value high enough to keep the inverter in a discharge state during the hob 'off' cycle.

I use a Geo pulse counting monitor which is 100% accurate (meter import = monitor import at the end of the month). The induction hob induced (pun intended) 'blips' on the import graph when the hob is used at midday on a dull day are higher than those when it is used after sunset. That would seem to back up what I see on the monitor.

I class the above as interesting rather than a problem as most of the time she has the oven on at the same time as the hob, so the demand is more than the ME3000 can handle anyway. Also, if the charge state isn't enough to make it through to sunrise, it makes no overall difference.

Please treat the above observations as a 'trend' rather than scientific evidence as I've not had batteries long enough to be sure about anything yet.;)

onomatopoeia99

1961Nick wrote: »

I wonder how many FMs are avoiding joining this discussion for fear of having their posts microscopically examined.....or does everyone run for the safety of the TAMG thread instead?

I would be one.

My battery should be coming at the end of May as the installer will have the new version of the gateway that allows it to supply the house in the event of a mains failure. My motivation for having a massive battery attached to the wall of my house is less money saving and more "monster sized UPS", though the considerably larger eventual payback of solar pv + battery vs just solar pv over the course of a couple of decades was a significant factor for going with the battery.

I haven't done the sort of minute analysis that has been seen over the course of the last few pages and have no wish to. I use a fairly flat 500kWh / month which is largely due to running a rack mount server (like are found in data centres) 24 / 7 with a bonus having a TV that consumes 500W (one of the last Pioneer 50" plasmas).

What's interesting to me are things like the tariff Octopus offer for EV charging as it could be used to charge the battery up over night and combined with even a single 'O' of generation during the day be enough to meet my entire electricity needs without troubling the grid. A few years down the line I'll think about further battery storage in the form of a car, when they can be made to seamlessly integrate with the home electrics when connected, so both charging overnight or from solar and also acting as a supply to the house in preference to the grid. The technology / flexibility doesn't quite seem to all be there so it will fit together yet.

Zarch

Forgive me for not keeping up with the entirety of the thread...... it looked pretty in-depth.

What price per kWh are batteries right now? What price / install size have folks been putting in recently?

Ta.

pile-o-stone

onomatopoeia99 wrote: »

A few years down the line I'll think about further battery storage in the form of a car, when they can be made to seamlessly integrate with the home electrics when connected, so both charging overnight or from solar and also acting as a supply to the house in preference to the grid. The technology / flexibility doesn't quite seem to all be there so it will fit together yet.

I was thinking about this and wondered why I'd want to run the house from the car battery, risking a situation where the car would not have enough juice for a journey the next day? I guess software could allow you to set a minimum threshold on the car battery so that it held enough charge for your normal commute. The issue would be if you had an unplanned/emergency trip to make and didn't have the range.

Though I guess you could just drive to the nearest rapid charge point. You'd pay more to charge the car than you saved when you discharged it into to the home, but this would be a rare occasion.

onomatopoeia99

pile-o-stone wrote: »

I was thinking about this and wondered why I'd want to run the house from the car battery, risking a situation where the car would not have enough juice for a journey the next day? I guess software could allow you to set a minimum threshold on the car battery so that it held enough charge for your normal commute. The issue would be if you had an unplanned/emergency trip to make and didn't have the range.

This is part of the whole "not quite all joined up yet" aspect of it, for me its an emerging technology and this part hasn't emerged far enough yet.

Though even if I bought an electric vehicle for the daily commute and local trips, I'd keep a petrol powered one for long journeys though.

1961Nick

pile-o-stone wrote: »

I was thinking about this and wondered why I'd want to run the house from the car battery, risking a situation where the car would not have enough juice for a journey the next day? I guess software could allow you to set a minimum threshold on the car battery so that it held enough charge for your normal commute. The issue would be if you had an unplanned/emergency trip to make and didn't have the range.

Though I guess you could just drive to the nearest rapid charge point. You'd pay more to charge the car than you saved when you discharged it into to the home, but this would be a rare occasion.

I guess that if there was a lot of uncertainty you could buy a hybrid vehicle rather than a BEV?

Get it wrong & the following could happen; PV > PV inverter > Battery inverter > Battery > Battery inverter > EV inverter > EV battery > EV inverter > Consumer unit.

orrery

pile-o-stone wrote: »

I was thinking about this and wondered why I'd want to run the house from the car battery, risking a situation where the car would not have enough juice for a journey the next day?...

You're thinking this through from the wrong end.


The massive battery capacity that the fleet of EVs brings to the grid will provide 'peaker' fill-in generation at busy times. This will provide a similar function to pumped storage like Dinorwig in Wales, only much much bigger and much faster. The grid will be prepared to pay for this.


The real demand will be between about 5 and 7pm each evening, and prices will fall overnight when there is little demand thus ensuring that you supply power when it is expensive and use it when it is cheap.


As for supplying the house as well, bog standard EVs are increasing in battery capacity 24 > 30 > 40 > 60kWh over the years and that will continue to increase. Your daily household peak demand will only be a fraction of that and you will recharge at off-peak times.


You will have control over when you discharge or charge to suit your usage and - as you say - if you get caught out then use a rapid (you'll have lots of income from the grid to cover it).

ASavvyBuyer

orrery wrote: »

You're thinking this through from the wrong end.

The real demand will be between about 5 and 7pm each evening, and prices will fall overnight when there is little demand thus ensuring that you supply power when it is expensive and use it when it is cheap.

As for supplying the house as well, bog standard EVs are increasing in battery capacity 24 > 30 > 40 > 60kWh over the years and that will continue to increase. Your daily household peak demand will only be a fraction of that and you will recharge at off-peak times.


You will have control over when you discharge or charge to suit your usage and - as you say - if you get caught out then use a rapid (you'll have lots of income from the grid to cover it).

Agreed, the use at the peak time/all evening, would be less than 20% of the EV battery capacity, even with the current EV we have of 30kWh. That would easily be replaced overnight on a cheap tariff.

That is why I am hoping that V2H will be available in the UK at a reasonable cost in the near future. As far as I am aware, they already have it in Japan, etc, but not approved for the UK yet.

zeupater

1961Nick wrote: »

It's great to have this thread back on track.:cool:

The induction hob/ME3000SP lag is something I've also observed. There appears to be 2 scenarios. The first is where the inverter is already discharging - the ramp up time is around 5 or so seconds every time the hob cycles 'on'. The second is when the inverter is in standby or charge mode - the switch over to discharge can take 20 - 30 seconds. The amount of actual import during the second scenario is noticeably higher & could be largely eliminated by increasing the base load to a value high enough to keep the inverter in a discharge state during the hob 'off' cycle.

I use a Geo pulse counting monitor which is 100% accurate (meter import = monitor import at the end of the month). The induction hob induced (pun intended) 'blips' on the import graph when the hob is used at midday on a dull day are higher than those when it is used after sunset. That would seem to back up what I see on the monitor.

I class the above as interesting rather than a problem as most of the time she has the oven on at the same time as the hob, so the demand is more than the ME3000 can handle anyway. Also, if the charge state isn't enough to make it through to sunrise, it makes no overall difference.

Please treat the above observations as a 'trend' rather than scientific evidence as I've not had batteries long enough to be sure about anything yet.;)

Hi

The monitoring I referenced wasn't your household monitoring, but the clamp based current import/export monitoring that the storage solution uses to manage charge & supply from the batteries ... for this to have full effect the voltage & PF at the time of a sampling must be considered, else there's an increased likelihood of unplanned imports or exports creating a charge loss to the grid or unnecessary imports. The frequency of the sampling is also quite important ... I did have a scan through the specification but either missed the relevant detail, or it simply wasn't there ... that's why I'm interested as it could explain the reaction lag that both yourself & Solarchaser have observed, for example a 5-10 second sampling frequency on a regularly switched high load such as a resistance hob may (or may not on averaging logic!) have a significant effect ... all depending on the interaction between the two frequencies!

I can understand the logic behind solution cost reduction through employing relatively cheap current based sensors as opposed to accurate measurement within a control environment, but the logical extension of inaccuracy would be for the designers to protect against unnecessary imports or exports by building a power switching threshold into the management system, which could account for the approx 200W mentioned .... on a 3000W maximum charge/discharge, 200W would represent around a 6-7% minimum threshold, which would adequately cover a 10V grid variance (4%) and have a little in reserve to cover PF issues, which would be one way of approaching a potential cost related issue ....

Just for context & comparison, monitoring import/export using the SMA metering solution samples every second and operates within a 1% accuracy tolerance ( link ) which could equate to a considerable performance difference in marginal generation/demand conditions! ....


HTH
Z

gefnew

hi
A little light reading with some techie bits about batteries and solar and wind.
https://batteryuniversity.com/learn/article/charging_with_solar_and_turbine
regards