Battery or thermal storage for HP timeshifting

[Marcus]

I lean strongly towards a battery for various reasons:

One is that it allows you to run the hp at the warmest time of the day (generally lunchtime to midafternoon), which in turn reduces the hp energy consumption. If you're running it at optimum efficiency then the power consumption should be well under 3kw.

Two:
I don't believe google.

60kwh of thermal storage = 1000 x 60 x 60 x 60 = 216,000,000 joules

Assuming you can achieve delta t of 50°C with a hp(?) and your heating system (underfloor heating?):

Shc of water is about 4200j/kg/°c (or j/l/°c)

So you would need 216000000 / 50 / 4200 = 1,028 kg (or litres) of water (assuming I've got my maths right).

But if you need 25°c for ufh, your hp would have to push it up to 75°c for a delta of 50, and most hps have a much higher COP at 30° than 60°, and relatively few will reach 75° if any.

Edit: actually google may have been about right after all.

[AlBargey]

Interesting comparison.

Round trip efficiency of a battery is quite a lot better than 75% I would think?

It's made me look at my battery efficiencies, which are likely worse for me being off grid and having the whole barge running off a larger 8kVa inverter, and topping up the solar from a diesel generator.

Anyway last year (2023):

4,092 kWh Solar
597 kWh Generator
3,419 kWh Consumption

= 73% System Efficiency (with FLA batteries)

This was using Forklift type FLA batteries, with a recording since installing a SmartShunt of:

4,453 kWh Discharged
5,683 kWh Charged

= 78% FLA Battery Efficiency (This was an older bank and is sold now as it wasn't giving the nameplate capacity any more)

Since Installing my DIY LFP batteries in December:

272 kWh Solar
293 kWh Generator
460 kWh Consumption

= 81% System Efficiency (With LFP Batteries)

The LFP battery efficiency:

418 kWh Discharged
427 kWh Charged

= 98% LFP Battery Efficiency

Quite a difference, and I expect the 81% to get a little better as I was testing and balancing the LFP's for a while when there wasn't much sun, It also varies throughout the year, but for someone with a grid connection and smaller inverter it should be a fair bit more efficient.

[nowty]

Some years ago at St Elsewhere, some one posted this,

Which was based on absolute peak efficiency levels which are not achievable in real life settings.



This made me do my own based on my real life data from a Forklift Lead Acid Battery in a 24V system.



I now have lithium batteries with a 48V system and I find my average system efficiency is now about 80% if I do grid battery charging during cheapslot and discharge during peak times.

But it gets complicated depending on whether you AC or DC couple your charging sources, I now do both.

If you DC couple form Solar PV you miss out on the DC to AC then back to DC losses which can be quite significant. There is a DC to DC conversion in the charge controller but its quite small if you design it right.

If you AC couple and use the power (or export) as its generated you miss out on the AC to DC, DC round trip in battery and the DC to AC losses.


My rule of thumb for average efficiency of modern stuff is,

AC to DC efficiency 95%.
DC to DC round trip of lithium batteries 98%.
DC to DC round trip of Lead Acid batteries 90%.
DC to AC efficiency 90%.

[AlBargey]

But it gets complicated depending on whether you AC or DC couple your charging sources, I now do both.

If you DC couple form Solar PV you miss out on the DC to AC then back to DC losses which can be quite significant. There is a DC to DC conversion in the charge controller but its quite small if you design it right.

When I went on a Victron training course a few years ago, they were very keen to emphasize the importance of the efficiency gains from using DC MPPT's for charging a DC battery vs a Fronius AC coupled solar charge controller. I was originally thinking to go for the Fronius, but it makes sense, and all our Solar is now on DC MPPT's to the batteries, but if exporting or a high base load then AC coupled would also make sense.

[robl]

We've got a 3.6kW Sunsynk hybrid inverter and 15.5kWh of LiFePo batteries from Fogstar. There's no PV yet attached to the sunsynk, but there is a good mid cert meter which shows the "go" import elec, and also the battery->home use electricity.
For the last 70 days the hybrid inverter used 1027kWh and output 788kWh, which is an average of 76.7% efficiency, a long way short of what the datasheet would make you expect! Still, it does a brilliant job for us. All elec house, we maybe import 1kWh of peak elec, the rest is "go"

Looking into the losses, they amount to (1027-788)/(70*24) = 142W average over the time. I think the Sunsynk takes somewhere around 60W just idling - that is, inverter running with tiny power in or out. It has a lower power mode than that they brag about, but the inverter isn't running then - so it's not relevant.
I think the battery itself is around 95% round trip. It doesn't loose any coulombs, but it has a series resistance which causes the I^2*R efficiency drop. We charge fast in the 4 hour window - it would be lower loss charging slower in a bigger window.

Sometime soon I'll stick some PV onto the sunsynk too

[CharlieB]

I lean strongly towards a battery for various reasons:

One is that it allows you to run the hp at the warmest time of the day (generally lunchtime to midafternoon), which in turn reduces the hp energy consumption. If you're running it at optimum efficiency then the power consumption should be well under 3kw.

Two:
I don't believe google.

60kwh of thermal storage = 1000 x 60 x 60 x 60 = 216,000,000 joules

Assuming you can achieve delta t of 50°C with a hp(?) and your heating system (underfloor heating?):

Shc of water is about 4200j/kg/°c (or j/l/°c)

So you would need 216000000 / 50 / 4200 = 1,028 kg (or litres) of water (assuming I've got my maths right).

But if you need 25°c for ufh, your hp would have to push it up to 75°c for a delta of 50, and most hps have a much higher COP at 30° than 60°, and relatively few will reach 75° if any.

Edit: actually google may have been about right after all.

Thanks Marcus. This post has mostly become a discussion about li-ion efficiencies - not a surprise really, and still interesting!

This response from you is just what I was looking for though. I can’t say I follow all the maths but if ~1,000 litres of water could store around 60kwh then this is very possible. The kerosene tank we’ll be getting rid of is 2,000+ litres so a thermal store could sit on exactly the same plinth.

Good point about running the hp during the day, but I’m imagining an overnight and a daytime running cycle in any case so it wouldn’t make that much difference. (I think most people with not-huge batteries do that too. Pv would change the game there of course.)

[richbee]

We've got a 3.6kW Sunsynk hybrid inverter and 15.5kWh of LiFePo batteries from Fogstar. There's no PV yet attached to the sunsynk, but there is a good mid cert meter which shows the "go" import elec, and also the battery->home use electricity.
For the last 70 days the hybrid inverter used 1027kWh and output 788kWh, which is an average of 76.7% efficiency, a long way short of what the datasheet would make you expect! Still, it does a brilliant job for us. All elec house, we maybe import 1kWh of peak elec, the rest is "go"

Looking into the losses, they amount to (1027-788)/(70*24) = 142W average over the time. I think the Sunsynk takes somewhere around 60W just idling - that is, inverter running with tiny power in or out. It has a lower power mode than that they brag about, but the inverter isn't running then - so it's not relevant.
I think the battery itself is around 95% round trip. It doesn't loose any coulombs, but it has a series resistance which causes the I^2*R efficiency drop. We charge fast in the 4 hour window - it would be lower loss charging slower in a bigger window.

Sometime soon I'll stick some PV onto the sunsynk too

What type of meter are you using, and where is it located in the circuit? - I have the same inverter with PV and batteries, and the import readings from the inverter CT coil are pretty far out - almost always ~2kWh per day higher than the smart meter readings (although the export readings from inverter vs smart meter are much much closer). I then doubt the inverter Load figures too, although I don't have another way to measure that.

[Marcus]

I lean strongly towards a battery for various reasons:

One is that it allows you to run the hp at the warmest time of the day (generally lunchtime to midafternoon), which in turn reduces the hp energy consumption. If you're running it at optimum efficiency then the power consumption should be well under 3kw.

Two:
I don't believe google.

60kwh of thermal storage = 1000 x 60 x 60 x 60 = 216,000,000 joules

Assuming you can achieve delta t of 50°C with a hp(?) and your heating system (underfloor heating?):

Shc of water is about 4200j/kg/°c (or j/l/°c)

So you would need 216000000 / 50 / 4200 = 1,028 kg (or litres) of water (assuming I've got my maths right).

But if you need 25°c for ufh, your hp would have to push it up to 75°c for a delta of 50, and most hps have a much higher COP at 30° than 60°, and relatively few will reach 75° if any.

Edit: actually google may have been about right after all.

Thanks Marcus. This post has mostly become a discussion about li-ion efficiencies - not a surprise really, and still interesting!

This response from you is just what I was looking for though. I can’t say I follow all the maths but if ~1,000 litres of water could store around 60kwh then this is very possible. The kerosene tank we’ll be getting rid of is 2,000+ litres so a thermal store could sit on exactly the same plinth.

Good point about running the hp during the day, but I’m imagining an overnight and a daytime running cycle in any case so it wouldn’t make that much difference. (I think most people with not-huge batteries do that too. Pv would change the game there of course.)

Your welcome but whilst a 1000l tank can store 60kwh of thermal energy, you would still need to change its temperature (delta T) by 50°C to achieve that.

I'm doubtful that you could actually use that temperature range with a HP. If i were to try this with my system for e.g.: My heat pump will heat to 60°C and my UFH could use water down to 25°C providing it's not sub-zero temps, but that's a delta T of just 35°C so a 1000l store would only give 42kwh.

If using rads and need a flow of, say 35°C, then the 1000l tank storage capacity drops to 30kwh.

Plus running the hp to 60°C puts it in it's worst performance zone (cop <3) meaning it will use much more energy than using it to heat the house directly (cop of 4.5 or 5).

Realistically then, if you have a hp and rads you should double or even triple that thermal store size to usefully store 60kwh.

Edit: that's supposed to be a regular smiley face, but for some reason this particular forum seems to exaggerate them into a cheering emoji

[GarethC]

Wouldn't just going for a big battery also have the advantage of allowing you to use off peak, cheap, low carbon electricity for all of your electricity needs. I. E. You kill two birds with one larger stone, rather than having to find space for additional thermal store?

[CharlieB]

I lean strongly towards a battery for various reasons:

One is that it allows you to run the hp at the warmest time of the day (generally lunchtime to midafternoon), which in turn reduces the hp energy consumption. If you're running it at optimum efficiency then the power consumption should be well under 3kw.

Two:
I don't believe google.

60kwh of thermal storage = 1000 x 60 x 60 x 60 = 216,000,000 joules

Assuming you can achieve delta t of 50°C with a hp(?) and your heating system (underfloor heating?):

Shc of water is about 4200j/kg/°c (or j/l/°c)

So you would need 216000000 / 50 / 4200 = 1,028 kg (or litres) of water (assuming I've got my maths right).

But if you need 25°c for ufh, your hp would have to push it up to 75°c for a delta of 50, and most hps have a much higher COP at 30° than 60°, and relatively few will reach 75° if any.

Edit: actually google may have been about right after all.

Thanks Marcus. This post has mostly become a discussion about li-ion efficiencies - not a surprise really, and still interesting!

This response from you is just what I was looking for though. I can’t say I follow all the maths but if ~1,000 litres of water could store around 60kwh then this is very possible. The kerosene tank we’ll be getting rid of is 2,000+ litres so a thermal store could sit on exactly the same plinth.

Good point about running the hp during the day, but I’m imagining an overnight and a daytime running cycle in any case so it wouldn’t make that much difference. (I think most people with not-huge batteries do that too. Pv would change the game there of course.)

Your welcome but whilst a 1000l tank can store 60kwh of thermal energy, you would still need to change its temperature (delta T) by 50°C to achieve that.

I'm doubtful that you could actually use that temperature range with a HP. If i were to try this with my system for e.g.: My heat pump will heat to 60°C and my UFH could use water down to 25°C providing it's not sub-zero temps, but that's a delta T of just 35°C so a 1000l store would only give 42kwh.

If using rads and need a flow of, say 35°C, then the 1000l tank storage capacity drops to 30kwh.

Plus running the hp to 60°C puts it in it's worst performance zone (cop <3) meaning it will use much more energy than using it to heat the house directly (cop of 4.5 or 5).

Realistically then, if you have a hp and rads you should double or even triple that thermal store size to usefully store 60kwh.

Edit: that's supposed to be a regular smiley face, but for some reason this particular forum seems to exaggerate them into a cheering emoji

Yep, I'm pleased to say I reached the same sort of conclusion before I saw your post. But even so, 3-4,000 litres of water is still less than the exterior volume of the existing tank, and there's no real problem with making a tank much bigger than that. I'm assuming - though I might be wrong - that even if very heavily insulated and extra plumbing is a hassle that a tank like that would cost less than a big lithium battery.