Posted by: @bobtskutter

↑

That doesn't sound right, my calcs might be wrong

I think they are! The DT (delta t) isn't 25 degrees, its the difference between the LWT and the RWT which is a lot less and changes all the time, and in a defrost cycle is goes negative for while, reflecting the 'negative' flow of energy, ie instead of going from heat pump to house, it is going from house the heat pump (aka operating in heat thief mode). I'm also not sure what formula you have used, I use energy out (delivered to the building) in kWh over an hour = mean flow rate over the hour x (mean LWT - mean RWT) x specific heat of circulating fluid.

Posted by: @bobtskutter

↑

Please can someone share some data on how much energy their ASHP uses during a defrost cycle?

This is actually a tricky question (how much energy their ASHP uses), because, as noted above, what happens during a defrost is the heat pump itself draws less energy (from the mains) and instead it draws energy from the house (ie operates in heat thief mode). Consider this defrost cycle (the left hand one) from the 11th January (the upper plot is minute data, the lower plot is for the hour):

The amps (and so energy in) actually go down during the defrost, and the LWT/RWT delta t (and so energy out) goes negative, reflecting the fact the 'energy out' is going the wrong way ie from house to heat pump. The overall effect is the defrost cycle severely impairs net 'correct' energy out, ie energy delivered to the house. In the hour shown 0600 to 0700, the energy in was 4.48 kWh, the energy out was 6.69 kWh (too low, it needs to be over 8 at these OATs) and the COP was a dire 1.49.   

Posted by: @johnr

↑

I contacted Tesla who changed the grid code in my Tesla gateway (battery controller) from G99 to G98 and the charging rate has reduced to approx 7.5kW which is still sufficient to recharge the empty batteries within the 5 hour Octopus Go tariff period.

That sounds a sensible response from Tesla @johnr

Do you or @toodles happen to know if a Powerwall suddenly starts recharging from the grid, giving rise to a surge-demand?
Or does it slowly ramp up over a few seconds/minutes?

Here's an example of a ramp-up (and down again) from my own batteries last night.

The red line shows the Command from the Controller to one of the three inverters.
It sets the charge-current which will be contributed from that inverter.

The blue line is superimposed, and shows the battery voltage as it responds to the charge-currents from all three inverters.

You can see the Controller gradually increasing the charge current taken from the grid in 8-steps, starting at 10A.

This is the sort of 'grid-friendly' strategy which we need to have implemented across all LCT (Low Carbon Technology) devices.

@transparent Sorry, I don’t think my apps help in this regard as the resolution is Farrrr to low to be of any use in this regard. When I watch the app at Tesla charge start up, the display just suddenly announces ‘charging’ and the rate shows 10 kW. Whether it ramps kindly or jumps inconsiderately, I know not. Regards, Toodles.

@transparent I've got some 5 minute data which includes the battery charge level. The increase in charge over the first 5 minutes of charging is about 65% of the increase in charge over subsequent 5 minute time steps (until the charge rate reduces as the battery charge exceeds 90%). This lower initial increase in charge implies that there is either some initial ramping up of the current from the mains or there is a time delay after the start of the charging period before the battery starts to charge.

@cathoderay thank you for sharing the most resent set of data, I'll work through that.  Are you able to create a table of numerical data, or can you only get charts?

Others, I'm not planning on installing battery storage.  Please set up your own thread to discuss battery storage so this one can stay on topic (a small ASHP for garage workshop).  Thanks.

Regards

Bob

Posted by: @bobtskutter

↑

Are you able to create a table of numerical data

It already exists, you can't have a chart without numerical data behind it! I'll post it shortly, after I've added the kWh columns. The way my data collection works is I have one minute sample data file with the raw data, and then another one with the calculated hourly data, so normally I don't see the minute calculations, just their hourly totals. You also need to know I use a 'correction factor' for the energy in to allow for the fact the minute data (which ultimately comes from Midea sensors) produces energy in values that are slightly under what an independent kWh meter shows, and I incline towards believing the independent meter.

Posted by: @bobtskutter

↑

Others...Please set up your own thread to discuss battery storage so this one can stay on topic

Please rest assured Chairman Mars will invite you to be a moderator just as soon as he thinks you are ready to be one. But fair point, and @transparent might even agree with you!

OK here the data for 0600-0659 in the chart I posted  earlier today at 10:13 (actually its 0559-0658 because each row is the trailing minute, but what's that between friends...). I've added some notes to I hope explain the additions/calculated values which are not normally in this data/file. Some of the apparently inexplicable columns are to do with things like deciding whether each row is space or DHW heating - ignore them, along with some of the never change columns eg room_temp which is Midea's idea of room temp! Ambient is OAT of more accurately AIT (air intake temperature), MD02_tmp is the independent room temperature. Also note this is 'working data from the furnace room', not polished results for publication other than here on a 'for interest' basis. E&OE etc etc  

OK... pulling this back on topic:

@bobtskutter - have you come across the Mitsubishi MSZ-AP20VGD ASHP?

It's mainly sold in Australia & NZ where they want cooling as well as heating.

It can provide 2kW heating and 2.5kW of cooling, and might therefore be preferable to the 4kW units you've referred to at the top of this topic.

@cathoderay thank you for the data.  It includes information on the fan speed, that's helpful!  The data show's your room temperature is 25C and never changes, is that correct?

I've worked through the data and estimate there's about 3kg of ice on the ASHP fins.  Ive attached a spreadsheet with the analysis, hopefully it works, I'm using OpenOffice.

There's a 9min period of steady operation which I've used to calculate the performance of your radiator system, then used that number to estimate it's performance as the flow temperature drops.  The flow temperature is always higher than the room temperature, so the room isn't warming the water inside the radiators.

The calculation takes each 1min period as an energy balance.  Think of it as having a bucket of water cooling down, Total energy out = melting ice + rads.  The energy out in each 1min depends on the temperature change in that minute (T_start - T_end), it's all in the spreadsheet.

Regards

Bob

@transparent Thank you for that suggestion.  I had seen that type of unit.  They fall under the F-Gas regulations in the UK, because there's a refrigerant line between the outside and inside unit.  I'm not F-Gas certified so wouldn't be able to install it legally.  I could pay a contractor to install one but then I'd also be paying the contractors overheads and I'm trying to do this cheaply.

Thanks

Bob

Posted by: @bobtskutter

↑

The data show's your room temperature is 25C and never changes, is that correct?

No! I mentioned this in the post that has the download ("ignore them, along with some of the never change columns eg room_temp which is Midea's idea of room temp!". Room temp is MD02_temp (an independent sensor).

Posted by: @bobtskutter

↑

I've worked through the data and estimate there's about 3kg of ice on the ASHP fins.  Ive attached a spreadsheet with the analysis, hopefully it works, I'm using OpenOffice.

Very interesting. If base load (8.4kW) is the total heat loss as derived from this data you have, it is as expected a bit on the low side because at these OATs my heat pump can't quite cope, that's why the room temp is around 17.7, rather than 19 degrees (my preferred room temp) or 21 degrees MCS/design room temp. My empirical heat loss (ie from observations rather than using a whatiffery calculator) is probably around 9kW at -4 degrees. Note the day the data you have been using comes from the day when it was very cold, and for the first time I really did crank up the heat pump to max, ie it has never worked harder (but was still within its operating range, for obvious reasons). 

Going back to your original question ("how much energy [a] ASHP uses during a defrost cycle"), I think my first observation has to be that all the energy comes from the hot water in the pipework/rads, none from the rooms (which makes sense, the MD02_temp doesn't from during the defrost) or the heat pump itself (note also amps in drop very low during the defrost). The amount then, surely, is the amount that flows back into the heat pump from the rads and pipework, ie the negative energy out flows?

The reason defrosts matter so much is not just what happens during the defrost, when the heat pump is not producing any new heat for the house, it is also because of the long recovery period, with up to a half or more of the cycle running with an actual LWT below the set LWT, sometimes well below. If I can borrow your chart from your spreadsheet: 

The set LWT during the period shown is 58 degrees, but the actual LWT only approaches that during the second half of the cycle.  

@cathoderay, darn it.  I missed your comment about room temperature.

I've also worked out your primary system volume, 20L.  That's small.  You have a plate heat exchanger for system separation don't you?  Once the LWT starts to drop the plate exchanger will start to pull energy from your radiator system - you did mention this in one of your posts.  But you will only get good heat transfer over the plate exchanger if there's a large DT between hot and cold.  I think that explains the rapid increase i LWT after the defrost, from 27C to 38C in one minute, then the slow recovery in LWT because it's transfering heat into the radiators.

You can change Base!N14 0/1 to specify 1=platehex, 0=no plate hex.

With a plateHex i get 7.7kg ice, no hex 4kg.  I'm going to need to think a bit more about how to model this.

Bob