I don’t use LWT and I do use Madoka.
In the spirit of leaving as many facilities “in place”, I decided to leave the WC curve active so that I could still trim its values as I went along. As you see, my WC is pretty much a flat flow temp anyway.
Regarding the Madoka, I felt that keeping it leaves some opportunities for scheduling and setbacks that might be out of reach if you selected LWT? I’m not using schedules or setbacks at the moment, but they are easy to dial in at any moment using the Onecta app.
There are folk on here who prefer not to run the heatpump 24/7 - especially if it is oversized.
With no modulation and a Madoka target temp presently a little higher than my heatpump settings are likely to achieve, the Madoka is not having any influence on the pump.
I agree with @John Gouk that you should do some trial and error to get your WC curve down as low as your house will allow. For every degree you can reduce the flow temperature (and also DHW tank target storage temperature) there are significant rewards in CoP and efficiency.
I haven’t noticed much difference in performance when switching between fancoil and radiator settings. The most significant effect has been when increasing the target flow/return DT setting (MMI 2.B.1).
Perhaps you don’t need your WC to be as high as 45@-3. Why not try running it at, say, 37 @ -3? Look for two outcomes.
1 If this lower flow temperature creates cycling, it means your rads can’t support this low temp and you should increase the 37 a couple of degrees (and maybe a couple more) until the cycling stops.
2. If your house stays warm enough with 37, great! If not, increase the 37 a couple of degrees until the heat pump is balancing the house heat loss.
Right, im back. So the odd issue is solved with the power difference. I’ve gone to take the cover of the box to take photo’s I double checked that I was monitoring the main feed to the fuse box and I was. But just by moving the clamp and I guess the proximity of the cable in the camp i’m now getting 883W. So going back to the other issue, that the pump cycles a lot above 7degress. Since Friday it’s cooled down to around 4 degree and the pump is running so much better, the house is 21.5 degrees, so a little to warm and that’s with a flow temp of 34/35 degrees. I think i’m going to push for Octopus to downsize to the 8Kw version.
Good luck there. I have been speaking to Octopus about a change to my pump but they appear to be standing their ground more. Unless you are operating your pump at the settings stated in your contract ( for the Daikin 9kW in my contract it is WC 50c @ -2 ) they will not consider a change.
Hi Stefan. I think this makes sense. As the ambient temp increases I think we will both need to increase our flow temperatures to prevent cycling, and at the same time be prepared to schedule some periods in the day when the pump is off.
It’s been occurring to me that there is a case to be made for having a “reverse” WC curve to achieve having the flow hotter as the weather warms up (and then heating intermittently).
In fact, the space heating COP was 4.26!! Dragged down by the 2.16 DHW CoP
There’s nothing wrong with cycling, it’s doing it every few minutes that’s bad. At about 9c mine was doing 5 cycles in 8 hours…
The only way to get the run rate to be less than 850W is basically to use Pulse Width Modulation, or cycling, so a 50% duty cycle would be 425W. That’s true for any device that won’t modulate to 0.
I had mentioned earlier a plan to drop the target flow temp after detecting a defrost, and then raising it slowly. On a similar vein, for this cycling issue, I had wondered about something similar. The problem is that at low flow temperatures, radiators cannot give out enough heat. Allowing overshoot means it can go higher before shutting off.
I’d been pondering detecting the shutoff and, again, dropping the target LWT a little, so that the HP is off for longer. Then once it resumes, and the LWT inevitably starts to overshoot, increase requested LWT to catch up. A difference is that, in defrost case, raising the target would have to lead the actual LWT, whereas when cycling, I’d be following the rise, as it overshoots.
I’ve got a few automations in Home Assistant (using jwillemsen’s Daikin Onecta plugin) that are helping make my ELDA06 more efficient (e.g. selecting the warmest time in the off-peak window to reheat the hot water tank, creating a more flexible but stable modulation of LWT and a couple of others).
What I’d point out is that there is a per-day and per-minute rate limit on the Daikin Onecta API, so if you are doing quite a few changes in a short space of time, you may find your system isn’t able to react in the way that you intend, if your request is rate limited by the API.
If you are using the HA plugin you could reduce the number of data updates your system does to ensure your changes are not rate limited. Or, if you were going to create your own script to interact with the Onecta API, you could do only write commands and not read any data (or do only minimal reads to verify changes). That would help in keeping the number of interactions with the API below the rate limit.
Hey, So as the temp go over 6 degrees I have raised the flow temp and it does reduce the cycling, but in the same token our house get to nearly 23 degrees.
I keep thinking of a large 4 port buffer tank that have a mixer on the CH side, so the CH flow temperature can be kept at the weather compensation defined level while the buffer tank is allowed to get as hot as is required to limit cycling.
To keep heatpump on lowest output the LWT will need to be increased slowly as the buffer tank overheats, then callForHeat removed until buffertank is below temperature WC requires.
But such a tank may cost nearly as much as a new heatpump.
yeah, 200/day or 20/min. Reading every 10 mins leaves 50 requests for writes. I have been logging using 10-minute intervals during the day, and 15 overnight. I think I may add some temperature sensors on the pipework so that I can sample locally rather than having to use the API for that.
I can get power consumption from Zappi cloud - not sure if that’s rate limited, but I’m going for every 2 minutes there. I can also ask my inverter locally, more frequently - that would be whole-house consumption, but if it’s below 800W, I can be pretty sure the HP has turned off.
Having read the advice provided on the Forum regarding control of the HP I have made some changes. Have switched to LWT as the control, offset on Madoka to -2 and left HP running 24hrs. I have the power down to 900 - 1000kW most of the time. This has given me a steady flow chart and the property is continually warm, if not too warm at times.
This is the reading I am getting now:
The blower door test is about £300 and was very worthwhile in our case since our actual leakage was way below the defaults for our house age. We have solid concrete floors and have recently re-plastered throughout so less scope for leaks even if the floor is obviously a very poor insulator.
