Thanks Tim
The design document says: Recommended emitter output in property (Delta T50): 20200W. The document includes a table showing that the new radiator output adds up to slightly more than that.
We have 11 emitters. Some of them are turned down a bit. None of them is turned right down. I haven’t systematically measured radiator temperatures, but none of them has ever felt noticeably hot. Even at LWT=55.
I’ll follow up on the other points you and Dave raise shortly.
Hello All
Thanks to Tim, Dave and David for posts so far on this. This is to summarise
where I think we’ve got to, and what I might do next.
The situation:
My heat pump’s minimum output appears high relative to mild-weather demand
In mild weather the system as a whole operates less efficiently at LWT = 50C than at
55C. This is shown by the plot of the heat pump’s dedicated consumer unit
electricity use vs Delta T (inside / outside)
We have a buffer tank. No ABV as far as I can see, and none is shown in
the design document.
The heat pump’s minimum heat output is 4.4kW, as indicated by Dave in post
12
One other point I didn’t mention before. With outside temperature roughly 10 to
12 C, the house cools at about 0.26 degrees C per hour with the thermostat
set very low, so that the heat pump switches off. With the thermostat set
so that the heat pump comes on, the house heats at about 0.43 degrees C
per hour. See the spreadsheet I’ve attached.
Cooling and Warming Rates.xlsx (8.6 KB)
Possible cause of inefficiency: Cycling through the following steps:
Thermostat in the kitchen indicates heating needed
Heat pump starts
On a mild day, the required average heat input is relatively low. For
context, with the old gas boiler, at around Delta T = 10°C our total
energy use was about 75 kWh/day (green line on the Energy per Day vs Delta
T plot), i.e. an average of about 3.1 kW over 24 hours.
Energy per Day vs Delta T.pdf (24.3 KB)
One of the following then stops the run:
In either case, the heat pump turns off; if the house still needs heat,
it will restart again later. A higher minimum output makes either
stopping condition occur sooner (faster room warming in case A, faster
LWT rise in case B).
This can lead to frequent short runs and repeated restarts, particularly
in mild weather. Overall efficiency suffers because each start/stop has
overheads (ramp-up, pumps/controls, and other fixed losses).
What next (experiment):
I’ll get more plot points at LWT = 45C (Currently I have only three). This
will confirm whether it is still less efficient than at LWT = 55 C
Staying at LWT = 45, I will try to encourage the system to run
for longer periods as follows:
I will do this for a day and then make my usual measurements and
add a point to the Consumer Unit electricity vs Delta T (internal -
external) plot to see if there is any improvement in efficiency. If
necessary I’ll do this for another day or two
If this works, then I would want to look for a way to simplify it so
that I don’t have to keep adjusting the thermostat manually. Ideally by
automating it with a schedule if possible.
Any comments or suggestions will be very welcome.
You might find it helpful to plot delivered heat energy per day vs delta, as this will be linear most of the time, excepting very windy or sunny days, and assuming constant indoor temperature. Heat delivered will match heat loss, which is a function of temperature differential, whereas electricity consumed will depend on the coefficient of performance.
It is possible to get real time data out of the Daikin heat pump, with some technical expertise. This topic describes what’s involved: Daikin Altherma, ESPAltherma & Home Assistant with OpenEnergyMonitor
Thanks Tim
How accurate would you expect the Daikin heat delivered to be? As we discussed earlier in this thread, there is a question over the accuracy of the energy input reading from the Daikin unit.
I’m hoping to get my hands on an emonPi before too long!
I don’t know, but it’s likely to be good enough to work out the actual heat loss of the property. Accuracy of heat measurement is independent to that of electrical consumption.
If you can find a period when the primary pump was circulating with the compressor switched off, check that flow and return temperatures match - that will give some confidence about the accuracy.
Knowing if there’s glycol in the loop is also useful.
As per my previous calculations, the heat pump should have no problem running at 45° flow on its minimum 4.4 kW output, possibly even lower.
Thanks Tim
That is 105.6 kWh per day. But last year, when we still had the gas boiler, at Delta T = 14 our total daily energy input was less than that. Building heat loss would probably have been a bit less due to less than 100% efficiency of the boiler, and hot water sent to the drain.
So it seems inevitable that the length of heat pump runs will be limited because the heat pump will be producing heat faster than it can be dispersed. Run length will be limited due to either mechanism A or mechanism B above. I think it is these short runs that are making the system run less efficiently than it should. I think there is an overhead every time the heat pump starts, so that more starts mean less heat produced for the same electricity.
My plan is to test this theory by adjusting the thermostat up and down throughout the day to so that when it switches off it has further to fall, and when it switches on it fas further to rise before it hits a limit due to either mechanism A or B.
Any thoughts on this theory would be very welcome. At the moment it’s the only way I can make sense of the data I’m collecting. Latest plot attached.
Energy per Day vs Delta T.pdf (24.4 KB)
So far as the building heat loss goes, I’ll look at the Daikin heat loss data (I’ve been collecting it on a daily basis for some while). Heat loss can probably also be extracted from the green ‘Floor insulated, double glazing complete’ line with some adjustments. The building fabric is unchanged since then. I expect the result would be something like the 6.4 kW you estimated earlier.
Chris