Heat Pump – not so good for water heating

Published July 16, 2012 by Sean

OK, here goes for a fairly complex posting. My Air-Source heat pump has been running for 6 months or so now, with a few glitches like a blown fuse (inside the unit) and initial teething problems which seem to stem from the heat supply being quite high in comparison to the load provided by at least the hot water cylinder.

Now that I’m able finally to turn off the heating aspect (manually at the moment), I’ve been able to check the efficiency for water heating. The Hitachi controller is set up to keep the tank temperature in a specified range between 45-55 degrees, and has no timing controls for DHW (Domestic Hot Water) other than provision for a ‘no heat’ input. This control is currently not used in my set up. I’m expecting to see some cost of running the system 24/7 (the compressor heater for example) and the storage loss of the cylinder.

Here is the plot of recent months electric in/heat out for the heat pump (kind of system COP). I was expecting around 2-3 during winter, but hoped for an improvement in warmer weather. The result is surprisingly low, so further investigation and instrumentation was justified.

Manual investigation of the water heat cycles suggested that the heat pump (Hitachi Yutaki RHUE 3AVHN) was being commanded to run at far too high an operating point in water heating mode. The interface between the indoor control unit and the external compressor/heat exchanger unit is simple – on/off and target output temperature using a 4-20mA current loop. The control panel inside the compressor unit will show the relevant temperatures it is running to; the indoor controller has sensors for flow and return temp, as well as tank temperature. My hypothesis is that there is a closed-loop feedback between flow temp sense and the set control current. Fortunately, it is possible to manually override the control outputs from the controller, and I managed to capture a day’s running with the loop current set to 17mA (about 51 degrees from the heat pump’s perspective). Settings for this run are:

Cfg Default Now Desc
P10 45 45 Hot Water set temperature (default is 45, I think mine was higher after installation)
P11 5 1 Hot water range
P12 10 1 Hot Water offset above P10 for setpoint
 P31 3 1 Sensor offset

Temperature is sampled every 30 sec using DS18S20 one-wire probes. Electric power is sampled by counting pulses from a dedicated meter (pulse counting using an LPC1769 microcontroller). The plot demonstrates heat loss over the day, hot water take-off causing the pipework to heat up (and ambient temperature having an impact on pipework temperature), in-tank turbulence during re-heat, and some periodic switching as the ASHP is idle.

Calculated from the graph (more accurate numbers can probably be automated):

Static heat loss 6degrees x120litres:
Daily pump runs (no usage): ~2
Energy to replace heat loss: 0.8kwH
Pump run for shower:  22 min, 1kWh
Idle power for compressor: 2.6 kWh/day

Zooming in on the shower run shows how the heat pump is able to modulate it’s output whilst the heat demand is high (water flowing), but at it’s lowest operating point it’s clear that flow temperature slowly increases. The heat pump seems to stop after about 1 minute at 4 degrees over it’s setpoint (which would correspond to 55 degrees C in this run). This is independent of the tank temperature, and demand may continue from the controller in which case the heat pump may come back in in 3 minutes (provided the loop temperature has dropped to 46 degrees – if not the system may lock out when it can’t reach the water target in 1.5 hours). There might be scope for modifying the demand profile, but the heat pump only seems to sample the control loop current every 10 minutes.

The profile of a heat cycle due to static losses is slightly different, the pump ramps up to temperature quite quickly. However, the periodic 200w (presumed heater) pulses are suppressed just after the run.

This data gives a reference to compare against so that the parameters can be tuned. It is also detailed enough to help assess the benefit of extra insulation and cold incomer pre-heat (combined with a buffer tank for space heating operation) which has yet to be installed. A further option would be to use solar thermal for hot water heating during the summer (or even something wacky like solar-thermal to supplement the compressor heater and pre-heat) once the maximum achievable cost saving can be calculated. Already, I know that the heat pump is taking ~20kWh per week in the summer (out of a total house consumption of 80-100 kWh) so this is the limit of what can be saved (£50 per 6 months); clearly any small improvement to winter costs will rapidly swamp this.

Filed under Heating

Comments (1)

Comments RSS - Trackback - Write Comment

  1. Hitachi Yutaki Heat Pump: Hot Water – Broken by design | Sean Houlihane says:

    […] identified earlier in this post, my heat pump is not following it’s control settings for hot water heating, resulting in it […]

    Posted April 2, 2013 @ 9:13 am (UK)

Write Comment