Well, by definition the heat leaving the house is equal to the heat being put into it by the heating system in the constant temperature steady state. If we didn’t have those losses there would be no need for heating.
I actually think both measurements are important. When I have a poorly performing unit, it could be due to a defect in the unit or due to a bad installation (uninsulated exterior pipework etc.). I need to have measurements at both points to be able to say which one it is for certain.
Every system will of course have its nuances and I agree it is important to capture and share relevant information to assist people looking to compare system performance.
The size and layout of my house necessitates long primary runs. The first 6-7m of the run is external (mostly underground), the next 2-3m (with 25mm insulation) is in my integrated garage space (unheated). The bulk of the remaining ~20m is in the insulated ceiling void between ground and 1st floor.
Previously my heat meter was installed at the end of the runs so did not account for the heat generated and lost through the primary pipework. My heat meters are now in the garage and therefore give me a better idea of heat pump appliance performance and include heat loss through 2/3rds of the primary pipework.
Watching this tread with interest as I try and work out where any of the inefficiencies of my system are.
I’ve found the following DHW run in coolish outside conditions but warm enough that there is no heating straight after (so any heat in the primary pipe work would be wasted):
Is there any way from this to work out what effect the primary ducts are playing?
We have about 5-6m of external pipework (quality of insulation install could be better) and about another 14-16m in the cellar (on this day it was about 16 C in the cellar and 9 degrees outside).
I assume the large first drop is the cold water from the primary pipework going through the flow temperature sensor and the warmer water sitting in the coil going through the return (looking like heat has been taken out of the system) and this would cancel out a bit as it all goes round the loop again and starts to stabilise as the HP adds heat.
It just looks like the first negative power output swings are very large compared to the positive ones.
I’ve recently been fooling around with some theoretical piping and vessel heat losses, based on adding the heat flow resistances due to convection and conduction. (This is applying comprehensive chemical engineering equations for inside/outside convection coefficients, and cylindrical surface heat loss for vessel wall and insulation.)
With ambient at -2degC, 2m/s wind, LWT at 40degC and RWT at 35degC, ~1m/s water flow in 28mm OD copper piping, and 25mm thick polyurethane foam insulation (k = 0.025W/m/K), the heat loss was only about 6W/m, i.e. about 60W over 10m piping.
A volumiser vessel 350mm ID x 550mm high at 40degC in an unheated outbuilding (5degC ambient assumed) insulated with 25mm foamglass (k = 0.08W/m/K) lost about 40W.
Even 1m completely uninsulated piping (pumps, valves etc) at 40degC only lost about 30W.
So in theory, 200W should cover all external heat losses for a typical house (even with dodgy insulation) - not negligible, but not hugely significant either.
I looked into this a bit - it seems Openmodelica would be perfectly suited to model this and also more complex systems. There are a few examples with heating included.
I noticed this on my boiler driven tank years ago. The solution is to put a in diverter/dump valve just before the tank that only lets the flow into the tank when it reaches a certain temperature and until then sends it back to the boiler.
I always wonder why a valve is not used to blend the Return using the Flow so it is always closer to the required output temp less the DT. So the Flow would reach its target temperature quickly, but stay there for longer. If you only wanted half a tank at 50°C you would manage that quicker.
Surely that isn’t correct as you do not instantly jump to the 50°C output. Actually, you heat the same bit of water multiple times (system volume/flow rate = time to circulate total volume).
It’s a simplified thought experiment to try and clarify the concept. You could go the next step up from that simplified concept and imagine successive layers of slightly hotter water move around the circuit and then closer together until you get to the point where it’s continuous.
I’d caution against comparing only hot water cycles, as there may be other factors that affect the system efficiency.
For example, my system (#1) fits a carnot factor of 0.5 when space heating, yet can barely manage 0.4 during DHW. Not sure why, maybe it’s the plate heat exchange, or the different pump it uses, or maybe higher primary pump speed?
This is very good observation you made. I’m monitoring the energy consumption of large relatively old hospital here in Sofia and in the summer period heat losses from poorly insulated DHW network are above 30% and sometimes with low hot water consumption they go over 50%…My own 200 liters hot water tank is situated at the attic and from my calculations last winter heat losses were in the range of 400 to 600 W/hour depending on outdoor temperatures and temperature of water in the tank although I have installed additional heat insulation of 10cm of glasswool around the tank.
I was able to measure the “losses” of a DHW run in the primary pipes, i.e. the energy stored in them after the DHW run finishes. They get used in the cold season for heating or are simply lost in summer. My approach was this:
I have sensors on flow and return lines right after the hydraulic station on my Vaillant, which is where the 3-way-valve is located. Outdoor temperature was above the heating threshold, so there was no heating demand. I let the DHW cycle start and finish. At the end of the DHW run, I raised the heating threshold temperature so the heating would turn on. However, this does not turn on the compressor at the same time but just circulates the hot water contained in the primary pipes into the CH circuit - right past my sensors.
This would not work with sensors right at the heat pump, as these would not fully “enclose” the stored hot water. The sensors right af where the pipes enter my house register around 1/3 of that, the rest is stored within the pipes after those sensors.
Integrating that power peak yields around 0.75 kWh - around 12% of the total heat delivered during that DHW cycle. In other words, around 16 liters of water could be heated from room temp to 58°C (mean between flow & return temperature at the end of the cycle). Note that this does not include the energy stored in the primary pipes on the DHW side of the 3-way-valve. I guess I could calculate that based off the exact length of pipework, but I guesstimate the overall loss to be 20%, which sounds reasonable.
Resolution of the data isn’t as good, but the numbers for “lost” heat are almost identical.
Primary loop is about 20-something meters, most of which is between heat meter sensors.
I do wish the heatpump would just keep the primary pump running until the dT had dropped.
