Why isn’t the return lower than the flow if the radiators are losing heat inside the house?
I wasn’t saying the heat is lost outside due to poor insulation, I think it is being lost to the outside as it flows through the heat pump which is outside.
Why isn’t the heat meter recording the heat going into the house?
If heat is going into the house and it is unrecorded then that must mean that the reported COP is lower than it actually is.
8.3lpm for 70 minutes (in the example) is 581 litres cooled down by 10c.
How much heat has gone unrecorded into the house?
If the system contains 100l then losing 10c in 70 minutes is about 1kWh of heat not recorded?
You would see return > flow if you measured directly at the radiator pipe connections. You basically measure whether heat is extracted or added between the flow/return measurement points. At the normal heat meter location the water going out through the return measurement location is essentially “short circuited” to the flow temperature location - there is typically very little pipe volume between the two and the flow and return temperatures measure the same value because nothing has happened to the water on the shirt way between those two points.
Now why could you measure something directly at the radiator? The difference is that you have a whole lot of system volume behind the radiator. Warm water comes in, is slightly cooled and after it leaves the radiator mixes with the remainder of the system volume to a temperature warmer than the radiator return temperature but overall the total temperature of the water in the system cools down. Basically the flow temperature coming into the radiator is the temperature of the whole system reservoir while the return from the radiator is just a fraction of that volume having cooled down by going through the radiator.
I have only wrapped my head around this very recently and the key is not just looking at it from the flow/return temperature perspective but also taking the enclosed volume into account. I hope that makes at least a bit of sense .
The heat meter recorded all of it. It records heat energy going into the water volume. During the off-cycle this heat is being distributed from the water to the house via the radiators. In case of cycling, the heat transfer from water → house is slower than the transfer heat pump → water, which is why you have the cycling in the first place. At first the heat pump dumps all its energy into the water and then the water gradually cools down during the off phase and releases this stored energy. Quite similar to a stone oven that you heat up with a fire and that still heats the house long after the fire is out
Thinking things through does the following make sense?
Ideally the capacity for heat transfer water → house ≥ pump → water as then you don’t get cycling.
This isn’t the case when radiators aren’t large enough or the heat loss is low (mild weather) so water → house < pump → water.
If I add a volumiser in that situation I would be increasing the capacity for the water to keep the pump → water heat transfer going for longer.
The cycle would still eventually have to end as the water can’t get rid of the heat to the house at the same rate but after a longer time.
However the off time would also be prolonged (assuming the pump is still on) while the water → house transfer continues for longer as there is more heat in the water.
I now think I’m satisfied that there would definitely be a benefit in this in reducing cycling but I still can’t get my head around what effect, if any, this would have on efficiency.
I think your line of thought is sound. I’m in basically this situation as my heatpump is oversized but I also have a 100l volumizer. Longer on-cycles means less impact of transient inefficiency when the compressor is just turned on. I think it really depends on how the heat pump handles the cycling. Vaillant is really good there and the COP I get matches their datasheet values for different outdoor/flow temperature quite well.
I’ve also been scratching my head with this recently, I thought I would add system volume and minimum modulation to the dynamic heat pump simulator tool that I started building a while ago, just got these added this morning, here’s the simulator: https://openenergymonitor.org/tools/dynamic_heatpump_v1.html.
The results suggest very little impact on performance from system volume, but this assumes that:
The heat pump has no delay in getting stable efficiency out of the compressor/fridge circuit. Something we dont see in the real world with certain heat pump models e.g the Daikin 9-16 kW range.
The heat pump does not over-ramp up at the start in order to hit a higher flow temp target. If compressor comes on during cycling at minimum modulation it will spend more time at lower flow temperatures than if the compressor goes to max in order to hit a set point target…
Im starting to think that beyond a certain minimum volume, the benefit of further volume in terms of performance is very much make/model of heat pump specific and how well the internal control algorithms and hardware works…
I wish Vaillant didn’t do this over-ramp. I had exactly the same thought as you - I could have longer cycles without that particular behavior. It definitely impacts COP of shorter cycles because during half the cycle the compressor runs harder than it needs to.
Yes I guess if my heat pump decides because the volume is taking longer to heat up it will just throw more power at it then that could defeat the purpose.
Although I wonder if the Ecodan will learn that this isn’t required as long as the room temperature remains stable if it does it less aggressively.
I’ve found that Auto Adapt is pretty good at ramping up slowly, whereas in Flow or Curve modes it would race to get up to target as quickly as possible before modulating down.
Nerd alert : If you really want to get stuck into the mathematics of heat pump cycling frequency, you can roughly estimate it from heat balance as follows.
If your heat pump is switched on and off through weather compensation, define
QHP = instantaneous heat pump duty (W) (e.g. from your manufacturer’s operating table for given LWT and ambient temperature)
Qemitters = instantaneous effective total emitter duty (e.g. rated from the vendor design duty at design dT using the well known 1.3 exponent equation)
(m.Cp)water = effective thermal inertia of circulating fluid (including containment system) (J/degC)
Then heat pump cycle time = “on” time + “off” time where
“on” time = (m.Cp)water x controller hysteresis (degC) / (QHP – Qemitters)
“off” time = (m.Cp)water x controller hysteresis (degC) / Qemitters
If your heat pump is switched on and off solely through roomstat operation, where (in addition) define
Qloss = house heat loss (W) (e.g. prorated from your installer’s calcs at design ambient temperature)
(m.Cp)house = thermal inertia of house fabric (walls, furniture etc.) (J/degC)
“on” time = (m.Cp)house x roomstat hysteresis (degC) / (Qemitters – Qloss)
“off” time = (m.Cp)house x roomstat hysteresis (degC) / Qloss
You can estimate (m.Cp)water from its initial warm-up rate with known QHP, and (m.Cp)house from overnight house cooling rate (with zero heat pump input).
My circulation fluid containment system (pipes, radiators, buffer tank) contributes ~25% extra thermal inertia above the water, so not insignificant).
By “effective”, just include those emitters and water actually in use (i.e. exclude TRV-isolated ones).
Do you feel a spreadsheet coming on… ?
Edit: The above may be conservative as it assumes that your heat pump operates flat out until WC or roomstat targets are reached - the more advanced HP controllers will anticipate setpoint by reducing output as it is approached, thus extending cycle times somewhat…
There is just so many factors involved I can both see how adding volume might not make any difference or it could make a significant difference and a lot of it might come down to how the controls react to it rather than theoretically what should happen.
If they weren’t so expensive or I could pick up a cheap 2nd hand one I’d give it a go and see what happens.
We’re working on a project in Oxford and we’ve used volumisers because no matter how much you advise people not to turn off radiators, they will turn off radiators. So you can’t count radiator volume of any radiator with a TRV in the minimum volume.
The Ecodan controller keeps the pump on between cycles I assume to both monitor the flow temperature when deciding when to bring the compressor back on but also to allow the heat in the water to be emitted more effectively.
.
: If you really want to get stuck into the mathematics of heat pump cycling frequency, you can roughly estimate it from heat balance as follows.
?
