The heat loss from your home is dependent upon the temperature difference between inside and outside. This can vary quite considerably throughout the day.

Maximum efficiency is normally achieved when the heat pump is supplying just the correct quantity of heat energy, at the lowest LWT, to meet the heat loss. In most systems the LWT is set as a fixed value or by means of weather compensation, with a few offering indoor temperature modulated control.

Let's consider what happens in weather compensation mode if the WC curve is a little too high. The vast majority of heat pumps are designed to operate with a DeltaT of 5C between LWT and RWT.

If the required LWT is say 35C, with a corresponding RWT of 30C, at a water flow rate of say 25 litres per minute, this will meet the present heat loss and keep the indoor temperature at approximately 20C. But if the actual LWT is 38C, with a RWT of 33C, and water flow rate still at 25 litres per minute, the heat emitters will produce more heating than required, and the indoor temperature will start to increase. As the indoor temperature increases, the difference between the heat emitters temperature and the indoor air temperature will decrease, so the quantity of heat energy actually transferred from the heat emitters will start to decrease. There is too much heat energy going to the heat emitters, but the heat emitters cannot dissipate all of this heat energy into the rooms, so the RWT will start to increase.

There are two ways in which the system could be re-balanced, one being to lower the LWT, but this is fixed by the WC curve, the other method is to lower the water flow rate, so that there is less heat energy being transferred each minute.

I'm afraid the statement below is not fully correct. 

My understanding is to maxi mise efficiency I need to minimise flow temperature, which means maximising flow and minimising delta T.

Lowering the LWT should indeed improve efficiency, but this should be at the minimum permissible water flow rate, at the designed DeltaT of 5C.

I appreciate that this is a difficult concept to understand, and equally difficult to explain, since there are quite a number of different parameters which are continually balancing and re-balancing.

The important thing is to adjust the WC curve to match the heat loss of your home as closely as possible. The heat pump controller will then be in a better position to control the LWT, and also vary the speed of the water pump to maintain a DeltaT of 5C. If the heat pump controller cannot actually control the water pump, then the DeltaT could vary outside the designed limits.

Sorry if everyone’s read this but on the question of deltaT this post says it all….

Nice one Derek

My Ecodan isn’t the preplumbed variety so the primary pumps are just connected via a power cable to the FTC board. Therefore a fixed speed with no PWM. 

I’m wondering if using different pumps wired in differently would allow them to modulate rather than being fixed speed and therefore allow the system to have more control over the DT

Anyone with experience of Ecodan know if this is possible when using 3rd party pumps?

@benseb I think I read some earlier posts of your installation. The biggest problem I recall was the long line of rads off a 15mm pipe you also have some microbore I thought? So there’s another side to the deltaT equation - can the pipe circuit carry enough heat freely to all the heat emitters so that the return pipes come back cool enough?

I have plastic piping, I have micro bore but when I installed it all I followed 2 simple rules 1. No more than 2 large rads per 15mm and 2. Only one rad per 10mm microbore and make sure it’s no longer than 5 metres and copper.

all other pipes were 22mm plastic.

When we received our HP installation it had a lousy COP of 1.49 we regulated our pump speed down to 12 or 13 LPM (which is within the pump range) which gave us a deltaT of between 5 and 8c. We now have a COP of 3 at -4c to 3.75 at+4 over the past 2 weeks. We are still looking at some improvements because we’ve got a low loss header fitted which might be a drain on our COP but, One step at a time…

Do you have a picture of your pump? Many have push button speed controls or older ones sometimes have sliding speed controls. And have you managed to do anything about any of your pipework?

Posted by: @sunandair

↑

@benseb I think I read some earlier posts of your installation. The biggest problem I recall was the long line of rads off a 15mm pipe you also have some microbore I thought? So there’s another side to the deltaT equation - can the pipe circuit carry enough heat freely to all the heat emitters so that the return pipes come back cool enough?

I have plastic piping, I have micro bore but when I installed it all I followed 2 simple rules 1. No more than 2 large rads per 15mm and 2. Only one rad per 10mm microbore and make sure it’s no longer than 5 metres and copper.

all other pipes were 22mm plastic.

When we received our HP installation it had a lousy COP of 1.49 we regulated our pump speed down to 12 or 13 LPM (which is within the pump range) which gave us a deltaT of between 5 and 8c. We now have a COP of 3 at -4c to 3.75 at+4 over the past 2 weeks. We are still looking at some improvements because we’ve got a low loss header fitted which might be a drain on our COP but, One step at a time…

Do you have a picture of your pump? Many have push button speed controls or older ones sometimes have sliding speed controls. And have you managed to do anything about any of your pipework?

 

 

Good one SUNandAir, I forgot to consider pipe sizes and lengths, which can also have a profound affect on the overall efficiency of a heat pump system.

So the 'Holy Grail' would appear to be obtaining the lowest LWT, with a low water flow rate, whilst achieving a DeltaT of around 5C.

@derek-m Perhaps not necessarily a delta T of 5 deg C. I’ve seen @heacol state that 8 deg C delta T is most efficient. My ASHP is most efficient, although consuming more, in the first 30 minutes of starting up, which is the time when delta T is at its highest. This is from looking at the passivuk graph of consumption and production on MMSP. 

@sunandair yes good shout. 

we’re planning on retrofitting the two rooms which are at either end of the housr

i think lounge will get UFH so we can run a dedicated 22mm to that. Then the 2 bedroom rads which shared the 15mm with the lounge will have more flow. 

similarly with the utility we don’t want UFH due to cost but will likely try to run a 22mm to it. Should help a lot I hope. 

Posted by: @mike-h

↑

Perhaps not necessarily a delta T of 5 deg C. I’ve seen @heacol state that 8 deg C delta T is most efficient.

I am still trying to get my head around this.  There is no deltaT parameter in the carnot equation so what specific factors make a dT of 8 degC optimal?

Given the RWT has a lower limit, take a simple example of UFH.  Your RWT can never be less than the temperature of the floor (in heating mode), and is likely to be a couple degrees above it.  So assuming you have a minimum RWT, then the only way to reduce the LWT is to reduce the deltaT. 

So if a house with UFH has a min achievable RWT of 27, I can't see how a deltaT of 8 would be more efficient than one of 3 since the LWT for the latter would be 30 vs 35 for the former.

Posted by: @william1066

↑

So assuming you have a minimum RWT, then the only way to reduce the LWT is to reduce the deltaT. 

I don’t see it as the only way.... how about lowering the ambitious set flow temperature? So that it matches the radiant capabilities of an inefficient emitter - the concrete slab. If your slab cannot release the heat quick enough to get a good deltaT then you are wasting energy heating it up above a certain temperature.

it must be quite difficult to balance the LWT temperature to the heat loss of the slab And guess that might be the unique challenge of under floor heating. But from a purly theoretical perspective, adding more heat when it cannot be released doesn’t sound like a good strategy.

I guess you need the weather compensation curve to set the flow temperature within the limit of the radiant floor but I don’t know if There is ever a situation that you cannot reach a comfortable room temperature without over heating the floor area.

PS I don’t have UFH so I might be talking Total nonsense apologies if I am. 

Posted by: @mike-h

↑

@derek-m Perhaps not necessarily a delta T of 5 deg C. I’ve seen @heacol state that 8 deg C delta T is most efficient. My ASHP is most efficient, although consuming more, in the first 30 minutes of starting up, which is the time when delta T is at its highest. This is from looking at the passivuk graph of consumption and production on MMSP. 

The most efficient operating parameters will undoubtedly vary from system to system, and will also probably vary on the same system as the weather conditions change.

An increased DeltaT will supply more heat energy at the same flow rate, so to maintain a constant supply of heat energy would require a reduced flow rate. At the heat emitter end of the equation, the quantity of heat energy being absorbed will be dependent upon the average water temperature, so an increased DeltaT would involve both lowering the RWT and increasing the LWT. As I think we have established by now, increasing the LWT reduces the efficiency of a heat pump.

In practice the system is performing a continuous balancing act, which is in fact dependent to a great extent upon the equipment installed, how it has been configured to operate, and the programming of the control algorithms.

In systems that do not have water pump speed control, the DeltaT will vary with changes in LWT, and the quantity of heat energy required. In systems that do have water pump speed control, the DeltaT should be more stable, but the water flow rate will vary. This could have detrimental effects if the system has hydronic seperation, and hence multiple water pumps, the speeds of which may, or may not, be controlled.

At the end of the day it is what works best for the homeowner, the problem being establishing what IS best, under varying operating conditions. 🙄 

Posted by: @william1066

↑

Posted by: @mike-h

↑

Perhaps not necessarily a delta T of 5 deg C. I’ve seen @heacol state that 8 deg C delta T is most efficient.

I am still trying to get my head around this.  There is no deltaT parameter in the carnot equation so what specific factors make a dT of 8 degC optimal?

One relevant factor will be that the purely DT based maths doesn't allow for the kinetic energy required to move the water i.e. the work the pump(s) is/are having to do. 16kw DT5 means 0.75 l/sec. in 28mm pipe this is 1.41 m/sec velocity. most rule-of-thumbs I've read say don't go over 0.9m/sec. 16kw DT3 is 2.35m/sec velocity in 28mm pipe which is seriously high. head loss, which is what determines how hard the pump has to work, comes from the hazen-williams equation, which is proportional to flowrate^1.852. If you double your DT, you're halving your flow rate, and so the head required drops by a factor of 3.61 [ 1 / (2^1.852)]. halve your dT, the reverse happens, your head requirement (and thus pumping energy) goes up by factor of 3.61. So, a very low DT on a system with high output requirement (without the pipes having been super-sized) would only be able to heat the premises by pushing huge amounts of kinetic energy in through the pumps to overcome head loss. so for larger systems there'll be a trade off somewhere in the middle , finding a DT that the heat pump prefers vs a DT that the circulating pumps can do without being firehoses, and that might be more than 5. whereas for smaller systems where the flow rates remain inside the recommend flow velocities for "normal domestic heating pipe sizes" regardless of DT, I suspect it matters less. 

Posted by: @iancalderbank

↑

One relevant factor will be that the purely DT based maths doesn't allow for the kinetic energy required to move the water i.e. the work the pump(s) is/are having to do.

I have been wondering about this piece.  My view is that from an efficiency point of view it should be marginal, at least at the high end of the output (16kw)  My pump can vary from about 28watts to 153watts (it is behind a relay, btw).   At max power this is likely 2 to 3 % of the consumption of the heat pump.

I think that real world, yes a higher design dT is cheaper to install, easier to design for and install (smaller pipes, smaller pumps, bigger spacing in your UFH pipes so less pipes and less labour, smaller radiator etc) probably easier for the control system to manage etc.

If however you focus very narrowly just on efficiency, then I am still struggling to understand the "engineering maths" here.

I have only had this Daikin EDLA 8 kW ASHP going for a few days and it is very mild at present; the installers left me with the pump running at 50 degrees C and the circulating pump on full speed. Within hours, I started reducing the set point and eventually left it at 35 degrees as this left the bathroom at a cosy 24.5 - 25 degrees but other rooms were far too warm!. I have gradually being shutting some of the lock shield valves whilst monitoring the radiator temperatures and am now much closer to desired temperatures. My aim is to allow the heat pump set on WC to run 24/7 with no or very little intervention from TRV’s. (As per Heat Geek advice)

I think I am getting close in some rooms but some lock shield valves are only perhaps 1/8th turn open. I observed a Delta T of approx 3 degrees on the largest unrestricted radiators as best I could measure it with an infra red thermometer measuring inlet and outlet points on these radiators. I have been reducing the circulating pump speed in small steps, now much quieter than when at full speed - and the Delta T was less than a degree at full speed.

The wind has increased noticeably this afternoon and this seems to have elevated the water temperature a little and the living room rose from approx 23 degrees to 24 degrees so I have closed the lock shield valve a little more for now. I don’t think mild weather is the best situation for setting up such systems - but it is at least a start! My hope is to have the system run steadily and just set TRV’s about 2 degrees above required temperature as a control.

I have carried out a cop measurement based on what the MMI control console reports to me and including the very over-the-top setting at commissioning, so far the cop is 3.22. I am wondering how low I might set the circulating pump, based on the Wilo on-line manual, I still have some latitude to slow it down further but don’t want to risk overdoing this at this early stage, it is a lot quieter if a TRV is starting to control the flow through a radiator than when at full speed and, as I said the Delta T seems better.

Any advice or comments would be much appreciated please. I appreciate that the installers should (if they were keen) be carrying out a lot of this initial setting up but I appreciate that I have more time to finesse the finer points myself! BTW, the heat loss calculation indicated 6.5 kW and the water heating is totally separate and supplied by a Sunamp Thermino ePV210 (I have 8.1 kWp of solar PV and a Tesla Gateway and Powerwall battery.) Regards, Toodles