@bontwoody

I also don't claim to be a heat pump engineer, but we probably know as much, if not more, about heat pumps than many of those who do claim to be heat pump engineers.

I feel it is also good that we can explore the pros and cons relating to heat pump sizing and optimal operation.

I have suggested that a heat pump should be sized at 1.25 to 1.5 times an 'accurate' heat loss calculation for two basic reasons.

1) Because not all heat pumps are created equal, with some maintaining their rated output at low temperatures, while others struggle.

2) To provide the extra capacity required after defrost cycles and DHW heating, both of which require an extra boost to the output to recover the thermal energy lost.

A further reason would be if owners are contemplating an overnight temperature setback, particularly during colder weather conditions, which again would require some extra capacity being available to recover the IAT within an acceptable timescale.

As has been suggested previously, prospective heat pump purchasers should carry out some tests with their present fossil fuel heating system, if at all possible. If possible lower the LWT setting to see how the present heat emitters cope with a lower water temperature. Monitor the energy consumption under differing weather conditions to assess the probable heat loss and thermal capacity of their home.

Gather information of the probable average OAT in their area and also the extremes that may occur.

Question the need for such items as a buffer tank or plate heat exchanger, and if absolutely necessary then make allowance for these in the sizing of the heat pump.

I suspect that confusion will still prevail, since there is no simple formula that can be applied to all properties.

A larger heat pump may be more prone to cycling during milder weather conditions, which may or may not lead to higher energy consumption. It could be that cycling reduces the need for defrost cycles and therefore actually reduces energy consumption. There are also possible techniques that can be employed to reduce the frequency of cycling, such as increasing the WC curve offset by 1C or 2C, and then using an on - off type thermostat to start and stop the heat pump. A thermostat with an operating deadband of say 1C, such that it starts the heat pump when the IAT falls to 20C and stops the heat pump when the IAT increases to 21C. Obviously such an arrangement would need testing on individual systems, but I feel it should reduce the possibility of cycling, and may even reduce overall energy consumption, since it is possible that the heat pump would be stopped for lengthy periods of time.

@jamespa

What was even more interesting was the article stating that Ofgem are at last tightening the rules around the balancing mechanism, to try to prevent some of the generating companies from manipulating the market to make excessive profits.

Posted by: @derek-m

↑

I have suggested that a heat pump should be sized at 1.25 to 1.5 times an 'accurate' heat loss calculation for two basic reasons.

1) Because not all heat pumps are created equal, with some maintaining their rated output at low temperatures, while others struggle.

Equally some achieve greater output than their name plate rating.  Choice of model should be based on the capacity table and design min OAT, not an arbitrary uplift to adjust for the worst vagaries of manufactures approach to naming.  You can still over or undersize if you choose, but based on actual output at your design OAT (taken from the capacity table) not nameplate output.

Posted by: @derek-m

↑

A larger heat pump may be more prone to cycling during milder weather conditions, which may or may not lead to higher energy consumption. It could be that cycling reduces the need for defrost cycles and therefore actually reduces energy consumption. 

If cycling is occurring when defrost cycles are running, the heat pump is well oversized at least for much of the UK so I'm not sure this argument is valid in most UK cases.

Posted by: @derek-m

↑

There are also possible techniques that can be employed to reduce the frequency of cycling, such as increasing the WC curve offset by 1C or 2C, and then using an on - off type thermostat to start and stop the heat pump. A thermostat with an operating deadband of say 1C, such that it starts the heat pump when the IAT falls to 20C and stops the heat pump when the IAT increases to 21C. Obviously such an arrangement would need testing on individual systems, but I feel it should reduce the possibility of cycling, and may even reduce overall energy consumption, since it is possible that the heat pump would be stopped for lengthy periods of time.

Whilst it's indeed possible to reduce the cycle frequency, these techniques won't change the on/off ratio.  I think therefore that you maybe ignoring the basic thermodynamics.  Even if there are no energy losses due to cycling, its still less efficient because, to deliver the same amount of energy from the emitters the flow temperature must be higher.  Extending the cycle time reduces wear and tear and start up losses, but does not mitigate the loss of thermodynamic efficiency.  A heat pump which is cycling at 7C with a 50% on off ratio and radiators with a design ft of 45C will will be very roughly 20% less efficient thermodynamically than one which is running 100% of the time.

Posted by: @jamespa

↑

 

If cycling is occurring when defrost cycles are running, the heat pump is well oversized at least for much of the UK so I'm not sure this argument is valid in most UK cases.

 

Our A2A heat pump still performs defrost cycles when the OAT is quite mild, it is more to do with humidity rather than OAT.

Whilst it's indeed possible to reduce the cycle frequency, these techniques won't change the on/off ratio.  I think therefore that you maybe ignoring the basic thermodynamics.  Even if there are no energy losses due to cycling, its still less efficient because, to deliver the same amount of energy from the emitters the flow temperature must be higher.  Extending the cycle time reduces wear and tear and start up losses, but does not mitigate the loss of thermodynamic efficiency.  

Raising the WC curve will indeed change the on/off ratio and will reduce the likelihood of cycling.

If that is the case, why are some bothering with overnight shutdown, which is basically the same thing. The quantity of energy difference during milder weather conditions I feel will be minimal, my objective was more to reduce the cycling frequency which many owners find of concern.

The other unknown is 'how much of the electrical energy input is actually being converted into thermal energy output'? Any losses will be greater while the heat pump is actually running, so does a larger heat pump use more electrical energy running part of the time, as against a smaller heat pump running constantly, both producing the same thermal energy output, or is it the other way around?

Posted by: @derek-m

↑

Raising the WC curve will indeed change the on/off ratio and will reduce the likelihood of cycling.

Yes but at the expense of running at a higher flow temperature (and thus lower efficiency) than is necessary to achieve the power output required from the emitters.   This will in fact make things even worse than keeping it as low as possible to achieve the required room temperature.

Posted by: @derek-m

↑

If that is the case, why are some bothering with overnight shutdown, which is basically the same thing. 

Because they are happy for the room to be colder overnight in return possibly for saving some energy. 

(the reason people argue against nighttime set back, even though the user is happy with the temperature drop, is because it isn't obvious that it saves energy.)

Posted by: @derek-m

↑

The quantity of energy difference during milder weather conditions I feel will be minimal, my objective was more to reduce the cycling frequency which many owners find of concern.

Sorry but you are still ignoring the basic thermodynamics.  If I want to maintain a given indoor temperature at a given OAT I need to deliver a certain amount of energy from the emitters to the house, let's say 5kW if I heat 100% of the time.  But if I am only heating 50% of the time I will need to heat at 10kW to deliver the same amount of energy on average.  To do that the delta t between my emitters and the room will have to be higher, which can only be the case if the flow temperature is higher.  For radiators where the emitter to room deltaT is typically about 20C, that means the flow temp will need to be roughly 10C higher, which means the heat pump will be roughly 20% less efficient.  It's all in the Carnot equation.

Posted by: @derek-m

↑

'how much of the electrical energy input is actually being converted into thermal energy output'? Any losses will be greater while the heat pump is actually running, so does a larger heat pump use more electrical energy running part of the time, as against a smaller heat pump running constantly, both producing the same thermal energy output, or is it the other way around?

Hopefully you get roughly 3 times (between 2 and 5 depending on OAT and ft) as much energy out as you put in, that's the whole point of a heat pump.  It's not about 'losses', it's about the relative efficiency of the _gains_ due to the fundamental properties of a heat pump.

Posted by: @derek-m

↑

my objective was more to reduce the cycling frequency which many owners find of concern.

I'm not sure reducing the cycling _frequency_ helps, unless it is initially very high ('short cycling').  In fact I have a feeling it makes things worse for any given required on/off ratio (which of course is determined by hp min output vs heat loss from house)

Sorry but the fundamentals of the thermodynamics really do matter in the argument over sizing.  A cycling heat pump, even if it is perfectly engineered with lossless components, MUST be less efficient than one on constantly (for any given OAT and target indoor temperature combination).  Of course it is _possible_ that the engineering effects outweigh the unavoidable thermodynamic variations in efficiency, but in the absence of evidence to support this it can be assumed that the thermodynamics must be the primary driver, an assumption which the capacity tables you use in your model do not contradict.

@jamespa I hope you will eventually get your own toy to play with in real time, so you can come to a conclusion based on facts rather than virtual data.

There is a difference in say overnight off period and the subsequent period of catch up as you put it in the example of 5kW continuous load and 10kW catch up.

Versus

A heat pump that is cycling because the heat load is 1kW but the 5kW sized has a minimum 2kW output.

If you want a heat pump not to cycle at low load, bellow minimum output, you need not let it come on automatically but control it to be 1h on 1h off or something similar, thus there's no difference in flow temp to give your theory a base.

In regards to cycling efficiency, data from heatpumpmonitor  doesn't show a tragic loss of efficiency, but in my view a tragic unnecessarily barrage of cycles of start/stop by letting heat pumps do it automatically.

Starting with Glyn's 5kW Samsung, data shows that after the initial heat up longer run, eventually the cycling starts if allowed to. Zooming in to see the COP from the long run vs cycling the COP doesn't differ that much, as it is being pictured on the forums.

Second 5kW Samsung it's a new system on the site not much past data but enough to show a similar picture of the COP, even with berserk cycling.

A side by side of the two systems

https://emoncms.org/app/view?name=Panasonic&readkey=d64d47045794fd42f3a7b9c0b0fa4019

@Derek M @JamesPa Thanks both for an interesting and illuminating discussion. By way of a practical example Ive attached my heat pump graph for yesterday.

Im happy with things up to 14:22 but then we start to see cycling getting particularly frequent about 17:00. Should I reduce the lower value of my WC curve to lower the heat input to my house in the shoulder season? My delta T is also somewhat small at about 3C, is this due to my emitters being undersized? (I did reuse radiators as much as possible to keep costs down, but also have an extension planned for next year which will increase the emitter volume)

My opinion is that oversizing is not a problem until you let the heat pump run in weather compensation/automatically when the load is bellow minimum.

Data on heatpumpmonitor of a 5kW vs 7kW Vaillant running in winter, when matching the loads ish, the 7kW defrosted less often whilst the 5kW was running flat out.

With the 7kW cycling from around 4C vs 5kW at 6-7C

Posted by: @bontwoody

↑

@Derek M @JamesPa Thanks both for an interesting and illuminating discussion. By way of a practical example Ive attached my heat pump graph for yesterday.

Im happy with things up to 14:22 but then we start to see cycling getting particularly frequent about 17:00. Should I reduce the lower value of my WC curve to lower the heat input to my house in the shoulder season? My delta T is also somewhat small at about 3C, is this due to my emitters being undersized? (I did reuse radiators as much as possible to keep costs down, but also have an extension planned for next year which will increase the emitter volume)

-- Attachment is not available --

I hope you don't mind I've just used your newly installed system as an example whilst you were posting your own post.

Initially I did question why you run at 38 flow, so yes, tweaking the WC to be at flow 30C when 15C outside should be a start.

You might very well not have enough radiator surface vs Glyn that runs it at 30, but then we don't know your WC vs his WC settings vs the rest of the house.

HeatGeek would say that the WC should be set when it's cold, and from the bottom up, when 2C outside set it such that the indoor temp doesn't increase and raise it until you become comfortable with the indoor temp.

In terms of your cycling, you need to find the lowest flow temp possible for your house, and then if it still cycling, but less often, you need run it without WC when warm outside. I mean whatever we say, try and see what works for you.

So yours yesterday was running for 11 hours at minimum, what was your indoor temp doing, was it steady, any signs of increasing or decreasing?

Your delta T is normal vs what I can see 2C on Glyn, are you aware that you water flow rate is variable? and Glyn is fixed, such that if yours didn't reduce you would have the same delta T 2C ish as Glyn, more or less. The Dt is driven by the heat loss of the radiators, so if in time are able to replace a few/swap them around then that's better, as it's never too much.

@fazel I dont mind at all! All help greatfully accepted 🙂 Is that Trystan? I havent had a cold day yet, so I will try and lodge the task in my brain for when we do.

I will adjust my WC down a bit now and monitor the results.

I was aware that I have a variable speed pump as I bought the wrong model initially and asked Glyn whether he thought it would be worth swapping. In hindsight, I think it was and I managed to recoup £50 by selling the old one. As you say I noticed the delta T increase after fitting because the flow rate dropped.

I will probably hold off on the radiators until I do the extension and see how that affects things.

thanks for the input!

@fazel Just noticed your question on indoor temperature. I have the thermostat on the controller set to 23C in the hall area (we do like it warm, hence my initial high flow temp) but it was starting to get too warm towards the end of the night and the missus made me turn it off 🙂

Posted by: @bontwoody

↑

 Is that Trystan?

no, sorry, random internet guy