Posted by: @iantelescope

↑

Could I improve the performance of my Heat Pump by fitting Anti-freezing Valves to replace the Glycol .

The short answer, on paper, is yes, but it is not that simple. In practice, glycol is simpler, whatever happens, as long as it is there, and the temps don't get absurdly cold, it just does its thing, job done. It is also, by the way, a (positive) reason for having a heat exchanger, which separates the glycol circuit from the rad circuit. Anti-freeze valves basically rely (at least in principle) on freezing the balls off a brass monkey, and as they are mechanical, they might fail. I have also still yet to find a satisfactory account of what happens after the balls pop off the brass monkey. Presumably the heat pump senses it has taken a leak, and shuts down, but what then? There is still fluid in the system, and it is both static, and does not have antifreeze in it. Might in due course other more expensive balls start popping off other brass monkeys inside the heat pump itself? Assuming that doesn't happen, what has to happen to get the system up and running again? What happens if there is no one at home? Too many unanswered questions for me at the moment to go down the brass monkey route...

Posted by: @iantelescope

↑

Could I improve the performance of my Heat Pump by fitting Anti-freezing Valves to replace the Glycol .

Glycol reduces the Specific Heat of Water from 4.3 to 4.0 , a reduction of 7.5 %.

I would do the fitting myself.

The change is specific heat capacity shouldn't directly affect the performance of the HP.  It simply means that the fluid has to flow about 10% faster through the system filled with glycol than if it is filled with water.  This does of course put a bit more strain on the pumps, but not materially so.

However the use of glycol externally is often the reason to fit a plate heat exchanger (so you don't have to fill the whole system with glycol).  If your plate heat exchanger is plumbed correctly, whilst it will reduce performance a bit, it sholdn't reduce it by more than a few %.  Remember 2% for every degree lost between HP and radiator and this should be no more than 5C possibly less, with a correctly plumbed system).

My very strong advice is to focus first on the major issue, which is the buffer tank bypassing the emitter circuit.  As you, for very good reasons, don't want to replumb at present, then just switching it out of circuit and ensuring engaged sufficient system volume by opening up some TRVs is the first step.  Until you have done, and observed the effects of, this, tweaking, or even worrying about, anything else is (IMHO) pointless.

Posted by: @iantelescope

↑

However, the NIC are saying that " my Installer should be given another chance to fix the problems."

Unfortunately I don't think you will get a choice over this unless you disengage completely with the contract.  The remedy for a civil tort is damages, in this case the 'damages' are the effects of the system being plumbed incorrectly, and the simple and lowest cost fix is to plumb it correctly (and possibly compensate you for the extra costs you have incurred).  BTW if you take someone to court suing for damages, you are obliged to make reasonable efforts to minimise those damages. 

However I can see no reason not to demand that the installer first explain in writing what he proposes to do, that he gives you time to run it past both Samsung and the members of this forum, and then he does that and that only unless you give express permission.  I would seek to persuade NIC that, in view of the history, this is in the interests of all parties.

Posted by: @iantelescope

↑

@derek-m 

Hi Derek ,

All of my Radiators, with ONE exception, were replaced during the installation.

Also, All of the pipes were replaced with a 22 mm Bus Backbone running the length of the house.

Finally, all of the new radiators were connected by 15mm pipes to the 22mm Bus Backbone.

So, I think that you and James are correct...............................the Black ink with the magnetic particles may be coming from the sole veteran Radiator.

However, I am also suspicious about the condition of the 50 l "Volumising tank" being covered in warnings about "annual service ".

What "Maintenance" does a 50 L Tank need?

Many thanks for your patience ............

ian

 

 

 

When water is open to the atmosphere the various gases within the atmosphere become dissolved into the water in a ratio of 35:1. Since the normal atmosphere contains approximately 20% Oxygen, over time approximately 0.57% Oxygen will become dissolved into the water.

When an unvented central heating system is filled with water, even after all the air has been removed, there will still be dissolved Oxygen present. This Oxygen reacts with some of the metals inside the heating system and causes oxidisation to occur.

There are chemicals that can be added to the water that help reduce oxidisation, but if the system is drained and flushed, these chemicals would need to be added once more.

With regards to the Volumiser, I would suggest Googling the manufacturer to see from what materials it has been manufactured, and what actions may be necessary to prevent oxidisation or corrosion.

Posted by: @iantelescope

↑

What "Maintenance" does a 50 L Tank need?

or, more to the point, what maintenance is it possible to do to a 50l tank?  Check the connections aren't leaking I suppose!

Posted by: @jamespa

↑

The change is specific heat capacity shouldn't directly affect the performance of the HP.  It simply means that the fluid has to flow about 10% faster through the system filled with glycol than if it is filled with water.  This does of course put a bit more strain on the pumps, but not materially so.

But does the heat pump know it has glycol in it? Isn't output normally controlled by the weather compensation curve (water law in Samsung) - if it is this temp outside, then the LWT is this? And this is achieved by altering the compressor output? Recall that at least on my Midea the flow rate is pretty constant. I am not sure whether this is normal for Midea, or just a quirk in my particular system. The bottom line is then that the heat pump itself (compressor not the circulating pump) then has to work harder, and that means a performance hit? I ask, because the general assumption seems to be that given the energy out equation (specific heat x flow rate x delta t), if the specific heat drops, then energy out drops (in proportion) - and this, apart from cost, is the main reason not to have glycol. If it doesn't incur a performance hit, then why not use glycol, and forget about brass monkeys? 

Posted by: @cathoderay

↑

Posted by: @jamespa

↑

The change is specific heat capacity shouldn't directly affect the performance of the HP.  It simply means that the fluid has to flow about 10% faster through the system filled with glycol than if it is filled with water.  This does of course put a bit more strain on the pumps, but not materially so.

But does the heat pump know it has glycol in it? Isn't output normally controlled by the weather compensation curve (water law in Samsung) - if it is this temp outside, then the LWT is this? And this is achieved by altering the compressor output? Recall that at least on my Midea the flow rate is pretty constant. I am not sure whether this is normal for Midea, or just a quirk in my particular system. The bottom line is then that the heat pump itself (compressor not the circulating pump) then has to work harder, and that means a performance hit? I ask, because the general assumption seems to be that given the energy out equation (specific heat x flow rate x delta t), if the specific heat drops, then energy out drops (in proportion) - and this, apart from cost, is the main reason not to have glycol. If it doesn't incur a performance hit, then why not use glycol, and forget about brass monkeys? 

 

I don't see why it has to work harder.  The energy transferred from house to outside is unaffected by the flow liquid.

The energy transferred from rads to room, which must match this, is determined by delta t rad to room, and is independent of shc.  So the WC curve is not affected.

The energy transferred from hp to rads must match the energy transferred from rads to room, and is determined by flow rate, delta T (flow to return) and specific heat capacity.  Flow rate will need to be a bit higher so the water pump has to work harder, but it's power consumption is negligible compared to the compressor which, since it is delivering the same amount of energy at the same temperature, will have the same work to do.

As long as the flow rate is adjusted for roughly the design delta t flow return, the compressor will work the same.  Small differences will sort themselves out as the hp adjusts it's output to match demand.

Some HP's do ask you to tell them whether there is water or glycol so that they can calculate delivered energy from flow rate.

As far as I know some hps control flow rate and some don't.  If flow rate isn't controlled then delta t flow-return will inevitably vary a bit as a function of load.  Actually this helps a bit with the WC curve for radiators, and means that a linear WC curve gives a closer match to the ideal WC curve than would otherwise be the case.  But ultimately the heat delivered by the hp must match the load and the efficiency with which it delivers it is, to first order, dependent on OAT and flow temp not flow rate or small changes in shc.

@cathoderay 

Anti-Freeze valves ?

What does happen  after the Anti-Freeze valve fires ?..................................

Excellent Question!

Does the Heat Pump shut down when the Anti-Freezing Valve activates?

I will , for the moment ....stick with Glycol.

ian

@jamespa 

Hi James,

My system has two Two Water Power control options :

Water Power Control by varying the Water Flow Rate using the Grundfos PWM motor socket.

OR 

Water Power Control by mains switching  the Grundfos motor Mains socket.

PWM Water Flow rate Control of Water Power

My system has Two Pulse width modulated, PWM, Grundfos Motors.

The PWM motor on the Heat Pump water circuit is controlled by a PWM signal from my Samsung Controller board.

This PWM signal is faulty , being unable to supply a signal to the PWM motor.

The Radiator Water circuit PWM motor socket has no connection to any source.

Both PWM motors work flawlessly when sourced from a "Arduino" .

OR

Mains Switching Control of Water Switching

In this power control option the Water power is Pulse Width controlled by the cycle time , typically turning ON for between 10 to 30 minutes.

Which should I use?

PWM  or Mains Switching ?

If PWM should I use

One PWM signal , the Samsung, when , and if repaired

OR

Two  PWM controls controlling the flow rates on both water circuits?

Modelling

Given that any given desired outcome for a heat pump system causes so many unwanted side effects :

Can we model , in arithmetic , a complete Heat Pump system?

Can we test our model against a physical system with pipe external measurements of Water  Temperature, Pressure and Water flow rate?

 Contracts

Again, you know much more than I about contracts , the law......et al.

I will take your advice and be more circumspect with respect to the NIC/MCS

ian

@jamespa - I am still stuck in an uncertainty loop about the effects of glycol on system performance. Starting at the beginning, why add glycol in the first place? To prevent frost damage, just as we add antifreeze to a car's cooling system. That much is clear, but then we soon come up against the received wisdom that adding antifreeze reduces performance, and so perhaps adding it is not such a good idea after all. The rationale for the alleged performance hit is in the energy transfer equation, kWh out = specific heat of the circulating fluid x flow rate x LWT/RWT delta t. Going back to goods train analogies, specific heat corresponds to wagon size, flow rate to how fast the trains run, and delta t to how much they off load. If any or all of these factors decrease, and nothing else changes (ie there is no compensation by way of the other factors), the total amount of energy delivered decreases. Thus, if flow rate and delta t remain the same (as by and large they should, and usually do, at least in my system, see below), then a system with added glycol must deliver less heat.

But is this a performance ie efficiency hit? Maybe the system remains at a constant efficiency level (lets say 300%) but instead it becomes less capable: instead of being able to deliver say 3kW with no glycol (I'm using the units for power rather than energy, as this is 'instantaneous power'), it can now with glycol only deliver say 2.7kW, but still at 300% efficiency. In this scenario, what happens is that the lower specific heat in effect throttles the system a bit, and so the demands made on the heat pump actually decrease. Using these hypothetical figures, without the glycol the heat pump would have used 1 kWh to generate 3 kWh out; with glycol, it would use 0.9 kWh to generate 2.7 kWh out. What this implies in the real world (to me at least) is that adding glycol isn't so much a performance hit per se (efficiency remains constant), instead it is like fitting a heat pump with a slightly smaller output. For example, my 12kW (without glycol) heat pump becomes a 11kW heat pump when glycol is added. This may feel like a performance (efficiency) hit, but in reality it isn't, because although my house may be a little cool, I am also using less energy.

At the same time I am also stuck on how a heat pump might adjust to having glycol in the system. First, it has to have some way of knowing that there is glycol in the system either because it has been told that this is the case, or by some proxy. In the heat pumps that can be told they have glycol in the system (not sure which ones have this), what do they then adjust, if anything? Maybe, as you say, they just use the information to fine tune their output figures. On units that don't know they have glycol in their circuit, or indeed how much - 5%? 10% (like mine) - how can they possibly adjust to something they don't know about? Instead, the loop described above happens: the train delivers a little less at the destination, and as a result, needs to collect slightly less at the source.

We are of course in kWh delta t soup here. As I see it, at least in my Midea based system, the heat pump adjusts for outside ambient temp (used as a proxy for demand, as ambient temp goes down, demand goes up) solely by adjusting LWT (by way of the weather compensation curve), leading to a higher rad/room delta t (and so more heat delivered to match the higher demand), while flow rate and delta t remain relatively constant, which is by design. The latter does fluctuate (often a lot in the short term), but it is, as far as I can see, passive reactive variation eg on cycling the delta t will increase a bit as the compressor comes back on (until the RWT catches up with the LWT), and decrease a bit when the compressor is off. This can be seen in this chart (which you have seen before):     

The flow rate (blue line at the bottom) remains pretty constant. The delta t (gap between the top two lines) does vary, but mostly in response to defrost cycles (on the left) and cycling (on the right). When it is in a steady running state (the two flat line periods on the left), it keeps the delta t to about 5 degrees, as expected.

As an afterthought, I had a look at Headroom Heat Pump's heat pump tool kit/calculator, to see if it included any adjustment for glycol in the system, for surely that is where to make the adjustment (because glycol lowers capability: a 12kW without glycol heat pump only delivers 11kW with glycol, so you need to uprate the heat pump to maintain a 12kW output when glycol is added). Glycol is mentioned in the toolkit, but it seems only as a guide to the amount required, based on the system volume. There don't appear to be any adjustments made to the calculations because there is glycol in the system.

The practical upshot of all this is that I can see a very good reason for having glycol in the vulnerable primary circuit, and only one downside, my system is a little less capable (but not less efficient, for the reasons given above), and that is a hit I am prepared to accept in return for the assurance that my rather expensive to repair heat pump is protected against frost damage in all but the most exceptional circumstances. I also, because this is a generic Midea unit with its output controlled by dip switches, do have the option, should I wish to do it, of flipping the dips, to increase the output (and so compensate for the glycol hit). But experience to date suggests that 95% or more of the time my heat pump is capable enough, and the solution for those cold spells when it is insufficiently capable is another layer of clothing and/or a supplementary heater. It is also possible that as I add more secondary glazing and plug the remaining leaks (they are everywhere, and not always easy to access and plug) the heat pump will get to the point where it is capable enough almost all of the time, if not all the time.                

Posted by: @iantelescope

↑

@jamespa 

 

Hi James,

My system has two Two Water Power control options :

Water Power Control by varying the Water Flow Rate using the Grundfos PWM motor socket.

OR 

Water Power Control by mains switching  the Grundfos motor Mains socket.

 

PWM Water Flow rate Control of Water Power

My system has Two Pulse width modulated, PWM, Grundfos Motors.

The PWM motor on the Heat Pump water circuit is controlled by a PWM signal from my Samsung Controller board.

This PWM signal is faulty , being unable to supply a signal to the PWM motor.

The Radiator Water circuit PWM motor socket has no connection to any source.

Both PWM motors work flawlessly when sourced from a "Arduino" .

OR

Mains Switching Control of Water Switching

In this power control option the Water power is Pulse Width controlled by the cycle time , typically turning ON for between 10 to 30 minutes.

 

Which should I use?

PWM  or Mains Switching ?

 

If PWM should I use

One PWM signal , the Samsung, when , and if repaired

OR

Two  PWM controls controlling the flow rates on both water circuits?

 

Modelling

Given that any given desired outcome for a heat pump system causes so many unwanted side effects :

Can we model , in arithmetic , a complete Heat Pump system?

Can we test our model against a physical system with pipe external measurements of Water  Temperature, Pressure and Water flow rate?

 

 Contracts

Again, you know much more than I about contracts , the law......et al.

I will take your advice and be more circumspect with respect to the NIC/MCS

 

ian

 

 

OK at this point I need to post a health warning - I'm not a heat pump expert, I'm just an enthusiastic amateur with a physics degree from 40 years ago and an engineering background, who has been studying heat pumps for 2 years plus because I want to fit one in my own home and realise that the public, and even the industry, understanding is nowhere near mature. 

Also I'm not a lawyer.  I do have experience of contracts and some aspects of consumer law, but if you want chapter and verse on your rights go to Citizens Advice or a solicitor.  You may well have rights under consumer protection legislation that are not available in the general law of contract (for example the absolute right to replacement of faulty goods or refund which applies when goods are purchased), however given that the system has been 'operating', the component parts are generally sound, and there is a reasonably low cost way to fix up the obvious faults, I think you would struggle to argue that this could reasonably apply to the system as a whole.  But I'm not a lawyer!

Health warning over, now onto PWM vs mains switching.  The answer to this depends on Samsung's algorithims which I don't know and probably nobody else does other than possibly the person who wrote them.  However in general terms I would expect PWM to better match the pump speed to the requirement so that the heat pump and the emitters operate continuously at the lowest possible temperature/flow rate to achieve the required heat output, whereas mains switching is on/off so, while the average throughput will be the same the peak will be higher which will make more noise/stress the heat pump more etc.  However the difference may be marginal depending on your system volume and house characteristics. 

Given that you don't have a volumiser which works, and system volume is clearly a concern because a volumiser was suggested, Id go for PWM.  Id anyway go for PWM unless there are reasons not to on the grounds that it should better match the pump speed to the requirement.

Re modelling, doubtless a heat pump system could in principle be modelled in its entirety given enough data, but I for one don't have enough data.  However, even with enough data about the heat pump there are some additional problems which don't lend themselves to easy modelling:

• houses are not a simple load and are part of the 'system' so any modelling of the heat pump will be to an extent limited by the assumptions made about the house. 
• people are complex beings and what is an acceptable temperature depends on perception and individuals not an absolute

These limit the usefulness of any model!

I have created a weather compensation model and a model of the heating of a hot water cylinder to explore these factors both of which have been posted here.  I would like to model timed vs continuous operation because I think that's an interesting question for houses that are unoccupied during the day and also at either end of the season even in continuously occupied houses.  I have made a start and have an idea how to do it, but lack of time and also lack of data on the 'start up' penalty has prevented progress.

Posted by: @cathoderay

↑

But is this a performance ie efficiency hit? Maybe the system remains at a constant efficiency level (lets say 300%) but instead it becomes less capable: instead of being able to deliver say 3kW with no glycol (I'm using the units for power rather than energy, as this is 'instantaneous power'), it can now with glycol only deliver say 2.7kW, but still at 300% efficiency. In this scenario, what happens is that the lower specific heat in effect throttles the system a bit, and so the demands made on the heat pump actually decrease. Using these hypothetical figures, without the glycol the heat pump would have used 1 kWh to generate 3 kWh out; with glycol, it would use 0.9 kWh to generate 2.7 kWh out. What this implies in the real world (to me at least) is that adding glycol isn't so much a performance hit per se (efficiency remains constant), instead it is like fitting a heat pump with a slightly smaller output. For example, my 12kW (without glycol) heat pump becomes a 11kW heat pump when glycol is added. This may feel like a performance (efficiency) hit, but in reality it isn't, because although my house may be a little cool, I am also using less energy.

I think you are right, there may be, in principle, a small capacity hit either at the heat pump level or because of the max speed through the pipework in the rest of the system (which, in principle, is less with glycol than with water because glycol is more viscous - another reason for separation using a heat exchanger).  However this hit may not in practice exist, it depends on what the limiting factor is on capacity.  In many heat pumps it appears to be a software imposed limit!

However even if there is a hit its most likely lost in the uncertainty of the specs (although I do think I remember seeing it in at least one heat pump spec)

Posted by: @cathoderay

↑

Instead, the loop described above happens: the train delivers a little less at the destination, and as a result, needs to collect slightly less at the source.

I think that's spot on.  The heat pump doesn't need to know, the feedback loop sorts it out.  Some do ask though, but I suspect this is so that they can calculate delivered heat from flow rate and delta T to give a measure of COP.

Posted by: @cathoderay

↑

The practical upshot of all this is that I can see a very good reason for having glycol in the vulnerable primary circuit, and only one downside, my system is a little less capable (but not less efficient, for the reasons given above), and that is a hit I am prepared to accept in return for the assurance that my rather expensive to repair heat pump is protected against frost damage in all but the most exceptional circumstances.

Its a fair view although others take the opposite view.  I have never seen any stats at all on heat pump failures due to freezing and the reasons behind them.  There are countries much colder than the UK where heat pumps are used, their experience should inform us!

Boilers of course are often located in garages or outbuildings so in principle suffer the same problem, albeit in slower time.  Mine (located in a garage) has a frost stat function which overrides everything else to switch on the boiler if the measured water temp goes below 5C.  This will also fail if the electricity or gas fail, which is essentially the same scenario as the heat pump failure.  Of course there is comfort in that a boiler in a garage will take many hours, perhaps days, to cool to freezing whereas an external heat pump will do so in a few hours or less.

Im not sure what would fail in a heat pump if it does freeze.  Most likely either a water pipe would split, the heat exchanger would blow, or the pump (if there is one) would seize.  None of these are particularly expensive.  The refrigerant side and the electronics should be fine, unless of course the water thaws and pours over the electronics.  Probably no more expensive than a similar failure in a gas boiler.

@jamespa - I agree about the need for stats on real world numbers on system failures caused by freezing, but I am not sure they exist, meaning we have uncertainty. I suspect I have antifreeze in my system mostly as a relatively low cost way of treating my anxiety that something might happen if I don't have it in the system, rather than it will happen. I don't fancy sitting in a freezing cold house with my fingers crossed during a prolonged cold spell plus a power cut (they do happen here, as it is fairly rural, and the overhead power lines are vulnerable in bad weather, mostly gales, but heavy snow can also cause problems). The other consideration for me is that a cold spell is definitely the worst time to lose the heating, having antifreeze in the system reduces the chance of that happening (at least in theory... because the real, absolute risk may be in fact already be very low, perhaps I am reducing a one in a million event to a one in two million event - but the risk is still (probably) present).

All that said, that is just my current thinking. There are perfectly good arguments for not using antifreeze solution, and using antifreeze valves instead, with probably, for me, the fact that a water only system is much easier to maintain than a water plus glycol system, followed by the better heat transport and less viscous fluid advantages. But it still means having to refill the system and bleed it if the balls do pop off the brass monkeys, albeit not that onerous to do if there is a preinstalled filling loop, but still a chore, and one needs to be present. Swings and roundabouts, horses for courses...