Altherma flow temp issue

[Jon Glanfield]

I wonder if anyone could shed some light on this scenario that we are having trouble with:

We are the installers, we have installed an Altherma sized by Daikin onto UFH supplied and designed by Polypipe and installed by the client and his plumber.

Property is 700 sqm, SAP came out at 21 watts a sqm though, due to levels of insulation specced, hence Daikin calculated heat loss at 14kw from coefficient and supplied a 16kw Monobloc with 285 (300) cylinder

Since then it transpires that the SAP was wrong, client forgot to include conservatory which now takes heat loss to over 18kw.

However the unit does not achieve set point flow temp on space heating (45 degrees), even if we isolate off all but 7kw of UFH. This was tested again last week when ambients were circa 0 degrees. It does not have an issue on hot water side.

We have installed a low loss header which holds 15 litres and a 4 litre make up vessel, header is circa 1.5m long. Heat pump f & r offset to UFH f & r and so on. We have a pump on the UFH side of the header feeding to 7 manifolds, all with their own circulators.

Delta T across f & r on heat pump side is 5 degrees and same on f & r to UFH. UFH was designed with a 43 degree flow temp and a 7 degree delta t. When we installed the header Daikin Tech advised to retard the Daikin pump to achieve the delta T, due to it's high flow characteristics and the low static losses to and from the header.

On Monday with a Daikin service engineer on site, we could not get the space heating flow temp above about 38 degrees (with 7kw of load) and the BUH would come in to achieve this, then drop out, followed by the compressor and then the flow temp would decay by 10 degrees before the compressor kicked back in to start the process over.

Some rooms are at set point but it's costing a fortune to run, circa £30 a day. We recognise that the sizing is an issue due to the incorrect SAP, but the client who is switched on, is blaming the unit because it does not reach set point and recruits the BUH too frequently.

We are not using the weather compensation and interestingly there are 2 sensors o/s, 1 on the unit and 1 linked to a Smart Home system that drives the space heating controls. The Altherma sensor is constantly 3 degrees lower than the smart home sensor, possibly due to the air wash.

Daikin could not see anything wrong with our install, Polypipe have been on site to help with balancing, but when I have measured delta t across the manifolds there is only 1 which has a delta t of 5, the others have delta t's of 1-2 degrees. Losses are minimal circa 2 degrees at furthest mf.

Daikin technical are now saying the Altherma pump should be on full chat, provided we can feed enough out of the header to avoid recirc and cycling due to high return temp.

Their view is that the flow is too low to clear the sensors of the excess heat that occurs when the BUH kicks in, causing a shut down and refrig recovery/balance. However from their manual 16kw/5/4.2=0.76l/s or 45.9lpm which is nominal.

We have the EP set at o degrees, and again the manual suggests that the BUH should be recruited when the ambient is less than set point, which it wasn't when we were testing on Monday.

Unit has not gone igloo over winter and all other manual requirements have been met.

I am just not convinced by the flow rate issue yet and wonder if there is a software issue.

Does anyone have any views?

[Contactor]

What, if any, are the bypass arrangements for the monobloc units?

[Jon Glanfield]

The header and the volume of water in it, the make up vessel,and the pipework satisfies the bypass arrangement.

If there is no header the instruction is to install an auto bypass as far away as possible from the unit to create 20 litres of system volume (for the larger monoblocs) and then shut down any actuators and/or TRVs and open the ABV until any 7h faults clear to prove min flow in a satisfied condition.

Jon

[Bigfreeze]

The header and the volume of water in it, the make up vessel,and the pipework satisfies the bypass arrangement.

If there is no header the instruction is to install an auto bypass as far away as possible from the unit to create 20 litres of system volume (for the larger monoblocs) and then shut down any actuators and/or TRVs and open the ABV until any 7h faults clear to prove min flow in a satisfied condition.

Jon

At 21w a sq m the house would need to be passive standard. Is it? Seems like a spectacularly flawed ufh system if it requires 45C. What pipe spacings and lengths are there per circuit? Are there stats in each zone. If not you don't require a low loss header or a bypass. If there is, a low loss header is incorrect design in any case, especially in a property of this size. A buffer tank is essential. You would require about 30L per kw. Also, if you're running stats you're not really running weather compensation.

[mad fridgie]

Excuse my ignorance, what is SAP (some sort of average heat load over a season) Thats a monster of a house, (heating requirment does seem small)
Why would you want to increase the split over the heat pump by reducing the flow rate, this only increases discharge pressure (increasing elecrtrical input and reducing energy output)
why are you concerned with theoretical flow temperatures, you stated that some areas are at temperature.
In my opinion very small splits are all positive (even heat distribution). Is this on a timer or does the system run 24-7 (the ability)
I do not understand how you are getting approx 2C split across the manifolds, yet getting a 5C split between the F&R of the circuit feeding the manifolds.
Are the feed temps to all mainfolds the same or different.
Is the system new (how long has in been running)
plus you heat pump will not give is rated output when at the present working conditions.
I would run 24 -7, reduce water set point temperature, (look at how the elements are controlled, not sure about your unit), never reduce flows.

[Jon Glanfield]

Hi Big Freeze some comments below:

Passivehaus is 10 watts per sqm as far as I know, but 21 is still very impressive and there may be an issue with the SAP ratings.

UFH was designed by Polypipe using the u-values from the SAP and has spacings between 100 and 200mm. Al rooms are zoned.

Daikin have never suggested the need for a buffer, on the basis that the inverter overcomes the requirement by ramping down as zones satisfy, we always fit buffers on zoned GSHP but not inverter driven ASHP's, why do you say that it needs a buffer?

Not sure I understand the point about weather comp either, because a room can still require heat as ambients rise but will need less flow temp as the ambients rise, hence the weather compensation to reduce consumption. Or am I misunderstanding something?

[Gary]

Delta T across f & r on heat pump side is 5 degrees and same on f & r to UFH. UFH was designed with a 43 degree flow temp and a 7 degree delta t. When we installed the header Daikin Tech advised to retard the Daikin pump to achieve the delta T, due to it's high flow characteristics and the low static losses to and from the header.


If the design output is 16kW with 7 delta-T, but you really need 18kW, wouldn't it make sense to increase the flow and drop the delta-T to make up the difference?

[Bigfreeze]

Hi Big Freeze some comments below:

Passivehaus is 10 watts per sqm as far as I know, but 21 is still very impressive and there may be an issue with the SAP ratings.

UFH was designed by Polypipe using the u-values from the SAP and has spacings between 100 and 200mm. Al rooms are zoned.

Daikin have never suggested the need for a buffer, on the basis that the inverter overcomes the requirement by ramping down as zones satisfy, we always fit buffers on zoned GSHP but not inverter driven ASHP's, why do you say that it needs a buffer?

Not sure I understand the point about weather comp either, because a room can still require heat as ambients rise but will need less flow temp as the ambients rise, hence the weather compensation to reduce consumption. Or am I misunderstanding something?

Jon,

Anything below 20W would be considered passive. If the house has that type of heat requirement then the flow temp of 45 would be extremely excessive. 200mm is too wide a spacing to use with a HP but at that level of insulation you should still get away with flow temps of no more than 32-35C.

Daikins policy on installation does not inspire me with any confidence whatsoever. They try to make their systems as pricefriendly as possible by recommending installation techniques that are designed towards oil/gas boilers. If the system has stats it needs a buffer. You are cutting down the flow through the system everytime a actuator closes. If a buffer is installed, all that will happen is that the buffer will come up to temp and the unit will then switch off. If you install a bypass or low loss header, everytime an actuator closes, you end up forcing more water from the flow side directly back into your return. This skews the actual return temp, pulls down your efficiency and causes the machine to turn off even if the work is not yet complete.

The whole point of weather compensation is to run at as low a temp as the heatloss of the house requires. If you are using stats you must cater to the highest common denominator i.e a stat could be set to 23 calling for heat while all other stats could be off. This is incorrect use of weather compensation as you cannot cater to the lowest temp possible. Also if you are using a low loss header, the return temp will heated by the returning flow temp so this will throw out any possibility you have of using the weather compensation properly.

The invertor should really be used to match the heatloss of the house as a whole. It is incorrect use of the system to have it match zones as they come on and off line. Lets say you get to a situation where only 2 zones remain on and you have 4-5l a minute returning through your UFH. The rest of the water is pushed up a bypass valve or low loss header. What effect does this have on your heatpump. Where do you dissipate the extra heat you've created. You're just opening yourself to a world of problems.

BTW at 14kw and 21w per sq m. the heatpump is very undersized. The 14kw unit installed will only give 14kw at an air temp of 2 and water outlet of 35C. You're requirement is 14.7kw going on your SAP, so no consideration has been given to the oversizing required with an ASHP (usually in the region of 20%) and also the higher design temps required by polypipe which drags down your efficiency further. At the temps you mentioned of outside temp 0C and flow of 45C that unit will have an output in the region of 11kw.

[MikeHolm]

If the design output is 16kW with 7 delta-T, but you really need 18kW, wouldn't it make sense to increase the flow and drop the delta-T to make up the difference?

Gary, as the heat transferred depends on both the flow rate and the dT, reducing dT would require a considerably higher flow rate to change the output. Increasing the flow rate that much will result in a much higher head loss in the pipe and the pump may not be able to handle it (don't know in this case) and pumps are not usually oversized that much for other reasons.

The bypass that is used must open up as the pump head will be so high trying to fit all the flow through 20% (for example) of the tubing area (2 or 3 - 1/2" tubes rather than maybe 10 -1/2" tubes). Looking on any pump curve you will see that this will come close to "dead heading" the pump and it won't last very long.

A quick note from memory of the Slant Fin convectors (3/4" copper tube and aluminum fins) that we see in cheap installations shows if a convector has an output of 500btu/foot at 1gpm it will have perhaps 650btu/foot at 2 gpm. The change in output will not be linear with fluid flow. I haven't thought of these for years so my numbers may be a bit off.:confused:

[Gary]

If i understand the original post correctly, they had 5 delta-T and reduced the flow to match design (7 delta-T). I am suggesting they take it back to the original settings.

[MikeHolm]

They seemed to have reduced both HP and floor heating flow rates and I cannot speak to the HP side as well as you can but if they can defy the odds, speed up the floor heat flow rates enough to get a lower delta-T over the whole floor it may even out things for the HP but actual heat transfer to the floor will go down. It remains to be seen if they will meet the heat loss of the building under those conditions.

[Gary]

Why would an increase in flow rate/decrease in delta-T cause heat transfer to the floor to go down?

[Gary]

They seemed to have reduced both HP and floor heating flow rates and I cannot speak to the HP side as well as you can but if they can defy the odds, speed up the floor heat flow rates enough to get a lower delta-T over the whole floor it may even out things for the HP but actual heat transfer to the floor will go down. It remains to be seen if they will meet the heat loss of the building under those conditions.

If we start with a low water flow and gradually increase it, I would expect that the heat transfer to the floor would increase. I have been told that there is a point where this reverses, such that an increase in flow results in less heat transfer, but I haven't actually seen this or tested for it, so I can't confirm or deny it. I would assume this would be the result of some sort of laminar flow effect. The way to test this would be to monitor the temp difference between the floor surface and the incoming water. If this TD increases (floor temp drops without water temp dropping), then there is less heat transfer.

As to the HP pump, crank up the flow. If the head pressure drops, it's all good. If the head pressure doesn't drop, or starts to rise, then again we are looking at the laminar flow thing. :)

[MikeHolm]

If we have a 5F delta T at 1gpm then we have a heat transfer of 2500btu (500 x 5 x 1gpm). If we increase the flow to about 1.5 gpm for example, and reduce the delta T to 2F we will only transfer 1500 btu. It does depend on our assumptions and the degree of change we want so there are times when we could transfer more heat by increasing the flow rate quite a bit and reducing the delta T maybe only 1-2 deg but we will be increasing pump power a lot.

Time in the tube is important for heat transfer in a floor heating system and most residential loop lengths are a max of 250ft and fluid velocities of 4-5ft/sec(although this will vary a lot when there are zone valves in the system). The insulative characteristic of PEX tubing over copper means we need a long length at those flow rates to get the heat transfer especially with lower input temps. Much of the flow is laminar as well which impedes transfer and I have seen the reduction in transfer that you mentioned. It screws up a lot of installers who keep thinking they need to put in a bigger pump. It is interesting that when floor heating with HPs in Europe the delta T is kept as low as possible 5C where with boilers we would be looking at 10-12C (or even up to 15C) here in North America.

[Gary]

If we have a 5F delta T at 1gpm then we have a heat transfer of 2500btu (500 x 5 x 1gpm). If we increase the flow to about 1.5 gpm for example, and reduce the delta T to 2F we will only transfer 1500 btu.

How about for example, if we increase the flow to 1.5 GPM and reduce the delta-T to 4F? Then we will transfer 3000 btu. It does indeed depend upon our assumptions.

[mad fridgie]

Increasing the flow and decreasing the flow temperature, reduces the heat transfer to the concrete and thus to the air (the heat transfer properties of the water have little to do with the exchange process), however by increasing the flow you reduce the diff between the flow and return, and end up with a better overall even heat transfer to the concrete.
The energy relased by the concrete to the air will be even. With low flows and and high splits you tend to get concentrated heat areas

[Gary]

I can see that as the delta-T gets closer to zero, it is going to take a progressively greater increase in flow to bring down the delta-T, but we are not talking about dropping from 5F to 2F delta-T, we are talking about dropping from 12.6F (7K) to 9F (5K) delta-T... and initially the delta-T was at 9F (5K).

[Jon Glanfield]

I been back on site and increased the flow, but it has not made any difference in terms of achieving and maintaining set point temperature. Generally the delta t on the HP side of the header remained at between 3 and 5 degrees.

The unit would ramp to set point, then allow the flow temp to decay by sometimes as much as 7 degrees before recovering but only to a point 4-5 degrees below original set point as per the controller.

We tried with different volumes of floor loop available and once it had settled at a flow temp we would then change the controller to match that set point and the same activity would be observed.

Big Freeze I am interested by your comments about weather comp and buffers on inverted driven heat pumps, which as you say fly in the face of manuf instructions. In fairness though we have had no issues with Ecodans installed on UFH with multiple zones.

I have always thought of the inverter as being similar to the modulation of a gas boiler, which ramps up and down gas usage depending on the load, have I got this wrong? Obviously any heat source has maximums and minimums of heat out put at different ambients but it is uncommon to see a buffer used with a modulating gas boiler, so why use one with an inverter driven heat pump that monitors flow and return in the same fashion?

Also on weather compensation my understanding was that this is a curve of flow temp against o/s ambient, so if a stat is calling it is o/s ambient that determines the temp of the flow and the temp required in the room is still determined by the stat but the heat generated to the flow is reduced as ambients rise and therefore heat loss decreases.

Is there not a break point too where the losses from a standing buffer which then have to be recovered, match any losses from a heat pump which is operating close to its lowest kw parameter due to zones shutting down?

We have used buffers before on GSHP's but as guided by the manufs installed a header in this case and again my belief was that this would provide hydraulic separation by creating a neutral point. So as the flow on the UFH reduces due to actuators closing down, the flow would remain constant to the HP, but gradually as less heat is drawn off the header the return temp rises and signals a shut down at the HP as thermal equilibrium is achieved, or a ramp down of the inverter as the load decreases (signalled by a reduced delta t)?

I agree there is a sizing issue and we did initially set the set point at 35 because of the high levels of insulation, SAP values and so on, it was only after the problems were encountered that we engaged set point and increased it. Interestingly though another UFH designer we work with routinely specs 200mm centres with HP's and we have had no issues even on plate and suspended timber when balancing is correct. That is not to say obviously that performance/efficiency is enhanced and our preference is for tighter spacings and recently alupex pipe to maximise the output from the pipe itself.

Anyway this is a bit off topic, we have submitted condensing and leaving water temps to Daikin, because my belief is that notwithstanding the sizing issues, the unit should cruise closer to set point than it is doing and until we can get to that point we can't begin to convince the client to obtain an alternate independent SAP. Does anyone know what the variance from set point should be on a Daikin monobloc? Ecodans are usually within 1 degree of required flow temp up or down.

[MikeHolm]

Part of the reason for the LL header in boilers (modulating or otherwise) is that many of the flow measure devices, including the negative pressure gas valve cannot measure properly at the very high and low flow rates. High flow rates in some boilers (Viessmann Vitodens for one) will cause the controls to lock out. Is it not the same with the HPs where the operating limits are even more strict?

I assume that the pump in the Ecodan is a single speed unit and does not vary with delta T and the flow rates I see would promote a lot of mixing in a header or buffer (correct me if I am wrong) and therefore the tight delta T on the floor side needs to be maintained lest it drive down the HP temp.

The floor temp in floor heating is less about heating air than providing radiant heat to the objects in the room. Air temp is of less importance with floor heat than with convectors for example and is typically a couple of degrees lower (I get lots of comments from homeowners thinking there is something wrong with their thermometers).

That said, with the heat load as specified, you should not need more than 35C into the floor so I think there is an undersized HP for the conditions or there is some extra heat load not accounted for such as slab edge losses or perhaps an aquifer close to the some area that drags down the floor temp. I had to be a witness in court over one of those before.

[Bigfreeze]

Jon,

Mike has it pretty much spot on. A boiler and heatpump are worlds apart in terms of the way they operate, so using the same control techniques does not work. Boilers burning at high temps actually have increased efficiencies over those at lower temps. Complete opposite with a heatpump. With a boiler you can produce 70C and mix it down to 45C in order for it to suit a UFH system. The fact that you get water at that temp coming back to a boiler has little to now effect on its running. Neither will the fact that the flow rate is cut when a actuator closes down. A boiler simply does not care what temp is being returned to it once it is below the cut out point on the boiler stat.

With heatpumps you have to ensure your return temp is as low as possible. A false return temp, because you are generally working with DT's of 5-7K has immediate implications as to what the machine is deciding to do. As I mentioned before, if you reduce the flow through the ufh and artificially bypass it through a LLH, you end up doing two things. 1. Giving a false reading to the heatpump of the true temp in the floor (your weather compensation will have given the unit a desired floor temp for the current outside) and 2. Pulling down your efficiency as a high return temp, means a higher condensing temp, higher electrical load.

Modulating units are not really designed to work with Buffer tanks, but they are most certainly not designed to operate with thermostats and restrictions in flow. Ideally a modulating unit will have a modulating circulating pump which will maintain your DT across your condenser depending on heat output at any particular outside temp. But that is the point of modulation. It is not a method of matching flow rates and room loads as they come on and off line. You're not dealing with an A/C unit where variable loads are dealt with in a completely different manner so the same principals do not apply. Do yourself a favour and ignore the "best practice" of Daikin, Mitsubishi etc. There are many manufacturers out there along way ahead of these guys when it comes to heatpumps and they will tell you the same thing I have. Modulation when used correctly allows the heatpump to have a 5kw output on a day you have 5kw heatloss and be a 15kw heatpump on a day you are losing 15kw, nothing more, nothing less.

Regarding the weather compensation, the only true way to run wc is when the house is regarded as one zone. You control your floor temp against your outside temp on any given day and that, once set up properly, will give you a perfectly stable temp across the house and the max efficiency possible at that particular time. You will always be producing the minimum flow temp possible in order to match the room temp required.

We've just finished a house, 650sqm, B1 standard, 25kw modulating GSHP, no stats, no buffer. That house will come in under €1k to heat next year. Its simple, no added complications, straight forward plumbing. Heatpumps work well with the simplest forms of plumbing possible. The more gadgets you add into the system the more heartache they cause. In your case you're probably stuck with the stats, if so, you have no alternative other than run a buffer. It will also give you clarity as to what the machine is actually producing. How long does it take to heat up the buffer? You'll have a pretty good idea of your output. Does the buffer reach temp before the stats reach temp? That will indicate flow problems further down the line. At the very least it will always provide you with the perfect flow rate in order to defrost properly and you will always have the lowest return temp possible which means your machine will be at its most efficient.

[Jon Glanfield]

Hi Big Freeze, when we first started with GSHP's we always used to work on the basis of open zones on underfloor heating and no TRVs if on rads, which gave a big thermal mass for the heat pump to work against. Gradually though I guess the market has moved more towards zonal control and we were also advised by manufacturers of inverter driven ASHP's that zonal control was fine with their units.

What you say though makes perfect sense, but how do you reconcile your designs with current Building Regs that require UFH to have stats in every room and time and temp control to the heat pump?

Functionally then do you balance the UFH as per design and let it run through the heating season, controlling the heat pump on return temp rather than on off via a room stat? Would this be the same protocol with an ASHP which generally are looking for either volt free or 230 volt signals on the units we use?

Conversely with a buffer is on off on the heat pump controlled by a set point/sensor on the buffer, but do you also then have a stat to interlock with the space heating circulator, to avoid overheating the house or degrading the heat in the buffer?

We have a number of projects in the pipeline where we are revising our designs on the strength of the issues thrown up by this project. On some though we are committed to using Althermas. From what you are saying we would be best placed using these with weather comp into open zones, or use a buffer and no zonal control so that we can guarantee flow rates and avoid flow switch issues when there are pumps downstream.

To date we have used IVT/WB GSHP's and been happy with them and their capabilities, but are with NIBE for our level 2 training in March and are also meeting with Stiebel Eltron next week. ASHP have always been Mitsubishi and Daikin, but we are intrigued by the Grant too. Are you allowed on the forum to advise us which products you use? If this needs to be taken offline my email is please use the PM system and thankyou for everyone's help so far.

[Bigfreeze]

Jon,

I'm not sure how strict they are in the UK but here in Ireland, once you explain why you are not using zonal controls and how it actually improves efficiency, its usually not a problem and the project will get signed off. These regs were compiled with gas and oil burners in mind, but as I said before, there is no one size fits all when it comes to heating systems anymore. Controls from most HP manufacturers would be superior to any thermostats currently available anyway.

Yes, you just balance the ufh and let the heatpump run on weather compensation throughout the heating season with your setback periods for night-time and times of inoccupancy. I can't speak for every heat pump as controls vary from manufacturer to manfacturer but if you're running true weather compensation you will be working on return temp and not a stat. If you must run a stat, it should be a common stat to control all zones, otherwise you're back in buffer territory. Strictly speaking, anytime a ASHP is installed a buffer should be used regardless due to the need for defrost.

In a buffer setup the HP would be linked to the buffer by stat and would only care about the buffer. What happens downstream would have no consequence to the heatpump and no info would be returning to it. The HP will have one purpose, get the buffer to temp. You can fit as many zones/pumps/controls as you see fit after the buffer. I have never encountered a situation where you would have either heat degradation of the buffer or overheating of the house with a setup like this. It just doesn't happen.

I'll pm you info regarding the other info you requested but at the moment you are not accepting pm's. Just check that and come back to me.

[MikeHolm]

What you say though makes perfect sense, but how do you reconcile your designs with current Building Regs that require UFH to have stats in every room and time and temp control to the heat pump?

Dummy stats in each room....works for some people. Damn, where have the smiles gone?

[MikeHolm]

BF and Jon,

Keep me in the loop, This is of huuuuggggeee interest to me.

Mike

[MikeHolm]

I have wished that the authorities here would even outlaw mid efficiency boilers and furnaces. Lots of talk but no action let alone getting as in depth as individual house zoning regulations. On the opposite side of the street is the mandated HRVs where I think they are superfluous.

[Bigfreeze]

I have wished that the authorities here would even outlaw mid efficiency boilers and furnaces. Lots of talk but no action let alone getting as in depth as individual house zoning regulations. On the opposite side of the street is the mandated HRVs where I think they are superfluous.

They never get it right Mike. They're always pushing too heavily one way or the other and its usually the vested interests that have their say. You never see guys like us being consulted on whats best practice when it comes to regulations. It'll never change, sometimes you just have to circumvent the rules because you know its the best thing to do. Any building inspector worth his salt should recognise that too.

[Jon Glanfield]

Thanks again Big Freeze, I have looked at my settings and can't see why I wouldn't be able to accept PM's, could one of the moderators help at all please?

[Bigfreeze]

Had the same problem Jon. Contact Webram and he'll be able to help you out

[yinmorrison]

Its simple, no added complications, straight forward plumbing. Heatpumps work well with the simplest forms of plumbing possible. The more gadgets you add into the system the more heartache they cause.

Hear Hear to that comment Bigfreeze

[yinmorrison]

Ok you guys , I am looking into this MCS Scheme and since you are obviously already in the scheme installing GSHP's etc maybe you can enlighten me. We are in the HVCA and are inspected for QA system already so I really cannot see why we should have to go through a seperate scheme for registration for MCS. Is it not just a case of another level of bureaucracy to line Govt coffers?

[Jon Glanfield]

Hi Big Freeze, who is Webram and where can I find him!?

Following your sage words of advice I have dangerously done some research into buffers and wondered whether I could pose some questions on the strength of it?

I believe the calc we need to perform to size them correctly is as follows:

V=t(Qh-qload)/(500xdelta T)

1. V=vol in litres of buffer
2. t=heat pumps on cycle
3. Qh=KW output of heat pump
4. qload=rate of extraction from buffer
5. delta t=temp rise of tank between settings (on vs off)

How do I calculate 4 and 5, and is 3 equal to max potential KW output or is it equal to max output at the lowest design ambient?

As far as as weather compensation is concerned when using a buffer, how does this work? If the buffer has a stat or sensor to signal the HP off when satisfied, how does it satisfy if the flow temp ramps down due to the WC slope at higher ambients?

What sort of temp should we aim for in the buffer for say UFH, should it be the design flow temp?

When designing a buffer into a system, which is better a 4 tapping unit which would require pumps up and downstream, or a 2 tapping version which T's into the space heating flow and return, and only starts to heat when static pressure increases as zones close and forces water into the vessel? This could mean a reduced number of pumps and hence power consumption and less heating of the thermal mass until zones satisfy.

We have also been speaking to some manufacturers who recommend a thermal store with instantaneous hot water take off along with space heating, but which works out as being most efficient? We believe it would be to use a buffer for space heating and separate cylinder for HW or a sling tank if there is a large HW demand.

If we choose to use an inverter driven ASHP and use a buffer for the defrost and or zoning, I am surmising that we run it on weather comp as per above, but will the inverter increase efficiencies further or cause complications when used with a buffer?

I apologise for my ignorance and volume of questions but we are now revising a bulk of projects which are imminent in the wake of what has happened on the OP project and on the strength of logic that you have presented.

Many thanks Jon.

[Bigfreeze]

Ok you guys , I am looking into this MCS Scheme and since you are obviously already in the scheme installing GSHP's etc maybe you can enlighten me. We are in the HVCA and are inspected for QA system already so I really cannot see why we should have to go through a seperate scheme for registration for MCS. Is it not just a case of another level of bureaucracy to line Govt coffers?

I'm not too familar with the regulations in the UK yin, but from what I do know, your government is really stunting the HP market there. You're at least 5yrs behind Ireland in terms of take-up and product range and thats saying something. The registration rules for HP manufacturers in the uk is very strict and very expensive so alot of good companies opt out as they can't avail of grants without the cert. I presume its the same case with installers. They did something similar here about 2yrs ago. Got everyone "certified" (3 day course) at a cost of €400 per worker, then promptly pulled the grants scheme so its basiclly useless. The only ones who benefit are the unscrupulous guys who would whack a HP in to any old house, regardless of insulation level, whether its just standard rads etc.

[Jon Glanfield]

Hi Yin I am not sure whether it lines the govt coffers but our experience of the MCS has been disappointing in so far that it focuses heavily on office and admin quality, which apparently is what is important to the customer.

The technical detail which is what is so critical in terms of efficiencies, cost savings and carbon reduction is miniscule in comparison. Unfortunately though in order to be able to sign off an installation for the RHI (when it finally gets agreed) you have to pay to play and there are on costs in terms of the prep for it and ongoing administration that will inevitably get passed on to customers.

Jon

[Bigfreeze]

Jon,

Webram runs the site. If you search through the members you'll find his contact details.

When using a buffer and weather compensation the HP will assume the temp of the sensor in the tank as your return temp and as such will shut down the pump once this temp is reached. Normally the sensor will be in the center of the tank so the water you will be extracting from the top will be 2/3C warmer so you will be pretty much on point for the temp you need at any particular time. Its unlikely that the desired temp would ramp down very quickly as most wc take their temp as an average over 1hr. Even if it does, it just means you have slightly higher temp than you need in the tank which will be no issue. You will normally set the temp slightly higher than if you were running staight to screed incase the homeowners sets some stats slightly higher.

The temp you will require in the buffer will depend on the design of the ufh and you will need to set your wc to suit. I would recommend that you start installing the ufh yourself if you're not already doing so. It is the only way you can guarantee the performance of your system and thats what your reputation depends on. You could have a HP with a COP of 10 but if the ufh is crap then you're on a hiding to nothing. We aim for flow temps of between 27 - 30C at -5C outside temp depending on the insulation levels.

We always run 4 tappings on a direct feed cylinder, so no coils. You should really consider the buffer as a kind of meeting point for two seperate circuits, the heat pump circuit (entering and leaving on the left of the tank for example) and the ufh circuit (entering and leaving on the right). Generally once piping is sized correctly the pump required for the heatpump will draw minimal power. Its almost not worth considering in terms of energy saving throughout a year.

I always prefer 2 seperate cylinders. One for DHW and one for ufh. The ufh will run most frequently but at the lowest temp so you'll be getting max efficiency on your longest running cycle. Hot water should generally take a very short period of time but it will be at your machines lowest efficiency so you really want to keep those seperate.

Yes, if running an ASHP thats exactly how you should run it. The invertor will not cause any problems. It will lengthen the life of the HP because you won't be shutting down constantly because you are heating the buffer very quickly in warmer weather.

Regarding sizing a buffer, I really wouldn't get too deep into calculating an exact design. Its very hard to predict exactly how quickly heat will be removed from the tank when stats are calling at different times. The best advice I've received and that I could give you in this respect, is that you should allow 30L per kw of output power for the HP. It will give the HP a decent store of water so it does not stop as soon as all zones shut down and more than enough for defrost applications. You could put in a massive buffer on any installation but then you are running into costs and problems with stratification so its just not worth it. Your main incentive in installing a buffer is to ensure a proper defrost and limit the influence of stats on your flow rates. Any situation you can avoid a buffer and go staight to screed on wc will be the ideal scenario

[Jon Glanfield]

Thanks Big Freeze, so in terms of sizing the 16kw we are having problems with may need a 480 litre buffer if we end up going down that route.

We are currently installing an 11kw IVT GSHP where the suppliers (themselves MCS accredited) have designed in a 110 litre buffer and Kensa recommend 10 litres per kw for a single compressor and 5 litres per kw for a twin compressor. Where did you come up with the 30 litres per kw and what is the benefit of the extra size, is it longer running cycles to avoid compressor burn out over time?

Have I also read elsewhere that if we use a buffer on an ASHP we won't need glycol? If so why not given that there are still external pipe runs?

Cheers

Jon

[Bigfreeze]

Thanks Big Freeze, so in terms of sizing the 16kw we are having problems with may need a 480 litre buffer if we end up going down that route.

We are currently installing an 11kw IVT GSHP where the suppliers (themselves MCS accredited) have designed in a 110 litre buffer and Kensa recommend 10 litres per kw for a single compressor and 5 litres per kw for a twin compressor. Where did you come up with the 30 litres per kw and what is the benefit of the extra size, is it longer running cycles to avoid compressor burn out over time?

Have I also read elsewhere that if we use a buffer on an ASHP we won't need glycol? If so why not given that there are still external pipe runs?

Cheers

Jon

The 30 litres is to allow for a decent running period on the HP once it starts. You do not want to get into a stop/start cycle too frequently on a HP. If you go too small there will be no temp differential in your buffer and you're back to where you started and as you mentioned the compressor life will be shortened. Minimum buffer we would ever fit would be 300L, so even if the unit was a 7kw we would fit a 300L. From 10kw up you would calculate at 30L per kw. For your 16kw machine, fit a 500L buffer.

If you have water going to outside from your ufh you will need glycol. It all depends on the heatpump you run as to weather that is the case. We have done both but prefer where all heat is produced and remains indoors. If you use a HP which is producing heat externally and circulating it to the buffer, then I would consider using a coil on the HP side of the buffer, so only that small section of pipe (the circuit between HP and Buffer) need be filled with glycol. It'll save you money in terms of extra glycol and also the extra power that will be consumed by the ufh pumps when pumping glycol

[MikeHolm]

If you must have the HX outdoors you must use glycol. To get -30C temp use 40% propylene glycol to water. There will be about a 6-7% decrease in heat transfer so that needs to be taken into consideration. You can use a lesser concentration to have a lesser reduction pump power and heat transfer but the lucky thing is that, opposed to a solar system, you may get away with the glycol lasting 20 years as it doesn't get so hot as to degrade.

[Jon Glanfield]

The whole glycol things makes sense now I was just a bit confused, internal ASHP's are very rare in the UK.

To reduce the glycol volume though how do we size the coil, ordinarily it is a 3 square metre coil up to a 300 litre DHW cylinder, so as a rule of thumb do we go 1 square metre per 100 litres on the coil?

Jon

[mad fridgie]

I do agree generally with big freeze, how ever buffer tanks, it would seem that his method of sizing is the best of what has been indicated, 30l/Kw, however, with a TD of 5C this is only going to give approx 10 mins, as most good units should already include anti cycle timers, that colud be set to 10 mins. I do not see the point of the extra cost.
(not to be confused with defrosting and water balancing requirement). If you are looking at a genuine thermal buffer, then much bigger (many 1000s of litres), then you can use some smarts, for example increased running during the day (higher ambients for air source) or increased running at night for ground source (cheaper power rates)

[Bigfreeze]

The whole glycol things makes sense now I was just a bit confused, internal ASHP's are very rare in the UK.

To reduce the glycol volume though how do we size the coil, ordinarily it is a 3 square metre coil up to a 300 litre DHW cylinder, so as a rule of thumb do we go 1 square metre per 100 litres on the coil?

Jon

Your coil size will depend on the heatpump size and the type of coil. A 3m2 coil with good heat exchange capabilities should be able to cater for a heatpump of up to 12kw. You will really have to go to the manufacturers specs in order to see what the coil is capable of. You should be looking for a 5k differential across the coil ideally.

Regarding the buffer size, if you put in a 100L buffer as was suggested by IVT you will have no differential of temp in the tank at all. So the water you put out is almost instantaneously being drawn back through the return. A 300L buffer of about 2m in height will give you about 4-5C difference between top and bottom. This is the difference you would expect across your condenser anyway so is ideal for application. Imo a 100L buffer is no better than a low loss header and as thats the cheapest you'd be better off sticking with that. In my experience with heatpumps installation proceedure is everything and it is very black and white, its either right or its wrong. Especially where ASHPs are concerned, small buffers and LLH's are wrong

[Bigfreeze]

I agree MF, the buffers I mention are not designed for any real storage capacity, merely for reasons of defrost and/or flow variance. Getting into buffers for storage is a whole different ball game and would not be applicable to your average british house

[mad fridgie]

In the UK, is the coil in your DHW (when used with a heat pump system) for the potable water or the fluid running around the heat pump loop?

[Bigfreeze]

In the UK, is the coil in your DHW (when used with a heat pump system) for the potable water or the fluid running around the heat pump loop?

Depends on the tank MF, some on the HP side, some have both.

[MikeHolm]

He would then be better off with a side arm HX or a BPHX if he cannot get a big HX in the tank. The biggest standard internal HX I have seen is from Viessmann. Most manufacturers assume you are using solar or boilers and don't size them up for HPs

[mad fridgie]

He would then be better off with a side arm HX or a BPHX if he cannot get a big HX in the tank. The biggest standard internal HX I have seen is from Viessmann. Most manufacturers assume you are using solar or boilers and don't size them up for HPs
I agree, i see many who run the heat pump water circuit through the coil to heat the DHW, with a static coil, the heated hp water has to be really quite hot, driving up head pressure/ reduce efficiency and longevity.
For santitry DHW, i use vented double wall, water directly into the cylinder

[MikeHolm]

What change in HP performance would we expect if we change the deltaT from 5 to 4 or 3 or 2. I am assuming that, as BF said, having the feed/return at the top/bottom rather than both at the bottom with a diffusion pipe would be preferable so what happens at a lower deltaT? In a boiler it would just shut off.

[MikeHolm]

I guess what I am asking is what would be the generally acceptable minimum deltaT in this situation

[MikeHolm]

MF, do you have regulations there for the use of double wall HX with DHW

[Bigfreeze]

It depends on the quality of coil used and its configuration in the tank. It does not necessarily have any implication on performance.

[mad fridgie]

Hi Mike, yes regulations on DHW,

[MikeHolm]

I thought it was only parts of North America that dictated double wall HX for DHW. It depends, of course, on what is on the other side of the HX. Municipalities have always been paranoid about litigation so some demand it when non toxic glycol is used, a hold back from when ethylene glycol was used which almost never is used now. It is relevant here because, in solar, using a double wall makes the financial effectivness of the system questionable due to lowered performance. It must have issues with HPs as well.

[yinmorrison]

Thanks Guys I thought as much. I will have to give the matter more thought.

[Jon Glanfield]

So what we need to do for ASHP's, that generally in the UK, are simple external units is identify a good supplier who can confirm coil size with a 5 degree delta t to match the size of the buffer and the flow temp we intend to spec?

I have mailed Webram so hopefully soon I will be able to receive PM's because I would be interested in hearing which suppliers you use BF.

Just to be clear too, the 100 litre buffers that we have used with the IVT pumps are not IVT designs, they are from well known Heat Pump company based in the UK with a wealth of experience, whether they are just passing these sizings on from IVT though I don't know. In fairness we have not experienced any dissatisfied customers but the units may not be operating at an optimal level.

BF how do you deal with a situation where space simply precludes the use of a buffer on an ASHP project, do you just elect not to take it on for reputational reasons? Also do you find that it makes you non-competitive if an alternate quote excludes a buffer which could also confuse customers who may already be a bit baffled given the new nature of the technology?

Out of interest too, have you arrived at the 30 litres per kw from trial and error, empirical means or manufacturer advice?

Cheers

Jon

[Bigfreeze]

Jon,

Ideally that is what you should be looking for from the buffer. Once you find equipment that works correctly, settle on it and do as little deviating from it as possible.

In a situation where a buffer is simply not an option, we will only install where we can go straight to screed on weather compensation. That way you are guarantee the flow rate through the machine on heating. On defrost, on the machines we work on we have an option to energise another valve/pump once the water temp gets below a certain temp. We usually set it up that once the water temp drops below 20 on defrost, the two way valve on the dhw tank opens in order to increase flow and increase energy available to complete the defrost.

We turn down about 20 installations a year on the basis that they are not suitable for HP applications i.e on standard rads, poor insulation levels in houses etc. It doesn't really matter that they may run for slightly cheaper than oil. The fact is that, these customers will expect dramatic drops in bills and when that does not materialise, your company and your products will be blamed as being shoddy (no matter how well you explain that it won't be the case) . We work on the basis of installing HP's in HP orientated houses i.e good insulation, ufh where we can guarantee running costs 1/4 to 1/5 the cost of oil. It took a while to build up the reputation but the return custom when you work like that gets better year on year and heavy investment in advertising etc is not required.

I don't really care about being uncompetitive. If it means compromising the efficiency of an installation by cutting corners to match others I'll walk away. I've found down through the years that the customers who don't heed advice and don't really see the benefits of why you're doing what you doing, are generally not worth having. Its the guys who are really interested in what you're installing and why that give you the least hassle after sales. People who chase the bottom line I have no time for. If you're buying a Hilti or a Ryobi, you understand why you're paying the money for the Hilti. It should be the same with heat pumps.

The 30L kw was a manufacturers recommendation. But also I've seen the benefits of it down through the years. We do alot of troubleshooting on poor installations and its then you really see why you should install heat pumps in a vey specific way.

[MikeHolm]

[When designing a buffer into a system, which is better a 4 tapping unit which would require pumps up and downstream, or a 2 tapping version which T's into the space heating flow and return, and only starts to heat when static pressure increases as zones close and forces water into the vessel? This could mean a reduced number of pumps and hence power consumption and less heating of the thermal mass until zones satisfy. ]

I'm a bit confused. In an above post it was mentioned that pumps for the UFH be both up and downstream.... on both feed and return????, if I read that right it means that each zone needs 2 pumps. I see no purpose for this.

If I am wrong please tell me what the above means.

Thanks

[Bigfreeze]

Mike, you put your pump on your return to the heat pump (pumping through the resistance) and the pumps on your ufh would be on your flows (again pumping through the biggest resistance), either just one pump or one pump per manifold depending on set up. The HP would be considered a seperate circuit. Which post had the configuration you mentioned in it, as, as you say, it serves no purpose.

What Jon was proposing there was that the buffer be used almost like a receiver, where excess flow could be diverted until the loops in the system reopened. It would work, but again would compomise efficiency

[Jon Glanfield]

BF, we contacted various manufacturers today interestingly and we got responses ranging from 10l per KW from IVT to 30l per KW from Stiebel Eltron with 20l per KW from Nibe.

I think we will be aiming for the 30l per KW though that you have suggested, I have also approached Mitsubishi for their views on buffer usage generally and will be doing the same with Daikin, although they do always say LLH's but we'll see.

Mike are you thinking of the standard Ecodan set up which has a pump on the flow from the HP and a pump on the return to it to increase flow rates by pushing and pulling around the system. These are before any 2-ports and a flow setter on the return to provide a read out of the flow with a retarder in case it is too high, which it never is!

These set ups generally do not incorporate any buffers due to manuf's instructions recommending against their inclusion due to the inverter and so on. Recently though they have changed their literature to mention the potential need for a LLH on UFH to overcome high static pressures and anecdotally today we heard of another big Ecodan supplier using buffers on twin Ecodan installations.

Cheers

Jon

[MikeHolm]

Being from Canada, we don't have any Ecodans yet. Actually, none of the Japanese makers sell an air to water HP yet and are only looking at it now. Also, I think all the air to air ones sold here are 407 or 410, no CO2 either. I am not referring to anything I have experienced here at all. The way the post was written it seemed there would be one pump on the feed and one on return of a zone and that is dumb. One going to the HP and one to each zone is understandable.

[mad fridgie]

In NZ we do also use 2 pumps (push/pull) but this is for economic reasons. There is not a great use for inline pumps, so in many cases it is a lot cheaper to install 2 pumps in series (if pressure is required) or 2 in parrallel (greater flow), than using one correct sized pump (selections made using pump curves).
The buffer tank, I do not think you should use a blanket rule for all applications, but you do need to understand your system.
1: no controlled UFH (fixed water flow rate) i do not see the requirement either a buffer or LLH. (assumed flow rate is suitable for all aspects of the system) The floor is the buffer
2; partial control of the UFH (limted control of some zones, where the flow rate passing through the underfloor never drops below 50% of design, not to be confussed with by-pass valves) The floor becomes the thermal buffer. I would install a LLH Dedicated water pump for HP loop
3; Totally control of UFH and rad systems, where heating loop flow could drop to close to zero, I would install a buffer tank. dedicated pump to HP loop.

Re the size of the buffer tank, if we accept that its primary use is to complete defrosting, then the size should be based upon the flost loading of the evap. (face area * coil depth) this would give the max amount of ice formation.
You then calculate the energy required to melt/remove the ice. Then use this figure to determine the tank size,
You have choose want temperatures you use. It would be a fair assumbtion that the water enetering the cond (evap when in defrost) should not be below 10C to ensure the heat exchanger does not freeze, you should also assume that the house is not up to temperature, and that the water being circulated is also not up to temperature, this being the case the water may only be around 20C. So size you buffer based upon a 10C drop during defrost. (very conservative sizing method)

[MikeHolm]

Mad, I would like to know what type of pumps you use in the push pull arrangement. I am going to do this all from memory so forgive me if some of the numbers are wrong but If you need 40L/m, for example, through a typical large BPHX a Grundfos UP 26-99 should suffice. It may even be a bit big. Its wattage if memory serves is about 250w and should be good for 7m head. Given the above condition would you use two UPS-15-58 pumps with one on either side of the HX? Where does the economy happen when there are twice as many pump flanges and electrical hook ups (I know that the wattage on a 15-58 is less than 100w)? Is it mainly the reduction in wattage?

In many cases where pumps are placed in series one pump will fail before the other as the load on each pump is not exactly the same, which of course, will cause the rest of the system to eventually fail and i would expect the stress on each pump to be different depending on which one is pulling (closer to cavitation on the curve) or pushing. ( I have seen the failures with series pumps many times in drainback solar systems where a high lift is needed to overcome the air in the system until flow is established) Heat pumps are under less thermal stress than solar so I may exaggerate that a bit in this case.

BTW, Unless forced I don't use anything other than wet rotor circs either.

[mad fridgie]

Hi Mike, Grundfos great pumps, hardly never use them, a small fortune here for inline pumps (very limited market)
Tend to use Wilo, OK, better than most but not as good as Grundfos.
We purchase the pumps with the fittings (so no extra costs) a 100 watt pump (25/6 i could be wrong,) from memory costs about $130 (US$100) and thats a lot compared to what i see in the UK, but I if i require a 250Watt pump (more flow and slight more pressure) i pay approx $600 (US$480), thats a big difference in our neck of the woods. Wiring is about 10mins, so about $10. It is a closed loop so i do not have any techincal issues, if one fails then system fails OK i can live with that, as if i only installed one pump and it failed then the system would fail.
I do not worry about cavitation to much on my systems, 1.5 bar standing head and only really working at temps no more than 40C and hopefully no air?????. (Higher for DHW but do not use this method)

[MikeHolm]

That's understandable. Wilo only came to North america a few years ago and matched cost with Grundfos, less 5-10%. The cost for a Wilostar as you noted above is about the same as you pay but the Grundfos is close to that as well. I assume you speak of cast pump and not bronze. The 250w pump will be $250 so for us, I cannot see the issue.

The one bug bear I have is that with a couple of exceptions, we must buy a 2 bolt flange on all pumps. They are bulky and the bolts are difficult to remove at times. Unions are generally not available here except on SS pumps.

[MikeHolm]

Given the cost of pumps in Europe, I still cannot see the reason for a push/pull arrangement on either side of the HP HX unless....a pump is not available with the flow rates needed and I cannot believe that would happen.

[Bigfreeze]

Wilo are about 60% the cost of Grundfos here Mike and for my money they're more reliable. I've had alot of Grundfos units split around the head. Never have any problem with Wilo. We use Lowara on larger pumps. Find them very good aswell. We would also use pump unions and shut off valves rather than pump flanges (screw on type) as the isolators fail to easily and they restrict flow more. We would usually get away with a 25/6 up to 12kw because of pipe sizing etc. That pump is only about €40

[MikeHolm]

I haven't got a pic of our pump flanges to send to you but if you use google.com and look for grundfos in the usa you will see the flanges. They are nothing like ones you mentioned and are far less versatile in some ways than yours. I am jealous of your pump prices and Mad should be even more jealous.

[Bigfreeze]

I know the ones, horrible things. We only really move to flanges on 40mm pumps. Grundfos are ridiculosly expensive here. A 25/80 could set you back €300 plus

[MikeHolm]

BF, would you be using two of those 26/6 pumps in a push pull arrangement on a 12kw machine?

[Bigfreeze]

No just the one Mike. We usually come off the machine in 1 1/4" pipe and manifold out to inch depending on the number of runs. Ufh circuits are usually no more than 80m in length so the resistance in the system is kept low and we can use a small pump in order to keep 5k across the coil. Most 12kw installs I see are using 25/80's but thats just bad pipework design that requires that as far as I see