Home  5  Books  5  GBEzine  5  News  5  HelpDesk  5  Register  5  GreenBuilding.co.uk
Not signed in (Sign In)

Categories



Green Building Bible, Fourth Edition
Green Building Bible, fourth edition (both books)
These two books are the perfect starting place to help you get to grips with one of the most vitally important aspects of our society - our homes and living environment.

PLEASE NOTE: A download link for Volume 1 will be sent to you by email and Volume 2 will be sent to you by post as a book.

Buy individually or both books together. Delivery is free!


powered by Surfing Waves




Vanilla 1.0.3 is a product of Lussumo. More Information: Documentation, Community Support.

Welcome to new Forum Visitors
Join the forum now and benefit from discussions with thousands of other green building fans and discounts on Green Building Press publications: Apply now.




  1.  
    Can anybody point me in the direction of how to calculate the flow rate for a pump for domestic heating?
    We have an open vented heat store, wood stove with back boiler, 60m2 UFH on the ground floor, rads on the first and second floor, and the SAP calculations say we should need 2.5kw at -3º. I'd like to install a pump to distribute the heat around the whole house. Thanks very much!
    • CommentAuthorgoodevans
    • CommentTimeDec 3rd 2018
     
    Pilgrim, we need a bit more detail. Your question implies that you need only one pump. However if you want to run the rads at a higher temperature than the the UFH you'll need two pumps I think.

    That aside for the sake of argument lets assume one pump with 50 % headroom for the pump delivery, and the temperature difference on the flow and return needs to be at 5 Deg C.

    Water will carry around 4200Joules of energy per litre of water per deg C. And 1J/second = 1 Watt.

    5000/(5 * 4200) = 0.23 litres of water per second = 14.3 Litres of water/minute, 0.86M3/hour.

    Obviously if more than one pump is required smaller pumps can be used. You will need to check that the pump can deliver the head required to overcome the head loss at the flow rate you require.

    You may wish to find a pump (or pumps) that have a variable control so that the pressure/flow rate can be adjusted to suit your final situation - In my house the heating engineer has suggested a Wilo pump. (I think it's because the range of settings available on the pump is such that he can always make it work quietly and efficiently.)

    I think this pump family ... (https://wilo.com/gb/en/Products-and-expertise/Series-Finder/Wilo-Yonos-PICO_175.html) will get you out of trouble provided you buy a big enough one which won't be difficult. This pump has a single dial but allows you to select from 3 fixed speeds (like pumps have done for years) but in addition allows you to select a maximum head (typically from 0.5m to the maximum of the pump) with either variable differential pressure or constant differential pressure adjustment.

    It sounds like an advert but I have no connection with these people at all - nor do I know the quality - I just like the simple - get out of jail free options on the pump.
  2.  
    Thanks @goodevans!

    I'm assuming that we can run the system with the single pump, though the UFH will have its own pump to push round the UFH circuits. I was thinking that the heat store would mix down to whatever temperature the radiators needed, and then the UFH manifold would mix down further to whatever (lower) temperature the UFH needed. What further information do you think I would need?

    I spoke to Grundfoss who seemed to think that all they needed to know was the flow rate. Presumably these modern modulating pumps can increase the pressure to achieve the required flow rate?

    The idea of a pump that can accommodate a lot of different settings sounds very useful in case we realise our heat demands are different to that which the SAP etc predicts.
  3.  
    Here's the formula that I'm assuming you're using:

    https://uk.grundfos.com/service-support/encyclopedia-search/calculation-of-flow.html

    But I guess I'll also need to work out what the heat loss is across the radiators and UFH?
    • CommentAuthorandyman99
    • CommentTimeDec 3rd 2018
     
    I'm just in the process of replacing all CH/HW. I was recommended a Grundfos Alpha 3, which is a smart pump that as I understand it will modulate according to demand (flow resistance?). It has fixed flow settings too. Unfortunately the one I have been sent appears faulty - it is working but has no display and wont communicate with the app so I cant set it up properly. A colleague who has one recommends it highly, as yet I cant comment, except to say that the theory seems to make a lot of sense. I don't really understand pump head/headroom though?
  4.  
    Here's quite a good explanation of head: http://www.pumpfundamentals.com/what%20is%20head.htm

    Reading this, it seems I should also calculate friction losses. In fact, I think these will be more important than head.
  5.  
    Posted By: ComeOnPilgrimReading this, it seems I should also calculate friction losses. In fact, I think these will be more important than head.

    Plumbers don't calculate anything - they put in a variable speed/head e.g. the Wilo pump family suggested above and crank up the flow rate until the system works. From the user point of view you crank down the flow rate (and so the energy consumption of the pump) to the minimum that works properly.
    • CommentAuthorgoodevans
    • CommentTimeDec 4th 2018 edited
     
    Sorry been away a couple of days.
    Posted By: ComeOnPilgrimHere's the formula that I'm assuming you're using:
    https://uk.grundfos.com/service-support/encyclopedia-search/calculation-of-flow.html
    No - I started with the specific heat capacity of water that I used at school in O level physics (4200 J/kg.DegC) - and knowing that 1 litre of water has a mass of 1Kg, and moving/burning/releasing energy at 1 joule per second is the definition of 1 watt of power. That formula in the link was horrible - with its use of Kcal and magic conversion numbers.

    So we have identified two pumps:
    One which will be part of the UFH system - presumably this is the traditional setup with a blending valve, pump, and manifold (with flow control valves and possibly actuators also).
    The other pump will be for the radiators. The Grundfos Alpha3 is a grand pump - and sophisticated - and efficient - but for my liking it's just a little bit too 'automatic' for me - the wilo pump I referred to seemed more controllable with enough flexibility to cope with 99% of situations. But the Alpha range will be just fine if you prefer. In all likelihood with two pumps you will probably end up with the smallest pump in the range.

    Would I be right in thinking the backboiler will use a gravity feed to the heat store? How will you control the temperature to the heat store, Does the heat store supply the DHW, are there coils in the store etc, - several queries here not related to the original question - but may be worth checking with us here. (and for more information a diagram is best, almost essential)

    Also worth checking how much heat is going to be released by your wood burning stove and flue directly into the room it is situated in (i.e. not via the back-boiler). It is likely that you should be looking for a stove where the water jacket surrounds the entire stove - if not the room the stove is in will be unbearably hot - but you'll still need to to keep it lit to get the heat store up to temperature to heat the rest of the house. In fact for this site (www.stovesonline.co.uk/stoves_with_backboilers.html) all of their back boiler stoves release more heat into the room than you need for your entire house on the coldest day!

    You may temperature control the flow to your radiators (like the UFH system) or maybe the radiators will run at the temperature of the top of the heat store. I don't know what temp is considered dangerous to send to a radiator.

    I have already had the difficult discussion with SWAMBO as to why we can't have a stove (it wont be lit often or for long, need an unwanted air vent or expensive self contained boiler sealed from the room, MVHR may cause smoke to be drawn into the room every time the stove door is opened, It'll be me that has to clean and set it 'cus it always is) - which was difficult discussion as she thought (not unreasonably) that being a self build she could have what she liked! 2.5kW is not a large heating load - if you have an alternative heat source for hot water and you still want a stove in the house look at a simple, very small room stove and save yourself the grief of steam vents, pressure valves etc etc, and be prepared that the flue will cost several times the cost of the stove.
    • CommentAuthorTimSmall
    • CommentTimeDec 4th 2018
     
    FWIW, I use a single pump for ufh and rads, which all run at the same temperature. This gets a reasonably low complexity, and max efficiency out of the gas boiler, and allows for future upgrade to a heat pump.
    • CommentAuthorgoodevans
    • CommentTimeDec 5th 2018
     
    Yes Tim - that can work depending on the size of the radiators and the heating requirement of the rooms/floors the radiators are in. It would be nice to know what Pilgrim has Plans to do, what kit has been ordered and what is built/existing so far.
  6.  
    +1 what GE said,

    Your heating demand of 2.5 kw is very low (good) compared to many c.h. installations, and that's only on very cold days, and only part of that needs to go to the upstairs rads. You might just find that picking the smallest pump in the range is already more than enough for the rads.

    For the ufh, sounds like the store will supply hot water which will be mixed down and returned at floor temperature, so the delta-t will be more than the normal 5deg used in GE's calculation, so less flow will be required.

    Conversely if there is a pumped circuit between the WBS and the store, that pump needs to be big enough to dissipate the whole output of the stove (many kW?) so needs to be beefier.
  7.  
    Posted By: goodevansSorry been away a couple of days.

    Would I be right in thinking the backboiler will use a gravity feed to the heat store? How will you control the temperature to the heat store, Does the heat store supply the DHW, are there coils in the store etc, - several queries here not related to the original question - but may be worth checking with us here. (and for more information a diagram is best, almost essential)


    Yes, a gravity feed. It works well so far. The DHW is supplied by coils.


    Also worth checking how much heat is going to be released by your wood burning stove and flue directly into the room it is situated in (i.e. not via the back-boiler). It is likely that you should be looking for a stove where the water jacket surrounds the entire stove - if not the room the stove is in will be unbearably hot - but you'll still need to to keep it lit to get the heat store up to temperature to heat the rest of the house. In fact for this site (www.stovesonline.co.uk/stoves_with_backboilers.html) all of their back boiler stoves release more heat into the room than you need for your entire house on the coldest day!

    The stove releases 5-6kW to the room on full power. However, there is so much mass in the house, and the whole of the ground floor is open plan, with an open stairwell to the rest of the house, that it only heats up the house quite slowly (conversely, the temperature drops very slowly once the stove is off).


    You may temperature control the flow to your radiators (like the UFH system) or maybe the radiators will run at the temperature of the top of the heat store. I don't know what temp is considered dangerous to send to a radiator.



    I'm planning on mixing the temperature of the heat store down with an ESBE 4-way mixer valve. I'd like it running pretty cool, as we don't need much heat.



    I have already had the difficult discussion with SWAMBO as to why we can't have a stove (it wont be lit often or for long, need an unwanted air vent or expensive self contained boiler sealed from the room, MVHR may cause smoke to be drawn into the room every time the stove door is opened, It'll be me that has to clean and set it 'cus it always is) - which was difficult discussion as she thought (not unreasonably) that being a self build she could have what she liked! 2.5kW is not a large heating load - if you have an alternative heat source for hot water and you still want a stove in the house look at a simple, very small room stove and save yourself the grief of steam vents, pressure valves etc etc, and be prepared that the flue will cost several times the cost of the stove.


    Our stove works well, we have a passive heat exchanger from Ventive, so we don't need a sealed air supply specifically for the stove. We got a deal on the stove, and the flue wasn't terribly expensive.
  8.  
    Here's an old schematic
      20170630 Plumbing Schematic copy.jpg
  9.  
    Posted By: TimSmallFWIW, I use a single pump for ufh and rads, which all run at the same temperature. This gets a reasonably low complexity, and max efficiency out of the gas boiler, and allows for future upgrade to a heat pump.

    But presumably the gas boiler has a pump - which would be the equivalent of our pump?
  10.  
    Posted By: goodevansYes Tim - that can work depending on the size of the radiators and the heating requirement of the rooms/floors the radiators are in. It would be nice to know what Pilgrim has Plans to do, what kit has been ordered and what is built/existing so far.

    So far, we have a stove, heat store, circuits for the UFH and manifolds and circuits for the rads. The fire and heat store are working well, though it hasn't been that cold so far.
    • CommentAuthorTimSmall
    • CommentTimeDec 6th 2018 edited
     
    Posted By: ComeOnPilgrim
    But presumably the gas boiler has a pump - which would be the equivalent of our pump?


    There is a total of one pump in the system. This is in the boiler, but it needn't be (i.e. instead you could run an analogous setup with a boiler which doesn't include a pump, plus a pump external to the boiler).

    I was trying to say that, it is not always necessary to run the rads at a different temperature to the ufh (an arrangement which nearly always requires an extra pump, and also causes a drop in system efficiency due to the increased flow temperatures required for those smaller radiators).

    A final point is that you may need to ensure that your underfloor heating flow rates are sufficiently high to cause turbulent flow in the ufh pipes. I'm not completely sure what the reduction in heat transfer would be if that wasn't the case though...
    • CommentAuthorgoodevans
    • CommentTimeDec 7th 2018 edited
     
    Much depends on how warm you need upstairs.

    When the WBS is on you will have a large pocket of warm buoyant air form above it - this will float through open doors and up stair wells. (I'll be posting in another thread soon to quantify).

    when the WBS is not on then the UFH will generate warm-ish air on the ground floor which, if upstairs is cooler - will bubble up through the house also but to a lesser extent than the WBS because the temp difference will be less.

    Assuming that you do not want to open doors upstairs to heat the rooms then radiators is a reasonable solution. As you have no requirement to run the rads at a low temp (as the store will have plenty of hot water and you do not need the low temps for heat pump efficiency) regular sized rads will be fine.

    The difficulty with using one pump for both UFH and rads is that the pump needs to be on the heat store side of the UFH mixer - but if / when the UFH mixer is letting in only a little bit of hot water into the UFH loops the flow rate will be so slow only the first part of each UFH loop will get the heat.

    I think you have two choices - large rads and the UFH and rads at the same temp, one pump solution, 3 way mixer valve. (however possibly very difficult to get the rads to match the output of 60m2 of UFH).
    or
    Two independent circuits one each for rads and UFH. UFH with 3 way valve and pump in traditional setup, rads heated with flow from the middle of the heat store plus a "check valve" to prevent rads warming up by gravity feed.

    Mixing valves that mix down to the sort of temperatures you need for your ground slab are difficult or expensive to get. One solution is to get a regular UFH mixing valve and a separate 2 port zone valve - continually circulate the water around the UFH loops and measure the slab or return temperature with a thermostat that controls the 2 port valve. Use the mixer valve as a safety measure to prevent cooking the slab in case the 2 port valve fails. I'm not sure what your 4 port valve gives you.
    • CommentAuthorTimSmall
    • CommentTimeDec 8th 2018 edited
     
    Posted By: goodevansMuch depends on how warm you need upstairs.

    As you have no requirement to run the rads at a low temp (as the store will have plenty of hot water and you do not need the low temps for heat pump efficiency) regular sized rads will be fine.



    I found in my case, that due to the way the individual room heat losses worked out, I ended up with rads which were "regular sized" despite the 34℃ flow temperature (at -5℃ outdoor temp).

    ... so I'd still recommend seeing if running the rads at low temperature is viable (i.e. calculate how large the rads would need to be in this case), since:


    • Store usable capacity is increased

    • Plumbing is simplified

    • Two fewer moving parts in the system which will require periodic replacement (pump, blender valve)

    • More comfortable rooms because of less air stratification and more radiant heat

    • Future proofed system



    If you think the resulting radiators would be too big (for aesthetic reasons) or the setup is uneconomic, then go with the dual temperature multi-pump setup...
  11.  
    Posted By: goodevansThe difficulty with using one pump for both UFH and rads is that the pump needs to be on the heat store side of the UFH mixer - but if / when the UFH mixer is letting in only a little bit of hot water into the UFH loops the flow rate will be so slow only the first part of each UFH loop will get the heat.

    Thanks both! I'm still trying to get my head around this idea of a single pump- if there is only one mixing valve (off the heat store), then surely there won't be another mixing valve at the UFH manifold - thus Tim's idea of one pump for the whole system might work?

    I'd be willing to give it a go and see if I can pull it off with having both the UFH and radiators running at a low temperature. From what I've seen, I don't think we'll need very much heat at all in the radiators. However, how to size the pump in that case? Surely I'd need something potentially very powerful that is able to simultaneously push the water around the radiator and UFH circuits?
    •  
      CommentAuthorDamonHD
    • CommentTimeDec 9th 2018
     
    You'll surely only have flow demands everywhere when rooms are cold. Once up to temperature your rad TRVs (or better, Radbots!) will close and they will not require flow for example.

    Rgds

    Damon
  12.  
    Posted By: WillInAberdeenFor the ufh, sounds like the store will supply hot water which will be mixed down and returned at floor temperature, so the delta-t will be more than the normal 5deg used in GE's calculation, so less flow will be required.

    Thanks Will! Not sure I understand this. I was thinking that the flow to the UFH would be a certain temperature, and the return would depend on how cold it was - when the weather is cold the return will be cooler than when the weather is warm?
  13.  
    Posted By: WillInAberdeenConversely if there is a pumped circuit between the WBS and the store, that pump needs to be big enough to dissipate the whole output of the stove (many kW?) so needs to be beefier.

    No, there's no pump between the stove and the store. It just works on siphoning. The stove manufacturer was a bit skeptical that it would work unless the pipe diameter was large (75mm they were suggesting), but 28mm seems to work very well.
  14.  
    Posted By: WillInAberdeenFor the ufh, sounds like the store will supply hot water which will be mixed down and returned at floor temperature, so the delta-t will be more than the normal 5deg used in GE's calculation, so less flow will be required.

    Thanks again @Will and @goodevans! Can I ask how the delta-t is chosen to be 5º? I guess this is some kind of heuristic that balances lower flow rates (so lower cost) with sufficient flow for UFH to reach all parts of the underfloor loops.

    I assume that radiator circuits would have a larger delta-t and lower flow rate?

    If so, how to choose a delta-t that is suitable for both UFH and radiators?
  15.  
    Posted By: DamonHDYou'll surely only have flow demands everywhere when rooms are cold. Once up to temperature your rad TRVs (or better, Radbots!) will close and they will not require flow for example.

    I guess the calculation above is a worst case scenario, either when the weather is cold, or for when the house is unusually cold (say after a winter holiday!). I'm presuming that the pump will always pump at the same speed, but have less head / use less energy when most of the circuits are closed.
    •  
      CommentAuthordjh
    • CommentTimeDec 9th 2018
     
    The delta-T depends on the flowrate and the rate that heat is dissipated from the 'radiator', whatever that may be. Bigger flowrate reduces delta-T. Greater dissipation increases it. Hence the desirability of balancing radiators etc to have approximately the same delta-T. Which in turn requires the areas to be designed correctly in advance.
  16.  
    Posted By: djhThe delta-T depends on the flowrate and the rate that heat is dissipated from the 'radiator', whatever that may be. Bigger flowrate reduces delta-T. Greater dissipation increases it. Hence the desirability of balancing radiators etc to have approximately the same delta-T. Which in turn requires the areas to be designed correctly in advance.

    Thanks @djh, I understand that principle, but I was wondering why a delta-t of 5º was normally chosen for the whole system, rather than say 3º, or 10º or whatever?
  17.  
    Hi CoP, you mentioned you would take water from the thermal store (I guess this will be about 50-60-70 degC?), pump it round the UFH blend loop and it will return at floor temperature? So your delta-T will be something like 60-25 = 35degC ish.

    In other words, each litre of water will give up about 35degC worth of heat, so you will need to pump far fewer litres than the standard assumption, which is that each litre will only give up 5deg of heat.

    The standard assumption is based on heating water in a boiler to say 30 degC and getting it back at 25 ish degC, which is more efficient for condensing boilers and heat pumps, but irrelevant for your wbs.

    As you said, 5degC is a rule of thumb for typical gas boiler systems, that balances the size/cost of the pump, pipes, ufh etc.
  18.  
    Posted By: WillInAberdeenHi CoP, you mentioned you would take water from the thermal store (I guess this will be about 50-60-70 degC?), pump it round the UFH blend loop and it will return at floor temperature? So your delta-T will be something like 60-25 = 35degC ish.

    In other words, each litre of water will give up about 35degC worth of heat, so you will need to pump far fewer litres than the standard assumption, which is that each litre will only give up 5deg of heat.

    The standard assumption is based on heating water in a boiler to say 30 degC and getting it back at 25 ish degC, which is more efficient for condensing boilers and heat pumps, but irrelevant for your wbs.

    As you said, 5degC is a rule of thumb for typical gas boiler systems, that balances the size/cost of the pump, pipes, ufh etc.

    Thanks Will. Isn't a gas boiler likely to have a larger delta-t of say 25-30º to ensure that it is able to condense?

    My understanding was that a lower delta-t is better for UFH as it means that the water is circulated more rapidly, which is less likely to lead to cold spots. However, one way of achieving a high delta-t for the system, but a low delta-t for the UFH would be to have a second mixer valve at the UFH manifold, which mixes the main central heating loop (also used for the radiators) down to a lower temperature. However, this would mean that the one temperature, one pump system recommended by @TimSmall would not be possible.

    I'm assuming that a higher delta-t will be more economic, as it would entail less water being pumped around the system. But there are other advantages and disadvantages, so I guess it's a question of balancing all these out. I should probably try and make a list of pros and cons of each.
    • CommentAuthorgoodevans
    • CommentTimeDec 10th 2018 edited
     
    Pilgim, I simply chose a delta T to answer the original post -at that stage the question was just a "how much heat can a pump deliver".

    A one pump solution is going to be potentially tricky in your situation. The argument goes like this...
    1)The pumped water will be the same temperature in the UFH and Rads.
    2)You have 60m2 of radiator on the ground floor (assume effective UFH rad size of 30m2 with ceramic tiles, 15m2 with carpet).
    3)You have 2 further floors to heat so we need to get around half the heat output upstairs through radiators.
    4)Can you get 15 to 30 m2 of radiators upstairs.

    It's a question of balancing the heat throughout the house.

    Radiators at v. low temperatures really do work as radiators but as they get warmer they act more like convector's as well and output proportionally more heat (particularly the multi panel radiators).

    HOWEVER - you have a supply of water that is nice and warm and that make it easier.

    One way to sort out the problem of only having one pump is to run the circulating water at say 40 degC to allow for a reasonable output to the radiators but is safe for the UFH. It would obviously be necessary to cut off the flow to the UFH when downstairs is warm enough.

    Alternatively 2 or more heating zones (thermostat and 2 port zone valve for each zone perhaps one on each floor) If a thermostat demands heat in one or more zones then the pump runs and the necessary zone valves open. Simple and easy to understand.

    manifold actuators could be used instead of zone valves if even more zones required. At those temps a small pump can easily deliver 2.5KW
  19.  
    Thanks @GoodEvans! It seems inevitable that I'll need to go for a 2 pump solution. That way I can run the UFH at a lower temperature and higher flow rate.

    I guess I really need to work out what delta-t I should be using for the main / radiator circuit. The particular temperature that the circuit runs at does not seem to be totally crucial (that said, presumably, the higher the circuit temperature is compared to the ambient temperature, the more heat will be transferred), but I suppose the delta-t is more important. I would assume that for the main pump, a large delta-t (say 20?) would be good as it would mean that the flow rate of the pump could be relatively low, and therefore save energy.
   
The Ecobuilding Buzz
Site Map    |   Home    |   View Cart    |   Pressroom   |   Business   |   Links   
Logout    

© Green Building Press