Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.

Heat pump to radiant ceiling and floor flow rates

Willits
Willits Member Posts: 4

Hello All,

I am looking for some advice about rule of thumb regarding flow rates for radiant heating and cooling.

I have 2 things I am trying to figure out:

Is more or less flow going to give me greater heat/cooling transfer to the floors and ceilings

Should the flow from the heat pump to the buffer/inertia tank be more, the same or less than the flow to the system from the tank, and is it the same for both hot and cold.

With this knowledge bracket the flows and see what best optimizes the heat pump operation and gives the best output.

The system is made up as follows:

I have an air to water heat pump which circulates into a 100l/26 gal buffer/inertia tank (4 pipe layout). The house has radiant floors in the bathrooms and radiant ceilings in everywhere else. The radiant volume is about 57l/ 15 gal. There is about 1,200m/4,000 ft of 10mm Pex run in 50 m/160 ft loops into distributors that then run in 19mm /¾” pipe to radiant manifolds. I am covering around 140m2/1,500 ft2and am in Portugal.

There are 7 radiant circuits 6 for the radiant ceiling and 1 for the radiant floors and the heat pump also heats DHW.

I have the option of setting heat the heat pump circulator from the max of about 30lpm/8gpm down to 8lpm/2gpm. The circulator after the buffer/inertia is rated at 30lpm/8gpm and has 3 speeds and Radiator heating mode (proportional-pressure curve) and Underfloor heating mode (constant-pressure curve).

Currently I have the flow meters in the manifolds to about 2.2 lpm/.6 gpm each for a total of 15.4 lpm/4gpm.

Many thanks,

Peter

Comments

  • Jamie Hall
    Jamie Hall Member Posts: 24,735

    The heating power which your ceilings and floors will provide is a function only of the temperature at which they are operating. Flow rate comes into the picture only with regards to the drop in circulating fluid temperature from the source to the return.

    That said, one does want to have the radiant surfaces have reasonably even temperatures everywhere — no hot spots or cold spots. This is accomplished by the layout of the piping and the length of the individual loop circuits — and by changing the flow rate. Usually one targets a flow rate which will give around a 10 Celsius of 20 Fahrenheit drop between the source and the return; this gives a reasonable balance between pumping power (and at very high flow rates, noise) and even temperatures in the radiation.

    The flow rate from your heat pump into your buffer tank is set differently: here you want the flow to be related to the boiler power, and the return temperature is set by the temperature in the buffer tank. The heat pump loop flow rate is not related to the heating loop flow rates at all. It needs to be set so that the heat pump is operating at its maximum efficiency output for the specific input temperature.

    Br. Jamie, osb
    Building superintendent/caretaker, 7200 sq. ft. historic house museum with dependencies in New England
  • DCContrarian
    DCContrarian Member Posts: 667

    It helps to understand the purpose of the buffer tank.

    The heat pump will modulate and try to match its output to the actual heating load. It has a minimum modulation, about 25% typically, below which it won't run. The purpose of the buffer tank is to absorb the excess output when the actual heating load is below the minimum, to prevent the compressor from short-cycling. When the load is low the compressor will cycle on and off, when it's off heat will flow out of the buffer tank to continue meeting the heating load.

    The heat pump will work best when it can sense the actual load, when the flow through the emitters is the same as the flow through the heat pump and the buffer tank isn't involved at all. The closest way to approximate that is to use 3-pipe plumbing on the buffer tank and a variable flow/constant pressure circulator with the pressure set to approximate the flow of the heat pump in the middle of its range. You also want to enable outdoor reset (water temperature varies depending on outdoor temperature) so that you have more or less constant flow with the temperature drop varying depending on the outside temperature.

  • hot_rod
    hot_rod Member Posts: 23,275

    To answer the question, yes higher flows will move more energy, bot heating and cooling flows.

    The system should have been designed around a specified delta T, temperature difference. the circulator pump is sized and adjusted to that flow rate. Radiant typically design around a 10- 15°∆

    It sounds like you have a delta P type circ. When adjusted properly it will help keep the flows correct as various zones open and close.

    The flow from HP to buffer should be what the HP manufacturer specifies, properly piped the two flows are independant of one another an will not necessarily match, especially with a variable speed circ on the distribution side.

    Are you cooling with the radiant ceiling?

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • DCContrarian
    DCContrarian Member Posts: 667

    >Is more or less flow going to give me greater heat/cooling transfer to the floors and ceilings?

    More flow will always give you more heat transfer. However, there are limits.

    Imagine a heat pump producing water at 110F (sorry, I'm going to use US units here). An emitter has a flow of 1 gallon per minute and water returns from the emitter at 100F. That's a delta of 10F, at 1 GPM that's 5000 BTU/hr. OK, let's say you increase the flow and the water is now returning at 105F. The output of the emitter is going to be determined by the temperature difference between the water and the room, let's say the room is at 70F. The average water temperature in the first example is 105F, in the second it's 107.5F. The difference rises from 35F to 37.5F, so the heat output rises from 5000 BTU/hr to 5357 BTU/hr. To get that heat flow the water flow has to increase to 2.14 GPM. So to get a 7% increase in heat output you have to more than double the flow.

    With water being delivered at 110F the return water can never be hotter than 110F, so the average temperature of the emitter can never be more than 110F, or 40F above room temperature. That would give an output of 5714 BTU/hr, that's the theoretical maximum. But that would require a delta of zero, which means infinite flow, which isn't achievable. What if you wanted 99% of maximum? You'd need a return water temperature of 109.6 or a water delta of 0.4F . To get 5657 BTU/hr with a delta of 0.4F you need a flow of 28 GPM. You're going to need a bigger pipe!

    What if you want to reduce the output by slowing the flow? Let's say you want water returning at 90F instead of 100F. That gives an average water temperature of 100F, which gives an output of 4286 BTU/hr. To get that with a 20F delta requires a flow of 0.43 GPM. So cutting the flow by 57% reduces the output by 17%.

    The summary is that big changes in flow result in relatively modest changes in output. You're much better off changing the water temperature to modulate the output. In the first example, raising the water temperature to 115F results in an output of 6428 BTU/hr, an increase of 28%. Generally heat pumps have outdoor reset which makes that simple to implement.

  • hot_rod
    hot_rod Member Posts: 23,275

    Changing output by varying flow is very non linear as shown on these graphs. A bigger difference at the low end, but the "hockey stick" shaped curve shows a point where it doesn't make sense to keeping upping the flow.

    2-14 shows a typical 250' loop in a concrete slab 12" OC, 110F supply. Notice the output change at the various flows.

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • Willits
    Willits Member Posts: 4

    Hello All,

    Thanks so much for all the input. This is really great.

    From the feedback I feel like I need to be focusing the delta T more than the flow.

    In cooling mode between the heat pump and the buffer am seeing around a 9⁰F delta T and around 6⁰F delta T from the manifolds to the emitters with the pumps running at maximum. I regret that I did not measure these during the winter when it was in heating mode so don’t have a reference yet.

    The heat pump to buffer flows is at 8 GPM and buffer to emitters around 4 - .6 GPM for each of the 6 emitter circuits.

    I am going to switch the heat pump from maximum output to delta T (it is a 9kW LG Therma V and there are a lot of pump options) and the buffer to manifold to the constant pressure setting (Grundfos Alpha 1L). I will see how that effects the performance

    I wonder if the switching from cooling to heating for the DHW has an effect on the cooling. It seems to switch about every 30 minutes. I haven’t yet connected the DHW tank internal electric heater yet so am only heating the water with the heat pump.

    Thanks again,

    Peter

  • hot_rod
    hot_rod Member Posts: 23,275

    With chilled water managing the supply and return temperatures becomes crucial. Typically keeping thje SWT 3° above dewpoint. You need to know of be able to calculate and watch that dewpoint with chilled radiant or panels.

    This is a deep read on hydronic cooling, the same design principle and formulas work for residential systems.

    For heating control the delta T can move up or down from design without causing issues. I'm of the opinion the delta should move around as the load changes, not be constrained by a control or circulator. Measuring or knowing the delta give you an idea of how much energy is being transfered into or out of a space.

    All of this control logic could or should be in a control that is going to manage heating, cooling and DHW functions.

    Another poster mentioned the ThermAtlantic control seems to be sophisticated enough to control HP/ radiant systems.

    https://idronics.caleffi.com/magazine/28-contemporary-hydronic-cooling-commercial-buildings

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • Willits
    Willits Member Posts: 4

    I did think about this when putting the system together.

    I have a dew point sensor on the supply pipe which turns off the circulation pump if it starts to have condensation.

    All the piping is in the attic which is warmer than the interior of the house. So, my idea is that the dew point sensor will shut off the pump before the supply temperature drops below the dew point inside the house.

    The emitters I have are from a German company with 10mm pex tubes in aluminum heat transfer plates sitting on top of a suspended ceiling.

    I only see a little sweating on the stainless manifolds which are not insulated. The insulated pipes do not show signs of dampness.

    With my supply temperature of 55⁰F at the manifold, throughout the day I measure the surface of the ceiling drywall at around most 66⁰F.

    I have also been running the chilled water in the bathroom floors to get an idea of what the pex tube temperature is and its around 61⁰F.

    I think I should be OK on the condensation side, but this is the first cooling season so time will tell.

  • Willits
    Willits Member Posts: 4

    My buffer tank is plumbed as a 4 pipe, suppl and return on each side. The supply from the heat pump connects to a 3-way valve which switches back and forth between the buffer and DHW tank. It is horizontal tank if that makes a difference.

    If I understand it correctly, I can add a T in the heat pump supply (after the 3-way valve) and relocate the system supply there and cap off the supply outlet from the buffer. This would not be too hard to do.

    I assume that when the heat pump is heating the DHW the system circulation pump then draws only from the buffer tank in heating or cooling mode.

  • hot_rod
    hot_rod Member Posts: 23,275

    You need to be very diligent when insulating chilled water piping. Any exposed metal will sweat, then run inside the insulation and cause possible corrosion, wet insulation, etc.

    All the seams and joints need to be glued, especially around tee fittings, like that manifold.

    A 3 piped buffer does provide a bit more efficiency as return water flows directly to the heat pump, and directly to the load, without much tank interaction. Certainly not worth repiping at this point.

    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
    Rich_49