Click here to Find a Contractor in your area.
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/.

Add radiant ceiling to get return temp way down in a low-load house?

3IMOH
3IMOH Member Posts: 7
Hi All-

I am looking for advice in trying to make our existing hydronic system more efficient in a super-insulated house that I have been building for the last 4 years. We have been living in it for 2 years.

All of our DHW and heat comes from a CO2 based heat pump from Sanden. We were involved in a WSU/BPA research study that tested out this unit on 10 homes. Here is a link for more info on the study.

The house is near Portland Or. It is 2 story, 2000 sf and has a heat load of 6500 btu/h. We have 2 panel radiators, one on each floor that are fed by a Taco X-pump block pulling from the Sanden DHW tank. We keep the setpoint at 71-72 and it has never gone below 70 in winter...we certainly don't have any complaints about comfort.

The issue is the efficiency of the Sanden heat pump when used for heating loads. The unit was designed to heat DHW, where any hot water used would be replaced by cold water at the bottom of the tank. It draws this cold water and heats it in one pass to 160 degrees (depending on setpoint), storing it in and 80 gallon tank. The outdoor compressor depends on getting cold water to operate efficiently. Here is a chart showing what happens to the unit at various input water temps:

| Inlet Water Temp °F | Capacity kw/ Btu/h | Unit COP | Flow Rate (GPM)
| 50 | 4.5 / 15,400 | 4.5 | 0.32
| 80 | 3.4 / 11,600 | 3.3 | 0.38
| 100 | 3.0 / 10, 200 | 2.7 | 0.45
| 115 | 2.6 / 8,900 | 2.4 | 0.49
| 120 | 2.4 / 8,100 | 2.0 | 0.58

When we are heating, the storage tank eventually gets mixed together and the water in the bottom of the tank going to the heat pump gets to around 110 degrees. As you can see, that temperature is really hurting the performance of the unit.

Not only is the COP affected but our first year we occasionally ran into issues with cold(ish) showers when the HP capacity was reduced when we were also calling for heat. The HP would eventually catch up but with 2 teenage daughters who really like showers but were not very interested in why they got cold, I installed a seisco booster on our DHW line. This only comes on briefly in situations where the DHW temp from the tank gets below 108 or so.

The house stays warm and the water stays hot but it is clear that using this particular unit from Sanden for DHW and heat has challenges. The other houses in the study ran into the same thing. No knock against the company or the unit itself...it was never designed to be used as a heater and it is only feasible on houses with really small loads. In fact, Sanden made some changes to the UL approved unit specifically to take care of a defrost issue with warm return water (which the unit was never expecting to see in a DHW only situation). Support from John Miles at Sanden has been great for this "experiment".

So what to do? I have briefly thought about just putting a mini-split in for heat and just letting the HP take care of DHW like it is supposed to. I don't want to go that route though given the time and energy I have put into this system. I also really prefer that silent heating of hydronics...well, except for that one squeaking 008 cartridge I had to replace.

So I am committed to making this system work the best it can but I need to get the heating loop to run at a much cooler temp. I can't do that now because my 2 radiators were sized tightly to the load expecting 130 degree input water.

So there is no other option except adding more radiation...a lot more. Fortunately, (depending on your perspective) we are still living in a 80% completed house. I have some key ceilings that are still open and I have access to the existing radiator loops. My plan is to add as much radiation in the ceilings as I can after the current radiators. My goal is to have the return water as close to ambient as possible...as you can see from the chart above, 70 degree water would make the HP a lot happier.

I have about 320 square feet of ceiling that I could easily work with downstairs, There are also some other areas I could add loops but they would take more work. I also have 24 sqft upstairs in a soffit too near the rad there. My current plan is to use the heavy thermofin C plates on 6" centers. I am also considering doing part of this in copper though it would be great if I could pull it off without it.

I am looking for any and all advice on this...has anyone tried to run really low water temps in a low-load situation? I can provide a lot more detail on my existing setup but I figured this is plenty to digest right now. If you have read this far, thanks much!

Howie
Portland, Or.

Comments

  • hot_rod
    hot_rod Member Posts: 15,065
    It could be simply how the buffer tank is piped. For heat pumps a 3 pipe buffer helps assure the lowest possible return, minimal blending, an example on the cover of this application guide..

    I believe this application manual is available at the SpacePak website. It has some of the best, most current design information on buffer tank use with A2WHPs
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
    Gordy
  • 3IMOH
    3IMOH Member Posts: 7
    Thanks Bob-
    The return location was discussed a good bit in that study paper. They found that for sites with radiant slab the return temp could get low enough (80 or so) to return to the tank at the bottom where the HP draws (similar to the drawing you posted). One of the issues with my setup is those nice blue lines in the drawing going back to the tank are not very blue here. Initially, high return temps were sent to the top of the tank. (there is no middle fitting on this tank) This caused issues with cold showers during heating as the water mixed at the top and the temp dropped. The researchers made what they called a "diversion fitting" which is a tube that mounted in the top of the tank that directs return water down to hopefully end up more towards the middle of the tank. This is what I have installed now. I am considering removing it and just having the return mix with the water at the top since I now have the booster on the DHW line to deal with the cold shower issue.
  • 3IMOH
    3IMOH Member Posts: 7
    Does anyone have an idea of the btu/sf using thermofin C plates with copper at 6" in a ceiling? I am trying to do some rough calculations to see what is feasible and all the formulas for radiant ceilings are geared towards using pex at 8".

    I can already see trying to utilize the existing radiators on the same loop as the ceiling radiant is going to be an issue. They really will not be providing any output at the temps I am looking at.

    The other problem is that I would need to keep the flow way below 2ft/sec in my existing 1/2" pex loops. I may need to look into re-piping in 3/8".



  • hot_rod
    hot_rod Member Posts: 15,065
    I think Dale has some output graphs at the Radiant Design website.

    Really the output comes down to surface temperature and ambient room temperature, .71 multiplier is about right

    110F ceiling temperature- 70F room X .71= 28 BTU/ft output. Insert your own numbers.

    The copper would not increase output, but may allow a bit lower SWT being a better, quicker conductor.

    Tighter tube spacing would have the same effect allowing lower SWT, but not increasing output so much, it has to do with consistent surface temperature the tight spacing and transfer fins afford.
    I doubt much gain going tighter than 6" OC spacing?
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • Gordy
    Gordy Member Posts: 9,514
    To give you an idea, my old ceiling radiant which was 3/8 copper tube imbedded in plaster 6” o.c would run about 87 degrees surface with 72 Setpoint, and a supply temp of 105 degrees on a -5 OAT day.

    50’s construction.
    SENWiEco
  • 3IMOH
    3IMOH Member Posts: 7
    So the .71 multiplier would not increase with 6" spacing instead of 8"? I am not understanding how that works.

    I took a stab at calculating things using the available 320 sqft of ceiling with the .71 multiplier. If I hit the existing radiator first with 106 entering water and a .2 gallon a minute flow then my return would be right near 70 degrees after the ceiling loop. I run ODR so that radiator is really not going to be doing much anymore except when I get near design.

    I see that people have run flows this slow in 1/2" pex but it is way below the "official" minimum. It seems the concern is getting air out. I am thinking I could setup some valves to bypass the flow meters at the manifold so I could purge the lines at full flow if it became an issue. I am already running a microbubbler on the return after the manifold.

    I guess the other issue is laminar flow? Should I look into using hard ells to try and keep things turbulent? I know you would normally try and keep the pressure loss down but with such a low flow is that really an issue here?

    The total length of all of this at 6" OC is going to be around 800 feet. I guess I am going to need to break this up. At such a low flow, could I get away with 2 400' loops?


  • SENWiEco
    SENWiEco Member Posts: 137
    Hi @3IMOH, Yes I believe the .71 multiplier is ONLY valid for the 8" O.C. design shown in John's Book. It should increase with 6" spacing I think. I am still trying to track down how much it would increase.
    Sean Wiens
  • 3IMOH
    3IMOH Member Posts: 7
    Thanks @SENWiEco

    Let me know what you find. Turns out there is no way I can run this loop at .2 gpm as I thought because the flow will be laminar. I found a site to calculate the reynolds numbers and I am for sure going to have to go to 3/8 pex. Even then at the lowest flow to keep things turbulent, I can't dump enough btu through the ceiling to get the return water to 70 now. If I can find out the output of 6" centers I will have a better idea if I can make this work.

    SENWiEco
  • hot_rod
    hot_rod Member Posts: 15,065
    I think Siggy crunched some numbers years ago about how much if any advantage to spacing tighter than 8 or 9" in a slab.

    I think it comes down to how consistent the temperature spread across the surface. Slabs always stripe to some degree, cooler between the loops, warmest over the tube.

    Ideally that is the job of the aluminum transfer plate or covering, it to evenly move the energy cross the entire surface. The width of the aluminum transfer plate limits how closely you could space the tube, and 6" is about as tight as the loop end could be with 1/2" tube.

    Dale at Radiant Design did some FEA and created isotherms drawings that showed temperature spread at different spacings for his ThermoFins, you might check his site.

    And of course is the end of the loop and ceiling gets lower that the room temperature, you have a radiant cooling panel :)
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • SENWiEco
    SENWiEco Member Posts: 137
    Yes there is diminishing returns on reducing spacing in high mass systems (such as slab), but on low mass, it makes a significant difference to decrease teh spacing. I have been told by others as well, that 6" is as tight a turn you want to make with 1/2" Pex.
    Sean Wiens
  • 3IMOH
    3IMOH Member Posts: 7
    Along those lines I really wonder what using copper would do to the output. I have a set amount of ceiling I can work with and I am now constrained to around .4 GPM to keep the flow from going laminar in 3/8" tube, so any changes in heat flux can make a big difference for me.

    Speaking of laminar flow...I could I combat this by putting some hard bends in my tubing runs? I am not sure if that would allow me to use some lower flows. A couple of days ago I did not know a Reynolds number from Reynolds wrap, so I am trying to learn as much as I can here.

  • Gordy
    Gordy Member Posts: 9,514
    You can go tighter by spooning the returns.
  • SuperJ
    SuperJ Member Posts: 605
    What about running the panel Rads on constant circulation, and putting a 3 or 4 way electronic control valve controlling the return temperature to about 70-80degF depending on the outdoor temperature?

    Can you just get bigger or thicker Rads? It's probably cheaper to up size your Rads then add a whole new radiant ceiling panel. My panel rads often run with a leaving water temperature basically equal to room temp.
    You want a tightly sized heat source, but oversized emitters can be more efficient and more flexible with heat sources. And when you bring the temp down they aren't oversized any more.

    Also, a modulating control on each panel rad like a TRV will let you maximize the temp delta, since it doesn't allow flow unless the room needs the heat. Also, make sure the balancing adjustment is set to an appropriate CV.

    My rads are often hot on top and feel cold to the touch on the bottom because the TRV is only letting a trickle of hot water in. By the time it comes out the bottom it's cold.

    Make sure you're piped so the hot water goes in the top and out the opposite bottom side for max delta. Many panel Rads are piped in and out the bottom which leaves some capacity and delta on the table.

    Another idea is to run all your return thru another emitter. I'm think of getting a couple cast iron Rads to sit in my basement. I'm going to run the return water from the rest of the house through them to increase my delta.

    I've seen runtal radiators on the ceiling before. The kind with flat bars.
  • Gordy
    Gordy Member Posts: 9,514
    edited February 2019
    As @hot_rod eluded to its pretty hard to get the panel return temp below room Setpoint.

    If set point is 72 that’s it. Below that if possible you add load to the space.
  • SuperJ
    SuperJ Member Posts: 605
    edited February 2019
    > @Gordy said:
    > As @hot_rod eluded to its pretty hard to get the panel return temp below room Setpoint.
    >
    > If set point is 72 that’s it. Below that if possible you add load to the space.

    Yes it's impossible to go below ambient. But, a 5-10deg offset from ambient should be achievable depending on the load.

    I've observed it can approach room temp pretty closely with TRVs, properly set balancing valves, proportional pressure controlled pumps and generously sized panel Rads. It's really about not letting any extra water move beyond what you need to satisfy your load.

    I mentioned beefing up the panel Rads and revisiting the control/piping strategy since the heating load is only 6500btu. It's hard to believe you couldn't fit some panel rads with a 120degf or lower design point and a 40degf delta target with such a small load.

    Some panel rad manufactures have de-rating charts with just about every entering water temp and delta you could ever want. Once you have the right Rads installed you need to control them closely to avoid losing your exergy.

    Here is a rad sizing calculator. I've attached an example. 6500 btu is about 2000watts. So I ran with 1000w (assuming you'll do two). 54C supply and 25C return water temp, and 20C room temp. If you raise the SWT, or RWT the recommended panels will shrink accordingly.

    https://www.purmo.com/int/products/3.htm
  • 3IMOH
    3IMOH Member Posts: 7
    Thanks @SuperJ

    I appreciate the thoughts. Unfortunately, I can't really add longer/larger radiators as the space would not support it. My walls are already finished it is just the ceiling I have not completed yet...it is going to be t&g wood and since the joists are still exposed I was hoping to put the radiation there.

  • SuperJ
    SuperJ Member Posts: 605
    edited February 2019
    3IMOH said:

    Thanks @SuperJ

    I appreciate the thoughts. Unfortunately, I can't really add longer/larger radiators as the space would not support it. My walls are already finished it is just the ceiling I have not completed yet...it is going to be t&g wood and since the joists are still exposed I was hoping to put the radiation there.

    Ok then, it make sense to use what's available and fits in your space.

    Keep in mind you can get "type 33 rads", that fit in the same footprint, but are thicker (instead of longer/taller), should be an easy remove and replace.
    Or higher end Jaga style rads might fit in the footprint you have with a built in high efficiency blower to maximize heat output at low temps. These can output a lot of heat at low temps (there charts go down to 95f in and 85f out). They are pricey though, but maybe cheaper than installing a radiant ceiling.
    http://www.jaga-canada.com/jaga-products


    Here's a view of the type 33 rads