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.

radiant cooling: how much output?

liquid cooled cpus are all the rage these days with gamer geeks. i don't know if this would be practiacal but you could parallel a bunch of cpu HX's together and have a radiator outside somewhere.

i once saw someones mod where they ran a pex line off the cpu to a panel rad with a bronze grundfos recirc pump!

google cpu radiator and see whats out there.

Comments

  • per sq ft.

    how much cooling output could i expect to get per sq. ft. out of 1/2" pex 9" OC embedded in 2" concrete with 55° F water @ 10° delta t? oh, and 75° room air temp.

    my hydronics programs choke when i input these numbers. ;(
  • Mike T., Swampeast MO
    Mike T., Swampeast MO Member Posts: 6,928


    This is a weird (and overly simple) way to answer the question, but under those conditions you would be removing 5,000 btus of heat for each GPM of flow. Presuming a nicely insulated slab-on-grade, nearly all of those btus would be removed from the space you're trying to cool.

    Again, assuming 10F delta-t and those conditions:

    Let's say the room is 200 square feet requiring 260' of tube @ 9" spacing.

    You move 0.6 gpm through this tube.

    5,000 * 0.6 = 3,000 btus of heat removed.

    3,000 / 200 sq.ft. = 15 btu/hr of heat removed per square foot.

    In a 75F room the slab would be about 67F.

    I used the "Rad Pad" for these estimates and haven't even attempted to account for latent vs. sensible heat removal. I do know that unless the structure were built like an ice chest with extremely low infiltration and very little solar gain via windows that such would be utterly inadequate in my area of the country yet there would still be the real possibility of condensation on the floor without separate mechanical dehumidification.

    Of course if you're even considering radiant floor cooling, you probably don't live in the mid-Mississippi valley...
  • GMcD
    GMcD Member Posts: 477
    No latent for radiant cooling

    Radiant cooling cannot, and should not be doing any latent cooling- that's called condensation, and is to be avoided when radiant cooling. I am also concerned that the 55F entering water temperature may be below dewpoint and should be avoided. All of the radiant cooling manuals and design guidelines recommend the entering water temperature to be at least 1.5F above ambient dewpoint in the space to avoid condensation in the first place.

    You cannot "assume" a 10F delta T either, normally most of the radiant cooling systems I've seen and also designed operate with a high flow/low delta T operations, with maybe 6F delta T as a maximum, and more like a 4F -5F as a design delta T. Based on actual measurements from local radiant cooling slab projects I have here, the surface temperature of the concrete will be about 2F above the average fluid temperature in a room at 76F, with tubing covered by 2" of concrete. There are some design charts at these links:

    http://www.h2ohtg.com/TWA/RadiantCoolingInfo/design.html

    http://radiantpanelassociation.org/i4a/pages/index.cfm?pageid=140

    http://doas.psu.edu/panels.html

    http://www.zentfrenger.de/Download/prospekte/Englisch/Zent_Frenger_Batiso_E.pdf

    Som capacity charts in that last link there, too, and the radiant floor cooling chart in the attched document.
  • thanks

    thanks guys, a lot of good ideas. some that i overlooked, it's all coming together i think.

    the actual situation i have is this: i installed a ground source heat pump with three floors of radiant for a customer about 2 years ago. they came to me recently saying that their upper floor is unbearably hot at night right now (and it's not even summer yet) they are very adamant that they want me to set up cooled floors for them, i explained to them that although i've been researching the subject lately i've never actually done a system yet. they replied that they really want me to do it and they don't mind being the guinea pig. i already reccommended that they get some kind of forced ventilation to purge the hot air from the attic space.

    they already have an HRV with a humidistat so that should keep the humidity in check as long as the floors are within the limits. and i had an engineer contact say that in our climate (south central British Columbia) if i have a return temperature of 65°F i should be above the dewpoint.

    so my control strategy thus far is: setpoint control keeping my buffer tank at an as yet undetermined temp.(60°-68°?) tekmar 512 1H/1C thermostat on each of the three pump zones with both relays wired in parallel to control the pumps in heating and cooling. and an aquastat on the returns of the three zones to shut off the pump if the return temperature drops below a pre-set point. (i had thought of having a humidistat in each zone to serve this duty but there arent enough wires roughed in) does this sound feasable or am i way off?

    if it ends up having problems its just a matter of re-doing the controls the way they originally were.
  • GMcD
    GMcD Member Posts: 477
    I'm in Vancouver

    South central BC should be fine with a 61-63F entering cool water temp and use the HRV for the touch-up cooling using the air side- depends how close you are to the lakes and the local micro-climate. Is there de-humidification in the HRV? Chilled water cooling coil from the heat pump? The primary cause of heating and cooling problems in the space is envelope, envelope, envelope. Are there excessive solar gain causing the top floor overheat problems? Is there any envelope modifications that can assist?

    You can find me at www.omicronaec.com
  • envelope

    well, the house in question is two years old, and i can't really figure out why they have such an overheating problem to begin with: they don't have any ridiculously large south facing windows and the ones they do have are double pane low-e. attic insulation should be R-40, but i haven't actually seen it yet. i think they have inadequate ventilation in their attic space however and like i said earlier i recommended they get some forced ventilation going.

    the HRV they have has dehumidification functions but no cooling coil, though i suppose one could be installed.

    they don't expect the floor to handle the whole cooling load, just take the edge off a bit, ie: 78°-80° instead of 90°
  • btw

    Geoff, just been perusing your site, very good work! its great that your sharing all those papers, and on the new green section too. i can feel my brain swelling with every article i read, thank you very much!
  • kal_2
    kal_2 Member Posts: 60
    65f water is more like it...

    you try 55 in the ceiling and you will get showers on a humid day.. radiant cooling absolutely depends on some dehumidification source keeping the air dry – and the 65f in the ceiling will feel positively cold!! – in dry climates (read Arizona) you can use desiccant wheels with outside air – in humid climates – you need to have refrigeration based drying
  • GMcD
    GMcD Member Posts: 477
    Company website

    I'm just "staff" - we're a fully integrated AE/Construction Management company with a few dark green folks.

    For this hot upper floor issue, I'd start by checking the envelope with a surface temperature gun on a hot sunny day to see if there are hot spots from thermal bridging, lack of insulation, or maybe it's an open concept house where solar loads on the ground floors create warm air that rises up and isn't ventilated out properly. Maybe use ceiling fans to de-stratify? At least add the cooling coil to the HRV off the chilled water source and get some supplemental cooling and dehumidification from that as a beginning. And then set up a secondary pump and control valve to send the 62F water to the floor tubing. Depending on your local micro-climate a dewpoint sensor on the cool water supply to the floor tubing could be used as a control point to reset the cool water supply temperature up by 2F if it senses condensation starting on the supply pipe.
  • ceiling

    i'm dealing with floors not ceilings, and i originally thought i'd be able to have a wider delta t, thus if i had 55 in i'd have 65 out, but not the case apparently. i have pump zones so i'll be able to get a fairly narrow delta t and a higher average temp. my climate is fairly dry and we do have some dehumidification with the HRV. enough?: maybe not, i may be adding a water coil to increase it.

    does anyone know of any stand alone coils (ie: without blowers) that i could install inline in the HRV system?
  • GMcD
    GMcD Member Posts: 477
    Local HVAC sales

    While I deal with commercial HVAC stuff- any of the local suppliers that deal with small reheat coils or small water coils should be able to get one- what you need is a water coil that will be a face area such that the maximum air velocity across it is not more than 300 feet per minute, and will likely be a two-row, 10-12 fins per inch, selected for your incoming water temperature and flow range available. I usually call the local Olympic International office in North Vancouver, or the local Trane office.
  • Mike T., Swampeast MO
    Mike T., Swampeast MO Member Posts: 6,928


    Thanks Geoff. GREAT info!!!! The only places I've lived are Swampeast MO and Washington DC and DC is typically better on my "swamp" scale... I'll never forget the first time my wife (who had only lived in the NE) flew into St. Louis on an August evening. She said, "How do you breathe?" Hard for me to separate the latent load from the sensible with night-time dewpoints around 70F quite common from July-Sep.

    In the places I've visited during the summer my personal opinion is there's not much need for A/C unless the temperature is extreme and you can use evaporative cooling.
  • GMcD
    GMcD Member Posts: 477
    Exactly

    In a humid climate, you can achieve good comfort simply by dehumidification, and very little lowering of the dry bulb temperature. ASHRAE-55-2004 has some good human comfort data that supports this.
  • Troy_3
    Troy_3 Member Posts: 479
    panel radiant cooling

    > While I deal with commercial HVAC stuff- any of

    > the local suppliers that deal with small reheat

    > coils or small water coils should be able to get

    > one- what you need is a water coil that will be a

    > face area such that the maximum air velocity

    > across it is not more than 300 feet per minute,

    > and will likely be a two-row, 10-12 fins per

    > inch, selected for your incoming water

    > temperature and flow range available. I usually

    > call the local Olympic International office in

    > North Vancouver, or the local Trane office.



    Geoff: Here's my project- if I could please pick your brain or anyones. I have a 20'x8' computer room. it has racks of computers. When we originally put 10 tons of cooling in this room 7 years ago all was good. we had excess cooling capacity. since then the equipment is constantly upgraded and now even with two extra koldwave water cooled unit 2tons each the room gets close to shutting down at times. This is a critical room and can not shut down. Room is a premium and almost impossible to put an outside condensing unit anywhere. My thout is to install a flattube radiator on the wall behind the rack of computers and pipe it to a water cooled chiller. Humidity is already contrlled by the existing a/c units. How do I size a runtal radiator to give me 5 tons of cooling. any thoughts? Am I way out there or what?
  • Troy_3
    Troy_3 Member Posts: 479
    panel radiant cooling

    Geoff: Here's my project- if I could please pick your brain or anyones. I have a 20'x8' computer room. it has racks of computers. When we originally put 10 tons of cooling in this room 7 years ago all was good. we had excess cooling capacity. since then the equipment is constantly upgraded and now even with two extra koldwave water cooled unit 2tons each the room gets close to shutting down at times. This is a critical room and can not shut down. Room is a premium and almost impossible to put an outside condensing unit anywhere. My thout is to install a flattube radiator on the wall behind the rack of computers and pipe it to a water cooled chiller. Humidity is already contrlled by the existing a/c units. How do I size a runtal radiator to give me 5 tons of cooling. any thoughts? Am I way out there or what?
  • GMcD
    GMcD Member Posts: 477
    Computer room cooling

    Unfortunately radiant cooling isn't a good match to computer room cooling, since almost 100% of the computer heat load is spit out as hot air from cooling fans in the individual equipment in the racks. You'd need to get right into the processors to get at the heat exchange point. Is the room air movement and air balance set up in "cold corridor/hot corridor" underfloor paths? Lots of cold air is the only current conventional approach for computer rooms, until the new liquid cooled processors start being used, where you can pipe chilled water directly through the processor racks.
  • cpu cooling

    liquid cooled cpus are on the rage these days with gamer geeks. i don't know if this would be practiacal but you could parallel a bunch of cpu HX's together and have a radiator outside somewhere.

  • LarryC_4
    LarryC_4 Member Posts: 3
    Computer room cooling

    Do you whole rack cooling fans? If so, perhaps hanging a cooling coil infront of or after the fan can cool the whole rack.
  • Mike T., Swampeast MO
    Mike T., Swampeast MO Member Posts: 6,928


    Granted I haven't been in a computer room since the early 1980s (and a good desktop could probably replace everything there), but 49 kw/hr of heat removal in a 20' x 8' space sounds incredible.

    This paper suggests that Adding More Cooling Isn't Best

    This paper gives some Best Practices/Concepts

    Granted I can't see the space, but it almost sounds as though you have plenty of cooling potential but it's not particularly well directed to removing the heat from the actual sources.

    This Website seems to have LOTS of options for computer rooms.
This discussion has been closed.