NEW IN-FLOOR SETUP IN THE SHOP

  • Active since 1995, Hearth.com is THE place on the internet for free information and advice about wood stoves, pellet stoves and other energy saving equipment.

    We strive to provide opinions, articles, discussions and history related to Hearth Products and in a more general sense, energy issues.

    We promote the EFFICIENT, RESPONSIBLE, CLEAN and SAFE use of all fuels, whether renewable or fossil.
Status
Not open for further replies.

jimdeq

Member
Apr 23, 2010
205
northeastern wisconsin
I am starting the second phase of my boiler system. I want to get my shop infloor up and running. My Tarm and 1140 gallons of pressurized storage are in my home with the storage in the basement. My heat loads in the house run off a Grundfos Alpha 15-55 which will also push water to a heat exchanger on the back wall of the house where I have 1-1/4" Thermopex. The Thermopex runs 150' to the shop. In the shop I have 1200 square feet of heated space. The Grade Beam slab has 1.5" high density foam underneath and 1" high density around the perimeter. The walls are also spray foamed and then batted. Ceiling will be blown in fiberglass. I installed 6 loops with a average of 275' of Wirsbo Hepex O2 barrier in the floor. My question is how large of a pump do I need to pump 150' from the house to the shop? I was thinking I would put the mixing valve in the shop instead of the house so I could still add a high temp unit heater if needed. Also what size pump would run the infloor and can the same pump supply the high temp unit heater? I also ran CAT 5 wire in a conduit from the house to the shop for control wiring. The system will be filled with glycol. I am looking forward to any suggestions and imput.
 
Sounds like you're golden, Jim.

With radiant heat you can expect nice big deltaT, call it 70 degF. At 2 gpm that's 70000 btu per hour. Any small steep curve pump should work for pumping the 300 feet round-trip, a 008, 15-58 for instance. If you've got a 007 laying around I'd even try it before springing for a new pump.

As a point of reference on the slab pump, I just hooked up a three-speed 15-58 -- because that's what I had laying around -- on a 1000 square feet of radiant flooring an it's doing a nice job on low speed with four 300 ft 1/2" PEX loops pulling through a 3/4" mixing valve. The mixing valve should definitely go in the shop.

--ewd

[Edit:]

Had another thought on this. For as much flow as it will take to feed 1200 sq ft, a single circ pulling though a mixer and all the way from storage could do the trick perfectly. Later another pump in parallel could feed the high temp unit when you're ready for that.

This would eliminate the problem of balancing the transfer flow with the demand flow and it would guarantee minimum return temperature to storage to prevent mixing of storage.
 
ew: With radiant heat you can expect nice big deltaT, call it 70 degF. At 2 gpm that’s 70000 btu per hour.

Is this really correct in a radiant slab application? While the math is right, isn't the btuh calculation based on delta-T between slab supply, slab return, and gpm? The mixing valve is not likely going to be set 70F higher than the slab. For example, I supply the six loops of 1/2"pex (275-300') in my floor at 100F, return is at about 70F, delta-T=30, gpm 2.75 based on the little flowmeters on the manifold, btuh = 41,000.

Of course, if the supply to the shop has a 70F delta-T with the mixing valve in the shop, the extra btu's on the slab return could be used in a unit heater, so 70,000 btu's are available; or the unit heater could have optional plumbing to allow it being supplied first with it's return to the slab and also a bypass to the slab, which I think is the way I would go. That would give immediate high heat boost if needed, and I bet the unit heater return is still high enough (100F+) to adequately supply the slab. In this scheme system delta-T could be 100F+ and > 100,000 btuh.

jimdeq, what's the size of your Tarm? How much extra btu's do you think you have available after heating the house? Your 1200 sq ft shop is a little smaller than mine (1500 sq ft, six loops) and by your description is better insulated than mine. With your location NE WI and mine in northern MN, our weather may not be too different, unless your temps are moderated by Lake Michigan or Lake Superior. Based on last winter, perhaps a little colder than the current "normal," my highest 24 hour heat load in the shop was about 17,000 btuh, more normal average winter heat load was 12-14,000 btuh. I keep the floor temp at 61F and 1F differential. The floor sensor shows a floor temp range 61-64F.
 
... or the unit heater could have optional plumbing to allow it being supplied first with it’s return to the slab and also a bypass to the slab, which I think is the way I would go.

Which is the way I did my plumbing. I have stub-outs in place for a unit heater which I have but have not installed. If installed, unit heater first, slab mixing valve and slab second. With the floor set at 61F, I never have needed an extra heat boost.
 
jebatty said:
ew: With radiant heat you can expect nice big deltaT, call it 70 degF. At 2 gpm that’s 70000 btu per hour.

Is this really correct in a radiant slab application?

I think the OP is proposing two pumps, a transfer pump and a slab circ. The transfer pump would have the advantage of working with a large deltaT, you're right about the slab part.
 
I missed that, good observation. Might one circ work anyway?
 
jebatty said:
... Might one circ work anyway?

You bet, I was editing my original post with this suggestion as you were posting yours. I think a single pump pulling through a mixing valve and all the way from storage for the slab would be ideal. Should run the numbers, but I wouldn't be surprised if a single 15-58 would do it, or maybe a pressure controlled ECM Wilo Stratus or Grundfos Alpha, they're getting more and more affordable.
 
jimdeq said:
I am starting the second phase of my boiler system. I want to get my shop infloor up and running. My Tarm and 1140 gallons of pressurized storage are in my home with the storage in the basement. My heat loads in the house run off a Grundfos Alpha 15-55 which will also push water to a heat exchanger on the back wall of the house where I have 1-1/4" Thermopex. The Thermopex runs 150' to the shop. In the shop I have 1200 square feet of heated space. The Grade Beam slab has 1.5" high density foam underneath and 1" high density around the perimeter. The walls are also spray foamed and then batted. Ceiling will be blown in fiberglass. I installed 6 loops with a average of 275' of Wirsbo Hepex O2 barrier in the floor. My question is how large of a pump do I need to pump 150' from the house to the shop? I was thinking I would put the mixing valve in the shop instead of the house so I could still add a high temp unit heater if needed. Also what size pump would run the infloor and can the same pump supply the high temp unit heater? I also ran CAT 5 wire in a conduit from the house to the shop for control wiring. The system will be filled with glycol. I am looking forward to any suggestions and imput.

Don't over look that the water should be pulled threw the valve and not pushed.
 
Do you really think you need glycol? It reduces the efficiency significantly. If you had the backup heater you could fire the slab if temps got too low (say 40F). You even just circulate the water in the loop if your supply is below the frost line. With the thermal mass of the slab and the insulated underground pipes I think your freeze risk is very low. If you are concerned about the above slab piping, build a box around it, insulate well, and add a thermocube controlled heater.
 
Hunderliggur said:
Do you really think you need glycol? It reduces the efficiency significantly. If you had the backup heater you could fire the slab if temps got too low (say 40F). You even just circulate the water in the loop if your supply is below the frost line. With the thermal mass of the slab and the insulated underground pipes I think your freeze risk is very low. If you are concerned about the above slab piping, build a box around it, insulate well, and add a thermocube controlled heater.

+1

If it were a hunting shack used twice a season, that's one thing, but integrated with your house system it would take tiny amounts of heat to keep above freezing, even out in the windswept frontiers.

--ewd
 
I live in central Wisconsin. My neighbor lost his garage in floor to freezing his first year. I am not sure of the details but it is heated by a lp boiler. The first winter cold snap it froze near the overhead doors. He no longer has in floor in his garage. He had other issues with his system so my guess was his installer did something wrong.

Around here it is really tough to find people that are knowledgable about radiant heat. Several people I know that have it complain about their heat bills and ability for house to keep up on temp swings. I believe that it is usually related to poor installs.

gg
 
If you're talking about using glycol throughout the entire system (1200+ gallons) price out a few barrels of glycol before committing yourself to it.

You could use a heat exchanger for the shop circuit and use glycol only out there. Perhaps that is what you're considering.

Contrariwise, you could just put one of those freeze alarm thermostats out there and see if it ever gets close to freezing and if it just isn't a problem, ignore it. You could always add glycol if it turns out to be a threat. Gotta think of the power outage scenario, though. If you had no electricity for a few days during a cold snap would you be screwed?
 
Hunderliggur said:
Do you really think you need glycol? It reduces the efficiency significantly. If you had the backup heater you could fire the slab if temps got too low (say 40F). You even just circulate the water in the loop if your supply is below the frost line. With the thermal mass of the slab and the insulated underground pipes I think your freeze risk is very low. If you are concerned about the above slab piping, build a box around it, insulate well, and add a thermocube controlled heater.

I would agree that glycol is probably an expense not worth it in this case especially if you have a backup heater. I pondered this when I installed my slab radiant and was a bit 'scared' not to but after researching it seemed logical to me not to because of the constant circulation of the floor circuits. If you do outdoor reset with your floor circuits then you could treat your supply circuit from the home as an injection mixing circuit using a controller such as a Tekmar 356 or equivalent and a standard shallow curve pump like a 007 or 15-42 etc. and get constant circulation of that loop with a high delta T. Your shop is 80sq/ft bigger than mine (I don't know your ceiling height so volume might be different) but I find that for the most part I supply the floor with 75-80deg water with a delta T of 10. Last year I had a much higher delta T (less flow) for the floor circuits and found this to be an inefficient method for slab radiant, because with more flow and a lower delta T you can have a much lower supply temp and get the benefit of that huge emitter being the same temp which also lowers distribution losses. I have 6 loops approx. 250 ft long and they each flow at approx. 1/2 gal/min. for a total of 3 gal/min now. Last year I had 0.2gal/min per circuit and while I didn't have to shovel in front of the garage doors, made for an overall inefficient system in my case. Make sure you get a mixing valve that has the same or as close a Cv as your total flow into your floor circuits for me this was a 3.5Cv mixing valve. And yes always pull through a mixing valve.

Find out your heat loss calculations for your building and you can then find out what size pumps or possibly only one pump you will need. I think if you only go with one pump you will need a high head low flow pump to accomplish this because you will be pulling through the mixing valve with it's psi drop the friction loss of the supply loop and the friction loss of the floor circuits but I could be way off base so now I gotta run the numbers to see, because its snowing like heck here and I don't feel like plowing.
 
Ok, so one more cup of coffee and I hope my math is correct.

I making some assumptions (therefore its usually ME that is made the one).

25,000 BTU/hr heat loss.
For slab radiant delta T of 10.
275' of 1/2" pex (you didn't state the size but it is pretty standard).
6 loops.
No glycol.
a = 0.055 avg water at 85 deg
c = 0.786 for 1/2pex and 0.01668 for 1-1/4pex

Ft=Q/k x deltaT = 25,000/500 x 10 = 5 gpm
Fl = Loops/Ft = 6/5 = 0.83 gpm/loop

Hloop = (acL)(f)^1.75 = (0.055 x 0.786 x 275loop) (0.83)^1.75 = 11.88825 x 0.72175 = 8.6 ft of head + fittings = 9.5 more realistic.

Hsupply = (0.055 x 0.01668 x 300supply) (5)^1.75 = 0.27522 x 16.72 = 4.6 plus fittings and psi of heat exchanger = 6 more realistic

Htotal = 15-16

So one pump with 5gpm at 15-16 head loss should do it.

Otherwise one loop pump with 5gpm at 9.5head and one supply pump at 5gpm at 6head. Although if you did injection mixing the supply pump calcs would be different because of the higher delta T lower flow and so forth.

Please check my math as I've only had 2 cups of coffee.
 
For freeze protection and no glycol, constant circulation is the key I believe. Although I do have constant circulation I used a pump exercise add-on board for my pump relay which allows me more piece of mind if the slab sensor or t-stat gets satisfied. If either of those does gets satisfied and the pumps turn off, the add-on board starts a timer and will exercise the pumps up to 4min/hr with no call for heat. Lets me sleep better at night and was very inexpensive insurance.

Time to plow for me.

Good luck and keep us posted with your final design.

I also use a UPS for my shop pump/relay in case of power outages.
 
If you already have your mind made up about glycol then replace the constant 'a' above with 0.07 for an average water temperature of 85deg, or 0.065 for an average water temperature of 100deg. You'll see that more flow hence higher head pumps will be necessary to extract the same amount of BTU's, but it might also let you sleep at night.

Another benefit of near-continuous circulation is the ability of the floor panel to redistribute heat within the slab like the interior portions to the areas of higher heat loss like overhead doors, to prevent freezing for many hours or days because of the high mass system. I think where most failures occur, if they don't have glycol, is because of the lack of near-continuous circulation and/or the delta T being too high, therefore too low of flow, so that the whole slab is not acting as one giant radiator with relatively the same temps everywhere. I would think this problem of too high a delta T or on/off circulation for the slab could also be compounded if the hottest water were not directed around the perimeter first or where the highest heat loss is expected like overhead doors for example. To me it just makes sense to add outdoor reset especially in this application to get the added benefits of near-continuous circulation with a lower delta T.

Anyone have any other thoughts? This stuff is interesting as heck.
 
On shop floor radiant you can get away with a larger ∆T t in the loops. You probably will not be barefoot out there. Just guessing :)

You need to know the pressure drop thru the mixing device as Da mentioned to select the correct pump. Both the grundfos 15-58 and 26-99 are now 3 speed so you have plenty of adjustability.

Remember on first, cold start up you will have a huge ∆T until the slab ramps up in temperature. The thermal mass of that slab will over power your boiler and probably the temperature in the buffer.Don't try to adjust it away for a day or so.

When we fire large commercial slabs in winter time temperatures it can take 3 days or more for the slab to show a "feel-able" temperature, be patient.

Also be sure you protect the boiler against the cold slab return temperature. If that return can get to the boiler, maybe your buffer tank will mix it up first.

hr
 
:) yes very true not many people go barefoot in their shop. So hr what is a reasonable delta T for shop floors? I know that last year I had delta T of 40-50 and that didn't seem to work very well. This year it's at 10 but that is probably too low especially for shop floors. Don't mean to hijack thread but it seems to fit. I keep my shop relatively warm, 65, since I'm in there most of the day.
-Bob
 
DaBackBurner said:
Fl = Loops/Ft = 6/5 = 0.83 gpm/loop

Won't flow rates this low result in flow velocities less than 2 fps which are considered the lower limit for moving air, sediment, etc., through the system?

I am trying to size my circulators at the moment and am running into this issue....my flow rates at a 10F delta T and a 35000 heat load into 8 loops of 250' are around 0.875 gpm at the design point. They would be much less than this most of the time....The minimum recommended flow rates through 1/2" pex to keep velocities above 3 fps are around 1.1 gpm......

What am I missing????
 
Yes they will be lower than 2fps at that flow rate with that size pex. I think that's why it becomes even more important and harder to bleed all the air out of the loops and to make sure you have air/dirt seperators before the loop circuits. The alternative is to increase your loop circuit flows which would decrease your delta T. This is what I did initially to help purge air before dropping flow rates to maintain my delta T of 10. Adjustable flow rate/indicator manifolds really help in this area. If all the circuits are balanced or the same flow rate you would be able to see if there was a restriction in one loop at a glance. It is not 'out-of-the-norm' to have your loop flow rates with that size pex.

When you say your flow rates most of the time...is this for something other than slab radiant like a home-run for panel rads?
 
Those are the flow rates at the design point heat load of 35,000BTU/hr (-30F day here in Wi). The heat load may be half that or even less on a more typical winter day and thus the flow rates would be way less than 0.875 gpm.....

It seems that the most common in slab pex size is 1/2"....is it common for flow rates to be 0.25-0.5 gpm in these systems?
 
I wouldn't expect your flow rates in your loops to change Chris, except for warm weather shut down or slab or t-stat satisfaction (off). Your mixing valve should vary the supply temperature either manually or motorized with a controller, based on heat demand. I could see if you had an injection mixing pump that varied the flow rate of an injection loop therefore taking the place of a mixing valve, but I've never seen or heard of variable flows in the slab loops. They are usually supplied by a fixed speed circulator hence only one constant flow rate. But heck, I'm always open to new ideas if there is a better mouse trap I'm all ears. Now you may want to install a 3 speed pump for your floor loops, in case your calcs are off for tweaking purposes, but once that is done you will have one flow, with variable temperature supply. I would say a fairly common flow rate per loop with 1/2" pex that is between 250-300 ft long would be 0.4gpm -1.5gpm, but it really depends on your heat loss.
-Bob
 
To answer everyones questions, my tarm is "Solo 60". So far above 25 degrees equals one fire per 24 hrs. Lower than 25 degrees equals one fire every 12 hrs. I have figured out to not let the bottom of storage get below 120 due to termovar function upon restart. Jebatty, I liked the option you described, because I don't always want my shop at 60 degrees. If I leave for a week or go on vacation I would probably turn it down and let the unit heater due a quick recovery when I'm home. I understand thermal mass but I want to keep it easy for my wife and kids to do the boiler. I want to use gylcol for the shop loop. I already have about 40 gallons from a friend. The heat exchanger I have in my basement has 1- 1/4 supply and return from boiler, but the problem is it has a 3/4 inch restriction where it ties into the 1 -1/4 inch thermopex. It is a stainless tube in shell pool exchanger.

DBB: I don't understand outdoor reset or injection mixing circuit. If you could explain to me what type of mixing valve has a CV close to total flow? I don't understand what all that means? I have a WATTS 3/4 inch mixing valve, but don't know if that is correct. Also, if I go one pump, will I burn more electricity with the high head pump vs. two small pumps?

Thanks for all the imput. Jim
 
kuribo said:
Those are the flow rates at the design point heat load of 35,000BTU/hr (-30F day here in Wi). The heat load may be half that or even less on a more typical winter day and thus the flow rates would be way less than 0.875 gpm.....

It seems that the most common in slab pex size is 1/2"....is it common for flow rates to be 0.25-0.5 gpm in these systems?

.6 gpm is what the Radiant Panel Assoc suggests for a 1/2" loop. 300' loops are about right.

Really the starting point should be the heat loss of the building. Then, knowing the available "un-encumbered" floor area you build the radiant installation. Divide out any floor area under tool chests that sit on the slab, cabinets, etc. Radiant energy, like sunlight needs to "see" the load, line of sight.

Typically 12" on center, 1/2" tube 300' lengths, at .6 gpm will provide about 25 BTU/ square foot output. Flowing .6 gpm of water thru 1/2' 300' loops would have a pressure drop of 3.8 feet of head, very reasonable.

If the load is higher you can decrease tube spacing, increase supply temperature, or increase flow rate. It's best to make changes to the supply temperature to change floor output. Once the tube is in the slab, that's hard to change.

That is what outdoor reset control accomplishes. Think of it as a "cruise control" for your heating system. It tries to supply the exact amount of energy (water temperature) to match the ever changing load of the building. The load being the outdoor temperature, & infiltration.

To have a variable temperature control you need a smart component. This could be a motorized 3 or 4 way valve, or a variable speed circulator feeding into a manifold. A 3 way thermostatic valve would not be considered a reset-able valve, it only reacts to temperature supplied to it, not the outdoor temperature input.

As always how much control and how accurate you want it depends on how much $$ you want to spend.

hr
 
Status
Not open for further replies.