First of all, sorry for the long post.  I’ve been browsing this and other forums for many months, I’ve bought the Siegenthaler book, and I think I’ve finished up the first part of my design…1) the design heat load and 2) the tubing layout.  I was hoping some of you experts out there can take a quick look and see if I am in the ballpark.  I’ve attached a couple of my design files if anyone cares to take a peek.  (The design sheet is an Excel file, but you need to add the .xls extension to it to open it.)

 

I am going to be pouring a concrete slab in my garage/shop this summer and am planning on putting a radiant heat system in it.  It is a 36x64 “pole barn”, and I plan on framing off a portion of it (1150 SF), insulating it, and heating it to 45-50 degrees consistently, with no bumps in temp.  I live in Alaska (no permafrost right here) and I’ve figured the HDD for each base temp.  Long story short, I came up with a heat demand of between 25k and 36k btu/hr, depending on if I go with 0.5 or 1.0 air exchanges and 45 or 50 degrees.  This gave me a required heat flux of between 17.5 and 32 btu/hr/SF.  Assuming ½” PEX on 12” spacing and a delta T of 20 degrees, I found a supply water temp of between 75 degrees (for a 45 degree base with 0.5 air exchanges) and 96 degrees (for 50 degrees with 1.0 air exchanges).

 

And now for the questions:

1. 75 degree supply doesn’t seem hot enough to give me 45 degree interior temp, but I’ve checked the calcs twice.  Did I mess up somewhere?
2. I really don’t know if I should be using 0.5 or 1.0 for air exchanges.  I plan on taping the vapor barrier seams and making it as tight as possible, but I am still working with a fairly rough structure, and I think the two overhead doors will end up leaking quite a bit.  Anybody have a recommendation, or should I just average it?
3. My design flow rate through the tubing is between 0.6 and 1.1 gpm.  Siegenthaler’s book (Fig 11-4) calls for a minimum flow rate of 1.2 gpm for ½” PEX in order to maintain 2 fps flow velocity.  I suppose I can just choose a circulator that will give me 1.2 gpm and the result will be a delta T of less than 20 degress, right?  Or is 0.6 gpm OK?
4. What is the lowest realistic supply temperature feasible?  Can I even supply at 75 degrees, or will I be supplying much higher by default?  (See heat source ideas below.)

 

The btu requirements for this design seem to be fairly small for an oil-fired appliance.  (Gas is not available, and electricity is 20 cents/kWh).  I think I have two options for a heat source: 1) oil-fired tank hot water heater or 2) oil-fired tankless hot water heater.  I picked up a used Bock 32e last year that I would like to use if it will work for me.  Otherwise I think I am looking at something that is available up here, like a Toyo OM-180 tankless, but those are spendy.

 

Will a Bock 32e work for me?  Can it handle low temps, or will it have condensing problems like a conventional boiler?  Is it too oversized?

 

Finally, does anyone have any tips for me as I try to figure out the control system for this thing?  I am planning on one zone, so that should make it somewhat easier. 

Anything I missed so far?

 

do you have a domestic demand in this building? if so, the bock with a side mounted heat exchanger is fine.

design for the higher scenario here. it won't hurt you much.

you do not need 1.2 GPM per loop. we design right down to 0.2 GPM per loop based on load requirements.

it's pretty easy to maintain 45 degrees. 75 degree water is not surprising.

best of luck!

It needs to be heated all the time?

No, there is no domestic demand for this.

Yes, I want to keep this at a constant 45-50 all winter long.  I plan on using glycol just in case, but the plan is to keep it heated.

I know that 45 isn't a big number to warm to, but when the design temp is -30, that is the same as heating to 70 when the design temp is -5.  That is why I was surprised at a 75 degree source temp.

I can't find an aquastat that will let me control the Bock at 75 deg, so if I keep the Bock at 120 deg with the aquastat, my pump is going to cycle more as it heats the slab with 120 degree water and then shuts off while the slab cools back to the pump-on temp.  But it probably isn't a big deal if the temp fluctuates some since it isn't living space...I am OK with the temp varying some as the pump cycles.

Does anyone know of a lower aquastat I can use?

Does anyone know if I can control the water heater more like a boiler where it comes on with the thermostat heat call rather than having 32 gallons of water heated to 120 degrees constantly?

You must keep a minimum temperature in any storage type water heater, so a mixing valve with outdoor reset is important. It will protect the tank, lower the heat load and fuel bill while keeping the building more comfortable.

Over-sizing the heat source is less an issue as water storage increases. Glycol is expensive and counter productive. I recommend a freeze alarm or a burst protection mix if you must.

3/4 barrier PEX in one zone systems fewer loops, smaller manifolds and pumps.

Huh, I'm surprised to see that recommendation from you MA. I would never recommend 3/4" pex in slabs unless the slabs were truly huge, just to add a dissenting opinion to that. I don't see any benefit. the tubing costs more than loops do and there is no need to upsize pumps either if a proper distribution design is done... which I know you know how to do, so why would you go this way? In a shop you can easily just raise the delta-T you design for, lower flow rates, and call it a day.

Thanks for the input guys. A couple things:

Fist, I understand glycol has drawbacks in efficiencies and cost (I'm going to need 20 gal for a 50/50 mix), but a freeze alarm isn't going to do me any good if either a) I am gone for two weeks over Christmas or b) we lose power for 4 days (hence the woodstove in the house). However, I know nothing of burst protection. MA, can you point me in the right direction on burst protection?

Second, if I do decide to stay with glycol, I have more issues. I have no way to temper the 120 water down to 75. I can't add cold water to the system since it is glycol, and there is no way to get a large enough delta T to use the return water for tempering without tiny flow rates and large unevenness in the heat. For flow rates that low I might just as well turn the circulator off I think. Right now it seems like my options are to switch to some kind of a condensing or tankless heat source, accept fluctuations in slab temp and high standby losses, or just heat to a warmer temp.

How would I use outdoor reset, or how would that help me in my situation?

you don't add cold water. you blend with the return water from the slab. in a water heater w/heat exchanger, it's easiest just to add a variable speed controller to the water heater side pump.

yeah, I thought about that after I posted...but my return water is still hotter than the 75 degree design source temp. So if my return water is 90 degrees (30 degree delta T), how do I get back to my design temp of 75?

To illustrate the point, If you have a tub, 100 gallons of 75 degree water, and you dump a cup of 100 degree water in it... the tub is not 100 degrees.

the circulation through the tubing is the tub. the hot water from your source is the cup. the question is, how many cups and how fast do you dump them in the tub? If you dump more hot water into the loop, you raise its temperature, if you dump less, you raise it less.

NRTRob-

Thanks for your patience...I'm sometimes a little slow on the uptake. I was assuming the return water was circulated past a heat exhanger, which wouldn't work since the heat exhanger will always have 120 degree water from the source side and will always be heating the supply water too hot.

But if a 3-way mixing valve is used rather than a heat exchanger, the water in the floor would circulate independently of the 120 degree source water and would eventually come down to 75 degrees. At that point the 3-way mixing valve would only allow enough 120 deg water in (a cup per tub as you so elegantly put it! ) to keep the supply at 75 deg.

Am I on the right track now?

almost there. you are mistaken to assume the heat exchanger is always a fixed temp. with a variable speed pump on the water heater side, the heat exchanger is your tub, and and you are dumping your cups of hot water into it. that is, you are injecting a little or a lot of heat into the heat exchanger to modulate the radiant side water temp. this requires a variable speed injection controller, but eliminates the need for a 3-way valve on the other side of the HE.

OK, so even though the water in the tank is 120, the water at the heat exchanger wouldn't be that hot unless it is circulated. Makes sense. Is the variable speed circulator on the heat source side controlled automatically by a temp probe in the supply side then? Will the 3-way mixing valve idea work anyway? Seems a bit less complicated. Thanks again.

the variable speed injection controller is no more complicated than a 3-way mixing valve controller. both require the same sensors and programming. VSI (Variable Speed Injection) is simper in this case because it requires no additional components or piping. In your particular case you could skip both if you have a good thermostat, but you'd have to put in a "dumb" 3-way mixing valve which is half the cost of a VSI controller and more piping. for the relatively few extra dollars, the VSI is a nice trade up. If you want really simple (not much cheaper, but simpler) just put in a dumb 3-way set to 100 degrees and use a good thermostat like a tekmar 508 or equal. with a high mass shop slab you'd be unlikely to be able to tell the difference... maybe a degree or two of extra "swing" but that would be about it.

and.. the water coming out of the tank, and into the heat exchanger, is the same temp no matter how fast you circulate. The temperature going BACK is what changes, and it's the AVERAGE temperature across the heat exchanger that determines heat transfer to your radiant fluid.

Freeze alarms may call your neighbor, brother, heating contractor and you until the alarm is satisfied.

A 35% mix will greatly improve heat transfer and viscosity issues while protecting from freezing to zero, flow to -15 and burst to -60°F.

Professional help is worth it.

I am a designer AND a contractors. All construction related contractors are in the labor business. Longer loops, fewer manifold connections less labor. Add to this smaller pumps and folks who want anti-freeze and any single zone slab over 1000 sq.ft. will probably get 3/4" barrier PEX.

But I get your point.

I've done plenty of slab installs. I just don't get it; adding loops to a slab manifold isn't much labor. Are you sweating together manifolds on site?

Hello Gentlemen,
Well my 2 cents,
I am afraid I have to agree with Rob on the pipe issue here, we find 3/4 pex to be unruly in most applications under about 3,500-4,000 feet. Your labor and effort to lay the pipe will go up, cost is a wash.
A 2,300 sq ft space is easy with 1/2 pex. I would run 6 loops 330, 1 loop 300 and call it a day, 12 inch pattern staple to foam insulation. Were I to want larger pipe I would consider 5/8 but not likely.
3/4= more work on a small slab.
Might look at the TACO I series valve for the mixing / out door reset concept,
Here is a link to there document, http://www.blueridgecompany.com/documents/iSERIES_100-19.pdf it has a schematic that would work in your application with a water heater.
Dan

It is not a matter of output as I often design and install residential snow melting systems using 1/2" PEX. As for unruly, it depends on the pipe and the ambient temperature.

Single zone, balanced loops may get cheap copper manifolds, but where flow meters and actuators are involved, labor and material can quickly eat up the "savings" of smaller tube.

Now I would naturally promote 1/2" PEX for everything (and make it work) if I were buying and selling 1/2" PEX by the truckload (but that was more than a decade ago.

As you both cater to the DIY crowd, labor is not an issue, but pumps may be larger depending on the load.

I don't decide until the radiant floor heat load specific analsis is done.

I would never upsize a pump to use a smaller pipe. and I've never seen a slab that needed 3/4" to make that a reality. I'd chalk this up to designer's preference, I wouldn't say you were wrong to do it, I guess I'm just not seeing any benefit.