Hi all, first post. I have been doing a lot of research (on this site and others), talking with manufacture's, etc. about my setup. Nothing too complicated, I have a new 36x30 detached garage. Walls are 2x4, 22' of the 36' is attic space with 2x8 joists, the remaining 14' is cathedral 2x12 for a lift someday. I have calculated my heat load (albeit 4-5 months ago), and if I recall it is in the range of 30-35K BTU/hr. This might actually be a little on the high side (but sticking with it as this is a garage with overhead doors, etc.), but I did put into the software as much accurate info as I could for insulation, windows, doors, ceiling, slab insulation, etc. The 6" slab floor is enveloped with 2" XPS on bottom and sides. I have 4 loops (single zone) of 1/2" oxygen barrier PEX around 240' per loop, terminated at a stainless mainfold with 3/4" fittings to/from heatsource. Everything came from BlueRidgeCompany. I know guys are getting away with tankless hot water heaters. I actually have a very good friend that has his up and running, and it seems to be working great. I understand that boilers are purpose-built for this application, but at 4x the cost I am seriously considering going this route as well. Takagi's are rated for radiant (albeit a lowered warranty period). I am looking at the Jr2 & K4. K4 has more max BTU (unneeded/overkill) but it also can modulate to a lower BTU (more interesting to me). I understand the head loss associated with tankless, and that a beefier circ pump is required. These particular units have a min head loss of around 10ft, which is the *max* that the Taco 007 can handle. So, immediately that pump is out. More to the point, what I am struggling with is that these heaters modulate both the water flow as well as the burner. I think boilers only modulate the burner (assuming you have a boiler that can modulate to begin with). I don't know this for sure about boiler operation, but it makes sense. I want to be able to keep a fairly consistent flow rate, but worry that this won't be possible with the tankless. The other is on the high end, the BTU output is way, way overkill for my space. I understand this, these things were designed to bring 50 degree water up to 120 degrees instantly. That requires a lot of heat. For a radiant floor with a delta T of say 20-30 degrees, is it possible for me to dial a tankless down so it is not short cycling and causing large temp swings? Barring the head loss, some basic questions -- can the flow modulation be disabled so that it can be consistent (i.e more like a boiler)? Can the water temp output set point be set low enough so that the unit runs at a lower BTU longer, instead of high BTU shorter? I am starting to look at a primary/secondary design, as it appears that this could overcome some of my concerns. Thoughts? Question on tankless heater efficiency/longevity. Is it better to run them under full BTU output for a longer period of time? In other words, are they designed to have maximum life running at 50% rated BTU, or? I didn't mention tank style hot water heaters, but I have sorta thrown them out due to the space requirements (although I do have an attic...). There would be no head loss, and the BTU is more in-line with what I really need. But, longevity is a concern. Thank you!

One normally selects the pump(s) and heat source AFTER properly accomplishing a good heat loss analysis and hydronic radiant floor heating design and BEFORE procuring/installing the hydronic radiant floor heating system components. It is important that all the components are compatible and function well with each other. It sounds like you have already accomplished the heat loss analysis, accomplished the hydronic radiant floor heating design, and procured/installed many of the hydronic radiant floor heating system components. It is usually best to continue working directly with your designer/installer/supplier(s) until your hydronic radiant floor heating system is fully operational and operating well.

Thank you for the reply. I have not purchased/procured any of the mechanicals. That is why I am researching this phase. Heat loss done, PEX/design done, slab design done, concrete is in. The only thing "above slab" that I have are the manifolds so that I could terminate and pressure test the PEX and make sure no damage occurred during the pour.

A 30-35,000 BTU/hr heat load for a 1080' building is ~32 BTU/hr per square foot of conditioned space. Unless this building is basically uninsulated AND leaks air like screen door it's unlikely that the true heat load is anywhere near that high with the windows & doors closed. I'm dubious, at best. (Incredulous is more like it.)

This needs at least a napkin-math sanity check.

Where are you located (for reasonable outdoor design temp estimates) and what is your inside design temperature?

How many square feet of window, with what U-factor?

What is the wall construction & insulation level?

What is the ceiling or roof construction & insulation level?

Propane, or natural gas (and at what price)?

What are your local electric rates?

Dana, after thinking about this a few times today, yes, I was overshooting the heat load (by a lot). However, with overhead door losses/leaking I want to make sure my "fudge" is sufficient to cover real world. I did this back in August, and have seen a lot of number since then. Anyway, see attached photo with the screen shot of the PEX layout (not dead on, but very, very close to how it laid out). The heat load calc is there too. To answer your questions to make sure we are talking apples to apples: - NorthEast Ohio, 45 Mins South of Cleveland - 1 single window is u-factor of 0.470, 3x4' in size - Walls are 2x4 studs, 16OC, R13. I have about 12" of cinder block exposed above slab before the framing starts. So *insulated wall* is 9', with 1' of block under. - 22' of building is attic truss, 10'1" ceiling from floor. Joists are 2x8s, planning on R30. - 14' remaining is cathedral 2x12 stick & carrier ridge beam. R38 is the plan. To the peak in this area it is around 19' - Natural gas, already trenched - Electric, 6.5-7 cents/KW/hr, I would need to look at last bill to be 100% - 1 service door, ufactor of .600. I took the worst of the choices. It is steel/wood core door. - 1-16x8' overhead door & 1-10x8' overhead door. U factor of .059. I bought really nice overhead doors, insulated, thermal break. The garage is not a perfect rectangle, the 14' lift bay (10' door) is offset only 28' deep, to give some curb appeal/architecture (less boxy). When I loaded loopcad with this stuff I fudged the ceiling detail due to it not being the same across the entire building. I am not very familiar with loopcad, but I think everything is close. Let me know what else might be required. Thank you so much for the review! (P.S. I cannot seem to get line feeds/breaks in this forum, sorry not intentionally running everything together!)

Attachment... okay, I need to review how to use this forum. Attachment was image from loopcad. Data is as follow: 14,641 BTU/hr @ 1030 ft/sq. Room temp - 65 degrees. Heating type - radiant. Panel Type - Embedded Slab. Surface Temp - 73F. Water Temp - 95.9F. Supplemental - 14,641 BTU/hr. Spacing - 12".

FWIW, been a while since I messed with loopcad. I opened the project and adjusted a few things a little more accurately. I am now at around 16k BTU/hr heat loss and 98 degree water temp. I just want to make sure I have a pretty good idea of where things are to make sure I don't undersize. But, if I end up with a tankless, undersizing is pretty hard to do.

The heat load in my 2400 sq.ft steel building is below 30k.

Real Manual 'J' people don't add-on. They use proper accepted software and stick to their guns. Under-sizing a boiler (electric boilers excepted) will only increase comfort and efficiency with the worst case dropping room temp during design condtions--coldest week of the year.

A garage door, mine is a true R-19, is the least of your worries as long as it is properly installed and the seals maintained it is likely better than your walls for real R-value and infiltration.

Naturally the PEX is the easy part. The heat source, pump and controls separate the experienced hydronic designer from the Pseudo-Engineer...

The line breaks get lost when using some web-browsers. Try FireFox or Chrome.

Seriously, with 7 cent electricity a 2- ton Chilltrix air-source inverter drive heat pump/chiller (with perhaps a tiny electric boiler as backup, if needed) is probably much lower cost heat than condensing natural gas. With 95-100F output water temps it's good for about 14,000 BTU/hr down to -4F, and can still deliver 12,000 BTU/hr @ -13F.

http://www.chiltrix.com/chiller-technology.html

What was the outside design temp on which that load number was based? Your 99% outside design temp is about +6F, your 99.6% outside design temp is about +3F. See:

https://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf

https://www.ruppams.com/CatalogContent/Fans/Sup_MPU/doc/Winter_Summer_Design_Temps_US.pdf

And what are you paying for natural gas?

If the cinder block walls (CMU) are not insulated above slab, you have a big long skinny R1.5-R2 hole in your wall that's worth insulating. If that wasn't included in your heat load calc, it should have been. A continous layer of 2" thick polyiso on the interior side would make it perform about as well as your 2x4/R13 wall. Alternatively, 2.5-3" of EPS on the exterior side down to about 2' below grade (if you don't already have the below-grade portion insulated as slab-edge insuation) would work too. It would need Z-flashing to keep rain moisture from getting between the foam an CMU, and you can finish the EPS with a cementious EIFS such as QuiKret Foam Coat. If taking the inteirior approach the foam needs a fire barrier such as half-inch gypsum board or OSB. In a shop building I'd use OSB tight to the foam, through-screwed to the CMU with 3.5" TapCons for that function.

Tankless hot water heaters would pretty much suck as a gas fired solution with a sub-15K design load. Sure, you couldn't possibly undersize the thing- it would be RIDICULOUSLY oversized. Your average wintertime load would be well under the minimum modulated output of the thing. A condensing hot water heater would probably work out better- still oversized, even ridiculously oversized, but inherently self buffered, and harder to short-cycle (though a creative hacker can short-cycle most anything. :-) )

Also note, a 5" slab is a pretty unwieldy amount of thermal mass to use as a radiator. Even 3" slabs sometimes have under/overshoot problems, but if you're not too picky about the shop temps locking dead onto the thermostat setting you'll be OK.

A 1-ton Fujitsu mini-split would also be capable of covering your heat load at comparable efficiency to the Chilltrix, with less design risk, and it can air-condition the place too.

http://www.fujitsugeneral.com/PDF_06/Submittals/12RLS3HSubmittal.pdf

At the 2030 Clean Power Plan goals for OH, a heat pump solution with a COP efficiency of 3 or better would have lower carbon emissions than a condensing gas solution too.

http://epa.ohio.gov/Portals/27/111d/CPP%20Fact%20Sheet%209-24-2015%20final.pdf

Hi, thanks to both of you!

The browser thing, I use Chrome 99% of the time, I am doing this post with IE in compatibility mode. This forum is finicky it seems.

...anyway, the design temp is set to 6 degrees F. I did all of this with LoopCad. I had not loaded the project for 6 months until last night to review this stuff. Badger, thanks on the overhead door info. My doors are a true (at least as far as marketing / specification literature) 16.9 R-value. I can list the door model if it matters. They are CHI.

Dana, I have seriously considered a mini split, but not specifically what you have detailed. I'll need to look at that. I don't have my gas rates in front of me, but in general this area of the country has moderate/low utility rates.

Badger, I am not trying to manipulate the numbers in the software to push an agenda one way or another. I am not an expert with the software, but my 14K-16K BTU figures don't appear to be too far off.

I realize I am posting and requesting info on a green building forum. So, you guys are going to give what is likely the most "green" of an answer in return. Again, this is a garage, not looking to live out there (yet!), so optimum comfort/efficiency is not the goal. I would say my goal is reliability of whatever system I choose over anything else.

My old garage, very small at 18x24, was heated with a small electric forced air heater. R13 in walls, R19 in ceiling. It was simple and effective. I am a KISS kind of guy. The drawback was cold floors and doing restorations got to be uncomfortable (and I assume as I get older that won't get any better!), which is why I at least have this new garage setup with PEX/radiant, to give me that option.

Thanks again.

Looking at EIA retail gas pricing data history in Ohia the wintertime residential gas rates run about $7.75 per 1000 cubic feet (=~1.02 MMBTU, so it's ~ $7.60/MMBTU) https://www.eia.gov/dnav/ng/hist/n3010oh3m.htm

Burned in an 95% burner, and ignoring pumping power use, that's $7.60/0.95=$8.00 /MMBTU for the heat delivered into the heating system, plus the cost of the pumping & contol power.

A single kwh is good for 3412 BTU used in a resistance heater at a COP of 1, and at a COP of 3 would deliver 10,236 BTU out of a heat pump. That would take (1,000,000/10.236=) 97.7 kwh per MMBTU. At 7 cents/kwh that's $6.84/MMBTU.

That's a bit cheaper heat than heating with condensing gas.

At a COP of only 2.5 (which means you didn't size the heat pump right) you'd be looking at about $8.55/MMBTU, which is probably comparable to heating with gas if you account for the power used in the gas system.

At a COP of 2 (meaning you TOTALLY screwed up the heat pump sizing) it would be about $10/MMBTU, but still only 25% more expensive than heating with condensing gas, and comparable to heating with a non-condensing gas-burner.

Heating the slab to say, 65F with an electric boiler and floor thermostat for comfort, but using a mini-split to heat the space to 68F is probably cheaper and more comfortable than using an inappropriate non-condensing tankless to heat the room to 65F.

I will go along with most of it, but ROI, not there my friend.

First, you won't have radiant floors, read; no real winter time comfort.

For a single zone, I would go for a dedicated condensing boiler in the 50mbtu input range with outdoor reset any slab will be reigned in nicely. ECR/Dunkirk has I nice looking 50mbuth input SS condensing boiler with built-in pump, outdoor reset and low-water cut out control.

http://www.dunkirk.com/products/boilers-condensing-gas/helix-vx-vertical-laser-tube-wall-hung-stainless-steel-modulating

Dana, I really appreciate you going all in on the calculations. You certainly know this stuff. However, I am concerned that with the up front cost of the more complex system(s), configuration/installation, and inevitiable maintenance costs, I will never realize an overall savings (or will take a very long time). I think this is what Badger is saying regarding ROI.

There are many folks online that are "happy" with a very simple radiant system for a garage space. I won't argue that it is not correct, ideal, most effecient, etc. Nor will I argue that the tankless will have a shortened life. This is the aspect I am trying to understand. Can I buy 4 tankless units for the price of a boiler (or whatever system we might come up with)? Will it last 5 years (less?) or 10 years? There is not much real-world data that I can find on this. There are a lot of "don't do it!" comments from the experts, but then there are at least the same amount of success stories. Dunno.

Badger, any experience with the Dunkirk boiler you linked to? That would make for a nice setup, pump integrated, etc. really clean install.

Thanks again!

DIYers often successfully use hot water heaters and tankless units for small scale or part-time use hydronic radiant heating. However, you are not going to find too many licensed HVAC companies that will publically endorse or recommend using heat sources that are not rated for space heating applications. There are hundreds of boilers that are rated for space heating that will meet your hydronic radiant heating needs:

SupplyHouse Boilers

The proper steps are to first determine what size boiler you need and then select a boiler manufacturer/supplier that you believe will stand behind your purchase in the event that you have installation or warranty issues. Full disclosure, SupplyHouse is one of our affiliates. However, the boiler section of their website amply illustrates the variety of mainstream brands, BTU output, fuel type, efficiency, vent type, features and price range that is readily available. There are many suppliers and you should always shop around for the best deal.

Sailawayrb I understand. Thanks. Considering a tank style hot water heater, is 30 gallons the smallest natural gas fired I am going to find? My research so far is looking that way. Packaging and placement is one of the other concerns. Thanks!

For the record, I heat my own house with a Takagi tankless hot water heater, but my heat load is more than 2x his real heat load, and I have a 48 gallon buffer tank between the Takagi and the zones (of diverse heat emitter type.)

For the DIYer it's safer to use a condensing tank for a slab-radiant system than a cheap standard gas fired tank. With the latter chronic condensing in the center-flue heat exchanger can ruin them fairly quickly.

Most domestic hot water heaters are now approved for space heating and in many cases make perfect sense. I have one heating my own garage and all of my DHW. It is a power-vented water I inherited and requires outside combustion air to be safely operated. My new shop, and most of my field work, will be driven by sealed combustion equipment. This makes sense no matter where you are if your envelope is tight.

A tank-less is fine and may not need a buffer if it is set to low temperature and pumped properly. You may use a Vertex or a Polaris, as I often do in my own design work, but for a single zone in makes little sense. You will still be lacking the fuel saving and comfort creating outdoor reset now standard on all condensing boilers and you will not be able to modulate output.

In the long run a good ModCon, properly sized, installed and maintained is a good investment. Buying a ModCon on-line is rarely advisable since local support can be a key factor for the novice. When I design a heating system for out-of-state clients I always consider the area in which they live. Even the best ModCon will need competent service and if there is a warranty issue most manufacturers of gas-fired equipment void all DIY warranties. Naturally many local heating technicians are reluctant to work on DIY equipment, either from lack of experience with your particular ModCon or fear of complications do to operator error.

The VLX is a nice machine and one of the few available with a low 50mbtuh input.

40 gallon WH is most common but if you use a conventional water heater you need to know how to pipe it and provide proper venting and combustion air. I have never seen a conventional tank water heater properly installed for radiant floor heating, though it can be done.

Posted By BadgerBoilerMN on 13 Jan 2016 08:27 AM
Most domestic hot water heaters are now approved for space heating and in many cases make perfect sense. I have one heating my own garage and all of my DHW. It is a power-vented water I inherited and requires outside combustion air to be safely operated. My new shop, and most of my field work, will be driven by sealed combustion equipment. This makes sense no matter where you are if your envelope is tight.

A tank-less is fine and may not need a buffer if it is set to low temperature and pumped properly. You may use a Vertex or a Polaris, as I often do in my own design work, but for a single zone in makes little sense. You will still be lacking the fuel saving and comfort creating outdoor reset now standard on all condensing boilers and you will not be able to modulate output.

In the long run a good ModCon, properly sized, installed and maintained is a good investment. Buying a ModCon on-line is rarely advisable since local support can be a key factor for the novice. When I design a heating system for out-of-state clients I always consider the area in which they live. Even the best ModCon will need competent service and if there is a warranty issue most manufacturers of gas-fired equipment void all DIY warranties. Naturally many local heating technicians are reluctant to work on DIY equipment, either from lack of experience with your particular ModCon or fear of complications do to operator error.

The VLX is a nice machine and one of the few available with a low 50mbtuh input.

40 gallon WH is most common but if you use a conventional water heater you need to know how to pipe it and provide proper venting and combustion air. I have never seen a conventional tank water heater properly installed for radiant floor heating, though it can be done.

"A tank-less is fine and may not need a buffer if it is set to low temperature and pumped properly."

Almost all DIY installations I've reviewed on tankless driven slab-radiant systems were severely over-pumped and over-fired, some with a primary/secondary, others pumped direct. Without reviewing the tubing layout & calculating the pumping head, I suspect this COULD be pumped direct with a programmable Bumblebee or Viridian at 1-1.5gpm (to guarantee that the tankless fires up), and adjust the firing rate by the temperature setting. But there's a lot of "ifs".

I took the centralized buffered approach to be able to micro-zone the place due to dramatically differing heat loss characteristics in different parts of the house. The smallest zone has a design load of ~3300 BTU/hr, well below the min-fire output of any mod-con, while the biggest is ~13,100 BTU/hr, but it all runs at a single ~125F average water temp.

HTP's UFT-80 is a decent stainless fire-tube wall-hung boiler with a 10:1 turn down ratio (~7600 BTU/hr out at min-fire), that would be able modulate at least some despite the low heat load. They advertise it as "No Primary/ Secondary Piping Needed", but I've never pulled the installation manual on it.

http://www.htproducts.com/literature/UFT-brochure.pdf

Badger, when you say "will not be able to modulate", the K4 (as an example) has an output between 11K BTU and 190K BTU. Are you stating that I won't have direct control over the burner modulation?

The plan was to keep the water output temp as low as possible, to ensure that the unit does not attempt to restrict GPM/flow, and also to keep the burner modulated down to a lower BTU. Would a condensing unit be worth the 500 or so extra dollars? (if looking at the Takagi product line like I have been).

FWIW, I have attached the PEX layout, again BlueRidge helped me with this. I had almost exactly the same PEX layout myself in LoopCad as well.

A tankless won't modulate up/down with outdoor temp the way a mod-con will. The firing rate in a tankless is determined by the return water temp, the output temp, and the flow rates only. The output temp you can program, the flow rate you COULD program with a smart-pump, and the return water temp won't change a whole lot. It'll simply duty-cycle on/off at whatever you've tweaked the parameters to be, with only a very small change in modulation levels based on the duty cycle. A higher duty cycle will result in a modestly higher return water temp, lowring the firing rate a bit.

That's about all the firing rate control you can get with a tankless unless you get fancy with outdoor reset mixers and pumps, etc., not exactly a K.I.S.S. type of installation.

But a mod-con with outdoor reset control can deliver 80F water when it's not very cold the loads are low, and automatically higher temp water when it's colder outside. Still,your design load at the very low modulation range of most mod-cons, and average load below the minimum firing range of most mod cons. I'm not convinced it's going to be worth it for a shop building, especially at the small design load.