DIY Hydronic Floor Heating

[Mobile Master Tech]

No-the incoming cold water goes through the loops before getting to the water heater. They aren't heated unless a circulator pumps hot water through them. The only disadvantage to an open system is that zones calling for heat only see warm water instead of hot water during a DHW draw. Since that doesn't happen for very long, it doesn't matter and the benefits outweigh this drawback.

Here's a site with a schematic and a good explanation of open systems:

The Open System | | DIY Radiant Floor Heating | Radiant Floor Company

I have a single water heater, an Eternal GU145S. Once I finish the buffer tank/GSHP/etc, it will be replaced by a 100ft 1" corrugated stainless coil in the buffer tank. Everything now and in the future other than that coil will be 3/4", only dropping to 1/2" for individual plumbing fixtures and the hydronic loops. My biggest zone has 4 300ft 1/2" loops, fed by a 3/4" pipe and manifold.

I originally thought I would need a smaller water heater to bump up the temp as I described in my "don't waste that heat" thread, but now it looks like my coil will be able to get the water within 5F of the tank temp, so I probably won't need it. If they were only 1/2" ports, I would parallel two.

[Mikesolar]

Originally Posted by Mobile Master Tech

I've been away a while! Here are some updates and thoughts after having my system a few years and reading recent posts:

I see a great deal of back and forth over pipe diameter, insulation/radiant barrier effectiveness, Reynolds numbers, pipe routing, required supply temp, reset to compensate for outdoor temp, etc. Don't overthink it. To keep it simple, I have learned through experience (some of this is reiterated or I have previously posted):

1: 1/2" pex is used because it is relatively cheap, stupendously durable, easy to work with and a cost effective way to get hot water to a floor with no more than 12" spacing between pipes for even heating. Go smaller and flow isn't enough, go larger and cost goes way up and installation is an absolute bitch. 1/2" works.

Rule of thumb for "normal" heat load houses using gas. 250ft length, 12" in slab, 8" max in gypcrete. I've never been a fan of staple up. I've seen too many failures and too many complaints with noise/ uneven heating. Pipe is cheap. If you use a heatpump for floor heat or want to use solar, get that spacing down as close as possible. You only have to do it once.

2: Pipes in a slab may be best, but staple up with cheap stamped heat transfer plates works fine and staple-up without plates is less than worthless. My not very efficient house envelope is heated fine with 12" on center 1/2" pex and transfer plates with 125F supply (I leave it at 130F minimum for Legionella safety since my system is open loop). I cannot imagine a climate where 8" on center with staple up plates and a minimum of envelope improvements (that should be done anyway) wouldn't be sufficient.

125F water is not what I would call efficient (sorry) but efficient is 90-100F and this means you could use a heat pump if you wanted. 130F for legionella is only necessary once every couple days for a couple hours. It takes time for the bacteria to develop.

3: I have had no failures or problems with the hydronic system. I use regular wall mounted air thermostats telling my controller to turn each zone pump on or off, and one-way bronze valves allowing water flow in only one direction. No reset, no mixing valves, no high wattage noisy pumps, no zone valves, etc. If the house needs more heat, each zone stays on longer and off less. K.I.S.S.

I don't know how long your system has been in place but bronze pumps on open systems will fail. It is a question of when, not if. I have 30 year old pumps on a closed system that are running just fine.

4: I have gotten a few more creaks/pops as the pipes expand when a zone has fully cooled and comes back on-not obtrusive, but I would like it quiet. Instead of silicone caulk as a transfer compound between the transfer plate and the pex, I would use silicone plumbers grease. It won't get runny and saturate anything nearby and would act as a lubricant for the pex in the plates to keep it from gripping/releasing-a very tiny amount does the trick.

There is a much higher possibility of a system needing repair vs a cement/gypcrete system. Again, it is over time and you may have a good working system for 5-10 years but these should be around for 50 years+.

5: I would make each loop a max of 250 feet or so for lower pumping head and route the pipes the simplest way, including pulling a single loop through to make 2 pipes in each joist bay as I described in previous posts. I wouldn't go through extra effort trying to make the outside wall areas hottest or keeping loop lengths nearly the same. Most of the outer floor area is covered with furniture and such, it's closer to the outer wall to lose heat through, plus I REALLY like feeling the warmth of the floor with my bare feet! If the loops are less than 250 feet but more than 150 feet, there isn't much difference in head. You could always crimp in a valve to partially throttle the zippiest loop in a zone if you have a large return temp difference. If that means crossing one pipe with another, just isolate the contact point with a sheath of larger pex or pipe insulation. A cleaner install is better than one with no crossover points or undue effort to evenly heat the perimeter or keep the loops the same.

6: I wouldn't buy Grand Hall's Eternal Water Heater again. I've had numerous different nuisance tripouts cured by powering off then on, the exit mixing valve go bad (under warranty but a hassle to get them to send me a new one), and a potentially catastrophic failure of the water pressure sensor. It's made of plastic and had gotten cooked from heat. I noticed a slow water leak from the heater one day, so I shut the water off and removed the covers. I barely even touched the sensor and it broke completely off, spraying residual water pressure everywhere! I didn't want the same problem again-imagine if it broke while I was on vacation for a month and the water was turned on! It's just a simple on/off pressure switch with a metric thread very close to 1/8" NPT, so I just retapped the hole and installed a metal engine oil pressure switch that fits numerous old GM cars. Done and ok now, but I wouldn't wish these problems on someone else. Great heat exchanger inside, I may repurpose someday....

Can't speak for the Eternal water heater as I don't think they are sold up here so I will shut up on that one, haha.

7: I have learned that copper is bacteriocidic like silver is. Having copper/brass here and there in a plumbing system is a good thing, but it isn't necessary to be all copper.

There is some benefit to some copper in the system and it's hard to get rid of anyway.

8: Having an AC system with an intermittent fan function and high efficiency filters is great for air filtration and moving heat around from one area to another. Air quality isn't a problem.

9: Realtors tell me what I have already confirmed: a warm floor is one of the best features a house can have! Aaaannndd, I get to post one of the benefits I can't believe no one else has posted on ecorenovator: Kotatsus are easy! Here's a link: https://en.wikipedia.org/wiki/Kotatsu

A Kotatsu usually has a heat source under the table and a comforter style perimeter draped down that you put your lower body under. The warm floor is the heat source! I've draped a blanket over a coffee table or draped it over the edge of the couch to create a cocoon of extra heat coming from the floor. 100% of the homo sapiens, canines and felines I surveyed agree that it's awesome!



Lastly, it makes sense to design a system to work at DHW temps so you can combine systems and heat sources and not have to work so hard at making lower exergy sourcewater work. I've updated my plans to fire my energy companies and ask all of you for some Phase Change Material help at post 15 of THIS THREAD

If you can find a good source for stable phase change material, please let me know, I want it for some solar storage.

Bumping up the temp of the heat source for a GSHP improves efficiency just as much as lowering its discharge temp. The actual temps don't matter much, what matters is the amount of lift. Taking heat from a 70F source to make 140F water is essentially the same work as taking heat from a 35F source to make 105F water.

Ummmm, no. For the same amount of lift, your COP will be much better with lower temps. Also, with higher temps you run up against the limits of the refrigerant.

I am also insulating and stuccoing the exposed exterior parts of my concrete basement walls so they are part of the thermal mass and not losing such an ungodly amount of heat through them. 1/2" R3.8 polyiso board is $10 at Lowes. Using dollops of adhesive and some pushpins to keep the radiant barrier side of the board 1/4" of an inch from the concrete is almost as good as the recommended 1" airgap for radiant barriers. Sealing the board at the top of the wall will prevent the barrier from getting dirty and also prevents an hidden path for termites to make it to the wood a few feet above.

Pay close attention to where your dew point is in that wall. I am all for lots of insulation on the outside of thermal mass but you won't get much benefit until you hit at least 2". You must have at least 2/3 of your insulation value on the outside of the mass or there will be lots of condensation.

I don't want to sound too critical but I have to put my 2 cents in. I'm doing a few passive houses at the moment with heat pumps and it gets quite tough to make a good heating system with a heat load of only 12Mbtu

[Drake]

Originally Posted by Mikesolar

it gets quite tough to make a good heating system with a heat load of only 12Mbtu

That is exactly what I am working on.

[Mobile Master Tech]

Mike, I beg to differ on a few.

I am not knocking slab or advocating "staple up" per se, but rather systems with heat plates holding up the pipes and using a bead of heat transfer compound. I originally just had staple up. With 150F supply water it couldn't cope with a 40F night! I removed the staples and added the stamped u-groove heat plates, now it can keep up with at least 15F nights using 130F water.

Pipe is cheap, and so are plates; completely remaking a house to add a slab is not. New flooring, new trim, adjusting door heights, resetting cabinets, less headroom, possible structural mods, etc. If I were designing from scratch I would probably plan for in-slab on all floors. Not as easy on an existing house.

There may be complaints with staple up, but I and my pets love having slightly warmer areas of the floor-I find that a positive, not a negative. There is enough temp difference to notice but it is not obtrusive, especially with a radiant barrier hanging freely just below the pipes to create a "cocoon" of warm air under the entire surface and conditioned space below. The underside of the barrier won't ever get dirty, so the barrier remains effective. The noise isn't a big problem and I believe a change of thermal compound wil address most of it. Both these problems pale in comparison with my larger complaint with retrofit slab systems: cost and effort!

Failures and problems are far more likely to be caused by improper design or install rather than system type. Failing to radius the end of the plate channel so the edge chafes the pex is a prime example. Do you have some examples to back up your claim that a properly designed underfloor system is inherently less reliable than slab?

Recall that I said bronze valves, not pumps. These one way check valves are just a loosely hinged slab of bronze falling by gravity against a seat. They don't have to seal perfectly, just stop the bulk of the flow. Those will probably never fail. I'll report again on my system more than 10 years down the road instead of just 4 so far.

As for pumps, are you familiar with the Laing/Bell & Gossett e3 design? No seals, no bearings, and the only moving part is a freefloating rotor, positioned and driven by magnetic force, spinning on a ceramic "pinhead". Bronze isn't why other style pumps fail, it's their moving/sealing parts, just like any conventional pump. Could you please back up your statement with examples of the Laing style pumps failing? Bronze or stainless is what allows pumps to be used in potable systems, and it has the benefit of not corroding.

For fun, check out hvac-hacks.com, especially the screw-ups and burnt n'crispy categories. Plenty of "don't do it this way" examples.

125F water IS efficient and so is 130-145F, if your source heat temp is high enough. Being able to combine systems for less complexity or points of failure is also efficient and elegant, which is why I am combining 6.1kw of PV, GSHP with a 70F+ source temp due to the annualized heat storage, hydronic with DHW, etc. I referenced my other thread to explain where I was going with this-completely fire my energy companies and eliminate most of my transportation energy with a Chevrolet Volt, under $15k for a 2013 model these days.

Look at any heatpump COP chart that shows multiple inlet temps and you will see that the COP for any combination of higher inlet/discharge temps is 95% or better of the COP for lower temp combinations with the same spread, assuming you don't try to run near the supercritical temp of your refrigerant. Again, proper system design. R410a with a supercritical temp of 163F would be difficult in a system designed to be capable of 145F output. Hydrocarbons or their blends like the R600 series, R134a, even R22 would all give better "fudge factor".

The insulation recommendations from IECC and others for climate zone 4 (which I am on the edge of) for abovegrade mass walls is R-5 when more than half the insulation is on the exterior. The total R value of the assembly is 6.7-7.6: wall and drywall (R0.4-R0.7), air on both sides (R0.6 interior, R0.2 exterior) , radiant barrier with 1/4" gap (R 1.9-R2.5), and R3.6 of polyiso. The exterior is at least R5.5 (barrier, airgap, polyiso). I could go thicker, but I don't want it to stick out further than the existing siding above it. I don't plan on having insulation on the interior so moisture won't be a problem, plus the belowgrade areas will be part of my passive annualized heating system described in my other thread.

R10 is recommended if more than half is on the interior, so you are correct if the insulation was on the inside.

There is more than one good way to do something. I'm trying to show that underfloor with plates (I'm going to quit calling it staple up-what a hotbutton word!) can be successful and much easier than slab when done properly, and there are ways to combine systems for simplicity and a sum greater than parts.

[Mobile Master Tech]

Yes it is difficult for a heatpump system to be designed for 12million btu/month, hence the storage buffer tank with PCM described in my other thread. It can run during the day while the PV panels are warm, PV output is easily available to run it, and source water is likely to be 80F+, only running at other times during high demand periods.

[jeff5may]

In Atlanta, I don't imagine dew point would be much of a problem in heating mode. Winter and Spring temps are all over the place, but rarely stay below freezing for more than 48 hours at a time. It gets into the cold range often, freezing range occasionally, and frigid range hardly ever. Freezing rain and sleet are much more common during winter than snowfall.

For example, on any given Sunday in January, the daily high may reach 65 degF. A cold front rolls in, and rain turns to sleet on Monday. The temp bottoms out at 28 when the sleet stops. Tuesday morning, school is called off. The black ice on the roads Tuesday morning causes wrecks everywhere, due to driver inexperience. By noon Tuesday, the ice on the roads has melted. Wednesday's high temperature reaches 52 degF.

Due to the lack of deep temperature troughs of frigid weather in the region, most housing infiltration and insulation efforts would be classified as "severely lacking" by Canadian standards. More effort is put into dehumidification, as RH regularly rises into the 90's between February and October, especially during after-work hours through the night. Most of the year, the weather conditions could be described as "cold and clammy, muggy, or tropical" in nature. Not surprisingly, most modern homebuilders are more concerned with tight vapor barriers than raw insulating power.

[buffalobillpatrick]

I have done 2 systems as follows & love it:

A large (buffer tank / DHW tank) that does both jobs.

Mine is 80gal & my Son's is 115gal

There is a flat plate heat exchanger between the boiler and the tank, pumps on both sides.

Heating system is on boiler side of heat exchanger.

System can heat during a boiler cycle or after, by pulling heat back out of the tank via the flat plate heat exchanger (tank acts like a buffer here)

This is KEY: there is no short cycing of boiler. Zones calling for heat can sip out of the tank.

The boiler is controlled by a Tekmar 256, Universal sensor on tank + outdoor reset sensor, continuous call for heat is jumpered on with a 25* delta programmed.

This works great, long efficient boiler cycles heating the tank with ODR control of water temp. The boiler is not controlled by system, only the Tekmar 256 controls the boiler.

[Mobile Master Tech]

Jeff, add that upon any news of frozen water in the forecast, the majority of the population here rushes to the stores picking the aisles clean, stocking a month's worth of perishables that will go bad in a week! Other than that, you nailed it!

BBP, you are absolutely right on preventing short cycling. I'm sure yours works well. The simplest way to do it though is to use an Energy Star conventional tank water heater (most are certified for hydronic/DHW combinations) and an open system (safe and approved in most locales). A 40Kbtu burner on a 40 gal 12 year heater for $600 is plenty because of the "buffer factor" of the tank AND floor/house unless your house and household is huge. I'm assuming anyone would already have a decent building envelope and water efficient fixtures such as Water Sense.

The water heater will have less work to do, since it will never see inlet water temps below your interior temp. The incoming water replacing the DHW being used is drawn through the floor first. More simplicity, less cost. You would have to have a LOT of DHW use combined with a peak heating time before you would overwhelm the burner and heat already in the plumbing and house. Energy Star's have a 67% or greater energy factor, which is probably about an 80% efficient burner.

Tankless, tankless hybrid and boiler systems have trouble if they can't throttle down to the lowest demand level (say, a faucet trickling warm water, or a single 0.5gpm heating zone). Conventional tankless has an awful lot of head loss in it's coil.

My Eternal water heater is a condensing "hybrid" with low head loss and a few gallons onboard storage but it still can't throttle lower than 26kbtu. It's 96% efficient unless short cycling or when warm loop water is returning. At 110-115F return temps, it does still condense some, but I imagine the efficiency is more like 85% then. The newer, smaller Eternal can throttle down to 14kbtu, but I'm not sure even that is low enough, and I would replace the plastic pressure sensor with a metal one before I would even consider an Eternal again.

A condensing boiler won't condense much as you have it installed since it is circulating already warm water, but condensing and noncondensing boilers will still work fine. It will operate most of the time with an efficiency not much better than a conventional Energy Star tank style though.

Of course, I'm advocating and switching to a solar/geothermal hybrid system to get heat for "free".

[buffalobillpatrick]

Originally Posted by Mobile Master Tech

Energy Star's have a 67% or greater energy factor, which is probably about an 80% efficient burner.

Gas hot water heaters loose convective heat up the stack constantly and are not nearly as efficient as advertised.


Heating requirements at high altitude in Colorado are MUCH different than Southern states.

My Son's house is in Crested Butte Colorado at 8,888' altitude, climate zone 6, as is much of Alaska (read too damn cold).

IBC 2009 states that the heating system has to be able to keep every room at 70* at design temperature (-20*F)

It was designed & specified by a great architect who used REScheck to calculate the UA = 1,194

The Design Delta between 70* -> -20* = 90*

The output of the boiler must be at least: 1,194 x 90 = 107,460 btus/hr

I decided on a Burnham ESC6NI-TH (high altitude version) Sealed combustion via side wall.
Efficiency=85.2%, DOE=130,000, IBR=113,000

A proven design cast iron boiler like this costs about 1/2 of a good ModCon boiler.
But it is way more reliable, will probably last 30-40 years, & requires MUCH less maintenance.

This is why plumbing/heating shops love ModCons, they weigh a lot less & there is way more profit $$

A large boiler needs a large buffer tank to prevent short cycling, especially if only 1 zone is calling for heat & it's 50* outside.

[buffalobillpatrick]

MMT, your post #228 in this thread has lots of good info:

Originally Posted by Mobile Master Tech

Concerning Legionella, that is why I am designing for minimum 130F supply temp, preferably 135F. I plan on having a small electric instant water heater at the exit of my 30 meter 1" dia corrugated stainless steel heat exchanger pipe located in the top of my 2000 gal storage tank. The heater will make up the last few degrees just in case the geo/desuperheater/solar can't keep up, and I will put some thin silver rods in the storage tank since silver is bacteriostatic (bacteriocidic if the contact time is long enough). The silver is easy to wipe down should it ever become corroded to maintain efficacy.

130F is hot enough to kill Legionella bacteria. Any hotter, they die faster. That is the reason tank style water heaters can be a serious Legionella risk. If the top of the tank is 130-140F, the bottom of the tank will definitely be below 122F due to stratification-right in their favorite growing temperature sweet spot. The bacteria also need a still place to breed-sediments on the bottom of the tank provide a perfect breeding sponge maintaining that perfect temperature and sheltering the bacteria from being flushed out with the next water draw. Even a gas heater warmed from the bottom by the burner will often retain areas of lower temperature that won't kill off the colony. A tankless system that does not allow any water to stagnate at a warm temp never gives the bugs a foothold. The water is either starting out hot enough or turbulent enough to kill/prevent growth when the system is in use, and quickly cools below their sweet spot once shut off and no longer turbulent. That's why a properly designed "year round flow through" open loop system is important-water can't stick around long enough to get stagnant. How about the water sitting in the DHW plumbing leg that serves the back part of the house you seldom use? An open loop radiant system prevents the water from ever sitting around like that. EVERY time ANY hot water is used ANYWHERE in the house, fresh water is drawn through the radiant loops to replace the water that was used.

It's interesting that Canada requires the 140F water heater setting, but apparently prevents open loop systems that by nature are less risky than unused portions of the DHW system. Open loop systems are ok by our national building codes and most local codes.

I read some reports from the CDC that said improperly maintained chiller equipment is the largest source of legionella infections, followed by tank style water heaters. One of the next largest sources? Windshield washer fluid in vehicles! Since only northern climate washer fluid (the stuff that says "good to -20F") has enough methanol in it to kill the bugs, most people are driving with low or no methanol washer fluid. Add warm summer temperatures, a cozy engine compartment that retains heat, and a way to atomize the bacteria laden fluid (the washer nozzles) into a respirable mist, usually in front of the fresh air inlets at the base of the windshield, it's no surprise. Truckers have a 4x risk of legionella infection compared to the general public since they put on so many miles.

I always recommend putting a capful of chlorine bleach in your reservoir 1-2x per year to prevent this. I see lots of cars with washer nozzles and pumps plugged up with growth in the reservoir. I wonder how many have legionella? I always throw some bleach in there and give them some contact time before removing them to clean them out.

Craig
The MMT