Need advice: Insufficient Output From Hydronic Floor Heat

TimH

We've got a 2-zone hydronic in-slab heating system in our shop, and the smaller zone is not producing enough heat. It was installed in 2008 when the shop was built. Any advice on solving the problem?

The building is 60' x 102' with 16' ceiling. The small zone is used to heat an office and restroom area that's 33' x 11' with 8' ceiling. The original installer had used Taco 007-F5's on both zones. The small zone has only one 1/2" pex line which I'm guessing is around 300' long. Around 3 years ago, the circulator on the smaller zone failed, so we replaced it with a 008-F6. Neither zone has check valves.

The small zone heat output seems to be gradually getting less over the last number of years. Whenever the outside temp drops below about 20 degrees F, the small zone will run round-the-clock without getting to set room temp of 70-72. The larger zone heats the rest of the building effectively to our set temp of 60. The zone switching relay has the option of giving priority to the small zone, which I've tried; but it seems to have little effect on the heat output of the small zone. It just prevents the large zone from heating at all, so it's currently set that either or both zones will run. There is no gurgling sound to indicate air in the system, and we checked for an air trap by the circ. Pressure gauge on the boiler shows 17 PSI. When the large zone comes on, all 4 boiler heating stages will run, with boiler temp staying around 70.

Here are some of the temps after a number of hours of running only the small zone:

Boiler Gauge: 135 F (range 120-155 with only 1 boiler stage cycling on/off)

Vertical copper supply tube from boiler: 114 F

Horizontal tube above circs: 94 F average (varies quite a bit from top to bottom of tube)

Black supply pex: 90 F

Black return pex: 72 F

Vertical return tube to boiler: 73 F

Large zone supply manifold: 66 F

Large zone return manifold: 72 F (Note that it's higher than supply, indicating reverse flow?)

Motor housing on 008-F6 circ: 144 F

I notice 2 potential problems: (1) With only the one 1/2" pex, a lot of head is required to force even a low GPM through. (2) We seem to be getting reverse flow through the large zone when only the small zone is running.

What I'm thinking of trying next is (1) replace the 008-F6 with a 009-F5, and (2) install an IFC valve into the 007-F5 on the large zone. Does this make sense? Is there some other problem that I'm missing, or something that needs checked out first? Any advice would be appreciated.

Should I get an 009-F5 with or without IFC? I couldn't think of a reason to put a check into the small zone. The large zone obviously draws the full heating capacity out of the boiler, and even if a little heat was pushed backwards into the small zone it would be fine as that zone is set to a higher temp and is generally hard to heat.

Ironman

Is that Entran tubing on the small zone?

hot_rod

Is there a lot of furniture or built ins in the office? Anything that covers the floor reduces the output. Floor coverings also.

There needs to be check valves in both pumps.

What size is the boiler to manifold piping?

20 loops at .65 gpm per loop would be 13 gpm for the main zone, at least 1-1/4" copper to supply that. Looks like the boiler S&R has been reduced a few sizes?

It is reduced to 1" copper at the boiler?

Without seeing the actual design, I'd guess you want to see around 90F at the supply manifold, or pump discharge.

That small zone may short cycle the boiler? Does it modulate down?

6100 square feet total with a 122,000 boiler gives you 20 btu/ sq ft. Where is the shop located? Well insulated including overhead doors?

TimH

https://forum.heatinghelp.com/discussion/comment/1827478#Comment_1827478

I can't read the printing on it, but I'm thinking it might have been a remnant from another building we did in 2007, and that tubing says "Comfort Heat Systems" and "pex-al-pex". Does that tell you anything?

GroundUp

I actually can't believe the rest of the building is being heated sufficiently with that many loops on an 007. Either way, in order to perform as anticipated, both zones need check valves to prevent backflow from one to the other.

With that said, if the boiler gauge is reading 135* while the copper near the circ is reading 114* and the PEX is reading 90*, there is minimal if any flow and that needs to be addressed. It's very possible that the loop is air bound and needs to be purged.

TimH

https://forum.heatinghelp.com/discussion/comment/1827481#Comment_1827481

Thanks for the response.

There is no floor covering except a few floor mats. We also have some desks, bookcases, walls, built-in counter. Total floor area covered is about 65 sq ft out of the 363 total. There's 1 outside walk-through door and 1 window.

Yes, the main supply and return lines are 1" copper. Does that have a direct connection to the problem of low heat output in the small zone?

You asked, "Does it modulate down?" Not sure I understand the question, but I believe the boiler has 4 equal heating stages. When the small zone is running by itself, only 1 of those stages will heat, turning on at 120 F and off at 150 F.

The shop is at 45 degrees north latitude in the Midwest. Yes, it's well insulated, including 3 overhead doors.

hot_rod

turn the small office zone on, turn the shop zone off

Then close the ball valve above or below the shop pump. Give it an hour or so and check supply and return temperature on the small zone

TimH

https://forum.heatinghelp.com/discussion/comment/1827564#Comment_1827564

I actually tried that the other day. Measuring the temps on the black pex supply and return, I got a temp differential around 25, so even higher than the 15-18 differential otherwise. My idea was that it might have even less flow because of the slightly higher head required to push the full volume up to the boiler, instead of allowing a portion of it to reverse flow through the large zone.

But I wasn't very confident in the accuracy of my temp results, so would like to try that again to verify.

DCContrarian

https://forum.heatinghelp.com/discussion/comment/1827558#Comment_1827558

I would eliminate the possibility of it being air bound before investigating anything else. OP had written, "There is no gurgling sound to indicate air in the system." That's not how it works, if it's air bound you just won't get any water flow.

I'd shut off the valve to the big zone to isolate the little zone. Put a hose on the hose bib and stick the end in a bucket of water, and open it up with the little zone running until you get hot water coming out. If you get some bubbles along the way you've found your problem.

hot_rod

when you measure temperature difference, ∆T, across a loop like that you need to wait until you reach what we call thermal equilibrium.

When a zone first starts up in a space the delta will be high, could be a 30 delta or more.

As the surface warms that delta will close up, because the space is warming up. So as it reaches room temperature in that zone the delta could be 10° or less. It goes to 0 when flow stops.

When you design a hydronic system you select an operating delta. For comfortable residential slab maybe 10- 15∆

For commercial slabs it could be a high as a 30∆, as consistent floor temperature is not as critical in a work or storage space.

You will know your system, or zone is at thermal equilibrium when the supply and return temperature stabilizes, it is no longer going up or down. That is when you want to take the temperature readings.

Additionally if you know or calculate the flow in that zone, this formula

500 X flow X (supply- return)

Lets call it .65 gpm in that single loop. 500 X .56 x 15∆ = 4,875 btu/ hr is going into thespace. If you have a heat load number for that zone, it should be around that number?

Assume that room is 363 sq ft? You are delivering 13 btu/ sq ft. That might be a tad light on design days, especially if you have a heat flux of 298.

298 is the actual square footage you have as the radiant panel, subtracting out cabinets, etc.

That single loop with its own pump should get plenty of flow.

Brings me back to flow reversal by not having checks on both pumps, this is what happens. Which also explains the odd temperatures you are seeing.

Screenshot 2024-12-18 at 8.22.41 AM.png

DCContrarian

To add on to what @Hot_water_fan said, the flow of heat into the room is determined entirely by the surface temperature of the floor relative to the room temperature. The formula is you get two BTU/hr per square foot for every degree difference between the floor and the room. So let's say you're getting 13 BTU/hr per square foot, that means a floor temperature 6.5F above room temperature. If the room is at 60F, the floor would be at 66.5F, which is a temperature that would not be perceptibly warm.

So the system could be pumping that 4875 BTU/hr into the room, but you don't notice because the floor is cool and the room is even cooler because that heating level isn't enough to meet the needed heating load.

Measuring the floor temperature with an IR thermometer can be informative.

hot_rod

Also, if you want to increase the output of a radiant floor, open the doors and windows :) Ambient temperature drops and floor ∆ increases.

Don't try this at home, as the room will get colder not warmer.

For the average consumer, they look at the thermostat on the wall to gauge comfort. It it is set a 70 and the room is actually 66, they want answers. They typically don't want or care about BTU, ∆s, gpm… just make it warmer in here.

For many of the locals on HH, the answer is in the numbers. Gather data to troubleshoot, compare it to design or expectations.

TimH

Thanks to everyone for the insights.

@hot_rod Thanks for the education. I'm a numbers-loving DIYer, but don't know the technical terms or formulas, so that's very helpful.

Here's some more data now. I closed the ball valves for 2-3 hours to isolate the small zone from the large and let the small zone run while I checked temps. From the boiler cut-off at 150* to the heat call at 120* took about 1/2 hour. Took 3-5 minutes to heat boiler from 120 to 150. The boiler overshoots the cut-off to 155.

Because the supply temp from the boiler was constantly varying, I didn't know how to determine "thermal equilibrium", so I measured the range of temps. On the black supply pex, it ranged from 93-113, spiking to the peak within a few minutes of boiler peak, then slowly dropping. The black return pex stayed fairly stable, varying 1-2* and climbing slowly over time from 76-78 to 79-81. Would you figure the delta T by using average supply temp in this case?

To get a more accurate comparison, I opened the valves again immediately after getting the above temps and checked the temp ranges again over an hour or so. Black supply pex was 85-99. Black return pex dropped to 75-77, and again climbed slowly over time. This seems to confirm that a check valve would definitely improve heat output by eliminating flow reversal, right?

Outside temps this forenoon were lower 20's. By noon the office room temp had climbed to 72 without the large zone running at all. With the data above, isn't it obvious that the system isn't air bound? Would you still recommend purging the system? The contractor who installed the circ 3 years ago had purged it then. Recently we checked the circ for air by shutting off the system, closing the valves above and below the circ, cracking the upper flange, then opening the lower valve. No air seemed to escape before water.

hot_rod

I don’t think you have an air issue.
The office is 363 sq ft? It would be nice to know that loop length. If it is 300 that is not a lot of coverage. But that 008 should be moving plenty of flow.

The concerns I have is the undersized boiler piping to handle all the loops. If it is 1”, that is good for around 8 gpm. It would be nice to flow 12 gpm or so with 21 loops.

A quick easy option would be to squirt a hydronic cleaner in the system, run it for a week, flush and refill. That would eliminate the question of dirt or sludge build up. If you do, flush into some plastic buckets to see what comes out. If performance has tapered off over the years, that is one indication of scaling

Several brands of these squirt cans to inject a cleaner.

How well is the slab edge insulated? That is a huge energy loss area, especially if it is uninsulated and exposed.

The more you can insulate, seal any air leaks, the less heat you need to pump in, and the $$of that energy.

Overhead door seals can leak a lot of heat out also.

IMG_0891.png

TimH

I believe the system has glycol in it. If I flush everything, what's the best way to get glycol back in?

The slab has 1 or 2' wide Styrofoam insulation angled out and down, around the perimeter.

hot_rod

Yes, glycol hampers heat transfer also, and when it goes bad it can sludge up the works. Check the ph of the glycol. When it is new it could run 10- 10.5. If it drops into the 8 ph range, you need to dump, flush, run a cleaner, flush, and fill with premixed glycol from 5 gallon containers.

Do you know what type of glycol? It should be a PG hydronic glycol. If it is an EG or automotive glycol you need to put it in a drum and take it to a recycler.

Do you really need glycol? Is the shop left unattended for periods in freezing weather? It is expensive, maintenance prone, and less heat transfer compared to water.

On my small shop, with plain water, I run the pump constantly and just kick the boiler on when it needs heat. So water is always moving in the tubing.

If you have 21 -300' loops ( 6300' of tube) at .92 gallon per 100' of tube, 5 gallons in the boiler and piping. So you are looking at a maybe 20 bucks a gallon for hydronic glycol $$$.

If you do go with glycol, a small pump like this, drop it right into the bucket to pump it into the

Screenshot 2024-12-19 at 8.26.55 AM.png

system

DCContrarian

If you can heat the office to 72F when it's 20F out I think it's safe to conclude that this loop is working as designed.

What we're left with is the observation that the office loop only seems to work when it is isolated from the rest of the building. The solution would seem to be to isolate it. A check valve would be the usual way to do that.

I'd have to think about where it should go.