Bad closed loop radiant heat installation? (low temp, no flow gauge, noisy check valves)

AlisonYoung

I moved into this house a year ago and I've been trying to diagnose the radiant floors taking a long time to heat up and it's inability to get warmer than 70ish degrees despite being on all day. I've done tons of digging around and hopefully have enough info such that potential issues could be pointed out.

First here's a picture of the main plumbing from the tankless water heater to the radiant closed loop system.


The circulator pumps are Astro 250ss set to the highest speed. I'm seeing the water heater output 140 degree water and it intakes water that's about 90-95 degrees. So it's exchanging the heat to the closed loop system, but it never gets higher than that. Does that mean the temp drop in the closed loop system is too much? If so, how is that remedied? Any other possible causes for the temp in the closed loop system not getting hot enough?

This next picture is of a cabinet with all of the tubing going into the home.


I've seen lots of other pictures where a metal manifold is used and they typically have a flow gauge as well as air eliminators. Was the installation done cheaply or is this still a decent installation? Is it easy to retrofit this installation with flow gauges and air eliminators? I recently did a flush of the system to get out the air that was in there and it's running very quite now. But it would have been nice to use an air eliminator.

The last thing is that I've noticed that there's a fig 233ab check valve right after the hot water output of the heat exchanger. When only 1 zone is on it can make rattling sounds like the valve is flapping open and closed rapidly. Is that anything to worry about? Should there have been a different type of check valve used?

Thank you for taking the time to read and help out! I really appreciate it!

gaabbee

Yea you have some issues. Lets start here first. You should not be using any setbacks with radiant floors. If you want it at 70, set it there and leave it. Next lets look at what kind of insulation you have. Any exposed radiant tubes below in the joists? If so how are they connected to the floor above? Have you adjusted any settings on the system yet? Pump speed, mixing valve, hot water heater temp?

ewang

Some of those EP multiports are way too close together. Uponor likes a certain distance between expansion rings.

Also, I don't think Uponor makes an oxygen barrier pex in red, meaning thats likely Aquapex for the supply/return piping.

AlisonYoung

I've set my temp to 69 and left it there. During the day it reaches the desired temp but by the time I wake up, the temp is usually 64-65. It usually takes 4-5 hours to reach back up to 69. I wonder if the nest thermostat isn't kicking it off again appropriately since it's a "learning" thermostat. I'll check into that.

I've set all pumps to the max speed. I think the hot water supply from the heater goes straight to the heat exchanger, and there's no additional mixing valve for the closed loop system. So it should be at 140 degrees.

I'll need to pull up some of my wood flooring to figure out how the tubes were installed. What should I be looking for? Should there be insulation on all of the tubing below the floors?

In regards to the Uponor multiports, what sort of issues would arise from them being too close? Also is there a place to read documentation about proper installation?

Thanks for the replies! I've got a few things to check out now.

Jamie Hall

Agh. The Nest thermostat is about the worst possible choice for radiant floors. See if you can defeat all the learning/occupancy features and set it to hold a constant temperature. It may take some fiddling in the instructions and menus.

ewang

At this point, if its not leaking, I wouldn't mess with the multi-port installation. You can look up Uponor's design manuals for pluming and radiant heating. Here's one for plumbing that shares the recommendations on distances between connections.
https://www.etnasupply.com/ASSETS/DOCUMENTS/ITEMS/EN/Uponor_LF2935050_Manual.pdf

BTW, where does the connection go thru the wall on the multi-ports where it appears they disappear to the right?

Petro55

Can you provide what size boiler you have?

How big of space you are heating?

Location your house is at?

How many zones?

Any idea how long the pex runs are? Looks like 1/2 pex.

AlisonYoung

Jamie Hall said:

Agh. The Nest thermostat is about the worst possible choice for radiant floors. See if you can defeat all the learning/occupancy features and set it to hold a constant temperature. It may take some fiddling in the instructions and menus.

I've disabled the learning portion of the nest now. So hopefully it'll just stick to the temp that I've scheduled it for.

ewang said:

BTW, where does the connection go thru the wall on the multi-ports where it appears they disappear to the right?

They don't go anywhere. Those are the ends of the tubes. I guess they just ran out of space and carved out a little bit in the cabinet to fit them.

Petro55 said:

Can you provide what size boiler you have?

How big of space you are heating?

Location your house is at?

How many zones?

Any idea how long the pex runs are? Looks like 1/2 pex.

The water heater is a Navien NPE-240A. 199k BTU.
Heated space is ~3k sqft.
House is located in the Seattle area.
There are 3 zones. 1 zone is for the basement which I have turned off pretty much all the time since I like it cold down there.
The pex lines are 1/2 inch. I've got no idea how long the pex runs are unfortunately. I'm trying to get ahold of the installer to see if they have schematics for the installation.

gaabbee

Ok lets see if removing the learn feature helps. Setting your pumps to the fastest speed does not make your house heat up faster. It will lower the temperature delta (difference between supply and return temps). We can't see what your return temp is but say your supply is 80* and the pumps are returning 80* then not much heat is being emitted to your house and you are just wasting electric. There are formulas for this, but I would lower the pump speed and see what happens. Take note of the supply temp before doing this. If you pump out 100* and return 90* the system will operate more efficiently and feel warmer, just don't let the supply temp drop from where it is. Increase the speed if it drops too much.

Second, the floors. The pex should be attached to your floors using aluminum plates. This helps transfer the heat to the floors. I would also have some insulation in the stud bay to keep the heat rising. Like others above, if it's not leaking I wouldn't touch much yet. There are things you can do to improve it but lets try the simple stuff first.

Petro55

Can you clarify the temperature after the heat exchanger? Looks like you said it was 80-100. That temp should be 110-120.

Can you adjust the mixing valve to increase the temp going to the exchanger?

gaabbee

@Petro55. We don't know what the system design temperature is. You can't assume this is their problem. They may have been using a setback feature on the nest. If it is designed for 80-100* then lets see if we can make it work at its designed efficiency. Raising the temp just burns more fuel. If it can't hold 70 through the night then there are still other options.

Ironman

@AlisonYoung,
What may be the biggest flaw in that system is the attempt to use a water heater where a boiler should have been used. Though an instantaneous water heater may look the same as a mod/con boiler from the outside, there are serious internal differences. The water heater is not designed, controlled or approved for space heating. It responds to flow and inlet water temp to fire and modulate its burner. It’s designed to have 50* inlet water with a minimum of 1/2 gpm flow. It’s heat exchanger has a very high resistance to flow (think 60 psi and an open faucet going out).

A mod/con boiler, on the other hand, has a heat exchanger that has a low resistance to flow since a hydronic circulator can only produce about a 5 psi pressure differential. It’s looking for an inlet (return) water temperature that’s about 20* below the supply water temp.

Since your hydronic circulator can only produce about 5 psi difference, it can’t overcome the high resistance of the water heater’s HX and therefore can’t move enough water through it to get it to supply enough btus to the plate HX that transfers heat to the floors.

A much larger circulator, like a Grundfos ups26-150, may help to correct this to a large degree, but you’ve still got the wrong heat source. The ups26-150 circulator would have to be stainless steel since it’s in the potable water circuit.

AlisonYoung

Some updates:
I think I may have fixed 1 pretty big issue. The wiring in the Taco 3 zone switch had an error. The basement zone C wire was put into the wrong slot and seemed to cause it's pump to start up even when that zone wasn't active.

Disabling the nest learning schedule and setting it to 70 seemed to help as well. I think the nest was just not running at night and starting up a few hours before wakeup which caused it to have a lot of catchup to do.

Reading about heat exchangers it seems like the supply for the hot and cold should happen at opposite ends. Mine is configured where they enter at the same end. Not sure if that's hurting efficiency of the heat exchange as well.

As for how well the heat exchange is working, it seems like the water heater intake temp and the temp in the closed loop system match up within a few degrees. So if the closed loop system is running at 90 degrees, the intake temp is usually 91 or 92. So I think the heat exchanger is equalizing the temp between the 140 degree water heater and the closed loop system.

I tested closing all but 1 loop and the temp for that loop was able to get over 120 before I opened more loops back up. That leads me to think that the heater isn't able to keep up with the temp drop from the closed loop system. I'd either have to increase the flow of the closed loop system so that there's less temp drop, increase the heat of the heater, or increase the flow out of the heater to transfer more btus. The water heater is showing 1 gpm while only the hydronic system is running. I'd need a more powerful circulator to force a higher flow?

AlisonYoung

Found some info about heat exchangers being more efficient when the hot and cold flow in opposite directions.
thermex.co.uk/news/blog/605-why-counter-flow-heat-exchangers-are-more-efficient
tldr: smaller heat exchangers only yield a 1-2% improvement but much larger ones can get 10% increased efficiency. So not that big of a deal for my use case.

gaabbee

You need to stop thinking that faster flow provides more heat. I'll show you some calculations. Btu/480(delta)=gpm. So lets plug in
10k btu heat loss/480(10*)= 2.08 gpm
10k btu heat loss/480(15*)= 1.38 gpm
10k btu heat loss/480(20*)= 1.04 gpm
So by changing the temperature delta (changing your pump speed) you don't increase or decrease the amount of btu's because the radiant pipes are a fixed length. If your house is maintaining the temperature now with the thermostat and wiring changes I would start to lower the pump speeds on the closed loop side. Shoot for a 10-20* delta for comfort and efficiency. Also look at improving your insulation for added comfort.

hot_rod

Hmmm, maybe I'm misreading your post @gaabbee

Faster flow does in fact increase heat output. This is true of any heat emitter, radiant, fin tube, forced convectors.

The heat output is based on the average temperature of the heat emitter.

Higher flow = higher average heat emitter = higher BTU/ hr output.

Here are some graphics to illustrate.
Identical loops all supplied with the same SWT. Add the SWT and RWT, divide by 2 to get average temperature.
The color of the loop shows the higher out in the higher flow loop.

Fin tube output charts show output at 1 and 4 gpm usually, higher flow =higher output, same with fan coils, etc.

The trade off is the extra pumping power required to run tighter ∆T. With todays ECM circulators that number is not so frightening, stay below 5 fps velocity.

For radiant floor comfort target a 10- 15 ∆ across the loop. Panel rads, since you don't stand on them, generally, can run a wider ∆.

gaabbee

@hot_rod the op’s system does not have a return temp so it’s hard to tell somethings here. They say the pumps are on the fastest speed setting and if the delta is already at a hypothetical 0 my understanding is that adding gpm via a bigger pump will not improve his heat output. I think they said Seattle for location. 

If it was a colder day and his deltas increased to say 20 or 30 degree then yes a bigger pump would be able to decrease the deltas and bring up the avg temperature. Sorry it’s late and my brain is fried right now to try and explain my thoughts more. 

AlisonYoung

Just judging from the delta in temp at the heat exchanger, the return temp is well below 95 degrees. The heater outputs 140 degree water and gets a return temp of 90-95 after going through the heat exchanger. So it's transferring quite a bit of heat over to the closed loop system.

The biggest issue I'm seeing right now is when the 2 main zones are both active at the same time. That means all 24 loops are flowing and the loss of heat through the system is so much that it takes forever to actually heat everything. In the meantime, I've closed the loops that go to rarely used spaces (closets and spare bedroom). I'll see how 8 less loops helps with the deltaT.

Eventually I might need to just get more powerful pumps to be able to increase the flow.

Ironman

The universal hydronic formula:
Btus= (delta Tx500)/gpm.

You have a delta T of 45* (140 - 95). 
45x500= 22,500 / 2.6 gpm = 8,653 btus. 

8,653 btus! That’s what’s being DELIVERED to your house. It doesn’t matter how large the heat source is if the train that carries the load can’t haul any more than that.

The 2.6 gpm is taken from Navien’s pressure drop chart below.

That water heater has a small internal circulator which is intended for domestic recirculation which now has an external circulator in series with it. You can’t put a much larger external circulator in series there.

As much as you may not wanna here this, the only correct solution is to replace the water heater with the correct appliance: a boiler.

The sad thing is that a combi boiler would not have been that much difference in cost than that water heater, but the installer tried to cheap out. Also, had he used O2 barrier pex on the loops, the more expensive bronze circs and the plate heat exchanger would not have been necessary.

In the long run, it pays to do the job right to start with.

AlisonYoung

@Ironman Thanks for all of that explanation and information.

One thing that isn't adding up is that the water heater is showing a 1 gpm output which would mean the circulator pump isn't generating the actual 5 psi pressure differential. When I turn the pump down to the lowest setting, the water heater outputs 0.6 gpm.

Could something be restricting flow in that loop? Maybe the heat exchanger that's outside of the water heater is needing even more of a pressure differential to flow faster? I'll try to look up more info on that.

Another thing I've noticed is that the check valve between the heat exchanger and circulator pump tends to flap open and closed instead of staying all the way open. So the flow does seem to be really poor through that loop.

BillyO

when you looked at house, was boiler room door locked?

AlisonYoung

BillyO said:

when you looked at house, was boiler room door locked?

No. I didn't know anything about radiant heating systems, so I've been learning about it now to make sure I understand how it's suppose to work so that if I need to call someone to come change or fix something, I know what they are doing.

Zman

When I look at these photos, I keep asking myself "what is done correctly?" The answer is absolutely nothing!
You would think that even a bumbling idiot would have accidentally done some thing correctly once or twice. Even a broken clock reads the correct time twice a day...

Even if you chose to keep the tankless water heater as a heat source (not a great plan), the pipe sizing, pumping away, and overall schematic is just plan wrong. My advice is to have someone qualified redesign your system and then take a sawzall to the whole mess. You will never get that mess to function like a comfortable, trouble free heating system.

Tinman

@Zman @Ironman @HotRod couldn't have said it better. I think it would be so much harder to do a job THAT badly than to do it the right way. Zero thought went into that mess.

HomerJSmith

Very, very interesting! I thought I'd seen everything!

Those black manifolds are really for home run domestic water sys where street pressure is higher than 12 psi..

AlisonYoung

Honestly, after flushing the system of bubbles, re-pressurizing it, fixing one of the pumps being wired incorrectly, and stopping the use of the nest learning schedule, things are working decently well. It's holding a temperature of 70 degrees all through the night and day now.

All of this has been a great learning experience tho and I have really appreciated all of the knowledge being shared. Now that I understand how things should work, I can weigh the option of having it corrected without blindly trusting whoever comes to look at it.