Help Needed: Radiant Floor System Temperature Drop Between Boiler and Manifolds

pstutcher

New DIY radiant floor guy here. Running into a little problem that I don't understand and I need some help. I'll ask the question first and follow that up with some details of my system to answer as many of your questions that I can foresee as possible.

Question: Why is the temperature drop between firings dropping 25-30F at my output manifolds when my boiler temperature is only dropping 5F?

Background:

Early 70s-built house with baseboard heat with a boiler that I suspect is original. It worked great though I think greatly oversized. It is an 80% efficient gas boiler with a rating of 165k/132k BTU/hr.

We were tearing out all of the flooring in the house - poorly installed kitchen tile and nasty old carpet everywhere else - and replacing it all with hardwood. We didn't care much for the baseboards not allowing us to push things against wall and such and have always wanted in-floor radiant.

On inspection, we found that our house was built with a 3/4" plywood subfloor topped everywhere with a second 3/4" MDF subfloor and made the decision we would slice the MDF up and use it to embed the 1/2" pex required.

I went about the tasks of educating myself on all of the requirements, calculations and designing the system, ordering components and starting to build it out.

For various reasons, my intention was to continue to use the existing boiler and pump but design the system in a way that allows an upgrade to a high efficiency condensing boiler with smaller variable speed pumps later on.

Basics of the house: 1840 square foot ranch, brick exterior, modern windows, well insulated attic. Using different calculators produced differing results but centered around 60-70k BTU/hr heat loss.

I'm located in the Metro Detroit area, so it gets plenty cold here in the winter so -20F design temp was used with a 90 degree rise requirement.

Knowing the boiler was rated at 132k BTU/hr I was confident that it could keep up.

I went for full control and turned the 3 zone main floor system into 6 zones to be able to control the bedrooms individually plus the bathrooms.

I dialed the high limit control aquastat on the boiler to down from 175 for the baseboards to 120F with a 5 degree differential. My secondary thermometer readings confirm the temperature is staying within that range while the system is running.

I've got most of the system installed and operating and so far it's doing up fairly well. All of the wiring, controls, valves and the boiler/pump controls are behaving the way they were designed to.

But it has stayed quite warm late in the year for this region and I'm worried when it does get substantially colder it might not be able to keep up. So far, on days that it gets down into the 30s, my thermostat setpoint of 69F isn't being achieved, it seems to stall out at 67F. As it turns out, 67F is pretty comfortable for us so we're not overly concerned, but like I said, I'm not convinced it will keep up when the serious cold sets in.

Yes, I'm aware that my boiler is short cycling, it's kind of unavoidable given what I've asked it to do and I know that I'm shortening the boiler life with a too cold water return temp but like I said, it only needs to go a couple years or less before it's being replaced anyhow.

But what I am seeing is a little strange. As I said, the boiler temp is staying in the 115-120F range, but the temperatures measured at my zone manifolds are very different. They are about 10-15 feet away down a 1.5" diameter main copper feed line away from the boiler and pump.

When the boiler fires, it only runs about 90 seconds and it does this about every 8-10 minutes when there's a call for heat. The temperature at the supply manifolds rises shortly thereafter to about 130F. I find this strange as my boiler temp has been limited to 120 and they all basically agree, so it does not seem to be an instrumentation problem.

But what really puzzles me is that after the initial temperature rise to 130, in the interim time to the next boiler firing, the manifold temperatures drop to 100-105F and it does it fairly quickly.

What this tells me is that the system is spending most of its time pumping sub-120F water through the system that has been designed to operate at 120 and I feel like this could be the reason that my overall temperature isn't getting to where I want it.

Initially I thought that the aquastat was faulty, so I replaced it with a new one but it behaves exactly the same.

Clearly I still have a lot to learn on much of this, but why am I seeing such a dramatic and prolonged drop in temperature at the manifolds between firings when seemingly my boiler water temperature isn't dropping below 115?

Any and all help is appreciated. I'll be happy to add further details if necessary.

hot_rod

How did you install the pex tube? Slots in the MDF? Any transfer plates?
Got pics of the boiler piping?
The heat load seems high at 38 btu/ sq ft? If it is that high, I doubt radiant floor heat alone will cut it on cold days.

pstutcher

Yep, I cut 5/8 slots into the MDF using a jig and circular saw and yes, I used AL transfer plates.
I will get some pictures loaded tomorrow.
How did you get to 38 BTU/SF?

hot_rod

pstutcher said:

Yep, I cut 5/8 slots into the MDF using a jig and circular saw and yes, I used AL transfer plates.
I will get some pictures loaded tomorrow.
How did you get to 38 BTU/SF?

You mentioned a 70,000 btu/hr heatload÷ 1840 square feet of home = 38.04 btu/sq ft.

About high 20's (btu/ sq. ft) for a comfortable radiant floor 25- 28 btu/sq ft.

pstutcher

Ahh, yes. Well I agree that it's probably too high in which case my boiler is way too big. I did forget to mention the basement is heated, that is to say it is fitted up to heat and is undoubtedly included in that giant boiler size but in the almost 4 years we've been there we've never turned it on except to make sure it works

hot_rod

The heated basement should have an impact on the heatload for the main level. Did the load calc program take that into account.

pstutcher

I don't recall, but I think so . I'd have to dig it all out and check. In any case what are your thoughts on that temperature drop between the boiler and manifolds?

pedmec

A picture of your set up is the only way to get a good answer. Everything else is just an educated guess.

Dave Carpentier

Wild-a$$ guess..
Every time the boiler cycles off, it turns off the boiler's circulator, and you piped in a primary secondary kinda thing with closely spaced tees ?

pstutcher

Dave, thank you but no. No primary/secondary loops and the pump runs when any zone is calling. 

GGross

Would be interested to see what you have, and how it is piped. If your load is really that high the floor likely would not keep up with the load. That being said you mentioned designing to -20f and the design temp for Detroit is actually +7f , and it has been fairly mild in Michigan. I don't think off hand that the
actual load is greater than what the floor should be able to output. We need pictures of everything

pstutcher

Here are some pics. the zone board is split into two pieces in a 90 degree corner. The boiler is at my back in the pic that shows the expansion tank (right board) and the 1.25" copper pipes run up from the boiler, across the underside of the joists above my head and come down to the right board where i show the red and blue arrows. The pictures were taken at various stages of construction and not all rooms had been connected up yet, but the distribution system is how it is shown here

hot_rod

Looks like you are running directly from the boiler to the various manifolds? No mixing valves, just boiler temperature? What temperature do you run the boiler?

GGross

How much insulation do you have under your pex? I wonder if you are transferring some heat down via those plates and heating the space below the floor.

Some more pictures of the boiler set-up would help, the picture of the pump looks like you are pumping at the fill valve and that can't be right.

Kind of a side note but where you show the partially completed manifold set up, with 5-loops going to pex, with inline ball valves and then zone valves. That is a bit of a goofy set up and might create quite a bit of head to pump against, the crimp valves restrict flow, the manifold has pressure drop, and then added more from the zone valves. they make devices to attach to the manifold that will act as zone valves should you ever need to do a set up like that again, it would save you the restriction from the ball valves, as well as the zone valves. Check out cross manifolds for a nice DIY manifold set up as well.

pstutcher

Peakedtoosoon said:

Find it hard to picture what you have installed without seeing some pics, or a sketch of the piping around the boiler. From your description, you have an older, conventional boiler, with it's own return pump, serving a mix of baseboards and underfloor heating loops, with their own zone pumps or zone valves? You also have no hydraulic separation, between the boiler and the distribution loops. Is that correct?

Peaked, Yes, older conventional boiler with a single circulator pump on the return line. At the moment, yes, it is serving the new in floor systems and one existing baseboard* that will be replaced soon. Each zone has its own valve, but no secondary pumps. I think I understand what you're asking about hydraulic separation (but I'm not certain) - the answer is no, the water running thru the boiler is the same water flowing through the zones, just as it has for the last ~50 years.

*I know the baseboards won't work at the water temperature I'm using for the floor heating but it's temporary.

pstutcher

hot_rod said:

Looks like you are running directly from the boiler to the various manifolds? No mixing valves, just boiler temperature? What temperature do you run the boiler?

120F, 5 degree differential. No mixing valves.

GGross

This is a cast iron boiler running at 120F?

pstutcher

GGross said:

How much insulation do you have under your pex? I wonder if you are transferring some heat down via those plates and heating the space below the floor.

Some more pictures of the boiler set-up would help, the picture of the pump looks like you are pumping at the fill valve and that can't be right.

Kind of a side note but where you show the partially completed manifold set up, with 5-loops going to pex, with inline ball valves and then zone valves. That is a bit of a goofy set up and might create quite a bit of head to pump against, the crimp valves restrict flow, the manifold has pressure drop, and then added more from the zone valves. they make devices to attach to the manifold that will act as zone valves should you ever need to do a set up like that again, it would save you the restriction from the ball valves, as well as the zone valves. Check out cross manifolds for a nice DIY manifold set up as well.

No insulation under the pex, not sure how that would work. The pex is laid above the subfloor.

Thanks for the advice on the ball valves, I see that now that you point it out and I'll be removing them.

Here's a picture of the boiler in its old state (when it was feeding the baseboards that have been/are being removed). The blue arrow is the return feed going into the pump and below that it turns 90 degrees away from the camera and into the boiler. Pretty standard setup I would imagine. That portion of it has been there almost 50 years.

Another angle as well showing the infeed to the boiler. Again, all of the old plumbing is now gone and it's just the out and return (and of course the city fill feed).

pstutcher

GGross said:

This is a cast iron boiler running at 120F?

yes.

pstutcher

Peakedtoosoon said:

There's a couple of things that don't seem right to me. Not sure if they relate to your issue, but here they are, for what it's worth.....

1) The expansion tank seems to be located on the discharge side of the pump. Should really be located on suction side.
2) No insulation on any of the copper pipes, so some heat loss there.
3) No sub-circuit balancing valves on the returns from each return manifold. Are the zones nearer the pump, warmer than the farthest circuits?

How is the pump switched on/off?

1. Agreed. As this is being done in stages the boiler will be removed entirely in a year's time and with it the current pump location will move as shown here. For now I have the new and old expansion tanks still in place.


2. Yeah, I considered that, but it's a pretty small room and while it's warm in there, it's not overly warm. Also, if the heat loss on the pipe were the culprit, the manifold temperatures would be lower all the time, but they do spike up to 130 shortly after the boiler fires.
3. Not sure I understand because each circuit has a valve I can adjust to manage the flow in each individual tube. (I mounted my manifolds upside down because I wanted to the tubes to enter and exit towards the ceiling not the floor so I didn't have to reverse them).



4. No, there isn't any difference between near and far zones.
5. Pump is turned on by the Taco zone controllers - on when any zone calls for heat and the zone valve confirms open.

hot_rod

Are you planning to have a two temperature system at some point? Some radiant and some fin tube?

pstutcher

hot_rod said:

Are you planning to have a two temperature system at some point? Some radiant and some fin tube?

Originally no, it was all to be floor. However, we're considering changing strategy based on the severe difficulty in pulling up and putting back down some really nice and very expensive Brazilian cherry flooring in the back room that doesn't have basement underneath it.

So for the time being, there is 120F water flowing through the baseboards in that room and we know it's not keeping up but it's only for this season until we make a decision.

I do like the notion of setting that boiler back at 175 and running the back room properly and doing either a heat exchanger or mixing valves to correct the floor temp.

Also, since this will ultimately be done with a new high efficiency boiler, there's the possibility of installing two and running the floor and baseboard systems separately.

hot_rod

I'm not sure how long before you upgrade to a mod con type boiler, but the boiler you have and the flue piping will not like those low operating temperatures. Best to boost up to the fin tube temperature requirement, 170- 180-, and throw in a thermostatic mixing valve or other type of mixer for the radiant.

pstutcher

hot_rod said:

I'm not sure how long before you upgrade to a mod con type boiler, but the boiler you have and the flue piping will not like those low operating temperatures. Best to boost up to the fin tube temperature requirement, 170- 180-, and throw in a thermostatic mixing valve or other type of mixer for the radiant.

Just need to get thru this season.

I'll take any advice you have on that.

I'm not an expert, but it seems to me that I need two of them, one to temper the outflow to the floors to 120 and the other to ensure the return temp to the boiler doesn't drop below 135. Can one device do this? It seems to me an indirect system for the floors would make sense, but I don't yet know what that looks like either.

hot_rod

here is one simple option, primary secondary with 3 way mix valve

pstutcher

hot_rod said:

here is one simple option, primary secondary with 3 way mix valve

OK, I see that. TY, this simplifies things for me. Can you recommend a mixing valve?

hot_rod

The most common, accurate, simplest to set and service is a thermostatic 3 way valve. A step up would be a motorized version that could be modulated via outdoor reset control.

If you knew what type of boiler is in your future, your piping changes now could accommodate that upgrade.
For instance, many mod con boilers need hydraulic separation. Many of the new boilers have a two temperature control on board that would maybe change decision on the mix valve you put in now.

Ideally to size a mixing valve, any valve really you need to know the gpm flow rate of the circuit you are connecting it to.

A common 3 way thermostatic mix valve, 3 Cv is good for 5- 6 gpm. Caleffi 521 series for example.

Looks like you have 9 radiant loops planned? Call it 1/2 gpm per loop, so a 4.5 gpm requirement. A 3 Cv thermostatic valve will cover that.

If you have a higher flow requirement, maybe a motorized type is a better option. the Taco I-valve seems to be a common valve for both higher flow rate and outdoor reset function.

As far as boiler return protection, if you crank it to 180F for the fin tube, and the radiant load is only 30- 40,000 btu, that 132,000 boiler you have should be able to stay above condensing mode. A rule of thumb being if the low temperature load is around 20% or less of the total load return protection is not a must have.

Adding another protection valve would look like this.

pstutcher

I appreciate all of the advice, critiques and recommendations but can anyone take a crack at answering the question I've posed?

Jamie Hall

On the original question -- I'll take a WAG at it, since the rest of the folks have covered most of the bases here.

You have one pump? And everything goes through that when the pump is running? Now when the pump turns off, you longer have any circulation, right? The boiler contains a fair amount of water, so will retain its temperature -- but the floor doesn't, and its temperature will drop quite quickly. That may be what you are seeing.

pstutcher

Thanks, no that's not it, the pump is still running. Basically because the one baseboard loop can't keep up in the back room, it's calling for heat almost continuously so the pump is running accordingly.  I could easily understand the temperature drop at the manifolds when there's no call for that or no zones.