Boiler Flow Rates

LST3 · July 2019

Hey guys, I am going to apologize in advance for the long post, but I know people hate when threads don't include enough information. Feel free to skip to the bold part for my actual question.

I am currently building a cabin in NY near the Adirondacks. We are utilizing in-floor radiant to heat the building (with provisions for adding a low temp fin tube loop to the 2nd floor loft if needed.) The building will have a wood stove in it for supplemental heat, but I don't want to factor that into the design and be forced to use it.

I have been researching everything and seem to have narrowed down my design to use an HTP UFT-80W boiler and a SSP-40 indirect HW tank for DHW.

I am unsure whether I should pipe the boiler direct, or with P/S piping. The boiler manual shows it being piped in either fashion, but I am not sure that my flowrates support this (or that it would be the best way to go considering we may add fintube in the future.)

In the basement slab, we have 4 loops of 1/2" pex (250', 250', 230' and 230') laid on top of ~2 1/2" Creatherm insulation panels (R10 according to spec sheet). LoopCAD seems to calculate an output of 13,400 BTU/Hr @ 100° supply temp and 1.5 GPM (for a 20° delta.)

We will be installing the radiant tubing in the 1st floor joist bays in a few weeks using heat transfer plates and R19 fiberglass insulation below (pushed loosely up against the plates.) I have designed the tubing layout with 8 loops, approximately 195-205' each. For a total output of 23,900 BTU/Hr @ 158° supply temp and 2 GPM (for a 20° delta.) I know that the supply temp is not ideal in this scenario, but the heating load is pretty big for this floor and I haven't figured out a better way to lower it (we can't justify the extruded aluminum plates, and the floor/walls have already been built, so we don't want to install tubing above the floor.) I am hoping that outdoor reset will still keep us condensing the vast majority of the time.

Does it seem like the best route to pipe the boiler directly, with one circulator for each zone? I would install a flow meter on each zone to verify adequate flow through the HX at all times (the manifolds have those flow meters for each circuit, but I don't know if I want to trust those.) The manual lists a minimum flow rate of 1.3 GPM at minimum output. We are slightly above that, so in theory it seems as though everything should work fine with direct piping, and it would save on having an additional boiler loop pump running all the time. But I keep going back to thinking that I should just stick with a tried and true P/S setup.

Also, if I were to go with a P/S design, has anybody used a variable speed circulator for the primary loop with this boiler? I see that the boiler has a 0-10v output, but the manual doesn't mention anything about using that to drive a circulator pump. And I am not sure if it would be a good idea to use a variable speed circulator driven by a delta T. Just seems like a waste to have the boiler pump circulating away at 3-5 GPM or whatever even when the secondary loop is only flowing at 1.5 GPM (say if only the basement loop is calling for heat.) In my mind, this equates to high return water temps, which could keep the boiler from condensing adequately like we want. I have also heard that it is not a good idea to have the primary loop running at a higher GPM then the secondary loop, but I don't see any way around this besides using a variable speed circulator.

In either case, I would use the boiler controls to vary the 1st floor loop between minimum/maximum temp (based on outdoor reset) and just use a mixing valve to supply the 100 degree water to the basement loop (I have considered using one of those Taco mixing valves with built in outdoor reset, but I am finding it hard to justify bothering with that since the supply temps are so low on this loop even at design conditions anyways.)

Building Info
The building is 1000 S.F. per level (levels include a below grade basement, a first floor, and then a partial 600 S.F. second floor). Insulation will be 2" of closed cell spray foam with fiberglass batts over top to meet the R21 code requirement, and 3-4" of closed cell spray foam on the roof with fiberglass batts to get us to R49 for the roof. Zoning will be pretty simple, one zone for the basement slab and one zone for the 1st floor, controlled by thermostats mounted on the wall (or potentially a floor sensor of some sort if needed.) We will be installing pex in the walls (and adding an extra "zone" connection to the boiler piping) to utilize at a later time for low temp baseboard heat if we decide that the 2nd floor is not adequately heated (we would like to avoid this, but don't want to shoot ourselves in the foot if it turns out the building doesn't heat like we planned.)

Building heat loss is as follows (calculated using Trane Trace 700, -8° DD),
Basement - 14,500 BTU/Hr
1st Floor - 20300 BTU/Hr
2nd Floor - 9,400 BTU/Hr

Thank you in advance for any insight or advice that is given. I am open to any suggestions for the design. The basement loop is the only thing "set in stone" at this point, but the 1st floor loop will be installed in a few weeks. Nothing else from the heating system has been purchased yet.

GroundUp · July 2019

Minimum input on the UFT80W is 8,000 BTU, so with your small zone being 13,400 you would be at 20% fire with some excess, requiring ~2.3 GPM minimum flow across the HX if my calculations are correct. Being your heat loss in the lower level is a bit higher than that, you may wish to raise the flow rate and tighten the delta a bit to meet the 20% firing rate but given the low load I would still probably pipe it P/S to be safe. There is nothing wrong with the primary (boiler) loop flow rate being higher than the secondary, that's how P/S is supposed to operate. What arrangement of transfer plates do you have figured for the 1st floor that the 23,900 output needs 158 degree supply water? Sounds like you need more plates to lower that SWT, possibly more tubing- or both. This would potentially allow you to use the same temp water for LL and 1st and skip the mixing valve ordeal by using the ODR within the boiler. This is merely a personal opinion, but if I understand the system correctly and were doing it in my own home, I would pipe the system P/S with either a Grundfos 15-58 or Alpha 15-55 on the primary and an Alpha on the secondary with Taco Sentry zone valves and do whatever it takes to make the SWT all the same so you're always taking full advantage of the condensing boiler.

LST3 · July 2019

Good point about the higher minimum flow rate at 20% fire.
I also forgot to mention that we are going to be using 40% glycol. I should have included that in the original post.
I haven't decided which plate to use yet, the extruded plates are not in the budget, but I am open to suggestions on what is a good middle of the road plate.
I would like to use a lower water temperature for the 1st floor loop, but I am having trouble getting the amount if output out of the floor area at a lower temperature. 8 loops of 200' each fill all of the 1,000 S.F worth of floor joist cavities with a supply and a return in each joist cavity, with "heavy gauge heat transfer plates" only seems to put out about 22,000BTU/Hr.

STEVEusaPA · July 2019

You don’t think you have enough room to use 3/4” warmboard on top of the subfloor?

LST3 · July 2019

STEVEusaPA said:
You don’t think you have enough room to use 3/4” warmboard on top of the subfloor?

We have already framed the interior walls, and with it only being 1000 S.F. to start with and the floor plan layout being what it is, it seemed like it would be extremely hard to get the amount of tubing laid out in equal loops. The headroom isn't necessarily the issue, more the logistics of laying out the floor after the interior walls have been framed. The walls were only framed with a single bottom plate (which could be worked around by simply nailing in short boards where needed) but it shortens the door and window heights as well.

Do you really think that getting it above the subfloor would net that much gain in reduced temperatures VS heat transfer plates below the floor? How much better would it be then the extruded transfer plates, if I managed to get ahold of enough of those to mix them in with say, every other plate? Or even use those in the high heat requirement areas of the 1st floor and omega style sheet metal plates in the remaining areas?

I would obviously like to use as low of temps as possible, but I should already have relatively low temps for the majority of the heating season anyways due to the outdoor reset.

Zman · July 2019

Where are you getting the output numbers for omega plates? I am surprised that you need 158 degrees. Unless you have huge heat losses, I think it should run at 130-140. You might consider a non-thermostatic mixing valve for the lower slab which will give you supply temps that run a parallel outdoor reset curve to the upper level. A thermostatic valve will give you a fixed temp and a smart valve may add excessive complexity to the system.

Consider running the slabs at a 10 degree delta. You will get more even heating and will have less air lock issues. With your relatively short tubing lengths, most common circs will easily move water at that rate with minimal increase in cost. This would be a good application for delta T zone circs.

If you desire to chase a few extra efficiency percentages by varying the flow to the boiler, the safe move is to go primary/secondary and vary the flow rate according to firing rate. Lochinvar has products that does this automatically. I don't believe HTP has this available on board, @Rich_49 has tons of experience with HTP, I think he has found a way to do this using delta T boiler circs.

I generally prefer delta P circs in systems that have zone valves with a variety of emitters. Delta T is the way to go for individual zone circs and boilers.

LST3 · July 2019

> @Zman said:
> Where are you getting the output numbers for omega plates? I am surprised that you need 158 degrees. Unless you have huge heat losses, I think it should run at 130-140. You might consider a non-thermostatic mixing valve for the lower slab which will give you supply temps that run a parallel outdoor reset curve to the upper level. A thermostatic valve will give you a fixed temp and a smart valve may add excessive complexity to the system.
>
> Consider running the slabs at a 10 degree delta. You will get more even heating and will have less air lock issues. With your relatively short tubing lengths, most common circs will easily move water at that rate with minimal increase in cost. This would be a good application for delta T zone circs.
>
> If you desire to chase a few extra efficiency percentages by varying the flow to the boiler, the safe move is to go primary/secondary and vary the flow rate according to firing rate. Lochinvar has products that does this automatically. I don't believe HTP has this available on board, @Rich_49 has tons of experience with HTP, I think he has found a way to do this using delta T boiler circs.
>

> I generally prefer delta P circs in systems that have zone valves with a variety of emitters. Delta T is the way to go for individual zone circs and boilers.

I am using LoopCAD to calculate my zone output. It is actually using their version of "heavy gauge transfer plates" in the modeling to give me that 22,000 BTU/Hr heating output at 158 degrees and a 20 degree delta. I am hoping that in the real world scenario will have it running at a good bit lower supply temps the majority of the time, but I can only base my actual assumptions off of what the program is outputting. If there is other data available that would show me getting that much heat output out of the floor with lower supply water temps then I am definitely all ears. This is the biggest Grey area in the design. I haven't found much for concrete data available that would give me a heat output for a given temperature with the heat transfer plates (besides the LoopCAD software). I would love to know right now exactly what I will be able to achieve for a flow rate, heat output, and delta t in my design, but it seems that data is hard to come by.

I will have to research more into applying a delta T circ for the primary boiler loop (any and all real world input from @Rich_49 would be greatly appreciated on this subject). I was concerned that this could interfere with the boilers internal temp sensors or programming logic (aka I am worried the boiler could have trouble calculating the actual system loads at a given time if I am constantly varying the circulation flow rates). This seems like an efficient way to do it, I just want to verify that it would work as intended.

As for the thermostatic mixing valve, I have considered the non-thermostatic mixing valve, I just need to hash out the flow rates and temps first. The same thought had crossed my mind about creating a sort of hybrid outdoor reset for the basement loop based on the varying loop temperature available for the 1st floor.

Boiler Flow Rates

Comments

Categories