Should I direct pump this boiler...radiant floor with one zone.

nickh1 · December 2016

I have a 7000 sq ft shop built in the mid 80's with radiant floor heat. It was heated with 2 40 gallon water heaters until this fall. I just put in a HTP UFT-80W boiler and it is humming along great. It was suggested by the boiler supplier and many things I read that it should be piped primary/secondary. The supplier spec'd out Grundfos 15-58 pumps.

There are 9 loops of 3/4" pex, but I honestly don't know how long they are. What I do know, is the pipe is only in the center 3500 sq ft. My assumption and by feeling the floor as it warmed up is they are on 2' centers, which would make them 200'ish loops. With the secondary pump on the lowest speed, I am getting about 10-15 degree's delta T. I currently have the boiler set at 115 and the return water is around 100. Both Primary and secondary pumps are one the low speed.

I'm content how it is operating, my question is if I'm wasting power with the 2 pumps and I would get the same results with one. It would not be a big deal to take out the extra pump/tee's and I have a need for a pump on another project so it would save me some cash to do so.

Mark Eatherton · December 2016

Unless you have the ability to prove whether or not the existing distribution system has a low head pressure drop, I'd stick with the two pump plan.

The parasitic cost of operation of these pumps are minimal, and it ensures peak performance from the heat source. If the heat source has it's own circulator, then it knows what it is supposed to be doing, and can "see" the influence of the connected loads and respond accordingly.

Will it work with just one pump? Probably, but not as thermally efficient as it would with two separate pumps.

I'm certain others will express different opinions, but based on personal experience, these are mine.

ME

Zman · December 2016

It looks like you have a pressure gauge on both sides of the secondary circ. With that pressure difference and the circ performance curve (15-58?), you should be able to determine the flow.
Can you post that info with the circ at the 3 different speeds?
I agree with Mark that if you don't know, you should play it safe.

Paul48 · December 2016

I don't see any means of balancing the loops. When you say you're getting a 10-15* DT, I assume you mean across the manifolds. The water will take the path of least resistance. With some electrical tape on each supply and return, and a cheap IR thermometer you can check the DT of each circuit. You can determine the length of each circuit, by flip-flopping the numbers for the Universal Hydronic Formula, in conjunction with the pex manufacturers info on head loss , and the pump curve. GPM= BTU / ( DT x500)

Bob Bona_4 · December 2016

Best to decouple boiler from an unknown system. P/S.

nickh1 · December 2016

Thanks for the responses. I will get the pressure readings and post. I also have a thermal camera, I haven't used it on the loops but thought about it. I charge it up and take some pictures.

I debated about replacing the old manifold with a newer one that would allow balancing. In my brief search, I didn't find any that had adapters to 3/4 pex. I also figured that I would have to buy two for sizing and to deal with where the pipe comes out of the floor. The original setup had a fairly large pump on each water heater that each ran half the loops. The current manifold has I believe 1" female straight pipe thread on the pex adapter with a washer. Would there be any modern manifolds that use 1" npt vs bsp?

I was going to fill the loops with water and blow it into a bucket to measure the volume of each loop, but figured in the end I can't change the outcome so I didn't' take the time. But if they were unbalanced, it would make a case for a new manifold.

I personally would be surprised if they are much over 200'. This shop was built in the middle of the 80's farm crisis years and my family has a genetic history of being minimalists. 7000 sq ft is small by today's standards, but I'm very thankful to have it, even more when the floor is warm.

Zman · December 2016

I do believe that primary/secondary can increase the efficiency of a system.

By controlling the flow across the heat exchanger, you can achieve efficiency gains. Lowering the return water temp by controlling these flow rates will increase efficiency. Not really sure if this is considered "thermal efficiency".

In the system in question, I would be a bit surprised if these efficiency improvements would be worth the additional operating and installation costs associated with the additional circ. In reality, it would take a pretty thorough study to figure it out and it probably doesn't matter much.

In the end these debates usually end up being an exercise in hair splitting.

Paul48 · December 2016

Everything being proper..... nothing can be more efficient than using the most btus produced by the boiler. Directly pumping the boiler allows for that. How is that even debatable?

Solid_Fuel_Man · December 2016

@Zman I've fo7nd that unless the primary (boiler circ) is flowing less than the secondary (radiant circ) than the return Temps to the boiler will be higher and reduced condensing the result. If the secondary flows are sufficient, then wouldn't a direct pipe guarantee the lowest return Temps back to the boiler. Not short circuiting and warming up the return Temps.

Taylor

nickh1 · December 2016

I check the psi on the different secondary speeds. It didnt change much.

Low was 20/18.5, medium was 20.5/19 and high was 21/19.

I have a 12 psi pressure reducing valve on the makeup water and filler/flushed it with cold water. Is 20ish psi in the ballpark?

Gordy · December 2016

In P/S pumped system. The system side flow must be greater than boiler side to avoid btu constipation at the boiler.

Gordy · December 2016

6 gpm on low, 11.5 gpm on med, and 13 gpm on high

nickh1 said:
I check the psi on the different secondary speeds. It didnt change much.

Low was 20/18.5, medium was 20.5/19 and high was 21/19.

I have a 12 psi pressure reducing valve on the makeup water and filler/flushed it with cold water. Is 20ish psi in the ballpark?

If loop lengths are identical divide values by 9 for individual loop flow rates. Rough guess at best.
Measuring loop deltas is most accurate when done correctly.

Zman · December 2016

There is some inherent inaccuracy in measuring slight variations with the gauges you are working with. You should absolutely check that they both read the same with the system off.
Based on the info provided, you could eliminate the primary secondary and leave the remaining circ on low. You would maintain minimum flow rates, increase delta t and save electricity.

Gordy · December 2016

The pex dia of 3/4" and assumptions of loop lengths with the above psi differential readings would, and do indicate low system pressure drop.

However as zman said accurate gauges 0-30 psi, and accuracy between the two gauges can change the readings a lot.

Gordy · December 2016

One question does the UFT cycle?

Henry · December 2016

it has 10 to1 modulation

nickh1 · December 2016

I checked it with it off. They appear very similar, but they are 0-60 so it is hard to tell.

I checked the return lines with my flir camera and as I suspected, some were around 94, others were 97-98. Would I be able to radiate more heat if I slowed flow down? Or would it just even out the floor temp more. My assumption is if I increase delta t at the expense of gpm btu's stay the same.

I also looked at the floor and the pipe is 24-36" on center. The floor temps range from 60 between and 75 on top of the line. Not ideal, by today's standards, but we aren't wearing t shirts either. The sides of the building without pipe had a floor temp of 55.

Thanks for the thoughts guys.

nickh1 · December 2016

No it is not cycling. The ignition counter counts 10x and it still reads 0. The unit has been running for 3 days. It is currently modulating at 52%.

There is a backup modine heater in the building, so it wasn't completely cold in here, but the floor definitely wasn't warm.

Solid_Fuel_Man · December 2016

Also, I'm guessing from the look and the age of that install that is PB tube. No oxygen barrier, the circulators and the little bit of iron fittings may not fair well long term. Keep it all nonferrous including the expansion tank. I'd also give each loop a good flushing until the water runs clear.

Taylor

Paul48 · December 2016

Two to 3 ft spacing is not good.

Rich_49 · December 2016

I've always wondered what might happen in the long term when a P/S system was piped incorrectly and the boiler circ was placed on the supply . Will you keep me informed of what happens ? Can it even be considered P/S when piped the way this boiler is ?

Supplies are on the left and returns are on the right , top and bottom . RTFM !

nickh1 · December 2016

I flushed everything for 20-30 minutes.

I wondered if it was barrier pex. I guess time will tell.

That diagrahm showed the pump on the supply line, but it is P/S it should be on the return?

The spirovent instructions said to place it after the boiler, and I was told to pump away from the tank, so that is what I did.

Gordy · December 2016

It appears the primary is using the top supply , and bottom return in this case. However im not sure why other than room for x tank, and air removal.

The manual for the UFT shows a definitive approach of indirect bottom set, and ch top set. Or vice versa. Not sure if using top and bottom s/r in this case is okay.

Rich has a deep understanding of this boiler.

Seems cycling is non existent.

Rich_49 · December 2016

All the diagrams published show the circ on the supply on direct piped examples . All P/S examples show circ on return . All examples whether P/S or direct show PONPC on system piping .

To make what you have correct , you must remove tees and lower circ , replace tees with 90*s . That would then be direct piped .

If you want P/S you must move the air eliminator , make up water and expansion tank to the system piping , move the boiler circ to the return .

Instructions are tricky things . We all must determine which instructions are most important , in your case choosing the instructions that come with the most costly and dangerous piece of equipment would have been prudent .

kcopp · December 2016

Solid_Fuel_Man said:
Also, I'm guessing from the look and the age of that install that is PB tube. No oxygen barrier, the circulators and the little bit of iron fittings may not fair well long term. Keep it all nonferrous including the expansion tank. I'd also give each loop a good flushing until the water runs clear.

Taylor

I noticed the Poly Butylene straight away. You are a at a slight advantage that's its a slab install and not an above grade install. (Less tubing exposed to free air) I would certainly hit it w a cleaner, flush and add a good inhibitor.... This may effect your warranty if you don't.
Fernox, Sentinel and Rohmar are reputable products.

Zman · December 2016

Rich makes good point about PONPC. If you got rid of the lower circ and the tees and just connected the pipes those issue will go away.
The grey PB tubing generally does not have an O2 barrier. It is less of an issue when embedded in concrete.Running low temps will also help. If the 15-58 does not hold up, switch to stainless.

nickh1 · December 2016

Once again, thanks for the input. I'll work on some changes and post a sketch.

Gordy · December 2016

This becomes a good discussion for the OP as to why?

Sometimes pictures only paint part of the story.

nickh1 · December 2016

I'm not sure I follow what you mean on flow going back to the supply. It is pumping away from the boiler on the supply and I dont intend on reversing it.

As to why I plumbed it like I did, I dont have a good reason, other than I came out the top thinking having the air vent and blead valve high would help get rid of the air. I returned on the bottom just because I didnt think there was a real need to return to the top. The top and bottom fittings are connected by a copper tube inside that is teed in the middle to the heat exchanger.

Gordy · December 2016

The manuals can be confusing.

Usually the boiler manufactures want the circ pumping into the hx so the circs pressure differential is raising the psi in the hx preventing the possibility of the water in the hx flashing to steam.

Pumping away from the hx the circs differential would lower the pressure in the hx. This was more of an issue with higher head loss hx designs.

nickh1 · December 2016

I'm fairly certain it faces down, as was my intent. I did turn the motor so the cord came out the bottom however.

Gordy · December 2016

That explains it.......

nickh1 · December 2016

I know a lot of acronyms, but you got me with PONPC.

Gordy · December 2016

Point of no pressure change. (The expansion tank.)

Zman · December 2016

Point of no pressure change.
In a closed system, the expansion tank is the PONPC.

nickh1 · December 2016

That makes sense. Thanks.

Gordy · December 2016

Do you understand why the X tank is the PONPC?

Rich_49 · December 2016

Just to clarify . The UFT does not care how it is piped and what goes where so long as checks are located so flow is directed through the HX . I have one that is actually piped top supply and bottom return for CH and vice versa for DHW . That unit operates flawlessly .

Our Op does not need the second circ and I recommend , due to circulator selection that he make the modest changes to go direct piped . The circ probably be more than sufficient on Medium speed also , possibly even Low .

Did you also turn the cartridge on the system circ Nick ?

nickh1 · December 2016

Yes, I installed both pumps with the arrows facing the right way and turned the motors on each.

Just for curiosity. If I left it P/S is there a better sized circulator I should have used.

Mark Eatherton · December 2016

Paul48 said:
Everything being proper..... nothing can be more efficient than using the most btus produced by the boiler. Directly pumping the boiler allows for that. How is that even debatable?

Paul, in this case, he has full flow always occurring on the load side. Rare condition in the field. If each of the 3/4" tubes were a micro zone, then the boiler would short cycle if piped direct. This is what I call driving a tack with a sledge hammer. WHAM BAM thank you ma'am...

The flow through the boiler could be so low, and the delta so high, that it will continually bounce off of its ODR high limits. Having the boiler set up as P/S allows the boiler to "see" what the load looks like (a variable), because it knows what its flow rate (a constant) is, and can determine the best burner operation based on the true connected load. It will still cycle due to output exceeding connected load, but not as drastically as it would if piped P/S.

Also, you stated that 24 to 36 inch centers was a problem. How can you determine that? The rules of 12" OC maximum for residential space heating do not apply to commercial/industrial applications. If we did our heat losses and heat gains on a per square foot basis of the building, being nearer to the outside walls, the loads increase significantly. Just as they drop significantly near the interior.

With the conductive heat loss formula being A/R*delta T, with an R 30 ceiling, and a 70 degree differential, 1/30* 70 = 2.3 BTUs/square foot per hour. Infiltration would need to be aded to the conductive losses.

I am not 100% sure, but based on what I've read, this is either a storage barn or a possibly an equipment storage shed on a farm with heat only being delivered to the center core. People are not walking around in their bare feet doing the ooh ahh dance on warm/cold spots. They are conditioning this floor and the connected space. Also, from an overall perspective, the physical plant is only able to deliver around 20 btu's /sq foot per hour (72,000 divided by 3,500 sq. ft. conditioned space). Radiant floors, with constant human contact, are limited to 30 btu's/sq ft/hr (85 degree F surface temperature).

As it pertains to thermal efficiency, across the boiler, a large delta is not conducive to good net thermal efficiency. The lower the delta across the heat source within reason, the lower the average log temperature between the fire and the fluid, the better the thermal performance of the heat exchanger/burner assembly, which equates to less stack loss.

People have a tendency to get hung up on the CON of a MOD CON appliance, when in reality, much of its efficiency comes from the MODulation capabilities of the appliance to match its capacity to the load, eliminating excess air in the process, which is second only to short cycling for killing the thermal output of a given appliance. Large heat exchangers and small flames, combined with low entering water temperatures provide good thermal efficiency. I've witnessed this through the use of a combustion analyzer with the equipment under fire, intentionally varying flow rates through the appliance. Higher delta T equals more stack loss.

To the original poster, unless my eyes are playing tricks on me, I count 10 supplies and 9 returns. Am I missing something or did they split one of the loops out in the field?

Mechanically speaking, you do have the system set up for good and optimal pump/boiler/PONPC operation. If it were me, I'd cut a ball valve in between the P/S tweener tees. Close the valve, disable the load pump and you're in direct series mode. Open the valve and you're in P/S mode. See which appears to be the most efficient and run it that way. Might even need to partially close it.

Also, not being familiar with this particular product, you'd said there was an internal bypass? What controls it? You usually only have one P/S bypass unless you are trying to intentionally control the return water temperatures to keep an appliance (non condensing) return water temperature from being too low.

To @gordy, I find it hard to believe that the pump is experiencing nearly the same delta P on all three speeds. I guess anything is possible.

Try generating a "system performance curve" and plot it on the manufacturers pump performance curves and see where the chips fall. Take the known flow and its associated pressure differential, and double the flow rate. Plot a 3.3 times head increase at that point on the chart. Now halve the load (GPM wise) and plot the head pressure at 1/3 of the original.

Now using a French curve, draw the system performance curves for all three speeds.

Hope this helps all concerned...

ME

Gordy · December 2016

Yes Mark, 9 on the supply, and 10 on the return.

Those tridictator gauges are less than desirable for accuracy in performing a delta p circ test.

This boiler is not cycling, and modulation is mid %. I also see the primitive condensate drain. Might want to tend to that. How much condensate is the boiler producing?

Should I direct pump this boiler...radiant floor with one zone.

Comments

Welcome

Quick Links

Categories