Trying to work out the wrinkles on my new boiler

[babaji]

Hello Guys,

First post , new to this, it's my first boiler. It's heating the radiant heat floor in my 3100 sq-ft warehouse. Simple Primary loop through a 12 port manifold.

Boiler is a NTI TFT-110 , had a plan made, contractors stuck to it for the most part. (ended up with an extra loop, and one slightly long loop at 320 ft')

I am finding it to be very enjoyable observing it learning about how it works.

A few bugs to be worked out, flame is weak on lowest modulation 1625 rpm, and the exhaust is sloping outwards dumping condensate outside. These will be looked at soon.

I seem to have some popping , hissing, gurgling noise at start up. From all the articles, forums posts and videos I have seem, leads me to believe I have a problem with the circulator fighting the expansion tank.

The remedy, i suppose is to move the circulator so that it is located downstream from where the expansion tank piping meets the discal air separator. I am hoping it is as simple as that.

On the other hand, some say it's better to leave the pump at the inlet before the boiler in the cold stream and re-route the expansion tank piping to the inlet side of the circulator.

I would love to hear some of your thoughts and recommendations on this issue or any other things that could be improved.

Something else i have wondered about, could the Dirt Mag be too close to pump inlet, causing any sort of turbulence for the pump? Haven't found any info about that online. (maybe it's a non issue)

Thanks
Scott


PS EWT is about 67 -70 degrees F seems kinda low, read that may be symptomatic of weak flow due to the pump being on the wrong side of PONPC.

HERE IS THE PIC
https://drive.google.com/open?id=1LoJeUOrdr51JujR0Q0zsFhwUeqnfb-rj

 

[Dana]

Is the "... popping , hissing, gurgling noise at start up..." coming from the boiler, or somewhere else? Flash-boil on the heat exchangers due to low pressure or low flow will often be audible at the boiler.

What is the system pressurized to? Was the expansion tank's air charge pre-charged to the system pressure, or slightly higher? The minimum operating pressure of this series is 12 psi, measured at the outlet/supply side of the boiler.

Is this system being pumped-direct, or plumbed primary/secondary (as prescribed in Figure 10-5, p48 in the manual)? It doesn't appear to be. (More pictures might be useful here.) The picture of your system looks like Figure 10-1(c) Supply Top / Return Bottom on p.41, but one pump, no hydraulic separation. Unless the pump is specified properly and total head calculated, the minimum flow requirements of the boiler might not be met, and the max delta-T across the boiler may be exceeded. It's a bit easier to design it with a primary pump to set the boiler flow and a secondary pump to drive the radiation flow if there's much mis-match in flow requirements.

See Section 10.0 starting on p.38 where it reads:

The fire tube design of the Trinity Tft heat exchanger results in minimal head loss, however it must be considered when sizing system piping and circulators. Furthermore, the low mass of the Tft heat exchanger requires a minimum flow rate anytime the burner is operating. To maintain efficient and reliable operation of the heat exchanger, and to avoid heat exchanger failure, it is critical to ensure the rules and guidelines in this section are followed.

Circulators are never really "...fighting the expansion tank..." but if properly pre-charged and the expansion tank is near the intake side of the pump it reduces the chance of cavitation on the impellers. With the low pumping head of fire-tube boilers like the TFT series it doesn't affect the pressure at the boiler much whether the pump is driving toward the boiler from the return side or pumping away from the boiler on the supply side, but for water tube boiler toward the boiler is usually better. But whether the pump is on the supply side or the return side, the pump is better protected if the connection to the expansion tank is a fairly short and straight distance to the intake side of the pump.

Babaji desi hain, Scott na. Aap... Punjabi-Canadian?

 

[Sponsor]

 

[babaji]

Hello Dana,

To answer the most important question, Babaji, is a nick-name i acquired on a trip many moons ago.

I believe the system is pressurized to 15psi, i get this from the pressure gauge located to the right of the expansion tank. It is 15psi at rest and does not waver while the system is running.

Yes, the noise its coming from the boiler. The system is primary only, no secondary loop. The engineer designed it that way. Although spec'ed a BCI (sl14-115 G3) boiler, but wasn't too worried that the contractor wanted to use a NTI as long as it was similarly sized. The pump is an ALPHA2 15-55 exactly as specified by the engineer. It seems to have all kind of options, Right now I have it on constant pressure setting 3 (highest).

Maybe i have been reading up on this too much, and watching too many Caleffi videos, but I now have it it my head that my pump is located on the wrong side of my expansion tank. The outlet side of my pump is only about 1 boiler length and a little pipe away from the expansion tank tie in.

I found this article very helpful (or maybe it's got me in a lather about nothing?) https://www.industrialcontrolsonlin...how-avoid-problems-your-hydronic-system-pumps

I was wondering if perhaps I were to heed it's advice and have my pump's inlet just after the PONPC, would it be best to

a) move the pump physically to the hot side coming down the wall just after the caleffi discal unit (above and to the left of pressure tank mounting location). There is a long 4ft section of straight pipe that looks very suitable. In this scenario the dirtmag would be left by it's own, just before the boiler inlet more or less where it is now.

b) just reroute where the line from the expansion tank ties into the system. Right now it goes to the bottom of the discal air separator up top (which i understand is a good location for the air separator). In this scenario, i would remove the expansion tank piping from the bottom inlet on the separator and plug that inlet. Then aim that expansion tank piping for the section of vertical pipe just before the dirtmag and pump, just after the cold manifold valve. That way the expansion tank would be adding it's pressure very close to the inlet side of my pump.

Scott

 

[Dana]

Yes, the expansion tank is on the "wrong" side of the pump, but that's not a disaster unless the pump is cavitating constantly (which you would be able to hear if it was bad enough to affect flow), and it has nothing to do with the noise coming from the boiler.

That noise could be pointing to a real problem, and that's what's most important to sort out.

Different boilers have different minimum flow & pumping head specs, and often different max delta-Ts- pull the specs on both boilers to compare them. If the engineer approved the change, he should be on the hook for debugging it.

The pumping head of a fire-tube boiler is likely to be well below the pumping head of a radiant floor, so pumping away from the boiler isn't going to lower the pressure at the boiler enough to create a flash-boil problem, but the description of the symptom is consistent with flash-boil. I suspect the pumping head of the radiation is too high for the pump to deliver the minimum flow spec at the boiler when pumping direct, or that it's marginal.

What is the EWT to supply-out temperature difference when it's running? (You mentioned 67-70F EWT, which is fine if the supply out is running only ~80F, but a potential disaster if it's 140F out.)

 

[babaji]

Hi Dana,

I suspect the pumps ability to cope is borderline. Things have improved since i set it to 'Constant pressure mode 3' . It has quieted down a little, but will want to observe it for a while. I do have a recording of the noise i was hearing if you would find that helpful.

Yes I agree, i want to sort this out, i don't like that sound. I may talk to the engineer.

The delta T is usually about 45-50 f , with the EWR ranging from 67-71 f and the output water 109 f on up to about 131 f when it's running hard (which is about 5700-6500 rmp, out of the max of 7000 rpm) I suspect the machine is throttling the running speed so the differential never gets over 50f.

My thinking is if it is a borderline flow is 'marginal' then relocating to expansion tank's input to the intake side of the pump, will help rather than hinder and perhaps just give the pump the extra help it needs.

 

[Dana]

In your system moving the expansion tank is most likely make NO difference in the pumping rate (none!).

Do you hear the pump cavitating? They are QUITE noisy if cavitating and will self-destruct in a matter of weeks if it's to the point where it's affecting total flow. And it will happen constantly, not just under start-up conditions.

If you're going to move one or the other, move the tap the expansion tank to the intake side of the pump rather than moving the pump to the outlet side of the boiler. Keep the pump driving toward the boiler, maximizing the pressure at the boiler, which is probably keeping it from doing even more flash-boil. I'm highly skeptical that those changes will affect either the flash boil or the big delta-T issue. Go ahead and fix it (I would), but any improvement in flow from that change would be miniscule unless you've been hearing major cavitation noise whenever the pump is running.

This boiler apparently has a max delta-T lockout that may show up if you're on the margin (and it sounds as if you are). See the bottom of page 67 of the manual under" Lockout or Hold 81 – Delta T limit OR Appliance making banging or hissing sounds " near the bottom of the left column.

Table 10.2 in the manual specifies a minimum of 4 gpm for the TFT-110. In Table 10.4 they're specifying Grundfos UPS15-58 for the primary pump, even if plumbed primary/secondary:

At max-speed that pump is capable of driving 4 gpm through 15'-16' of head.

That's the most the Alpha2 can muster too:

At the 5' of head indicated in your engineer's tables either of these pumps should be able do deliver way more than 10 gpm, plenty of margin than the 8 gpm indicated, and more than twice the boiler's minimum requirement.

You should be seeing MUCH lower delta-Ts. A flow of 4 gpm at a delta-T of 50F is 100,000 BTU/hr, pretty much the entire high-fire output of that beast. Your boiler flow isn't just marginal, it's probably substantially below spec. The question is "why"? It's not because the pump is cavitating (or would sound like it's full of marbles or rocks), and not because the pumping head of the boiler is so high (it's negligible at 4gpm). There may have been a math error on calculating the pumping head of the radiation, or a plumbing implementation/restriction error inducing much higher head.

 

[babaji]

Thanks for taking the time to give your input on this. It is immeasurably helpful to talk this out with you. I may have to take this up with the HVAC guys or the engineer and knowing you agree with me that there is an issue that needs looking at will help strengthen my resolve.

No the pump is not cavitating, it's nice and quiet and you have to get fairly close to hear it at all.

I think you may not have fully grasped the implications of having your expansion tank so close to the discharge side of the pump. It's not only that it can cause cavitation in pumps, it can (and I think this is what I am dealing with here) rob your pump of flow.

It is possible i am confused and do not totally grasp this, it took me a while to 'get it' but i did finally had an aha moment while watching this video.

(start watching at 7 mins and 06 seconds)

I have tried using the pump recommended by NTI the grundfos ups 15-58, I switched to the alpha2 15-55f as originally spec'ed by the engineer, it did not fix the problem, but it consumes half the power , quieter and has some interesting modes and don't regret it at all.

I agree with you that leaving the pump at the boiler intake would be better, that way it can make sure your boiler gets 'it's all'.

I totally agree with you that as you said "Your boiler flow isn't just marginal, it's probably substantially below spec." I hope I can solve this.

The first article I linked to mentions putting a pressure gauge before the pump intake and one again after the pumps output side. Where I can get the pressure differential , multiply it by 2.3 to get the head (feet) and then look up on the grundfos pump curve to see what that translate to in GPM flow. Then at least I would know if the pump is not putting out enough.

 

[Dana]

Whether the tank is on the intake or output side of the pump, the pressure difference across the pump (and thus the induced flow) is the same, even though pressures on the system relative to the atmospheric pressures all change. In the video's first example with the tank positioned correctly, the pump is driving from 9 psi at the intake to 18 psi at the outlet, a delta-P of 9 psi. In the "wrong" example it's driving from -3 psi at the intake, 6 psi at the outlet, a delta-P of 9 psi. The fact that the system's pressures at different points on the loop change relative to atmospheric pressure has no effect on flow (until/unless the pump is cavitating.)

Correcting it so that there is now a higher pressure at the boiler relative to atmospheric pressure is likely to reduce (but not necessarily eliminate) the clang-bang-sizzle but it isn't going to change the flow or the delta-T across the boiler appreciably.

If the pump were cavitating there could be a significant change in flow by fixing the expansion tank tap placement but it isn't cavitating, so expect the delta-Ts to stay pretty much the same. Even if you're lucky enough that the change fully suppresses the flash boil, the delta-T at the boiler isn't going to change.

We absolutely know that the pump is not putting out enough, even if we don't know why. If it were delivering the specified design flow of ~8 gpm (7.83 gpm in the chart) it would be impossible for a boiler with ~100K of output to have a delta-T much greater than ~25F at the boiler, whereas the measured delta-T is about twice that, indicating flow of 4 gpm or lower.

There's is apparently enough static pressure on the system to avoid cavitation issues despite the higher than spec head, but not enough to suppress the flash boil from having the boiler at barely above the static pressure. But since the static pressure is within spec at the boiler, the fact that there's any flash boil at all is an indication that the flow is too low.

The manual's recommendation for the 15-58 primary presumed that it was a primary pump in a primary/secondary configuration. A typical primary loop's pumping head in a low-head fire tube boiler is much lower than typical radiation's pumping head. Since it's being pumped direct rather than primary/secondary the pump spec has to account for the full pumping head of the radiation too, so it's surprising that it might need more pump. If it were only driving the loop between the boiler and hydraulic separator the pumping head of the radiation & pressure differences induced by the secondary pump would have only a minor effect on boiler flow.

 

[babaji]

Thanks again for your response Dana.

Ok using the example of a pump that should be a 9psi boost on the output side, but is actually only putting out 6psi and -3 on the intake does this create a situation where water will boil at a lower temperature ie flash boil in the boiler? Perhaps if it was getting all the benefit of the expected pressure it might not make such a commotion?

Is the flow check valve that came with the grunfos pump really needed in a single pump system like this with all the water in the floor below? They say it may cause a 10% handicap on the flow, worth removing that check valve?

Pump has a digital display of GPM, (yes i know not all that accurate). When the pump starts in Constant pressure mode I (6ft), the gpm starts at about 6gpm and climbs slowly to 8 or 9 gpm. Is this normal, is it just building momentum, does the glycol mixture present more resistance on start?

Scott

 

[Dana]

Thanks again for your response Dana.

Ok using the example of a pump that should be a 9psi boost on the output side, but is actually only putting out 6psi and -3 on the intake does this create a situation where water will boil at a lower temperature ie flash boil in the boiler? Perhaps if it was getting all the benefit of the expected pressure it might not make such a commotion?

Yes- the boost in pressure will knock back the flash boil potential, but only slightly. In the current configuration the pressure at the boiler is still measuring at or above the specified pressure in the boiler even when the pump is running, so it should not be experiencing flash boil if the flow rate is fully up to spec.

Is the flow check valve that came with the grunfos pump really needed in a single pump system like this with all the water in the floor below? They say it may cause a 10% handicap on the flow, worth removing that check valve?

Yes- totally worth it! Lose that check valve, given that it's a one-pump, one zone system.

Pump has a digital display of GPM, (yes i know not all that accurate). When the pump starts in Constant pressure mode I (6ft), the gpm starts at about 6gpm and climbs slowly to 8 or 9 gpm. Is this normal, is it just building momentum, does the glycol mixture present more resistance on start?

Scott

ECM drive pumps do have a slower spin-up ramp than their low efficiency cousins, which may be why it's only getting the flash boil at the beginning. Glycol does affect the pumping head and heat transfer efficiency, but SFAIK it doesn't affect flash boil.

If the pump's display is showing 8-9 gpm it should really is doing at least 6-7 gpm, which would be good, but that doesn't quite explain the huge delta-T. I don't know the details of how they're sensing/measuring the flow with that pump. Even with the reduced heat transfer efficiency and lower specific heat of the glycol mix it seems unlikely that if it's really delivering the design ~8 gpm you'd be seeing 45-50F delta-Ts on the boiler.

 

[babaji]

Great I will remove the check valve soon.

The other pump UPC 15-58 also had this problem grugling / hissing / pingiing plunking noise on startup).

I may be crazy but i am wondering if it is getting better with time. Or could it also be weather dependant, (temperature / load dependant)

Could the problem be related to air in the system? That is starting to work it's way out?

Scott

 

[Dana]

Air does aggravate the condition, but usually isn't the primary factor. (Attributing the gurgle to air, yes but to hissing, pinging, plunking, no.)

Given the high delta-T I'm not surprised that it still exhibited a flash boil problem when running with the 15-58.

Raising the static system pressure a few psi can also help suppress flash-boil, but do the check valve flow restriction fix and expansion tank fix first.

 

[babaji]

Hey Dana,

Did the check valve first. Flow definitely increased a little according to the digital display. It did not however 'fix' the noisy startup.

Yesterday my neighbor came over with his torches and assorted Tees and elbows, and showed me how to solder while I watched him create a temporary rerouting of the expansion tank.

here is what the band-aid fix looks like: https://drive.google.com/open?id=1JJ37thwNC8kr7l6VTOBQon9eQ5MBxopa

I will still have to plug the air separator at the bottom, and remove all the left over piping. Also I will need to reroute the glycol injector's path in. But that can now be done lower down just above the manifold. I can get a shorter hose and really tidy this up. Maybe also put a bypass valve in somewhere down low to make bleeding / filling easy.

So what's changed. Well first thing i noticed, and it was so obvious i am kicking myself for not realizing this earlier!..... wait for it...... The pressure gauge on the hot water pipe coming down the wall, now shows a pressure increase when you run the circulator pump. Hallelujah! Funny, it always read 15 psi before, with or without the circulator pump running. (that should have been my first clue) Now it reads 19 - 20 psi when the circulator pump runs. A gain of 4-5 psi over the static pressure, and that reading is being made after running through the boiler. ( i assume it would be much higher if it were read straight away after the pump)

As for noise, it did not fix all the moaning and groaning straight off, but I did notice right away that it ran much quieter overall (less boomy, like it wasn't so stressed or it could be my imagination and or placebo effect). We did a real botch job of filling / bleeding and had to mop the floor a little! I have been running it for about 24 hours, and think I must have the lions share of the air / entrained air out by now?

Did not eliminate 100% of the tink'ing noises, it seems to make a much softer 1 or 1.5 tinks when it initially puts the heat to it at the start of a burn, but it's much better now, and no longer does a 10 second dance at the start with a series of plink , plonk, plunks. I am best to just leave it for a while and observe, it may get better with time. It's going down to below zero f. tonight, hopefully it'll get a workout, heat up some and puff out some air...

Scott

 

[Jeremy7835]

Why isn’t it piped primary secondary?

 

[Jeremy7835]

Why isn’t it piped primary secondary?

 

[babaji]

That's the plan the engineer came up with, primary only.

 

[Jeremy7835]

That is a pretty stupid plan! Can we see the plan?

 

[babaji]

Why are you calling it a stupid plan?

 

[Dana]

Why are you calling it a stupid plan?

Just guessing here (can't read his mind), but probably because some the issues you're seeing here are foreseeable. Other designers might have specified more pump for a system configured to be pumped direct rather than simple a pump that filled the bill for the steady-state pumping head. (The term "primary only" isn't a term commonly used anywhere I'm aware of- "pumped direct" is more generally understood, but there may be other regional terms for simple one loop system.)

When pumped direct the raw inertia of getting all the water mass in the system flowing can mean a few seconds where the flow rate through is well below spec, even if the pump is fully capable of delivering the steady-state flow. With just mass in a short primary loop to get moving the acceleration ramp is much shorter, and a smaller pump can get there without flash boil.

There are other advantages and a few disadvantages (such as lower supply temp to the radiation, higher entering water temp at the boiler) for going primary/secondary, but it's a generally more robust approach, easier to tweak delta-Ts, with fewer "gotcha" problems. In a very low temp slab situation the combustion efficiency hit from going primary/secondary is negligible a tiny fraction of a percent, but the electricity use can often be higher with two pumps than with just one perfectly sized pump for pumping direct.

On systems with lot less water mass in it than yours pumping direct makes a lot of sense when using low head fire tube boilers, but it's inherently risky with high head water tube boilers. Some high mass systems a competent designer wouldn't even think about pumping direct, but your system may be in a bit of a gray zone.

 

[babaji]

Dana

I think the issues for the most part have been resolved in moving the inlet of the expansion tank so that the circulator pumps away from the expansion tank. I had last reported 24 hours after making the modification, that it was better. At 48 hours after the piping change, virtually all symptoms I had reported were gone.

I think this NTI model no. 110 TFT is a low head fire tube design.

I think the engineer was spot on with his calculations. I am getting what appears to be very efficient operation with the pump in it's lowest setting CP I (constant pressure setting of 6 ft head with a resultant flow of 8-9 GPM ). On the coldest days it holds the delta T to a max of 55 f. with an EWT of about 72 f. If I turn up the pump to the next setting CP II (constant pressure setting of 9 ft head with a resultant GPM flow of 10-11 GPM flow) the boiler maintains the inlet and outlet temperature about the same by firing the boiler significantly harder (I assume in order to maintain the high Delta T of 50, due to the excess cooling resulting from running the circulator on CP setting II). Not only is less fuel used, the pump only uses about 27-30 w on CP setting one, vs 43 w on CP setting two)

The boiler has a very (at least it appears to me) sophisticated reporting system. I have it set to plot firing rate, inlet, outlet, and outdoor temperatures. The resulting graphs from the collected data are extremely configurable and the X-AXIS (time) can be viewed by the second, minute, hour or day. While the left hand side Y-AXIS (temperature) has adjustable MIN and MAX values so you can set the graph to a range of say (-10f to 130f) or adjust for whatever the conditions are. The Right hand Y-AXIS shows RPM which can also be adjusted, but was set to to the Boiler's default of 1,625 to 7,000 rpm. I will [post a sample graph later showing a comparison two burns, one at CP I pump speed and the other at CP II. It made it very apparent to me that's raising the pump speed just burned more propane to achieve the same results.

I am astounded by this feature and I am not sure how common such reporting is, but it seems like a very useful tool.