Heating Help

Simple System, Impossible Problem.

This one is driving all of us nuts.

This is a simple system: no primary/secondary. Pump from Munchkin, into buffer tank, past a bypass valve, into radiant manifold, back to munchkin.

Problem is, we are seeing at least a 15 degree temperature drop from boiler across the buffer tank, even when the system has been running at continuous temperature for over an hour. Flow rates SHOULD be more than enough to "turn over" the whole 30 gallon tank in no more than 10 minutes (really, at least 4 GPM) at which point temp out should be very similar to temp in.

We're at a loss:

1. Flow rate is confirmed by visual loop flow meters on the manifold. All my good explanations involve insufficient flow, yet the loops show proper flow.

2. Temperature readings are redundant in both locations , radiant manifold supply/return and boiler output. we have switched thermometers and checked and temps are right. Temperature in and out of buffer tank are noticeably different.

3. At this delta-T with any flow rate like we expect, this would mean we are basically losing 20kBTUs/hr or MORE through the buffer tank.... but that would mean the area would be very warm. At the very least, the tank would be really hot! and it's not.

4. Fill valve is closed, so cold water intrusion/leaks are not a possible solution.

5. Closing the bypass valve results in negligable changes, so "back mixing" through the bypass, while physically impossible, has also been ruled out.

6. Also we have run the bypass turned all the way down which should give us over 10 GPM, and closed the radiant manifold, to ensure all flow is just into the tank and out again... and while we do see creep upward in tank temp, it's far slower than expected.

7. Also we have run the bypass turned all the way down which should give us over 10 GPM... and while we do see creep upward in tank temp, it's far slower than expected.



I have no idea what is going on. it really appears we have a huge temperature drop across the buffer tank but no explanation for why, or where the heat is going. Again all my explanations would be low-flow related, except we can see our flow is proper.

any thoughts would be appreciated. this system is way too simple to be this much of a head scratcher!

Hey Rob,

Are you saying that if the water leaving the boiler is 90*, and you are not heating the home at all, after an hour of continous circ from the boiler through the tank, the temp will be approx. 75*  at the tank and does not get any higher?

I am missing something I know.

If I were to pipe this I would have piped the boiler in and out one side of the buffer tank and piped the system in and out the other with its own circulator. I do not do many radiant systems and none with buffer tanks so far.

Munchkin output?

Is the Munchkin capable of rated output? Have you clocked the gas meter to determine actual firing rate?

How well are you heating the 80 gal Superstor?

Low flow would be my first guess, but since you have checked that so the only thing I can think of is not enough BTU input to raise the temperature of 40 gallons of water. If you are bypassing the rest of the system and still cannot raise the temp of the tank rapidly, then it just seems you don't have enough BTU's going in.

If it were me;

I'd suspect the flow is diminished (clogged impellor?) and the boiler is not ramping up to full bore. Hopefully there is not another load (like Domestic Hot Water) with a leak somewhere.

Rob

Clarity Needed.  Is the temp drop from the boiler to the tank or from the buffer tank inlet to the buffer tank outlet?
To rule it all out, I would check the DHW system.  Even though circ has a flow check (?) and flow is isolated, could a pin-hole in the indirect coil let enough cold in to affect the temperature drop you describe?  Seems like an increase in pressure would then be apparent.  Can you close the ballvalves to isolate the buffer tank and see if you have any increase in pressure?
Paul

 

thanks guys

responses so far:

-Water is leaving boiler at 115 or so but is about 100 when it gets to the radiant... for a LOOOONG time, much longer than just lag because of the buffer tank *should* take.

-No need for primary/secondary on this one, thus the one pump and only 2 taps used on the tank.

-Firing rate of the munchkin is low, and it IS maintaining its calculating outgoing target, AT THE BOILER. it's this mysterious temperature drop across the buffer tank that is kind of weird. It just never goes away....

-The temperature drop does appear to be across the buffer tank. Temperature across the DHW tees doesn't appear to change significantly.


Has anyone else seen any *really* extended lag times across buffer tanks?

Or, has anyone seen loop flow meters that seemed to be reading fine but that were really, really inaccurate for some reason?

Problem is less when we ONLY go through the bypass turned way down... so it SEEMS a lot like a flow issue. but those meters are all reading well and indicate we should have at least 4 to 6 GPM at all times. Which SHOULD turn this tank over in 5 to 10 minutes. Yet here we are, an hour into a heat demand and still a 15 degree drop across the boiler.

if we really had the GPM, then we're losing something stupid like 30kBTUs/hr from the tank! But of course that's crazy.

I've never seen the laws of physics broken quite like this.

Rob,

Could you put a guage, either strap-on or digital, to the pipe leaving your buffer? I would be interested to know if the temp leaving the tank is 15* less then the boiler temp, or is the temp dropping over the length of the piping from the buffer to the distribution point. Also, what size is the tubing?

Trying to se if the loss is within the buffer, or within the supply tubing. As for low flow, maybe, but with 80K heating a 30 gal tank, it should be able to get the temp up even with a less then adequate flow.

Rob, all things being equal

, the only thing that could be dumping cool water into the system by the supply distribution branch would be if your pressure by-pass is mounted in reverse.....

what about the SuperStor

does it have check protection? Could you have un-wanted flow through it?

I could

but even then, it couldn't possibly take that long to heat up and cease to be an issue, you know?

I'm actually starting to wonder if the flow meters are bad. It's really, really improbable, but every single issue we've seen here points to low flow, and the only reason we don't think that's the case is that they are reading good flow.

anybody had bad flow meters recently?

my 2 cents

Curious, why you did not use the other tappings for system side?
I know it would have meant another pump.
Also it seems like the supply from boiler is top tapping and bottom return from tank is supply to manifolds.
My thoughts are, stratisfied hotter water is not leaving the the tank, and no matter how much flow thru tank you will always get the cooler temperature from tank supplying manifold.
Maybe boost tank temp, and add a three way valve for fixed temp before manifold..

I wondered about that too

we only used two taps because we most definitely only need one pump on this system and we do not use extra pumps... we're fanatics for low energy pumping. I'm pumping houses like this all the time, so I'm pretty sure it's not a pump sizing problem, though a bad pump maybe.

I do admit the top in bottom out method may have some drawbacks, but it has some benefit too... I'm pretty sure this tank doesn't "mix up" real fast when we're loading/unloading it like a hot in the bottom method might.

I thought about the "not overcoming stratification" solution... and worried about it... but the fact is, if it doesn't heat up eventually, it's because you're losing heat, and if my flow rates and temp drops are right, it would have to be losing heat 100x faster than this tank ever should. Otherwise, eventually it would just have to heat up, the energy can't just disappear.

? heat rises

Stratification, is it possible the BTUs are just staying at the top of the tank?
Can't imagine it if the whole house is calling, ,,,full flow.
Could you mount temp gauges in those other taps, top and bottom, and see the temp layering right at the tank?
pardon the simplicity here, I am NOT a pro.
neil

I hate spending other peoples money but

If you do not need the primary secondary can you by pass the buffer tank? I am sorry but its function is not getting through to me asside from being more water to try and keep warm. Then again if it quacks like a duck it must be a duck. Swap the pump as it is the cheapest and fastest fix to try.

Piping

Did I understand right the system goes back to the boiler. Don't you need to circulate from the boiler into the tank and back to the boiler then from the tank to the system then back to the tank.If I understand your piping scheme Your system is always taking heat out.

WP

This set up is used to help with boiler cycling issues when there are smal loads in the system.

Pictures?

Do you have any pictures to post of the job?

Simply / Impossible:

Complicated systems create complicated problems with simple solutions. If and when you find the problem.
Something installed is not as the drawing shows.
You and all know far more about this stuff than I and I would never do it like you do. You claim success. IMO, you have not piped the radiant system as Primary/Secondary. The "buffer tank" is just a place to increase water volume in the radiant loop. My pea brain sees a "Buffer Tank" as a Hydraulic Separator but it isn't because you have a thing about using "extra circulators" and wasting electricity. I think your problem is what you see in the tank and what you think you see.
I see people try to use water heater tanks on tankless water heaters, Some use something like this. The hot water goes in and gets stored. The hot goes out. There's no cold saved. I do it differently. Looking at how you drew this, the tank becomes a hydraulic separator.
Are you sure that the circulator isn't installed on the wrong place of your drawing? You are running 110' supply water temperature and you see a 15' drop? So, you are only sending 95' water in to the system after the buffer tank? If you raise the radiant system to 130', does it still have a 15' drop into the system after the buffer tank? 115' supply water?
Someone mentioned the Diff. By-pass valve.
Far too complicated and costly.

Top in bottom out?

Rob, what is the reason for the top in, bottom out connection of the buffer tank? To me, it would seem to encourage constant stratification, so you would always have cooler water at the bottom. I would think that you would have better convection mixing in the tank by piping it bottom in, top out?

What do you see as the advantage with the current setup?

The expected advantage would be longer boiler on cycles

Since the thermostats would be less likely to be satisfied prior to the tank coming fully up to temperature. So, in other words, maximizing buffer capacity.

stratification

I had a problem with a DHW solar storage tank. It was a piping issue similar to what a few others have mentioned in this thread. When I switched the piping to cold in the bottom and hot drawing from the top, my problems were solved.

See, the problem is...

If you don't know what the starting temperature of the water in the tank was, then you don't really know how fast you should expect the outlet temperature to rise. If, for instance, there are no flow checks, then a DHW call could draw from the buffer tank through the bypass valve (or even through the radiation) and during a relatively short draw this could result in the water in that buffer being a lot colder than those 75 F you see coming out. If your flow rate through the tank results in a short circuit through the tank, relatively speaking, then those 20000 BTU/hr you're missing could be heating the "backwater" portion of the tank. If you were truly losing those BTUs, the outlet temperature wouldn't even rise after a long time. At 20000 BTU/hr heat input, a 30 gal tank would be expected to rise at the rate of 1.3 F per minute. To compound the issue, you could even be back-drawing through the indirect during a heat call, so the boiler supply could be tempered down with cold water from the bottom of the indirect coil, and at the same time the flow through the boiler would be less than the flow through the buffer tank/system, so you could be overestimating the 20000 BTU/hr.

Parallel paths...

The DHW tank is in parallel to the heating circuit. It could be forward drawing water through its branch. The pump could also be installed backwards on the DHW circuit, which would allow the main pump to draw water through it. Easy enough to test. Close an isolation valve on the DHW circuit while you are charging the tank/space heating system. If recovery of the buffer straightens up, then you've found your problem.

At the tank, by using the two tappings from the same side you are not utilizing the full mass potential of the tank, and it could be absorbing heat until it becomes completely saturated. Go in the bottom, right hand and come out the top left hand so that you flow diagonally through the tank.

Other than air/dirt separation, I see no distinct thermal advantage to flowing thru the buffer backwards... What was your thinking there?

Also, aren't the two pumps supposed to be facing each other on the inlet to the boiler with check valves to avoid unwanted back flows? Water does things in the field that we can't envision in our minds eyes..

I'd move the PB pump to the boiler return, just to the lower left of the two disconnect switches, then move the PAB to be connected down stream of the AS, and tied to the inlet of the PB pump. This way, you maintain flow across the boiler, and the two pumps (checks assumed) face each other per the manufacturers recommendations.

As for the flow meters, you can check them by closing off flow, and if they drop to zero, then pick back up with flow, then they are working as well as can be expected.

ME

Thanks again all.

this is a very competent professional install using a control panel we built and verified here in our shop. So I'm 99% sure the install is pretty much exactly as we drew it. I've also verified key details over the phone with the installer on site.

We could bypass the buffer tank but that does require some repiping and we do want the buffer tank there to reduce boiler cycles during low load situations: this is low mass radiant with some small zones. Cycling control will improve boiler lifespan and longevity and that's why it's there.

We have played with piping tanks like this because you really have to heat the whole tank before you satisfy a thermostat, unless you have a lot of start/stop cycles that allow stratification to form. We don't have a lot of start/stop though, since we use aggressive reset curves, and so with several zones on a reset system, there should almost always be flow through the system. You could make the argument though that it shouldn't matter at all which way it is piped in that case as long as you aren't "straight through" the tank, and you'd probably be right. But at the end of the day we do want the tank to saturate fully before it satisfies a load.

Thanks for all the ideas. I'll post what we find as we continue on!

used simalar set up

Rob i have used a similar set up for a panel rad system but i used all for tapping's on the boiler buddy and used a wilo circ for the system .i installed the system sensor in the boiler buddy .This was using a peerless purefire my system pump runs constant and the boiler pump cycles only on a drop in buffer tank temp.It is hard for me to image that the boiler sees the proper flow after pumping through the tank and radiant loops.I think i would have mounted the boiler pump pumping into the boiler to ensure proper flow and as it is  shown in most of all boiler with that type of heat exchanger.I have seen other pump out of the boiler like in your diagram and there where issues with ghost flow  but the rest of the piping was also not up to snuff and those who where smarter than i chose to flow check on the returns (no flow check pumps or checks on supply either) this did not fix anything and i stopped giving any advice and they stopped returning to the job also.I still look at your piping diagram and think that the boiler needs its own pump  with check valves and the radiant loop side needs it's own pump with a system sensor installed in the buffer  and set with a good differential to lessen any short cycling  from mirco zoning .at least this is what i would do for sure if i was called in to get it working properly .Peace and good luck clammy

I do appreciate the feedback

But we do primary pumping on at least half of our projects. the concept works extremely well when you know your flow rates and head losses. In this case, the *design* flow and frictional loss through the loops and boiler are well within the capacity of this pump. I may be missing something, but it's not that.

and it wouldn't answer the question of why my flow meters say I have adequate flow already.

A real photo is needed

or a dozen. or a youtube tour of the piping. something is amiss and if this crew is scratching its head then either theres fleas or the piece of the puzzle is being left out.

System Impossible:

As I learned, the purpose of Primary/Secondary piping/pumping was to allow different flow requirements to work together.
I see a boiler that likes and needs high flow. I see a system that has a lot of restriction and potential low flow. I understood that a "buffer tank" is really just a hydraulic separator. I would think or I would have piped a circulator from the boiler through the buffer tank and piped the system through the buffer tank on the other side with a circulator. Most guys that I see driving on the road will blow me away at a stop light. While excessively obsessing on the amount of BTU's wasted by a 2 degree rise in temperature and the electrical cost of an additional circulator, have no concern over matting their DuraMax Diesel to get ahead of someone they don't want to be behind. Me? I just laugh at them while I still get 15 MPG in my 2000 E-150 Ford Van. With no excess tools in the truck. 
In my experience, tanks will stratify at a moments notice. Especially high recovery heaters like Bock oil fired water heaters. I found the absolute solution was to run a pumped return return circulation line into the bottom of the tank through the drain. I just don't see how piping it backwards will help.
IMO, if YOU personally, can't put your little round eyeball upon it and your hot little hand upon it, you are depending on someone else to DX the problem. Just because you or someone else drew this up and someone installed it makes it right. If this worked before (as you claim) and this application doesn't work, what changed? Or, what isn't installed right.
I've been in positions of having to fix some incredibly messed up systems. I never would have found a solution like this one by piping it as you did.
I also find that that Munchkin piping diagram with the DHW tank on P-1, and the system pump on P-2 or whichever is a drawing waiting for bad things to happen.

Seriously

"As I say", we pipe primary only routinely. It works, if you do the numbers and the boiler's flow requirements are in range for your heating flow requirements. The stock piping diagrams are so people who do not do the numbers do not have to call HTP with their boiler lockout problems. The price is an extra pump. HTP doesn't care about that.

I do, since I know it's close to $10/mo in electricity for our nearly constant circulation systems to run that pump, plus the original labor, for no benefit at all unless it's needed. I'd prefer to keep that money in my clients' pocket. Also, I drive a subaru outback, or a mercury villager.

Really guys, the "it's not primary/secondary" thing is barking up the wrong tree. Maybe we could pipe the buffer tank in a different direction, I'm not confident that piping it upside down is a huge improvement over hot in the bottom, but I'm pretty confident that wouldn't settle this issue.... if that were the problem the basement would be 90 degrees as we "lost" 30kBTU/hr from the tank to the basement, and that's just not happening, so obviously we're not *really* losing that heat.

Obviously something is not right here. But adding a pump would not change the fact that all the issues seem to point to a low flow problem, except the flow meters, which indicate excellent flow. So any solution has to explain why our flow meters say we have over 4 GPM going and yet we're at minimum fire with a 20+ DT across the boiler and a 15+ dt across the tank, sustained for long periods of time, with just one pump.

Contractor will be back on site tomorrow. we'll try some of the ideas posted here, and test the flow in a few conditions, and we'll post what we come up with.

I can absolutely guarantee, however, that the answer will not be to add a second pump.

Re: buffer and temp drops

Rob, are you piping hot in off boiler in the top and out the bottom? I was wondering if you did this and opposite with system piping if that would give you more accurate system temps and less stratification?  Just wunderin. Tim

install manual

from the install manual from munchkin they show pumping into the boiler and primary secondary  on all the installs

boiler plate

i am president of the aacc and we have an old fitzgibbons boiler ,it is oil fired.
because we are opened to the public we have to have it inspected yearly. we had it inspected and it does not pass cause the boiler plate that describes what it is is missing. how can i get one?   

Late to the party but

may I chime in?
I disagree that the buffer tank is a hydraulic separator in this case. If you had two circuits drawing through it, then yes. But being in-series and being an incompressible fluid, what goes in, comes out. Nothing lingers there.

I do not see this as a flow issue because if you had a degraded impeller, you would have less flow. If less flow than you would have a wider, not narrower delta T (if I grasped the problem correctly). In other words, if this were an OVER flow condition, throttling would be the test protocol. So the suspicion is an underflow condition? I just cannot see that.

Instead, I suspect a T-mix happening and we have to find it.

Rob, maybe I missed it, but where are you measuring the temperatures? Is it "in" at the tank and "out" at the tank? Or are you measuring all at the boiler?

Reflecting on what ME said, about the possibility of the DWH circuit influencing that, mmmm.

How I would approach this: Start measuring temperatures at each tee or juncture and map them on your (very nice by the way!) diagram. Sort of like hunting for pressure drop, find where the big drop happens.  Find where the big drop occurs (and it could be in a very discrete place).

Absent other information, that is what I see.

we are measuring

at the boiler outlet with a piped in thermometer, at boiler inlet and outlet with the boiler's onboard sensors, and at the radiant supply and return manfild with piped in thermometers... that we can pop out and switch around, and they both read the same in both locations.

a major drop is definitely occurring across the buffer tank though, hands on the pipes determined that and the dial thermometers are located exactly as they are on the drawing (as I said, we built the panel ourselves).

I See

the temperature gauges you designed, at:

1) The Boiler Outlet
2) Manifold Inlet
3) Manifold Outlet

Understood re: on-board boiler sensors and knowing you, you have checked those vs. Ohms and other instruments. Good.

Given this,
a) I do not know what the actual tank inlet temperature is (possible influence by the DHW circuit it passed by on the way- or at least rule that out.)

b) I do not know what the actual tank leaving temperature is (but it may well be the same as the manifold supply line, less any minor line losses or reverse flow/influence at the bypass valve). At least rule those factors out. 

So, are there any positive temperature readings at the tank itself? The hand-feel thing is fine, fairly subjective but qualitative. An actual number would help though.

Any infrared shots of the system? Thinking tank surface losses, insulation voids? Although for your conditions, I would think "no tank insulation at all"!

sadly

no other positive readings possible at this time. and you overestimate my rigor with sensor confirmation.. when they agree with hardware, I just trust 'em!

We are going to absolutely double extra plus make sure nothing is migrating from domestic tank or backwards through the bypass by closing off the valving.

Tank is a standard 1 to 1.5" of insulation, I forget which. It's a Boiler Buddy, and nothing seemed unusual from a jacket temperature perspective.

Thanks for the brain cycles Brad!

So Rob,

 have you phsyically seen the install? Your guys may have built the panel, but the installers may have the set-up different to what you drew. I agree with some of the others, we need actual photo's of the install.

This system is way to simple to have the energy being transported to Alpha Centauri within the buffer tank!  :)

you know

I have not yet ruled out the "wormhole theory"!

we built this for the contractor with the boiler on the board. so the only thing "off board" is the dumb pipes to the buffer tank and indirect, which I walked through with the installer, and this installer I hold in the highest possible regard as well. so pumps and wiring and all that stuff we did ourselves. which is not to say that we couldn't have messed it up, we're human and mistakes have been made before.

but I have not yet seen it in person. it's only about an hour from me so it may come to that if the installer and I can't isolate the issue tomorrow.

Well good luck

hopefully it will be something simple that a different pair of eyes will see.

Reread Dans first rule of trouble shooting last night

We are human after all. It may be a bad thermostat now I think about it. Maybe it is working as it should and the thermostat is making someone think it is not. I still respectfully disagree with the piping of the tank. I would consider that consumer very lucky to have you on the case though as I know it will get fixed. Last word from me is this Infrared thermostat confirmation.

thanks

for the kind words and the thoughts guys.

update will come as soon as I have one to give. enjoy your weekends!

THe OOHHhhh AHHHhhh test...

Eyes and hands on test during recovery will find it every time. It MUST be done while the whole system is coming out of a DEEP recovery, or you will burn your hand (DUHHhh).

I call it "Feeling Hydronics",

Warning: Hot metal pipes may be hotter than they appear in your hand held IR Home Depot thermometer. With laser :-)

Handle with EXTREME caution...

ME

SS, IP:

Rob,
I installed a Munchkin a couple of years ago. My son needed some work so I let him pipe it up because I was busy. He had been working for a HVAC guy doing all his hydronic boiler installs. My install was exactly like you have except yours is radiant and I had three zones of baseboard with zone valves. I drew up a diagram of how it had to be done. There was no question about how it had to be done. There was no other way. He understood what to do and how to do it.
The owner mentioned that the water was very hot at times. I looked at it and found a tee on the wrong side of another tee. When the heat zone was running, it was sending hot water into the SuperStor. Sometimes, the only way to find a problem is to cast your eye upon it. It took me a while to find the tee. It took me even longer to take it apart and re-pipe it.

unknown problem

I do not see where you have a pump listed but I would suspect based on what I read that you need a bigger pump for the application. Are the manifolds opened up all the way? Is the PDBV working or set correctly, try closing the isolation valve forcing all flow through the loops? Have you ruled out any restrictions within the boiler coils?

eluv8

if you knew me at all you would know that I consider the phrase "bigger pump" to be about the worst insult you could throw at me ;)

it's a 15-58 grundfos. medium speed is all we need, though it's on high at this point to try to jack up performance.

more soon!

Details, details, details....

I down loaded the beautiful drawing that Rob posted, and was able to zoom in on it for better detail and drawing clarity. Try it yourself and see what you are missing.

THis is about as straight forward and simple an installation as it can be. Does it go against the manufacturers recommendations. Maybe, but knowing that Rob knows what he is doing, and has done this before, successfully, I don't worry about the design too much.

The tank is nothing more than a fat spot in the line. And one might think that by running fluid backwards through the tank goes against conventional wisdom, he's NOT trying to enhance stratification. In fact, keeping the tank thoroughly mixed, giving the extra BTU's a place to sit around, waiting to be used should work just fine in my minds eye. I really don't think that part makes any difference.

I think that what IS going to be found is that a check valve on the DHW pump has become fouled, is not seating correctly, and is allowing cooler water to mix with the boilers discharge water, causing a drop in temperature. Eventually, it will manifest itself with the DHW tank short cycling.

Remember, energy can't be created or destroyed. The buffer tank is not an energy consumer. The dilution is coming from somewhere, and there is really only one place it could be coming from.

If the boiler were underfired, it would not be capable of achieving the required discharge temperature.

This is not rocket science. It's basic hydraulics at its best.

Close a ball valve on the DHW circuit and see if your temps don't come up to par.

The oooh aaahhh test will prove it.

ME

I agree with ME

The notion of an inadvertent T-mix is the only thing I can see causing a true temperature drop in that system. That is why we need to know temperatures right at the tank vs. at the boiler and out of the tank.

If the T-mix is happening at the DHW interchange, the temperature change would be happening upstream of the tank, even if detectable downstream of the tank.

I also agree, this is a great drawing and I would like to know what program you use to create it. Very professional, which is splitting an infinitive, I know. But it does set a high standard.

thanks Brad

we draw in a mac native CAD program, PowerCADD. Same one I think Siggy liked for a long time. though I have to gripe that a mac guy like siggy releases his own software for windows only!

I would prefer to use an active modelling software of some sort, but here we are.

Piping

Rob,
I would say the boiler buddy is piped wrong. Your pulling from the bottom of the tank which is the coldest water. Tank sould be piped just like a low loss header. The radiant return should be coming back to the bottom of the tank not back to the boiler. You radiant supply should be coming off the top of the tank.

If your only supplying 6gpm from the boiler to the tank and only pulling 2-3gpm from the bottom you are always going to see that temp drop. Even if you piped it like a LLH your going to have a 6-10 degree temp difference betwenn the boiler supply water and the system supply water.

The ideal solution would be to get a sensor on the radiant supply side of the tank and get the boiler to satisfy the systems need not the tanks. Unfortunaltey that Munchkin doesn't have that capability like the Lockinvar, Viessmann and others. 

You may have to live with this unless you re-pipe.
 

Resolution

Well, we worked it out this morning.

When the contractor got to the site, he measured the temps at Boiler out, then radiant in (which is after a pass through the buffer tank), then radiant out.

-Those temps were 120 to 104 to 82... just about a 20 degree drop at each step.
-Flow meters read 4.2 GPM total for all the loops...
-so if this was real, we just exceeded the maximum output of the boiler (4.2 GPM x 500 x 40 dt = 84,000 on a Munchkin T80M)!
-Yet, boiler modulation was at a low level.

So it's conclusive: either my temperature readings are all bad, or the flow isn't what it says it is. Or, we just figured out how to make a 74k net output boiler put out an extra 10kBTUs/hr on low modulation! Can't wait to see what it would do on high fire :D!

So the contractor just opened up all the loops as much as possible. They were now reading a total of almost 7 GPM.

Within 15 Minutes the temperatures were now 119 boiler out 114 radiant out 94 radiant back. So the formerly 20 degree drop across the buffer tank is now nearly gone... just as it should be. I'm still not happy with a 5 degree drop, but we got impatient and moved on to another test.

Closed the bypass... within 4 minutes boiler out and radiant out were the same temp, within a degree or so, exactly as it should be in steady operation.


So, the problem is definitely flow, and my flow meters are definitely lying to me. They are reading much higher than they should. I would have called this impossible previously, but now I suspect that we have a batch of bad springs out there to deal with.

Bummer, but the system appears to be operating fine now even though I am running a higher pump speed than I wanted to and can't flow balance exquisitely until we replace the manifold.

Thanks for all the thoughts and consideration. This was a real head scratcher. But as my partner likes to quote from our hero, Mr. Spock, "When you have eliminated the impossible, whatever left, however improbable, must be the answer".

Enjoy your design days tomorrow (at least, those of us up in the northeast!)

Impossible buffer tanks:

You know, that buffer tank, piped like it is is like a DHW heater is really piped. The dip tube on the cold inlet outs the cold water to the bottom and the heated water rises. If you feed the hot water in the top and take the hot water out of the bottom, the hot water rises to the top and as the hot water goes in, the hot level drops vertically to the bottom, there it finally comers out hot. Like ME said, it is like a bubble in the line.
What I have to say to you though is that no matter how experienced an installer is, they may have no idea how flow works. They just pipe by rote and follow the dots on the plan. Had you gone there, you would have started closing and opening valves. Like you did. Like I do. There's something about getting close to the beast and feeling it work.
My old boss used to say to never trust a gauge. It might be wrong. I always like to put gauges on either side of a circulator and take the differential pressure.
Glad you found it. 

Yeah

site time is good. But we are pretty experienced at remote troubleshooting... been doing it for more than ten years now... and it *usually* works out pretty well.

this was a real curve ball though, for sure. the idea that 9 separate gauges on 2 separate manifold bodies would ALL fail seemed pretty impossible. Unless there is a bad batch of flow meters out there!

certainly we do occasionally run into "yeah, you didn't really follow the plan, did you". and those are definitely real headaches. I think the calls for pictures were very sound, and if we were working with anyone other than the contractor I was working with, I would have required it too. But this one has all the trust we could ever muster to do a great and thorough job. and we did confirm critical details over the phone as well.

wild. I just hope there aren't too many of these flow meters out there!

Uh.... I don't think you are done yet....

Having ONE bad flow meter, is a real possibility.

Having TEN defective flow meters on TWO different manifolds.... Nahhh... You're not done yet.

Keep looking :-)

Before you waste your fitters time to replace both manifolds, have him feel the system out during cold recovery and see where the dilution is coming form. Even the use of an Infrared camera probably wouldn't help much in this case. Too many shiny metal pipes.

Just my $0.02 worth.

ME

PS, I am sure you have all your tube flow dynamic numbers. Close off the PAB bypass and turn off the boiler, and see if the numbers you are seeing match the numbers you calculated to be able to select your pump. Easier than ripping and replacing manifolds...

ME

no insult intended

You are right I know very little about you other than what I read once on your website. I have not frequented this site for some time and left the hydronics industry about a year ago and thought I would catch up a little and saw this post. I had no idea asking what pump you used, or if the DPBV had been checked would be considered insulting.

Based on the information you supplied I simply posted what I would look at first when troubleshooting a job I was unfamiliar with that presented the same concerns you stated in your post. It seemed to me like the thread was focusing in the wrong direction.

There are those who would say that a Grundfos 15-58FC in this application is no where near enough to be piped in this arrangement in medium or high speed. I have not seen your calculations, do not know the length or size of the tubing runs and am in no way in a position to question your methods. I am personally of the belief that in the areas that I have designed and troubleshooted for that a 26-99FC would be a better choice in this application but again I have not seen your calculations.  Again that is my own personal design criteria based on what I have seen in my own calculations.

It cant be fun when something you put your heart into and take much pride in doing does not work the way you intended and you do not understand why. I am glad you are getting things figured out. As  Mark E. I think the possibility of that many flow meters not working correctly at the same time is a stretch, unless they have proven to be inaccurate in the past. I personally never trusted or installed them. I always set up my systems by supply and return temperatures.

I wonder what the calculations look like when you remove the DPBV altogether in this application with a 15-58 in the system. You might find it to be beneficial as zones close....

I will go back to my lurking. By the way, you don't know my background either.......

I had a smiley face in there

because I was poking fun a little. I'm a small pump/minimal pumping fanatic. Telling me to add/increasing pump power is "fighting words" around this office. winky smiley ;)

designing a loop field to 4-5 feet of head isn't that hard and allows primary pumping like this in many "typical" homes, like this one. we do it all the time. many times with some loops on a system closing in on 500 feet even... when you know you only need 0.2 to 0.3 GPM, the loop lengths can get kind of silly.

I thought the same thing mark

but I really don't see any other solution, and a bad batch of flow meters seems at least like it doesn't break the laws of physics.

if I'm really pushing 4 GPM of 120 degree water and it's 40 degrees colder coming back, and the only thing wrong is I'm pulling cold water and leaking it out somewhere, and even if I really do have the near 20 design dt I designed for (even nowhere close to design) in the loop field, then I would need more than 1 GPM of incoming ground water to dilute.

Also, the boiler modulation would be much higher. 40 dt at 4 GPM would be max fire on this boiler. but we were nowhere close.

I'm not sure what you're suggesting with the bypass though. we hit target flow rates with the bypass set more or less as specified, according to the flow meters. without trusting the flow meters I don't know how I would be able to gauge flow in a system with the boiler off?

Sorry, I meant to say turn off the burner...

Keep the water at around room temp if you can, it is more dense at that point.

I don't think you have a hole in the system. Easy enough to figure out. Turn off the make up and see if the pressure drops. If it does, you need to find the leak and fix it, especially on well water.

You should be able to read the delta T across the boiler on the remote readout pad that came with the boiler. At least, I can on mine.

ME

I don't follow

if the water is room temp I'll have a near zero delta T. what would that prove?

I think I see a way what you're alluding to could work though. If the flow was through the domestic circuit instead of the boiler, flow through the boiler could be very small, but present, while flow through the indirect would set the meters properly.

we will eliminate that possibility.... I think we already have, but I'll make double sure.

Thanks Mark.

edit: though this would not explain the temp drop across the buffer tank, as in that case we'd have a "primary secondary" circuit.. hot boiler, lukewarm everything else.

My 2C

I'd put a pump on the lower buffer tank outlet to the radiant manifold, and a check valve on the DHWT circuit.  If you can raise the temperature of the buffer tank, but not the manifold, it would seem to be a flow issue. Just my 2C.

See my reply from awhile ago.

Given where your thermometers are, you wouldn't seem to be in the position to say whether there is or is not a temperature drop across the buffer tank - or, at least, how much of it there is. If your flow through the boiler is lower than you think, then you are overestimating your BTU input. Also, the boiler is likely to be on low modulation, just like you say it is. Combine that with the thermal mass of the buffer, and you could be "recovering" that buffer for a long time, which would account for some temperature drop.

I also mentioned for the possibility of draining the heat from the buffer during a DHW call, further aggravating the "recovery." Isolating the DHW circuit completely, as Mark also suggested, would quickly identify whether your BTUs (and flow) are getting "lost" to the indirect.

some truth

but the pipe does not drop temperature at the domestic circuit connections, as it should if we were diluting hotter boiler output water with a cooler "domestic/radiant circuit". stays hot until it gets to the tank, cooler coming out, via touch test.

isolation

Just found this thread, and my first thought on looking at the diagram was to isolate system portions to reduce/eliminate interacting from different flow paths.   In this case, I think the only one you can is the DHW circuit.
you can also isolate the DPBP loop and turn on all the zones, and watch the temperature increases thru the system. 
Bad flow meters would be no fun, but at least they are on unions.
 I come from mostly a service (of all sorts of things) background, and that's my first thought on any system I have to service/troubleshoot.  Isolate whatever you can and test individually, and see if isolating any part fixes the problem.  I'm a hands on pipes kinda guy, and have the smooth fingertips to prove it. :)

karl

At the risk of sounding stupid...

A loooong time ago (like 25 years - almost before electricity) I experienced a similar mystery problem - drove me nuts and killed a massive amount of time.

Symptoms are the same but mine had a Phalen shell and mini tube heat exchanger (agreed a different animal than a buffer tank).  The solution after days of messing around - the exchanger was not piped for counter flow.  Switched and problem solved.

You guys think this might be a possibility?

Steve

All of us here were guilty of overpumping at some point.
Thanks to Wilo, Siggy, and others, we now seem to be moving in a better
direction when it comes to properly applying pumps.

NRT however, IMHO, has taken the over pump issue to an extreme here and decided to under pump the system. Just as bad.

Maybe it is bad meters, maybe the math works, not fighting words, just seems fuzzy to me here without looking at the install hands on.

I, and others would have had a secondary ECM pump out of the other side of the tank, let the boiler pump load up the tank. The ECM will supply the micro loads. I doubt piping it that way, and using another pump would have added to , if any, more $$ spent on the electric bill vs. a pump on speed 3 operating as it is now.

My 2cents, not worth a dime.

Look, this is getting kind of old.

since we opened the flow meters further the system performed to expectations through a period of below design condition weather.

therefore, conclusively, it's not under-pumped. Never was. Case closed on the "under pump" theory, ok?

not ok

This thread has generated a lot of hits, and all are welcomed to express their views, opinions, and comments, unless the moderator decides to close this thread, so far it has been civil, sorry if we hit a nerve.

Munchkin's preferred piping according to the manual has two pumps, one for the system, one for the boiler. Piped to the manual , the pump chart in the manual shows speed 2 on a 15-58 pumped into the boiler, not away. That is why I , and others here say you are underpumped.

Kudos to you for getting this to work. But professional installers are the ones getting the no heat calls, and are looking the owners directly in the eye. We don't have the luxury to experiment , not unless we are upfront. Like saying, hey Mr. and Mrs. Smith , I want to try to save you guys $2 a month on your electric bill so I'm going to use one pump instead of two, it may some tweaking, please be patient.
My guess is since they hired you through the internet, you furnished a pre-piped board, and you had flows, loops, etc all mapped out, everyone thought this would work right .

I'd rather not get that call, so I would have design-build the system per manufacturer preferred reccomendations, and used two pumps, and that is a fact, not a theory. What you had designed was based on theory. Now it's ok, good luck.

Munchkin with One Pump

I was told by HTP (Munchkin), that it is entirely possible to use a single circulator for the Munchkin (any Giannoni type for that matter), IF you do not vary the flow through the boiler. The problem occurs when you have TRV's or other means of varying flow through the emitters when this also varies the flow through the boiler.

HTP does highly recommend two circulators or at least one dedicated to the boiler, and do show this on their diagrams, sure.  But notice too how all of those diagrams have multiple zones with varying flows.

The HTP recommendation for P/S is because (my words here but paraphrasing), too many people do not understand the concept or cannot size a circulator. By promoting P/S piping, HTP has fewer calls into their help line. Also they cannot control someone modifying the system later, such as adding TRV's to radiators. No one  would recall the original simple design premise, they figure and they would have a flash steam boiler issue.

But a simple directly-connected constant flow system will not be a problem.

exactly

we never void a warranty, and our design strategies meet manufacturer approval, even if they aren't in the manual.

one pump

Sure muchkin, TT Solo's, etc, can have a single pump. As shown. But the 800 pound gorilla is the 250 pounds of water in a buffer tank. That's a little more than just a "fat" spot in the loop.

I doubt if any maunufacturer or engineer would design a buffer tank in such a way. And while we are on the subject. Do we really need a buffer tank on a mod/con? Really?? I've yet to see one short cycle, even with a small , single load such as a panel rad. With reset, and low modulation , seems a little wasteful.

at system flow rate

What does water mass have to do with pump size? I use the same pump on radiator conversions with hundreds and hundreds of gallons of water in them and it works just fine. it really it just a fat pipe. gallons in equal gallons out.

there are a few single loop zones in this system, bathrooms, and it's low mass. You can have a very legitimate argument about whether the buffer tank is cost justified, and neither of us would have any science behind us to prove it, but I know this boiler will probably have short cycling episodes at least some of the time, and the buffer tank ensures that will never happen. It's a "boiler protection insurance policy". I think it's worth it. You're free to disagree, and you might even be right.

Some others, including some who I have great respect for, disagree with me, so you wouldn't be alone.

Siggy and the boiler MFG's, if you ask them to put a number on an ideal target cycle time, say ten minutes! That requires buffer capacity in microzoned systems. I think that's crazy. But I don't mind hedging more in their direction.

this is not experimentation

we do it routinely. as I've said plenty of times. You can have what opinion you want about how we design, but if you call this "under pumped", you're wrong, plain and simple. You might not do it this way, that's fine. But "Under pumped" is "the pump can't push enough flow to satisfy the load or the heat extraction needs of the boiler". That's a qualitative measure, not an opinion. Either a system has enough pump energy to do the job, or it does not. this one does. that's a fact, not theory.

Unfortunately, this job had some other issue we had to figure out and it's a weird issue. but the *same thing* would have happened if I had a 15-58 on speed two on a secondary circuit as is happening here. the pressure drop across the boiler just *isn't that big*. It did not cause this problem. I know this because I know the numbers behind what is happening there. They disagree with the flow meters. That would happen no matter what circuit the manifold was installed on.

those are the facts. Have whatever opinion you like, that's fine. You are more than welcome to prefer to do things to the manual. But you don't get to pretend that we're just throwing darts here.

Darts

I was'nt pretending to throw darts.

So let's agree to disagree on this one.

I also hope you compensated the contractor for his additional time.

Cheers.

what would you think is fair

if a manifold had bad flow meters on a project you installed?

fair question for anyone. this wasted an awful lot of my time as well.

Fair.

As a homeowner, and not a heating professional, I would expect at least         
that the manufacturer would replace the things free of charge. He might wish you to demonstrate that they are defective or have you return them for them to test. It would be nice if they would compensate you for your time, but that is probably too much to expect. Not too much to ask for, though. Maybe your distributer would consider this as a goodwill courtesy for a good customer.

It seems difficult for them to be bad; they are very simple devices. A weak calibrated spring, a slightly conical cylinder, and a slightly undersize piston. Were they erratic (e.g., dirt in the mechanism)? Or just calibrated wrong (wrong spring, wrong size piston)?

it appears

they just read more flow than they really have. consistently. by perhaps 40% or more.

though, I have to own up, on a DIFFERENT job I thought this was the problem on, it does appear to be a problem, but on THAT one.... we are, in fact, underpumped, and the flow meters just hid that fact from us. That one was actually primary/secondary as well.

First time for everything! Bummer.

Now now NRT

Of course you're not under pumped. That would be impossible, because what you use are circulators!  :)
I like your thinking beyond the box myself.

P/S with ECM Vs. direct connected W/ DPBV

hey rob,

Sorry to revive this but I'v just noticed it.
I don't understand why the dpbv wasn't opening sufficiently to allow your buffer tank to come up to temperature.

With all due respect to your efforts at reducing Electrical consumption, I think you should be considering ECM pumps and P/S as an alternative to your direct connected system with bypass. The cost of a 15-58 and a good quality DPBV is close to an alpha. An alph eliminates the DPVB and uses significantly less energy even at it's highest output. In fact you can run two Alphas wide open for just a bit more energy than it takes to run one 15-58.

Finally if you are really committed to direct connected systems why use a boiler with such a high presure drop and strict minimum flow requirements. We've seen the pictures of overheated Giononni's , these HX's are beter protected from thermal stress when over pumped a bit. What happens if your DPBV is miscalibrated (as it seems to have been) what if it becomes stuck or needs adjustment? How do you confirm your boiler flow rates without yet another flow meter. How long could a flow sensitive boiler run before this problem is noticed? I'd save the direct connect systems for boilers that are more tolerant of low flow rates and or single zone systems that have consistent flow rates. BTW I have seen a stuck flow meters on new Viega manifolds, but this was easy to spot and came free with a bit of tapping. What kind of manifold are you using?

for the math we've done

unless you can ditch a PBV, it is not yet cost effective to use ECMs for constant flow apps.

having to use two alphas would be not at all cost effective compared to this setup. If we want to save electricity, we could trade the 15-58 for a single alpha, run constant pressure and keep the PBV and still be ahead by a couple hundred bucks and some electricity.

The Munchkin actually has one of the lowest pressure drops across the HE of any mod/con except the solo, but in this case, as in most cases, we use the boiler the owner/contractor wants to use and do what we have to to make it work. For most homes, you only need 3-6 GPM and we can handle that flow rate through almost any mod/con out there, honestly, with a single pump if you design your loop field well.

You make a decent point about not setting the PBV correctly, but on the other hand, it's not a complicated device and we do clearly specify what it should be set to. That's really a fair point though and I wouldn't fault someone for choosing the P/S instead of direct pump method, there are pros and cons. As long as they don't attribute all kinds of problems to direct pump systems that aren't real, either ;)

our clients typically want it "done right" though, and are paying attention, and we allow them to reap the benefits of that. Cookie-Cutter or Dummy-Proof situations have benefit too, but also a parasitic cost in upfront and ongoing expense.

so it's a matter of what floats your boat, I think.

the manifolds we are using are the Mr. Pex Brand, brass. This appears to be one run of the manifolds as we have seen no other such problems in other years, just two on systems commissioned this year with manifolds shipped early last year.