A question for the engineers...

Mark Eatherton

Specifically, Brad White, but if you feel you can answer my question, and you are not Brad WHite, by all means, swing away.



It has to do with the popular tank within a tank design being promoted by Triangle Tube.



In Siggys class at Shamrock last week, we were discussing why the GSHP people use plastic instead of copper and if copper would work to a greater degree than plastic due to its higher inherent conductivity, and Siggy said it has more to do with the Reynolds numbers and breaking up the surface boundary layer to promote better heat transfer. Fair enough.



Then, I got to thinking about a problem that I had had with 2 different TT tanks in different boiler rooms, but with the same physical configurations, that being Buderus atmospheric boiler with Buderus 2107 control, and the fact that the DHW tank stayed in demand mode for so freaking long, that it caused shortages of space heat.



Essentially, the tank within a tank design would also suffer from low velocity issues in the annulus between the tanks that would affect the ability of the inner tank to receive heat transfer, thereby prolonging the time necessary to transfer heat. Does this make sense?







I've always just thought that there was so much surface area in these tanks that caused them to work well, but there is more at play than just tank material conductivity. The rep suggested I reverse the flow through the shell side of the tank. Not sure if that would do the trick or not. I'm thinking I can NOT generate enough flow to raise the Reynolds numbers high enough without having to put a BIG pump on line.... Its already running a Grundfos 4075.



What thinks ye bretheren and sisteren of the fluids? Yea or Nay?



Anyone else ever experienced this lag in heat transfer with this style of tank?



It really sucks when property manager is standing there staring at you, staring at a tank that just doesn't want to get satisfied...



Curious minds want to know.



ME

LarryC

Not professional, not Mech Eng reply

Mark,

 

Handycapped by being an electrical engineer, I'll take a stab at the questions by asking some more questions.

 

Does the tank within a tank design have multiple ports when it comes to plumbing the boiler side?  By changing the location of the inlet or outlet, can you introduce turbulent flow inside the outer tank?

 

Is it possible to introduce turbulent flow into the tank by the use of a hard elbow at the entrance port?

 

Larry C

Ironman

Definitely not an Engineer...

In fact, they won't even let me blow the whistle. But I've had the same experience with these tanks and have had to up-size the circ to get faster heat transfer. Sometimes going to 1.25' line also to keep velocity below 5 ft. per sec. then neck back down about 1' from indirect to create some turbulence at tank.



I'm having another thought (rare moment): How about making a primary/secondary right at the indirect with a higher velocity circ pumping through the tank and the smaller circ on the boiler side. I know, it's more work and $$$, but I'm thinking in terms of a solution to an existing problem.



My $.02 worth.

Mark Eatherton

Larry...

Interestingly, the tank boiler tappings are both on the same side of the tank. I presume they have some baffling inside the annulus to force the water around the tank, but do not know for a fact that that is the case.



Thanks for the input Sparky. It's all relevant. :-)



ME

Mark Eatherton

Well...

I've already got a large pump with large cajones working in a short loop, with very little pressure drop through a cast iron boiler. Not sure there would be any advantage, but is worth considering.



Thanks



ME

Charlie from wmass

What is the delta T

of the water flowing through the heater?

hot_rod

could it be

the inside of the inner tanks are scaled up. I know how hard that mountain water is and I have seen it completely plug up a finned coil HX within a years time. High boiler water temperatures really attract the minerals to the HX surfaces.



I'm thinking how low the flow is through indirect tanks and solar tanks when connected to one or two panels, usually 2 gpm or less, and they still transfer the heat well, even the large diameter HX coil tanks.



hr



hr

Ironman

Mark,

Charlie and Hot Rod brought up a valid point, but I'm assuming that you would have covered those bases already.



I have one other suggestion: How about putting a heat exchanger in the return line that pre-heats the incoming cold to the tank, or if there's a re-circ, re-heats it?



Also, if there is a re-circ and it's running intermittently, wire it to run during a DHW call to create turbulence on the other side of the tank?

LarryC

Boiler problem not tank problem?

 

I've already got a large pump with large cajones working in a short loop, with very little pressure drop through a cast iron boiler.

 

 

What if the problem is not at the tank but at the boiler? 

 

Can you pump too fast and not be picking up enough heat from the boiler?

 

SPARKY

Steamhead

What size are the circs

and which TT tanks are they? 

haventseenenough

a question for engineers...

from a non engineers stand point is the aqua stat cap tube inserted all the way and is there heat conductive compound on the sensing bulb ?

Mike Kusiak_2

Turbulent flow and heat transfer

I don't know if the need for turbulent flow is really valid in this case. First, as you and others have already said, there is so much surface area involved, it doesn't seem likely that it would really be a limiting factor.



There is also the argument that there are really two heat transfer interfaces involved here. One being the outer surface of the inner tank and second being the inner surface in contact with the DHW. Even if the outer surface had turbulent flow, the heat transfer would still be limited by the inner surface where there would never be any turbulent flow.



I would tend to agree with hot rod that it might be something like scaling, which is limiting the heat transfer.

Mark Eatherton

The delta T...

is high under a good load, but drops to almost nothing when it is close to "topping out" the tank, which is the problem, because it can't quite get to the top of the hill, if ya know what I mean...



ME

Mark Eatherton

The old square 120 commercials...

mated to a Grundfos 4075, with 1-1/2" pipe between the boiler and the tank (short pipe runs). I am certain we have more than the recommended minimum flow going through the tank.



FWIW, there is a circ return pump moving water through the inside of the tank, os stratification is not present.



ME

Mark Eatherton

Brand new tanks HR...

BTW, I guess you are teaching the AM session and I am doing the PM session in Fort Collins. Look forward to seeing you.



No lime scale to speak of in Denver water (7 grains).



ME

Mark Eatherton

Brand new boilers...

they are 2 stage, but 170 degrees F is 170 degrees F regardless of the source. These are Buderi 2 stage atmospherics and they are perfectly set up for efficiency.



ME

Mark Eatherton

TIght fitting sensors....

Actually had to grind the body of the thermistors to get them to fit into the skinny sensor well tube that comes on these tanks, so sensor has good contact.



It is almost as if the tank is seriously stratifying, because if I pull the sensor up inside the well, it satisfies the tank setting and goes back to doing space heating, but pulling the sensor up will short the building hot. Tank set point is around 130 degrees F.



ME

Mark Eatherton

Mike...

See my responses above. I too thought that there should be more than adequate surface area to allow good heat transfer, but now I am being told that there is more to heat transfer than just surface areas. Boundary layers are a major restriction to heat transfer, and you are right, it would be on both sides of the heat exchange surface, and that may well be the issue here. These tanks are tapped on the same side for boiler water, which to me is a no no. I've tried to find a cut away for the tank to see if there is a baffle system within the tank to cause it to have to follow a circuit other than the obvious short circuit, but found nothing on the internet.



BTW, this is only an issue during periods of little to no load. The tanks handle the loads being imparted no problem. It is just when it is trying to top off that it becomes problematic, and the system is sized such the DHW priority is a must. They obviously have tested them under full load considerations, and it had to put out what they said it puts out, or they would be the target of a federal investigation. It's just near the end of the load that it becomes an issue. It would be interesting to see an actual time based chart of temperatures during load and non load recovery.



ME

Steamhead

Sure the flow isn't short-circuiting

from the boiler inlet to the boiler outlet connection, in between the inner and outer tanks?



Try throttling the 43-75 a bit and see if that helps. That circ can pump about 39 GPM against a 5-foot head. There shouldn't be that much resistance thru this thing, certainly not as much as in a coil-type indirect, so the flow may be missing some of the heat-transfer surface. This might explain why the heat transfer drops off so much as the ΔT across the heat-transfer surface decreases.

Mark Eatherton

Yeah but....

Other than convenience to the installer, why would they put the S&R tappings for the tank on the same side? Common sense tells me that is not a good idea, unless there are directional baffles that can't be seen...



It would be like taping a radiator high and low, on the same side. Not too effective at heating up all the surfaces, and my Ex Brother In Law syndrome kicks in (wet, lazy and stupid) and messes up everything.



ME

NRT_Rob

have to check the obvious

since this only happens at the end of the demand.



how hot is the DHW tank setting?



Any chance stratification, misreading the average tank temperature, or an unnaturally high temperature otherwise in the DHW tank is severely limiting heat transfer? does this occur after an "active" demand, at the end, or ONLY if the tank is sitting there unused and doing a "make up" call?

meplumber

Are you using the DHW sensor in the 2107?

I haven't seen that problem before.  I use the TT tanks, both with the Phase 3 and Weil Mclain jacketing (same tank though).  On the 40 gallon versions, with the buderus sensor, I have had wicked stratification problems.  140+ at the top and 90 deg at the bottom.



I have a college dorm with 3 of the older square Phase 3's + (2) 80 commercial amtrols, heated by a pair of Vitorond 200's, oil.  Not a single glitch.  I installed the boilers, tanks were existing.  You are way over my feeble intelligence with this one.



I am reading the posts with interest.  Good question Mark.

Mark Eatherton

Rob...

THat is one of the issues I have with this manufacturer. Their DHW thermostat is in color, not numeric, and I have not been able to find an equivalent chart that interprets the colors into an actual temperature setting. If someone has that information, it would be helpful, but the discharge temps are around 130 to 135.



And yes, this is a non load standby loss recovery issue, not a peak demand issue.



Baffeled....



ME

Mark Eatherton

Yes, using the 2107 sensor...

Am thinking about eliminating it and going with the onboard bimetal control tho...



ME

NRT_Rob

in that case

it sure sounds like stratification, other than perhaps the discharge temp. Is there a tempering valve on this system limiting the discharge temp to 135 or is it straight tank to faucet?

Jean-David Beyer

I do not have a problem, but ...

my indirect has bot hthe supply and the return of boiler water on the same side. Tank within a tank type. I wondered why the boiler water should bother to go to the other side. The hot goes into the bottom, and the return comes out the top. Now maybe the flow is fast enough that inertia makes it hit the inner tank and curve around to the other side and eventually get up to the top and back to the boiler. There is a Taco 007-IFC in that loop; the flow resistance of the heater must be pretty low, and the resistance in the Ultra-3 heat exchanger is probably not very high (though non-zero). So maybe a little over 6 gallons per minute of boiler water.



I do not really know what is inside, but the drawing by the manufacturer shows a stainless inner tank that is sort of like an accordeon inside the outer tank. Since it is an artist's rendition, there is no indication of any baffles inside.



But as I said, I always get enough hot water, so I suppose it is not a problem. On the other hand, my hot water needs are low.

Jean-David Beyer

Non-numeric aquastat

Mine is like that. There is an arrow where they suggest to put it, and my contractor left the knob at that position. When I first turned on the hot water, it seemed awfully hot. I put a thermometer in the water from the nearest faucet to the hot water heater and it went off the top of the scale. So I turned it down a few gradations at a time until the water coming out was about 125F which about as hot as I wanted it. (No kids in the house.) That is quite a ways down from the arrow. I wish I had noticed if the contractor put thermal grease on the sensor before pushing it into the well. The thing does hold the temperature well, though, so it is precise, whether it is accurate or not.



It is my understanding that those aquastats are not all that accurate, so they do not want to mislead anyone by putting numbers on.

cattledog

Stratification

"It is almost as if the tank is seriously stratifying, because if I pull the sensor up inside the well, it satisfies the tank setting and goes back to doing space heating, but pulling the sensor up will short the building hot. Tank set point is around 130 degrees F."



I concur with the stratification hypothesis.



Perhaps there is a sweet spot for the sensor position which puts you between long runs with no space heating, and not enough hot water?



Stratification would also put the focus on the inner tank recirculation. Its as likely to be subject to the brother in law principle as the jacket flow.

heatboy

Raising my hand with questions, Perfesser.

What size tank?  Even the smallish 60 needs 20 GPM.  What size is the S&R piping and how far from the boiler?  The 43-75 will handle the flow needed, up to 15' of head, if piped correctly.



Does the boiler run on reset?  If it does, it might take some time to get the boiler up to 160 - 180 F once the control sense a need for DHW recovery.  On DHW, does the control start the DHW pump immediately or does it allow the boiler to heat up first, then run the pump?  I don't think these indirects lend themselves well to reset systems.



I think it would be very difficult to coat either side of the tank with sludge (boiler side) or calcium (domestic side) due to the way the inner tank flexes with temperature changes and there is really nothing for debris to grab onto a smooth surface like that.



I would check incoming water temperture of the potable side and measure Delta T across the boiler loop both static and draw periods.  Could the water be short circuiting from the supply to the return since both ports are in very close proximity to each other?



I am assuming the supply is at the bottom and the return is near the top of the tank, per TT's instructions?

Leo_G

Hey Prof,

I have installed exactly one of these style tank within tank systems. Am not impressed with it. It seems to me that stratification is a very likely outcome of this design. The one that I installed seems fine to fill a tub or allow a shower, but to my mind, seems to need to reheat way to soon and to often.



I decided to stay with the inner coil design. Seems to be more precise.

Leo_G

Hmmm

thought expirement -



Heating of outer jacket produces an out to in flow with the DHW portion. i.e. in my minds eye I see the inner heated water flowing up the side of the tank, rolling across the top to displace the cold water downwards and outwards. Following this flow would it not tend to bathe the top mounted aquastat with hot water and inform that the tank was satisfied much before it really was?  Then would not the hotter water, being lighter settle up on top of the coller/cold water, thus allowing maybe only half a tank of hot water?



Also in this thought expirement, as a high use occurs, the cooler water would get to the centre mounted top aquastat earlier, thus bring the heating cycle on, and no doubt, that with all that surface transfer area and the rounding of the top of the inner tank to promote quicker access to the hot water for the potable system, would in fact be very effeciant for high use times.



I am thinking that this may be the very thought process used by the developers of this style of tank. They would be far more efficient for high use systems, say appartments, but would give up the efficiency on the lower use parameters, say small single/double occupancy homes.



Interesting post Prof, thank God I had my coffee before reading it!

Mark Eatherton

Jeffrey...

It's 1-1/2" copper (ProPress) and some B.I.P., the tank sit right next to the boiler. Less than 5' of piping betwixt them. I honestly don't think a lack of flow is the issue, and I can't see how too much flow would degrade its performance. The only thing that keeps sticking in the back of my brain is the 2107's ability to "coast" up to the set point and that logic is based on using the Buderus Storage tanks. Maybe IT is the confused one...



As for stratification, there is a full time circulation return pump running on the potable side, so I can see the potable stratification as being an issue.



Timmy Mac gave me Guy Wollards email, and he responded and said he'd drop by later an comment. Said he's out Christmas shopping. Crazy man :-) Thanks Tim.



ME

Mark Eatherton

Leo...

My 80 gallon reverse indirect does the same thing after shut down. I use it, it heats up the water and hits set point, and post purge pump runs about thirty seconds, then shuts down, then 1 or 2 minutes later, it is back on again, for a shorter duration. I think what it is doing is de-stratifying or turning over after the pump shuts down, with as you explained, hot rising, and cold settling, and it triggers the controls to fire again. I wish I had the ability of ignore, or postpone the second, immediate, subsequent call. Not sure why other tanks don't have the same issues tho. My R/I is an Amtrol 80 with the extraction coil in the top, instead of the bottom where it usually lives. And I am running boiler water through the tank, and potable water through the coil. Must say, it works like a CHAMP.



I honestly do not know why they (Amtrol) don't produce it that way regularly. They be one of the only N.A. suppliers doing so.



ME

Gordy

Being simple minded

 But an Analogy comes to mind here. Thinking of cooking a roast in an oven from the outside in......Prime Rib to be exact. Verses cooking in a microwave from the inside out.......YUK.



 But anyway my thought is the sensor placement in the tank in a tank design not maybe deep enough in the tank. If the meat thermometer probe in a roast in an oven is not in the middle of the roast, it says well done when in fact actually it is rare in the middle if the probe is not deep enough in the roast. I like rare anyway!



With a coil design being in the middle of the tank essentially the probe being on the outer surface is not getting the hottest reading in the middle of the tank in this situation. Like cooking  with a microwave from the inside out. So the probe is always seeing the cooler water until satisfied.



I don't think pumping, more turbulent flow, ect will change the out come

unless stirring the tank from with in while heating has a benefit.



Just a thought



Gordy

bob_46

Information

Mark, We need more data. I'll take a stab at it. When the domestic is relatively cold there

is a large ∆T between boiler supply temp and domestic temp so lots of heat transfer. As

the domestic approaches set point, i think you said 135º, and boiler supply, I think you

said 170º not near as much temp difference and less transfer. Have you measured the actual boiler supply temp entering the TT? You mention hi lo control. Say the hi lo is set to

run the boiler at high fire up to 170º and then switch to lo and the limit is at 180º. Let say the control has a 10º diff. The boiler runs to 170º and switches to lo fire, but it can't quite match the load on lo fire. The boiler temp drifts down all the way to 160º before it goes back to hi fire. I think you need to keep that boiler between 190º and 200º until the call for DHW is satisfied.

Jean-David Beyer

My tank within tank does not do this.

I have a W-M ultra 3 boiler and one of their indirect hot water heaters. The well for the aquastat goes down fairly near the middle of the tank and goes quite a long ways down. Apparently you can control things by how far down you push the aquastat, but mine is wherever the contractor put it; probably most of the way down.



Mine does not restart soon after the aquastat is satisfied. One reason is that the circulator that drives the indirect is (now) set to run for 2 minutes after the aquastat is satisfied to drain much of the heat from the boiler. The other reason I let it go so long is that the next zone to run might be the radiant slab and I would not like to run 180 degree water through it. Since it would be only a short time, maybe I am just being paranoid about this.



Since the aquastat is set to run the domestic at a bit over 120F, and since the boiler supplies 170F water to heat it, the outer tank is full of 170F hot water for a while (until the circulator pumps most of it out), this may supply enough heat to deal with the domestic side "destratifying." Maybe? The outer tank holds about 6 gallons of boiler water.

Dave Yates (GrandPAH)

sounds like (tugging on my ear)

Is there a DHW draw contributing to those long runs? The recirc would tend to add some kick to the current & possibly cause incoming cold water to be sent round the tank to "fool" the aquastat.



The DHW recirc pump could be adding enough additional load to cause looonnnggg end-of-cycle run-times or, perhaps, given the TT tank's design, that flow might be sweeping cooler water currents over the aquastat well.



I'd be curious to observe what happens if you disable the recirc for one or several cycles while there's also no DHW being used.



I'd also be curious to know the recirc & DHW & hydronic GPM flows along with Delta-T's across all four ports under existing conditions.

tim smith

Re: not engineer but ,TT indirect and not so good heat transfer

Mark, I am going to babble for a minute, don't mind me. We know heat transfer drops dramatically as temp differential gets closer. I wonder if during reheat of tank, would shutting off recirc would come in handy. That would create a much wider differential at lower portion of tank during load times as cold water is entering bottom of tank. I just wonder if using the recirc is bringing your differential so close on the cold side that your exchange goes to heck. Just an off the cuff idea. Its nice to have the recirc most the time to get the tank fully hot but it may be messin us up a bit on transfer.  "not an engineer"

Glen

hang on ther' perfesser!!!

170 deg F is 170 deg - but your dhw don't care about that. what they care about is horsepower or the btu's behind the 170 deg. so what I have read all already on this thread has a lot of validity - 1 gpm @ 170 is one thing but with two tanks to satisfy you might need 100 gpm @ 170. so look at everything, pipe sizing, circ sizing and then the real question does my boiler(s) have enough jam to make this happen. what if my boiler can't make 170 with 150 coming back??? at the flow I need - what then??? what was siggy's quote "more and more boiler system (design) are being driven by the dhw needs rather than the space heating" yes you still want to design around the 2-4 fps but that is just one of the criteria - I think you need to go back to the drawing board and re think the calcs.

Mark Eatherton

Glen....

These are two completely separate, but identical boiler rooms. One boiler and tank per room. Not two tanks and one boiler.



The boilers are 300K btuH each. PLENTY of horse power. The boiler is 2 stage atmospheric. It is bouncing off of its high limit, so I know everything on the fire side of the system is in good shape.



There has never been a shortage problem during an actual load, so it is obvious that as far as generating hot water is concerned, no problem. It is during little to no load conditions that it has trouble getting over the top of the hill and satisfying the call.



I am going to write a general synopsis at the bottom of this post.



Thanks for your input.



ME

Mark Eatherton

Bottom line synopsis...

To all, thanks for the valid inputs and questions. It sounds as if this (these) is (are) isolated cases. Triangle tube obviously has a good handle on heat transfer, or they wouldn't be in the heat transfer business. I think I am being fiddle faddled by the fickle finger of fate. That, or diblbled by the dirty digit of destiny, as my father , rest his soul, would say.



Regarding flow through the tank, I am assured that the flow around the tank, goes all the way around the tank, and if properly flowed, the Reynolds numbers on the source side of the system are in good shape. However, what goes on on the inside of the tank is significant stratification, which is warranted and wanted in the European scheme of things. As for scaling, due to the accordion design of their tank, and the associated expansion/contraction, the scale doesn't get a chance to accumulate on the heat exchange surface (theoretically). Not sure where all the stuff that is breaking off ends up going to though... I'd recommend the use of a magnetic water conditioner if you suspect hard water lime scale situations. But I digress.



I was told that by pulling the sensor higher in the well, that it would cut the load off faster, which I did and it did. Done and done. Just seems kind of silly to me to have a well that long, if it isn't necessary.



I know of numerous indirect tank manufacturers that keep their aquastat position high in the vertical profile, and I understand the need for completely immersing their coil in cold water to effect the greatest delta T to drive the heat across the process, but it seems to me that it would have a tendency to short the load hot water during demand, but based on the lack of complaints regarding shortages of hot water during peak demand (there have been none) I guess it works just fine.



I think the real crux of the problem is the use of the Buderus 2107 control with the solid state DHW sensor. I think there is some borderline artificial intelligence onboard their control logic. I have witnessed this control with a Buderus DHW tank actually shut down the burners before it had achieved storage tank set point, and it had not hit what I would consider a high limit condition (180 degrees F). The tank actually coasted up to its set point without having to hit the boilers high limit. Could have knocked me over with a feather... I guess their logic sees a rate of rise, and knows the target and shuts down early to allow the mass momentum to coast to avoid unnecessary over shoot. Too smart for me :-)



So, in closing, the tanks works as advertised under load, but it works differently under no or part load, i.e. standby loss considerations. If one were to use a conventional boiler control logic that is capable of generating high (180 to 190 degrees F) temperature, the tank would be able to hit the lower (130 degrees F) set point at the tank, and shut down and go back to doing space heat.



BTW, the Buderus control does not limit DHW priority, like some of the newer controls do. Also, the logic of the 2107 can not be forced to generate a higher hot water temperature than what it feels it needs to do to satisfy the load.



Again, thanks to all who contributed. I guess the simple solution is to raise the elevation of the sensor in the well.



Personally, I would LOVE to see a video of a tank with a stripped insulation package, showing in infrared, what happens during cold start and recovery. I just can't see the flow patterns the factory says is happening with the S&R tappings on the same side of the tank, as opposed to being diagonally opposed, top and bottom. But it wouldn't be the first time, nor would it be the last time that my minds eye was incorrect.



ME