Garn water temp - difference between front and rear?

[foxt]

I've had a GARN 2000 operating for about a month now, and am curious to know if others see the same thing that I see with temps. The Dectra-supplied temp gauge on the front of the unit consistently reads 5-6 degrees hotter than my own temp gauge I have mounted in-line on the black iron pipe hanging off the top supply fitting in the rear.

My gauge is installed about three feet away from the unit, and that black iron pipe is insulated, so I can't believe I am losing 5-6 degrees in the pipe. My circ runs 24/7 at the moment, so I would have assumed I don't have much stratification in the tank. I've considered that my gauge could be at fault (Letro/Pentair dial gauges with well fittings), but I've swapped out between 4 different gauges and they all show about the same results. Could be an effect unique to the gauge design (the probe is immersed in the well which is mounted in a T fitting, but the dial is hanging out there in unheated space and I wonder if that affects the accuracy). The Dectra-supplied gauge could be reading high, but I won't be able to swap that out unless I frain down the tank.

I wonder, do other folks see the same difference? I'm wondering if the location of that front gauge (being right above the combustion chamber) yields a difference in temp between front and rear? With the circulator running I would think there wouldn't be much difference between the front and back, especially hours after the fire is out, but there is a bit more thermal mass at the front of the unit, and heck, I'm just guessing here. I have been planning on swapping out the dial gauge for an electronic model with a probe that will fit in the existing well, but before I do that I figured I ask here ...

Tom

 

[Jim K in PA]

Tom - I also have seen a difference in temp between the front gauge and my thermowell sensor gauge at the supply pipe. I don't see quite that high a difference, and mine balance out to within a degree or two after the burn is done. The bungs on the front wall are very close to the air collar, and I see a cool down on the front gauge when starting a new fire with the fan pulling cold air in. This reverses during the burn when the upper portion of the air collar gets mighty hot.

My pumps are also running 24/7, but hopefully they won't by the end of the week.

 

[TCaldwell]

tom, there might be a calibration screw on the back of the thermometer, most bimetal types have one. there is one on the flue gas thermometer, i found that the flue temps were about 25 deg off ,low untill it reached 350deg+. i switched to a type k thermocouple. i would think jim's theory might affect it also to some degree.

 

[foxt]

Thanks for the replies guys. No adjustment screw on my gauge at the back of the unit. I do notice a bit higher temp differential (maybe as high as an extra 3 degrees) when the fire is at it's peak, that could be due to the effect that Jim suggests. But when the fire's been out and I'm just drawing heat, I see a consistent 5 or 6 degree difference. It's got to be the design of my gauge. Gonna be cold tonight, I'll wrap the whole thing in fiberglass and see what I have in the morning.

Tom

 

[foxt]

Sure enough, wrapping the face of my gauge at the rear of the unit with some fiberglass seemed to work. Both gauges are reading within 1-2 degreess of each other at this point, so mystery solved ...

Tom

 

[Rick Stanley]

While the fire was burning and after it had burned down to only coals, I had 200* on gauge on front of Garn and 199* at the hx 170 ft away in the house. About 16 hours later with the fire completely out, I have 176* on the garn gauge and 148 getting to the house. This has been a pattern since I started. Seems like the turbulence of the hot firebox while burning and shortly after, is causing some nice mixing. After the water and the firebox are the same or closer temps, there is some short cycling in the back of the garn. Make sense??

 

[foxt]

I'd love to compare notes on this. I am still seeing something similar, although my temp drop is not as significant - more like 20*. I didn't have time last heating season to test the effect or dropping a submersible pump into the tank and letting that mix the tank, but I intend to do that in the next few days.

I'm still guessing that this is due to short-circuiting between the in/out at the back of the unit, perhaps related to the fact that I only have a single loop plumbed to the GARN, and it consists of a total length of about 300' (most of that doubled-up 1" pex with 60' under ground) and a heat exchanger. I could be wrong but, for some reason, I don't think I've read of this phenomenon when people have a primary/secondary system set up with a relatively short primary right at the GARN. How is yours plumbed?

Tom

 

[Rick Stanley]

Mine is plumbed the same way Tom. The 1-1/4'' underground dual-pipe Insulpex is 170' long for a round-trip of 340'. When you add the piping in each building it's a round-trip of about 380' to a flat-plate hx and back. Just a single loop like yours.

The pump is a 15-58 grundfos and I've tried different speeds and it doesn't change anything as far as temps. When I first noticed it, it scared the crap out of me. I was thinking, "great, I'm losing heat into the ground". Sinking feeling. But, as long as the firebox is hot I only see a 1-2 degree difference.

I've been burning late afternoon because this is a two family set-up and the biggest demand is in the evenings. Four showers, sometimes two at once, laundry, dishwashers, heat on everywhere, there are seven zones, you name it. So while that's going on, I have approx 200* water reaching my hx. In the morning, the fire has been out completely for hours, the heat has run all night and is still running and I have 145-150 water reaching the house, which is fine, I can run on low temps with the radiation I have, but when I go check the garn it's 160-170 or higher most mornings.
So, it keeps pumping as needed with those temps and slowly draws down the garn temp throughout the day. Then I'll build a fire later in the afternoon and within an hour or two the gauge on my hx will reflect the garn temp again within a couple degrees.
My needs are being met as it is, right now. I feel it may be different when it gets cold, so I'd be real interested in hearing how the submersible works. Good idea.

 

[foxt]

I even have the same pump, and have also not noticed any affect on the temp behavior based upon pump speed. Where do you have your supply plumbed - top or bottom of GARN? Mine is on the top per the older install manual, and I have wondered if swapping the connections would make a difference.

Tom

 

[foxt]

I should have also referred to this other thread where there was some exchange on this starting on page 2.

 

[Rick Stanley]

I even have the same pump, and have also not noticed any affect on the temp behavior based upon pump speed. Where do you have your supply plumbed - top or bottom of GARN? Mine is on the top per the older install manual, and I have wondered if swapping the connections would make a difference.

Tom

Mine is the same. Supply from the top.

 

[Gooserider]

Have you tried looking at the PIPE temps leaving the Garn, as opposed to the heat that's showing on the tank gage? That should help narrow things down to being either a problem in the Garn or something strange happening in your piping to the HX...

Not having been in a Garn, I'm wondering do the pipes on the supply and return have any sort of diffusers in the tank to spread the returning water / suction around? Or do the pipes just go through the end wall and dump into the tank? I'm wondering if adding something of the sort might help if it doesn't exist already, I know that it appears to make a significant difference in maintaining stratification for the storage tank folks...

Gooserider

 

[foxt]

Goose,
Yes, I can measure/observe a 10-15 degree delta between the front and back of the tank via thermometers installed in thermowells at the front bushing and at the top supply port.
There are no additional fittings or diffusers on my GARN at the supply and return ports - the water just mixes via whatever turbulence is created by the geometry of the tank fittings themselves ...

Tom

 

[Gooserider]

Goose,
Yes, I can measure/observe a 10-15 degree delta between the front and back of the tank via thermometers installed in thermowells at the front bushing and at the top supply port.
There are no additional fittings or diffusers on my GARN at the supply and return ports - the water just mixes via whatever turbulence is created by the geometry of the tank fittings themselves ...

Tom

OK so it does sound like a Garn problem... I am kind of surprised that they are just dumping the water into the tank w/o providing some way of spreading it out... Probably no way to do it w/o draining, especially on the bottom port, but I'd think it would really help to put some sort of diffuser on those pipes, or possibly even just an extension on the top pipe so that it is pulling from the front end of the tank, diagonall opposite from the return... One of the approaches I've heard being used that struck me as good is to put a long pipe inside the tank with a cap on the end, and a series of relatively small holes drilled in one side - put the holes facing up on the top and down on the bottom pipes, this is supposed to minimize mixing and help maintain the stratification...

Gooserider

 

[Rick Stanley]

I just checked my temps. I have only a 16* difference this morning, which is more in line with what you are seeing Tom. It has occurred to me that the large delta-t's I have seen might be because I have no way of 1- seeing both temps at once without walking up and down stairs and across the yard and in and out of two buildings (guess I should run) and 2- knowing how long the pump has been running when I do check the temps. Maybe when I see the large difference, the pumps had just started and the temp in the house was on the rise and I didn't notice it.

Kind of a long story, but I can't read the temps on the piping on the back of the garn very well either. I have two different brands of gauges and wells that aren't interchangeable AND seem to read 10* different from each other. Also, the supply water leaving the back of the garn goes through about 15' of uninsulated copper/iron piping, in an unheated uninsulated space before reaching the gauge. I don't know how much the water can cool between the garn and the gauge but maybe more than I think, especially when there's only a 1'' board between it and 26* outside overnight temps.
So after I get things tightened up and fully insulated and maybe buy an infared gun, I'll better see what's happening.

Jim K,
Your garn is a sister-ship to mine. Am I imagining it or was there a pipe on the lower return port extending into the tank a couple of feet??

 

[foxt]

I had about a 20* difference this morning, and decided it was time to drop a submersible pump into the tank and see what happens. I don't recall offhand what size the submersible is. I dropped it in the tank and plugged it in, and also had the circ that serves my supply/return loop running.

Within about a minute or two, the temp at the top supply port of the GARN had jumped 6-7*, and had dropped about 4* at the front gauge. After about 5 mins, the front temp had dropped a total of 7-8*, and the rear temp had no further change. The front/rear were now about 4-5* apart. I walked away at this point, and came back about 30 mins later. At that point, the front/rear had equalized within a degree or two (I'm calling that even, what with different gauges and all).

Interestingly enough, overall the front temp eventually dropped by more (11-12*) than what the rear rose (7-8*), indicating that the front was more of a hot-spot than the rear was a cold spot, if that makes any sense (eventual steady state temp was closer to the colder vs. warmer starting points).

So, there is clearly not enough mixing going on within the tank in this configuration. I can only think of 4 options (maybe others will have more), in order of simplicity:

1. leave the submersible in the tank and have it cycle on/off (5 mins on was enough to mix, I don't know what the off cycle would have to be), or let it run when the GARN circ is running. I'm sure this could be set up with an even more complex control at some point to trigger off delta-T at front/rear.
2. Swap the supply/return ports and see if that makes a difference.
3. Switch to a primary/secondary config, I don't think people see this effect in that config.
4. Drain the tank and extend the supply or return further into the tank (how long of an extension, and which port to extend are all great questions that I have no answers for)

I know #1 works and is easy enough to do. I don't know if #2 would have any impact, and would love to know if anyone has a GARN plumbed with only one loop (not p/s) with the supply coming off the bottom port and doesn't see this effect. #3 seems to be the "right" solution given that folks don't see the front/back delta-T in this config, but I didn't go that way in the first place because I didn't want the extra complexity and the install manual indicated that what I was doing was ok. I really don't want to tackle #4 at the moment, if ever.

So, what do you all think? Is it worth a question in to Dectra at this point as well?

Tom

 

[heaterman]

Question for Fox

Is this situation causing you problems as far as system operation and heat delivery?

 

[heaterman]

A couple other comments..........

The Garn does have an extension on the return line in the bottom of the tank. It terminates approximately 3-4' inside the unit as I recall. The purpose is to help eliminate "short circuiting", (for lack of a better term) whereby the return water gets drawn directly back to the supply. Can't happen to a great extent the way it's made. The pipe is also angled slightly downward to help keep sediment off the bottom of the tank.

I have noticed this occuring on every Garn I have checked whether it is run constant circ or not. Circulation would have to be really high to get blending in a 1500 or 2000 gallon tank because if you think about it, most residential applications are running about 12-15gpm through the Garn. at that rate it would take about 100 minutes plus just to turn over the tank one time. Add another 20 minutes or so for a 2000.

What I have seen is good stratification in the unit when the fire is down. I checked on this past week with an infrared and it showed 125-130 at the bottom (return) and 165 at the supply port. The temp gauge on the front was reading about 172 IIRC with a freshly started fire in the combustion chamber. When I checked it after about an hour and a half of work in another part of the building, it was pretty clear to see that the temp was equalizing in the tank as there was less than 15* difference return to supply. The temp gauge at that point was reading about 185* and the return port was close to 170*. So, with no fire in the Garn and the heating load drawing off what it needs, the hottest water obviously is at the top of the tank. As heat is used up, cooler water begins to fill the tank. You'll find that with an infrared thermometer you can see a marked difference from top to bottom with a distinct band about a foot deep where the temps begin to blend. When the Garn has been sitting for a long time with heat being used out of it that band of blended water eventually reached the supply port. At that time you can see quite a bit of variation from the front mounted temp gauge to the actual temp at the supply port.

 

[foxt]

heaterman, yes, it is causing me a problem, or at least a perceived one. The control for the GARN circ uses a sensor mounted in a thermowell next to the front gauge. When there is a 20* difference between the front and back temps, the system is delivering water that is 20* cooler than it thinks it is. To correct for this, I need to artificially raise the setpoint 20* above the low end of my operating range. For instance, I can take water temps as low as 120* for my distribution system, but I have to set the control for 140* (not considering loss in the pipes, fphx, etc). Isn't the net that I am effectively reducing my operating range by 20* (burn to 200*, and then run down to 140* instead of 120*)?

Thanks for your other thoughts, a couple of questions:

1. I took pictures of the inside of my GARN while I was in there cleaning it out prior to first fill. I need to dig them up, but I don't recall the extension that you mentioned. I could be wrong, but I do recall thinking it was odd that there wasn't one. An extension makes sense to me. I have been thinking that I don't have one, hence my thinking that I am seeing short circuiting.

2.

You’ll find that with an infrared thermometer you can see a merked difference from top to bottom with a distinct band about a foot deep where the temps begin to blend. When the Garn has been sitting for a long time with heat being used out of it that band of blended water eventually reached the supply port.

Makes sense to me, I expect vertical stratification in the tank.

At that time you can see quite a bit of variation from the front mounted temp gauge to the actual temp at the supply port.

I got confused there. Is the front-mounted gauge at a different elevation than the top supply port? I've been thinking that they are roughly at about the same height. Plus, the front is always warmer than the back, never the other way around. Given that there is vertical stratification, does this difference between front/back mean that there is also horizontal stratification? I suspect that the geometry and location of the combustion chamber might be contributing to what I am observing, maybe preventing the water that sits above the chamber from mixing with the rest of the tank?

3. A final question: IIRC the newer (or was it older?) GARN manual calls for the supply to come off the bottom of the tank, right? Have you installed any that way, and is there enough experience with the different config to know whether it makes any difference in this front/back delta-T? Could one potential difference be that there might be less resistance to mix with the water over the top of the combustion chamber if the return is pushing water to the top front of the tank vs. the bottom?

My net on all of this is that I'm potentially having to fire the GARN more frequently than necessary, because I am not getting all of the stored BTUs out of the tank when I choose to run it to the low end of my operating range. If I am wrong about that, I'll just accept that this is the way it works ...

Tom

 

[TCaldwell]

tom, gary switzer builds a simular boiler to the garn, he has 2 extra bungs, one high and one low that he uses a for tank mixing when the boiler is firing, i think they are 2 inch and 40gpm mix rate to homoginize tank temps

 

[heaterman]

"supply to come off the bottom of the tank",

You know..........I think I have one of the first ones we did hooked up that way......I have not been back to that install lately but the next time I'm there I will do a little checking and see. .......Interesting. I will drop an e-mail to Martin regarding this subject tomorrow.......to tired right now.

TC. Have you noticed anything like this with yours? BTW I got your message yesterday evening but didn't know how late you were up. Left for a job early this AM and finally sitting in my chair undisturbed now. So how late are you up and when can I return your call? I'll get in touch.

 

[brad068]

I guess I will have to check my tank temps too. I welded 4 extra heavy 2" couplings in the front plate of my unit. Two high and two low. Being that they are full length and installed all the way through, I put 5' extension on the return and draw right short from the front. My solar uses the other two via a plate heater and two pumps.

Its to early too tell, but I use to pull off the bottom and return on top, I also seen older garn manuals. I switched it this year and will see if it is a improvement.

 

[TCaldwell]

my wood circ used to run 24/7, i noticed a moderate difference from the front of tank thermometer and the temp gauge on the supply line that pulls off the bottom bung, it would equalize a few hrs after the fire was out, circ delivers 14gpm. now the wood circ is only energized when a zone calls, so it probably never runs more than 10 minutes at a time, annoyingly after a burn the temp gauge will read 200, and the supply[stagnant] will read 155, and increase when a zone calls. this i can live with, but i believe it would be beneficial to run the wood circ when the inducer is running, to mix the tank?

 

[leaddog]

I would think that with the big tank that it would be benificial to have statification. Why don't you take the hot off the top and return to the bottem like we all do with storage. You would get more btu's at the higher temp that way. The garn shouldn't work any different than our propane tanks!! we might have 180* on the top and 120* on the bottem and can still feed the baseboards with hot water.
leaddog

 

[heaterman]

I would think that with the big tank that it would be benificial to have statification. Why don't you take the hot off the top and return to the bottem like we all do with storage. You would get more btu's at the higher temp that way. The garn shouldn't work any different than our propane tanks!! we might have 180* on the top and 120* on the bottem and can still feed the baseboards with hot water.
leaddog

You are absolutely correct there Leaddog. At least for the actual heating part of the equation. You want stratification when you are in heating mode with any storage system. I think the problem that this guy is encountering is the result of insufficient flow during the firing mode. You want to encourage blending while firing so you can bring the entire storage volume up to temp. I was just talking with TCaldwell on the phone about that very issue. We found that a plugged up HX was severely limiting flow on the Garn side of the system in this case. What was happening was that the tank thermometer would show 180-190* but then drop off very rapidly a short time after the burn was done. No turnover of the storage was happening during firing so the temp gauge would pop right up there quick and show that everything was up to temp. The reality of the situation was that with very limited flow (probably less than 5 gpm) the bottom of the tank never heated.