Garn Stratification vs. Homogenization

[gregwhs2000]

I've read a lot of the threads on this board about stratification with a Garn. I think I understand the concepts pretty well and I know that during the off-cycle the water stratification is your friend. It keeps the hottest water in your supply while the coolest stays in the bottom of the tank. I understand that the cold water rolls upward during the firing cycle, although I have noticed the amount of cold water being returned while firing seems to greatly affect the amount of water homogenized by the roll. The more cold water returned during firing, the less the roll seems to homogenize the tank.

So my issue is this unwanted stratification during the burn-cycle. I've noticed that one or two Garn owners have went so far as to install a mix loop to activate during a burn-cycle so that they can maximize the total BTU storage by homogenizing the temperature of the water in the tank during the burn cycle. I'm wondering if this is something I should consider? I try not to take the controller temp above 190* or so, but it really bothers me knowing that so few gallons of the total tank are at that temp. I would really love it if I could get the entire tank to an even 185* or so.

In my situation, I'm using only a water to air exchanger to heat my home this winter. By next winter I will add about 2000SF of retrofit under-floor aluminum plates, and some radiators and I expect to have less of an issue with water temps. But even at that point, I can't help but think it's a major advantage to be able to homogenize the tank temp during firing.

Can someone with some hands-on experience with using a mixing loop post some comments? I'd be curious to know the details on how you installed your loop, what size pump you use, etc? I'm also wondering if you feel it was worth the effort?

 

[S.Whiplash]

If Garn thought that a mixing loop was beneficial they would have incorporated it into the design. Stratification is desirable at all times so that the supply water can be used within minutes of firing to generate the useable btu output siphoned off the top of the tank. By mixing it with the cold water at the bottom you would need to wait for the entire 2000 gals. to come up to a useable temperature which might take an hour or more.

 

[gregwhs2000]

If Garn thought that a mixing loop was beneficial they would have incorporated it into the design. Stratification is desirable at all times so that the supply water can be used within minutes of firing to generate the useable btu output siphoned off of the top of the tank. By mixing it with the cold at the bottom you would need to wait for the entire 2000 gals. to come up to a useable temperature which might take an hour or more.

Not sure I agree with the opinion that Dectra has exhausted all the possible design improvements that can be made to a Garn and therefore homogenization is a bad idea.

I'm proposing a method to homogenize the tank once the lowest temp at the bottom is already acceptable for use as supply ... but the temp at the very top has reached the 190* mark that you are advised not to exceed. If a big pump were to homogenize the whole tank at that point in time, firing could continue and a rather large amount of additional BTU's could be stored while all the water in the tank is still well below 190*.

I'm not saying this would be a good idea for everyone with a Garn. It may not even be worthwhile for my situation. I just wanted to gain some insight from someone who has already done it. All comments are welcome though.

 

[TCaldwell]

I did it 2 years ago and believe that my 4 temp sensors indicate it was worth while. Without it the size of the wood load and starting temp decide how much turnover occurs on it's own. It's common to start with a 40+ degf , within 15 minutes into a burn it's within 5 degf till end of burn. As a drop in supply temps, it doesn't happen as your firing rate outstrips you demand exponentially. My mix loop is comprised of 1.5 inch bi, 2 nipples welded into the top of the tank, one towards the front low pulling up cold water and one towards the rear, about 8 inches below the water line discharging. I used a 0011 and a caleffi dirt cal in line. Not sure what the gpm is, the 0011 is energized from the same circuit as the blower motor. I have heard of some incorporating into the existing supply and return lines. If you decide to go through the top, I think a company called rhino makes high temp bulkhead fittings that are water tight. Running the pump in the other direction works also.

 

[ewdudley]

first registered garn on hearth, first o2 controlled garn anywhere

If Garn thought O2 control was worthwhile they would have incorporated it into their design long ago. Lack of O2 control is desirable at all times because Dectra has not decided to do it.

 

[gregwhs2000]

I did it 2 years ago and believe that my 4 temp sensors indicate it was worth while.

I'm wondering if this could be engineered somehow with a modified man hole cover? Or do you think it's that important that the high/low ends of the black iron be further apart? Just thinking about the least drastic method to get a loop in-place. I know from reading here that you've thrown caution into the wind long ago as far as hesitating to make mods ... I'll probably never get to the same level of expertise and confidence.

So I guess the advantage of the bulkhead fittings is that they would eliminate the need for welding?

 

[jebatty]

Stratification vs mixing has a lot to do with what your system usable or needed temperature is. If your demand is all low temperature radiant, then there is little or no benefit to stratification because you have to mix down to the radiant supply temperature. For example, for my in-floor pex I supply at 100F. So long as the supply from the tank is above 100F, all is good. Stratification does not help at all. In fact supplying at a higher temperature also returns higher temperature water back to the storage tank from the mixing valve, and this works to reduce/eliminate stratification.

With regard to the Garn, mixing will depend in large measure on the circulator flow rate through the Garn. At DP with the Garn WHS 3200, the circulator flow rate is 69 gpm. Mixing is extreme. System usable temperature can be as low as 140F and maximum target temperature, depending on outside temperature, is about 165F. The Garn is burned to as needed to maintain the desired supply temperature.

 

[TCaldwell]

yes the bulkhead fittings would eliminate welding, my four temp sensors are surface mounted equidastant on the boiler front. Before a burn I add them up, divide by 4, giving the average water temp before mixing. This allows for easy calculation of wood needed to reach 185degf for example.

 

[maple1]

If Garn thought that a mixing loop was beneficial they would have incorporated it into the design. Stratification is desirable at all times so that the supply water can be used within minutes of firing to generate the useable btu output siphoned off the top of the tank. By mixing it with the cold water at the bottom you would need to wait for the entire 2000 gals. to come up to a useable temperature which might take an hour or more.

Do you have a Garn?

 

[heaterman]

If Garn thought O2 control was worthwhile they would have incorporated it into their design long ago. Lack of O2 control is desirable at all times because Dectra has not decided to do it.

Not necessarily EW. The added cost would be a significant factor in a boiler that is already at the upper end of the price range.
Worthwhile from a clean burn standpoint.......maybe.
Worthwhile from a cost/benefit standpoint........not so much.

 

[jebatty]

Before a burn I add them up, divide by 4, giving the average water temp before mixing.

I've done the same thing to get a good estimate of the average. This calculation of "average" can be off by a fair amount, however, unless the tank is being mixed well during the loading. With my horizontal 1000 gal tank, and loading storage only, no system draw and no mixing, the stratification line is very distinct as it moves down the tank. I would assume the same on a vertical tank. This can result in the bottom sensors reading quite a bit lower than the hot stratification line just above the sensors. For example: T - 190; M1 - 190; M2 - 185; B - 140; average is 176. But the M1-190 water may be just above the M-2 sensor, and the M2-185 line may be just above the B sensor. The actual average may be much closer to (190 +190 +190 +185) / 4 = 189. Not a deal killer, but need to be conscious of how the stratification line moves in the storage tank.

 

[TCaldwell]

Curious if the garn tank shape and size affects natural turn over, my 1900 is basically rectangular with 45 degree corners, not too much different than a jr. I'd think the round style might turn over better. I'm not promoting my changes, but sometimes they do get noticed such as the turbulator!

 

[maple1]

Not necessarily EW. The added cost would be a significant factor in a boiler that is already at the upper end of the price range.
Worthwhile from a clean burn standpoint.......maybe.
Worthwhile from a cost/benefit standpoint........not so much.

You may have overlooked a subtle zinger in that post.

Or maybe I mis-saw one.

 

[TCaldwell]

It was not missed, however Steve's point is well taken as he's a Dectra rep, and not about to jump on the train unless it burns pellets. The garn still holds a good market due to its design and always will. All of mods, o2, turbulator and mix loop though not part of the garn philosophy, still pique a viable interest. It's all good!

 

[maple1]

Yes, I think a lot of us play with our boilers & have made some improvements in their operation using things not incorporated in the original design - and just the fact that they weren't incorporated in the original design does not mean they were not worth doing or do not provide a real benefit. It's a pretty minor mod/add, but I keep replacing the fireplace grating over my nozzle that keeps burning up, I have convinced myself it's worth doing - not sure if they considered something like that at the Varm factory or not, it does look kinda Red Green.

(The nozzle liner/overlay thing is another item that seems to be a pretty popular try - an easy one to play with, and I think with real benefits. Could go on likely...)

 

[flyingcow]

Red Green=duct tape. In a nozzle?

 

[maple1]

Red's a pretty diversified guy - he's so much more than duct tape. lol...

 

[varadhammo]

I have my Garn set up with a mixing loop. I dropped a string of five ds18b20 sensors into the tank at the manhole cover, sensors are every 16" from bottom to top of tank (the bottom sensor about 1-2" from the bottom, top about 1-4" from top of water level (depending on tank temp). I wanted to check out the stratification/mixing behavior. Most of the time the load is drawing a low gpm from the Garn (radiant floors). The mixing pump is an Alpha, which reports 13 gpm. I have found in my case that the Garn does not mix well unless it's being pumped. This seems to run contrary to others' experience, and I'm not sure why.

Here is one graph I saved. You can see that the firing starts with the very top of the tank at 171F, 16" down (approximately supply level) it's 164F. The mixing pump is OFF (wanted to see what happened). So midway through the firing, the top of the tank reaches 208F (boiling here at 1900' elevation) while the middle levels continue to climb. So the top is boiling away while the middle is still pretty cool, and the bottom is cold!. At this point I turned on the mixing pump. Within about 20 minutes, the tank is mixed, and the top begins to drop slightly while the rest of the tank continues to charge. (The spikes on the bottom sensor you see after that are probably one of the larger zones turning on & off. Then when the bottom sensor drops to ~90F and stays there, that is when I turned the mixing pump off at a little after 6AM (I still have to rewire some things to turn it on/off with the blower fan.)

The second graph shows an entire charge/discharge cycle with the mixing pump on (at present the "mixing pump" is actually the GARN primary loop). You can see that there is a layer of water at the top of the tank which is somewhat "stranded" (stays above the supply tapping and doesn't get mixed down even with the pump running). Maybe pumping more gpm that top layer would get disturbed, but with my 13 gpm (pumping from supply to return tapping) it doesn't really get touched.

 

[varadhammo]

Here is another version of that first graph but showing the flue temperature plot as well.

Edit: Just noticed that these don't show up with a white background for some reason, I guess the PNG has alpha transperency. Anyway if you can't see it, right click and view image in new tab.

 

[varadhammo]

For example, for my in-floor pex I supply at 100F. So long as the supply from the tank is above 100F, all is good. Stratification does not help at all. In fact supplying at a higher temperature also returns higher temperature water back to the storage tank from the mixing valve, and this works to reduce/eliminate stratification.

I don't understand how this would be the case; doesn't the return temp always = system return temp? AFAICS, only the flow rate would change based on the storage temp (I'm assuming you're talking about a thermostatic mixing valve, and not a fixed ratio valve).

 

[TCaldwell]

Good conclusive evidence supporting the garn mix loop benefit of tank charging, what model and water capacity is your garn?

 

[varadhammo]

Garn WHS2000.

Also, I did speak with the good folks at Garn and it was indicated that they are playing with the idea of making a mix loop and sidestream filter setup standard at some point...

 

[jebatty]

I don't understand how this would be the case; doesn't the return temp always = system return temp? AFAICS, only the flow rate would change based on the storage temp (I'm assuming you're talking about a thermostatic mixing valve, and not a fixed ratio valve).

Surprised me at first also. Assume 175F supply temp to the mixing valve from storage, mixing valve provides 100F to the in-floor pex, pex return temp is 70f. That means that 70F pex return is mixed in the mixing valve with 175F supply to bring the temp down to 100F. It also means that some 175F supply water is being returned to the storage tank. Stratification in the storage tank is immaterial so long as the tank can supply 100F water.

Also, supply from the tank is drawn from the top and return to the tank is to the bottom via a diptube. That means that the tank is constantly mixing, and stratification is greatly reduced. If the tank draw continued long enough, the tank would mix to a relatively uniform temperature.

Keep in mind that in my system the 1000 gal storage tank only supplies the in-floor pex. There are no other draws on storage.

 

[eauzonedan]

I have been planning just that addition this spring. I often have a 40+ degree stratification between the top of the tank and the bottom as I get close to time to light a fire. I've got three webrelay sensors in there plus the one for the controller and be advised the temp readings on the controller do NOT tell the whole tale. Mike at Precision indicated a by-pass filter would be a good idea for my 1500's boiler water. I wanted to wait till spring to pull this off, but my idea is to run a circ/filter loop from one of the bottom drain ports up to the electrical ports near the top of the tank. I would try pulling power for this pump off the duplex outlet for the fan so it would only circ while I was pumping heat into the beast. I do not want to circ otherwise as in my situation more stratification - is a good thing when drawing heat out. It is however a bad thing if recharging as I can't get the full btu load stored. I kicked this around with Karl at Northwind and he suggested I consider pulling off the top and discharge to the bottom......opposite pumping direction of my original thought.....something still in play. Your thoughts would be welcome. I was also going to fly this by the big guys at Dectra before I started down this road.
Dan

 

[TCaldwell]

Garn WHS2000.

Also, I did speak with the good folks at Garn and it was indicated that they are playing with the idea of making a mix loop and sidestream filter setup standard at some point...

What a surprise!