Garn Stratification vs. Homogenization

gregwhs2000 said:

I looked at that pump as well, but decided the few bucks extra might be well spent in buying the heavier cast iron version. I'm thinking black iron piping may be better too, since more weight may be better ... you won't want to flip the power on and have this baby going on a cruise of the inside of your tank!

Maybe I'm overestimating it, but it seems like a good idea to keep it heavy if you can so that the lateral flow does not cause it to move around too much.

Click to expand...

Yes, good point.

Did you find anything for a cycle timer yet?

I'm getting ready to modify my setup as well, and I'm thinking about going the same route as you... With my current setup, the way we've been doing it is manually flipping valves and switches to go between "mix mode" and "stratify mode". I'm going to change over the primary/secondary plumbing behind the Garn to simple supply & return headers with no primary pump, and add a mixing pump separately.

I've had good luck with SESTOS timers similar to this one:

http://www.ebay.com/itm/100-240V-AC...971?pt=LH_DefaultDomain_0&hash=item3cc5dbd7eb

Contacts are only rated for 3 amperes resistive, so probably should use as a pilot relay for a motor relay/contactor.

With 2000 gal at 50 gpm that's 40 minutes for a turnover. Not sure how much on/off during a burn you will need. If the x 300 does differential measurement, one sensor at the bottom and one on the top, when the differential is less than 5 deg the pump is off and on when over during a burn, just a thought.

If you are triggering off the inducer through the garn controller that has sensitive temp wiring, might try a solid state relay for isolation.

I was just realizing that having that pump and the blower fan both plugged into the duplex GFCI on the controller will probably overload the circuit, especially on startup when they both power on simultaneously. The Liberty sump pump draws 12A full load, the 3/4 hp blower fan is ~13.8A. The relay in the Garn controller box may be rated for 30A, will have to check. I only have 12ga romex running to mine though, wouldn't be enough to run 30A. The relay in the Garn controller box is DPDT but only one pole is used for that GFCI, so you could use the other pole to trigger the timer relay that Tom suggested...

Karl_northwind said:

But you'd only want to run it while firing, to be sure.
I wonder if they make smaller ones. I've looked for such things, and not found them. time to look again.
You could definitely filter. High temperature bag filters can be found pretty easily, and as long as you can let the temp get low enough to reach in and change it, you'd be OK.

k

Click to expand...

Karl I'm going to jump in here and offer my $.02 worth regarding when to run it. And maybe this will affect how some of this gets implemented.

I have not seen any issues at all, in any size Garn concerning stratification while firing. The hottest HX tubes are located at the bottom of the tank and from what I have observed/measured, any stratification existing in the Garn is pretty much gone within 15 minutes of starting a fire. The tank temp measured on the LWCO stem will in many cases drop after a burn gets going due to the mixing occurring in the tank when those bottom tubes get rocking.
If you take the man hole cover off you can actually see the water moving up from the bottom of the tank.
Personally I don't think stratification DURING firing is an issue.


The stratification problem I see (in some installations) is not so much top to bottom but front to back.
Under low flow conditions, which I will arbitrarily call out as under 10GPM, the front of the tank above the fire box seems to "stack" for lack of a better term. It just sits there and doesn't move. This robs the system of available btu's that are basically held in storage but not available for use.
Addressing that issue with something like the high temp sump pump mentioned above "stirring" the tank intermittently would keep those btu's in circulation.

I don't think a constant mix is a good solution as you want to maintain some stratification top to bottom in order to keep highest available water temp at the supply outlet on the tank.

gregwhs2000 said:

Not yet. I found some countdown timers, but not really what I'm looking for.

I did just order the pump, but not sure how quick it will get here. I may just run it manually for a bit and then sort it out. On the flip side, I have been thinking about this device which would be used for doing much more than just controlling the mix pump:

http://www.controlbyweb.com/x300/

Overkill to just control the mixing, but I've got some other uses for it and it would be great to check the Garn temps without walking out there.

Click to expand...

Here is a much cheaper option.
http://www.automationdirect.com/adc...kable_Micro_Brick)#sort=item_code+asc&start=0
(Or which ever variant you choose)
You would need to also buy a power supply. I'm pretty sure the software for programming is still free. I suspect the internal relays can't handle the amps of the pump so external relays would be required. Guessing, I would say you could do this for around $125 and control the pump to your hearts content. You could also incorporate water temp sensors to signal pump on and off.

Take a minute and check out that site. If you want to automate something they probalby have it, and have it cheap.

heaterman said:

I have not seen any issues at all, in any size Garn concerning stratification while firing. The hottest HX tubes are located at the bottom of the tank and from what I have observed/measured, any stratification existing in the Garn is pretty much gone within 15 minutes of starting a fire. The tank temp measured on the LWCO stem will in many cases drop after a burn gets going due to the mixing occurring in the tank when those bottom tubes get rocking.
If you take the man hole cover off you can actually see the water moving up from the bottom of the tank.
Personally I don't think stratification DURING firing is an issue.

Click to expand...

Yeah, from experience I'd have to disagree as well. I am intrigued by the difference in what @heaterman has observed and what I'm seeing in my own unit. What I've observed (see datalogger graphs above in this thread, https://www.hearth.com/talk/threads/garn-stratification-vs-homogenization.141118/#post-1901377), seems quite the contrary to what he describes. I wonder what gives! In the first graph I posted above, I fired the Garn without the mixing pump running and the top was boiling at 208F while the bottom was 90F, middle was ~140F... With the pump on I can get the top to 190F, bottom to 180F.

This is not quite the homeowner situation. With the Garn 3200 at Deep Portage, and a 69 gpm circulator, which basically runs all the time when the heating system is active, the 3200 gallons are always mixed. When the Garn is used, it is usually fired to maintain 140-160F water supply to the system. It's 3200 gallons also doubles as shared system storage, along with 4000 gallons external storage which primarily serves a Wood Gun E500. The unpressurized Garn is isolated from the system and supplies the system via a large plate HX.

Deep Portage uses fan coil units for space heating, it has no use for low temperature supply water, so a mixed tank at 140-160F provides maximum btu's to the system.

heaterman said:

Karl I'm going to jump in here and offer my $.02 worth regarding when to run it. And maybe this will affect how some of this gets implemented.

I have not seen any issues at all, in any size Garn concerning stratification while firing. The hottest HX tubes are located at the bottom of the tank and from what I have observed/measured, any stratification existing in the Garn is pretty much gone within 15 minutes of starting a fire. The tank temp measured on the LWCO stem will in many cases drop after a burn gets going due to the mixing occurring in the tank when those bottom tubes get rocking.
If you take the man hole cover off you can actually see the water moving up from the bottom of the tank.
Personally I don't think stratification DURING firing is an issue.


The stratification problem I see (in some installations) is not so much top to bottom but front to back.
Under low flow conditions, which I will arbitrarily call out as under 10GPM, the front of the tank above the fire box seems to "stack" for lack of a better term. It just sits there and doesn't move. This robs the system of available btu's that are basically held in storage but not available for use.
Addressing that issue with something like the high temp sump pump mentioned above "stirring" the tank intermittently would keep those btu's in circulation.

I don't think a constant mix is a good solution as you want to maintain some stratification top to bottom in order to keep highest available water temp at the supply outlet on the tank.

Click to expand...

My recommendation to only mix when firing was mostly just so you don't mix the tank too much (50 GPM) during heat storage periods. and somewhat for energy consumption reasons. and 50 GPM is plenty to mix/filter even a 2000 several times during a several hour burn.
the GARNs I've installed all have the cold return piped at least halfway to the front of the unit. I would think this would help the front to back stratification.

Backing up a step or two here regarding stratification. Let's talk about when you want it and when you don't.
These are the operating characteristics I like to see when setting a system up.

When you are charging heat into your storage, regardless of what type of storage you have, you want total mixing of the entire volume. This will allow maximum btu's to be held in your "battery". In any type, shape or size of vessel, this would mean same temp top to bottom and side to side.

When pulling heat from storage you would ideally like to have a situation where you are pulling the highest temperature water in the vessel (always at or near the top) out to your loads and returning the coldest possible fluid back to the bottom. The greater temperature difference you can create, the better it will stratify all other things being equal.
This format will provide maximum drawdown of the heat before having to refill.

Is there agreement on those points? Or am I missing something?

Now speaking specifically to some points brought up about the Garn tank vs supply temp sensors........I have found that the actual water temperature is closer than what they read. The main factor in play there is that the tank is an immersion sensor in direct contact with the fluid and the tank sensor is a surface mount/contact type sensor. The supply sensor is always going to be lower than the tank sensor reading. I'm at the point where I automatically assume the supply is 10* warmer than the reading shows. That has been my observation in the field when I measure the temps with my own meter. (Fluke 116 with thermocouple)

What I have observed with some Garn installs, specifically with the 2000, is that after the tank has been drawn down for a while, I can just about draw a diagonal line starting about midway up the loading door and ending at the back of the manhole, which will have noticeably warmer water in it than the rest of the tank. Sometimes as much as 25*.
It seems that hot water will stack in that area and somehow due to currents or flow in the tank, not find its way back to the supply opening on the unit. Some of this might be due to the residual heat in the fire box or maybe just the fact that the firebox itself impedes internal circulation. I don't know and have never heard a good explanation of why it might be so either. To me this particular phenomena is not an ultra critical issue because the volume is relatively low in comparison to the total capacity.

Bear in mind that every Garn we have ever installed has been constant circulation in the tank itself.

Gotta run.

heaterman said:

Now speaking specifically to some points brought up about the Garn tank vs supply temp sensors........I have found that the actual water temperature is closer than what they read. The main factor in play there is that the tank is an immersion sensor in direct contact with the fluid and the tank sensor is a surface mount/contact type sensor.

Click to expand...

Agreed, in my case I have the "supply" and "return" sensors taped to back face of the Garn at about halfway down the tank and at the level of the return bung, respectively. They are useful to be able to read at the controller, but compared to the readings from my string of immersion sensors (see above) they can be about 10 degrees lower. So I don't give much credence to the surface readings except as a rough (and useful) guide.

I have just found somewhat conclusively (with my setup at least) that without a mixing pump, the top of the tank can reach maximum temperature while the middle and bottom are not yet up to temperature, which significantly limits the amount I can "charge the battery". With the pump on, I can get a more or less full charge top to bottom. Without the pump, I can't. Like I said above, I know this is at variance with what @heaterman and perhaps others have found (i.e. that the heat from the firebox & hx tubes causes the Garn to mix itself to a significant extent while firing), and I do wonder why... anyway, I second Greg's sentiment in appreciation of the elegant simplicity of the design...and also in reserving the right to keep tinkering

Tinkering and the ingenuity of the forum community with respects to the garn is unprecedented. Rightly so their reluctance to change has kept the product true. However today's market no mater what the product demands optimization.

varadhammo said:

Agreed, in my case I have the "supply" and "return" sensors taped to back face of the Garn at about halfway down the tank and at the level of the return bung, respectively. They are useful to be able to read at the controller, but compared to the readings from my string of immersion sensors (see above) they can be about 10 degrees lower. So I don't give much credence to the surface readings except as a rough (and useful) guide.

I have just found somewhat conclusively (with my setup at least) that without a mixing pump, the top of the tank can reach maximum temperature while the middle and bottom are not yet up to temperature, which significantly limits the amount I can "charge the battery". With the pump on, I can get a more or less full charge top to bottom. Without the pump, I can't. Like I said above, I know this is at variance with what @heaterman and perhaps others have found (i.e. that the heat from the firebox & hx tubes causes the Garn to mix itself to a significant extent while firing), and I do wonder why... anyway, I second Greg's sentiment in appreciation of the elegant simplicity of the design...and also in reserving the right to keep tinkering

Click to expand...

First of all do not interpret anything I say as a warning against tinkering. Tweaking, fiddling with or just messing around with one's stuff, especially mechanical stuff is the American way!

I'm wondering if some of the difference in the tank blending depends on the temperature at which you stoke thing up again?
Thinking back to what particular installation I observed the rapid convective mixing taking place I recall that specific instance as being re-fired with the return sending water back to the unit at about 90* and the tank thermometer (analog at that time and located about 12" down from the top, on the face) as being in the 120* range.
It could be that rapid mixing of the tank is more pronounced when there is a larger T going on in the tank rather than if one began charging at say 140 or 150*.

ewdudley said:

I've had good luck with SESTOS timers similar to this one:

http://www.ebay.com/itm/100-240V-AC...971?pt=LH_DefaultDomain_0&hash=item3cc5dbd7eb

Contacts are only rated for 3 amperes resistive, so probably should use as a pilot relay for a motor relay/contactor.

Click to expand...

But why not control to a setpoint?

PID controller in cooling mode with [for example] a three minute PWM bang-bang cycle. PID controller would automatically adjust on-off timing to maintain temperature at top of tank to setpoint.

http://www.ebay.com/itm/AC-110-240v...818?pt=LH_DefaultDomain_0&hash=item5d4d9155b2

heaterman said:

It could be that rapid mixing of the tank is more pronounced when there is a larger T going on in the tank rather than if one began charging at say 140 or 150*.

Click to expand...

That could very likely be the case. I'll have to do that experiment some time. It makes sense that a larger delta T between the HX tubes and the surrounding water, will create more convective mixing.

ewdudley said:

PID controller would automatically adjust on-off timing to maintain temperature at top of tank to setpoint.

Click to expand...

I don't get it...how does this provide tank mixing?

Is this discussion, and especially Heaterman's "Is there agreement on these points?", intended to be limited to the Garn or does it include other gasification boilers with an external storage tank? The agreement may be different depending on the scope of this discussion.

ewdudley said:

But why not control to a setpoint?

PID controller in cooling mode with [for example] a three minute PWM bang-bang cycle. PID controller would automatically adjust on-off timing to maintain temperature at top of tank to setpoint.

Click to expand...

varadhammo said:

I don't get it...how does this provide tank mixing?

Click to expand...

If the mixing pump was over-sized like the one above, the PID controller could adjust the duty cycle to maintain the temperature at the top of the tank more or less at a constant setpoint temperature. In cases where there is no pressing need for hot water during the burn then no need for temperature control.

Some real world numbers on the last 11 burns in my 1500 Note that my Garn side HX pumping rates bounce all over the map as my different emitters can call for supplying either 140 or 115 degree water and depending what's calling and the number of them - the variable speed pump on the Garn side of the HX ramps up and down to hit that selected set point on whatever flow is on the building side of the HX . At this point I have no good way to correlate that variable pumping rate as it relates to any stratification numbers. These averages are generated from three wet thermometers in the tank located at the mid point as well as near the top and bottom - all below the manhole opening. One odd thing I see is that I get up to a 10 degree difference between my top thermometer read and the Garn controller value at the end of a draw down, but they are spot on after a recharge. My thought is that there is a extremely sharp temperature change per elevation change at the end of a draw down and even an inch of elevation of the thermometer can skew the numbers pretty well.

Here are the average temperature difference between the top sensor and either the mid or bottom one averaged over my last 11 burns

Top to mid of tank before a burn 33.5 degrees
Top to bottom of tank before a burn 39 degrees

Top to mid of tank after a burn 7 degrees
Top to bottom of tank after a burn 33 degrees

It looks like Steve is pretty spot on (imagine that!) about the mixing that goes on during a burn and that the post burn stratification is not as bad as I originally thought. The bulk of the volume is at the mid point of the tank - so the bottom sensor represents a much smaller volume of water, and thus btu's and % or storage volume. It still would be nice to homogenize and stretch out the burn a bit more........... that 7 degrees looks to be maybe another 10-15% of effective storage in my case.

Anybody kick around the idea of an internal bubbler? Works great to keep docks from freezing up. Might need to draw the air from inside the boiler and then return it to keep from inducing "fresh air" that could lead to come corrosion issue? more brain flatulence......l