Garn WHS3200 Performance Experience - Part 1 & 2 merged

[heaterman]

Chris@FHS, GARN Rep -- thanks for the info, which I will forward on to the owner. Do you have a page reference in the Garn literature for

“burn rate numbers based on 24” oak with 20% moisture content and a reloading once an hour”

? I obviously missed that, and I wish that the local rep would have brought that info to the attention of the owner. I assume the literature you are referring to was provided to the owner.

You know Jim........I have to say in all honesty, your statement above indicating the awareness of such basic information makes me wonder a little what is going on here.
If a person is troubleshooting anything in a sincere manner, it's pretty basic practice to refer back the the I&O manuals for information relative to the equipment. Agree?


If you want to determine the output of that Garn in a highly accurate manner you have a perfect opportunity to do so right now. Fire it up with no external loads connected and you eliminate any variables that may be system introduced. Temp rise x weight of water in pounds = net btu's delivered to the water/load. I'm guessing it will be very enlightening for all involved to experiment a little, and fire it as the operator was doing and then fire it for full capacity output as detailed in the manual. If you need a manual set I'm sure that a quick call to Garn would get them on the way in short order.

 

[emesine]

Turning off the pumps and lighting a fire is a very simple way of measuring heat output, but highly inaccurate. A GARN with the pumps turned off is likely going to have significant stratification. If the top of the tank is 180F, the bottom is almost certainly going to be much lower. You'd have to account for that with multiple temperature sensors. Even then, it would be very hard to estimate how much water is at what temperature, as there is going to be more or less water at each level based on the shape of the vessel, location of the firebox, etc.

Andrew

 

[heaterman]

Turning off the pumps and lighting a fire is a very simple way of measuring heat output, but highly inaccurate. A GARN with the pumps turned off is likely going to have significant stratification. If the top of the tank is 180F, the bottom is almost certainly going to be much lower. You'd have to account for that with multiple temperature sensors. Even then, it would be very hard to estimate how much water is at what temperature, as there is going to be more or less water at each level based on the shape of the vessel, location of the firebox, etc.

Andrew

On the contrary. A Garn will turn over the entire tank within minutes of the flue gas reaching operating temps. The thermal mixing is far greater than I imagined until I personally observed it. Temps at the end of a burn will be very close top to bottom, usually less than 5 degrees from what I have seen.

 

[emesine]

If you want to go to the trouble, Solodyne makes a delta-T thermostat that does exactly what you are doing here. It will interface with a grundfos flow meter to give you water flow. There are multiple temperature sensors to give you the delta t. the device itself will calculate BTU's for you. It's not cheap, somewhere around $400 for the unit itself. However, it would do the job beautifully. It interfaces directly to your computer.

I have only two problems with your data above:

1. Is your heat exchanger big enough? A small HX could kill your Garn output by impeding heat transfer to the system. The wood gun would't have this problem, as I assume it is piped directly.

2. What is the actual water flow. That's critical, as I am sure you know. On the other hand, based on pump performance curves, you can't be more than 50% off. Even an error of 100% wouldn't make 375 Kbtu into 800 Kbtu.

Thanks for the data- I love it.

Andrew

 

[Jim K in PA]

Jim - since my first response I have been reading your and others responses. I think that whomever was handling the project decisions on this system during the design and decision phase shortchanged the institution/owner and the vendors. You say that GARN specified the HX, but based on what data? What were they told regarding the heat load and transfer rates? If either vendor were suggesting two of their units then this is clearly a large load, but just how large needs to be calculated to a reasonable degree of certainty. Yes, calculated loads and calculated flow rates can be wrong, but without ANY reasonable analysis of the loads and distribution means and methods, neither unit may be apropriate for the application. Although your attempt at data acquisition is valiant, it is defeated by the large voids in the rest of the analytical process to be used for any definitive conclusions.

As has been elicited above, the operator was clearly not familiar with the best practices for operating the GARN, and perhaps for the Woodgun as well. That is why I noted the 3-hour dwell time between loadings in my first response. If the demand on the GARN was near or at it's rated capacity, it cannot meet the demand if reloading does not take place until the fire is well past the peak burn rate.

Performing a single batch burn now during the off-season will help the operator understand how the GARN works best, and will give you a better perspective on the true output of the unit. IF you are really trying "to objectively analyze the system as installed to achieve the best performance reasonably possible" then you will need to pursue the data needed and perform the tests suggested.

 

[jebatty]

Heaterman:

You know Jim….....I have to say in all honesty, your statement above indicating the awareness of such basic information makes me wonder a little what is going on here. If a person is troubleshooting anything in a sincere manner, it’s pretty basic practice to refer back the the I&O manuals for information relative to the equipment. Agree?

Absolutely agree. I also relied upon the owner/operator info that the Garn was being loaded as Garn recommended. This likely was true as to the volume of the wood load but not true with respect to the required hourly loading to obtain full rated output. The three hour loading schedule for the Garn had credibility because it also roughly matched the 3-4 hour loading schedule I use on my Tarm, so it appeared to be valid based on Tarm experience. Also, based on Tarm experience, I did not expect the Garn to provide full rated output; rather about 70% of rated output over the course of a burn, or about 640,000 BTUH. Regardless, I believe my data is accurate and reasonably describes Garn WHS3200 performance when loaded on a three hour schedule. This is confirmed by Chris-Garn, who stated

Your numbers are just about right for the way your unit is being fired. Unfortunately your operator is grossly under-firing it... So you averaged somewhere around 350,000 btuh with a third of the loadings for that period specified by the manufacturer.

Not an excuse, but reality, in that I also am a citizen-volunteer. I failed to commit to memory an asterisked footnote on the specifications. Mea culpa on that.

As to test firings, etc., the owner is working with Garn on the performance questions, and I hope that resolves the questions to the owner's satisfaction.

Andrew:

Is your heat exchanger big enough? A small HX could kill your Garn output by impeding heat transfer to the system. The wood gun would’t have this problem, as I assume it is piped directly.

Based on the data, the issue is not with the HX, as Garn HWS temperatures were falling, indicating that more BTU's were being drawn from the Garn than the Garn was producing. If hourly firings of the Garn were tested, the data may be different and there could be an issue with the HX.

Jim K: your points make sense, but even so, it was well known that the Garn could not meet the heat demand of the system. It never was purchased with the expectation that it would meet full demand. And the owner/operator believed that the Garn was being loaded in volume and frequency as Garn recommended. In that scenario I collected data and reported it as accurately as possible.

Perhaps a batch test burn will be performed. And if so, I hope I will be given the opportunity to collect data. The owner remains in consultation with Garn and I do not have the authority to step in and direct any actions. I believe Garn and the owner will cooperate to do what needs to be done to verify that the Garn can perform to specifications.

 

[jebatty]

I have had several private PM's with Chris Holley and also a phone call. In the phone call Chris triggered a thought that I pursued, and was wondering whether this makes sense. This calculation is quite rough, as my data collection was not designed to provide high accuracy over such a short period. The data points are every 5 minutes, and error here could be quite high.

In any event, I did a quick recalculation based on the ideas Chris told me over the phone. This is a recalculation applying only during a single high burn period, data points every 5 minutes (10 data points), and likely error. Also, this is based on the assumption of no material stratification in the Garn, which I believe is true given the high flow rate through the Garn and several sources of information.

Chart for 0319-0320, Garn HWS - HWR Temperature
1. Look at the time period from 395 to 440 minutes, a 45 minute period following loading the Garn and the Garn in high burn.
2. HWS starts at about 144 and ends at 155, average 149.5
3. HWR starts at about 136 and ends at 140, average 140
4. Delta T = 9.5 average
5. Water density in this temp range is about 8.20 lbs/gal
6. BTUH to system via the HX = 75 gpm x 492 x 10.5 / 0.75 = 467,400 BTUH
7. Garn HW temperature for 3200 gal of water then also has risen from 144 to 155
8. Delta T = 11
8. Garn stored BTU'S during high burn period = 3200 x 8.2 x 11 / 0.75 = 384,853
9. Garn total BTUH output during high burn period = 467,400 + 384,853 = 852,253 BTUH

I think the point Chris and others are making is that the Garn is performing properly (spec is "burn rate" of 950,000 BTUH), as the specs are based on maximum output. Maximum output is only during the high burn stage, which is about 45 - 60 minutes shortly following a full loading of wood.

Unless the Garn is kept in constant high burn state, and fully reloaded every 45-60 minutes, then real world, operational performance, is less, and is much less if a rough 3 hour loading schedule is used, as shown by the charts.

In the particular system to which my posts relate, a number of questions still might remain, but do you think this methodology is correct?

 

[Der Fuirmeister]

Someone may have already mentioned this, but to test the Garn just fire it without circulating and monitor the rise in tank temps. Add up the btus. I've done this with my Garn 1500 and it ends up at roughly the advertised btu. If I remember correctly Garn does list the average btu output for a 3 hour burn on their chart. I used that figure (210,000?? don't remember) to size the Garn, not the 350,000 btu peak rate. I end up with the best output when reloading every 2 to 2-1/2 hrs. Heat transfer is greatest at lower tank temps.

System requirements as a whole need to be considered when deciding which boiler to use.

What is the design temp of the schools heat system? Is it low temp radiant or higher temp air / water HX, or??? The heat transfer rate is greatest at low tank temps with the Garn. The WG requires much higher return temps. than the Garn.

These are both great boilers. I own them both. They are installed at different locations and the decision to purchase each one (and not two of one brand) was based on the different operating conditions of those two heat systems.

Just my two cents.....

 

[heaterman]

You are halfway there with the following statement Jim.

"Unless the Garn is kept in constant high burn state, and fully reloaded every 45-60 minutes, then real world, operational performance, is less, and is much less if a rough 3 hour loading schedule is used, as shown by the charts."

To be 100% accurate for the Garn and any other wood burner for that matter, you would add something like this; "(insert any brand name here) output varies with the input provided because output can never exceed input from the fuel load."

It's one of the elementary laws of thermodynamics.

If you want 150K from an Econoburn 150 you have to feed it enough fuel to get there. If you want 240K from an EKO 80 it's not going to do it with 20 or 30 pounds of wood. A Wood Gun E500 has to be fed the appropriate amount of fuel to develop its maximum rating.
There's no magic involved, it's simply btu's in and btu's out. The only variable is efficiency of the fuel conversion.

 

[jebatty]

Heaterman:

There’s no magic involved, it’s simply btu’s in and btu’s out. The only variable is efficiency of the fuel conversion.

And this statement is now where I will be focusing, but not until the 2010-2011 heating season. Although it might be somewhat interesting, I see little gained by doing a static test burn on the Garn. The calculation above, plus statements of others, plus Garn itself, lead to the conclusion that the Garn, depending on the wood supply and MC, will provide output into the 800,000+ BTUH range. That's reasonable.

My primary goals this coming heating season are 1) to better learn the best operation of the Garn and the Wood Gun to obtain as best a heating result in this system as is possible and 2) under as close identical conditions as I can, to attempt to ascertain the efficiency of the Garn and the Wood Gun. The wood supply and MC will be the same for each. Tests will include weighing the wood burned, measuring stack temperature, and HWS/HWR return temperature. The circulators on the Garn and the Wood Gun are the same, piping also is very close in length and diameter, etc., so even with a calculated pump head, any error will be about the same for both -- thus, relative efficiency determination will be useful.

In retrospect, I think some things learned include, all of which take into account that system demand is considerably in excess of the output of the Garn or the Wood Gun:

1. Garn 3200 gal of storage is not particularly useful, other than as a buffer, but even then, if boiler output does not exceed demand, then the storage is useless. Assuming boiler output would exceed demand, I think storage at least in the 10,000 gallon range would be needed to provide useful heat to the fan coil units. With 10,000 gal of storage and delta-T of 40 (180 to 140 water), only 3.3 million BTU'S would be stored, which may result in just a couple of hours of stored BTU's. This same statement also relates to the Wood Gun, which currently has no storage or buffer.

2. The 3 hour firing schedule for the Garn was particularly problematical because of the slow response time between firing and the delivery of usable heat. As the sharp spikes in the chart show, when the system demands heat, the system demands lots of heat right now. This will need to be worked through with both the Garn and the Wood Gun in an attempt to provide reasonable response to large spikes in demand.

3. The information from Garn that efficiency drops off after HWS is above 130F raises questions, because the fan coil units essentially always need water in excess of 130F. It would be useful to know Garn efficiency at HWS at various temperatures above 130F as compared to 130F. At this time I think only Garn may have that information.

4. The Wood Gun is designed to always operate at HWS temperature of 160F or above. Based on this alone it is likely that the Wood Gun is more suited to this heating system than is the Garn, and it is possible that the Wood Gun design also is more efficient at supplying hot water in the 160+F range than the Garn. At the same time, the Wood Gun needs some buffer and/or storage capacity to reduce or eliminate any idling. More attention will be directed to this, and possibly using the Garn as the buffer for the Wood Gun.

5. I expect experience this second heating system will produce results more in line with the owner's expectations, yet still realizing that these two boilers were never intended to meet 100% of demand.

 

[Der Fuirmeister]

Jim-

I think you're on the right track. And I would agree with most of your thoughts.

I only suggested the static test of the Garn because you seemed to believe that it did not have the btu output as advertised.

I would like to offer some helpful suggestions in the operation of the Garn. Most of which I learned first hand and not from Garn.

In order to run the temps up, I tend to reload every 2 to 2.5 hours. Because the fan acts to cool the tank (by pulling outside air through the tank) I found that I get more out of the burn by only running the fan for 2-2.5 hours, even if I'm not reloading. Otherwise the exhaust temps cool off to just above the tank water temp, and the air being run through the tank cools as much as the bed of coals heats. I added a "chimney" on the Garn exhaust. It amounts to a 15' section of uninsulated steel 6" pipe. This allows the coals to heat without excessive air flow.

You will get longer burns / more efficiency by filling it up (to the point of puffing), vs small loads when the demand is low. Again, the fan pulls in X amount of cold air over a given time span. If your fire is small, the net output will be lower. Uninsulated pipe since I don't want to to draw air when the unit not firing. The steel pipe cools way down and loses it's ability to draw air when the coals die out. Some "tinkering" as you say, but it works well.

I'm in the process of modifying how our Garn works to maximize the efficiency even more. When I complete the design and am satisfied that what I have in mind works well, I'll post that here.

In general I'd have to agree that the WG is the better machine for your system.....if you could only use one. I suspect the method you mention in #4 should work very well. That would solve both problems with very little added cost. With the Garn "pre-heating" (if your intent is to connect in series) the Garn pre-heats the tank water and the WG runs it up another 20-30 degrees before feeding to the system. By doing this you're running each in a water temp range where they run most efficiently, and providing the added btus needed, and utilizing the storage as a buffer. In affect maximizing both units.

I'm very interested in your results this winter. Please keep us posted.

 

[Como]

For a system of this size, then a wood chip boiler is a much more logical solution.

Storage with the current setup will only save the labour costs of somebody feeding it on a regular basis, is this a factor?

There was comment about a comparable system that used 2 Garn's but that was a Prison so labour was not an issue.

 

[Como]

Jim-

In order to run the temps up, I tend to reload every 2 to 2.5 hours. Because the fan acts to cool the tank (by pulling outside air through the tank) I found that I get more out of the burn by only running the fan for 2-2.5 hours, even if I'm not reloading. Otherwise the exhaust temps cool off to just above the tank water temp, and the air being run through the tank cools as much as the bed of coals heats. I added a "chimney" on the Garn exhaust. It amounts to a 15' section of uninsulated steel 6" pipe. This allows the coals to heat without excessive air flow.


I'm very interested in your results this winter. Please keep us posted.

And I am interested in your results!

I take it that your Garn vents horizontally?

 

[jebatty]

Como -- For a system of this size, then a wood chip boiler is a much more logical solution.

I don't think so. The facility is in the heart of heavily wooded, and logged, federal, state and county lands. There is plenty of firewood available in semi-load quantities at very reasonable prices; plus, staff is on hand to tend the boilers. Wood chips would be a much more costly fuel due to processing, transportation, drying, and specialized storage, and unless kept in a heated space would face freezing issues in the cold MN winters.

Der Fiur Meister -- we've noticed the cooling effect of running the draft fan as the fire burns down; and we think it is significant. I tend to think the rationale for keeping it running is to eliminate the possibility of creosote collecting in the secondary burn tubes, which could pose a problem to clean. We're discussing ideas to reduce this cooling effect.

We're also reviewing some modifications to increase the Garn efficiency and minimize cooling. Unfortunately, Garn has stated to the facility that the warranty may or will be voided if certain modifications are added.

 

[Der Fuirmeister]

Jim-

In order to run the temps up, I tend to reload every 2 to 2.5 hours. Because the fan acts to cool the tank (by pulling outside air through the tank) I found that I get more out of the burn by only running the fan for 2-2.5 hours, even if I'm not reloading. Otherwise the exhaust temps cool off to just above the tank water temp, and the air being run through the tank cools as much as the bed of coals heats. I added a "chimney" on the Garn exhaust. It amounts to a 15' section of uninsulated steel 6" pipe. This allows the coals to heat without excessive air flow.


I'm very interested in your results this winter. Please keep us posted.

And I am interested in your results!

I take it that your Garn vents horizontally?

Yes.

 

[Der Fuirmeister]

Der Fiur Meister -- we've noticed the cooling effect of running the draft fan as the fire burns down; and we think it is significant. I tend to think the rationale for keeping it running is to eliminate the possibility of creosote collecting in the secondary burn tubes, which could pose a problem to clean. We're discussing ideas to reduce this cooling effect.

We're also reviewing some modifications to increase the Garn efficiency and minimize cooling. Unfortunately, Garn has stated to the facility that the warranty may or will be voided if certain modifications are added.

If you set the fan to run just long enough so there are only small coals left there shouldn't be any creosote. When the exhaust temp gets down to 200-225 there's not much heat being added by the coals.

From what I read Garn was concerned about sparks / embers coming out of the exhaust pipe, so they require you to run it down into a barrel. That's a liability issue. The mod I'm talking about would not be added to the Garn, but outside, and won't void the warranty.

I cleaned the tubes a couple times worried that they would be plugged, but they were very clean. It was recommended not let the water temp get below 80 or you could get creosote. I typically run water temps between 90-100 as a low and 150-170 as the high.

 

[Como]

The Garn.Woodgun combination is clearly a cheaper option and of course we have limited information on the situation.

But an ideal solution for a load this size, 1m btus plus has got to be chips.

My solution to storage would be to chip weekly rather than try and store for the season.

 

[heaterman]

From someone who's been there done that.

RE: efficiency testing, Every Garn that I have done a flue gas analysis (and this is dozens) under real world conditions has shown these operating parameters. Save yourself the work and look elsewhere for the problem.

Water temp 110*-130*: efficiency will be from 86-91%. A Garn will actually go into condensing mode due to the heat exchangers ability to transfer heat. Condensate will run from the exhaust opening at temps below 120 depending in wood characteristics

Water temp 130*-150*: Efficiency will be from 83-87%

Water temps 150-180*: the range will be from 79-84% This would be normal operating temperature range.

Water temps 170-200*: Efficiency routinely tests at 77-82%

These efficiencies would be the US method of testing which is results in a lower value than the European method of rating. You can take them to the bank.

I think the fundamental flaw in this system is the assumption that someone along the chain of events made regarding what happens when you connect a heating appliance not capable of carrying the load and anticipate that it will "still help". True, it will "help" but depending on the piping/pumping arrangement of the system you could very well find yourself in a situation that consumes a lot of wood but does little to actually carry the load.
I have no idea how these wood boilers are piped in as there has never been anything useful shared regarding that but suffice to say if not done and controlled correctly either or both boilers could be a near total waste. Basically, if neither unit is capable of carrying the load, one would have to simply tee them into the return of the existing boiler, add whatever btu's could be extracted and let the LP unit supply the rest.
For example, Assuming the system works with a 20* delta T and an operating temp of 180*, you could use the wood units to add maybe 5-10*, picking the return up to 170 and then let the LP boiler work with a 10* split instead of 20*. This of course has implications for the operation of the LP boiler because it will now short cycle to a certain extent and that brings its own set of complications that would have to be dealt with. Obviously a completely different topic of discussion and I won't get into it here.
Also, designing a system that attempts to use wood fired units of disparate design such as the WG and the Garn brings its own set of complications and I think whoever did that should be brought back into the loop and asked to lend a hand to the problem solving at the very least. I really see that issue as part of the problem here. It should have been one type or the other, not both.
I would have to say that both of these boilers are severely handicapped by what appears to be the system design. The Wood Gun, or any non storage type boiler for that matter, cannot cycle on/off at the rate demanded by the LP fired system and the Garn would need a heat exchanger large enough to obtain a very, very close temp split to obtain any type of operating range at all. Using a normal 20* split for HX sizing would be ineffective at best. Along with that, neither unit can carry the load by itself and the issues surrounding that are obviously apparent at this point. Being that both the system and the Wood Gun need 150+ water temps to operate satisfactorily the amount of storage would need to be huge to serve that load for any length of time at all. In shoulder seasons where lower water temps could be used the Wood Gun would be useless due to the 150* minimum water temp that it is capable of working with.
For what you are trying to do there it's pretty simple to see that both the storage capacity and the firing rate have to be elevated significantly from what I can glean in this thread. Inadequate storage would seem to be the main deficiency and that could be solved with the addition of a tank or another Garn. You could also help the system out by adding an outdoor reset to allow for a more constant heat demand rather than cycling the air handlers on/off with high temp water. This would allow for better and more useful heat transfer through the HX on the Garn and let the Wood Gun work with a higher delta T.
I also have to think reading between the lines, that it seems you are dealing with some heat loss between the wood fired units and the actual heating load. If you could share some details on the length and location of that piping it may be useful information.



.

 

[jebatty]

Heaterman -- regardless of all of the possible faults, which may start from the decision to purchase both a Garn and a Wood Gun, and may include faults in design, installation and controls, and then may end with inexperienced operation of the boilers, a couple of things stand out after one season of operation:

1) The Garn alone, even on a 3 hour firing schedule, met system demand at outside temperatures generally above 21F; and, 2) the Wood Gun alone, on an unknown to me firing schedule, met system demand at outside temperatures generally above 0F. This is significant because it means that from April through October the Garn can meet 100% of system demand and that in all but January the Wood Gun can meet 100% of system demand based on average high/low temperatures in the area. The potential problem remains the extremes, which do occur, because while average January low is -3F, the area in January usually gets about a 1-2 week period of lows in the -20 to -40F range, and briefer periods of 0 to -20F can occur in December, February and sometimes early March.

I believe with the experience gained last winter and all the input from the posts on this thread, that improved operation of the Garn and the Wood Gun will result in these two units, alone or in combination, meeting 100% of system demand in all but about 20 days of Minnesota winter, and even then will supplement demand which will require LP boiler operation to provide the heat boost needed. It appears highly probable that the goal of 75% reduction in LP usage can and will be met or exceeded. That will result in a very substantial $ saving to the owner. Wood likely will be inexpensive in this area for a long time to come, and LP pricing is a wild card which only impacts the amount of the savings.

...how these wood boilers are piped...

I will attempt to put together a diagram of the system, along with stats provided by the contractor.

LP short cycling

This may or may not be a problem. In general the system is controlled to give priority always to the wood boilers, which never are "off" if there is demand. If demand is not met within a specified time frame (if I recall correctly, about 15 minutes), then the LP boilers fire. The LP boiler all are rotationally staged, so if the first does not meet demand within the specified time frame, a 2nd, then a 3rd, etc. LP boiler fires until demand is met.

Garn heat exchanger

I didn't post the chart on Garn HWS/HX HWS approach temperature, but under the operating conditions shown in my first post, approach temp was about 5-7F. I will monitor approach temp again under improved firing conditions of the Garn this coming heating season, and if the HX becomes a significant issue, replacement could be considered.

storage capacity

This was a known issue from the start but there was insufficient funding to permit adding storage. The boiler building has space available, and the facilities being heated also have space available, and when funding is obtained, storage will be further addressed. I see storage needed to buffer the Wood Gun to keep idling at a minimum, and possibly to allow sufficient heat for a couple of hours with the boiler wood load burned out. As you mention, substantial storage is needed, and that is costly, especially if pressurized, and ASME standards need to be met for this facility.

...pretty simple to see that both the storage capacity and the firing rate have to be elevated significantly...

How true! Sometimes I wish this was "my" system with freedom to operate it as I chose. And then again I'm glad it's not. I think the learning curve would have closed a lot faster. I'm hopeful that I get this opportunity come next winter.

...help the system out by adding an outdoor reset to allow for a more constant heat demand rather than cycling the air handlers on/off with high temp water.

To be referred to the heating contractor.

...seems you are dealing with some heat loss between the wood fired units and the actual heating load. If you could share some details on the length and location of that piping it may be useful information.

This will come with a piping diagram when I can get it put together. My gut tells me the heat loss probably is not significant.