Gassifier Draft Control

[jebatty]

An idea that may address the variety of questions concerning draft fan and stack temperature. Would not a K-type temperature controller using a stack probe with on/off control of the draft fan work to maintain a set stack temp and control draft at the same time? Example: set the controller at 380F with hysteresis of 20F. Draft fan "on" until stack temperature = 400F, draft fan "off," stack temperature falls to 380F, draft fan "on," etc. K-type temperature controllers and K-type probes are quite inexpensive, and this would be an easy modification to a system.

An extension of this idea revolves around the barbecue temperature controller/ethernet/web control offered by Rock's. This product also will provide Twitter update reports in addition to web control. Other applications of this controller spill into many areas. Plus, a grill fitted into the gasification chamber will sizzle that steak to perfection in seconds!

 

[ewdudley]

An idea that may address the variety of questions concerning draft fan and stack temperature. Would not a K-type temperature controller using a stack probe with on/off control of the draft fan work to maintain a set stack temp and control draft at the same time? Example: set the controller at 380F with hysteresis of 20F. Draft fan "on" until stack temperature = 400F, draft fan "off," stack temperature falls to 380F, draft fan "on," etc. K-type temperature controllers and K-type probes are quite inexpensive, and this would be an easy modification to a system.

An extension of this idea revolves around the barbecue temperature controller/ethernet/web control offered by Rock's. This product also will provide Twitter update reports in addition to web control. Other applications of this controller spill into many areas. Plus, a grill fitted into the gasification chamber will sizzle that steak to perfection in seconds!

Yes, I've got a '7100' PID controller controlling stack temperature by switching from low-speed-fan to high-speed-fan on a 100 second cycle. It works quite well and maintains gasification, but I'm still looking for something more elegant.

--ewd

 

[jebatty]

What is your more elegant desire?

 

[ewdudley]

What is your more elegant desire?

With my 'standard' kindling and 'standard' splits and 'standard' draft damper setting, the low-speed-fan is low enough to lower the flue temperature and high-speed-fan brings it up, so normally everything works very nicely. But at startup I could use a little less draft damper, and also larger/uglier/higher-moisture pieces could use more air. So I'd like something that leaves the fan set at medium-low and controls an inlet damper.

I'm working with a 7100 controller switching an SSR to control a Taco 555 power head, which looks promising. The power head only has about a 0.100" throw so I need to develop a linkage that will rotate a butterfly through 70 degrees or so.

--ewd

 

[DaveBP]

But if you change fan speed to control your desired flue temperature won't that throw out your exquisitely tuned air adjustments?

Which parameter (if you must choose to control only one) is more important to overall efficiency?

 

[ewdudley]

But if you change fan speed to control your desired flue temperature won't that throw out your exquisitely tuned air adjustments?

Which parameter (if you must choose to control only one) is more important to overall efficiency?

Dunno, I set mine to the factory specs and on low speed or high speed I see a lean blue flame, probably sub-optimal.

I'm assuming for best system efficiency I want the lowest flue temperature possible while maintaining gasification and generating enough heat to keep up with the load.

I'm fairly indifferent that it may be burning somewhat leaner than optimal, which is to say I doubt I would ever see a payback on installing and maintaining a lambda control system.

 

[DaveBP]

So can you, at a given constant fan speed, change your air adjustment and notice any change in flue temp?

And same question at high and low fan speed?

 

[ewdudley]

So can you, at a given constant fan speed, change your air adjustment and notice any change in flue temp?

And same question at high and low fan speed?

My flue temperature is very sensitive to inlet damper.

The boiler has a mickey-mouse flap over about a 12 cm inlet port, with a set screw that determines minimum flap opening. A change of on-the-order-of 0.5 mm in the set screw setting will have an on-the-order-of 20 kelvin effect in flue temperature. It is of course less sensitive with low-speed-fan and more sensitive with high-speed-fan.

 

[Singed Eyebrows]

I used a cartridge thermostat to control stack temps on a non gasser boiler & it worked well./// My Atmos is similarly sensitive to small changes in intake air. It is much better now that I bent 2 of 6 vanes on the impeller flat as the cast iron manual damper would show it was dropping vacuume, I do not believe it was affecting under nozzle vacuume much at all. The German GSE Atmos uses a Belimo servo valve that the controller uses flue gas temps to regulate. This is a large ball valve that is rotated for air intake. Randy

 

[jebatty]

Put the idea to work. I had a K-type electronic temperature controller, jerry-rigged it into place, and first run this morning I set it to turn the fan off at 430F, back on at 415F (this is adjustable). No change in primary or secondary air settings. Except for the very high burn period, temp stays under 430F. At high burn, fan off at 430, takes about 5-10 seconds for the stack to cool 15F, fan back on, etc. During the off period, still a natural draft and gasification continues. I don't see any benefit or reason to change the air settings for this very limited "off" period. What adverse effects might be anticipated?

First impression is that this works very well to clip the high temperature peaks. I will give it another couple of tests to see if I notice any adverse effects by way of any buildup in the HX tubes or anywhere else. Between roughly 380-430F my Tarm purrs very nicely with high output, and this seems to be the ideal operating temperature.

 

[ewdudley]

Put the idea to work.
....

During the off period, still a natural draft and gasification continues.

....

Nice.

I was assuming I had to have some fan running all the time, which made the inlet damper setting critical, which led to a less-than-robust control in some situations. I'm going to try it now with bang-bang control and see what happens.

Does anybody know if the fan motor is horribly inefficient when it is spinning up from a dead stop?

--ewd

 

[RobC]

I had just added a Barometric damper to my set up. I mention this because on my second wood loading, to finish getting storage up to temperature. The boiler started to idle. With the fresh load of wood and the extra smoke going by damper there was quite a back-flash at the damper. This may be something to consider in choosing the bang-bang vs, high low speed control.

 

[kielka]

Hello there. I have a Scandtec (formerly tarm) solo plus 40 I installed in 2008. I'm only really getting into this forum lately. After reading this thread my mind is boggled by the way you guys are tweaking and controlling your boilers. I'm very impressed. Is there a thread that I can read up on to better my burns? are the gains that good? I've always thrown the wood in and let the baby burn flat out until it reaches water temp. I'm now wondering if there is an aspect of using a gassifier that I'm completely missing.

 

[RobC]

Welcome Kiel. Do you have storage ?

 

[kielka]

No I do not. I just made a post asking about some DIY ideas being as I've been thinking about it since I got my boiler but cost's through the dealer were exceptionally high. Also I have an older home without a means to easily install a large water storage tank. I'm open to ideas and would love to look at options.

 

[jebatty]

Rob C: The boiler started to idle. With the fresh load of wood and the extra smoke going by damper there was quite a back-flash at the damper. This may be something to consider in choosing the bang-bang vs, high low speed control.

I think the introduction of a barometric damper changes things, and that the flashback is related to the BD in that it allows air (oxygen) into the stack which allows for the possibility of unburned gases in the stack to ignite, and in your case, flashback. This might be a particular issue on a reload, if the boiler has not burned down to low coals and a good fire or substantial coals remain, followed quickly by idle, as the reload results in an immediate high burn and rush of combustible gases, probably more than the boiler can consume, draft fan shuts down, high combustibles still going up the stack, mixing with air from the BD, and ignition/flashback. It probably would be good practice to avoid a reload if an idle situation is imminent. A reload followed by a good burn period without idle would not present the same problem, as the initial rush of gases from a fresh load of wood have then substantially burned out.

If there is no BD present, bang-bang control of the draft fan does not allow air into the stack. So far as boiler operation is concerned, this is similar to the draft fan shutting down as the boiler goes into idle. Some air may still be pulled through the system by the natural draft, but that air is coming through the firebox, then the gasification chamber, and is consumed by combustion where the combustion is supposed to occur. Even if uncombusted gases go up the stack, there is no air present to cause combustion, and the situation is the same as the initial idle process. But I see an exception, at least in my situation reported above, and the exception being that the draft fan is off only for a few seconds, and then operation continues, although this may repeat several times. Yet, the issue of unburned gases going up the stack remains, and some good thinking and experience may be needed to see how great a problem, if any, this might be. It seems to me, however, that it is little different than a lambda control to maintain a desirable rate of burn. Shutting down the draft fan starves the fire of oxygen, as a lamda control also would do. What do others think?

 

[RobC]

Jim, your description of my experience is accurate. My scenario was adding more wood to the burn to finish getting storage fully charged. When my return temps start to climb the 60 will start to cycle, but well before one would expect.

 

[Singed Eyebrows]

Rob C: The boiler started to idle. With the fresh load of wood and the extra smoke going by damper there was quite a back-flash at the damper. This may be something to consider in choosing the bang-bang vs, high low speed control.

Shutting down the draft fan starves the fire of oxygen, as a lamda control also would do. What do others think?

I had thought that the Lamda regulated primary & secondary air to get a clean burn. Does it also just shut off the draft fan? Why would it need to do this if it can adjust the flaps way down? I believe the Froling has a variable speed fan & this with the control over the 2 flaps & I wouldn't see any need to start stop the fan. Randy PS, Only if boiler & tank were up to temp would it need to shut off fan?

 

[TCaldwell]

What you are discussing really requires 2 closed loops. loop one if you will would need to control draft pressure in the firebox to maintain a slightly negative pressure with respect to ambient air pressure. the result of no control is the example of the garn puffing at startup, it actually vasilates from a negative normal state, to a positive pressure caused by the accumulation of unburned gasses supplied o2 and they ignite. this scenario is caused by a imbalance of primary/secondary control and total air volume with respect to the ammount of fire present. The concept is to maintain a set neg pressure in the firebox no mater how large or small the fire is, this can be done 2 ways. One with a vfd/3ph motor a differential pressure controller. beginning and end of burn would require less air to maintain a neg pressure because there is less fire fo consume o2, as the fire ramps up so would the 3ph motor to supply more air to maintain the same neg pressure. The second way is to use a fan damper in the flue, no vfd and 3 phase, but the same differential pressure controller. It would modulate the damper to close to restrict airflow at beginning and end of burn. The pitfalls to this method for a wood boiler are periodic cleaning of the damper and flue in a dirty flue stream, the wasted energy of a fan and damper fighting each other and the non linearity of a round damper with respect to airflow. you probably have noticed a round damper does most of its control in 30% of its travel. the preferred but more costly way is the vfd, but you also get if the need arises the availibility for more than 60hz of fan speed and the vfd is more linear in air volume control. Once you have that loop set, the second loop is the actual o2 control with a insitu fluegas analyser with a sensor in the fluestack, monitoring residual o2 after the p/s air control. The o2 range of 0-20.9% is your process variable, the setpoint you want to attain is approx 7%, stiometeric value for wood at 20%mc. The primary air damper burns the wood creating woodgas, the secondary air damper burns the woodgas hopefully to maintain setpoint. Pid is one method of control by using the o2 signal as your input and output signals to the p/s dampers. pid requires a defined variables, to create a model for controller settings for a given situation, or set of given situations. Fuzzy logic is based on user defined rules, more leniant criteria for varying process. mfa, Model free adaptive control uses a neural network , has memory and only objective is to get the process value as close to setpoint as possible, regardless how non linear the process. the first 2 are commonly used by oem manufacturers in biomass control, they have extensive r and d into controller tuning. Also by maintaining a set differential pressure in the firebox it makes the primary/secondary air control more linear and predictable, enabling the use of pid and fuzzy sucessfully, secondally you are able to maintain setpoint longer because the p/s dampers are working longer in there range, with controlled differential pressure. This whole control practice is used to varying degrees by european mfgrs, they simply call it draft pressure and o2 control. There are many mfgrs here that encorporate all of the above with a burner controller and other peripheral equipt. I am working with one now micromod in rochester ny to control my garn.

 

[jebatty]

There is plenty of technology and science behind draft and O2 control. TCaldwell has a good handle on this.

Perhaps back to one original question. Will bang-bang control of the draft fan (no other air adjustment) like I initially described, and the experience on which I reported, likely to cause issues and, if so, what are they? When I watched the burn yesterday, the on-off draft fan control appeared to work very satisfactorily. It maintained a maximum high stack temp of 430-415, which continued for about 30 minutes, and then stack temp settled under 430 for the rest of the burn. During the "off" period (5-10 sec), hot burning gases continued to swirl in the gasification chamber, and natural draft, albeit much less than the fan forced draft, continued as well, as the fire and stack cooled 15F and the draft fan came back on. So, what are the things to be concerned about?

 

[Nofossil]

I've been following this thread. As long as bang-bang doesn't result in excessive / prolonged idling I don't see a problem.

Where I get curious is the relationship of fan speed to flue gas temp. Since I run many short fires, I'm particularly interested in the beginning and end of the fire.

It seems to me that once secondary combustion is well established at the beginning, there's a strong positive correlation with a lot of gain. A little fan speed increase gives you a lot more flame and heat.

Once the surface if the wood is charred, it appears to me that the amount of wood gas generation slows considerably, and flue temp is much less affected by fan speed.

At the end, there's not much wood gas at all, and increasing fan speed just pushes more excess air through the system.

I'm a bit surprised and very interested in the idea that you can close the loop around stack temperature - very cool. How well does it work towards the end of the fire, and how do you decide it's time to give up and shut the fan off?

 

[ewdudley]

There is plenty of technology and science behind draft and O2 control. TCaldwell has a good handle on this.

Perhaps back to one original question. Will bang-bang control of the draft fan work
...
So, what are the things to be concerned about?

You've demonstrated that it works, and if you're not running too rich or extraordinarily lean you're good. Extending the burn with lower flue gas temperature equates to better thermodynamic efficiency assuming you're not suffering offsetting losses due to the boiler wasting more heat though its skin by sitting there hot longer or other such secondary effects.

It's possible you'd be promoting more unburnt hydrocarbons and whatnot on one end or more oxides of nitrogen and so forth on the other end, but if you're happy with the amount of wood going in and the amount of heat coming out you can externalize those concerns for now. At least until it's no longer legal not to have lambda control on wood boilers, even if such concerns have no significance.

--ewd

 

[Singed Eyebrows]

There is plenty of technology and science behind draft and O2 control. TCaldwell has a good handle on this.

Perhaps back to one original question. Will bang-bang control of the draft fan (no other air adjustment) like I initially described, and the experience on which I reported, likely to cause issues and, if so, what are they? When I watched the burn yesterday, the on-off draft fan control appeared to work very satisfactorily. It maintained a maximum high stack temp of 430-415, which continued for about 30 minutes, and then stack temp settled under 430 for the rest of the burn. During the "off" period (5-10 sec), hot burning gases continued to swirl in the gasification chamber, and natural draft, albeit much less than the fan forced draft, continued as well, as the fire and stack cooled 15F and the draft fan came back on. So, what are the things to be concerned about?

Jim, I think efficiency & emissions probably suffer under the non running fan. Even though you have good stack temps you have way lower cfm through the stack. You might have a 1000 degree torch flame under natural draft & still have good 400 degree flue gas because of the lower cfm. I think much of the efficiency of a gasser is that hot(2000 degree) flame that is easily lost(transfered). I would guess much more unburned particulates are going up the chimney on natural draft, Randy

 

[pybyr]

Put the idea to work. I had a K-type electronic temperature controller, jerry-rigged it into place, and first run this morning I set it to turn the fan off at 430F, back on at 415F (this is adjustable). No change in primary or secondary air settings. Except for the very high burn period, temp stays under 430F. At high burn, fan off at 430, takes about 5-10 seconds for the stack to cool 15F, fan back on, etc. During the off period, still a natural draft and gasification continues. I don't see any benefit or reason to change the air settings for this very limited "off" period. What adverse effects might be anticipated?

First impression is that this works very well to clip the high temperature peaks. I will give it another couple of tests to see if I notice any adverse effects by way of any buildup in the HX tubes or anywhere else. Between roughly 380-430F my Tarm purrs very nicely with high output, and this seems to be the ideal operating temperature.

This is a very intriguing set of ideas and experiments, and I am looking forward to following continued results, especially given Jim's knack at instrumentation and methodical documentation-- it'll be interesting to see if this boosts net btu capture per pound of wood.

I do wonder whether combustion efficiency during the no-fan intervals may drop enough that it counterbalances the benefit of lower average flue temps, but that's just a curiousity, not a criticism.

Re: adverse effects, as far as I know, all of these units use capacitor run motors. While I don't have expertise in such items, I wonder if frequent cycling may create more strain on the cap/ shorter lifespan (most electronic components seem to experience more stress at turn-on than when steadily running). Not a reason to not try this, just something to perhaps be aware of.

Also- rather than the expense of a VFD, I wonder if something approximating variable speed could be achieved by pulsing the motor at a fairly rapid rate (rather than letting it coast all the way to zero rpm), but varying the length of the pulses. Could that maybe even be roughly achieved via fine tuning the various parameter settings in a PID controller (I'm more at home with analog parameters so have not really become acquainted with all the ways those devices can work)?

 

[jebatty]

Thanks for the continuing input.

NoFo: I’m a bit surprised and very interested in the idea that you can close the loop around stack temperature - very cool. How well does it work towards the end of the fire, and how do you decide it’s time to give up and shut the fan off?

What I have described has no impact on the time to shut the fan off. I'm only impacting the very high burn output which occurs, as you described, at the beginning of the burn and before the wood is charred. This seems to be the time that the highest volume of volatile gases are emitted. Other than on-off I am not changing the speed of the draft fan at any time.

The Tarm has an adjustable lo-limit control with manual reset based on smoke-box temperature (area above the HX tubes). I have set this at 100C. After the burn is finished and when the smoke-box temp falls to the lo-limit setting, the lo-limit shuts the draft fan off because at that low temp there is nothing left to be burned.

I think also I agree generally with your description of the burn process and the diminishing impact of air volume as the burn progresses.

ewdudley: Extending the burn with lower flue gas temperature equates to better thermodynamic efficiency assuming you’re not suffering offsetting losses due to the boiler wasting more heat though its skin by sitting there hot longer or other such secondary effects.

It’s possible you’d be promoting more unburnt hydrocarbons and whatnot on one end or more oxides of nitrogen and so forth on the other end, but if you’re happy with the amount of wood going in and the amount of heat coming out you can externalize those concerns for now.

Getting better thermodynamic efficiency was my goal, as higher than needed or desirable flue gas temp wastes considerable btu's through failure of heat transfer to the water. I would agree that unburnt hydrocarbons likely are being emitted. The net effect on wood burned will be a tough one to measure as there are plenty of other variables at work in any burn scenario, other than the laboratory. Prior measurements of heat transfer to storage water show the Tarm operating at 80%+ efficiency, based on 6050 btu/lb of wood input. I will attempt to see if this changes with the new burn procedure, but that will have to wait a couple of weeks because my old temperature controller quit on me and I just ordered a new one.

Boiler waste through the skin is of no concern. My boiler is in the heated space and all skin "waste" heat just means less need for stored heat later.

Randy: Jim, I think efficiency & emissions probably suffer under the non running fan.... I think much of the efficiency of a gasser is that hot(2000 degree) flame that is easily lost(transfered). I would guess much more unburned particulates are going up the chimney on natural draft.

On the one hand I agree in loss of burn efficiency and increase in emissions during the "off" period and first few seconds of the "on" until gasification is fully reestablished. But on the other hand I don't think I'm losing efficiency in heat transfer to the water; rather I think I may be gaining some efficiency in heat transfer. I'm measuring stack temp, which I have limited at this time to 430F maximum. I would think I'm losing heat transfer efficiency as this temp increases, at the expense of burn efficiency and increase in emissions. The 2000F gases your mention quickly cool as they travel to the HX tubes and up to the stack, exiting in my controlled case at 430F, rather than perhaps a uncontrolled 460-480F, which I have seen on extreme high burn occasions.