Beehive burner

OK I'm digging for ideas. There are burner programs out there, but not available in public domain. I have one mill sending a program but not sure when. I was thinking of using a PID loop, but the control is so slow in response Im thinking it will be faster to program and easier to modify if I used LIM instructions.

Details: Beehive burners burn mill waste. This is not constant, production determins waste. This changes during each shift. The burner has dome dampers (4) 12 volt winch motors. There are (4) underfire fans, with motorized damper motors, 12 Volt DC. There is (4) overfire fans with damper motor intake control also. The main temperature control is the underfire fans and the damper control. The fans run at a constant rate, the air flow changes with the dampers. The temperature must not get below 550 Deg F. There are 5 program levels. Low burn, bark only. Medium burn bark and sawdust. High burn bark, sawdust and chips. Production finished, energy saver. Weekend clean-out. Using the dome dampers and the underfire fans, the temp is regulated within the specs of the program. The chipper operator can intervine and send edgings and trimblocks to the burner, but the lag time is so bad control with fuel is a last resort. If the operator sent edgings to the burner for 15 min. it would increase the temperature about 450 Deg F. He will not know this for about 20 min. when the fuel was being consumed. This slow response or lag time is a concern to me. A PID block is a waste here. The PLC is an AB 5/20. I will be implementing some program control on a new PV 1250. To maintain enviromental compliance I must over-ride the operator to feed the burner with trimblocks. All programs Ive seen to date use PID and are fed fulltime by production and yard crew personel. There is no yard crew available to feed this burner. Based only on mill production I have a problem. They will expect enviormental compliance but we are at the mercy of the mill production. I think LIM instructions are a good option. PID functions would work, but in both cases the timing factor is a problem. Adding too much fuel, like over 20 min. would over heat the burner. I see that using temperature and time are important. Both must be used because of the lag time between the fuel addition and the resulting temperature. Any ideas out there????

Bruce

Bruce99,

I have some experience with wood-fired furnaces and wood-fired boilers. I also have a wood stove in my home. The problem with wood burners is that (as you mentioned) you cannot make a change and then expect an immedicate response. There is a long time lag while you wait for the new fuel to ignite, produce heat, and then finally have enought time to heat up the air in the furnace.

You might want to set up a "predictor" equation that continually calculates the predicted temperature, say 30 minutes from Now, based on the air and fuel factors. Basically, for wood burners,

Temperature at some Future Time = (C1 X Quantity of Fuel + (C2 X Type of Wood) + (C3 X Quantity of Air) X (C4 X Time Since Fuel Added).

This will be an equation derived from observing and measuring the actions of your specific burner. The C1, C2, C3, and C4 coefficients practically have to be derived by experiment, because every burner is different, and even the type of wood and the dryness and shape of the wood makes a huge difference in the amount of heat produced.

Anytime the predicted 30-minute temperature drops below 550 degrees, you must take some action (add fuel, air, or both).

This is probably way over my head but I had a thought. I would think waste would be monitored or at least an average quantity known per shift...I know in the paper industry waste is a big issue and they keep up with it.

Seems there should be several factors that you could use to make it more consistent.
Base waste quantity per day/shift???
Monitor waste infeed...maybe determine amount by use of time fed or something along those lines.
Maybe waste could be placed in a holding stage and fed at a monitored varying rate...That would help maintain consistency and production loads it whenever

I know I am over my head but seems you would have to be able to get an idea of how much "waste" is involved even if erratic.

You have a multi-variable system here, and a variable process response. I would avoid PID like the plague!

Air flow has two effects. First is dillution - more air will tend to slightly decrease the stack air temp. Second is combustion - more air will increase combustion rate (to a point).

I agree with your idea of using a high and low limit function to incrementally move the damper. I'd make both the movement increment and the time delay between movements adjustable.

Thanks

Thanks Guys,
Keeping the temp above 550 and the pile in the burner below 20 feet is going to be fun. I wish there was a way to monitor the pile hight. Do you think acoustic probes would work? There maybe too much debries inside to get accurate readings. I may just sit back and think about the KISS principle here, the operator will still be involved. Im not sending a probe to mars.. Bruce.

K Factor somebody?

I think that you should really look into what Lancie1 has written. Thats the way we used to do it.

We called this the K factors.

In every setings we studied the curves after the fact and found Constants for every mixes we through in the beast.

You don't need to be exactly on the bulls-eye you only need to be in the ball park. Then you refine the equation.

Run the sequence in parallel in the processor and log some data to see what you equation would have given you for the realtime scenarios. Some call this "ghost code" (TW). Ain't they crazy?

Anyway it works pretty good.

After some times all your K factors become one.

As Tom Jenkins mentionned, don't even think about PID. You'll get burned!!!

We allways based our mix in Methane-equivalence" (if this means anything in englsih???) so we never used chips or used oil but Energy bearing materials which where equal to some volume of gaz.

Of course, we also use the 3 Ts in order to meet EPA compliance.

Do you think acoustic probes would work?

Click to expand...

No, experience says two reasons.

1) Acoustic probes would have to be top mounted and if I visualize this correctly a top mount would see a high temperature, in your case a minimum of 550°F. Acoustics typically use organic compounds (like cork) internally that just won't handle the 550° temperature.

2) The swirling hot gases in a combustion environment are not conducive to a medium that carries the measured velocity signal. Too much variation resulting in distortion.

However, radar level probes might because radar doesn't care about the medium (air or vacuum makes no difference) and I think they higher temperatures. I used a Siemens LR400 on solids that form an angle of repose and it worked OK at 200°F.

Dan

danw said:

However, radar level probes might because radar doesn't care about the medium (air or vacuum makes no difference) and I think they higher temperatures. I used a Siemens LR400 on solids that form an angle of repose and it worked OK at 200°F.

Click to expand...

You can install your radar in an elbow pipe farr away from the medium. The wave will follow the path and the returning wave also.

update

OK We've decided no to PID loops, and are working on the idea of calculating average shift waste. Radar sounds good. Aimed from the side, if the beam is narrow enough, we can tell if the pile is getting too high. I was looking at some high power through beam photo eyes also. We are interested if the pile reaches the 20 foot mark. 25 feet is the bottom of the infeed conveyor chute. At 20 feet we will open the underfire intake louvers and raise the temperature to 700 deg F. The dome dampers will close at first, till the temp starts to rise, then open as the temp raises. Once fully open the temp will be controlled by the underfire intake. The time constant we calculate vs the average waste input will determin if we have to further increase the burner temp to reduce the pile size. This will take some time to get a good cross section of fuel vs time. The production here can vary from nil to record each shift. Im now going to relax and have a beer. When I start dreaming about work problems I have to slow down. (Grin...) Thanks guys. Bruce

thoughts

Hi Bruce99 here`s some food for thought or maybe waste for thought. Put a set of scales on the belt going to your burner and weigh the material going in the burner. Put a thermocouple some where up top to sense the temp. maybe a couple they have infared that put out 4 to 20 cheap. Or if you know how many tons of logs the mill is using find out what your convertion factor is ton of logs to board feet of lumber the rest is waste. Go to your sorter take board feet times specific gravity of the lumber your cutting and figure weight of waste. We used to have trouble firing boilers when the fuel was wet so i guess you`ll need a moisture dector on your belt also. Boy this could turn into a fun project. bonkhead
And by the way what happened to all the SMILES

:site:

I thought those bee-hive burners were outlawed years ago. Spectacular viewing at night - like little volcanoes, spewing glowing cinders and ash into the atmosphere...

Sounds like a candidate for some fuzzy logic.

License to burn

Gerry you are very right. There are at my last count less than 6 left in BC. The only ones left are outside towns and limited in the time frame. Our License expires in 4 years. We are 60Km outside any town and 2Km off the highway. Most people dont know its there. After the license expires, all of our Hog waste will be sold to a pulp mill building a Co-Gen plant. It will generate enough power to feed back to the hydro grid, and run the mill. There are two other mills also suppling waste for this project. It is to start construction late november 2005. The government is involved so you can add a few months or a year to that date. Till then, we burn.... .. Bruce