Material Handling - Feeder Motor Speed Control VIA Blower Pressure

[mellis]

Let me see if I have some key points right...

1. The original problem was the feeder valve went too fast and caused line plugging.

2. This was solved by using a VFD to slow it down to an acceptable feed rate.

3. Everything works fine if you leave the feeder valve at this speed.

4. You have a problem with bags running out.

Conclusions

1. Do not try to control the feeder valve speed. It doesn't do anything good for you. If it's too fast, the line plugs (the orginal problem) so why install controls that let it go too fast?

2. The problem with bags running out is completely unrelated to feeder valve speed. I can see where a beacon light that comes on if the discharge pressure is low and the feeder valve is running for a couple minutes would be helpful to let the operators know that the bags have run out. I can also see shutting down the system automatically if say 10 minutes go by and no one hangs another bag.

3. What happens if the silo fills up? You may need a high level switch so you can shut the feeder valve off before you plug the line. Or if the operators know the level, they can just avoid hanging enough bags to fill it. That's probably OK if the silo is relatively large.

 

[Anvilsoup]

Mellis has a point

4. You have a problem with bags running out.

Sometimes one of the most important parts of the control system is overlooked by the controls guy. Specificly the operators. What do the operators do when the line is plugged? Do they go on break? Watch the millwrights unplug the pipe? Or do they get sweaty, and dirty, or perhaps get blisters on their hands from shoving 3/8" rebar in the pipe to unplug it. If an operator knows that if he can keep the hopper full it will keep him from getting blisters, he will keep the hopper full.

 

[DickDV]

In order to have an effective control system, you have to be able to detect and react to the correct variables. In my view, blower discharge pressure is not the correct variable.

It would be nice to have a constant velocity blower but it seems clear that solids loading of the tube to the top of the silo will impact velocity. Actually, velocity will fall exactly when it should rise so that is less than ideal.

That leaves us with two variables as far as I can see----discharge blower pressure and blower motor torque.

Robert has already described the pressure as varying in relationship with solids loading. Maybe he could advise us on what the blower torque or current (next best thing!) does with varying solids loading.

 

[Gerry]

I've programmed & commissioned several similar systems for flour and milk powder. The blowers and rotary valves are all constant speed, DOL. Control was based on two pressure switches.
Low pressure = line empty
High pressure = line blocking/blocked

On starting delivery, the blower starts and a few seconds later, the rotary valve starts.
On ending delivery, the rotary valve stops and the blower runs until low pressure is achieved.
During delivery, if high pressure is reached, stop the rotary valve until pressure drops below high for several seconds.
...end of story.

I endorse mellis's conclusions/recommendations

1. Do not try to control the feeder valve speed. It doesn't do anything good for you. If it's too fast, the line plugs (the orginal problem) so why install controls that let it go too fast?

2. The problem with bags running out is completely unrelated to feeder valve speed. I can see where a beacon light that comes on if the discharge pressure is low and the feeder valve is running for a couple minutes would be helpful to let the operators know that the bags have run out. I can also see shutting down the system automatically if say 10 minutes go by and no one hangs another bag.

3. What happens if the silo fills up? You may need a high level switch so you can shut the feeder valve off before you plug the line. Or if the operators know the level, they can just avoid hanging enough bags to fill it. That's probably OK if the silo is relatively large.

 

[milldrone]

In order to have an effective control system, you have to be able to detect and react to the correct variables. In my view, blower discharge pressure is not the correct variable.

I'll agree with that statement.

Maybe he could advise us on what the blower torque or current (next best thing!) does with varying solids loading.

I tried this once. What I ended up with was almost the same as air pressure. The two things current was better at was resolution and responsiveness. The pressure sensors I have previously had any contact with were magnehelic brand, slow expensive, and not sensitive in the range we needed. This particular system was able to load 70,000 lbs of paper quality chips into rail cars in 25 minutes (summer time).

To automate the vfd on the feeder I had to roll my own PID. This included a slow speed increase and a fast speed decrease. One of the problems with these systems is that almost nothing can be tollerated in the material flow area in the pipe. This makes proving flow next to impossible. In older systems the feeder wears out allowing air to exit. Sometimes allowing you to build pressure but with almost no flow at the discharge end of the pipe.

I had heard of some one that used a flat "target" at the discharge end of the pipe with a strain gauge connected to it. The air and the product were supposed to deflect the target and show as strain on the gauge. Don't know if it worked

 

[xlr8]

Hi All
We had a similar problem with conveying PE (polyethylene pellets and grind) from hoppers on the main floor to hoppers unto the 5th floor hoppers above our extruders. The pneumatic lines were plugging with too much material. The pressure blower runs at full speed (not sure the rate scfm) and the rotary feeder ran at a constant speed. There is a manual slide valve above the rotary feeder which the operators would adjust for flow rate. The pressure blower has a thermocouple on the exit pneumatic line and a pressure switch on the exit line. We monitor both in the PLC, shutting down the rotary feeder on hi temp or hi press.
The system was changed to use a VFD on the rotary feeder with a PID control from the PLC. We put a CT on the blower motor and used it as the process signal. The setpoint is set and not adjustable by the operators, eliminating the possibility of the lines plugging. The destination hopper has a high level switch to control the stopping and starting of the rotary feeder. The only problem I had was starting the rotary feeder at slow speed, it could not over come the shearing of the pellets. Therefore I start the rotary feeder at full speed and the PID takes over.
I hope this helps.

 

[R_Stone]

I want to thank all of you for your input and suggestions. I printed the thread and we are having a meeting this afternoon to decide which way to fly. I will let you know how it goes.

Robert

 

[R_Stone]

Thanks guys, the torgue output seems to be the ticket for determining if material is present in the absence of a level detector.

As far as overfilling the silo, I have 4 load stands under the silo which feed a kistler morse scale. I am using an output from he scale to protect from overfill.

I have convinced the bosses to give me a few more sensors to make my life easier. I will have a level switch located about 18" inches from the transfer hopper discharge.

As far as controlling the speed of the rotary valve, I feel that this will be necessary down the road. I guess I did not make this clear, but we may change to a different densities of materials down the road and if there is an appreciable difference in weight/particle size, it will require different speeds.

I tried the torque output back to the PLC and it seems to do exactly what I need.

Once again guys THANKS..

Robert

 

[milldrone]

I tried the torque output back to the PLC and it seems to do exactly what I need.

Glad something worked. I'm presuming your torque feedback is from the rotary feeder drive and not from the blower. As you already have one strain gauge setup for the silo weight is there a posibility of another at the discharge end of the pipe? As I mentioned I have no evidence of this working but it seemed to be such a natural fit.

The target I'm envisioning would be shaped like the paddles of a pelton water wheel. Think two tea cups siameezed together at the lip edge.

The diameter would need to be about 3 to 4 pipe diameters and placed about 2 pipe diameters away from the disharge end of the pipe. And aimed so the septum (line between the cups) splits the flow from the pipe. Obviously the "target" would be suspended by the strain gauge so the axis of the thrust would deflect the gauge axis. In this case more strain = more flow. An auto zero may have to be invoked to account for material build up on the target. Some slope sheeting that was steeper than the angle of repose for your material might also help.

Another option would be to have the whole wheel mounted at the discharge and monitor the rpm via a simple prox setup. In this case higher rpm = more flow

Have fun

 

[R_Stone]

My biggest concern would be the abrasiveness of the mica. Abrasiveness, is that even a word. Anyway, even it a powdered form this mica is very abrasive. We have had to put wearbacks on piping turns and transitions. All the target type flowmeters that I have used have been in fluid or steam flow applications.

I am glad I found this place. For years I have talked about this stuff to people and they just look at me with a blank stare.

Robert

 

[milldrone]

My biggest concern would be the abrasiveness of the mica. We have had to put wearbacks on piping turns and transitions.
Robert

As I mentioned earlier I'm in the wood fiber industry. I've had to live with wear in these types of systems for 35 years. Under some conditions (think about the role moisture plays in metal destruction)I found that some surprising materials work for wearbacks. One thing I do know is welding patches on top of patches is a recipe for disaster. Wood would get forced into the seam between layers and begin to force what was left into the pipe thus reducing the diameter.

Have fun!

 

[Tom Jenkins]

It may not work for your application, but here is an interesting product:

http://www.auburnsys.com/homepage.htm

 

[R_Stone]

I know what you mean. We have a preference for chrome alloy integral wearbacks. My first experience with this type of piping was in flyash transport systems for TVA back in the mid '80s. The biggest drawback is the excessive weight. A sweeping 90 degree turn on a 6" diameter pipe would run close to half a ton.

As far as moisture is concerned, our mica is very dry and fairly easy to transport. Moisture content is about .1 to .2%. On the cenosphere material the moisture content is 10 to 50% depending upon where in the world we get our raw materials. If it is barged in from China it is on the higher end. We typically belt convey this material to the dryers and after it gets down to an acceptable level then I can transport it through piping.

Have fun yourself,

Robert


This stuff is fun if you are wired for it. I have been a gearhead all my life. My 72 year old Mom lives with us and she loves to tell stories of the alarm clocks, lamps and other stuff I used to take apart as a kid.

You have fun too.

Robert

 

[milldrone]

This stuff is fun if you are wired for it. I have been a gearhead all my life. My 72 year old Mom lives with us and she loves to tell stories of the alarm clocks, lamps and other stuff I used to take apart as a kid.
Robert

You left the door open for a smart aleck response. What percentage of the stuff you took apart went back into working order?

 

[R_Stone]

You left the door open for a smart aleck response. What percentage of the stuff you took apart went back into working order?

I thought of this one. My skills (cough) increased through the years. Mom said my first alarm clock reassembly consisted of putting all the parts back in the case. It did not work. That was age 5 or 6. By the time I was 17 I was building digital clocks and timers from scratch. Then at about age 40, 4 large tentacles were accidently fused to my spine.... No wait, that was Doc Ock..

Robert