Material Handling - Feeder Motor Speed Control VIA Blower Pressure

I have an application on which I would like a few comments/ideas etc. I am not looking for specific programming here, just a concept analysis sort of thing. There are two schools of thought on this project. I will not say which side of the fence each of us is on so as not to prejudice your comments (like it really would LOL).

I have a material handling system that feeds a Bepex grinding mill. Here is a link to the mill equipment.

http://www.bepex.com/pulvocron.htm

Material is fed to the mill from an 80ft. silo which supplies a Smoot Rotary Feeder at the bottom of the silo. Here is a link for the feeder.

http://www.premierpneumatics.com/alockHD.htm

Finally we are moving the material with a Sutorbilt Model 4MP Blower which will develop about 15 PSIG using a TECO model MA7200 AC vector drive.

So what we have is a blower which goes to a transfer hopper with the rotary feeder at the bottom. From there the material blows up to the top of our silo where it is stored until the mill demands material.

We want the transmitter at the outlet of the blower to produce 7 PSIG (setpoint). We can do this by varying the speed of the motor drive. More material dumped into the lines means more back pressure. When the lines are empty the blower is about 2.14 PSIG. At 25hz on the drive we have 6.8 PSIG. There are no sensors of any type on the hopper. There are no level switches, bindicators or anything to determine that material is actually present in the hopper. Here is what one person had to say:

“You need the PSI on the blower to tell the rotary feeder when it is empty. Meaning when psi drop from 6 to 1-2 psi the feeder should sound a alarm. Let the operator know its empty or after 5 min the system automatic shuts down. For as plugging, the moment it see 8 psi the feeder stops feeding material. Let me know if this works.”

The other school of thought is that some type of level sensor is essential to this system. The reason being is that when the hopper goes empty, the controls will naturally call for more and more feed which is not there. Within a short period of time the feeder will spin up to high speed and when material is fed into the hopper, it will plug the lines.

Plugging of these lines is critical. It just can’t happen. It is too damned expensive to clean out the lines and get the system running again.

As far as feedback on the motor, the drive has several outputs that could confirm a fault or that the motor has been asked to start, but not that it is actually running.

OK, I need some spitballing here. You can probably guess where I fall on this, but I need some really good suggestions. BTW, this all goes to a Siemens 315 processor, S7 with a MP270B Siemens HMI.

Thanks in advance,

Robert

Robert,

Let me see if I understand your system correctly. You seem to be describing a dilute phase conveying system that feeds an undisclosed material from a silo to a hopper. You are using pressure (rather than vacuum) to do the transfer. The signal to begin and end feeding is completely manual (operator start/stop only). Your proposed system appears to have the blower turn off and on between feeds.

Some things you haven’t told us include the air/material separation method at the hopper (bag house, cyclone, etc.), the proposed timing between starting the blower and starting the rotary feeder and the form and make-up of the material (powdered limestone for example). You also have not told us if you are working with an existing system that is being modified or a new installation.

Even without this additional information, and assuming my system understanding is reasonably correct, it is my humble opinion that you will be in for a whole lot of trouble with lines plugging if you try to use pressure as your only feedback. The base problem is that dilute phase systems depend more on flow than on pressure. What this means is that pressure only tells you about overall restriction. It tells you nothing about how the material is moving.

If you have the freedom to do so, I suggest investigating a closed loop delivery system where the material is constantly being delivered and the unused material blown back into the silo. This type of system lets you drop material as required into your hopper (using another rotary feeder) without having to worry so much about material dropping out and plugging the line. These types of systems are especially nice when feeding multiple drops with a powdered, non-abrasive material.

Steve

Let's see, we are talking about MICA with a weight/density of 35 lbs per cubic foot. The boss says this is dense phase and the material is mica which is in a flake form and we grind it to a very fine powder. Yes it is completely manual. The operators start the system and then hang bags of material over the transfer hopper. It falls into the hopper and is fed into 3" lines by the rotary feeder. It is then blown up into the silo by the blower. The material is transfered from the hopper to the silo, then when the mill needs it it is transported by a similar system into the mill. As far as timing is concerned I have the basic system operating. However at this stage we manually select the speed on the motor drive. The boss wants to control the drive automatically.

Robert

I think you may be both re-inventing the wheel here, and mis-interpreting the mechanics of the system.

A quick search of the web turned up these guys http://www.dynamicair.com/ who may be able to offer a canned system that is over all more cost effective than doing your own. I suggest you also contact Gardner Denver (they now own Sutorbilt) and see if they have any application guidelines that you can use.

I don't know a whole lot about conveying systems, but I tend to agree with Steve that velocity is more critical than pressure in this application. Controlling pressure by cutting speed and air flow could easily result in material dropping out in your piping.

One last caution - with a PD blower you can get into trouble with cutting the speed below about 50%. Not only will the motor, a constant torque load, start to overheat, but as speed drops the discharge temperature of the blower rises, and this can cause blower mechanical problems.

Actually the vendor did spec the blower, piping etc . and that part works fine. It is the control of the speed of the rotary feeder and thereby the pressure in Auto that is the issue. They would like a PID setup to control the blower.

Robert

R_Stone said:

We can do this by varying the speed of the motor drive. More material dumped into the lines means more back pressure. When the lines are empty the blower is about 2.14 PSIG. At 25hz on the drive we have 6.8 PSIG.it will plug the lines.

Click to expand...

More detail please. Does this mean you are varying the speed of the blower drive, or the feeder drive, or both? I'm having trouble understanding why there is a VFD on the blower drive. Perhaps you can explain the concept.

R_Stone said:

Plugging of these lines is critical. It just can’t happen. It is too damned expensive to clean out the lines and get the system running again.

Click to expand...

I've worked in the wood fiber industry and experenced this phenomenom.

R_Stone said:

As far as feedback on the motor, the drive has several outputs that could confirm a fault or that the motor has been asked to start, but not that it is actually running.

Click to expand...

Is this the feeder drive or the blower drive? On page 3-63 of the operator's manual (TECO model MA7200)it looks like "00 drive running" and "11 torque detection" are avalible for "configurable output" contacts.

Questions: What is the minimum feeder speed that will trigger a change in pressure? What type of comunication exists between the plc and the VFD's? Profibus, modbus, hard wired inputs and outputs?

I absolutely agree with Tom that a canned system from a company that does this for a living would be the best way to go. I really never intended to suggest a roll-your-own system, but I do see how it sounded that way.

In any case, I tend to question the statement that you are working with a dense phase system. Unless I mis-read you here, having material drop out of the air flow while in the line is undesireable. In fact, the fact that a rotary feeder is used seems to lend credibility to the dilute phase idea.

None the less, I still don't favor the pressure feedback method over some sort of level sensor. It looks to me like you will, at best, only be able to detect plugs after they happen.

Steve

Ahh, but the system is installed. It mwas all speced and decided months before they hired me so I gotta live with it. The blower is not variable. In the original design there was no speed control. When we started up the first time, the motor turned way too fast and plugged the lines. so we put a VFD on there to slow it down and get a better handle on the control.

OK, here is what happens:

The operator presses start on the HMI. The baghouse attached to the hopper starts to avoid dusting. Then, the transport blower starts. When the blower is up and stable and everything lines up (40 sec.) then the rotary feeder starts turning dropping material into the line. The material is blown up into the silo. When we shut down, the rotary feeder stops first, then I give it 2 minutes to purge completely then I shutdown the blower and the baghouse.

Robert

Actually, this company has been doing material handling/crushing/grinding/transport for 40 years. This is an issue of controls, not the basic mechanical concepts of the system. The system transports the material just fine.

The point is that one faction wants to control the speed of the ROTARY FEEDER to control the line pressure and the other faction believes that you need a few additional basic things. A level switch like a vibrasonic or maybe even a radar xmitter, and a prox on the motor gearing to confirm motor rotation. Whaddaya think?

Robert

OK - some data is required. First, has someone actually tested and recorded data to verify that the pressure drops if the rotary feeder is empty? Has some one been able to correlate pressure to material flow rate?

I'm confused by references to "drive". Is this a new VFD on the rotary feeder, and not the VFD you added to the blower motor?

Finally, I don't understand the concept that running the blower too fast, i.e. too much air flow and velocity, caused plugging of the lines. This seems counter-intuitive.

Let's see, I am not great at expressing myself through the keyboard.

The VFD is on the motor for the rotary feeder at the bottom of the transfer hopper. The blower has no vfd, just a starter. It blows air through the rotary feeder to blow the material to the silo.

OK, say we are transporting material. The operators get busy and don't hang a bag over the transfer for a while. The hopper runs empty. Since there is no level sensor, the PLC does not know this.

With the blower running and all lines empty, the pressure is 2.14 PSIG. When the drive is at 25hz the pressure is right at 7 PSIG which is about optimum.

So, we a running and everything is good. The hopper runs empty. The pressure drops. The PLC sees this and increases the motor speed to try to bring the pressure up by dropping more material into the line. But, there is no material to drop. Pressure stays low and the controller keeps increasing the speed trying to increase the pressure. The operators finally hang a bag on the hopper. Since the motor is running fast, as soon as they drop material into the hopper the lines will plug.

BRB. I gotta go look at a scale.

Robert

By they way if anyone would like to talk voice, here is my info:

Robert Stone
Sphere One, Inc.
423-629-7160 ext. 114

Thanks y'all

Here is a Protool pic of this area of the plant. Hope this works OK.

Robert

Looks to me like there might be another issue/need here behind the scene...

Is the system sized such that the rotary feeder continuously feeding material while running continuously at its lowest "acceptable" speed would feed faster than the silo is using the material? If so, why would you need to speed it up? If not, wouldn't that indicate an incorrectly sized system?

By any chance are we also looking to detect possible bridging problems inside the hopper?

Steve

Ok now it's better defined.

I also see in the TECO model MA7200 operator's manual that (page3-63)

"23 Torque Detection 1, Contact B Tq. Detect 1 NC_Cont ON : Torque detection 1 detected, (Contact B)
24 Torque Detection 2, Contact A Tq. Detect 2 NO_Cont ON : Torque detection 2 detected, (Contact A)
25 Torque Detection 2, Contact B Tq. Detect 2 NC_Cont ON : Torque detection 2 detected, (Contact B)"

Are also available for counfiguring outputs, so you can detect the movement of the rotary feeder.

Do you have some reliable data on feed speeds?
1. lowest speed that will cause a measureable pressure increase over an empty pipe.
2. The highest speed a "full" rotary airlock that will yield 7 psi. ( You appear to have this one already, 25 hz)

Can you roll your own PID loop that will "go to sleep" when the pressure drops to 2.14 psi, and limit the feeder speed to #1 above. Then when the pressure rises to that safe level go back to an auto pid loop. Also can you limit the feed speed to that 25 hz you mentioned earlier?

I see Steve has posted about bridging. Are you sure the plugging is not a result of flow after bridging?