RS5000 process control...fail

Hello everybody,

New guy here. I have been a long time reader but finally found the end of the internet looking for the anwser to my question.

My current system is a dry product pneumatic conveyance system (4") that uses a rotary airlock to feed the product in closed loop system (product can return to feed tank).

The problem: System is not smart. It just turns the Rotary airlock (vfd) on full speed and when pressure spikes (15psi) it stops feeding. Waits for the pressure to fall (10psi) then hammers at it again. the result is a system that is probably not delivery at maximum eff.

Current hardware: CLX-L62. Pressure sensor. VFD on airlock. Feed tank weight, delivery tank weight.

I would like to figure out a way to use some sort of "smart" method to control this where I can feed the product at some speed (or maybe vary it slighty based on pressure) to maintain a somewhat level pressure below the set amount.

I am new to 5000, plcs and process control. I have a good handle on how to make program changes within the safety aspects of the system. What I dont have is the ablity to use the process features of function blocks. (I understand fucntion blocks). Example, PIDE, POSP, SRTP and so on.

I have probably not given enough information but apprecaite any help yall are able to offer.

If you have an extra analog output on your CLX-62, and an unused analog input in the VFD, then you can use the CLX analog out to send a speed setpoint to the VFD Rotary Airlock. Then it will run at whatever speed you choose.

For your 2nd trick, you can send the pressure to a PID instruction, scaled to be the speed setpoint, and use the PID output to control the analog output to the VFD.

The first thing I would do is create and run a trend of the pressure and the on/off state of the VFD. See how the pressure builds up, and you might get an idea of how slowing down the VFD can keep the pressure from reaching a "spike".

I'm not very familiar with pneumatic conveying. Does the pressure spike indicate a blockage ?

Is the pressure measurement taken before the airlock ?

Just from a nuts-and-bolts standpoint, how is the speed of the VFD controlled ? Is it wired to a PLC analog output, or set from the keypad, or sent over a network connection ?

Lancie1,

I currently have the pressure sensor pulled into an analog card on the rack. The drives communicate via ethernet/ip.

The problem I for see with your suggestion is that pressure will spike if too much product enters the line. If I understand you correctly, what your suggestion would further plug the line with product as they would be directly proportional. If they were inversely proportional that would be a start.

My first stab at this was with a PIDE instruction. The problem is there are large time constants between product added and pressure spiking (its builds in the line). Pressure is good, to much is bad (indicates line blockage).

I am not sure how exactly to handle large time delays.

Pressure is taken before the airlock yes.

I have run a trend and tried manually adjusting rotary speed. The problem is, to get it low enough to not spike severely impacts lb/min delivery. The pressure spikes regardless, it just does not happen to be above the "cooldown" setpoint.

The problem I for see with your suggestion is that pressure will spike if too much product enters the line.

Click to expand...

When using pressure, you must filter the input, or average it over some time period. Also you must set limit conditions, so if the pressure signal is too low or too high, you peg it to the Minimum Allowed and the Maximum Allowed. These will be numbers that you find usually by trial and error.

If I understand you correctly, what your suggestion would further plug the line with product as they would be directly proportional. If they were inversely proportional that would be a start.

Click to expand...

No problem, there is a parameter "Control Mode" on the PID that makes it reverse-acting.

The problem is there are large time constants between product added and pressure spiking (its builds in the line).

Click to expand...

That is why I suggest setting it up first with manual speed setpoint entry. When using a PID, you can adjust the P, I, and D parameters to offset some of the time lag. But the problem is that your pressure is not only proportional to the speed, but also the amount of accumulation in the line over some time period. The pressure-only parameter does not give the full picture, so it will not result in good control. Probably the line is 3/4 full before the pressure rises very much. If you had an analog level sensor in the worst-plugging spot in the line, that might give better control than the pressure.

Beware of letting the pressure get too high. This reminds me of a long-ago project using a blower and airlock to move PVC powder from a silo to a weigh hopper. The pressure got too high, line filled with power, then blew out. The white powder coated me from head to toe, and spread out in the plant in a 50-feet circle. I remember having to sweep and clean for an hour.

Blue is commanded VFD speed
Red is recieve tank weight
Purple is line pressure in PSI
White is my poor attempt at a moving average of lb/min

You can see that pressure build slowely peak by peak and them "plugs" and causes the system to pruge the line



That is pretty tough to read, but there it is.

This product is a lot like flour (sticks to everything), except if it covers you from head to toe....lets just say you will not enjoy the result....

You say that slowing down your rotary valve will impact lbs/min, but stopping your feed altogether does the same thing. Ideally, you should be able to slow the feed rate down enough that you don't stop due to high pressure. I would tend to think that if your rotary valve runs continuously it would delivery at least as much as it does while stopping.

Trying to control the rotary valve rate based on pressure will be very difficult due to the lagtime of pressure vs feed. I think most systems run at a set speed that is determined through testing. Different products will need to be run a different speeds.

Does the program has some sort of filter or delay for the high pressure shutdown. You don't want to stop based on a microsecond blip in pressure.

I would suggest you run some simple tests:
1. Run the way it is and time how long takes to convey a certain amount.
2. Set your VFD speed a little bit slower and do the same
3. Repeat and keep slowing down your speeds until the system does not stop due to high pressure.
4. Compare your lbs/min average
5. In the end, your system is only physically capable of conveying so much. Programming can maximize that amount, but cannot exceed the physical limitations.

we have done pnuematic conveying for sugar and experienced the same issue. Solution was to put the air lock on a duty cycle. So many seconds ON, so many seconds OFF. Prevents product choking the line.

There is a filter on the pressure transducer. Its a MAVE of (i think) 10 samples accross .5 sec total time.

Controlling the airlock with JUST speed does not work. I cant explain why, but the line almost requires a little build up to move product. This could be from improper sizing of blowers or piping. I dont know. when the rotary valve is slowed down, you get the same affect its just longer between each high pressure interval.

My thoughts is to attempt as Curly suggested and some sort of duty that will help keep line line at some load. I am going work with a sine wave and a filter to attempt some duty. I have also been working diligently on the PIDE to see what happens. I will report back when I know more.

Thanks again,