Pressure/Flow Control Application

I am working on a project that is utilizing a controllogix plc to control gas pressure in a gas metering station. There are several other functions other than controling gas exit pressure but they are mainly alarms and gas production related items.

The control system is using a fisher pneumatic control valve with an I/P converter and a downstream presssure sensor for feedback or process variable. The gas in coming in varies from 30 to 70 bar pressure and the setpoint is 35 bar (yes the incoming pressure can be less that the setpoint).

Now, in addition to the primary pressure control there are a few other considerations as follows:

1. If the downstream gas demand is high enough to lower the outlet pressure below 35 bars the set point needs to be lowered to maintain constant pressure for gas available (this condition occurs when an upstream disturbance occurs which limits flow to gas metering station).

2. If the differential pressure of the filter/scrubbers exceeds a preset value the set point needs to be lowered to limit the gas flow to a point where there is an acceptable pressure drop across the filter/scrubbers (the filter/scrubbers are protected by an independent differential pressure switch which will shut down the station if preset diff pressure is exceeded).

Anyone have any suggestions/recommendations as how to approach this control application?

Hello Dave, and Welcome!

This seems to be a fairly well-defined function. You should be able to handle it with some simple comparison rungs.

1. Using Less Than instruction: If Incoming Gas Pressure is Less Than 35, then Move [Incoming Gas Pressure value] to Setpoint.

2. First, Define Preset Value for Filter Differential Pressure. Then define some increment X, by which you want to reduce the Setpoint each scan until the differential pressure is adequate. Then create a rung with a Greater Than comparison. If Preset Value is Greater Than Setpoint, then Setpoiint = Old Setpoint - X.

Yes that's a way I had not thought of - and it's simple and should work fine.
The reason I posted the question was that I was looking a cascading the the loops to create a moving setpoint based on conditions and it was getting very involved.

Thanks for you help.

Davemacx

Although you seem to have a project design already, here are some thoughts.

For my benefit, your post indicates converted inlet pressures in the range of 435 psig to 1015 psig, with the setpoint normally being 508 psig.

Being in the gas pipeline industry, I can say that for maximum reliability our company's typical meter station with regulating duties does not have a PLC control the pressure regulator directly via a PID loop. We provide natural gas with inlet pressures up to 1200 psig, and outlet pressures are set per specific need of the equipment using the gas (such as our own turbine engine or natural gas compressor), or the agreement with the customer (for a meter station serving a local gas distribution company).

The biggest concern using a PLC is power and/or hardware failure, and how to maintain a safe, controlled gas flow upon failure.

In our later installations, we use Bristol 624 II - RSP pneumatic controllers if we want a movable setpoint. The RSP (Bristol 5457) is a remote setpoint mechanism which accepts an analog input or a pulsed "raise" or "lower" discrete output from a PLC to mechanically drive the setpoint arm of the controller until the desired setpoint has been reached. The controller also has a 4-20mA setpoint feedback loop for the PLC to confirm the setpoint. Upon failure of the PLC and/or associated hardware, the controller is minimally affected.

The PLC is only "monitoring" the system until it needs to make a setpoint change. If there is a problem during a setpoint change from the PLC with the analog setpoint model, the Bristol will respond per the failure mode selected on the Bristol, but it needs power to do this. We use the pulsed "raise/lower" model, which works by driving the setpoint arm a specific selectable amount per input pulse received from the PLC. Failure of the PLC/hardware in this case results in a last-state failure mode, since the setpoint just stops changing.

For fixed setpoints, we use piloted regulators.

Using your CLX PLC for PID control of the meter station pressure, please clarify the following:

A) You write that you want to "maintain constant pressure for gas available". Do you mean you want to maintain a constant inlet/upstream pressure, or outlet/downstream pressure during the condition #1?

B) Is the station being designed for minimal or no operator input, that is - depending on the PLC for most if not all of the setpoint control?

C) After the PLC recognizes a limited inlet pressure and adjusts the setpoint accordingly, do you want the PLC to also return to previous "normal" setpoint when the lower inlet pressure condition has ended, or will you require operator input to return to higher setpoints?

I have some suggestions waiting for your answers...

Kent

Thanks for your comments Kent they were very informative.

I am not that up on pneumatic controllers, although I understand that is what has been mainly used to control pressure/flow in gas plants.

Additional information on this project:
The PLC was selected by the customer as the primary means of controlling the gas pressure. This metering station will serve a number of customers which will in turn have their own metering pressure control systems.

A) The PLC control system for this metering station is designed to maintain a constant outlet/downstream pressure of 35 bars.
The conditions that would cause the controls to limit the downstream pressure are the upstream pressure (not high enough to maintain the required downstream pressure - due to pipeline disruptions - which actually limit flow) and the pressure drop across the filter/scrubbers. Normal operation would be 35 bars outlet pressure.

B) There will operators around the clock, however, we do not anticipate they will be a lot of help in controlling the station in the initial stages. Ater some time and training we feel the operators will become familiar and comfortable with the controls and will be more interactive. The station will start out a very low flows and gradually increase as more gas users are put online.

C) In the beginning we intend to have the PLC reset outlet pressure setpoint after a given time period. As operators become more involved they will be the ones resetting the setpoint.

Hopefully this has answered you questions, if not, please respond and I will attempt to provide additional information/clarification.

Dave

After reading you post again the following is some additional information about the system we are proposing.

The PLC system or as AB calls it the PAC system is an Allen Bradley Controllogix Redundant System using ControlNet commnunications.
For those not familiar with this setup see graphic below copied from AB website.



The system we are proposing is very similar to the above except we have dual HMI's on the ControlNet network and we have only the main I/O chasis.

In addition, to having redundant PAC's and ControlNet Channels we split the I/O up so if a card failed we would still be fully functional.

The metering system includes two pressure control trains with dual pressure controllers (active and monitor units). The control valves are pneumatically operated with I/P's.

The control hardware is all operated on a 24 VDC UPS system with battery pack and dual balanced charger systems. In the event of power loss the system has a 2 hour run time minimum. The system also has a standby natural gas powered gen-set. The fuel gas system for the standby power gen-set and the instrument air system have dual source gas supplies.

After the control scheme was submitted to customer, the customer has came back and is requesting a backup pneumatic control unit to be utilized if the PLC system fails or is inoperable. I would be intersted in hearing your comments on this.

Looking forward to hearing from you.

Thanks for you help,

Dave

Sorry for the late reply. By all means, we can continue this discussion via private messages if it seems to be too much banter for the forum...


The problem as I see it for maintaining a constant outlet pressure during an upstream/inlet flow disruption is you will not be able to continue to control the outlet at a constant value UNLESS the downstream load (i.e. your customers) is reduced during the same event.

Say you start out with 70 bar inlet pressure and you are controlling the outlet at your normal pressure of 35 bar. You then have a process upset on the inlet side which limits the flow available through your station. If you reach the point where you need to limit the outlet pressure because there isn't enough flow, in your case this is 30 bar, you have already exhausted your "cushion" on the inlet side.

So your PLC reduces your setpoint, trying to get your system to provide a constant pressure. Problem is, the pressure will keep dropping and your valves will remain 100% open trying to maintain your last setpoint, because the setpoint is higher than your inlet. Without a corresponding reduction in the downstream gas usage, your pressures will continue to drop in this manner.

If having a constant, or "stable" pressure is your primary concern, you must either 1) have downstream customers reduce consumption, causing your system to return to control or 2) limit your gas flow to your customers by preserving your inlet pressure at a certain setpoint. The first would be a "manual" response, while the second is an automatic one.

The second answer is easily done in the PLC by having a second PID loop also trying to control the same pressure, using a non-changing setpoint which would be the lowest pressure you would want your inlet to ever see. You then compare the two PID outputs, and either the highest or lowest (depending on your system design - air opens or air closes your valves) controls the output to send out to your valves. The "normal" PID loop would always "win" this comparison unless you have dropped to your preset inlet limit which will cause your limiting PID loop to overrule the original and preserve your inlet pressure. Eventually, your customers will have to reduce their consumption.

If you are interested in a specific example of this second option, we need the following answers.
Are your control valves are air to open, or air to close?
0-100% PID output equals what output to your I/P, 4-20mA?
The RsLogix 5000 PIDE function might also have the ability to accomplish this, I don't know.

As for the pneumatic backup you mentioned, I think it is a good idea. For further discussion on that, again we need some more details:
Are you using instrument air for the control valves?
What will your client want to use for the pneumatic backup - instrument air also (easier) or would they want a natural gas source (harder, but doable)?

Kent

Kent Hostetler said:

Sorry for the late reply. By all means, we can continue this discussion via private messages if it seems to be too much banter for the forum...



Kent

Click to expand...

Heck no! We thrive on this stuff. The whole point of the forum is for everyone to learn and share. This stuff is on topic and even if someone such as myself doesn't work in this industry it is still and interesting read.

Sorry I can't contribute but carry on anyway.

Got real busy here also.

Thanks for the input Kent.

If we have to reduce pressure because of upstream conditions, you are right the customers will have to reduce there consumption or the pressure will drop to some point of equilibrium which will be far below the orginal set discharge pressure.

This whole plan of reducing flow by reducing outlet pressure was a knee jerk reaction to a couple of the larger customers which would not voluntarily reduce consumption during these upset conditions. The smaller customers were willing to reduce consumption, however the large consumer was using the majority of the gas. So i'm not sure how this plan will work - since the main problem is the large consumers who will go along with the plan.

I has not been determined yet, but the owner feels that the smaller consumers are less pressure sensative than the larger consumer.

The owner is suggesting a scheme reducing flow as a percentage of normal flow - say 75% or some adjustable set percentage - Operator would look at inlet pressure change and make percentage adjustment. Your suggestion of picking a minimum inlet pressure and basing outlet pressure on that preselected conditon may be a better solution. Definitely simpler. I will have to think about that scheme.

We are using dual pressure control valves - the active and monitor scheme where the active valve is controlling and the monitor valve is standby. The active valve fails open and uses air to close - instrument gas with I/P from PLC. The monitor valve fails close and uses air to open - instrument gas with I/P from PLC. From what I have seen this scheme is fairly common. The pressure reduction trains and controls are redundant.

We use instrument gas - system includes an instrument gas system that utilizes dual gas sources - normally gas from after the scrubbers would be used and in emergency conditons gas upstream of plant would be used.

Dave Mc

Could you set up three loops and a 3 way low select on the output to the valve(s).

Loop 1- Downstream pressure control
Loop 2- Upstream pressure control
Loop 3- Differential pressure control

Gone are the days of the 4160 I guess...

Hi Mike and Thanks for your input.

Probably, however, I have not set up that kind of control would you mind elaborating a little or direct me to a location where I can get more information about that kind of setup.

Dave Mc

Another control question - same setup
Setup:
Redundant Pressure control lines - each line has dual pressure control valves with fisher air actuators and I/P converters.
The two control valves are operated as an active control valve (FO) and a monitor control valve (FC) the monitor valve has a slightly set point (pressure) than the active valve which causes the monitor valve to move to the fully open operation unless the the active valve fails.
The question is:
What control steps (logic/scheme) would provide the smoothest transition when switching from one pressure control train to the other?

Anyone have any suggestions, at present we are looking at keeping the active valve PID loops in both trains in auto and putting the monitor valves in manual and incrementing one open while incrementing the other valve on a time based step rate.

Dave Mc

Kent Hostetler said:

If having a constant, or "stable" pressure is your primary concern, you must either 1) have downstream customers reduce consumption, causing your system to return to control or 2) limit your gas flow to your customers by preserving your inlet pressure at a certain setpoint. The first would be a "manual" response, while the second is an automatic one.

The second answer is easily done in the PLC by having a second PID loop also trying to control the same pressure, using a non-changing setpoint which would be the lowest pressure you would want your inlet to ever see. You then compare the two PID outputs, and either the highest or lowest (depending on your system design - air opens or air closes your valves) controls the output to send out to your valves. The "normal" PID loop would always "win" this comparison unless you have dropped to your preset inlet limit which will cause your limiting PID loop to overrule the original and preserve your inlet pressure. Eventually, your customers will have to reduce their consumption.

If you are interested in a specific example of this second option, we need the following answers.
Are your control valves are air to open, or air to close?
0-100% PID output equals what output to your I/P, 4-20mA?
The RsLogix 5000 PIDE function might also have the ability to accomplish this, I don't know.

Click to expand...

Kent, Rslogix PIDE has this option and is called Override i´m attaching an example. Sorry it is in spanish but is OK.



Hope this will help

Thanks for the input Bill.

I am going to try that setup with some modifications. This application has two items (scrubber PDIT and upstream pressure) that will override the main pressure control loop so I want to incorporate both of them in the evaluation.

More on this as the controls develop.

Dave Mc