Hi all,

In our plant, there are installed pressure compensated temp. controller in a column. It is cascaded to steam condensate flow control to control reboiler heat input to column in targeting desired bottom product composition. I found this control loop didn't work and never used by operators for a long time. They said when it worked cascaded, the condensate CV was frequently unstable. I am sure, there is something wrong with this control scheme. I hope for brief description of this control system work mechanism and how to troubleshoot the problem.

Regards,
R. Lakota

 

Hi there,

First you need to understand how the temperature in both the bottom and the top section of the column are controlled and what the effects are of these controls.

The temperature in the top section of the column is controlled by regulating the reflux flow to the column. A higher reflux flow to the column will lead to lower temperatures in the top section of the column.

The top column temperature is regulated by means of a cascade control. a TIC provides the setpoint for a FIC which monitors the reflux flow to the column and adjusts a FCV in the discharge line of the column reflux pumps until the reflux flow matches the flow remote setpoint on the FIC.

Temperature reflux control can be very difficult, because it is effected by the temperature control in the bottom section of the column. If a stable control cannot be achieved, disconnect the temperature cascaded signal to the FIC. a Setpoint for the FIC can then be provided by the CCR operator manually.

The temperature in the bottom section of the column is controlled, by controlling the steam flow through the column reboilers. Higher flow of steam will lead to higher heat transfer to the column sump and thus to a higher temperature in the column.
The temperature is regulated by means of another cascade control. a TIC provides the setpoint for a FIC which monitors the steam flow to the reboilers and adjust a FCV located in the steam feed lines of the reboilers.

If I understand you correctly this is the control that doesn’t work cascaded. I think the same thing is happening at the bottom and the two temperature controllers are fighting each other, so again if a stable control cannot be achieved, disconnect the temperature cascaded signal to the FIC. a Setpoint for the FIC can then be provided by the CCR operator manually.

Ideally you should change the control for the top and bottom to a cascaded self tuning adaptive controller setup or at least do some a balancing act with the current PID tuning sets of both top and bottom controls. To tune a cascade control loop can be very difficult and to try and tune two slow temperature controls at the same time while they keep on fighting each other might be a job for someone with a bit more control and tuning experience than you and me.

So to keep things going until you can get a tuning and control specialist out there to help you, work with the manual setpoints on the flow controllers to achieve stable temperatures. It will also be a balancing act that have to be done by the operators but temperature control is very slow so that is why it is quite possible to keep the temperature stable with manual setpoints inputs.

I have worked with one of these tuning specialist before and what they can do is very impressive and they also have all the right software and models to do the job. I would not even get close to setup a advance control system the way they can do it so forget about trying to do this yourself.

Something else from a different perspective you might want to consider, is that you can propose to management to get a tuning expert out but if production is satisfied with the way they can control the plant, you should not push the issue to hard. If this is really a problem to them and they cannot control the plant due to this problem they will start jumping up and down soon enough and you and everyone else will then know how critical this cascaded controls are to them. So basically, don’t scratch where it doesn’t itch or try and repair something that isn’t broken, since your efforts, well intended as they are, sometimes could cause bigger problems, and possibly production losses, than the current problem.
Also keep in mind the second highest priority, after safety, on any plant, is production so if production can carry on as is management will not be interested in anything that could jeopardize that, like a tuning expert that will come and upset the stability of the plant and possibly cause production losses just to make life a bit easier for the production operators.

Obviously you will know best which route to follow so these are just some issues you might want to take into consideration, which I have personally experienced on various on and offshore plants around the world, and I am sure it will not be different on your plant.