Control loop modification

We recently completed a major debottlenecking project on one of our plants. This entailed installing new equipment in certain areas of the plant to enable a 10% increase in plant production capacity. The reactor is core to the plant and it was not modified as part of the project. Equipment on either side of this reactor was either upgraded or replaced as applicable. Since the plant was recently re-commissioned, we have been experiencing significant reactor instability in certain reactor operating modes over the last month. This instability leads to a sub-optimal reaction and reactor blockage in the worst. Production suspects that the peroxide additive system on this reactor is not working correctly. This system was not modified as part of the debottlenecking project. A concern has been raised that the peroxide control valves are not controlling accurately. In some instances, too much peroxide is dosed into the reactor despite a control room request to dose a lower quantity.

You suspect that something might be faulty with the control loop, the flow meter and/or the control valves. Two valves are involved with this control. One valve is 75% smaller than the other valve. You have also questioned if the current system is the correct technology for this application since a metering pump system (without control valves) is used on another peroxide
dosing system at Sasol. Despite this difference, the existing peroxide system did not appear to give problems before the plant was debottlenecked.

Production is expecting that this matter be resolved by the control and instrumentation team. A worst case reactor block up scenario will cost the business at least 2 days of production loss with an associated value of $10million. How would you approach this matter? Who would you involve?

You would involve your textbook!

Serious though, If you have a specific question about your homework problem ask that.

Alt answer - SCRAM the reactor and reverse the polarity on the bussard collectors to invert the tachyon pulse.

badurally said:

Production is expecting that this matter be resolved by the control and instrumentation team. A worst case reactor block up scenario will cost the business at least 2 days of production loss with an associated value of $10million. How would you approach this matter? Who would you involve?

Click to expand...

You never explained what you are doing. What does this reactor do? I am familiar with nuclear reactors but I know there are chemical ones. So what is the chemical process?

The best thing you can do is log everything you might feel is important. Control signals, valve position etc. If you are using a Rockwell PLC then make a trend and share it with the forum.

You can't expect us to guess.

removed

In general terms, if it worked before and you haven't changed anything then it should still work. So... Either you did change something (maybe without knowing it) or something is broken.

If you slowed the process down again does the reactor become stable?

Has your new equipment changed the dynamics of something critical to your process? Air pressure? cooling water flow?

Using a dosing pump for metering is always preferable to opening a valve for a time because volume flow is dependent on differential pressure as well as time. If you care about volume dosed, control volume dosed. I've seem other processes where people try and save some money by omitting the expensive dosing pumps and it's almost always a mistake.

Nick

I'd hand in my resignation so that management can employ a controls engineer who actually did his own work and learnt something at University.

And just incase your English isn't up to scratch due to lack of effort in high school, here's some helpful words in 3 of the 11 official languages back in the Motherland:

Do Your Own Work
Doen Jou Eier Werk
ukwenza umsebenzi wakho

My Zulu is a bit rusty but that should do it.