5
$\begingroup$

In an industrial jacketed reactor, we have two temperature probes:

  • one on a baffle the extends down about to the 50% fill line
  • one built into reactor, very close to the bottom valve.

When is it more reliable and practical to use one location over the other when monitoring and controlling the temperature in a chemical process? Please ignore obvious situations, such as the liquid level is too low for the baffle so the bottom must be used.

I've heard some reasons for each location. For example, baffle temperature is best when reacting at a constant temperature, or when a slurry presents possible solids buildup on the bottom probe.

Are there any solid industry best practices relating to this?

$\endgroup$
4
  • $\begingroup$ Is the jacketed reactor using steam, water, or another fluid? $\endgroup$ – jmac Jan 21 '15 at 6:59
  • $\begingroup$ Most of the time it's a special heat transfer fluid (-40 to 85 C) controlled by a TCU. If we need to go hotter, we can blow it out and inject steam directly into the jacket. $\endgroup$ – tralston Jan 21 '15 at 7:02
  • 1
    $\begingroup$ Are you bringing the heated fluid through the top of the jacket or the bottom? Also, is a higher reading or a lower reading more likely to cause quality issues with the batch (e.g. if liquid is introduced from the top, and circulated to the heater from the bottom, the bottom reading will be a lower temperature -- if you are concerned about overheating, this is a bad place to put a sensor) $\endgroup$ – jmac Jan 21 '15 at 7:08
  • $\begingroup$ The heated fluid always enters at the bottom of the jacket (which is why the bottom temperature probe often reads colder, even if just by a few degrees). When using a control loop, we'll often control it via the jacket temp (which is the jacket outlet temp, or rather the TCU return temp, out the top of the jacket, and sometimes several yards from the reactor itself, on the way back to the TCU). $\endgroup$ – tralston Jan 21 '15 at 7:17
2
$\begingroup$

When there is a change of temperature of the reactor contents, which sensor detects the change first?

The quicker the effects of a change can be detected, the more responsive the control system and (generally speaking) the smoother the control. Following this principle using whichever sensor is closest to the jacket inlet is preferred, as it will reduce the lag caused by heat transfer through the fluid bulk.

However, you should consider employing cascade control. This uses two controllers, one for the reactor temperature (primary controller) and one for the jacket temperature (secondary controller). The output of the reactor temperature is used as the setpoint for the jacket controller, and the output from jacket controller is used to control the position of the jacket valve.

This eliminates much of the heat transfer lag, and would provide more responsive control.

$\endgroup$
2
  • $\begingroup$ What about in a batch reactor? $\endgroup$ – tralston Aug 8 '15 at 2:32
  • $\begingroup$ Cascade control is frequently used for both batch and continuous tank reactors. This white paper describes a few examples. My comment about which of the two sensors in the tank you want to use is also equally valid for batch reactors. $\endgroup$ – Jim Hargreaves Aug 8 '15 at 9:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.