DCS temperature trend
Thread Content
For HaliTech’s DCS system, can programming be used to determine whether the temperature is on an upward or downward trend? How is that determined? One of our company’s batch reactors has severe heat insulation around the thermometer probes to prevent corrosion, which results in a delay of 10–20 minutes in the temperature readings. How can we control the temperature when heating is required to supply additional heat for the reaction? For example, the normal kettle temperature should be between 65-75℃ ； After the feeding process is completed, the temperature of the reactor needs to be increased. At this point, the instrument shows an upward trend in temperature; when the instrument indicates 65°C, the actual temperature inside the reactor has already reached 80°C℃ ； Then another cold material is added. Due to the lag of the thermometer, the temperature gauge in the reactor shows 75°C at this point; in reality, the temperature inside the reactor has already started to drop. When the temperature inside the reactor drops to the same level as what the gauge shows, and the gauge continues to show a decreasing temperature, for example, if the gauge reads 68°C, the actual temperature inside the reactor is already below 65°C.
The lag in temperature is a problem; it’s not due to delays in the instruments, but rather in the reaction systems. Each reaction system is different, so if there is experience with manual operations, it should be possible to program it accordingly. For example, with the heating system, it needs to be heated to 70 degrees; once the gauge shows 70, heating stops. Of course, the temperature will rise again to over 70 after a while. If, based on experience, the gauge reads 67 at that point, I reduce the heating intensity, and after some time the temperature will reach 70. This can be achieved in DCS, whether through simple settings or complex programming.
What was said upstairs is correct. Also, check whether the delay setting has been adjusted properly
First, the temperature response should actually be sensitive; the issue in the situation you described is mainly related to the placement of the temperature sensing points. Secondly, it can be achieved through program settings, that is, the method of program control mentioned above, to meet your actual control requirements.
However, if the temperature is on the rise, the displayed temperature will be lower than the actual temperature； The downward trend appears to be on the high side; there are bound to be fluctuations when controlling the temperature! It’s not possible to simply set a fixed temperature; now, manual intervention needs to be eliminated in order to achieve automatic temperature control
In the case of conventional PID control, the derivative time (TD) can be increased.
It is best to modify the tank by welding a ceramic thermometer to its body, but this requires significant investment. If you don’t want to invest more, then you have to develop your own algorithm for making corrections; this can only be done based on experience, and the accuracy will be limited.