Temperature difference in a PID temperature control loop 2

[RNvd]

Hi Guys,
Our facility uses hot oil to transfer heat to several jacketed tanks which are used to heat up resin and sand mixture. Tanks have PID temperature control loops.

In one of our hotoil jacketed tank, temperature set at controller for tank mixture differ from the actual mixture (coming out of tank)by 35 degree. The temperature is sensed by a thermocouple inserted in a thermowell, which sticks out inside lower half of tank..

What could be the possible causes of this Temperature Differential.Please advise.
Thanks,

Regards,

Raza Naveed

 

[jsummerfield]

Several potential sources of error exist. The thermocouple type could be different than the cable type. One could have copper instead of thermocouple table. The thermocouple type may not be selected for the maximum accuracy for your range of temperature measurement. Thermowells add time lag. The T/C may not be bottomed in the well. The elements can burn out or have a poor junction between the different wire types. BTW, thermocouples are not as accurate as resistance temperature elements.

John

 

[sfi]

And don't forget about the cold junction temperature.

 

[Mike2468]

Also:
is the mixture at a uniform temperature? A heavy viscosity sand/resin mixture will be warmest near the tank's wall (where the heating jacket is) than elsewhere.

If the tank thermocouple has another junction (maybe a nicked cable, maybe an improper connector, using copper extension wire instead of thermocouple wire), then you can read the average of two temperatures (example: tank at 160, ambient at 80 = reads 120)

A thermowell may be heatsinking the temperature from the nearby heating jacket and not providing the real temperature of the material.

Pat