skin temperature of reactor tubes?


Thread Starter


Hi i am an Imperial College student designing a furnace to heat reactor tubes and my task is to control the furnace heat( by fuel: air ratio).
My problem is that i am unable to estimate the outer skin temperature of these tubes. the outer temperature of the flames is about 1300c and so no
temperature sensor will stand this temperature. there is assumed to be some coking on the inside of the tubes and we dont know how much coking?

if anyone has any suggesstions please contact me on : "[email protected]", mailto:[email protected]
Type R and Type S thermocouples can be used up to 1600C, type C can go well over 2000C. Make sure your TC sheath can withstand the
temperature though. A better choice would be an optical pyrometer. 1300 C is really not all that hot for an industrial process.
If I understand it correctly, the objective of the designed system is to control the temperature of the fluid in the tubes or in a certain point down the flow (for some reason you mention the cocking in the tubes that makes me think in a different direction: maybe you are
designing the system, which estimates the cocking through the temperatures and the flow measurement?). If you are designing the temperature control system there are some points I would like to mention.
1. When talking about the furnace heat control by varying the fuel-air ratio you probably mean the control by varying the fuel (and air)
supply, within which there is a loop that sustains the required fuel-air ratio. The generated heat can be varied by the fuel-air ratio within a certain range but not very much, as a rule. Please clarify.
2. You want to measure the temperature of the tubes skin. Probably the purpose of this is to obtain an intermediate process variable that
can be used within the developed control strategy, for example, for creating an internal loop, which controls the tube skin temperature.
This can be helpful for the overall control but is not a must. You can design a temperature control system without measuring this intermediate variable. However, the quality of this system, generally speaking, would be lower as knowing intermediate variables is usually
3. I am not an expert in the temperature sensors but if no thermocouple is available for that purpose you can try to measure the flame
temperature with an optical (spectral) device. However, it may be too slow for the control loop and it will measure the flame temperature,
not the tube skin.
4. As I mentioned, you can design the temperature control system without measuring the tube skin temperature (again, if I correctly
understand the function of the device itself) but you have to note that this system would have a fairly large dead time (even two dead times: one is due to the heat transfer through the tubes and the other one is due to the fluid flow). Another bad factor for the control
design is that the dynamics of the fuel combustion is faster than the dynamics of the heat transfer. Here you have a system with a large dead time. When designing a PID-controller or another compensator you have to keep this fact in mind. There were some postings on such systems design on this web site.
5. Usually a certain logic is added to a fuel-air ratio control strategy (for increasing/decreasing modes).
Hope this would help.

Dr. Igor Boiko
Consulting in Control is available
(including modeling, simulations and control design)
[email protected]
Tel: 1-403-294-2745