Precision Temperature Control 1

[brent199]

We have a 25,000 gal. tank with an aqueous solution that will be recirculated through a heat exchanger (not yet installed). Would it be feasible to control the temperature of the solution to +/- 0.5 deg F in the tank? Can anyone suggest literature that would confirm or disprove this feasibility?

Thanks

 

[ash9144]

Should be fairly easy to control at that temperature unless you adding or removing substantial heat in batch processes. Overall heat balance will give you your answer.

 

[sshep]

brent,

Perry's used to have a good process control chapter in the 5th and 6th edition, which was changed for the worse (in my opinion) starting in the 7th edition. The old chapter rightly defined a control design procedure which concentrated on defining the nature and magnitude of disturbances (and deadtime) expected. It will be very possible to control to +/-0.5F with the right instrumentation and design if the process is not subject to excessive disturbances- I think this is along the lines of ash's comment. For example, if you are trying to control temp using steam that varies unpredictably in pressure or quality then you are unlikely to be successful regardless of how much cost and effort is spent on design and instrumentation.

Once the disturbances are quantified, you will be in a better position to assess ability to meet your objective. You can also gain significant advantages using calculated variables (like duty control) and feed forward action in addition to conventional PID cascades.

More detail of your process and design proposal could help in getting an opinion from the forum.

best wishes,
sshep

 

[hacksaw]

disturbances are part of it, for the most part you have to decide how uniform you want the temperature to be. That is not a function of control accuracy, but of the degree of mixing, tank insulation, etc.

for a continuously stirred tank and with more sensors than it would be worth, +/- 0.5 F is no problem.

 

[sshep]

This is not a large tank but hacksaw makes great comments. I calc approx 8ftDx8ftH should do it- but check my calcs. The control objective determines both the equipment and control design. For example a disturbance of 1MMBTU/hr is calculated to change the pv by 5F/hr. The exchanger needs to be bigger than the disturbance, but not too big to control. Will you need to heat and cool? A bit more about the system would be helpful.

A well mixed tank would be essentual, and insulation will reduce the disturbance of ambient loss/gain. I envisoned that a pumparound system will control the temp on the pump suction using the recycle flow and exchanger outlet temp. The heat input can be reasonably calculated and controlled by such a system, but the he tightness of control is ultimately reflected in the cost. Hacksaw is also right to question the worth of setting a +/-0.5F objective.

best wishes,
sshep