Heat Exchanger Control

Hi all,

Looking for some basic information on the control of a heat exchanger via a PLC.

I am using the Siemens S7-200 PLC...

Its the first time I am working with Heat Exchangers...

Thanks Much
BlackSlipper

The first part of basic information must be supplied by you. What do you want to 'Control' ? What are your process variables? What are your feedback devices? What are your controlled devices? What do you want to do?

The aim is to control the temperature at the output of the heat exchanger, based on the temperature if to cold then another loop should be implemented which has heated water to reduce the temp of this over cold water at the output

Soooo... you have a temperature probe, a 'Heat' Solenoid, and a 'Cool' Solenoid.

Sounds simple, just compare and outputs. Maybe with a bit of deadband.

If Temp_Out < (Desired_Temp - DeadBand) then Heat_Sol_ON
If Temp_Out > (Desired_Temp + DeadBand) then Cool_Sol_On

You have been very specific.

However, there is a thread called contrl guru www.controlguru.com. A simple heat exchanger is used for an example. There is about1 new thread a week having to do with some small part of controlling a heat exchanger. It is not plc specific.

Are you proficient with the tricon, triconex PLC

blackslipper said:

Are you proficient with the tricon, triconex PLC

Click to expand...

If you are asking me, no.

Rdrast. Your solenoids would get a workout. Does any body actually control heat exchangers this way?

Peter Nachtwey said:

Does any body actually control heat exchangers this way?

Click to expand...

If you have enough time constant (thermal mass) and a deadband in the controller it isn't that bad. After all, this is essentially the way your home thermostat works.

I was recently studying heat exchangers

I got into the things like the log mean temperature difference and the heat balance. I was looking at doing a more advance model for the controlguru site.

For ONE set of conditions the temperature band would probably work well enough. However, I would be tempted to make the dead band a varible depending on the rate of change in the temperature.

I would think the deadtime would be a killer. You couldn't tell the temperature drifted outside the deadband until the dead time has elapsed. The amount of error would be dependent on the rate of change.