Control Engineering

Peter Welander

Changing the value of a process variable
July 9, 2007

Our August issue will have a special supplement on wireless applications, including process plants. As I have been writing a large scale article for this, I have been considering larger aspects of instrumentation applications.

Let us not forget, wireless is nothing but a communication technology. It doesn't determine what we communicate. (Actually it does in many respects, but that's another discussion.) The same things that apply to hard wiring and fieldbus networking also apply to wireless. Those things are the cost vs. the value of adding another piece of instrumentation to obtain another process measurement.

Instrumentation providers would have you believe that the more data you can collect, the better you can understand and optimize your processes. How much will having one more pressure sensor tell you? What is the value of that reading when measured against the cost of getting it?

One extreme end of the cost range would be to put a mechanical gage at the top of a tower where it is 100+ steps above ground level and you have to send an operator up every five minutes to read it. That would be very expensive, and you would have to REALLY WANT that bit of data to justify it.

Replacing that mechanical gage with a sensor and running a 4-20 mA cable back or attaching it to a fieldbus network would be far less expensive, although if you had to add all the wiring could get pricey.

Wireless providers say, correctly most of the time, that they can make it even cheaper by eliminating the cost of the wiring. They are trying to say that when the cost of gathering data goes down, less valuable data becomes more interesting. If that one more pressure reading can generate $20 per day in higher production or reduced costs, you might consider spending $1000 to get it. How about $10 per day?

The point is, lower cost instrumentation and lower cost communication makes lower value measurements more attractive. But there will ultimately be a cut-off point. The interesting thing will be to determine where that point is. Assigning values will be the tricky part and bring out real creativity among engineers.

Posted by Peter Welander on July 9, 2007 | Comments (0)