TABLE OF CONTENTS When I first went to work in the process industry in 1964, computer control was in its infancy. The use of computer modeling of distillation towers was just beginning. I believe I was the first person in the oil refining industry to model a crude distillation tower with:
Multiple pumparounds
Side-stream steam strippers
Multiple side-product draw-offs
Two-stage overhead system
In 1968, I was the technical support engineer on the crude distillation unit at the Amoco Refinery in Whiting, Indiana. My job was to implement a new computer control program for the automated operation of the unit. Basically, I was to convince the operators that automated, direct, on-line computer control could produce more consistent product specifications than an experienced panel board operator. One of the factors that led to my failure in this assignment was that in 1968 I was not sure if this was true. Now, in 2002, I m still not convinced that direct computer control of refinery processes is always the best approach.
The situation regarding computer modeling is somewhat more straightforward. No engineering calculation can be better than the assumptions used as a basis to carry out the calculations.
I had been hired to review the design for a new 12,000 BSD propane-propylene splitter for a chemical plant in Convent, Louisiana. The tower was intended to be a replacement for an existing 10,000 BSD propylene-propane splitter. The incremental capacity of 2000 BSD would economically justify the $3,000,000 investment for the new splitter.
