The relative importance of the steps in control system design and the order in which they should be taken depend
on the kind and size of the facility, the project components to which the design tasks are assigned, and the time and money constraints imposed No matter what the design procedure, experience has shown the importance of formu­lating and documenting certain design procedures A schedule and a set of rules must be established early in the project, even if they will be changed many times later

The design of a computer-based data-acquisition system or a computer-based control system differs in many ways from that of their analog counterparts The main reason for these differences is that a large part of the system design resides jn the computer programs This fact goes a long way toward explaining why the schedule, vendor responsibility, costing, and specification are so unlike those for analog equipment 1 3

8- 5.1 Schedule

One aspect of the difference between the two systems is apparent from examination of the control-system part of a

CONTROL EQUIPMENT Preliminary design Develop specification Bid review Manufacture

Checkout (acceptance tests)

COMPUTER PROGRAMS Develop specification Design structure Programming training Coding Debugging

Checkout (programs and process) project schedule A typical schedule (Fig 8 5) is developed by taking the target date for reactor start-up as the completion time for the control system The previous 12 to 18 months are then reserved for the major programming effort, which is on-site program development and debugging after equipment delivery Coding starts as soon as the computer model is chosen and finishes when the programs have been tried out on the computer after it has been installed in the plant Still working backward in time, a year or two is allowed for equipment manufacture and check­out The length of this period depends on the complexity of the system and how much of nonstandard hardware is to be put together by the supplier Finally, a few months are added for vendor selection and contract negotiation and a few more for writing the specification. At this point the control engineer finds that he is already behind schedule or, at best, that he will reach the specification deadline having
dangerously little information about the processes that his system is going to control

Then begins the task of shortening the various segments of the chart to produce a properly timed, but realistic, set of goals To be done right, this job requires familiarity and experience with both computer systems and reactor projects Factors that must be considered are

1 The minimum amount of plant design data needed to develop a sound specification

2 The extra control-system capacity that should be specified as a contingency against further plant design changes

3 The anticipated delays because of prevailing procure­ment policies

4 The probable system suppliers and their reputations for making accurate delivery-schedule estimates

5 The available programming personnel

6 The likelihood of a postponement of the plant start up date and the estimated number of months exten­sion

These and many other factors influence the control — system schedule, which becomes a firm guideline for acquiring, programming, and installing the equipment It will, of course, be modified from time to time as the project progresses toward completion