Page 261 - Design of Simple and Robust Process Plants

6.6 Summary 247
± record keeping of failures and repairs
. Perform a Failure Mode and Effect Analysis (FMEA)
± To achieve a good understanding of the system, a FMEA is performed, whether
qualitative or quantitative. This provides a good insight into the operational/
automation procedures required, and identifies potential failures which are an
input for the spare parts needed to cope with availability requirements.
. Maintainability has affects in three ways:
± the time to repair, and its impact on availability
± cumulative costs
± repair quality

The different issues need to be covered by both the vendor and user, and the respon-
sibilities determined for each partner documented.
During the development of a RAM specification it should be realized that there is
no ªfree rideº. A 100% reliability and availability does not exist, and a very high
requirement is something that the user has to pay for. When we examine the power
supply, everybody prefers 100% reliability and availability. Such a requirement will
certainly be very expensive, as it would require high investment in redundancy for
generation, back-up supply, transformers, cables, and switching stations. The level
of reliability and availability is an overall optimum covering both user and supplier,
while cost will always be the final product.

6.6
Summary

Currently, reliability engineering receives an increasing amount of attention in pro-
cess design, the objective being to develop reliable yet simple and robust designs.
This can be achieved by the application of reliability engineering techniques based
on historic failure data, and the evolution to more robust equipment designs. The
following points have been discussed in this chapter:

. The basic theory of reliability
. The development of reliability flowsheets (RBDs) and analytical and Monte
Carlo solving techniques. Using these techniques, the reliability and availabil-
ity of processes can be quantified based on the historic data of the compo-
nents.
. The methodology for the development of an optimized design of a process
plant.
. An application for a process plant, with process alternatives being compared
in order to obtain an improved reliable design.
. An application of a utility plant, where several alternatives were compared.
. A good partnership between a vendor/supplier and the user is a primary
requirement to develop and maintain a RAM specification for optimal pro-
cess design.