Vh. Edwards et Wb. Howard, UTILIZE PLANT-DATA FROM EMERGENCY RELIEF EVENTS TO TEST DESIGN METHODS A PROPOSAL TO THE AICHE DIERS USERS GROUP, Process safety progress, 16(3), 1997, pp. 147-151
Advanced methods for the design of emergency relief systems were devel
oped under the auspices of the AlChE Design Institute for Emergency Re
lief Systems (DIERS). Rapid progress has continued, but most testing h
as been done at laboratory or small pilot plant scale. Full scale test
s of design methods for emergency relief systems are much needed, but
would be prohibitively expensive. Proposed here is a practical alterna
tive: Use existing production data from operating plants to test desig
n methods for emergency relief systems. Fortunately, although most eme
rgency pressure relief events represent plant upsets and lost producti
on, the relief system prevents a catastrophe in most process upsets. H
owever, the qualitatively successful operation of an emergency relief
system in a given upset does not represent proof of the design method
employed. More importantly, it does not guarantee that the existing re
lief system will perform successfully in every credible upset that the
process might experience in the future. Advocated here is an industry
-wide program to use data from selected process upsets to systematical
ly evaluate the effectiveness of current emergency relief system desig
n methods to quantitatively predict the performance of existing emerge
ncy relief systems. Products of this effort would include confirmation
of a significant number of existing design methods, identification of
deficiencies in current methods, and a published compendium of docume
nted and evaluated case histories of full-scale relief system performa
nce. In addition to its instructive value to workers in emergency reli
ef the compendium could be used to test new design methods. This propo
sal is offered to the AIChE DIERS Users' Group. The DIERS Users' Group
, which is comprised of over one hundred member companies, has represe
ntatives qualified to define, organize, sponsor, and execute a multi-c
ompany research program of this type. (NOTE: At the semiannual meeting
of the AIChE DIERS Users' Group in Calgary, Alberta, Canada on Septem
ber 9, 1996, this proposal was presented to and adopted by the DIERS U
sers' Group and assigned to the Case Histories Committee).