4.6 Article

Application of loss functions in process economic risk assessment

Journal

CHEMICAL ENGINEERING RESEARCH & DESIGN
Volume 111, Issue -, Pages 371-386

Publisher

INST CHEMICAL ENGINEERS
DOI: 10.1016/j.cherd.2016.05.022

Keywords

Loss functions; Inverted normal loss functions; Inverted Beta loss function; Hierarchical Bayesian loss function; Markov Chain Monte Carlo

Funding

  1. Vale Research Chair Grant
  2. Natural Sciences and Engineering Research Council (NSERC) of Canada

Ask authors/readers for more resources

Loss functions describe the economical consequences of the deviations from the target values. In recent years they have been used in wide range of application including process safety assessment. This paper provides a novel analysis to assess potential loss due to process deviation. The assessed losses help to better estimate process economic risk, which in turn assist in effective process system design and operational decision-making. The analysis is presented in four different development stages: (i) loss functions focusing on simple functions; (ii) loss functions with estimated maximum loss; (iii) loss functions focusing on probability distributions; and (iv) loss functions in which both distributions of variables and their dependencies are considered (i.e., hierarchical Bayesian based loss functions). Details discussion on development stages three and four are presented with case studies. First case study demonstrates application of inverted probability distribution and while second case study provides application of the hierarchical Bayesian loss functions. Advantage and disadvantages of different types of loss functions are also discussed. Finally, future research directions have been proposed. (C) 2016 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.6
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available