SUNConferences, 29th Annual Conference of SAIIE

Font Size: 
THE EFFECT OF AVERAGING; WHY STOCHASTIC OPTIMISATION APPROACHES ARE IMPERATIVE
Philip van Zyl Venter, Stephanus Esias Terblanche, Martin van Eldik

Last modified: 2018-10-09

Abstract


A common engineering practise is to implement averaging of production-related data during decision-making processes. Average values of a data set are typically applied to determine an operating set-point or the sizing of equipment for investment purposes. In a production environment, where production-related outputs are within a sufficient small standard deviation, making use of average parameter levels may prove to be satisfactory. This may, however, not be suitable towards engineering production plants where more erratic fluctuations are observed. 
This paper demonstrates how the averaging of production data in a fluctuating process environment may result in sub-optimal design and investment decisions. A case study is presented where results from an investment optimisation model, which incorporates historic profiles, are compared to possible investment outcomes based on average parameter levels. Results indicate that typical averaged production profiles may yield sub-optimal investment decisions with over-optimistic expected incomes.


Full Text: PDF