An investigation into the normalisation of water and energy usage in the brewery industry

Kirstein, J. C. ; Brent, A. C. (2017)

CITATION: Kirstein, J. C. & Brent, A. C. 2017. An investigation into the normalisation of water and energy usage in the brewery industry. South African Journal of Industrial Engineering, 28(4):14-31, doi:10.7166/28-4-1726.

The original publication is available at http://sajie.journals.ac.za

Article

ENGLISH ABSTRACT: Benchmarks are often used to assist brewers in identifying improvement opportunities; but a comparison of water and energy performances in breweries is deficient without normalising for differences between facilities. The normalisation of water and energy use was subsequently investigated, using SABMiller breweries as a case study. Drivers of water, electricity, and thermal energy usage obtained from the literature were selected, rationalised, and ranked in a Delphi survey of industry experts, and correlated with data from 64 SABMiller sites. The main drivers identified, and the data from 58 SABMiller sites, were then used to develop multi-variable linear regression (MVLR) models. The models, tested with data from six separate SABMiller sites, were able to predict water, electrical, and thermal energy usage to within a seven per cent error. By eliminating the variability in drivers within the control of brewery staff, the MVLR models were used to normalise the performance indices, and enabled direct comparisons between plants.

AFRIKAANSE OPSOMMING: Maatstawwe word dikwels gebruik om brouers te help met die identifisering van geleenthede vir verbetering; maar ’n vergelyking van water en energie vertonings in brouerye is gebrekkig sonder normalisering vir verskille tussen fasiliteite. Die normalisering van water en energie gebruik is dienooreenkomstig ondersoek, deur die gebruik van SABMiller brouerye as ’n gevallestudie. Drywers van water, elektrisiteit en hitte-energie gebruik, verkry uit die literatuur, is gekies, gerasionaliseer en ingedeel in ’n Delphi-opname van kundiges in die bedryf, en gekorreleer met die data van 64 SABMiller fasiliteite. Die belangrikste drywers is geïdentifiseer, en data van 58 SABMiller fasiliteite, is toe aangewend om multi-veranderlike lineêre regressie (MVLR) modelle te ontwikkel. Die modelle, getoets met data vanaf ses aparte SABMiller fasiliteite, is in staat om water, elektrisiteit en hitte-energie gebruik te voorspel binne ’n sewe persent fout. Deur die uitskakeling van die variasie in drywers binne die beheer van brouery personeel, is die MVLR modelle gebruik om die prestasie-indekse te normaliseer, en is die direkte vergelykings tussen fasiliteite moontlik gemaak.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/103019
This item appears in the following collections: