Outcomes from a diversity of models need to be deliberated by policy makers of developing countries to evaluate the methods of the decision-making. Various energy models have been employed in energy studies. These models include IDA (index decomposition analysis), ANN (artificial neural network) and DEA (data envelopment analysis). This paper presents a technique based on the strengths of the mentioned energy models to assist policy makers and stakeholders in industrial energy management. Changes in energy consumption patterns can be attributed to three factors, i.e. activity, structure and intensity effects. Our investigation is based on these factors. The study successfully combines IDA, ANN and DEA for an appraisal of energy consumption of 11 energy intensive industrial sectors of South Africa between 1971 and 2008 for possible industrial energy saving. (C) 2013 Elsevier Ltd. All rights reserved.