468313 Novel Data Envelopment Analysis Approach for Sustainability Assessment: Application to Electricity Generation Technologies

Thursday, November 17, 2016: 12:55 PM
Union Square 3 & 4 (Hilton San Francisco Union Square)
Angel Galán-Martín1, Gonzalo Guillén-Gosálbez1,2, Laurence Stamford3 and Adisa Azapagic3, (1)Departament d’Enginyeria Química, Universitat Rovira i Virgili, Tarragona, Spain, (2)Centre for Process Systems Engineering, Imperial College of Science, Technology and Medicine, London, United Kingdom, (3)School of Chemical Engineering and Analytical Science, The University of Manchester, Manchester, United Kingdom

Moving towards sustainable development implies implementing actions, projects, programs, plans and policies, which involve the balanced integration of economic, environmental, and social objectives. However, setting sustainability goals requires some knowledge and understanding of the current level of sustainability. To this end, sustainability assessment approaches can be applied for measuring the extent of progress towards sustainability and for defining quantitative targets that ensure a more sustainable development.

In this work we explore the use of data envelopment analysis (DEA) to quantify the level of sustainability attained by a system and identify targets for improvements. DEA is a non-parametric linear programming (LP) technique that measures the efficiency of a set of entities, termed decision-making units (DMUs), each transforming multiple inputs into multiple outputs (Charnes et al., 1978). By solving the DEA LP problem, it is possible to distinguish between ‘efficient units’ (i.e., efficiency score of 1) and ‘inefficient units’ (i.e., non-frontier units with an efficiency score strictly lower than 1), for which DEA provides in turn specific guidelines to become efficient by projecting them on the efficient frontier.

In the context of sustainability assessment, DEA presents two major limitations particularly critical when a wide range of economic, environmental and social indicators are considered in the analysis. First, the standard DEA makes no distinction between the units deemed efficient, which makes it difficult to select a final alternative in the absence of a ranking scheme. Secondly, DEA is very sensitive to the set of inputs and outputs, leading to very different results depending on the sustainability criteria considered in the analysis.

To overcome the limitations of the standard DEA as applied to sustainability studies, this work proposes an enhanced DEA methodology tailored to carrying out sustainability assessments. The methodology proposed integrates DEA with the concept of “order of efficiency”, originally introduced to rank Pareto optimal solutions from a set of many Pareto points (Das, 1999). In essence, the main idea behind the new methodology consists of repeating the DEA calculations iteratively for each and every possible combination of inputs/outputs, to subsequently aggregate the results into efficiency metrics that assess each sustainability dimension separately and finally determine an overall sustainability level integrating the three scores (obtained in each dimension).

The capabilities of the methodology proposed are illustrated through its application to the assessment of the electricity-generation technologies expected to play a major role in the future electricity mix of United Kingdom (i.e., nuclear, gas, coal with and without carbon capture and storage, wind, solar photovoltaics and biomass with wood and Miscanthus pellets). Each technology is defined as a DMU that is characterised by a set of economic, environmental and social sustainability indicators determined using a life cycle assessment approach by Stamford and Azapagic (2014). We consider 18 sustainability indicators: three economic (capital cost, operation and maintenance cost, and fuel cost), nine environmental (freshwater ecotoxicity, marine ecotoxicity, global warming, ozone depletion, acidification, eutrophication, photochemical smog, land occupation, and land ecotoxicity) and six social (direct employment, worker injuries, human toxicity potential, radiation, depletion of elements, and depletion of fossil fuels).

The results generated with our approach show that gas, nuclear and wind technologies are efficient in each dimension of sustainability when all the indicators within each such dimension are considered. Gas is also the most economically and socially efficient (highest economic and social efficiency scores), while nuclear is the best in the environmental dimension. From an overall sustainability point of view, the gas technology presents the highest overall sustainability efficiency (mainly because of its very good performance in the social dimension), followed by nuclear and wind. On the other hand, coal is found economically inefficient; coal, solar PV and both biomass options are environmentally inefficient; and coal and biomass with Miscanthus are found socially inefficient. For these technologies deemed inefficient in a particular sustainability dimension, an inefficiency assessment is performed which provides targets that should be attained to reach the efficiency in each specific dimension.

The enhanced DEA proposed can facilitate the transition to a more sustainable future, since it enables ranking alternatives according to the extent to which they adhere to sustainability principles; and also pinpoints in a systematic manner the main sources of inefficiency and sets targets for improvements.


Charnes, A., Cooper, W.W., Rhodes, E., 1978. Measuring the efficiency of decision making units. Eur. J. Oper. Res. 2, 429–444. doi:10.1016/0377-2217(78)90138-8

Das, I., 1999. A preference ordering among various Pareto optimal alternatives. Struct. Optim. 18, 30–35.

Stamford, L., Azapagic, A., 2014. Life cycle sustainability assessment of UK electricity scenarios to 2070. Energy Sustain. Dev. 23, 194–211. doi:10.1016/j.esd.2014.09.008

Extended Abstract: File Not Uploaded
See more of this Session: Going to a Decision Point in Sustainability Analysis
See more of this Group/Topical: Environmental Division