Consistent multi-level energy efficiency indicators and their policy implications
Introduction
Although the energy efficiency indicator (EEI) has become a fundamental tool for understanding the impacts of various energy efficiency policy options, measuring the impact and effects of energy policies, and making meaningful cross-sector and cross-economy comparisons, the determination of commonly-agreed definitions of energy efficiency and what is meant by the term ‘energy efficiency indicators’ still forms the core of related studies. One important study on the definition of EEI is Patterson (1996), who pointed out that at least four kinds of EEIs can be used to measure energy efficiency, namely, the thermodynamic indicator, the physical-thermodynamic indicator, the economic-thermodynamic indicator, and the economic indicator. Each of these has its advantages and disadvantages (Patterson, 1996, Ang and Zhang, 2000). Among them, the physical-thermodynamic indicator and the economic-thermodynamic indicator are the two most popular tools that have been widely used in EEI-related studies, such as Ang (1994), Ang (1995), Ang et al. (2003), Ang and Lee (1996), Ang and Liu (2001), Ang and Pandiyan (1997), Ang and Zhang (1999), Ang et al. (1998), Boyd and Roop (2004), Choi and Ang (2003), Freeman et al. (1997), Greening et al. (1997), Hoekstra and van den Bergh (2003), Howarth et al. (1993), International Energy Agency (1997), Farla et al. (1997), Mukherjee (in press), Murtishaw and Schipper (2001), Natural Resource Canada (2001), Ozawa et al. (2002), Park (1992), Sun (1998), Sun and Ang (2000), Sun and Malaska (1994), Steenhof (2006), Worrell et al. (1997), and Zhou et al. (in press).
Within this vast literature, most studies have adopted the decomposition technique as the analytical tool, for which Ang (1994) and the International Energy Agency (1997) can serve as models. Two types of indices that are most frequently chosen to decompose the data (Ang 2004) are the Laspeyres index (Ang and Zhang 2000) and the Divisia index (Hulten 1973). Basically, these two types of index numbers are used in aggregation comparisons of the time-series data of economic models. The most important property concerning the use of index numbers in econometric modeling is that of consistency in aggregation, or the ‘exact property’ (Diewert 1978). However, in energy efficiency analysis, the focus is to realize the causes of changes in energy use and its performance as an input in economic production. The decomposition technique is an application of the total differential method (Stewart 2001). Due to the fact that policy-makers may not know the true pathway of energy consumption efficiency change from time A to time B in a country/region, the total differential decomposition technique has two major problems: an error in terms of missing the true solution and an error in the residual term.
Some of the studies on EEIs (e.g., Howarth et al., 1993, International Energy Agency, 1997) have focused on the cross-sectional decomposition of general indicators such as the output effect, structural effect, and intensity effect. Such studies have often failed to take into consideration the consistency between upstream1 and downstream industries during the decomposition and aggregation of effects, and they have often been unable to provide viable explanations for the residual items thus created. While computational methods cited in other studies (e.g., Ang, 1995, Greening et al., 1997) have provided comparisons of the decomposition effect and have avoided the residual items, their application to the tracing of efficiency changes between different industry levels remains problematic. In addition, the computational techniques of nearly perfect decomposition are interesting only in pure mathematics because they are of no help in identifying the true pathway of change in the use of energy. Without identifying the true pathway, all numerical approximation methods used to reassign the weights of the factors or index numbers become meaningless.
The present paper first defines and derives the economic-thermodynamic and physical-thermodynamic EEIs of secondary energy (i.e., end-use energy). The economic-thermodynamic EEIs are computed based on real GDP, while the physical-thermodynamic EEIs are computed based on the physical output volume index. Note that economic welfare is favored by most of the resource and environmental economists as the target measure of economic-thermodynamic EEIs. However, for simplicity and due to data limitations, the present paper uses real GDP (gross domestic product without an inflation effect) as a proxy for economic welfare. The two types of indicators are computed separately to avoid any possible interference from price inflation or fluctuations in the economic-thermodynamic EEIs of end-use energy efficiency. This paper also proposes the multi-level decomposition of changes in the end-use energy efficiency between the upstream and downstream industries, on which the changes and the composition thereof of indicators across multiple upstream and downstream industries are analyzed. This methodology is innovative in the sense that it echoes and embodies the IEA's pyramid concept (Fig. 1) of energy efficiency indicators in a hierarchy of industry levels. That is, the purpose of the present paper is to develop the energy efficiency relationship between upstream and downstream levels of sectors (see Fig. 1).
With the multi-level decomposition of changes in energy efficiency developed in this study, a relationship between the upstream sector and the downstream industries in terms of energy efficiency indices can be established through the factor of a ‘contributing weight’ — which is similar to the partial derivatives (weights) in a total differential equation in calculus. When a change occurs in the end-use energy efficiency of an upstream sector, the impacts of respective downstream industries on that sector's energy efficiency may be derived by multiplying the energy efficiency indices of respective downstream sectors by their contributing weights. Government officials can then use this method to identify those downstream industries with a significant influence on overall energy efficiency and focus their regulatory efforts on those industries with significant changes and impacts in terms of energy efficiency in accordance with the standards for energy audit and guidance for the purpose of policy review and improvements.
Section snippets
Mathematical presentation for end-use EEIs
The principal function of the end-use energy efficiency indicators lies in the evaluation of the secondary energy usage performance of a nation or a sector, as well as the estimation of energy conservation potential. Basically, energy can be divided into the two categories of primary and secondary energy. Since primary energy does not enter the final energy consumption market, it is very difficult to obtain compatible economic data to match the primary energy data. Therefore, some suitable
Developing multi-level end-use energy efficiency indicators
Although Eqs. (2), (3) provide a consistent base for end-use EEIs at each level in the energy efficiency indicator diagram (Fig. 1), there is a missing link between the upstream and downstream industries' EEIs. Among the studies referred to in Section 1, Ang (1995) has also specified that most studies only deal with energy efficiency decomposition within one level, which is insufficient for energy policy. Therefore, Ang (1995) proposed a multi-level method, featuring the Divisia index, where
Data description
Based on the standard classification of the production sectors in Taiwan (DGBAS, 2006), the present study divides productive activities into twenty detailed categories: fishery; other agricultural products; mining; food products; textile mill products; wood products; paper and printing products; petrochemical products; petroleum and coal products; non-metallic mineral products; basic metallic products; fabricated metallic products; electrical and electronic machinery; other manufacturing
Energy efficiency indicators and policy implications in Taiwan
End-use energy efficiency indicators are a set of analytical tools for the measurement of secondary energy usage in various industries. This paper refers to standard industry categories whereby all industries in Taiwan are divided into the three major sectors of agriculture, industry, and services, under which 9 major industries consisting of 20 sub-industries are included (see Section 4 above). Due to the limitations of the available data, the computation of the end-use energy efficiency
Conclusions and remarks
This paper has proposed the adoption of end-use energy efficiency indices and the weighted vertical effect decomposition of changes in energy efficiency indices between upstream and downstream industries, which enable policy-makers to trace and identify those downstream industries that lead to significant changes in energy efficiency in the upstream sector. Among the three major sectors in Taiwan, the industrial sector accounts for the highest percentage of overall final energy consumption in
References (37)
Decomposition of industrial energy consumption—the energy intensity approach
Energy Economics
(1994)Multilevel decomposition of industrial energy consumption
Energy Economics
(1995)Decomposition analysis for policymaking in energy: which is the preferred method?
Energy Policy
(2004)- et al.
A new energy decomposition method: perfect in decomposition and consistent in aggregation
Energy
(2001) - et al.
Inter-regional comparisons of energy-related CO2 emissions using the decomposition technique
Energy
(1999) - et al.
A survey of index decomposition analysis in energy and environmental studies
Energy
(2000) - et al.
Decomposition of industrial energy consumption: the energy coefficient approach
Energy Economics
(1996) - et al.
Decomposition of energy-induced CO2 emissions in manufacturing
Energy Economics
(1997) - et al.
Perfect decomposition techniques in energy and environmental analysis
Energy Policy
(2003) - et al.
Factorizing changes in energy and environmental indicators through decomposition
Energy
(1998)
Decomposition of aggregate energy intensity changes in two measures: ratio and difference
Energy Economics
Measuring industrial energy intensity: practical issues and problems
Energy Policy
Comparison of six decomposition methods: application to aggregate energy intensity for manufacturing in 10 OECD countries
Energy Economics
Comparing structural and index decomposition analysis
Energy Economics
Energy use and CO2 emissions in Mexico's iron and steel industry
Energy
Decomposition of industrial energy consumption
Energy Economics
What is energy efficiency? Concepts, indicators and methodological issues
Energy Policy
Some properties of an exact energy decomposition model
Energy
Cited by (29)
Data and energy efficiency indicators of freight transport sector in Mexico
2021, Case Studies on Transport PolicyDo indicators have politics? A review of the use of energy and carbon intensity indicators in public debates
2020, Journal of Cleaner ProductionCitation Excerpt :First, the availability of GDP statistics contributes to the diffusion of the use of energy intensity in cross-sector and cross-economy comparisons. According to Bor (2008), for example, energy intensity is ‘a fundamental tool for understanding the impacts of various energy efficiency policy options, measuring the impact and effects of energy policies’. Second, intensity indicators may be the basis for climate change pledges that do not fundamentally question the imperative of economic growth.
Assessment of economic and environmental energy performance of EU countries using CV-TOPSIS technique
2019, Ecological IndicatorsEnergy efficiency performance-based prognostics for aided maintenance decision-making: Application to a manufacturing platform
2017, Journal of Cleaner ProductionCitation Excerpt :Furthermore, they are may be time-dependent. It should be noticed that several approaches have been propose to support the weighting factors consideration for energy efficiency applications (energy consumption and output) at multi-level of industrial sectors (Bor, 2008; Mobtaker et al., 2012). For example, Data Envelopment Analysis (DEA) approach, (Mobtaker et al., 2012), can be used to calculate these weighting factors.