assessing the energy efficiency performance in the german and colombian
TRANSCRIPT
ASSESSING THE ENERGY EFFICIENCY PERFORMANCE IN THE GERMAN AND
COLOMBIAN FOOD INDUSTRY
Clara Ines Pardo Martinez
University of Wuppertal, Wuppertal Institute and University of La Salle
Abstract
This study conducts a cross-country and cross-sector analysis of energy consumption and
energy efficiency in the German and Colombian food industries to a 3-digit level of
aggregation. Changes in energy efficiency were monitored using both economic and
physical indicators, which showed that the food industries of both countries improved their
energy efficiency performance. Also, the results indicated considerable variation in energy
efficiency across countries and sectors. To explain the observed variation in energy
efficiency during the sample period, it employees regression analysis, which reveals that
the variables of economical factors such as energy cost and index of production had a
positive influence on energy efficiency performance. The index of production variable has
played an important role in the increase of energy efficiency in the German food industry,
whereas the size of enterprises variable was key for the improvements in energy efficiency
in the Colombian food industry. Moreover, the technical variables factor showed that the
labour productivity variable had a positive influence in the Colombian food industry and that
capital input and electricity were key variables for the improvement of energy efficiency in
the German food industry.
Keywords: Energy efficiency, German and Colombian food industries, economical factor
and technical factor.
1. Introduction
Energy efficiency has become the first step to controlling and stabilising greenhouse gas
concentrations because it is the most cost-effective and fastest option. Hence, it slightly
improves the energy system by reducing losses and overload; it could reduce the
investments in energy infrastructure; it will help mitigate energy price increases and
volatility by easing short- and medium-term imbalances between demand and supply; and it
will also help reduce CO2 emissions and increase energy security. Additionally, energy
efficiency offers non-energy benefits, such as reducing operating costs; growth in
productivity; improvements in product quality, capacity utilisation, and worker safety; waste
reduction and pollution prevention (Pye et al., 2000; Boyd et al., 2000; UNF, 2007). Some
worldwide declarations that have recognised the importance of promoting energy efficiency
include: the 2005 Gleneagles Declaration, which expressed support for specific energy
efficiency activities and policies; the 2006 St. Petersburg declaration, which reiterated
support for existing proposals and extended discussions to improve efficiency to the energy
supply sector; and the Group of Eight (G8) countries’ commitment to a collective goal of
doubling the global historic annual rate of energy efficiency improvement to 2.5 percent per
year from approximately 2012 through 2030 in their 2007 Summit in Germany (UNF, 2007).
The measurement of energy efficiency plays an important role in the formulation,
application and evaluation of energy policy due to the fact that its measurement allows
energy use to be described, potentially saving energy, and can demonstrate the impact of
various instruments by an increase or decrease of the energy consumed.
Generally, energy efficiency is measured through energy intensity indicators, which assess
the quantity of energy required to perform an activity in physical or monetary units. Studies
at the micro-levels have been focused especially on energy intensive sectors (e.g., Ramirez
et al. (2007) studied energy efficiency trends in the Dutch energy intensive sector; Schwarz
(2008) explained the driving forces and barriers to technology diffusion in the metal
industries, with application to the German aluminium smelting industry). These studies have
analysed specific factors to explain energy efficiency performance as the impact of energy
price, and the impact of technology, among others.
Several research studies have demonstrated that the physical energy intensity indicator is a
better indicator of energy efficiency than is economic energy intensity, and physical energy
intensity is often mentioned as the most reliable indicator for providing estimates of change
in energy efficiency, (Freeman et al., 1997, Phylipsen et al., 1998, APERC, 2000).
Assessing energy intensity in terms of energy per unit of physical output in the industrial
sector has concentrated on energy-intensive sectors with a low level of aggregation, such
as steel, paper, chemical and cement (e.g., Larsson et al., 2006, Neelis et al., 2007, Farla
et al., 1997 and Azadeh et al., 2007).
In sectors with a high level of aggregation, such as the food industry, the studies of energy
intensity using physical output are limited. Hence, the studies made about energy
consumption and energy efficiency in the food industry have focused on the analysis of
energy conservation technologies (Amon et al., 2008), the adoption of industrial best
practices (Wang, 2008), the environmental implications of the food industry (Dalzell, 2000),
and the application of energy management and clean production (Muller et al., 2007, Kumar
et al., 2003, Hyde et al., 2001, Henningsson et al., 2001, Kramer et al., 1998) again
indicating the lack of attention paid to the analysis of energy use across sectors of the food
industry as well as the lack of studies that determine the factors that have affected energy
consumption and energy efficiency performance with cross-country and cross-sectoral
comparisons. In order to address deficiency; this chapter has two main goals. It first seeks
to examine in detail the energy efficiency performance using traditional indicators (energy
intensity in terms of economic and physical units) and from a production-theoretic
framework through Data Envelope Analysis (DEA) by the German and Colombian food
industries (ISEC 15). The chapter then seeks to explain the variations in measured energy
efficiency through regression analysis in terms of economic and technical factors in the
German and Colombian food industries.
The structure of this chapter is as follows. The first section briefly describes the food
industry and its importance in both countries’ economies; the second section offers a
description of the methodology and data used; the third section provides the main results of
the indicators used to measure energy efficiency; the following section contains discussion
on and results from the empirical application. Finally, conclusions are drawn in the last
section.
2. The German and Colombian food industries and energy use
In Germany and Colombia, the food sector represented about 7% and 19%, respectively, of
the total energy consumed by the manufacturing sector in the year 20051 (Destatis, 2007
and UPME, 2007). In the same year, with a total of 4,958 establishments in Germany and
1,553 in Colombia, this sector accounted for about 10% and 22%, respectively, of industrial
employment and 7.3% and 26.4%, respectively, of the industrial value added. In terms of
costs, however, energy only amounted to about 2% to 3% of gross production in the food
manufacturer industry. The food industry can be broken down into 10 three-digit ISEC2
industry sectors in accordance with raw materials (generally of animal or vegetable origin)
and their processing into food products. This industry is highly diversified and dominated by
large-scale and capital-intensive firms. Figure 1 shows the distribution of energy demanded
by the food sub-sectors in both countries.
Energy is an essential input to ensure that processes function properly and that food and
beverages are safe and can be preserved and stored under controlled conditions.
Approximately half of all energy end-use consumption is used to change raw materials into
1 It does not include agriculture and mining. 2 ISEC classifies data according to the kind of economic activity; German and Colombian statistical data are reporting with this classification.
products (process use). Boiler fuel represents nearly one-third of end-use consumption
(boiler fuel can be used to produce steam, which can have two end-uses). Moreover, food
preservation is dependent on strict temperature controls; safe and convenient packaging is
extremely important in food manufacturing and is also energy intensive (Okos, et al., 1998).
The energy sources used during the period of study by food industry were relatively
constant except for electricity and natural gas, which in both countries increased while fuel
oil and coal decreased, e.g., in 2005: in Germany and Colombia, 44% and 10%,
respectively, of the energy used by factories came from natural gas, 32% and 18%,
respectively, from electricity, 21% and 16%, respectively, from fuel oil, and 3% and 45%,
respectively, from other sources3.
Figure 1: Distribution of energy demanded by the German and Colombian food sub-sectors,
2005. (According to ISEC classification of economic activities at the 3-digit levels of
aggregation)
3. Data and methodology
3.1 Data
German energy data were taken from the annual energy balances for the food industry
published in the Use of the Environment and the Economy Report by Statisches
Bundesamt Deutschland (German Bureau of Statistics), and Colombian energy data are
3 In 2000, 29% of the German food sector’s energy came from electricity, 41% from natural gas, and 26% from fuel oil. In Colombia, 16% of the food sector’s energy came from electricity, 6% from natural gas and 15% from fuel oil.
Other food products (158)36%
Beverages (159)15%
Production of meat (151)
12%
Dairy products (155)13%
Fruit and vegetables
(153) 7%
Grain mill products, starches (156) 8%
Vegetable and animal oils ‐ fats (154) 6%
Prepared animal feeds
(157) 3%
Processing of fish (152)
1%
Germany
Other food products (158)39%
Beverages (159)15%
Meat (151) 8%
Dairy products (155)9%
Fruit and vegetables
(153) 2%
Grain mill products,
starches (156) 12%
Vegetable and animal oils ‐ fats (154) 7%
Animal feeds (157) 5% Processing of
fish (152) 2%
Colombia
published by Departamento Nacional de Estadística (Colombian Department of Statistics,
DANE) and Unidad de Planeación Minero Energética (Unit of Mines and Energy Planning,
UPME).
The data of physical production were assessed at a three digit-level of aggregation in both
countries. In Germany, the data were calculated taking into account two data sources: the
first is the annual report of Statisches Bundesamt Deutschland (German Bureau of
Statistics) for Produzierendes Gewerbe (Industrial Production), and the second was the
industrial production survey Prodcom. In the Colombian case, the data were calculated
from the annual survey of manufactures published by the Departamento Nacional de
Estadística (Colombian Department of Statistics, DANE) and through observation of the
Colombian agro-chain of the Ministerio de Agricultura y Desarrollo Rural and Instituto
Interamericano de Cooperación para la agricultura IICA (Ministry of agriculture and rural
development and Inter-American Institute for cooperation on agriculture). The advantages
of these data sources in both countries are that the reporting of data uses a uniform
methodology, the products are classified using the same coding as for the ISEC
classification of economic activities, they cover all industrial enterprises with 20 or more
employees and describe more than 4,000 products of the manufacturing sector with
independent statistics regarding the unit production of goods (Destatits, 2005 Eurostat,
2007 and DANE, 2005)4.
3.2 Methodology
Changes in energy efficiency can be monitored by examining energy use by unit of activity
(traditional measures) and energy efficiency based on energy input minimisation from a
production-theoretic framework and the use of Data Envelopment Analysis (DEA). This
chapter provides three indicators of energy efficiency and the DEA method using one of the
models developed by Mukherjee (2008) in the context of the U.S. manufacturing industry.
4 This study excludes the processing and preserving of fruits and vegetables (153).
The first indicator (EIi) measures energy use per euro of gross production (Equation 1); the
second indicator (SECi) is defined as the energy used to produce one unit of physical
product (equation 2); and the third indicator (CEIi) measures the carbon emission intensity
of generated greenhouse gases (in terms of CO2 emissions) by ton produced by each
sector i of the food industry (Equation 3)5. These three indicators enable the detailed
analysis of energy efficiency from an economic, technical and environmental approach
across sectors of the food industry.
(1)
€⁄
. . ,
€
(2)
⁄
(3)
⁄
.
Energy efficiency based on energy input minimization uses DEA analysis6 which considers
an industry producing a single output y from a vector of n inputs x = (x1, x2,…,xn). Let yi
represent output and the vector xi represent the input package of the ith DMU. Suppose
that input–output data are observed for m DMUs. Then the technology set can be
completely characterized by the production possibility set S = {(x, y):y can be produced
5 This indicator is important for the food industry because this sector is a point source of atmospheric emissions originating from fossil fuel combustion operations; 82.6% of emissions are directly linked with the use of energy (AATCC, 1999 and Maxime et al. 2006). 6 Charnes et al., (1994) and Coelli et al., (2005) may be consulted for further details and bibliographies about DEA.
from x} based on a few regularity assumptions of feasibility of all observed input–output
combinations, free disposability with respect to inputs and outputs, and convexity. If, in
addition, a constant return to scale is assumed, then it implies that all radial expansions as
well as (non-negative) contractions of the feasible input–output combinations are also
considered feasible.
The CCR DEA7 model can be used to measure energy efficiency for a DMU with input-
output package (x0, y0), through the model developed by Mukherjee (2008), where the input
vector x0 is divided to explicitly every input component –In this study: Labour (L), materials
(M) and energy (E)–. Moreover, inequalities (4b) and (4d) ensure that the other inputs not
be increased at the optimal solution and inequality (4e) ensures that the output produced is
no lower than what is actually being produced.
DEA Model:
, 4
Subject to
λ 4
λ 4
λ ß 4
λ 4
λ 4
λ 0, 1,2, … , 4g
:
λ :
7 The first development of non-parametric approach DEA was by Charnes, Cooper, and Rhodes (CCR, 1978) to measure the efficiency of individual DMUs.
4. Energy efficiency development in the German and Colombian food industries
During the period of study, energy consumption in the German food industry increased by
an average of 1.3% per year, largely due to the manufacture of other food products and
dairy products, whereas the Colombian food industry decreased its energy consumption by
an average of 1.9% per the year, mostly due to the beverages and vegetable and animal
oils and fats sectors. Energy efficiency performance was assessed using the indicators
described in the methodology section in order to analyse the relationship between energy
consumption and output, with different alternative measurements across sectors of food
industry at 3-digit levels of aggregation during the sample period. Table 1 provides average
results for energy intensity, carbon emission intensity and from the DEA model for the
German and Colombian food industries.
Table 1 Average results of energy intensity, carbon emission intensity and DEA model in
the German and Colombian food industries (3-digit level).
1998 1999 2000 2001 2002 2003 2004 2005 Average Energy intensity (E/Y = MJ/€1998)
Germany 2.61 2.56 2.43 2.43 2.36 2.32 2.31 2.28 2.41 Colombia 5.06 4.85 4.45 4.40 4.76 5.24 4.98 4.52 4.78
Energy intensity based on physical production (E/Y = GJ/Ton)Germany 1.91 1.88 1.77 1.80 1.83 1.83 1.87 1.84 1.84 Colombia 5.38 5.14 4.54 4.40 4.72 4.60 4.39 3.96 4.64 Carbon emissions intensity based on physical production (Ton CO2/Ton production) Germany 0.121 0.121 0.119 0.125 0.129 0.128 0.123 0.118 0.123 Colombia 0.166 0.159 0.136 0.131 0.140 0.138 0.134 0.123 0.141
Energy efficiency based on energy input minimisation (DEA model) Germany 0.33 0.31 0.31 0.34 0.31 0.30 0.29 0.29 0.31 Colombia 0.33 0.35 0.33 0.35 0.35 0.35 0.36 0.36 0.35
The first and second indicators, called energy intensity (EIi) and specific energy
consumption (SECi), reflect the amount of energy required per unit of output or activity. In
this study, energy intensity was measured using both economic (gross production)8 and
physical units as the output denominator.
8 Energy intensity measured as gross production is appropriate for studies that include other variables as the intermediate inputs like energy, labour and materials or that measure efficiency in terms of inputs and their relation with outputs, as is done in this study (U.S. Department of Energy).
The results of the German energy intensity indicator as gross production showed that the
manufacture of beverages and the manufacture of other foods increased in this measure
while the other sectors decreased in this measure by an average of 17%. In the Colombian
case, production, processing and preserving of meat and meat products and the
manufacture of other foods increased in this indicator whereas the other sectors decreased
this indicator by an average of 12%.
The energy intensity indicator for physical production showed by German food industry that
the manufacture of other food products (4.91 Gj/ton in 2005) and the processing and
preserving of fish (3.84 Gj/ton in 2005) were the most energy-intense sectors. For the
Colombian food industry, the processing and preserving of fish (12.53 Gj/ton) and the
processing and preserving of meat (5.21 Gj/ton in 2005) were the most intense sectors.
The results of the energy intensity indicators may suggest that the differences among the
main sectors of the food industry depend on raw materials and new trends in food
consumption. In the first case, the thermal properties of foodstuffs are key variables for
determining the process duration, the energy consumption, quality controls, hygiene
requirements and the design of equipment and process optimisation (Milles, et al. 1983 and
Earle, 2004). In the second case, New trends in food consumption have also played an
important role in the development of energy use and energy efficiency, as these changes
could increase or decrease per-unit energy consumption. According to FAO (2006) meat
and cereal consumption and the global use of fish and dairy have increased dramatically9,
and this trend concurs with the results of energy intensity in the food industries of both
countries, where these sectors are clustered as energy-intensive sectors in the food
industry.
9 World cereal consumption has more than doubled in the last 30 years, meat consumption has tripled since 1961 and the consumption of fish and fish-related products has risen by 240 percent since 1960 and is increasing at a linear rate (Mathews et al., 1999).
The CEIi indicator and energy efficiency based on energy input minimisation are alternative
measurements for assessing energy efficiency. The first measurement analysis as the fuel
mix used or inter-fuel substitution has contributed to the improvement of energy efficiency
and a decrease in CO2 emissions whereas the second measurement allows the analysis
within a production theory framework of the role of input substitutions in achieving energy
efficiency. These two measures provide additional insights to understand and determine
key factors that have affected energy efficiency development in the German and Colombian
food industries.
The results of the CEIi indicator showed that all the sectors of the food industry decreased
in this indicator during the sample period. These results concur with the results of the
energy intensity indicator regarding physical production, which demonstrates the narrow
relationship between improvements in energy efficiency and carbon emissions reduction.
It is important to note that DEA analysis clearly divides food industry between energy-
intensive sectors (EISs) and non-energy-intensive sectors (NEISs). It is also important to
note that this division concurs with the notion that the sectors requiring a cooling chain and
the application of strong and especial hygiene and quality measures10 (e.g., meat and fish)
or a process of separation or mixing that requires higher time or significant features of purity
(e.g., other food and grain mill products11) are clustered as EISs in the food industry
whereas the sectors that only require a drying process, conventional mixing or process
heating and the application of hygiene and quality measures are less stringent (e.g., the
manufacture of vegetable and animal oils, beverages and prepared animal feeds) could be
clustered as NEIS within the food industry.
10 Stringent hygiene requirements in meat and fish companies include the variable temperature as a fundamental factor in increasing food safety where cold and hot treatments (e.g., drastic change of temperature between -50°C to 150°C) are effective processes to control microbiological growth and eliminate pathogens (Dwinger et al., 2007) and bringing as a secondary consequence changes in the patterns of energy consumption in these industrial sectors. 11 Typical processes of these sectors linked to higher energy consumption include the extraction of coffee and sugar products, the air classification in cereal production and the screening and separation of various fractions of flour (Ciras et al., 2005).
5. Discussion of results
The results showed that energy efficiency in several sectors of the food industry has
improved in the period of study in both countries and that in Germany, energy intensity in
economic terms decreased at a significantly faster rate, while in Colombia, the energy
intensity in terms of physical production decreased significantly in all sectors. To determine
the causes and differences of energy efficiency results for the German and Colombian food
industries, multivariate regression analysis was conducted. The four alternative measures
of energy efficiency obtained in this chapter are used as dependent variables in the various
regression models, which use economic and technical factors as independent variables12
(see Equations 5 and 6). In the case of traditional measures of energy efficiency, the
energy intensities (measured as monetary units, physical units and CO2 intensities) are
obtained and computed from OLS. On the other hand, the energy efficiency measure
obtained from DEA analysis is estimated by the Tobit procedure, which is the appropriate
method when the dependent variable is censored.13 The results from the regression models
are reported in Tables 2 and 3.14 For each energy efficiency measure, an initial regression
was run with all economic and technical explanatory variables. A second model was then
run, retaining only those variables that were significant at the 10% level or better.15
• Economical factor:
Δ (Eq. 5)
• Technical factor:
Δ (Eq. 6)
12The variables are created from Destatis data (German case) and DANE data (Colombian case). 13 The observed efficiency score by DEA analysis is right censored at 1, as it is equal to the actual (latent) score whenever the actual score is <1. When the actual score is ≥1, the observed efficiency score =1. 14 The regression analysis was also estimated for the German and Colombian food industries together. However, the results were not robust, mainly due to the differences in the results of indicators of energy efficiency and independent variables between both countries. Therefore, in comparisons across countries with significant differences in their indicators and variables, regression analysis should be estimated for each country in order to understand the main factors that could determine energy efficiency performance. 15 Note that the Tobit procedure shows an opposite sign of the coefficients due to the DEA model’s assessing energy efficiency whereas traditional measurement measures energy intensity. This is because the former is an inverse measure of energy efficiency.
Δ ,
,
5.1 Relationship between economic factor and energy efficiency performance in the
German and Colombian food industries
Sustainable economical growth is generally considered an important precondition for the
further improvement of industry with respect to environment and energy efficiency. Further
improvement of the economy is widely expected (EC, 2005). Therefore, the analysis of the
relationship between economic factor and energy efficiency performance is important to
understand differences of results across sectors of the food industry.
Tait (2000) states that energy cost is a key variable to improve energy efficiency in the food
industry. The energy cost variable (EC) is used to determine the relationship between
energy efficiency measure and changes in energy prices. For each sector of the food
industry the share of energy cost of gross production was used. It would expect a higher
energy cost to be associated with less energy intensity and more energy efficiency.
The enterprise size variable ENSI measures the share of the manufacturing output in small
and medium enterprises (SMEs). This has an important relationship with improvement in
energy efficiency because the majority of measurements have focused on big industries,
despite small to medium enterprises’ (SMEs) having good opportunities to improve energy
efficiency, not only to save money but also to aid the image of their companies as energy-
and environmentally-responsible companies (EC, 2009). Therefore, one would expect that
a higher production in SMEs should be associated with less energy efficiency.
Another important economic variable in the analysis of energy efficiency performance is the
production level that has a relationship with economies of scale, where the growth of an
industry and the increase of its production units will have a better chance to decrease its
costs and energy consumption and increase its productivity. The index of production (IP)
variable is measured as output index for every food sector during the sample period; one
would expect that a higher value of this variable is associated positively with improvements
in energy efficiency.
Table 2 shows the estimation results for the economic factors for four measures of energy
efficiency (traditional measures with OLS and DEA measure with Tobit procedure) and for
the explanatory variables energy cost (EC), enterprise size (ENSI), and index of production
(IP) in the German and Colombian food industries. The results obtained are robust across
all four energy efficiency measures and are very similar for every country.
Table 2: Estimation results – economic factor in German and Colombian food industries
Parameter German Food industry
EIi SEC1 CEIi EE-DEA (1) (2) (1) (2) (1) (2) (1) (2)
Intercept -2.75 (0.43)
-3.46* (0.89)
-0.89** (2.46)
-0.92** (2.85)
1.27 (0.99)
EC -0.09* (1.59)
-0.11** (2.92)
-0.013** (3.59)
-0.013** (3.83)
0.01* (1.95)
ENSI 0.04* (1.73)
0.026** (2.31)
0.0002 (0.18)
-0.006* (1.69)
IP -0.019* (1.99)
-0.04*** (3.75)
-0.004*** (4.82)
-0.002* (2.18)
0.005* (1.74)
R2 0.36 0.58 0.57 0.54 Durbin Watson 2.68 2.66 2.54 2.52
F static 3.11 9.10 5.37 8.16 3.50 Log
likelihood 4.62
Parameter Colombian Food industry
EIi (€) SEC1 (Ton) CEIi EE-DEA (1) (2) (1) (2) (1) (2) (1) (2)
Intercept 9.91*** (3.76)
9.42** (3.72)
1.88 (0.29)
-4.04*** (6.90)
0.039 (0.19)
-0.12*** (6.92)
-0.29*** (0.44)
EC -0.067** (2.76)
-0.062** (2.69)
-0.055 (0.93)
-0.002 (0.83)
0.012* (1.89)
ENSI 0.058*** (7.69)
0.059*** (8.06)
0.32*** (17.39)
0.33*** (20.92)
0.01*** (17.3)
0.01*** (20.8)
-0.016*** (8.52)
IP -0.001 (0.68)
-0.009*** (4.32)
-0.01*** (4.86)
-0.0003*** (4.35)
-0.0003*** (4.87)
0.0005** (2.40)
R2 0.64 0.64 0.87 0.87 0.87 0.87 Durbin Watson 1.54 1.51 2.43 2.42 2.39 2.37
F static 40.55 61.06 149.8 224.8 147.7 222.2 46.79 Log
likelihood 32.94
t-statistics (EIi, SEC1, CEIi) and z-statistics (EE-DEA) in parenthesis. *, **, *** imply significance at the 10%, 5%, and 1% level, respectively.
In the German food industry, energy cost has a strong positive influence on energy
efficiency, but the influence of the index of production is strongly negative and that of
enterprise size positive towards energy intensity. On the other hand, in the Colombian food
industry, enterprise size variable has a strong positive influence on energy intensity, but the
influence of index of production and energy costs variables is negative.
The results of the EC variable in the German and Colombian food industries could suggest
that in this sector the energy prices have not contributed to decrease energy intensity, likely
due by the low share of energy cost (2-3%). As a result, increases in energy price should
not generate effective mechanisms to improve energy efficiency; this strategy ought to
consider which sector of the food industry should generate an impact able improve energy
efficiency. These results concur with Broder et al., 1981 in the context of food processing in
the U.S. and Patel et al., 2005 in the context of non-energy-intensive sectors in
Netherlands.
Enterprise size variable has been a key variable in the energy efficiency performance in
developing countries, whereas developed countries have not played an important role,
meaning in particular that in industrialised countries, technological levels are similar
between great enterprises and SMEs, while in developing countries, there is a higher
technological gap between great enterprises and SMEs in the food industry, demonstrating
the potential of these food industries to improve energy efficiency16.
Likewise, the results of the IP variable showed that in the German and Colombian food
industries, this variable achieves improvement in energy intensity measured as physical
output, demonstrating that several sectors of food have increased their production and have
16 SMEs in the food processing sector of developing countries have opportunities and challenges with respect to the production of non-traditional products and that of improvements in productivity, quality and technology that could indirectly increase energy efficiency performance (Wilkinson, 2004).
decreased costs and energy consumption due to the economies of scale that are quite
significant for this industry.17
It concludes the following for the economical factor in food industry: It found a positive effect
of energy cost on energy efficiency in both countries; the index of production variables is
shown to have an enhancing influence on energy efficiency in the German food industry
whereas size enterprises have played an important role in energy efficiency performance in
the Colombian food industry.
5.2 Relationship between technical factor and energy efficiency performance in the
German and Colombian food industries
The food industry involves defined production environments and changing temperature
zones, which fluctuate according to energy demands and the supply with diverse kinds of
media that have direct influence with energy consumption. Hence, it is necessary to
analyse the impact and influence of technical factors behind energy efficiency performance
in this sector. These mainly include production features, machinery used, levels of
technology implementation, characteristics of raw materials, requirements of unit
operations, and energy sources, among others.
In the food industry, it is possible to increase energy efficiency through technological
change, inter-fuel substitutions, more efficient production methods and the implementation
of best energy management practices (Persson, 2000 and EPA, 2007). These factors could
also influence the results of energy efficiency across sectors of the food industry. It includes
the variable productivity (LAPRO), measured as physical output per worker and would
expect that the higher the productivity, the higher the energy efficiency. Furthermore, to
17 Several studies have identified the importance of scale economies in the food industry, e.g., Dalzell (2000) and Wijnands et al., (2007) in the context of European food industry, Gervais et al. (2006) on Canadian food processing, and Reardon et al., 2008, in the food industry of developing countries.
evaluate inter-fuel substitutions, the variable ELE is used as the share of electricity in total
energy (fuel) consumed in every sector of the food industry.
The European Commission (2007) states that the decrease in energy consumption could
be influenced by a substitution effect caused by changes in the industrial structure and the
capital stock towards higher productivity as well as by the substitution of energy for labour
and/or other input factors. Therefore, this chapter analyses capital input as the capital-
labour ratio KL in each sector of the German and Colombian food industries; and this
variable could have either a positive or a negative coefficient.
Table 3 provides a synopsis of the estimation results with OLS and Tobit procedure for four
energy efficiency measures and for explanatory variables’ productivity (LAPRO), electricity
(ELE), and capital input (KL) in the German and Colombian food industries.
In the German case, the results showed that electricity (ELE) and capital input (KL)
variables had a strong positive influence on energy efficiency, whereas the productivity
variable (LAPRO) was statistically insignificant for the energy efficiency of the food industry.
On the other hand, in the Colombian food industry, the LAPRO variable had a strong
positive influence, whereas capital input had a strong negative influence on energy
efficiency.
Likewise, the productivity variable showed a positive influence on the energy efficiency
performance in the food industries of both countries. However, in the German food industry
it was not significant, whereas in the Colombian food industry, it was strongly significant,
chiefly due to the alignment of strategic or market competencies between national
companies and multinational companies that established themselves in Colombia during
the sample period, compelling food companies to improve their production, management
and technology models;18 this was reflected in the results of energy efficiency indicators.
Therefore, in developing countries, the increases in productivity should close the
relationship with improvements in energy efficiency, which concurs with ICC (2007) and
UNIDO (2007), which deal with the strategies to promote energy efficiency in the industrial
sectors of developing countries.
Table 3: Estimation results – technical factor in the German and Colombian food industries
Parameter German Food industry
EIi SEC1 CEIi EE-DEA (1) (2) (1) (2) (1) (2) (1) (2)
Intercept 4.06** (5.55)
3.92** (8.16)
6.19*** (11.81)
6.15*** (11.85)
0.28*** (8.47)
0.399*** (9.42)
-0.15 (0.83)
0.008 (0.098)
LAPRO -6.3E-6 (0.02)
-0.0001 (0.69)
-4.2E-6 (0.35)
0.45 (0.95)
ELE -0.072** (3.55)
-0.071** (3.64)
-0.046** (3.17)
-0.047** (3.26)
-4.2E-5 (0.05)
1.1E-5 (0.23)
KL -0.012 (0.26)
-0.30*** (8.92)
-0.30*** (9.05)
-0.017*** (7.81)
-0.022*** (8.619)
0.038*** (3.42)
0.031*** (3.69)
R2 0.26 0.26 0.55 0.55 0.49 0.48 Durbin Watson 2.16 2.23 2.84 2.85 2.44 2.45
F static 4.33 13.29 27.89 41.91 21.69 66.75 4.22 12.00 Log
likelihood 13.50 12.51
Parameter Colombian Food industry
EIi (€) SEC1 (Ton) CEIi EE-DEA (1) (2) (1) (2) (1) (2) (1) (2)
Intercept 2.83** (2.61)
-3.47 (0.63)
-2.98*** (0.56)
-0.16 (0.95)
-0.15 (0.90)
-0.039* (1.94)
LAPRO -0.004*** (10.22)
-0.014*** (6.45)
-0.02*** (9.134)
-0.0004*** (6.44)
-0.0004*** (6.61)
0.54*** (6.92)
ELE 0.15** (2.39)
0.64* (1.93)
0.63* (1.93)
0.023* (2.07)
0.022* (2.26)
-0.002* (1.66)
KL 0.017* (1.53)
0.018 (0.32)
0.001 (0.34)
-0.006* (1.68)
R2 0.63 0.57 0.56 0.50 0.50 Durbin Watson 1.58 2.33 2.37 2.34 2.37
F static 39.09 15.31 23.22 15.69 23.78 620.47 Log
likelihood 115.48
t-statistics (EIi, SEC1, CEIi) and z-statistics (EE-DEA) in parenthesis. *, **, *** imply significance at the 10%, 5%, and 1% level, respectively.
The effect of capital input was positive and significant on energy efficiency in the German
food industry, indicating what is likely a close relationship between technical progress and
capital in this sector, which has also achieved improvements in energy efficiency. German
18Iregui et al. (2006) and Villamil (2003) analyse the differences in productivity in the Colombian industrial sector, finding that the food industry has achieved important improvements in productivity due to economic liberalisation and market competence.
research institutes and centres specialising in the food industry19 have reported that during
the period of study, the food industry underwent technological changes, mostly having to do
with the compressed air system, cogeneration, the pumping system, the refining of raw
materials, pasteurisation and sterilisation techniques, use of renewable energy, extrusion
procedures, automation and check processes, all of which are in line with the results found
in both indicators’ assessments and regression analysis.20
On the other hand, in the Colombian food industry, the effect of capital input was negative,
meaning that the energy-intensive sectors of the food industries tended to be more capital-
intensive and that technical changes to save energy had secondary importance, with capital
instead generally used to save labour costs (Kander et al., 2007). However, some
Colombian food companies during the sample period made technical changes, such as
converting their boilers to natural gas, engaging in some projects of cogeneration and
renewable energy, condensed recovery, and acquisition of new factories and equipment, all
of which with the aim to achieve consolidation in the international market21.
The electricity (ELE) variable had a positive and significant influence in the German food
industry where the increasing use of cogeneration (CHP)22 is mainly considered a strategy
to improve energy efficiency in this sector,23 because this technology decreases the amount
of electricity bought but not electricity consumption. This is because the electricity
generated with this technology is consumed with higher efficiency than when the electricity
is produced or utilised from other sources. On the other hand, in the Colombian food
19According to Bundesministerium für Ernährung, Landwirtschaft und Verbraucherschutz, www.initiative-energieeffizienz.de, FEI: Forschungskreis der Ernährungsindustrie Projektdatebank: www.fei-bonn.de/projekte/projektdatenbank.html and Max Rubner - Institut: www.mri.bund.de 20 These technologies have mainly been applied in the dairy industry, the production of meat, the manufacture of grain mill products, among others; these are mainly the sectors that have managed to improve their measures of energy efficiency. 21 These change technologies have largely been made in the sectors whit the highest improvements in energy efficiency e.g., manufacture of beverages, oils and dairy products, which showed almost 50% improvement in energy intensity measured as physical output. 22 Combined heat and power (CHP) systems involve the combined production of electrical and useful thermal energy from the same energy source. 23 Cogeneration offers a substantial potential gain in efficiency with a market share of 6%, the 4th position in German, after natural gas (47%), oil (25%) and electricity 11.5% (Schulz, 2006).
industry, this variable had a negative and statistically insignificant influence on energy
efficiency, likely showing that the patterns of electricity consumption have generated no
improvements in energy efficiency. However, inter-fuel substitutions have increased the use
of natural gas due to its competitiveness in price and efficiency, the increase in
environmental regulations, and the decrease of CO2 emissions, suggesting that the
industries do not decrease electricity consumption because their new energy source had a
higher efficiency in production and cost.
It concludes for the technical factor: labour productivity had a positive influence on energy
efficiency, especially in the Colombian food industry, whereas the capital input and
electricity variables were key variables for the improvement of energy efficiency in the
German food industry.
6. Conclusions
This chapter analysed the development of energy efficiency in the German and Colombian
food industries in the time period 1998-2005 using a production-theoretic perspective and
traditional measures (e.g., energy intensity) with economic and physical production data.
The results showed that Germany increased its energy consumption by an average of 1.3%
by the final year, largely from the manufacture of other food products and dairy products,
whereas the Colombian food industry decreased its energy consumption by an average of
1.9% by the final year, mostly due to the sectors of beverages and oils.
German and Colombian food industries improved their energy efficiency and decreased
CO2 emissions, demonstrating that the trend of this sector is “make more with less energy
consumption.”
In order to determine the effects of economic and technical factors in energy efficiency
performance across sectors and countries, a regression analysis was performed in terms of
several key characteristics of the food industry. This analysis reveals that the variables of
economical factors, such as energy cost and index of production, had a positive influence in
energy efficiency; the concentration process is shown to have a positive influence on
energy efficiency in the German food industry, whereas the size of enterprises has played
an important role in energy efficiency performance in the Colombian food industry; and it
cannot identify a significant influence of investment on energy efficiency in either country.
The technical factor variables showed that the value added had a positive influence on
energy efficiency performance in the food industries of both countries; labour productivity
had a positive influence in energy efficiency, especially in the Colombian food industry,
whereas the capital input and electricity variables were key variables for the improvement of
energy efficiency in the German food industry; and the electricity variable was not
statistically significant in the Colombian food industry with regard to achieving energy
efficiency.
Finally, it concludes that in the analysis of energy efficiency in a sector with a variety of
products and a high level of aggregation, such as the food industry, requires large amount
data, that energy efficiency indicators based on physical amounts of output are preferable
to assess energy efficiency in comparisons across countries and sectors; and that the
traditional measures of energy efficiency alongside measures from a production theoretic
perspective allows additional insights into what determines energy efficiency performance
in an industrial sector.
Acknowledgement The author would like to thank Professors Dr. Werner Bönte and Dr. Irrek Wolfang for their
helpful suggestions and comments. The author is grateful for the support provided by the
Wuppertal Institute, DAAD and the University of La Salle. Any remaining errors are the
responsibility of the author.
References
Agriculture and Agri-Food table on climate change (AATCC), 1999. Final report of the food
processing industry sub-group. Competitive Analysis Centre Inc. (CACI).
Amon R., Sugar J., Hall V., Jones M., 2008. California’s Food Processing Industry Energy
Efficiency Initiative: Adoption of Industrial Best Practices. Staff Report California Energy
Commission. CEC 400-2008-006.
Asia Pacific Energy Research Centre (APERC), 2000. Energy efficiency indicators. A study
of energy efficiency indicators for industry in APEC Economies. APEC # 00-RE-0.1.7.
Azadeh, A., Amalnick M., Ghaderi S., Asadzadeh M., 2007. An integrated DEA PCA
numerical taxonomy approach for energy efficiency assessment and consumption
optimization in energy intensive manufacturing sectors. Energy Policy 35, 3792–3806.
Boyd, G.A., Hanson, D.A., Sterner, T., 1988. Decomposition of changes in energy intensity-
a comparison of the Divisia index and other methods. Energy Economics 10, 309–312.
Broder J., Booth J., 1981. Energy efficiency in food processing in the southern region.
Southern Journal of agricultural economics, 53-59.
Colombian Department of Statistics (DANE), annual publication, Survey of manufacturer
sectors. (In Spanish).
Dalzell J., 2000. Food industry and the environment in the European Union. Ed. 2.
Springer.
Earle R., 2004. Unit Operations in Food Processing, Web Edition. The New Zealand
Institute of Food Science & Technology (Inc.). www.nzifst.org.nz/unitoperations/index.htm
Dwinger R., Golden T., Hatakka M., Daelman W., 2007. A brief overview of food hygiene
legislation. Deutsche Tierärztliche Wochenschrift 114, 294-298 (In German).
Environmental Protection Agency (EPA), 2007. Energy Trends in Selected Manufacturing
Sectors: Opportunities and Challenges for Environmentally Preferable Energy Outcomes.
Final Report. Office of Policy, Economics, and Innovation Sector Strategies Division.
European Commission (EC), 2005. Energy. ATLAS Project. Market factors.
www.ec.europa.eu/energy/atlas/html/market_barriers.html
European Commission (EC), 2009. Key information for Small to Medium Enterprises
(SMEs). Directorate-General for Energy and Transport. www.managenergy.net/smes.html
Eurostat, 2007. Prodcom: Statistics on the Production of Manufactured Goods. Statistics
by product. Luxemburg. http://epp.eurostat.ec.europa.eu/portal
Farla J., Blok K., Schipper L., 1997. Energy efficiency developments in the pulp and paper
industry. A cross-country comparison using physical production data. Energy Policy 25,
745-758.
Freeman S., Niefer M., Roop J., 1997. Measuring industrial energy efficiency: practical
issues and problems. Energy Policy 25, 703-714.
Gervais J., Bonroy O., Couture S., 2006. Economies of Scale in the Canadian Food
Processing Industry. MPRA Paper No. 64.
Henningsson S., Smith A., Hyde K., 2001. Minimizing material flows and utility use to
increase profitability in the food and drink industry. Trends in Food Science & Technology
12, 75-82.
Hyde K., Smith A., Smith M., Henningsson S., 2001. The challenge of waste minimization in
the food and drink industry: a demonstration project in East Anglia, UK. Journal of Cleaner
Production 9, 57-64.
International Chamber of Commerce (ICC), 2007. Energy efficiency: a world business
perspective. The Commission on Environment & Energy. Document 213/40 rev.4
Iregui A., Melo L., Ramirez M., 2006. Regional and sectoral productivity in Colombia: an
analysis using panel data. Special Edition Productivity and Growth. Essays about Economic
Policy 25, 18-65 (In Spanish).
Kander A., Schön L., 2007. The energy-capital relation—Sweden 1870–2000. Structural
Change and Economic Dynamics 18; 291–305.
Kramer K., Moll H., Nonhebel S., Wilting H., 1999 Greenhouse gas emissions related to
Dutch food consumption. Energy Policy 27, 203-216.
Kumar S., Senanayake G., Visvanathan C., Basu B., 2003. Desiccated coconut industry of
Sri Lanka: opportunities for energy efficiency and environmental protection. Energy
Conversion and Management 44, 2205-2215.
Larsson M., Wang C., Dahl J., 2006. Development of a method for analysing energy,
environmental and economic efficiency for an integrated steel plant. Applied Thermal
Engineering 26, 1353–1361.
Maxime D., Marcotte M., 2006. Development of eco-efficiency indicators for the Canadian
food and beverage industry. Journal of Cleaner Production 14, 636-648.
Ministry of agriculture and rural development MARD and Inter-American Institute for
cooperation on agriculture IICA, 2006. Observatory of Colombian Agro-Chain.
www.agrocadenas.gov.co
Mukherjee, K, 2008. Energy use efficiency in U.S. manufacturing: A nonparametric
analysis. Energy Economics 30, 76–96.
Muller D., Marechal, F., Wolewinski T., Roux P., 2007. An energy management method for
the food industry. Applied Thermal Engineering 27, 2677-2686.
Neelis M., Patel M., Bach P., Blok K., 2007. Analysis of energy use and carbon losses in
the chemical industry. Applied Energy 84, 853–862.
Okos M., Rao N., Drecher S., Rode M., Kozak J., 1998. A Review of Energy Use in the
Food Industry. American Council for an Energy-Efficient Economy. IE.981. www.aceee.org
Patel M., Ramirez C. Blok K., 2005. The non-energy intensive manufacturing sector: An
energy analysis relating to the Netherlands. Energy 30, 749-767.
Persson C., 2000. A Sustainable Food Supply Chain. A Swedish Case Study. Oikos
Foundation for economy and ecology. www.oikos-stiftung.unisg.ch
Phylipsen, G., Blok, K., Worrell, E. 1998. Handbook on International Comparisons of
Energy Efficiency in the Manufacturing Industry, Department of Science, Technology and
Society, Utrecht University.
Pye M., McKane A. 2000. Making a stronger case for industrial energy efficiency by
quantifying non-energy benefits. Resources, Conservation and Recycling 28, 171–183.
Ramirez A., Neelis M., Patel A., Farla J., Boonekamp P., Blok K., (2007). Energy efficiency
developments in the Dutch energy-intensive manufacturing industry, 1980–2003. Energy
policy 35; 6112-6131.
Reardon T., Barret C., Berdegue J., Swinnen J., 2008. Agrifood Industry Transformation &
Small Farmers in Developing Countries. Working paper. www.aem.cornell.edu/faculty_
sites/cbb2/workingpapers.htm
Schulz E., 2006. What’s the importance of cogeneration? www.energie–fakten.de
Statistisches Bundesamt (Destatis), annual publication-a, Production in industry. Serie 4 /
serial 3.1. (In Germany)
Schwarz H., 2008. Technology diffusion in metal industries: driving forces and barriers in
the German aluminium smelting sector. Journal of cleaner production 16; 37-49.
Statistisches Bundesamt (Destatis), annual publication-b, Statistical Yearbook for the
Federal Republic of Germany. (In Germany)
Statistisches Bundesamt (Destatis), 2007. Reporting period 1991-2005-a. Use of the
Environment and the Economy. (In Germany)
Tait J., 2000. Energy efficiency and improved profitability. Chapter 3. Food Industry and the
Environment in the European Union: Practical Issues and Cost Implications. Second
Edition. Springer
Unit of mines and energy planning (UPME), National association of industry (ANDI), 2003.
Comparative analysis of international electricity prices. Industrial sector. (In Spanish)
United Nations Foundation (UNF), 2007. Realizing the Potential of Energy Efficiency.
Targets, Policies, and Measures for G8 Countries. Executive summary.
United Nations Industrial Development Organization (UNIDO), 2007. Policies for Promoting
Industrial Energy Efficiency in Developing Countries and Transition Economies.
Commission for Sustainable Development (CSD-15)
Villamil J., 2003. Productivity and technological change in the Colombian industry.
Economy and Development 2, 151-167.
Wang L., 2008. Energy Efficiency and Management in Food Processing Facilities. CRC
Press.
Wijnands H., Meulen B., Poppe K., 2007. Competitiveness of the European Food Industry.
An economic and legal assessment. European Commission. Enterprise and Industry.
Reference no. ENTR/05/75.
Wilkinson J., 2004. The Food Processing Industry, Globalization and Developing Countries.
Journal of Agricultural and Development Economics 1, 184-201.