Journal of Cleaner Production xxx (2013) 1e8

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Implementation of a Cleaner Production Agreement and impact

analysis in the grape brandy (pisco) industry in Chile
Mauricio Castillo-Vergara a, b, *, Alejandro Alvarez-Marin a, b, Susana Carvajal-Cortes a,
Sebastián Salinas-Flores a
a
Departamento de Ingeniería Industrial, Universidad de La Serena, Benavente 980, La Serena 1720170, Chile
b
EGEU Business School, Santa Beatriz N
100 of. 604, Providencia, Santiago 7500515, Chile

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this study was to determine the impact generated by the Cleaner Production Agreement
Received 12 April 2013 signed in 2004 between the Pisco Sector of the Region of Coquimbo and the National Council for Cleaner
Received in revised form Production. For this study, the pisco sector is represented by the Sociedad Agrícola Hacienda Mal Paso.
6 August 2013
The agreement considered actions associated with clean practices used in the production process, mainly
Accepted 27 September 2013
Available online xxx
the handling of waste products and the efficient use of water, among other variables. An econometric
model was used to determine the incidence of the data gathered between the years 2006 and 2011. A low
incidence of solid waste was found directly related to water resources during processing, demonstrating
Keywords:
Cleaner Production Agreement
that the handling of liquid industrial waste, along with efficient use of water, are the factors having the
Pisco industry greatest impact.
Pisco waste products Ó 2013 Elsevier Ltd. All rights reserved.
Chilean pisco
Pisco sector

1. Introduction processing, products and services, so as to increase efficiency and

A new production practice, called Cleaner Production, has been
implemented for a number of years in the region of Coquimbo, with
the aim of maximizing resources while at the same time decreasing
emissions generated by production processes.
This research project was conducted in Sociedad Agrícola
Hacienda Mal Paso, a pisco-producing company that has been
implementing these cleaner practices stipulated in the Cleaner
Production Agreement, an agreement signed between the Region of
Coquimbo and the National Council for Cleaner Production. Eval-
uation of the agreement was conducted after the year 2008, only
considering compliance of actions and goals, which have not yet
been quantified for our purposes. It should be noted that only pisco
will be analyzed, without considering other grape by-products, and
only during the manufacturing process.
Cleaner production has been defined as “the continuous appli-
cation of a preventive integrative environmental strategy for
* Corresponding author. Departamento de Ingeniería Industrial, Universidad de
La Serena, Benavente 980, La Serena 1720170, Chile. Tel.: þ56 51 2334645; fax: þ56
51 2204200.
E-mail addresses: mhcastillo@userena.cl (M. Castillo-Vergara), aalvarez@
userena.cl (A. Alvarez-Marin).
reduce human and environmental risks” (Askounes, 1994). It ad-
vocates for a voluntary emphasis in the reduction of environmental
damage, while at the same time reducing costs (Van Hoof and Lyon,
2013).
There have been many efforts to develop new models to moti-
vate cleaner production (Dieleman, 2007), as well as to promote
and implement cleaner practices (Berkel, 2007). Positive impacts
have been recorded when these cleaner practices are implemented,
both in efficiency and financial performance (Zeng et al., 2010).
Cleaner production practices represent an important ally in the
optimization of raw materials, and thereby in the conservation of
economic resources (Varón et al., 2011), while at the same time
achieving a decrease in supplies used and waste, resulting in the
improvement of operations, the economy and the competitiveness
of businesses (Medina and Medellín, 2006).
A number of countries have adopted the concept of cleaner
production as a strategy to improve environmental performance
(Kjaerheim, 2005). In Colombia, policies have been established in
order to make cleaner production practices a driving force for
development (Van Hoof and Herrera, 2007), identifying good
practices in processes to reduce risks to human health and the
environment (Pareja et al., 2010). Centers have also been created
that are responsible for technology transfer to support sustainable
production (Grutter and Egler, 2004).

0959-6526/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jclepro.2013.09.048

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
2 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8

In order to accurately assess the real costs and benefits of
implementing cleaner production practices, it is fundamental to
use the proper evaluation methods that support decision-making
processes (Bonilla el al., 2010). Diverse world-wide studies and
analyses have been conducted in various industries, including
footwear manufactures in Colombia (Londoño, 2010), a group of 71
Lithuanian companies in 18 regions (Kliopova and Staniskis, 2006),
milk farms in Chile (Nahuelhual et al., 2009), and there have been
many simulations using fuzzy logic to identify the efficiency status
of cleaner production within a company (Rodríguez, 2009). In
Denmark, fish processing plants have yielded reductions in the use
of water as well as in wastewater (Thrane et al., 2009). In the
province of Zhejiang, China, a program has been developed aimed
at pilot companies whose main goal is to reduce pollution costs,
improve environmental performance, as well as to build organi-
zational skills. The implementation of this program has produced
savings upwards of 1.1 million dollars (Hicks and Dietmar, 2007).
In New Zealand, one of the main research aims in the wine in-
dustry, which also uses grapes as raw material, has been to measure
the environmental impact of the industry by analyzing the main
waste products (pomace, lees and wastewater) (Barry, 2011). In
2011, Bosco et al. conducted a study in the wine industry to un-
derstand greenhouse gas emissions in Italy. In Spain, environ-
mental impact was evaluated using the life cycle assessment,
including economic and environmental indicators. Water use as a
variable was not included in the study, since the author did not
consider this factor to be important (Vázquez-Rowe et al., 2012).
Rugani et al. (2013) has stated that recent research activity in the
industry has been focused on improvement strategies regarding
consumption of water and energy, pesticide use, and possible
subsequent pollution effects. In Chile, a cleaner production agree-
ment in the frozen fish industry in the Region of Bio-Bío, has yiel-
ded results that have led to the analysis of the main environmental
issues (Bezama et al., 2012).
In 1997, Chile began the development of its national cleaner
production strategy, originally headed by the Ministry of Economy,
and subsequently by the National Council on Cleaner Production
(CPL for its acronym in Spanish) in 2001. Recently, the CPL estab-
establishing environmental commitments through the environ-
mental policies of each participating company, shown on Table 1.
Since the complete implementation and certification of the
agreement, two reassessments have been conducted with 100%
compliance. The pisco industry continues to comply, which has
made way for a second CPA currently in the diagnostic stage to
gather baseline information about the industry, and contemplates
the inclusion of a carbon footprint and efficient water use (Herrera,
2011).
2. The pisco industry in Chile
Pisco production is an emblematic and significantly relevant
activity in the Regions of Atacama and Coquimbo, with the majority
of grape plantations located in the Region of Coquimbo, where 94%
of the planted surface area is destined for pisco production (Banfi,
2010).
The Government of Chile has registered the Appellation of
Origin of Pisco with international organizations (Cortés, 2005),
hence the implementation of a cleaner production agreement will
significantly contribute to not only the development of the in-
dustry, but also to the development of pisco grape producers.
The economic pisco sector is currently represented by 10,500
productive hectares in the Regions of Atacama and Coquimbo,
grown by 2726 farmers, of which 85% are considered small, and
whose main means of organization is through cooperatives
(Instituto de Investigaciones Agropecuarias, 2012). The number of
pisco grape producers is 3,190, with an average farm size of 3.1 ha.
Available information additionally shows that 90.4% of producers
are located in the Region of Coquimbo, with an average farm size of
3.3 ha. Therefore, the region of Atacama, which accounts for 9.6% of
producers, represents an average farm size of 1.9 ha (Banfi, 2010).
In terms of pisco grape varieties produced, four varieties
represent over 91% of the total surface area, with three of these
Table 1
Participating production plants in the Cleaner Production Agreement.
Name of plant Company Location Region

lished its cleaner production agenda towards 2020 (CPL, 2012a). PL. Pisco Elqui Compañía Pisquera de Paihuano Coquimbo
The National Council for Cleaner Production is defined as a Chile S.A.
PL. Monte Cooperativa Agrícola Paihuano Coquimbo
platform for open dialog and collaborative action between the
Grande Pisquera de Elqui Ltda.
public sector, companies and their workers, with the aim of PL. Río Elqui Soc. Agrícola Pisquera Vicuña Coquimbo
disseminating and focusing environmental management on the Río Elqui Ltda.
prevention of pollution rather than on controlling the end result. A PL. Vicuña Cooperativa Agrícola Vicuña Coquimbo
cleaner production agreement (CPA) is defined as a voluntary Pisquera de Elqui Ltda.
PL. Hurtado Cooperativa Agrícola Hurtado Coquimbo
management tool, based on an agreement between a specific pro- Pisquera de Elqui Ltda.
ductive sector and the public sector, and looks to implement PL. Sotaquí Cooperativa Agrícola Sotaquí Coquimbo
cleaner production through specific goals and actions within a Pisquera de Elqui Ltda.
determined period of time (CPL, 2012b). PL. Sotaquí Compañía Pisquera de Sotaquí Coquimbo
Chile S.A.
The cleaner production agreement in the pisco industry came
PL. Monte Patria Compañía Pisquera de Monte Patria Coquimbo
into effect in La Serena in the year 2004, between the Pisco Sector, Chile S.A.
represented by the Association of Pisco Producers (APP) in their PL. Huamalata Soc. Agrícola Hacienda Ovalle Coquimbo
capacity as a union, and diverse institutions such as the Regional Mal Paso y Cía. Ltda.
Government, the Ministry of Economy and Energy, the Ministry of PL. Varillar Agroproductos Bauzá y Monte Patria Coquimbo
Cía. Ltda.
Health (MINSAL), Chile’s Agriculture and Livestock Service (SAG), PL. Punitaqui Cooperativa Agrícola Punitaqui Coquimbo
the Office of the Superintendent of Sanitary Services (SISS), the Pisquera de Elqui Ltda.
National Council on Cleaner Production (CPL), the Corporation for PL. Salamanca Compañía Pisquera de Salamanca Coquimbo
Production Development (CORFO), and the National Commission on Chile S.A.
PL. Salamanca Cooperativa Agrícola Salamanca Coquimbo
the Environment (Association of Pisco Producers AG, 2008), with the
Pisquera de Elqui Ltda.
aim of implementing cleaner production measures and technology PL. San Felix Agricola e Inmobiliaria Alto del Carmen Atacama
in order to increase productive efficiency and to decrease point San Félix S.A.
source pollution, as well as to promote the implementation of good PL. El Rosario Soc. Agr. Y Comercial El Alto del Carmen Atacama
operation practices, reutilization and recycling (Herrera, 2011). Rosario Ltda.
PL. Alto del Cooperativa Agrícola Alto del Carmen Atacama
The duration for the CPA was set as 31 months, during which Carmen Pisquera de Elqui Ltda.
time the pisco industry worked on the implementation,

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8
varieties belonging to the Muscat group, and according to experts,
exhibit excellent properties for the production of high-quality
pisco. The Pedro Jiménez variety, considered of lesser value for
pisco production, still occupies almost 25% of the total area
destined for pisco production. This is due to the high productivity
level, increasing the profitability level of the plantations (Catastro
Vitivinicola Nacional, 2008).
There are 13 Pisco-producing companies in the country, and
two, Compañía Pisquera de Chile and Cooperativa Agrícola y Pis-
quera Elqui Ltda. (Capel), predominate 95% of the national market
in terms of sales. There are other important mid-level and niche
market pisco companies, such as Mal Paso, Bauzá, Pisquera Río
Elqui, Horcón Quemado, El Rosario “Bou Barrueta”, Chañaral de
Carén, Ovalle, ABA, Pisquera Tulahuén, Los Nichos and Tres Cruces
(CPL, 2012c). The cleaner production agreement was comprised of
17 plants belonging to 5 companies.
The production process is regulated by Regulation 521 of the
Designation of Pisco Origin, where the basic standardized re-
quirements to be fulfilled are established, such as zone, varieties,
vinification and distillation processes, volatile acid content, among
others (Diario Oficial de la Republica de Chile, 2000).
Following is a description of the main stages in the production of
pisco starting from the cultivation of grapes to the manufacturing of
pisco and subsequent bottling (Instituto de Investigaciones
Agropecuarias, 2012).
Care and Cultivation: Productive systems used in the production
of pisco grape varieties are generally those traditionally used in the
viticulture industry, having greater care given to the watering
systems, given the semiarid conditions and lack of nutrients in the
soil that are characteristic of the geographical area. The overhead
trellis system is generally used, and the main phenological stages
are sprouting, flowering, development of fruit, ripening and
harvest.
2.1. Harvest
Harvesting season extends from February to mid-May, starting
in the Region of Atacama with Muscat varieties and ending in the
Provinces of Limari and Choapa with Pedro Jimenez and Torontel
varieties.
2.2. Grape reception
Upon reception, processing plants identify the grapes (supplier,
quantity, grade, variety and health status) in order to subsequently
transform them into juice under the required guidelines. The best
quality grapes exhibit a good balance between sugar content,
aroma and weight.
2.3. Vinification
During this stage, the sugars that are found in the grape juice are
transformed into alcohol. Prior to fermentation, grapes are crushed
and sulfur dioxide and pectolytic enzymes are generally added at
low temperature for different amounts of time depending on the
characteristics of the raw materials and the infrastructure of the
vinification plant.
2.4. Distillation
The distillation process is probably the most characteristic stage
in the production of pisco, and consists of transforming wine into
pisco alcohol. There are different types of copper stills with an
approximate capacity of 2500 L, which are heated by vapor circu-
lating through a coil. This type of distillation is not continuous,
Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
3
meaning that separation of the different volatile constituents is
done in batches.
2.5. Storage and ageing
The purpose of storing and ageing is to obtain a high-quality
pisco in terms of chemical and organoleptic properties, while
avoiding the precipitation of alcohols. It is necessary for all the
volatile components obtained in the distillation stage to combine
and harmonize.
2.6. Bottling
In the bottling stage, the resulting alcohol from the aging
process is placed into bottles. This is generally a semiautomatic
process that starts with the rinsing of new bottles which are
subsequently filled, capped and finally labeled and placed in cases
or pallets. Cases are then stored in warehouses for their subse-
quent shipping.
Throughout the pisco manufacturing process, several waste
products can be identified at each stage, as outlined in Fig. 1
(Pisquera de Chile, 2013).
3. Methods
To analyze the aforementioned factors, an econometric model
was developed that allows the visualization of any relationships or
dependence among the variables, to thereby determine the impact
generated by the final product and the effect on the environment.
Equation (1) represents the linear model of the proposed sce-
nario (Campos-Aranda, 2011).
Y ¼ b0 þ b1 X1 þ b2 X2 þ . þ bk Xk þ ε (1)
Coefficients b1, b2,., bk represent the magnitude of the effects that
explanatory variables X1, X2,., Xk may have on explanatory vari-
able Y. The coefficient b0 is the constant variable in the model. In the
model, ε is the error term (Pérez, 2006). This model is able to
predict a response in function of a broader group of variables
(Pértega and Pita, 2000). This model aims to explain the behavior of
a dependent variable (Pérez, 2006), which in this case is the con-
sumption of water, both for the plant as well as for the production
of pisco.
To determine the significance of the model, the F test of good-
ness of fit was used. The hypothesis is as follows:
H0: s2R ¼ s2E vs H1: s2R > s2E; The hypothesis H0 is rejected if
F0 > F1-a;k;n-k-1 (Montgomery and Runger, 1996).
Once the significance of the model is known, the inference test
must be conducted in order to determine the significance of each
individual parameter. The hypothesis is tested using a Student’s t
test, as follows:
H0: bj ¼ 0 vs H1 : bj s 0 for j ¼ 1,2,3,4,5; H0 is rejected if H0 > t 1-
a/2,n-k-1 (Montgomery and Runger, 1996).
Several techniques can be used to measure the adequacy of the
multiple linear regression model. The coefficient R2 is a measure of
magnitude in the reduction of variability of Y, and it is often used to
judge the adequacy of a regression model, and can vary between
the values of 0 and 1 (Montgomery and Runger, 1996). The purpose
of the DurbineWatson test is to provide a numeric parameter to
establish autocorrelation among waste products, and its main use is
to determine if the use of a single independent variable is sufficient
for a good model (Fahidy, 2006). If the value falls between 1.7 and
2.3, with a confidence interval of 95%, then there is a weak corre-
lation (Rutledge and Barros, 2002). The Akaike and Schwarz crite-
rion aids in choosing the appropriate size of the model, allowing for
4 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8

Fig. 1. Waste products in the production process.

the adjustment to a determined group of observations (Schwarz, previously obtained between 2006 and 2011 in the Huamalata
1978). processing plant by the Sociedad Agrícola Hacienda Mal Paso S.A.,
The selection of the variables used is shown in Table 2, and are during which time the cleaner production agreement was in effect.
considered to be productive indicators for the pisco industry. On The same data was requested from the remaining companies that
the other hand, the wine industry also considers these variables in participated in the agreement; however, no data was forwarded.
environmental impact evaluations (Barry, 2011; Bosco et al., 2011; Therefore, the sample consists of 24 observations, shown in Table 4.
Vázquez-Rowe et al., 2012; Gazulla et al., 2010). Water resources In order to develop a multiple linear regression model, the
are of extreme relevance, since drought is one of the main chal- variables X1, X2,., Xk must be linearly independent and must obey
lenges facing the planet in light of climate change (Terry et al., the independence or multicollinearity hypothesis (Zárate de Lara
2009). and Diaz de Zarate, 1985), that is, there must be lack of an exact
In the case of the regression model, the values to be estimated linear relationship among them. The variable X6, which corre-
are related to the variables shown in Table 3. The fundamental sponds to the amount of grapes received in kilograms, and X7,
problem was to determine if there is a relationship between water corresponding to the number of liters of vinasse applied to the
use (dependent variable) and the generation of solid and liquid fields, were excluded. Variable X6 exhibits a direct relationship with
waste (independent variables), their application in the field, as well solid and liquid waste from processing, while variable X7 was
as the amount of grapes used. Trimestral measurements were directly related with the variables vinasse and alcoholase, since the
irrigation process is dependent on these liquid waste products.
Table 2 For the treatment and modeling of the data, the Eviews 5 soft-
Description of waste products as variables. ware program was used.
Name Description Unit of
measure 4. Results
Grapes received Refers to grapes that are treated Kilograms
for the production of wine and (Kg) Regarding the analysis of the results, determination coefficient
pisco. The variety used is Muscat R2 has an adequate goodness of fit (Guerra et al., 2003). As can be
Rose and Muscat of Alexandria. seen in Fig. 2, 91.87% of the variation in water consumption can be
Application of vinasse Refers to incurred costs in fuel for Chilean Pesos
in the fields liquid waste transport (vinasse ($)
explained with the linear regression model, therefore the model is
and alcoholase), and subsequent explanatory.
road spraying. For the case in point, the value of the DurbineWatson statistic is
Lees Sediment produced during the process Kilograms 2.08, which indicates there are no problems with autocorrelation.
of fermentation due to yeast (Kg)
Using a significance level of 5%, and considering a sample size of 24
development.
Stems Vegetable fraction from the grape Kilograms observations and 5 variables in the analysis (k), an F statistic value of
bunch after removing the berries. (Kg)
Table 3
Pomace Skin from grapes left after pressing, Kilograms
Model variables.
(skins and seeds) removing all juice and pulp. (Kg)
Vinasse Industrial liquid waste produced in Liters (Lt) Variable Description
the distillation stage. Alcohol content
in wine is evaporated and condensed, Y Water consumption (m3)
and the remaining liquid is rich in X1 Stems (Kg)
nutrients. X2 Pomace (skins and seeds) (Kg)
Alcoholase Industrial liquid waste corresponding Liters (Lt) X3 Lees (Kg)
to distillation of impurities. It exhibits X4 Vinasse (Lt)
characteristics similar to those of X5 Alcoholase (Lt)
vinasse, with a solid content and X6 Reception of grapes (Kg)
BOD5 less than those of vinasse. X7 Application of vinasse in the fields ($)

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8 5

Table 4
Data obtained in the years 2006e2011.

Year Semester Grapes [Kg] Stems [Kg] Pomace [Kg] Less [Kg] Vinasee [Lt] Alcoholase [Lt] Water [Lt]

2006 1st 2,120,940 55,739 146,977 0 0 0 10,474

2nd 10,992,410 310,429 761,750 85,630 1,269,700 17,000 35,153
3rd 0 0 0 60,650 2,012,810 257,465 47,788
4th 0 0 0 0 1,681,410 239,590 41,466
2007 1st 3,013,770 81,673 203,740 0 0 0 1778
2nd 8,774,280 237,783 593,141 89,640 1,487,500 149,500 35,571
3rd 0 0 0 90,850 1,252,250 204,750 35,919
4th 0 0 0 45,000 191,000 50,000 9789
2008 1st 0 0 0 0 141,000 86,000 9384
2nd 13,493,020 365,661 912,128 21,690 358,000 0 9656
3rd 186,460 5053 12,605 0 655,000 182,000 22,451
4th 0 0 0 95,000 1,207,000 126,000 29,785
2009 1st 2,293,430 62,152 155,036 0 153,000 0 4454
2nd 8,850,030 239,836 598,262 0 442,000 0 12,910
3rd 0 0 0 30,000 218,000 403,000 22,123
4th 0 0 0 55,900 0 145,000 7787
2010 1st 1,140,120 30,897 77,072 0 479,000 0 13,590
2nd 10,385,520 281,448 702,061 30,000 440,000 107,000 16,917
3rd 0 0 0 90,000 1,477,000 282,000 37,069
4th 0 0 0 0 0 0 19,848
2011 1st 379,420 10,282 25,649 0 0 0 2323
2nd 10,202,030 276,475 689,657 0 0 0 3367
3rd 0 0 0 120,000 1,060,000 300,000 38,433
4th 0 0 0 10,000 112,000 580,000 32,557

2.77 is obtained, shown in Fig. 3. The F (F0) value is 40.680, therefore,
since the F (F0) value is greater, H0 is rejected (F0 > Ftable), and it can be
concluded that the model is significantly explicatory at the 5% level.
For the inference test, a significance level of 5% was used, and a
value of 2.101 for the t test was obtained from the Student’s t dis-
tribution table, as can be seen in Fig. 4. Table 5 shows the com-
parison of the t test value for each parameter.
Therefore, on first instance, the model can be expressed as fol-
lows in Equation (2).
Y ¼ 6250:805 þ 0:7404  Stems  0:2979  Pomace þ 0:0111
 Lees þ 0:0159  Vinasse þ 0:0360  Alcoholase
(2)
The results of the hypothesis test for the individual significance
tests indicate that solid industrial waste products are not significant
for the model, and hence should not be included. The following
model is definitive (3)
Y ¼ 6250:805 þ 0:0159  Vinasse þ 0:0360  Alcoholase
(3)
Considering only the variables that resulted significant using the
significance test for the parameters, a second more assertive model
was analyzed including only the variables vinasse and alcoholase.
These show a minimal variation in comparison to the previous
model, since the values were 0.0159 and 0.0360 respectively, and
the current values are 0.0167 and 0.0365. The value of R2 adjusted
increased slightly, while the determination coefficient R2 exhibited
a minimal decrease, indicating that the model is more significant
and is 90.61% explanatory with only these two variables. The Dur-
bineWatson statistic remained at 2.079, confirming the absence of
autocorrelation problems, and the Akaike and Schwarz criteria are

Fig. 2. Linear regression model. Fig. 3. F distribution.

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
6 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8

Fig. 4. t distribution.

use of water, with constant generation of alcoholase. Similarly, a

Table 5 unit increase in the cubic meters of alcoholase resulted in an in-
Student’s t test values. crease of 6.3 L in the trimestral use of water, keeping vinasse
Coefficients Std Student’s t Student’s Decision generation constant. In terms of overall consumption and taking
error. test statistic t test table the year 2006 as baseline, at which time the water consumption per
value
ton of processed grape was on average 10.29 cubic meters, a
b1 0.7405 0.8814 0.8401 2.101 H0 is not rejected, therefore the decrease of 29.55% was recorded, resulting in a consumption of 7.25
variable Stems is not significant cubic meters per ton processed.
b2 0.2979 0.3546 0.8402 H0 is not rejected, therefore the
The cleaner production agreement signed by the pisco industry
variable Pomace is not
significant had savings in water consumption during the productive process as
b3 0.0111 0.03192 0.3465 H0 is not rejected, therefore the one of the main objectives. Once the econometric analysis was
variable Yeast Sediment is not conducted and evidenced those processes having the greatest
significant
incidence of water consumption, it was possible to determine that
b4 0.0159 0.00201 7.9109 H0 is rejected, therefore the
variable Vinasse is significant
the stage of distillation, where the liquid waste vinasse and alco-
b5 0.0360 0.0074 4.8396 H0 is rejected, therefore the holase are produced, is where the greatest water impact is seen. It
variable Alcoholase is significant should be noted that during the initial stages, water is only used for
cleaning, maintenance of work areas, personal consumption, and
less than the in previous model, indicating that this new model has only slightly influences the actual productive process, having a
greater explicatory ability (4). greater influence in the latter stages of the productive process, as
other authors studying the wine industry have also recorded.
Y ¼ 5891:47 þ 0:0167  Vinasse þ 0:0365  Alcoholase (4)
(Barry, 2011; Bosco et al., 2011; Vázquez-Rowe et al., 2012; Gazulla
et al., 2010).
5. Discussion The main effects observed were in the disposal of liquid in-
dustrial waste, which are associated to the investment by the
A unit increase in cubic meters in vinasse produced during the pisco industry in general, where the main investment indices
productive process produces an increase of 16.7 L in the trimestral point to the implementation of liquid handling systems, with an

Fig. 5. Investment in the implementation of the CPA.

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8
investment of 6.11% with respect to other investments, as can be
seen in Fig. 5. During the period of 2006e2011, a 30% decrease
was seen in the consumption of water per ton of grape processed
(Asociación de Productores de Pisco AG, 2008). When water
consumption and tons of grapes processed during the period are
compared, a 43% decrease is seen in water intake, with a 30%
reduction in processed tons, confirming the benefits obtained in
the reduction of water consumption following implementation of
cleaner practices.
6. Conclusions
The current study evaluated the existing relationship between
the waste products generated by the pisco industry and their
relationship with the plant’s water consumption following the
implementation of the cleaner production agreement in Chile using
a multiple regression model.
In terms of the costs associated with the implementation of
the agreement, the savings generated by the efficient use of
water and handling of liquid waste show a balance between the
cost and benefits resulting from the implementation of the
agreement. On one hand, the reduction in water consumption
significantly reduced the costs incurred by the company in
treatments applied to water prior to utilization in the plant,
and on the other hand, the new liquid industrial waste
handling system and its direct use in the maintenance of com-
pany’s roads generated a 600% reduction when using vinasse for
irrigation.
As a consequence of these results and to better understand the
benefits obtained by implementing such agreement, participating
entities should consider alternatives regarding vinasse recycling or
its treatment, which could include: use of vinasse concentrators
and reutilization of remaining solid waste; implement forced
evaporation through dispersers; install treatment plants, and us-
ing these for irrigation or for third parties for reutilization in other
processes.
Successful compliance of the requirements of a cleaner pro-
duction agreement represents productive quality and efficiency,
linking participating entities with high environmental standards
and technological innovations that serve to minimize waste
products, preserve resources and decrease pollution. The country’s
current policies aim in the right direction to motivate companies
to participate in cleaner production agreements. Some of these
policies serve to diagnose, follow up and control, develop skills in
CPA, create public-private agreements in cleaner production,
communicate and disseminate.
In terms of the future for this industrial sector, it is suggested
that subsequent agreements also include previous stages in the
process, particularly harvesting, evaluation of the environmental
impact using the Life Cycle Assessment (LCA), and Water Footprint
Assessment.
One of the main limitations of the current work is that the study
is focused on Sociedad Agricola Hacienda Mal Paso y Cía. Ltda, and
hence the results cannot be considered representative of the pisco
industry as a whole. Although the remaining participating com-
panies were willing to collaborate, we were unable to access their
information.
Acknowledgments
The authors would like to thank Sociedad Agrícola Hacienda Mal
Paso y Cía Ltda for their support in this study. They would also like
to express their gratitude to the University of La Serena, particularly
to DIULS and to DII-ULS.
Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048
7
References
Askounes, N., 1994. Strategies and Policies for Cleaner Production. United Nations
Enviroment Programme Industry And Enviroment, France.
Asociacion de productores de pisco AG, 2008. Acuerdo de Producción Limpia Pro-
ductores de Pisco y Procesadores de Uva Pisquera III y IV Región.
Banfi, S., 2010. Pisco: Producción y Mercado. Chile. Available at: http://www.odepa.
gob.cl/odepaweb/publicaciones/doc/2439.pdf (accessed December 2012).
Barry, M.T., 2011. Life Cycle Assessment and the New Zealand Wine Industry: a Tool
to Support Continuous Environmental Improvement. Master thesis. Massey
University, Wellington, New Zealand.. Available at: http://muir.massey.ac.nz/
handle/10179/2920 (accessed June 2013).
Berkel, R., 2007. Cleaner production and eco-efficiency initiatives in Western
Australia 1996e2004. J. Clean. Prod. 15, 741e755.
Bezama, A., Valeria, H., Correa, M., Szarka, N., 2012. Evaluation of the environmental
impacts of a Cleaner Production Agreement by frozen fish facilities in the Biobio
Region, Chile. J. Clean. Prod. 26, 95e100.
Bonilla, S., Almeida, C., Giannetti, B., Huisingh, D., 2010. The roles of cleaner pro-
duction in the sustainable development of modern societies: an introduction to
this special issue. J. Clean. Prod. 18, 1e5.
Bosco, S., Di Bene, C., Galli, M., Remorini, D., Massai, R., Bonari, E., 2011. Greenhouse
gas emissions in the agricultural phase of wine production in the Maremma
rural district in Tuscany, Italy. Ital. J. Agron. 6, 93e100.
Campos-Aranda, D., 2011. Transferencia de información hidrológica mendiante regre-
sión lineal múltiple, con selección óptima de regresores. Agrociencia 45, 863e880.
Catastro Vitivinicola Nacional, Servicio Agrícola y Ganadero, 2008. Servicio Agricola
Ganadero. División de Protección Agrícola y Forestal, Subdepartamento de
Viñas y Vinos, Chile.
Cortés, H., 2005. El origen, producción y comercio del pisco chileno 1546e1931.
Revista Universum 20, 42e81.
CPL, 2012a. Ministerio de Economía, Fomento y Turismo. Chile. Available at: http://
www.produccionlimpia.cl/medios/Agenda_PL_2020.pdf/ (accessed June 2013).
CPL, 2012b. Ministerio de Economía, Fomento y Turismo. Chile. Available at: http://
www.produccionlimpia.cl/link.cgi/QuienesSomos/ (accessed December 2012).
CPL, 2012c. Ministerio de Economía, Fomento y Turismo. Chile. Available at: http://
www.produccionlimpia.cl/medios/Agenda_PL_2020.pdf/ (accessed December
2012).
Diario Oficial de la República de Chile, 27th May 2000. Fija Reglamento de la
Denominación de Origen Pisco. Decreto N
521, pp. 17e21 (Chile).
Dieleman, H., 2007. Cleaner production and innovation theory; social experiments
as a new model to engage in cleaner production. Revista internacional de
contaminación ambiental 23, 79e94.
Fahidy, T., 2006. An application of DurbineWatson statistics to electrochemical
science. Electrochim. Acta 51, 3516e3520.
Gazulla, C., Raugei, M., Fullana-i-Palmer, P., 2010. Taking a life cycle look at crianza
wine production in Spain: where are the bottlenecks? Int. J. Life Cycle Ass. 15,
330e337.
Grutter, J., Egler, H., 2004. From cleaner production to sustainable industrial pro-
duction modes. J. Clean. Prod. 12, 249e256.
Guerra, W., Cabrera, A., Fernández, L., 2003. Criterios para la selección de modelos
estadísticos en la investigación científica. Revista Cubana de Ciencia Agrícola 37,
3e10.
Herrera, F., 2011. Somos una industria que piensa limpio. Revista Producción Limpia
Coquimbo. Revista de la Secretaría Ejecutiva de Producción Limpia, Región de
Coquimbo, 8. Chile. Available at: http://www.produccionlimpia.cl/medios/
Revista_PL_3_Coquimbo.pdf (accessed November 2012).
Hicks, C., Dietmar, R., 2007. Improving cleaner production through the application
of environmental management tools in China. J. Clean. Prod. 15, 395e408.
Instituto de Investigaciones Agropecuarias, 2012. Chile. Available at: http://
nuestropisco.cl/esp/index.php?option¼com_
content&view¼category&layout¼blog&id¼18&Itemid¼18/ (accessed December
2012).
Kjaerheim, G., 2005. Cleaner production and sustainability. J. Clean. Prod. 13, 329e
339.
Kliopova, I., Staniskis, J., 2006. The evaluation of Cleaner Production performance in
Lithuanian industries. J. Clean. Prod. 14, 1561e1575.
Londoño, L., 2010. Implementación del programa de producción más limpia, orden
y aseo en Calzado Kondor Ltda Antioquia e Colombia. Producción þ Limpia 5,
113e121.
Medina, A., Medellín, P., 2006. La experiencia de adopción de la producción más
limpia en el sector de la fundición de México. Innovar 16, 173e186.
Montgomery, D., Runger, G., 1996. Probabilidad y Estadística Aplicadas a la Ingen-
iería. McGraw-Hill Interamericana Editores.
Nahuelhual, L., Engler, M., Carrillo, B., Moreira, V., Castro, I., 2009. Adoption of
cleaner production practices by dairy farmers in southern Chile. Ciencia e
investigación agraria 36, 97e106.
Pareja, C., Alexander, R., López, F., Matilde, L., Arroyave, S., Milena, S., Pérez, A.,
Cristina, I., 2010. Fortalecimiento del desempeño ambiental empresarial, a
través del programa de producción más limpia y consumo sostenible del Área
Metropolitana del Valle de Aburrá. Producción þ Limpia 5, 1e23.
Pérez, C., 2006. Problemas Resueltos de Econometría. Thompson, Madrid.
Pértega, S., Pita, S., 2000. Técnicas de regresión: regresión lineal múltiple, Unidad de
Epidemiología Clínica y Bioestadística, Complexo Hospitalario Juan Canalejo, A
Coruña. Cad Aten Primaria 7, 173e176.
8 M. Castillo-Vergara et al. / Journal of Cleaner Production xxx (2013) 1e8

Pisquera de Chile, 2013. Chile. Available at: http://www.ccu.cl/nuestros-negocios/
pisco-y-otras-bebidas-espirituosas/ (accessed July 2013).
Rodríguez, M., 2009. Lógica difusa como herramienta para interpretar datos de
producción limpia en el sector agrícola. Idesia (Arica) 27, 101e105.
Rugani, B., Vázquez-Rowe, I., Benedetto, G., Benetto, E., 2013. A comprehensive
review of carbon footprint analysis as an extended environmental indicator in
the wine sector. J. Clean. Prod. article in press. Available at: http://dx.doi.org/10.
1016/j.jclepro.2013.04.036 (accessed June 2013).
Rutledge, D., Barros, A., 2002. DurbineWatson statistic as a morphological esti-
mator of information content. Anal. Chim. Acta 454, 277e295.
Schwarz, G., 1978. Estimating the dimension of a model. Ann. Stat. 6, 461e564.
Terry, E., Pino, M., Salomón, J.L., Dell Amico, J.M., Suárez, Y., Chaveco, O., Peña, R.,
Wrigh, J., 2009. La innovación local como alternativa para atenuar el impacto de
la sequía. Cultivos Tropicales 30, 121e126.
Thrane, M., Nielsen, E., Christensen, P., 2009. Cleaner production in Danish fish
processing e experiences, status and possible future strategies. J. Clean. Prod. 17,
380e390.
Van Hoof, B., Herrera, C.M., 2007. La evolución y el futuro de la producción más
limpia en Colombia. Revista de Ingeniería 26, 24e33.
Van Hoof, B., Lyon, T., 2013. Cleaner production in small firms taking part in
Mexico’s Sustainable Supplier Program. J. Clean. Prod. 41, 270e282.
Varón, L., Sierra, D., Bedoya, L., 2011. Indural: un aporte significativo a la pro-
ducción más limpia y la construcción sostenible. Producción þ Limpia 6,
128e135.
Vázquez-Rowe, I., Villanueva-Rey, P., Moreira, M.T., Feijoo, G., 2012. Environmental
analysis of Ribeiro wine from a timeline perspective: harvest year matters
when reporting environmental impacts. J. Environ. Manage. 98, 73e83.
Zárate de Lara, G.P., Diaz de Zarate, B.L., 1985. Técnica de diagnóstico y manejo de
puntos de influencia y multicolinealidad en regresion lineal. Agrociencia 61,
41e58.
Zeng, S.X., Meng, X.H., Yin, H.T., Tam, C.M., Sun, L., 2010. Impact of cleaner pro-
duction on business performance. J. Clean. Prod. 18, 975e983.

Please cite this article in press as: Castillo-Vergara, M., et al., Implementation of a Cleaner Production Agreement and impact analysis in the
grape brandy (pisco) industry in Chile, Journal of Cleaner Production (2013), http://dx.doi.org/10.1016/j.jclepro.2013.09.048