Energy Consumption Characterization based on a Self-analysis Tool:

A Case Study in Yarn Manufacturing

Samuele Branchetti, Carlo Petrovich, Gessica Ciaccio, Piero De Sabbata,

Angelo Frascella and Giuseppe Nigliaccio

ENEA – National Agency for New Technologies, Energy and Sustainable Economic Development,

Via Martiri di Monte Sole 4, Bologna, Italy

Keywords: Benchmarking, Energy Consumption, Energy Efficiency, Sustainable Economy, Textile Industry, Yarn

Manufacture.

Abstract: Even if energy efficiency represents a crucial issue for the sustainability of the manufacturing industry, the

companies need to be encouraged in investing their resources for this goal. One of the means to facilitate

this effort is the comparison of the energy performances with similar factories. Nevertheless, since the

enterprises are very heterogeneous, these performance values have, even within a specified manufacturing

sector, a high variability and therefore risk not to be representative. The dispersion of these data has to be

decisively decreased. This goal is pursued here by means of an energy consumption characterization model

based on: 1. a self-analysis software tool collecting energy consumption data in a simple and homogeneous

way; 2. the clustering of the factories; 3. the separation of the auxiliary energy uses from the production

process energy consumption. The method is here applied to textile industry with a focus on the electrical

consumption in yarn factories. The outcomes show a correlation with some production variables, such as the

raw materials, and allow to reduce the relative errors of the energy performances of different factories from

about 80% to about 25-40%. In this way, energy reference indicators can be built in an acceptable and

representative way.

1 INTRODUCTION

Energy efficiency has become a crucial issue for

the manufacturing industry because of the need of

reducing greenhouse and pollutant emissions, of

optimizing fuel resources and of increasing

economic market competitiveness.

In this paper the textile industry, and in

particular the yarn manufacturing, is chosen as a

case study for an energy consumption

characterization, and the approach which will be

outlined here can be applied also to other

manufacturing sectors.

In 2015 textile industry in Europe counted

61,000 companies and produced a turnover of

around 79,000 million € with an energy cost of

about 1,900 million €. The energy cost contribution

was higher than the average energy contribution

cost of manufacturing sector (EUROSTAT, 2015).

For yarn manufacturing the incidence of the

energy costs can vary a lot, depending on many

variables, such as the kind of raw materials, the

involved processes and labor costs. According to

different countries and for 20 tex carded open-end

rotor cotton yarn, the share in energy cost

represents the 5-18% of the total mill costs

(Kaplan, 2010; Alkaya, 2014) and 10-25% for ring

and rotor spinning (ITMF, 2014).

In 2016, China, European Union, India and the

U.S. were the four largest textile importers and

exporters in the world. The EU textile sector

represented 23% of the world textile exports

(WTO, 2017). In China and in the U.S., the

contribution of the textile industry sector to the

national final energy use in manufacturing is

respectively of the 4% and of the 2% (Hasanbeigi,

2012a). Even if the textile industry is not a very

high energy-intensive industry, it involves a large

number of plants, consuming together a significant

amount of energy.

The purpose of improving energy efficiency in

textile industry is a common declared

governmental strategy for many regions, such as

Branchetti, S., Petrovich, C., Ciaccio, G., De Sabbata, P., Frascella, A. and Nigliaccio, G.

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing.

DOI: 10.5220/0007680300400050

In Proceedings of the 8th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2019), pages 40-50

ISBN: 978-989-758-373-5

Europe (Scheffer, 2015), Turkey (Alkaya, 2014)

and Taiwan (Hong, 2010). The Best available

techniques REFerence document (BREFs),

prepared by the European IPPC Bureau, describes

how to implement the best available techniques to

use natural resources in an efficient way and to

minimize pollution emissions (IPPC, 2003).

Notwithstanding these government goals and

the fact that the payback time is often less than few

years, the companies, especially SMEs, are not

usually prone to invest time and relevant resources

to face a comprehensive energy consumption

assessment and to implement the adequate

measures.

A dissemination campaign of the several energy

saving opportunities can help to overcome the

limited information and to make aware of the

possible benefits of these measures. Another way

to promote and facilitate investments in energy

efficiency measures is to provide some reference

performance indicators for the energy

consumptions of the production processes. This

would represent a fast way, even if of preliminary

nature, for the factory to understand whether its

energy consumptions are efficient or not.

Comparisons and benchmarks for the energy

consumption of a plant can be based on the past

performance of the same plant or on other plants’

performance (or average performance) of the same

plant group or more general plant groups. Another

possibility is the comparison with the performance

of “best practice” plants of the same group or

performance of “best technologies” in the industry

(Hasanbeigi, 2012b). Unfortunately, even factories

of the same group are heterogeneous and differ

greatly from each other because of different

processes involved, different machine setting and

kinds of products and different machines and

components (Hasanbeigi, 2012a). Even within the

same textile sub-sectors, such as yarn

manufacturing, fabric manufacturing and finishing,

a similar fragmentation occurs.

Moreover, the main available references (e.g.

data from EUROSTAT (2015)) are few and too

general to be meaningful for all the different kind

of companies of the textile sector. In literature

benchmarking, no homogeneous description of the

plant characteristics and of the implemented

processes exists (because of confidentiality issues,

besides being a time-consuming activity).

Therefore the problem of energy performance

comparison, of energy and economical

characterization is still unsolved because of the

great variability of the production processes.

On the other hand, more accurate estimations of

the energy consumption of a textile mill can be

provided using more precise equations but referring

only to very specific production phases, machines

and setting parameters, such as yarn counts and

twists. This approach is followed for example by

Koç (2007), but it needs a huge amount of time and

detailed data, such as the yarn properties, the power

and number of the machines, the load factors, the

efficiencies and the number and type of processes.

This paper proposes an energy consumption

characterization model which allows, by means of

self-analyses, to obtain approximate reference

values for the comparison. The data for the

construction of the model were collected through a

set of software tools, named SET Tool, developed

within three European projects (ARTISAN, SESEC

and SET), which were focused on the Textile and

Clothing sector (Branchetti, 2016). Furthermore, an

European informative campaign, named Energy

Made to Measure (EM2M) (EM2M, 2016), led by

the European industry association (EURATEX),

has allowed, by means of these software tools, to

retrieve many information data from the factories

and thus to build a quite large database.

The collected data have been clustered

according to the kind of production and the used

raw materials. Then, the data have been elaborated

using regression analysis methodologies and

indicators, such as the Specific Energy

Consumption (SEC), which represents the total

energy consumption of the whole mill for unit of

product (kWh/kg). Here we refer to the “electrical

SEC” and “thermal SEC” when representing

electrical and thermal energy consumption

respectively. The present paper focuses on part of

the textile production chain, in particular on the

whole yarn manufacturing processes.

The main goals of this paper are:

 to make available to the public literature some

energy consumption indicators of the textile

industry, as collected by the SET tools,

respecting industrial data confidentiality;

 to decrease the range of variability of the SEC

for yarn production mills, without taking into

account production details; in this way a

textile factory can more easily compare its

energy consumption with reference values;

 to validate a general approach to compare

energy performances in manufacturing

industry avoiding detailed, time-consuming

and demanding measurement campaigns.

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing

In section 2, energy consumption in yarn

manufacturing is discussed, together with a

literature review of this issue and of the SEC

values. In section 3, the SET Tool and its database

are presented along with the used indicators. After

examining the relationship between energy and

production, the main results are presented in

section 4 and discussed in section 5.

2 ENERGY CONSUMPTION IN

YARN MANUFACTURING

The textile industry is a complex manufacturing

industry because it represents a fragmented and

heterogeneous sector, dominated by Small and

Medium Enterprises (SMEs) (Hasanbeigi, 2012a).

The textile production process is composed of

many and different production phases and sub-

phases, hierarchically identifiable. Starting from

the NACE rev. 2 classification, three main kinds of

production can be highlighted:

 yarn production;

 fabric production;

 finishing (of yarn and fabric).

Even if yarn production is made up of various

processes, which can be in sequence or alternative,

it is possible to define some main production

phases, as depicted in Figure 1: opening and

preparation of the raw materials (such as cotton and

wool), carding, combing, drawing, spinning,

winding, rowing, steaming, etc. The dyeing and

other finishing processes are not included here in

the “yarn production” category. A factory can

implement many of these phases or only just one.

Moreover, it is possible that these phases occur at

different facilities of the same company.

In yarn production most of the energy

consumption is due to spinning processes. In case

of middle count, carded, ring yarn, the spinning

and winding processes represent about 55-80% of

the energy consumption per kg of single yarn (Koç,

2007). The spinning systems can be classified

according to different technologies into ring

spinning, compact spinning, rotor spinning, air-jet

spinning and other spinning machines, which are

characterized by different energy consumption

behaviour. On the other hand, the scientific

literature on the energy consumption is not

systematic, often outdated and not well

documented (Van der Velden, 2014).

Comprehensive studies for a specific textile

mill, facing energy consumptions, water

consumptions, waste and pollutant generations are

reported in Ozturk (2015). Potential energy

reduction applying the Best Available

Technologies list (BAT) are evaluated there as

about 10-30%, with a total potential reduction up to

70% with a payback period up to 4 years (by means

of energy monitoring and control, insulations, heat

recovery, substitution of electric motors) (IPPC,

2003).

In Hasanbeigi (2012b), the energy consumption

of 13 textile plants in Iran have been analyzed and

audited, regarding 5 different sub-sectors

(spinning, weaving, wet-processing, worsted fabric

manufacturing and carpet manufacturing).

Therefore there are only 2 or 3 plants in this study

in each manufacturing sub-sector, for each of

which the energy intensity has been analyzed and

then a range of energy consumption has been

estimated. Hence, benchmark values for other mills

in the same sub-sector are attempted, even if the

indicators are deduced from a very limited number

of factories. Moreover, the highest share of energy

consumption in the textile industry worldwide turns

out to be due to spinning, weaving, and wet-

processing.

In Lin (2016) Chinese regional differences in

the total energy efficiency of the textile industries

are analyzed for the period 2000-2012. There, the

technology gap is taken as the most important

parameter to explain the disparity in energy

efficiency in eastern, central and western regions.

Considering the distance from the frontier

technology, a huge energy saving potential is

highlighted for the Chinese textile sector.

Other studies focused their investigations on the

whole life cycle of the textile industry using the

LCA (Life Cycle Assessment) approach

(Steinberger, 2009; Zamani, 2014; Van der Velden,

2014). In these studies the complete chain is taken

into account, from the cultivation up to the

production, textile use, dress washing and even

ironing. The approach is comprehensive, but, on

the other hand, only a rough estimation of the

production energy consumption is used and a

discussion of its high spread is missing.

SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems

Figure 1: Typical yarn processes.

Table 1: Electrical and thermal energy consumption for yarn production mills according to the analyzed literature.

Description

Electrical

energy

consumption

[kWh/kg]

Thermal

energy

consumption

[MJ/kg]

Reference

Spinning (1972) 5.4 (Kim, 1983)

Spinning (1980) 4.9 (Kim, 1983)

Range for spinning mills 2.7 – 4 1.1 - 4.7 (Tarakçıoğlu, 1984)

Range for textile mills 0.5 - 7.5 11 - 65

(Kumar, 1999)

(UNIDO, 2010)

Range of spinning plants 0.55 - 7.3 0.14 - 0.73 (Visvanathan, 2000)

Ring yarn (combed)

20 tex

3.5 - 3.6 (ITMF, 2003)

1.8 - 5.1 (Dahllöf, 2004)

Cotton spinning (1997) 5.1 (Dahllöf, 2004)

- 11.6 (Ellebæk Larsen, 2007)

Spinning mill

(mix ring and open-end, cotton)

3.2 - 3.8 (Kaplan and Koç, 2010)

Open-end spinning mill (calculated,

20 tex)

3.0 (Kaplan , 2010)

Specific spinning plant 3.2 - 3.5 (Palamutcu, 2010)

Spinning plant 3.2 - 3.5 (EMS)

Ring spinning mills 6.6, 4.7 12.4, 7.1 (Hasanbeigi, 2012b)

Open-end Spinning mills

3.6 8.1

(Hasanbeigi, 2012b)

SimaPro 7.2 5.1 (Van der Velden, 2014)

SimaPro 7.2 3.4 (Van der Velden, 2014)

Table 1 reviews and summarizes the data found

in the abovementioned references and in other

literature. The data are presented in chronological

order and reported in Figure 2. The bars show the

ranges of values declared by the respective

references. Figure 2 confirms the wide distribution

of the data, but also a general consistency among

them. The ranges of the data is 0.5-11.6 kWh/kg and

the average turns out to be 4.5 kWh/kg. Reference

Van der Velden (2014) claims that the wide range of

the literature data is mainly due to the mixture of

data coming from very different textile product

characteristics, the most relevant of which is the

yarn count.

Anyhow, as abovementioned, the range of the

data is too high to be useful for benchmarking

purposes. As shown in Figure 2, SEC in yarn

manufacturing can vary by a factor of 20. Values of

SEC, as a matter of fact, depend on many factors

and choices, such as the raw materials characteristics

(Hasanbeigi, 2012b), type and number of processes

taken into account, type of spinning system, yarn

count (Van der Velden, 2014; Koç, 2007), yarn twist

(Hasanbeigi, 2012b), energy efficiency of the

machines and machine time utilization (the

workload), geographical location (Hasanbeigi,

2012b), production capacity (Palamutcu, 2010).

Moreover, the technology evolution has to be taken

into account as well, because it drives a decrease of

the SEC with time. This issue has been investigated

for Germany and Colombia from 1998 to 2005 in

Pardo Martínez (2010).

Even the detailed energy consumption analysis

for a specific mill is not easy. Direct measurements

in 5 plants in Palamutcu (2010) show discrepancies

between estimated and actual energy data of the

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing

order of 10-30% due to the variety of different

processes, the efficient use of the equipment,

processes steps, the discontinuity of machine use by

cause of maintenance periods.

From this description, it is clear that an unique

reference number representing the energy

consumption of the whole textile industry is poorly

representative for a specific mill, because the textile

factories have very different features. On the other

hand, this paper does not take into account all these

variables, because this would have required a deep

analysis of the machines and costs from the textile

plants. Rather, the approach proposed here is to

consider only few more significant variables, such as

the type of production process and the main raw

materials.

Figure 2: Range of specific electrical energy consumptions

(kWh/kg) for yarn mills from the analyzed literature

reported in Table 1.

3 METHODS

The proposed methodology is based on the so-called

SET Tool, that is a self-analysis standalone tool

which provides feedback on energy efficiency

measures and energy indices to the textile mills

interested in investigating this issue. Enterprises are

encouraged by the tool to provide information on:

the yearly and monthly production amount, the

electrical and thermal consumptions of the mill, the

number and the type of processes involved, the

turnover, the number of employees, the main used

raw materials, the product market segment and the

market segment application (clothing, home textile,

technical textile, etc.).

The SET Tool was designed together with a

web-based application, named SET Web, which is

able to retrieve and filter data, to check their

faithfulness and to calculate customized energy

benchmarks. Enterprises can access SET Web and

its services, obtaining performances comparison

with energy benchmarks, by anonymously sending

their data through the SET Tool. As a matter of fact,

all these data are provided by the textile factories

themselves and are automatically recorded and

organized to build up a centralized and growing SET

database. With respect to other benchmarking

methodologies (Andersson, 2018), the energy

benchmarks obtained by the SET Tool are built

dynamically through company self-profiling.

Therefore, the companies themselves contribute to

improve the quality of the benchmark.

3.1 The Factory Database

At the end of 2017, the factory database was

composed of 204 datasets, which were provided by

136 companies in relation to 140 factories. Pruning

the data referring to the same production but to

different years and those being inconsistent or not

complete, 123 datasets have been selected, regarding

4 main areas: yarn production, fabric production,

finishing processes and factories involving a

combination of these productions (yarn and finishing

of yarn, fabric and finishing of fabric or the overall

production processes). Table 2 shows characteristics

and consumption indices of the mills, which have

uploaded the data. The data mainly refer to the years

2014-2015.

From the SET datasets it is possible to provide a

quite general view of the energy consumption in

yarn production, fabric production and finishing in

Europe: the numbers of the textile factories which

have been collected with the SET Tool are about 25-

40 for every sector. Even if some outliers and

incomplete datasets have been excluded, the range

and the spread of the total energy consumption

values are still very high (0.09 – 16.5 kgoe/kg)

because different kinds of textile factories have been

collected.

3.2 Electrical and Thermal Energy

In textile industry, both electrical energy and

thermal energy are used. The former one is used

mainly for operating machines (e.g. spinning,

weaving and knitting) and facilities such as air

compressors, air conditioning and lathing, while the

latter one mainly for heating and production

SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems

processes such as fixation of yarns, steam for

autoclaves and hot water for dyeing.

According to the different types of processes and

products, thermal energy or electrical energy can

prevail.

In UNIDO (2010) the electrical consumption

rate in the total consumed energy for individual

textile production stages are reported to be 93% for

spinning, 85% for weaving, 43% for wet processing

and 65% for clothing manufacturing. The rest of the

energy is provided as thermal energy by other

energy fuels (natural gas, coal, etc.). According to

three spinning plants analyzed in (Hasanbeigi,

2012b), 60-70% of the used energy is electricity (for

machinery, humidification systems, compressed air

systems and lighting). The spinning step accounts

for 56% of total energy use in the yarn

manufacturing process (Hong, 2010).

Table 2: Characteristics of the SET datasets with min. and max. values. Primary energy (*) is expressed by kilogram of oil

equivalent [kgoe] using a conversion factor of 0.000215 toe/kWhe (Table 4 of Commission decision 2007/589/CE and

Annex II of directive 2006/32/CE).

Yarn Fabric Finishing

Yarn+Finishing

Fabric+Finishing

Yarn+Fabric+Finishing

Factories 26 49 29 36

n. of mills

consistent and

complete

21 44 27 31

Date for reference

year

2013 - 2015 2013 - 2015 2012 - 2015 2013 - 2015

European

countries involved

BE, CZ, HU, IT, PT, RO

BE, BG, CZ, DE, HU,

PT, RO

CZ, DE, FR, HU, IT, PT

BE, DE, HR, HU, IT, CZ,

LT, PT, RO

Number of

employees

5 - 535 3 - 410 6 - 200 15 - 1000

Turnover

[Millions of €]

0.15 - 90 0.10 - 35 0.19 - 24 0.71 - 116

Product market

segment

From low target market

to luxury market

From low target market

to luxury market

From medium target

market to luxury market

From medium target

market to luxury market

Market segment

application

Clothing, Home textile,

Technical textile,

Protective textile

Clothing, Home textile,

Technical textile,

Underwear

Clothing, Home textile,

Technical textile,

Underwear, Protective

textile, Other

Clothing, Home textile,

Technical textile,

Underwear, Protective

textile

Main raw

materials used

Cotton, Wool, Acrylic,

Polyamide,

Polypropylene, Linen,

Other natural fibres

Cotton, Wool, Acrylic,

Polyester, Polyamide,

Polypropylene, Linen,

Other natural fibres

mainly Cotton, but also

Polypropylene, Polyester,

Polyamide, Silk, Wool and

Other synthetic fibres

Cotton, Wool, Polyester,

Polyamide, Acrylic,

Linen, Acetate, Other

natural fibres, Other

Electrical energy

consumption

[kWhe/kg]

0.44 – 14.55 0.49 – 25.14 0.49 – 32.98 1.11 – 17.87

Thermal energy

consumption

[kWhth/kg]

0.01 – 20.30 0.00 – 44.40 2.43 – 109.69 2.39 – 54.24

Total energy

consumption

[kgoe/kg]*

0.088 – 3.89 0.11 – 9.22 0.560 – 16.52 0.61 – 7.35

% Electrical

energy

from 52% up to 100%,

with average of 85%

from 15% up to 100%,

with average of 73%

from 13% up to 76%, with

average of 28%

from 21% up to 80%,

with average of 47%

% Thermal energy

from 0% up to 48%,

with average of 15%

from 0% up to 85%, with

average of 27%

from 24% up to 87%, with

average of 72%

from 20% up to 79%,

with average of 53%

Annual product

amount

[tonnes/year]

30 - 32781 5 - 1767 60 - 4800 55 - 15165

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing

Table 3: Electrical consumption rate with respect to the sum of electrical and thermal consumption for each type of

production in textile mills.

Type of production Literature SET database

Distinct

textile

areas

Spinning (yarn production) 93% (UNIDO, 2010)

60-70% (Hasanbeigi, 2012b)

56% (Hong, 2010)

85% (average of 21 mills)

Weaving (fabric production) 85% (UNIDO, 2010) 73% (average of 44 mills)

Wet processing (finishing) 43% (UNIDO, 2010) 28% (average of 27 mills)

Composite

textile

areas

Yarn production and finishing - 56% (average of 4 mills)

Fabric production and finishing - 42% (average of 22 mills)

Yarn, fabric and finishing - 61% (average of 5 mills)

In wet textile processing thermal energy prevails,

because of high temperature processes (Hong, 2010)

and this is confirmed by the SET database analysis.

Table 3 shows some data found in literature

compared to data of the present paper (obtained by

the SET database). From the SET datasets (Table 2),

the electrical energy consumption prevails on

thermal energy consumption for both yarn and fabric

production with an average of 85% and 73%

respectively.

On the other hand, the thermal energy

consumption prevails in finishing factories

(electrical energy consumption has an average value

of 28%). The energy consumption in factories of

composite textile areas depends on the different mix

of production processes and thus the splitting of

energy consumptions between electrical and thermal

appears more balanced with respect to distinct textile

areas, showing an average of electrical energy

consumption ranging from 42% to 61% (Table 3).

3.3 Relationship between Energy and

Production

The SEC is the main indicator to express the energy

efficiency of a factory and it represents the energy

consumption of the whole factory per unit of product

(kWh/kg).

In order to obtain further information about the

energy management of the companies, the

Incremental Energy Consumption (IEC) is here

introduced, representing the energy consumption to

produce an additional unit of product (kWh/kg).

This indicator is obtained by investigating the

relationship between the monthly production and the

monthly energy consumption, since the energy

consumption is expected to be related to the

production (Palamutcu, 2010; Branchetti, 2016).

With a regression analysis method, the energy

consumption is estimated as:

y = m·x + q (1)

Where, y denotes the whole energy consumption

[kWh], m is the energy consumption to produce each

additional unit of product [kWh/kg], x is the

production amount [kg], q is the consumption when

the production is zero [kWh].

The model parameters (the slope m and the

interception q of the best fit line) were estimated for

all the available factories of SET database

correlating the monthly production (independent

variable) with the electrical and thermal energy

consumptions (dependent variables). In this way, m

represents the IEC and it can be calculated both for

the electrical and thermal energy.

The strength of the relationship between

production and energy consumptions (respectively,

electrical and thermal) for the SET database

factories has been checked by means of the

correlation coefficient R

When R

is close to 1, then the model fits the

data with good agreement and the energy

consumption appears strongly correlated to the

production. In these cases, it is possible to evaluate

the “base energy consumption”, which is the portion

(percentage) of the total energy not related to the

production (Branchetti, 2016). The “base energy

consumption” represents the energy auxiliary uses

(such as lighting, air conditioning, heating and

ventilation) and contributes to the increase of the

SEC, whereas it does not affect the IEC indicator of

the company.

On the other hand, in factories with a low R

and then with a low correlation between energy

consumption and production, it is not possible to

evaluate the “base energy consumption” and, in

SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems

these cases, the factories may have not an adequate

management of the energy consumption or the

production might be composed of a wide mix of

different products and raw materials.

4 RESULTS

In the regression analysis for the SET datasets, the

values of the model parameters have been accepted

only when R

>0.5. Focusing on the 21 textile

factories for yarn production (Table 2), the linear

regression analysis shows that the electrical R

greater than 0.5 for 13 factories, whereas the thermal

is greater than 0.5 only for 1 factory.

The subsequent analysis is then focused on

electrical energy only, because of the prevalence of

electricity uses with respect to thermal energy in

yarn manufacturing and of the better correlation of

electricity with production.

The electrical SEC of the 13 factories ranges

from 1.4 to 14.5 kWh

/kg with an average of 5.6

kWh

/kg, a standard deviation (std) of 4.6 kWh

/kg

and a relative error (std/average) of 83% (Table 4).

On the other hand, the electrical IEC of these

factories ranges from 1.1 to 7.9 kWh

/kg and has an

average of 3.5 kWh

/kg with a standard deviation

(std) of 2.2 kWh

/kg and a relative error of 64%.

Table 4: Electrical SEC and IEC for factories with R

>0.5.

Average

[kWh

/kg]

std

Relative

error

Electrical SEC 5.6 ± 4.6 83%

Electrical IEC 3.5 ± 2.2 64%

Clustering the electrical SEC based on the kind

of raw materials (Table 5 and Figure 3), the factories

producing wool yarn show an electrical SEC ranging

between 6 to 14 kWh

/kg, which is higher than the

electrical SEC for factories producing yarn starting

from raw materials composed mainly of cotton or

“other materials” (i.e. linen, polyamide, acrylic and

polypropylene). The latter, as a matter of fact, ranges

between 2 and 4 kWh

/kg.

Table 5: Electrical SEC and related std clustered by the

kind of raw materials for factories with R

>0.5.

Electrical SEC

Average

[kWh

/kg]

std

Relative

error

WOOL 10.4 ± 4.0 39%

COTTON 2.4 ± 0.6 26%

OTHER 2.7 ± 1.0 37%

Figure 3: Electrical SEC clustered by the kind of raw

material for factories with R

>0.5.

Similarly results are obtained for electrical IEC

(Table 6 and Figure 4). The factories producing

wool yarn show an electrical IEC ranging between 4

to 8 kWh

/kg, higher than the other clusters, which

ranges between 1 and 3 kWh

/kg.

Table 6: Electrical IEC and related std clustered by the

kind of raw materials for factories with R

>0.5.

Electrical

IEC

Average

[kWh

/kg]

std

Relative

error

WOOL 5.8 ± 1.7 29%

COTTON 1.8 ± 0.7 38%

OTHER 2.2 ± 0.8 37%

Figure 4: Electrical IEC clustered by the kind of raw

materials for factories with R

>0.5.

5 DISCUSSION OF THE RESULTS

The dataset extracted from the SET database for

yarn manufacturing confirms the prevalence of

electrical energy consumption with respect to the

thermal consumption and shows an electrical SEC

ranging from 1.4 to 14.5 kWh/kg with an average of

5.6 kWh/kg and a high relative error of about 83%.

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing

This range of values and its spread are compatible

with those found in literature (Figure 2) and are due

to many factors and choices already described in

section 2.2.

The regression analysis method applied to the

monthly data of each yarn manufacturing dataset

allows to calculate the electrical consumption per

each additional produced unit and the base electrical

energy consumption (auxiliary energy uses). The

latter ranges from 6.2% to 50.3%, with an average of

28%, and it is compatible with the energy

consumption values found in literature (Kaplan,

2010). Concerning the electrical energy

consumption per each additional unit of product

(electrical IEC), the outcome shows that it ranges

from 1.1 to 7.9 kWh/kg with an average value of 3.5

kWh/kg and with a lower relative error (64%) with

respect to the electrical SEC. The reason is that the

auxiliary energy uses in SEC added a further

variability in the electrical consumptions.

Clustering the data by raw materials, we have

found that the production of yarn based on wool is

more energy consuming than the production of yarn

starting from cotton or other fibers. This result has

been verified for both electrical SEC and electrical

IEC indicator. On the contrary, the clustering based

on market segment application of products (e.g.

clothing, home textile, etc.) or product market

segment (luxury, top, medium and low), does not

show satisfactory results.

The thermal electrical consumption is less

significant in yarn manufacturing, because in

average it contributes to about the 15% of the total

energy consumption of the factory (see table 2).

Moreover, it is not strictly related to yarn production

(only one yarn manufacturing mill of the SET

database shows a correlation coefficient R

>0.5).

6 CONCLUSIONS

In order to provide a fast and easy method to the

factories to preliminary evaluate their own energy

consumptions, a reference indicator would be useful

for a comparison. This would deliver a fair trade-off

between too general or too specific approaches and

between too naïve or too demanding methodologies.

Unfortunately the factories have very different

features: the textile production chains, even in their

sub-sectors segments, combine very heterogeneous

and fragmented processes and type of products.

Consequently, the energy indicators appear highly

variable and poorly representative.

The difficulties were tackled using a factory self-

analysis approach which has allowed to retrieve and

analyze wide and detailed sets of data. The SET

database, obtained by the SET tools, has been

presented in this paper and the yarn manufacturing

data have been investigated with the final goal of

obtaining valuable references for energy

consumptions.

The SET database counts 204 sets of data

regarding 140 factories and 4 main textile areas:

yarn production, fabric production, finishing

processes and a combination of them.

The results for yarn manufacturing are

complementary and consistent with respect to the

available literature and public data. They confirm a

high variability of SEC values within the textile

sector, but also a correlation with some production

variables, such as processes and raw materials. The

outcomes enrich the available data for the textile

industry and in particular for yarn manufacturing.

Clustering the datasets on the base of raw

materials allows to decrease the relative error from

83% to 25-40%

The separation of the auxiliary energy uses from

the production process energy consumption allows

the comparison of different energy contributions

among similar factories. The factories producing

wool yarn show electrical energy consumption per

each additional unit of product ranging from 4 to 8

kWhe/kg, while it ranges from 1 to 3 kWhe/kg for

factories producing yarn composed mainly of cotton

fibres or “other materials” (i.e. linen, polyamide,

acrylic and polypropylene).

The self-analysis approach allows to create a

self-growing dataset with indicators which are

supposed to become more and more representative

along with increasing factory involvement. From

this perspective, the SET database is meant to be a

starting point to build up comprehensive and

consistent models to depict energy consumptions for

textile industry. Moreover, the approach followed in

this paper can be implemented also in other

manufacturing sectors, even if clustering choices and

their usefulness depend on the particular chosen

sector and have to be verified case-by-case.

ACKNOWLEDGEMENTS

Financial support has been provided by the EU for

the development of the SET tools within the

ARTISAN, SESEC and SET projects.

SET was a project funded in 2014-2016 by the

“Intelligent Energy – Europe” programme (grant no.

SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems

IEE/13/557/SI2.675575) with the goal of improving

the energy efficiency of at least 150 European

Textile SMEs with tangible and quantifiable

benefits.

SESEC was a co-funded project in 2012-14

within the European Programme Intelligent Energy

Europe by EASME (grant no.

IEE/11/827/SI2.615931).

ARTISAN was a R&D project co-financed in

2011-14 by the European Commission 7th

Framework Programme (FP7-ICT-2011-7 Grant

agreement 287993).

The results of this paper cannot be intended as

official reference values to be used for applications

of national subsidies for energy efficiency measures.

REFERENCES

Alkaya, E., Demirer, G.N., 2014. Sustainable textile

production: a case study from a woven fabric

manufacturing mill in Turkey. In Journal of Cleaner

Production. 65, 595-603.

Andersson, E., Arfwidsson, O., Thollander, P., 2018.

Benchmarking energy performance of industrial small

and medium-sized enterprises using an energy

efficiency index: Results based on an energy audit

policy program. In Journal of Cleaner Production.

182, 883-895.

Branchetti, S., Ciaccio, G., De Sabbata, P., Frascella, A.,

Nigliaccio, G., Zambelli, M., 2016. Energy Saving and

Efficiency Tool - A Sectorial Decision Support Model

for Energy Consumption Reduction in Manufacturing

SMEs. In SMARTGREENS: Proceedings of the 5th

International Conference on Smart Cities and Green

ICT Systems. 2016 Apr 23-25, Rome, Italy.

SCITEPRESS, pp. 330-339.

Dahllöf, L., 2004. LCA methodology issues for textile

products [Licentiate thesis]. Chalmers University of

Technology, Göteborg.

Ellebæk Larsen, S., Hansen, J., Knudsen, H.H., Wenzel,

H., Larsen, H.F., Møller Kristensen, F., 2007.

EDIPTEX — Environmental assessment of textiles.

Danish Ministry of the Environment, Environmental

Protection Agency. Working Report No. 24, Denmark.

EM2M, 2016. www.em2m.enea.it (accessed 9 May 2018).

EMS - Energy Management Standard in Textile Project -

EC European Intelligent Energy report, Section 4 -

Energy Consumption.

European Commission (EU), 2003. Integrated Pollution

Prevention and Control (IPPC) - Reference Document

on Best Available Techniques for the Textiles Industry,

Seville.

EUROSTAT, 2015. http://ec.europa.eu/eurostat/data/data

base (accessed 20 February 2018).

Hasanbeigi, A., Price, L., 2012a. A review of energy use

and energy efficiency technologies for the textile

industry. In Renewable and Sustainable Energy

Reviews. 16, 3648–3665.

Hasanbeigi, A., Hasanabadi, A., Abdorrazaghi, M., 2012b.

Comparison analysis of energy intensity for five major

sub-sectors of the Textile Industry in Iran. In Journal

of Cleaner Production. 23, 186-194.

Hong, G.B., Su, T.L., Lee, J.D., Hsu, T.C., Chen, H.W.,

2010. Energy conservation potential in Taiwanese

textile industry. In Energy Policy. 38, 7048-7053.

ITMF (International Textile Manufacturers Federation),

2003. International Production Cost Comparison -

Spinning/Weaving/Knitting. Zürich.

ITMF (International Textile Manufacturers Federation),

2014. International Production Cost Comparison –

Spinning/Texturing/Weaving/Knitting. Zürich.

Kaplan, E., Koç, E., 2010. An Investigation of Energy

Consumption in Yarn Production with Special

Reference to Open-End Rotor Spinning. In FIBRES &

TEXTILES in Eastern Europe. 18(79), 7-13.

Kim, S.Y., Grady, P.L., Hersh, S.P., 1983. Energy

consumption and conservation in the fibre-producing

and textile industries. In Textile Progress. 13(3), 1-14.

Koç, E., Kaplan, E., 2007. An Investigation on Energy

Consumption in Yarn Production with Special

Reference to Ring Spinning. In FIBRES & TEXTILES

in Eastern Europe. 15(63), 18-24.

Kumar, S., Visvanathan, C., Priambodo, A., 1999. Energy

and environmental indicators in the Thai textile

industry. Asian Inst. Technol. Thailand.

Lin, B., Zhao, H., 2016. Technology gap and regional

energy efficiency in China's textile industry: A non-

parametric meta-frontier approach. In Journal of

Cleaner Production. 137, 21-28.

Ozturk, E., Karaboyacı, M., Yetis, U., Yigit, N.O., Kitis,

M., 2015. Evaluation of Integrated Pollution

Prevention Control in a textile fiber production and

dyeing mill. In Journal of Cleaner Production. 88,

116-124.

Palamutcu, S., 2010. Electric energy consumption in the

cotton textile processing stages. In Energy. 35, 2945-

2952.

Pardo Martínez, C.I., 2010. Energy use and energy

efficiency development in the German and Colombian

textile industries. In Energy for Sustainable

Development. 14, 94-103.

Scheffer, M.R., 2015. In-depth Assessment of the Situation

of the T&C Sector in the EU and Prospects - task 7 -

Synthesis report for the European textile and clothing

sector. Final report ENTR/2010/16, 2012; Ref.

Ares(2015)2204860. Saxion Universities, The

Netherlands.

Steinberger, J.K., Friot, D., Jolliet, O., Erkman, S., 2009.

A spatially explicit life cycle inventory of the global

textile chain. In Int J Life Cycle Assess. 14, 443-455.

Tarakçıoğlu, I., 1984. Energy Consumption and

Conservation of Textile Finishing Mills. Uludag

University Press, Bursa. Turkish.

UNIDO (United Nations Industrial Development

Organization), 2010. Global Industrial Energy

Energy Consumption Characterization based on a Self-analysis Tool: A Case Study in Yarn Manufacturing

Efficiency Benchmarking - An Energy Policy Tool

Working Paper. Vienna.

Van der Velden, N.M., Patel, M.K., Vogtländer, J.G.,

2014. LCA benchmarking study on textiles made of

cotton, polyester, nylon, acryl, or elastane. In Int J Life

Cycle Assess. 19, 331-356.

Visvanathan, C., Kumar, S., Priambodo, A., Vigneswaran

S., 2000. Energy and Environmental Indicators in the

Thai Textile Industry. In Sustainable Energy and

Environmental Technologies. Proceedings of the Third

Asia-Pacific Conference. Hong Kong, 2000 Dec 3-6,

China, pp. 524-528. Hu X, Yue PL editors.

WTO (World Trade Organization), 2017. World Trade

Statistical Review, Geneva (Switzerland).

Zamani, B., Svanström, M., Peters, G., Rydberg, T., 2014.

A Carbon Footprint of Textile Recycling - A Case

Study in Sweden. In Journal of Industrial Ecology.

19(4), 676-687.

SMARTGREENS 2019 - 8th International Conference on Smart Cities and Green ICT Systems