Social life cycle assessment case studies from the textile and energy sectors

Generation in a Solar Power Plantin Spain —A Life Cycle Perspective Blanca Corona and Guillermo San Miguel Abstract This publication demonstrates the practical application of Social Life

Environmental Footprints and Eco-design

of Products and Processes

of Products and Processes

Series editor

Subramanian Senthilkannan Muthu, SgT Group and API,

Hong Kong, Hong Kong

of products, development of environmental and ecological indicators and eco-design

of various products and processes Below are the areas fall under the aims and scope

of this series, but not limited to: Environmental Life Cycle Assessment; Social LifeCycle Assessment; Organizational and Product Carbon Footprints; Ecological,Energy and Water Footprints; Life cycle costing; Environmental and sustainableindicators; Environmental impact assessment methods and tools; Eco-design(sustainable design) aspects and tools; Biodegradation studies; Recycling; Solidwaste management; Environmental and social audits; Green Purchasing and tools;Product environmental footprints; Environmental management standards andregulations; Eco-labels; Green Claims and green washing; Assessment of sustain-ability aspects

More information about this series athttp://www.springer.com/series/13340

Social Life Cycle Assessment

Case Studies from the Textile and Energy Sectors

123

Subramanian Senthilkannan Muthu

SgT Group and API

Hong Kong, Hong Kong

ISSN 2345-7651 ISSN 2345-766X (electronic)

Environmental Footprints and Eco-design of Products and Processes

ISBN 978-981-13-3232-6 ISBN 978-981-13-3233-3 (eBook)

https://doi.org/10.1007/978-981-13-3233-3

Library of Congress Control Number: 2018961219

© Springer Nature Singapore Pte Ltd 2019

This work is subject to copyright All rights are reserved by the Publisher, whether the whole or part

or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.

The use of general descriptive names, registered names, trademarks, service marks, etc in this

the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made The publisher remains neutral with regard to

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

The lotus feet of my beloved Lord Pazhaniandavar

My beloved late Father

My beloved Mother

My beloved Wife Karpagam and Daughters —Anu and Karthika

My beloved Brother

Social Performance of Electricity Generation in a Solar Power

Plant in Spain—A Life Cycle Perspective 1

Blanca Corona and Guillermo San Miguel

Socio-Economic Effects in the Knitwear Sector—A Life

Cycle-Based Approach Towards the Definition

of Social Indicators 59

Maria Ferrante, Ioannis Arzoumanidis and Luigia Petti

Social Life Cycle Assessment of Renewable Bio-Energy

Products 99

A Saravanan and P Senthil Kumar

vii

Generation in a Solar Power Plant

in Spain —A Life Cycle Perspective

Blanca Corona and Guillermo San Miguel

Abstract This publication demonstrates the practical application of Social LifeCycle Assessment (S-LCA) methodology in the analysis of a 50 MWeConcentrating Solar Power (CSP) plant located in Spain The assessment makes use

of two complementary analytical approaches: (1) a generic social hotspot analysisbased on the social risks related to financial flows generated by the provision ofgoods and services taking place during the life cycle of the power generationsystem, and then (2) a site-specific analysis focussing on the social performance ofthe construction/energy company involved in the construction and operation of thepower plant The site-specific analysis followed the procedures proposed by UNEP/SETAC but included a new classification/characterization model suited to theparticularities of the project and the energy sector The analysis considered fourstakeholder categories (workers; local community; society; and value chain actors)and used the number of worker hours as activity variable for the quantification ofsocial risks Worker hours attributable to each of the stages of the life cycle of theCSP system were calculated using input-output (IO) analysis The impact assess-ment phase of the S-LCA was carried out using a Social Performance Indicator(SPI), which required the estimation of performance reference points for a series ofindicators/subcategories proposed by the UNEP/SETAC Guidelines The SPI cal-culated for the CSP plant (+0.388 for a±2 range) suggested that the use of solarpower results in an increase of social welfare in Spain, primarily with regards tosocioeconomic sustainability and fairness of relationships The inventory data used

in the social hotspot analysis were monetary flows attributable to each of theprocesses considered in the life cycle of the power system These flows wereassigned to the corresponding sector of the producer country The Social Hotspot

B Corona ( &)

Copernicus Institute of Sustainable Development,

Utrecht University, Utrecht, The Netherlands

e-mail: b.c.coronabellostas@uu.nl

G San Miguel

Department of Chemical and Environmental Engineering,

ETSII, Universidad Polit écnica de Madrid, ES28006 Madrid, Spain

e-mail: g.sanmiguel@upm.es

© Springer Nature Singapore Pte Ltd 2019

S S Muthu (ed.), Social Life Cycle Assessment, Environmental Footprints

and Eco-design of Products and Processes,

https://doi.org/10.1007/978-981-13-3233-3_1

1

Database (SHDB) was used to link these demand values to social risks andopportunities The results showed that the life cycle phase contributing the most tothe social risk of the solar power system was operation and management This isdue primarily (over 75% of the weighed risk) to the social risks associated with thesupply chain of the natural gas used as auxiliary fuel For Spain, the main socialrisks associated with the solar power plant were related to gender inequality andcorruption, and to a lesser extent to injuries and immigrants Some of these riskswere confirmed in the site-specific assessment The paper ends with a discussionabout the application of Multi-Criteria Decision Making (MCDM) for evaluatingthe results obtained in this Social-LCA in combination with environmental andeconomic oriented LCA.

Keywords S-LCA
Electricity
Social performance
Spain
Social risksStakeholders

1 Introduction

The UN World Commission on Environment and Development (WCED), alsoknown as the Brundtland Commission, developed between 1983 and 1987 thegrounds for the modern interpretation of sustainability In its final report “OurCommon Future”, the Brundtland Commission produced a definition of SustainableDevelopment that is still widely accepted today:“the development that meets theneeds of the present without compromising the ability of future generations to meettheir own needs” (WCED1987) That report also stated that the concept of sus-tainability rests on three elements: economic growth, environmental protection andsocial equality

At present, the Sustainable Development Goals (issued by the United Nations in

2015, and a continuation of the Millennium Development Goals) (Biermann et al

2017) are in the front line of international, national and local agendas Publicadministrations and customers are exerting pressure on companies to ensure that theprinciples of sustainable development are incorporated into the goods and servicesthat they supply to the market The practical application of this ambition necessarilyentails the use of a systematic methodology capable of quantifying the sustain-ability of specific goods and services in an objective manner

A holistic methodology referred to as life cycle sustainability assessment(LCSA) is currently under development with the purpose of integrating the threepillars of sustainability under a coherent life cycle approach UNEP/SETAC LifeCycle Initiative states in its report “Towards a Life Cycle SustainabilityAssessment” that LCSA may be seen as the summation of three analysis tools:Environmental Life Cycle Analysis (E-LCA), Life Cycle Costing (LCC) and SocialLife Cycle Analysis (S-LCA) (UNEP/SETAC2011) This concept is illustrated inequation SLCA = E-LCA + LCC + S-LCA

A more advanced and flexible approach to LCSA was developed under theCoordination Action for innovation in Life Cycle Analysis for Sustainability(CALCAS) (2006–2009) project (Heijungs et al 2009) This new conceptualframework relies on expansion of the scope of conventional E-LCA to incorporatethe economic and the societal dimensions of the system under consideration.The CALCAS project approach provides the practitioners with moreflexibility inthe selection of the analytical tools employed to evaluate different aspects of thesystem and provides an integrated framework where the results may be evaluated as

a whole (Guinée et al.2011)

The ultimate purpose of S-LCA is to assess the effect of a given product onhuman wellbeing As the name suggests, the analysis applies a life cycle approachthat takes into consideration social and socio-economic effects associated with theextraction and processing of raw materials required for the fabrication of the pro-duct, manufacturing activities, transportation and distribution, utilization and anyend-of-life actions that may be associated with the product (reuse, recycling andfinal disposal) These effects considered in S-LCA are primarily those generated bythe companies participating in the different stages of the life cycle of the productunder consideration.1This performance has an effect (positive or negative) on thewellbeing of a series of stakeholders, which typically include Consumers, Workers,Local Community, Value Chain Actors and Society

S-LCA may be used on its own or, as described above, it may be part of abroader Life Cycle Sustainability Assessment (LCSA) (Guinée et al.2011; UNEP/SETAC 2011) The scientific community recognizes E-LCA and LCC as maturemethodologies, while S-LCA is usually regarded as being at an early stage ofdevelopment in terms of methodological harmonization and acceptance (Cinelli

et al.2013)

1.1 Key Methodological Issues in S-LCA

Since its inception in 2002, the UNEP-SETAC Life Cycle Initiative has guished itself as a key promoter and developer of S-LCA methodology TheGuidelines for Social Life Cycle Assessment of Products (from now on the S-LCAGuidelines) have become a landmark and a key reference in the field (UNEP/SETAC 2009) This methodology operates on the principles of ISO 14040 and

distin-14044, with the typical four interrelated phases: (i) identification of goal and scope,(ii) inventory analysis, (iii) impact assessment and (iv) interpretation The practicalapplication of these guidelines is facilitated with the Methodological Sheets forSub-Categories in Social Life Cycle Assessment (S-LCA) (UNEP/SETAC2013)

In the identification of goal and scope phase, the S-LCA practitioner needs to set

up the basis of the investigation including identifying the objectives of the

1 In addition, the ultimate utility of the product may also be considered in the analysis

assessment, describing the system under investigation and identifying the socialissues of concern that would be evaluated These social issues can be referred to asImpact Categories, and the S-LCA Guidelines suggest some of them, includinghuman rights, working conditions, health and safety, cultural heritage, governance,and socio-economic repercussions.

The S-LCA Guidelines also define five groups of stakeholders including:(i) worker, (ii) consumer, (iii) local community, (iv) society, and (v) value chainactor Each of these categories is linked to a series of sub-categories describingsocial aspects that may have an effect on these stakeholder For instance, thestakeholder category“workers” includes the following sub-categories: (i) Freedom

of Association and Collective Bargaining, (ii) Child Labour, (iii) Fair Salary,(iv) Hours of Work, (v) Forced Labour, (vi) Equal Opportunities/Discrimination,(vii) Health and Safety, (viii) Social Benefit/Social Security These subcategoriesmay be characterized by a series of indicators The Methodological Sheets produced

by UNEP/SETAC provides indications about the most suitable indicators that may

be used to evaluate these sub-categories (UNEP/SETAC2013)

Depending on the goal of the S-LCA, but also on the availability of data, timeand economic resources, the assessment may be carried out following two differentapproaches The generic S-LCA approach relies on generic data describing thesocial risks or opportunities associated with the country specific sectors whereactivities or unit processes of the life cycle take place This generic approach can becarried out using databases, such as the Social Hotspot Database (SHDB)(Benoit-Norris et al 2013) (http://www.socialhotspot.org) and the Product SocialImpact Life Cycle Assessment (PSILCA) (https://psilca.net/) The site-specificS-LCA approach is carried out at a company level and involves an investigation ofthe social performance of the organizations involved in the life cycle of the systemunder investigation (Dreyer et al 2006; Macombe et al 2013; Martínez-Blanco

et al 2014) Company site-specific inventory data refers to the life cycle of thesystem under investigation (company, product and location), making this step verytime consuming and very demanding of time and human resources

The aim of the impact assessment phase in the S-LCA is to transform theinventory data into a set of social performance indicators This may be achievedusing one of two methodological approaches (Parent et al.2010): thefirst is usuallycalled the Taskforce approach (referred to as Type 1) and is aimed at assessingsocial performance; and second is called the Impact Pathway approach (referred to

as Type 2) and it is aimed at assessing social impacts In the characterization step,the Taskforce approach relies on the use of Performance Reference Points (PRP) toquantify the importance of the data collected throughout the inventory phase In thiscase, an activity variable can be used to reflects the relative importance of specificprocesses in the life cycle of the product

Different authors state their preferences regarding these two methodologicalapproaches The main argument in favour of the Taskforce approach is that

“cause-effect relationships are not simple enough or not known with enough cision to allow quantitative cause-effect modelling’’ (Chhipi-Shrestha et al 2014,UNEP-SETAC Life Cycle Initiative 2009) Regarding the impact pathway

pre-approach, Dreyer et al (2006) discusses the inability of social damage indicators(QALY—Quality Adjusted Life Years) to measure the social performance ofcompanies and organization.

Despite its early stage of development, most authors agree that S-LCAmethodology is already at a point where it may used to address the social area ofLCSA, primarily in simple systems Further testing and refining will be required inorder to increase precision and permit the analysis of systems that are more complexand sophisticated (Ekener-Petersen 2013; Ekener-Petersen and Finnveden 2013;Macombe et al.2013)

1.2 Social and Sustainability Assessment

of Solar Power Plants

According to a review by (Petti et al.2014), the amount of publications describingthe social life cycle assessment of goods and services is very limited This authoridentifies 7 publications dedicated to the S-LCA of energy products and tech-nologies, 7 papers on information and communication technologies, 7 more onproducts from the agri-food sector, 5 on waste management and a few others onother varied subjects Mattioda et al (2015) has also published a review on S-LCAidentifying 99 publications related to S-LCA, 13 of which describing the application

of the S-LCA methodology to specific products or services, while the others wererelated to theoretical and methodological issues The case studies described by thisauthor focus on the energy sector (3 on biofuels and 1 on diesel and petrol) while therest of papers where related to the manufacturing sector (4 papers), agriculturesector (2 papers), packaging (2 papers) and waste management (2 papers) At thepresent, there have been five S-LCA publications assessing energy systems, inparticular, three on biofuels, 1 on diesel and petrol and 1 on photovoltaic systems.Five S-LCA specifically dedicated to energy systems include those published byEkener-Petersen et al (2014), Macombe et al (2013), Manik et al (2013a, b),Traverso et al (2012)

The term Concentrated Solar Power (CSP) is employed to describe a range oftechnologies designed to produce electricity using direct solar radiation Theoperating principles of these plants are well documented in various publicationsincluding (Fuqiang et al.2017; Heller2017; Lovegrove and Stein2017; San Miguel

et al.2015) CSP plants have two components: the solarfield and the power block.The solarfield consists of an array of mirrors designed to concentrate the radiationfrom the sun into a receiver A thermalfluid captures this radiative energy in theform of thermal energy, thus increasing its temperature as it makes its way throughthe solarfield A heat engine (usually in the form of a Rankine cycle) transformsthis thermal energy into electricity using a generator The form of radiation mosteffectively utilized by CSP plants is direct normal irradiance (DNI) (Meyer et al

2012)

Depending on the geometric nature of the receiver, CSP plants are usuallyclassified into two broad categories CSP plants based on linear receivers (parabolictroughs and Fresnel collectors) are the most commercially proven (primarily theformer) CSP plants based on point receivers (including central tower solar plantsand dish/engine systems) are less widespread, despite the higher concentrationratios and temperatures that may achieve (NREL2018).

The world leaders in CSP technology are Spain and the USA, accumulatingmore than 90% of the installed capacity worldwide Only Spain has 50 commercialCSP power plants totalling 2300 MW of installed capacity (PROTERMOSOLAR

2018) Other countries with high solar resources in the form of DNI (such as India,Chile and South Africa) already have or have announced the construction of newCSP plants

A key problem with solar energy is that it is intermittent by nature One way ofsolving this issue is by incorporating thermal energy storage (TES) systems Thesesystems are charged during the day using an extended solarfield, allowing the plant

to extend its operating hours TES systems are usually based on the use of moltennitrate salt Additionally, CSP plants may be hybridized with auxiliary fuels thatsupplement the solar radiation when it is not available This may be done byincorporating a combustion system to the HTF circuit The nature of the auxiliaryfuel has a notorious effect on the economic, environmental and social performance

of the CSP plant

The fact that CSP uses solar radiation as energy resource does not mean that itdoes not produce any negative impacts on the environment The environmentalperformance of CSP plants has been investigated in various publications using lifecycle methodology (Burkhardt et al.2011,2012; Corona et al.2014,2016c; Desideri

et al.2013; Klein and Rubin2013; Lamnatou and Chemisana2017; Lechón et al

2008; Piemonte et al.2011,2012; San Miguel and Corona2014) The results aresubjected to some variability due to differences in plant configuration, availability ofsolar resources and LCA methodology In general, the analyses have shown verylow global warming potential (between 25 and 50 kg CO2eq/MWh for plantsoperating with solar energy only, and higher values for hybrid plants depending onits solar factor)

The economics of these installations has also been the subject of various lications Information about the viability and economic costs of the technology may

pub-be found in (IRENA2018,2012; San Miguel and Corona2018) The application oflife cycle costing methodology to CSP plants may be found in Corona et al.(2016a) The social performance of this type of technology is also evaluated inCorona et al (2017)

This chapter is describes the application of S-LCA methodology to evaluate thesocial performance of a solar power plant, representing those deployed in Spain inthe past decade The methodological approach of this assessment has been carriedout in coherence with earlier life cycle based investigations covering the environ-mental and economic dimensions of the same solar power plant, and has beenpreviously described in Corona et al (2017) This chapter has been structured intofive sections as follows Section1 describes the state of the art of S-LCA

methodology and of the solar power system investigated Section2 explains theobjectives, methodological approach and operational decisions taken to carry outthe generic and the site specific S-LCA Section3 describes the results of the twoS-LCA and provides a discussion describing in combination the outcome of thisassessment Finally, Sect.4 provides a set of conclusions focusing on themethodological objectives of the exercise and also on the description of the socialperformance of the solar power plant.

2 Methodology

This section describes the practical implementation of S-LCA to assess the socialconsequences of the solar power plant The investigation has been carried out intwo steps following two different, but complementary, methodological approaches.Thefirst one is a generic approach using the SHDB aimed at evaluating the exis-tence of social hotspots in the life cycle (value chain) of the plant, while alsohelping to prioritize data collection for the second approach Social hotspots are

defined in the S-LCA Guidelines as “specific situations within a region that can beregarded as a problem, a risk or an opportunity in terms of social concern”(UNEP/SETAC 2009) The second is a site-specific approach following the rec-ommendations stated in the S-LCA Guidelines aimed at evaluating the socialperformance of the organizations involved in the life cycle of the solar plant.Both methodological frameworks were based on the principles of ISO 14040,which was adapted to the particularities of the social assessment approach, thespecific characteristics of the system under investigation (primarily in terms ofinventory data accessibility) and the limitations of the analysis team in terms oftime/budget availability

This section has been structured following the four classical steps described inISO 14040 for life cycle assessment: Definition of objectives and scope; social lifecycle inventory analysis; social life cycle impact assessment; and interpretation

2.1 De finition of Objectives and Scope

2.1.1 Definition of Objectives

The main objectives of this investigation are as follows:

• To explore the practical application of the Social Hotspots Database (SHDB) toproduce a generic assessment of social risks associated with the life cycle of thesolar power plant

• To explore the practical application of the S-LCA Guidelines to produce asite-specific S-LCA of the solar power plant

• Based on the generic and the site-specific S-LCA analyses, to evaluate the socialand socio-economic performance of the solar power technology in Spain.

• To evaluate the integration of S-LCA results into a broader sustainabilityanalysis covering the environmental, social and economic dimensions

The solar power plant investigated in this chapter has also been analysed for itsenvironmental and economic performance using life cycle based methodology.Environmental Life Cycle Assessment (E-LCA) was used to evaluate the envi-ronmental dimension (Corona et al.2014; Corona and San Miguel2015) and LifeCycle Costing (LCC) and Multiregional Input/Output (MRIO) were used to eval-uate the economic dimension (Corona et al.2016a,2017) Since the ultimate goal ofthis series of investigations is to analyse the overall sustainability of the system, theapproach employed in this S-LCA was consistent with the decisions taken inprevious investigations in aspects such as system characteristics, system boundariesand functional unit

2.1.2 Characteristics of the System

Figure1shows an aerial view of the system investigated in this publication, and amap showing its geographical location The system is a commercial hybridConcentrating Solar Power (CSP) plant based on parabolic trough (PT) technologybased in Ciudad Real (Spain) The plant represents the CSP configuration mostwidely deployed in Spain over the past decade As a reference, 96% of the CSPcapacity installed in Spain (45 of the 49 plants) are based on PT technology (SanMiguel and Corona2018) Other CSP technologies less mature and widely repre-sented include Linear Fresnel Reflector Systems, Power Tower Systems and Dish/Engine Systems (NREL2018)

The CSP plant investigated in this publication entered into operation in 2011, ithas a nominal capacity of 50-MWe, it extends over 200 ha of unproductive ruralland and has a lifetime expectancy of 25-year Table1 describes the technicalcharacteristics of the solar plant and Fig.2 illustrates its main components: solar

Fig 1 Aerial view and

location of the CSP plant

investigated in this S-LCA

Source aerial image: BSMPS

2009

field, heat transfer fluid (HTF) circuit, thermal energy storage (TES) system, iliary natural gas boiler and power block.

aux-The solar field is made of 624 SENERTROUGH parabolic trough collectorsassembled into 156 loops providing a total aperture of 510,120 m2 The collectors

Table 1 Technical characteristics of the CSP plant under investigation

Thermal ef ficiency of the cycle (η) 35 %

Normal direct irradiance 2030 kWh/m2
year

Thermal storage capacity 7.5 hour

NG input (for power generation) 3.01E+08 MJ/year

NG input (for maintenance) 6.28E+06 MJ/year

Total NG consumption 7.87E+06 Nm3/year

Full load equivalent operation 3290 h/year

Gross electricity generation 194,926 MWh/year

Electricity self-consumption 31,188 MWh/year

Net electricity generation 163,738 MWh/year

Direct water use 988,660 m3/year

Fig 2 Flow diagram describing the different components and operation of the hybrid CSP plant

are mounted on stainless steel structures with sun tracking systems that maximizethe concentration of direct solar irradiation into a tube receiver The Heat TransferFluid (HTF) circulating inside the receiver absorbs the radiating energy from thesun raising its temperature from 285 °C at the entrance of the solarfield to 395 °C

at the exit In the power block, the hot HTF circulates through various heatexchangers to generate superheated steam at 100 bar/375 °C This steam drives aturbine associated with a generator for electricity generation, as in conventionalpower plants The Rankine cycle is cooled using forced-draft evaporative tech-nology, resulting in a thermal efficiency (η) of 36.8% (San Miguel et al.2015).The plant also incorporates a thermal energy storage (TES) system based ontwo-tank molten salt (nitrate) technology This technology stores thermal energygenerated in the solar field during the day for use during periods of reduced irra-diation (at night or during cloudy episodes) in order to increase stability andaugment the operating time of the plant Additionally, the CSP plant incorporates anauxiliary boiler operating on natural gas that provides heat for maintenance activ-ities such as daily start-up operations, to avoid the freezing of the HTF and moltensalts during cold periods and to reduce system instability caused by transientclouds Natural gas consumption for these maintenance applications typicallyrepresent around 1% of the thermal requirements of the plant This auxiliary fuel isalso used as a complement to solar energy to extend with operation of the plant andgenerate additional electricity when the solar radiation is not available The Spanishlegislation regulating the generation of electricity from sustainable resourcesallowed CSP plants to produce up to 15% of their electricity from auxiliary fuels.This additional electricity was entitled to the same subsidy assigned to solar power(feed-in tariff of 26.9 c€/kWh—Royal Decree 661/2007) (San Miguel and Corona

2018)

As explained, the plant operates in hybrid mode with natural gas for a full-loadcapacity of 3290 equivalent hours per year and a gross electricity output of194,926 MWh/year Natural gas consumption amounts to 7.87 106

Nm3/year(equivalent to 3.01 108

MJ/year of auxiliary energy) Only 2.0% of this fuel isused for maintenance activities while the remaining 98.0% is used for extendedpower generation Net electricity sales (after subtracting power losses due to gridinefficiencies and onsite consumption) amount to 163,738 MWh/year Onsite wateruse is rather high due to the evaporative cooling technology employed in theRankine cycle at 988,660 m3/year

2.1.3 Description of the Life Cycle of the Solar Power Plant

Figure3shows the four stages in the life cycle of the CSP plant investigated in thisS-LCA including: (i) Extraction of raw materials and Manufacturing of components(E&M), (ii) Construction of the facility (C), (iii) Operation and Maintenance of thepower plant (O&M), and (iv) Dismantling and Disposal (D&D) A complete list ofall the unit processes considered in the analysis may be found in the economicinventories of Annex 1, with Table6 corresponding to the raw materials and

manufacturing of components phase, Table7to construction processes, Table 8tothe operation and maintenance phase and Table9 to end-of-life activities.

2.1.4 Scope of the Analysis

The scope of the generic S-LCA followed a cradle to grave approach, covering allfour stages in the life cycle of the solar power plant The transmission, distributionand utilization of the electricity were out of the scope of the analysis due to the factthat impacts associated with these elements are not affected by the characteristics ofthe power generation technology For all the unit processes included within theboundaries of the system, inventory data was available regarding economicflowsand country specific sectors where the transactions take place

The aim of the site-specific S-LCA is to explore the social and socio-economicperformance of the organizations responsible for the activities that make up the lifecycle of the system The promoter company is, without a doubt, the most importantorganization in the life cycle of the system, known to be responsible for the projectdevelopment, construction of the power plant, operation and maintenance, andend-of-life activities Key unit processes in the life cycle of the solar plant notassociated with the promoter include those associated with the extraction of rawmaterials (primarily natural gas employed as auxiliary fuel but also steel andconcrete for the solar collectors, glass and silver for the mirrors and nitrate salts forthe thermal energy storage system) and the manufacturing of certain plant com-ponents (e.g absorber tubes, steam turbine, solar tracker and electronics)

Gathering primary social data from every activity and supplier involved in theCSP life cycle would have required more time and economic resources than the

Fig 3 Life cycle diagram of the CSP power plant including economic, material and energy flows

available by the analysts at the time of the study Therefore, the scope of thesite-specific analysis was narrowed considering the findings in the social hotspotanalysis and the accessibility of data.

With the exception of natural gas, the generic S-LCA had shown that most of thesocial risks associated with the life cycle of the plant were associated with unitprocesses carried out by the promoter company As explained above, the solar plantinvestigated in this analysis operates in hybrid mode with natural gas, which isresponsible for 15% of the power generated This natural gas originates fromcountries (primarily Algeria but also Nigeria and Qatar) whose gas and energysectors are associated with high social risks (see Sect.2.2.2)

Based on this information, and being aware of the weaknesses associated withthis pronouncement, the analysis team decided to leave the life cycle phase “ex-traction of raw materials and manufacturing of components” out of the scope ofsite-specific S-LCA Furthermore, the system boundaries for other life cycle phases

in the solar power plant were limited to those unit processes carried out directly bythe promoter company, which was the only organization investigated at a companylevel

2.1.5 Function and Functional Unit

For the purpose of this investigation, the function of the solar power plant is toproduce electricity and the functional unit considered was 1 MWh of electricitypoured into the Spanish electricity grid This is consistent with the functional unitsemployed to evaluate from a life cycle perspective the environmental and economicperformance of the system

However, it should be discussed at this point that this functional unit does nottake into consideration some aspects of power generation that are essential to definethe adequacy of a power generation technology In other words, different powergeneration technologies are not necessary interchangeable solely on the basis oftheir capacity to generate electricity For example, aspects like dispatchability(ability to adapt power output to the required demand at any hour of the day withoutwasting primary energy) and firmness (ability to supply electricity during peakhours) (Servert et al.2016) may be essential to determine the ability of a plant toadapt effectively to a certain demand curve These attributes, which are not readilyquantifiable, may also affect the price at which electricity is sold to the market andthe revenues earned by the plant operator

The thermal energy storage in the solar power plant under investigation providesthis technology with a certain degree of dispatchability, which may not be attributed

to other renewable power generation technologies like photovoltaic or wind power.However, its capacity to generate on demand is not as good as that achieved bynatural gas in combined cycles, for instance The integration of these aspects intothe function and functional unit of the plant should be taken into consideration infuture LCA investigations

2.1.6 Selection of Impact Categories, Sub-Categories and Indicators

Regarding the generic S-LCA, the social risk assessment has been carried outconsidering 17 impact categories (Child Labor, Forced Labor, Excessive WorkingTime, Injuries & Fatalities, Toxics & Hazards, Poverty Wage2, Poverty Wage1,Migrant Labor, Collective Bargaining, Indigenous Rights, Gender Equity, HighConflict, Legal System, Corruption, Drinking Water, Improved Sanitation, HospitalBeds) grouped intofive damage categories (labour rights, human rights, health andsafety, governance and community infrastructure), as considered by New Earth intheir Social LCIA Method 1 v.1.0 (Benoit-Norris et al.2013) (Fig.4)

As illustrated in Fig.5, the site-specific S-LCA analysis was based on a series of

27 social impact sub-categories classified into the following five impact categories:Labour rights and decent work, Health and safety, Cultural and natural heritage,Fair relations and Socio-economic sustainability The sub-categories describedabove represent social attributes susceptible to be affected by the system The state

of these attributes were evaluated using 24 social indicators, including 11 tative, 10 semi-quantitative and 3 qualitative

quanti-This selection of impact categories, sub-categories and indicators was based onthe recommendations stated in the S-LCA Guidelines (UNEP/SETAC2009) andthe Methodological Sheets (UNEP/SETAC2013) but also took into considerationthe availability of inventory data, the results from the generic S-LCA, the particularcharacteristics of the system (importance of the unit processes making up the solarplant) and the potential vulnerability of the stakeholders involved Twenty six of thesub-categories selected are explicitly mentioned in the S-LCA Guidelines Oneadditional sub-category (product utility, ascribed to the stakeholder category

Fig 4 Social impact categories and sub-categories selected for the site-speci fic S-LCA

Society) was included due to the significance that the availability of electricity mayhave on the social wellbeing of a given community Table2 shows these 27sub-categories and their associated indicators, classified into the four stakeholdercategories proposed in the S-LCA Guidelines.

2.1.7 Critical Review

A critical review of the S-LCA was carried out by members of the Spanish NGOIngeniería Sin Fronteras (Engineers Without Borders) The reviewers involved inthe review had experience in the international implementation of sustainability anddevelopment projects related to the energy and electricity sectors

2.2 Social Life Cycle Inventory Analysis

2.2.1 Inventory Analysis for the Generic S-LCA

Background inventory data employed in these analyses were obtained from theSocial Hotspot Database (SHDB), which was integrated into Sima Pro v8.4 togetherwith the Social LCIA Method 1 v.1.0 The Social Hotspot Database (SHDB) is anextended input/output life cycle inventory database which contains data about the

Fig 5 Diagram showing the sub-categories and indicators considered to evaluate the impact category “Labour rights and decent work” (Adapted from Corona et al 2017 )

Table 2 Stakeholder categories, sub-categories and indicators employed to carry out the site-speci fic S-LCA

Subcategories Indicators

Workers Freedom of association

and collective bargaining

Existence of trade unions in the organization is adequately supported and workers are free to join them

% of af filiates of total employees Child labour Presence of child labour

Fair salary Wage inequality (average salary compared to

highest rank executive salary) Average annual salary Lowest paid worker Hours of work Hours of work

Forced labour Existence of forced labour

Equal opportunities/

discrimination

Employment rates of people with special needs with respect to the total employed people Men/women occupation ratio in the company Men/women executive managers ratio in the company

Health and safety Education, training, counselling, prevention and

risk control programs in place to assist workforce members

Presence of a formal policy concerning health and safety

Accident ratio per employee 2008 versus 2013 Social bene fit/social

resources

Based on information provided by local sources, it has been considered that the social attributes associated with the stakeholder category “local community ” are not affected by the solar plant This is so because the plant is far away from population centres (6 km from the closest village) and that the potential interactions with local people (except for workers, which are assessed in the workers stakeholder category) are very limited

social risk associated with 113 geographical regions (mainly countries) and 57economic sectors (Benoit-Norris et al.2013).

Input data is in the form of monetary units (2002 US$) spent on country-specificsectors (CSS) throughout the life cycle of the system under investigation Thesemonetaryflows are transformed into labour intensity data (worker hours), which isactually the activity variable employed to weigh the importance of the unit pro-cesses considered in the analysis

The economic inventory data regarding the construction, operation and mantling of the solar power plant was provided by an energy engineering consul-tancyfirm specialized in CSP technology (IDIE S.L.) This included informationabout the magnitude of the economic transactions associated with each of theelements comprising the value chain of the solar power plant and information aboutthe economic sectors and regions (countries) where this activity takes place.These economicflows were first converted from €2013 (data supplied by con-sultancyfirm) to US$2002 (units employed in SHDB) using Market exchange ratesand the OECD CPI index (OECD 2014) No discount and inflation rates wereconsidered infinancial transitions occurring at different times, since the magnitude

dis-of social issues is not necessarily related to the variation dis-of the value dis-of money overtime Annex1provides full details about the economic inventory with informationabout the specific SHDB dataset employed for each unit process The informationhas been grouped into four tables (Tables6,7,8and9) each one corresponding to adifferent stage in the life cycle of the system

national and international projects Investment in

R + D Prevention and

basic needs Value

One problem when gathering this inventory data was to trace with precision theorigin of each of the numerous elements and components that make the life cycle ofthe system Most of the components in the solar power plant are known to originatefrom Spain An exception to this is the heat transferfluid (produced in Belgium),the absorber tubes and the power block components including gas turbine (inGermany), and the nitrate salts for the TES (from Chile) Country-specific sectorSHDB datasets were used in the generic analysis of these items.

However, some other raw materials (e.g natural gas, steel, aluminium) andelementary plant components (e.g solar tracking systems, electronics) may alsooriginate from other countries Tracing this information is not a simple task due tothe diversity of providers and the confidential nature of this information Due to theimportance of natural gas in the social performance of the solar plant, the origin ofthis energy resource was evaluated in detail for the generic S-LCA

Total expenses associated with NG consumption in the solar plant were lated at 3,595,400 US$2002/year, assuming a fuel input of 7.87E+06 Nm3/year(see Table1) and a market price of 3.3875 c€/kWh, as reported by the SpanishMinistry of Industry and Energy (Ministry of Industry 2013) This source alsoinforms that 86.18% of industrial NG costs are attributable to the raw material(2.9194 c€/kWh) and 13.82% to its transport and distribution to the final user(0.4681 c€/kWh) Hence, in order to model the natural gas supply, the distributionprice was assigned to the sector Gas manufacture, distribution/ES (496,829 US

calcu-$2002/year), and the raw material component (3,098,571 US$2002/year) wasassigned to the country specific gas sector of each exporter country, considering thefollowing mix (MINETAD 2018): Algeria (37.1%), Nigeria (13.6%), Norway(9.39%), Qatar (9.65%), Trinidad & Tobago (6.03%), Peru (5.60%), Egypt (1.47%)and the Netherlands (23.19%) (MINETAD2018)

Monetary expenses associated with the consumption of industrial water marily for evaporative cooling in the Rankine cycle) was based on a unit price of0.50€/m3

(pri-, as reported by the Spanish Association for Water Supply and Sanitation(Ciudad Real, Spain) (AEAS2014) For the purpose of the generic S-LCA, Spainwas assumed to be the producer of all other raw materials and components whoseorigin was unknown

2.2.2 Inventory Analysis Site-Specific Assessment

The site-specific assessment was conducted in order to analyse at a company levelthe potential risks detected in the generic assessment Regarding the scope of thisinvestigation, it has been discussed above that it will only cover the activitiesundertaken by the promoter of the solar plant, who is also responsible for thedevelopment and construction of the installation, its operation and dismantling atthe end of its useful time

The promoter company belongs to a holding of companies operating primarily inthe Construction and Industrial Services sectors The promoter company also has anumber of subsidiary enterprises that operate in specific areas of this sector

Site-specific inventory data relating to the promoter company was obtained bysearching the internet, from direct communications with company members and byrevising certain corporate reports that were made available to the S-LCA analysts asfollows:

• The annual Corporate Social Responsibility (CSR) Report of the holding ofcompanies to which the promoter belongs, drafted following the premises of theGlobal Reporting Initiative (GRI) (year 2014)

• The annual Corporate Report of the promoter company (year 2013)

• The Collective Bargaining Agreement (CBA) drafted by a company that issubsidiary of the company responsible for the construction and operation of thesolar plant (year 2010)

Regarding the data quality of the site-specific inventory, it was decided that thedata employed would need to have been produced 5 years prior to the com-mencement of the solar power project (between 2008 and 2013)

2.3 Social Life Cycle Impact Assessment Modelling

The impact assessment stage of the S-LCA is used to transform inventory data intosocial impact values This section describes the characteristics of the impactassessment methods employed to carry out the generic and the site-specific socialassessments

2.3.1 Life Cycle Impact Assessment Method for the Generic Social

Analysis

The impact assessment method Social LCIA Method 1 v1.0 (Benoit-Norris et al

2013), based on New Earth’s Social Hotspots Index and adapted to SimaPro 8.4software, was used to carry out the hotspot analysis The input data for this mod-elling phase was in the form of monetary units (US$2002) spent on country specificsectors This method was also used to calculate worker hours associated with each

of the elements in the supply chain of the solar power plant Worker hours wasemployed as activity variable to aggregate the social performance of unit processes(and life cycle stages) that make the life cycle of the solar power plant

2.3.2 Life Cycle Impact Assessment Method for the Site-Specific SocialAssessment

Although the S-LCA Guidelines provide information about impact categories,sub-categories and indicators, it also recognizes that“there are no characterization

models between subcategories and impact categories that are generally accepted

by S-LCA practitioners” The impact assessment method employed in thesite-specific analysis was developed ad hoc for this investigation following therecommendations of the S-LCA Guidelines It follows a Type 1 approach and wasdesigned in order to be consistent with the approach employed in the genericS-LCA and also with Environmental-LCA and Life Cycle Costing (LCC) carriedout previously as part of a broader sustainability analysis of solar power plants Theaim was to offer a simple and transparent procedure, easy to understand bystakeholders, capable of transforming the site-specific inventory data into numericalvalues describing the social performance of the system on a selection of socialimpact categories

The impact assessment method includes a meaning assessment step, whichcombined the classification, characterisation, normalisation and weighting of theindicators to produce an aggregated value of social wellbeing This meaningassessment step was developed using as a reference the one proposed by Ciroth andFranze (2011a,b) However, our model includes the following improvements:– It represents both positive and negative impacts in the results,

– all the subcategories proposed in the S-LCA Guidelines have been included inthe analysis (even if they are not significantly affected by the system),– an aggregation step to combine characterized values throughout the life cyclestages of the system has been included using worker hours as activity variable,– the impact category “socio-economic sustainability” includes the new subcate-gory“product social utility”

In the classification phase of the meaning step, the 27 sub-categories selected forthis investigation (as described in Sect.2.1.6) were grouped intofive impact cat-egories As an example, Fig.5 illustrates the structure of subcategories and indi-cators used to characterize the impact category“Labour rights and decent work”.Table3 describes the classification of social impact sub-categories (correspondingindicators can be found in Table2) into impact categories, with reference to thestakeholder category affected in each case

The characterisation phase in the meaning step was aimed at transforminginventory data for each of the impact sub-categories considered into PerformanceReference Points (PRP) This was done using as a reference the average socialperformance of the sector where the activity under consideration takes place Sincethe only organization investigated at a company level was the promoter company,all the social performance values used to calculate PRP referred to different eco-nomic sectors in Spain This reference data was obtained from authoritative sourcesincluding the national statistics bureau, governmental reports, reports from rep-utable organizations and the national media The meaning assessment step wasoperated following the set of rules described below and illustrated in Fig.6:

Table 3 Classi fication of social impact sub-categories in impact categories, with reference to the stakeholder category affected in each case (Adapted from Corona et al 2017 )

Categories Subcategories Corresponding

Stakeholder category Labour rights and

decent work

Freedom of association and collective bargaining

Workers Child labour Workers Fair salary Workers Working hours Workers Forced labour Workers Equal opportunities/Discrimination Workers Delocalization and migration Local community Health and safety Health and safety Workers

Social Bene fit/Social security Workers Safe and healthy living conditions Local community Secure living conditions Local community Cultural and natural

heritage

Access to material resources Local community Cultural heritage Local community Respect of indigenous rights Local community Prevention and mitigation of armed

con flicts SocietyAccess to immaterial resources Local community Fairness of

relationships

Corruption Society Fair Competition Value chain actors Supplier Relationships Value chain actors Respect to intellectual property

rights

Value chain actors Promoting social responsibility Value chain actors Public commitments to sustainability

issues

Society Community engagement Local community Socio-economic

1 If the indicator value produced by the company was twice as good (or more) asthe Spanish average, the company was rated as“much better” and assigned ascore of (+2) social-performance points (s-pp) in that matter.

2 If the indicator value produced by the company under investigation was betterthan the Spanish average, the company was rated as “better” and assigned ascore of (+1) s-pp

3 When the indicator value was similar to the national average, the company wasrated as“similar” and assigned a score of (0) s-pp in that matter

4 If the indicator value was worse than the Spanish average, the company wasrated as“worse” and assigned a score of (−1) s-pp in that matter

5 If the indicator value was twice worse (or more) than the Spanish average, the

social-performance points (s-pp) in that matter

6 When there was not information available about a given indicator, but the socialrisk determined from the Social Hotspot Analysis was low for that subcategory,the company was rated as “similar” and assigned a score of (0) s-pp in thatmatter

7 When a subcategory was represented by more than one indicator, the score ofthe subcategory was calculated as the average of the s-pp assigned to eachindicator

Fig 6 Algorithm describing the characterization step in the life cycle impact assessment phase of the S-LCA

8 The score of each impact category was determined as the average of the scorescalculated for each of the sub-categories included, and considering equalweights.2

As described above, the operator company is the only organization investigated

in the site-specific S-LCA This company operates different unit processes thatbelong to different stages in the life cycle of the system (Construction, Operationand Maintenance, and Demolition stages) and to different sectors Hence, the sameorganization may obtain different characterized values for each unit process, as theyare affected by the performance of the organization in that specific element and also

by the average national performance used as a reference

The importance of each of the unit processes considered in the life cycle of thesolar plant was pondered using“worker hours” as activity variable The amount ofworker hours associated with these processes had been calculated previously to thisinvestigation using Multiregional I/O methodology (Corona et al.2016b) and usingthe economic inventory data described in Tables6,7,8 and9

Finally, the scores attained in each impact category were aggregated in aweighting step, resulting in afinal social score Since the S-LCA Guidelines pro-vide no indication about how to carry out the weighting step, the same importancewas allocated to all the impact categories considered, although it is acknowledgedthat different stakeholders may regard certain social categories as being moreimportant than others

This simplified impact assessment method estimates whether the system underconsideration (solar power plant) has a positive or a negative influence in the socialwellbeing of the country where the activity takes place (in this case, Spain)

A positive weighted final score means that the system is beneficial to the socialwellbeing of the country, due to the fact that the overall social conditions con-tributed by the system are better than the national average A negative weightedscore signifies that the social performance of the country would be damaged by theintroduction of the system into the national economy These results must be anal-ysed with care, and always presented together with the results obtained by eachsub-categories indicator It should also be noted that a characterized score close tozero for any given impact category or sub-category does not necessarily mean thatthe social performance of the organization is not detrimental (or beneficial) to thewellbeing the stakeholders affected, but that its performance is similar to theaverage weighted performance of the country in that specific social issue

2 Although the analysts considered equal weights for this assessment, an evaluation of the weights for each subcategory within each category would be necessary to properly represent the stake- holders ’ preferences.

2.4 Interpretation and Aggregation of Sustainability Results for Decision Making

As explained above, this S-LCA has been carried out as part of a more extensiveinvestigation aimed at describing the overall sustainability of different solar powertechnologies and configurations The investigation covered all three dimensions ofsustainability and it was carried out using a coherent life cycle approach that wouldallow integrating the result into a complete life cycle sustainability assessment(LCSA) Information about the environmental and economic analysis of the solarpower plant has been published elsewhere (Corona et al 2014, 2016a, b; SanMiguel and Corona2018)

The sustainability of the system is quantified through a series of indicators thatdescribe the economic, environmental and social consequences of different solarpower plant technologies and configurations The environmental dimension wasdescribed using the following indicators: Climate change (kg CO2eq/MWh), Waterstress (m3/MWh), Energy Payback Time (EPBT) (months) and Single score envi-ronmental impact (pt/MWh) All these indicators were calculated using conse-quentional environmental LCA The economic dimension was described using thefollowing indicators: life Cycle cost (€/MWh), Cost balance (€/MWh), Valueadded (%) and Multiplier effect Thefirst two indicators were calculated using LifeCycle Costing methodology and the latter two using I/O methodology Finally, thesocial performance was evaluated using the following indicators: social risk(pt/MWh) as determined using SHDB generic S-LCA, company social performance(Pt/MWh) as determined using site-specific S-LCA and employment generation (h/MWh) as determined using I/O analysis

The same analyses were carried out for three solar power technologies: CSP PTC(50 MWe based on parabolic trough collectors, like the one investigated in this casestudy), HYSOL BIO and HYSOL NG, both representing an innovative CSP con-figuration that delivers improved efficiency and power dispatchability (Corona et al

2016d; Nielsen et al.2016; Servert et al.2015,2016) The suffix NG indicates thatthe power plant uses natural gas as auxiliary fuel while the suffix bio indicates thatthis auxiliary gas is actually biomethane The objective was to provide information

in order to facilitate decision making on the most sustainable solar powerconfiguration

Aggregating the indicators obtained for the three sustainability dimensions into asingle sustainability score was a controversial issue due to the uncertainty accu-mulated by thisfinal score, and the subjectivity in weighting the different indicators.Instead, the procedure proposed by the analysis team involved applyingMulti-criteria Decision Making (MCDM) in the form of Analytic hierarchy process(AHP) This would be carried out by representatives of key stakeholders that should

be informed of the results obtained in the sustainability analyses This information

is facilitated by the distribution of clear, simple and unambiguous diagramsdescribing the technical results obtained for the social, economic and environmentaldimensions for each of the technology scenarios considered in the investigation

Figure12 shows the sustainability diagrams (sustainability crowns) proposed,which would be made available to stakeholder representatives to facilitate decisionmaking during the AHP exercise One diagram would need to be produced for each

of the technology scenarios considered in the analysis The diagrams are circular inshape and are they divided into three equal size sectors, each one corresponding to asustainability dimension The sectors contain the absolute numeric value calculatedfor each indicator, a colour indicator (from red to green) and a relative indicator (%)representing the relative performance of the technology scenario represented by thediagram compared to the average of all the alternatives The colour code and therelative value (%) assigned to each indicator represents its percentage differencerelative to the average of this indicator in all the alternatives/scenarios evaluated, asdetermined in Eq (1):

% change¼Alternative value Average value

The colour scale follows the traffic light code where green represents superiorand red inferior than average, in terms of sustainability performance Yellow colour(0% change) is assigned to values that are similar to average in the correspondingindicator If the sustainability performance of an alternative/scenario is better thanaverage, the colour of this indicator turns from yellow to green and its relative value

is positive If the performance is worse than average, the colour of the indicatorturns from yellow to red and its relative value is negative For instance, a tech-nology performing 100% better than average for a given indicator is assigned anintense green colour and if its performance is 100% worse, it receives intense red.The colour grade assigned depends on the relative value % assigned A moredetailed description about the construction of these sustainability crowns may befound in (Corona and San Miguel2018)

3 Results and Discussion

3.1 Generic Social Risk Assessment: Hotspot Analysis

Figure7 shows the characterised risks of the analysed system as calculated in thesocial hotspot analysis The results describe the contribution of each life cycle phase

to the social risks in 17 social impact categories The results evidence that most ofthe social risks are attributable to the O&M phase, especially due to risks related toForced Labour, Indigenous Rights, Poverty wage, and Gender equity The secondhighest life cycle phase contributor to social risks is E&M, mainly due to risksrelated with Toxics & Hazards (25%) and Improved Sanitation (20%)

The consumption of natural gas contributes to most (between 70 and 97%,depending on the social issue) of the weighted social risks observed in the O&Mphase

Figure8represents the social risks determined for each of thefive social impactcategories evaluated in the SHDB These values are determined by weighing andaggregating the results corresponding to each of the social issues considered Theresults suggest that the category with highest social risks is Health & Safety, fol-lowed by Labor Rights & Decent Work These highest risks are originated by thesupply chain of the NG and also by the banking services associated with thefinancing of the power plant.

When an economic sector is assigned a high social risk, this may be caused byone of two reasons: (i) a large amount of money spent on that sector throughout thelife cycle of the system and, (ii) a high level of social risk associated with thatcountry specific sector In the first case, a high amount of money spent is associatedwith a high number of worker hours in that sector When a sector presents socialrisks, an increased amount of worker hours means increased exposure to socialrisks, which is translated into high characterised social risks In the second case, ahigh level of risks for social issues would be characterised as very high, and present

Extraction of raw materials and Manufacturing Construction Dismantle & Disposal

Fig 7 Characterised social risks per life cycle phase of producing electricity with a CSP plant

Dismantle & Disposal Construction Extraction of raw materials and Manufacturing Operation & Maintenance

Fig 8 Weighted social risks per life cycle phase of producing electricity with a CSP plant

a high relative share of social risks even when the amount spent is not so high.These two cases are explained in more detail in the following paragraphs.

A closer look into the economic sectors providing goods and services to the lifecycle of the CSP plant evidenced that most of the social risks of the systemoriginate from four countries: Algeria, Spain, Peru and Egypt Except for Spain, thesocial risks attributable to these countries are associated, directly and indirectly,with the life cycle of the natural gas consumed in the solar plant In particular, theAlgerian commerce (Commerce/DZ) is by far the country specific sector con-tributing the most to the overall social risks of the system (29% of the totalweighted risk) Even though there is not a high amount of money spent in thissector, the share of weighted risks is very high due to the very high risks in multiplesocial issues in the sector

Although the CSP plant is located in Spain, and most of its components ufactured in the same country, there are only three Spanish economic sectorscontributing to more than 1% to the weighted social risks of the system: Financialservices nec/ES (8.2% of the total risks), Construction/ES (3.6% of the total) andBusiness services nec/ES (1.9%) The high risk attributable to the Financial servicesnec/ES and the Construction/ES are primarily due to the very large amount ofmoney spent on those sectors (128 M$2002 and 58 M$2002 respectively), which istranslated into a higher relative exposure of social risks to worker hours in thissectors According to the SHDB, these sectors in Spain present“very high risks” ofcorruption, injuries, unfair conditions for migrant workers, HIV and unemployment(only in construction) and high risks of forced labour and gender inequality.Figures9 and 10 show the characterised and weighted social risks associatedwith the life cycle phase Extraction of raw materials and Manufacturing (E&M).The power plant component presenting the highest weighted risk is the solarfield(34% of the social risks in E&M phase), followed by Thermal storage (31%), Powerblock (17%), HTF system (14%) and Facilities (3.4%)

man-The solar plant component contributing the highest to the social risks of thesystem is the solarfield (34% of the social risks in E&M phase) due primarily to theuse of metallic elements in the structure of the solar collectors (mainly steel but alsoaluminium) This is followed by the components in the Thermal storage system(31%) due to economic activities in Chile related to the purchase of nitrate salts.Comparative lower social risks are associated with the Power block (17%), the HTFsystem (14%) and the auxiliary Facilities (3.4%) of the CSP plant

The country economic sectors contributing the most to the weighted social risks

of the power plant belong to Spain, Chile, Angola and Mozambique The rence of social risks in thefirst two countries is due to the sectors directly involved

occur-in the manufacturoccur-ing of several components occur-in the CSP plant For occur-instance, thesystem uses molten salts extracted in Chile for the thermal storage system Thepresence of Angola and Mozambique in this list was somehow unexpected and ithas been traced back to indirect economic activities in the value chain of the plantcomponents Even though the amount of goods and services demanded fromAngola is relatively low (only 7090 $2002spent in the Commerce sector), the socialrisks in that country specific sector are very high in multiple social issues Angola’s

commerce sector is not directly linked to the manufacturing of components of theCSP plant, but according to the SHDB database, the Chilean Minerals nec sector(which is directly providing molten salts to the CSP plant) is importing mineralsfrom Angola However, it is known that all the molten salts employed in the CSPplant are extracted directly in Chilean territory (not imported from any othercountry with which the Chilean mineral sector may have other connections).Therefore, in this specific case, the social risks observed are an result of theaggregation of activities within the sectors of the SHDB, and the Commerce sector

Facilities Power Block Thermal storage Solar field HTF System

Fig 10 Weighted social risks for the extraction and manufacturing phase of producing electricity with a CSP plant

in Angola is surely not affected by the consumption of molten salts in the CSPplant.

Mozambique shows high risks via its Metals nec, Commerce, and Electricitysectors These risks are mainly due to the 43,000 $2002 demand from theMozambican sector Metals nec/MZ This sector is connected with the CSP plantthrough the Spanish sectors Metals nec, Electronic equipment and Machinery andequipment, that are importing metals from the aforementioned African country.According to reports from the Spanish Ministry of Foreign Affairs, the main metalthat Mozambique exports to Spain is aluminium (Ministry of Foreign Affairs2015),that is consumed in very limited amounts in the CSP plant Therefore, the riskscalculated by the hotspots analysis for Mozambique are probably overrated due tothe low contribution of this country to the actual components of the power plant.The country specific sector Metal products in Spain also shows a high share ofsocial risks attributable to the use of metals within the CSP plant According to theSHDB, the social profile of this sector is comparable to that of the Financialservices nec sector, except for unemployment (rated as very high risk) and genderinequality for workers (rated as high risk)

The results of the generic assessment are useful to identify social hotspots thatshould be further investigated in a site-specific assessment According to the resultsobtained in this analysis, the suppliers providing metal products, machinery andequipment in Spain to the CSP plant are the ones that should be investigated in asite-specific assessment The specific issues to be further investigated includecorruption, gender inequality, injuries and immigrants

3.2 Site-Speci fic Assessment

This section shows the results of the site-specific assessment of the solar powerplant based on the social performance of the promoter company Full informationabout the site-specific inventory data, the national data employed as a reference, andthe characterization terms (as much worse, worse, similar, better, much better)selected for each indicator may be viewed in Annex3(Table11Inventory data andcharacterisation of the Construction and Dismantling phases and Table12

Inventory data and characterisation of the O&M phase)

3.2.1 Meaning Assessment

The results for the characterization and weighing of the inventory data are described

in Fig.11 and Table5 respectively, following the eight rules of the meaningassessment step (see Sect.2.3.2)

As observed in Fig.11, each impact category analysed is depicted in a spiderdiagram containing the characterized values for every sub-category classified in thatcategory The characterized values for each sub-category are represented in a scale

from−2 to +2 s-pp (from worst performance to best performance with respect tothe national average performance) The impact category Cultural and NaturalHeritage was not depicted due to its low relevance in the case under study Twodifferent lines can be observed in the diagrams, one representing the characterized

C&D phases O&M phase

Fig 11 Site-speci fic characterisation of social issues for the life cycle phases considered in the site-speci fic assessment (C&D = Construction and demolition phases, O&M + Operation and Maintenance phase)

Table 4 Meaning assessment step: weighted results for each social category and life cycle phase analysed in the site speci fic assessment (every sub-category was weighted equally)

Impact categories Construction and

dismantling phases

Operation and maintenance phase Labour rights and decent work 0.14 −0.02

Health and safety 0.25 0.25

Cultural and natural heritage 0 0

Fairness of relationships 0.29 0.29

Socio-economic sustainability 1.38 1.38

values for the Construction and Demolition phases, and the other one the values forOperation and Maintenance of the power plant.

The weighted aggregated values (Table4) represent the social performance ofthe system in each of the impact categories considered, and they are calculated asthe average of the scores corresponding to each of the sub-categories in that impactcategory

The results evidence that the characterised values for most social impact categories were very close to zero or above zero, which suggests that the companyunder investigation rated similar or better than the Spanish average in terms ofsocial performance The only subcategories performing below the national averagewere Fair salary, Discrimination, Fair competition, and Corruption

sub-Fair salary is performing worse due to wage inequality, which is evidenced bythe large difference of salaries reported between the executive managers andaverage workers in the promoter company (771% higher compared to 134% higher

in the construction sector, INE 2014) Regarding Discrimination, the promotercompany has a ratio of 7.92 men to women in the workforce (Annual CorporativeReport), which is higher than the average ratio of 6.34 in the corresponding sector(INE 2015a, b) In addition, the ratio of men to women for executive managerspositions was rated as much worse, with a 22 men to women executive managersratio (Annual Corporative Report) compared to the 2.75 Spanish average ratioconsidering every sector (INE2015a,b) Since gender inequality was found to be asocial risks according to the hotspot analysis, and it also performed badly in thesite-specific assessment, it was further investigated by revising the companyreports However, the CSR Report of the business group to which the promotercompany belongs did not provide any indicator on this issue, even though the citedreport included a GRI certification providing information in many other indicators.The Fair competition category was measured by accounting for Legal actionstaking place during the reporting period related to anti-competitive behaviours Theinternet research carried out by the S-LCA analysis indicated two legal actionsrejected by the Comisión Nacional de los Mercados y la Competencia (CNMC)

Table 5 Weighting step and social performance results according to impact categories for the production of electricity with a CSP plant

Weighting

Worker hours 394,357 1,440,256 82,261 1,916,874 Weighting factors 0.21 0.75 0.04 1

Labour rights and decent work 0.0294 −0.015 0.0056 0.02 Health and safety 0.0525 0.1875 0.01 0.25 Cultural and natural heritage 0 0 0 0

Fairness of relationships 0.0609 0.2175 0.0116 0.29 Socio-economic sustainability 0.2898 1.035 0.0552 1.38 Total (mean average) 0.087 0.285 0.016 0.388

(National Commission for Markets and Competition) involving the promotercompany and one legal action executed against the business group of the promotercompany Since two out of three legal actions were rejected by the CNMC, thisindicator was rated as worse (and not as much worse).

Corruption is evaluated using a semi-quantitative indicator: “Have there beenany legal actions related to corruption during the reporting period?” The internetsearch showed several legal actions between 2010 and 2014, which have not beenresolved yet The company was accused of embezzlement of public funds andfraud Although this situation may be regarded as similar to the national average(very high social risk of corruption for Spain according to the SHDB), thissub-category has been rated as worse, since corruption represents a breach of law.The best ranked sub-categories were Public commitments to sustainabilityissues, Contribution to economic development, and Product social utility TheContribution to economic development was measured by the increase in thenational income as a result of the demand of goods or services generated by theproject, and was represented by the multiplier effect This effect was calculatedpreviously by an Input Output analysis (Corona et al 2016a, b, c,d) Since themultiplier effect of the power plant was 2.60, this indicator was ranked as muchbetter

The assessment of most of the subcategories affecting the local community didnot present significant results This is so because the CSP plant is located far frompopulation centres and its interaction with the local communities is limited (exceptfor employment, which is evaluated under the impact category Labour rights anddecent work) A visit to the power plant, which included informal conversationswith members of the local community about these matters, and revision of localnewspapers appear to confirm this limited interaction

Based on these results, it may be concluded that the social performance of thecompany could be improved by increasing salary and gender equality The pro-moter company should also improve their performance regarding fair competitionand legality, e.g by increasing transparency

3.2.2 Weighting and Aggregation of Site-Specific Results

Characterized site-specific results were weighted and aggregated to produce a singlesocial performance indicator of the system Worker-hours was used as activityvariable to evaluate the importance of each of the three life cycle phases considered.This was calculated using I/O methodology, as described elsewhere (Corona et al

2016b)

As shown in Table5, the total amount of worker hours associated with the lifecycle of the power plant amounts to 1,916,874 h, of which 75% correspond to theOperation and management life cycle phase, 21% to construction and 0.04% todemolition Worker hours were not calculated for the life cycle phase Extraction ofraw materials and Manufacturing (E&M) because this life cycle phase was notconsidered in the site-specific S-LCA The weighting factors were assigned in

accordance with the labour intensity through the estimation of the worker hourscorresponding to each life cycle phase The total aggregated result of the system is0.388 s-pp, which means that the solar power plant is beneficial to the socialwellbeing in Spain (note the range between 2 and −2 s-pp) The category pre-senting better social performance is Socio-economic sustainability (1.38 s-pp),followed by Fairness of relationships (0.29 s-pp) and Health and Safety(0.25 s-pp) The categories performing worst are Cultural and natural heritage andLabour Rights and Decent work The cultural and natural heritage is presenting aneutral performance (around 0 points) due to the similar ranking attributed to thecompany’s performance in this category (very similar to the national average inevery sub-category).

3.3 Interpretation of Results and Decision Making

As explained above, this S-LCA has been carried out as part of a broader tigation aimed at evaluating the sustainability of a range of solar power generationtechnologies Figure12 illustrates the sustainability diagrams (crowns) producedfor each of the solar power alternatives considered in the LCSA

inves-As observed in the diagrams, the solar power plant codenamed HYSOL BIOexhibited the best results in terms of environmental and social performance, with allthe indicators marked in green and representing better than average performance(except for company social performance, which produced the same value in all thealternatives) The solar fraction of this plant (55%) is significantly lower than in theconventional CSP plant (85%) and uses biomethane as auxiliary fuel However, thistechnology produced the worst results in terms of economic performance, partic-ularly in the indicator Cost balance

The substitution of biomethane with natural gas in HYSOL NG penalized theenvironmental and social performance of the technology, as demonstrated by thepredominance of red colours in the environmental and social sectors of the sus-tainability crowns HYSOL NG is similar to HYSOL BIO but using natural gas asauxiliary fuel The only exception is water stress, due to the higher water demandsgenerated by the biomethane life cycle In contrast, the HYSOL NG technologyexhibited improved economic performance primarily due to superior cost balancesand life cycle costs

The conventional CSP PTC investigated in this paper performed slightly betterthan average in all the environmental indicators (except for water stress due to use

of wet cooling instead of dry cooling in the Rankine cycle) and very similar toaverage in the social dimension Conventional CSP PTC performed worse thanHYSOL BIO but better than HYSOL GN in the environmental and social indica-tors, while the opposite was observed in the economic dimension

The results described in this section show that none of the solar power natives performed best in all the sustainability dimensions and indicators evaluated.Hence, selection of the most sustainable alternative is not straightforward, requiring

alter-the application of a Multi-Criteria Decision Making (MCDM) analysis, such as alter-theAnalytic Hierarchy Process (AHP) The MCDM analysis may be facilitated by theuse of the sustainability crowns described above, which could be used by stake-holders to interpret the results obtained in the LCSA.

4 Conclusions

This case study describes the practical application of Social—Life CycleAssessment (S-LCA) methodology for the analysis of a commercial solar powerplant located in southern Spain The analysis has been carried out following thestandard procedures described in ISO 14040, which have been applied in such a

Fig 12 Sustainability diagrams employed to facilitate decision making describing the mental, economic and social performance of producing electricity with three different types of solar power plants