This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

Cite thisDOI 101039c7ee00158d

Environmental life cycle assessment andtechno-economic analysis of triboelectricnanogeneratorsdagger

Abdelsalam AhmedDaggerab Islam HassanDaggerbc Taofeeq Ibn-MohammedDaggerde

Hassan Mostafafg Ian M Reaneyh Lenny S C Kohde Jean Zub andZhong Lin Wangai

As the world economy grows and industrialization of the developing countries increases the demand

for energy continues to rise Triboelectric nanogenerators (TENGs) have been touted as having great

potential for low-carbon non-fossil fuel energy generation Mechanical energies from amongst others

body motion vibration wind and waves are captured and converted by TENGs to harvest electricity

thereby minimizing global fossil fuel consumption However only by ascertaining performance efficiency

along with low material and manufacturing costs as well as a favorable environmental profile in

comparison with other energy harvesting technologies can the true potential of TENGs be established

This paper presents a detailed techno-economic lifecycle assessment of two representative examples of

TENG modules one with a high performance efficiency (Module A) and the other with a lower efficiency

(Module B) both fabricated using low-cost materials The results are discussed across a number of

sustainability metrics in the context of other energy harvesting technologies notably photovoltaics

Module A possesses a better environmental profile lower cost of production lower CO2 emissions and

shorter energy payback period (EPBP) compared to Module B However the environmental profile of

Module B is slightly degraded due to the higher content of acrylic in its architecture and higher electrical

energy consumption during fabrication The end of life scenario of acrylic is environmentally viable given

its recyclability and reuse potential and it does not generate toxic gases that are harmful to humans and

the environment during combustion processes due to its stability during exposure to ultraviolet

radiation Despite the adoption of a less optimum laboratory manufacturing route TENG modules

generally have a better environmental profile than commercialized Si based and organic solar cells but

Module B has a slightly higher energy payback period than PV technology based on perovskite-

structured methyl ammonium lead iodide Overall we recommend that future research into TENGs

should focus on improving system performance material optimization and more importantly improving

their lifespan to realize their full potential

a School of Materials Science amp Engineering Georgia Institute of Technology Atlanta Georgia 30332-0245 USA E-mail zlwanggatechedub NanoGenerators amp NanoEngineering Laboratory School of Mechanical amp Industrial Engineering University of Toronto Toronto M5S 3G8 Canadac Design amp Production Engineering Department Faculty of Engineering Ain Shams University Cairo 11535 Egyptd Centre for Energy Environment amp Sustainability The University of Sheffield Sheffield S10 1FL UKe Advanced Resource Efficiency Centre The University of Sheffield Sheffield S10 1FL UKf Department of Electronics and Communications Faculty of Engineering Cairo University Giza Egyptg Center of Nanoelectronics and Devices (CND) at Zewail City and AUC Egypth Departments of Materials Science amp Engineering University of Sheffield Sheffield S1 3JD UK E-mail imreaneysheffieldacuki Beijing Institute of Nanoenergy amp Nanosystems Chinese Academy of Sciences Beijing 100083 China

dagger Electronic supplementary information (ESI) available See DOI 101039c7ee00158dDagger A Ahmed I Hassan and T Ibn-Mohammed contributed equally to this work

Received 17th January 2017Accepted 22nd February 2017

DOI 101039c7ee00158d

rscliees

Energy ampEnvironmentalScience

ANALYSIS

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article OnlineView Journal

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

Broader contextThe ability of triboelectric nanogenerators (TENGs) to convert mechanical energy from various sources into useful electrical energy has drawn attention inrecent years Given their potential for low cost energy generation for self-powered applications it is important to assess their environmental profile and costviability by carrying out a detailed techno-economic lifecycle assessment This will provide an indication as to whether they constitute new environmentalchallenges or not In this paper a robust environmental life cycle assessment within a techno-economic framework is carried out for two TENG modules in thecontext of other energy harvesting technologies 11 lifecycle environmental metrics as well as the energy payback periods and carbon dioxide emission factorsare determined Although the environmental impact of both modules is lower compared to traditional PV technologies the higher quantities of acrylic in one ofthe modules along with energy-intensive fabrication led to a slightly higher environmental burden Material optimization focusing on reduced materialutilization as well as better fabrication processes should however improve their environmental profile Uncertainty sensitivity analysis is conducted to providedeeper insights into TENGs The current work therefore lays the foundation for future investigations into the profile of TENGs for environmentally friendlyinnovation in the energy sector To have a significant impact technological solutions capable of harvesting electricity from mechanical energy must also becompetitive within the marketplace

1 Introduction

The burning of fossil fuels is responsible for 480 of primaryenergy demands and current profiles reveal that the worldremains highly dependent on carbon-based power generationresulting in the emission of record levels of carbon dioxide (CO2)1

The growth of the world economy coupled with industrializationof the developing world has resulted in a demand for energy thatcontinues to increase2 Given the growing demand for energy anddwindling oil reserves the development of alternative sustainableenergy is of paramount importance Energy from solar wind andtidal waves has the potential to be integrated with electrical powergrids to meet mega- to gigawatt power requirements3 The overallrequirements for harvesting these forms of energy are based ona number of factors including low-cost high stability and highefficiency3

An increasingly wide range of mobile electronic devicesoften connected to the Internet of Things (IoT) have not onlymodified our way of life but also have created the need for ahighly diversified energy platform3 For applications such asmedical care4 healthcare monitoring infrastructure monitoringenvironmental protection and security many sensors computercontrol circuits and antennas are required Although the powerfor driving each miniature system is relatively small (from microto milli-Watt range)3 the collective number of units is forecastedby Cisco (the worldwide leader in information technology) to betrillions by the year 20205 The use of batteries to power theseunits is currently the default solution but this is not sustainablegiven the large number required and their limited lifespanMoreover the concept of the IoT will be rendered meaninglesswithout the inherent ability of devices to be self-powered Thischallenge has prompted the development of nanogenerators thatharvest mechanical energy from the surrounding environmentNanogenerators were first developed based on two effects namelypiezoelectricity6ndash12 and triboelectricity13ndash15 with the intention ofharvesting energy from activities such as walking talking typingand breathing A string of groundbreaking research advanceshave subsequently been reported since the landmark publicationsby Wang and Song13

The concept of the triboelectric nanogenerator (TENG) isbased on the use of the electrostatic charges created on thesurfaces of two dissimilar materials when they are brought into

physical contact the contact induces triboelectric charges andgenerates a potential drop when the two surfaces are separatedby a mechanical force causing electrons to flow between thetwo electrodes built on the two surfaces316 Following the firstpublication on TENGs in 2012 huge progress has been recordedFor instance by the year 2015 the areal power density hadreached 500 W m217 and the volume power density attainedwas 15 MW m3 with an instantaneous conversion efficiency ofaround 7018 TENGs boast a wide range of applications giventheir capability to harvest mechanical energy from a variety ofsources including body motions vibrations wind and waves19

Additionally the successful application of TENGs in self-poweredchemical sensors has recently been demonstrated20ndash22 for drivingelectrochemical processes23ndash25 and commercial light-emittingdiodes (LEDs)26ndash30

Several fabrication processes for TENGs have been describedin the extant literature Specifically four modes of operation ofTENGs including vertical contact-separation mode in-plane slidingmode single-electrode mode and free-standing triboelectric-layer mode were extensively described by Wang et al3 In thispaper attention is focused on two fabricated modules The firstis a thin-film-based micro-grating triboelectric nanogenerator(MG-TENG) The operation principle of the MG-TENG relies onthe coupling between electrostatic induction and the triboelectriceffect1731ndash34 Consisting of two sets of complementary micronsized electrode gratings on thin-film polymers the MG-TENGharvests energy by sliding these surfaces17 Based on previousresearch on this technology a 06 g MG-TENG with an overall areaof 60 cm2 and a total volume of 02 cm3 achieves an averageoutput power of 3 W (50 mW cm2 or 15 W cm3) and an overallconversion efficiency of roughly 50 which is sufficient to powerregular electronics such as light bulbs17 These performanceparameters highlight that MG-TENGs are a promising and effi-cient solution for harvesting energy from mechanical energyunder ambient conditions The second module is a TENG basedon two radially arrayed fine electrodes that generates periodicallychanging triboelectric potential and induces alternating currentsbetween the two electrodes As presented in previous work at arotation rate of 3000 rpm (rotation per minute) a TENG with adiameter of 10 cm achieves an output open-circuit voltage (Voc) ofaround 850 V and a short-circuit current (Isc) of around 3 mA at afrequency of 3 kHz Additionally with a load of 08 M the TENG

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

provides an average output power of 15 W (19 mW cm2) and theefficiency to an external load reaches 2426 Here we fabricatedthe same structure using a copper electrode on both the rotator andstator layers instead of a gold electrode which was used as the statorelectrode in the previous work26 In addition other devices with thesame structure fabricated based on copper instead of gold areused2535 and the average output power density of the fabricateddevice used in this study is the same as the previous one Moreoverthe small volume light weight low cost and high scalabilitycharacteristics make the TENG a suitable solution for harvestingmechanical energy for both small-scale self-powered electronics andpotentially in future larger scale energy generation

Given the potential of TENGs for low cost energy generationfor self-powered applications it is important to assess theirenvironmental profile and carbon footprint by carrying out adetailed lifecycle assessment (LCA) This will provide an indicationas to whether they constitute new environmental challenges or notA great deal of work has been published on the LCA of energyharvesting technologies However to the best of our knowledgeother than the comparative LCA of lead zirconate titanate (PZT) vspotassium sodium niobate (KNN) both potential materials forpiezoelectric energy harvesters36 no LCA work currently exists onmechanical energy harvesters such as the TENG Given the limitedenvironmental information on TENGs LCA is undertaken withinthe context of other energy harvesting technologies LCA involvesthe evaluation of the complete environmental impact of a materialor product from the raw material extraction phase through theprocessing as well as the usage phases and the final disposal37 It isan important technique that should be adopted to highlight theenvironmental hotspots in the production of consumer goods andtheir global environmental impact38 The use of LCA thereforedefines and addresses environmental sustainability issues that areessential for future development and upscaling Significantly per-haps it steers us clear of paths that will create new environmentalproblems while providing the necessary information with respect tothe consequences of material or device substitution

We live in a world dominated by networked product supplychains complex production technologies and nonlinear con-sumption patterns3940 It is essential therefore for consumersindustries and policy makers to have the right information inthe course of evaluating the environmental consequences ofsubstitute materials (from extraction design and fabricationprocesses to usage)36 To date a detailed cost estimation andtechno-economic evaluation and analysis of TENG modules hasnot been carried out Such an evaluation is vital regarding thefuture of TENG technology due to the urgent need to build acost-efficient industry that can survive with minimal governmentintervention41 Accordingly the power conversion efficienciesand the ensuing financial costs of two TENG module designswere analyzed and compared with existing energy harvestingtechnologies

In light of the above the rest of the paper is structured asfollows In Section 2 a brief description of the fabrication processesof both TENG modules under consideration is presented Details ofthe overall methodological LCA principles and the techno-economicframework for comparative cost-benefit analysis with existing

energy harvesting technologies are presented in Section 3 InSection 4 the key findings from the LCA and techno-economicanalysis are discussed leading to the summary and final con-clusions in Section 5

2 Fabrication route of a micro-gratingtriboelectric nanogenerator (MG-TENG)

To manufacture the TENG modules roll-to-roll (R2R) processing isused R2R processing is a cheap and fast substrate-based manufac-turing process4243 which can build structures in a continuousmanner and has become an important manufacturing technologyfor a wide range of new environmentally friendly and energy-efficientproducts Roller-based R2R lines consist of a series of sequentialprocessing steps which begin by feeding input materials andculminate in winding of the finished material It is often chosenbecause it can make a sheet or roll at high volume and relatively lowcost a desired attribute for the concepts discussed in this paper Inaddition it is used globally to fabricate high volume commercialproducts such as flexible electronics chemical separationmembranes and multilayer capacitors44

Fig 1(AndashC) shows the architectures of Modules A and Bwhich were assembled with series connections Module A17 wasdeveloped using a new type of electricity-generation method thattakes advantage of triboelectrification a universal phenomenoncreated upon contact between two materials Based on polymerthin films that have complementary linear electrode arrays theMG-TENG (Module A) effectively produces electricity that issufficient for powering regular electronics as the two contactingsurfaces slide with respect to each other The shape-adaptivedesign of Module A suggests that it may be ideal for harvestingenergy from a wide variety of mechanical motions Given its highelectric output power and other significant advantages in termsof weight volume cost scalability and adaptability Module A isa practically promising approach in harvesting mechanicalmotions for self-powered electronics

Module B was developed with a new type of planar-structuredelectricity-generation method to convert mechanical energyusing the triboelectrification effect Based on a statorndashrotatorstructure that has arrays of micron sized radial sectors Module Bproduces output power sufficient for conventional consumerelectronics It also has the potential to harvest energy from avariety of types of ambient energy from motions such as air flowwater flow and even body motion The fabrication of Module Brequires a series of finely controlled processes and production ofpatterns with lasers while DC sputtering is used to produce Cuelectrodes The high precision of the fabrication processes mayresult however in a prohibitively high manufacturing cost

The main functional differences between Modules A and Bare their mode of operation performance efficiency andpotential applications Whereas Module A operates in a slidingfree standing mode Module B operates in a rotating freestanding mode The performance efficiency of A was experi-mentally determined to be 50 with a resulting power output of500 W m2 and an area of 60 cm2 (see Table 1) For Module B

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

the calculated conversion efficiency is 24 (7895 cm2) with acorresponding power output of 190 W m2 (see Table 1) Interms of their applications Module B offers more robust andreliable applications regarding energy harvesting from water

bodies wind and body motion under ambient conditions Onthe other hand Module A boasts higher conversion efficiencycompared to Module B but offers less practical applicationscompared to TENG B1726

Fig 1 (A) Structural design of TENG Module A (B) structural design of TENG Module B and (C) fabrication steps for both TENG modules

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

3 Materials and methods

In the preceding sections the phenomenon of triboelectricity asa potential effect for energy harvesting is highlighted Againstthis backdrop a detailed environmental profile evaluation andtechno-economic analysis of TENG modules are carried outbased on the framework schematically illustrated in Fig 2

31 Life cycle analysis framework

LCA can be used as a decision-making tool for the systematictracking of a wide spectrum of environmental impacts across theentire value chain of the development of a product45 identifyingbaskets of interventions for reducing the environmental impactwithout burden shifting3846 LCA entails the gathering andevaluation of the inputs outputs and potential environmentalimpacts of a product system throughout its lifespan and involvesfour key steps namely (i) goal and scope definitions wherequestions such as what how and why regarding the LCA workare asked and where the systems boundaries and functional unitare set (ii) inventory analysis where input and output data ofeach process in the life cycle are collated adding them across theentire system (iii) evaluation of the environmental effectsdetailing LCA results through classification and characterizationfor comparative analysis (iv) the interpretation of the inventoryand impact assessment of results and the identification of issuesthat are of significant importance374748

The goal of this study is to assess the potential life cycle impactsof two TENG modules (A and B) The overall assessment includesfive main steps (i) gaining an understanding of the TENGtechnology in terms of raw material requirements and productionand fabrication processes of the modules (ii) characterization ofthe system (ie establishing systems boundaries the functionalunit modular components material composition operationalefficiencies etc) (iii) construction of the system inventory (ieinput requirements (physical units) process flow energy flowmaterial flow and reference flow) (iv) overall impact assessmentand environmental profile evaluations across multiple sustain-ability metrics and (v) performance evaluation and techno-economic analysis

In this work the functional unit is set as 1 m2 of the TENGmodule and all of the inventories generated are converted byaligning them to conform to the functional unit based on thedefined system boundaries as schematically illustrated inFig 3

The TENG module is assembled by depositing the componentsonto the substrate The manufacturing process consumes energyand produces emissions After the TENG module is utilizedand decommissioned the waste modules are landfilled in thedisposal stage Disposal mechanisms including incinerationand waste recycling are not taken into consideration within thesystem boundaries drawn due to the dearth of data regardingcombustion processes or waste recycling for TENG modulesModular use phases and transportation are also excludedfrom the system boundaries in line with assumptions madein a number of LCA studies for energy harvesting technologiessuch as photovoltaics49ndash51 Although inputndashoutput data can beaugmented with process-based data within a hybrid LCA frame-work36 to complete the system boundaries based on missingdata such an approach is not considered in the current workThe balance of system (BoS) is omitted as part of the overall

Fig 2 Schematic representation of the overall framework for life cycle assessment (LCA) and techno-economic analysis (TEA) of TENG modules

Table 1 Differences between two TENG modules

Module parameters Module A Module B

lsquolsquoTENGrsquorsquo module size 60 cm2 7895 cm2

Distance between TENG unit 1 cm 1 cmModule efficiency 50 24Power output of one piece of module W 3 15Power output of one piece of TENG W m2 500 190

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

system boundaries to ensure fair comparison with those ofother energy harvesting technologies

311 Life cycle inventory The construction of the life cycleinventory (LCI) is central to any LCA work Based on the systemboundaries described in Fig 3 we classified the LCI of eachmodule into two categories namely material inventory andenergy inventory A material inventory table consists of themass of raw materials direct emission during manufacturingand disposal materials per functional unit of the module Inthis analysis the focus is on two TENG modules (ie Module Aand Module B) used as representative solutions for TENGmodules The major differences between the two modules arelisted in Table 1

The material inventory of a 1 m2 functional unit of Module Ais shown in Table 2 The active area ratio and the moduleefficiency are 90 and 50 respectively17 The masses of thecleaning solvents PTFE and acrylic are obtained from theliterature1726 The masses of the electrode layer copper andtitanium are derived based on the thickness of the corres-ponding layers the active area ratio of the module and thematerial utilization efficiency Since the material utilizationefficiencies are not reported for TENG modules we assumethat the material utilization efficiencies for laser cutting andsputtering are 30 and 75 respectively The mass of directemission is determined as the mass of the cleaning solvents ofethanol acetone and deionized water

The energy inventory of 1 m2 of the TENG Module A is shownin Table 3 As shown all the operations are performed usingelectric equipment Therefore energy consumption is evaluated

Fig 3 System boundary considered in the LCA showing the material composition and energy flows associated with the fabrication steps capturedwithin the inventory

Table 2 Material inventory of 1 m2 of TENG Module A with 90 activearea

Raw materials Mass (kg) Usage

Substrate patterningAcrylic sheet E 118 10 SubstrateEthanol 100 105 Substrate cleaning solventDeionized water 100 105 Substrate cleaning solventAcetone 100 105 Substrate cleaning solvent

Grating patterningPTFE film E 110 101 3Layer thickness 25 mmEthanol 600 105 Grating cleaning solventDeionized water 600 105 Grating cleaning solventAcetone 600 105 Grating cleaning solvent

Electrode depositionCopper ETH U 224 102 5Layer thickness 500 nmTitanium I 443 104 5Layer thickness 20 nm

Electrode wiresLead ETH U 230 103 Wire diameter 00100 length 4 m

Direct emissionEthanol 700 105 Cleaning solventAcetone 700 105 Cleaning solventDisposal materials 132 10 To landfill

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

by multiplying equipment power by corresponding operating timeWe apply the same energy consumption as that evaluated byEspinosa et al50 The total electricity consumption for manufacturing1 m2 of the TENG module is 114 kW h We translate the electricityconsumption in manufacturing the TENG modules to the equivalentprimary energy consumption assuming that the electricityapplies to the average electricity mix in the US52 The end-of-life primary energy consumption accounts for the energy usageinvolved in landfilling the waste modules

Tables 4 and 5 summarize the material and the energyinventories of 1 m2 of Module B respectively The mass of ModuleB is evaluated from the data reported in the literature26

312 Life cycle impact assessment modelling The overallimpact assessment based on the LCI above is performed followingthe guidelines provided in the International Organizationfor Standardization (ISO) 1404053 and 1404454 This allows forthe appropriate data management of life cycle inventory andassessment of environmental impacts stemming from each of

the materials used for the fabrication of the TENG modulesover their life cycle Each entry of life cycle inventory developedfor this work is matched with an appropriate unit process inconformity with the functional unit Using life cycle inventoriesthe environmental impacts are calculated as follows3655

Bj frac14XI

ifrac141bji xi j frac14 1 2 J (1)

Ek frac14XJ

jfrac141ek j Bj k frac14 1 2 K (2)

where bji is the environmental burden j per unit activity i withburdens constituting raw materials and energy consumedwithin the system and emissions to air land and water Theseparameters are obtained from LCA software and databases suchas SimaPro and Ecoinvent56 xi is the mass or energy flowassociated with unit activity i ekj is the relative contributionof the total burden Bj to impact Ek as defined by the CML 2001method57

The overall focus of the current work is on global warmingpotential (GWP) However the need to consider multiple sustain-ability metrics when analyzing the environmental profile of aproduct or process was demonstrated by Ibn-Mohammed et al36

This will for environmental trade-off analysis ensure that green-house gas (GHG) emissions are not minimized at the expense ofother indicators including human toxicity acidification eutro-phication material use fossil fuel and ozone layer depletion

32 Techno-economic evaluation of TENG modules

321 Module cost estimation To assess the cost of fabricat-ing the modules we assumed the production capacity of bothroutes to be 100 MW per year As shown in Fig 4 the module costconsists of the capital the materials and the overhead cost Thecapital cost is calculated based on the depreciation of capitalinvestment (CI) Given that the complete process of Module A wasbased on the fabrication steps in Fig 1 the CI is taken to be $7million for a production capacity of 100 MW (see Tables S1 andS2 in the ESIdagger) Module B has an efficiency of 24 which is lowerthan that of Module A (50) as such the capital investment for

Table 3 Energy consumption for manufacturing 1 m2 of TENG Module Awith 90 active area

Power (W) Time (S) Electricity (MJ)

Substrate cuttingLaser cutter machine 150 103 850 101 128 101

Grating patterningLaser cutter machine 150 103 255 102 383 101

Electrode depositionTitanium coatingsputtering 150 103 400 102 600 101

Copper coatingsputtering 150 103 200 103 300 10

Total 114 kW h

Table 4 Material inventory of 1 m2 of TENG Module B with 78 activearea

Raw materials Mass (kg) Usage

Substrate patterningAcrylic sheet E 227 10 SubstrateEthanol 300 105 Substrate cleaning solventDeionized water 300 105 Substrate cleaning solventAcetone 300 105 Substrate cleaning solvent

Grating patterningFEP film E 705 102 Layer thickness 25 mmEthanol 300 105 Cleaning solventDeionized water 300 105 Cleaning solventAcetone 300 105 Cleaning solventAdhesive 250 104 Epoxy resin

Electrode depositionCopper 5374 103 Layer thickness 200 nm

amp layer thickness 100 nmTitanium 886 105 Layer thickness 10 nm

Electrode wiresLead 230 103 Wire diameter 00100 length 4 m

Direct emissionEthanol 600 105 Cleaning solventAcetone 600 105 Cleaning solventDisposal materials 235 10 To landfill

Table 5 Energy consumption for manufacturing 1 m2 of Module B with78 active area

Power (W) Time (s) Electricity (MJ)

Substrate cuttingLaser cutter machine 150 103 350 103 525 10

Grating patterningLaser cutter machine 150 103 200 101 300 102

Drilling 220 103 200 101 440 102

AirO2 plasma 100 102 600 101 600 103

Electrode depositionTitanium coating 150 103 500 102 750 101

Copper coating 150 103 500 103 75 10136 101

Total 378 kW h

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

Module B (CI Module B) for a 100 MW capacity per year isestimated to be $14 million per year (see Tables S1 and S3 ESIdagger)Details of how the cost estimates are carried out are presented inTables S2 and S3 of the ESIdagger

The depreciation of the facility results in a decrease of thecapital investment from year to year according to eqn (3)41

(CI)n = CI bn (3)

where n is the number of years after construction and b is thedepreciation ratio which is assumed to be 05 based oninformation from the nascent industry of TENG developersDepreciation of an investment should cease when bn o 01After four years there was no further depreciation of theinvestment because (05)4 = 0063 The capital costs of ModulesA and B are based on the ratio of capital investment to theoutput power which changed from US$ 007 per W to US$0004375 per W and from US$ 014 per W to US$ 000875 per Wrespectively during the first five years (see Tables S4 and S5 inthe ESIdagger for details) The module cost is calculated by summingthe capital amortization materials and overhead costs Thecapital amortization costs for Modules A and B are taken to beUS$ 0016 per W and US$ 0032 per W respectively based onthe annual worth of CI (16 million USD for Module A and 32million USD for Module B) they were equated to

i eth1thorn iTHORNn CI

eth1thorn iTHORNn1 (4)

where i is the annual interest and n is the 5-year equipmentlifetime An annual interest of 5 is assumed for 2020 basedon current low global interest rates The costs of materials forModules A and B are estimated to be US$ per 0617 W and US$per 256 W respectively based on the ratio of investment inmaterials to the output power with a material usage of 80The overhead costs consist of labor renting facilities andutilities The labor cost of US$ 00304 per W was estimatedbased on the flexible electronics industry average (see Table S7ESIdagger for details) Based on a similar industry of DSCs and thin-films silicon solar cell manufacturing lines the rents forModules A and B are estimated to be US$ 000792 per Wand US$ 0022 per W respectively The costs of utilities forModules A and B are estimated to be US$ 000792 per W andUS$ 0022 per W respectively After adding 1 of the capitalcosts for maintenance fees (US$ 00016 million per year andUS$ 00016 million per year for Modules A and B (Table S8ESIdagger)) the overhead costs of Modules A and B are calculatedto be US$ 004784 per W and US$ 0075 per W respectively(Table S8 ESIdagger)

The resultant module costs calculated based on ourassumptions are US$ 068084 per W and US$ 2667 per W forModules A and B respectively (Table S9 ESIdagger) These are thebaseline values used in the sensitivity analysis (Section 452)Estimations of the levelized cost of electricity are based on thetotal cost of a solar cell system including the costs of themodule balance of systems (BoS) land support structures

Fig 4 Cost parameters considered for the techno-economic analysis of TENG modules detailing the relevant materials overhead costs capital costsand levelized cost

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

wiring power conditioning and installation41 and summedaccording to eqn (5)5859

LCOE frac14 ICC 1000CRF

CF 8760thorn 0ampM (5)

where ICC is the Installed Capacity Cost ($ per W DC) = BoS cost +module cost CRF is the Capital Recovery Factor expressed as

CRF frac14 ieth1thorn iTHORNneth1thorn iTHORNn 1

(6)

where i is the discount rate and n is the useful lifetime (ielifetime of the system) CF is the alternating current capacityfactor calculated as 08 the renewable energy source (iewind energy8760 hours) This factor is reduced by 37 due tothe losses in the conversion process from direct current toalternating current OampM is the operation and maintenancecost expressed in $ per kW h

The following assumptions are made BoS is US$ 75 per m2

based on the projected long term goal of silicon based solarcells in 202041 BoS costs at an efficiency of 50 and 40 forModule A are US$ 015 per W and US$ 01875 per W respec-tively For Module B with efficiencies of 24 and 20 thecorresponding costs are US$ 0394 per W and US$ 0474 per Wrespectively By using BoS cost = 75 US$ per m2 OampM =$0001 per kW h i = 5 and n = 20 (no tax credits and noaccelerated depreciation) from these values CRF (i = 5 n = 15) =01 In order to derive the energy produced per year due to 1 W ofinstalled TENGs a CF of 37 is assumed

4 Results and discussion41 Primary energy consumption and carbon footprint

Primary energy demand and correspondingly the carbonfootprint due to the fabrication of both TENG modules is thefocus of the current LCA work with a view to identify hotspots inthe entire supply chain of these modules Based on the con-structed LCIs in Tables 2ndash5 the primary energy consumptionand the corresponding carbon footprint distributions for TENGmodules A and B are evaluated and depicted in Fig 5 and 6respectively As indicated in Fig 5 about 90 of primary energyconsumed in both modules is attributed to raw material require-ments A disaggregation of the material embodied energy high-lights the key differences between the TENG modules Forinstance in Module A acrylic (7818) and polytetrafluoroethylenePTFE (2048) are the major contributors to the material embodiedenergy Similarly the distribution of embodied material energy isdominated by acrylic fluorinated ethylene propylene (FEP film)and copper with each contributing shares of 9688 287 and025 respectively As indicated in Tables 2 and 4 the quantitiesof acrylic in the materials composition of both modules A and Bare 118 kg and 227 kg explaining their dominance in the totalmass of the modules (7818 for Module A and 9688 forModule B)

In terms of electrical energy consumption (also expressed inMJ m2 to conform to the unit of material embodied energy)electrode deposition of copper coatingsputtering consumedthe largest amount of energy (B73) due to the length of timeassociated with carrying out such a process during the fabricationof Module A Electrical energies consumed by sputtering fortitanium coating deposition and the laser cutting machineconstitute roughly 15 and 13 respectively Overall electrode

Fig 5 Distributions of the primary energy consumption for fabricating two TENG modules

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

sputtering consumes B85 of the electrical energy for the fabrica-tion of Module A Adoption of alternative deposition techniques forthe copper and titanium coating would go a long way in minimizingthe overall electrical energy consumption As for Module B theincreased number of operations involved in its fabrication results inhigher electrical energy consumption compared to Module A Aswith Module A sputtering of titanium and copper consumes B62of the electrical energy and the laser machining and associateddrilling activities consume 38 Sputtering as a means of depositingthin films of the metals in the modules guarantees high quality butcomes at the expense of high cost60 Overall Module B consumesmore electrical energy during fabrication compared to Module A

Fig 6 shows the distribution of carbon footprint from whichthe major contributors of the substrate the copper electrodesputtering and laser cutting can be established The distributionof primary energy consumption during fabrication indicatessimilar patterns to the carbon footprint because different fabri-cation operations consume only electricity and their conversionto carbon dioxide equivalent (CO2-eq) is based on appropriatecharacterization factors in the evaluation process Not only dothe distributions of primary energy consumption and the carbonfootprint exhibit similar patterns but also the impact of othercategories remains identical provided that the steps involved inthe fabrication process remain constant A resemblance can befound between the distributions of the material embeddedprimary energy consumption and the carbon footprint whichsuggests similar strategies for optimizing both modules forimproved environmental performance should be adopted

42 Environmental profile assessment of contributingcomponents of TENG Modules A and B across multiple indicators

Fig 7 and 8 show the environmental profiles of 1 m2 of TENGModule A and 1 m2 of TENG Module B respectively All 11

environmental impact metrics are normalized to 100 with theview that the sum of the impact of each of the contributingprocesses or materials is 100 As indicated in Fig 7 the acrylicis the most significant contributor for carcinogens (82)respiratory organics (85) respiratory inorganics (73) climatechange (74) acidificationeutrophication (76) fossil fuels(81) and ecotoxicity (33) Although the intensity of materialembodied energy and CO2-eq of copper lead and titanium arenumerically higher than that of acrylic given that the quantity ofacrylic in the material composition is the largest its overallimpact across the aforementioned impact categories outweighsother materials Sputtering due to electrical energy consumptionalso has a great influence on radiation (96) the ozone layer(83) and land use (83) The use of acrylic however offers anadvantage in the fabrication of the modules For instance acrylichas very good structural properties such as lightweight ease offabrication impact resistance and ability to withstand poorweather conditions Its high strength and durability are alsoimportant advantages Furthermore acrylic sheets are fabricatedusing fabrication processes in facilities that are certified byISO-14001 More importantly the scenario of their end of lifeis environmentally viable given their recyclability and reusepotential Additionally compared to other plastics that producetoxic gases that are harmful to humans and to the environmentduring combustion processes acrylic does not pose such threatsdue to its stability during exposure to ultraviolet radiation

As shown in Fig 8 for TENG Module B the presence of acrylic asin Module A also constitutes the major influence across a number ofindicators For instance acrylic is the most significant contributorfor carcinogens (831) respiratory organics (B92) respiratoryinorganics (808) climate change (81) acidificationeutrophica-tion (795) and fossil fuels (88) The reason for this is similar tothat of Module A (ie the quantity of acrylic used dominates those of

Fig 6 Distributions of the carbon footprint of TENG Modules A and B

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

other materials in the structure) Sputtering due to electrical energyconsumption also has a great influence on radiation (88) theozone layer (70) and land use (70)

Comparative life cycle impact assessment results betweenthe two TENG modules are depicted in Fig 9ndash11 Module A isused as the standard for normalization In Fig 9 Module A

Fig 7 Environmental profile of 1 m2 of TENG Module A

Fig 8 Environmental profile of 1 m2 of TENG Module B

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

performs better environmentally than Module B except in oneimpact category minerals This is attributed to the higherquantities and triple layer thickness of polytetrafluoroethylene(PTFE) used in Module A compared to the single layer thicknessof fluorinated ethylene propylene (FEP film E) used in ModuleB PTFE is generated through polymerization of tetrafluoro-ethylene using free radicals and hence has high mineralresource requirements The uniformity of its material struc-ture (ie PTFE) its excellent chemical electrical and physicalproperties its tightly controlled thickness and its inherentcapabilities to serve as a semi-permeable membrane render itapplicable for TENG and biomedical applications61 On theother hand the compatibility of FEP with various chemicalsits reliable electrical properties its mechanical toughnessand its broad thermal range make it suitable for TENGapplications62

Fig 10 displays proportions between impacts of the twotypes of TENG modules with respect to Eco-indicator 99 underhuman health resources and ecosystem quality As shownModule B results in more damage compared to Module ASingle score comparison by impact category based on Eco-indicator99 is depicted in Fig 11 where the environmental impact ofModule B also surpasses that of Module A For further comparativeresults of the environmental profile of TENG modules we referreaders to the ESIdagger

43 Comparison with existing energy harvesting technologies

431 Eco-indicator Eco-indicator 99 results across eco-system quality resources and human health for eight variantsof energy harvesting technologies notably PV modules arecompared with the TENG modules as depicted in Fig 12 Inall three damage categories both TENG modules achieve the

Fig 9 Comparison per damage category by summation of individual impacts the higher impact set equal to 100 using Eco-indicator 99 Europe EEmethodology

Fig 10 Endpoint comparison after weighing using Eco-indicator 99Europe EE methodology

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

lowest points and are one order of magnitude lower than thoseof c-Si a-Si ribbon-Si CdTe CIS OPV TiO2 and ZnO PVmodules This clearly demonstrates the overall environmentaledge of the TENG modules when compared to PV technologiesTherefore a more environmentally sustainable energy harvest-ing technology could potentially be developed based on TENGmodules although this requires switching to greener sub-strates and reducing the consumption of organic solvents aswell as the use of efficient fabrication processes

432 Energy payback period In this section the energypayback periods (EPBPs) of Modules A and B are compared withexisting PV technologies (ie silicon technologies thin-film

technologies organic solar cells and perovskite solar cells) TheEPBP is given by

EPBP frac14Embodied energy kWhm2

Energy output kWhm2 year1eth THORN (7)

The result of the comparison is shown in Fig 13 As shownModule A has a shorter nominal EPBP than the other technologiesat 005 years Module B also has a shorter EPBP compared totraditional PV technologies but higher than those of organic andperovskite solar cells The reason for TENGs outperformingsilicon and CdTe based PV cells is because their fabricationdoes not have high energy intensity requirements associatedwith silicon or rare element purification and processing thatcauses a higher environmental impact64 This is largely due tothe efficient fabrication routes based on R2R processing It isimportant to note that the EPBP of Module B is higher thanthose of OPV and perovskite solar cells attributed to its lower

Fig 11 Single score comparison by impact category using Eco-indicator 99 Europe EE methodology

Fig 12 Eco-indicator 99 results for 1 m2 of each module The data forc-Si a-Si ribbon-Si CdTe and CIS are extracted from the study ofLaleman et al63 The data for OPV are extracted from the study of Espinosaet al50 The data for the TiO2 perovskite module and the ZnO perovskitemodule were based on the work of Gong et al64

Fig 13 Comparison of energy payback time for 7 PV modules with TENGmodules The data for the energy payback period of all the PV moduleswere based on the work of Gong et al64

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

output energy compared to the aforementioned technologiesNevertheless in the future by leveraging optimal and efficientprocessing technologies the EPBP of TENGs can be furtherreduced significantly Overall the favorable environmental pro-file and EPBP of the TENG modules compared to the traditionalenergy harvesting technologies suggest that in the future theywill challenge the existing technologies whilst contributingimmensely towards addressing global energy problems

433 CO2 emission factor The CO2 emissions factor (CEF)is given by

CEF frac14 Carbon footprint ethkgCO2 eqTHORNEnergy output across the lifespan kWheth THORN (8)

To apply eqn (8) the lifespan of the TENG system underconsideration should be established The lifespan of other existingPV technologies is already well-established Likewise assumptionshave been made about the lifespan of perovskite solar cells Giventhat TENGs are still in their infancy no exact value in terms oflifespan has yet been reported for them Fig 14 shows the compar-ison of CO2 emission factors for existing energy harvesting techno-logies to TENG modules As indicated the CO2 emission factor forTENG Module B is higher similar to that of CdTe Ribbon-Si andP-Si (TENG Module A shows a significantly lower CEF) This suggeststhat currently the associated cost of CO2 is currently high due to theirshorter lifespan (assumed to be 2 years) In the future it is expectedthat the lifespan of TENGs will increase considerably due toadvancement in material optimization thereby lowering their CEFThese results deliver an important message for the development ofother energy harvesting devices such as TENGs as potential envir-onmentally viable energy harvesters The TENG is the youngestamong the energy harvester technologies with enormous potentialfor better manufacturing processes with improved efficiency morestable performance and a longer operational lifetime

44 Sensitivity analysis

The probability distributions of the two forecasts for the TENGmodules are shown in Fig 15 and 16 Both distributions

demonstrate a wide range with the highest bars representingthe values of the highest probabilities The asymmetric profileof both distributions results from the nonlinear relationshipbetween the input parameters and the sustainability indicatorsThe simulation results are shown in Fig 15 and 16 The singlecores in both cases are comparatively robust when the key speci-fications of the modules are subject to uncertainty The low singlecore points for the entire 95 confidence regions demonstrate thatTENGs are already environmentally competitive

A sensitivity analysis is also conducted to estimate how theenvironmental performance of Modules A and B alter if theconsumption of materials and energy during manufacturing isvaried given the dominating influence of some input para-meters across all the considered impact categories For eachparameter two scenarios were modeled and then compared withthe baseline ie a 10 variations in the total consumption

Fig 14 CO2 emission factor for selected PV modules and 2 TENGmodules A and B The data for the CO2 emission factor of all the PVmodules were based on the work of Gong et al64

Fig 16 Probability distributions for the single core impact category ofModule B

Fig 15 Probability distributions for the single core impact category ofModule A

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

As shown in Tables 6 and 7 the variation of the acrylicconsumption has the highest influence on the Resp organicsFor instance a 10 variation of cement consumption leads to84 and 91 changes in the Resp organics impact inModules A and B respectively As expected the mineral impactis most sensitive to the variation of electrode deposition con-sumption and a 10 decrease of electrode deposition leads to97 and 48 corresponding decreases of this indicator forModules A and B The fluctuation of manufacturing duringconstruction and operation leads to the largest value change ofradiation and about 97 and 10 variation occurs forModules A and B respectively if the former changes by 10

45 Techno-economic analysis

451 Estimation of costs of TENG modules Fig 17 showsthe cost of Modules A and B in the 1st and 5th year andamortization capital cost over 5 years The module cost canbe divided into the materials the overhead and the capitalcost The capital costs for Modules A and B are calculated basedon the capital costs of TENGs fabricated using the data inTables S2 and S3 (ESIdagger) respectively The cost of the materialsis estimated based on the amount used The overhead cost isestimated based on reasonable assumptions (see Table S8ESIdagger) The details of the calculations are shown in the Methodssection and the ESIdagger The relatively high module cost in the firstyear is due to the high depreciation rate (50) of the capitalinvestment The calculated capital costs in the first year are 007and 014 US$ per W for Modules A and B respectively The initialcapital cost of Module A is lower because the capital investmentassociated with the use of large efficiency is higher than that inModule B However the capital cost rapidly decreased because ofdepreciation and there is a monotonic decrease of the totalmodule cost during the first 5 years (Tables S4 and S5 ESIdagger)

After that time the contribution of the capital cost to the totalcost is lowered so that the module cost is determined mainly bythe overhead and the materials costs Fig 18 presents thedistribution of the materials cost for TENG production routesDSM layers represent device structural materials DEM repre-sents electrode dielectric materials and LW represents electrodewire Other materials costs in Fig 17 include the expense ofTiCU deposition The total calculated cost of materials forModule A is 0617 US$ per W which is lower than the cost forModule B that is 256 US$ per W (Table S6 ESIdagger) The higher costof materials for Module B is because both the output power andefficiency are higher in Module A than in Module B

Based on thin film silicon solar cell production6566 ModuleA and Module B have total overhead costs of US$ 004784 per W

Table 6 Sensitivity analysis results for Module A

Input Variation () CarcinogensRespOrganics

RespInorganics

Climatechange Radiation

Ozonelayer Ecotoxicity

Acidificationeutrophication

Landuse Minerals

Fossilfuels

Acrylic 10 +82 +84 +74 +75 00 00 +33 +75 00 00 +81+10 82 84 74 75 00 00 33 75 00 00 81

PTFE 10 00 +07 +05 +05 00 00 +09 +05 00 00 +05+10 00 07 05 05 00 00 09 05 00 00 05

Electrodedeposition

10 +03 +01 +03 +01 +03 +09 +16 +02 +04 +97 +01+10 03 01 03 01 03 09 16 02 04 97 01

Manufacturing 10 +14 +08 +18 +19 +97 +91 +28 +18 +96 +02 +13+10 14 08 18 19 97 91 28 18 96 02 13

Table 7 Sensitivity analysis results for Module B

Input Variation () CarcinogensRespOrganics

RespInorganics

Climatechange Radiation

Ozonelayer Ecotoxicity

Acidificationeutrophication

Landuse Minerals

Fossilfuels

Acrylic 10 +85 +91 +80 +83 00 00 +36 +81 00 00 +89+10 85 91 80 83 00 00 36 81 00 00 89

FEP 10 00 +03 +02 +02 00 00 +03 +02 00 00 +02+10 00 03 02 02 00 00 03 02 00 00 02

Electrodedeposition

10 00 00 00 00 00 +01 02 00 00 +48 00+10 00 00 00 00 00 01 02 00 00 48 00

Manufacturing 10 +14 +07 +17 +15 +100 +99 +51 +17 +10 +49 +09+10 14 07 17 15 100 99 51 17 10 49 09

Fig 17 Calculated module costs of TENGs for the first year the fifth yearand amortizing over 5 years by taking depreciation and amortizing capitalcost into consideration The depreciation rate was 50 per year and thecapital cost was assumed to remain constant after the five-year period

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

and US$ 0075 per W respectively (see Table S8 ESIdagger) There-fore the total production costs of Modules A and B are similarIn order to compare the costs between different energy harvest-ing technologies and to calculate the costs for electricity gen-eration and amortization module costs are also calculated byamortizing total capital cost by the lifetime of the devices Theresults show that Module Arsquos amortization cost is US$ 068084whereas Module Brsquos amortization cost is US$ 2667 as shown inFig 18 These results are used in the sensitivity analysis and theestimation of the leveled cost of electricity to obtain an estimateof the cost of electricity generation

452 Sensitivity analysis of module cost It is noteworthythat these cost estimates are based on assumptions about thetwo kinds of TENG structures However the assumed para-meters might vary when TENGs are commercialized Hence weperformed further sensitivity analyses to consider the effect ofTENGs on module costs The module costs increase exponen-tially as their module efficiency decreases (Fig 19) The solid

line corresponds to the efficiency of the present research statusThe efficiency of Module A is assumed to be 20ndash50 based ona current device efficiency of 40ndash50 The correspondingestimated module cost is 08308ndash086834 US$ per W And theefficiency is assumed to be 15ndash24 based on a current deviceefficiency of 20ndash24 for Module B The calculated module costis 4731ndash4811 US$ per W If we further extend the solid line themodule costs of Module B decrease dramatically while ModuleA decreases only slightly This result reveals that the moduleefficiency acts as an important factor for module cost no matterwhich route is used for the manufacturing process Improve-ment of the TENG efficiency and active area by upgrading theprecision of deposition methods will further increase themodule efficiency and therefore it will be an effective way toreduce the cost of Module B

453 Levelized cost of electricity produced with TENGsThe LCOE is typically used to compare system costs of electricityproduced using different sources of energy The LCOEs oftraditional energy sources are 704ndash1190 US cents per kW hand the costs of solar PV technologies are 978ndash1933 US centsper kW h as reported in Levelized Cost and Levelized AvoidedCost of New Generation Resources in the Annual Energy Outlook201567 The LCOE is calculated according to eqn (5) (Section 3)and the output is affected mainly by the module cost efficiencyand lifetime In our module cost analysis both Modules A and Bare estimated to produce TENG energy harvesting modules ata cost in the range of 068084ndash2667 US$ per W We calculatethe LCOE of a TENG energy harvesting module by assuming amodule cost of 068084 US$ per W for Module A and 2667US$ per W for Module B and a lifetime of 15 years The LCOEsare 2569 US cents per kW h and 2681 US cents per kW hcorresponding to module efficiencies of 50 and 20respectively for Module A On the other hand the costs are9198 US cents per kW h and 943 US cents per kW hcorresponding to module efficiencies of 24 and 20respectively for Module B which are lower than those ofthe traditional energy sources (Fig 19) for Module A and inthe same range of wind power for Module B Details of thecalculations are shown in the Methods section and Table S10(ESIdagger) This analysis indicates that the module efficiency has asignificant influence on the LCOE

Fig 18 Cost of material distribution for Module A (left) and Module B (right) The values of material costs are assumed by the real amount of materialused in both structures and the wholesale price The 80 material usage ratio has been considered

Fig 19 Module cost of TENGs as a function of module efficiency Exceptfor the independent variables in these figures the other parametersassociated with Module A and Module B were fixed The solid lines werecalculated based on the range of reported efficiencies the dashed linesare based on calculations assuming high module efficiencies that areexpected but not yet achieved

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

Fig 20 shows the effect of lifetime on the LCOE of TENGs forwave energy harvesting The LCOEs estimate 50 and 40efficiency for Module A and 24 and 20 for Module B buteach decreases exponentially with the extension of the systemlifetime in the range of 10ndash30 years For high efficiency (50)modules a lifetime of 10 years can lead to an LCOE of 342 UScents per kW h The low-efficiency (40) modules require ashort lifetime (12 years) to achieve a similar LCOE A conserva-tive estimate of the discount rate of 5 is used above Based onthe above analysis the module efficiency and lifetime are themost sensitive factors for the LCOE of TENGs The ultra-lowLCOE of TENGs is achieved to be 2569ndash268 US cents per kW hwith 15 years of lifetime surpassing the United States lsquolsquoSun ShotInitiativersquorsquo target of 60 US cents per kW h Hence improvements

of the efficiency and the lifetime of TENGs are urgent tasks fromthe perspective of cost and more efforts should be devoted tothis field

The LCOE is calculated according to eqn (5) (Method part)and it is affected mostly by the module cost efficiency andlifetime In our cost analysis the module cost is estimated to bein the range of 068ndash2667 US$ per W corresponding to TENGModule A and TENG Module B respectively We calculate theLCOE of TENG Module A by assuming a module cost of068 US$ per W and a lifetime of 15 years While TENG ModuleB is calculated by assuming a module cost of 2667 US$ per Wand a lifetime of 15 years The TENG Module A LCOEs are2569 US cents per kW h and 2681 US cents per kW h forefficiencies of 50 and 40 respectively On the other handthe TENG Module B LCOEs are 9198 US cents per kW h and943 US cents per kW h corresponding to module efficiencies of24 and 20 respectively which are lower than other energysources (Fig 21) Details of the calculation are shown in theMethods section and Table S10 (ESIdagger) Consequently moduleefficiency has a significant influence on the LCOE

5 Summary and concluding remarks

Mechanical energy is available in abundant quantities every-where around us and is completely independent of weatherdaynight or even season This abundant source of energyremains largely untapped but with continuous and improvedpower conversion efficiencies reported in the past few yearstriboelectric nanogenerators (TENGs) have been touted ashighly promising sources of electricity generation frommechanical energy In this paper a cradle-to-grave life cycleassessment of two TENG modules is performed The life cycleenvironmental impact assessment involves 11 midpoint impactcategories and an endpoint evaluation by following the Eco-indicator 99 methodology We shed light on two importantsustainability indicators and find that TENG modules have theshortest EPBT among existing PV technologies In addition wefind that the environmental hotspots come from the use ofacrylic (both Modules A and B) PTFE (Module A) and FEP(Module B) As such for future development of this technologymaterial optimization should be advanced Moreover we eval-uated the sustainable indicators considering the uncertaintiesof the major input parameters The resulting probability dis-tributions demonstrate that for TENGs at the current stageEPBTs are stable and competitive while CO2 emission factorsare less stable Lastly through sensitivity analysis we find thatTENG modules are potentially one of the most environmentallysustainable energy harvesters if future development confirms alarger performance ratio and a longer lifetime To this end acomparative techno-economic analysis of the TENG moduleshas been performed based on an annual capacity of 100 MWWe find that the cost of Module A is much lower than othertechnologies when fully operational while the cost of Module Bis found to be comparable to the cost of hydropower technol-ogies The results of the sensitivity analysis show that improved

Fig 20 The relationship between the LCOE and the lifetime A systemlifetime o10 years was not considered in our analysis

Fig 21 The comparison of LCOEs based on coal natural gas nuclearwind commercialized solar PV hydropower PSC and TENG modules TheLCOE values are referenced to the Levelized Cost and Levelized AvoidedCost of New Generation Resources in the Annual Energy Outlook 2015reported by the United States Energy Information Administration

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

Energy Environ Sci This journal iscopyThe Royal Society of Chemistry 2017

performance efficiency reduces significantly the module costThe fabrication of high-efficiency modules through the adop-tion of high precision fabrication processes is the most promis-ing approach for further cost reduction The results indicate anestimated levelized cost of Module A and Module B to be US2681 cents per kW h and US 943 cents per kW h respectivelyThe LCOE of TENGs is also very sensitive to the moduleefficiency and is expected to be lower than that of other energytechnologies if the module efficiency and lifetimes exceed 25and 15 years respectively To achieve these targets more effortsshould be made to improve the lifetime and the efficiency ofTENGs rather than to identify cheaper materials and fabrica-tion processes

Competing financial interest

The authors declare no competing financial interest

Acknowledgements

This research was supported by KAUST and the HightowerChair foundation The support provided for completing thisresearch is gratefully acknowledged In addition a helpfuldiscussion with Professor Heather MacLean is gratefullyacknowledged

References

1 EIA US Energy Information Administration Washington DC2013

2 IPCC Climate change 2014 mitigation of climate changeCambridge University Press 2015

3 Z L Wang J Chen and L Lin Energy Environ Sci 2015 82250ndash2282

4 G-T Hwang Y Kim J-H Lee S Oh C K Jeong D Y ParkJ Ryu H Kwon S-G Lee and B Joung Energy Environ Sci2015 8 2677ndash2684

5 D Evans How the Next Evolution of the Internet is ChangingEverything Whitepaper Cisco Internet Business SolutionsGroup (IBSG) 2011 vol 1 pp 1ndash12

6 Y Gao and Z L Wang Nano Lett 2007 7 2499ndash25057 Y Qin X Wang and Z L Wang Nature 2008 451 809ndash8138 X Wang J Song J Liu and Z L Wang Science 2007 316

102ndash1059 S Xu B J Hansen and Z L Wang Nat Commun 2010 1 93

10 S Xu Y Qin C Xu Y Wei R Yang and Z L Wang NatNanotechnol 2010 5 366ndash373

11 R Yang Y Qin L Dai and Z L Wang Nat Nanotechnol2009 4 34ndash39

12 G T Hwang M Byun C K Jeong and K J Lee AdvHealthcare Mater 2015 4 646ndash658

13 Z L Wang and J Song Science 2006 312 242ndash24614 Z L Wang G Zhu Y Yang S Wang and C Pan Mater

Today 2012 15 532ndash54315 Z LinaWang Faraday Discuss 2014 176 447ndash458

16 F-R Fan Z-Q Tian and Z L Wang Nano Energy 2012 1328ndash334

17 G Zhu Y S Zhou P Bai X S Meng Q Jing J Chen andZ L Wang Adv Mater 2014 26 3788ndash3796

18 W Tang T Jiang F R Fan A F Yu C Zhang X Cao andZ L Wang Adv Funct Mater 2015 25 3718ndash3725

19 G Zhu B Peng J Chen Q Jing and Z L Wang NanoEnergy 2015 14 126ndash138

20 Z H Lin G Zhu Y S Zhou Y Yang P Bai J Chen andZ L Wang Angew Chem Int Ed 2013 52 5065ndash5069

21 Z Wen J Chen M-H Yeh H Guo Z Li X Fan T ZhangL Zhu and Z L Wang Nano Energy 2015 16 38ndash46

22 H Zhang Y Yang T-C Hou Y Su C Hu and Z L WangNano Energy 2013 2 1019ndash1024

23 Y Yang H Zhang Y Liu Z-H Lin S Lee Z LinC P Wong and Z L Wang ACS Nano 2013 7 2808ndash2813

24 Z Li J Chen J Yang Y Su X Fan Y Wu C Yu andZ L Wang Energy Environ Sci 2015 8 887ndash896

25 S Chen N Wang L Ma T Li M Willander Y Jie X Caoand Z L Wang Adv Energy Mater 2016 1501778ndash1501787

26 G Zhu J Chen T Zhang Q Jing and Z L Wang NatCommun 2014 5 3426ndash3435

27 K Y Lee J Chun J H Lee K N Kim N R Kang J Y KimM H Kim K S Shin M K Gupta and J M Baik AdvMater 2014 26 5037ndash5042

28 L Zhang B Zhang J Chen L Jin W Deng J TangH Zhang H Pan M Zhu and W Yang Adv Mater 201628 1650ndash1656

29 Z-H Lin G Cheng X Li P-K Yang X Wen and Z LWang Nano Energy 2015 15 256ndash265

30 Z L Wang L Lin J Chen S Niu and Y Zi Green energy andtechnology 2016

31 F-R Fan L Lin G Zhu W Wu R Zhang and Z L WangNano Lett 2012 12 3109ndash3114

32 G Zhu C Pan W Guo C-Y Chen Y Zhou R Yu andZ L Wang Nano Lett 2012 12 4960ndash4965

33 G Zhu Z-H Lin Q Jing P Bai C Pan Y Yang Y Zhouand Z L Wang Nano Lett 2013 13 847ndash853

34 G Zhu J Chen Y Liu P Bai Y S Zhou Q Jing C Pan andZ L Wang Nano Lett 2013 13 2282ndash2289

35 S Chen C Gao W Tang H Zhu Y Han Q Jiang T LiX Cao and Z Wang Nano Energy 2015 14 217ndash225

36 T Ibn-Mohammed S Koh I Reaney A Acquaye D WangS Taylor and A Genovese Energy Environ Sci 2016 93495ndash3520

37 T Ibn-Mohammed R Greenough S Taylor L Ozawa-Meida and A Acquaye Energy Build 2013 66 232ndash245

38 S Hellweg and L M I Canals Science 2014 344 1109ndash111339 S Koh T Ibn-Mohammed A Acquaye K Feng I Reaney

K Hubacek H Fujii and K Khatab Sci Rep 2016 6 3951440 A Acquaye K Feng E Oppon S Salhi T Ibn-Mohammed

A Genovese and K Hubacek J Environ Manage 2016 187571ndash585

41 M Le Solar Energy Technologies Office U S Department ofEnergy 2015 pp 1ndash121

42 F C Krebs Sol Energy Mater Sol Cells 2009 93 394ndash412

Analysis Energy amp Environmental Science

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online

This journal iscopyThe Royal Society of Chemistry 2017 Energy Environ Sci

43 R R Soslashndergaard M Hosel N Espinosa M Joslashrgensen andF C Krebs Energy Sci Eng 2013 1 81ndash88

44 K Hwang Y S Jung Y J Heo F H Scholes S E WatkinsJ Subbiah D J Jones D Y Kim and D Vak Adv Mater2015 27 1241ndash1247

45 G Rebitzer T Ekvall R Frischknecht D Hunkeler G NorrisT Rydberg W-P Schmidt S Suh B P Weidema andD W Pennington Environ Int 2004 30 701ndash720

46 T Ibn-Mohammed R Greenough S Taylor L Ozawa-Meida and A Acquaye Build Environ 2014 72 82ndash101

47 A A Acquaye T Wiedmann K Feng R H Crawford J BarrettJ Kuylenstierna A P Duffy S L Koh and S McQueen-MasonEnviron Sci Technol 2011 45 2471ndash2478

48 D Collado-Ruiz and H Ostad-Ahmad-Ghorabi J CleanerProd 2010 18 355ndash364

49 R Garcıa-Valverde J A Cherni and A Urbina Prog Photo-voltaics 2010 18 535ndash558

50 N Espinosa M Hosel D Angmo and F C Krebs EnergyEnviron Sci 2012 5 5117ndash5132

51 D Yue P Khatav F You and S B Darling Energy EnvironSci 2012 5 9163ndash9172

52 J J Conti P D Holtberg J A Beamon A Schaal J Ayoub andJ T Turnure United States of America Department of EnergyInformation Office of Integrated and International EnergyAnalysis Available at httpwwweiagovneicspeechesnewell_12162010pdf 2011

53 I ISO London British Standards Institution 200654 I ISO International Organization for Standardization Geneva

Switzerland 2013

55 A Azapagic C Pettit and P Sinclair Clean Technol EnvironPolicy 2007 9 199ndash214

56 Ecoinvent Ecoinvent database httpwwwecoinventorgAccessed 20th November 2016 2016

57 J B Guinee Int J Life Cycle Assess 2002 7 311ndash31358 W Kellogg M Nehrir G Venkataramanan and V Gerez

IEEE Trans Energy Convers 1998 13 70ndash7559 K Branker M Pathak and J M Pearce Renewable Sustain-

able Energy Rev 2011 15 4470ndash448260 D L Smith and D W Hoffman Phys Today 1996 49 6061 Polyflon Technology Limited 100 PTFE Film (Polytetrafluoro-

ethylene) httpwwwpolyfloncoukproductsfluoropolymer-filmptfe-film Accessed 16th December 2016

62 Teflon Fluoroplastic Film-Properties Bulletin httpswwwchemourscomTeflon_Industrialen_USassetsdownloadsteflon-fep-film-propertiespdf Accessed 16th December2016

63 R Laleman J Albrecht and J Dewulf Renewable SustainableEnergy Rev 2011 15 267ndash281

64 J Gong S B Darling and F You Energy Environ Sci 20158 1953ndash1968

65 J Poortmans and V Arkhipov Thin film solar cells fabricationcharacterization and applications John Wiley amp Sons 2006

66 C Becker D Amkreutz T Sontheimer V Preidel D LockauJ Haschke L Jogschies C Klimm J J Merkel and P PlocicaSol Energy Mater Sol Cells 2013 119 112ndash123

67 Levelized Cost and Levelized Avoided Cost of New Genera-tion Resources in the Annual Energy Outlook US EnergyInformation Administration 2015

Energy amp Environmental Science Analysis

Publ

ishe

d on

22

Febr

uary

201

7 D

ownl

oade

d by

Geo

rgia

Ins

titut

e of

Tec

hnol

ogy

on 0

303

201

7 21

27

02

View Article Online