Techno-economic analysis and optimization of ... › smash › get › diva2:1247170 ›...

Techno-economic analysis

and optimization of

electrochemical energy

storage solutions

GUIDO LORENZI

KTH Royal Institute of Technology

Industrial Engineering and Management

Department of Energy Technology

Heat and Power Division

Brinellvägen 68, 100 44, Stockholm,

Sweden

IST Instituto Superior Técnico

Department of Mechanical Engineering

IN+ Center for Innovation, Technology

and Policy Research

Avenida Rovisco Pais 1, 1049-001,

Lisbon, Portugal

Printed by Universitetsservice US-AB Drottning Kristinas väg 53B SE-114 28 Stockholm

Sweden.

TRITA-ITM-AVL 2018:43

ISBN: 978-91-7729-940-0

© Guido Lorenzi, 2018

To my grandparents Maria, Giuseppe and Mara

“On ne reçoit pas la sagesse, il faut la découvrir soi-même après un

trajet que personne ne peut faire pour nous, ne peut nous épargner, car

elle est un point de vue sur les choses.”

“We do not receive wisdom, we must discover it for ourselves, after a

journey through the wilderness which no one else can make for us, which

no one else can spare us, for our wisdom is the point of view from which

we come at last to regard the world.”

“La saggezza non la si riceve, bisogna scoprirla da soli, dopo un

tragitto che nessuno può fare per noi, né può risparmiarci, perché è un

punto di vista sulle cose.”

Marcel Proust

i

Abstract

The need to integrate the rapidly growing share of variable renewable

energy sources in the power sector requires solutions that are capable of

mitigating the intermittent nature of these sources. They are expected to

constitute the backbone of the electricity generation system in the coming

years in order to reach ambitious goals in terms of energy security and

reduction of environmental impact. Energy storage appears a promising

solution to improve the capability of variable renewable plants to meet the

energy demand at all times, mainly through the re-allocation of the

generation surplus.

Among the several options that can favor the integration of variable

renewables, electrochemical storage technologies - batteries and

electrolysis cells - constitute the focus of this dissertation. These

approaches can be used to defer substantial quantities of energy for

medium to long time intervals, are largely location independent, and are

considered a strategic part of the decarbonization pathway in the European

Union, which represents the geo-political framework under investigation.

Battery storage systems are analyzed in both small- and large-scale

settings to quantify the energetic and economic benefits deriving from a

more efficient use of renewable energy. A small-scale battery system

connected to a residential PV plant is analyzed and the results are

compared to a demand response strategy for load shifting. The integration

of a large-scale battery facility in the energy system of an island is also

simulated and the results show a much lower level of renewable energy

curtailment. In both the situations the projected costs of the battery

technologies are used to assess their techno-economic performance.

Solid oxide electrolysis cells (SOECs) as employed in power-to-gas

upgrading of biogas constitute the second electrochemical energy storage

pathway that was studied. The upgrading process sought to increase the

methane content in the biogas by directly converting the embedded carbon

dioxide through high-temperature electrolysis and methanation. This

process showed energy conversion efficiencies higher than 80%. However,

its economic viability depends on the cost of electricity, the cost of the core

components, and the price of natural gas.

ii

Keywords:

Electrical Energy Storage, Battery Energy Storage, Lithium-ion Batteries,

Vanadium Flow Batteries, Power-to-Gas, Electrolysis, Biogas Upgrading

iii

Sammanfattning

Behovet för integrering av förnybar energis ökande andel inom

elkraftssektorn innebär lösningar som kan hantera dessa energikällors

oregelbundenhet. Förnybara energikällor förväntas att bidra till elkrafts

ryggrad inom de kommande åren för att kunna nå de ambitiösa målen

gällande energisäkerhet och förbättrad miljöprestanda. Energilagring är

en lovande teknik för att underlätta förnybara energikällors möjligheter att

tillgodose energibehoven dygnet runt, primärt genom omallokeringen av

överskottsel.

Bland de olika alternativen som kan öka integreringsgraden av

förnybara energikällor är elektrokemiska lagringstekniker – batterier och

elektrolysörer – i fokus i denna avhandling. Dessa enheter kan användas

för att flytta en del av energin inom mellan- och långtidsintervaller, är i

huvud sak oberoende av placeringen och är en strategisk del av EU:s plan

för minskning av koldioxidutsläpp. Det senare omfattar avhandlingens

geopolitiska ramverk.

Batterilagringssystem är analyserade i både små- och storskaliga faller

för att kvantifiera de energimässiga och ekonomiska fördelarna med en

mer effektiv användning av förnybar energi. Ett småskaligt batterisystem

kopplat till villabaserade solceller analyserades och resultaten jämfördes

mot en efterfrågans-baserad strategi. Integrering av en storskalig

batterianläggning för ett energisystem p en ö simulerades och resultaten

visar en lägre nivå av slösad elkraft. I både fallen används projicerade

kostnader för att utvärdera den tekno-ekonomiska presentandan.

Fastoxid elektrolysörer (SOFC) inom ramen av elkraft-till-gas

uppgradering av biogas omfattar det andra energilagringssätt som

studerades. Uppgraderingsprocessens mål var att öka metanhalten hos

biogasen med direktomvandling av koldioxid genom högtemperatur-

elektrolys och -metanisering Denna process visade verkningsgrader för

energiomvandling högre än 80%. Däremot är den ekonomiska

konkurrenskraften en funktion av elkostnaden, kostnaderna hos de olika

komponenterna och gaskostnaden.

Nyckelord

Elenergilagring, Batterilagring, Litiumjonbatterier,

Vanadinflödesbatterier, Elkraft-till-gas, Elektrolys, Biogasuppgradering

v

Resumo

A necessidade de integrar o rápido crescimento dos recursos renováveis

no setor da geração da energia elétrica exige soluções capazes de mitigar a

intermitência destes recursos. As energias renováveis têm o potencial para

se tornar a espinha dorsal da produção elétrica nos próximos anos para

alcançar os objetivos ambiciosos que dizem respeito à segurança energética

e à reducão do impacte ambiental. O armazenamento da energia parece

uma solução promissora para melhorar a capacidade das renováveis

variáveis (eólica e fotovoltaico) de satisfazerem continuamente a procura

de energia elétrica, principalmente através da disponibilização do excesso

de energia gerada.

Entre as diferentes opções que podem favorecer a integração dos

recursos renováveis intermitentes, as tecnologias para o armazenamento

eletroquímico – baterias e células de electrólise – constituem o foco desta

tese. Estas abordagens permitem adiar o uso de grandes quantidades de

energia a médio-longo prazo, são independentes da localização e são, além

do mais, consideradas uma componente estratégica do plano para a

descarbonização na União Europeia – a entidade geo-política sobre a qual

o trabalho se concentra.

Os sistemas de armazenamento com baterias são analisados para casos

de estudo de pequena e grande escala para quantificar os beneficios

energético e económico que derivam de um uso mais eficiente dos recursos

renovaveis. Para a pequena escala apresenta-se a análise de um sistema de

baterias conectado a um sistema fotovoltaico residencial e os resultados

são comparados a uma estratégia de demand response para o

deslocamento da carga. Para a larga escala, simula-se a integração de uma

instalação de baterias no sistema de energia de uma ilha e os resultados

mostram um nível muito menor de curtailments de energia renovável. Em

ambas as situações, usaram-se os custos projetados das tecnologias para

avaliar o seu desempenho técnico-económico.

As células de eletrólise de óxido sólido (Solid Oxide Electrolysis Cells –

SOECs) foram analisadas para um processo power-to-gas mediante o

upgrading de biogás. Esta tecnologia constitui o segundo tipo de

armazenamento eletroquímico de energia que foi estudado. O processo de

upgrading serve para aumentar o conteúdo de metano no biogás,

convertendo diretamente o dióxido de carbono contido nele, por meio de

vi

eletrólise a alta temperatura e metanação. A eficiência de conversão de

energia deste processo mostrou-se superior a 80%. Contudo, a viabilidade

económica deste conceito depende maioritariamente do custo da

eletricidade, do custo dos componentes principais e do preço do gás

natural.

Palavras Chave

Armazenamento de Energia Elétrica, Armazenamento de Energia em

Baterias, Baterias de Iões de Lítio, Baterias de Fluxo de Vanádio, Power-

to-Gas, Eletrólise, Upgrading de Biogás

vii

Preface

This doctoral thesis was completed at the Center for Innovation,

Technology and Policy Research (IN+), Instituto Superior Técnico

(Lisbon), and at the Heat and Power Technology (HPT) Division,

Department of Energy Technology (EGI), KTH Royal Institute of

Technology (Stockholm). The PhD was conducted in the framework of the

SELECT+ Doctoral Program (Environomical Pathways for Sustainable

Energy Services), financed by the EACEA (Education, Audiovisual and

Culture Executive Agency). The main supervisors are Prof. Carlos Augusto

Santos Silva (IST) and Andrew Read Martin (KTH). The supervision team

also includes a Co-supervisor, Prof. Massimo Santarelli, affiliated

professor at KTH. The research at IN+ focused mainly on the economic

assessment of battery energy storage solutions for small- and large-scale

applications. At HPT the evaluation of the techno-economic performance

of power-to-gas pathways was performed.

The topic of electrical energy storage is addressed both in general terms

and in connection with some practical applications, highlighting the

potential and the barriers of the analyzed solutions. The main motivation

of such work is to provide a reasonable estimation of the economic

performance that can be expected by storage systems in a medium-term

scenario. The outcomes of the investigation aim at quantifying the progress

in terms of efficiency and cost reduction that storage systems should

achieve before being considered economically viable as low-risk

investments.

The present document is a paper-compilation thesis. The thesis

presents a synthesis of the motivation, the methodology and the key-

findings published in the form of research papers in several international

journals. These papers can be found in Appendix B with the same order as

they are referred to in the dissertation.

Stockholm, October 2018

Guido Lorenzi

ix

List of appended publications

This thesis is based upon the following appended scientific articles. The

contribution of each author is summarized after the paper list.

Paper I

Guido Lorenzi, Carlos Augusto Santos Silva (2015), “Techno-

economic comparison of storage vs. demand response strategies in

distributed generation systems”, Proceeding of the 12th International

Conference on the European Energy Market (Lisbon 20th – 22nd May 2015).

Paper II

Guido Lorenzi, Carlos Augusto Santos Silva (2016), “Comparing

demand response and battery storage to optimize self-consumption in PV

systems”, Applied Energy, 2016, 180, 524-535.

Paper III

Guido Lorenzi, Ricardo da Silva Vieira, Carlos Augusto Santos Silva

and Andrew Martin (2018), “Techno-economic analysis of utility-scale

energy storage in island settings”, submitted to Journal of Energy Storage

Paper IV

Guido Lorenzi, Andrea Lanzini, and Massimo Santarelli (2015),

“Digester gas upgrading to synthetic natural gas in solid oxide

electrolysis cells”, Energy & Fuels, 2015, 29 (3), 1641-1652

Paper V

Guido Lorenzi, Andrea Lanzini, Massimo Santarelli and Andrew

Martin (2017), “Exergo-economic analysis of direct biogas upgrading

process to synthetic natural gas via integrated high-temperature

electrolysis and methanation”, Energy, 2017, 141, 1524-1537

x

Other publications not included

Paper A

Guido Lorenzi, Andrea Lanzini, Massimo Santarelli, Carlos Augusto

Santos Silva (2015), “Thermoeconomic analysis of a biogas upgrading

process in solid oxide electrolysis cells (SOEC)”, Proceeding of ECOS 2015

- The 28th International Conference on Efficiency, Cost, Optimization,

Simulation and Environmental Impact of Energy Systems (Pau 29th June

– 3rd July 2015).

Paper B

Lucio De Fusco, Guido Lorenzi, Hervé Jeanmart (2016), “Insight into

electric utility business models for high-share renewables and storage

integration”, Proceeding of the 13th International Conference on the

European Energy Market (Porto 6th – 9th June 2016).

Paper C

Guido Lorenzi, Patrícia Baptista (2018), “Promotion of renewable

energy sources in the Portuguese transport sector: A Scenario analysis”,

Journal of Cleaner Production, 2018, 186, 918-932.

Paper D

Guido Lorenzi, Marco Gorgoroni, Carlos Augusto Santos Silva, Massimo

Santarelli (2018), “Life Cycle Assessment of Biogas Upgrading Routes”,

10th International Conference on Applied Energy 2018 (Hong Kong 22nd –

25th August 2018)

Paper E

Guido Lorenzi, Leopoldo Mignini, Baldassarre Venezia, Carlos

Augusto Santos Silva, Massimo Santarelli (2018), “Integration of High‐Temperature Electrolysis in an HVO Production Process Using Waste

Vegetable Oil”, 10th International Conference on Applied Energy 2018

(Hong Kong 22nd – 25th August 2018)

xi

Contributions to the appended papers

In Paper I and II the model optimization and the choice of the

technologies to compare were developed by the author of this thesis. The

same author was also in charge of the manuscript drafting and of the

presentation of the results. Paper I was publicly presented to the 12th

International Conference on the European Energy Market, held in Lisbon

on May 20th-22nd 2015. Co-author C. Silva was involved in the definition of

the topic and of the methodology, and in the data collection. He also proof-

read the papers and participated in the review process.

In Paper III the development of the unit commitment model, its

implementation and the data collection were performed by the author of

this thesis. Similarly, the manuscript was also written by the author of this

thesis. Co-author Ricardo Vieira was involved in the definition of the

assumptions to describe the energy system and provided useful insights for

the environmental analysis of the power system. Co-author C. Silva

supervised the definition of the model for storage. Co-author Andrew

Martin contributed to highlight the relevance and the originality of the

selected case study.

In Paper IV the model development and implementation, as well as the

analyses, were performed by the author of this thesis. Similarly, all sections

of the paper were written by the author of this thesis. Co-author A. Lanzini

was involved in the study by providing relevant literature, useful

suggestions for the model definition, and by participating in the review

process. Co-author M. Santarelli gave his support along the study by

providing feedback on the structure of the paper and the comparison with

other technologies.

In Paper V the model was adapted from Paper IV and complemented

with the design of the heat exchange network that was necessary to satisfy

the thermal needs of the plant with the heat produced by the process itself.

The calculations for the exergo-economic analysis were done by the author

of this thesis. Similarly, the estimations of the benefits deriving from the

accounting of the saved CO2 and from the economic exploitation of oxygen

were performed by the same author. All sections of the paper were written

by the author of this thesis. Co-author A. Lanzini was involved in the

determination of the cost of the components, especially the sulfur polisher.

Co-author M. Santarelli supported the work with his expertise in exergo-

economic analysis. Co-author A. Martin gave useful insights concerning

xii

the structure of the paper, by improving its clarity and readability. All the

co-authors participated in the discussion during the paper review to

formulate responses to the reviewers.

xiii

Acknowledgements

I would like to thank my supervisors for their inspiring support during

the course of my PhD. In particular, prof. Carlos Silva, who was with me

since the first day of this adventure, has always encouraged me to pursue

my research ideas and has taught me that people are more important than

deadlines. The precious guidance and the fundamental contribution to the

review of the doctoral dissertation of prof. A. Martin are gratefully

acknowledged, as well as the support that I received from prof. M.

Santarelli through his scientific supervision.

I would like to acknowledge the Erasmus Mundus Joint Doctoral

Programme SELECT+ and the VULCANO Project that have financed the

research that I conducted. I am really grateful to Chamindie that has

always done her best to help me overcome the numerous difficulties of a

double degree program. I would also like to acknowledge the Innoenergy

PhD School for their financial support and for all the opportunities of

professional growth that they offered me.

I want to express my sincere gratitude to Ana Barbosa. Her support in

practical and administrative issues, and her positive attitude helped me get

through the numerous daily problems that were continuously popping up

during my first months in Portugal.

A heartfelt thanks goes to my whole family because during these four

years spent away from home they have never stopped providing me their

full practical and moral support. Un grazie speciale ai miei genitori che mi

han sempre sostenuto nelle difficoltà di questo percorso e si sono sempre

spesi per aiutarmi coi miei problemi, malgrado la lontanza. A mio fratello

Giorgio che mi ha spinto a dare sempre il massimo. Ai miei nonni che mi

hanno sempre pensato in questo periodo di vita all’estero e a cui, quasi

certamente, sono mancato di più. Un ricordo particolare per lo zio Hassan

che avrebbe sfruttato volentieri la possibilità di venirmi a trovare nelle mie

nuove città.

A special thought is reserved to the friends of the ACcidenti group that

helped me keep a bond with my home town, giving me a reason to go back

home for their weddings.

I would like to thank my friend Marco most profoundly. He was with

me in Lisbon when this wonderful story began and shared with me some

unforgettable experiences and an amazing piece of life. A cherish memory

goes also to Cecy and Elisa.

xiv

Thanks to Lucio for his friendship and for the important collaborations

that we have had during these years of PhD.

Thanks to Sérgio and Teresa that have been perfect flat mates in the

charming Lisbon.

Thanks to Damiano and Andrea for their friendship was one of the

reasons why I never suffered too much the Swedish winter.

Thanks to the friends of the Missione Italiana a Stoccolma and

especially to don Furio that has made me feel at home. Thanks also to

Giovanni and Silvia who opened the doors of their house to me, and

everybody knows how expensive the houses in Stockholm are.

Thanks to my colleagues at IN+ and in particular to João, Nathália,

José, Patrícia, Felix and Paulo for their cheerfulness and their fundamental

language support.

Thanks to my colleagues at EGI that made my year at KTH joyful and

pleasant. Thanks to Monica, José, Mauricio, Rafael, Monika, Justin, Jorge,

Tobias and Hanna.

Thanks to the STEPS research group at Polito and especially to Andrea,

Marta, Davide and Domenico for their insightful scientific suggestions.

Finally, an infinite thanks to Irene who has shared every day of

enthusiasm, joy and satisfaction, but also every moment of sadness,

downheartedness and difficulty that I have had during this incredible

experience. Without you this journey would have been much more solitary

and tasteless.

xv

Nomenclature

Abbreviations

AC Alternate Current

BEC Bare Erected Cost

BES Battery Energy Storage

CAES Compressed Air Energy Storage

CAPEX Capital Expenditure

CC Cycle Cost

CEPCI Chemical Engineering Plant Cost Index

CF Capacity Factor

DC Direct Current

DoE Department of Energy

DHW Domestic Hot Water

DPB Discounted Payback

DR Demand Response

DSO Distribution System Operator

ERSE Entidade Reguladora dos Serviços Energéticos (Energy Services

Regulation Entity)

FIT Feed-in Tariff

HP High Pressure

GHG Greenhouse Gas

INV Investment

LHV Lower Heating Value

LiB Lithium-ion Battery

LP Low Pressure

MS Membrane Separation

NG Natural Gas

NO STO System without any storage device or flexibility measure

NPV Net Present Value

O&M Operation and Maintenance

OPEX Operational Expenditure

P2G Power-to-Gas

PHES Pumped Hydroelectric Energy Storage

PSA Pressure Swing Adsorption

PV Photovoltaic

ROI Return on Investment

SNG Synthetic Natural Gas

SOEC Solid Oxide Fuel Cell

SP Specific Productivity

TASC Total As Spent Cost

xvi

TES Thermal Energy Storage

TPC Total Plant Cost

TSO Transmission System Operator

USD United States Dollar

VFB Vanadium Flow Battery

W2E Waste-to-Energy

WS Water Scrubbing

Latin Symbols

�̇� Exergy Flow [W] C Cost [€] d Discount Rate [-] E Energy [J] G Gibbs Free Energy [J] F Faraday Constant [C/mol] H Enthalpy [J]

�̇� Enthalpy Flow [W] L Methane Loss [-] �̇� Mass Flow Rate [kg/s] n Number of electrons involved in the reaction [-] N Plant Lifetime [y] P Price [€] S Entropy [J/K] T Temperature [K] V Voltage [V] W Electrical Power [W] Y Molar Fraction [-] Z Exergo-Economic Cost [€/s]

Greek Symbols

Δ Delta/Difference η Efficiency [-]

𝜃𝑠 Sulfur Coverage [-] Ф Thermal Power [W]

Subscripts

cap Capacity ch Charge CoE Co-Electrolysis disch Discharge el Electric elect Electrolysis ex Exergy exec Exergo-Economic F Fuel

xvii

in Inlet out Outlet OVN Overnight pl Plant RG Raw Gas sh,in Shifted-in sh,out Shifted-out sep Separation sp Specific STO Storage TN Thermoneutral

xix

Contents

Abstract......................................................................................... i

Sammanfattning ......................................................................... iii

Resumo ........................................................................................ v

Preface ....................................................................................... vii

List of appended publications ................................................... ix

Other publications not included ................................................. x

Contributions to the appended papers ..................................... xi

Acknowledgements .................................................................. xiii

Nomenclature ............................................................................ xv

Contents .................................................................................... xix

1. Introduction ............................................................................. 1

1.1. Objectives and Methodology ............................................................... 6

1.2. Research questions .............................................................................. 7

1.3. Limitations ............................................................................................. 8

1.4. Structure of the thesis .......................................................................... 9

2. Context and literature review................................................ 11

2.1. Classification of energy storage pathways ...................................... 12

Battery Energy Storage ........................................................................... 14

Electrolysis cells for chemical energy storage ......................................... 16

2.2. Battery energy storage: literature review ......................................... 21

Battery storage in residential PV systems ............................................... 21

Battery energy storage in island settings ................................................ 24

xx

2.3. Power-to-gas: literature review ......................................................... 26

Carbon sources for Power-to-gas............................................................ 27

Biogas upgrading .................................................................................... 28

2.4. Summary of the main contributions to the state of the art ............. 31

3. Optimization of electricity dispatch with battery storage

systems (Papers I, II and III) ..................................................... 33

3.1. Battery energy storage vs. demand response in residential PV

plants (Papers I and II) ................................................................................ 33

Preliminary study: Boolean logic approach (Paper I)............................... 34

In-depth study: optimization approach (Paper II) ..................................... 39

Assumptions for storage and demand response operations ................... 42

Energy analysis results ........................................................................... 45

Economic analysis results ....................................................................... 46

3.2. Battery energy storage to promote renewable integration and grid

support: the case of Terceira (Paper III) ................................................... 49

Case study description and main assumptions ....................................... 50

Energy and environmental results ........................................................... 56

Investment analysis ................................................................................. 57

Sensitivity analysis .................................................................................. 60

4. Techno-economic analysis of a biogas upgrading process

via electrolysis and methanation (Papers IV and V) ............... 63

4.1. Process description ............................................................................ 63

Plant layout ............................................................................................. 63

Sulfur effects on SOEC nickel catalyst .................................................... 66

4.2. Energy and exergo-economic analyses............................................ 67

Energy and exergy analyses ................................................................... 67

Assumptions for the economic analysis .................................................. 68

Exergo-economic analysis ...................................................................... 69

4.3. Energy analysis results (Paper IV) .................................................... 70

Comparison with other biogas upgrading routes ..................................... 71

4.4. Exergo-economic analysis results (Paper V) ................................... 75

xxi

Exergetic results ..................................................................................... 75

Economic results ..................................................................................... 77

Exergo-economic results ......................................................................... 79

Economic valorization of the by-products ................................................ 82

Sensitivity analysis .................................................................................. 84

4.5. Economic profitability with net present value ................................. 88

5. Conclusions ........................................................................... 93

5.1. Future Work ......................................................................................... 96

References ................................................................................. 99

Appendix A – Exergo-Economic Analysis ............................. 113

Appendix B - Publications ...................................................... 115

INTRODUCTION | 1

1. Introduction

The paradigm shift in electricity generation from fossil fuel plants to

renewable alternatives that has taken place during the first two decades of

the third millennium faces challenges in the way electricity is managed.

Increases in renewable energy have been very visible in the European

Union, where the supply of electrical energy from renewable origin

doubled between 2004 and 2016 – passing from 467 to 960 TWh – while

the total electricity generation remained stable at around 3250 TWh [1].

Additionally, in 2016 the annual wind production (311 TWh) almost

reached the same level provided by hydroelectric sources (350 TWh), and

solar PV supply increased more than tenfold up to 110 TWh in the period

2004-2016 [1]. The electricity supply of wind and solar is heavily

influenced by sudden changes in weather conditions (e.g. wind speed and

direction, or cloud cover for solar plants). Therefore, the variable nature of

these sources has introduced more uncertainty on the supply side of the

electricity dispatch, jeopardizing the instantaneous balance between

electricity supply and demand [2,3]. Moreover, the oversupply during

production peaks can lead to energy generation curtailments for lack of

demand, resulting in a waste of low-carbon energy. All these issues can be

handled with the introduction of energy storage systems that are able to

provide a buffer for the fluctuations of electrical energy supply.

The oldest and most widespread technology for electrical energy

storage is pumped hydroelectric energy storage (PHES), a technique that

exploits the reversible conversion of gravitational potential to electrical

energy via water flow between reservoirs at different altitudes. PHES is

mainly used for bulk energy storage, that is the storage of considerable

quantities of energy that can be gradually released when the demand is

high. So far, more than 96% of the storage facilities by installed capacity

operate with this technology [4]. However, the construction of a reservoir

is a capital-intensive and complex process that requires, besides the

selection of an appropriate site, a long approval procedure due to its

considerable impact on the surrounding environment [5]. These

constraints limit the expansion potential of pumped hydro and, for this

2 | INTRODUCTION

reason, it cannot represent the only solution for energy storage. Therefore,

research on alternative storage mechanisms that can help integrate

variable renewable sources and provide bulk storage in the energy system

has received a growing interest.

Storage pathways are typically classified according to the form into

which the energy is stored, as shown in Figure 1. The rate at which energy

can be stored or retrieved is a critical aspect of the particular storage

pathway. Figure 2 (a variant of the Ragone plot [6]) shows this relationship

for the various storage methods. Among the technologies included in

Figure 1, those that can perform similar tasks to PHES (i.e. bulk storage)

are compressed air energy storage (CAES), batteries and flow batteries,

and electrolyzers. Generally, the minimum charge or discharge time at full

power of all these technologies is one hour, which allows for the storage of

considerable quantities of energy. A closer look at the related storage

pathways reveals that electrochemical devices are at the core of all of these

technologies, excluding compressed air storage. On the other hand,

flywheels, capacitors and supercapacitors are options that address

primarily the issues and challenges related to the use of storage for short-

time-scale ancillary services (e.g. frequency regulation and voltage control)

[7]. The application of thermal energy storage (TES) solutions for

electricity management are mainly three: heat storage in form of sensible

heat (usually molten salts) to allow night operation of concentrating solar

plants [4]; production of ice (latent heat) or chilling of water (sensible heat)

to smooth the air conditioning load of residential and industrial premises

[8]; and use of low-cost electricity to produce and store heat in order to

reduce the electrical peak load for space heating and domestic hot water

supply in residential and commercial premises [9–11].

Another critical aspect of linking renewables-based electricity supply to

energy storage is its context. The European Union (EU) serves as a useful

example, since it is one of the few geo-political entities that have set clear

and ambitious goals in terms of decarbonization in the use of energy

connected to human activities. The ambition to “become the world number

one in renewable energy” has been clearly stated by the European

Commission [12]. Among the various energy storage pathways analyzed in

the Strategic Energy Technology plan [13], electrochemical energy storage

technologies are explicitly mentioned among those that are strongly

promoted for a transition towards low-carbon energy systems. The number

of projects carried out worldwide in electrochemical storage (almost

INTRODUCTION | 3

exclusively with batteries) is three times larger than those for PHES. Total

installed capacity is growing fast (from 0.1 to 1.5 GW between 2011 and

2017) [4], although the cumulative power is still one order of magnitude

lower than PHES.

In general, batteries are characterized by modularity, and they have a

reduced impact on the ecosystem compared to the construction of a

reservoir, provided that their disposal is properly managed. This can favor

their global diffusion, also in view of the fact that they can accompany the

growth of renewable installed capacity both for small-scale (at building

level to reduce locally the mismatch between supply and demand) and

large-scale facilities (at utility level to reduce the curtailments of renewable

energy and the risk of supply failure). In particular, battery technologies

are suitable to improve the profitability of residential renewable plants

under grid-parity conditions [14], favoring self-consumption and reducing

the exposure to high electricity prices [15]. This translates in huge

opportunities for market growth [16,17], especially if the energy exported

to the grid is poorly compensated. However, other options are available to

achieve similar results in terms of savings, exploiting the difference in

electricity prices under time-of-use rates. These alternatives go under the

name of demand response (DR) and consist in deferring part of the load to

hours when the electricity price is lower. Quantifying the utility and

convenience of the approach(es) from an end user perspective is

paramount towards finding optimal solutions for small-scale systems.

4 | INTRODUCTION

CLASSIFICATION OF ELECTRICAL

ENERGY STORAGE PATHWAYS

Mechanical ChemicalElectro

chemicalThermal

· Pumped hydro (PHES)

· Compressed air (CAES)

· Flywheels

· Electrolyzers (Hydrogen and Synthetic Natural Gas)

· Batteries · Flow

batteries

· Sensible heat storage

· Latent heat storage

Electrical

· Capacitors· Supercapaci

tors

Figure 1: Classification of the electrical energy storage pathways. Adapted from [7,8].

TES

PHES

Chemical Energy Storage

(Hydrogen, Synthetic Natural Gas,

Higher Hydrocarbons)

Ch

arge

/Dis

char

ge T

ime

Power

1 kW 10 kW 100 kW 1 MW 10 MW 100 MW 1 GW

1 m

s1

s1

min

1 h

ou

r1

day

FOCUS OF THIS WORK

CAES

Flywheels

(Super)

Capacitors

Battery Energy Storage

Figure 2: Map of the energy storage technologies.

In contrast, the presence of multi-MW renewable plants has introduced

additional variability in the electricity supply promoting the need for large-

INTRODUCTION | 5

scale energy storage solutions [18]. Batteries appear to have a large

potential because they do not present significant constraints on the

geographical location, unlike PHES and in-ground CAES [8]. The use of

battery technologies has been investigated in island settings, as part of so-

called hybrid systems [19,20], to mitigate the effects that fluctuations of

electricity supply entail on micro or mini electrical grids. Although some

studies already examined the economic viability of multi-MW battery

facilities in remote energy systems [21], none of them compared two

different technologies in terms of investment profitability.

In addition to the previously mentioned issues, high penetrations of

variable renewables in the power system are likely to produce large

curtailments of renewable electricity due to a lack of demand or

transmission capacity [22–25]. The need for additional flexibility to

accommodate large quantities of energy for a relatively long time is offered

by a second electrochemically-linked pathway, namely chemical energy

storage through electrolysis devices. Chemical energy storage (also known

as power-to-chemicals [26]) refers to the processes and technologies that

use electrical energy for the synthesis of chemical compounds that can be

easily stored and released at a later time. The energy density of the

produced substances is higher 1 than that of other bulk storage options

(pumped hydro, compressed air and batteries) and, for this reason, can

facilitate the storage and the transport of very high quantities of energy

[7,27]. This route is called power-to-gas (P2G)2 if the final product is in

gaseous form. The production of hydrogen through water electrolysis and

of synthetic natural gas through CO or CO2 methanation are the most

studied processes in this category of storage [28]. Then, the produced gas

could be injected into the natural gas grid, acting both as reservoir and as

distribution system for the stored energy. Hydrogen can be directly

produced in an electrolyzer and then used for electricity production in a

fuel cell, in a so-called power-to-power process, like what happens in

pumped hydro and battery energy storage. However, this process is

characterized by a low round-trip efficiency (around 30%) in comparison

with the other competing technologies (generally higher than 70%), and

this constitutes the main obstacle for its adoption [7]. On the other hand,

1 Typical values for energy density per unit volume are: PHES 1-2 Wh/l, CAES 2-6 Wh/l, lithium-ion 200-500 Wh/l, lead-acid 50-90 Wh/l, vanadium flow batteries 25-35 Wh/l, hydrogen 500-3000 Wh/l, other synthetic fuels 500-10000 Wh/l [8]. 2 Power-to-gas can be abbreviated as P2G or PtG. In this thesis the first acronym is preferred.

6 | INTRODUCTION

the synthesis of hydrocarbons requires a substantial feed of carbon

compounds, mainly carbon monoxide and carbon dioxide, which can be

obtained from the processing of another renewable resource, namely

biomass [29,30]. Moreover, if the produced fuels were used for

applications that do not include the reconversion into electricity, this could

be a way to decarbonize other energy sectors. The intersection between the

challenges of energy storage and transport and their possible synergies

could favor reduction of fossil fuel dependence. The promotion of system

integration has been defined as the “core of the low-carbon transition” in a

roundtable organized by the European Commission on these topics [31].

Addressing all of these challenges – i.e. mismatch between supply and

demand, and curtailment reduction – requires significant efforts in

technology development and in integration with existing facilities and

infrastructures. Electrochemical storage pathways would necessitate

considerable investments because their diffusion is currently quite limited.

However, they can be positively affected by other present technological

trends, as well as be integrated with other practices for the reduction of the

energy sector’s environmental impact. Particularly, the price of batteries is

likely to drop thanks to the mass production of electric vehicles [32],

although supply chain risks of materials like Cobalt and Nickel could

reverse the trend, while chemical energy storage could be coupled with

biomass conversion processes to increase their yields [33]. Therefore,

techno-economic analyses are a critical first step towards identifying

efficient combinations of technologies and process layouts.

1.1. Objectives and Methodology

The overall aim of this thesis is to provide specific knowledge on the

opportunities for the adoption of electrochemical energy storage systems

in the transition towards a low-carbon energy framework. The goal is to

assess if these technologies can take a cost-effective advantage of the excess

electrical energy supply that variable renewables are likely to cause.

Specific objectives are linked to case studies examining the following two

lines of investigation:

· Assessment of the profitability of battery storage at residential and

utility scales

· Evaluation of the techno-economic conditions that can make

electrolysis a viable alternative for power-to-gas biogas upgrading

INTRODUCTION | 7

This work intends to show that different electrochemically-linked

energy storage pathways can co-exist in high RES shares energy systems.

The rigorous presentation of opportunities and obstacles for the

conversion of electrical energy into chemical energy, for different scales

and addressing diverse challenges of a low-carbon energy supply,

constitutes the main general novelty of this investigation.

The method of investigation of this thesis focuses on the determination

of the techno-economic performance of a restricted set of storage options,

as identified through a literature study (Chapter 2). Techno-economic

evaluations are based on functional system models taken in the perspective

of mid-term energy planning, meaning that most of the assumptions are

realistic for 2030. Then, the obtained results are compared to alternative

solutions. Various methodological approaches are employed for the

specific case studies in order to fill the gaps pinpointed in the literature

review (Chapter 2):

· Optimization algorithms for modeling electricity dispatch and

evaluating operating and fixed costs for (i) single-family home

equipped with PV panels and buffering systems and (ii) electricity

supply of islands featuring large-scale battery storage.

· Physical and chemical modeling of electrolysis and methanation

processes for analyzing biogas upgrading, accompanied by exergo-

economic analyses.

1.2. Research questions

In alignment with the aim and objectives listed above, a number of

research questions were defined, grouped according to the two main

objectives. The following list contains the main research questions as

related to the appended papers:

Assessment of the profitability of battery storage at residential and utility

scales

1. What is the most cost-effective way of reducing the mismatch

between energy production and demand in a small-scale

residential PV system: battery storage or demand response

through management of the domestic hot water load? (Papers I-

II)

8 | INTRODUCTION

2. What are the benefits that large-scale battery energy storage

systems can create in island settings in terms of reduction of

curtailed energy and fuel cost savings? Do these benefits justify an

investment decision according to common investment evaluation

methods (net present value, payback time and return on

investment)? (Paper III)

Evaluation of the techno-economic conditions that can make electrolysis

a viable alternative for biogas upgrading

3. What is the energy performance of a plant that can upgrade biogas

into synthetic natural gas via high-temperature electrolysis and

methanation, and what is the configuration that produces the best

results? (Paper IV)

4. Is biogas upgrading through electrolysis and methanation

economically profitable? Which are the factors (e.g. cost of

components, cost of utilities, price of natural gas) that have the

most influence on its profitability? (Paper V)

1.3. Limitations

The virtual impossibility to address the numerous applications of all the

energy storage pathways forced the analysis to be limited to some specific

aspects of the integration of variable renewable sources by means of

storage systems. The choice of the geographical context, i.e. the European

Union, influenced the selection of relevant technologies and case studies,

which were defined based on the pertinence to the topic, the possibility to

provide original knowledge and the availability of data. A detailed

description of the state of the art and of the gaps in the literature that led

to the selection of the case studies is reported in Chapter 2. As described in

the Introduction, electrochemical energy storage technologies are

considered strategic by the European Union for the transition towards low-

carbon energy systems [13]. On the other hand, other technologies could

represent a technically viable alternative to electrochemical storage. For

instance, compressed air energy storage (CAES), both in the in-ground and

modular variants is theoretically suitable to defer large quantities of

electrical energy. However, efficiencies lower than 70% [8] and the need

for an appropriate geographical location (especially for in-ground plants)

restrain the CAES expansion potential. These reasons, together with the

INTRODUCTION | 9

limited presence of CAES facilities in Europe3, explain why this technology

is not part of this investigation.

Furthermore, the focus of this dissertation is on storage processes and

technologies having a charge and discharge time framework in the order of

hours. Therefore, this excludes from the scope of this work all the energy

storage technologies that deal with short-time issues and are designed with

high power and low energy capacity (e.g. flywheels, capacitors and

supercapacitors).

Thermal energy storage can be used for power applications mainly in

concentrating solar power plants, in order to allow such plants to produce

continuously also during nighttime [34]. However, this application is

specific for a particular power technology, which is not investigated in this

work. Alternatively, thermal storage solutions can be adopted in buildings

or industrial processes where there is a consumption of electrical energy to

satisfy thermal needs [35]. Although this does not constitute the main

focus of the thesis, this kind of storage has been studied as a form of

demand response in small-scale PV systems, as an alternative to battery

storage (see Chapter 3).

In addition to the production of synthetic natural gas via electrolysis

and methanation, liquid fuels can also be obtained by using products

resulting from electrolysis processes [36]. In particular, the conversion of

CO or CO2 into liquid hydrocarbons, through the addition of hydrogen

(Fischer-Tropsch synthesis) has been investigated as a pathway for

biomass utilization [37,38]. However, these processes would entail higher

thermodynamic losses [39,40] and for this reason they have been

neglected in the dissertation.

1.4. Structure of the thesis

This thesis intends to offer an organized and coherent framework in

which the author’s publications serve as building blocks. The collection of

papers published throughout the course of the doctoral studies is

contained in Appendix B.

Chapter 2 contains the general context of the thesis and the literature

review that highlights the previous works and the gaps that are addressed

3 Two plants are operating in Europe, one in Germany (Kraftwerk Huntorf) and one in Switzerland (ALACAES demonstration plant). Some more plants are present in the US [4].

10 | INTRODUCTION

in this dissertation. It also contains a section with the summary of the main

contributions that the appended papers brought to the state of the art.

Chapter 3 includes the studies about battery energy storage. Section 3.1

reports the techno-economic comparison between battery storage in

alternative to demand response for the integration of a small-scale

residential PV plant in Lisbon (Portugal). Section 3.2 presents the

investment analysis of a large-scale battery plant on the island of Terceira

(Azores), comparing two different technologies (i.e. lithium-ion and

vanadium flow batteries) in terms of net present value and return on

investment.

Chapter 4 describes a power-to-gas process that uses electricity to

upgrade biogas to synthetic natural gas, via electrolysis and methanation.

It contains the description of the process, the layout of the analyzed plant

and presents its energy and economic performance.

Chapter 5 collects the main conclusions of the thesis and the possible

future works to expand the knowledge on the integration of storage

technologies in energy systems.

Appendix A contains a more detailed description of the exergo-

economic analysis methodology, which was used to quantify the costs of

the biogas upgrading plant.

CONTEXT AND LITERATURE REVIEW | 11

2. Context and literature review

The growing environmental concern connected to the use of non-

renewable resources promoted the massive growth of intermittent

renewable energy sources over the last 20 years. This has profoundly

changed the electricity dispatch paradigm that was in place before. Not

only the generation is distributed and the flow is bi-directional, but the

uncertainty has affected also the supply side, in addition to the demand

side. The challenge to increase the penetration of variable renewable

sources, can be won by introducing flexibility measures [41] that can

mitigate the variability attributable to intermittent sources.

Among the European countries, Germany has experienced a

tremendous surge in RES capacity, especially wind and solar (Figure 3),

and this has caused profound changes in the way electrical energy is

managed. In 2016 the renewable content of the German electricity was 32%

of a total supply of more than 590 TWh [1]. Other countries in Europe (e.g.

Austria, Sweden, Portugal) have higher shares of renewables, but they are

characterized by a lower absolute electricity demand.

Figure 3: Evolution of RES in Germany between 2000 and 2016 [42,43].

12 | CONTEXT AND LITERATURE REVIEW

The case of Germany is relevant because it is the country with the

largest power system in Europe and, in spite of the interconnections with

ten countries and a considerable hydro-pumping capacity (6.8 GW in 2016

[44]), has been experiencing problems in integrating the massive amount

of intermittent renewable energy produced by the 56 GW installed power

of wind farms (onshore and offshore) and 43 GW peak power of solar

panels (2017 figures [45]). The most updated data show that the curtailed

renewable energy reached its maximum in 2015 with 4.7 TWh and

decreased in 2016 to 3.7 TWh [46]. The same situation is present, though

with lower figures, in other countries that have to integrate large quantities

of wind energy, such as Ireland, UK and Spain [47]. This is mainly due to

constraints in the transmission lines that connect neighboring countries or

different parts of the same country. The presence of storage facilities in the

areas where renewable plants are more present reduces the need for

investments in new transmission capacity, which are usually more difficult

to realize because they affect a larger area and can encounter the

opposition of the public opinion [48].

2.1. Classification of energy storage pathways

Given the complexity of the electricity generation system, the roles that

storage can play within this technical environment are various. They are

mainly: power quality support, bridging power and energy management4

[49]. They span between multiple time and energy scales (Figure 4) in

order to fulfil the multiplicity of services that electricity generation and

dispatch require.

Power quality support involves the issues related to the minimization of

the deviations between the instantaneous values of frequency, current and

voltage, and their target values or wave forms [50]. The characteristics of

the storage systems that operate in this range are very fast response time

(ms) and discharge time up to approximately 15 minutes [49].

Bridging power is an expression used to indicate the role of storage in

guaranteeing the reliability of the power supply or the possibility to restart

the operation in case of black-out. Some of the services included in this

4 In this context the expression “energy management” characterizes the energy storage processes that offer a continuous energy discharge for a prolonged time [49].


group are contingency reserve (spinning and non-spinning), load following

and forecast uncertainty [51].

Energy management encompasses all those actions that are aimed at

using energy for time intervals longer than 30 minutes to perform mainly

renewable energy arbitrage, peak shaving and seasonal storage.

LEGEND

- Frequency stabilization

- Voltage control

- Reactive power control

- Contingency reserve

- Load following

- Forecast uncertainty

- Peak shaving

- Renewable energy arbitrage

- Seasonal storage

Power Quality Support

Bridging Power

Energy Management

Ch

arge

/Dis

char

ge T

ime

Power

1 kW 10 kW 100 kW 1 MW 10 MW 100 MW 1 GW

1 m

s1

s1

min

1 h

ou

r1

day

Figure 4: Applications of energy storage systems in the electric grid. The bottom side of each rectangle indicates the response time, while the top border is the maximum discharge time at full power (adapted from [49,51,52]).

The most common categorization of energy storage solutions is done

with reference to the form into which electrical energy is stored. Several

papers attempt to organize the knowledge around these topics [7,8,49,53]

and they extensively describe the different technologies for energy storage.

A possible taxonomy is the following:

- Electrochemical energy storage systems

- Mechanical energy storage systems


- Electrical energy storage

- Thermal energy storage

As explained in the Introduction, only electrochemically-linked energy

storage (namely batteries and electrolyzers) is analyzed in detail, while

more information on the other types can be found in the literature

[7,8,49,53].

Battery Energy Storage

Battery energy storage includes those devices that enable the reversible

conversion between electrical and chemical energy. Batteries and flow

batteries constitute the main categories of devices that belong to this

storage option [53]. Both of them can be employed for bridging power and

energy management [49,54].

A battery system is composed by electrochemical cells that generate

electricity at a particular voltage level or store electricity into chemical form

through reversible electrochemical reactions. The main cell components

are the positive and negative electrodes, where the chemical energy is

stored, and the electrolyte, that is responsible for the flow of ions that

closes the circuit (Figure 5) [8]. If the energy is stored in two redox couples

(electrolyte) dissolved in a solution, and the electrodes have mainly the

function of electrons collectors, then the electrochemical device is called

flow battery. This technology shares the same working principles of the

other batteries, but the main difference resides in the fact that the

electrolyte is stored in a separate tank and it is continuously pumped in the

half chamber where the reaction takes place (Figure 6). This configuration

allows the power and the capacity to be decoupled since the power is

proportional to the number of cells that collect the current, while the

capacity is related to the size of the electrolyte tank.

The distinction among batteries is usually made according to the

electrode/electrolyte materials. The main technologies that are considered

for energy storage are:

- Lead-acid batteries

- Sodium-sulfur batteries

- Lithium-ion batteries


a)

ELECTROLYTE

ELECTRODE 1

(ANODE)

ELECTRODE 2

(CATHODE)

LOADe- e-

ions

b)

ELECTROLYTE

ELECTRODE 1

(CATHODE)

ELECTRODE 2

(ANODE)

POWERe- e-

ions

Figure 5: Operation of a storage system during a) charge and b) discharge.

Lead-acid batteries (cathode material PbO2, anode material Pb,

electrolyte sulfuric acid) are the most mature and widely used battery

technology. They are mainly used in conventional cars because of their low

cost and satisfactory round-trip efficiency, while their adoption in utility-

scale applications is limited by their low cycle life (< 2000

charging/discharging cycles) [8,49].

The sodium-sulfur technology (electrode materials molten sodium and

molten sulfur, electrolyte in β-Al2O3) features high energy density and low

idle discharge, although the working temperature is around 600 K, which

increases the operating costs.

Lithium-ion batteries (cathode in Li-oxide, anode in graphite,

electrolyte non-aqueous solution of Li-salts) are the most versatile

technology because of their fast response time (ms), high energy density

(200-500 Wh/L) and high cycle efficiency (up to 97%) [8].

Flow batteries are instead categorized in redox flow batteries and

hybrid flow batteries. The latter have one of the substances participating to

the electrochemical reaction at the solid state. In this thesis only redox flow

batteries are considered because they are the most technologically mature

[55,56] and they have a good potential for large-scale storage applications

[57].


V2+

V3+E

lecctr

od

e

ElectrolyteElectrolyte

V4+

V5+M

em

bra

ne

Ele

ctr

od

e

Load

e- e -

H+

Figure 6: Working principle of a Vanadium Flow Battery (discharging phase).

An extensive description of the working principle, the reactions, the

materials and the applications for vanadium flow batteries can be found in

Doetsch and Burfeind [58]. Several articles focus on the optimization of the

design and on the search for new materials [59–61], while only a limited

number on their integration in energy systems [57,62].

Electrolysis cells for chemical energy storage

Chemical energy storage embodies the processes in which electrical

energy is used to produce chemical compounds via electrolysis. The high

energy density of the produced substances is the main reason why they are

used as storage media. The most studied options for storage applications

are: hydrogen, methane and liquid fuels. Other non-hydrocarbons are

alcohols and ammonia [7].

Hydrogen is obtained through water electrolysis, while the other fuels

require the addition of carbon, which is usually available in the form of

CO2. The options to produce hydrocarbons are methanation of syngas and

Fischer-Tropsch synthesis if the resulting compound is methane [63] or a

liquid hydrocarbon [38,64], respectively.

Gaseous compounds can be used directly in fuel cells to produce

electricity or they can be injected into the natural gas grid, which serves as


storage capacity. Hydrogen can be tolerated by the gas infrastructure up to

10% on volume basis. However, the presence of a compressed natural gas

filling station or of specific gas turbines and engines restrain this value

under 2% [65]. The synthesis of synthetic natural gas solves the problem

of compatibility with the distribution infrastructure and with the end-use

appliances.

Electrolysis of water is the key process to connect power systems with

the production of gaseous fuels. The main technologies that are available

for this process are: alkaline cells, proton exchange membrane (PEM) cells,

and solid oxide electrolysis cells (SOECs). The first two options have an

operating temperature around 60-100°C and they are referred to as low-

temperature electrolysis. The third one instead works at 600-1000°C and

is considered high-temperature electrolysis. Many review papers describe

the main features of these three technologies, which are summarized in

Table 1. All of them are already commercially available, however, alkaline

cells technology is the most widespread, while solid oxide electrolysis cells

have only been used in research projects so far [66].

The working principle, the materials and the advantages of SOECs are

described in e.g. Laguna-Bercero [67]. In the study the author explains that

the required electrical power is reduced with increasing temperature, since

the Joule effect of the flowing current produces heat that is used in the

water splitting process. The minimum electrical energy contribution for

the electrolysis reaction is expressed through the Gibbs free energy (Δ𝐺),

while 𝑇Δ𝑆 is the heat demand:

Δ𝐺 = Δ𝐻 − 𝑇Δ𝑆 ( 1 )

where Δ𝐻 is the enthalpy change, ΔS is the entropy change and 𝑇 is the

temperature expressed in Kelvin. In Figure 7 the behavior of Δ𝐺 is reported

as a function of temperature and it is observable that the electric need has

a decreasing trend. Thereby, high-temperature operation can decrease the

hydrogen production cost, because a larger portion of the energy is

obtained through Joule effect. This is the reason why these devices tend to

be operated at the thermoneutral potential (𝑉𝑇𝑁 – equation ( 2 )), that is

the condition at which the heat produced by the cell electrical resistance is

equal to the thermal need of the electrolysis reaction ( Δ𝐻𝑓 is the total

energy required for the electrolysis, 𝑛 is the number of electrons involved

in the reaction, i.e. 2 for water electrolysis, and 𝐹 is the Faraday constant).


𝑉𝑇𝑁 =Δ𝐻𝑓

𝑛𝐹 ( 2 )

Under this condition the efficiency of the cell 𝜂𝑆𝑂𝐸𝐶 (equation ( 3 )) is 100%

because all the energy flow for the electrolysis (Δ�̇�) is supplied as electrical

power (𝑊𝑒𝑙) and the external heat need (Φ) is zero.

𝜂𝑆𝑂𝐸𝐶 =Δ�̇�

𝑊𝑒𝑙 + Φ ( 3 )

Figure 7: Specific energy demand for water and steam electrolysis. Adapted from [68].

In addition, high-temperature electrolysis devices, and especially

SOECs, are the only ones that can perform co-electrolysis of water and

carbon dioxide (i.e. reactions ( 4 ) and ( 5 )) where water and CO2 are both

split on the cells’ active surface. Such a possibility enables the conversion

of CO2 into hydrocarbons, e.g. methane, and is the main reason why this

technology was selected to be studied for power-to-gas applications.

𝐻2𝑂 → 𝐻2 +1

2𝑂2 ( 4 )

𝐶𝑂2 → 𝐶𝑂 +1

2𝑂2 ( 5 )


All the considerations on electrical and thermal needs, thermoneutral

voltage and SOEC efficiency are valid for co-electrolysis. The energy

required for the electrolysis of CO2 is reported in Figure 8.

Figure 8: Specific energy demand for H2O and CO2 electrolysis. Adapted from [69].

ΔH

ΔH


Table 1: Comparison of the main technologies for electrolysis [66,70].

Alkaline PEM SOEC

Electrolyte Alkaline

solution of KOH or NaOH

Polymeric membrane

(Nafion)

Yittria Stabilized

Zirconia (YSZ) Temperature [°C]

60-80 50-100 600-1000

Voltage [V] 1.8-2.4 1.7-2.0 0.9-1.4 Current density [A/cm2]

0.2-0.3 1-2 0.4-1

Catalyst material

Ni, Co, Fe Pt, Ir Ni

Charge carrier OH- H+ O2- Efficiency [%] 45-70 40-80 70-100 System life span [years]

20-30 20-30 10-20

Operation range [% of nominal capacity]

20-100 5-100 d.n.a.*

Current investment cost [€/kW]

600-2600 400-900 5000-10000

Projected investment cost in 2030 [€/kW]

400-900 300-1300 400-1000

*data not available


2.2. Battery energy storage: literature review

The size of variable renewable energy source plants ranges from few kW

to several MW. The small-scale plants are often referred to as distributed

energy sources because they are directly connected to the distribution grid.

These units do not have constant electricity generation profiles and their

widespread presence is likely to introduce problems for the energy

management to maintain the reliability of the service. Therefore, storage

systems should be adopted to mitigate the effect of the mismatch between

supply and demand. The fact that variable renewables are suitable for

small- and large-scale applications, suggests that also storage systems

should follow this classification.

Battery energy storage has acquired an increasingly important role in

electricity management mainly because of its adaptability to small- and

large-scale applications to reduce the negative effects introduced by wind

and solar for distributed and centralized applications. Moreover, the mass

production of electric vehicles has triggered enormous efforts in research

and development that caused the price of batteries to drop dramatically

after 2010. According to a study conducted by McKinsey and Company

[71], between 2010 and 2016 the average price of a battery pack passed

from 1000 to 227 $/kWh. This trend has benefitted also from the

application of battery for stationary application as it is proved by the very

high number of installations around the world [4].

Small- and large-scale battery energy storage have received a growing

attention as evinced by the number of scientific publications (6-fold

increase between 2010 and 2017)5 and industrial investments [72]. For this

reason, both of these opportunities were assessed under a techno-

economic perspective in relevant case studies as described in Chapter 3.

The next two sections present the state of the art and the research gaps that

oriented the investigation.

Battery storage in residential PV systems

The progressive phase-out or reduction of the incentives for solar

photovoltaic (PV) in different European states has changed the economic

paradigm of small distributed generation systems. The significant decrease

in the cost of PV panels and the increasing retail price of electricity has

5 Based on a search in Scopus [203] with the query “battery energy storage”.


made the concept of grid parity true in many European countries [14,73].

This means that, when the PV plant is working, the cost of generating

electricity is equivalent or lower to buying electricity to the grid. Thus, in

those places where it is more favorable to install a PV system to supply

electricity, the economic feasibility should be guaranteed by a high level of

self-consumption and this explains the reduction of direct subsidization

schemes (i.e. feed-in-tariffs).

The power supplied by residential solar PV installations is a function of

the instantaneous solar irradiation that reaches the panels, which is in turn

affected by numerous factors concerning the season, the time of the day

and the meteorological conditions. Therefore, such installations suffer

from the problem of mismatch between production and demand, because

the peak in demand often occurs when the PV production is low (e.g. early

morning and late night). The approaches that can be adopted to mitigate

the discrepancy between the production and the demand patters are

mainly two: (i) equipping the PV installation with a storage means, i.e.

adjusting the production to the demand; or (ii) setting up a load deferral

strategy capable of adapting the demand profile to the renewable energy

availability. The latter option is usually referred to as demand response

(DR). The presence of a flexibility measure can also be used to consume

more energy at times when the price is lower, like in time-of-use tariffs (i.e.

dual tariff, where the energy has a peak and an off-peak price).

A number of studies considered the integration of storage devices with

renewable plants in grid connected systems. Gitizadeh and Fakharzadegan

[74] suggests a procedure for the sizing of batteries that favors their

optimal scheduling. Dufo-López and Bernal-Agustín [75] proposes a

methodology for the profitability assessment of battery storage systems to

reduce consumers’ electricity bills when variable purchase prices of

electricity are in place. This study explicitly considers lead-acid and

lithium-ion batteries, but does not include any renewables-based power

plant. Ratnam et al. [76] quantifies the operational savings deriving from

the addition of a battery system to a residential PV system with time-of-use

electricity prices, without considering the battery technology and the

equipment cost. Mulder et al. [77] studies how the increasing price of

electricity along the years impacts on the economic attractiveness of

different sizes of the storage. Nottrott et al. [78] calculates the net present

value for the investment of different sizes of lithium-ion batteries, in

connection with their ability to reduce the peak load consumption of a large


consumer. Assuming the conditions present in 2011, the study concludes

that only a decrease in the system cost around 55% would improve the

profitability of batteries to an acceptable level.

On the other hand, demand response has been studied primarily from

the perspective of distribution system operators [79,80], with the aim of

reducing dispatch cost and increasing power quality performance. Demand

response strategies at residential level encompass alternative options

which can be subdivided into postponable appliances and buffered

appliances. The first category includes all those appliances (e.g. washing

machines, tumble dryers and dishwashers) that offer flexibility since their

operating cycle can be shifted within a time interval defined by the user.

The second group involves appliances that have a built-in buffer in which

energy is stored (e.g. electric water heaters with a tank). The flexibility

potential of each appliance is determined by the size of the buffer (e.g. the

water storage tank) [81]. The role of voluntary demand response on a

residential aggregate electrical load, under a dual-tariff pricing structure,

is assessed in Muratori et al. [82]. The study finds that DR contributes to

increasing the load level during the night, but creates an even steeper load

peak during the transition between the peak and the off-peak rate,

concluding that the local optimum does not coincide with the global one.

Residential demand response constitutes the focus of a Belgian pilot

project with the scope of quantifying the flexibility potential of domestic

appliances [81,83]. The project defined the potential to modulate the

power consumption of a household as a function of the time of day for

different typologies of appliances. Water heating was found to offer the

highest savings in terms of energy bills since the domestic hot water load

is larger than that of postponable appliances [83].

A few attempts to study the effect of the combination of battery storage

and demand response have been reported. Wang et al. [84] optimizes the

operation of a stand-alone hybrid renewable energy system that includes

PV panels, a wind turbine, a diesel generator, and a battery pack in order

to minimize the operating cost. The possibility to defer loads of particular

postponable appliances (dishwashers and tumble dryers) is also

considered in the study. Huang et. al [85] proposes an optimization

algorithm that minimizes the expenditure for electricity of a house using

both household appliances and the battery of an electric vehicle as

flexibility measures.


None of the previous studies performed a comparison between the

operational savings obtained with one flexibility measure against those

obtained with the other, as applied to the same system. Also, the equipment

cost has never been taken into account when quantifying the profitability

of the two options in an energy system directly connected to a renewable

energy plant. Therefore, one part of the present study (Chapter 3) contains

a direct techno-economic comparison between the integration of batteries

and the implementation of a demand response strategy to a small-scale PV

plant. This constitutes the main original contribution of the study. The two

options are ranked according to their potential to reduce the electricity bill,

with limited adjustments to the original equipment.

Battery energy storage in island settings

Battery energy storage systems can provide support to the operation of

the electrical grid by facilitating the integration of intermittent renewable

generators, which are characterized by steep variations in their production

patterns [86]. This in the main reason why grid operators (TSOs and

DSOs)6 have started investigating the possibility of employing batteries to

compensate the fluctuations introduced by utility-scale wind and solar

power plants in order to avoid congestion issues. Their variable behavior

can lead the grid operator to curtail the energy coming from renewable

plants if the demand is not high enough [87]. The problems connected to

intermittent renewable plants are more apparent in remote systems, e.g.

islands, with isolated mini-grids (with tens or hundreds MW of installed

power) [87]. The effects deriving from a significant number of RES

generators are amplified by the small size of the network, hence, the

occurrence of curtailments is higher. Therefore, islands represent a

framework where the economic and environmental benefits of storage

systems are more visible. The replacement of fossil fuels for electricity

generation with less polluting technologies in island settings has been the

object of some review articles (e.g. [88,89]) that affirmed that hybrid

energy systems (i.e. where the integration of renewable and storage

technologies with conventional thermoelectric plants occurs), can improve

the reliability and the environmental impact of the electricity generation

system [90]. In these contexts, energy storage systems are considered

necessary to reach renewable shares above 50% of the total electrical

6 TSO: Transmission System Operator. DSO: Distribution System Operator.


energy consumption [88]. Among the available technological options,

batteries appear the preferred alternative if pumped hydro solutions

cannot be implemented [89]. For this reason, economic and environmetal

performance of utility-scale battery energy storage in micro-grids has been

analyzed in several articles [21,91]. These studies considered lithium-ion

batteries as the most suitable technology to support the management of the

electricity grid. In fact, the tasks that battery systems can provide in

support of grid operation range from reserve capacity [92,93] to energy

arbitrage [94]. Although most of the studies analyzed lithium-ion batteries

as an option for energy storage, other alternatives might also be cost-

effective, e.g. vanadium flow batteries (VFB). Vanadium flow batteries are

an interesting technology for large-scale energy storage since they have

long cycle lives (up to 20,000 charge/discharge cycles) [7] and they allow

for the decoupling of the design of power and energy specifications [58].

Moreover, this type of flow batteries has to keep an operating temperature

between -5°C and 40°C to maintain the stability of the electrolyte, making

it suitable for warm climates [95]. Some studies investigated the behavior

of VFB both in connection with small-scale solar PV systems [96,97], and

for utility-scale energy storage [98,99]. However, there is room for

additional studies that assess the profitability of these kinds of solutions in

comparison with other storage technologies in large-scale applications.

A number of research institutes and associations have studied how to

improve the environmental performance of energy systems in islands. In

particular, the National Renewable Energy Laboratory (NREL) analyzed

the possibility to integrate wind and solar plants to decarbonize the power

system of the Hawaiian archipelago [100,101]. The International

Renewable Energy Agency (IRENA) dedicated some studies on the

development of renewable energy and storage in island settings [102,103].

Moreover, several research efforts have been conducted to favor the

integration of renewables in the electricity generation sector in the

archipelago of the Azores [104,105], which is politically an autonomous

region of Portugal.

In this thesis, the investments that characterize the installation of

lithium-ion and vanadium flow batteries in island settings are compared.

The degradation of the battery performance along the years is also

considered in the economic analysis, an aspect that until now has been

omitted in a comparison between large-scale storage projects in islands.

These elements constitute the main original contributions of the study.


2.3. Power-to-gas: literature review

The key concept of power-to-gas lies in the possibility to transform the

excess of electricity production – which cannot be dispatched because of a

lack of demand – into a gaseous fuel that could be more easily stored,

transported and used. The obtained gas can be re-converted into electricity

or used for other purposes. The main strength of this approach is the

simplification of the electrical energy management, which would pass from

a paradigm in which an imbalance between production and demand is not

permitted – that is, the two quantities have always to be equal to one

another – to one in which the energy can be produced at one point in time

and easily used at another. Moreover, the high energy density of gaseous

fuels permits a higher quantity of energy to be transported and this has a

crucial role in countries were the centers of consumption are distant from

the bulk of the intermittent generation [21].

The idea of power-to-gas was born in the framework of the hydrogen

economy, that had the objective to utilize hydrogen as the preferred vector

for energy transformation. This is confirmed by Geithner [106] in a review

study of the P2G pilot projects. Until 20127, the majority of the projects

involving electrolysis with low-cost renewable energy were focused on the

production of hydrogen, both for mobility and for electricity generation

applications. Around the first years of the 2010s, a growing interest

towards the production of synthetic fuels through electrolysis of water and

CO2 can be noticed. The study produced by Graves et al. [107] presents the

perspectives and the technologies of CO2-recycling to produce non-fossil

hydrocarbons. Special relevance is given to the pathways that include the

electrolysis of water and CO2 in high-temperature cells and the subsequent

methanation of the produced syngas, to obtain a methane-rich gas.

During the last 7-8 years, the number of studies and pilot projects

involving P2G showed a marked increase. A part of them still focuses on

the production of hydrogen [23,108], while others have broadened the

perspective through the analysis of power-to-methane and power-to-liquid

processes, where the final product is synthetic methane or a liquid fuel

produced mainly via Fischer-Tropsch synthesis [24,37], and the hydrogen

needed for the reactions comes from the electrolysis of water. A new

7 G. Geithner published her paper [106] at the end of 2012 and, therefore, that is considered the time limitation of the study.


expression, electrofuels, has been coined [70,109] to indicate the crucial

role that electricity plays in their synthesis.

Carbon sources for Power-to-gas

The availability of cheap or zero-cost electricity is one of the main

factors influencing the economic viability of P2G systems. However, the

production of synthetic hydrocarbons also requires the presence of a

source of carbon to recycle, which is relatively cheap and concentrated. The

possibilities for carbon collection, especially in the form of CO2, are

thoroughly presented in Reiter and Lindorfer [110]. In particular, the

processes are:

· CO2 from combustion in power plants

· CO2 as a by-product in production processes

o CO2 with biologic origin (biogenic)

o CO2 from industrial processes8

· CO2 from the air

The concentration of CO2 influences the ease and the profitability of

recycling carbon. Most of the exhaust streams deriving from the listed

routes have a concentration below 15% vol.; however, in biotechnological

processes (e.g. bioethanol production, biogas upgrading and biomass

fermentation) the streams can reach concentrations higher than 95%. CO2

recovery from combustion offers room for improvement with respect to

capture efficiency and energy intensiveness, which increase the fuel

consumption. This performance handicap can be quantified in terms of

additional emissions of CO2 per quantity of captured CO2. This so-called

CO2 penalty ranges between 142 and 615 gCO2/kgCO2,captured according to the

fuel (natural gas, coal or solid biomass) and the capture technology (Post-

combustion, Pre-combustion, Oxy-combustion) [110]. Fermentation and

biogas upgrading through CO2 separation are able to yield a purge gas

(considered a by-product) with very high CO2 concentrations, that can be

used with minor further treatments and, consequently, very little

8 The biggest anthropologic single source of CO2 is a synthetic liquid fuel production plant (operated by SASOL in Secunda, South Africa) that performs a coal-to-liquid process. However, the number of these plants is very limited in the world and they are not present in Europe. Therefore, they are not explicitly considered.


additional energy or monetary expenditures9. Refineries, cement factories,

and steelworks are large CO2 emitters and with appropriate technologies

they could become suitable for carbon capture. The capture efficiency

spans from 59 to 90%, and the CO2 penalty lies in the range 116-487

gCO2/kgCO2,captured.

Finally, the presence of CO2 in the ambient air makes it theoretically

possible to isolate this compound for fuel synthesis. Nevertheless, the

process, usually chemical absorption, is expensive because of the very low

concentration in the atmosphere (around 400 ppm) [111], leading to

energy requirements in the order of 1.5–2.5 kWh per kgCO2 and specific cost

around 150 – 320 € per kgCO2. In spite of the great potential of this concept,

the techno-economic conditions appear prohibitive for a massive diffusion.

Biogas upgrading

Among the possible carbon sources for power-to-gas, the biomass-

related ones appear to be the most cost-effective, as pointed out in Section

2.3.1. In particular, biogas upgrade can be considered one of the most

convenient routes because raw biogas contains a substantial quantity of

CO2 that can be converted into hydrocarbons. Furthermore, biogas is

usually obtained from residues treatment processes, which already

contribute to reduce pollution and to minimize landfilled waste. Therefore,

coupling biogas upgrading processes with power-to-gas applications

results in a carbon-negative environmental impact.

Prior to the diffusion of the power-to-gas concept, biogas had already

gained large attention as a renewable fuel. Biogas is obtained through the

anaerobic digestion of residual organic matter, such as sewage sludge,

manure and agricultural waste. This gaseous mixture is principally

composed of methane (50-70% vol.) and carbon dioxide (30-50% vol.)

[112]. In Europe the production of this energy carrier started to be

incentivized at the beginning of the 21st century in several European

countries such as Germany, Sweden, Austria and Italy [113]. At the

beginning it was mainly used in boilers or CHP plants because the

relatively high content of non-combustible components limited its use as

fuel. Therefore, the removal of CO2 is necessary to turn it into a natural gas

substitute.

9 The only process that might be required is the drying of the CO2-rich stream.


Several processes have been developed to perform this operation and

they all focus only on CO2 separation, without paying much attention to its

re-utilization. The most important ones are water scrubbing, membrane

separation and pressure swing adsorption.

Water scrubbing (WS) is the most widespread technology both for the

cleaning and the upgrading of biogas and it exploits the different solubility

that carbon dioxide and methane have in water, according to Henry’s law

[114]. The process takes place in a scrubbing column where the contact

between water and biogas is favored by properly shaped internal packing

elements (i.e. Raschig rings, Berl saddles, pall rings, Lessig rings) [115].

The CO2-rich water is then sent to a stripping column in which an air

stream removes the CO2 and regenerates the water stream.

Membrane separation (MS) processes remove CO2 and other pollutants

from biogas thanks to the selective permeability of porous media. The

materials used for the membranes are usually polymers (e.g. cellulose

acetate and polyimide) [116]. The cleaning of sulfur compounds and water

should be done prior to CO2 separation, in order to prevent corrosion from

degrading the porous medium [114], unless the membrane is specifically

treated to withstand an aggressive environment [117].

Pressure swing adsorption (PSA) is based on the different affinity that

the chemical species present in the biogas have with an adsorbent material,

usually zeolites or activate carbon. In this process, the pressurized biogas

enters a vessel filled with a high surface area material that selectively

adsorbs the compounds that need to be separated, while methane can be

collected from the top. When the adsorbent pack reaches saturation, the

internal pressure drops, the gases desorb and can be removed [118]. The

desorbed gases usually contain a non-negligible amount of methane and

they should be recycled to increase the methane recovery. H2S is

recommended to be removed before the PSA takes place since it is strongly

retained by the absorbent material and interferes with the CO2 separation

[114].

The methane content in the upgraded biogas is usually higher than 95%

in all the aforementioned technologies, while the separated CO2 is usually

vented, without being further utilized.

The conversion of CO2 into synthetic fuels has been extensively

investigated in connection with high-temperature electrolysis devices (e.g.,

solid oxide electrolysis cell (SOEC)) [69,119,120]. The findings allowed

techno-economic analyses of power-to-gas processes using pure CO2


streams to be developed [121,122]. These processes included the injection

of pure CO2 into the solid oxide electrolyzer where syngas was produced.

Subsequently, a series of methanation reactors, placed downstream from

the SOEC, transformed hydrogen and carbon monoxide into methane. An

available commercial catalytic process that can be used to obtain this result

is named TREMP (“Topsøe Recycle Energy-efficient Methanation

Process”) [123]. This process allows the methane content in the product

gas (called biomethane or synthetic natural gas) to reach a molar fraction

greater than 90%, with a consequent improvement of the net calorific

value. The resulting product gas has to be compatible with the

specifications required for grid injection: gas gravity, Wobbe Index and

content of non-hydrocarbons. The conditioning is generally performed by

the admixture of a heavier compound (e.g. propane or nitrogen).

The high CO2 content present in biogas (around 40%) that is usually

lost in separation-based upgrading processes could be used as carbon

source for the synthesis of hydrocarbons. Based on this consideration,

Collet et al. [124] analyzed the possibility of upgrading biogas using excess

electricity in a life cycle assessment perspective. However, the processes

that are described in the paper only consider the production of hydrogen

and the subsequent reaction with biogas, instead of feeding biogas directly

to the electrolyzer. The direct upgrading of biogas through co-electrolysis

of water and carbon dioxide avoids the separation of CO2 from CH4 that,

instead, characterizes all the other upgrading techniques. In addition to

that, electrolysis produces high-purity oxygen that is a high value added

substance, with a relevant economic value. The simulation of an upgrading

process where biogas is directly injected in the electrolyzer, its exergy and

economic analyses, as well as the comparison with other upgrading

techniques, constitute the original contributions of this thesis to the state

of the art.


2.4. Summary of the main contributions to the state of the

art

The present thesis is structured as a compilation of published articles

and each of them added a piece of knowledge to the state of the art. In the

previous sections a literature review is performed to identify the research

gaps. The original contributions of this dissertation were stated at the end

of sections 2.2.1, 2.2.2, and 2.3.2, and Table 2 links them to the appended

papers.

Table 2: Summary of main contributions of the papers appended in this thesis work

Papers I and II

Research Topic Integration of residential PV systems with cost-effective

flexibility measures

Contributions to the state of

the art

· Comparison between the monetary savings produced

by two buffering options (battery storage or demand

response) separately applied on the same small-scale

grid-connected residential system powered with a PV

plant.

· Quantification of the equipment cost to implement

demand response

Paper III

Research Topic Integration of large-scale battery systems in island

settings


the art

· Comparison between a lithium-ion and a vanadium

flow battery investment in island settings

· Incorporation of battery degradation issues in the

investment analysis of the two large-scale battery

technologies


Table 2 continued

Papers IV and V

Research Topic

Thermodynamic and techno-economic analyses of a

biogas upgrading process via electrolysis and

methanation


the art

· Thermodynamic model of the studied process for

biogas upgrading with the calculation of energy and

exergy efficiency of the process for different

configurations

· Comparison of the analyzed plant with other

techniques for biogas upgrading

· Determination of the cost of synthetic natural gas

resulting from the biogas upgrading process through

electrolysis and methanation

· Quantification of the benefits deriving from the

valorization of by-products (e.g. oxygen)

OPTIMIZATION OF ELECTRICITY DISPATCH WITH BATTERY STORAGE SYSTEMS (PAPERS I, II AND III) | 33

3. Optimization of electricity dispatch with battery

storage systems (Papers I, II and III)

In this chapter, two possible applications of battery energy storage are

presented. One addresses the reduction of the mismatch between

electricity production and consumption in a residential system equipped

with a small-scale PV plant. The other concerns the installation of a large-

scale battery storage system to integrate large-scale renewable plants in an

island setting. Lithium-ion batteries have been considered in both the

cases due to their versatility, and they have been compared with other

alternatives (i.e. lead-acid and vanadium flow batteries) on the basis of the

economic benefits that they produce. For both the applications the

simulations have been performed using Matlab and its routines.

3.1. Battery energy storage vs. demand response in

residential PV plants (Papers I and II)

The installation of distributed and intermittent RES in residential

contexts, where the electricity load profiles are driven by end-user

behavior, leads to a mismatch between demand and production of

electricity. The solution to this issue is expected to reduce the quantity of

electrical energy exchanged with the grid with positive repercussions on

the electricity costs borne by the end users. Two approaches can be used to

face this challenge: adapting the production profile to the demand (energy

storage), or shaping the demand profile in order for it to be more similar

to the production schedule (demand response). For the purpose of this

study, demand response entails only the alteration of the electric load

connected to domestic hot water (DHW) supply, on the basis of time-of-

use electricity prices and renewable energy availability. The study is

focused on a Portuguese household that is connected to the electricity grid

and buys electricity in a dual-tariff regime.

In Section 3.1.1, key qualitative results are reported based on a

preliminary study. This investigation addressed the reduction of the

mismatch based on the allocation of the surplus of production according to

34 | OPTIMIZATION OF ELECTRICITY DISPATCH WITH BATTERY STORAGE SYSTEMS (PAPERS I, II AND III)

a Boolean logic approach, following an “if-then-else” rule with two

mutually exclusive options. Lead-acid batteries were chosen as storage

means, due to their low cost and technological maturity.

Section 3.1.2 presents a linear programming optimization algorithm

that was developed to minimize the cost of the system in a constrained

environment, by taking into account the cost contribution of the

equipment. This method was chosen because all the components of the

physical energy model were assumed to have a linear behavior and because

it guaranteed the existence of an optimal solution. Therefore, the use of

more complex techniques was considered unnecessary.

Lead-acid and lithium-ion batteries were compared in terms of cycle

cost [94]. Lithium-ion appeared to be the most convenient option. Demand

response was implemented through a smart-electric-boiler control, which

is able to heat the water when the electricity is cheaper.

Preliminary study: Boolean logic approach (Paper I)

In the following pages the preliminary techno-economic analysis of a

household with small renewable plants and buffering options is performed.

Two weeks, representative for different periods of the year, are considered

in order to simulate different production levels of the renewables. The

storage plant is constituted by a deck of batteries, while the demand

response consists in the possibility of deferring part of the load.

System layout and general assumptions

The data for the renewable production are obtained by the

meteorological station located in the main campus (Alameda) of Instituto

Superior Técnico (IST - Lisbon) and they have a 1-hour resolution [125].

The main information about the PV panels, wind turbine and battery pack

is reported in Table 3, while the schematic representation of the system is

illustrated in Figure 9. The renewable installations and the batteries

reproduce the micro-generation unit present in the IST campus.

The load profile was derived from the typical consumption patterns for

residential customers available on the Portuguese TSO website [126]. For

the demand response situation, the maximum hourly load that was allowed

to be deferred was equal to 500Wh, equivalent to the minimum load value.

The load could be shifted continuously, as for hot-water boilers and electric

heaters. The dual tariff for electricity is characterized by a peak price of 16


c€/kWh – active from 8 am to 10 pm – and an off-peak price of 9 c€/kWh

for the remaining hours. These prices are the same for selling and for

purchasing electricity.

Table 3: Summary of the equipment specifications [127].

Equipment

type

Name Number of

units

Total

Power/Capacity

Wind Turbine Southwest Wind-

Power Air 40 1 400 W

PV

modules LDK 235Wp 6 1410 W

Battery

system Autosil EC3 240Ah/5h 3 720Ah/5h

Inverter SMA Sunny Island SI

2224 1 230 V; 2200 W

Figure 9: Layout of the renewable generators, storage system and load [127].

The control logic for the two cases is reported in Figure 10 and Figure 11.

The operating parameters of the batteries are shown in Table 4.


IFRES POWER

> LOAD POWER

IFSTORED ENERGY

< MAX. STORED

ENERGY

IFSTORAGE POWER

> UNFULFILLED

LOAD

THEN ELSE

INJECT INTO STORAGE

INJECT INTO GRID

WITHDRAW FROM

STORAGE

WITHDRAW FROM GRID

THEN ELSE THEN ELSE

Figure 10: Control logic for the storage case [128].

For the demand response case the residual shifted load, i.e. the load that

has still to be processed at the end of the day, is executed at the midnight

of the new day, by buying electricity from the grid at a constant power of

500W for the number of hours necessary to complete the residual load.

This expedient simulates the reallocation of the loads into the off-peak

hours.

Table 4: Operating parameters of the batteries.

Maximum depth of discharge (DODmax) 80%

Capacity (kWh) 17.28

Maximum charge/discharge power [W] 2200

Charging efficiency (ηch) 0.9

Discharging efficiency (ηdisch) 0.9

Inverter efficiency (ηinv) 0.95


IFRES POWER

> LOAD POWER

IFSURPLUS POWER

< MAX. SHIFTABLE

LOAD

IFDEFICIT POWER

< MAX. SHIFTABLE

LOAD

THEN ELSE

FEED LOAD FEED LOAD + INJECT INTO

GRIDSHIFT LOAD

SHIFT LOAD + WITHDRAW FROM GRID

THEN ELSE THEN ELSE

Figure 11: Control logic for the demand response case [128].


Qualitative results

The simulations showed that the electricity supply from the renewable

energy generators is below 40% of the weekly load both for winter and

summer conditions. This entails that the majority of the production is

already in phase with the consumption and that the surpluses are very

limited, especially in winter. This fact does not favor the presence of the

batteries, which are seldom charged. Also demand response does not have

a large impact because there are few occasions to reabsorb the shifted load.

For this reason, the system was also studied with a larger PV plant having

a size double than the previous one. The results in this case are favorable

to demand response because this technique is able to reduce the

consumption during peak hours, spreading it over the off-peak ones,

independently of the availability of surplus PV electricity. Instead,

batteries’ operation is restrained by the energy previously stored, which is

directly proportional to the excess of solar energy. In addition to that, the

electricity is sold and purchased at the same price and this fact does not

promote self-consumption. Moreover, batteries are penalized by the round

trip efficiency, which further reduces the amount of stored energy that is

reallocated for consumption.


In-depth study: optimization approach (Paper II)

The approach adopted in the previous sections is characterized by a set

of limitations that need to be addressed to enlarge the validity of the

results. Firstly, in the preliminary study the strategy for the dispatch of

energy had not been optimized considering the different prices of

electricity. Secondly, the cost of the equipment to mitigate mismatch had

not been estimated. Therefore, a more detailed comparison should be

performed, as described in the next sections.

Case study description

The study was conducted on a typical Portuguese single-family dwelling

in the area of Lisbon with 3 occupants [129], 4 rooms and 3.45 kW of

contracted power, equipped with an electric water boiler. The household

was assumed to be connected to a PV plant.

The load profiles were obtained from the measurement campaign

performed in the framework of the project “Net zero energy school -

Reaching the community”, where residential buildings were characterized

according to their electricity consumption [130].

Four weeks (one per each season) were selected and the load

distribution between peak and off-peak hours is represented in Table 5.

Table 5: Load distribution for the four representative weeks (adapted from [94]).

Winter Spring Summer Autumn

Weekly load consumption [kWh] 111 124 138 123

Peak load share in dual tariff 58% 55% 59% 55%

Off-peak load share in dual tariff 42% 45% 41% 45%

The purchase price of electricity was defined according to a dual-tariff

weekly cycle [131,132], including VAT (23% in Portugal). The selling price

was equal to 90% of the average market price (on the Day Ahead Market)

of the previous month [133] (Table 6).

The solar irradiation data were obtained from the same weather station

that was used for the preliminary study, with the same resolution. These

values were used to determine the electricity production of the PV panels

that provide electricity to the house, using the three-parameter model

described in Crispim et al. [134]. The calibration of such model was done


with the data reported on the datasheet of the panels (polycrystalline LDK

250P-20 [135]). The three-parameter model leads to satisfactory

estimations10 of the PV production for polycrystalline panels, if datasheet

values are used for its calibration [136]. The analysis was carried out with

three different sizes of PV, namely 1.5, 3 and 4.5 kWp, in order to correlate

the PV installation size with the economic profitability of the proposed

flexibility measures.

Table 6: Selling and purchase prices of electricity. Adapted from [94].

[€/kWh] Purchase price - weekdays 7 a.m. – 0 a.m. 0.234 0 a.m. – 7 a.m. 0.123 Purchase price - Saturday 10 a.m. – 1 p.m. / 6 p.m. – 10 p.m. (winter time); 9 a.m. – 2 p.m. / 8 p.m. – 10 p.m. (summer time)

0.234

Remaining hours 0.123 Purchase price - Sunday All day 0.123 Selling price Winter 0.0427 Spring 0.0388 Summer 0.0493 Autumn 0.0467

The presence of a feed-in-tariff (FIT) scheme in the past led the PV

system to be often oversized in order to maximize the energy production.

Oversized systems are likely to produce a substantial excess of electricity

during low-consumption hours. If this surplus is not used locally, it must

be transferred to the grid, though being poorly compensated (90% of the

wholesale market price for electricity).

Dispatch optimization

The design of the algorithm for the optimal dispatch has been oriented

to reach the cost minimization, which constitutes the objective function

(Equation ( 6 ) for battery storage and ( 7 ) for demand response,

respectively). The decision variables are the quantities of energy crossing

10 Relative error below 5% [136].


the system borders (corresponding to purchased and sold electricity), and

those relative to the buffering option (charged/discharged energy for

batteries - 𝐸𝑐ℎ /𝐸𝑑𝑖𝑠𝑐ℎ - or energy shifted in/out for demand response -

𝐸𝑠ℎ,𝑖𝑛/𝐸𝑠ℎ,𝑜𝑢𝑡11). The upper and lower boundaries of the previous variables

were established in accordance with the contracted power and the

technical specification of the equipment. Figure 12 shows a block structure

of the system under evaluation.

𝑂𝐶𝐵𝐸𝑆 = ∑((𝑃𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑𝐸𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑,𝑖 − 𝑃𝑠𝑜𝑙𝑑𝐸𝑠𝑜𝑙𝑑,𝑖)

24

𝑖=1

+ (𝐶𝐵𝐸𝑆𝐸𝑐ℎ,𝑖))

( 6 )

𝑂𝐶𝐷𝑅 = ∑ ((𝑃𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑𝐸𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑,𝑖 − 𝑃𝑠𝑜𝑙𝑑𝐸𝑠𝑜𝑙𝑑,𝑖)

24

𝑖=1

+ (𝐶𝐷𝑅𝐸𝑠ℎ,𝑖𝑛 + 𝐶𝐷𝑅𝐸𝑠ℎ,𝑜𝑢𝑡))

( 7 )

The objective function (OC – operation cost12) contains two groups of

terms: the first one, includes the value of the electricity purchased

(𝐸𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑) and sold (𝐸𝑠𝑜𝑙𝑑) with the relative prices (𝑃𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑑 and 𝑃𝑠𝑜𝑙𝑑);

while the second one corresponds to the cost of using the equipment (𝐶𝐵𝐸𝑆

for batteries or 𝐶𝐷𝑅 for demand response) and it considers the cost related

to the actual use of the system. Linear programming does not allow the cost

of the equipment to be inputted as a lump sum – which is how the payment

transaction is usually conducted –; therefore, the investment cost should

be spread over the operation of the components, in a levelized-cost-like

approach.

11 The energy shifted-in represents the additional load that is executed when the conditions for consumption are favorable, while the energy shifted-out is the load deferred when the conditions are unfavorable. 12 The expression operation cost groups all the cost components that determine the operation of the system.


Figure 12: Schematic structure of the system with the indication of the energy flows. The buffering option is the battery system or the water tank connected to the boiler controller.

For storage and load deferral systems this concept is named cycle cost

(CC) [137], and it is referred to the number of charge cycles for BES and of

load shifting cycles for DR ( 8 ).

𝐶𝐶 =𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝐶𝑜𝑠𝑡

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑐𝑦𝑐𝑙𝑒𝑠 𝑖𝑛 𝑙𝑖𝑓𝑒 ( 8 )

It can be also referred to each kWh of energy that is handled by the

system and in this case it is called specific cost (𝐶𝑠𝑝) ( 9 ).

𝐶𝑠𝑝 = 𝐶𝐶 ∙1

𝐸𝑛𝑒𝑟𝑔𝑦 𝑝𝑒𝑟 𝑐𝑦𝑐𝑙𝑒 ( 9 )

For the case of storage, the specific cost is assigned totally to the

charging process. On the other hand, the cost of DR is attributed both to

the energy shifted in and out. The detailed description of the algorithm is

available in Paper II.

Assumptions for storage and demand response operations

The choice of the battery technology was made thanks to a comparison

between the cycle cost of lead-acid and lithium-ion batteries. Two

equivalent options were simulated and lithium-ion batteries were found to

have the lowest cycle cost (Table 7). The option that was selected for the


storage system is represented by the lithium-ion battery Tesla Powerwall13,

manufactured by Tesla Motors© and launched on the US market in early

April 2015. The battery is coupled with an inverter capable of handling the

bidirectional power flow. In this case, the inverter is Fronius Symo Hybrid

5.0-3-S [138], which is considered optimized for the chosen battery, due to

the collaboration between the two companies [139]. However, the

computed battery cost appears very high in order to allow the profitability

of the storage system. Therefore, the cost for the storage system has been

decreased by 65%, resulting into a capacity cost lower than 300

€/kWhinstalled, which reflects an optimistic learning curve for residential

storage systems [140].

Table 7: Operating and economic parameters for the two storage options. Adapted from [94].

Li-ion Lead-acid Specs Capacity [kWh] 6.4 6.2 Maximum depth of discharge (DODmax)

80% 60%

Roundtrip efficiency (ηRT) 0.92 0.81 Expected Life [cycle number] 3500 800

Cost (€) Battery deck 2730 [141] 14 1486 [142] Battery inverter 2208 [143] 2241 [144] Installation (5% of the total) 247 186 Total Cost of Storage 5185 3913 Cycle Cost (CC) [€/cycle] 1.48 4.89 Specific cost (Csp) [€/kWhcharged] 0.289 1.317 Capacity Cost [€/kWhinstalled] 810 632

The presence of an electric boiler for domestic hot water supply and the

nature of the DHW load makes it suitable for a load deferral strategy

[104,145]. The load connected to the consumption of domestic hot water,

which has been estimated equal to 4.19 kWh/day15, corresponds to a daily

13 In this work the first model of Tesla Powerwall is considered. By now, the only option available has 13.5 kWh of usable capacity (at DoD = 100%) and a round-trip efficiency of 90% [141]. 14 The conversion factor is €/$ = 1.1. 15 Domestic hot water is intended as the drinking water that must be supplied at a temperature of at least 45ºC [204]. However, to exclude the proliferation of legionella bacteria in the tank the minimum temperature should be 60 ºC. The grid water temperature is considered equal to 15 ºC.


consumption of 40 l/person [146]. This value has been incremented by 5%

to account for the heat loss through the pipes and the tank shell, resulting

in a value of 4.40 kWh/day. The cost of a smart boiler controller has been

set at 200 € after a market analysis of the product that could perform the

task of shifting the DHW load [94]. The functional life has been considered

equal to 10 years.

The European eco-labeling directive [147] was used to model the energy

load profile for DHW (Figure 13). The variations of the proposed DHW load

profile are buffered by the presence of a storage tank, and this justifies the

use of a constant profile.

The installation of a so-called “smart boiler” or of a controller that turns

a traditional electric boiler into a smart one is necessary for the

implementation of a DR strategy. A smart boiler is an appliance that can

shift in time the heating of water according to the instantaneous electricity

price, without compromising the consumers’ needs.

Figure 13: DHW daily load profile.


Energy analysis results

The generation of electricity of the PV system was simulated for 4

weeks, one per season, and three sizes of the installation (1.5, 3 and 4.5

kWp). In the majority of the cases the weekly renewable production is not

enough to quantitatively satisfy the weekly electrical consumption of the

house. The share of solar energy over the total load spans from 29% (in

January with 1.5 kWp) to 126% (in July with 4.5 kWp). The renewable

production is likely to surpass the load consumption for particular periods

of the year (e.g.: spring and summer) only in the case of 4.5 kWp. However,

even if the weekly production is lower than the demand, the instantaneous

renewable production can be higher than the corresponding load, and the

resulting energy surplus has to be handled.

In Table 8, the comparison of the two strategies for enhanced self-

consumption (BES and DR) shows that, with the increase of the installed

PV power, the percentage of the renewable production which is sent to the

grid increases more for DR than for storage. This means that the exchange

of electricity with the grid is more intense when DR is adopted, because the

buffer capacity of DR (4.4 kWh/day) is lower than that of the batteries (5.12

kWh/day). The highest percentage of self-consumed (i.e. charged) energy

over the total surplus energy appears in the case of 3.0 kWp, because for

larger installed capacity the battery size acts as the limiting factor for

batteries to uptake more electricity (Figure 14).

Table 8: Summary of the energy flows for the considered week in summer for the cases of PV production without buffering strategies (PV-only), with battery storage (BES) and with demand response (DR) (adapted from [94]).

PV

only BES DR

PV only

BES DR PV

only BES DR

PV peak power [kW] 1.5 3.0 4.5

RES production [kWh]

58 116 174

Load consumption [kWh]

138 138 138

Import from grid [kWh/week]

92 83 80 80 52 57 76 48 55

Export to grid [kWh/week]

12 2 0 59 24 36 113 80 91


Figure 14: Percentage of stored energy over the total renewable production.

Economic analysis results

The different weekly savings16 for DR and BES are shown in Figure 15.

The implementation of demand response measures appears more

convenient in all the cases because of the considerably lower specific cost

(𝐶𝐷𝑅 = 0.62 vs. 𝐶𝐵𝐸𝑆 = 10.1 c€/kWh). In other words, every kWh of surplus

electricity can be used either to heat up water for future demand with a

price of 1.24 c€/kWh (2𝐶𝐷𝑅); or it can be stored with a cost of 10.1 c€/kWh

(𝐶𝐵𝐸𝑆), of which only 88% is available to cover the load due to the round-

trip and inverter efficiencies (𝜂𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 = 95%).

If each week is considered to be representative for an entire season (13

weeks); then, the annual costs and savings can be roughly estimated.

Referring the annual savings to the installed PV capacity, the configuration

with 3 kWp appears the one which gives the best results for storage (25

€/kWp/year), while the one with 1.5 kWp produces the highest specific

savings for demand response (102 €/kWp/year). The total annual savings

for DR reach a maximum in the 4.5 kWp configuration, but the specific

savings are the highest for the smallest installation size (Table 9).

The annual equipment cost is calculated by dividing the initial cost of

the equipment by the functional life in years (see Section 3.1.3 for the

16 Difference between the cost for electricity in the PV-only case and the cost in presence of a flexibility measure.

0%

5%

10%

15%

20%

25%

30%

35%

WINTER SPRING SUMMER AUTUMN

Sto

red

En

erg

y /

RE

S p

rod

uc

tio

n

RES Production 4.5 kWp RES Production 3 kWp RES Production 1.5 kWp


equipment cost definition). The lower specific cost of demand response

allows the system to purchase energy during off-peak hours to cover part

of the peak load. For this reason, demand response strategies reach higher

relative savings for smaller PV systems. The cost of storing energy is higher

than the difference between peak and off-peak electricity, hence it is not

convenient to store off-peak electricity, restricting the storable energy to

the overproduction. This means that the new installations of PV systems

for self-generation, which are generally small compared to the contracted

power would benefit from demand response, while storage solutions will

be more interesting for oversized systems installed when the feed-in-tariff

supporting scheme was in place. This conclusion is valid if the house is

equipped with an electric boiler, because the application of automatic

demand response strategies to other appliances could present additional

challenges due to the interference with people’s behavior [81].

Table 9: Total and specific annual savings on the electricity bill for the considered cases (adapted from [94]).

1.5 kWp 3.0 kWp 4.5 kWp

BES DR BES DR BES DR

Annual PV-only cost [€] 879 879 729 729 606 606

Annual electricity cost [€] 800 706 472 522 321 407

Annual equipment cost [€] 181 20 181 20 181 20

Annual savings [€] -102 153 75 187 103 179

% of PV-only cost [%] -12 17 10 26 17 30

Specific annual savings [€/kWp]

-68 102 25 62 23 40


Figure 15: Percentage weekly cost savings compared to the condition without flexibility measures (PV-only) for different PV peak powers. The columns delimited by thicker borders and filled with a texture represent the weekly savings when the storage option is concerned, the solid-fill columns refer to demand response (adapted from [94]).

0%

5%

10%

15%

20%

25%

30%

35%

40%

WINTER SPRING SUMMER AUTUMN

% w

eekly

savin

gs

RES Production 4.5 kWp RES Production 3 kWp RES Production 1.5 kWp BES


3.2. Battery energy storage to promote renewable

integration and grid support: the case of Terceira (Paper

III)

Islands’ energy needs are usually fulfilled through the importation of

fossil fuels, both for electricity generation and for transport. Therefore,

some studies [88,89] address the issue of replacing fossil fuels with less

polluting alternatives.

In the archipelago of the Azores, an autonomous region of Portugal

located 1,500 km away from the mainland, the local utility (Eletricidade

dos Açores – EDA) has been actively committed to increase the quota of

electricity produced by renewable energy sources, especially in the biggest

islands (i.e. São Miguel and Terceira). The island of Terceira17 was chosen

as case study since the renewable installed power on the islands has been

growing considerably in the period 20o8-2018 with the installation of two

wind farms (12.6 MW in total), a geothermal plant (3.5 MW) and a waste

to energy plant (2.3 MW) [148,149]. Moreover, the investment plan for the

period 2018-2022 states that in Terceira the investments in fossil-fueled

electric power plants will come to an end in 2020, while those in renewable

capacity expansion are supposed to rise, with more than 9.5 M€ in 2020

and almost 14 M€ in 2021 [150].

The need for storage to enable higher RES shares in the electricity mix

of isolated energy systems is pointed out in Section 2.2.2 and, among the

possible storage technologies, battery systems are able to perform different

tasks to support the grid. In addition to this, their modularity and their

site-independence (unlike pumped hydro and in-ground compressed air)

make them suitable for different applications and geographical settings, as

thoroughly described in Castillo and Gayme [92], in which grid-scale

battery applications are reviewed in connection with their services to the

electrical system. In this study, batteries are utilized to implement

renewable energy time shifting (also called renewable energy arbitrage)

[151] and to provide reserve power [93].

Lithium-ion (LiB) and vanadium flow batteries (VFB) are the

technologies considered for this study. According to the database of the US

DoE [4], LiB is the technology in stationary energy storage with the highest

17 Terceira is the second most populated of the Azores archipelago with nearly 56,000 inhabitants (2014) [205].


number of storage projects (665), 68 of which can be classified as very

large-scale (>10 MW or 10 MWh). On the other hand, the long cycle life

(up to 20,000 cycles) [7] and the decoupling between energy and power

levels makes VFBs a promising option for large-scale storage [58]. They are

suited for mild climates – like the one present in the Azores – because the

operating temperature should fall in the range -5°C – 40 °C in order to

avoid precipitation in the electrolyte solution [95].

The assessment of the economic performance of a battery storage

facility to reduce RES curtailment and limit the use of thermoelectric

power plants is the main aim of the study. The objective is to model

accurately the unit commitment of the generation system in Terceira and

test the viability of the investment connected to energy storage, according

to common evaluation criteria (i.e. net present value, payback time, return

on investment). Different cost scenarios for LiB and VFB are considered,

in order to deal with the uncertainty that will affect the cost reduction trend

for battery systems in the coming years. This study simulates the behavior

of the system for a whole year and not only for typical days, including also

an estimation of the battery degradation in the economic evaluation.

Case study description and main assumptions

The structure of the market for electricity generation in Terceira is

monopolistic and the price of electricity is set by the Portuguese Energy

Services Regulator (ERSE). The island’s electrical energy production for

the years 2011-2017 is reported in Table 1, in which a decreasing trend in

electricity production can be observed and it is mainly due to the effects of

the economic crisis. On the other hand, the share attributable to renewable

energy sources is growing, with clear implications on the electricity

management. The time horizon for a storage facility to be built has been

assumed between 10 and 15 years, because of the long decision-to-

implementation time that characterizes these projects. Therefore, the

investment analysis is performed starting from the year 2030. The year

2013 is taken as a reference, because it is the year with the most recent and

sufficiently detailed data about the electricity production (i.e. with a half-

hour resolution and the indication of the power produced by each

generator).

The electricity production in 2030 is assumed to be 5% higher than the

corresponding value in 2013, due to the growing electrification trend

taking place both in the residential [152] and transport [105] sectors. The


shape of the electricity profile reproduces that of 2013 and the profile of

wind generation in 2013 is assumed to be sufficiently representative of the

weather conditions of the island for the following years, according to the

approach presented in Varone and Ferrari [24].

Table 10: Electricity production by source and population for the island of Terceira (2011-2017) [149,153].

GWh 2011 2012 2013 2014 2015 2016 2017

Thermoelectric plants 194 181 172 168 168 156 144

Wind 19.0 29.2 34.0 34.6 31.3 33.3 31.3

Waste-to-Energy 0 0 0 0 0 8.5 8.7

Hydro 1.5 2.4 2.8 0.5 0 0 0

Geothermal 0 0 0 0 0 0 9.8

Total RES 20.5 31.7 36.8 35.1 31.3 41.8 49.8

Electricity Production 214 212 209 203 200 198 194

RES Share [%] 9.6 14.8 17.2 16.4 14.6 19.5 25.6

Population (103 ) 56.6 56.6 56.6 56.4 56.1 56.0 56.6

The structure of the generation system by type of plant in 2030 and the

operating assumptions are shown in Table 11. The cost function of each

thermoelectric generator is extrapolated from the data of specific

consumption, reported by the system operator (Equation ( 10 )) [154]. The

dispatch optimization algorithm that is used to simulate the unit

commitment in Terceira is extensively described in Paper III.

𝐶𝐻𝑃,𝑖 = (𝐺𝐻𝑅,𝑖 ∙ 𝑊𝑒𝑙,𝑖 + 𝐺𝑁𝐿,𝑖) ∙ 𝐶𝐹,𝑖 ( 10 )

𝐶𝐻𝑃 represents the hourly production cost of the ith committed thermal

generator, 𝐺𝐻𝑅 is the incremental heat rate, 𝐺𝑁𝐿 is the no-load

consumption and 𝐶𝐹 is the fuel cost.


Table 11: Operating assumptions for the electricity generators in Terceira.

Type of plant

Maximum power [MW]

Operating conditions

Notes

Diesel Engines

12.1 According to the dispatch optimization

Fuel Oil Generators

49 According to the dispatch optimization

Geothermal 7

Continuously at 75% of the maximum power for 11 months (1 is for maintenance)

Not enough data on real behavior. Assumption in accordance with IEA [155]. Rated power value higher than current installed capacity because of planned capacity expansion.

Waste-to-Energy

2.3

Continuously at 80% of the maximum power for 11 months (1 is for maintenance)

In the first year of operation the availability factor was lower than 50%. These plants can work up to 8000 hours per year [156].

Wind 12.6 According to wind availability profile

2013 used as base year.

In electricity generation systems the spinning reserve18 should be able

to counterbalance the loss of the largest generator, without introducing

excessive fluctuations in the power flows or frequency level, thus, ensuring

the continuity of the supply in case of sudden outage [21,157]. In this study,

the spinning reserve was assumed equal to half of the power injected by the

largest online generator. This hypothesis was adopted to tune the model so

that it matches baseline data. Figure 16 shows the relative difference

between the results of the model and the baseline data in terms of

operating cost for electricity generation, and GHG emissions. One week for

18 Spinning Reserve: “Generation capacity that is online but unloaded and that can rapidly respond to compensate for generation or transmission outages” [157].


each month has been sampled and used for the purpose of validation. The

baseline data values are obtained by calculating the result of the cost

function defined in Equation ( 10 ) applied to the values of power of each

generator, made available by the system operator. It can be observed that

the difference is always below 2.5%. This outcome is satisfactory for the

purpose of this study because the uncertainty introduced by the model is

much lower than the uncertainty on the battery price, that will be

addressed in Section 3.2.4.

a)

b)

Figure 16: Percentage difference between model results and baseline data for a) GHG emissions and b) operating cost.

The parameters for the techno-economic analysis of the two battery

plants are reported in Table 12.

-1.5%

-1.0%

-0.5%

0.0%

0.5%

1.0%

% Δ

GH

G e

mis

sio

ns

-2.5%

-2.0%

-1.5%

-1.0%

-0.5%

0.0%

0.5%

1.0%

% Δ

op

era

tin

g c

ost


Table 12: Technical parameters and costs for lithium-ion (LiB) and vanadium flow batteries (VFB)

LiB Ref. VFB Ref.

Capacity MWh 20 – 30 - 20 - 30 -

Maximum Charge/Discharge Power

MW 10 - 10 -

Depth of Discharge % 0.9 [158] 1 [159]

Round-trip efficiency % 95 [160] 85 [7]

Efficiency (end of life) % 80 o.a.19 75 o.a.

Capacity Loss (end of life)

% 20 [161] 10 o.a.

Cycle life 5000-10000

[8] >10000 [8]

Capacity Cost (CAPEX) €/kWh 296 [32] 243 [32]

Operation and

Maintenance (OPEX)20 €/year

3% 𝐼𝑁𝑉𝑂𝑉𝑁

o.a. 5%

𝐼𝑁𝑉𝑂𝑉𝑁 o.a.

Investment life years 20 [91] 20 [91]

Discount rate % 3 [162]21 3 [162]

The estimation of the cash flows for the NPV analysis and the

quantification of the Return on Investment (ROI) are performed according

to Equations ( 11 ) and ( 12 ).

𝑁𝑃𝑉 = ∑ (𝐺𝑎𝑖𝑛𝑆𝑇𝑂 − 𝐶𝑆𝑇𝑂,𝑂&𝑀

(1 + 𝑑)𝑦)

𝑁𝑌

𝑦=1

− 𝐼𝑁𝑉𝑂𝑉𝑁 ( 11 )

𝑅𝑂𝐼 =𝑆𝑎𝑣𝑒 − 𝐶𝑆𝑇𝑂,𝑂&𝑀

𝐼𝑁𝑉𝑂𝑉𝑁

( 12 )

The positive cash flows are represented by the annual savings on the

cost for electricity generation due to the storage installation (𝐺𝑎𝑖𝑛𝑆𝑇𝑂). The

negative ones are due to the operation and maintenance expenditures

attributable to the storage plant (𝐶𝑆𝑇𝑂,𝑂&𝑀). 𝑁𝑌 is the life of the project, 𝑑 is

19 Own assumption. 20 The O&M cost for LiB is lower than that for VFB due to the moving parts needed for the electrolyte flow in the second option. 21 This value corresponds to the average of the cost of capital in the period 2008-2016 that the local utility has declared in their annual report (Relatório e Contas 2016 [162]).


discount rate and 𝐼𝑁𝑉𝑂𝑉𝑁 is the overnight initial investment. The

discounted payback is calculated by equating the NPV to zero.

The ROI is the percentage of the initial investment that is recovered

each year, on average, along all the project. 𝑆𝑎𝑣𝑒 is the average yearly saving

throughout the life of the project.

The performance of the batteries suffers from degradation due to cycle

and calendar life [163–165] and this appears to have always been neglected

in similar studies. The performance loss of the flow batteries is linked to

the mechanical degradation of the membrane between the two electrodes,

which can have implications on efficiency and useful capacity [96]. This

phenomenon affects also the economic benefits produced by the storage

facility. In order to quantify this effect, a series of simulations with

degraded values of efficiencies and useful capacities was performed and the

annual savings under these new conditions were computed. Then, the

positive cash flow was assumed to decrease linearly between these two

values.


Energy and environmental results

The reduction of curtailed energy and of fossil-generated electricity are

considered as the indicators to quantify the contributions of energy storage

to the grid integration of renewables. For the island of Terceira, lithium-

ion and vanadium flow batteries are simulated in two different

configurations, in order to minimize the curtailed energy: with a rated

capacity of 20 MWh (Case A) and of 30 MWh (Case B).

The reduction of the thermoelectric energy supply and the related GHG

emissions for the effect of the battery storage facility is remarkable (Table

13). The thermal production is reduced by 9% compared to the case without

storage (NO STO), as well as the curtailment of wind energy, while the

GHG emissions experience a slightly lower decrease (around 7%) because

of the effect of the emissions deriving from the waste-to-energy (W2E)

plant.

Wind production has a lower dispatch priority, because the renewable

energy coming from plants that would suffer more from being cut out (e.g.

geothermal and W2E) is privileged. Although the majority of the curtailed

energy comes from wind, the operation of the geothermal and waste-to-

energy plants can also be affected by the constraints about spinning reserve

when the demand is low22. The presence of the batteries, thanks to their

possibility to provide spinning reserve, allows the system operator to

disconnect also one of the two fuel-oil generators, so that all the energy can

be dispatched. This option considerably reduces the surplus, which passes

from 13.5 to less than 1 GWh per year. The losses of the storage cycle

slightly increase the total energy production, but this is totally

compensated by the large reduction in curtailed energy.

22 The geothermal and waste-to-energy plants can be forced to reduce their energy production because a fossil plant has to run to guarantee the minimum reserve power.


Table 13: Energy and environmental benefits deriving from the installation of a storage facility.

NO STO LiB A LiB B VFB A VFB B

Electrical Energy Supply [GWh/y]

220.5

Thermoelectric Plants [GWh/y]

132.1 119.6 119.6 120.0 119.6

Geothermal + W2E [GWh/y]

56.0 56.8 56.8 56.8 56.8

Wind [GWh/y] 32.3 44.3 44.7 44.4 44.8

Fossil Share 60% 54% 54% 54% 54%

BES Discharged Energy [GWh/y]

4.30 4.60 4.11 4.38

Fuel Consumption [1000 m3]

31.1 28.2 28.3 28.5 28.3

Percentage Reduction 9.4% 9.2% 8.6% 9.0%

Energy Surplus [GWh/y] 13.5 0.7 0.3 0.6 0.2 Total Production [GWh/y]

220.5 221.4 221.4 221.8 221.4

Total Emissions [ktCO2,eq] 104.0 96.6 96.3 96.9 96.5

Percentage Reduction 7.1% 7.4% 6.8% 7.2%

Investment analysis

The profitability of the investment is assessed using NPV and ROI, as

mentioned in Section 3.2.1. The values of the positive and negative cash

flows are reported in Table 14. The flow battery option is favored by the

lower specific capital expenditure, which produces a faster recovery of the

invested capital and a higher NPV after 20 years. This happens in spite of

the worse round-trip efficiency and the higher maintenance costs for VFB.

Under these conditions, all the projects have a discounted payback time

shorter than 10 years (Figure 17 and Figure 18), which means that they

generate positive economic value for more than half of their considered

lifetime and this can be an acceptable condition for the project realization.

In the system under analysis the storage facility with 20 MWh of

capacity produces better financial results for both the technologies,

because most of the capacity remains underused in summertime due to low

wind production. Therefore, the integration of a solar PV plant appears as

a favorable complement in order to improve the economic viability of the

project, owing to its peak production in summer months.


Table 14: Overnight investment, positive cash flows for the first and last year, and operation and maintenance cost in the four analyzed cases.

LiB A LiB B VFB A VFB B

Investment (𝐼𝑁𝑉𝑂𝑉𝑁) – negative cash flow [M€]

6.2 9.3 5.1 7.6

Savings first year – positive cash flow [M€]

1.50 1.56 1.45 1.50

Savings last year – positive cash flow [M€]

1.37 1.44 1.39 1.41

O&M annual cost – negative cash flow [M€]

0.19 0.28 0.25 0.38

In addition to the Net Present Value and the payback time, the Return

on Investment (ROI) can be calculated for each project as reported in Table

15. In any scenario, the values are very high, which is a good indicator of

the economic benefits of the project.

Table 15: Return on Investment after the first year (ROI first year) and Return on Investment as defined in Eq. ( 12 ) (ROI average).

ROI (first year) ROI (average)

LiB A 21.3% 20.2%

LiB B 13.8% 13.2%

VFB A 23.5% 22.9%

VFB B 14.7% 14.1%

OP

TIM

IZA

TIO

N O

F E

LE

CT

RIC

ITY

DIS

PA

TC

H W

ITH

BA

TT

ER

Y S

TO

RA

GE

SY

ST

EM

S (P

AP

ER

S I,

II AN

D III) | 5

9

Figure 17: Net Present Value along the life of a lithium-ion battery system for two values of capacity.

Figure 18: Net Present Value along the life of a vanadium flow battery system for two values of capacity.

-15

-10

-5

0

5

10

15

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

NP

V [

M€]

Time [y]

LiB A

LiB B

-10

-5

0

5

10

15

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20NP

V [

M€]

Time [y]

VFB A

VFB B


Sensitivity analysis

Vanadium flow batteries show better economic performance than

lithium-ion batteries, however, this is strongly influenced by the

assumption that their CAPEX is 20% lower than the corresponding value

for LiB. A dramatic reduction of battery system costs is predicted by several

studies [166–168], but high uncertainty characterizes the extent of this

reduction. Therefore, given that the base-case cost is already quite low, the

base-case CAPEX was doubled in order to cover a larger part of the cost

range proposed by Few et al. [166] for remote systems. The analysis was

performed for both the technologies considering only the 20 MWh capacity

case. In addition, the OPEX was sensibly increased take into account the

uncertainty on this factor, and the results are reported in Table 16. In all

the cases, vanadium flow batteries have higher ratios between the net

present value after 20 years and the capital invested (NPV20/INVOVN)

compared to lithium-ion batteries. However, the impact of a double capital

cost on the profitability is remarkable, with an NPV20 that is approximately

half of the base case. Doubling the OPEX has a slightly lower impact on the

net present value, which for both technologies is visibly lower than the base

case, but remains approximately 1.5 times higher than the initial

investment.

OP

TIM

IZA

TIO

N O

F E

LE

CT

RIC

ITY

DIS

PA

TC

H W

ITH

BA

TT

ER

Y S

TO

RA

GE

SY

ST

EM

S (P

AP

ER

S I,

II AN

D III) | 6

1

Table 16: NPV and discounted payback (DPB) for increased CAPEX and OPEX referred to LiB and VFB (the subscript 0 indicates the base-case conditions; the subscripts 10 and 20 indicate the number of years after which the NPV is calculated).

LiB 20MWh VFB 20MWh

INVOVN [M€]

DPB [y]

NPV10

[M€] NPV20 [M€]

NPV20/ INVOVN

INVOVN [M€]

DPB [y]

NPV10

[M€] NPV20 [M€]

NPV20/ INVOVN

CAPEX0 - OPEX0

6.18 6 4.79 12.54 203% 5.09 5 5.02 12.33 242%

1.5*CAPEX0 - OPEX0

9.27 9 1.70 9.45 102% 7.64 8 2.47 9.78 128%

2*CAPEX0 - OPEX0

12.36 12 -1.39 6.36 51% 10.18 11 -0.08 7.24 71%

CAPEX0 - 1.5*OPEX0

6.18 6 4.00 11.16 180% 5.09 6 3.93 10.44 205%

CAPEX0 - 2*OPEX0

6.18 7 3.21 9.78 158% 5.09 7 2.84 8.54 168%

1.5*CAPEX0 - 1.5*OPEX0

9.27 9 0.91 8.07 87% 7.64 9 1.38 7.89 103%

2*CAPEX0 - 2*OPEX0

12.36 15 -2.97 3.60 29% 10.18 14 -2.25 3.45 34%

TECHNO-ECONOMIC ANALYSIS OF A BIOGAS UPGRADING PROCESS VIA ELECTROLYSIS AND METHANATION (PAPERS IV AND V) | 63

4. Techno-economic analysis of a biogas

upgrading process via electrolysis and

methanation (Papers IV and V)

This chapter presents a process that enables the upgrading of biogas to

synthetic natural gas through its direct injection in an electrolyzer. The

concept is based on a US patent [169]. The information contained in the

patent was used to model the components of the plant and its operating

conditions. A detailed simulation model allowed the energy efficiency of

the process to be calculated, which constituted the basis for the comparison

with other competitive options. The economic evaluation was conducted

through the exergo-economic analysis methodology, supplemented by a

sensitivity analysis and net present value determination for a further

examination of the investment viability of the concept.

4.1. Process description

The following pages contain the description of the biogas upgrading

process via electrolysis and methanation. In addition, a specific

explanation of the passivation effect that sulfur has on the SOEC catalyst is

provided because this is a specific feature of the modeling of co-electrolysis

in presence of methane.

Plant layout

The biogas upgrading system comprises three main sections: reactant

conditioning, electrolysis, and methanation (Figure 19). The modeled plant

is fed with liquid water and biogas. The latter is assumed to have an average

molar composition of 60% CH4 and 40% CO2 on a volume basis. All the

reactants are supplied at 20°C and atmospheric pressure. The feed streams

have to be heated up to be suitable for the operation in the electrolyzer.

This operation can be realized by using external sources of heat (e.g.

electricity or combustion processes) or recovering the heat made available

by other processes in the plant (i.e. cooling of electrolysis products and

64 | TECHNO-ECONOMIC ANALYSIS OF A BIOGAS UPGRADING PROCESS VIA ELECTROLYSIS AND METHANATION (PAPERS IV AND V)

methanation). The electrolysis reactions (see Section 2.1.2) take place in

solid oxide electrolysis cells (SOECs), which represent the component with

the highest consumption of electricity, which is supplied in form of direct

current. This component can operate at different pressures, ranging from

1 to approximately 30 bar. The reactants enter the SOEC with a

temperature of 800°C, while the electrolytic reactions are considered to

take place at 850°C. The temperature increase from 800 to 850°C is

accomplished inside the stack through the Joule effect produced by the

current flow. This fact allows the plant to reach a complete thermal

integration, meaning that all the heating processes are realized through the

utilization of recovered heat. The network of heat exchangers that connects

the hot and cold flows, according to the principles of the Pinch analysis

[170], is depicted in Figure 19. The energy performance of the plant is

simulated for two different pressure levels of the electrolyzer and for two

different values of sulfur coverage of the nickel catalyst (see Section 4.1.2).

TE

CH

NO

-EC

ON

OM

IC A

NA

LY

SIS

OF

A B

IOG

AS

UP

GR

AD

ING

PR

OC

ES

S V

IA E

LE

CT

RO

LY

SIS

AN

D M

ET

HA

NA

TIO

N (P

AP

ER

S IV

AN

D V

) | 65

Figure 19: Plant layout of the electrochemical biogas upgrading process. The cleanup unit is optional and it is present only if a controlled quantity of sulfur is fed to the SOEC stack (Adapted from [171]).


Sulfur effects on SOEC nickel catalyst

The high-temperature electrolysis reactions (Equations ( 4 ) and ( 5 ) –

Section 2.1.2) occur on the solid oxide cell’s cathode, which is usually made

of Yttria-stabilized Zirconia (YSZ) with nickel particles dispersed on top,

acting as catalyst for the electrolytic conversion.

When the stack of cells is fed with a gaseous mixture containing

significant quantities of methane, the nickel promotes the reaction of

steam reforming ( 13 ). This reaction is exothermic and it is

thermodynamically favored at low pressure and high temperature. Hence,

the reduction of the methane content in the inlet gas is inversely

proportional to the operating pressure of the stack, creating also a non-

uniform temperature distribution, if the produced heat is not properly

managed through an effective cooling system.

CH4 + H2O → 3H2 + CO ( 13 )

Such methane depletion in the biogas could be inhibited by co-feeding

a controlled amount of H2S to the inlet stream, since sulfur covers the

active sites of the catalyst and reduces the rates of the surface reactions

[172]. Biogas usually contains sulfur in the form of H2S, which is the

product of the reactions involving the organic sulfur compounds present in

the feedstock to the anaerobic digester. Hydrogen sulfide should be

completely removed from biogas in a cleanup unit [173] to avoid

combustion problems; however, in this case a small quantity could be kept

to inhibit steam reforming. This phenomenon is called sulfur passivation

effect and it can be simulated by multiplying the reaction rate by a

correction coefficient of (1 − 𝜃𝑠 )3, where 𝜃𝑠 is the sulfur coverage23 [174].

The catalyst active sites for heterogeneous catalysis that favor the steam

reforming reaction were found to be separate from those promoting the

electro-catalytic reactions [175], thus it seems that sulfur prefers to interact

with those of the first type. Therefore, the presence of sulfur inhibits the

decomposition of methane more than the electrochemical conversion of

water and carbon dioxide. However, the electrochemical reactions are

partially penalized by the reduction in the number of the active sites, the

effect of which can be modeled as an increase in cell resistance.

23 The coverage expresses the ratio between the number of covered active sites and the total number of active sites for a catalyst.


Although H2S can have a beneficial effect for the electrolyzer, it is surely

detrimental in the methanation reaction, since it inhibits the methane

formation reaction (the reverse of steam reforming). Therefore, an

additional cleanup unit (also called sulfur polisher) has to be placed

between the two sections to ensure that the gas entering the series of

methanation reactors is sulfur-free.

4.2. Energy and exergo-economic analyses

The determination of the energy performance (mass and energy

balances) is done through the implementation of the plant model in the

chemical engineering software Aspen Plus [176], that allows one to

quantify the mass and energy streams involved. The techno-economic

evaluation is based on the functional model and it is developed with the

methodology of the exergo-economic analysis, a brief description of which

is reported in the next sections and in Appendix A. This methodology

integrates an exergetic evaluation with an economic analysis of the process,

where the economic value is linked with the exergy content of each stream.

In other words, the economic value assigned to all the mass and energy

streams is based on the amount of exergy and capital that is needed to

produce each specific stream. Therefore, this approach provides a way to

distribute the production cost among the different outputs of a multi-

product process, as the one under evaluation. Such analysis supplies

information that traditional economic analysis based on NPV or ROI are

not able to capture, because they assess the profitability of a project for

different market conditions. Instead, they fail to provide a complete

indication of the techno-economic modifications that can be implemented

to improve the profitability of the process.

Energy and exergy analyses

The energy analysis includes the calculation of all the enthalpy values

associated to the mass streams and the quantification of the electrical and

thermal needs of the different components of the process, in order to define

the efficiency of the process. This step is complemented by the

computation of the physical (Equation ( 14 )) and chemical exergy

(Equation ( 15 )) of all the mass and energy streams of the plant (Figure

19). The calculations are done following the approach described in

Gandiglio et al. [177]. The dead state condition (that is the zero exergy


state) are T0 = 298.15 K, p0 = 1.0 atm, and the environment composition is:

N2 75.67%, O2 20.35%, H2O 3.12%, CO2 0.03%, other 0.83% [178].

𝑏𝑝ℎ − 𝑏𝑝ℎ0 = (ℎ − ℎ0) − 𝑇0 ∙ (𝑠 − 𝑠0) ( 14 )

𝑏𝑐ℎ = ∑ 𝑦𝑖 ∙ 𝑏𝑐ℎ,𝑖

𝑖

+ 𝑅𝑇0 ∑ 𝑦𝑖 ∙ ln (𝑦𝑖)

𝑖

( 15 )

𝑏𝑝ℎ is the physical exergy, 𝑏𝑐ℎ is the chemical exergy, ℎ is the enthalpy (all

expressed in J kg-1) and 𝑠 is the entropy (in kJ kg-1 K-1), 𝑦𝑖 is the molar

fraction of the ith component of the mixture and 𝑅 is the universal gas

constant (8.31 J mol-1 K-1).

Moreover, a resource and a product are defined for each unit of the process

(see Appendix A), the behavior of which is described with particular

performance indices, which are:

· Exergy efficiency (𝜂𝑒𝑥,𝑖) – ratio between the product and the resource

of the ith component.

· Exergetic factor (𝑓𝑖 ) – ratio between the resource utilized by the ith

component and the total input exergy given to the plant as the external

resource.

Assumptions for the economic analysis

The economic evaluation is influenced by the different specific cost

between an SOEC operating at high pressure and one operating at low

pressure. The value of 4,000 €/m2 was considered for the SOEC stack

operating at low pressure, while 6,000 €/m2 when it operates at high

pressure. These estimations are similar to those proposed by Mogensen

[179] and slightly higher than those found in Giglio et al. [180]. When a

controlled quantity of sulfur is fed to the electrolyzer, a high-temperature

polisher filled with ZnO is used to yield a sulfur-free syngas to the series of

methanators. The cost of this component is related to the cathodic outflow

of the cells [36]. The methanation reactors are considered to have a cost of

400 €/kWSNG [63].

The costs of pressure changers, heat exchangers and vessels are derived

from Turton et al. [34] and they are expressed in 2001 USD. The values are


updated to 2014 USD via the chemical engineering plant cost index (CEPCI

- equal to 576.1 in 2014 and to 397 in 2001). The results are converted into

Euros through a fixed exchange rate equal to 1.1 $/€. In addition to the

purchase cost of the physical components (bare erected cost – BEC), the

cost estimation incorporates also all the sub-components that are

indispensable for the connection and the control of the various units (such

as piping and instrumentation), as well as the costs for the design and

construction of the plant. Their sum represents the Total Plant Cost (TPC).

The conversion to TASC (Total As Spent Cost) is done by applying the

correction factors described by Giglio et al. [180] 24 , according to the

methodology developed by the National Energy Technology Laboratory

(NETL) for energy conversion plants [181]. On top of this, the operation

and maintenance costs are added, in this case they are estimated as 4% of

the total plant cost.

Exergo-economic analysis

The exergo-economic analysis requires the computation of the exergo-

economic costs of the components (Z) and those of the inlet mass and

energy flows. These quantities are useful to define two indices of

performance that contribute to detail the behavior of the plant. In

particular:

· The cost of exergy destruction (CD), defined as the fuel cost

entering a component, multiplied by the rate of exergy destruction

of the same unit

· The exergo-economic factor (𝑓𝑒𝑥𝑒𝑐), defined as the ratio between

the exergo-economic cost and the sum of exergo-economic and

exergy destruction costs of each component (𝑍

𝑍+𝐶𝐷)

The exergo-economic factor determines to which extent the presence of a

component is responsible for the stream-cost increase. If 𝑓𝑒𝑥𝑒𝑐 is close to 1,

the majority of the cost increase of a stream is due to the equipment cost;

otherwise, the cost of exergy destruction is predominant.

24 The TPC is 30% higher than the bare erected cost (BEC) and the TASC is 30% higher than TPC.


4.3. Energy analysis results (Paper IV)

The upgrading process, the layout of which is shown in Figure 19, was

simulated in four different operating conditions (Table 17), combining two

pressure levels of the electrolyzer and two sulfur coverage values. For all

the cases, the inlet and the outlet energy and mass flows were computed,

using as design variable the current absorbed by the electrolyzer (10,000

A). The energy efficiency of the plant (𝜂𝑝𝑙 ) was obtained as defined in

equation ( 16 ).

𝜂𝑝𝑙 =𝐿𝐻𝑉𝑆𝑁𝐺 ∙ �̇�𝑆𝑁𝐺𝑜𝑢𝑡

𝐿𝐻𝑉𝐶𝐻4�̇�𝐶𝐻4𝑖𝑛

+ 𝑊𝑒𝑙,𝑝𝑙

( 16 )

Table 17: Main distinctions among the four cases analyzed in the study.

Case A Case B Case C Case D

SOEC Pressure Low

(ca 1 atm) Low

(ca 1 atm) High

(ca 30 bar) High

(ca 30 bar)

Sulfur Coverage (𝜃𝑠)

0% 95% 0% 95%

In all the four cases the efficiency is quite high (above 70%), however,

Case A is characterized by a much lower value compared to the others

(Table 18). This is due to the higher electrical consumption in the pre-

methanation compressor and in the SOEC stack. The compression of the

gaseous stream entering the series of methanation reactors is more energy-

intensive than the pressurization of the liquid water before it enters the

SOEC. Moreover, the endothermic reaction of steam reforming is favored

by the thermodynamic conditions present in Case A and this entails that

all the inlet methane is converted into carbon monoxide and hydrogen,

with an additional thermal need to maintain the temperature constantly at

850 ºC. Therefore, a larger part of the electrical energy is used to keep the

temperature constant and is not used for the electrolysis. This fact is

reflected by the lower oxygen production rate (Table 18).


Table 18:Global mass and energy flows for the plant in Figure 19.

Quantity Unit A B C D

Total AC electric need [kW] 15.5 15.9 14.2 14.3

Pumping Power [kW] 2.2 2.2 0.59 0.60

Stack Power [kW] 13.2 13.6 13.7 13.7

Upgraded Gas (SNG) [kg/h] 1.39 1.82 1.81 1.89

Methane inlet [kg/h] 0.82 1.08 1.07 1.12

Oxygen outlet [kg/h] 2.17 2.87 2.84 2.97

LHVSNG [MJ/kg] 49.2 49.4 49.4 49.4

SNG chemical energy flow

[kW] 19.0 25.0 24.8 25.9

Plant efficiency (ηpl) [%] 71 81 86 87

Comparison with other biogas upgrading routes

The process that has been described so far represents a tentative one for

coupling biogas upgrading with an efficient use of low-carbon and low-cost

electricity. However, there are other possibilities to remove CO2 from

biogas in order to improve its quality. The principal options are: water

scrubbing (WS), pressure swing adsorption (PSA) and membrane

separation (MS); the description and the specific energy consumption of

which can be found in Pertl et al. [182]. The efficiency and the specific

productivity (SP) of each process can be defined as in Equations ( 17 ) and

( 18 ).

𝜂𝑠𝑒𝑝 =𝐿𝐻𝑉𝑆𝑁𝐺 ∙ �̇�𝑆𝑁𝐺,𝑜𝑢𝑡

(𝐿𝐻𝑉𝐶𝐻4 ∙ �̇�𝐶𝐻4𝑖𝑛 + 𝑊𝑒𝑙 + 𝐸𝑒𝑥𝑡𝑟𝑎)

( 17 )

𝑆𝑃 = 𝑌𝐶𝐻4(

100 − 𝐿𝐶𝐻4

100) ( 18 )

where 𝐸𝑒𝑥𝑡𝑟𝑎 includes the additional energy flow required to obtain a grid-

suitable SNG, YCH4 [Nm3SNG/Nm3

RG] is the methane molar fraction in the

biogas and 𝐿𝐶𝐻4 is the percentage of methane lost during the process. For

instance, PSA produces an SNG that needs to be mixed with propane in

order to be suitable for grid injection and the propane consumption has


been estimated to be 0.013 [Nm3/Nm3RG], while the electrical energy

requirement is 0.053 [MJ/Nm3RG] [182].

The results for each technology are summarized in Table 19, assuming

that the outlet SNG has a methane content of 97% in volume25.

Table 19: Efficiencies and specific productivities for Water Scrubbing, Pressure Swing Adsorption and Membrane Separation (adapted from [183]).

PSA Efficiency (𝜂𝑠𝑒𝑝) 92.9 [%]

Specific productivity 0.576 [Nm3SNG/Nm3

RG]

WS Efficiency (𝜂𝑠𝑒𝑝) 94.2 [%]


RG]

MS Efficiency (𝜂𝑠𝑒𝑝) 86.6 [%]


RG]

The electrochemical upgrading process that has been analyzed in this

study is a one-step process, because the biogas is fed directly to the SOEC

and the product gas of the electrolysis undergoes methanation. The sealed

structure of the SOEC and of the methanation reactors drastically reduces

the carbon dioxide slip from the plant, which was neglected. Therefore, the

carbon dioxide present in the biogas is completely transformed into

methane throughout the SOEC and the methanation section. Moreover, CO

is almost absent after the water removal step and this implies that the

conversion of CO2 into CO, due to the water gas shift reaction, is negligible

in the methanation process.

However, in order for the one-step process to be comparable with the

other commercial options, a two-step process is considered. Namely, CO2

must be firstly removed from the digester gas and then a mixed H2O/CO2

stream is fed to the integrated SOEC-methanation plant (with an efficiency

– 𝜂𝐶𝑜𝐸 – of 81.4%26). A scheme of the two-step process is depicted in Figure

20. The two-step process efficiency (𝜂𝑡𝑠) is defined in Equation ( 19 ) and

this quantity is affected by the efficiency of the CO2 separation technique

(𝜂𝑠𝑒𝑝).

25 The productivity values refer to a molar composition of biogas equal to 60% CH4 and 40% CO2 vol. The lower heating value of methane is assumed equal to 35 MJ/Nm3 for the efficiency calculations. 26 The H2O/CO2 integrated SOEC-methanation plant was extensively studied in Giglio et al. [180], on the basis of a previous study [206].


𝜂𝑡𝑠 =(𝐿𝐻𝑉𝑆𝑁𝐺 ∙ �̇�𝑆𝑁𝐺,𝑜𝑢𝑡)𝑠𝑒𝑝 + (𝐿𝐻𝑉𝑆𝑁𝐺 ∙ �̇�𝑆𝑁𝐺,𝑜𝑢𝑡)𝑒𝑙𝑒𝑐𝑡

(𝐿𝐻𝑉𝐶𝐻4 ∙ 𝑚𝐶𝐻4𝑖𝑛 + 𝑊𝑒𝑙 + 𝐸𝑒𝑥𝑡𝑟𝑎)𝑠𝑒𝑝 +(𝐿𝐻𝑉𝑆𝑁𝐺 ∙ �̇�𝑆𝑁𝐺,𝑜𝑢𝑡)𝑒𝑙𝑒𝑐𝑡

𝜂𝐶𝑜𝐸

( 19 )

The efficiencies are clearly higher for the conventional methodologies

(Table 20). However, these methodologies do not recycle CO2, which is the

reason why they are characterized by much lower SNG yields. The coupling

with the electrolytic upgrade increases considerably the specific

productivity values, with a lower CO2 release. The atmospheric one-step

process has a lower energy efficiency than the two-step routes, but entails

a reduced plant complexity, since the CO2 flow is directly fed to the SOECs.

CO-ELECTROLYSIS +

METHANATION

BIOGAS:

0.6 CH4

0.4 CO2

0.4 CO2

0.6 CH4

0.4 CH4

H2O

CO2 SEPARATION

H2O

Wel1

Wel2

CONTROL VOLUME

Figure 20: Scheme of the two-step upgrading process with the indication of the mass and energy flows (adapted from [183])

The comparison between the different techniques shows that SNG

production from biogas can become a carbon negative process, able to

convert all the carbon present in organic waste into usable fuel.


Table 20: Summary of the efficiencies and specific productivities for all the digester gas upgrading routes (adapted from [183]).

Process Action(s) Efficiency Specific

productivity

[%] [Nm3

SNG/ Nm3

RG] Pressure Swing Adsorption (PSA)

CO2 separation 92.9 0.576

Water Scrubbing (WS)


Membrane Separation (MS)


PSA + H2O/CO2 CoE (Two steps)

CO2 separation + H2O/CO2 co-electrolysis

88.0 0.976

WS + H2O/CO2 CoE (Two steps)


90.7 0.991

MS + H2O/CO2 CoE (Two steps)


84.4 0.964

Digester gas CoE (One step) – atmospheric

H2O/CO2 co-electrolysis

71.0-81.0 1.0

Digester gas CoE (One step) – pressurized

H2O/CO2 co-electrolysis

86.0-87.0 1.0


4.4. Exergo-economic analysis results (Paper V)

The energy performance analysis oriented the subsequent economic

assessment, inasmuch as only the configurations that appeared more

relevant were further considered. Therefore, the Cases A and D described

in the previous sections were disregarded either because the efficiency was

much lower than the other configurations (Case A) or because the

efficiency improvement due to the steam reforming inhibition performed

by H2S was not high enough to justify the presence of a sulfur cleanup unit

downstream from the SOEC (Case D). Secondly, in order to reduce the

specific cost of the plant components (i.e. compressor, pumps, heat

exchangers and reactors), the plant size had to be increased. Therefore, the

direct current (DC) power absorbed by the electrolyzer that was used as

size variable was assumed equal to 5 MW. In the following sections the

results of the three methodological steps of the exergo-economic analysis

(see Appendix A), namely exergetic, economic and integrated exergo-

economic analyses are reported [177].

Exergetic results

The exergy efficiency of the plant in the two analyzed configurations

(Cases B and C) is calculated according to Equation ( 20 ) and it is equal to

83 and 87%, respectively.

𝜂𝑒𝑥,𝑝𝑙 =�̇�𝑆𝑁𝐺

�̇�𝑏𝑖𝑜𝑔𝑎𝑠 + 𝑊𝑒𝑙

( 20 )

The components that treat a large quantity of exergy (i.e. with a high

exergetic factor - fi27) have exergetic efficiencies greater than 90%, which

means that these components have very good performance and the margin

to improve the exergy management is very limited (Table 21).

The lowest exergy efficiency belongs to the low-temperature heat

exchangers, since there is a quite large temperature difference between the

two fluids involved in each heat exchanger. This is due to the large

availability of excess heat from the overall process, which may be used to

restrain the heat exchange area and, in turn, the equipment cost. The

exergy lost in these heat exchangers could be reduced, but it would imply a

27 See Section 4.2.1 for its definition.


higher investment cost for the plant, without a real benefit in terms of

efficiency, since the thermal needs of the plant are fully satisfied.

Table 21: Exergetic indices of performance for the main components of the low-pressure (LP) and high-pressure (HP) configurations ordered by decreasing exergetic factor (fi).

LP configuration

Component

Exergy

efficiency,

ηex,i

Rate of exergy

destroyed [kW]

Exergetic

factor, fi

Methanator 1 >99% 115.0 210%*

Cleanup unit 99% 56.8 89%

Methanator 2 >99% 45.9 87%

Drum 99% 86.5 84%

SOEC 90% 504.3 42%

Pre-methanation

Compressor 53% 375.8 7%

Oxygen Flow 0% 123.5 1%

Biogas Compressor 76% 18.1 1%

HP configuration

Methanator 1 99% 173 192%*

Methanator 2 >99% 46 90%

Drum >99% 42 87%

SOEC 91% 445 46%

Biogas Compressor 72% 52 1.7%

Oxygen Flow 0% 154 1%

Pre-methanation

Compressor 76% 5 ~0%

* 𝑓𝑖 values higher than 100% appear because some components are placed in a

recirculation loop in the methanation section (Figure 19), where the total exergy entering the

component is higher than the exergy entering the whole plant

The highest rates of exergy destruction are found in the SOEC, Pre-

methanation Compressor, Methanator 1 and Methanator 2, although some

of them have high efficiencies (i.e. SOEC and Methanator 1). However, they

process a large quantity of exergy and, therefore, they substantially impact

the exergy balance. Despite this, the satisfactory energy performance

makes them very expensive to improve and, thus, they are considered

optimized.


Moreover, the oxygen flow is responsible for the loss of a good quantity

of exergy if it is completely vented. Instead, it may be further exploited

because it has high purity and, if the SOEC is pressurized, it is available at

about 30 bar. In particular, in the latter case it represents 10% of the total

dissipated exergy. This, together with its relatively high pressure, makes it

more appealing for a further economic use than the product of the low-

pressure electrolysis, because the conditioning treatment to make it

suitable for sale quality is more expensive.

Finally, lower consumption of the auxiliary components, higher overall

exergetic efficiency and better thermodynamic quality of the outlet oxygen

stream suggest that the high-pressure configuration should be preferred

over the low-pressure one.

Economic results

Following the assumptions and the methodology described in Section

4.2.2, the total costs of the main components of the plant 28 have been

estimated and they are reported in Table 22. The equipment cost

contributes significantly to the cost of the products of the plant (see Section

4.4.3).

Another item influencing the economic performance of the plant is the

cost of the utilities, i.e. electricity, water and biogas. The electricity price

includes only the distribution and transmission cost (calculated according

to the parameters provided by the Italian TSO [184]). The plant size is quite

large for this kind of application and it requires a substantial biogas feed,

which is unlikely to be provided by the anaerobic digestion of agricultural

sources since these are usually small-to-medium scale plants (typically

below 500 - 600 Sm3/h of biogas production29). Instead, the biogas is

assumed to be available from a large wastewater treatment plant or a

landfill, with a concentration of H2S of 500 ppm. This value is

representative of biogas streams with considerable sulfur loads, while the

majority of the plants shows values around 100 ppm [173]. The biogas

produced in such plants can even be considered to have a zero cost, since

it comes as a by-product of the water sanitation process (or waste

28 The cost of the heat exchangers is not included, but it is available in [171] as well as the procedure for the estimation of the cleanup cost. 29 A biogas plant for electricity production with a rated electric power of 1 MW and an average electrical efficiency of 30% requires a biogas feed of 600 Sm3/h. Such size is usually considered medium to large scale [207].


management, in the case of landfills) [185]. Nonetheless, a cost is

associated to the inlet biogas feed since it has to be collected, stored,

pipelined across the plant and finally treated to remove contaminants. The

components of the biogas cost are, thence, ascribable to the production and

management of raw biogas and to contaminant cleanup. The cost of raw

biogas per standard cubic meter (at 15°C and 1 atm) is 0.09 €/Sm3 [39],

while the cost for pre-upgrading cleanup is calculated according to the

assumption made by Papadias et al. [186], thus resulting in a cost of 0.04

€/Sm3. The total biogas cost is therefore 0.13 €/Sm3.

Table 22: Total cost of the main components for the low- and high-pressure configurations. (adapted from [171]).

Components (Low Pressure)

Total Cost [k€]

% of total

Components (High Pressure)

Total Cost [k€]

% of total

Ref.

Biogas Compressor

155 1.8 Biogas Compressor

731 11.1 [187]

SOEC 2881* 37.1 SOEC 3058* 46.2 [188]

Cleanup unit 134 1.6 [186]

Pre-methanation Compressor

2375 27.8 Pre-methanation Compressor

56 0.8 [187]

Recirculation Compressor

14 0.2 Recirculation Compressor

6 0.1 [187]

Methanator 1 1213 14.2 Methanator 1 995 15.0 [63]



Total 8244 Total 6613

* includes the replacement of the stack after 10 years

The assumptions for the economic assessment and the specific costs of

the utilities are summarized in Table 23. A high discount rate (i.e. 20%) is

considered because of the risk premium that this kind of plant bears, due

to the limited commercial availability.


Table 23: Economic and financial parameters, and costs related to the utilities (adapted from [171]).

Economic and financial parameters

Utilities

Capacity factor, CF 70% Biogas price [€/Sm3] 0.13

O&M, fraction of the TPC 4% Electricity price [c€/kWh]

1.1

Capital recovery factor, CRF 0.21 Biogas price [c€/kWh] 2.54

Discount Rate, d 20% Demineralized Water [€/m3]

2.1

Plant Lifetime [years], NY 20 SOEC module lifetime [years] 10

Exergo-economic results

The value of the exergo-economic factor discriminates if the cost

increase is mainly attributable to the equipment cost or to the low

efficiency of the component. For the electrolyzer this value is very close to

1, which means that the component cost has a high impact on the cost of

the outlet streams. Pressure changers have relatively high exergo-

economic factors, because of their small-to-medium size, compared to

those used in large chemical plants (Table 24). On the other hand, a further

increase in plant size would be problematic since it would not favor the

modularity and the possibility to exploit the biogas produced in smaller

plants.


Table 24: Exergo-economic indices of performance for the main components of the low-pressure (LP) and high-pressure (HP) configurations (adapted from [171]).

LP configuration

Component

Cost of destroyed Exergy, CD [€/h]

Exergo-economic equipment cost, Z [€/h]

Exergo-economic factor, fexec

Methanator 1 5.69 53.52 90% Cleanup unit 2.08 5.40 72% Methanator 2 2.39 13.27 85% Drum 4.82 0.43 8% SOEC 5.55 128.31 96% Pre-methanation Compressor

4.13 105.79 96%

Biogas Compressor 0.20 6.91 97%

HP configuration

Methanator 1 6.94 45.10 87% Methanator 2 1.97 18.57 90% Drum 2.03 0.43 17% SOEC 4.90 136.18 97% Biogas Compressor 0.57 32.55 98% Pre-methanation Compressor

0.06 2.47 98%

The difference between the exergo-economic costs of SNG in the two

configurations (5.62 vs. 4.87 c€/kWh for LP and HP respectively) is

created mainly downstream from the SOEC, where the influence of the

compression of the cathodic gas, in terms of energy and capital

requirement is substantially higher in the LP case. The cleanup unit also

contributes to increase the exergo-economic cost of the final product.

However, the latter component is strictly necessary to avoid problems in

the operation of the subsequent methanation section.

The exergo-economic cost of SNG in both cases is broken down into the

different component in Table 25.


Table 25: Breakdown of the SNG thermo-economic cost for the two configurations (adapted from [171]).

Low-

pressure High-

pressure

Exergo-economic cost [c€/kWhexergy] 5.62 4.87

Equipment 64.1% 60.0%

Utilities 35.9% 40.0%

- Electricity 11.6% 12.2%

- Biogas 23.8% 27.2%

- Water 0.5% 0.6%

Energy cost [c€/kWhLHV] 5.77 5.04

Energy cost [c€/kWhHHV] 5.19 4.53

The equipment cost occupies the predominant share. In the LP case, it

is higher both in relative and in absolute terms, meaning that the negative

effect that sulfur has on the electrochemical performance, leading to a

larger active area of SOEC ( 𝐴𝐴𝑇𝑀~1.4 ∗ 𝐴𝑃𝑅𝐸𝑆 ) for the same electrical

consumption, is not counterbalanced by the lower cost of the equipment

operating at nearly atmospheric pressure. Therefore, with the assumptions

adopted in this study the pressurized configuration dominates the

atmospheric one both in terms of energy efficiency and in terms of

economic convenience.

The biogas cost represents approximately 25% of the total cost, while

electricity slightly more than 10%, since the specific cost is very low (1.1

€/kWh).

The specific cost of SNG reported in Table 25 can be compared with the

cost estimated for biogas upgrading through water scrubbing. This value

has been found to fall in the interval 2.8-5.8 c€/kWhLHV for a cost of biogas

up to 0.17 €/Sm3 [115]. Therefore, with a biogas cost equal to 0.13 €/Sm3

the resulting SNG cost would be 4.89 c€/kWhLHV. This estimate considers

a higher cost of electricity compared to the present analysis (18.7 vs. 1.1

c€/kWh), however, for water scrubbing the electricity cost represents 14%

of the SNG cost. Therefore, if the scrubbing process used low-cost

electricity, the SNG would result in 4.2 c€/kWhLHV, which is 15% lower

than the cost of the electrolytic upgrading.


Economic valorization of the by-products

In addition to synthetic natural gas, the plant produces high-purity

oxygen, an electrolysis by-product with a good market potential. The main

uses of oxygen are: steel industry, healthcare and gasification processes.

The most common way of producing oxygen is through cryogenic air

separation, which is generally energy-intensive and thereby the price of

electricity has a considerable impact on the oxygen production cost. The

energy consumption for large oxygen plants is reported to be around 300

– 325 kWh/tO2 and the final price can span between 20 and 50 $/t,

according to the size of the plant [189].

In the present case, the oxygen coming from the SOECs is cooled down

to preheat the mixture of biogas and water and then it is compressed up to

150 bar. The specific cost of the oxygen treatment components is carried

out considering the formulas defined in Turton et al. [187]. In the low-

pressure case, the higher pressure increase influences both the quantity of

electrical energy that is needed, and the sizing of the oxygen compressor,

resulting in a much higher cost of the oxygen. The cost per unit mass of the

oxygen is 54.6 €/tO2 for the LP configuration and 15.2 €/tO2 in the HP one.

Assuming a sale price of 50 $/tO2 (45.5 €/tO2), only the HP case would be

profitable, with an hourly gain of 32 €/h (0.89 c€/s) corresponding to 6.7%

reduction on the SNG thermo-economic cost. The latter would decrease to

4.54 c€/kWhexergy. However, these estimates do not consider the cost for

bottling and transportation, hence, it would be much better if the oxygen

was used onsite.

The electrochemical biogas upgrading prevents the CO2 from being

released, because it is recycled into a fuel. For this reason, the saved CO2

could be accounted as an avoided emission and, it might be eligible to be

traded in the carbon market. Three possible scenarios for the carbon price

are considered to calculate the reduction of exergo-economic cost of SNG

[190]. The results are summarized in Table 26.

The maximum saving per unit of exergy is around 8% of the SNG

exergo-economic cost. However, it is likely that the impact of cost

reduction will be lower, due to the several political and economic obstacles

which can prevent the market from reaching high CO2 prices in the next

decades. Despite this, the participation to the carbon market could be an

option to encourage the recycle of carbon dioxide in digestion processes,

which are already considered carbon neutral.


Table 26: Economic gains for avoided emissions in different scenarios of CO2 price for the period 2022-2050 (adapted from [171]).

CO2 Price

[€/ton] CO2 Gain [c€/s]

CO2 Gain

[c€/kWhexergy SNG]

LP HP LP HP

Low Price 23 0.45 0.49 0.17 0.17

Medium

Price 38 0.75 0.81 0.28 0.29

High Price 55 1.09 1.17 0.41 0.41

If the exergo-economic cost of the biomethane obtained with the

electrochemical upgrading process in the high-pressure case (Table 25) is

referred to its higher heating value, this results in a value of 4.53

c€/kWhHHV. This value is higher than the price of natural gas for medium-

sized consumers 30 both in 2013 (~4.1 c€/kWhHHV) and in 2015 (~3.5

c€/kWhHHV). However, the inclusion of oxygen sale and CO2 allowances

can bring the production cost closer to the natural gas price. The SNG

exergo-economic cost could be lowered up to 15%. This is equivalent to

3.84 c€/kWhHHV, which is still higher than the 2015 natural gas price, but

lower than the 2013 one (Figure 21).

30 Average EU-28 price for medium-sized industrial consumers (100 000 GJ < Consumption < 1 000 000 GJ) [208]. In 2013 the natural gas price reached its maximum and for this reason this year has been taken as reference.


Figure 21: SNG unit exergo-economic cost reduction for different options of by-product valorization with the indication of the percentage saving. The black markers illustrate the results in case of combined oxygen sale and CO2 credits. The horizontal lines report the price of natural gas for medium-size industrial consumers (average EU28) per unit of exergy. Note that the scale is on the y axis is partial (adapted from [171]).

Sensitivity analysis

Electricity has a fundamental role in power-to-gas applications and,

since its price affects the profitability of the whole process, a sensitivity

analysis is warranted. The minimum price has been considered zero, while

the maximum price is assumed to be equal to 4 c€/kWh, which is in the

range of the LCOE of capital intensive technologies such as coal and

nuclear [191]. The taxes and levies are excluded, since the electricity used

for power-to-gas applications could be exempted from taxation. Between

the electricity price and the exergo-economic cost of the synthetic natural

gas there is a linear relation with an increase of 0.59 c€/kWhSNG (LP) and

0.54 c€/kWhSNG (HP) per c€/kWhel, as shown in Figure 22. The percentage

variation of the SNG unit exergo-economic cost is steeper for the

pressurized case since the weight of the electricity cost on the total cost is

higher.

3.5

3.6

3.7

3.8

3.9

4.0

4.1

4.2

4.3

4.4

4.5

4.6

4.7

4.8

4.9

SNG cost O2 sale CO2 low CO2 medium CO2 high

2013 NG price

2015 NG price

6.7%3.5%

5.8%8.5%

SN

G u

nit

ex

erg

o-e

co

no

mic

co

st

[c€

/kW

h]


Figure 22: Sensitivity analysis of the SNG exergo-economic cost by varying the electricity price. The red marker indicates the base case. The labels indicate the costs expressed in c€/kWhexergy for the low- and high-pressure conditions, respectively. Adapted from [171].

Another important parameter which has a large impact on the

evaluation of the profitability of the plant is the capacity factor, that is the

equivalent number of hours of plant operation at full power in a year.

Theoretically, the plant should be operating when the electricity is

available at a low price. If power-to-gas was used exclusively to reduce the

curtailment of solar energy, the storage plant would work 26% of the time

at full load and 7% at partial load [192]. Varone and Ferrari [24] simulates

a situation in which the installed wind power in Germany is increased by 8

times and the electricity demand is reduced to 89% of the 2012 value. In

this scenario the number of surplus hours would be 53% of the total. These

two facts combined suggest that it is not easy to reach high capacity factor

values. Therefore, a capacity factor of 60 - 70% can be assumed as a target

for power-to-gas plants to achieve. The results of the sensitivity for

different capacity factors are presented in Figure 23. In this case the

slightly steeper variation occurs for the atmospheric case, since less

working hours mean less efficient use of the machinery, the cost of which

is predominant compared to the cost of utilities.

-20%

-10%

0%

10%

20%

30%

40%

0 0.01 0.02 0.03 0.04

Δ%

SN

G c

ost

Electricity price [€/kWh]

HP

LP

4.97 – 4.27

5.62 – 4.87

7.33 – 6.42

6.74 – 5.88

6.15 – 5.34


Figure 23: Sensitivity analysis of the SNG exergo-economic cost for different capacity factors. The red marker indicates the base case. The labels indicate the costs expressed in c€/kWhexergy for the low- and high-pressure conditions, respectively. Adapted from [171].

Since the project was considered very risky and with a high operational

uncertainty, a high discount rate was used for the evaluation of the project.

However, generally for energy project evaluation much lower values are

applied (3-5%). For this reason, a sensitivity analysis on this parameter was

done to estimate the effect of the investment risk on the exergo-economic

cost in the two cases. A considerable reduction is observable in Figure 24,

which causes the exergo-economic cost to fall below the 2015 price of

natural gas (3.81 €/kWhexergy) used in Section 4.4.4.

-20%

-10%

0%

10%

20%

30%

0.5 0.6 0.7 0.8 0.9

Δ%

SN

G c

os

t

Capacity Factor

HP

LP

6.21 – 4.87

7.04 – 6.04

5.17 – 4.51 4.99 – 4.36

4.83 – 4.22

5.62 – 4.87


Figure 24: Sensitivity analysis of the SNG exergo-economic cost for different discount rates. The red marker indicates the base case. The labels indicate the costs expressed in c€/kWhexergy for the low- and high-pressure conditions, respectively.

-50%

-40%

-30%

-20%

-10%

0%

0 5 10 15 20

Δ%

SN

G c

os

t

Discount rate [%]

HP

LP

3.43 – 3.13

3.64 – 3.30 4.24 – 3.77

4.90 – 4.30

5.62 – 4.87


4.5. Economic profitability with net present value

The methodology and the results illustrated in the previous sections are

focused on the determination of the production cost of SNG using exergy

as allocation criterion. However, an economic analysis using more

traditional methodologies can be beneficial to get an idea of the

profitability that such an investment could have under different

technological and market conditions. Therefore, the net present value

(NPV) of the biogas upgrading plant previously described in this chapter is

calculated, using the economic and financial parameters reported in

Section 4.2.2. Its formulation is given in Equation ( 21 ).

The positive annual cash flow is the revenue deriving from the sale of

SNG ( 𝑅𝑒𝑣𝑆𝑁𝐺 ), while the negative annual cash flows are the costs of

electricity, biogas production, demineralized water (grouped into

feedstock costs 𝐶𝑓𝑒𝑒𝑑𝑠𝑡𝑜𝑐𝑘 ) and operation and maintenance ( 𝐶𝑂&𝑀 ). The

overnight investment (𝐼𝑁𝑉𝑂𝑉𝑁) is partially discounted because the SOEC

pack is supposed to be replaced after 10 years (Table 23). The price of SNG

was set at 80 €/MWh which is a value in line with the incentive package

recently approved by the Italian government [193]. The results for the low-

and high-pressure configurations are shown in Figure 25, using a discount

rate of 20% like the one considered for the exergo-economic analysis.

𝑁𝑃𝑉 = ∑ (𝑅𝑒𝑣𝑆𝑁𝐺 − 𝐶𝑓𝑒𝑒𝑑𝑠𝑡𝑜𝑐𝑘 − 𝐶𝑂&𝑀

(1 + 𝑑)𝑦 )

𝑁𝑌

𝑦=1

− ∑ (𝐼𝑁𝑉𝑂𝑉𝑁

(1 + 𝑑)𝑦)

𝑁𝑌

𝑦=0

( 21 )


Figure 25: NPV comparison between the high- and low-pressure configurations of the plant for biogas upgrade.

However, the discount rate commonly used for the economic and

financial analysis of energy projects is much lower than the value of 20%

used in Section 4.4.5. For this reason, from now on a discount rate of 5%

will be considered. Some other factors that influence the profitability of the

investment are the cost of electricity, the specific cost of the electrolyzer

and the price that the SNG producer is going to be paid. The results are

shown only for the pressurized configuration since it is the most profitable

one (Figure 26). The baseline assumptions are: 30 €/MWh for the

electricity cost, 60 €/MWh as SNG price31, 6 k€/m2 of active area for SOEC

(CAPEX)32. The cost of electricity is slightly higher than that assumed so

far since this value is closer to the minimum price of electricity in Europe33.

The results varying the aforementioned parameters are reported in Figure

26. The highest impact on the profitability of the investment is attributable

31 This price is takes into account both the price that SNG would have on the natural gas market and the value of the incentive provided in the form of a subsidy for each MWh of produced gas. 32 6 k€/m2 is equivalent to approximately 470 €/kW for pressurized SOECs. 33 Sweden is the European country with the lowest price of electricity for large consumers (>150 GWh/year). These consumers paid 38 €/MWh (excluding all the taxes) in 2017. The minimum electricity price for consumers with a level of consumption comparable to the upgrading plant (between 20 and 70 GWh/year) is in Norway and it is equal to 40 €/MWh (in Germany 45.9, while in Sweden 47.1 €/MWh) [209].

-12

-10

-8

-6

-4

-2

0

2

4

6

8

0 4 8 12 16 20

NP

V [

M€

]

Time [y]

High Pressure

Low pressure


to the price of synthetic natural gas, while the cost of the electrolyzer has

the lowest impact because it constitutes just a part of the whole plant.


(a)

(b)

(c)

Figure 26: Net present value for different (a) electricity costs, (b) SNG price and (c) SOEC CAPEX.

-5

0

5

10

15

20

25

Electricity cost

NP

V [

M€

]

0 €/MWh11 €/MWh20 €/MWh30 €/MWh50 €/MWh

-10

-5

0

5

10

15

20

25

SNG price

NP

V [

M€

]

40 €/MWh

50 €/MWh

60 €/MWh

70 €/MWh

80 €/MWh

-6

-4

-2

0

2

4

6

8

SOEC CAPEX

NP

V [

M€

]

6 k€/m2

9 k€/m2

12 k€/m2

15 k€/m2

18 k€/m2

CONCLUSIONS | 93

5. Conclusions

In this dissertation an investigation of the techno-economic potential

for different electrochemical energy storage solutions has been presented,

covering different applications which have the potential to play a

significant role in future energy systems. In particular, small-scale and

large-scale battery energy storage was studied in order to mitigate the

uncertainty that variable renewables-based electricity supply brings to the

management of electricity. On the other hand, power-to-gas was

considered as a way to limit the curtailments of low-carbon energy by

transforming it into synthetic natural gas. The techno-economic evaluation

of the aforementioned processes represented the main research objective

of the thesis, that translated into four main research questions.

The first one focused on the most cost-effective strategy to reduce the

mismatch between energy production and demand in a small-scale

residential PV system. Batteries and demand response were compared in

terms of yearly savings for residential consumers. It turned out that

demand response produces better savings on the electricity bill, because

the cost of the equipment is much lower. Conversely, the implementation

of an automatic load shifting strategy, as the one addressed in this work, is

limited to cases in which thermal needs (e.g. domestic hot water or heating)

are directly satisfied with electricity. The present cost34 of the compound

battery+inverter (both for lithium-ion and for lead-acid) appears too high

to achieve a positive economic result. On the other hand, with a projected

cost that is 65% lower than the market price, the investment turns

convenient, although less attractive than DR implementation.

The second research question investigated the energy and economic

benefits of a large-scale battery energy storage plant in island settings,

adopting lithium-ion or vanadium flow batteries alternatively. The results

of the unit commitment model showed that both the technologies

considerably reduce the consumption of imported fossil fuels for electricity

generation. They have satisfactory performance in terms of NPV (>200%

34 The study was published in 2016, therefore the cost is referred to this year.

94 | CONCLUSIONS

after 20 years) and ROI (>20%), meaning that they are a meaningful choice

to promote renewables. The comparison reveals that vanadium flow

batteries produce slightly better results than lithium-ion because they are

assumed to have a lower capacity expenditure. In contrast, vanadium flow

batteries have a lower cycle efficiency, meaning that they waste more

energy in the storage process, but their lower projected cost compensates

for their poorer energy performance.

The third research question was related to the energy performance

analysis of biogas upgrading through high-temperature electrolysis and

methanation. The main outcome of the process is synthetic natural gas, but

also pure oxygen and excess heat are available as by-products. The

configuration with the electrolyzer operating at high pressure (around 30

bar) was the one characterized by the highest efficiency, even though

pressurizing the stack of SOECs increases the complexity and the cost. The

comparison with other techniques to upgrade biogas via carbon dioxide

separation revealed that the electrochemical upgrade reaches higher yields

for the same inlet biogas, even though it requires higher electrical inputs.

The profitability analysis of such process was part of the fourth research

question. Considering a low electricity price and a favorable estimation of

the price per ton of CO2 on the emission trading market, the production

cost of upgraded biogas was comparable with the price that medium

industrial consumers had to pay for natural gas in 2013, when the oil and

natural gas prices reached their maxima. Although, this analysis is referred

to a convenient future scenario, it shows that a viable pathway for CO2

recycle and synthetic fuel production is profitable, provided that a high

degree of process integration is adopted. Moreover, methanation-based

processes are exothermic and this constitutes a good basis for their further

integration with district heating or cooling network.

The adoption of electrochemical energy storage processes and

technologies proved to be able to mitigate some of the problems caused by

the increasing presence of renewables in the electricity generation mix.

Cost-wise the energy and environmental medium-term benefits have been

demonstrated for different storage applications and this validates the

efforts of a large part of the scientific and industrial community for the

improvement of such pathways. The main reasons that are restraining the

massive diffusion of storage solutions are: the still limited shares of

intermittent renewable energies in most of the world countries and the

high investment cost. In just a few of the advanced countries the share of

CONCLUSIONS | 95

variable renewable energy in power generation is above 20%, which is a

value for which the already present pumped hydroelectric plants and the

high degree of interconnection can handle the variability of renewables

[194]. Among the countries with higher share of renewables in the power

system, some of them, e.g. Portugal, have been able to incorporate large

quantities of renewable energies without curtailments, because of the high

number of hydro pumped plants [47].

The not-yet-competitive cost of storage solutions is an issue that has

been confirmed by all the case studies reported in this thesis, where the

economic evaluations have been performed based on projected costs. Both

batteries and electrolyzers produce good economic results if their capital

expenditures are strongly reduced compared with the current values and if

the price of fossil fuels is higher than the present level. The analysis of

small- and large-scale battery systems highlighted their high efficiency (up

to 95%), meaning that less energy is lost during the storage cycles in

comparison with other options, e.g. pumped hydro energy storage (PHES)

and compressed air energy storage (CAES)35.

The exergy analysis of the power-to-gas route through biogas upgrading

showed that the conversion efficiency of such a process can reach values

higher than 80%. This is favored by the use of the heat produced by the

process, which is in excess compared to the need of the process itself. This

fact favors the further integration with processes that have thermal needs

to satisfy (e.g. the anaerobic digester or the nearby buildings). On the other

hand, the economic evaluation revealed that, albeit with low-cost

electricity, a low price of natural gas, as the one in 2015, hinders

irremediably the profitability of the production of synthetic natural gas.

However, the combination of higher natural gas price, economic

exploitation of the by-products and participation to the carbon allowance

market can make this process competitive.

The main environmental advantage of renewable gaseous fuels lies in

the possibility to recycle residues and waste, which is in accordance with

the targets contained in the Circular Economy Package approved by the

European Parliament. In particular, 65% of municipal solid waste should

be recycled by 2035 and only a maximum of 10% could be landfilled, of

which the biodegradable fraction, useful for biogas production, will have to

be either collected separately or recycled at home through composting

35 The round trip efficiency of PHES is in the range 65-85% [7], while for CAES it can reach 70% [8].

96 | CONCLUSIONS

[195]. The integration of the disposal process with the fuel synthesis can

solve two problems at the same time, namely waste treatment and

decarbonization of large energy consumption sectors (e.g. transport), by

reducing their pressure on the environment. These strategies are

functional to follow the long-term roadmap that the European Commission

defined to promote the reduction of greenhouse gas emissions [196]. It is

true that the EU emissions account only for 9% of all the emissions related

to fossil-fuel combustion, cement production and gas flaring in the world

[197]; however, the fulfilment of the goals that have been established to

reduce the dependence on fossil fuels and to lower environmental impacts

of human activities, can be an inspiring example for others to follow.

Finally, the connection between the power and the transport sectors is

becoming more and more evident both for the diffusion of electric mobility

solutions (not only cars but also scooters and, to some extent buses [198]),

and for the rising interest in electrofuel production. Electric mobility seems

more suitable for urban contexts because it does not produce local

emissions and the issues connected to limited mileage can be solved

through a dense network of charging opportunities. On the other hand, for

long-distance travel and heavy-duty vehicles the higher energy density of

fuels is a feature that appears difficult to be replaced by electric vehicles.

Therefore, electrochemical energy storage solutions seem to have a large

potential to promote the development of greener and more sustainable

consumption patterns.

5.1. Future Work

This doctoral thesis started when the topic of energy storage was

considered a challenge that energy systems were going to face in a near

future, though this future was still quite far to come. After only four years

the interest in this topic has grown exponentially, especially driven by the

dramatic reduction in the cost of solar PV, wind, and battery systems. This

is proved by the huge amount of peer-reviewed researches published on

the topic of energy storage. They increased exponentially passing from

1500 in 2010 to 7249 in 201736 and this testifies the importance that the

scientific community is attributing to the topic. These favorable conditions

have triggered the growth of renewable power plants in many countries.

36 These numbers are based on a search in Scopus among the peer-reviewed papers and book chapters with the keyword “energy storage”.

CONCLUSIONS | 97

That being said, some of the problems connected to the integration of

variable renewables are still open and the possibility for the study of

interesting applications is unprecedented. Thereby, the development of

ecologic and cost-effective solutions to accompany the transition in the

direction of an energy production paradigm less dependent on fossil fuels

is one of the priorities of the current innovation trends.

In general, isolated or remote systems should be regarded as test beds

for innovative solutions. The exploitation of local energy resources is vital

to reach a higher degree of sustainability of the local energy system. Many

islands can count on high solar irradiation levels, on wind blowing for most

of the year, and on wave and tidal energy. All these resources are

intermittent and should be complemented with storage systems to fully

exploit their decarbonization potential for the generation system.

Therefore, additional efforts should be put into the design of effective

pathways that can promote the use of local resources, also through the

actual implementation of pilot projects to discover the challenges related

to the peculiarities of each case.

For power-to-gas processes using biogas, the main need is the scale-up

from the proof-of-concept stage to a more reliable system, which is able to

work in real-time conditions. Some projects of this kind are already

running [199], but most of them use alkaline electrolyzers, while an

additional advantage could be brought by the use of high-temperature

electrolyzers. The studies presented in this thesis are the result of a precise

of an accurate modeling effort, which would greatly benefit from an

experimental campaign to validate and fine-tune it. Moreover, market

mechanisms to make the use of surplus electricity profitable should be

studied, as a means to reward the activity of the processes that are able to

act as buffers during hours of extra-production. Some initial attempts have

been done as in Larscheid et al. [200], but more efforts are needed to define

how to efficiently couple the electricity and the natural gas markets to allow

power-to-gas to be profitable.

Finally, the integration of power system and transportation should be

further investigated both for islands and for continental settings. Strategies

to favor the rational management of mobility demand (i.e. vehicle and ride

sharing, carpooling, and public transport optimization) should be coupled

with a sustainable procurement of the necessary energy resources. In other

words, the systemic reduction of the energy consumption in transport is as

important as finding new low-environmental-impact energy carriers. In

98 | CONCLUSIONS

particular, the implementation of power-to-gas routes should be

considered as a possibility to pare back the quantity of organic waste that

have to be landfilled. Together with this, an analysis of the potential raw

materials locally available is important to define the extent up to which

local resources can serve the aim of decarbonization. This step is complex

because it needs an accurate inventory of crops, agricultural and industrial

residues and other material streams, the data of which are usually not

easily available. The choice of isolated systems facilitates the definition of

the boundaries and, therefore, can provide useful insights on the further

developments of the interrelation between energy storage and

transportation services.

REFERENCES| 99

References

[1] Eurostat. Eurostat Shares 2018. http://ec.europa.eu/eurostat/web/energy/data/shares (accessed January 28, 2018).

[2] Holttinen H, Milligan M, Ela E, Menemenlis N, Dobschinski J, Rawn B, et al. Methodologies to determine operating reserves due to increased wind power. IEEE Trans Sustain Energy 2012;3:713–23. doi:10.1109/TSTE.2012.2208207.

[3] Rejc ŽB, Čepin M. Estimating the additional operating reserve in power systems with installed renewable energy sources. Int J Electr Power Energy Syst 2014;62:654–64. doi:10.1016/j.ijepes.2014.05.019.

[4] DoE. Global Energy Storage Database 2018. [5] Rogner H-H, Aguilera RF, Archer C, Bertani R, Bhattacharya, S.C.,

Dusseault MB. Chapter7 - Energy resources and potentials. Glob. Energy Assessment—Toward a Sustain. Futur., Cambridge, UK and New York, NY, USA and the International Institute for applied systems analysis, Laxenburg, Austria. Cambridge University Press; 2012, p. 423–512.

[6] Christen T, Carlen MW. Theory of ragone plots. J Power Sources 2000;91:210–6. doi:10.1016/S0378-7753(00)00474-2.

[7] Aneke M, Wang M. Energy storage technologies and real life applications – A state of the art review. Appl Energy 2016;179:350–77. doi:10.1016/j.apenergy.2016.06.097.

[8] Luo X, Wang J, Dooner M, Clarke J. Overview of current development in electrical energy storage technologies and the application potential in power system operation. Appl Energy 2015;137:511–36. doi:10.1016/j.apenergy.2014.09.081.

[9] Kepplinger P, Huber G, Petrasch J. Autonomous optimal control for demand side management with resistive domestic hot water heaters using linear optimization. Energy Build 2015;100:50–5. doi:10.1016/j.enbuild.2014.12.016.

[10] Kepplinger P, Huber G, Petrasch J. Field testing of demand side management via autonomous optimal control of a domestic hot water heater. Energy Build 2016;127:730–5. doi:10.1016/j.enbuild.2016.06.021.

[11] Kim Y, Norford LK. Optimal use of thermal energy storage resources in commercial buildings through price-based demand response considering distribution network operation. Appl Energy 2017;193:308–24. doi:10.1016/j.apenergy.2017.02.046.

[12] European Commission. The European Union leading in renewables. 2015. [13] European Commission. The Strategic Energy Technology (SET) Plan.

Publications Office of the European Union,; 2017. doi:10.2777/48982. [14] Shah V, Booream-Phelps J. Crossing the Chasm. Dtsch Bank – Mark Res

2015:185.

100 | REFERENCES

doi:http://www.climatecouncil.org.au/uploads/ed4518226c655546cc529390c7cd4a8f.pdf.

[15] European Comission. Best practices on Renewable Energy Self-consumption. J Chem Inf Model 2015;53:1689–99. doi:10.1017/CBO9781107415324.004.

[16] TechNavio. Global Residential Solar Energy Storage Market 2017-2021, Technavio (2017). 2017.

[17] Ranaweera I, Midtgård OM, Korpås M. Distributed control scheme for residential battery energy storage units coupled with PV systems. Renew Energy 2017;113:1099–110. doi:10.1016/j.renene.2017.06.084.

[18] O’Dwyer C, Ryan L, Flynn D. Efficient large-scale energy storage dispatch: challenges in future high renewables systems. IEEE Trans Power Syst 2017;32:3439–50. doi:10.1109/TPWRS.2017.2656245.

[19] Lopes AS, Castro R, Silva C. Contributions to the Design of a Water Pumped Storage System in an Isolated Power System with High Penetration of Wind Energy. Technol Innov Cyber-Physical Syst 2016;470:378–86. doi:10.1007/978-3-319-31165-4.

[20] Ma T, Yang H, Lu L. A feasibility study of a stand-alone hybrid solar-wind-battery system for a remote island. Appl Energy 2014;121:149–58. doi:10.1016/j.apenergy.2014.01.090.

[21] Ellison JF, Rashkin LJ, Serio J, Byrne RH. The benefits of grid-scale storage on Oahu. J Energy Storage 2018;15:336–44. doi:10.1016/j.est.2017.12.009.

[22] Guandalini G, Campanari S, Romano MC. Power-to-gas plants and gas turbines for improved wind energy dispatchability: Energy and economic assessment. Appl Energy 2015;147:117–30. doi:10.1016/j.apenergy.2015.02.055.

[23] Lyseng B, Niet T, English J, Keller V, Palmer-Wilson K, Robertson B, et al. System-level power-to-gas energy storage for high penetrations of variable renewables. Int J Hydrogen Energy 2018;43:1966–79. doi:10.1016/j.ijhydene.2017.11.162.

[24] Varone A, Ferrari M. Power to liquid and power to gas: An option for the German Energiewende. Renew Sustain Energy Rev 2015;45:207–18. doi:10.1016/j.rser.2015.01.049.

[25] Mazza A, Bompard E, Chicco G. Applications of power to gas technologies in emerging electrical systems. Renew Sustain Energy Rev 2018;92:794–806. doi:10.1016/j.rser.2018.04.072.

[26] Peduzzi E, Boissonnet G, Maréchal F. Screening methodology for biorefinery and power to chemicals concepts. Comput Aided Chem Eng 2016;38:1893–8. doi:10.1016/B978-0-444-63428-3.50320-9.

[27] Gutiérrez-Martín F, Rodríguez-Antón LM. Power-to-SNG technology for energy storage at large scales. Int J Hydrogen Energy 2016;41:19290–303. doi:10.1016/j.ijhydene.2016.07.097.

[28] Blanco H, Faaij A. A review at the role of storage in energy systems with a focus on Power to Gas and long-term storage. Renew Sustain Energy Rev 2018;81:1049–86. doi:10.1016/j.rser.2017.07.062.

[29] Clausen LR. Energy efficient thermochemical conversion of very wet biomass to biofuels by integration of steam drying, steam electrolysis and gasification. Energy 2017;125:327–36. doi:10.1016/j.energy.2017.02.132.

REFERENCES| 101

[30] Qian H, Zhu W, Liu C, Lu X, Kontogeorgis GM, Gani R. Exergy efficiency based design and analysis of utilization pathways of biomasses. vol. 43. Elsevier Masson SAS; 2018. doi:10.1016/B978-0-444-64235-6.50150-9.

[31] Turk D. Energy storage and system integration – an international perspective. In: IEA, editor. Sect. Integr. Support. by Energy Storage Hydrog. High Lev. Roundtable, 2018.

[32] Schmidt O, Hawkes A, Gambhir A, Staffell I. The future cost of electrical energy storage based on experience rates. Nat Energy 2017;6:17110. doi:10.1038/nenergy.2017.110.

[33] Petrakopoulou F, Sanz-Bermejo J, Dufour J, Romero M. Exergetic analysis of hybrid power plants with biomass and photovoltaics coupled with a solid-oxide electrolysis system. Energy 2016;94:304–15. doi:10.1016/j.energy.2015.10.118.

[34] Liu M, Steven Tay NH, Bell S, Belusko M, Jacob R, Will G, et al. Review on concentrating solar power plants and new developments in high temperature thermal energy storage technologies. Renew Sustain Energy Rev 2016;53:1411–32. doi:10.1016/j.rser.2015.09.026.

[35] Sarbu I, Sebarchievici C. Thermal Energy Storage. 2017. doi:10.1016/B978-0-12-811662-3.00004-9.

[36] Buttler A, Spliethoff H. Current status of water electrolysis for energy storage, grid balancing and sector coupling via power-to-gas and power-to-liquids: A review. Renew Sustain Energy Rev 2018;82:2440–54. doi:10.1016/j.rser.2017.09.003.

[37] Pozzo M, Lanzini A, Santarelli M. Enhanced biomass-to-liquid (BTL) conversion process through high temperature co-electrolysis in a solid oxide electrolysis cell (SOEC). Fuel 2015;145:39–49. doi:10.1016/j.fuel.2014.12.066.

[38] Ail SS, Dasappa S. Biomass to liquid transportation fuel via Fischer Tropsch synthesis - Technology review and current scenario. Renew Sustain Energy Rev 2016;58:267–86. doi:10.1016/j.rser.2015.12.143.

[39] Herz G, Reichelt E, Jahn M. Techno-economic analysis of a co-electrolysis-based synthesis process for the production of hydrocarbons. Appl Energy 2018;215:309–20. doi:10.1016/j.apenergy.2018.02.007.

[40] Becker WL, Braun RJ, Penev M, Melaina M. Production of Fischer-Tropsch liquid fuels from high temperature solid oxide co-electrolysis units. Energy 2012;47:99–115. doi:10.1016/j.energy.2012.08.047.

[41] Lund PD, Lindgren J, Mikkola J, Salpakari J. Review of energy system flexibility measures to enable high levels of variable renewable electricity. Renew Sustain Energy Rev 2015;45:785–807. doi:10.1016/j.rser.2015.01.057.

[42] Ehlers N. The German energy transition & integration of renewable energy 2017.

[43] Eurostat. Infrastructure - electricity - annual data [nrg_113a] 2018. https://ec.europa.eu/eurostat/statistics-explained/index.php/Energy_from_renewable_sources (accessed August 30, 2018).

[44] IHA. Hydropower in Germany 2017. https://www.hydropower.org/country-profiles/germany (accessed April 25, 2018).

102 | REFERENCES

[45] Net installed electricity generation capacity in Germany. Fraunhofer ISE 2018. https://www.energy-charts.de/power_inst.htm (accessed April 4, 2018).

[46] Federal Network Agency. Monitoring report 2017. 2017. [47] WindEurope. WindEurope views on curtailment of wind power and its link

to the priority of dispatch. 2016. [48] Buijs P, Bekaert D, Cole S, Van Hertem D, Belmans R. Transmission

investment problems in Europe: Going beyond standard solutions. Energy Policy 2011;39:1794–801. doi:10.1016/j.enpol.2011.01.012.

[49] Akinyele DO, Rayudu RK. Review of energy storage technologies for sustainable power networks. Sustain Energy Technol Assessments 2014;8:74–91. doi:10.1016/j.seta.2014.07.004.

[50] Bollen MHJ. What is power quality? Electr Power Syst Res 2003;66:5–14. doi:10.1016/S0378-7796(03)00067-1.

[51] Denholm P, Ela E, Kirby B, Milligan M. The Role of Energy Storage with Renewable Electricity Generation. 2010. doi:69.

[52] Grid Scale Energy Storage Systems 2018. http://www.mpoweruk.com/grid_storage.htm (accessed March 15, 2018).

[53] Guney MS, Tepe Y. Classification and assessment of energy storage systems. Renew Sustain Energy Rev 2017;75:1187–97. doi:10.1016/j.rser.2016.11.102.

[54] Yang Y, Bremner S, Menictas C, Kay M. Battery energy storage system size determination in renewable energy systems: A review. Renew Sustain Energy Rev 2018;91:109–25. doi:10.1016/j.rser.2018.03.047.

[55] Divya KC, Østergaard J. Battery energy storage technology for power systems-An overview. Electr Power Syst Res 2009;79:511–20. doi:10.1016/j.epsr.2008.09.017.

[56] Chen H, Cong TN, Yang W, Tan C, Li Y, Ding Y. Progress in electrical energy storage system: A critical review. Prog Nat Sci 2009;19:291–312. doi:10.1016/j.pnsc.2008.07.014.

[57] Arenas LF, Ponce de León C, Walsh FC. Engineering aspects of the design, construction and performance of modular redox flow batteries for energy storage. J Energy Storage 2017;11:119–53. doi:10.1016/j.est.2017.02.007.

[58] Doetsch C, Burfeind J. Vanadium Redox Flow Batteries. Storing Energy, Elsevier Inc.; 2016, p. 227–46. doi:10.1016/B978-0-12-803440-8/00012-9.

[59] Cao L, Skyllas-Kazacos M, Menictas C, Noack J. A review of electrolyte additives and impurities in vanadium redox flow batteries. J Energy Chem 2018;0:1–23. doi:10.1016/j.jechem.2018.04.007.

[60] Choi C, Kim S, Kim R, Choi Y, Kim S, Jung H young, et al. A review of vanadium electrolytes for vanadium redox flow batteries. Renew Sustain Energy Rev 2017;69:263–74. doi:10.1016/j.rser.2016.11.188.

[61] Liu T, Li X, Zhang H, Chen J. Progress on the electrode materials towards vanadium flow batteries (VFBs) with improved power density. J Energy Chem 2018. doi:10.1016/j.jechem.2018.07.003.

[62] Vanýsek P, Novák V. Redox flow batteries as the means for energy storage. J Energy Storage 2017;13:435–41. doi:10.1016/j.est.2017.07.028.

[63] Götz M, Lefebvre J, Mörs F, McDaniel Koch A, Graf F, Bajohr S, et al. Renewable Power-to-Gas: A technological and economic review. Renew

REFERENCES| 103

Energy 2016;85:1371–90. doi:10.1016/j.renene.2015.07.066. [64] Rauch R, Kiennemann A, Sauciuc A. Fischer-Tropsch Synthesis to Biofuels

(BtL Process). vol. #volume#. © 2013 Elsevier B.V. All rights reserved.; 2013. doi:10.1016/B978-0-444-56330-9.00012-7.

[65] Judd R, Pinchbeck D. Hydrogen admixture to the natural gas grid. Compend Hydrog Energy 2016:165–92. doi:10.1016/B978-1-78242-364-5.00008-7.

[66] Wang L, Pérez-Fortes M, Madi H, Diethelm S, herle J Van, Maréchal F. Optimal design of solid-oxide electrolyzer based power-to-methane systems: A comprehensive comparison between steam electrolysis and co-electrolysis. Appl Energy 2018;211:1060–79. doi:10.1016/j.apenergy.2017.11.050.

[67] Laguna-Bercero MA. Recent advances in high temperature electrolysis using solid oxide fuel cells: A review. J Power Sources 2012;203:4–16. doi:10.1016/j.jpowsour.2011.12.019.

[68] Mingyi L, Bo Y, Jingming X, Jing C. Thermodynamic analysis of the efficiency of high-temperature steam electrolysis system for hydrogen production. J Power Sources 2008;177:493–9. doi:10.1016/j.jpowsour.2007.11.019.

[69] Sun X, Chen M, Jensen SH, Ebbesen SD, Graves C, Mogensen M. Thermodynamic analysis of synthetic hydrocarbon fuel production in pressurized solid oxide electrolysis cells. Int J Hydrogen Energy 2012;37:17101–10. doi:10.1016/j.ijhydene.2012.08.125.

[70] Brynolf S, Taljegard M, Grahn M, Hansson J. Electrofuels for the transport sector: A review of production costs. Renew Sustain Energy Rev 2018;81:1887–905. doi:10.1016/j.rser.2017.05.288.

[71] Knupfer SM, Hensley R, Hertzke P, Schaufuss P, Laverty N, Kramer N. Electrifying Insights: How Automakers Can Drive Electrified Vehicle Sales and Profitability. 2017.

[72] EASE – Delta-ee. European Market Monitor on Energy Storage - Second Edition (EMMES 2.0). 2018.

[73] Briano J, Báez M, Morales R. Photovoltaic Grid Parity Monitor - Residential. Berlin: 2015.

[74] Gitizadeh M, Fakharzadegan H. Battery capacity determination with respect to optimized energy dispatch schedule in grid-connected photovoltaic (PV) systems. Energy 2014;65:665–74. doi:10.1016/j.energy.2013.12.018.

[75] Dufo-López R, Bernal-Agustín JL. Techno-economic analysis of grid-connected battery storage. Energy Convers Manag 2015;91:394–404. doi:10.1016/j.enconman.2014.12.038.

[76] Ratnam EL, Weller SR, Kellett CM. Scheduling residential battery storage with solar PV: Assessing the benefits of net metering. Appl Energy 2015;155:881–91. doi:10.1016/j.apenergy.2015.06.061.

[77] Mulder G, Six D, Claessens B, Broes T, Omar N, Mierlo J Van. The dimensioning of PV-battery systems depending on the incentive and selling price conditions. Appl Energy 2013;111:1126–35. doi:10.1016/j.apenergy.2013.03.059.

[78] Nottrott A, Kleissl J, Washom B. Energy dispatch schedule optimization and cost benefit analysis for grid-connected, photovoltaic-battery storage

104 | REFERENCES

systems. Renew Energy 2013;55:230–40. doi:10.1016/j.renene.2012.12.036.

[79] Gutiérrez-Alcaraz G, Tovar-Hernández JH, Lu C-N. Effects of demand response programs on distribution system operation. Int J Electr Power Energy Syst 2016;74:230–7. doi:10.1016/j.ijepes.2015.07.018.

[80] Koliou E, Bartusch C, Picciariello A, Eklund T, Söder L, Hakvoort RA. Quantifying distribution-system operators’ economic incentives to promote residential demand response. Util Policy 2015;35:28–40. doi:10.1016/j.jup.2015.07.001.

[81] D’hulst R, Labeeuw W, Beusen B, Claessens S, Deconinck G, Vanthournout K. Demand response flexibility and flexibility potential of residential smart appliances: Experiences from large pilot test in Belgium. Appl Energy 2015;155:79–90. doi:10.1016/j.apenergy.2015.05.101.

[82] Muratori M, Schuelke-Leech BA, Rizzoni G. Role of residential demand response in modern electricity markets. Renew Sustain Energy Rev 2014;33:546–53. doi:10.1016/j.rser.2014.02.027.

[83] Vanthournout K, Dupont B, Foubert W, Stuckens C, Claessens S. An automated residential demand response pilot experiment, based on day-ahead dynamic pricing. Appl Energy 2015;155:195–203. doi:10.1016/j.apenergy.2015.05.100.

[84] Wang X, Palazoglu A, El-Farra NH. Operational optimization and demand response of hybrid renewable energy systems. Appl Energy 2015;143:324–35. doi:10.1016/j.apenergy.2015.01.004.

[85] Huang Y, Tian H, Wang L. Demand response for home energy management system. Int J Electr Power Energy Syst 2015;73:448–55. doi:10.1016/j.ijepes.2015.05.032.

[86] Zame KK, Brehm CA, Nitica AT, Richard CL, Schweitzer GD. Smart grid and energy storage: Policy recommendations. Renew Sustain Energy Rev 2018;82:1646–54. doi:10.1016/j.rser.2017.07.011.

[87] Song J, Oh SD, Yoo Y, Seo SH, Paek I, Song Y, et al. System design and policy suggestion for reducing electricity curtailment in renewable power systems for remote islands. Appl Energy 2018;225:195–208. doi:10.1016/j.apenergy.2018.04.131.

[88] Neves D, Silva CA, Connors S. Design and implementation of hybrid renewable energy systems on micro-communities: A review on case studies. Renew Sustain Energy Rev 2014;31:935–46. doi:10.1016/j.rser.2013.12.047.

[89] Kuang Y, Zhang Y, Zhou B, Li C, Cao Y, Li L, et al. A review of renewable energy utilization in islands. Renew Sustain Energy Rev 2016;59:504–13. doi:10.1016/j.rser.2016.01.014.

[90] Raju L, Morais AA, Rathnakumar R, Ponnivalavan S, Thavam LD. Micro-grid grid outage management using multi-agent systems. Proc - 2017 2nd Int Conf Recent Trends Challenges Comput Model ICRTCCM 2017 2017;117:363–8. doi:10.1109/ICRTCCM.2017.21.

[91] Stenzel P, Schreiber A, Marx J, Wulf C, Schreieder M, Stephan L. Environmental Impacts of Electricity Generation for Graciosa Island , Azores Location of the Azores and Graciosa Island Graciosa Island : J Energy Storage 2017;15:292–303. doi:10.1016/j.est.2017.12.002.

[92] Castillo A, Gayme DF. Grid-scale energy storage applications in renewable

REFERENCES| 105

energy integration: A survey. Energy Convers Manag 2014;87:885–94. doi:10.1016/j.enconman.2014.07.063.

[93] Lin Y, Johnson JX, Mathieu JL. Emissions impacts of using energy storage for power system reserves. Appl Energy 2016;168:444–56. doi:10.1016/j.apenergy.2016.01.061.

[94] Lorenzi G, Silva CAS. Comparing demand response and battery storage to optimize self-consumption in PV systems. Appl Energy 2016;180. doi:10.1016/j.apenergy.2016.07.103.

[95] Zhang C, Zhao TS, Xu Q, An L, Zhao G. Effects of operating temperature on the performance of vanadium redox flow batteries. Appl Energy 2015;155:349–53. doi:10.1016/j.apenergy.2015.06.002.

[96] López-Vizcaíno R, Mena E, Millán M, Rodrigo MA, Lobato J. Performance of a vanadium redox flow battery for the storage of electricity produced in photovoltaic solar panels. Renew Energy 2017;114:1123–33. doi:10.1016/j.renene.2017.07.118.

[97] Sadeghi S. Study using the flow battery in combination with solar panels and solid oxide fuel cell for power generation. Sol Energy 2018;170:732–40. doi:10.1016/j.solener.2018.05.091.

[98] Alotto P, Guarnieri M, Moro F. Redox flow batteries for the storage of renewable energy: A review. Renew Sustain Energy Rev 2014;29:325–35. doi:10.1016/j.rser.2013.08.001.

[99] Minke C, Kunz U, Turek T. Techno-economic assessment of novel vanadium redox flow batteries with large-area cells. J Power Sources 2017;361:105–14. doi:10.1016/j.jpowsour.2017.06.066.

[100] Eber K, Corbus D. Hawaii Solar Integration Study. 2013. [101] Woodford D. Oahu Wind Integration and Transmission Study. 2011. [102] Komor P, Glassmaire J. Electricity Storage and Renewables for Island

Power. 2012. [103] Sioshansi FP. National Energy Roadmaps for Islands. 2013. [104] Neves D, Silva CA. Optimal electricity dispatch on isolated mini-grids using

a demand response strategy for thermal storage backup with genetic algorithms. Energy 2015;82:436–45. doi:10.1016/j.energy.2015.01.054.

[105] Pina A, Baptista P, Silva C, Ferrão P. Energy reduction potential from the shift to electric vehicles: The Flores island case study. Energy Policy 2014;67:37–47. doi:10.1016/j.enpol.2013.07.120.

[106] Gahleitner G. Hydrogen from renewable electricity: An international review of power-to-gas pilot plants for stationary applications. Int J Hydrogen Energy 2013;38:2039–61. doi:10.1016/j.ijhydene.2012.12.010.

[107] Graves C, Ebbesen SD, Mogensen M, Lackner KS. Sustainable hydrocarbon fuels by recycling CO2 and H2O with renewable or nuclear energy. Renew Sustain Energy Rev 2011;15:1–23. doi:10.1016/j.rser.2010.07.014.

[108] Gondal IA, Masood SA, Khan R. Green hydrogen production potential for developing a hydrogen economy in Pakistan. Int J Hydrogen Energy 2018;43:6011–39. doi:10.1016/j.ijhydene.2018.01.113.

[109] Ridjan I, Mathiesen BV, Connolly D. Terminology used for renewable liquid and gaseous fuels based on the conversion of electricity: A review. J Clean Prod 2016;112:3709–20. doi:10.1016/j.jclepro.2015.05.117.

[110] Reiter G, Lindorfer J. Evaluating CO2 sources for power-to-gas applications – A case study for Austria. J CO2 Util 2015;10:40–9.

106 | REFERENCES

doi:10.1016/j.jcou.2015.03.003. [111] Elfving J, Bajamundi C, Kauppinen J, Sainio T. Modelling of equilibrium

working capacity of PSA, TSA and TVSA processes for CO2adsorption under direct air capture conditions. J CO2 Util 2017;22:270–7. doi:10.1016/j.jcou.2017.10.010.

[112] Ryckebosch E, Drouillon M, Vervaeren H. Techniques for transformation of biogas to biomethane. Biomass and Bioenergy 2011;35:1633–45. doi:10.1016/j.biombioe.2011.02.033.

[113] Vagonyte E. Biogas & Biomethane in Europe. 2011. doi:247896520. [114] Angelidaki I, Treu L, Tsapekos P, Luo G, Campanaro S, Wenzel H, et al.

Biogas upgrading and utilization: Current status and perspectives. Biotechnol Adv 2018:0–1. doi:10.1016/j.biotechadv.2018.01.011.

[115] Rotunno P, Lanzini A, Leone P. Energy and economic analysis of a water scrubbing based biogas upgrading process for biomethane injection into the gas grid or use as transportation fuel. Renew Energy 2017;102:417–32. doi:10.1016/j.renene.2016.10.062.

[116] Baker RW. Membrane Technology and Applications. third. Wiley; 2012. doi:10.1002/0471238961.1305130202011105.a01.

[117] Scholz M, Melin T, Wessling M. Transforming biogas into biomethane using membrane technology. Renew Sustain Energy Rev 2013;17:199–212. doi:10.1016/j.rser.2012.08.009.

[118] Augelletti R, Conti M, Annesini MC. Pressure swing adsorption for biogas upgrading. A new process configuration for the separation of biomethane and carbon dioxide. J Clean Prod 2017;140:1390–8. doi:10.1016/j.jclepro.2016.10.013.

[119] Graves C, Ebbesen SD, Mogensen M. Co-electrolysis of CO2 and H2O in solid oxide cells: Performance and durability. Solid State Ionics 2011;192:398–403. doi:10.1016/j.ssi.2010.06.014.

[120] Li W, Wang H, Shi Y, Cai N. Performance and methane production characteristics of H2O- CO2co-electrolysis in solid oxide electrolysis cells. Int J Hydrogen Energy 2013;38:11104–9. doi:10.1016/j.ijhydene.2013.01.008.

[121] Giglio E, Lanzini A, Santarelli M, Leone P. Synthetic natural gas via integrated high-temperature electrolysis and methanation: Part I—Energy performance. J Energy Storage 2015;1:22–37. doi:10.1016/j.est.2015.04.002.

[122] De Saint Jean M, Baurens P, Bouallou C, Couturier K. Economic assessment of a power-to-substitute-natural-gas process including high-temperature steam electrolysis. Int J Hydrogen Energy 2015;40:6487–500. doi:10.1016/j.ijhydene.2015.03.066.

[123] From solid fuels to substitute natural gas (SNG) using TREMPTM n.d. [124] Collet P, Flottes E, Favre A, Raynal L, Pierre H, Capela S, et al. Techno-

economic and Life Cycle Assessment of methane production via biogas upgrading and power to gas technology. Appl Energy 2017;192:282–95. doi:10.1016/j.apenergy.2016.08.181.

[125] Irradiation Data for Lisbon n.d. http://www.wunderground.com/personal-weather-station/dashboard?ID=ILISBONL3#history/s20150219/e20150226/mweek (accessed June 26, 2015).

REFERENCES| 107

[126] REN, Perfis de Consumo n.d. http://www.mercado.ren.pt/PT/Electr/InfoMercado/Consumo/Paginas/PerfisConsumo.aspx.

[127] Paleta R, Pina A, Santos Silva CA. Polygeneration Energy Container: Designing and Testing Energy Services for Remote Developing Communities. IEEE Trans Sustain Energy 2014;5:1–1. doi:10.1109/TSTE.2014.2308017.

[128] Lorenzi G, Silva CAS. Techno-economic comparison of storage vs . demand response strategies in distributed generation systems. 12th Int. Conf. Eur. Energy Mark., 2015.

[129] Pordata. Average Number of People per Household 2016. http://www.pordata.pt/en/Europe/Average+number+of+persons+per+household-1613.

[130] Pombeiro H, Pina A, Silva C. Analyzing Residential Electricity Consumption Patterns Based on Consumer’ s Segmentation. Procs. 1st Int’l Work. Inf. Technol. Energy Appl., n.d.

[131] ERSE (Entitade Regulatoria dos Serviços Energéticos). Caracterização da procura de energia elétrica em 2012 (Document in Portuguese). 2013.

[132] ERSE (Entitade Regulatoria dos Serviços Energéticos). Preços De Referência No Mercado Liberalizado De Energia Elétrica E Gás Natural (Document in Portuguese). 2015.

[133] Ministry of Environment Land Management and Energy. Decreed Law n.o 153/2014 - 20 October 2014. 2014.

[134] Crispim J, Carreira M, Rui C. Validation of photovoltaic electrical models against manufacturers data and experimental results. POWERENG 2007 - Int. Conf. Power Eng. - Energy Electr. Drives Proc., 2007, p. 556–61. doi:10.1109/POWERENG.2007.4380161.

[135] LDK 250P-20 n.d. http://www.ldksolar.com/uploadfiles/down/LDK_215P_250P_20_EN_V1_12_120229.pdf (accessed June 27, 2016).

[136] Dolara A, Leva S, Manzolini G. Comparison of different physical models for PV power output prediction. Sol Energy 2015;119:83–99. doi:10.1016/j.solener.2015.06.017.

[137] Mark D. Hopkins, Pahwa A, Easton T. Intelligent Dispatch for Distributed Renewable Resources. IEEE Trans Smart Grid 2012;3:1047–54.

[138] Fronius Symo Hybrid 5.0-3-S Datasheet n.d. https://www.fronius.com/cps/rde/xchg/SID-450DE065-4091E8DA/fronius_international/hs.xsl/83_35331_ENG_HTML.htm#.Vj-6LLerTIU (accessed June 27, 2016).

[139] Collaboration between Tesla and Fronius n.d. http://www.fronius.com/cps/rde/xbcr/SID-9AAC2BC5-F7233319/fronius_international/SE_MAIL_Newsletter_ROW_May_2015_579409_snapshot.pdf (accessed June 27, 2016).

[140] Kittner N, Lill F, Kammen DM. Energy storage deployment and innovation for the clean energy transition. Nat Energy 2017;2:1–6. doi:10.1038/nenergy.2017.125.

[141] TESLA MOTORS. Tesla Powerwall. Energy Storage a Sustain Home 2016. https://www.teslamotors.com/powerwall?redirect=no (accessed June 27, 2016).

108 | REFERENCES

[142] Wholesalesolar n.d. http://www.wholesalesolar.com/9905042/concorde-sun-xtender/batteries/concorde-sun-xtender-pvx-2120l-agm-battery (accessed June 27, 2016).

[143] Fronius Symo Hybrid 5.0-3-S Price n.d. http://www.europe-solarstore.com/solar-inverters/fronius/fronius-symo-hybrid/fronius-symo-hybrid-5-0-3-s.html (accessed June 27, 2016).

[144] SMA Sunny Island 3.0M Price n.d. http://www.mg-solar-shop.de/solar-inverter/sma-solar-inverter/SMA-Sunny-Island-SI-3-0-M-11-with-SRC-20.html (accessed June 17, 2016).

[145] Ruelens F, Claessens BJ, Vandael S, Iacovella S, Vingerhoets P, Belmans R. Demand response of a heterogeneous cluster of electric water heaters using batch reinforcement learning. 2014 Power Syst Comput Conf 2014:1–7. doi:10.1109/PSCC.2014.7038106.

[146] Fuentes E, Arce L, Salom J. A review of domestic hot water consumption profiles for application in systems and buildings energy performance analysis. Renew Sustain Energy Rev 2018;81:1530–47. doi:10.1016/j.rser.2017.05.229.

[147] European Commision. Supplement to the European Parliament directive regarding energy labelling of water heaters. 2013:77.

[148] Abrantes J. Avaliação técnica e económica da aplicação de sistemas Waste to Energy no tratamento de resíduos urbanos em aglomerados de média e pequena dimensão. Instituto Superior Técnico, 2016.

[149] EDA - Historical data. EDA 2018. http://www.eda.pt/Mediateca/Publicacoes/Producao/Paginas/Produção-de-Energia-Elétrica.aspx (accessed February 15, 2018).

[150] EDA. Multi-annual strategic plan and budget for 2018. 2017. [151] Wilson IAG, Barbour E, Ketelaer T, Kuckshinrichs W. An analysis of

storage revenues from the time-shifting of electrical energy in Germany and Great Britain from 2010 to 2016. J Energy Storage 2018;17:446–56. doi:10.1016/j.est.2018.04.005.

[152] Neves D, Silva CA. Modeling the impact of integrating solar thermal systems and heat pumps for domestic hot water in electric systems - The case study of Corvo Island. Renew Energy 2014;72:113–24. doi:10.1016/j.renene.2014.06.046.

[153] Pordata - Portuguese Statistics. Resident Population in Portugal 2017. www.pordata.pt (accessed January 20, 2017).

[154] EDA. Relatório Técnico 2007. 2008. [155] IEA. Renewable Energy Essentials: Geothermal. 2010. [156] European Commission. Reference Document on the Best Available

Techniques for Waste Incineration. 2006. doi:10.1002/0470012668.ch5. [157] Association ES. Spinning Reserve n.d. http://energystorage.org/energy-

storage/technology-applications/spinning-reserve (accessed March 20, 2018).

[158] Jaiswal A. Lithium-ion battery based renewable energy solution for off-grid electricity: A techno-economic analysis. Renew Sustain Energy Rev 2017;72:922–34. doi:10.1016/j.rser.2017.01.049.

[159] Li MJ, Zhao W, Chen X, Tao WQ. Economic analysis of a new class of vanadium redox-flow battery for medium- and large-scale energy storage in commercial applications with renewable energy. Appl Therm Eng

REFERENCES| 109

2017;114:802–14. doi:10.1016/j.applthermaleng.2016.11.156. [160] Abdon A, Zhang X, Parra D, Patel MK, Bauer C, Worlitschek J. Techno-

economic and environmental assessment of stationary electricity storage technologies for different time scales. Energy 2017;139:1173–87. doi:10.1016/j.energy.2017.07.097.

[161] Wang D, Coignard J, Zeng T, Zhang C, Saxena S. Quantifying electric vehicle battery degradation from driving vs. vehicle-to-grid services. J Power Sources 2016;332:193–203. doi:10.1016/j.jpowsour.2016.09.116.

[162] EDA. Relatório e Contas. 2016. [163] Käbitz S, Gerschler JB, Ecker M, Yurdagel Y, Emmermacher B, André D, et

al. Cycle and calendar life study of a graphite|LiNi1/3Mn1/3Co1/3O2Li-ion high energy system. Part A: Full cell characterization. J Power Sources 2013;239:572–83. doi:10.1016/j.jpowsour.2013.03.045.

[164] Sarasketa-Zabala E, Martinez-Laserna E, Berecibar M, Gandiaga I, Rodriguez-Martinez LM, Villarreal I. Realistic lifetime prediction approach for Li-ion batteries. Appl Energy 2016;162:839–52. doi:10.1016/j.apenergy.2015.10.115.

[165] Petit M, Prada E, Sauvant-Moynot V. Development of an empirical aging model for Li-ion batteries and application to assess the impact of Vehicle-to-Grid strategies on battery lifetime. Appl Energy 2016;172:398–407. doi:10.1016/j.apenergy.2016.03.119.

[166] Few S, Schmidt O, Offer GJ, Brandon N, Nelson J, Gambhir A. Prospective improvements in cost and cycle life of off-grid lithium-ion battery packs: An analysis informed by expert elicitations. Energy Policy 2018;114:578–90. doi:10.1016/j.enpol.2017.12.033.

[167] Nykvist B, Nilsson M. Rapidly falling costs of battery packs for electric vehicles. Nat Clim Chang 2015;5:329–32. doi:10.1038/nclimate2564.

[168] Berckmans G, Messagie M, Smekens J, Omar N, Vanhaverbeke L, Mierlo J Van. Cost projection of state of the art lithium-ion batteries for electric vehicles up to 2030. Energies 2017;10. doi:10.3390/en10091314.

[169] Hansen JB. Process for converting biogas to a gas rich in methane. WO2012003849 A1, 2012.

[170] Kemp I. Pinch Analysis and Process Integration. 2nd ed. Oxford: 2007. [171] Lorenzi G, Lanzini A, Santarelli M, Martin A. Exergo-economic analysis of

a direct biogas upgrading process to synthetic natural gas via integrated high-temperature electrolysis and methanation. Energy 2017;141. doi:10.1016/j.energy.2017.11.080.

[172] Rostrup-Nielsen J. Sulfur-passivated nickel catalysts for carbon-free steam reforming of methane. J Catal 1984;85:31–43. doi:10.1016/0021-9517(84)90107-6.

[173] Lanzini A, Madi H, Chiodo V, Papurello D, Maisano S, Santarelli M, et al. Dealing with fuel contaminants in biogas-fed solid oxide fuel cell (SOFC) and molten carbonate fuel cell (MCFC) plants: Degradation of catalytic and electro-catalytic active surfaces and related gas purification methods. Prog Energy Combust Sci 2017;61:150–88. doi:10.1016/j.pecs.2017.04.002.

[174] Rostrup-Nielsen JR. Catalytic Steam Reforming. 1984. [175] Rostrup-Nielsen JR, Hansen JB, Helveg S, Christiansen N, Jannasch a.-K.

Sites for catalysis and electrochemistry in solid oxide fuel cell (SOFC) anode. Appl Phys A 2006;85:427–30. doi:10.1007/s00339-006-3702-1.

110 | REFERENCES

[176] AspenTech Products n.d. https://www.aspentech.com/products/aspen-plus.aspx (accessed January 1, 2017).

[177] Gandiglio M, Lanzini A, Leone P, Santarelli M, Borchiellini R. Thermoeconomic analysis of large solid oxide fuel cell plants: Atmospheric vs. pressurized performance. Energy 2013;55:142–55. doi:10.1016/j.energy.2013.03.059.

[178] Moran, Michael J, Shapiro HN. Fundamentals of Engineering Thermodynamics. 5th ed. John Wiley & Sons, Inc; 2006.

[179] Mogensen MB. Sustainable fuels from renewable energy. Seminar activity in Politecnico di Torino, Torino, Italy, 21/05/2013 2013.

[180] Giglio E, Lanzini A, Santarelli M, Leone P. Synthetic natural gas via integrated high-temperature electrolysis and methanation: Part II-Economic analysis. J Energy Storage 2015;2:64–79. doi:10.1016/j.est.2015.06.004.

[181] Cost Estimation Methodology for NETL Assessments of Power Plant Performance. 2011.

[182] Pertl A, Mostbauer P, Obersteiner G. Climate balance of biogas upgrading systems. Waste Manag 2010;30:92–9. doi:10.1016/j.wasman.2009.08.011.

[183] Lorenzi G, Lanzini A, Santarelli M. Digester gas upgrading to synthetic natural gas in solid oxide electrolysis cells. Energy and Fuels 2015;29. doi:10.1021/ef5023779.

[184] Distribution components of the electricity tariff. Ital Auth Electr Energy, Gas Water Syst 2016. http://www.autorita.energia.it/it/elettricita/distr.htm (accessed July 18, 2016).

[185] Curletti F, Gandiglio M, Lanzini A, Santarelli M, Maréchal F. Large size biogas-fed Solid Oxide Fuel Cell power plants with carbon dioxide management: Technical and economic optimization. J Power Sources 2015;294:669–90. doi:10.1016/j.jpowsour.2015.06.091.

[186] Papadias DD, Ahmed S, Kumar R. Fuel quality issues with biogas energy - An economic analysis for a stationary fuel cell system. Energy 2012;44:257–77. doi:10.1016/j.energy.2012.06.031.

[187] Turton R. Analysis, Synthesis and Design of Chemical Processes. Third Edit. Prentice Hall; n.d.

[188] Jensen SH, Larsen PH, Mogensen M. Hydrogen and synthetic fuel production from renewable energy sources. Int J Hydrogen Energy 2007;32:3253–7. doi:10.1016/j.ijhydene.2007.04.042.

[189] Fu Q, Mabilat C, Zahid M, Brisse A, Gautier L. Syngas production via high-temperature steam/CO2 co-electrolysis: an economic assessment. Energy Environ Sci 2010;3:1382. doi:10.1039/c0ee00092b.

[190] Luckow P. Spring 2016 National Carbon Dioxide Price Forecast. 2016. [191] Projected Costs of Generating Electricity - 2015 Edition. 2015.

doi:http://dx.doi.org/10.1787/cost_electricity-2015-en. [192] Estermann T, Newborough M, Sterner M. Power-to-gas systems for

absorbing excess solar power in electricity distribution networks. Int J Hydrogen Energy 2016:1–10. doi:10.1016/j.ijhydene.2016.05.278.

[193] GSE. Procedure applicative DM 2 maro 2018 (document in Italian). 2018. [194] IEA. Renewables 2017. 2017.

REFERENCES| 111

[195] European Parliament. Circular economy: More recycling of household waste, less landfilling 2018. http://www.europarl.europa.eu/news/en/press-room/20180411IPR01518/circular-economy-more-recycling-of-household-waste-less-landfilling (accessed July 25, 2018).

[196] European Commission. 2050 low-carbon economy 2018. https://ec.europa.eu/clima/policies/strategies/2050_en (accessed July 25, 2018).

[197] Boden TA., Marland G., Andres RJ. National CO2 Emissions from Fossil-Fuel Burning, Cement Manufacture, and Gas Flaring: 1751-2014. 2017. doi:10.3334/CDIAC/00001_V2017.

[198] Lajunen A. Lifecycle costs and charging requirements of electric buses with different charging methods. J Clean Prod 2018;172:56–67. doi:10.1016/j.jclepro.2017.10.066.

[199] Bailera M, Lisbona P, Romeo LM, Espatolero S. Power to Gas projects review: Lab, pilot and demo plants for storing renewable energy and CO2. Renew Sustain Energy Rev 2017;69:292–312. doi:10.1016/j.rser.2016.11.130.

[200] Larscheid P, Lück L, Moser A. Potential of new business models for grid integrated water electrolysis. Renew Energy 2018;125:599–608. doi:10.1016/j.renene.2018.02.074.

[201] Valero A, Lozano M, Munoz M. A general theory of exergy saving, parts I, II and III. Comput Eng Energy Syst 1986;3:1–22.

[202] Tsatsaronis G, Winhold M. Exergoeconomic analysis and evaluation of energy-conversion plants — I. A new general methodology. Energy 1985;10:69–80. doi:10.1016/0360-5442(85)90020-9.

[203] Scopus n.d. www.scopus.com. [204] Decreto-Lei n.o 118/2013 (Decree-Law on the improvements of building

energy performance - Document in Portuguese) n.d. [205] SREA. Azores Statistics 2017. http://srea.azores.gov.pt/ (accessed January

20, 2018). [206] Hansen JB. Production of Sustainable Fuels by Means of Solid Oxide

Electrolysis. ECS Trans. Vol. 35, issue 1, 2011, p. 2941–8. [207] Reichhalter H, Bozzo A, Dal Savio S, Guerra T. Analisi energetica ,

ambientale ed economica di impianti a biogas in Provincia di Bolzano - Relazione conclusiva [document in Italian]. 2011.

[208] Eurostat. Gas prices for non-household consumers - bi-annual data 2017. [209] Eurostat. Electricity price for non-household consumers [nrg_pc_205]

2018.

112 | REFERENCES

APPENDIX A – EXERGO-ECONOMIC ANALYSIS| 113

Appendix A – Exergo-Economic Analysis

The exergo-economic analysis (EEA, also known as thermo-economic

analysis) is a methodology developed by Losano, Valero and Munoz [201]

which has the aim of attributing both an exergetic and an economic value

to each substance, according to its thermodynamic value.

The key concept is the exergetic cost which is the exergy that must be

spent to obtain a certain product. It is fundamentally different from the

exergy content of the same product as explained in [201]: “[…] if we were

to produce certain chemical substances from their elements, their

exergetic cost would be very high due to the inefficiency of current

processes. However, if these some substances are used as fuel, we cannot

get more than their exergy, and the maximum possible saving will be

given by the lost exergy which can be recovered in the process”. However,

the production cost does not determine the value of a substance, which is

linked with the purpose for which they are used.

The EEA methodology provides the tools to determine the cost of each

stream in a process, based on the exergy content and cost of the resources

and the cost for the components of the process. It is structured as an

integrated exergy and economic analysis which is based on the definition

of a productive structure for each subcomponent of the process under

analysis (Figure 27). Resources and Products can be mass or energy

streams as well as the difference of two streams. The definition of Resource

and Product is done through the definition of the Incidence Matrix (A)

which contains the information about the productive structure of the

process and specifies the exergetic cost balance [201].

Sub-comp.Resource Product

Figure 27: Productive structure of a generic subcomponent.

114 | APPENDIX A – EXERGO-ECONOMIC ANALYSIS

The described procedure is characterized by the following steps [202]:

a) Detailed exergy balance of the plant for the calculation of the

Resource and Product values, with reference to a dead state (zero-

exergy state). This step includes the determination of the exergetic

costs and the definition of the incidence matrix (A).

b) Comprehensive annual cost estimation for each component,

including investment, operation and maintenance and other

overheads. Definition of the exergo economic costs of the

components of the process and of the exergy flows entering the

process (all expressed in €/s). Each of these costs is a component

of the vector 𝑍.

c) Integration of the first two steps as shown in Equation ( A1 ).

𝐴 ∗ 𝐶 = 𝑍 ( A1 )

where C is the cost vector (€/s) containing the exergo-economic costs of

each mass/energy stream. The incidence matrix (step a) is coupled with

the cost estimation (step b) and the solution of the system (step c) provides

the cost distribution along the process, highlighting if the cost increase of

a stream is due to the exergy loss or to the capital expenditure for the

specific component.

APPENDIX B - PUBLICATIONS| 115

Appendix B - Publications

Paper I: “Techno-economic comparison of storage vs. demand response

strategies in distributed generation systems”, Proceeding of

the 12th International Conference on the European Energy

Market (Lisbon 2015).

Paper II: “Comparing demand response and battery storage to optimize

self-consumption in PV systems”, Applied Energy, 2016, 180,

524-535.

Paper III: “Techno-economic analysis of utility-scale energy storage in

island settings”, (Submitted to Journal of Energy Storage)

Paper IV: “Digester Gas Upgrading to Synthetic Natural Gas in Solid

Oxide Electrolysis Cells”, Energy & Fuels, 2015 29 (3), 1641-

1652

Paper V: “Exergo-economic analysis of direct biogas upgrading process

to synthetic natural gas via integrated high-temperature

electrolysis and methanation”, Energy, 2017 141, 1524-1537

Techno-economic analysis and optimization of ... › smash › get › diva2:1247170 ›...

Documents

Transcript of Techno-economic analysis and optimization of ... › smash › get › diva2:1247170 ›...