ThesisPDF Available

A life cycle analysis of storage batteries for photovoltaic water pumping systems in Sub-Saharan remote areas

Authors:

Abstract and Figures

Photovoltaic Water Pumping Systems (PVWPS) have shown evidence to be a promising solution for providing water access in the rural communities of Sub-Saharan Africa (SSA). These systems are normally accompanied by water tanks as the storage technology in order to extend the utilisation of the system and provide water at any time. Previous techno-economic studies have demonstrated that using lead-acid batteries as the storage technology could minimise the Life Cycle Cost (LCC) of the system. However, utilising batteries for these systems which are located in isolated areas could be a concern from an environmental point of view. Thus, this paper will carry out an environmental impact assessment amongst the different storage technologies for PVWPS. The four selected technologies are: steel water tank, plastic water tank, lead-acid battery and lithium-ion battery. Life Cycle Analysis (LCA) is the most common tool for environmental impact assessments and it has been the selected approach for this thesis. The present study has used the PVWPS installed in the village of Gogma (Burkina Faso) as case study. The analysis has been carried out in the three endpoint categories: human health, ecosystems and resources, and five selected midpoint categories: global warming, acidification, human toxicity, water eutrophication and mineral resource scarcity. The results show that for the PVWPS in Gogma, lithium-ion batteries and plastic water tank are the two most sustainable solutions, with a Global Warming Potential (GWP) of 365 and 780 kg CO2-eq, respectively. Nevertheless, when interpreting the results on a kg basis, the steel water tank presents the lowest environmental impact. Based on the obtained results, a methodology has been developed to easily estimate the GWP of the four storage technologies with a given life cycle without the need of been familiarised with the LCA approach. Beyond results, a discussion about the viability of the systems has been performed. Besides technical, economic and environmental aspects, a system must be practical and feasible. Depending on the storage technology, different Operation & Maintenance (O&M) activities are required. While water tanks need no replacements, lithium-ion batteries need one replacement and lead-acid batteries need four replacements. In light of this, it is essential that a proper maintenance network is arranged if a battery-PVWPS is installed.
Content may be subject to copyright.
MASTER THESIS
A life cycle analysis of storage batteries
for photovoltaic water pumping systems
in Sub-Saharan remote areas
Student
Fatima Hermosin Acasuso
1st September 2021
Supervisor
Dr. Judith A. Cherni
Co-supervisors
Dr. Vincent Reinbold
Dr. Simon Meunier
Dr. Loïc Quéval
A thesis presented to Imperial College London in partial fulfilment of the requirements for
the degree of Master of Science in Sustainable Energy Futures and the
Diploma of Imperial College
Energy Futures Lab
Imperial College London
SW7 2AZ
ii
Abstract
Photovoltaic Water Pumping Systems (PVWPS) have shown evidence to be a promising
solution for providing water access in the rural communities of Sub-Saharan Africa (SSA) [1].
These systems are normally accompanied by water tanks as the storage technology in order to
extend the utilisation of the system and provide water at any time. Previous techno-economic
studies [2] have demonstrated that using lead-acid batteries as the storage technology could
minimise the Life Cycle Cost (LCC) of the system.
However, utilising batteries for these systems which are located in isolated areas could be a
concern from an environmental point of view. Thus, this paper will carry out an environmental
impact assessment amongst the different storage technologies for PVWPS. The four selected
technologies are: steel water tank, plastic water tank, lead-acid battery and lithium-ion battery.
Life Cycle Analysis (LCA) is the most common tool for environmental impact assessments
and it has been the selected approach for this thesis.
The present study has used the PVWPS installed in the village of Gogma (Burkina Faso) as
case study. The analysis has been carried out in the three endpoint categories: human health,
ecosystems and resources, and five selected midpoint categories: global warming,
acidification, human toxicity, water eutrophication and mineral resource scarcity. The results
show that for the PVWPS in Gogma, lithium-ion batteries and plastic water tank are the two
most sustainable solutions, with a Global Warming Potential (GWP) of 365 and 780 kg CO2-
eq, respectively. Nevertheless, when interpreting the results on a kg basis, the steel water tank
presents the lowest environmental impact. Based on the obtained results, a methodology has
been developed to easily estimate the GWP of the four storage technologies with a given life
cycle without the need of been familiarised with the LCA approach.
Beyond results, a discussion about the viability of the systems has been performed. Besides
technical, economic and environmental aspects, a system must be practical and feasible.
Depending on the storage technology, different Operation & Maintenance (O&M) activities
are required. While water tanks need no replacements, lithium-ion batteries need one
replacement and lead-acid batteries need four replacements. In light of this, it is essential that
a proper maintenance network is arranged if a battery-PVWPS is installed.
iii
Acknowledgements
I would like to thank Dr. Judith Cherni, my first supervisor, for her professionalism, constant
support and guidance, especially in the cultural and environmental aspects of the thesis. I would
also like to thank my three French co-supervisors: Dr. Vincent, Dr .Simon and Dr. Loïc. Dr.
Vincent, thanks for your patience and availability during every weekly meeting. Thanks for
guiding me throughout the year and always encouraging me to give my best. Dr. Simon, thanks
for your constructive feedback, continuous advices and for transmitting me your enthusiasm
and motivation like if this project was yours. Dr. Loïc, thanks for always providing a new and
different point of view and for your helpful and valuable feedback during the year.
Thanks one more time to the four of you, for giving me the amazing opportunity of developing
this study and get closer to the harsh reality in which many people live in other parts of the
planet. I hope I have contributed to this beautiful project and hopefully one day I will be lucky
enough to visit Burkina Faso and experience it in person.
I would also like to thank my group of SEF friends. Blanca, my housemate for her daily support.
Ana, Arancha and Guille, my “Spanish gang”, for their continuous laughs and endless dancing
nights. Damien, Frapaul and Arthur, my “French athletes”, for our weekly runs in Hyde Park.
Finally, thanks to all the SEFers for making of this pandemic year such an unforgeable
experience. I look forward to seeing you in the near future succeeding in all the crazy plans
you had in mind!
Last but not least, thanks to the special and unique support received from my family during this
tough year. Thanks mum and dad for teaching me that things require effort, constancy and
discipline. Thanks Rafa for your patience and for believing in me more than I do.
iv
Table of Contents
Abstract .....................................................................................................................................ii
Acknowledgements ................................................................................................................ iii
Table of Contents .................................................................................................................... iv
List of Figures .........................................................................................................................vii
List of Tables ........................................................................................................................ viii
List of Abbreviations .............................................................................................................. ix
1. Introduction ...................................................................................................................... 1
2. Literature Review ............................................................................................................ 3
2.1. Water Access in Rural Areas of Sub-Saharan Africa ................................................ 3
2.1.1. Water and Electricity Access Worldwide and in SSA ........................................... 3
2.1.2. Progress of Water Extraction Approaches ............................................................. 4
2.1.3. Storage Technologies for PVWPS ......................................................................... 5
2.1.4. Advantages & Challenges of Water Tanks and Battery Storage for PVWPS ....... 7
2.2. Battery Technologies ................................................................................................. 9
2.2.1. Main Battery Technologies Available ................................................................... 9
2.2.2. Comparison of Battery Technologies in regards to Their Applications .............. 11
2.2.3. Environmental Performance and Challenges of Batteries ................................... 12
2.2.4. Battery Market Landscape and Outlook for 2040................................................ 13
2.3. Environmental Impact Assessment of Storage Technologies for PVWPS .............. 15
2.3.1. Introduction to LCA ............................................................................................. 15
2.3.2. Review of Studies ................................................................................................ 20
2.4. Lithium-Ion and Lead-Acid Batteries Life Cycle Overview ................................... 22
2.4.1. Operating Principle & Cell Architectures ............................................................ 22
2.4.2. Overview of Full Life Cycle ................................................................................ 23
2.4.3. Material Extraction & Material Processing ......................................................... 23
2.4.4. Transport .............................................................................................................. 25
2.4.5. Manufacturing ...................................................................................................... 25
2.4.6. Use Phase ............................................................................................................. 27
2.4.7. End-of-life ............................................................................................................ 29
2.5. Research Gaps & Objectives ................................................................................... 34
3. Methodology Design....................................................................................................... 35
3.1. Introduction .............................................................................................................. 35
3.1.1. Our Case Study .................................................................................................... 35
3.2. Overview of Methods .............................................................................................. 36
3.2.1. Literature Review................................................................................................. 36
v
3.2.2. Criteria Definition ................................................................................................ 36
3.2.3. LCA Modelling of Storage Technologies for PVWPS ........................................ 37
3.2.4. Scenario Approach ............................................................................................... 37
3.2.5. Results Representation ......................................................................................... 37
3.3. Data Collection ........................................................................................................ 37
3.4. Analysis of Results .................................................................................................. 38
3.5. Assumptions for LCA Modelling ............................................................................ 38
3.6. Electricity Mix ......................................................................................................... 39
3.6.1. Carbon Intensity ................................................................................................... 39
3.6.2. Electricity Mix of the Three Main Nations .......................................................... 39
4. Model Development ....................................................................................................... 41
4.1. Selection of Criteria for LCA of Storage Technologies for PVWPS ...................... 41
4.1.1. Functional Unit Definition ................................................................................... 41
4.1.2. System Boundaries Definition ............................................................................. 42
4.1.3. Impact Assessment Method Selection ................................................................. 43
4.2. Scenario Definition for LCA of Storage Technologies for PVWPS ....................... 44
4.2.1. Steel Water Tank Scenarios ................................................................................. 44
4.2.2. Plastic Water Tank Scenarios .............................................................................. 45
4.2.3. Lead Acid Battery Scenarios ............................................................................... 45
4.2.4. Lithium Ion Battery Scenarios ............................................................................. 47
4.3. Selected Information for LCA Modelling per Storage Technology ........................ 47
4.3.1. Selected Information for Steel Water Tank ......................................................... 48
4.3.2. Selected Information for Plastic Water Tank....................................................... 49
4.3.3. Selected Information for Lead Acid Battery ........................................................ 50
4.3.4. Selected Information for Lithium Ion Battery ..................................................... 51
5. Results ............................................................................................................................. 54
5.1. Results of the Environmental Impact per Storage Technology in relation to
Endpoint Indicators .............................................................................................................. 54
5.1.1. Steel Water Tank Endpoint Results ..................................................................... 54
5.1.2. Plastic Water Tank Endpoint Results .................................................................. 55
5.1.3. Lead Acid Batteries Endpoint Results ................................................................. 56
5.1.4. Lithium Ion Batteries Endpoint Results ............................................................... 56
5.2. Comparison of the Environmental Impact between Water Tanks and Batteries in
relation to Midpoint Indicators ............................................................................................ 57
5.2.1. Global Warming Potential of the Six Selected Scenarios .................................... 58
5.2.2. Terrestrial Acidification Potential of the Six Selected Scenarios ........................ 58
5.2.3. Human Carcinogenic Toxicity Potential of the Six Selected Scenarios .............. 59
vi
5.2.4. Freshwater Eutrophication of the Six Selected Scenarios ................................... 60
5.2.5. Mineral Resource Scarcity Potential of the Six Selected Scenarios .................... 60
5.2.6. Normalisation of Midpoint Categories ................................................................ 61
5.3. Sensitivity Analysis of the Environmental Impact per Storage Technology in
relation to Three Indicators .................................................................................................. 62
5.3.1. Sensitivity Analysis of Electricity Mix ................................................................ 63
5.3.2. Sensitivity Analysis of Lorry Transport .............................................................. 63
5.3.3. Sensitivity Analysis of Aircraft Transport ........................................................... 64
5.4. Environmental Coefficients ..................................................................................... 65
6. Discussion of Results ...................................................................................................... 66
6.1. Environmental Impact of Storage Technologies for PVWPS .................................. 66
6.2. Environmental Impact of Storage Technologies on a kg basis ................................ 68
6.3. Comparison of Results in regards to the Selected Functional Unit ......................... 69
6.4. Viability of Storage Technologies for PVWPS ....................................................... 70
6.4.1. Water Tank Viability ........................................................................................... 70
6.4.2. Battery Viability................................................................................................... 71
7. Conclusion ...................................................................................................................... 73
7.1. Future work .............................................................................................................. 74
Bibliography ........................................................................................................................... 76
Appendices .............................................................................................................................. 87
Appendix A .......................................................................................................................... 87
A.1. Life Cycle Inventory of Steel Water Tank ............................................................... 87
A.2. Life Cycle Inventory of Plastic Water Tank ............................................................ 88
A.3. Life Cycle Inventory of Lead-acid Battery .............................................................. 89
A.4. Life Cycle Inventory of Lithium-ion Battery ........................................................... 90
Appendix B .......................................................................................................................... 91
B.1. Global Warming Potential Results per Storage Technology per Phase ................... 91
B.2. Terrestrial Acidification Potential Results per Storage Technology per Phase ....... 91
B.3. Human Carcinogenic Toxicity Potential Results per Storage Technology per Phase
91
B.4. Freshwater Eutrophication Potential Results per Storage Technology per Phase .. 92
B.5. Mineral Resource Scarcity Potential Results per Storage Technology per Phase ... 92
Appendix C .......................................................................................................................... 92
C.1. Environmental Coefficients of Electricity Mix for Steel Water Tank ..................... 92
C.2. Environmental Coefficients of Electricity Mix for Plastic Water Tank................... 93
C.3. Environmental Coefficients of Electricity Mix for Lithium-ion Battery ................. 93
vii
List of Figures
Figure 1. Global Electricity access by percentage of population [6] ......................................... 4
Figure 2. PVWPS architecture with water tank storage [7] ....................................................... 6
Figure 3. PVWPS architecture with battery bank storage [27] .................................................. 6
Figure 4. Ragone Plot of Energy Storage Devices [49] ........................................................... 11
Figure 5. Global Cumulative storage deployments [57] .......................................................... 14
Figure 6. Life Cycle Assessment Framework .......................................................................... 16
Figure 7. Standard unit process [37] ........................................................................................ 17
Figure 8. Lead-Acid Battery Architecture [146]...................................................................... 22
Figure 9. Lithium-Ion Cylindrical Cell Architecture [147] ..................................................... 22
Figure 10. Full Life Cycle flow chart for lead-acid and lithium-ion batteries adapted from
[54] ........................................................................................................................................... 23
Figure 11. Flow Diagram for Lead-Acid Battery Manufacturing Process adapted from [87]. 26
Figure 12. Flow Diagram for Lithium-Ion Battery Manufacturing Process adapted from [57]
.................................................................................................................................................. 27
Figure 13. Relation between battery capacity, DoD and cyclic life for a lead-acid battery
[143] ......................................................................................................................................... 28
Figure 14. Flow chart of hydrometallurgical and pyrometallurgical LAB recycling processes
adapted from [100] ................................................................................................................... 31
Figure 15. Flow chart of pyrometallurgical LIB recycling process adapted from [84] ........... 32
Figure 16. Flow chart of hydrometallurgical LIB recycling process adapted from [84] ......... 33
Figure 17. Flow chart of direct LIB recycling process adapted from [84] .............................. 33
Figure 18. PVWPS installed in the village of Gogma ............................................................. 36
Figure 19. Electricity Mix of the Three Main Nations Used in the LCA Study ...................... 40
Figure 20. Overview of ReCiPe's structure adapted from [120] .............................................. 43
Figure 21. Steel Water Tank Scenarios.................................................................................... 45
Figure 22. Plastic Water Tank Scenarios ................................................................................. 45
Figure 23. Lead-Acid and Lithium-ion Battery Landfill Scenarios ......................................... 46
Figure 24. Lead-Acid Battery Recycling Scenarios ................................................................ 46
Figure 25. Lithium-ion Battery Recycling Scenarios .............................................................. 47
Figure 26. Endpoint Results for Steel Water Tank Scenarios ................................................. 55
Figure 27. Endpoint Results for Plastic Water Tank Scenarios ............................................... 55
Figure 28. Endpoint Results for Lead-acid Battery Scenarios ................................................. 56
Figure 29. Endpoint Results for Lithium-ion Battery Scenarios ............................................. 57
Figure 30. Global Warming Results in the Six Selected Scenarios ......................................... 58
Figure 31. Terrestrial Acidification Results in the Six Selected Scenarios ............................. 59
Figure 32. Human Toxicity Results in the Six Selected Scenarios.......................................... 59
Figure 33. Freshwater Eutrophication Results in the Six Selected Scenarios ......................... 60
viii
Figure 34. Mineral Resource Depletion Results in the Six Selected Scenarios....................... 61
Figure 35. Midpoint Normalisation Results in the Six Selected Scenarios ............................. 62
Figure 36. Sensitivity Analysis of Electricity Mix for the Four Selected Storage Technologies
.................................................................................................................................................. 63
Figure 37. Sensitivity Analysis of Lorry Transport for the Four Selected Storage
Technologies ............................................................................................................................ 64
Figure 38. Sensitivity Analysis of Aircraft Transport for the Four Selected Storage
Technologies ............................................................................................................................ 64
Figure 39. Net Results in the Six Selected Scenarios for the Five Midpoint Categories ........ 67
Figure 40. Results (per kg) expressed in Percentages in the Six Selected Scenarios for the
Five Midpoint Categories ........................................................................................................ 69
List of Tables
Table 1. Key characteristics of selected battery technologies (1Cycle Life at 50% DoD) ...... 10
Table 2. Description of the main players of the battery industry ............................................. 15
Table 3. Selected impact categories for the LCA study [63], [65] .......................................... 18
Table 4. LCA Software tools comparison................................................................................ 19
Table 5. Technical parameters of the selected LCA studies .................................................... 21
Table 6. Components and Materials of a lead-acid battery [35], [47], [67], [78] .................... 24
Table 7. Components and Materials of a Lithium-ion battery [47], [65], [68], [81] ............... 24
Table 8. Carbon Intensity of the Nations Used throughout the Study ..................................... 39
Table 9. Functional Unit Definition for LABs and LIBs ......................................................... 42
Table 10. Overview of Endpoint categories adapted from [120] ............................................. 44
Table 11. Transport Phase Overview for the Steel Water Tank Scenarios .............................. 48
Table 12. Transport Phase Overview for the Plastic Water Tank Scenarios ........................... 49
Table 13. Transport Phase Overview for the Lead-acid Battery Scenarios ............................. 50
Table 14. Transport Phase Overview for the Lithium-ion Battery Scenarios .......................... 52
Table 15. Normalisation Results of Midpoint Categories in the six selected scenarios .......... 62
Table 16. Environmental Coefficients for the GWP of Lead-acid Battery per Phase ............. 65
Table 17. Net Results (per kg) in the Six Selected Scenarios for the Five Midpoint Categories
.................................................................................................................................................. 68
Table 18. LCI of Manufacturing Process of Steel Water Tank ............................................... 87
Table 19. LCI of Recycling Process of Steel Water Tank ....................................................... 88
Table 20. LCI of Manufacturing Process of Plastic Water Tank ............................................. 89
Table 21. LCI of Recycling Process of Plastic Water Tank .................................................... 89
Table 22. LCI of Manufacturing Process of Lead-acid Battery ............................................... 89
Table 23. LCI of Recycling Process of Lead-acid Battery ...................................................... 90
Table 24. LCI of Manufacturing Process of Lithium-ion Battery ........................................... 90
ix
Table 25. LCI of Recycling Process of Lithium-ion Battery ................................................... 91
Table 26. GWP Results per Storage Technology per Phase .................................................... 91
Table 27. TAP Results per Storage Technology per Phase ..................................................... 91
Table 28. HTP Results per Storage Technology per Phase ..................................................... 91
Table 29. FEP Results per Storage Technology per Phase ...................................................... 92
Table 30. MRSP Results per Storage Technology per Phase .................................................. 92
Table 31. Environmental Coefficients for the GWP of Steel Water Tank per Phase .............. 92
Table 32. Environmental Coefficients for the GWP of Plastic Water Tank per Phase ........... 93
Table 33. Environmental Coefficients for the GWP of Lithium-ion Battery per Phase .......... 93
List of Abbreviations
ADP
ALOP
AREI
CED
CTG
DRC
ETP
EV
FDP
FEP
FU
GWP
HREWPS
HTP
HTPc
HTPnc
IRP
KPI
LABs
LLDPE
LCC
LIBs
MDP
MPPT
MRSP
NFs
NLTP
x
NPC
O&M
ODP
PE
PE
PEU
PMFP
POFP
PVWPS
RES
RESWPS
SEFA
SHS
SLI
SSA
TAP
ULOP
UPS
WDP
WEWPS
Net Present Cost
Operation & Maintenance
Ozone Depletion Potential
Per Person
Polyethylene
Primary Energy Use
Particulate Matter Formation Potential
Photochemical Oxidant Formation Potential
Photovoltaic Water Pumping System
Renewable Energy Sources
Renewable Energy Source Water Pumping System
Sustainable Energy Fund for Africa
Solar Home Systems
Starting, Lighting and Ignition
Sub-Saharan Africa
Terrestrial Acidification Potential
Urban Land Occupation Potential
Uninterruptible Power Supply
Water Depletion Potential
Wind Energy Water Pumping System
1
1. Introduction
Access to a safe and sustainable source of water is a growing concern in many regions of the
world, and the COVID-19 crisis has highlighted the importance of water access for hygiene
and sanitation to protect human health. Still, 2.2 billion people are lacking safe drinking water
and 4.2 billion people are lacking safe water sanitation [3]. The United Nations has set the goal
of ensuring availability and sustainable management of water and sanitation for all for 2030,
but if immediate action is not taken, targets would not be met [3]. High correlations have been
identified where countries with lack of water access also suffer from lack of electricity access,
and these inter-dependencies are predicted to intensify [4]. Furthermore, high correlations have
also been identified between off-grid areas and high solar irradiation, where solar water
pumping becomes competitive compared to diesel pumping for water access [5]. PVWPS have
proven to be a operationally and financially sustainable solution to provide water access in
remote communities [1], [6].
PVWPS are normally accompanied by a storage technology, usually water tanks, to improve
performance and reliability [5]. However, other storage technologies, such as batteries could
be used for PVWPS [7]. Indeed, [2] has demonstrated with a techno-economic analysis that
battery-PVWPS have a lower LCC compared to tank-PVWPS. This finding raises a new
opportunity to make PVWPS more economically affordable with a storage technology that has
not been fully explored yet. Nevertheless, no study has previously investigated the
environmental impact of these storage technologies, which must be considered in line with
technical, social and economic criteria.
This thesis aims to quantify the environmental impact of storage technologies for PVWPS and
identify the technology with the best environmental performance. Furthermore, a methodology
will be developed to facilitate the integration of these results into future studies and quickly
examine the environmental impact of a specific technology in conjunction with other
parameters. This study will use the PVWPS installed in the village of Gogma, located in the
centre-east of Burkina Faso as case study. This PVWPS provides water to the 250 inhabitants
of the village. Considering the environmental impact of PVWPS prior to their installation is
essential to ensure a sustainable and clean future for these rural communities.
In Chapter 2 a literature review about the different architectures for PVWPS and the
environmental impact of batteries is carried out. Previous studies on LCA are reviewed and it
is identified that environmental impact of batteries in the African continent and their end-of-
life phase is not explored enough.
2
Chapter 3 describes the methodology utilised for this study, from criteria definition and LCA
modelling to analysis of results. In Chapter 4 the methodology described in Chapter 3 is applied
to our case study in Gogma. The criteria used for the LCA model is outlined (see Section 4.1),
then the different scenarios are defined (see Section 4.2) and finally the information that has
been selected for our LCA model is explained (see Section 4.3).
Chapter 5 presents the environmental impact results of the storage technologies for the PVWPS
in Gogma. These have been analysed in endpoint (see Section 5.1) and midpoint (see Section
5.2) indicators. Normalisation results (see Section 5.2.6) and sensitivity analysis for electricity
mix and transport (see Section 5.3) are also included.
Finally, Chapter 6 presents a discussion of the results where the two analysed functional units
are debated, for the PVWPS in Burkina Faso (see Section 6.1) and on a kg basis (see Section
6.2). A discussion about the viability of both architectures is also carried out (see Section 6.4).
3
2. Literature Review
2.1. Water Access in Rural Areas of Sub-Saharan Africa
This first section starts with a description of current water and electricity access in SSA.
Conventional water pumping systems are explained as well as the challenges they involve.
Considering that Africa has one of the largest renewable energy potentials and which has not
yet been fully exploited, Renewable Energy Water Pumping Systems are discussed. Due to the
suitability that photovoltaic based systems exhibit in SSA, this will be the architecture of focus
of this thesis. The two main storage technologies for PVWPS, water tanks and battery banks
will be summarised. The section concludes with a comparison about the key advantages and
challenges that each architecture presents, highlighting the potential of battery storage to
significantly reduce the cost of water supply.
2.1.1. Water and Electricity Access Worldwide and in SSA
Domestic Water Access in SSA
Water scarcity in Africa is still becoming an important concern. Population in SSA is expected
to increase from 345 million in 2014 to 1.3 billion in 2050, and it has been pointed out that this
rapid urbanisation rate has not been accompanied by an adequate economic growth [8]. As
stated by [9], the water source can be considered improved if: (i) it is accessible to everyone,
(ii) water is available when needed, and (iii) water is free from contamination.
According to [10], 663 million people around the globe do not have access to improved
drinking water sources, of which eight out of ten live in rural areas. Furthermore, 2.4 billion
people do not have access to water sanitation facilities, of which seven out of ten live in rural
areas. Most of the countries with the lowest improvements are in SSA [10]. Access to improved
water sources in rural areas in Africa increased from 31% in 1990 to 63% in 2015 [11].
However, the number of people with unimproved water sources is still considerably high.
Unsafe water is responsible for 1.2 million deaths each year, mainly correlated with the
diarrheal diseases generated from drinking water from unimproved water sources [12].
Electricity Access in SSA
Percentages of electricity access all over the globe are represented in Figure 1. It can be
appreciated how majority of continents have 75% of their population with access to electricity,
while the African continent is the only one below 54%. While the average electricity access in
SSA is 47.7%, the average in Burkina Faso is only 14.4%, being the third lowest country after
Burundi and Chad [13]. Electricity consumption in Africa has risen from 286 TWh in 1990 to
723 TWh in 2018. However, the African continent still has a heavy fossil fuel based economy,
having an electricity mix of 30% coal, 10% oil and 40% natural gas [14].
4
SSA shows a large renewable energy potential which differs across the different geographic
areas, having solar resources everywhere, wind resources mainly in the North, South and East,
and biomass and hydropower in wet areas, mainly in southern regions. African energy
resources are amongst the best in the world. Thus effective policy measures are crucial to
ensure fast deployment of these technologies [15]. This is why the African Renewable Energy
Initiative (AREI) was created in 2015, which aims to foster and accelerate the African energy
transition by setting the goal of installing 300 GW of renewable energy sources (RES) by 2030.
2.1.2. Progress of Water Extraction Approaches
Conventional Water Pumping Systems
Because of the increasing effects of climate change and consequent uncertainty of freshwater
resources, groundwater reserves provide high quality water and have become an important
shield against climate variability. Current groundwater reserves in Africa are estimated to be
20 times larger than water accumulated in lakes or surface water reservoirs [16]. Key sectors
such as agriculture and industry rely heavily on water access. Proper management of water
resources can contribute to significant economic growth and consequently, human health and
productivity improvements [17], [18]. Groundwater reserves are usually accessed by boreholes
(drilled by a machine and with small diameters) and water is extracted either by hand-pumps
or motor-pumps.
Hand-pumps can be purchased or locally built, representing the most economical solution.
Even so, they require significant physical force for pumping water and regular maintenance
due to their moving parts [19]. Furthermore, hand-pump corrosion could have an impact on
water quality and incite groundwater pollution [20].
Figure 1. Global Electricity access by percentage of population [6]
5
On the other hand, motor-pumps require an energy source. Electricity is normally supplied by
a diesel generator or by the electric grid. Motor-pumps offer great advantages such as higher
pumping rates, reduced queuing times [9] and consequent contribution to social development
[21]. Nevertheless, high diesel prices and limited electricity access in SSA constitute the main
barriers to the expansion of this technology [22].
Suitability of Photovoltaic Water Pumping Systems
Water pumping systems powered by renewable energy sources will become progressively more
accessible to the African continent, due to the expected increase in the deployment of RES and
their independence from the electric grid. According to [7], the most common are powered
either by solar energy (PVWPS), wind energy (WEWPS) or a combination of both, known as
hybrid systems (HREWPS). While PVWPS require minimal maintenance, WEWPS require
periodic maintenance due to their multiple moving elements. In addition, PVWPS offer the
cheapest and simplest solution whereas HREWPS are the most expensive and complex solution
[23]. Therefore, PVWPS seems to be a suitable technology for the African continent [1], [5].
However, a detailed economic evaluation should be carried out for each location due to the
diversity of energy sources within the African continent.
PVWPS offer multiple benefits when compared to traditional diesel water pumping systems.
The initial investment of a PVWPS is 23% lower compared to a diesel water pumping system
[24], mainly driven by the significant decline in the price of PV panels [25]. PVWPS also offer
lower operation and maintenance (O&M) costs, due to no fuel dependence. Not using a diesel
generator has other benefits such as zero CO2 emissions during operation and noiseless
generation.
A PVWPS is made of a series of PV panels, a controller, a motor-pump and a fountain.
Depending on the system requirements, a storage technology to overcome intermittent water
supply can be incorporated. The two main available storage technologies are described and
compared in the following sections.
2.1.3. Storage Technologies for PVWPS
Because of their dependence on solar irradiation, PVWPS offer intermittent power. To avoid
the subsequent irregular water supply, and determined by the system requirements, a storage
device can be integrated. Storage-less PVWPS are lower in cost and require less maintenance,
but do not offer the key advantage of providing water during off sunshine hours [7]. These type
of systems are simpler and commonly used in smaller domestic or irrigation applications [7],
[23], [26]. The two storage technologies for PVWPS are summarised below: (i) water storage
with water tanks and (ii) electrochemical storage with batteries.
6
Water Tank
The architecture of a PVWPS with a water tank is illustrated in Figure 3. An example of this
architecture has been installed in the village of Gogma to provide domestic water access to 250
inhabitants. It is made of a PV array of 620 Wp and a water tank of 11.4 m3 [27], [28]. The
function of each component is summarised below [25]:
i. PV array: PV panels connected in series and/or in parallel to meet the current/voltage
requirements. They convert solar energy into DC electric energy .
ii. MPPT: electronic converter which optimises power extraction from the PV array.
iii. Motor-pump: normally consists of a submersible centrifugal pump driven by a DC
motor. Positive displacement pumps are another type of solar pumps which offer higher
hydraulic efficiencies but require more maintenance.
iv. Controller and float switch: monitors the water level in the tank.
v. Water Tank: stores the water and can be made of steel, PVC, concrete or masonry.
vi. Fountain: point of water collection.
vii. Pipes: path for the water to flow between motor-pump, water tank and fountain.
Battery
Figure 2 represents the schematic diagram of the battery architecture. During sunshine hours,
the motor-pump is powered by the PV array and simultaneously batteries get charged. During
off-sunshine hours, the motor-pump is powered by the energy stored in the batteries, and thus
allowing water to be extracted at any time [1]. An example of this architecture was installed by
[29] for an automated irrigation system in Saudi Arabia. The system consisted of two PV panels
of 32 Wp each, a 24V controller, and two 12V sealed LABs. The charge controller is
responsible for regulating the charging rate and/or depth of discharge of the batteries to use
them efficiently and expand batteries’ lifespan [23]. This architecture is not as common due to
the need of a more complex control system, increasing replacements and maintenance [30].
Figure 2. PVWPS architecture with water
tank storage [7]
Figure 3. PVWPS architecture with battery
bank storage [27]
7
2.1.4. Advantages & Challenges of Water Tanks and Battery Storage for PVWPS
In developing countries, economic factors are usually the main drivers when selecting one
architecture or another. Communities would normally choose the system which provides the
maximum benefit at the lowest cost [23]. As discussed by various authors [7], [25], [30] tank
storage is the most common architecture. They also defend that this architecture requires less
maintenance when compared to batteries. On the other hand, [23], [26] support the idea that
storage-less PVWPS offer the simplest and still reliable solution for applications like irrigation,
where irregular water supply is not a concern. Nevertheless, both of them agree that water
storage would be the 2nd simplest solution.
[31] carried out a techno-economic comparison of battery and tank storage by using the existing
PVWPS in Burkina Faso as case study. Data regarding irradiance, temperature, voltage,
current, water level in the borehole, and pumping and collected flow rates were collected [32].
The full Life Cycle Cost (LCC) of the battery architecture was found to be almost half of the
water tank architecture, being 6.4 k$ and 12.9 k$, respectively. The final PVWPS system
consisted of a battery bank of 2076 Wh and a PV array of 493 Wp. The higher LCC of the water
tank architecture can be justified by the high prices of steel water tanks in Burkina Faso, which
constitutes 59% of the total components costs, compared to 28% of the batteries [31]. Plastic
tanks can offer a cheaper solution, but since they are less reliable, steel tanks are usually the
selected choice [31]. Steel tanks also require maintenance due to the need for antirust
treatments [23]
Another important aspect is the social dimension. PVWPS can be satisfactory from a technical
point of view, because the given conditions and the sizing of the system are suitable to meet
the water demand. However, the feasibility of the system is also dependent on acceptability
from the community, as well as the ability and willingness to pay [23].
Another key finding from [31] which positively affects the battery architecture is that the rate
of collected water by the inhabitants for battery architectures is higher. The time to collect
water is 3.9 hours per day with battery architecture compared to 5.9 hours for tank architecture.
This relates to a 30% saving in time per inhabitant, which can also be translated to a positive
social impact [28]. Battery architectures also lead to smaller variations in the water level of the
borehole, contributing to the sustainability of the water resources and performance of the
motor-pump [31].
On the other hand, the battery architecture also presents multiple challenges. The main barrier
of batteries is their durability. The battery’s lifespan can be significantly reduced by
deregulated charge-discharge cycles and high ambient temperatures. Considering that
temperatures in Burkina Faso can reach up to 32C in cold months and 42C in hot months
8
[33], for a 20-year PVWPS, lead-acid batteries would require around six replacements [31] and
lithium-ion batteries one replacement [34]. A water tank architecture would require no tank
replacements [31]. This comparison in the number replacements required for each architecture
leads to O&M costs. As it could be expected, the battery architecture presents higher O&M
costs, representing 43% of the total LCC, and only 10% for the tank architecture [31].
From these findings, it can be concluded that due to the required replacements the performance
of battery-PVWPS is highly dependent on maintenance. Therefore, to provide a viable system
to these communities, O&M activities should be prioritised and investigated prior to
installation. It was raised by various assistants during the solar pumping workshop hold in
December 2020 [35], that there was a significant lack of maintenance of battery systems in
SSA, what made them not perform as expected, and thus decrease confidence from local
communities on this type of technologies. More information about what was discussed during
the workshop can be found in their YouTube channel [36].
The economic dimension also plays a crucial role during the decision-making process, and the
results on LCC between different storage technologies showed the potential of batteries to
significantly reduce the cost of water [2]. Another valuable opportunity that battery
architectures can offer if they get expanded across the continent, especially in rural areas with
no electricity access, is that they can be used for other purposes, such as lighting, and thus
contribute to the social and economic growth of these communities. It was pointed out by [37]
that due to the high cost of grid expansion, off-grid systems will contribute in a large amount
to the electrification of rural areas in developing countries. This will involve an increase in
battery storage and consequently, accelerate the development of the battery industry in Africa,
something very favourable for battery-PVWPS.
On the contrary, the environmental damage of batteries in SSA is a key concern, due to the
lack of regulation and recycling facilities across the continent [34]. It has been forecasted that
using batteries for PVWPS can be beneficial from a technical point of view [31], but before
these get installed, an environmental impact assessment should be carried out to quantify the
impacts that fitting batteries in these locations can have on human health as well as on the local
environment. Due to the environmental concerns associated with batteries [38] and the lack of
information about their environmental impact in remote areas such as the village of Gogma,
this thesis will aim to quantify their environmental burdens, as well as define the pathway that
can contribute to mitigate their impact.
9
2.2. Battery Technologies
This section gives an overview of the different battery technologies in the market, as well as
their common applications and technical performances. Environmental concerns associated
with Lead-Acid and Lithium-ion batteries will be outlined. Lastly, current and future battery
markets have been explored at a local and global scale, as well as the major manufacturers in
the battery industry.
2.2.1. Main Battery Technologies Available
Batteries can be split into two main types, primary and secondary, using primary for non-
rechargeable and secondary for rechargeable batteries. Only rechargeable batteries will be
considered. There are three main categories of rechargeable batteries in the market: Lead-Acid
(PbA), Lithium-ion (Li-ion) and Nickel (Ni). These categories also have sub-categories.
Lead-Acid Batteries (LABs)
This technology has been in the market for more than a century and it is the most used
rechargeable battery. There are two main types of LABs: flooded and sealed/valve regulated
(SLA/VRLA). The latter can also be split in Gel or AGM (Absorbed Glass Mat), which refers
to the state of the electrolyte in the battery. The main differences are that flooded batteries are
cheaper, require maintenance and have a lower specific energy compared to VRLA. Both offer
similar efficiencies and sensitivity to high temperatures [39]. LABs are characterised for their
low prices when compared to other technologies. Due to their ready availability in the African
continent [34], and [31]’s previous analysis, this thesis will focus on these type of batteries.
Lithium-Ion Batteries (LIBs)
They were introduced in the market in the late 90s and since then multiple chemistries have
been developed, which are mainly based on the composition of their cathodes, considering that
90% of Li-ion anodes are mainly made up of graphite [39]. Main chemistries are Lithium Iron
Phosphate (LiFePO4 - LFP), Lithium Nickel Manganese Cobalt Oxide (LiNiMnCoO2 - NCM),
Lithium Nickel Cobalt Aluminium Oxide (LiNiCoAlO2 - NCA), Lithium Cobalt Oxide
(LiCoO2 - LCO) and Lithium Manganese Oxide (LiMn2O4 LMO) [40]. Since the most
popular chemistries are LFP, NCM, and LMO, the majority of LCA studies cover these battery
chemistries, being 30% based on LFP, 24% on NCM and 16% on LMO [41]. Therefore, this
thesis will also focus on these three types.
Nickel Batteries (NIBs)
Nickel based batteries can also be split into two main sub-categories, Nickel-Cadmium (NiCd)
and Nickel Metal Hydride (NiMH). NiCd is one of the most mature technologies, in line with
LABs. They are slightly more expensive than LABs due to the high discharge rates and
10
temperature ranges they can be subjected to. However, since they use Cd for the anode which
is a toxic element, by the late 90s they were replaced by NiMH and nowadays they can offer
up to 40% higher specific energy compared to NiCd [42]. Since NIBs are commonly used in
applications where high power is required, such as power tools, LABs and LIBs are more
suitable for solar applications than NIBs [43]. Therefore, this thesis will be narrowed to the
first two battery technologies.
Battery Performance Indicators
The main properties that characterise batteries performance are as follows (Table 1) [44]:
i. Specific energy is the amount of energy a battery can deliver per unit weight.
ii. Open circuit voltage is the voltage when current is zero.
iii. Cycle life is the number of charge and discharge cycles delivered before the battery
fails.
iv. Depth of discharge (DoD) is the percentage of the battery capacity which has been
discharged relative to the total capacity of the battery.
v. Energy efficiency is the percentage of energy delivered (discharging) relative to the
energy used in charging the battery.
vi. Cost in US dollars per kWh of electricity delivered.
vii. Toxicity level in regards to the embedded materials.
Type
Technology
Specific
Energy
(Wh/kg)
Open Circuit
Voltage (V)
Cycle Life
(80% DoD)
Energy
Efficiency
(%)
Cost
($/kWh)
Toxicity
[36]
LAB
Flooded
30 [30]
2 [30]
1,200 [30]1
>80 [35]
120 [30]
High
VRLA
40 [30]
2
1,000 [30]1
>80 [35]
65
High
LIB
LFP
90 120 [36]
3.3 [36]
>2,000 [36]
85 95 [37]
580 [36]
Low
LMO
100 150 [36]
3.7 [36]
300 700
[36]
85 95 [37]
400 [31]
Low
NMC
150 220 [36]
3.7 [36, 30]
500 - 1,000
[36]
85 95 [37]
420 [36]
Low
NIB
NiCd
40 60 [35,
37]
1.3 [35, 37]
800 [35, 37]
75 [35, 37]
450 [31]
High
NiMH
75 95 [35,
37]
1.25 - 1.35
[35, 37]
750 - 1,200
[35, 37]
70 [35, 37]
450 [31]
Low
Table 1. Key characteristics of selected battery technologies (1Cycle Life at 50% DoD)
The cycle life of LIBs is considerably higher than LABs for deep discharge cycles. As a result,
LAB manufacturers usually recommend a DoD of 50%, whereas LIB manufacturers
recommend a DoD of up to 80% [45]. The cycle life of a battery mainly depends on the DoD,
temperature and discharge rate [46]. Since LABs are significantly more sensitive to any of
these factors, any alteration could dramatically decrease their cycle life. At room temperature
(25°C), LABs at 30% DoD could reach a similar cycle life as LIBs at a 75% DoD. However,
at 35°C, LABs cycle life decreases by 50% whereas LIBs could stay constant up to 50°C [39].
11
As previously mentioned, PbA batteries are characterised for their low cost when compared to
Li-ion, something essential to ensure battery-PVWPS viability in SSA. LFP is the most
expensive option, which is justified due to the excellent properties they offer such as longer
cycle life and no use of rare earth metals. An essential feature to consider is battery’s safety,
especially if these are located in remote areas, due to the high chances of batteries of going into
thermal runaway. A thermal runaway takes place when a cell reaches a certain temperature at
which the temperature will continue to increase on its own due to one exothermal process
triggering other processes. If the process becomes uncontrollable, this can result in flames and
gas emissions [47], [48]. LIBs are known to be more prone to reach higher temperatures due
to the large amounts of energy they can store in such small volumes.
2.2.2. Comparison of Battery Technologies in regards to Their Applications
An excellent way of comparing energy and power densities, two key characteristics when
analysing the performance of several storage devices, is by plotting them in a Ragone plot as
shown in Figure 4 [49]. In this 2-axis chart, specific energy (Wh/kg), which refers to the amount
of energy that can be stored per kilo of battery material, is plotted against specific power
(W/kg), which refers to how quickly a device can deliver the stored energy per kg of battery
material. As it can be observed, LIBs have significantly higher specific power and slightly
higher specific energy when compared to NIBs or LABs, making them more attractive for
applications where high charge/discharge rates are required, such as automotive applications.
Because of the high voltage, lightweight, temperature range and long cycle life they offer, Li-
ion batteries are commonly used in specific applications where these properties are required.
LFP is commonly used in stationary applications because of its higher level of safety compared
to other Li-ion technologies, NMC is commonly used in e-bikes and EVs, and LMO is
commonly used in power tools and electric powertrains [40]
Figure 4. Ragone Plot of Energy Storage Devices [49]
12
Although LABs do not offer the best choice in terms of specific power and energy, their low
prices, thermal stability and easy production have made it possible for them to still own a large
share of the market despite the new technologies that have emerged. They can be used for a
wide range of applications, varying from small scale like automotive applications (for starting,
lighting and ignition purposes (SLI)) to larger scale like grid energy storage systems [50].
The technical properties described above of the different battery technologies must be
considered when defining an adequate solution for PVWPS applications. This literature review
will focus on Lead-Acid and Lithium-ion batteries due to their extensive use and availability
in the market.
2.2.3. Environmental Performance and Challenges of Batteries
Since lead-acid batteries have been in the market for decades, their manufacturing and
recycling processes are very well standardised. This feature significantly facilitates the
assessment of their environmental impact. However, since Li-ion batteries are a more novel
technology and no standard recycling processes are fully defined yet [51], there is a limited
amount of literature on their environmental impact. Nevertheless, because of their ongoing
exponential growth, environmental concerns are rising, and the industry is trying to define the
path that could offer a sustainable development [51].
The main concern with LABs is the toxicity of lead, which can have dramatic human and
environmental impacts. The first cases of death related to lead poisoning started to be reported
in the late 19th century, and subsequently many countries started to implement regulations to
prevent lead exposure [52]. Since then, levels of lead in the population decreased by almost
80% from 1976 to 1991. Human exposure to lead can be either by the air, mainly due to
emissions from metal smelters, or by food contamination and subsequent ingestion. A
standardised way of determining lead pollution levels is by measuring grams of lead per m3 in
the air, and grams of lead per ml in people’s blood. Nowadays, air quality limits are 2 g/m3
according to EU Limit Value and 0.5-1 g/m3 according to WHO Guideline Value [53]. It is
estimated that an air level of 1.5 g/m3 leads to a blood concentration of 0.15 g/ml.
Lead emissions usually take place during the mining, manufacturing and recycling processes
[54]. LABs offer the key advantage that they can be easily recycled, and lead can be reused to
manufacture new batteries. Almost 99% of LABs are recycled, especially in countries like the
US, where the highest recycling rates can be found. However, the WHO report [55] stated that
in many developing countries, such as Senegal or Vietnam, where the recycling of LABs is
poorly regulated, health impacts due to lead exposure are still significant, especially for
children. It was found that blood Pb concentrations in children of a community in Senegal near
13
an informal recycling facility were ranging from 0.39 g/ml to 6.13 g/ml, considering that
levels above 0.45 g/ml indicate serious poisoning. A rate of 99% in recycling is impressive,
still, emissions from lead recycling must be strictly regulated all over the globe to avoid human
exposure, with a special focus on low and middle income countries [54].
On the other hand and as pointed out by [56], recycling is the main weakness of Lithium-ion
batteries, considering that less than 5% are recycled nowadays. LIBs are still not widely
recycled because some of its materials such as Li, Co, Mn or Ni are not considered as
problematic as Pb, so there is less social pressure, although they are still considered toxic heavy
metals. There are already few companies dealing with the recycling of LIBs. Most of them
recover Ni, Co and Cu, which can indeed decrease energy demand for primary production up
to 70% [57]. However, Li, Al and Mn are not recovered due to the high cost of the recovering
processes and not making the recycling process profitable [58]. Furthermore, LIBs contain rare
metals (RMs), such as Li, Co or Ni, and to avoid future shortage of these, recycling of LIBs is
needed [59]. In addition, mining processes of these metals require a large amount of energy,
and multiple environmental impacts are associated with them. It is estimated that Co, Li and
Ni comprise 39%, 16% and 9% of the economic value of a recycled LIB, respectively [60]
Another concern with LIBs is that some of its materials, such as Lithium or Cobalt, are usually
found in few places on the Earth. South America has the biggest Li reserves, containing around
72% of the total reserves. These are found in Bolivia, Argentina and Chile [61]. 70% of the
global mined Co comes from the Democratic Republic of the Congo (DRC), where it is mainly
mined by hand by local people and sold to foreign companies. The main reason why Co is
linked with ethical issues is because of the poverty, corruption and young and cheap labour of
the country [62]. Since a majority of the criticism for LIBs was driven by the use of Co, it
encouraged the industry to explore different chemistries and that is why many companies like
Tesla or Panasonic are moving away from using Co in their batteries [56].
It is of paramount importance that if batteries get installed in Gogma or any other place in SSA,
a clean and regulated disposal is defined. Batteries should ideally be recycled, and especially
if LABs are selected, lead emissions must be controlled and below the requirements.
2.2.4. Battery Market Landscape and Outlook for 2040
With the net-zero target that many developed countries have set for 2050, electrification of the
transport and power sectors have been set as one of the key priorities, because of being
responsible for 65% of global CO2 emissions [63]. In the transport sector, electrification will
be achieved by switching from fossil-fuel cars to EVs, and by an increase in the use of
intermittent renewables combined with storage technologies in the power sector [38].
14
According to [64], global demand for EV batteries is expected to increase from 110 GWh in
2020 to 5,910 GWh in 2040. According to their estimates, NMC Li-ion batteries will gradually
own the whole share for EVs by 2050, while LFP and NCA are expected to disappear by 2030.
Regarding batteries for storage applications, the global energy storage market is expected to
increase to 2,857 GWh or 942 GW by 2040 (Figure 5). This increase is significantly driven by
the massive decrease in cost that battery technologies will experience in the coming years,
which is expected to be about 52% from 2018 to 2030 [65]. Even though China and the US
will be leading the way, developing countries in Africa are expected to experience a rapid
growth in battery storage as well, because of the economical solutions that PV systems with
batteries can offer in isolated areas when compared to an extension of the grid.
Indeed, the World Bank Group has invested $1 billion to support fast deployment of battery
storage in developing countries. For example, a 150 MW solar farm with 200 MWh battery
storage has been installed in Burkina Faso [66]. However, a discussion about the environmental
impact or collection procedures when these batteries reach the end-of-life was inexistent.
Furthermore, as indicated by [34], Ghana had the target of installing 30,000 Solar Home
Systems (SHS) by 2020 and Nigeria has the target of 30,000 MW of PV by 2030, while many
other African countries have similar targets.
The key players and the ones responsible for making this transition occur are battery
manufacturers. Even though lithium-ion is expected to experience the biggest increase because
of being the main technology used for EVs, demand for lead-acid batteries is also expected to
grow significantly, mainly driven by Uninterruptible Power Supply applications (UPS), cars
(for SLI purposes) and e-bikes [67]. Main battery manufacturers have been summarised in
Table 2. As it can be appreciated, all major LIB manufacturers are based in Asia, whereas LAB
manufacturers are widely spread over the globe, mainly because of the maturity of the
technology.
Figure 5. Global Cumulative storage deployments [57]
15
Technology
Manufacturer
Country
Comments
Lead-Acid
Johnson Controls Int.
[60]
Ireland
Involved in the recycling of LABs.
EnerSys [60]
US
Involved in the recycling of batteries.
Exide technologies
[60]
India
Four major recycling facilities in US, Spain and Portugal.
East Penn [61]
US
Largest manufacturing site in the world. Involved recycling.
GS Yuasa Corporation
[61]
Japan
Manufacturing plant in the UK. Involved in the recycling of
batteries
Lithium-
ion
LG Chem [62]
Korea
Biggest LIB manufacturer. Development plants in China and
Poland.
CATL [62]
China
#2 global battery manufacturer and number 1 in China.
BYD [62]
China
#2 in Chinese production with huge expansion plans.
Panasonic [63]
Japan
4th largest supplier of LIBs. Partnered with Tesla, owns the world's
largest Li-ion battery factory in US
Ganfeng Lithium [63]
China
Covers a wide fraction of the Li-ion supply chain. Expected to
double capacity by 2025 from 2020.
Table 2. Description of the main players of the battery industry
With the information provided in this section, it can be said that battery-PV systems in Africa
have a high potential to play an important role in the future, mainly for off-grid applications,
and are highly favoured by the subsidies being provided by institutions for renewable energy
deployment in developing countries, such as AREI or SEFA (Sustainable Energy Fund for
Africa) [68].
2.3. Environmental Impact Assessment of Storage Technologies for PVWPS
This section covers the selected approach for this study, Life Cycle Analysis, to evaluate the
environmental performance of the chosen storage technologies. It provides a brief description
of the technical parameters that this approach implies as well as the different available tools to
carry it out. A detailed review of previous studies is also included.
2.3.1. Introduction to LCA
Life Cycle Analysis (LCA) is the most comprehensive and preferred approach when aiming to
quantify the environmental impact of a product or service. It is a standardised process defined
by the ISO 14040 and ISO 140144 [47], [62], [63]. It is widely used by governments, academia
and industry to support environmental policies and sustainable business decisions. It is
characterised by its holistic view and for considering all life cycle stages, from raw material
extraction to disposal [71].
LCA Process
According to the ISO standards, the LCA process is split into the four main phases, which have
been illustrated in Figure 6 and summarised below [70]:
16
Figure 6. Life Cycle Assessment Framework
LCA Phase 1: Goal and Scope Definition
The goal definition consists on determining different aspects, such as (i) the intended
application of the study, (ii) the purpose of the study, and (iii) the target audience. The scope
definition involves the description of assumptions and the methodology implemented to
perform the LCA study. The various features to be defined are: (i) function of the system, (ii)
functional unit (FU), (iii) system boundaries, and (iv) data quality requirements. FU consists
of the quantification of the function of the system to allow comparison with other products or
services that deliver the same function. There are three main system boundaries commonly
used: gate-to-gate, only includes the processes from the start of the production phase to the end
of the production phase; cradle-to-gate, includes the processes from raw material extraction to
the end of production phase; and cradle-to-grave, includes the processes from raw material
extraction to the end of life of the product.
LCA Phase 2: Life Cycle Inventory
The Life Cycle Inventory (LCI) aims at collecting and quantifying the inputs and outputs of
energy and materials for the analysed product system throughout the defined system
boundaries. As outputs and in parallel to products and co-products, there is generated waste
and emissions to air, soil and water. A unit process is the smallest unit for which inputs and
outputs can be quantified, as represented in Figure 7. Standard unit process [37]. Unit processes
are defined and then inventory data is collected and classified for each one of them [71]. A
product system is the compilation of all the unit processes required to deliver the defined FU.
This phase can be split in two sub-processes: (i) data collection and classification and (ii)
calculation of LCI. The latter is usually carried out thanks to a LCA software. Data of inputs
and outputs of standardised processes are stored in different databases which are used by these
software programs
17
Figure 7. Standard unit process [37]
LCA Phase 3: Life Cycle Impact Assessment
The Life Cycle Impact Assessment (LCIA) is split into compulsory and optional processes.
The compulsory processes can be split into four sub-processes: (i) selection of impact
assessment method, (ii) selection of impact categories, (iii) classification, and (iv)
characterisation. Some of the most common impact categories are: Global Warming Potential,
Fossil Depletion Potential, Ozone Depletion Potential, Acidification Potential, Eutrophication
Potential and Human Toxicity Potential. Classification consists of grouping the inputs and
outputs of the product system and assigning them amongst the selected impact categories.
Characterisation consists of the quantification of the LCI results by using the corresponding
characterisation factors. These are included within the impact assessment methods and are used
to convert LCI results into the reference units of each impact category. The optional processes
can be split into two sub-processes: (v) normalisation and (vi) weighting. Normalisation
consists of converting the characterisation results into a common unit to allow comparison
between the different impact categories, and weighting consists of multiplying each impact
category by a factor which is based on their importance. These results can then be aggregated
and a single result is obtained to allow comparison with other products or services.
LCA Phase 4: Interpretation
The interpretation can also be split into different steps: (i) identification of issues, (ii)
development of conclusions and recommendations, and (iii) construction of the final report.
Environmental Impact Categories
Environmental impact categories have to be selected based on the scope of the study and their
significance with the analysed product or service. As listed by [46], the top five analysed
impacts categories for battery LCA are Global Warming Potential, Terrestrial Acidification
Potential, Eutrophication Potential, Human Toxicity Potential and Mineral Resource Scarcity
Potential. Table 3 summarises the reference units and descriptions of the impact categories
listed above.
Product
Waste and emissions
Co-Product
Materials
Energy
Unit Process
INPUTS OUTPUTS
18
Impact Category
Acronym
Reference Unit
Description
Global
Warming
Potential
GWP
kg CO2-Eq.
It is based on climate change and GHGs are given in relation to
CO2. It has impacts on human and ecosystem health. Since the
residence time of gases varies, GWP can be calculated for 20 or
100 yrs.
Terrestrial
Acidification
Potential
TAP
kg SO2-Eq.
Acidification occurs when air pollutants are transformed into acids.
It is known because of its damages to ecosystems. The reference
substance is sulphur dioxide.
Freshwater
Eutrophication
Potential
FEP
kg P-Eq.
It is provoked by air pollutants, wastewater and fertilizers. In
water, an accelerated algae growth leads to a reduction in oxygen,
and as a result, fish die and CH4 is produced.
Human
Toxicity
Potential
HTP
kg 1.4-DCB
Effect of chemicals on human health based on three factors:
environmental persistence (exposure), accumulation in human food
and the toxicity of the chemical (effect).
Mineral Resource
Scarcity
Potential
MRSP
kg Cu-Eq
Amount of non-renewable natural resources consumed throughout
the lifetime of the product or service. It includes metal ores, crude
oil and mineral raw materials.
Table 3. Selected impact categories for the LCA study [63], [65]
Environmental Impact Assessment Methodologies
In the same way as the impact categories, the environmental impact assessment method has to
be previously selected. LCIA methods can have two different approaches: midpoint
level/problem-oriented or endpoint level/damage-oriented. While midpoint level refers to the
impact categories previously described, the end-point level is simply the classification of the
mid-point categories into three generic categories: damage to human health, damage to
ecosystem quality and damage to resources [70], [71]. According to [46], the three most
common methods used for the environmental impact assessment of batteries are ReCiPe, Eco-
Indicator 99 and CML.
i. ReCiPe is developed by PRé Consultants partnered with other institutions. It is a
damage-oriented method that involves eighteen midpoint categories and three endpoint
categories. It is characterised for using three different scenarios which represent
different perspectives: (i) Individualist (I), which refers to the short-term; (ii)
Hierarchist (H), which is the default model; and (iii) Egalitarian (E), which refers to the
long-term, and thus, greatest uncertainty. The Hierarchist default version uses
environmental models based in Europe for normalisation, therefore, it might show
limited validity for developing countries [73].
ii. Eco-Indicator 99 is a damage-oriented method. It includes ten midpoint categories and
three endpoint categories. It also includes the same three perspectives as the ReCiPe
methodology (I), (H) and (E). However, because of not being updated in a long time,
its use is not recommended.
iii. CML is developed by Leiden University in The Netherlands, so it is also a Europe-
based methodology. It is a problem-oriented method including fifteen impact categories
and its process goes all the way until the normalization step, ie. Phase 3 (iii).
19
LCA Software
The calculation of the LCI (Phase 2, ii) can be done manually, but since significant amounts of
data have to be compiled, LCA software are commonly used. These connect to large databases
which include information regarding standard material extraction and manufacturing
processes. They also contain the different impact assessment methods previously described,
and are able to undertake the classification (Phase 3, i), characterisation (Phase 3, ii) and
normalization (Phase 3, iii) steps. Each software contains different databases, but the most
employed database is Ecoinvent, based in Switzerland, and usually embedded in the software
if a professional license is acquired. The LCA software that will be used in this thesis is
SimaPro. It has been selected based on three criteria:
i. Identification of commonly used tools in the market: Based on literature, three LCA
software were identified as the most commonly used and with the most constructive
reviews: SimaPro and GaBi and OpenLCA [73][75].
ii. Accessibility: A second analysis was carried out to evaluate their accessibility. It was
found that Imperial College London (ICL) provides access to SimaPro but not to GaBi.
However, GaBi offers a free Education version for students but it only has access to a
limited number of databases. OpenLCA, as it names describes, is open-access software.
iii. Match with the current project: A clear answer cannot be provided considering that the
effectiveness of a software is dependent on the database it uses. SimaPro is the only
one with access to Ecoinvent and this is the main reason why it has been selected.
Majority of the information required for our case study can be gathered from this
database. Furthermore, it robustness considerably decreases running time when
performing large calculations.
A demo version of the three programmes was downloaded, and different tutorials were
followed to identify the strengths and weaknesses of each of them. Table 4 summarises the
technical information of the three different LCA programmes.
Category
SimaPro
GaBi
OpenLCA
Provider
Pré Sustainability
Thinkstep
GreenDelta
License
SimaPro Faculty (ICL)
GaBi Education (Free)
Free
Databases
Ecoinvent
Agri-footprint (food)
ELCD
US LCI
GaBi Database
US LCI
EF database
EF database
ELCD
Ecoinvent LCIA methods
Import/Export
Options
Import data in CVS
Export data in excel
Import data in excel
Export data in excel
Import data in excel, CVS, zolca
Export data in excel, CVS
Compatibility
Windows
Windows
Windows, Mac, Linux
Reference
users
Procter & Gamble, Caterpillar,
GE, Huawei, TUV, SABIC
Cepsa, Ford, General Motors, HP,
Hyundai, Renault, Siemens, Sony,
Toyota, Volkswagen, Volvo, Tesco
UNEP, SIG, Steelcase, Knoll,
CSC, USDA
Table 4. LCA Software tools comparison
20
2.3.2. Review of Studies
This section provides a compilation of papers which are based on the analysis of the
environmental performance of lead-acid batteries, lithium-ion batteries or both. Due to the
recent uptake of LIBs driven by the growth of the EV market, and due to the fact that there are
various chemistries available, a vast majority of papers are based on this battery technology,
and specially on automotive applications.
Lead-acid Battery LCA Studies
[76] carried out a study focused on the flow of Pb throughout the entire lifecycle (cradle-to-
grave), considering that 73% of a battery pack is lead, and most of the environmental burdens
are linked with its emissions. They quantified the environmental impact of LABs in twelve
midpoint categories, finding that material extraction and processing are responsible of about
80% of the environmental burdens. Since this paper is based in China and its lead recycling
rate is very low, they tried to incentivise the creation of new recycling regulations. [44] carried
out a comparison of different battery technologies PbA, NiCd, NiMH around a very specific
functional unit. The study was based on an EV of 1300 kg, during a 10 year lifetime, and
throughout a range of 20,000 km. It was a cradle-to-grave analysis. Considering these factors,
his results cannot be extrapolated for future use because of being based on such a specific
functional unit. Nevertheless, data about the recycling phase was provided on a kg basis.
Lithium-ion Battery LCA Studies
Papers based on LIBs usually present a comparison between different chemistries, normally
involving LFP, NMC and LMO. Two papers were based on PV stationary applications using
LIBs, [77] and [78], using 1 kWh of electricity delivered as their functional unit. On one hand,
[77] analysed a 100 MW PV system under three different irradiance scenarios, defining three
different battery bank sizes and using three different battery technologies (LMO, NMC and
LFP). On the other hand, [78] analysed a 10 kW PV system in Switzerland for three LFP battery
bank sizes. PV applications which integrate PV panels in their LCA are difficult to compare
since there are multiple external factors which considerably affect the results, such as solar
irradiance or PV panels efficiency.
Detailed Comparison between Six Selected Studies
A detailed analysis between six selected battery-LCA studies has been carried out (see Table
5). All studies have defined cradle-to-gate as their system boundary, except from [54] who goes
all the way to the end-of-life. [79] indicated that since recycling is not included, their results
could be considered as the worst case scenario because of not considering secondary metals.
Studies also differ on location. While [69] used broad data from various countries, [45] focused
their study in South Korea for the manufacturing process and in Norway for the assembly
21
process. Studies also vary in the source from which data is gathered. Whilst [80] and [59] used
a combination of previous literature and Ecoinvent, others such as [45] and [79] collected
primary data to get a more accurate and up-to-date inventory. Finally, studies also diverge in
the selected impact categories. While [69] only quantified CED and [80] quantified CED and
GWP, the rest of the studies quantified the impact in about fifteen midpoint categories.
Goal and Scope
Inventory
Impact Assessment
Reference
Battery
Technology
System
Boundary
Location
Functional Unit
Data Source
Methodology
/Software
Impact Categories
[80]
PbA, Li-Ion,
NaS, V-Redox
cradle-to-
gate
Germany
MWh of
electricity
Ecoinvent &
secondary data
ReCiPe/
SimaPro
CED
GWP
[59]
PbA, Li-ion,
NaS, NiCd
cradle-to-
gate
Europe
per kg of
battery
per MJ of
capacity
Ecoinvent &
secondary data
ReCiPe/
SimaPro
GWP, ODP, HTP,
POFP, PMFP, IRP,
ALOP, ULOP, NLTP,
WDP, MDP, FDP
[69]
PbA, NiCd,
NiMH, NaS,
Li-ion
cradle-to-
gate
Global
per kg of battery
per Wh
delivered
Secondary
Data
-
CED
[54]
PbA
cradle-to-
grave
China
per kWh
Primary Data
& CLCD
eBalance
(Chinese LCA
Software)
PEU, ADP, TAP, EP,
GWP, ODP, POFP,
HTPc, HTPnc, ETP
[45]
NMC
cradle-to-
gate
South
Korea
&
Norway
1 battery pack
for EV
expressed per kg
and per kWh
Ecoinvent,
primary &
secondary data
ReCiPe
GWP, FDP, ODP,
POFP, PMFP, TAP,
FEP, ETP, HTP, MDP
[79]
NiMH, LFP,
NMC
cradle-to-
gate
Europe
50MJ
per kWh
per kg
Ecoinvent,
primary &
secondary data
ReCiPe
GWP, FDP, ODP,
POFP, PMFP, TAP,
FEP, ETP, HTP, MDP
Table 5. Technical parameters of the selected LCA studies
From the compiled results from literature, we conclude that LABs demand lower energy during
the lifetime when compared to LIBs. At the same time, their GWP is lower as well. LIBs
present considerably higher metal depletion potential (MDP) due to the use of rare earth metals,
whereas LABs present higher human toxicity potentials (HTP) because of the use of lead.
The information described in this section will be considered and used as reference during the
development of our LCA model. This section has highlighted the research gap that this thesis
aims to fulfil within the battery LCA field, which consists of giving greater emphasis to the
end-of-life phase and performing the modelling within the African continent, according to our
case study of the village of Gogma. Therefore, this study will aim to examine the environmental
impact from cradle-to-grave of storage technologies for PVWPS within the African context,
something that to our knowledge has never been previously done.
22
2.4. Lithium-Ion and Lead-Acid Batteries Life Cycle Overview
This section starts with a description of the working principle of batteries, followed by an
overview of the lifecycle of batteries. From raw materials extraction and manufacturing, it
continues with transport and use phase, and concludes with the end-of-life and the different
recycling approaches. Various flow charts have been created to support the description of some
of the processes.
2.4.1. Operating Principle & Cell Architectures
A battery is a device that stores energy in the form of chemical energy and when discharged, it
is converted into electrical energy. This conversion occurs due to the chemical reactions taking
place between the electrodes, where one electrode releases ions (oxidation) and the other one
absorbs them (reduction), and this flow of electrons through a connected device generates a
current. A battery is a group of cells connected in series. In the discharging phase, the cathode
is oxidised and the anode is reduced, and in the charging phase, the cathode is reduced and the
anode is oxidised [44]. The full chemical reactions for a lead-acid and lithium-ion battery are
shown in Eq (1) and (2), respectively, where discharging takes place from left to right and
charging from right to left [81]. In Eq (2), M stands for any metal such as Co, Ni or Mn. The
structure of a Lead-Acid battery is represented in Figure 8. LIBs can have various shapes such
as coin, prismatic or cylindrical. The latter is illustrated in Figure 9, which is the most widely
used and the one for which the manufacturing process has been described [81].
     (1)
    (2)
Figure 9. Lithium-Ion Cylindrical Cell
Architecture [147]
Figure 8. Lead-Acid Battery
Architecture [146]
23
2.4.2. Overview of Full Life Cycle
As described in the previous chapter, a cradle-to-grave LCA involves all the processes from
raw material extraction to disposal. The flow chart in Figure 10 represents the different
processes that will be considered for the analysis of the environmental impact of this study. It
has been kept non-specific so that it is applicable to both, lithium-ion and lead-acid batteries.
Since no changes will be made to the cradle-to-gate system boundary and these processes will
keep fixed values, they have been represented with orange boxes. On the contrary, processes
from the end of the manufacturing phase to the end-of-life, will be given different values until
the solution with the lowest environmental impact is defined, and thus represented with blue
boxes.
2.4.3. Material Extraction & Material Processing
This section focuses on the different materials contained within the batteries as well as the
corresponding components they are used for. The manufacturing processes of the components
and their function within the battery will be covered in the following section.
Lead-Acid Batteries
A solid LCI was found in the literature for lead-acid battery composition with all the referenced
authors rigorously coinciding on the quantities [44], [59], [72], [74]. As it can be appreciated
from Table 6, Pb is the most predominant material, because it is used for both electrodes
anode and cathode. The negative active material is made of porous lead and the positive active
material is made of lead peroxide (PbO2). Both of them are supported by a lead-based alloy
(including antimony, copper and arsenic) grid which is used for conducting electric current
[82]. The separator is made of non-conductive materials, usually of mats of fibreglass with
polyethylene as a binder, and its aim is to keep the anode and cathode apart [69].
Figure 10. Full Life Cycle flow chart for lead-acid and lithium-ion batteries adapted from [54]
24
Lead-Acid Battery
Component
Material
Percentage (%)
Electrodes
Antimony (Sb)
0.71
Arsenic (As)
0.03
Copper (Cu)
0.01
Lead (Pb)
60.97
Oxygen (O2)
2.26
Electrolyte
Sulfuric Acid (H2SO4)
10.33
Water (H2O)
16.93
Separator
Glass
0.2
Polyethylene (PE)
1.83
Case
Polypropylene (PP)
6.73
Table 6. Components and Materials of a lead-acid battery [35], [47], [67], [78]
Lithium-Ion Batteries
Compiling this LCI was more complicated due to the number of different chemistries, as well
as the large dissimilarities in literature. Since all LCI presented variations, Table 7 compiles
the inventory of four different studies. The anode consists of a copper current collector covered
by a coat of negative electrode paste, which is a mixture of synthetic graphite and binder
(PVDF). The cathode consists of an aluminium current collector covered by a positive
electrode paste, which is a mixture of the positive active material (LiMO2) with carbon black
and binder. This is what determines the chemistry of the battery [45]. The binder in the
electrodes holds the active material together. The electrolyte is LiPF6 in a solution of PC, EC
and DMC [84].
Lithium-ion Battery
Component
Material
Ramirez
(LMO)
Matheys
(Li-ion)
Buchert
(LFP)
Gaines
(LFP)
Electrodes
Graphite (C)
14.96
14.96
21.22
15.30
Lithium metal (Co/Ni/Mn) oxide (LiMO2)
23.63
23.63
33.08
22.20
Copper (Cu)
15.55
9.45
11.47
13.80
Aluminium (Al)
20.73
12.6
6.50
13.30
Polyvinylidene fluoride (PVDF)
1.19
1.19
-
3.40
Styrene Butadiene rubber (SRB)
1.19
1.19
-
-
Graphite
-
-
-
15.30
Electrolyte
Propylene carbonate (PC)
1.19
1.19
16.25
14.20
Ethylene carbonate (EC)
6.3
6.3
Dimethyl carbonate (DMC)
3.15
3.15
Lithium hexafluorophosphate (LiPF6)
3.15
3.15
Separator
Polyethylene (PE)
5
0
8.22
4.60
Case
Steel
2
21.23
3.25
0.10
Table 7. Components and Materials of a Lithium-ion battery [47], [65], [68], [81]
25
2.4.4. Transport
This phase consists of the quantification of the energy required and the emissions associated
with the transportation of goods from one place to another. It is usually expressed in tonne-
kilometre (t·km), which is the product of the weight of the shipped goods times the travelled
distance. As it can be appreciated from Figure 10, transport has been considered between every
unit process. Due to the fact that the aim of this thesis is to minimise the environmental impact
of batteries, every aspect must be taken into account, so the travelled distance from the
manufacturing facility and to the recycling facility must also be considered.
Ecoinvent database contains two datasets within the materials and processes sections, called
transformation and market. The transformation dataset simply considers the activities required
to transform inputs into outputs, whereas the market dataset considers not only to the
transformation activities, but also the average market transport requirements [86]. Since the
market dataset is the most widely used and most of the values obtained from the literature have
used Ecoinvent, it can be assumed that most of the studies have included transportation in their
cradle-to-gate analysis. However, transportation from the manufacturing plant to the location
of use, from the location of use to the recycling plant, and from the recycling plant back to the
manufacturing plant, has to be defined and manually populated.
Few studies mentioned the importance of transport in their studies. [44] presented a LCI where
transport between each stage has been considered. [45] described the transportation from the
battery manufacturer in East Asia, to the battery pack assembler in Norway, by a combination
of 0.55 t·km of road transport and 4.8 t·km of ocean freight. [54] assumed all materials came
from domestic suppliers and used an average distance of 181 km by heavy-duty diesel trucks.
They also included the transportation to users as a separate phase. On the other hand, [79] just
pointed out that transport requirements were considered but these were kept generic.
2.4.5. Manufacturing
This section describes the different steps required throughout the manufacturing of lead-acid
and lithium-ion batteries. This process starts with processed materials entering the
manufacturing plant and finishes with batteries ready to be dispatched.
Lead-Acid Batteries
There are two types of manufacturing processes for LABs, dry-charged and wet-charged,
which are very similar except for the charging step. The latter comes with the electrolyte and
is ideal for immediate use, and thus the one described in this section [87]. The process can be
divided into the six steps, which have been summarised below and illustrated in Figure 11 [44],
[69], [87]:
26
i. Lead oxide production Pure lead is placed under high temperatures and the molten
lead is oxidised and forms lead oxide.
ii. Grid production lead alloy is melted and poured into moulds with the grid patterns.
iii. Paste Production and Pasting two types of pastes are manufactured for the positive
and negative plates. The pastes are then applied into the grids and dried in a flash drier.
iv. Plate curing and parting plates are cured in ovens and cut into the required dimensions.
v. Battery assembly the separator is placed in between the positive and negative plates
to create a stack. Tabs are welded and stacks are placed into the case.
vi. Filling and Charging the battery is filled with the electrolyte. Batteries are slow-
charged, a process also known as formation. They are then subjected to high rate testing,
labelled and be shipped.
Lithium-Ion Batteries
The manufacturing process of the different lithium-ion batteries technology is very similar. The
only feature that changes is the active material used for the cathode which will depend on the
chemistry being manufactured. The manufacturing process can be split into the five steps that
have been summarised below and illustrated in Figure 12 [57], [88].
i. Paste Production active materials, binder and solvents are added to make the wet
paste.
ii. Coating and Drying anode and cathode pastes are applied over the foils. The coated
foil is transferred to the dryer to reduce thickness by 25-40%.
iii. Calendaring and slitting to width the coated foils are compressed to achieve a uniform
thickness. These are then trimmed into several small electrode coils.
iv. Cell assembly and Filling - anode, cathode and separator are stacked and inserted into
cylindrical or rectangular cases. The electrolyte is then filled into the cell.
v. Charging and Testing cells are charged and discharged at specified voltages and
currents. Quality and end-of-life tests are carried out before the cells are delivered.
Figure 11. Flow Diagram for Lead-Acid Battery Manufacturing Process adapted from [87]
27
2.4.6. Use Phase
In our case, the use phase of the batteries consists of the period it is connected to the PVWPS.
They will be charged while PV panels are generating electricity and there is no water demand,
and discharged when there is no electricity generation but there is water demand. Batteries will
also store energy during the periods when there is no water demand. The following section
describes the key parameters which affect battery performance.
Key parameters affecting use phase
Most studies make assumptions about the electrical parameters of batteries, but these play a
crucial role in the battery performance during the use phase. Indeed, it has been demonstrated
that better performances can significantly decrease environmental damages [46]. These
parameters are specified by manufacturers, but good or poor maintenance can improve or
worsen them. The sub-sections below cover the parameters that have been identified as having
the biggest impact in the environmental performance of batteries during their use phase [46].
Cycle life
As it can be appreciated from Table 1, LABs and LIBs show large differences in the average
number of cycles they can offer throughout their use phase. If batteries get degraded, their cycle
life can be considerably minimised. The degradation is influenced by the depth of discharge,
charging rate and temperature [46]. Figure 13 shows how the optimal DoD for LABs is around
50%, and discharging beyond this fraction can significantly reduce the cyclic life of the battery.
On the other hand, LIBs can be subjected to a DoD of up to 80%. Indeed, [45] highlighted that
LIBs with a DoD of 100% can reach a cycle life of 1,000 cycles, whereas a DoD of 50% can
expand the cycle life to 5,000 cycles. Therefore, DoD of 80% for LIBs is usually the optimal
trade-off between DoD and number of cycles [46].
Figure 12. Flow Diagram for Lithium-Ion Battery Manufacturing Process adapted from [57]
28
Round-trip Efficiency
Batteries have efficiencies range around 80-95%, being these a bit lower for LABs. Various
factors can also have an impact on battery efficiency, such as the charging rate or the ambient
temperature [46]. Few studies take into consideration the efficiency of batteries, such as [76]
who used an efficiency of 82.5% for a lead-acid battery and [45] who used an efficiency of 92-
96% for a lithium-ion battery.
Electricity mix
Electricity mix refers to how energy sources are split to meet the electricity demand of a certain
region. Energy sources can be fossil fuels - coal, oil or natural gas - nuclear energy or renewable
energy - wind, solar, or hydro [89]. Electricity mix varies significantly from a country to
another, and they usually have a carbon intensity associated to them, which is dependent on
the proportions of the consumed energy sources.
The carbon intensity of the electricity mix plays a crucial role during the use phase of batteries
used in EVs, considering that these get charged using the electricity from the grid, so the CO2
content of the electric grid is transferred to the use phase of the vehicle [45]. However, in
PVWPS, batteries are charged by the electricity generated by the PV panels, and since they do
not emit any CO2 emissions during the generation of electricity, zero CO2 emissions are
associated with the use phase of batteries. Another option would be to consider the embedded
CO2 emissions of PV panels - CO2 emissions from the manufacturing and transport stages -
during this phase.
On the other hand, if we were analysing a diesel water pumping system, all the emissions from
the diesel combustion would be associated with the use phase. This diesel consumption would
also have an effect in other impact categories apart from GWP, such as fossil fuel depletion
potential and terrestrial acidification potential [90].
Figure 13. Relation between battery capacity, DoD and cyclic life for a lead-
acid battery [143]
29
Since the electricity mix is associated with a given geographic area, it will vary during the
different life cycle stages of batteries. The electricity mix must also be taken into account
during the manufacturing and recycling phases. Europe and China are the two most common
locations where batteries get manufactured or recycled, so their carbon intensity must be
considered when selecting one location or another. While Europe’s electricity is based on coal,
natural gas, nuclear and hydro and has an average carbon intensity of 296 gCO2/kWh [91],
[92], China’s electricity is mainly based on coal with a small fraction of hydro, leading to a
higher average carbon intensity of 555 gCO2/kWh [85, 86].
2.4.7. End-of-life
This section looks at the end of life of batteries, starting with the definition of the required
terminology and followed by an overview of the existing recycling processes for batteries. It
continues with a detailed description of the recycling processes of both lead-acid and lithium-
ion batteries, as well as a brief discussion about the main actors in the battery recycling
industry.
End-of-life Terminology
There are different pathways that can be followed when a product reaches its end of life. In
general, there are three main routes: reuse, recycle or disposal. The simplest form of reuse
consists of utilising a product for the same function without changing its nature. When a
product requires cleaning or repair to be used again, this product needs to be remanufactured
[95]. Examples of this approach are batteries from EVs, which can be disassembled after their
end-of-life in the EV industry and be reused for less demanding applications, such as stationary
storage [96]. On the other hand, recycling consists of the transformation of disposed materials
or products into usable components. Closed-loop recycling is when materials are used for the
same product. Direct recycling is a type of closed-loop recycling in which materials can be put
back into the same product with insignificant processing. A clear example of direct recycling
are LABs, in which lead can be easily recycled for new batteries with minimal treatment. Open-
loop recycling is when materials are used for a different type of product from which they were
recovered. When this product is of lower value, is called downcycling, and when it is of higher
value, is called upcycling. Finally, disposal is when consumed products just go to landfill
without any form of treatment [95].
Recycling of Batteries
The recycling can be divided into pre-treatment and recycling steps. Batteries must be first
collected and transported to a specific recycling facility. Once there, these are sorted by battery
chemistry and discharged to avoid any risks of the residual charges. These are then dismantled
to separate the different components, either manually or by mechanical means, and then sorted
30
by type. Finally, the different battery components are ready to be fed into their corresponding
recycling process. There are three main recycling approaches for batteries: pyrometallurgical,
hydrometallurgical and mechanical [55], [97][99].
Lead-Acid Batteries
Current Status
LABs are simple devices which contain few materials, two key features which considerably
facilitate their recycling process. The standardised recycling method across the industry and
the high collection levels have significantly contributed to the elevated recycling rates achieved
in the EU [38]. The majority of materials are recovered and reused in new LABs.
Recycling Process
i. After breaking the battery into pieces, these are placed in a tank where plastics will rise
to the top and metals and heavy components will sink, and thus resulting in the
separation of components. After this point, each material follows its own process [55],
[98]. Firstly, plastic pieces are cleaned and molten to obtain a liquid material. This is
then cooled and extruded to create uniform pieces, which are sold and reused in future
batteries. The sulphuric acid electrolyte is mixed with baking soda and converted into
water. After quality controls, the water is released into sewerage systems. Finally, lead
is recycled. The most common process is pyrometallurgical recycling, accounting for
>90%, although there have been recent developments in the eco-friendly
hydrometallurgical process [98], [100]. Figure 14 shows the different unit processes, as
well as inputs and outputs for each recycling method.
ii. Pyrometallurgy PbSO4 is desulphated with caustic soda. By mixing the mixture with
coke, all Pb compounds are reduced into metallic Pb [101]. No sulphur removal prior
to the smelting process (direct smelting) results in increased SO2 emissions. The direct
smelting procedure is still used in developing countries. Primary and secondary bullions
are produced with ~90% Pb and ~9% Pb, respectively. The last stage produces a slag
with ~2% Pb, discarded if Pb levels are acceptable [100]. Refined components are
casted and used in new LABs [55].
iii. Hydrometallurgy The key feature of this process is that it involves a discharge method
by which lead can dissolve at low temperatures, and thus resulting in a more sustainable
pathway for the lead recycling [98]. However, its high operating costs do not make this
process economically feasible [100].
31
Main Actors
Since Pb recycling has been in the industry for a long time now, there is a vast amount of
battery recyclers, especially in Europe and the US, where LABs recycling rates are
considerably high. ECOBAT is the world’s leading lead producer and recycler and offers
closed-loop recycling for LABs. They have twenty-nine sites across Europe, the US and South
Africa [102].
However, the African continent is strongly characterised by its high levels of informal
recycling. Unsound, backyard or informal recycling refer to the inadequate recycling of LABs
due to the unawareness of the toxicity of Pb. This leads to a massive contamination of the
environment and subsequent impacts on human health, triggered by the lack of precaution
measures [103]. Therefore, it is of paramount importance to ensure that batteries of PVWPS
are collected and sent to formal recycling facilities. There are various well-established
recycling facilities in Africa, such as First National Battery in South Africa [104]; Gravita and
Goldline in Ghana; Metal World and Lloyds in Nigeria; Gravita Recycling in Mozambique;
ASSAD in Tunisia; and Metafrique in Cameroon [105]. Other companies, such as Johnson
Controls, have taken a different approach by shipping LABs from Ghana to be recycled in their
German facility [106].
Lithium-Ion Batteries
Current Status
Opposite from LABs, the recycling of LIBs is not widely implemented. These are more
challenging to recycle due to the large number of blended materials and the lack of a
standardised battery design across the industry. The continuous development of electrodes are
making the standardisation process even harder [38].
Figure 14. Flow chart of hydrometallurgical and pyrometallurgical LAB recycling processes
adapted from [100]
32
Recycling Process
There are three main pathways LIBs can follow after their end-of-life. The first two are based
on chemical processes, pyrometallurgy and hydrometallurgy, and the last one is based on
reusing components with minor adjustments, known as direct recycling [58]. These three
processes are summarised below.
i. Pyrometallurgical with no pre-treatment requirements, LIBs are introduced into
furnace for the smelting process. Since the electrolyte and plastics burn and help to
maintain the high temperatures, they can be considered as recovered materials. This
process results in the recovery of an alloy made up of valuable metals like Co, Cu,
Ni and Fe, which are refined for later use in new LIBs, providing a closed-loop
recycling approach. On the contrary, Li, Mn and Al are contained within the slag
and are used as concrete additives, proving a downcycling approach. A flow chart
is illustrated in Figure 15, where the steps needed for the alloy recovery are
included. This process is being used in the current industry, but it might stop being
feasible as new battery designs decrease the amounts of Co and Ni [57], [58], [84],
[95], [107].
ii. Hydrometallurgical opposite to the previous process, hydrometallurgical
processes require a uniform feed, therefore, some pre-treatment is needed. Based
on the experiment carried out by [108] for a laptop battery, [84] adapted it for a EV
LMO battery. The different steps involved in the recycling process have been
illustrated in Figure 16. Once components have been separated [57], the cathode is
soaked into warm NMP solvent to separate the paste from the aluminium foil.
Active materials are then crushed and calcinated at 1300C. The leaching step is
based on the reduction of the Co and Li by the addition of hydrogen peroxide and
organic acids. Finally, metals are filtered and ready to be reused in new LIBs [84].
The main advantage of this process is that it requires lower temperatures compared
to pyrometallurgical processes and it offers higher extraction efficiencies [109].
Figure 15. Flow chart of pyrometallurgical LIB recycling process adapted from [84]
33
iii. Direct recycling consists of the recovering of battery materials with minimal
processing. The process has been illustrated in Figure 17. Discharged and
dismantled cells are placed in a container to which CO2 is added. Temperature and
pressure are increased to reach the supercritical CO2 point for the electrolyte to be
extracted. This can then be processed and reused in new LIBs. The rest of the
components are then reduced in size and separated by type to follow their
corresponding recycling process. The main advantage of this process is that most
of the materials can be reused, except separators. Nevertheless, there are questions
regarding the technical performance of recovered cathode materials [84], [107].
Main Actors
In contrast to LABs, a proper network for recycling lithium-ion batteries has not been
established yet. Umicore is a European company leading LIB recycling, which is known for its
pyrometallurgical process and for connecting cathode manufacturing with recycling [57], [69],
[84], [85], [109], [110]. Umicore starts its process by shipping spent batteries to Sweden, where
they are smelted and valuable metals are recovered. The resulting Ni and Co are sent to
Belgium, where CoCl2 is manufactured and sent to South Korea for LIBs manufacturing.
Umicore does not recover Lithium [57]. Other recycling companies in the EU are SNAM in
France, Redux in Germany and Batrec in Switzerland [110]. TOXCO is a Canadian company
which processes batteries and produces three products, which are rich in Co, Cu, Ni and Al.
Nevertheless, these compounds must be reprocessed to be used in new LIBs [57], [69].
Figure 17. Flow chart of direct LIB recycling process adapted from [84]
Figure 16. Flow chart of hydrometallurgical LIB recycling process adapted from [84]
34
2.5. Research Gaps & Objectives
The fact that batteries can considerably decrease the LCC of PVWPS in SSA is an option that
has never been previously explored. Taking this path could considerably unlock the value of
batteries in the African continent and bring plenty of opportunities which can contribute to its
development. As described in Section 2.1, battery storage for PVWPS could offer multiple
advantages from a technical point of view when compared with water tank storage.
Nevertheless, barriers such as maintenance, replacements and recyclability should be tackled
prior to installation. In line with the discussion in Section 2.2.4 concerning the current battery
market, it is concluded that only lead-acid and lithium-ion batteries will be analysed in this
study due to their high penetration in present and future markets.
As described in Section 2.3.2, no previous literature has carried out cradle-to-grave LCA
studies in the battery industry. The majority of studies are based in Europe or China, but no
LCAs are located in the African continent. Even though existing studies can be extrapolated,
electricity mixes and transport distances should never be underestimated, bearing in mind that
these can significantly influence final results. In addition, none of the mentioned studies was
focused on PVWPS applications. Authors commonly compare different battery technologies
but they rarely compare them with other storage technologies, such as water tanks. No previous
studies have looked at existing manufacturing and recycling facilities around the world.
Finally, Section 2.4 tries to compile all the data, processes and concepts which will be required
for the modelling of the LCA.
All the aspects described above that previous LCA studies are missing, make it hard to perform
an environmental analysis of batteries in SSA by only using secondary data. This is the reason
why the study carried out in the following chapters is tremendously needed to design
sustainable and feasible solutions for PVWPS in the future. Therefore, this thesis aims to
develop a methodology to define the storage technology for PVWPS which offers the
lowest environmental impact, whilst taking into account all the stages during their
lifetimes.
One of the main differences of this study is that it deals with real manufacturing and recycling
facilities and all compiled data comes from reliable sources such as previous studies and
Ecoinvent database. Therefore, the final choice from this study could ideally start to be
implemented at any time.
All results and conclusions gathered during this study will not only be useful for the PVWPS
in Burkina Faso. Since PVWPS are becoming more popular across the continent, it is expected
for this information to be widely shared and implemented in other projects, and contribute to
the development of sustainable and feasible ways of water access in remote areas.
35
3. Methodology Design
This chapter defines the methods and tools implemented during the development of the thesis,
as well as data collection and result analysis approaches. Section 3.1 gives a brief overview of
the methodology and the case study. Section 3.2 explains one by one the employed methods.
Section 3.3 and 3.4 describe data collection and result analysis approaches, respectively.
Finally, Section 3.5 outlines the assumptions taken during the LCA modelling process, and
Section 3.6 provides an overview of the utilised electricity mixes.
3.1. Introduction
PVWPS are traditionally accompanied by water tanks to store water and be able to provide
water at any time of the day. Because of the lack of resources and awareness in developing
countries, no other storage technologies have ever been explored for these applications. As
concluded by [2], it was found that the LCC cost of battery-PVWPS is $24.1k while it is $31.1k
for a steel tank-PVWPS. The finding brings new opportunities and the possibility to decrease
the cost of PVWPS and make them more affordable to the communities in SSA. However,
batteries are usually characterised for their environmental impacts, especially for the toxicity
and scarcity of some metals. Therefore, this thesis aims to quantify the environmental impact
associated to each of the storage technologies available for PVWPS to incorporate an
environmental factor in future decision-making processes. This study looks at four different
storage technologies: steel water tank, Linear Low-Density Polyethylene (LLDPE) water tank
(plastic water tank from now on), lead-acid battery and lithium-ion battery.
The study carried out in this thesis has never been previously performed (see Section 2.5). The
following sections describe the methods that have been followed to develop this study and
allowed the compilation of meaningful results.
3.1.1. Our Case Study
The PVWPS which was installed in 2017 in Gogma (see Section 2.1.3), a village located in the
centre-east of Burkina Faso, will be used as case study (Figure 18) [1]. This system consists of
620 Wp of multicrystalline PV modules, a steel tank of 11.4 m3 and a motor-pump SQFlex 5A-
7 [111]. This PVWPS provides around 10 m3 per day for domestic use to the 250 inhabitants
of the village. Data like solar irradiance and water collection have been gathered from this
PVWPS, and it has allowed the publication of other studies which aim to define an affordable
and sustainable PVWPS design for isolated areas.
36
Figure 18. PVWPS installed in the village of Gogma
3.2. Overview of Methods
3.2.1. Literature Review
As covered throughout Chapter 2, the first stage of this study consisted of a deep reading and
understanding of published studies regarding the topic of the thesis, environmental impact of
batteries. Other areas such as electricity and water access in SSA and current status of PVWPS
were investigated. This stage was crucial to be able to identify the methodologies that have
been previously implemented and define the research gap that this thesis was going to tackle.
Furthermore, some of the most relevant papers with useful information for LCA modelling
were labelled.
3.2.2. Criteria Definition
Prior to starting with the modelling, it was indispensable to define the criteria that was going
to be used during the thesis. As slightly mentioned above, four different storage technologies
have been analysed: two water tanks, steel and plastic, and two batteries, lead-acid and lithium-
ion. In order to carry out an objective analysis amongst the different storage technologies,
various scenarios have been developed representing the different locations where these
technologies are manufactured, landfilled and recycled. Europe, Asia and Africa have been the
three main areas of focus. To examine the environmental impact of the four technologies,
endpoint and midpoint categories (see Section 2.3.1) have been used as the main criteria.
Considering that there are up to 18 midpoint categories, only the most relevant five have been
analysed.
37
3.2.3. LCA Modelling of Storage Technologies for PVWPS
As explained in Section 2.3.1, LCA has been the selected approach for this study. The four
phases described in the section have been used as a guide during the thesis. Corresponding to
LCA Phase 1, definition of the functional unit, system boundaries, and methodology see
Section 4.1. Corresponding to LCA Phase 2, Life Cycle Inventory, see Section 4.3.
Corresponding to LCA Phase 3, Life Cycle Impact Assessment, see Chapter 5. Finally,
corresponding to LCA Phase 4, Interpretation of Results, see Chapter 6.
Even though OpenLCA was initially the preferred software to perform this study, due to
technical issues it was not possible to run some of the large calculations. Because of this, all
the scenarios were transferred to SimaPro and calculations were performed and exported.
External support from local people and LCA experts has been crucial for the development of
the study. LCA is characterised to be an iterative process. By the time final results were
obtained, lots of iterations had to be made because of data updates or calculation errors.
3.2.4. Scenario Approach
Due to the large amount of information that had to be analysed in this study, it was essential to
organise it with a logic behind it. Different scenarios have been created which classified the
information per storage technology, per location and per end-of-life treatment. How these
scenarios were created and the final outcome is described in Section 4.2.
3.2.5. Results Representation
SimaPro’s data representation tool was not as powerful as required for this study, considering
that results were represented in percentages and not in units. Therefore, since it was necessary
to accurately examine the different storage technologies, results were exported and imported
into Python for data representation. Python is a high-level programming language and very
useful for a wide range of applications. It was used to create graphs for endpoint and midpoint
categories as well as for the sensitivity analysis (see Chapter 5).
3.3. Data Collection
As previously mentioned, the literature review was very useful to identify useful data from
LCA studies for the development of the model. Nevertheless, further research has been carried
out because of still missing some key values. Another important contribution was Ecoinvent
Database. Data from here has not been directly used but many of its processes have been used
as templates for our model.
38
3.4. Analysis of Results
Thanks to the graphs created with Python, it was pretty straightforward to analyse the obtained
results. The scenario with the lowest environmental impact represented the best solution. By
using endpoint indicators as a first filter, the best scenarios from each storage technology have
been selected (see Section 5.1). The six selected scenarios have been further analysed with
midpoint indicators as a second filter (see Section 5.2). In order to get a better understanding,
midpoint results have been split per phase (manufacturing, transport and disposal).
In addition, a sensitivity analysis (see Section 5.3) has been carried out by using three key
performance indicators (KPIs): electricity mix, transport by lorry and transport by aircraft.
These results will give us a better estimation of how much the environmental impact of a
specific storage technology can vary when changing any of those KPI.
3.5. Assumptions for LCA Modelling
Various assumptions had to be made during the study, specially while using the software
because it was the moment when most of the uncertainties had to be solved. The most relevant
assumptions are listed below:
i. The LMO chemistry has been the one selected for the lithium-ion battery scenarios. It
was one of the outlined types (see Section 2.2.1) and also the one that Ecoinvent
database included.
ii. It was assumed a metal recovery of 70% from batteries. This assumption only applies
to recoverable metals, such as lead from lead-acid batteries and copper from lithium-
ion batteries. This value was obtained from a lead-acid battery recycling facility in
Sweden [112]. This hypothesis had to be made because of the lack of information in
the lithium-ion battery recycling industry.
iii. In the batteries scenarios it was assumed no recovery of plastics, only of metals.
iv. Since the LCI of the LMO LIB from Ecoinvent database included the manufacturing
facility (infrastructure), the same values have been used for the lead-acid battery
scenarios. Water tanks also include their corresponding infrastructure in their
inventories.
v. When transport distances were too long and no direct flights were available, real
commercial flights have been explored to use the same stops and acquire more accurate
data. Travel distances were calculated by using Google Maps.
vi. All utilised Ecoinvent processes followed the APOS system model. It stands for
‘Allocation at Point of Substitutionand it is Ecoinvent’s default model. It consists of
proportionally attributing the environmental burdens to specific processes [113].
39
3.6. Electricity Mix
This terminology has already been discussed from a theoretical point of view (see Section
2.4.6). In this section, the carbon intensity of the countries used in the different scenarios will
be presented as well as an overview of the composition of the general electricity mixes. As
previously explained, this information is not relevant during the use phase, but it is crucial
during the manufacturing and recycling phases, especially in energy-intense processes.
3.6.1. Carbon Intensity
Table 8 shows the carbon intensity of all the nations used during the LCA study. The rows
highlighted in grey (Africa, Europe and China) represent the average values of those regions.
Since the electricity mix of Burkina Faso was not present in Ecoinvent database, RAF’s value
(Africa’s average) has been used instead. These values have been taken from the Global
Warming indicator from the ReCiPe Midpoint (Hierarchist) category.
China has the highest carbon intensity out of the three main nations, followed by Africa and
then Europe. South Africa is the country with the highest carbon intensity and Belgium the one
with the lowest carbon intensity, with a difference of 0.232 kg CO2/MJ (Table 8).
Nation
Abbreviation
kg CO2/MJ
Africa
RAF
0.203
South Africa
ZA
0.296
Burkina Faso
BF
-
Nigeria
NG
0.139
Ghana
GH
0.128
Europe
RER
0.116
Lithuania
LT
0.102
Sweden
SE
0.016
Belgium
BE
0.064
China
CN
0.294
Hong Kong
HK
0.238
Table 8. Carbon Intensity of the Nations Used throughout the Study
3.6.2. Electricity Mix of the Three Main Nations
In order to get a better understanding of where these values are coming from, a deeper analysis
has been carried out by looking at the sources of generation of the three main nations. Data
were obtained by using the Cumulative Energy Demand (CED) indicator, developed by PRé
Consultants too. It is used to quantify the total energy demand throughout the lifetime of a
product or service, providing as well the amount of renewable and non-renewable sources
consumed. Therefore, this indicator could also be used to identify the different sources of
generation in 1 MJ of electricity (medium voltage) of a specific region.
40
Figure 19 represents the weight of each generation source in the three main regions Africa,
Europe and China. Non-renewable sources are represented in grey colours and renewable
sources are represented in bright colours. As it can be appreciated, China and Africa have a
very similar distribution, with about 90% of the electricity being generated from fossil fuels.
On the other hand, Europe has a greener mix with only 50% being generated from fossil fuels
and about 15% from renewable sources.
Figure 19. Electricity Mix of the Three Main Nations Used in the LCA Study
91.2%
2.6%
0.2% 0.7% 5.3%
Africa
48.3%
37.5%
3.5%
3.8%
7.0%
Europe
88.7%
2.8%
1.1% 1.2% 6.3%
China
41
4. Model Development
This chapter describes how the study has been developed by following the methodology
described in the previous chapter. This chapter is equivalent to phases 1 and 2 of an LCA study,
as described in Section 2.3.1. Section 4.1 describes the LCA criteria that has been employed
during the modelling process. Section 4.2 explains the scenario development and presents the
final outcome. Finally, Section 4.3 compiles the information that has been used for the LCI.
4.1. Selection of Criteria for LCA of Storage Technologies for PVWPS
This section is equivalent to LCA phase 1, where the key criteria for the LCA study is defined.
4.1.1. Functional Unit Definition
The definition of the functional unit is one of the key steps in order to properly develop an LCA
study (see Section 2.3.1). It will allow the comparison of results amongst the different storage
technologies as well as with other studies. The functional unit used to compare the four
different storage technologies was to meet Gogma’s water demand. In other words, the capacity
needed per storage technology to be able provide water at any time and never run out of storage.
These values were obtained from the previous study carried out by [31]. She found out that the
optimal sizes for the steel water tank and the lead-acid battery bank were 5m3 and 2160 Wh,
respectively. The optimal size used for the plastic water tank was the same as the steel water
tank. Nevertheless, the optimal size for the lithium-ion battery bank could not be the same as
the lead-acid battery bank due to the differences in the performance indicators (see Section
2.2.1), which significantly influence battery bank capacities.
Three studies that compared the performance of lead-acid and lithium-ion battery banks for the
same application were explored [114][116]. All of them showed a ratio in the rated capacity
of lithium-ion to lead-acid (Ah) of around 0.5. Following this procedure and as featured in
Table 9, the lithium-ion battery bank would have a rated capacity of 24.75 Ah.
Since this LCA study will be carried out for the lifetime of a PVWPS, which is around 20 years
(see [117]), the required replacements for the storage technologies must be taken into account.
Steel and plastic water tanks have an average lifetime of 15 20 years [118], so no
replacements would be needed. As covered in Section 2.1.4, battery banks would need
replacements. Average lifetime of battery technologies can be found in Table 9, as well as the
weight per battery. The number of batteries per set refers to the number of batteries required
per installation. Four batteries of 12 V connected in series are required per set to reach the 48
V requirement. The weight per battery bank is obtained by multiplying the weight per battery
42
times four. The number of replacements is obtained by dividing the lifetime of the PVWPS by
the lifetime of the battery.
Due to data availability and software requirements, the LCA had to be carried out per kg of
battery, so the total battery weight throughout the PVWPS lifetime was calculated to allow
comparison with the other storage technologies. It was calculated by multiplying the weight
per battery bank times the number of replacements.
Even though the functional unit to meet Gogma’s water demand is used throughout this study,
final results will also be represented per kg of storage technology to facilitate comparison with
other studies and allow future use (see Section 6.2).
Functional unit for batteries for PVWPS in BF
Feature
Lead-Acid
Lithium-Ion
Voltage (V)
48 [24]
48 [24]
Rated Capacity (Ah)
45 [24]
24.75 [114][116]
Battery Bank (Wh)
2160
1188
Lifetime (year)
4 [111], [112]
10 [111], [112]
Weight per Battery (kg)
15.3 [115]
2.9 [116]
Number of Batteries per Set
4
4
Weight per Battery Bank (kg)
61.2
11.6
Number of Replacements
5
2
Total Weight throughout PVWPS’s lifetime (kg)
306
23.2
Table 9. Functional Unit Definition for LABs and LIBs
4.1.2. System Boundaries Definition
One of the main differences and highlights of this study is its wide scope because of embracing
from material extraction all the way to disposal. This study has performed a cradle-to-grave
analysis (see Section 2.3.1) in the four selected storage technologies. Complementary
components such as pipes or electronic devices are outside the system boundary.
To simulate recycling processes, the system expansion approach has been implemented. As its
name indicates, it consists of expanding the system boundaries and substituting a process with
a different way of providing it [71]. To replicate that a fraction of metals from batteries are
recycled, the metal output from the recycling process was changed to a negative sign to
simulate its recovery. Figures and more detailed explanations of this procedure can be found
at the end of life sections in Section 4.3.
Sensitivity analysis with respect to the source of the electricity and transport means and
distances have been performed in order to assess how these factors affect the total
environmental impacts.
43
4.1.3. Impact Assessment Method Selection
Three impact assessment methods were described in Section 2.3.1 (ReCiPe, Eco-Indicator 99
and CML) and finally the ReCiPe methodology was selected because of being an improvement
of the other two. Currently, it is the most common and worldwide acceptable method [119].
The ReCiPe method determines indicators at two levels midpoint and endpoint - by using
two mainstreams of characterization factors [120]. There are 18 midpoint indicators (left hand
side) and 3 endpoint indicators (right hand side) (see Figure 20). The difference is that while
midpoint indicators focus on single environmental problems, such as global warming or human
toxicity, endpoint indicators sort environmental impacts in three higher aggregation levels.
Converting from midpoint to endpoint simplifies results but at the same time increases
uncertainty [120].
Figure 20. Overview of ReCiPe's structure adapted from [120]
The different characterisation factors that ReCiPe uses for each of the two mainstreams
endpoint and midpoint - can be found in the RIVM Report 2016 [120]. Midpoint and endpoint
indicators have been used to represent the results. Firstly, in order to get a quick overview and
identify the best and worst scenarios, endpoint indicators have been used for each one of the
24 scenarios. Endpoint categories are related to the three areas of protection and each of them
is expressed in different units, as described in Table 10.
44
Area of
Protection
Endpoint
Category
Unit
Meaning
Human
health
Damage to
human health
DALYs
(Disability Adjusted
Life Years)
Represents the years that are lost or that a person is disabled for
due to a disease or accident
Natural
environment
Damage to
ecosystem
quality
species.yr
(Species x years)
Loss of species integrated over time
Resource
scarcity
Damage to
resource
availability
USD2013
(US Dollar - $)
Extra costs dedicated for mineral or fossil resource extraction
Table 10. Overview of Endpoint categories adapted from [120]
To carry out a more detailed and accurate analysis, six scenarios have been selected from the
endpoint analysis: the baseline of steel water tank, the best