TH E EMP LOYME NT BEN EF IT S
OF AN E NERGY TRAN S IT ION
IN YEME N
APRIL 2023 // PREPARED BY THE WORLD BANK MENA ENERGY
i
©2023 The World Bank
International Bank for Reconstruction and Development
The World Bank Group
1818 H Street NW, Washington, DC 20433 USA
DISCLAIMER
This work is a product of the staff of The World Bank with external contributions. The findings,
interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World
Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not
guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and
other information shown on any map in this work do not imply any judgment on the part of The World
Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.
Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and
immunities of The World Bank, all of which are specifically reserved. In the case of any discrepancies
between this English version and any subsequent translations, the English version prevails. The report
reflects information available up to October 15, 2022.
RIGHTS AND PERMISSIONS
The material in this work is subject to copyright. Because The World Bank encourages dissemination of its
knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full
attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be
addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433,
USA; e-mail: pubrights@worldbank.org.
Please cite this work as follows: World Bank. 2023. “The Disruptive Energy Transition and Opportunities
for Job Creation in the Middle East and North Africa: Case Study—Yemen.�? World Bank, Washington, DC.
Cover: Architecture in Yemen
Photo credits: © Mohannad Khatib/Shutterstock
Back: Old city of Sanaa the capital of Yemen. View on the city from roof at sunrise
Photo credits: © Oleg Znamenskiy/Shutterstock
ii
CONTENTS
Acknowledgments ............................................................................................................................................................. vii
Executive Summary ............................................................................................................................................................. 1
1. Introduction ................................................................................................................................................................ 5
1.1 Country Context............................................................................................................................................. 5
1.2 Yemen’s Labor Market ................................................................................................................................. 5
1.3 Yemen’s Energy Sector and Electricity Consumption ........................................................................... 7
1.4 Solar PV in Yemen ...................................................................................................................................... 10
2. Evaluating the Employment Effect of Decentralized Solar PV .................................................................... 12
2.1 Research Methods ..................................................................................................................................... 12
2.2 Business Surveys......................................................................................................................................... 15
2.3 PV Value Chain Analysis and Solar Companies ................................................................................... 18
2.4 Market Sizing ............................................................................................................................................... 24
2.5 Employment Multipliers ........................................................................................................................... 27
2.6 Market and Employment Estimates ...................................................................................................... 37
3. Recommendations and Outlook ......................................................................................................................... 46
Annex 1. Exponential Growth of PV Markets ............................................................................................................ 50
Annex 2. Description of Sampled Companies ........................................................................................................... 52
Annex 3. Operations and Maintenance Mini Survey ............................................................................................... 55
Annex 4. Potential Development of Solar PV ............................................................................................................ 59
A4.1 Barrier Assessment ........................................................................................................................................... 61
A4.2 PV Investment Cases: Costs and Risks ......................................................................................................... 64
A4.3 Business Case Takeaways ............................................................................................................................... 90
Annex 5. Limitations and Future Research Needs .................................................................................................... 91
References .......................................................................................................................................................................... 94
iii
List of Figures
Figure 1.1 Summarizes the Status of Power System Assets in 15 Cities ................................................................ 7
Figure 1.2 Republic of Yemen, Annual Global Horizontal Irradiation (GHI) (kWh/m 2) ................................... 10
Figure 1.3 Solar Irradiation Values for the Yemeni Governorates and the Respective Shares of
the Population (kilowatt-hour per kilowatt-peak per year) .................................................................................. 11
Figure 2.1 Logistical Growth Model (Cumulative Growth)..................................................................................... 13
Figure 2.2 Location of Solar Companies in Yemen ................................................................................................... 17
Figure 2.3 Solar PV Value Chain in Yemen ................................................................................................................. 17
Figure 2.4 Actors Across Yemen’s Solar PV Value Chain......................................................................................... 19
Figure 2.5 Activities of Solar PV Companies (Most Popular Answers from the Online Survey) ................... 19
Figure 2.6 Key Activities of Solar PV Companies ....................................................................................................... 21
Figure 2.7 Solar Shop Activities ..................................................................................................................................... 22
Figure 2.8 PV Value Chain in Yemen—Upstream to Downstream Hardware Flow ......................................... 23
Figure 2.9 Measure Job Creation by Value Chain Element .................................................................................... 25
Figure 2.10 Full-Time Equivalent Units per Activity ................................................................................................. 28
Figure 2.11 Full-Time Equivalent Multipliers per Activity (Direct Formal Labor) ............................................. 29
Figure 2.12 Direct Formal Job Multipliers .................................................................................................................. 30
Figure 2.13 Skills of Interviewed Companies’ Employees ...................................................................................... 31
Figure 2.14 Direct Formal and Informal Job Multipliers ......................................................................................... 31
Figure 2.15 Multipliers: Benchmark and Comparator Study ................................................................................. 32
Figure 2.16 Job Profiles and Skills: “Needed and Scarce,�? Top Answers ............................................................ 33
Figure 2.17 “Easily Available�? Job Profiles and Skills: Top Answers from Survey ............................................. 34
Figure 2.18 Indirect Labor: Externally Purchased Products ................................................................................... 35
Figure 2.19 Indirect Labor: Externally Purchased Services..................................................................................... 35
Figure 2.20 Forecasted Annual PV Installations in Base-Case Scenario (S2), by Region and in
Total ..................................................................................................................................................................................... 38
Figure 2.21 Forecasted Cumulative PV Capacity in Base-Case Scenario (S2), by Region and in
Total ..................................................................................................................................................................................... 38
Figure 2.22 Forecasted Annual PV Markets in Three Scenarios: S1 (Best), S2 (Base), and S3
(Worst) ................................................................................................................................................................................ 39
Figure 2.23 Forecasted Cumulative PV Capacity under S1 (Best), S2 (Base), and S3 (Worst) ....................... 40
Figure 2.24 Details of PV Market Forecasts, by Scenario ....................................................................................... 40
Figure 2.25 Direct Labor, Upstream and Downstream, in Base-case Scenario ................................................. 41
Figure 2.26 Direct Labor Value Chain: Base-case Scenario .................................................................................... 42
Figure 2.27 Direct, Indirect, and Induced Labor: Base-case Scenario ................................................................. 43
Figure 2.28 Total Labor: Base-case Scenario ............................................................................................................. 43
Figure 2.29 Total Labor: Best-case Scenario .............................................................................................................. 44
Figure 2.30 Total Labor: Worst-case Scenario ........................................................................................................... 44
Figure A1.1 Historic PV Market Growth in the Middle East, 2000–18 ................................................................ 50
Figure A2.1 Overview of Primary Research ................................................................................................................ 52
Figure A2.2 Governorates Served by the Interviewed and Surveyed Companies ............................................ 52
iv
Figure A2.3 Solar PV Turnover and Employees Compared with Total Company Turnover and
Employees .......................................................................................................................................................................... 53
Figure A3.1 Responsibility for Battery Replacement ............................................................................................... 57
Figure A3.2 Battery Replacement Intervals ............................................................................................................... 58
Figure A4.1 Investment Climate for Solar PV ............................................................................................................ 60
Figure A4.2 Final Electricity Consumption (Share of Customer Segments), 2010 –18 ..................................... 64
Figure A4.3 Criteria for Financial Investment ............................................................................................................ 65
Figure A4.4 Residential Rooftop PV: Equity Cash Flow ........................................................................................... 69
Figure A4.5 Residential Rooftop PV: Specific Yield Sensitivity .............................................................................. 71
Figure A4.6 Residential Rooftop PV: System Price Sensitivity ............................................................................... 71
Figure A4.7 Residential Rooftop PV: Sensitivity of Generator’s Efficiency ........................................................ 72
Figure A4.8 Residential Rooftop PV: Fuel Costs ........................................................................................................ 73
Figure A4.9 Residential Rooftop PV: Fuel Cost Development ............................................................................... 73
Figure A4.10 Residential Rooftop PV: LCOE Development .................................................................................... 74
Figure A4.11 Health Center PV: Equity Cash Flow.................................................................................................... 78
Figure A4.12 Health Centers: Specific Yield Sensitivity ........................................................................................... 79
Figure A4.13 Health Centers: System Price Sensitivity ........................................................................................... 79
Figure A4.14 Health Center PV: Sensitivity of Generator’s Efficiency ................................................................. 80
Figure A4.15 Health Center PV: Fuel Costs ................................................................................................................ 81
Figure A4.16 Health Center PV: Fuel Cost Development ........................................................................................ 81
Figure A4.17 Health Center PV: Battery Replacement Intervals .......................................................................... 82
Figure A4.18 Solar-Powered Irrigation System: Equity Cash Flow (Loan Tenure, 2 Years) ............................ 85
Figure A4.19 Agriculture Water Pumping: Equity Cash Flow (Loan Tenure, 7 Years) ..................................... 86
Figure A4.20 Solar-Powered Irrigation System: Specific Yield Sensitivity .......................................................... 87
Figure A4.21 Solar-Powered Irrigation System: System Price Sensitivity ........................................................... 87
Figure A4.22 Solar-Powered Irrigation System: Sensitivity of Generator’s Efficiency .................................... 88
Figure A4.23 Solar-Powered Irrigation System: Fuel Costs .................................................................................... 89
Figure A4.24 Agriculture Water Pumping: Fuel Cost Development .................................................................... 89
v
List of Tables
Table ES.1: Actors Across the Republic of Yemen’s Solar PV Value Chain ............................................................ 2
Table ES.2: Educational attainment, Yemen overall and solar PV industry (large companies) ....................... 3
Table 1.1 Education Level of the Labor Force, 2013–14 ............................................................................................ 6
Table 1.2 Household Electricity Consumption and Generation Sources in Yemen ............................................. 9
Table 2.1 Educational Attainment Levels in Yemen Overall and in the Solar PV Industry (Large
Companies) ........................................................................................................................................................................ 21
Table 2.2 Key Assumptions (Market Forecasts) ........................................................................................................ 26
Table 2.3 Monthly Salaries Earned in the PV Industry ............................................................................................ 36
Table 2.4 Proportion of Total Direct Jobs to Total Productive Use Jobs ............................................................. 37
Table 2.5 Overview of Best-case, Base-case, and Worst-case Results, 2030 .................................................... 45
Table A3.1 O&M Activities, Work Split, and Frequency ......................................................................................... 55
Table A3.2 O&M Activities: Work Time by System Size .......................................................................................... 56
Table A4.1 Key Barriers to Investment ........................................................................................................................ 59
Table A4.2 Residential Rooftop PV System: Base-Case Assumptions ................................................................. 67
Table A4.3 Residential Rooftop PV: Financing Assumptions ................................................................................. 67
Table A4.4 Residential Rooftop PV: Savings .............................................................................................................. 68
Table A4.5 Residential Rooftop PV: Results ............................................................................................................... 69
Table A4.6 PV System at Health Center: Base-Case Assumptions ....................................................................... 75
Table A4.7 Financing Assumptions for PV System at Health Center ................................................................... 76
Table A4.8 Savings from PV System at Health Center ............................................................................................. 77
Table A4.9 Financial Results for PV System at Health Center ............................................................................... 77
Table A4.10 Solar-Powered Irrigation System: Base-Case Assumptions ............................................................ 83
Table A4.11 Financing Assumptions for Solar-Powered Irrigation System ........................................................ 84
Table A4.12 Solar-Powered Irrigation System: Savings and Revenues ............................................................... 85
Table A4.13 Solar-Powered Irrigation System: Financial Results ......................................................................... 85
vi
ACKNOWLEDGMENTS
Energy Sector Management Assistance Program (ESMAP) is a partnership between the World Bank and 24
partners to help low- and middle-income countries reduce poverty and boost growth through sustainable energy
solutions. ESMAP’s analytical and advisory services are fully integrated within the World Bank’s country financing
and policy dialogue in the energy sector. Through the World Bank Group (WBG), ESMAP works to accelerate the
energy transition required to achieve Sustainable Development Goal 7 (SDG7) to ensure access to affordable,
reliable, sustainable, and modern energy for all. It helps to shape WBG strategies and programs to achieve the
WBG Climate Change Action Plan targets.
This report on Yemen is one of three country case studies under the umbrella project The Disruptive Energy
Transition and Opportunities for Job Creation in the Middle East and North Africa. This project, made possible
by funding from the ESMAP, was initiated in 2019 in response to requests from various Middle East and North
Africa (MENA) governments to explore the nexus between the clean energy transition and employment. The
umbrella project is led by Tu Chi Nguyen and Ashok Sarkar (MENA Energy and Extractives Global Practice, World
Bank) and includes Abdellatif Touzani, Alona Kazantseva, Arslan Khalid, James Barrett, Kabir Malik, Laure Grazi,
Manjula Luthria, Mark Njore, Sarah Moin, Skip Laitner, and Yao Zhao.
This report was authored by Ulf Lohse, with research and modeling conducted by Ulf Lohse and Lucas Schimming,
both of the consulting agency eclareon GmbH, Berlin and Tawfik Al-Dhubhani and Mohammed Al-Khader
(consultants in Sana'a).
The task team wishes to thank senior management in the World Bank, particularly Paul Noumba Um (Regional
Director, MENA Infrastructure), Husam Beides (Practice Manager, MENA Energy and Extractives Global Practice)
and Erik Fernstrom (former Practice Manager, MENA Energy and Extractives Global Practice), for their continued
support and guidance.
Questions on this report may be directed to Tu Chi Nguyen (tnguyen19@worldbank.org), Ashok Sarkar
(asarkar@worldbank.org), and Yao Zhao (yzhao5@worldbank.org).
vii
EXECUTIVE SUMMARY
Renewable energy (RE) deployment is accelerating across the globe, contributing to renewables’ increasing
share of electricity generation. Renewable energy has the potential to support economies’ decarbonization
pathways, and to create significant employment opportunities. Global installed RE capacity is led by hydropower
at approximately 47 percent of installed RE capacity, followed by wind energy at 25 percent, and solar
photovoltaics (PV) at 23 percent (IRENA 2019a). Wind energy and solar PV play particularly important roles in
the current electricity supply of the war-torn Republic of Yemen.
The Republic of Yemen was already among the least electrified countries in the Middle East and North Africa
(MENA) region before the outbreak of civil war in 2014–15. The conflict has since destroyed important segments
of the country’s energy infrastructure, decreasing the availability of centralized electricity generation, and
hampering the country’s fuel supply. Many Yemenis have turned to distributed solar PV in their search for an
available and reliable energy source that can be installed relatively quickly. This development, although driven
primarily by a lack of energy-generating alternatives, is also supported by the country’s high solar potential and
the decreasing prices of solar components in the world market. The distributed PV solutions deployed in the
Republic of Yemen are off-grid systems ranging from pico-solar systems generating a few watts, to solar home
systems, most often with a capacity of less than 500 W, to installations generating one- to two-digit kilowatts.
Mostly, the PV systems are equipped with a battery to enable nighttime consumption. As shown by the World
Bank’s household surveys, as of 2019/2020, solar PV had come to occupy a significant share of the Republic of
Yemen’s electricity capacity, composing 75 percent of the electricity mix. The share was larger in the northern
than in the southern governorate regions,1 where more energy alternatives are available (World Bank 2020a).
To assess solar PV’s job creation potential, this report analyzes the value chain of the Yemeni solar PV industry
based on existing literature and primary research. The value chain encompasses both upstream and downstream
activities. In the upstream segment, work is created by importing the key solar components, such as the module,
the inverter, or the battery. These components are not yet manufactured on an industrial scale in Yemen, and,
oftentimes, manufacturing is viewed primarily as an activity related to installation as well as operation and
maintenance (O&M). The downstream segment includes work related to wholesale, distribution, installation,
and O&M. Today, in terms of the PV lifecycle, decommissioning of solar systems is not seen as a part of the core
business of solar companies, but some hardware recycling, especially of batteries, takes place.
The Yemeni PV industry is, broadly speaking, divided into two types of businesses: larger companies, with mostly
more than 10 employees, and solar shops with 1–3 employees. Solar shops can be further subdivided into shops
that are associated with the large companies and those that are not. While the large companies are active across
all value chain links, shops focus on selling directly to customers (business-to-customer distribution activities).
They may also be active across other parts of the value chain depending upon the skill set and the business
network of the shop owners and their employees.
1 The north includes 14 governorates located in the Azal, Janad, Saba, and Tehama regions. The south includes 8 governorates located
in the Aden and Hadramout regions.
1
Table ES.1: Actors Across the Republic of Yemen’s Solar PV Value Chain
Source: Original compilation.
Note: Solar companies and shops may be either associated or not associated with a large company. Activities that are performed as part
of the core PV business are in green. Activities that are performed occasionally but are not core are in yellow. Red indicates activities that
are not performed.
Although there are large companies with PV as their sole business line, the majority of large companies are also
active in other industries that are often indirectly related to solar PV. Within a company, the boundaries between
activities are often fluid, meaning that work is allocated based on skills and education rather than clearly
segregated departmental roles and responsibilities.
To quantify PV’s employment effect, two models were used: first, a market model was set up to forecast the
future size and development of the PV market in terms of megawatts (MW) installed; second, a socioeconomic
impact model was developed to estimate solar PV’s job creation potential. Input data for the models were
collected from existing literature on both the Republic of Yemen and international benchmarks as well as
through business surveys. The surveys, conducted both online and in person, gathered data about solar
companies in general and about their employment structures in particular. The results were used to identify
multipliers measuring direct jobs along the solar PV value chain. The identified multipliers were applied to three
different PV market growth scenarios to calculate the gross job creation by 2030. All three scenarios targeted
specific future electricity consumption levels in the northern and southern governorates. These consumption
levels were to be achieved after 5 years in the best-case scenario, after 10 years in the base case scenario, and
after 15 years in the worst-case scenario.
In addition to direct jobs, other job categories, namely, indirect, induced, and productive use jobs, were
estimated based on data points from the literature. Approximately 6,200 direct, 4,700 indirect, and 11,000
induced jobs are estimated to be created from solar PV in the base-case scenario by 2030. O&M accounts for
about 42 percent of direct labor, while other downstream activities account for about 47 percent, and PV
equipment imports account for about 11 percent of direct labor. An additional 10,000 jobs could be created by
productive use jobs. In a worst- and a best-case scenario, total jobs without productive use could range from
14,000 to 59,000. The identified employment potential of PV is even more impressive considering that before
the conflict, 7,000 people worked in “electricity, gas, and water supply�? jobs (International Labour Organization
(ILO) 2015).
Although the research undertaken focused on the quantification of jobs, some qualitative information on jobs
was obtained. During guided interviews with large solar companies, it was found that employees’ average
2
education level is apparently higher than that of the general populace in the prewar Republic of Yemen.
Table ES.2: Educational attainment, Yemen overall and solar PV industry (large companies)
Yemen Solar PV
overall companies
Primary education: 69% 6%
Secondary education: 23% 39%
Tertiary education: 8% 35%
Unknown/not answered: --- 20%
Source: ILO labor force survey 2013-14 and eclareon research, guided interviews
The primary research also assessed the qualifications needed in the Yemeni PV sector and those that might be
easily available.
Figure ES.1 Job Profiles and Skills: “Needed and Scarce,�? Top Answers
Source: Online survey conducted for this study.
Engineering and higher management skills were identified as being the most “needed and scarce,�? with 55
percent of the online survey’s respondents reporting engineers for project planning and design as scarce and 48
percent reporting engineers for quality assurance as scarce. Skills identified as “easily available�? were those that
required less formal education, such as support staff roles.
In the post-conflict Republic of Yemen, the number of jobs created will depend on the development of the
investment climate for PV in the future. The investment climate encompasses a number of factors that will
influence market actors’ decisions to invest in PV or not. Such decisions will be based on balancing perceptions
about the barriers, costs, and risks associated with a PV investment in the Republic of Yemen. The most
important barrier to be removed is the instability caused by the current conflict. This instability is pervasive
across all economic sectors and also has concrete negative impacts on solar PV: for example, the import of
electrical equipment, including solar panels, was temporarily blocked as the conflicting parties tried to (mis)use
the electricity infrastructure for military goals. In addition, equipment for solar has been subject to partially
informal transit duties levied by different parties (Ansari, Kemfert and al-Kuhlani, 2019). Abbreviations and
Acronyms
3
C&I commercial and industrial
CFM cash-flow model
CGE computable general equilibrium
FTE full-time equivalent
IEA International Energy Agency
ILO International Labour Organization
IRENA International Renewable Energy Agency
IRR internal rate of return
kWp kilowatt-peak
kWh kilowatt-hour
LCOE levelized cost of electricity
MFI microfinance institution
MW megawatt
O&M operation and maintenance
p.a. per annum
PEC Public Electricity Corporation
PV photovoltaic
RCREEE Regional Center for Renewable Energy and Energy Efficiency
RE renewable energy
S1, S2… scenarios 1, 2…
SEIM socioeconomic impact model
SHS solar home system
UN United Nations
UNOPS United Nations Office for Project Services
UNDP United Nations Development Programme
YRIs Yemeni rials
4
1. INTRODUCTION
In 2019, the renewable energy (RE) sector employed at least 11.5 million people worldwide. Of these, 3.8 million
worked in solar PV (Ansari, Kemfert, and al-Kuhlani 2019; IRENA 2019b). To date, RE employment opportunities
are mainly found in a limited number of countries, such as China, Brazil, the United States, India, and member
states of the European Union. Several factors—including the scale of national deployment, the magnitude of
industrial policies, changes in supply chains’ geographic footprint and in trade patterns, and industry
consolidation trends—determine how and where RE jobs are created. The transition from conventional energy
to RE and energy efficiency (EE) can lead to the creation of many more RE jobs. In its “Global Renewables
Outlooks 2020�? report, the International Renewable Energy Agency (IRENA) estimated that there could be 30
million RE jobs by 2030 and 42 million by 2050, and that EE jobs could employ up to 21.3 million people, doubling
the number of jobs in 2020 (IRENA 2020a). These new jobs are expected to exceed the jobs lost in fossil fuels
and nuclear energy, generating net job gains. However, since the case of each country is unique, country-specific
investigations are required to assess the effect of RE and EE growth on employment. In this context, the World
Bank, with funding from the Energy Sector Management Assistance Program (ESMAP), has conducted a series
of studies on job creation and replacement based on RE and EE development in three Middle East and North
Africa (MENA) countries: The Republic of Yemen, Morocco, and Egypt. This report presents the results for the
Republic of Yemen (hereafter, Yemen).
1.1 COUNTRY CONTEXT
The employment potential of renewables, especially solar energy, may be particularly important for Yemen
based on its high solar irradiation levels. Yemen has been in a state of conflict since 2014–15. It is classified as a
fragile country, and the many difficulties it faces today are described in several reports. The Organisation for
Economic Co-operation and Development (OECD) assesses countries’ fragility based on five dimensions:
“economic,�? “environmental,�? “political,�? “security,�? and “societal.�? In its 2020 report, OECD stated that
“Somalia and Yemen are an example of two extremely fragile contexts that are experiencing severe fragility in
all five dimensions�? (Desai and Forsberg 2020). The situation in Yemen was precarious before the conflict. It has
since become devastating: while about 47 percent of Yemenis were classified as being poor in 2014, poverty
among Yemenis was forecasted to reach 75 percent by the end of 2019. Seventy-nine percent of the population
would be living under the poverty line and 65 percent would be classified as living in extreme poverty (UNDP
2019).
Yemen has also been affected by the global trend of urbanization: while its rural population still constituted
approximately 71 percent of its total population in 2004, the share steadily declined to an estimated 63 percent
as of 2019. While population in urban areas is growing at 4 percent a year, growth is slower in rural areas, at 1.3
percent a year. In 2019, Yemen’s overall population was estimated to be growing at a 2.3 percent annual rate.
This indicates a decline from 2003, when it peaked at 2.9 percent (World Bank 2020b).
1.2 YEMEN’S LABOR MARKET
For a sustainable future and growth, Yemen needs both electricity and jobs. The country’s population has
increased tremendously in the past 50 years, and it is still growing. Yemen has a predominantly young
5
population: 58 percent are between 15 and 64 years old, with a further 40 percent younger than 15 years and
thus ready to enter the official workforce in a few years.
The majority of the information on the prewar Yemeni labor market is based on International Labour
Organization (ILO) reports, mostly “Demographic and Labor Market Trends in Yemen�? (ILO 2014a) and “Yemen
Labour Force Survey 2013-2014�? (ILO 2014b). The increase in labor market entrants means that the Yemeni
economy would have to create an average of 150,000 additional jobs a year to maintain a 75 percent
employment–population ratio (ILO 2014a). This is an important challenge to meet since many studies cited high
unemployment levels as a primary contributor to conflict (Drine 2012).
The ILO reported that before the war, about 50 percent of Yemen’s population were not in education,
employment, or training systems. Yemen’s labor force participation rate stood at just 34 percent. More than 50
percent of Yemen’s labor force had not completed compulsory education, and those with at least secondary
education were reported to represent around 30 percent (ILO 2014a).
Table 1.1 Education Level of the Labor Force, 2013–14
Total (%) Men (%) Women (%) Youth (15–24) (%) Adults(25+) (%)
Primary education 69 69 66 75 66
Secondary education 23 23 20 23 23
Tertiary education 8 7 13 2 10
Source: ILO 2014b.
A skills gap or deficit was identified as the most widely reported problem by employers in Yemen (ILO 2014a).
Of people in the labor force, almost 18 percent were unemployed, mostly those under 24. Youth unemployment
rates in Yemen were three times higher than adult employment rates—only 1 out of 5 young people was
employed, disaggregating to 1 in 3 young men and 1 in 40 young women. Poor job growth has led to a situation
in which the informal economy, characterized by irregularities and a lack of social protection (contracts, health
insurance, and benefits) makes up a large part of the country’s economy.
Yemeni women have some of the lowest labor participation rates in the region and the world. According to the
ILO report “Demographic and Labor Market Trends in Yemen,�? women constituted 12.7 percent of the country’s
labor force, 7 percent of the employed, and 73.7 percent of the inactive (ILO 2014a). The data do point to more
women gradually entering the labor market, but most often as unpaid family workers; part-time, low-wage
earners; but otherwise, they are unemployed or inactive.
Before the war, the Yemeni “electricity, gas, and water supply�? industry employed about 7,000 people (ILO
2015). This employment was classified as “informal�? for 1 in 7 employees. In the industry, women and men
worked, respectively, 27.3 and 38.1 hours per week on average, indicating that women work more often as part-
time employees.
Very little is known about Yemen’s labor market at the moment, although it has surely not improved since the
outbreak of war. The government could not pay many public employees. They have, therefore, de facto lost
their jobs, even though their contractual relationship with Yemen’s largest employer, the state, has not ended.
6
With regard to informal labor in Yemen, it should be noted that there are distinct definitions: Yemen’s Central
Statistical Organisation (CSO) defines employment in the informal sector as including employment taking place
outside the premises of establishments. This study uses the international definition used by the ILO, by which
informal labor “refers to the unincorporated nature of the enterprise, its employment size, and the non-
registration of the enterprise or its employees under specific forms of national legislation.�? According to this
definition, a study paper written for the ILO found that as much as 94 percent of all private establishments with
fewer than 5 employees would be part of the informal sector (Omeira 2013). This information, mentioned in a
2013 report, refers back to a 2002–03 ILO survey and is, therefore, only a very rough estimate of the situation
today. Nevertheless, the number can serve as an indicator of the magnitude of informal labor in Yemen.
1.3 YEMEN’S ENERGY SECTOR AND ELECTRICITY CONSUMPTION
Yemen was considered the least electrified country in the MENA region even before the current conflict broke
out: pre-crisis electricity access rates differ between various sources of information. One World Bank report
mentions a “pre-crisis access rate from all sources of only 55 percent�? (World Bank 2017a), while another speaks
of “66 percent [of the population] having access to public electricity and another 12 percent to private electricity
solutions [in 2014]�? (World Bank 2018).
It is relatively easy to measure the power that is generated centrally in Yemen in large power plants operated
by a single public entity such as the Public Electricity Corporation (PEC), but it is rather challenging to measure
decentralized energy generation. This is because the existing large and countable power plants are replaced by
a myriad of small generators, and there is no longer a central actor but multiple decentralized actors with
different sizes, backgrounds, and preconditions.
Yemen’s public power supply infrastructure, involving electricity generation, transmission, and distribution, was
partially or completely destroyed during the conflict and therefore can no longer supply electricity.
Figure 1.1 Summarizes the Status of Power System Assets in 15 Cities
100%
10% 14% 11%
90% 21% 20%
80% 22%
3%
70%
45% 20%
71%
60% 30%
50% 100%
40% 86%
30% 60% 67%
20% 45% 45%
29%
10%
0%
Power plant Distributed Substation - Substation - Tower Transformer Admin Office
generation Distribution Transmission
unit
No damage Partially damaged Completely destroyed Unknown
Source: Yemen Dynamic Needs Assessment (DNA): Phase 3—2020 Update, World Bank estimates 2020
7
Regarding the regional distribution of the cumulative damage, the southern and eastern regions of the country
were not as affected by the war as the northern and western regions (World Bank 2017b). This difference is also
a reason for the different electricity consumption levels between the northern and the southern regions, as
reflected in the household survey shown in Table 1.3. According to reports from Yemen, the situation of
centralized electricity supply has improved since 2017, especially in the southern governorates, but the situation
of energy supply from the national grid is still dire and a reason for the growth of solar PV.
Another factor that has prompted the growth of solar PV in Yemen has been the scarcity of fuel products, which
led either to high prices or to the complete unavailability of fuel. As a consequence, diesel- or fuel-powered
generators could not be operated at all or lacked sufficient reliability to meet the energy demand. In 2014, about
half of the population relied on electricity from private generators, which mostly used fuel-powered engine–
generator set (gensets) to generate electricity (UNDP 2014). These people were therefore independent from the
grid, although fuel availability was still a factor for them.
Since other sources of electricity could not be accessed, solar PV started to boom in the country after the conflict
broke out. However, the market did not emerge only in 2014–15: Solar PV had started to be used as a backup
electricity source back in 2010–11, when the Arab Spring ended Ali Abdullah Saleh’s 33-year presidency. This
forced government change and resulted in the energy infrastructure also coming under attack, leading to power
cuts and unstable supply. As a result, Yemenis started to invest in private independent power supply solutions,
for example, diesel generators but also solar PV.
The already low access to energy in general and electricity in particular declined further after the conflict between
the Houthi and the Yemeni government broke out. The UN reported in 2019 that “only an estimated 10 percent
of the population has access to reliable electricity and the consumption has decreased by about 75 percent�?
(United Nations 2019). A recent survey, conducted on behalf of the World Bank among 1,000 randomly selected
households, confirmed that at a national level, about 12 percent of the population receive electricity from the
national grid as the primary source of energy (World Bank 2020a). However, the survey also painted a more
differentiated picture of the current electricity supply situation and illuminated the different levels of electricity
access and power supply between the northern and southern regions: the households of the southern regions,
accounting for 20 percent of the population, have more energy at their disposal (680 kilowatt-hours [kWh] per
capita per year) than the northern households (267 kWh per capita per year). Moreover, in the southern regions,
63.5 percent of households used, or were able to use, the grid as the primary energy source, while in the North,
this was the case for only 1.2 percent of the population. Solar PV represents the primary energy source for about
30 percent of the population in the south and about 85 percent of the population in the north. In both parts of
the country, the evening availability of PV accounts for nearly 40 percent of the total availability, which shows the
importance of batteries for the Yemeni market. Table 1.3 lists the estimated shares of different electricity levels
for Yemen overall, in the north, and in the south.
8
Table 1.2 Household Electricity Consumption and Generation Sources in Yemen
For the
Access to
Residential Entire
Average Daily Residential Source as
Yearly Population,
Primary Electricity Availability Evening Estimated Yearly Share of a Primary
Consumption incl. HHs
Source (and share (Day and Availability Loads Consumption Generation or
per w/o
of total) Night) per Person Secondary
Householdc Electricity
Source
Access
(Hours) (Hours)a (Watts)b (kWh) (kWh) (kWh) (%) (%)c
National
12% 15.5 4.8 1,377.0 934.8 103.9 99.3 30.1
grid 32.0
Private grids 12% 19.6 5.7 896.0 769.2 85.5 81.7 24.7
Solar PV 75% 13.1 5.0 393.0 1,404.0 156.0 149.1 45.2 82.0
All sources
14.0 5.0 567.5 3,108.1 345.3 330.1 100.0 114.0
(average)
Evening
availability of 38.3%
PV
Northern regions
(Hours) (Hours)a (Watts)b (kWh) (kWh) (kWh) (%) (%)c
National
1.2% 21.1 5.8 2,291.0 211.7 23.5 22.5 8.4
grid 21.5
Private grids 12.8% 20.1 5.8 812.0 762.5 84.7 81.0 30.4
Solar PV 85.0% 12.9 5.1 384.0 1,536.9 170.8 163.2 61.2 91.5
All sources
13.8 5.1 457.8 2,511.1 279.0 266.7 100.0 113.0
(average)
Evening
availability of 39.5%
PV
Southern regions
(Hours) (Hours)* (Watt)** (kWh) (kWh) (kWh) (%) (%)***
National
63.5% 14.8 4.7 1,476.0 5,063.1 562.6 537.8 79.1
grid 79.5
Private grids 6.3% 11.5 4.8 940.0 248.6 27.6 26.4 3.9
Solar PV 30.3% 14.2 5.5 696.0 1,093.0 121.4 116.1 17.1 38.0
All sources
14.4 5.0 1,207.4 6,404.7 711.6 680.3 100.0 117.5
(average)
Evening
availability of 38.7%
PV
Source: Consumption and generation share calculated based on a World Bank household survey of a random sample of 1,000 respondents
(World Bank Group 2020a).
a Availability from 6pm to 12am.
b Based on standard load factors.
c Consolidated number for the national and private grids.
As a response to the lack of energy, international organizations such as the United Nations and the World Bank
have initiated programs to support the country’s reelectrification. Since the reconstruction of large-scale
electricity infrastructure already proved to be “close to impossible�? before the conflict, mainly due to inadequate
9
institutional capacity, and distributed solutions showed better results (World Bank 2018), the current strategy
focuses on the deployment of decentralized energy generation capacities and especially on solar PV. The
destruction of important parts of the public electricity infrastructure and the unavailability of fuels have made
solar PV the only available source of electricity for many Yemenis.
1.4 SOLAR PV IN YEMEN
Yemen’s irradiation values are among the highest in the world. This makes the use of solar energy very attractive
from a technical standpoint.
Figure 1.2 Republic of Yemen, Annual Global Horizontal Irradiation (GHI) (kWh/m2)
Source: © 2019 The World Bank, Global Solar Atlas 2.0; solar resource data from Solargis (World Bank 2020d).
The electricity that can be generated by solar PV depends on multiple factors, such as the irradiation at a specific
site; the system’s age; losses due to shading, soiling, and other factors; and the technical quality of the
installation and operation and maintenance (O&M) services, among other factors. This being said, it is still true
that the sunnier a region, the better it is for deploying solar PV. Figure 1.4 details the yearly solar irradiation of
the different governorates of Yemen.
10
Figure 1.3 Solar Irradiation Values for the Yemeni Governorates and the Respective Shares of the
Population (kilowatt-hour per kilowatt-peak per year)
Source: Original calculations based on the Global Solar Atlas (World Bank 2020d).
Note: Specific PV output assumes 5 percent losses.
Despite all governorates having favorable solar conditions, the difference between the governorate with the
highest irradiation (Al-Baida, 1,914 kWh per kilowatt-peak) and the lowest (Al-Hodeidah, 1,520 kWh/kWp)
amounts to approximately 20 percent. It must also be considered that some very sunny regions (such as Al-
Maharah) are home to only 0.5 percent of the overall population. Therefore, relatively few people can take
advantage of the high irradiation levels in such regions.
11
2. EVALUATING THE EMPLOYMENT EFFECT OF DECENTRALIZED SOLAR
PV
Since the objective was to be able to quantify the employment effect that solar PV could have on Yemen’s labor
market, this study had to use prewar data for the analysis: the development of every market asks for
predictability and stability as prerequisites, conditions that are rare in a war-torn country. The current conflict
has affected all economic sectors, including solar PV. For example, the import of electrical equipment, including
solar panels, was temporarily blocked, since the conflicting parties tried to use (or misuse) the electricity
infrastructure for their military goals (Ansari, Kemfert, and al-Kuhlani 2019). This example makes it clear that in
Yemen’s context, the most important prerequisite is to end the current conflict and stabilize the country, be it
for the development of PV or for any other development. This will be the ultimate enabler to substitute the war-
induced irrationality with rational behavior and replace a black market with a regular economy, switch from
short-term thinking to long-term planning, and make it possible for people to stop living in survival mode and
return to everyday life with an opportunity to make basic but fundamental choices, such as which type of
electricity they want to use and how they want it to be generated.
It has been this lack of choice that has caused solar PV to boom. As noted earlier, PV started to emerge in 2011,
and the breakdown of the centralized energy infrastructure and the increasing difficulties in securing fuel supply
for decentralized generators led it to become the sole reliable source of electricity for many when the war broke
out in 2014–15. Some of the destroyed infrastructure had been rebuilt by 2020 and some households were
receiving electricity supply from centralized generation once again. However, fuel supply remains inconsistent
and fluctuates, according to multiple factors, such as war activities, availability of infrastructure for transport,
and also political actions. For example, in June 2020, fuel imports declined substantially after political and
economic actors could not agree on roles and responsibilities with regard to fuel imports and clearing of ships
(World Bank Group 2020b). This shows at least two things about present-day Yemen: first, the predictability of
the energy supply situation is low due to the unstable conflict environment; second, the consequence of this
unstable environment negatively affects the availability of fuel, and, ultimately, energy in Yemen. This
unavailability has two main consequences, which will hit the least economically secure the hardest: Generation
will decline, because there is less availability of fuel that could be used to generate electricity. Moreover, the
supply shortage drives up prices, thereby making fuel again unaffordable for many.
2.1 RESEARCH METHODS
This study sought to quantify the potential employment to be created by decentralized PV in Yemen by following
two key steps:
1. Market size estimation using a market sizing model
2. Estimation of jobs to be created, using a socioeconomic impact model (SEIM)
First, information on per capita electricity consumption, population size, and the current PV market size was
used to forecast the PV market potential, which was defined as the megawatts required to instantly satisfy 100
percent of the demand for PV. Since demand cannot necessarily be satisfied instantly when it arises, a market
growth function was used to transform the market potential into a “realizable�? market size.
12
For the second step, job multipliers to estimate direct labor were identified based on primary research. Estimates
for indirect, induced, and productive labor were based on literature reviews.
ESTIMATING MARKET SIZE
Given the absence of PV market forecasts for Yemen, a PV “market sizing model�? was used to estimate the future
development of the Yemeni PV market in megawatts.
The most relevant input data, including current population size and growth, annual per capita electricity
consumption, installed PV at model start, and a hypothetical share of PV in Yemen’s future electricity mix, were
defined based on available data. The hypothetical share of PV in Yemen’s future energy mix takes the current
share of PV as a reference point. As shown in Table 1.3, PV is the primary energy source for about 85 percent of
the population in the northern governorates and 30 percent in the southern governorates. It was assumed that
as more generation alternatives become available in a more peaceful future, postwar shares will be lower.
As a first step to estimate the future market size, a “total PV potential�? was calculated based on the demand for
electricity generated by PV. Due to certain capacity constraints, an existing market potential cannot always be
realized fully and instantly when demand arises. Such constraints include, for example, a need to manufacture,
order, and import hardware in the upstream segment of the value chain, and a need for people to be hired and
trained in the downstream parts segment who are able to install, sell, or maintain a PV system. Whether instant
and complete satisfaction of demand is possible depends also on the maturity or size of the PV market. While a
1 megawatt (MW) market may well double in a year, this will become more difficult as the market matures, and
installed capacities have already reached a significant level. It is unlikely that a multi-megawatt PV market with
a significant discrepancy between installed capacity at model start (t=0) and a huge theoretical potential one
year later (at t=1) will be able to absorb the entire potential in one year. More realistically, PV markets grow
exponentially at first, then reach their highest growth rate at some point in time, and then continue to grow at
a slower rate when they approach an upper threshold. This upper threshold is the market potential. An ideal
market growth curve, representing a cumulative PV market size, would look like the S-curve in figure 2.1.
Figure 2.1 Logistical Growth Model (Cumulative Growth)
MW CUMULATIVE
-20 -17 -14 -11 -8 -5 -2 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49
TIME (YEARS), (T= 0 = THIS YEAR)
Source: Original compilation.
Historic market data from the International Renewable Energy Agency (IRENA 2020b) show that national PV
13
markets behave like this. Market growth curves for mature PV markets, such as Germany or Italy, resemble an
S-curve. In the SEIM tool, such a curve was modeled using an exponential growth function (for more details,
please refer to annex 1).
ESTIMATING JOBS CREATED
The literature describes different methods to quantify employment and describe green jobs—for example, jobs
in solar PV in a qualitative way. This section does not restate all the details of the existing literature, but simply
presents the most relevant aspects of estimating employment effects in order to provide a sound understanding
of what type of employment is assessed with what types of methods.
The ILO, in its practitioner’s guide, “Assessing Green Jobs Potential in Developing Countries,�? distinguished the
following three methods to estimate environment-related employment (ILO 2011):
• Business surveys: Employment numbers and qualitative employment criteria are assessed by surveying
companies.
• Employment factor or multiplier approach: The number of jobs created for a unit of a market size are
multiplied by the market size. Jobs are counted in full-time equivalents (FTEs), one of which corresponds
to one person working full-time. This approach relies on data collection and a literature review. The job
multiplier is applied to a specific market size.
• Input/output (I/O) tables: This is a statistical method to calculate the interdependencies between
sectors of an economy.
Moreover, there are also computable general equilibrium (CGE) models, a class of economic models that use
actual economic data to estimate how an economy might respond to changes in policy, technology, or other
external factors.
In addition to the selected method, other dimensions, which again may be different for each country, can be
considered, for instance, type of employment, gross and net employment, and units of measure (also called
employment categories).
TYPE OF EMPLOYMENT
• Formal employment: This is a measure of those who are “officially�? employed, earn a salary, and
contribute fully to a country’s social security systems.
• Informal employment: This counts those who do not have an employment contract and do not
contribute fully to a country’s social security systems, but may still earn an income based on an activity.
GROSS AND NET EMPLOYMENT
• Gross employment: Looks at the number of jobs created by solar PV.
• Net employment: Assesses the number of jobs created and the number of jobs displaced due to the
introduction or expansion of a specific technology, such as solar PV, compared with a counterfactual (for
example, a similar expansion of conventional energy).
UNITS OF MEASUREMENT (ALSO CALLED EMPLOYMENT CATEGORIES)
14
• Direct labor: Those persons who, if asked where they work, would answer, “I work in the solar PV
industry.�?
• Indirect labor: Those persons who, if asked where they work, would answer, “I don’t work in the solar
PV industry, but companies from the solar PV industry are my customers and buy products or services
from my company.�?
• Induced labor: Those persons who, if asked where they work, would answer, “I don’t work in the solar
PV industry, and I do not supply products or services to solar PV companies either. However, I earn parts
of my income by selling non-PV-related products or services to the people employed by solar companies
or by their suppliers.�?
• Productive use labor: Those persons who, if asked where they work, would answer: “I don’t work in the
solar PV industry and do not supply products or services to solar PV companies either, and also do not
necessarily sell any products or services to the people employed by solar companies and by their
suppliers. However, thanks to PV, I have access to (additional) electricity that I did not have before or
could not get, at least not for the same price, from another source. I do use this PV electricity to generate
additional income.�?
The boundaries between these units of measurement are fluid, and perspective plays a role in setting employees
in one category rather than another. For instance, a truck driver employed by a solar company who exclusively
transports solar PV equipment may be classified as direct labor, while a truck driver employed by a logistics
company that also transports other goods could be classified as indirect labor.
Other dimensions, such as the qualifications of the people working in the industry, the types of jobs that exist
along a PV value chain, and the quality of labor, may also play a role.
For this report, company surveys were conducted to find multipliers, which were then used to estimate direct
jobs created by solar PV (for more information, please refer to section 2.2). The formula used was this:
������������������������
������������������������ ������������������������ ������������������������������������������������������������������ (������������������������) = Market size (MW) ∗
������������
Indirect, induced, and productive use jobs were assessed based on information found in the literature. The
assessed effects are the gross employment effects.
2.2 BUSINESS SURVEYS
It is a challenge to obtain data that allow one to quantify a fragmented market in a war-torn country where
public data are mostly not available. Despite some available reports, the information from the literature was not
sufficient to quantify the PV labor market. Therefore, it was decided to gather primary data in Yemen to better
understand the country’s PV market. The following preparatory work was undertaken:
1. A review was conducted of available statistics and secondary literature, such as reports and websites
about the prewar situation of the energy market, and the current situation with regard to solar PV and
electricity supply.
2. Expert interviews were conducted to validate the data and other information from these reports in order
to better understand the Yemeni solar PV market and discuss its challenges, opportunities, and
prospects.
15
3. Based on steps 1 and 2, an online survey was designed, discussed, adapted, translated into Arabic, and
distributed to all known Yemeni solar PV companies that could be contacted by email:
• The data for this list of companies came from the United Nations Office for Project Services (UNOPS),
which sent a list of 57 solar companies. Forty-eight of these companies were based in Sana’a (city or
region), 3 in both Aden and Al-Hudaydah, 2 in Hadramout, and 1 in Taiz.
• The list was extended by adding more solar companies. These included companies mentioned
in reports or identified by a local consultant from Sana’a. The final list included 139 companies,
of which only 92 had a known email address. Of these companies, a large majority was still from
Sana’a city or the Sana’a region (60 percent).
• The main objective of the online survey was to gather information about the solar PV value
chain, the size of the companies, the number of solar employments, and the job profiles of those
working in solar PV. Moreover, the online survey attempted to gather information on job
multipliers that could complement the results of guided interviews conducted in Yemen.
4. A local consultant from Sana’a conducted guided interviews with 10 solar companies in Yemen to gain a
more detailed understanding of the solar value chain and gather information that would allow for the
calculation of job multipliers to quantify solar employment in the country. The guided interviews were
based on a detailed questionnaire, which included questions on the following topics:
• Situation of the (overall) electricity market, before and after the conflict
• Energy supply of households
• Energy supply of nonresidential (that is, commercial and industrial) consumers
• PV market and users
• PV systems and prices
• Obstacles to and drivers of PV market growth
• PV value chain
• Current structure of the Yemeni PV industry
• PV direct jobs (formal and informal)
• PV indirect jobs (formal and informal)
• Productive use jobs (households)
• Socioeconomic effects of PV (beyond employment)
In addition, the consultant also visited solar shops to gather information on this company type.
16
Figure 2.2 Location of Solar Companies in Yemen
Source: UNOPS, RCREEE, and research conducted for this study.
The questions in the online survey and the guided interviews were not identical. This was in order to account for
the fact that some questions could not be asked in an online survey without additional guidance. Furthermore,
that people would spend less time on an online survey than on a scheduled phone or face-to-face interview was
also considered.
Based on the preparatory work (steps 1 and 2) and then based on the research results, a value chain for Yemen’s
PV industry was developed, as depicted in figure 2.3.
Figure 2.3 Solar PV Value Chain in Yemen
Source: Original compilation.
The upstream part of the value chain defines how solar PV hardware, such as the module, the inverter, the
batteries, or controllers, flow into the market, while the downstream part describes how the PV systems are
sold, installed, and maintained. The solar PV value chain used for job estimations will be explained in more detail
in section 2.3 of this report. A description of the sampled companies can be found in annex 2.
17
2.3 PV VALUE CHAIN ANALYSIS AND SOLAR COMPANIES
At this moment, the Yemeni solar PV value chain is driven almost entirely by the private sector, which is
organized along an industry value chain that represents the aggregation of all core activities taking place at a
particular time in a specific industry in a particular country. When analyzing an industry value chain, it needs to
be remembered that every company that is active in the industry under review has its own individual activity
profile. Thus, the results of the primary research were used to analyze the value chain and come up with a profile
that best fits the Yemeni market.
When looking at solar PV employment studies, the only value chain elements are upstream, downstream, and
O&M activities. Sometimes these elements are detailed further through a number of more granular value chain
elements. When defining a value chain, the “right�? level of detail will depend on the available data about the
market and the value chain’s purpose. The level of detail that can reasonably be assigned to a value chain also
depends on the level of specialization attained in a given country at a specific point in time. The boundaries
between value chain elements (or links) are often fluid in the sense that people working for solar PV companies
often work across different functions and cannot easily be assigned to one specific value chain element.
This was also the case in Yemen. For example, the results of the interviews made clear that it was difficult for
the interviewees to distinguish between a technician working in installation and one doing O&M-related work
because the same people could and would do both jobs. For the purpose of this study, however, it is important
to be able to distinguish between these two activities, since the drivers for job growth are different for the two
value chain elements: installation jobs (FTEs) are driven by annual demand, whereas O&M jobs are driven by
cumulative market capacity. Moreover, some functions and value chain elements are often not separable:
system design in the downstream part of a value chain is an important activity for non-customized PV
installations. It may be included as a distinct building block in a value chain or as part of sales and installation.
Provided one has the hardware, one can sell and install PV systems without system design. However, the
opposite is not true: one would not design PV systems if there were no PV installations or sales.
The solar PV value chain that was used to structure the questions of both the online survey and the guided
interviews was based on the PV values and supply chains identified for Yemen.2
The companies that were interviewed and participated in the online survey were found to be formal private
companies, which are large by Yemeni standards, with mostly above 10 employees. These companies were
found to be active across all or at least many links of the value chain. Moreover, a local consultant also visited
solar shops selling PV equipment in order to learn about their business with regard to solar PV. The surveys and
the visits to small shops resulted in figure 2.4, developed to illustrate the type of solar companies and their
activities within the solar value chain. A traffic light system was used to visualize the typical activities of the two
typical company types, (large) solar companies and shops, either associated or not associated with a large
company.
2For example, RCREEE’s supply chain in Assessment of Solar PV in Yemen (RCREEE 2020) and those for decentralized solar—for
example, Gogla’s “Off Grid Solar Product Journey�? in “Off-Grid Solar: A Growth Engine for Jobs�? (GOGLA 2019), or Power for All’s “Local
Value Chain Scope�? in the “Powering Job Census 2019�? (Power for All 2019).
18
Figure 2.4 Actors Across Yemen’s Solar PV Value Chain
Source: Original compilation.
Note: Activities that are performed as part of the core PV business are in green. Red indicates activities that are not performed.
Activities that are performed occasionally but are not core activities are in yellow.
SOLAR PV COMPANIES
Large companies are active across many segments of the solar PV value chain. They typically employ more than
10 employees, although this may vary considerably by company, as was already described in section 2.2. Such
companies would be formally registered, employing a large share of formal employees.
In the 27 answers received in the online survey, 167 value chain–related activities were mentioned. Installation
(downstream component) and importing were the most widespread value chain activities (figure 2.5).
Figure 2.5 Activities of Solar PV Companies (Most Popular Answers from the Online Survey)
Source: Online survey conducted for this study.
The guided interviews yielded similar results: on average, every interviewed company was actively carrying out
5.9 activities related to the solar PV value chain. All 10 companies interviewed were importers, and nine were
also installers. In both the online survey and the guided interviews, decommissioning- and manufacturing-
related activities were the least developed. In the guided interviews, only two companies confirmed that they
were active in manufacturing and two confirmed that they pursue activities related to the decommissioning of
PV power plants. In the online survey, 3 out of 24 companies stated that they were active in decommissioning.
19
However, the interviewed companies did not assign any FTEs to decommissioning activities, which is a data
inconsistency that can be explained by fluid boundaries between activities, especially when it comes to O&M
and decommissioning. The companies that confirmed that they are active in decommissioning were potentially
not referring to entire PV systems but rather to single PV components, especially the battery, which needs to be
exchanged regularly and would then have to be counted as an O&M activity.
It is not surprising that the decommissioning of entire PV systems (as opposed to the exchange of single
components that may have a useful life of about 20 years) does not yet play a big role in Yemen, where the
market emerged around 2011. Because of these considerations and because of an absence of FTEs for
decommissioning, the activity was not mentioned in the value chain. This should, however, be reconsidered for
future research when the market becomes more mature and a significant number of PV systems will approach
the end of their useful life, which should be the case beyond 2030. Another interesting, albeit rather scientific,
question is to what extent the decommissioning of solar systems, in Yemen or elsewhere, should be assigned
directly to the solar PV value chain or whether it will be part of another industry like the circular economy. This
would then have to be counted as indirect labor for solar PV in employment studies such as this, since the
disposal and recycling of the old system would be a service provided to the PV industry. A preliminary answer to
a similar question about O&M activities related to batteries was already given by UNOPS: the collection,
recycling, and replacement are often done by scrap-collection shops (UNOPS n.d.), which have their own value
chain but are also linked to the solar PV industry (refer to section 2.5 of this report).
Regarding manufacturing, the information obtained was somewhat more concrete, although the interviewees
had still not given quantifiable FTEs for manufacturing activities. It can also be argued as to whether the
“manufacturing�? mentioned by the interviewed practitioners is not again, at least partially, an activity that could
also be seen as belonging to another element of the value chain, most notably, O&M. Regarding what types of
PV-related products are manufactured, the following were mentioned: assembly of baseboards and collection
boxes, solar panel bases, pump tubes, mounting structures for solar PV, batteries, inverters, cables, and also the
assembly of PV modules.
That there was an industrialized assembly of key components such as PV modules or inverters could not be
confirmed. The interviewees confirmed that key PV components are imported, but that eventually single or
multiple broken modules or inverters are eventually repaired to make them functional and, thus, valuable again.
The importance of imports over manufacturing activities in the upstream part of the PV industry also does not
come as a surprise considering that before the conflict, Yemen covered most of its needs across most industries
with imports, which were relatively cheap because of the high value of the Yemeni rial from oil exports (Nasser
2018). Still, recognizing that there are already some manufacturing-related activities, the activity was colored in
yellow, although it could not be fully quantified as a separate activity based on the collected data, since the
interviewees chose to assign the people who did manufacturing work to other parts of the solar value chain as
well.
Solar companies are the primary importers of PV hardware. They are active in sales, including both wholesale
(business to business, B2B) and distribution (business to customer, B2C), as well as the installation of larger
systems. These larger companies would also be the ones to offer O&M contracts for larger systems. The activities
of solar PV companies are summarized in figure 2.6.
20
Figure 2.6 Key Activities of Solar PV Companies
Source: Original compilation.
Analysis of the data on the educational attainment of the interviewed companies’ employees showed that their
average education level was higher than the prewar average for Yemen: 74 percent of those working in solar
jobs were reported to have completed secondary or tertiary education, whereas in prewar Yemen, this was only
the case for about 31 percent of the overall labor force.
Table 2.1 Educational Attainment Levels in Yemen Overall and in the Solar PV Industry (Large
Companies)
Source: ILO (2014b) and guided interviews conducted for this study.
SOLAR PV SHOPS
The second company type considered, solar shops, primarily sell solar PV products but may also offer other PV-
related services. Since shops were not as accessible as large companies during the primary research phase, the
local consultant visited two shops to learn about their structure and role in the solar PV market. Shops can be
categorized largely into two types: those that are associated with the large companies described above and
those that are not. Associated shops may receive more stock of solar components and may also be legally owned
by a large solar company, though this is not always the case. Very few shop owners are directly employed by
large companies, but most are associated with big PV suppliers.
Solar shops most often employ one to three persons, including, generally, the owner. Employees may be family
members or close relatives. The people working for solar shops do not always receive fixed salaries, and they
work, in most cases, from 8am to 8pm six days a week. In essence, shops combine both formal and informal
elements, and have mainly informal employees.
21
On average, small shops are estimated to have sold 25 kW of pico-scale solar3 and solar panels, as well as other
components, such as batteries, per year. The shops stated that sales of solar products had declined considerably
since 2018 and that their sales were 40–50 percent less than in 2016. Most of the shops source their products
from the wholesale market, with associated shops sourcing from their associated large companies but also
through inter-shop trading. Non-associated shops in particular were reported to team up to import equipment
themselves so that they would not have to purchase hardware from the large companies and were able to offer
lower prices by “cutting out the middleman.�? Most shops also sell other electrical appliances and electric
products and accessories. Those shops that exclusively sell PV products and related accessories spend up to 70
percent of their working time on sales. If PV products are only a part of their product range, they would spend
less than 40 percent time selling PV. Only a few shops sell full systems, except when it comes to pico-scale solar
systems, with an approximate size of between 5 and 30 W. Most small shops offer installation and repair services
as they also have common networks with each other and with PV technicians. This installation- and O&M-related
work represents an additional income source for small shops. The typical activities of solar shops are summarized
in figure 2.7.
Figure 2.7 Solar Shop Activities
Source: Original compilation.
A maximum of 100 large solar companies and approximately 2,000 shops more or less specializes in solar PV.
Based on the interview results and the PV volumes handled in each segment of the value chain, the flow of PV
components between large companies and solar shops is depicted in figure 2.8.
3Pico-scale refers to a small-scale and autonomous energy system that can be used for rural electrification. It can power efficient
appliances, such as table lamps, fans, radios, and phone chargers.
22
Figure 2.8 PV Value Chain in Yemen—Upstream to Downstream Hardware Flow
Source: Original compilation.
One hundred percent of (core) PV products are imported in the value chain’s upstream segment. Manufacturing
does not occur, with the exception of mounting structures, which are manufactured locally but are most often
seen as part of installation or purchased externally (indirect labor for PV). The interviewed solar companies
distributed about 14 percent of the imported volume to final customers on average. It needs to be remembered
that not all PV systems need to be installed by a professional, especially the widespread pico-scale solar
appliances and solar home systems (SHSs). Sixty-four percent of the imported equipment is installed by the large
companies themselves. These systems are larger SHSs or systems of greater than 1 kW, which require a
professional’s intervention. Approximately 22 percent of the imported volume goes into wholesale—and either
goes directly to wholesale or is put on stock first. The wholesale volume could be traded between solar
companies and shops and would eventually go into another company’s distribution or would be installed by
another company. Based on the interviews, it was not possible to assess the circulation of the wholesale volume
and the number of people involved in handling it before it reaches the final customer, either as distributed or
installed PV.
FUTURE DEVELOPMENT OF THE PV VALUE CHAIN
To identify a value chain today does not mean it will remain unchanged in the future. Value chains are a part of
an economy, and they may change over time just as every industry needs to react and potentially adapt to new
developments, both in and outside a specific market. For instance, decommissioning and how components are
recylced do not yet play an important role, but this will have to be addressed as the market matures and the
installed systems age.
An important question is that of local production: having no solar PV–specific manufacturing today does not
mean this will also have to be the case in the future. Module assembly lines exist in many countries and could
also emerge in Yemen. Whether manufacturing will or will not be established depends on many factors, such as
the appearance of public or private investors who seek benefits from hardware “made in Yemen.�? Investments
in local production will have to yield profits sooner or later. An important question is how soon profitability shall
23
be or needs to be attained. If manufacturing is seen as a prerequisite and a necessary part of research and
development for new products that cannot be found on the world market and present a comparative advantage
to existing solutions, investors would rather have a long-term expectation of profits with regard to making the
initial investment.
2.4 MARKET SIZING
There is no PV employment without a market for PV. This is intuitively true, but one also needs to define the
market under review, its segments, and its drivers. The PV market size, measured in megawatts, is the first of
two factors used for forecasting employment in this study. This factor is potentially even more important for
employment growth than the second factor: the employment multiplier (FTEs/MW). Mathematically, FTEs could
be increased by increasing both factors. However, a growth in FTEs/MW could also convey a message of
inefficiency. A greater value for FTEs/MW means more people employed, but it does not answer the question
of whether all these additional FTEs are truly required to install a PV system. An installation realized by
inexperienced staff will take more labor and time than the same installation realized by an experienced team. In
other words, it is not necessarily desirable to employ additional people who are only needed due to a lack of
expertise. Market size is a more unequivocal factor. Nevertheless, it is influenced by a myriad of developments
and conditions, which makes predicting its future development challenging, to say the least. Taking this into
account, a market forecast was developed based on these four steps (Barnett 1988):
1. Market definition.
2. Market segmentation—the division of the market into main components and segments.
3. Market drivers—describing the drivers.
4. Sensitivity analysis.
MARKET DEFINITION
The focus of analysis is the Yemeni market for decentralized PV systems, which are systems that directly convert
solar irradiation into electricity at, or very near, the site of electricity consumption. Since there is basically no
grid-connected PV at the moment, this criterion would fully describe Yemen’s PV market. This may change in
the future when, depending on the development of the electricity market and a postwar energy mix, grid
connection of PV becomes an option.
Returning to the first sentence of this section, it is true that there is no PV employment if there is no PV market
at all. But even if no new capacity were added in a specific year, there would still be a certain level of
employment, which does not depend on annual additions to the PV power park sector (large-scale grid-
connected PV power stations) but on the cumulative, previously installed base: existing systems would still need
to be operated and maintained, which holds the potential for creating or keeping jobs. Several studies define
jobs that are kept regardless of growth as “sticky�? or at least “stickier�? than jobs that depend on annual demand
(such as described in Bridge to India 2014).
24
Jobs are distributed across the value chain links as follows:
• Upstream activities based on annual installations:
o Manufacturing of key hardware components, such as the module, the inverter, or the
battery
o Import of key hardware components needed to build a PV system
• Downstream activities based on annual installations.
o Wholesale business: Business to business (B2B) trading of solar components.
o Distribution: Sales to the final customer
o Installation: Construction of PV systems
• Downstream activities based on cumulated installations;
o O&M-related activities
• Downstream activities based on installations to be dismantled:
o Decommissioning-related activities
Figure 2.9 Measure Job Creation by Value Chain Element
Source: Original compilation.
The market forecast calculated with the market sizing tool includes both annual and cumulative market sizes.
MARKET SEGMENTATION
The market sizing tool allows the entry of different input parameters for the northern and southern regions.
Such parameters reflected the availability of recent data about residential electricity consumption in both
regions (please refer back to section 1.2).
Another distinction is to be made between residential and other sectors, most notably, the commercial and
industrial (C&I) segment. Detailed recent data about electricity consumption were not available in this case.
Analysis was based on data from the International Energy Agency (IEA). It was possible to forecast market sizes
for the C&I segment by triangulating with residential data from the household survey (IEA 2020: at the time of
writing this report, C&I electricity consumption, including public services, corresponded to about 15 percent of
residential electricity consumption. However, taking the long-term average, this ratio is about 29 percent. Since
the model will look into a stabilized and more rational future, it assumed C&I consumption in both the northern
and southern governorates would account for 20 percent of residential consumption.
25
MARKET DRIVERS
As noted earlier, Yemen’s PV market growth in recent years was fueled by a lack of alternatives. In a postconflict
situation, this will hopefully change as new electricity sources will be created, and damaged energy
infrastructure will become available again. Still, there are market drivers that can be used to forecast future PV
market growth. Such drivers used in the market sizing model as input parameters are:
• The population today
• Future population growth
• Installed cumulative PV market base at model start
• Market share of PV in Yemen’s future electricity mix
• Electricity consumption today
• Future electricity consumption
While some of these drivers can be assessed more reliably (for example, population and population growth),
other numbers had to be estimated based on literature reviews and expert opinions.
The results of the market sizing forecast, which are presented in the upcoming sections of this study, are based
on unaltered, key input data for all three scenarios, as listed in table 2.2.
Table 2.2 Key Assumptions (Market Forecasts)
Assumption Source Values for North Values for South
Date of model start Original estimate January 1, 2020
Population CSO and World 23,897,347 5,981,407
Bank
Population growth rate World Bank 2.32% p.a.
Residential annual electricity consumption per capita World Bank 267 681 kWh/capita/
at model start kWh/capita/p.a. p.a.
Targeted residential annual electricity consumption Original estimate 850 kWh/capita/ 1,700 kWh/capita/
per capita p.a. p.a.
C&I annual electricity consumption per capita at Original estimate 53 kWh/capita/ p.a. 136 kWh/capita/
model start p.a.
Targeted C&I annual electricity consumption per Original estimate 150 kWh/capita/ 400 kWh/capita/
capita p.a. p.a.
Market share of solar PV in the future electricity mix Original estimate 60% 20%
Installed PV, residential (at model start) Original estimate 650 MW 130 MW
Installed PV, C&I (at model start) Original estimate 98 MW 20 MW
Annual production of a “new�? 1-kWp system Global Solar Atlas 1,886 kWh/kW 1,843 kWh/kW
Annual degradation of PV systems Original estimate 0.7%
Source: Market sizing model developed for this study.
Note: C&I = commercial and industrial; kW = kilowatt; kWh = kilowatt-hour; MW = megawatt; PV = photovoltaic.
MARKET GROWTH SCENARIOS
The possibilities to sensitize market drivers are endless. For this study, it was decided to fix a target for per capita
electricity consumption that appears to be attainable in the next 5, 10, or 15 years; 10 years was selected as the
26
base-case scenario, 5 years as the best-case scenario, and 15 years as the worst-case scenario. Per capita
electricity consumption was selected as a criterion for the scenarios because it was considered an appropriate,
albeit simplified, parameter for Yemen’s future development: the longer the situation remains precarious, the
longer it will take to attain a specific level of electricity consumption. The targeted electricity consumption to be
reached after these periods was different for the northern and southern governorates, accounting for today’s
differences in both electricity consumption and PV market share.
For the targeted electricity consumption per capita, the latest available “Electric power consumption (kWh per
capita) - Middle East & Northern Africa (excluding high income)�? data, as reported in a World Bank database,
was taken as a reference point: in 2014, this value was 1,702 kWh per capita. Considering that eight years have
since passed, the target values for the northern region would correspond to attaining around 50 percent of this
value in 5, 10, or 15 years, whereas the south region would attain 100 percent of this electricity consumption
level in the same period. The values for PV’s market share in an overall electricity mix were based on the results
of the household survey, which showed that PV was used less as a primary source of electricity in the south than
in the north. Whether these shares of PV in the overall electricity mix will remain the same in a stabilized Yemen
will depend on a variety of factors, such as the availability of affordable alternatives, the price development of
different energy sources, and the barriers that will exist in the future. More information about the barriers to
PV market growth and the future competitiveness of PV relative to other energy sources can be found in annex 4.
As regards the cumulative installed PV, the volume was estimated to be approximately 1,000 MW by 2020: the
information in the literature and media is scarce and dates back to 2017 or before. The values mentioned the
most are approximately 500 MW of cumulative capacity. However, the solar electricity consumption from the
World Bank’s household survey, as well as the business surveys conducted for this report, suggest that this value
should be much higher now. The challenge of estimating the correct installed PV volume not only lies in the
absence of official statistics, but also in the structure of Yemen’s PV market, which comprises many small systems
(solar home systems and pico-scale solar applications, each of low capacity), which are harder to count than
larger, megawatt-scale systems that do not yet exist in Yemen.
2.5 EMPLOYMENT MULTIPLIERS
An important objective of the primary research (if not the most important) was to identify the FTEs/MW
multiplier that could be used to quantify and forecast PV-related employment. Questions regarding the volumes
handled and the people employed across the value chain elements described earlier were asked in both an
online questionnaire and the guided interviews. However, only the FTE/MW calculated from the guided
interviews was considered in the end, since the data from the online survey proved to be too inconsistent,
especially for the PV volumes handled. To eliminate data inconsistency or implausible data from the guided
interviews, a correction run was conducted, wherein some of the questions were discussed again with the
interviewees.
As a first step toward identifying multipliers, the number of solar PV employees was counted. The interviewed
companies employed 138 full-time and 127 part-time employees. The hours worked by a part-time employee as
compared with a full-time employee ranged from 13 to 63 percent. Applying these shares, approximately 193
people worked full time in solar PV. The distribution of FTEs per value chain element is outlined in figure 2.10.
27
Figure 2.10 Full-Time Equivalent Units per Activity
Source: Based on guided interviews conducted for this study.
An exact statement as to whether these employees are formally or informally employed could not be made. The
interviewees were not willing to disclose the conditions of employment, but it was implied during the interviews
that the number of employees disclosed was at the low end and would only include those declared for official
purposes. Therefore, and based on the fact that solar companies are formal companies, all FTEs were categorized
as formal employment.
DIRECT FORMAL LABOR, UPSTREAM AND DOWNSTREAM
To calculate direct FTEs/MW for the value chain links dependent on annual market volumes in the SEIM tool,
the FTEs/activity were combined with the market volume that the surveyed companies installed or sold directly
to a final customer. Decommissioning and manufacturing could not be quantified since the interviewees did not
assign any FTEs specifically to these value chain elements. Nevertheless, since manufacturing and
decommissioning involve some work, these individuals would implicitly be included in other value chain
elements, most notably, installation and O&M.
As mentioned above, the direct job multipliers for both upstream and downstream activities (excluding O&M)
were based on the market volume that the surveyed companies sold to the final customer (distributed volume,
accounting for approximately 14 percent of the imported PV volume; refer also to Figure 2.8) or installed
themselves (approximately 64 percent of the imported PV volume). The volume of PV that was reported to
belong to the wholesale market (approximately 22 percent of the imported PV volume) was not used for
calculating the job multipliers because wholesalers’ business to business (B2B) transactions do not directly and
immediately contribute to electricity generation: Wholesalers put components on stock and trade with other
businesses before this PV volume reaches the final customer, where it is installed and contributes to the
electricity supply of households and businesses. In fact, a certain share of imported PV will also not be sellable
(due to, for example, low quality or damage suffered during transportation) or it will be in transit, thus not
contributing to electricity production. Therefore, the wholesale volume was not accounted for while calculating
the direct job multipliers. It could, however, be argued that the majority of the wholesale volume would
eventually be installed in Yemen, given that international trade is rather unlikely in the present situation. If the
wholesale volume were to be included in the calculation of the job multipliers, the multipliers would decrease.
Thus, the job multipliers and the resulting employment numbers presented in this study could rather be
interpreted as high values, especially when compared with other employment studies that do not distinguish
28
between different types of PV market volumes. The job multipliers generated by applying PV volumes to FTEs
are outlined in figure 2.11.
Figure 2.11 Full-Time Equivalent Multipliers per Activity (Direct Formal Labor)
Upstream Results
Import FTE/MW 4.4
Manufacturing FTE/MW
Upstream-total FTE/MW 4.4
Downstream
Wholesales FTE/MW 4.5
Distribution FTE/MW 4.7
Installation FTE/MW 7.0
Decommissioning FTE/MW
Downstream- total FTE/MW 16.3
Source: Original compilation using SEIM tool.
It needs to be noted that the multiplier for installation was still modified in conjunction with the O&M
multiplier’s definition, as will be explained in the next section.
DIRECT FORMAL LABOR, OPERATION AND MAINTENANCE
The O&M multiplier could not be derived by applying the PV volume assigned to O&M activities, since these
numbers were not plausible: the interviewees stated, for example, that four people worked full time over a year
to maintain 15 kW of PV, which is not a meaningful result. Although sizes and dimensions of PV modules vary, it
can be taken as a rough estimate that one needs a square meter to install 150 W of PV (Solar Power for Ordinary
People 2013). Fifteen kilowatts would therefore be installed on a surface of about 100 square meters (m2).
Imagining that four people would need to work full time for an entire year to service such a small area of PV is
not plausible, even if modules were not laid out on an even surface and even if other components, such as the
battery and the inverter, would also need to be serviced. When investigating the apparent mismatch between
FTEs and volumes stated for O&M work, it was found that companies generally do not distinguish between
installation and O&M activities, which are often performed by the same technicians. As mentioned earlier, the
boundaries between value chain elements and the jobs done by people are fluid in the Yemeni solar PV market:
jobs are performed based on qualifications and what is currently required. Job classification based on
department and value chain elements requires a higher level of specialization than observed in Yemen today.
Another problem regarding the quantification of O&M labor is that cumulative capacity is to be used instead of
annual volumes for calculating O&M labor. It was easier for the interviewees to assess or recall the PV volume
installed over the past 12 months, which is the determinant for the FTE multipliers of the other value chain
elements. This resulted in a mini survey being conducted among a limited number of previous interviewees
(see annex 2).
The additional information gathered, combined with different estimations for the market shares of different PV
system sizes, allowed for the calculation of an O&M multiplier of about 1 FTE/MW. This value was used as an
O&M target value. As a result of the target value calculation for O&M, the FTEs were mathematically
redistributed between installation FTEs and O&M FTEs based on the observation that the same people were
doing O&M and installation work. This led to an increase in the installation multiplier by 1.3. The final
29
employment multipliers for direct formal jobs are given in figure 2.12.
Figure 2.12 Direct Formal Job Multipliers
Driven by annual installation
Upstream Results
Import FTE/MW 4.4
Manufacturing FTE/MW 0.0
Upstream-total FTE/MW 4.4
Downstream
Wholesales FTE/MW 4.5
Distribution FTE/MW 4.7
Installation FTE/MW 8.3
Decommissioning FTE/MW 0.0
Downstream- total FTE/MW 17.5
Driven by cumulated installation
O&M total FTE/MW (cum.) 1.0
22.8 FTE/MW
Source: Original compilation using SEIM tool.
Note: The figures are rounded off.
New annual installations of 1 MW would lead to the creation of 21.8 formal direct full-time jobs. In addition, 1
formal direct full-time job would be created for every 1 MW of cumulative capacity.
INFORMAL DIRECT LABOR
As indicated earlier, the number of employees mentioned by interviewees most likely did not include those
without a formal labor contract, implying informal labor. The interviewees were not ready to disclose anything
that could eventually lead public authorities to investigate or collect additional taxes (even as the administration
is in dire need of tax revenue, as the war has severely strained public budgets).
Based on detailed (prewar) information about informal labor from ILO’s Labor Force Survey (ILO 2014b), a
percentage value for informal labor was calculated and added to the formal direct job multiplier. This factor was
based on a comparison of the ILO’s data on education and job profiles to the education level and job profiles of
the employees of the surveyed companies. This comparison resulted in a value of 40 percent, which was added
to the direct formal job multipliers for informal labor. This value does not appear to be very high compared with
the ILO’s finding that 94 percent of all private company employees in Yemen were employed informally.
However, this statement refers to small companies with fewer than five staff members, whereas the companies
interviewed for this report were larger. In addition, the detailed ILO statistics show that informal labor is more
widespread among those with a lower education level and skill level. The solar companies’ employees had higher
education levels (see also Table 2.1 Educational Attainment Levels in Yemen Overall and in the Solar PV Industry
(Large Companies)) than average, and most were also highly skilled, as shown in figure 2.13.
30
Figure 2.13 Skills of Interviewed Companies’ Employees
Source: Guided interviews.
Note: ILO terminology used; “professionals�? are mostly engineers.
The direct job multipliers, including both formal and informal labor, are, therefore, as listed in figure 2.14.
Figure 2.14 Direct Formal and Informal Job Multipliers
Driven by annual installation - formal & informal labor
Results Yemen
Upstream
Import FTE/MW 6.1
Manufacturing FTE/MW -
Upstream-total FTE/MW 6.1
Downstream
Wholesales FTE/MW 6.3
Distribution FTE/MW 6.6
Installation FTE/MW 11.6
Decommissioning FTE/MW -
Downstream- total FTE/MW 24.5
Driven by cumulated installation - formal & informal labor
O&M total FTE/MW (cum.) 1.4
Total FTE / MW 32.0 FTE/MW
Source: Original compilation using SEIM tool.
Note: Figures rounded off.
New annual installations of 1 MW would create 30.6 formal and informal direct full-time jobs. In addition, 1.4
formal and informal direct full-time jobs would be created per 1 MW of cumulative capacity.
Although the value added from comparing multipliers from different studies on different countries or regions
and across time can be questioned, other studies were consulted to see whether the values fell within a
reasonable range: for example, van der Zwaan, Cameron, and Kober (2013) identified job multipliers not based
on business surveys but on a review of studies of job multipliers in different RE industries, including, among
others, solar PV. The level of detail was limited to manufacturing, installation, and O&M. Figure 2.15 lists results
31
for the direct PV job multipliers next to the multipliers identified for Yemen.
Figure 2.15 Multipliers: Benchmark and Comparator Study
Driven by annual installation - formal & informal labor
Results Yemen Study-Min Study-Median Study - Max
Upstream
Import FTE/MW 6.1
Manufacturing FTE/MW -
Upstream-total FTE/MW 6.1 3.2 12.6 19.4
Downstream
Wholesales FTE/MW 6.3
Distribution FTE/MW 6.6
Installation FTE/MW 11.6
Decommissioning FTE/MW -
Downstream- total FTE/MW 24.5 3.9 15.4 23.6
Driven by cumulated installation - formal & informal labor
O&M total FTE/MW (cum.) 1.4 0.1 0.3 0.7
Total FTE / MW 32.0 FTE/MW 7.2 FTE/MW 28.3 FTE/MW 43.7 FTE/MW
Sources: Original compilation using SEIM tool; van der Zwaan, Cameron, and Kober (2013).
The consolidated multiplier for upstream activities (which are limited to “Manufacturing�? in the benchmark
study) is between the study’s minimum and median values, which would make sense given that PV modules or
inverters are not being manufactured in Yemen. One can see that the consolidated multipliers for downstream
activities and O&M for Yemen are in the range of the maximum multiplier values for the Middle East. It is
interesting that the employment factor calculated for downstream activities in Yemen (24.5) is even higher than
the maximum value reported by van der Zwaan, Cameron, and Kober (2013). Given that their study is from 2013
(with research being conducted in 2012), it could be assumed that with improvement in technology, the
requirement for labor for each activity decreased over the years. However, there are other possible explanations
for the relatively high multipliers. First, as explained above, the employment multipliers used in this study were
based on the market volume of PV only. The volume of PV that went first into wholesale before being sold to a
final customer or installed by a PV company was not considered. If the wholesale volume were factored in the
calculation of the direct multipliers, the multipliers used in this study would decrease. Second, the employment
multipliers in the van der Zwaan, Cameron, and Kober study were derived from a review of other studies and
are not based on surveys, unlike this study’s multipliers.
As regards the large range of employment factors, van der Zwaan, Cameron, and Kober explain:
The range of direct employment factors listed […] results from a large number of factors,
including: (i) differences in methodology used for calculating employment impacts, (ii) varying
coverage of what types of labour are accounted for under the notion of direct job, (iii) variable
country context in terms of the degree of local job content, (iv) diverging country context in
terms of average overall employment intensity, (v) different assumptions with regards to the
types of technology deployed, and (vi) especially for those studies in which no methodological
description is provided—differences regarding representing renewables favourably or
unfavourably in terms of job creation potential given possible vested commercial or ideological
interests of the institutions behind the publications.
32
Moreover, their study encompasses a number of countries in the Middle East4 and hence very different PV
markets. Yemen, especially wartime Yemen, will most likely be on the higher end of the employment factors’
range, given its many small PV installations, which are generally more labor intensive than the large-scale
systems dominant in other countries covered in their study. Finally, it can also be assumed that war affects labor
intensity and would thus lead to higher job multipliers: if we assume that, for instance, traveling time is counted
as working time and that destroyed infrastructure impairs traveling to sites where PV shall be sold or installed,
then it makes sense that system installation or sale requires more time in war-torn countries like Yemen than in
countries with intact infrastructure.
Although the primary research focused on job multipliers, some information about job profiles and skills was
also obtained. The online survey also asked about the qualifications needed in the Yemeni PV market and those
that would be easily available. Survey results, outlined in figures 2.16 and 2.17, may be considered when
organizing PV-related education programs and training courses.
Figure 2.16 Job Profiles and Skills: “Needed and Scarce,�? Top Answers
Source: Online survey conducted for this study.
As can be seen, mostly engineering and higher management skills were seen as being “needed and scarce�? in
the Yemeni solar PV market: 55 percent of the online survey’s respondents reported engineers for project
planning and design as scarce, and 48 percent reported engineers for quality assurance as scarce.
At the other end of the scale are the skills and jobs classified as “easily available,�? which are those requiring less
formal education.
4Which the authors have “defined as the region including Bahrain, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Oman, Palestinian
Territories, Qatar, Saudi Arabia, Syria, Turkey, United Arab Emirates and Yemen�? (van der Zwaan, Cameron, and Kober 2013).
33
Figure 2.17 “Easily Available�? Job Profiles and Skills: Top Answers from Survey
Source: Online survey conducted for this study.
INDIRECT LABOR
Next, the socioeconomic impact model (SEIM) tool was used to estimate indirect labor, meaning those that
supply products or services to the solar PV value chain. An input/output (I/O) model would be the standard
methodology suggested by literature for estimating indirect jobs. However, such a model is not available for
Yemen. Another method would be to survey the suppliers of products and services to the solar PV industry,
although this was beyond the scope of this study. Nevertheless, to provide a starting point for such an
assessment, the surveyed solar companies were questioned about the types of services and products they
purchased externally. Regarding “physical�? products, building materials play a role just as much as equipment
used for solar-powered pumping systems for water pipes, wells, and reservoirs (figure 2.18).
34
Figure 2.18 Indirect Labor: Externally Purchased Products
Source: Survey conducted for this study.
As regards services, a list of services and service providers was given and the following question was asked:
“What types of non-PV-specific services do you buy externally for the PV business of your company today?�?
Answers are listed in figure 2.19.
Figure 2.19 Indirect Labor: Externally Purchased Services
Source: Survey conducted for this study.
Accounting for the fact that products and services are supplied to the solar industry, indirect jobs were also
calculated by adding 75 percent to direct jobs as an average of typical values for indirect jobs. According to a
study from 2015, “the inclusion of indirect jobs would typically increase job numbers by 50 percent – 100
percent�? (Rutovitz, Dominish, and Downes 2015).
35
INDUCED LABOR
The further one moves from the industrial value chain, the less specific the assessment of labor becomes.
Induced labor would also typically be based on an I/O table. Since this was not available, induced labor
calculations were also based on a typical value found in the literature. As noted by one study, “the inclusion of
both indirect and induced jobs could increase job numbers by 100 – 350 percent�? (Rutovitz, Dominish, and
Downes 2015). In the SEIM tool, 100 percent was added to the calculated direct and indirect jobs. Induced labor
depends foremost on the income spent by those working directly in solar PV or by those supplying products and
services to the solar PV companies. The salaries earned by indirect employees could not be assessed and also
the money paid to the informally employed workforce is unknown. Moreover, interviewees were mostly
unwilling to disclose information about salaries, and so the information summarized in table 2.3 is based on a
very limited number of responses.
Table 2.3 Monthly Salaries Earned in the PV Industry
Job Profile Permanent Employees (YER) $ % of Respondents
Owner 350,000 583.33 6%
Chief executive 300,000 500.00 7%
Engineers 200,000 333.33 38%
Trainers – –
Technicians 100,000 166.67 36%
Office clerks – –
Support staff – –
Total 950,000 1,583.33 86.7%
Average income 153,017 255.03
Source: Original compilation.
According to the information provided, the average monthly income, weighted by the percentage of
respondents, is YRI 153,017 (approximately $255), depending on qualifications and position within the firm. If
this income were spent and reinjected into the labor market, induced labor would be created. The number of
induced FTEs would then depend on the quality of jobs being created, measured by their pay level. For example,
if an owner of a solar company with an income of YRI 350,000 saved YRI 50,000 and spent the remainder on
non-PV-related products, that owner could finance three induced FTEs with a monthly income of YRI 100,000 or
1.5 induced FTEs with a monthly income of YRI 200,000. The income earned by the induced employee would
then further trickle down as the income earned is spent again on other products and services.
PRODUCTIVE LABOR
Productive labor (or productive use jobs) is potentially the most difficult to assess and no standard methodology
has yet been defined. Household surveys could shed some light on this topic by assessing how the availability of
electricity generated by PV is related to household income. A survey that would only assess those households or
companies with relatively more PV capacity at their disposal and that also have a higher income would not be
sufficient because correlation does not explain causality: those with more PV capacity at their disposal could
very well be those who already had a higher income, allowing them to buy PV in the first place. Hence, a time
36
series could be conducted to see how household income develops after the installation of a PV system. In
general, it could also be discussed whether the same productive labor would also be created independent of PV,
with another source of energy generation, for example, diesel or fuel gensets (as productive labor is not specific
to the technology itself but to the kWh generated). In the case of Yemen today there are often no or not enough
alternatives to generate electricity, but this will hopefully change in the future. Competition stimulates business
and Yemen, already under-electrified before the war, needs all the electricity it can generate. The decision as to
which technology will prevail will be based on a number of considerations, which are different and individual for
every person, company, or institution that wants to install PV. Ecological and health-related considerations may
hopefully play a role but mostly, investment decisions will be based on the economic performance of one
technology compared to another. Annex 4.2 of this report shows, from an economic perspective, how PV
performs compared to other decentralized generation technologies in Yemen.
Very little evidence could be gathered with regards to productive use jobs in Yemen except for the proof that
they exist. For example, the local consultant contracted to conduct research for this study was able to take on
the assignment only because he had PV that allowed him to power the PC and phone used to communicate and
discuss the onsite information included in this report. In the SEIM tool, productive labor was included and based
on the proportions between direct jobs (including both formal and informal jobs) and productive use jobs found
in the “Powering Job Census 2019�? (Power for All 2019). The census assessed direct jobs related to decentralized
solar PV in three countries (India, Kenya, and Nigeria) and estimated productive use labor based on a literature
review. According to its findings, total productive use jobs accounted for between 115 and 260 percent of direct
jobs (table 2.4). The SEIM tool used the average value for all three countries—160 percent—which means that
every direct employee created 1.6 productive use jobs.
Table 2.4 Proportion of Total Direct Jobs to Total Productive Use Jobs
India Kenya Nigeria Total
Direct formal jobs 95,000 10,000 4,000 109,000
Direct informal jobs 210,000 15,000 9,000 234,000
Total direct (formal and 305,000 25,000 13,000 343,000
informal) jobs
Productive use jobs 470,000 65,000 15,000 550,000
Total productive use / 154 260 115 160
total direct jobs (%)
Source: Based on Power for All (2019).
2.6 MARKET AND EMPLOYMENT ESTIMATES
MARKET SIZES
Based on the assumptions and input parameters of the base-case scenario (see table 2.2)—in which targeted
electricity consumption would be achieved within 10 years of 2020—annual PV installations are outlined in
figure 2.20.
37
Figure 2.20 Forecasted Annual PV Installations in Base-Case Scenario (S2), by Region and in Total
Source: Original compilation.
As can be seen, the PV market is largely defined by the northern region. This is not surprising given that the
majority of Yemen’s population lives in the northern governorates, where there is less access to energy
alternatives than in the southern region. Annual PV installations in the south are forecasted to start with 13 MW
in 2020 to reach approximately 22 MW in 2030. In the north, approximately 67 MW are installed in 2020 and up
to 103 MW by 2030. Annual installations for the entire Yemeni market (around 80 MW in 2020 and 125 MW by
2030) were used to estimate PV-related employment (measured as FTEs) across the entire PV value chain, except
for jobs related to O&M, which were based on cumulative installed PV capacity (as presented in figure 2.21).
Figure 2.21 Forecasted Cumulative PV Capacity in Base-Case Scenario (S2), by Region and in Total
Source: Original compilation.
38
Overall installed capacity, at about 1,000 MW in 2020, could reach 1,900 MW by 2030 in the base-case scenario,
with the majority installed in the north.
However, depending on the development of the country and future barriers to PV, targeted electricity
consumption could be attained earlier or later. For this study, three scenarios were considered, in which the
time required to reach the targeted consumption level is the only parameter altered:
• Scenario 1 (S1) assumes that the target is achieved after five years (refer to table 2.2). This is the most
optimistic or the best-case scenario. Once the target value is reached, the market continues to grow at
the same rate.
• Scenario 2 (S2) assumes that the targeted electricity consumption is achieved after 10 years. This is the
base-case scenario. The numbers shown in figures 2.20 and 2.21 correspond to this scenario.
• Scenario 3 (S3) assumes that the targeted electricity consumption is achieved after 15 years. This is the
worst-case scenario.
Figure 2.22 shows the forecasted sizes of the annual PV installations for the three scenarios.
Figure 2.22 Forecasted Annual PV Markets in Three Scenarios: S1 (Best), S2 (Base), and S3 (Worst)
Source: Market sizing model compiled for this study.
The S2 (base case) line presents the total annual PV market for Yemen as shown in figure 2.20, which reaches
125 MW of annual installations in 2030. The worst case, however, forecasts an annual market of only about 68
MW in 2030. The best case shows a steep development and an annual PV market of 400 MW by 2030. The
cumulative PV capacities of the three scenarios are shown in figure 2.23.
39
Figure 2.23 Forecasted Cumulative PV Capacity under S1 (Best), S2 (Base), and S3 (Worst)
Source: Market sizing model compiled for this study.
The S2 base case line presents the total cumulative PV capacity for Yemen, as shown in figure 2.21, which reaches
1,900 MW of cumulative capacity in 2030. The worst case forecasts a cumulative capacity of only about 1,500
MW in 2030 while the best case scenario, with accelerated growth, would allow PV to reach 3,700 MW by 2030.
These main assumptions, as well as the results of the market forecast for the year 2030, are shown in figure
2.24.
Figure 2.24 Details of PV Market Forecasts, by Scenario
5 years until target (S1) 10 years until target (S2) 15 years until target (S3)
Residential Segment Residential Segment Residential Segment
Annual installations MWp 333 Annual installations MWp 107 Annual installations MWp 58
Cum. PV MWp 3,205 Cum. PV MWp 1,665 Cum. PV MWp 1,308
C&I + other Segment C&I + other Segment C&I + other Segment
Annual installations MWp 70 Annual installations MWp 18 Annual installations MWp 10
Cum. PV MWp 522 Cum. PV MWp 250 Cum. PV MWp 196
Total PV Total PV Total PV
Annual installations MWp 403 Annual installations MWp 125 Annual installations MWp 67
Cum. PV MWp 3,727 Cum. PV MWp 1,915 Cum. PV MWp 1,504
Source: Original compilation.
EMPLOYMENT
As seen in section 2.5. the multipliers for different job categories () were combined with the different market
scenarios. Applying a learning curve for all activities would lower the FTE/MW by 0.5 percent per year. This value
is rather low compared to the 2.7 percent per year indicated in a study of the Middle East (Rutovitz, Dominish,
and Downes 2015). However, considering Yemen’s lower labor efficiency, measured in terms of gross domestic
product (GDP) per capita, this learning rate value seems appropriate.
With regards to the base case that foresees a PV market growth from 78 MW in 2020 to 125 MW in 2030, direct
40
jobs would develop along the value chain.
Figure 2.25 Direct Labor, Upstream and Downstream, in Base-case Scenario
Source: Original compilation using SEIM tool.
The model starts in 2020 with a total of nearly 3,800 FTEs, of which the large majority—86 percent—are in the
downstream market. This overall total is also the best estimate for the number of employees working in Yemen’s
PV industry. Based on cumulative installations, the “sticky�? O&M jobs (see section 2.4 for a brief discussion of
the term) account for the remaining 36 percent, although their multiplier is only 1.4 FTE/MW. Until 2030, the
overall job market could grow to 6,236 direct FTEs, which is an increase of 65 percent within 10 years. Figure
2.26 shows how direct FTEs are distributed across the value chain in the period between 2020 and 2030.
41
Figure 2.26 Direct Labor Value Chain: Base-case Scenario
Source: Original compilation using SEIM tool.
O&M is the single most important element of the value chain. Beyond 2030, decommissioning could start to
grow in importance as the first systems installed around 2011 would approach the end of their useful life.
Manufacturing core solar components could begin before that date, provided that Yemen has regained enough
stability to attract investors who see a market for locally manufactured products. This interest could be
stimulated by politically imposed local content rules (refer to section 2.3).
Based on the factors mentioned above, figure 2.27 lists the number of jobs that could be created, including
indirect and induced labor.
42
Figure 2.27 Direct, Indirect, and Induced Labor: Base-case Scenario
Source: Original compilation using SEIM tool.
By 2030, the PV market could create indirect jobs of the magnitude of 4,677 FTEs and induced labor of 10,912
FTEs, adding up to a total of 21,825 FTEs. If productive jobs are also included, the total number of jobs created
would be substantially higher. According to the assumptions made in the model, productive use jobs can provide
labor to another 9,977 FTEs.
Figure 2.28 shows that adding up direct (both formal and informal), indirect, induced, and productive labor, a
maximum total of 31,801 jobs could be created by solar PV by 2030.
Figure 2.28 Total Labor: Base-case Scenario
Source: Original compilation using SEIM tool.
43
However, FTE growth can be substantially higher or lower, depending on the development of the market, which
in turn depends on the overall development of the country. In the best-case scenario, Yemen can reach the
electricity consumption levels of the market sizing model after five years, as shown in figure 2.29. In this case,
PV could provide labor up to 58,802 FTEs without productive labor or 85,682 FTEs if productive labor were
included.
Figure 2.29 Total Labor: Best-case Scenario
Source: Original compilation using SEIM tool.
If, on the other hand, development were slower and the target levels of electricity consumption were to be
reached only after 15 years, employment would also be lower (figure 2.30).
Figure 2.30 Total Labor: Worst-case Scenario
Source: Original compilation using SEIM tool.
44
The results and key assumptions of all three scenarios—the best case (S1), the base case (S2) and the worst case
(S3)—for the year 2030 are summarized in table 2.5.
Table 2.5 Overview of Best-case, Base-case, and Worst-case Results, 2030
5 year scenario (S1) - best-case 10 year scenario (S2) -base case 15 year scenario (S3) - worst-case
PV market in 2030 PV market in 2030 PV market in 2030
Annual installations MWp 403 Annual installations MWp 125 Annual installations MWp 67
Cum. PV MWp 3,727 Cum. PV MWp 1,915 Cum. PV MWp 1,504
Direct labor in 2030 Direct labor in 2030 Direct labor in 2030
Upstream Upstream Upstream
Import FTE 2,350 Import FTE 729 Import FTE 393
Manufacturing FTE 0 Manufacturing FTE 0 Manufacturing FTE 0
Upstream total FTE 2,350 Upstream total FTE 729 Upstream total FTE 393
Downstream Downstream Downstream
Wholesales FTE 2,444 Wholesales FTE 758 Wholesales FTE 409
Distribution FTE 2,525 Distribution FTE 783 Distribution FTE 423
Installation FTE 4,454 Installation FTE 1,381 Installation FTE 746
Decommissioning FTE 0 Decommissioning FTE 0 Decommissioning FTE
Downstream total FTE 9,423 Downstream total FTE 2,922 Downstream total FTE 1,578
Driven by cum. PV Driven by cum. PV Driven by cum. PV
Operations and Maintenance
FTE 5,027 Operations and Maintenance FTE 2,585 Operations and Maintenance FTE 2,030
O&M total FTE 5,027 O&M total FTE 2,585 O&M total FTE 2,030
Total direct labor FTE 16,800 Total direct labor FTE 6,236 Total direct labor FTE 4,002
Indirect and induced labor in 2030 Indirect and induced labor in 2030 Indirect and induced labor in 2030
Upstream Upstream Upstream
Import FTE 5,874.1 Import FTE 1,821.5 Import FTE 983.7
Manufacturing FTE 0.0 Manufacturing FTE 0.0 Manufacturing FTE 0.0
Upstream total FTE 5,874.1 Upstream total FTE 1,821.5 Upstream total FTE 983.7
Downstream Downstream Downstream
Wholesales FTE 6,111.1 Wholesales FTE 1,895.0 Wholesales FTE 1,023.4
Distribution FTE 6,312.6 Distribution FTE 1,957.5 Distribution FTE 1,057.2
Installation FTE 11,134.7 Installation FTE 3,452.8 Installation FTE 1,864.7
Decommissioning FTE 0.0 Decommissioning FTE 0.0 Decommissioning FTE 0.0
Downstream total FTE 23,558.4 Downstream total FTE 7,305.4 Downstream total FTE 3,945.3
Driven by cum. PV Driven by cum. PV Driven by cum. PV
Operations and Maintenance
FTE 12,568.7 Operations and Maintenance FTE 6,462.0 Operations and Maintenance FTE 5,074.9
O&M total FTE 12,568.7 O&M total FTE 6,462.0 O&M total FTE 5,074.9
Total indirect+induced FTE 42,001 Total indirect+induced FTE 15,589 Total indirect+induced FTE 10,004
Productive use jobs FTE 26,881 Productive use jobs FTE 9,977 Productive use jobs FTE 6,402
Total FTE w/o prod. use FTE 58,802 Total FTE w/o prod. use FTE 21,825 Total FTE w/o prod. use FTE 14,005
Total FTE incl. prod. use FTE 85,682 Total FTE incl. prod. use FTE 31,801 Total FTE incl. prod. use FTE 20,408
Source: Original compilation using SEIM tool.
To conclude, the Yemeni PV market has the potential to create up to 10,000 jobs, depending on the job
categories selected and the speed of market development. Even in the worst-case scenario with the slowest
market development, more than 14,000 people could generate income, thanks to solar PV. If productive labor
is included, the total goes up to more than 20,000 people. However, embedded in the overall country context,
the development of jobs created by PV will also depend on Yemen’s investment climate in the future, which is
defined by costs, risks, and market barriers. A detailed outlook of the potential future development of the PV
sector in postwar Yemen is presented in annex 4.
45
3. RECOMMENDATIONS AND OUTLOOK
This study has shown that solar photovoltaics (PV) has the potential to create tens of thousands of jobs in Yemen
along a solar PV value chain. These jobs are mostly created downstream because the manufacturing of PV-
specific hardware is currently very limited and nearly all major components are imported. Today’s value chain
could develop beyond its current state but we do not know how the future value chain will look as it depends
on a myriad interlinked factors. These factors will determine the investment climate for PV. The big question is:
what is required to create the jobs calculated and described in this report, regardless of specific business cases?
The answer is plain and simple: the war must end so the country can stabilize. Once a higher degree of
stabilization is reached, then educated decisions about Yemen’s energy future can be taken to create an
investment climate that can provide both employment and energy.
SOLAR PV: A GOOD FIT FOR YEMEN
Decentralized energy generation, such as solar PV, seems to be a good if not the natural choice for re-electrifying
sun-rich Yemen, even after considering the difficulties of adding centralized capacities that were already
insufficient before the current turmoil began. Although it is difficult to predict Yemen’s future in general and its
energy sector in particular, solar PV already plays an important role in Yemen’s current energy mix. When the
centralized power supply broke down and fuel supplies became scarce, a private PV market prospered. It can be
argued that it was out of necessity that PV has arrived in Yemen. There are good reasons to not only keep it but
also promote its development for a better future.
Yemen is endowed with three significant advantages when it comes to solar PV. First, Yemen is among the
countries with the highest solar PV potential in the world, overshadowed only by Namibia, Chile, Jordan, and
Egypt (World Bank 2020f). The second advantage is that most Yemenis are familiar with and feel positive about
PV, which means that grassroots campaigns to promote the technology are not required. Finally, the
development of PV in Yemen has been created by the private sector, which can be leveraged even more if further
support from public authorities were provided.
QUALITY ASSURANCE IS NEEDED
To capitalize on these advantages, several interlinked issues and challenges will have to be addressed in parallel
to make solar PV sustainable. The management of these issues and challenges will define the future of PV,
especially in a postwar environment. One pressing issue is the quality of PV products and systems. Low-quality
PV products not only have shorter lifespans and produce less energy but can also harm the reputation of PV as
a reliable energy source. Low-quality products may lead people to abandon PV as soon as power supply
alternatives become increasingly available again. Alternatives are needed but misrepresenting this technology
needs to be avoided. Quality installations require both quality equipment, identifiable by certifications based on
measurements and tests, as well as installers who are properly trained, educated, and ideally certified. To ensure
hardware quality and good knowledge about proper installation and maintenance, enforced regulations are
needed and technical test facilities should be established. Yemeni-specific norms and standards for PV products
should be defined and enforced while training centers should be founded or integrated into existing facilities,
such as universities or technical schools. Quality assurance and training can also create additional jobs.
46
FINANCING SOLUTIONS CAN IMPROVE BOTH ABILITY AND WILLINGNESS TO PAY
In parallel, the questions of PV investment costs and financing will have to be addressed to increase the number
of those who are able to pay for PV and convince those who are able but not yet willing to pay. With regards to
the purchase price and the associated costs, two main aspects need to be addressed: (i) how financing can be
accessed; and (ii) how the costs of PV can be further reduced. With regards to financing, households or
businesses either need to have enough financial resources on their own (equity based on income or wealth), or
they need to have access to lending institutions for the purchase of PV systems. Both equity and debt financing
are already available to some extent, otherwise, solar PV would not have been bought in the past.
Still, after years of war, poverty is widespread and growing, and there is neither enough capital nor sufficient
access to capital. A UN report states that “war has driven poverty in Yemen from 47 percent of the population
to a projected 75 percent by the end of 2019�? (UNDP 2019). Some financing is provided by international donors
but this will be insufficient to reestablish or expand electricity access in the country. More financing sources are
needed and will have to be generated in the long run by Yemenis, who can only do so efficiently if stability is
regained. An important, yet not clearly identifiable, financing source for PV investments could also come from
the remittances of Yemenis living and working abroad. If parts of the income earned by the Yemeni diaspora
could be channeled into solar PV, its investment burden could be lowered with positive impacts on both ability
and willingness to pay. It is still unclear how these payments could be tapped efficiently. Potentially, Yemenis
abroad could be addressed directly and a financial bonus could be offered for every dollar sent to Yemen that is
spent on solar PV.
THE PRICES OF ALTERNATIVE SUPPLY OPTIONS PLAY A ROLE
Another question to be addressed is how the profitability of PV can be further improved, which will be key for
people’s willingness to pay in a competitive energy market. This can be done by decreasing costs and increasing
revenues. Revenues will depend on the price levels of alternative energy supply, most notably fuel and diesel.
Cost decreases in solar PV will positively impact the payback time for the installation and, consequently, the
ability and willingness to pay for PV. With regards to hardware costs, Yemen will most likely continue to depend
on imports of key solar components, even if local manufacturing were initiated. The prices of these components
are defined mostly by the supply and demand of the world market and technological progress. With solar PV
scaling up at a global level, prices are likely to decrease. In turn, this will make solar PV systems more affordable,
provided that the purchasing power of Yemenis does not decrease faster than PV’s world market prices , and
that the customs and taxes levied on imported equipment do not distort these prices. In other words, besides
the levy of customs and taxes, Yemen has limited influence on hardware prices. However, the general global
development of the sector seems to work in favor of the affordability of solar PV.
DOWNSTREAM JOBS (AND MEGAWATT-HOURS) BEFORE UPSTREAM JOBS (AND MEGAWATTS)
To decrease dependence on world market prices, Yemen and Yemeni entrepreneurs, as well as investors, could
enter the manufacturing market to produce key solar components locally. This could create upstream jobs while
decreasing dependence on solar imports, but such a decision would have to be evaluated carefully.
Manufacturing locally may also reduce the number of people who work in imports. In addition, there is a risk
that locally manufactured, non-innovative components could be more expensive than imported goods because
new manufacturers still face the learning curve that global manufacturers have already undergone. If the price
difference cannot be compensated, local manufacturing risks increasing system prices, which reduces the ability
47
and willingness to pay for PV. If fewer people can buy a solar PV system, annual capacity would go down, and
downstream jobs in sales, installation, and operation and maintenance (O&M) risk being lost. The possibilities
for local manufacturing are more promising for a utility-scale market segment, which does not exist yet in
Yemen. This is because international investors (utilizing energy performance contracting) are more likely to
absorb the additional local manufacturing costs than buyers of small systems, provided that the overall business
case, considering the risk level of the investment, remains sufficiently attractive.
Ideally, local manufacturing would be coupled with innovation and learning effects that can spread across value
chains and different industries. Research and development projects in universities should be incorporated.
Copying existing products will have a limited effect on job creation, and possibly even negative impacts. The
manufacturing of PV components deserves more investigation and needs to be embedded in the broader
context of industrialization, responding to Yemen’s industrial strengths and comparative advantages. However,
from today’s perspective, if asked what is needed most urgently in Yemen—either a warehouse full of 100
megawatts of locally manufactured but unsold PV modules or the sale of a single imported solar light—the
answer would certainly be a single solar light. Unsold locally manufactured products do not power anything, and
what Yemen’s power sector urgently needs most is affordable energy.
For Yemen, today’s recommendation would be to create as much electricity as possible, which means installing
as much quality PV as possible, creating kilowatt-hour capacity that can provide jobs downstream. In an ideal
world, both scenarios could be coupled together and every light powered by locally manufactured, cost-
competitive hardware. However, this potential is still far from a reality in Yemen. Component manufacturing
could still make sense in the future if the primary objective is not to drive down costs and compete with
manufacturers from Asia, but to learn.
LARGER SYSTEMS AND REDUCED BATTERY STORAGE NEEDS CAN DRIVE DOWN COSTS
Another way to bring down the cost of solar PV systems, when dependent on international markets, is to
capitalize on the economies of scale, replacing or complementing many small systems with larger systems. When
doing so, the tradeoffs with regards to employment need to be considered by answering questions similar to
the ones discussed before: what is more important—direct labor per megawatt or access to electricity?
Another point to consider is the importance of storage for the Yemeni solar PV market as storage is another
factor driving solar PV systems prices up. A lot could be gained by reducing the need for battery storage.
However, this can be achieved only by using alternative sources of energy to deliver the required nighttime
electricity. How nighttime electricity supply is secured in the future will be different for every region, depending
on the available infrastructure after the war. PV could be complemented by a centralized electricity grid,
especially in the southern region. In the northern region and more remote areas, especially those that did not
have centralized electricity generation before the conflict, mini or micro grids can play a role. Encouraging
examples for such installations already exist (Mistiaen 2020).
All measures to establish a favorable environment for solar PV in postwar Yemen, whether related to quality
assurance, lower costs, or increased profitability, need to be accompanied by proper risk management. Proper
risk management will identify risks along the value chain and assign them to the market participants
(government, firms, households, etc.) best able to manage the risk impact and the probability of occurrence.
Ultimately, PV needs to be embedded in an overall energy and economic strategy that enables access to reliable
48
and affordable electricity for a growing population, increasing the use of clean energy, promoting local
industries, and creating as many jobs as possible, whether they are direct or indirect, induced or productive.
However, stability and security are the prerequisites for such a strategy to be developed and implemented.
49
ANNEX 1. EXPONENTIAL GROWTH OF PV MARKETS
When looking at historical PV market data for different countries and regions, exponential growth curves can
frequently be observed. Figure A1.1 depicts growth in the Middle East.
Figure A1.1 Historic PV Market Growth in the Middle East, 2000 –18
4,000
3,500
3,000
2,500
MW
2,000 Middle East
1,500
1,000
500
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Years
Source: Based on IRENA (2020).
However, as markets cannot grow forever, there are natural limits to growth such as the available space in the
energy sector, the electricity needed, and so forth. The exponential growth will slow and later stop when an
upper threshold for market growth is approached. An S-curve can be observed in mature PV markets like those
of Italy and Germany (figure A1.2).
Figure A1.2 Historic PV Market Growth in the Middle East, Italy, and Germany, 2000–18
50,000
45,000
40,000
35,000 Middle East Italy Germany
30,000
MW
25,000
20,000
15,000
10,000
5,000
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Years
Source: Based on IRENA (2020).
50
Considering the patterns across PV markets, a logistical growth function was used to project Yemen’s cumulative
PV market growth (see also PV Magazine 2019). The mathematical formula used is presented here:
������ ∗ ������
������ =
������ + (������ − ������) ∗ ������ −������������������
x = Cumulative market size in period t
a = The cumulative installed PV capacity at model start
S = Maximum value (upper threshold)
e = Mathematical constant e
k = Correction factor
t = Time (years)
Based on the calculated cumulative market figures, the annual PV market size can be determined as the
difference between years. Annual PV installations are expected to grow until market saturation is approached.
At this point, annual growth rates will begin to decline. By determining a maximum PV capacity in Yemen and
the speed at which we estimate the PV market will grow, the PV market sizes were forecast.
51
ANNEX 2. DESCRIPTION OF SAMPLED COMPANIES
A total of 27 answers were received from the online survey and 10 interviews were conducted by the local
consultant (figure A2.1). In the online survey, different people from the same company provided answers. In
addition, two companies that were interviewed also answered the online survey; in one instance the same
person who was interviewed had also answered the online survey.
Figure A2.1 Overview of Primary Research
37
35
32
27 26
24
10 10 10
Online & interviews Online Guided
Total answers received Number of respondents Number of companies
Source: Original compilation.
Although the vast majority of Yemeni PV companies are headquartered in Sana’a City, both the respondents to
the online survey and the interviewed companies are active in all Yemeni governorates (figure A2.2). With the
exception of the island of Socotra, all governorates were served by one or several solar companies.
Figure A2.2 Governorates Served by the Interviewed and Surveyed Companies
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
% of online survey % of guided interviews
52
Source: Original compilation.
It was found in both surveys that most PV companies are not only active in the solar PV business but also in other
industries. In addition, PV companies are usually active in several elements of the solar PV value chain. In the
online survey, participants claimed that, on average, around 51 percent of company turnover in the last 12
months was attributable to PV and the remainder was generated in other business segments. Equally, the ratio
of employees working in solar PV versus the overall employees of the company was stated to be around 50
percent. Although these averages match, every company has an individual profile with different business lines
among which PV may be more or less important. By and large, the number of employees working only in PV are
in line with the contribution of solar PV to the total turnover of the company but there is a variability in the data.
Figure A2.3 Solar PV Turnover and Employees Compared with Total Company Turnover and Employees
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
% solar turnover / total turnover % PV employees / total employees
Average share of solar turnover Average share of solar employees
Source: Original compilation.
Note: The x-axis plots companies/observations. Only observations with answers for both solar turnover and solar PV employees were
considered. The averages of the total answers in each category are shown.
The guided interviews provided additional information; however, the contribution of the solar PV sales to overall
company sales could not be assessed because most interviewees refused to provide financial numbers. The
share of solar PV employees to total employees was found to be 56 percent. On average, approximately 13
persons worked exclusively in the solar business, with a minimum of 9 solar employees in one company and a
maximum of 58 employees in another. With regards to non-PV-specific activities the companies were active in
different industries, such as automotives, cement and construction, computer sales, construction supporting oil-
field services, general trading, industrial control equipment, generators, and maintenance services.
With regards to the gender of employees working in solar, interviewees answered that about 95 percent were
male and 5 percent were female. A solar expert interviewed reported having been involved in training about
200 people in the installation of solar home systems in 2015–16, all of whom were men. According to the same
expert, training courses might be attended by women—but they would most likely have been trained separately.
It is not known what impact future PV market growth might have on female employment but the technology by
itself would most likely not be able to change fundamental societal values. When given the opportunity, women
53
have taken the lead in providing PV-based electricity to their communities. The UN financed three micro-grid
projects in Yemen that promoted the employment of women in running PV installations (Mistiaen 2020).
Investors could play a role in promoting women’s jobs in the industry.
54
ANNEX 3. OPERATIONS AND MAINTENANCE MINI SURVEY
A mini survey among solar companies was conducted to better understand the operation and maintenance
(O&M) business of solar companies and to identify a useable job multiplier for the socioeconomic impact model
(SEIM) tool. O&M for very small systems (up to approximately 100 watts) is exclusively done by the end users
without assistance from a company. For battery-related O&M work, several workshops offer reactivation
services that can extend battery lifetime. These shops buy old batteries for up to 10 percent of their original
prices. Some solar companies indicated that they resell the raw material extracted from old batteries by
exporting them, especially to China; this has become more difficult, however, amid a war-induced rise in
transport costs.
In the mini survey, instead of asking for the volumes of O&M, questions were asked about the hardware that
required most maintenance: batteries are at the top of this list, accounting for 50 percent of O&M time, followed
by inverters (20 percent), and charge controllers (15 percent). Solar companies repair customer batteries
electrically and chemically, but they do not repair damaged electrodes.
Other questions focused on how activities are split with system owners, and how much time is needed per
activity. The results can be found in table A3.1.
Table A3.1 O&M Activities, Work Split, and Frequency
Activity Typically Typically Number of Staff Frequency
Done by Solar Done by Hours (if done
Company Owner by company)
Solar system: cleaning of modules X According to the Once a month
number of
panels
Solar system: checking the array X Depending on Once a year
structure the size of the
system, 2–8
hours
Solar system: checking the cabling X 1–2 hours Four times a
(mechanical) year
Solar system: checking the cabling X 1–3 hours Four times a
(electrical) year
Solar system: checking the output X 2–10 hours Four times a
current and voltage year
Solar system repair work for any of X
the above
55
Activity Typically Typically Number of Staff Frequency
Done by Solar Done by Hours (if done
Company Owner by company)
Battery: checking connections and X Varies from Twice a year
case electrical connections system to
system
according to the
number of
batteries
Battery exchange X One time, only
during the
warranty period
Battery repair X 2–30 hours Once a year
Inverter: checks X 2 hours Twice a year
Inverter: repair X 2–18 hours Once a year
Battery charger checks X X 1 hour Once a year
Battery charger repairs X 2–12 hours Once a year
Switches/cables/checks X 2 hours Four times a
year
Control board of charger hybrid X X 2 hours Four times a
control checks year
Control board of charger hybrid X X 2–60 hours Once a year
control repairs
Source: O&M mini survey.
It can already be seen that work is split between a solar company and customer. In addition, it also becomes
clear that the range of time it takes to perform a specific O&M activity varies a lot and that the battery
maintenance depends on the type of system. Further questions were asked, to learn more about how system
types affect O&M time (table A3.2).
Table A3.2 O&M Activities: Work Time by System Size
Minimum Staff Maximum
System Average Staff Days
Days Staff Days
SHS system 150–300 W with battery 4 days a year (32 hours) 4 days a year 8 days a year
SHS system >300 W with battery 4 days a year 4 days a year 8 days a year
56
1 kW “standard�? PV system with battery 8 days a year 8 days a year 12 days a year
5 kW “standard�? PV system with battery 8 days a year 8 days a year 12 days a year
10 kW “standard�? PV system with battery 16 days a year 10 days a year 20 days a year
10 kW solar pumping system without a battery 8 days a year 4 days a year 12 days a year
20 kW system with battery 20 days a year 16 days a year 30 days a year
Source: O&M mini survey.
Note: One staff days is eight hours. kW = kilowatt; W = watt.
The battery plays an important role for solar systems in Yemen. According to the household survey, nearly 40
percent of PV capacity is used during the night (refer to chapter 1, Table). According to analysis conducted by
UNOPS (n.d.), scrap collection shops buy, sell, recycle, export, and trade (solar and other) batteries. This implies
that activity is an important part of solar O&M work is not done by solar companies, but by companies belonging
to another industrial value chain, related to the circular economy. Still, knowing that the boundaries between
industries are fluid and job segregation is not a rigid concept in Yemen, survey respondents were asked who
takes care of battery replacement. The answers differ by system size (figure A3.1).
Figure A3.1 Responsibility for Battery Replacement
Source: O&M mini survey.
Note: kW = kilowatt; PV = photovoltaics; SHS = solar home system.
Solar companies noted that they work in tandem with scrap collection/battery shops. Reported battery
replacement intervals are summarized in figure A3.2.
57
Figure A3.2 Battery Replacement Intervals
Source: O&M mini survey.
As replacement intervals for lead batteries should be around five years (Solar Reviews 2020), the most commonly
stated replacement interval of 1–2 years suggests that the batteries are either not fit for purpose, of low quality,
or not properly operated (for example, perhaps maximum depths of discharge are being disregarded).
58
ANNEX 4. POTENTIAL DEVELOPMENT OF SOLAR PV
The investment climate for a specific technology in a specific country encompasses a number of factors that
determine the willingness of market actors, both national and international, to invest in that technology
(“willingness to pay�?). Key assessment criteria for the investment climate include costs, risks, and barriers to
competition (World Bank 2005). These criteria are linked by several other factors, for example, regulations,
markets, institutional capacities, or access to finance.
Table A4.1 Key Barriers to Investment
Costs Risks Barriers to Competition
Regulation/government Regulatory burdens and Regulatory Barriers to market entry
administrative uncertainty and exit, incentives in
procedures, regulatory regulated sectors
fragmentation (utilities)
Market size, segments, Market fragmentation, Lack of standards Implementation of
and structure segments, value chains competition law and
policy
Public sector promoter Infrastructure, and public Weak planning and Possible unintended
constraints sector efficiency and project preparation consequences of public
capacity capacity procurement procedures
Liquidity/access to finance Cost of finance (debt and Financial instability, Limiting entry into new
equity) and liquidity unavailability of products and markets
instruments to
allocate risks
Source: Based on data from the European Investment Bank (2016).
When deciding whether to purchase PV, various market actors, both public and private, evaluate criteria in
different ways. In the case of PV, the perspective of not only corporations but also of private households needs
to be considered. While public agencies as well as private firms decide whether to invest in projects such as
building power plants or grid infrastructure, this is not the full picture in the case of decentralized energy supply
involving small systems, that center on the decisions of residential households. With decentralized energy,
former electricity consumers have an opportunity to become producers of their own electricity and will,
therefore, also play an active role in the future development of PV.
Today, the focus in Yemen needs to be on regaining stability, since “The outbreak of war or other widespread
violence spells the end of almost all productive investment, and a reasonable level of political and
macroeconomic stability is a threshold requirement for other policy improvements to gain much traction (World
Bank 2005).�? Figure A4.1 gives an overview of the actors, relations, and interacti ons defining the investment
climate for solar PV in Yemen.
59
Figure A4.1 Investment Climate for Solar PV
Source: Original compilation.
In the center of the figure is the (future) PV market. The government influences many aspects of an investment
climate and has the potential to leverage PV market growth in the future if policies supported and
complemented the existing private market initiatives. The market size and its segmentation will be defined by
investors’ interests, needs, and confidence. Investors can be both public and private, and both national and
international. There are market entry barriers for all investors, which define how difficult it is to enter the Yemeni
solar PV market. Specific regulations such as the requirement to follow specific administrative procedures, but
also local content requirements (for international investors), could be examples of such barriers. All potential
investors are influenced by the government and its regulations. In contrast to large utility-scale power plants,
which are mostly built by public companies, private firms, or by a combination of both (public-private
partnerships, PPPs), investment decisions for PV installations can also be taken by households and communities.
All investors will first evaluate whether an investment in PV is feasible for them, based on their liquidity and
“ability to pay�? as determined by the availability of self-owned funds (equity) and access to debt. This is the
minimum prerequisite to participate in the market. For those who do not have any alternatives to generate
electricity, the ability to financially cover the capital expenditure (CAPEX) needed to purchase a PV system is the
primary constraint to market participation. The ability to cover operational expenditure (OPEX) after the
system’s installation may also be accounted for when evaluating the ability to pay. However, since these costs
are lower for PV than they are for generators, primarily because fuel purchases are not required and there are
just a handful of moving parts, OPEX will play a smaller role for PV and market participants’ ability to pay. In
Yemen, battery exchange costs will be the biggest concern during the operational phase of most installations. If
60
a potential investor is able to pay for a PV installation, they will also evaluate whether they are willing to pay.
This will depend on the alternatives, which are defined by their availability and the expected profitability of the
PV investment. The profitability for an investor will depend on the costs of the PV system (CAPEX and OPEX), the
expected future revenues (savings from avoided fuel purchases), and the acceptable level of risk associated with
the investment. The more favorable these conditions are, the more investors would be willing to pay, the faster
the market would grow, and the more the capacity that would be installed. The effect of the PV market will be
the creation of jobs and electricity, which will contribute to poverty reduction and economic growth.
Since profitability, determined by revenues and the costs of a PV investment, is a decisive factor in the WTP of
different investors, the future investment climate for PV in Yemen will be looked at by presenting, first, the
barriers to solar PV development and, second, a profitability analysis of different PV market segments.
A4.1 BARRIER ASSESSMENT
The development of solar PV in Yemen will be influenced by a number of barriers, which are often interrelated
and include, but are not limited to, the following.
POLITICAL BARRIERS
It is obvious that being in a state of civil war has led to a disruption in energy supply in Yemen, and the
distribution of political power will continue to play a key role in the development of the country and its energy
sector. The PV market in Yemen today has been driven nearly exclusively by the private sector and a lack of
alternatives. Once stability and a sufficient level of security are regained, policy makers will have to define the
country’s overall energy strategy and energy mix to ensure that sufficient electricity is available. Electricity will
also have to be affordable.
In the past, tariffs for public electricity were heavily subsidized, but there were no direct subsidies for solar PV.
The average prewar consumer electricity price was 8 US cents per kilowatt hour (8¢/kWh), but these costs
covered only about 25 percent of the economic cost of supply, which would be at 32¢/kWh (World Bank 2017).
Clearly, it will be more challenging for PV, or any other technology, to compete with the grid electricity price of
8¢ than the economic cost of 32¢/kWh. It is, however, unclear whether this cost–price relationship is applied to
all customers, and even less clear whether this relationship would remain unaltered in the future: if new and
modern centralized generation technologies were built (a task that already proved difficult before the war) and
distribution losses were to be reduced, the competitive landscape for solar PV would be different.
Whether to subsidize grid electricity prices at 75 percent or redirect (parts of) this subsidy toward supporting,
for example, solar PV is a political decision for a future administration. Any decision to promote solar PV or other
energy sources, alternatively or in parallel, and any support in this regard, along with the extent of the support,
will influence solar PV’s growth in the country.
POWER SECTOR BARRIERS
As mentioned earlier, large parts of the Yemeni power supply infrastructure were damaged due to the conflict.
In addition, the organizational capacity of the Public Electricity Corporation (PEC) and other institutions will also
have to be reinvigorated. Several qualified PEC staff are reported to have left the country or had to seek other
61
job opportunities after the government could no longer pay the salaries of public workers (Alameri 2017). The
physical infrastructure will, therefore, have to be rebuilt, and staff will have to be recruited and educated.
The capacity of future infrastructure and staff will be shaped by political decisions about Yemen’s energy future,
especially with regard to solar PV, which did not play an important role in the public power sector before the
war. Any energy alternatives to PV that the power sector provides will influence the PV market’s growth.
Moreover, if grid electricity access is restored and ideally expanded to a larger share of the population, it may
become possible to connect PV to the grid. This would, in turn, decrease today’s reliance on solar batteries. Since
batteries are a very important and rather costly component of existing off-grid systems, grid connection could
reduce the costs of PV systems, because they would either not require a battery or require one with lower
capacity.
QUALITY-SPECIFIC BARRIERS
In the past, many low-quality PV products were sold on the Yemeni market. A 2019 study found that only about
22 percent of the surveyed products were certified or adhered to international standards (ITMAM 2019). The
study mentioned a lack of local standardization rules, a lack of customer awareness, and a short-term profit
focus of inexperienced solar companies as potential reasons for these low standards.
Low-quality products may harm solar PV’s reputation in the long run and deter future users once affordable
energy alternatives become available. Lower-quality equipment and services may not be easily identifiable
especially when labels and certifications are present but falsified. Certified products sold in Yemen are not much
more expensive than products without certification: for instance, while authenticated modules had an average
price of 60¢/W, modules lacking certifications by international standards were, on average, only 3¢/W cheaper
(ITMAM 2019). Low-quality hardware increases the risk of low performance, has a shorter useful life, and
increases O&M costs, thereby making PV less attractive.
Another factor determining the quality of a PV installation is of course the qualification of the personnel carrying
out the installation. Dedicated training and certification programs for solar installers would minimize the risk of
bad-quality installations due to a lack of know-how. As already indicated, low-quality installations could
paradoxically increase the FTEs/MW in the short run; the more O&M work that a system requires, the higher
the level of effort. If a battery exchange is required every 1–2 years, as appears to be the case in Yemen today
(see section 2.5, subsection Direct Formal Labor, Operation and Maintenance), instead of every 2–4 years, this
would double the level of effort and create jobs on a micro or single system level. However, if this system were
representative on a macro or market level, then the cost increase for this additional work would make PV less
attractive and destroy jobs in the long run, because the cumulative megawatts of PV would disappear from the
market long before the end of their “normal�? useful life of 20+ years, thereby reducing O&M jobs. Moreover,
annual installations would decrease as well, because, provided there are alternatives, additional efforts increase
costs, which would make PV less competitive than other technologies.
STRUCTURAL BARRIERS
At the moment, Yemen has no centralized organization like a solar association that could collect data on the PV
market and promote market growth by attracting investors from the country or, in a more stable environment,
62
from abroad. A central agency could introduce and monitor quality standards and provide the required training
to disseminate know-how across different elements of the PV market value chain. In addition, a central
organization like a Yemeni solar association could represent the solar companies’ interests at the government
level to reduce the political barriers described above.
FINANCE-RELATED BARRIERS
As regards PV financing, there are two primary challenges. First, a rather high up-front capital investment will
have to be financed by a mix of owner’s capital and external debt provided by financial intermediaries. This
financial aspect may be called the “capital expenditure challenge,�? and how PV systems can be financed for
those without sufficient means and no access to debt will have to be assessed. A 2017market assessment by the
World Bank found “that almost all systems are paid in cash and debt finance is not readily available to most
households,�? and that loans from financial institutions “are often mainly targeted at government employees and
customers able to provide guarantees�? (World Bank 2017).
International aid programs such as the Yemen Emergency Electricity Access Project improve access to finance,
as a project “leverages private financing for solar by providing a platform to extend commercial c redit to
households to make solar products affordable.�? In addition, microfinance institutions have also made solar
lending an important business line, with solar loans accounting for up to 20 percent of their credit activities
(World Bank 2018). Despite the efforts made, access to finance for solar installations remains critical for the
future growth of the PV market.
LOOKING AHEAD
Accounting for the above barriers, one needs to ask whether, and under which assumptions, decentralized solar
PV will be competitive in comparison with other decentralized generation technologies, most importantly, fuel
or diesel generators. Its degree of competitiveness will also be an important criterion for willingness to pay. To
assess PV’s profitability against other generation technologies, cash-flow-based analyses were conducted for
three relevant systems representing three typical PV customer segments. Based on International Energy Agency
(IEA) data, in 2018, residential consumption accounted for by far the largest share of final electricity
consumption (80 percent), whereas commercial and public customers accounted for 9 percent and industry
accounted for 3 percent (figure A4.2). The share of residential customers has been growing since the outbreak
of war in 2014–15, but has remained below 60 percent over the past 10 years.
63
Figure A4.2 Final Electricity Consumption (Share of Customer Segments), 2010–18
Source: Based on IEA (2020).
Of course, every system is unique and varies by size, configuration, location, and so on. Moreover, prices may
vary depending on the quality, the supplier, and the time of purchase and access to finance, as well as financing
conditions. Still, the reflections and results of economic analysis are relevant for assessing the prospects for PV
in Yemen.
As a result of market shares but also accounting for the importance of health services and agriculture for Yemen’s
economy, the following customer segments were assessed:
• Residential households
• Public health centers
• Agricultural water pumping stations
A4.2 PV INVESTMENT CASES: COSTS AND RISKS
Staying within the framework of evaluating the investment climate for solar PV, a profitability analysis was
conducted. The calculations are based on cash-flow models (CFMs). Sensitivity analyses provide an outlook of
changes to the payback period and the internal rate of return (IRR) based on changes in input parameters such
as system price, energy yield, and fuel savings. Sensitivity analyses reflect uncertainty regarding these
parameters’ future development and are, therefore, an important component of a risk assessment that investors
will formally or informally undertake before they decide whether an investment will be made.
For the business case simulations, solar radiation values were derived from the Global Solar Atlas (World Bank
2020): 2,300 kWh/kWp/year was used for the base case and a performance ratio of 0.8 was applied to arrive at
the specific yield of 1,840 kWh/kWp/year. This specific yield value is comparable to the PV output values shown
in chapter 1, Figure 1..
64
The analysis focuses on profitability, which is a key criterion for investors to decide whether they are willing to
pay for an investment. However, as mentioned above, profitability is also interlinked with at least two more
finance-related criteria that will determine whether a project will be realized or not. These criteria are (sufficient)
liquidity, which defines an investor’s ability to pay, and an acceptable risk level (figure A4.3).
Figure A4.3 Criteria for Financial Investment
Source: Original compilation.
Liquidity for a PV investment describes whether potential owners have access to the funds necessary to cover
PV investment costs (capital expenditure, CAPEX). They either possess the required money themselves or can
borrow it. Commercial banks in Yemen offer loans for PV investments, although these loans may not be available
to everyone. Liquidity will be decisive for an investor’s ability to pay. The business cases assume that the funds
for a PV investment are available so that the customer is able to pay.
Risk level refers to the possibility to incur losses based on changes made to the investment assumptions at the
beginning of a project. The assessment of risks’ investment criterion includes the time horizon for which
assumptions can be made with a reasonable degree of certainty. For the profitability analysis presented, it is
assumed that the input parameters known and defined for today will remain reasonably stable over time until
the end of a PV system’s useful life. However, if, for instance, the electricity supply situation or electricity prices
(also based on political decisions such as subsidies for specific energy generation technologies) fundamentally
change in the future, the outcome of the business cases would be very different. If potential investors cannot
reasonably be sure that the PV system will securely produce electricity and generate savings until the end of its
useful life or at least until the end of the payback period, then they would likely not invest in PV today but wait
until the situation can be more reliably assessed.
All three investment criteria play an important role in deciding the overall investment climate. The criteria are
interlinked and will vary according to the individual profile, the environment, and the preferences of a potential
PV investor. Policy measures have the potential to influence all three criteria, to incentivize investments in PV
and leverage market growth. Financial criteria are not the only ones factored in a decision to purchase a PV
system. The (current and expected) reliability of electricity supply, available electricity generation alternatives,
as well as environmental and health-related criteria all play a role.
65
BUSINESS CASE 1: RESIDENTIAL HOUSEHOLD
This case compares a rooftop-mounted PV system of a residential household with a privately owned small fuel-
powered (petrol-powered) electricity generator. Due to its small size, the generator used by the household
would be less efficient than larger generators.
DESCRIPTION OF THE CUSTOMER SEGMENT
A typical customer for this business case would be a residential household with 9 persons, as described in the
World Bank’s household survey (see also section 1.2, Table 1.3). The average yearly electricity consumption of
the household is estimated to be in line with the average values for Yemen overall: approximately 160 kWh per
person or 1,440 kWh for the entire household. Approximately 40 percent of the consumption occurs during the
night and needs to be covered by a solar battery. The customer also owns a small fuel-powered generator, but
since fuel deliveries were unreliable, the customer installed a 1-kWp SHS, potentially as a combination of
multiple smaller PV systems, ranging from pico-scale solar to a (larger) SHS. Considering only today’s electricity
consumption and the system’s PV production, 1 kWp would be slightly over-dimensioned given that it will
produce approximately 1,840 kWh/year. However, the customer knows that PV is a long-term investment, and
would like to buy new electric appliances in the future. The customer also knows that the PV system’s
performance will deterioriate slightly over time and that not all electricity generated can be used given that
there are times when the PV system produces electricity that is not immediately needed and cannot be stored
(for instance, because the battery is already fully charged and the surplus power can't be shared).
The customer wants to save as much fuel as possible to pay back the PV investment, only wanting to keep the
old generator as a backup system.
PV SYSTEM
The lifetime of the system was set to a rather conservative 20 years. An increase based on the lifetime of the PV
modules to 25 years could be reasonable provided the system was installed professionally and with equipment
of sufficiently high quality, which has often not been the case in Yemen.
The total turnkey PV system costs (CAPEX) for the above PV system were estimated to be approximately
$2,000/kWp without storage for a small system. In fact, a home system in Yemen may comprise different PV
systems of different sizes and quality. Most typical SHSs would be between 150 and 330 W, although there may
be large variations depending on the region and the type of household. Since smaller systems have a higher price
per kWp, this price is the highest among all three business cases analyzed in this report. The price for the battery
is estimated at $300/kWh. Total system costs per kWp, including battery storage, amount to $2,540/kWp. This
value is in the range of the prices reported by RCREEE (2020) for systems smaller than 330 W: between
$1,400/kWp and $2,945/kWp. Based on market research conducted for this study, the prices for SHS may even
be lower. The price estimate can thus be seen as conservative, and it should be possible to buy a system at a
better price. The effects of lower and higher prices on the payback time and the equity IRR will be shown later
in the sensitivity analysis.
The performance factor that determines which share of solar irradiation can be effectively converted into
“useable�? electricity was set to 80 percent. To achieve this rate, a professional-quality installation is required;
66
otherwise, the value would be lower. The resulting applied solar yield was 1,840 kWh/kW per year. The PV
system’s performance was assumed to decrease by 0.7 percent per year on account of the natural wear and tear
of the system components.
The operation costs for standard O&M were set to 1.5 percent of the system costs. This does not include battery
replacement, which was assumed to happen every three years, where battery costs were assumed to decrease
by 4 percent every year. Battery replacement depends on the battery type and the way it is operated. As was
presented earlier, batteries are replaced quite frequently in Yemen, with many batteries already exchanged after
one year. If this is the case, then it could be that the batteries are of low quality and/or they are not operated as
required, thereby shortening their lifetime. Advances in battery design are improving, with more efficiency and
longevity at a lower price. Table A4.2 lists the base-case assumptions for a residential, rooftop-mounted
PV system.
Table A4.2 Residential Rooftop PV System: Base-Case Assumptions
PV System System Operation - Savings
Project Duration Years 20
PV System Size kWp 1.0
Nominal Storage Capacity kWh 1.8
PV System Costs w/o Storage USD/kWp 2,000
Total PV System Costs incl. Storage USD/kWp 2,540
Total PV System Cost USD 2,540
Performance Factor % 80%
Degradation % p.a. 0.70%
Applied Solar Yield kWh/kWp/a 1,840
Average Yearly Generation kWh/a 1,712
Fixed Operation Costs PV % p.a. 1.50%
Battery Replacement Interval Years 3
Yearly Price Development Battery % (4.00%)
Source: Original compilation.
FINANCING OF THE SYSTEM
Many households do not have access to regular bank loans. Since debt financing is not currently an option for
many Yemenis, PV systems in the base case were financed with 100 percent cash (equity) from a household. As
the war continues, poverty spreads and the availability of funds to finance PV systems poses a major barrier (see
the financial barriers described above). The main motivation for a typical residential customer is not to earn a
high return on investment but to have access to electricity. Still, typical customers do not like to lose money
either and would therefore expect that their investment would recover the inflation rate. Therefore, the
business case was modeled with a modest discount rate for equity investment of 2 percent in line with the
modeled long-term 2 percent inflation rate. This is a simplification for Yemen, where systems, especially smaller
ones, will more likely be purchased in Yemeni rials. However, since the calculation is in US dollars, a hard
currency, a more stable setting is assumed. Table A4.3 summarizes the financing conditions for a rooftop PV
system.
Table A4.3 Residential Rooftop PV: Financing Assumptions
67
Financing
Debt (Gearing) - USD -
Loan Tenor Years -
Debt Interest Rate % -
Initial Equity USD 2,540
Additional Equity USD 257
Discount Rate % 2.0%
Longterm Inflation Rate % 2.0%
Source: Original compilation.
Additional equity is required during the operational phase of a project, when the fuel savings realized, thanks to
PV electricity, are insufficient to cover the operating costs. In this business case, battery replacement, which
takes place after three years of operation, requires the owner to invest additional equity.
SAVINGS AND REVENUES
The customer is self-consuming PV electricity, which leads to savings in fuel purchases for any generator. The
value of the savings depends on the fuel’s price as well as the generator’s efficiency, measured as the kWh
produced using 1 liter of fuel. In addition, the kilowatt-hours that can be saved depend also on the PV
consumption, meaning the percentage of useful electricity generated that can also be used at the time when
consumption occurs. A battery adds more flexibility in this regard because electricity can be stored in it at times
when PV generation exceeds electricity consumption, provided the battery is not fully charged already. Good
system configuration leads to proper dimensioning of the system and high direct consumption rates. For the
business case, it was assumed that 60 percent of the electricity generated by the PV system would be used
directly during the daytime, whereas 30 percent would be used for charging the battery and for nighttime use,
thus bringing the overall direct PV consumption to 90 percent.
Concerning generator efficiency, it was assumed that a generator needs 1 liter of fuel to produce 2.5 kWh of
electricity. The fuel is purchased at 50¢/liter. In addition, customers also save on O&M for the generator, most
importantly, for the exchange of the lube oil required to run it. The savings for the lube oil and other
consumables were estimated to account for 10 percent of the fuel consumption. Fuel prices were assumed to
remain stable and only increase by the long-term inflation rate of 2 percent. Given that a household already
owns a generator and also wants to keep it as a backup system, neither generator purchase nor replacement
costs were accounted for in this business case.
In summary, table A4.4 lists the savings of a rooftop PV system.
Table A4.4 Residential Rooftop PV: Savings
68
System Operation - Savings
Applied Direct PV Consumption % 60.00%
Applied Battery PV Consumption % 30.00%
Genset Efficiency kWh/ltr 2.5
Average Replaced Fuel Consumption p.a. ltr/year 585
Fuel Price (1st Ops Year) USD/ltr 0.50
Generator O&M costs as % of fuel costs % 10.00%
Fuel Price Escalation % p.a. 2.00%
Genset CAPEX fee saved USD p.a. -
Genset OPEX fee saved USD p.a. -
Generator related savings (average) USD/kWh 0.24
Source: Original compilation.
FINANCIAL RESULTS
The results for this business case are a positive net present value (NPV) of $1.476 and an IRR of 7 percent, which
mean the PV system is an economically viable investment under the base-case assumptions described above.
The equity investment is paid back after approximately 12–13 years, longer using discounted cash flows and
shorter if cash flows are not discounted. The levelized cost of electricity (LCOE) for the installation would be
18¢/kWh. Hence, the LCOE for rooftop PV systems is 6¢ lower than the average kWh generated by fuel-powered
gensets. The LCOE is relatively high primarily because of relatively high overall system costs, including the use of
a battery (table A4.5).
Table A4.5 Residential Rooftop PV: Results
Results
Net-Present Value USD 1,476
Equity IRR % 7%
Amortization - discounted payback period Years 13.19
Undiscounted payback period Years 12.13
LCOE (no subsidy) USD/kWh 0.18
Source: Original compilation.
Looking at this case in more detail, the equity cash flow for the base case is depicted in figure A4.4.
Figure A4.4 Residential Rooftop PV: Equity Cash Flow
69
Source: Original compilation.
As can be seen, the equity cash flow is stable each year due to a stable fuel price and, hence, stable savings.
Every third year of operation, batteries need to be exchanged, causing additional costs for system owners. For
the first three battery exchanges—in years 3, 6, and 9—fuel savings would not suffice to cover these costs, and
the owner of these PV systems would have to invest additional equity. For the exchanges from year 12 on, the
cash flow from fuel savings would suffice to cover the battery investment costs, which were also reduced by 4
percent a year to account for the expected future cost reductions for battery systems. The point where the
cumulated cash flows cross the x-axis shows the discounted payback period, which is approximately 13 years.
RISK INFLUENCE: SENSITIVITY ANALYSIS
Risks are uncertain future events that would adversely affect a project if they occur. Opportunities are the
opposite of risks: they are also uncertain events but with positive effects. A risk analysis would assign
probabilities to the effects to calculate a risk value:
������������������������ ������������������������������ ($) = Effect of risk ($) ∗ �����������������������?�����������?������������������������������ ������������ �����������?�����?�����������������������������?������ (%)
A sensitivity analysis is used to show the effect of modified assumptions on key economic performance indicators
only. This study shows how the discounted payback period (amortization) and the return on equity (equity IRR)
changed when some of the assumptions described above were modified. A probability of occurrence will be
defined by the project specifics and the investment climate when a project is undertaken in the future.
The first sensitivity analysis looks at the specific yield, which represents the kilowatt-hours produced by a PV
system per kilowatt-peak of capacity and per year (figure A4.5). It is calculated by multiplying the solar radiation
by a PV system’s performance factor. This factor includes the technical conditions for PV systems’ efficiency; the
efficiency, orientation, and inclination of PV modules; and possible losses due to shadowing, etc.
It can be seen that the financial results for PV installation increase when a system is built at a site that receives
higher irradiation. The equity IRR increases when more electricity can be harvested, and the payback period
70
decreases simultaneously. The electricity that can be generated will also depend on the quality of the equipment
used and the installation’s quality. Poor quality would increase the risk of generation of less electricity.
Figure A4.5 Residential Rooftop PV: Specific Yield Sensitivity
Source: Original compilation.
Another key factor to consider when assessing a PV system’s economic viability is its investment costs (CAPEX).
The higher these costs are, the less attractive an investment becomes. Costs have an important effect on a
potential PV customer’s ability as well as willingness to pay. Costs are influenced by the quality of the effect but
also by taxes, import duties, loan life cover ratio (LCR) requirements, and so on. It can be seen that the system
could still be repaid in the 20 years of its lifetime even if prices were higher than $3,500/kWp. It is also shown
that, with a system price of approximately $1,500/kWp, the payback period is reduced to approximately 7 –8
years (figure A4.6). Cost increases could, for example, stem from regulations enforcing the use of locally
manufactured equipment (see also section 2.3).
Figure A4.6 Residential Rooftop PV: System Price Sensitivity
71
Source: Original compilation.
Another parameter that affects PV’s economic attractiveness is the efficiency of generators, with which it is
compared, since the avoided fuel purchases determine the revenues from PV systems. The more kilowatts that
a generator can produce per liter of fuel, the more efficient it is and the less attractive a PV investment becomes.
In contrast, the payback life for a PV system will be shorter if it can generate fewer kilowatts per liter of fuel.
Larger and modern generators are indeed more efficient than smaller and older generators. If locally centralized
generators were used to supply an entire neighborhood, they would be more efficient, even though other
factors, such as the average load, also play a role. Moreover, the farther a generator is located from a household
means that there are added costs of electricity transmission and distribution that would have to be considered.
As can be seen in figure A4.7, a kWh–liter ratio of 3 instead of 2.5, as in the base case, would result in a payback
period of more than 18 years. In contrast, the payback period for PV systems would be reduced to 9.5 years if a
less efficient generator produced only 2 kWh of electricity per liter of fuel.
Figure A4.7 Residential Rooftop PV: Sensitivity of Generator’s Efficiency
Source: Original compilation.
72
Regarding the fuel price for generators, it is inherently clear that the higher the price of fuel, the more
economically viable a PV investment becomes (figure A4.8). However, if the initial fuel price were lower than
35¢/liter, a PV investment could not be financially justified, all other assumptions remaining equal.
Figure A4.8 Residential Rooftop PV: Fuel Costs
Source: Original compilation.
The next sensitivity analysis shows the future development of price for fuel products. In the base case, the annual
increase was set to 2 percent, meaning fuel prices will increase but only at the estimated long-term inflation rate
of 2 percent (in US dollars). If this annual increase were higher or lower, the payback period and equity IRR would
change accordingly: if fuel prices decreased every year, a PV investment would become less attractive, because
potential fuel savings would decrease. In the opposite case, with an increase in fuel prices, payback periods could
become significantly shorter (figure A4.9).
Figure A4.9 Residential Rooftop PV: Fuel Cost Development
Source: Original compilation.
73
Although this report focuses on decentralized solar PV, it can also be assessed how decentralized PV would
compare with electricity from the grid: the LCOE of PV systems varies in accordance with the parameters
presented above. Looking at the results of this business case, it can be seen that at the selected base case price
of $2,540/kWp, LCOE is 18¢/kWh. This would be higher than the average subsidized prewar electricity prices of
approximately 8¢/kWh but lower than the economic cost of 32¢/kWh. Hence, even though it would not be
attractive for a household to substitute a kilowatt costing 8¢ with a kilowatt costing 18¢, it would make sense
for a government to support the rollout of PV that costs 14¢ less than every kilowatt from a centralized system.
It also needs to be considered that if reliable grid electricity were available and PV electricity could be injected
into the grid, batteries would no longer be needed or could be smaller, which would decrease system prices and,
consequently, the LCOE of a grid-connected system.
Figure A4.10 shows how LCOE varies with system price. Even for a system price of $4,500/kWp, the LCOE of solar
PV would be 6¢ lower than the economic cost of centralized electricity supply of 32¢/kWh. On the other hand,
it is also shown that a very low system price would not allow the LCOE to match the subsidized electricity price
of 8¢.
Figure A4.10 Residential Rooftop PV: LCOE Development
Source: Original compilation.
BUSINESS CASE 2: HEALTH CENTERS
The second business case compares a rooftop-mounted PV system of a health center to a fuel-powered (petrol-
powered) electricity generator.
DESCRIPTION OF THE CUSTOMER SEGMENT
A typical customer for this business case would be a health center that would need to provide services around
the clock and cannot allow interruption of electricity supply. The health center also owns a fuel generator as a
backup system; however, given that diesel supply cannot be guaranteed throughout the year, it needs a PV
system that could be operated self-sufficiently at all times. This requirement can be met by a battery with a high
74
capacity. International organizations support the purchase and installation of PV systems for public health
facilities. Therefore, the center example is a 10-kWp solar system. In this scenarios, the public health
administration would like to know how a self-sufficient PV system compares with a diesel generator.
PV SYSTEM
The PV system’s lifetime was set to 25 years. The system is installed professiona lly by a “best-in-class�? solar
company. An O&M contract was also signed to assure high availability of the system. The system can thus be
considered to be of high quality.
The total turnkey PV system costs (CAPEX) for the above PV system were estimated to be approximately
$1,800/kWp without storage. This price is in between the prices of all three systems analyzed in this section.
The price for the battery was, as before, $300/kWh. The total system costs per kilowatt-peak, including storage,
amount to $2,376/kWp. Based on market research, the prices may even be lower. The price estimate can be
thus seen as conservative, and it should be possible to buy a system at a better price. The effects of lower and
higher prices on payback time and equity IRR will be shown later in a sensitivity analysis.
The performance factor that determines which share of solar irradiation can be effectively converted into
“useable�? electricity was set to 80 percent. To achieve this rate, a professional-quality installation is required;
otherwise, the value would be lower. The resulting applied solar yield was estimated at 1,840 kWh/kW per year.
The PV system’s performance was assumed to decline by 0.7 percent per year on account of its natural wear
and tear.
The operation costs for standard O&M were set to 2 percent of the system costs—a bit higher than in the other
business cases. This does not include battery replacement, which was assumed to happen every five years,
where battery costs were assumed to decrease by 4 percent every year. Battery replacement depends on the
battery and the way it is operated. Although batteries are replaced relatively frequently in Yemen, this is a high-
quality system that is operated professionally, and exchange periods are, thus, longer. Table A4.6 shows the
base-case assumptions for a PV system at a health center.
Table A4.6 PV System at Health Center: Base-Case Assumptions
PV System System Operation - Savings
Project Duration Years 25
PV System Size kWp 10.0
Nominal storage capacity kWh 19.2
PV system costs w/o storage USD/kWp 1,800
Total PV system costs incl. storage USD/kWp 2,376
Total PV System Cost USD 23,760
Performance Factor % 80%
Degradation % p.a. 0.70%
Applied Solar Yield kWh/kWp/a 1,840
Average Yearly Generation kWh/a 16,827
Fixed Operation Costs PV % p.a. 2.00%
Battery Replacement Interval Years 5
75
Source: Original compilation.
FINANCING OF THE SYSTEM
Given that installations at health centers are supported by international organizations, they can be financed
with long-term debt capital at interest rates below market levels; 60 percent of the investment costs are
financed by debt. In addition, the loan tenure is also relatively long, at eight years, which is only possible with
access to international finance. Yemeni banks would require a shorter payback period of two to four years. The
main motivation for the health center is not a high return on investment but rather access to reliable electricity.
Still, typical customers would expect that their investment would recover the inflation rate. Therefore, the
business case was modeled with a modest discount rate for the equity investment of 2 percent in line with the
modeled long-term inflation rate of 2 percent. This is, again, a simplification for Yemen where systems,
especially smaller ones, will be purchased in Yemeni rials. However, as explained above, as the calculation is in
US dollars, a more stable setting is assumed.
The financing conditions for a health center’s PV system are summarized in table A4.7.
Table A4.7 Financing Assumptions for PV System at Health Center
Financing
Debt (Gearing) 60% USD 14,256
Loan Tenor Years 8
Debt Interest Rate % 5%
Initial Equity USD 9,805
Additional Equity USD 6,461
Discount Rate % 2.0%
Longterm Inflation Rate % 2.0%
Source: Original compilation.
Additional equity is required during the operations phase of the project when the savings realized from fuel
saved are insufficient to cover operating costs. In this business case, the replacement of the battery, which takes
place after five years of operation, requires the owner to invest additional equity.
SAVINGS AND REVENUES
Customers are self-consuming PV electricity, which leads to savings in purchases of fuel for their generator. The
extent of savings depends on the price of fuel and also on the efficiency of the generator (that is, the kilowatt-
hours that can be produced with 1 liter of fuel). In addition, the amount of kilowatt-hours that can be saved
depends also on the PV consumption, meaning the percentage of useful electricity generated that can also be
used at the time when consumption takes place. A battery adds more flexibility in this regard because electricity
can be stored in the battery at times when PV generation exceeds electricity consumption, provided the battery
is not fully charged already. Good system configuration leads to a proper dimensioning of the system and high
direct consumption rates. For the business case it was assumed that 60 percent of the electricity generated by
the PV system would be directly used during the daytime, while 35 percent would be used for charging the
battery and nighttime usage, bringing the overall direct PV consumption to 95 percent.
76
With regards to the generator efficiency, it was assumed that the generator requires 1 liter of fuel to produce
3.5 kWh of electricity. The fuel is purchased for 50¢/liter. In addition, the customer also saves on O&M for the
generator, most importantly on the consumption of lube oil required to run the generator. The savings for the
lube oil and other consumables were estimated to account for 10 percent of fuel consumption. Fuel prices were
assumed to remain stable and increase only by the long-term inflation rate of 2 percent. Given that the health
center already owns a generator and also wants to keep it as a backup system, neither the generator purchase
nor replacement costs were taken into account for this business case.
Table A4.8 summarizes the savings for the PV system.
Table A4.8 Savings from PV System at Health Center
System Operation - Savings
Applied Direct PV Consumption % 65.00%
Applied Battery PV Consumption % 30.00%
Genset Efficiency kWh/ltr 3.5
Average Replaced Fuel Consumption p.a. ltr/year 4,351
Fuel Price (1st Ops Year) USD/ltr 0.50
Generator O&M costs as % of fuel costs % 10.00%
Fuel Price Escalation % p.a. 2.00%
Genset CAPEX fee saved USD p.a. -
Genset OPEX fee saved USD p.a. -
Generator related savings (average) USD/kWh 0.19
Source: Original compilation.
FINANCIAL RESULTS
The results for this business case are a positive NPV of $10,562 and an IRR on equity of 6 percent, which means
that the PV system is an economically viable investment under the base-case assumptions described above. The
equity investment is paid back after approximately 16–17 years; longer using discounted cash flows, shorter if
cash flows are not discounted. The LCOE for the installation would be 16¢/kWh. Hence, the PV system’s LCOE is
3¢ lower than the average kilowatt-hours generated by the fuel genset. The LCOE is relatively high mainly
because of relatively high overall system costs including the use of a large battery (table A4.9).
Table A4.9 Financial Results for PV System at Health Center
77
Results
Net-Present Value USD 10,562
Equity IRR % 6%
Amortization - discounted payback period Years 17.53
Undiscounted payback period Years 16.11
LCOE (no subsidy) USD/kWh 0.16
Min DSCR** x 1.01 x
Min LLCR*** x 1.05 x
Source: Original compilation.
Looking in more detail at this case, the equity cash flow is depicted in figure A4.11.
Figure A4.11 Health Center PV: Equity Cash Flow
Source: Original compilation.
As can be seen, the cash flow for equity is equal to zero in most years during the loan period of eight years. Every
fifth year of operation, the battery needs to be exchanged, which causes additional costs for the system owner.
Hence, during the loan repayment period, additional equity is required in year 5. The same applies for the battery
exchange in years 10 and 15. For the exchanges from year 20, cash flow from fuel savings would suffice to cover
the battery investment costs, which were also reduced by 4 percent a year in order to account for expected
future cost reductions of battery systems. The point where the cumulated cash flows cross the x-axis shows the
discounted payback period of approximately 17 years.
RISK INFLUENCE: SENSITIVITY ANALYSIS
The following figures show how two key economic performance indicators for the investment, that is the
discounted payback period (amortization) and a return on equity (equity IRR), change when some of the
78
assumptions described above are modified. The figures show how variations of assumptions influence
profitability.
The specific yield shows the kilowatt-hours produced by a PV system per kWp capacity and per year (figure
A4.12). It is calculated on the basis of the solar radiation multiplied by the performance factor of the PV system.
This factor includes the technical conditions of the efficiency of the PV system, the efficiency, orientation and
inclination of PV modules, possible losses due to shadowing, and so on.
It can be seen that the financial results for the PV installation increase when the system is built at a site with
higher irradiation: the equity IRR increases when more electricity can be harvested, and the payback period
decreases at the same time.
Figure A4.12 Health Centers: Specific Yield Sensitivity
Source: Original compilation.
Another important factor behind the economic viability of a PV system are its costs: the higher the costs, the
less attractive the investment. In figure A4.13, the system could still be repaid in the 25 years of its lifetime even
if prices were higher than $3,000/kWp. It is also shown that, with a system price of around $1,600/kWp, the
payback period is reduced to approximately 11–12 years.
Figure A4.13 Health Centers: System Price Sensitivity
79
Source: Original compilation.
Another parameter that affects the attractiveness of PV is the efficiency of the generator to which PV is
compared. The more kilowatt-hours that can be generated with 1 liter of fuel, the more efficient the generator
is and the less attractive the PV investment becomes. Inversely, if fewer kilowatt-hours can be generated with a
liter of fuel, the shorter the payback for the PV system will be. Larger and more modern generators are more
efficient than smaller generators. The health center example uses a more efficient generator than the other two
customer segments presented in this report.
As can be seen in figure A4.14, a kWh/liter ratio of 4 would result in a payback period of more than 21 years.
Inversely, if 1 liter of fuel only generates 3 kWh of electricity the payback period would be reduced to about 13
years.
Figure A4.14 Health Center PV: Sensitivity of Generator’s Efficiency
Source: Original compilation.
80
The higher the price of fuel, the more economically viable the PV investment, because the avoided fuel
purchases are the revenues of the PV system. However, if the initial fuel price were lower than 45¢/liter, the PV
investment could not be financially justified, all other assumptions remaining equal (figure A4.15).
Figure A4.15 Health Center PV: Fuel Costs
Source: Original compilation.
The next sensitivity analysis shows the future development of the price for fuel products. In the base case, the
annual increase was set to 2 percent, meaning that fuel prices would increase but only along the estimated long-
term inflation rate of 2 percent (in US dollars). If this annual increase were higher or lower, the payback period
and equity IRR would change accordingly. If fuel prices decreased every year, the PV investment would become
less attractive because potential savings would decrease. In the opposite case of increasing fuel prices, payback
periods could become significantly shorter (figure A4.16).
Figure A4.16 Health Center PV: Fuel Cost Development
81
Source: Original compilation.
Finally, we assess how the frequency of battery exchanges may affect the economic attractiveness of PV systems
in this business case. This business case is particularly suited to analyzing this sensitivity as the battery capacity
is bigger than the PV system for households. It is shown that the payback period increases substantially if the
battery needs to be exchanged more often than every five years (figure A4.17).
Figure A4.17 Health Center PV: Battery Replacement Intervals
Source: Original compilation.
BUSINESS CASE 3: SOLAR-POWERED PUMPING FOR IRRIGATION
The third business case compares a farm’s ground-mounted PV system to a privately owned generator powered
by diesel. The generator would be more efficient than smaller generators such as those used by households.
DESCRIPTION OF THE CUSTOMER SEGMENT
A typical customer for this business case would be a farmer or a group of farmers who need(s) to pump water
to irrigate their fields. The technology used is referred to as solar-powered irrigation systems (SPISs). The average
yearly electricity consumption of the pump depends on the quantity of water required as well as on the depth
of the groundwater reservoir from which the water needs to be drawn to the surface. The customer also owns
a diesel generator but as fuel deliveries were unstable, the customer installed a 20 kWp solar irrigation system.
A system of this size should have a positive influence on the price per kWp. The customer does not need a battery
as the water is either directed immediately to the fields or stored in a large water basin (mechanical storage).
PV SYSTEM
The lifetime of the system was set to a rather conservative 20 years. An increase based on the lifetime of the PV
modules to 25 years could be reasonable, provided that the system is installed professionally and with
equipment of sufficiently high quality, which is not always the case in Yemen today.
82
Total turnkey PV system costs (CAPEX) for the PV system part of the SPIS were estimated to be about
$1,600/kWp. According to a solar company from Yemen, system sizes for solar pumps vary mostly between 5
and 22 kWp. As larger systems and ground-mounted systems have a lower price per kWp than smaller rooftop-
mounted systems, this price is the lowest among all three systems analyzed. The fact that no battery is required
is also positive for the economic evaluation. Based on market research, the prices for solar pumps may even be
lower, hence the price estimate can be seen as conservative, and it should be possible to buy a system for a
better price. The effect of lower and higher prices on payback time and equity IRR will be shown later in a
sensitivity analysis. The costs are limited to the costs of the PV system and do not include the prices for the
pump, the water reservoir, or pipes and valves needed to distribute the water to the fields. It is assumed that
these costs would also have to be covered by the competing diesel-powered system.
The performance factor that determines what share of the solar irradiation can effectively be converted into
“useable�? electricity was set to 80 percent. The resulting applied solar yield was 1,840 kWh/kW per year. The
PV system performance was assumed to decrease by 0.7 percent per year, accounting for natural wear and tear
of the system. The costs of standard O&M were set to 1.5 percent of the system costs. The base-case
assumptions are listed in table A4.10.
Table A4.10 Solar-Powered Irrigation System: Base-Case Assumptions
PV System System Operation - Savings
Project Duration Years 20
PV System Size kWp 20.0
Nominal storage capacity kWh -
Total PV system costs USD/kWp 1,600
Total PV System cost USD 32,000
Performance Factor % 80%
Degradation % p.a. 0.70%
Applied Solar Yield kWh/kWp/a 1,840
Average Yearly Generation kWh/a 34,235
Fixed Operation Costs PV % p.a. 1.50%
Battery Replacement Interval Years -
Source: Original compilation.
FINANCING OF THE SYSTEM
As previously noted, many Yemenis do not have access to regular bank loans; but this may not be true for
commercial applications, especially solar water pumping. A financial institution in Yemen confirmed that it
finances up to 100 percent of the investment in solar water pumps. The interest rate (in US dollars) was set to
10 percent. The loan tenure would be rather short, at two years. As a result, the PV system in the business case
was financed with 80 percent debt financing and 20 percent cash from the farmer. For a larger farm, commercial
considerations do play a role. The water pumped is an important factor of production, and so any investors
would expect to recover more than just the inflation rate. Therefore, the business case was modeled with a
discount rate of 12 percent for the farmer’s equity investment (table A4.11).
83
Table A4.11 Financing Assumptions for Solar-Powered Irrigation System
Financing
Debt (Gearing) 80% USD 25,600
Loan Tenor Years 2
Debt Interest Rate % 10%
Initial Equity USD 7,255
Additional Equity USD 18,698
Discount Rate % 12.0%
Longterm Inflation Rate % 2.0%
Source: Original compilation.
Additional equity is required during the operations phase of the project, when the savings realized from diesel
saved are insufficient to cover operating costs. In this business case, the rather short debt tenure of two years
requires additional equity to cover the loan repayments in years 1 and 2 of the operating life of the PV system.
If the loan repayment could be distributed over a longer period of time (say, seven years), the cash flows
generated by the project alone would eventually suffice to cover the payment.
SAVINGS AND REVENUES
A solar PV customer self-consuming PV electricity saves the costs of purchasing diesel for a generator. The extent
of savings depends on the price of fuel and also on the efficiency of the generator (that is, the kilowatt-hours
that can be produced with 1 liter of fuel). In addition, the amount of kilowatt-hours that can be saved also
depends upon the PV consumption, meaning the percentage of useful electricity generated that can be used at
the time of consumption. A good system configuration leads to a proper dimensioning of the system and high
direct consumption rates. For the business case it was assumed that 60 percent of the electricity generated by
the PV system would be used in the absence of other electricity-consuming appliances. Water can be stored
when it is not being used in a tank (reservoir) but the irrigation of plants may not require constant water supply
year-round. Instead, depending on a number of parameters (most importantly crop type), the water may be
needed only at specific periods of time in alignment with the growth cycle of a specific crop. During periods
when water is not required for irrigation and no other uses are foreseen, there is a risk of overextraction. This
is simply because the system can be operated with no or very limited extra costs as long as the sun is shining
(Siepman n.d.). As a consequence of wasting water, groundwater depletion may be accelerated and cause
severe environmental damage. Moreover, if the pump can no longer extract water due to lowered groundwater
levels, it may also lose its productive use.
With regards to generator efficiency, it was assumed that the generator needs 1 liter of fuel to produce 3 kWh
of electricity. The diesel is purchased for 70¢/liter. This accounts for the fact that diesel fuel is more expensive
than the petrol fuel used in the other two business cases, and that costs for transport to rural areas may have
to be added to the market price. In addition, the customer also saves on O&M for the generator, most
importantly for the exchange of lube oil required to run the generator. The savings for the lube oil and other
consumables were estimated to account for 10 percent of fuel consumption. Diesel prices were assumed to
remain stable and increase only at the long-term inflation rate of 2 percent. Given that the farm already owns a
generator, neither the generator purchase nor replacement costs were taken into account for this business case.
84
The savings generated by a solar-powered irrigation system are listed in table A4.12.
Table A4.12 Solar-Powered Irrigation System: Savings and Revenues
System Operation - Savings
Applied Direct PV Consumption % 60.00%
Applied Battery PV Consumption % -
Genset Efficiency kWh/ltr 3.0
Average Replaced Fuel Consumption p.a. ltr/year 6,847
Fuel Price (1st Ops Year) USD/ltr 0.70
Generator O&M costs as % of fuel costs % 10.00%
Fuel Price Escalation % p.a. 2.00%
Genset CAPEX fee saved USD p.a. -
Genset OPEX fee saved USD p.a. -
Generator related savings (average) USD/kWh 0.19
Source: Original compilation.
FINANCIAL RESULTS
The results for this business case are a positive NPV of $10,631 and an IRR of 18 percent, which means that the
PV system is an economically viable investment under the base-case assumptions described above (table A4.13).
The equity investment is paid back after approximately 6–10 years—longer using discounted cash flows, shorter
if cash flows are not discounted. The LCOE for the installation would be 13¢/kWh. Hence, the ground-mounted
PV system costs 6¢ less per kWh, on average, than running the diesel genset. The debt service cover ratio (DSCR)
is lower than 1, which means that the project would not be acceptable to a bank (“bankable�?) based on project
cash flows alone. Banks issue loans based on the creditworthiness and securities of the borrower, or farmer in
this case.
Table A4.13 Solar-Powered Irrigation System: Financial Results
Results
Net-Present Value USD 10,631
Equity IRR % 18%
Amortization - discounted payback period Years 10.26
Undiscounted payback period Years 6.61
LCOE (no subsidy) USD/kWh 0.13
Min DSCR** x 0.36 x
Min LLCR*** x 0.37 x
Source: Original compilation.
**DSCR = Debt service cover ratio.
***LLCR = Loan life-cycle ratio.
Looking at this business case in more detail, the equity cash flow is depicted in figure A4.18.
Figure A4.18 Solar-Powered Irrigation System: Equity Cash Flow (Loan Tenure, 2 Years)
85
Source: Original compilation.
As can be seen, the cash flow for equity is stable each year due to a stable fuel price and hence stable savings.
In the first two years of the lifetime of the PV system, the loan needs to be paid back, which implies additional
costs for the system owner that need to be covered with additional equity. If the same case were calculated with
a longer loan tenor of seven years, the equity cash flows would be as depicted in figure A4.19.
Figure A4.19 Agriculture Water Pumping: Equity Cash Flow (Loan Tenure, 7 Years)
Source: Original compilation.
The point where the cumulated cash flows cross the x-axis shows the discounted payback period of
approximately 10 years. This does not change with the redistribution of the loan. The main change is that now
86
project cash flows suffice to repay the loan and the farmer no longer has to invest additional equity. Moreover,
the DSCR increases to greater than 1, which means that the project could become bankable based on this ratio.
The availability of long-term finance will play a role in the development of PV and hence PV-related employment
in the future. If PV projects were financed based on their projected cash flows and to a lesser degree based on
the securities of the borrower, greater access to PV would be facilitated.
RISK INFLUENCE: SENSITIVITY ANALYSIS
The following figures show how two key economic performance indicators for the investment, that is the
discounted payback period (amortization) and the return on equity (equity IRR), change when some of the
assumptions described above are modified. The figures show how variations of assumptions influence
profitability.
The specific yield shows the kilowatt-hours produced by a PV system per kWp capacity and per year. It is
calculated on the basis of the solar radiation multiplied by the performance factor of the PV system (figure
A4.20). This factor includes the technical conditions for the efficiency of the PV system—the efficiency,
orientation, and inclination of PV modules, possible shadowing, etc.
It can be seen that the financial results for the PV installation increase when the system is built at a site with
higher irradiation: the equity IRR increases when more electricity can be harvested and the payback period
decreases at the same time.
Figure A4.20 Solar-Powered Irrigation System: Specific Yield Sensitivity
Source: Original compilation.
The economic viability of a PV system also hinges on costs. The higher the costs, the less attractive the
investment. In figure A4.21, it can be seen that the system could still be repaid in its 20-year lifetime even if
prices were higher than $2,000/kWp.
Figure A4.21 Solar-Powered Irrigation System: System Price Sensitivity
87
Source: Original compilation.
Another parameter that affects the attractiveness of PV is the efficiency of the generator to which PV is being
compared. The more kilowatt-hours that can be generated with 1 liter of diesel, the more efficient the generator
is and the less attractive the PV investment becomes. Inversely, if fewer kilowatt-hours can be generated with a
liter of diesel, the shorter the payback for the PV system will be. Larger and more modern generators are more
efficient than smaller and older generators. Using locally centralized generators could boost efficiency, but other
factors such as the average load also play a role. Moreover, the farther the generator from the customer, the
more that losses during transmission and distribution of electricity have to considered.
As can be seen in figure A4.22, a kWh/liter ratio of 3.5 instead of 3 as in the base case would result in a longer
payback period. Inversely, if 1 liter of fuel generates only 2.5 kWh of electricity, the payback period would be
reduced to 7–8 years.
Figure A4.22 Solar-Powered Irrigation System: Sensitivity of Generator’s Efficiency
Source: Original compilation.
88
Regarding the price of diesel for the generator, it is inherently clear that the higher the price, the more
economically viable the PV investment becomes. However, If the initial diesel price were lower than 55¢/liter,
the PV investment could not be financially justified, all other assumptions remaining equal (figure A4.23).
Figure A4.23 Solar-Powered Irrigation System: Fuel Costs
Source: Original compilation.
The next sensitivity analysis shows the future development of the price for fuel products. In the base case, the
annual increase was set to 2 percent, which means that diesel and lube oil prices would increase but only along
the estimated long-term inflation rate of 2 percent (in US dollars). In case this annual increase were higher or
lower, the payback period and equity IRR would change accordingly. If fuel prices decreased every year, the PV
investment would become less attractive as potential savings decrease. In the opposite case of increasing diesel
and oil prices, payback periods could become significantly shorter (figure A4.24).
Figure A4.24 Agriculture Water Pumping: Fuel Cost Development
Source: Original compilation.
89
A4.3 BUSINESS CASE TAKEAWAYS
Although there are multiple other business cases and customers for the application of solar PV in Yemen, the
three exemplary business cases led to the following conclusions with regards to PV market growth and hence,
the ability of solar PV to create jobs in the future:
1. Solar PV is competitive in different business segments against the usage of diesel or fuel powered
generators.
2. The usage of batteries to cover nighttime electricity consumption has a significant influence on costs. As
long as batteries, depending on their size, are the most expensive component of a solar PV system, nighttime
use from batteries should be limited to drive down system costs. This does not mean that batteries should
not be purchased but rather that nighttime electricity usage is covered to a larger extent by other,
alternative energy sources if they exist. Uses such as solar pumping are a great application in this respect
because nighttime use is nonexistent or very limited. However, such systems should ideally be coupled with
other electricity consumers or uses to avoid the risk of water overextraction.
3. Costs can be decreased and affordability increased by building larger installations and reducing financing
costs.
4. The CAPEX issue of not having enough cash to afford a PV system even if it proves to be a profitable
investment can be overcome by bringing in the debt finance of both smaller (microfinance) and larger
financial institutions. Moreover, it could be evaluated to what extent remittances sent from Yemenis abroad
to Yemen could be channeled into PV.
Last but not least: all the business cases require an acceptable risk level, which also means a certain degree of
stability and predictability in order to allow the savings to materialize. No business case will yield positive results
if a solar system is destroyed after two years of its useful life.
90
ANNEX 5. LIMITATIONS AND FUTURE RESEARCH NEEDS
The following limitations with regard to the data used and the calculations presented in this report should be
considered when interpreting the results and planning future investigations of solar PV employment in Yemen.
Market sizing: It is not possible to predict with 100 percent accuracy how electricity consumption in Yemen will
develop in the future and what role solar PV will play in a future electricity mix. All of this will depend on many
factors that can only be better assessed once the situation in Yemen has stabilized, which will only be the case
after the current conflict situation has been resolved. Therefore, the estimated market size should be revisited
as soon as new information becomes available.
Underestimation of job multipliers, especially with regards to installation and distribution : The distribution list
used for the online survey and for the identification of interview partners is the most complete list of Yemeni
solar companies known today, but it only includes those companies reachable by email. Smaller companies
without email addresses were not included in the observations.
The most recent available information about the general structure of companies in Yemen —according to the
2004 census of the Central Statistical Organisation and used in the World Bank’s Economic Memorandum on
Yemen (World Bank Group 2015)—found that 92 percent of private enterprises are small companies with four
employees at most. But the companies who participated in both the guided interviews and the online survey
conducted for this study are substantially larger. Meanwhile, smaller companies such as solar shops that are
active in the solar PV business are underrepresented. As previously described, solar shops were visited by a local
consultant to learn about their business model but no quantification that could be used for the employment
multipliers could be obtained.
If the FTE/MW multipliers found through research on the larger companies are the same for these smaller shops,
the estimation of direct jobs would not vary. However, it can be assumed that, given the absence of economies
of scale (MW/shop) and scope (products/FTEs), these shops have a higher FTE/MW ratio than the larger
companies. Therefore, the direct job multipliers identified can be assumed to be underestimated, especially in
those parts of the value chain where the shops are active (so, mainly in distribution but also, albeit to a lesser
degree, in installation and aftersales, that is, O&M, work). The impact of these parts of the value chain is less
significant than in distribution because the shops are less active in these segments than the larger companies.
For instance, the shops focus on smaller systems that often can be installed or maintained by the customers
themselves. It is also possible that imports were underestimated if the volumes imported by small shops are an
important part of the overall market, and the solar shops would be less efficient in imports than larger
companies. Future research on solar PV employment could investigate the role and the employment factors of
the solar shops as well as the role of private labor (such as when a customer installs a PV system without the
assistance of a paid professional).
Overestimation of job multipliers, especially with regards to imports and wholesale: It is also possible that
some job multipliers were overestimated, taking into account that the solar companies surveyed represent a
large share of both the imported volume and the wholesale market. The FTE/MW in the industrial value chain
could be overestimated if the MW volume of imports and wholesale to small shops was very small or nonexistent
or if these shops used less FTE/MW than the larger companies surveyed.
91
Overestimation of upstream and downstream job multipliers (excluding O&M): As previously explained, the
direct job multipliers upstream and downstream were calculated based on the annual volume of solar PV sold
directly to the final customers off-grid distributed PV and the volume of solar PV installed by surveyed
companies. However, the volume of solar PV systems that first went to the wholesale market before installation
was not used in the calculation of the multiplier. Even though some share of imported PV will always be in
transition—it is unclear how much of the PV wholesale volume will be installed or when it will be used—it could
be argued that most will be installed at some point. Therefore, it could be used in the multiplier calculations as
well. If the wholesale volume was used for the multiplier calculation of direct jobs, this multiplier would
decrease. Therefore, the employment numbers presented in this report could be interpreted as high numbers
compared to other studies that do not distinguish between different types of market volumes.
Uncertainty around O&M multiplier: Based on the interviews, it was possible to calculate employment factors
for all job categories, except for O&M. Except for one interview, the O&M volume stated by interviewees did
not come close to matching the information provided by people who worked in O&M, even after the
interviewees were contacted again and corrections were made. The resulting O&M job multipliers varied
between 6.5 jobs/MW and 886 jobs/MW and only one company had a “reasonable�? 0.5 jobs/ MW multiplier.
There seems to be a big overlap between people working in installation and O&M, and so respondents did not
distinguish between O&M and installation work. Moreover, the concept of “cumulative capacity,�? which
determines the number of O&M jobs, seemed to be more difficult to capture than the concept of annual
installations, which is the basis for all other job categories along the PV value chain.
Some interviewees were contacted again, and the outcomes of their O&M-related mini-survey are presented in
section 2.5 of this report. The resulting O&M multiplier was based on a calculation which regrouped people
between installation and O&M, based on a calculated target value. However, this value was based on a limited
number of interviewees. Therefore, there is a higher level of uncertainty associated with the O&M job multiplier
than with the others. Since O&M jobs are long term (“sticky�?) in nature, future research should take a more
detailed look at the labor intensity of O&M in Yemen.
Limited data on informal direct labor: A percentage value based on the educational attainment and job profiles
of the interviewees, as compared to ILO data (2014a and 2014b), was applied to the direct formal job multipliers
to estimate informal labor. However, despite the ILO data being the latest available, the importance of informal
labor along the solar PV value chain could be reassessed. It can be assumed that informal labor will become even
more important for small companies that are not formally registered and employ fewer people than the
companies surveyed for this study. This would be true for solar shops, associated or not.
Missing data for indirect, induced, and productive use jobs: FTE numbers for these labor categories were based
on secondary literature reviews but were not assessed based on market data from Yemen. These numbers
should therefore be seen as rough first-level estimates only, which need to be interpreted with care. Future
research should look deeper into the indirect, induced, and productive labor of the Yemeni solar PV market.
No assessment of the quality of jobs: This report estimated the number of jobs in different categories but did
not assess the quality of the jobs created. In reality, not every FTE is of the same quality, whether in terms of
salary, job security, health-related aspects, or other criteria. These differences could have an impact on other
job categories as well, such as induced labor. For instance, a job that is better paid increases the purchasing
92
power of the worker, which should also have a larger impact on induced labor. In the future, a more detailed
assessment of the quality of labor created by solar PV in Yemen should be undertaken to ensure that not only
do job numbers increase but also that the quality of these jobs is sufficiently high.
Potential Sana’a bias: The overwhelming majority of companies identified and interviewed were in the city of
Sana’a. While Sana’a is the economic center of the country, it accounts for only 11 percent of Yemen’s overall
population (15 percent if the Sana’a region is included as well). As can be seen in this report, the companies from
Sana’a are active across all Yemeni regions, hence regional coverage has been achieved. Still, it needs to be
verified that companies from Sana’a truly dominate the market, as available information seems to suggest.
Another possibility is that there is simply not enough information from other regions available or accessible from
outside Yemen.
93
REFERENCES
[1] Amal Nasser, 2018: Rethinking Yemen’s economy, private sector engagement in post-conflict Yemen,
https://devchampions.org/files/Rethinking_Yemens_Economy_No3_En.pdf
[2] Basic Solar Energy Math, 2013: Solar Power for Ordinary People,
https://livingonsolarpower.wordpress.com/2013/03/23/basic-solar-energy-math/
[3] Bob van der Zwaan, Lachlan Cameron, Tom Kober, 2013: ‘Potential for renewable energy jobs in the
Middle East’ Energy Policy Journal 60, 296-304.
https://publications.tno.nl/publication/34633659/06X4Fw/w13062.pdf
[4] Bridge to India, 2014: “Beehives or elephants? How should India drive its solar transformation?�?
[5] Central Statistical Organisation, 2019: Statistical Year Book for 2017, http://www.cso-
yemen.com/content.php?lng=english&cid=131, last accessed on September 15th, 2020
[6] Dawud Ansari, Claudia Kemfert and Hashem al-Kuhlani, 2019: Solarenergie im Jemen-Konflikt:
Entwicklungen, Herausforderungen, Chancen [online]. Berlin: DIW Berlin.
https://www.diw.de/documents/publikationen/73/diw_01.c.678514.de/diwkompakt_2019-141.pdf
[7] European Investment Bank, 2016: Breaking Down Investment Barriers at Ground Level
https://www.eib.org/en/publications/breaking-down-investment-barriers-at-ground-level
[8] Filip Bryjka, 2016: ‘North and South Yemen as a Theatre of a Proxy War During the Cold War’, Journal
of Science of the Gen Tadeusz Kosciuszko Military Academy of Land Forces, 180(2), 19-34.
[9] Gogla, 2019: Off-Grid Solar, A Growth Engine for Jobs,
https://www.gogla.org/sites/default/files/resource_docs/gogla_off_grid_solar_a_growth_engine_for
_jobs_web_opt.pdf
[10] Harsh Desai and Erik Forsberg, 2020: OECD, MULTIDIMENSIONAL FRAGILITY IN 2020,
https://www.oecd-ilibrary.org/docserver/b4fbdd27-
en.pdf?expires=1610037945&id=id&accname=guest&checksum=82C013D26EE82DB25E67D24D1B4A
23CA
[11] Imed Drine, 2012: Youth Unemployment in the Arab World
https://www.wider.unu.edu/publication/youth-unemployment-arab-world
[12] International Energy Agency (IEA), 2020: Yemen: Key energy statistics,
https://www.iea.org/countries/yemen
[13] International Labour Organization (ILO), 2011: Assessing green jobs potential in developing countries,
A PRACTITIONER’S GUIDE, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/---
publ/documents/publication/wcms_153458.pdf
[14] International Labour Organization (ILO), 2014: Regional Office for Arab States, DEMOGRAPHIC AND
LABOUR MARKET TRENDS IN YEMEN, https://www.ilo.org/wcmsp5/groups/public/---arabstates/---ro-
beirut/documents/publication/wcms_358144.pdf
[15] International Labour Organization (ILO), 2014: Yemen Labour Force Survey 2013-2014,
https://www.ilo.org/wcmsp5/groups/public/---arabstates/---ro-
beirut/documents/publication/wcms_419016.pdf
[16] IRENA, 2019: Renewable Energy and Jobs – Annual Review 2020 [online}, ,
https://www.irena.org/publications/2020/Sep/Renewable-Energy-and-Jobs-Annual-Review-2020
94
[17] IRENA, 2019: Statistics Time Series, Trends in Renewable Energy,
https://www.irena.org/Statistics/View-Data-by-Topic/Capacity-and-Generation/Statistics-Time-Series
[18] IRENA, 202: Statistics time series, ‘Trends in Renewable Energy’
https://www.irena.org/Statistics/View-Data-by-Topic/Capacity-and-Generation/Statistics-Time-
Series, accessed July 2020
[19] IRENA, 2020: Global Renewables Outlook [online}, https://www.irena.org/-
/media/Files/IRENA/Agency/Publication/2020/Apr/IRENA_Global_Renewables_Outlook_2020.pdf
[20] ITMAM, 2019: “Product Quality Assessment Report�?
[21] Jay Rutovitz, Elsa Dominish, Jenni Downes, 2015: CALCULATING GLOBAL ENERGY SECTOR JOBS, Institute
for Sustainable Futures, UTS,
https://opus.lib.uts.edu.au/bitstream/10453/43718/1/Rutovitzetal2015Calculatingglobalenergysector
jobsmethodology.pdf
[22] Mansour Omeira, 2013: From informality to decent work in Yemen, paper for the ILO,
https://www.researchgate.net/publication/317552336_From_informality_to_decent_work_in_Yeme
n
[23] Power for All, 2019: Powering Jobs Census 2019,
https://www.powerforall.org/application/files/8915/6310/7906/Powering-Jobs-Census-2019.pdf
[24] PV Magazine, 2019: From the editor: Solar growth – linear, exponential or accelerating? https://pv-
magazine-usa.com/2019/09/23/from-the-editor-solar-growth-linear-exponential-or-accelerating/
[25] RCREEE, 2017: Assessment of the status of solar PV in Yemen,
https://www.rcreee.org/content/assessment-status-solar-pv-yemen
[26] RCREEE, 2020: Assessment of solar PV in Yemen (Phase II), final report
[27] Rua’a Alameri, 2017: ‘No income for a year: Yemeni professionals endure war without salaries’, Al
Arabiya News, https://english.alarabiya.net/en/features/2017/09/12/-No-income-for-a-year-Yemeni-
professionals-endure-war-without-salaries
[28] Solar Reviews, 2020: Are lithium ion solar batteries the best energy storage option?
https://www.solarreviews.com/blog/are-lithium-Ion-the-best-solar-batteries-for-energy-storage
[29] Stefan Siepman, Solar-powered groundwater pumping https://www.un-igrac.org/stories/solar-
powered-groundwater-pumping
[30] The World Bank, 2005: World development report 2005: a better investment climate for everyone,
http://documents1.worldbank.org/curated/en/554071468182337250/pdf/288290WDR00PUB0r0inve
stment0climate.pdf
[31] UNDP, 2014: Policy Note: Prospects of Solar Energy in Yemen
[32] UNDP, 2019: Prolonged conflict would make Yemen the poorest country in the world, UNDP study says.
https://www.undp.org/content/undp/en/home/news-
centre/news/2019/Prolonged_conflict_would_make_Yemen_poorest_country_in_world_UNDP.html
[33] United Nations Yemen, 2019: The power to save lives , https://yemen.un.org/en/20057-power-save-
lives, last accessed on September 14th, 2020
[34] UNOPS: YEMEN INTEGRATED URBAN SERVICES EMERGENCY PROJECT, Battery Disposal and Recycling
Analyses
95
[35] Veronique Mistiaen, 2020: ‘Yemen’s ‘microgrid girls’ power community amid war and COVID -19’, Al
Jazeera, 18.02.2020. https://www.aljazeera.com/features/2020/10/18/amid-war-and-covid-19-
yemeni-rural-women-set-up-solar-microgrid
[36] William Barnett, 1988: ‘Four Steps to Forecast Total Market Demand’, Harvard Business Review, July
1988. https://hbr.org/1988/07/four-steps-to-forecast-total-market-demand
[37] World Bank Group, 2015: Economic Memorandum on Yemen, “THE REPUBLIC OF YEMEN - Unlocking
the Potential for Economic Growth�?,
http://documents1.worldbank.org/curated/pt/673781467997642839/pdf/102151-REVISED-
box394829B-PUBLIC-Yemen-CEM-edited.pdf
[38] World Bank Group, 2020: Yemen Electricity Access Household Survey In Short, May 20th, 2020
[39] World Bank Group, 2020: Yemen Monthly Economic Update – June 2020,
http://pubdocs.worldbank.org/en/933471597048500989/Yemen-Monthly-Economic-Update-June-
2020-eng.pdf
[40] World Bank, 2016: Defying darkness Scaling up Sustainable Electricity Development in Yemen
[41] World Bank, 2017: Republic of Yemen: Restoring and Expanding Energy Access Reengagement Note,
http://documents1.worldbank.org/curated/en/655811496412539032/pdf/P158449-Output-Yemen-
Power-Sector-Reengagement-Note.pdf
[42] World Bank, 2017: TOWARD A BLUEPRINT FOR THE RECOVERY AND RECONSTRUCTION OF YEMEN
[43] World Bank, 2018: Report No: PAD2589: INTERNATIONAL DEVELOPMENT ASSOCIATION PROJECT
APPRAISAL DOCUMENT ON A PROPOSED GRANT
http://documents1.worldbank.org/curated/en/845691523844015143/pdf/PAD-04042018.pdf
[44] World Bank, 2020: Electric power consumption (kWh per capita) - Middle East & North Africa (excluding
high income) https://data.worldbank.org/indicator/EG.USE.ELEC.KH.PC?locations=XQ
[45] World Bank, 2020: ESMAP, Solargis, Global Solar Atlas, https://globalsolaratlas.info/map
[46] World Bank, 2020: Population growth (annual %) – Yemen Rep.,
https://data.worldbank.org/indicator/SP.POP.GROW?locations=YE, last accessed on February 11th,
2021
[47] World Bank, 2020: Rural Population (% total population) – Yemen Rep.,
https://data.worldbank.org/indicator/SP.RUR.TOTL.ZS?locations=YE;
https://data.worldbank.org/country/yemen-rep, last accessed on September 16th, 2020
[48] World Bank, 2020: Yemen Dynamic Needs Assessment (DNA): Phase 3—2020 Update,
http://documents1.worldbank.org/curated/en/490981607970828629/pdf/Yemen-Dynamic-Needs-
Assessment-Phase-3-2020-Update.pdf
[49] World Bank, ESMAP, 2020: GLOBAL PHOTOVOLTAIC POWER POTENTIAL BY COUNTRY,
http://documents1.worldbank.org/curated/en/466331592817725242/pdf/Global-Photovoltaic-
Power-Potential-by-Country.pdf
96
97