Rwanda Renewable Energy Fund Case Study
Rwanda Renewable Energy Fund Case Study
INVESTMENT
FUNDS
© CIF 2022
[Link]
The author would like to acknowledge the valuable discussions with members of
the project implementation unit (PIU), the Government of Rwanda, solar compa-
nies, financial institutions, and the World Bank team.
Photo Credits
All photos by CIF or under license to CIF unless otherwise noted.
Design
Art Direction: Andrea Carega
Graphic Design: Karlien Truyens.
DISCLAIMER
© CIF 2022
[Link]
This publication was produced by the Climate Investment Funds (CIF);
however, the findings, interpretations, and conclusions expressed in this
work do not necessarily reflect the views of CIF, its governing bodies, or
the governments they represent. While reasonable efforts have been
made to ensure that the contents of this publication are factually correct,
CIF does not take responsibility for the accuracy or completeness of
its contentsand shall not be liable for any loss or damage that may be
occasioned directly, or indirectly, through the use of, or reliance on, the
contents of this publication.
This report is based on research carried out in March 2022. More recent
data may have become available since the research was completed.
CIF encourages the use, reproduction, and dissemination of this text
for use in non-commercial products or services, provided that CIF is
appropriately acknowledged as the source and copyright holder.
2
PROJECT
DATA
COUNTRY Rwanda
SECTOR/S Energy
TOTAL PROJECT COST USD48.94 million, fully financed from SREP and executed as a
World Bank investment project
3
KEY
MESSAGES
EXECUTIVE
Market actors should
have access to multiple
sources of financing.
SUMMARY
Subsidies may be
required to further
expand solar home This case study examines the first four years of the planned six-
systems and mini grids year implementation of the Rwanda Renewable Energy Fund (REF)
in population segments project — the country’s largest off-grid expansion program — with a
with low income levels. focus on its delivery challenges and solutions.
Projects to promote The case study aims to provide lessons from the project on how
private sector teams have addressed delivery challenges in climate projects. The
participation and foster delivery challenges are the problems that hinder development
market development interventions and prevent practitioners from translating technical
should rely on extensive solutions into results on the ground. The case study explores
consultations with the major challenges during implementation, the solutions that
market participants and the government put in place in response to challenges, how the
solutions were arrived at, and key lessons. The Rwanda Renewable
near real-time market
Energy Fund (REF) project was selected for a case study because
monitoring.
of its innovative design and the proactivity with which the team
has addressed the multiple delivery challenges the project
Projects to promote
encountered.
private sector
participation and foster
The project’s design was motivated by the need to accelerate off-
market development
grid access to electricity in rural areas. The REF aims to promote
need flexible
private sector-led, off-grid renewable energy (RE) development,
mechanisms to respond
by providing lines of credit and creating an enabling environment
to market developments
for off-grid solar electrification. The project was implemented
and market participants’ by the Government of Rwanda through the Development Bank of
concerns. Rwanda (BRD), with management support from the World Bank
and financing from the Climate Investment Funds’ (CIF) Scaling Up
There is need for Renewable Energy Program (SREP).
continued dialogue of
multilateral development The project set up a fund for financing private sector engagement
banks and government in off-grid electrification. The core intervention at the heart of
to address delivery the project lies in the creation of lines of credit in local currency
challenges stemming for financial institutions to finance solar home systems and mini-
from policies and grids, and solar companies offering those solutions. Local financial
regulations and changes
to these.
4
institutions were expected to use these lines of intended users could not afford the solar systems.
credit to finance private solar companies to provide Many of the project’s intended end-users are
households with off-grid solutions. smallholders with few and irregular income sources
and lack of cash during the planting seasons. This
This case study traces the delivery challenges the can make it hard for them to make timely payments
project encountered during its implementation on pay-go systems and there were defaults.
and the government’s adaptive management as it
identified and addressed the delivery challenges. A project restructuring in 2020 introduced subsidies
to improve the affordability of solar home systems
CHALLENGE ONE: Skill and Human Resource to low-income households. The subsidies, which are
Misalignment. Building an effective and qualified results-based, are targeted at low-income households
project implementation unit (PIU) proved to be more in a progressive fashion. Solar companies are paid
challenging than expected. The BRD were unfamiliar for each installation made for eligible customers,
with World Bank procedures, lacked experience in upon the verification of the system installation
managing off-grid energy projects, and faced some and operation. Extensive monitoring of the market,
initial capacity issues. The challenge was resolved, dialogue with the solar companies, and the existence
and project management improved, after BRD of a pilot results-based project facilitated this
was trained in World Bank procedures, acquired solution.
experience, and brought onboard the necessary
human capital and instituted weekly calls for However, no progress has been made in the area of
monitoring, coordination and problem-solving. mini-grids. Essentially, mini-grid electricity tariffs are
unaffordable in the absence of grant funding and no
CHALLENGE TWO: Private Sector Limitations. REF’s grant financing is currently available. Furthermore,
design contained multiple financing windows, as it the project’s financing window to provide a line of
was based on the idea that solar companies would credit to mini-grid developers is insufficient.
be free to choose if they wanted to seek financing
from savings and credit cooperatives (SACCOs), CHALLENGE 4: Adverse changes in regulations from
banks and microfinance institutions, or BRD. One of revisions to off-grid targets and technical guidelines.
the credit lines, direct financing through BRD, was In 2021, challenges stemming from revisions to the
initially inactive because the government wanted to national off-grid targets and from Rwanda’s technical
develop sustainable sources of domestic financing standards for solar home systems affected the
for the off-grid sector. However, providing credit lines project. Although the government’s original National
at competitive terms did not sufficiently incentivize Electrification Plan of 2018 had set an off-grid
Rwanda’s financial institutions to lend to the off-grid provision target of 48 percent, it updated the plan
sector. In fact, the financial institutions showed little in the middle of 2021 with a lower off-grid target in
interest in lending to the solar companies during light of the faster-than-expected progress with grid
the initial years of the project, as they perceived this expansion and slower-than-expected off-grid uptake.
new business to be risky. Therefore, the movement of
funds was limited during the first three years of the Periodic updating of national electrification plans
project. The solution was to activate the window for is a good practice in general. However, some of the
direct financing from BRD to the solar companies. solar companies were concerned that the revision
to a lower off-grid target would leave too small a
CHALLENGE 3: Unaffordability to Target Population. market for solar home systems. A solution was found
Although the strengthened PIU was implementing in the form of an upward adjustment of the new
project activities and the solar companies had better off-grid target, thus making the provision of solar off-
access to finance, they were still not making sales grid solutions a financially worthwhile proposition.
and installations at the targeted rate, because the
5
The solution was arrived at through constructive institutions to finance the sector. Five key lessons
policy dialogue between the government and the that may be useful for similar projects are presented
World Bank. below:
Another issue emerged surrounding quality y Market actors should have access to multiple
standards for solar home systems. The government sources of financing. Though there were valid
formulated national quality standards for solar home reasons for steering the solar companies to
systems, specified in ministerial guidelines, because borrow from local financial institutions, it proved
it observed that some substandard systems were limiting. Direct lending from the BRD to the solar
being imported, with adverse consequences for end- companies through window 4 turned out to be
users and companies selling quality products. more effective.
y Subsidies are required to further expand the
However, the guidelines had become more penetration of solar home systems in population
demanding than the international norm, few segments with low income levels.
manufacturers met Rwanda’s standards, and imports y Projects to promote private sector participation
slowed as the guidelines made it hard to get products and foster market development should rely on
through customs. The government resolved the issue extensive and continued consultations with
by aligning its standards to international norms. market participants and near real-time market
monitoring.
As of 2022, BRD’s changes to the project, in response y Projects to promote private sector participation
to the challenges, have produced positive effects. and foster market development need flexible
Disbursements are up and the project is meaningfully mechanisms to respond to market developments
engaging the private sector in selling and installing and market participants’ concerns.
solar home systems, as well as working with financial y There is need for continued dialogue of
multilateral development banks (MDBs) and
government to address delivery challenges
stemming from policies and regulations and
changes to these. MDBs with ongoing sector
engagement, policy dialogue, and policy lending
in the same sector as an investment project are
well-positioned for this dialogue.
6
LIST
OF
ABBREVIATIONS
7
TABLE
OF
CONTENTS
ACKNOWLEDGMENTS 2
PROJECT DATA 3
EXECUTIVE SUMMARY 4
LIST OF ABBREVIATIONS 7
1. Introduction 10
1.1. Rwanda’s Investment Plan 10
2. Context 12
2.1. Intervention: Lines of Credit 13
3. Delivery Challenges 15
4. Tracing the Implementation Process 16
4.1. Challenge 1: Skill and Human Resource Misalignment 16
4.2. Challenge 2: Private Sector Limitations 16
4.3. Challenge 3: Unaffordability to Target Population 18
4.4. Challenge 4: Adverse Regulatory Changes 21
4.4.1. Off-grid electricity targets 21
4.4.2. Standards for solar home systems 22
4.5. COVID-19’s Impacts 22
5. Results 23
6. Conclusions and Lessons 24
ANNEX 1: LIST OF STAKEHOLDERS INTERVIEWED 26
ENDNOTES 27
8
LIST
OF
EXHIBITS
9
1. INTRODUCTION
Millions of people across Africa have no electricity. It 1.1. Rwanda’s Investment Plan
will take many years before electric grids reach all the
unconnected households. As such, solar products and Rwanda’s investment plan for SREP aimed to establish
mini-grids offer faster and cheaper “off-grid” solutions. the conditions for growth in off-grid electricity
Solar home systems are targeted for one household access for households, firms, and institutions.1 The
or a micro-business. Mini-grids are off-grid electricity investment plan was developed under the leadership
distribution networks that can generate electricity on a of the Government of Rwanda, with support from the
small scale. Nonetheless, the question remains: how to African Development Bank, the World Bank, and the
best make off-grid solutions available to the typically International Finance Corporation, as well as through
low-income households who need them? consultations with stakeholders in the energy sector.
This case study examines the first four years of the At the time that the investment plan was finalized
planned six-year implementation of the Rwanda (November 2015), the government had set an
Renewable Energy Fund (REF) project — the country’s ambitious target of increasing electricity access from
largest off-grid expansion program — with a focus on around 22 percent of the households in 2014 to 70
its delivery challenges and solutions. The aim of the percent by 2018. The government’s Energy Sector
project is to increase electricity access and facilitate Strategy Plan called for 48 percent of the country’s
private sector engagement in off-grid electrification households to be connected to the grid, and 22
through the establishment of a fund. With financing percent, or about 550,000 households, to be covered
from the Climate Investment Funds’ (CIF) Scaling by off-grid solutions.
Up Renewable Energy Program (SREP), the project
was implemented by the Government of Rwanda Though many shops and companies distributed solar
through the Development Bank of Rwanda (BRD), with products, off-grid coverage was considered to be at a
implementation support from the World Bank. low base; therefore, reaching the electrification target
required major investments. The investment plan
The project, which came into effect in November 2017, indicated that private companies were in a better
was restructured in March 2020, though the original position to deliver off-grid solutions, provided that
objective and end-date of September 2023 remained certain barriers to private sector participation could
unchanged. CIF visited Rwanda in March 2022 to be addressed. Therefore, it focused on helping the
collect the data for this case study. country to meet the 22 percent off-grid electricity
10
Photo: Sarine Arslanian / [Link]
access target through private sector-led off-grid and Poverty Reduction Strategy II and the current National
mini-grids. SREP provided the government with the Strategy for Transformation for 2017–2024.
initial capital for REF — the energy fund it had set up
to help finance the necessary investments. REF was Rwanda has identified energy as an essential
the only project to be financed under the investment condition for sustainable growth and development,
plan, though other donors provided complimentary recognizing the importance of providing reliable and
support to the sector. affordable energy for all, if the country is to achieve
its aspirations. Rwanda’s Economic Development and
The Rwanda investment plan is aligned with the Poverty Reduction Strategy II aimed for 70 percent
country’s national and energy sector priorities at the access to electricity by 2018. Subsequently, the
time of approval and to date. Vision 2020, adopted National Strategy for Transformation for 2017–2024
in 2000 and revised in 2011, envisages a primary revised the target to 100 percent access — to be
role for the private sector to serve as the engine of achieved through a combination of on-grid and off-
growth and poverty reduction for the country. Rural grid connectivity. The government also formulated a
development and reduction of rural poverty are major rural electrification strategy that emphasized private
themes in Rwanda’s Economic Development and sector participation and the use of solar home
systems and mini-grids.
11
2. CONTEXT
The REF’s project development objective is to government and the project preparation team were
increase electricity access in Rwanda through off- optimistic about the solar products’ growth potential.
grid technologies and facilitating private sector Spurred by new technologies, the cost of solar
participation in renewable off-grid electrification. systems had come down and new pay-as-you-go
REF aims to promote private sector-led, off-grid business models had emerged.
RE development by providing lines of credit and
creating an enabling environment for off-grid solar The project design also reckoned with barriers to
electrification. Essentially, REF is a fund that has been the expansion of the private off-grid sector. More
established to provide credit lines to support off- than 20 off-grid solar companies were active in the
grid electrification. Uganda and Tanzania have set up country at the time of designing the project, with
similar funds, also with World Bank support. more expressing interest. Optimism was expressed
in the project document about the private sector’s
The project design was motivated by the need to capacity to increase investment, provided that key
accelerate off-grid access to electricity in rural areas. barriers could be addressed. These barriers consisted
At the time the project was designed, electricity of financial and commercial constraints; the nascent
access was rising extraordinarily fast in Rwanda. It state of the off-grid market; inadequate technical
grew from about six percent (110,000 households) standards; capacity constraints, such as a shortage of
in early 2009 to 24 percent (600,000 households) by qualified technicians; and institutional constraints,
mid-2016, according to the REF’s Project Appraisal including the lack of clarity on the geographic area
Document. However, electrification was concentrated or socioeconomic strata targeted by off-grid services.
among grid-connected urban households in the top Mini-grids also came with additional barriers, such
income quintile. Achieving the national electrification as a lack of experienced companies and regulatory
target would require an aggressive expansion of the barriers, including the lack of clarity regarding the
market for off-grid systems — an expansion that the eventual connection to the grid.
government expected would involve both solar home
systems and mini-grids. The government and the World Bank’s project team
envisaged the outcomes as contributing to increased
However, the project design recognized the access to renewable energy and increased energy
affordability of off-grid solutions as a major security as well as helping create the necessary
barrier. The off-grid market was nascent in Rwanda. conditions for Rwanda’s off-grid energy markets to
Although the prices of solar systems had come down take off. This involved removing barriers to private
significantly worldwide, few Rwandan households sector participation in the market for solar products.
were able to afford the upfront costs of purchasing The intent was for the mechanism of REF to continue,
solar systems. Some earlier interventions, which had even after the close of the SREP-financed project.
involved the distribution of free systems, produced The project document foresaw follow-on financing to
negative experiences. For example, some households REF from the International Development Association
sold their systems. The government and the World (IDA) — the World Bank’s soft loan facility and other
Bank also did not want to give the systems away for development partner contributions — which did
free because their goal was to involve the private materialize.
sector in off-grid electrification. Ultimately, the
12
2.1. Intervention: Lines of Credit in the off-grid sector and thus ensure sustainable
financing for the sector even after the close of the
The core intervention at the heart of the project project. In the original design, the project was focused
involved the creation of lines of credit in local on developing the financial sector’s ability and
currency for financial institutions to finance willingness to finance off-grid expansion.
solar home systems and mini-grids for off-grid
electrification, and solar companies offering those. However, due to the continuous lack of interest on
Local financial institutions – banks, microfinance part of local financial institutions in lending to the
institutions, and savings and credit cooperatives off-grid sector, a credit line for BRD to lend directly
(SACCOs) – were expected to use these lines of to the off-grid sector was subsequently activated.
credit to finance private solar companies to provide Furthermore, an additional change to the project
customers with off-grid solutions. Prior to the project, was introduced: to address a lack of project progress
banks, microfinance institutions, and SACCOs were not in uptake, subsidies were provided, ensuring the
actively lending to the off-grid sector because they affordability of solar home systems to the targeted
preferred lending to traditional sectors with short- low-income households.
term financing needs and readily available collateral
(World Bank 2017). The project supported the sales, installation, and after-
sales services of solar home systems of good quality
The government expected that the credit lines — (Box 1). Specifically, the project supported Tier 1, or
essentially dedicated financing in local currency with multi-light point solar systems. This was intended to
long tenures and affordable rates — combined with avoid financing substandard solar home systems so as
technical assistance to address the financial sector’s to ensure that customers will have effective access to
lack of experience and understanding of the off-grid electricity for a period of time. Companies, which sell
market, would overcome these issues. The credit lines these products via two types of contracts (upfront cash
were expected to engage the domestic financial sector purchase and pay-as-you-go), are free to set prices.
13
Photo: Wirestock Creators / [Link]
Window 4. Direct financing of solar companies. At restructuring, in 2020, a fifth window was added:
This window would provide direct financing to
eligible, locally registered solar companies offering Window 5. Results-based financing for off-grid
Tier 1-and-above solar home systems and ongoing access. This window offers partial grants for the
maintenance services to its clients through delayed sale of Tier 1-and-above solar systems to poorer
payment options. Essentially, the window would households, pending the verification of installation
provide working capital assistance in the pay-as-you- and after-sales services. The partial grants are
go market. Though part of the original REF design, the targeted for lower-income households.
window was inactive at project inception. The project
implementation unit (PIU) activated Window 4 in
March 2019.
14
3. DELIVERY CHALLENGES
The project encountered several different types of 3 Unaffordability to target population: Solar
delivery challenges — both expected and unexpected off-grid solutions turned out to be largely
— during its implementation: unaffordable to many of their intended end-users,
without grant financing.
1 Skill and human resource misalignment: Building
an effective and qualified PIU proved to be more 4 Adverse regulatory changes: Policy and regulatory
challenging than expected. uncertainties around the national off-grid targets
and Rwanda’s technical standards for solar home
2 Private sector limitations: Providing credit systems emerged.
lines at competitive terms did not sufficiently
incentivize Rwanda’s financial institutions to The next section examines how the PIU team at BRD
finance the off-grid sector. addressed these challenges during implementation.
15
4. TRACING THE
IMPLEMENTATION PROCESS
This section examines how the PIU at BRD BRD lacked experience with lending to the off-grid
implemented the project with the World Bank’s sector. There were also some challenges in ensuring
support. Particular attention is paid to how the adequate project management capabilities of the PIU
PIU addressed and overcame each of the delivery upon project inception.
challenges by detailing the solutions and the process
of identifying those solutions. Solution 1: Hiring of appropriate skills. The
recruitment of a Senior Financial Advisor helped
The core project team comprised a project manager strengthened the project management capabilities
and staff in the PIU. A team of World Bank staff and of the PIU, though not to adequate levels. The
consultants supported the implementation. The World World Bank flagged the project implementation
Bank’s team leadership changed a few times because as unsatisfactory and highlighted the risk of not
of staff rotations. The World Bank handled the achieving the project development objectives. The
changes in team leadership well, via handovers and BRD resolved these challenges successfully by
co-Task Team Leader arrangements. making some personnel changes, which lent further
momentum to project implementation.
The project had a slow start. Based on its records,
only 2,853 people gained electricity access between Solution 2: Weekly calls for closer project monitoring
November 2017 and March 2020, compared to the and communication. Another adaptive management
targeted rates of 280,000 and 1,800,000 by the second response was to introduce weekly calls between the
year and the end of the project, respectively. By PIU and the World Bank.
March 2020, the project had only disbursed 12 percent
(USD5.65 million) to BRD.
4.2. Challenge 2: Private Sector
Electricity connections, disbursements, and other Limitations
results only picked up after the project team
responded to the delivery challenges that emerged REF’s project design relied on the use of a variety
by initiating a formal restructuring in March 2020. of financing windows, based on the idea that solar
Subsequently, connections reached 282,543 and companies would be free to choose where to seek
340,704 people in October 2021 and November 2021, financing from: SACCOs under Window 1, banks and
respectively. microfinance institutions under Window 2, or BRD
under Window 4. However, though part of the project’s
original design, was initially inactive. The logic behind
4.1. Challenge 1: Skill and Human having Window 4 closed initially stemmed from the
Resource Misalignment government’s wish for domestic financing for the
off-grid sector to be sustainable, before tapping
Staffing the PIU, who would be housed at BRD, proved into the direct lending option from BRD. The hope
to be more challenging than expected. As this was was that domestic financing development could be
the first World Bank project BRD had implemented, accomplished via the support of Windows 1 Window
the World Bank procedures were new to BRD. The 2 for SACCOs and banks, respectively, to lend to solar
16
companies, with the expectation that these sources in the absence of sufficient collateral in land and
would continue to finance the off-grid sector after the real estate. Under Window 4, BRD has piloted the use
project’s closing. of receivables as part of the collateral to show the
commercial banks that other forms of collateral can
However, although Windows 1 and 2 were designed be used; yet on-lending from Window 2 remains well
to support financial institutions in on-lending for below target.
the solar off-grid sector, the financial institutions
showed little interest in the initial years. This was At the same time, there were challenges that had
because the solar off-grid sector was a new business been foreseen and the solutions were, therefore,
which the financial institutions perceived to be risky. built into the project design in the form of technical
SACCOs — basic institutions with low capacity — were assistance and the direct financing window 4. Another
designed to introduce rural communities to finance by part of the solution was provided through a donor-
providing access. With regard to Rwandan banks, they funded guarantee facility that covered some of the
tended to lend to traditional sectors, with land and financial institutions’ risk from lending to the sector.
buildings used as collateral. Therefore, neither banks
nor SACCOs had experience in lending to the sector, or Solution 1: Technical assistance. The financial
to small and medium enterprises, without traditional institutions’ low interest in financing the off-grid
forms of collateral. sector was partially foreseen; as such, the project
design included the provision of technical assistance
Entering the world of off-grid sector financing is a to the financial institutions. However, while the
strategic decision that every bank needs to make. project’s technical assistance could hope to create
The commitment to enter off-grid financing would awareness and some capacity, for example by training
therefore need to come from the top. Some banks staff in how to handle loan applications, banks tend
preferred to maintain a more passive approach and to be conservative, with a predisposition toward
wait until another bank succeeds in the sector or the staying in their areas of comfort. The banks in Rwanda
risks and business dynamics of the off-grid sector did not make the strategic decision to lend to the
otherwise become clear before making the decision to off-grid solar sector, in part because of the issue of
enter. Further, once a bank decides to enter, it would collateral requirements that lie outside the project’s
also need to develop measures and accommodations scope.
relevant to solar companies, which are small
enterprises with few years in business and little or no Solution 2: Direct financing vs. on-lending. The
credit history. project design included a formal milestone one year
after project effectiveness to inform the activation
Window 1 and 2 moved few funds in the first three of Window 4 in the form of a detailed assessment of
years or so. The solar companies, being unable to the performance of Window 1, 2, and 3, that would
access finance, could not pay for the necessary stocks trigger a review of the need to activate Window 4.
to supply customers and time was wasted. The milestone specified the performance metrics
that would be used. The data-driven process helped
Two lessons can be derived from this experience. isolate the decision-making process from other
The first is that market actors should have access to factors. The team showed foresight in including the
multiple sources of financing. Though there were valid inactive Window 4 and the milestone for triggering a
reasons for steering the solar companies to borrow review of the need to activate Window 4, as part of the
from local financial institutions, it proved limiting. The original design. These measures thus allowed BRD to
second lesson is that there are limits to how much activate Window 4 through a technical, data-driven
an energy sector project can hope to address with process, without formally restructuring the project.
regard to systemic issues in the financial sector. In
Rwanda, commercial banks find it hard to on-lend,
17
Once activated by the PIU in March 2019, Window 4 growing the customer base. Dialogue with the solar
immediately attracted interest from solar companies. companies active in the market helped the PIU
Furthermore, the PIU also simplified the operation reach this understanding. Furthermore, many of
manual’s rather stringent eligibility requirements the areas demarcated for off-grid expansion by the
for solar companies. As a result, the PIU was soon electrification plan were dominated by subsistence
reviewing applications and negotiating loan terms agriculture, with the likely implication that weak
with at least five of the major solar companies. affordability would remain an issue for a long time.
Many of these changes were made in response to the Solution 1: Subsidy scheme. BRD and the World Bank
demand from the companies themselves. Ongoing formally restructured the project in 2020, bringing in
dialogue between the PIU and the companies, its the new Window 5 with a subsidy scheme. Whereas
openness to listening and revising eligibility criteria, Windows 1–4 provided financing, Window 5 offered
along with the regular calls between the PIU and the grants to improve the affordability of solar home
World Bank, helped the project team to identify these systems to benefit end-consumers. The grants are
solutions. channeled through solar companies, which have
to demonstrate that they are passing the grants on
to customers. The restructuring kept the project
4.3. Challenge 3: Unaffordability development objective unchanged.
to Target Population
Solution 2: The subsidy scheme was progressive. It
The project document discussed affordability of off- was tied to the welfare level of the household (see
grid solutions while also pointing to data showing Box 2 below for further detail).
that thousands of systems were being sold and
installed in the years prior to the project’s start. This Solution 3: Results-based model. These grants within
data suggested the existence of a large market for the subsidy scheme were results-based. The PIU pays
solar home systems through pay-as-you-go contracts. solar companies, upon the verification of system
The project design’s rationale was to address installation and operation, for each installation they
affordability by supporting pay-as-you-go models, make for eligible customers (see Box 2). The project
which the project did with its focus on expanding withholds a part of the grant, with the subsequent
solar companies’ access to finance. disbursement to the solar companies subject to the
system still being operational two and three years
However, during 2018–20, it became clear that after installation.
something was not working as expected. At this
juncture, the strengthened PIU was implementing Introducing the results-based financing model was a
project activities, Window 4 was now active, and major course correction. While the project focused on
providing financing to some of the solar companies. financing solar companies before the restructuring,
Non-project sources also provided an additional it would, henceforth, both finance solar companies
means of financing for the other solar companies. and subsidize the sale and installation of solar home
Yet sales tapered off and the solar companies systems (see Box 3 for further detail).
were making nowhere near the targeted rate of
installations. All indications are that the diagnosis of low
affordability was correct and that the subsidy scheme
The PIU was concerned and studied the issue in has accelerated sales once Window 5 became effective
dialogue with the solar companies, again supported in October 2021. Solar companies interviewed for this
by the World Bank’s team. The PIU concluded that case study described the results-based, progressive
the better-off customer market segment had been subsidy as a “game changer” and essential for their
saturated and that affordability was now key to ability to make sales to lower-income households.
18
Substantial dialogue and engagement had paved
BOX 2. How the subsidies work the way for the pivot to a results-based model.
The World Bank and the government had a broader
The subsidy levels are progressive, depending on dialogue on the off-grid sector for several years, also
the Ubudehe level of the household. Rwanda’s
Ubudehe system is a household registry used to outside the REF project. A pilot results-based project
classify households according to their welfare level. with GIZ/ EnDeV (Pro Poor RBF program) showed
The system is used, for example, in social protection good results, and later the World Bank supported the
programs. The project used this system to determine
government in developing its own version of the pilot
subsidies, because it provided a ready-built way to
categorize beneficiary households and determine model under an energy sector development policy
the subsidy level: the lower the Ubudehe category, credit. Eventually, they agreed to use REF to finance
the higher the subsidy. the results-based model. To make this happen, the
Sales to households in Ubudehe 1 receive a PIU needed to restructure the project and design the
90-percent subsidy of the cost of the system, new Window 5 in collaboration with the World Bank.
followed by 70 percent and 45 percent to Ubudehe
2 and Ubudehe 3 households, respectively. The
subsidies are capped, as indicated in the third
column below. In practice, this means that the
absolute subsidy levels apply to systems with an BOX 3. How the results-based model work
end-user price of USD120 and above, and that the
percentage subsidy levels apply to systems with an The project delays part of the subsidy to incentivize
end-user price of below USD120. This subsidy design after-sales service by solar companies. How the
has the effect of making basic systems affordable to subsidies are delayed depends on the contract
poorer households, while making less of a dent in types.
the price of larger systems.
For cash sales, the project reimburses companies 80
percent of the subsidy, upon the verification of the
HOUSEHOLD CATEGORY SUBSIDY % SUBSIDY sale and installation, and the remaining 20 percent
CAP, IN $ after three years.
For pay-go sales, the project reimburses companies
Ubudehe 1 (poorest) 90 100
for 45 percent of the subsidy upon the verification
Ubudehe 2 70 80 of the sale and installation, 45 percent in the second
year upon the verification of payment made by the
Ubudehe 3 45 50 customers, and the finall 10 percent after three years
upon verification that the system is operational. This
Ubudehe 4 (best off) Ineligible n/a
disbursement schedule seeks to ensure adequate
for subsidy
after-sale services from the companies.
Source: Project documents.
19
Designing Window 5 required substantial additional helped project team members to address the
technical work. The PIU had to determine the affordability delivery challenge:
households’ ability to pay based on a study
done by the Ministry of Infrastructure, which the y The slow progress was clear from the sector’s
government of Rwanda used to decide the subsidy data and the project’s data on disbursements and
levels in consultation with the World Bank. This connections.
work also required understanding Rwanda’s system y Policy dialogue helped the government to
of classifying household’s income levels — the recognize the need to address the affordability
Ubudehe system — that the government decided to issue and the development policy credit
use for setting the subsidies. The PIU also specified supported the design of a results-based model.
household eligibility criteria; devised reimbursement y A pilot off-grid results-based project in six
mechanisms for cash and pay-go contracts; and set districts of Rwanda, supported by Energising
up a verification mechanism (see Box 4). The World Development (EnDev), provided proof-of-concept.
Bank’s energy team and other practices provided That project, which involved subsidies, recorded
support for the PIU’s technical work on estimating good uptake. The government then tweaked the
the necessary subsidy levels and developing the pilot project’s design to arrive at its own model.
verification tool. y The team had an active dialogue with the solar
companies to understand the market perspective.
The policy dialogue and troubleshooting process y Some members of the PIU team also went on an
that resulted in the diagnosis of low consumer exposure visit to Bangladesh and Nepal, where
affordability and the subsidy design took into they learned about those countries’ experiences
account multiple perspectives and multiple sources with subsidized off-grid solar electrification.
of data and knowledge. According to interviews
and the restructuring paper, the following elements
20
There are some concerns regarding whether the grid developers was to access the financing to bring
subsidy scheme’s current design gives sufficient the project to commissioning. Window 3 was intended
incentives for after-sales service and whether it treats to provide a line of credit to mini-grid developers,
cash and pay-go contracts equitably. The project which would act as bridge financing to bring the
recorded instances where solar companies repossess project to commissioning before grant (subsidy)
pay-go customers who fail to make payments. Based financing from that project or other sources become
on the logic of results-based projects, companies available. However, the donor project has since
that repossess systems are not supposed to get closed. Without those grants, the rationale for Window
reimbursed, because the sale has not resulted in a 3 disappeared, as the mini-grid electricity tariff is
sustainably electrified household. Some stakeholders unaffordable in the absence of grant funding. At the
expressed concern that the risk of subsidy loss could time of writing, the future of mini-grids in Rwanda,
drive companies toward cash contracts and weak being dependent on grant funding, remains uncertain.
after-sales service. This would occur if the company
is satisfied with receiving 80 percent of the subsidy;
it will then ignore after-care sales and forgot the
4.4. Challenge 4: Adverse
remaining 20 percent of the subsidy. Stakeholders Regulatory Changes
also raised concerns that some companies may be
created solely to take advantage of the program 4.4.1. Off-grid electricity targets
by aiming for the 80 percent subsidy under cash Another challenge emerged when the government
contracts and exiting the market without providing of Rwanda revised downward the national target for
after-sales service. There are also fraud risks, with off-grid electricity provision. The original National
companies repossessing a system and then claiming Electrification Plan from 2018 had set a target of 48
the subsidy a second time when they resell the percent for off-grid electricity provision. However,
repossessed system. the government updated the plan in the middle of
2021 with a lower provisional target of 10 percent of
Issues around the incentive design and a need for villages.
continuous tweaks are normal under a results-based
financing model. With its commitment to continuous Periodic updating of national electrification plans
improvement, the PIU will, therefore, review, from is a good practice in general. The revision was
time to time, both the subsidies’ levels for the motivated by the government’s desire to accelerate
different Ubudehe categories and the incentives they electrification and reach 100 percent coverage in 2024.
give rise to. Sources explained that the government observed
faster-than-expected progress with grid expansion
In the case of mini-grids, no progress has been made, and slower-than-expected off-grid uptake for reasons
as they remain financially unviable. Although there is discussed above: low affordability of low-income
still potential for mini-grids in some remote places in population segments, the banks’ risk aversion, and
Rwanda, they are rarely economically viable without the REF project’s slow progress. The government
substantial subsidies. This is, in part, because most revised the national off-grid target downward because
mini-grids only support household consumption of the grid’s fast expansion. However, the downward
rather than productive uses in light industries of revision of the targets could discourage solar
which there are few in remote areas designated for companies from entering the market to provide off-
mini-grids. grid solutions.
At the time of the project design, there was a donor- Solution: Multi-stakeholder dialogue toward a win-
funded project implemented by EnDev that was win solution. A process of dialogue and problem-
providing grants to mini-grid developers, but only solving ensued. The World Bank’s energy team
upon commissioning. The challenge faced by mini- pointed out that the targets were not underpinned by
21
historical grid expansion performance, that financing According to the members of the World Bank team,
for grid electrification was not fully secured, and that Rwanda’s national quality standards for solar home
it would be risky to jeopardize the off-grid segment systems, as specified in its ministerial guidelines,
amidst ambitious grid expansion expectations. The had become more demanding than the international
World Bank’s energy team was well-positioned for norm. Consequently, only a few manufacturers met
this dialogue because of its extensive engagement in Rwanda’s standards. The import of solar products
Rwanda’s energy sector and strong relationship with slowed down because of confusion regarding how
the government. The Technical Working Group on customs agents at the border should interpret the
Off-grid Electricity — a forum that brings government ministerial guidelines.
and donor partners together— also facilitated the
dialogue. Ultimately, the discussions were successful Solution: Data-driven dialogue to align with
in producing an upward revision of the final off-grid international benchmarks. The government resolved
target to 30 percent, while the grid target was set at 70 the issue by aligning its standards to international
percent, as of late 2021. norms. Data-driven dialogue helped identify the
challenge. The project had collected data on the
The revised plan gives clarity and a sufficient companies’ stocks of solar home systems needed by
customer base for the solar companies. The 30- the PIU to administer the companies’ credit lines. This
percent off-grid target corresponds to around 1 data indicated that the stocks of solar home systems
million households, of which the government reckons were running low, primarily because of the regulatory
around half are already connected. That still leaves uncertainty surrounding the imports. Furthermore,
more households to connect than the project funds the World Bank Group’s lighting global team provided
can cover. Moreover, the electrification plan’s map technical support to the Rwanda Standards Board
of villages targeted for off-grid electrification helps to help Rwanda in aligning its national standards
provide clarity to the market. To meet the revised to international norms, according to interviews. The
targets, companies will, in fact, need to further issue was resolved as of March 2022: solar companies,
accelerate the rate at which they connect new interviewed in March 2022, indicated that they were
households, beyond the acceleration observed since able to import products and had stocks available for
the project restructuring. installations.
22
5. RESULTS
The PIU team faced a number of implementation The project covers 30 districts targeted for off-grid
challenges during the first approximately four years solar systems. It has trained 68 SACCOs and lent
of REF’s planned six-year implementation period to 56 SACCOs. Between them, these SACCOs have
covered in this case study. At the time of writing, the drawn down USD1,576,880 from Window 1 and on-
responses by the PIU team to those challenges are lent USD317,178 to households and enterprises, as
producing positive effects: disbursements are up, and of November 2021. Thirty-five percent of the SACCO
the project is engaging the private sector in selling on-lending went to women. Under Window 2, the
and installing solar home systems and working with project has lent USD2,722,104 to two commercial banks
financial institutions to finance the sector. However, and two micro-finance institutions (MFIs), which
since the project will be continuing until September have, in turn, on-lent USD201,215 to households and
2023 as planned, this case study provides only interim enterprises. Though the project has reached its target
conclusions and lessons. for the number of districts covered and exceeded the
target for participating SACCOs, the actual amounts of
The number of people provided with new or improved on-lending delivered by the SACCOs and banks are far
electricity service from solar home systems rose below the targets of USD4 million and USD8 million
significantly after Window 5 came into effect in for SACCOs and banks, respectively.
October 2020, reaching 340,704 by November 2021 (see
Table 1). A little more than half of these beneficiaries In the case of Window 4, the project has on-lent
are women. The project has also provided new or USD3,344,923 to off-grid solar companies, as of
improved electricity service to 1,309 enterprises, November 2021, with its end-target being almost
according to the project’s results monitoring. Although USD14 million by September 2023. With Window 5,
the acceleration in results is impressive, the project it has provided USD1,333,707 in result-based grant
remains far from achieving its targets of 1.8 million financing against a targeted amount of USD15 million.
people and 27,500 enterprises. As mentioned, Window 3 has not financed any mini-
grid developer, nor does the prospect look likely for
the duration of the project.
TABLE 1. People provided with new or improved
electricity service from the project REF — as a mechanism for financing private sector
involvement in off-grid solar — will outlast the project.
DATE NEW OR IMPROVED A follow-on World Bank IDA-financed project — the
ELECTRICITY SERVICE Rwanda Energy Access and Quality Improvement
June 2017 (baseline) 0 Project — came into effect in March 2021, with a
planned closing date at end-2026. The project is
October 2019 2,853
providing additional finance to REF, which will allow it
October 2020 3,180
to continue to provide results-based grants for solar
May 2021 114,813 home systems. It also aims to promote clean cooking
October 2021 (window 5 282,543 solutions, among other goals.
becomes effective)
23
6. CONCLUSIONS AND LESSONS
This case study examined the first four years of the from BRD to the solar companies — and a milestone
planned six-year implementation of the Rwanda for reviewing the need to activate the window.
Renewable Energy Fund (REF) project — the country’s
largest off-grid expansion program — with a focus on The first lesson from this experience is that it is
its delivery challenges and solutions. As explained better to give the market actors, the solar companies
below, the project encountered several challenges, in this case, access to multiple sources of financing
both expected and unexpected, which the government and offer choice. It did not prove successful to use a
addressed proactively in dialogue with the World sector intervention, such as REF, to address issues in
Bank. The challenges and solutions point to five key the financial sector, as the constraints stemmed from
lessons that may be useful for similar projects: collateral requirements and central bank regulations
outside the remit of the energy sector. This situation
y Market actors should have access to multiple resulted in weak access to finance for the off-grid
sources of financing. Though there were valid sector.
reasons for steering the solar companies to
borrow from local financial institutions, it proved Affordability issues were partially foreseen. In
limiting. Direct lending from the BRD to the solar the case of mini-grids, the project had envisaged
companies through window 4 turned out to be providing financing to mini-grid developers, based on
more effective. the expectation that other donor-funded programs
y Subsidies may be required to further expand the would provide the necessary subsidies. However,
penetration of solar home systems and mini grids this expectation did not materialize. In the case of
in population segments with low income levels. solar home systems, the project paper discussed
y Projects to promote private sector participation affordability issues while also referencing high sales
and foster market development should rely on data, leading to a project design that supported pay-
extensive and continued consultations with as-you-go models via credit lines to ensure adequate
market participants and near real-time market working capital for the solar companies.
monitoring.
y Projects to promote private sector participation Therefore, the second lesson highlights the need for
and foster market development need flexible subsidies to further expand the penetration of solar
mechanisms to respond to market developments home systems in population with low affordability
and market participants’ concerns. levels. The project shows that targeted subsidy
y There is need for continued dialogue of support to improve affordability can be done by using
multilateral development banks (MDBs) and results-based financing mechanisms, in partnership
government to address delivery challenges with the private sector, without undermining private
stemming from policies and regulations and sector-led, off-grid development.
changes to these. MDBs with ongoing sector
engagement, policy dialogue, and policy lending The unexpected challenges the project encountered
in the same sector as an investment project are include the downward revision of its off-grid
well-positioned for this dialogue. electrification target in 2021. This change in the
off-grid plan could potentially have undermined
Engaging the financial institutions in off-grid lending the REF project, had it not been for the productive
was probably the most expected challenge. The project technical dialogue involving the World Bank, the
design, therefore, included Window 4 — direct lending technical working group on off-grid electricity, and the
24
government that resulted in an upward revision of the y With its agile implementation support, the World
off-grid electrification target. Bank was able to pay close attention to the
day-to-day implementation details, including
Finally, though the initial capacity challenges in the the status of the contracts of the individual
PIU were not foreseen, it is common for entities companies and financial institutions with the
without prior experience in implementing MDB project. The World Bank provided implementation
projects to experience an initial learning curve and to support via frequent calls with the PIU team,
need some capacity strengthening. technical support through staff and consultants,
and country visits. It also provided hands-on
The identification of the solutions to these challenges support to the verification process, while awaiting
also changed the project’s focus in important ways, the hiring of an independent verification agency.
even while the project objective was maintained.
While the original project design had been focused The third lesson, therefore, concerns the need
on engaging the financial sector in providing off-grid for projects to rely on extensive and continued
finance, the more consequential solutions adopted consultations with market participants and near real-
by BRD led to the project’s provision of support to the time market monitoring, in order to promote private
off-grid sector via subsidies and direct lending. sector participation and foster market development.
Several factors facilitated the PIU team’s adaptive Furthermore, a related lesson — lesson 4 — is the
management: need for flexible mechanisms to respond to market
developments and market participants’ concerns.
y The PIU and the World bank worked In the REF project, the ability to design and adjust
collaboratively to detect, diagnose, and resolve incentives and the financing windows’ eligibility
challenges. criteria allowed for that flexibility.2
y Active dialogue with companies and financial
institutions helped the PIU team to understand The project has also revealed the significance of
the market, its challenges, and the corrective complementarities between project and policy
actions needed, so that effective incentives for the interventions. In this context, the World Bank’s
corresponding stakeholders could be designed. broader engagement in Rwanda’s energy sector has
y The project’s design with separate financing been helpful for the project, as its dialogue with the
windows also proved to be flexible. BRD could government and the donors on the energy sector
activate direct financing (Window 4) with relative development have allowed it to engage on issues of
ease and add new results-based grants (Window 5). policy and strategy beyond what would have been
y Predetermined formal milestones helped trigger possible had the project been its only engagement
reviews and course corrections. The World Bank in the sector. For example, the energy sector
and the Government of Rwanda inserted three development policy-financing operation supported
formal milestones in the project design document the development of the results-based financing
because they anticipated the need for learning by model that eventually became REF’s Window 5.
doing and course corrections. They included (1)
a review of the need to activate Window 4 after The final lesson is therefore that there is need for
12 months; (2) a review about one-third of the continued dialogue to address delivery challenges
way into project implementation, that is, after stemming from policies and regulations and changes
two years; and (3) the standard mid-term review. to these. MDBs with ongoing sector engagement,
The first two reviews were highly consequential, policy dialogue, and policy lending in the same sector
as they triggered the activation of Window 4 and as an investment project are well-positioned to help
restructuring, respectively. unlock delivery challenges stemming from adverse
changes in policy and strategy.
25
ANNEX 1: LIST OF
STAKEHOLDERS
INTERVIEWED
NAME POSITION ORGANIZATION
Denis Rugamba Manager Renewable Energy Fund Project BRD
Alida Ikuzwe Investment Manager — energy portfolio BRD
Umesh Prasad Acharya Renewable Energy Financing Expert BRD
Uwera Rutagarama Director, Off-grid and Alternative Energies EDCL, Rwanda Energy Group
Alex Gready Finance Manager, BBOXX BBOXX
Innocent Mitali Retail Projects Coordinator BBOXX
Agnes Shop manager, Musanze BBOXX
Kayumba Polepole Director SME finance CoqeBanque
Peace Kaliisa External links and Donor Coordination Officer, eSWAP Ministry of Infrastructure
Secretariat
Ndayishimiye Dieudonne Managing Director BUIM Ltd.
Samuel Bimenyimana Managing Director Hello Renewables Ltd
Federico Querio Task Team Leader (TTL), Senior Energy Specialist The World Bank
Arun Singh Energy Specialist The World Bank
Chiara Rogate Former TTL, Senior Energy Specialist The World Bank
Joern Huenteler Former TTL, Senior Energy Specialist The World Bank
Sarah Melissa Leitner Advisor EnDev
26
ENDNOTES
27
THE
CLIMATE
INVESTMENT
FUNDS
The Climate Investment Funds (CIF) is one of the
largest multilateral climate funds in the world. It
was established in 2008 to mobilize finance for
low-carbon, climate-resilient development at scale
in developing countries. 14 contributor countries
have pledged over US$10 billion to the funds. To
date CIF committed capital has mobilized more
than $62 billion in additional financing, particularly
from the private sector, in 72 countries. CIF’s large-
scale, low-cost, long-term financing lowers the risk
and cost of climate financing. It tests new business
models, builds track records in unproven markets,
and boosts investor confidence to unlock additional
sources of finance.
CLIMATE
INVESTMENT
FUNDS
@CIF_action
Telephone: +1 (202) 458-1801 CIFaction
Internet: [Link] Implementing MDB
CIFaction CIFaction
CIFaction CIFaction
CIFaction @CIF_action