Greening Digital Infrastructure in Eastern Africa World Bank Group June 2025 Report No. 202871 © 2025 The World Bank Group 1818 H Street NW, Washington, DC 20433 Telephone: +1 202-473-1000; Internet: www.worldbankgroup.org Some rights reserved 1 2 3 4 21 20 19 18 The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the World Bank Group, its Board of Executive Directors, or the governments they represent. The World Bank Group 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 Group, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2025. Greening Digital Infrastructure in Eastern Africa. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by the World Bank Group and should not be considered an official World Bank Group translation. The World Bank Group shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by the World Bank Group. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by the World Bank Group. Third-party content—The World Bank Group does not necessarily own each component of the content contained within the work. The World Bank Group therefore does not warrant that the use of any third party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org 2 Acknowledgments T his report has been prepared by Thomas Birk (Consultant, DAEDU), under the supervision of Tim Kelly (Lead Digital Development Specialist, DAEDU). Funding for this study The information contained in this document has been compiled by the consultant and may include material from third parties which is believed to be reliable but has not been was generously provided by Public Private Infrastructure independently verified or audited. The findings enclosed in Advisory Facility (PPIAF) under a Trust Fund to support this document may contain predictions based on current the Kenya Digital Economy Acceleration Project (KDEAP; data and historical trends. Any such predictions are subject P170941), the Eastern Africa Regional Digital Integration to inherent risks and uncertainties. It is important to note Series of Projects (EARDIP, P176181; P180931), and the that this risk analysis should not substitute for consultations Digital and Energy Connectivity for Inclusion in Madagascar with local experts or the integration of observational project (DECIM, P178701). data. While every effort has been made to ensure that the information in this document is derived from reliable The report was prepared for the purpose of assessing climate sources, the World Bank disclaims responsibility for any risks and providing high level guidance on climate resilience errors or omissions, as well as for the outcomes resulting and other greening measures (e.g. renewable energy) from the use of this information. to digital infrastructure in Eastern Africa. Specifically, this study focuses on Eastern African countries Djibouti, The opinions expressed in this document are valid only for Ethiopia, Kenya, Madagascar, Somalia and South Sudan. The the purpose stated herein and as of the date stated. No final report was subject to a World Bank review, conducted obligation is assumed to revise this document to reflect virtually in May 2025, in a decision meeting chaired by Isabel changes, events or conditions, which occur subsequent to Neto, Practice Manager, Digital Development, Eastern and the date hereof. Southern Africa. We wish to thank Edward Oughton for providing technical assistance with the risk analysis in the form of georeferenced maps of infrastructure and hazards, as well as exposure and vulnerability data for fiber networks and cell sites. We also wish to thank [insert reviewers] for their insightful comments which have been incorporated into this revised version, Victor Kyalo, Senior Digital Development Specialist (DDAEU); Himmat Singh Sandhu, Digital Specialist (DCADU). We wish to thank also Martha Oringo for her work in page composition of the final report. 3 Contents Acknowledgement 4 Figures, tables and boxes 6 Abbreviations and acronyms 7 1. Introduction 10 Key take aways 11 2. Risk analysis 16 2.1 Floods 16 2.2 Landslides 19 2.3 Cyclones, storms and coastal flooding 19 2.4 Other hazards 20 3. Climate resilience of digital infrastructure 21 3.1. Technical measures for network resilience 22 3.2. Institutional capacity for resilience 24 4. Resilience standards and performance requirements 26 4.1 Resilience specifications for different hazards 34 4.2 KPIs to enhance resilience outcomes 35 4.3 Risk allocation management 37 4.4 Enhancing Resilience: Costs and Benefits 39 5. Greening of energy for digital infrastructure 41 5.1 Greening standards 45 5.2 Greening Key performance indicators (KPIs) 48 6. Procurement 50 7 Key recommendations 53 Annex A: Country case – Djibouti 56 Annex B: Country case - Ethiopia 62 Annex C: Country case - Kenya 68 Annex D: Country case - Madagascar 75 Annex E: Country case – Somalia 83 Annex F: Country case - South Sudan 91 Annex G: Technical note on data analysis and risk assessment 97 Annex H: Project-specific climate adaptation and mitigation objectives and policy alignments 100 Project alignment with national agenda on climate goals 103 4 Figures Fig 2.1: Examples of underground fiberoptic cable exposure to climate hazard impacts 14 Fig 4.1: Risk Management 37 Fig A1: Projected Change in Temperature and Precipitation in Djibouti 56 Fig A2-3: Mobile cell locations and fiber optic networks routes 57 Fig A4: Projected riverine flooding, RCP4.5 climate scenario, severe flooding event 58 Fig A9: Landslide risk exposure for medium and high-risk areas 59 Fig A10: Estimated landslide exposure for mobile cells and fiber assets in Djibouti, reported by risk category 60 Fig A11: Historical cyclone tracks for Djibouti 61 Fig A12: Projected coastal flooding, RCP8.5 climate scenario, 2080 61 Fig B1: Projected change in temperature and precipitation, Ethiopia 62 Fig B2-3: Mobile cell site and fiber optic network locations 63 Fig B4: Projected River flooding, RCP4.5 climate scenario, severe flood event 64 Fig B5-8: Estimate riverine flooding damage and damage costs 65 Fig B9: Landslide risk exposure for medium and high-risk areas 66 Fig B10: Estimated landslide exposure for mobile cells and fiber assets, reported by landslide category 67 Fig C1: Projected change in temperature and precipitation, Kenya 67 Fig C2-3: Mobile cell and Fiber-optic network locations 67 Fig C4: Projected river flooding RCP4.5 climate scenario, severe flood event, 2080 70 Fig C5-C8: Estimate riverine flooding damage and damage costs, Kenya 71 Fig C9: Landslide risk exposure for medium and high-risk areas 72 Fig C10: Estimated landslide exposure for mobile cells and fiber assets in Kenya, reported by risk category 73 Fig C11: Projected coastal flooding RCP4.5 climate scenario, severe flood event, 2080 74 Fig D1: Projected temperature and precipitation 75 Fig D2-3: Mobile cell site and fiber-optic network locations 76 Fig D4: Projected river flooding RCP4.5 climate scenario, severe flood event 76 Fig D5-D8: Estimate riverine flooding damage and damage costs, Madagascar 78 Fig D9: Landslide risk exposure for medium and high-risk areas 80 Fig D10: Estimated landslide exposure for mobile cells and fiber assets in Madagascar, reportet by risk category 81 Fig D11: Historical cyclone tracks for Madagascar 81 Fig D12: Cyclone exposure Source: ACAPS thematic report 82 Fig D13 Projected coastal flooding, RCP8.5 climate scenario, severe flooding event 83 Fig E1: Projected temperature and rainfall, Somalia, Somalia 83 Fig E2-3: Mobile cell site and fiber network locations 84 Fig E4: Projected river flooding, RCP4.5 climate scenario, severe flood event 85 Fig E5-E8: Estimate riverine flooding damage and damage costs, Somalia 86 Fig E9: Landslide risk exposure for medium and high-risk areas 87 Fig E10: Estimated landslide exposure for mobile cells and fiber assets Somalia, reported by risk category 88 Fig E11: Historical cyclone tracks Somalia 89 Fig E12: Projected coastal flooding RCP8.5 climate scenario, 2080 90 Fig F1: Projected temperature and precipitation 91 Fig F2-3: Mobile cell site and fiber optic network locations 92 Fig F4: Projected river flooding, RCP4.5 climate scenario, severe flood event 93 Fig F5-F8: Estimate riverine flooding damage and damage costs, South Sudan 94 Fig F9: Landslide risk exposure for medium and high-risk areas 95 Fig F10: Estimated landslide exposure for mobile cells and fiber assets in South Sudan, reported by risk category 96 Fig G1 Damage curves by hazard type 98 5 Tables Table 1.1: Greening considerations for climate adaptation and mitigation in the three project 11 Table 1.2. Categorized summary of resilience and mitigation objectives included in project 11 appraisal documents (PADs). Table 2.1: Summary of hazards risk levels in six Eastern Africa countries 14 Table 2.2: Summary of country specific hazard risks to cell sites and fiber networks 15 Table 2.3: Flood and landslide risk to current and planned fiber optic cable routes and cell sites 17 Table 3.1: Different categories of resilient infrastructure planning 23 Table 4.1: Key industry standards on digital infrastructure resilience and disaster risk reduction 27 Table 4.2: Performance requirements 28 Table 4.3: Asset specifications for risk reduction 30 Table 4.4: Key Performance Indicators (KPIs) for resilience 36 Table 4.5: Risk allocation matrix template 38 Table 5.1: Solar resource and PV power potential maps for Kenya and Madagascar 42 Table 5.2: Calculations of emission and cost reduction from solar installed at 120 tower sites in Madagascar 43 Table 5.3: International standards relevant to greening telecom 45 Tabel 5.4: Sustainability indicators 46 Table 6.1: Procurement approaches 48 Table H.1. Summary of resilience and mitigation objectives included in project appraisal documents (PADs) 100 Boxes Box 2.1. Risk analysis approach 16 Box 3.1: Key advantages of resilient digital infrastructure 21 Box 3.2. Recent technological advancements in resilience 23 Box 3.3: Kenya and Somalia ICT policy frameworks with resilience objectives 25 Box 4.1: Country ICT standards - Kenya 28 Box 4.2: Enhancing redundancy cost-benefits 39 Box 4.3: Risk allocation requirements in Kenya PPP Contract 40 Box 5.1: Kenya, Solar energy for base stations 41 Box 5.2: Technological advancement in energy savings for digital infrastructure 44 6 Abbreviations, Acronyms and Definitions AU African Union DECIM Digital and Energy Connectivity for Inclusion in Madagascar project EAC East African Community EARDIP Eastern Africa Regional Digital Integration Project IGAD Intergovernmental Authority on Development KDEAP Kenya Digital Economy Acceleration Project MNO’s Mobile network operators TORs Terms of reference RFPs Request for proposals HoA Horn of Africa RCP Representative Concentration Pathway PPIAF Public Private Infrastructure Advisory Facility Adaptation The process of adjustment to actual or expected climate and natural hazards and its effects. Green digital transformation refers to the process of reducing the environmental and climate impacts of the growing digital sector, while at the same time enhancing its resilience to climate-related and natural hazards. This transformation also involves harnessing digital technologies and data to address the wide range of challenges posed by climate change, natural disasters, and other environmental shifts. Green digital On one side, this means implementing measures such as improving energy efficiency, transitioning transformation to renewable energy, adopting sustainable materials, managing e-waste responsibly, strengthening infrastructure, and ensuring redundancy in networks and data backups. On the other side, it entails leveraging digital tools for solutions such as early warning systems, digital financial platforms, and other innovations that enhance disaster preparedness and response capacity. Climate hazards are physical events or processes-such as extreme weather, shifting rainfall, or oceanic changes-that can harm people, property, infrastructure, and services. These hazards may be chronic (gradual temperature or precipitation changes) or acute (floods, cyclones, and storms). They can have Climate hazards direct effects (for example, cable cuts) or indirect impacts (financial costs from loss of connectivity). Hazards are shaped by natural factors as well as human activities like urbanization, environmental degradation, and greenhouse gas emissions, which can increase their frequency and severity. An operation’s exposure to relevant climate hazards is based on two main factors: (i) whether the operation is in a location and setting where (directly or indirectly) the relevant climate hazards are Exposure expected to occur, and (ii) whether the assets, systems, beneficiaries and/or vulnerable groups might be exposed to these hazards The potential for consequences from climate and natural hazards where something of value is at stake and the outcome is uncertain. Often represented as the probability that a hazardous event or trend Risk occurs multiplied by the expected impact. Risk results from the interaction of vulnerability, exposure, and hazard. Residual risk from climate hazards is the risk that remains following the integration of climate risk reduction (adaptation) measures Probabilistic risk analysis uses scientific evidence to simulate future disasters that are likely to occur. This method helps to overcome the limitations of relying solely on historical data. Exposure can be calculated to determine if the infrastructure route, site, or asset is situated in a hazard-prone area. Risk analysis In addition, data on the severity of the hazard (e.g. flood depth or wind speed) and a corresponding vulnerability model can be utilized to calculate the potential impacts (or damage costs) on exposed assets. 7 Improving the availability and fault tolerance of a system or service by duplicating one or more components of the system. Network redundancy focuses on diversifying the network routes to Redundancy eliminate single points of failure and dead ends. By creating alternative paths for data transmission, MNOs can minimize the impact of disruptions caused by cable cuts, equipment failures, or hazards. The ability of telecom infrastructure and network operators to provide and maintain an acceptable level of service in the face of various faults and challenges to normal operation. Resilience is provided Resilience by having a set of defenses (including redundancy) that reduce the probability of a fault leading to a failure and reduce the impact of an adverse event on network service delivery. Once an operation’s exposure to relevant climate hazards is known, their impact on each activity financed by the operation must be systematically assessed. Sensitivity is the degree to which an asset, system or species may be affected (either adversely or beneficially) when exposed to climate hazards. Vulnerability After the impacts of relevant climate hazards on the activities being financed through the operation have been assessed, their vulnerability to these hazards can be determined by considering their ability to cope with these impacts (i.e., adaptive capacity). 8 Chapters 9 Introduction 1 World Bank digital transformation projects play a crucial Kenya - awareness is growing. Countries are increasingly role in Eastern Africa countries by expanding connectivity aligning digital strategies with climate goals, supported to underserved areas and driving economic growth and by World Bank projects, as well as regional bodies like social development. In the six countries of Kenya, Djibouti, the East African Community (EAC), Intergovernmental Ethiopia, South Sudan, Somalia, and Madagascar, initiatives Authority on Development (IGAD), and African Union (AU) such as the Kenya Digital Economy Acceleration Project Commission, which are developing frameworks to green (KDEAP), the Eastern Africa Regional Digital Integration digital infrastructure and harmonize standards based on Project (EARDIP), and the Digital and Energy Connectivity for best practices Inclusion in Madagascar Project (DECIM) are instrumental in advancing digital transformation. These projects facilitate The three World Bank projects KDEAP, EARDIP 1 and significant investments in digital infrastructure and provide DECIM incorporate various greening initiatives aimed essential technical assistance, thereby improving access to at reducing climate risks, improving energy efficiency, affordable and reliable digital services across the region. and promoting renewable energy use. These greening They also aim to bridge the digital divide between urban and objectives are embedded in both project objectives, goals rural areas, enhance cross-border integration, and empower and implementation strategies, reflecting a shift toward communities to participate fully in the digital economy sustainable digital development. Regional and continental organizations are fostering cross-country collaboration Incorporating environmental sustainability into digital to ensure digital infrastructure investments contribute to transformation is essential to protect investments, climate goals while expanding digital access. This integrated optimize their impact, and reduce climate footprints. This approach is critical as Eastern Africa countries strive to involves integrating “greening” considerations related to accelerate digital inclusion while addressing urgent climate climate resilience and energy consumption into the design, challenges. The table below summarizes climate adaptation deployment, and operation of digital infrastructure. While (resilience) and mitigation objectives for the projects Kenya these aspects have historically been underprioritized (KDEAP), Somalia/South Sudan (EARDIP), Ethiopia/Djibouti and underfunded in the region - with national climate (EARDIP II), and Madagascar (DECIM). A more detailed policies rarely addressing digital infrastructure except in table can be found in annex H World Bank digital transformation projects Country Project KDEAP EARDIP 1 EARDIP II DECIM (Kenya) (Somalia & South Sudan) (Ethiopia & Djibouti) (Madagascar) 10 Table 1.1: Greening considerations for climate adaptation and mitigation in the three projects Project Climate Adaptation / Resilience Climate Mitigation - Climate risk surveys and resilience recommendations for - Energy-efficient digital infrastructure (fiber optics) infrastructure - Use of solar/wind for towers and schools KDEAP - Contractor ToRs require site-specific risk assessments - Policies for minimizing ICT environmental impact (Kenya) - Policy development for climate-resilient ICT sector - Support for e-waste management - Digitizing public services for continuity during shocks - Climate-informed network design and technical specs - Energy efficiency in network design and tenders EARDIP - Embedding climate-proofing in infrastructure - Use of energy-efficient fiber optics (Somalia & - Regional/national guidelines for resilience - Parallel deployment with other infrastructure to South Sudan) - Resilient data centers and hybrid hosting solutions reduce emissions - Guidelines for greening digital infrastructure - Climate-resilient cross-border links and backbone - Replace copper with energy-efficient fiber optics - Feasibility studies considering climate risks - Solar-powered cell towers EARDIP II - Redundant network design (e.g., buried cables) - Energy-efficient IT equipment (Ethiopia & - Training on digital tools for adaptation - Procurement of low-emission network equipment Djibouti) - Climate emergency response facilities - Guidelines for GHG reduction and e-waste management - Feasibility studies for climate risks and mitigation - Solar power and batteries for towers and DECIM - Monitoring for climate-resilience compliance communities (Madagascar) - Climate-resilient design specs in tenders - Energy-efficiency guidelines in tenders - Shift to energy-efficient network technologies The project design documents clearly incorporate multiple greening objectives across various initiatives, as illustrated in the preceding table. The following table summarizes these greening elements by digital asset category, as well as categories policies, regulations, and capacity-building efforts. To move from project design to implementation, these greening objectives must be systematically integrated into relevant policies, guidelines, and technical specifications. This integration should also extend to procurement frameworks, including terms of reference (TORs), requests for proposals (RFPs), bidding documents, and contracts. Table 1.2. Categorized summary of resilience and mitigation objectives included in project appraisal documents (PADs). Asset Type Adaptation/Resilience Efforts Mitigation Efforts Projects Towers / Site risk surveys, resilient design, robust Solar/wind power, energy-efficient EARDIP I, EARDIP II, Cell Sites construction, solar/battery backup operation DECIM Linear infra Climate-resilient routing, buried vs. aerial, Replace copper with fiber, energy- KDEAP, EARDIP I, (Cables) parallel deployment, climate risk assessment efficient deployment, dig-once principle EARDIP II Data Resilient design, disaster recovery, climate- Energy efficiency, renewable energy, KDEAP, EARDIP II Centers safe backup green ICT standards Equipment Robust equipment specs, redundancy, Energy-efficient hardware, e-waste KDEAP, EARDIP I, & Digital EARDIP II climate-resilient procurement management Assets Policies & Greening strategies, energy/resource Climate-informed policies, resilience KDEAP, EARDIP I, Capacity efficiency standards, e-waste protocols, EARDIP II, DECIM guidelines, adaptation training Building mitigation training This report offers guidance on the technical aspects of greening to facilitate effective implementation. It covers risk analysis related to climate exposure and vulnerability, resilience-building strategies, greening measures and standards, as well as recommendations for green procurement. Drawing on a geospatial analysis that overlays hazard risk data with digital infrastructure across six countries, the report provides valuable insights into climate and natural hazard risks in the region. It also summarizes essential measures, standards, and metrics designed to enhance resilience and mitigate disaster risks. Additionally, the report highlights various greening interventions aimed at improving energy efficiency and increasing the use of renewable energy sources. These initiatives not only help reduce operational costs but also lower the climate footprint of digital infrastructure in the region. Finally, the report concludes with practical recommendations for implementing green procurement practices. 11 Key take aways The impact of climate change on digital infrastructure in the six Eastern Africa countries is an escalating concern that demands targeted resilience and disaster management strategies. Understanding the specific hazard geography of each country and its subregions enables tailored responses to distinct climate risks. Increasingly severe climate events, such as floods, landslides, and storms, pose significant threats to the integrity and functionality of digital connectivity and data systems. Fiber optic cables buried underground are vulnerable to flood damage, while cell towers face threats from landslides and strong winds. The financial impact of severe flood events varies considerably across countries. For example, Djibouti faces relatively low damage costs to fiber networks (around $0.04 million) but incurs higher costs for cell site damages ($2.9–$3.3 million). In contrast, Kenya faces the highest potential losses, with estimated damages to cell sites ranging from $57.5 million to $79.2 million and fiber networks between $4.8 million and $11.2 million. This disparity also reflects differences in infrastructure scale and capacity: Kenya’s extensive network leads to higher absolute costs, but the relative risk per site may be lower compared to smaller countries such as Djibouti, Somalia, and Madagascar. Overall, floods, landslides, and storms impact digital infrastructure differently depending on the physical characteristics and vulnerabilities inherent to each type of infrastructure. The analysis highlights the critical need to enhance resilience and broader greening efforts for digital infrastructure throughout the region. As these countries expand their digital capabilities, prioritizing investments in climate-resilient infrastructure is essential to safeguard connectivity and data services against climate change impacts. This requires a multi-faceted approach, encompassing technical upgrades, capacity building and emergency planning. Furthermore, improving energy efficiency and accelerating the transition to renewable energy sources present significant opportunities, not only to enhance resilience but also to reduce operational costs and decrease the sector’s carbon footprint. Greening digital infrastructure delivers multiple benefits: it minimizes potential service interruptions, limits physical damage and energy expenses, and aligns infrastructure investments with national sustainability and climate objectives. Private sector engagement is critical to achieving both project-level and national objectives in greening the digital sector. Mobilizing private capital and leveraging private expertise are key to effective implementation, as private actors often possess the operational capacity to manage infrastructure projects efficiently. However, green investments typically require higher upfront and maintenance costs, which can deter bidders despite the long-term operational savings and reduced disruption costs. To overcome these financial barriers, the report recommends designing financial and contractual incentives that encourage private sector participation. Effective strategies include targeted subsidy schemes, private ownership, risk-sharing mechanisms, and economic support from public partners. Such approaches foster collaboration and mitigate the initial cost challenges that may hinder participation in greening efforts, enabling more widespread adoption of climate-resilient and sustainable digital infrastructure. 12 Risk Analysis 2 The challenges to infrastructure resilience in East Africa This situation underscores the urgent need for resilient are diverse, given that service disruptions have several digital infrastructure strategies to safeguard connectivity underlying causes. Common causes of communication and support sustainable development in the region. Failure disruptions across the six countries include hardware failure, to ensure stable and reliable digital connectivity undermines software bugs, power cuts, faulty software modifications/ business continuity, impedes disaster response and recovery, updates, and cable cuts. System failures and power outages and threatens socio-economic stability and community are the leading underlying causes of disruptions, followed by resilience in Eastern Africa countries, which already face damage from malicious activities, human errors, and natural numerous development challenges. events like heavy rains, floods, and storms. The impact of climate and natural hazards varies across countries, but Enhancing the resilience of digital connectivity and data recent severe floods and storms have clearly exposed the infrastructure to withstand climate-related and natural vulnerability of communication networks throughout the hazards, and other risks should be a top priority. To achieve region. this, it is crucial to understand the specific risks involved. East Africa is increasingly exposed to intensified The climate risk analysis conducted provides valuable climate risks, including acute and chronic hazards such insights by quantifying the exposure of mobile cells and as floods, droughts, and storms. These events expose fiber to flood and landslide events under various climate the vulnerabilities of digital infrastructure, including scenarios. Additionally, it assesses the risk levels and fiber cables, telecom towers, and data centers, and their potential impacts from tropical storms, coastal flooding, and supporting systems like power supply networks and access earthquakes, thereby informing strategies to mitigate these roads. Such disruptions frequently result in connectivity threats effectively. The primary objectives of the climate loss, adversely affecting economic activities and essential hazard risk analysis for digital infrastructure in six Eastern services, including emergency communications. The Africa countries—Djibouti, Ethiopia, Kenya, Madagascar, growing interdependence and -connectedness of various Somalia, and South Sudan—are as follows: systems via digital infrastructure implies that damage and disruption to the network will have cascading effects on • Highlight the use of geospatial hazard and infrastructure services such as power, financial transactions, transport, and data to enhance effective planning and infrastructure livelihoods. Digital infrastructure is crucial not only for daily development tailored to local contexts. operations but also for post-disaster response, recovery, • Identify and evaluate the risk and vulnerabilities and reconstruction efforts. When physical infrastructure of digital infrastructure to various climate hazards, is compromised due to climate-related events or natural including floods, landslides, and cyclones. disasters, the impacts can be felt by individuals, businesses, • Evaluate the potential costs associated with climate and governments alike. These impacts can manifest directly change impacts on digital infrastructure assets. through asset loss and disconnection from vital services, • Encourage the prioritization and implementation of affecting government agencies and first responders. resilience measures to mitigate damage risks. For instance, in Kenya, flash floods are known to cause The following pages provide a summary of the risk analysis cables cuts, leading to prolonged service interruptions in results. For more detailed information on each country, see rural areas that affect economic opportunities and access Annex A (Djibouti), Annex B (Ethiopia), Annex C ( Kenya), to critical services - see photo documentation below (fig Annex D (Madagascar), Annex E (Somalia) and Annex F 1). With global temperatures rising, these climate-related (South Sudan) risks are projected to worsen, compounded by the ongoing expansion of digital infrastructure and the growing reliance on its services. 13 Fig 2.1: Examples of underground fiberoptic cable exposure to climate hazard impacts Note: The pictures demonstrate the impacts of flood and erosion on underground fiber cables in different parts of Kenya. Powerful riverine and surface floodwaters can wash away protective layers of soil, concrete, and stone dams (gabions), leaving the cables vulnerable to exposure and physical damage. With sudden and significant rainfall, water is quickly funneled through rivers and down slopes, exacerbating the risk. Source of images: Anonymous. The risk analysis assesses overall hazard-related risks to digital infrastructure, specifically telecom cell sites and fiber networks-across six Eastern Africa countries. Key hazards include floods, landslides, cyclones/storms, coastal floods, and earthquakes. The findings, summarized in two tables, show risk levels categorized as low, medium, or high (Table 1), and detailed exposure percentages of fiber cables and cell sites to floods and landslides (Table 2). Where quantitative data is lacking, risks are qualitatively estimated. Results indicate varied but significant risk levels across countries and infrastructure types. This analysis focuses on potential damage and does not account for infrastructure resilience or local disaster response capacity, though lower asset exposure may reflect strategic site and route planning. Table 2.1: Summary of hazards risk levels in six Eastern Africa countries Hazards risks to current and planned digital infrastructure Country Flood Landslide Cyclone/Storm Coastal flood Earthquake Djibouti High Moderate Moderate Low-moderate Low-moderate Ethiopia Moderate Moderate Low N/A Low-moderate Kenya Moderate Moderate Low-moderate Low-moderate Low-moderate Somalia High Low Moderate Low-moderate Low South Sudan High Low Low N/A Low Madagascar Moderate Moderate High Low-moderate Low 14 Table 2.2: Summary of country specific hazard risks to cell sites and fiber networks Risk Description Estimated Damage Cost Quantitative (%) or qualitative risk Country Hazard Type USD million result Cell sites Fiber network Cell sites Fiber network Flood 37.5 % 13.7 % 2.9-3.3 0.04 Djibouti Landslide 4.3 % 16.8 % - - Cell sites total = 725 Cyclone/ storms Moderate Low - - Fiber network Coastal flood Low-moderate Low-moderate - - total = 401 km Earthquake Low-moderate Low-moderate - Ethiopia Flood 1.1 % 4.6-5.3 % 1.7 - 2.4 1.8 - 4.0 Cell sites Landslide 8.8 % 15.9 % - - total = 12,702 Cyclones/storms Low Low - - Fiber network total = 12,292 Earthquake Low-moderate Low-moderate - - km Kenya Flood 7.9 - 8.2 % 5.6 - 6.5 % 57.5 - 79.2 4.8 - 11.2 Cell sites Landslide 2.2 % 4.7 % - - total = 80,325 Cyclones/storms Low-moderate Low - - Fiber network Coastal flood Low Low-moderate - - total = 27,890 km Earthquake Low-moderate Low-moderate - - Somalia Flood 21.1 % 17 – 19.9 % 68.2 - 92.3 3.5 -7.3 Cell sites Landslide 0.5 % 3.2 % - - total = 5,994 Cyclone/storms Moderate Low - - Fiber network Coastal flood Low Low - - total = 2,933 km Earthquake Low Low - - South Sudan Flood 9.8 - 16.4 % 11.6 - 15 % 10.9 - 24.6 1.1 - 2.8 Cell sites Landslide 0.1 % 1% - - total = 5.994 Cyclones/storms Low Low - - Fiber network total = 2.943 Earthquake Low Low - - km Madagascar Flood 4 - 5.5 % 8.1 - 9.4 % 20.5-21.6 3.18-3.34 Cell sites Landslide 10.4 % 14 % - - total = 25,265 Cyclone/storms High High - - Fiber network Coastal flood Low-moderate Low-moderate - - total = 7,185 km Earthquake Low Low - - Note: The table shows the level of risk for both cell sites and fiber network to various hazards in different countries. The risk is presented both quantitatively, as the percentage of assets located in hazard prone areas, and qualitatively, indicating whether the overall risk for assets can be considered high, moderate, low, or none. Landslide risk description column include proportion of assets located in medium and high-risk zones. Only estimated damage costs for floods have been provided, as calculations for the other hazards are subject to too much uncertainty. Floods are extensively monitored, with records of extents, depths, damages, and detailed hazard maps, enabling reliable risk models and damage cost estimates for exposed infrastructure. Data for other hazards are sparse, inconsistent, and rarely cover digital assets, forcing reliance on assumptions and increasing uncertainty in cost estimates. Furthermore, the cost estimates does not reflect the full replacement cost, nor indirect costs (see box 2.1). 15 Box 2.1. The cost of internet disruption In East Africa, internet shutdowns and outages significantly disrupt essential digital activities, including mobile payments, online trade, and critical government e-services. The increasing interdependence of modern systems through shared digital infrastructure means that disruptions to network connectivity can trigger cascading effects across vital sectors such as power supply, financial transactions, transportation, and overall livelihoods. Although comprehensive data on the full scale of economic losses from such outages is limited, estimates are provided by tools like the NetBlocks Cost of Shutdown Tool (COST), which draws on data from the World Bank, ITU, Eurostat, and the U.S. Census Bureau. For example, the tool estimates that a nationwide internet shutdown in Kenya could result in GDP losses of approximately 1.2 million USD for every two hours of downtime. Digital infrastructure plays a vital role not only in sustaining daily operations but also in supporting post-disaster response, recovery, and reconstruction. When physical infrastructure is damaged by climate-related events or natural disasters, the ripple effects extend to individuals, businesses, and government agencies. These impacts can occur directly—through asset losses and severed access to essential services—and can hinder the ability of government agencies and first responders to act effectively. 2.1 Floods Floods are the most significant hazard across all countries, In summary, floods threaten fiber networks and nodes, presenting risk levels ranging from medium to high. Table especially those that are buried underground. In addition 2 includes quantitative measures of the proportion of fiber to riverine floods, which are included in the risk calculation and cell site assets that are located in flood-prone areas and and shown on the country-specific maps later, surface floods at risk during moderate to severe flood events. According - also known as flash floods -pose a significant risk that is to the data, approximately 5% of fiber networks in Kenya not accounted for in the open-source flood hazard data. The and Ethiopia are at risk of being impacted by floods. The damage caused by flash floods to digital infrastructure, such percentage is much higher in other countries, with Somalia as fiber optic cables, and supporting infrastructure (such as having the highest risk of 19.9% when including planned fiber roads, power installations, and other structures) is sparsely routes. In contrast, only about 1% of cell sites in Ethiopia face documented. However, conversations with stakeholders a risk of flooding, while Madagascar and Kenya have slightly in Kenya revealed that flash floods lead to frequent breaks more at 5-8%. The number is significantly higher in Somalia, in fiber optic cables. These risks will only increase across South Sudan, and Djibouti compared to other countries. all countries as precipitation events become more severe due to increased temperatures, and therefore additional The estimated direct damage costs have been calculated assessment and protection should be carried out in areas for flood hazards. These vary significantly based on the that are prone to flash floods. country and infrastructure type. Djibouti shows minimal damage costs for fiber networks ($0.04 million) but higher potential costs for cell sites ($2.9-$3.3 million). In contrast, Kenya has the highest estimated damage costs with $57.5- $79.2 million for cell sites and $4.8-$11.2 million for fiber networks. Other countries like Somalia and South Sudan have moderate damage estimates ranging from $1.1-$2.8 million. The variation in estimated damage costs reflects both the scale of infrastructure and the specific vulnerabilities each country faces. The infrastructure capacity across the HoA region varies greatly. Djibouti has a total of 725 cell sites and 401 km of fiber, while Kenya leads with 80,325 cell sites and 27,890 km of fiber. This disparity in infrastructure suggests that while Kenya may face higher absolute risk costs due to its extensive network, the relative risk per site may be lower compared to smaller nations like Djibouti or Somalia. 16 Table 2.3: Flood and landslide risk to current and planned fiber optic cable routes and cell sites ETHIOPIA Flood Exposure Landslide Risks KENYA SOMALIA 17 SOUTH SUDAN DJIBOUTI MADAGASCAR Note: The maps on the left depict a high-probability (1/25-year) riverine flood risk event under a ‘best-case’ climate scenario (RCP 4.5), whereas the maps on the right illustrate landslide risk exposure for areas classified as medium and high risk. Cell site locations are crowdsourced data from OpenCellID. Fiber routes in orange include live/current fiber and planned fiber – see details for fiber in supply maps in annexes. 18 2.2 Landslides Landslides can also be a major hazard, with varying levels is expected to intensify tropical storms, with a greater of risk depending on the terrain. The greatest landslide risks likelihood of storms reaching cyclone wind speeds. It is are found in mountainous regions and valleys, especially for crucial to closely monitor available data on wind speeds, infrastructure assets (cables, masts and structures) situated as well as climate changes and projections for the region, on or below steep slopes. If a landslide occurs in these areas, particularly when selecting sites and designing digital it can result in significant damage due to the powerful forces infrastructure near coastal areas. This consideration is involved. essential for mitigating risks associated with storms and high • In Djibouti, the exposure of telecom assets to landslide winds. As discussed in the following chapter, implementing risk is relatively low at 4.3% for cell sites and 16.8% for various resilience measures can help reduce these risks. fiber networks, indicating a moderate concern for fiber infrastructure compared to cell sites. Coastal flooding is primarily driven by a combination of • In Ethiopia, the exposure to landslide risks is higher, storms and high tides. However, the risk is exacerbated by with 8.8% for cell sites and 15.9% for fiber networks, rising global temperatures, leading to more intense storms suggesting significant exposure and potential damage and higher sea levels.3. As shown in annexes, the geospatial to both types of infrastructure. data overlays of cell sites and fiber networks with coastal • Madagascar also exhibits a notable landslide risk flooding hazard information reveals that the risk is generally of 10.4% for cell sites and 14% for fiber networks, classified as low to moderate across six countries. This indicating a considerable risk. suggests that only a small portion of fiber network and cell • The exposure to landslide risk in Kenya is lower at 2.2% site assets are situated in areas vulnerable to coastal flooding. for cell sites and 4.7% for fiber networks, indicating Despite the relatively low risk levels, significant disruptions a lesser immediate threat compared to Ethiopia and can still occur, particularly in coastal regions hosting critical Madagascar. infrastructure such as submarine cable landing stations. • In Somalia and South Sudan, the exposure to landslide Risks are considered higher in isolated coastal areas with risk is minimal, with values of 0.1 to 0.5% for cell sites high exposure and a high density of telecommunications and 1 to 3.2% for fiber networks, highlighting a relatively assets. Cities like Djibouti City (Djibouti), Mogadishu and safe environment concerning landslides. Berbera (Somalia), Toliara and Mahajanga (Madagascar), and Mombasa (Kenya) may face increased risks, especially in the long term. 2.3 Cyclones, storms and coastal flooding Data availability challenges is an issue when determining risks of infrastructure assets to storms and coastal flooding. Tropical storms and cyclones are generally classified as low Determining the risks of infrastructure assets to storms and to moderate risk across the listed countries, suggesting coastal flooding is hindered by a lack of comprehensive data. a reduced likelihood of these hazards causing extensive The current cyclone risk assessment relies on open-source damage. However, Madagascar and, to a lesser extent, data that provides historical cyclone tracks. Consequently, Somalia stand out as exceptions. Madagascar’s eastern this assessment reflects past exposure rather than and southwestern coastal regions are frequently impacted accurately predicting current and future risks. To achieve a by tropical storms and cyclones, with wind gusts often more precise risk evaluation, it is essential to incorporate exceeding 150 km/h.1 These events cause substantial additional detailed storm data. This includes assessing local damage to infrastructure, including prolonged power maximum wind speeds and analyzing historical and projected outages and disruptions to communication systems. Somalia, storm return periods under various climate scenarios. While while less exposed than Madagascar, experiences cyclones such data has recently become available globally, it requires or their remnants in its northern regions such as Puntland adaptation to specific country-level contexts for effective and Somaliland.2 Cyclones not only bring destructive winds application. Although some data on coastal flooding has been but also heavy rainfall due to their low-pressure systems. utilized, its quality is insufficient for effective downscaling or This rainfall can lead to surface flooding, river flooding, and quantifying current and future risks. Furthermore, existing landslides in vulnerable areas. The resulting impacts often data do not adequately account for the combined impacts of include communication blackouts and service interruptions, rising sea levels and storm events. exacerbated by extended power failures. Climate change 1 Acaps 2024, Madagascar cyclone exposure and vulnerabilities. Source: https://www.acaps.org/fileadmin/Data_Product/Main_media/20241019_ACAPS_Thematic_Mad- agascar-cyclone_exposure_and_vulnerabilities.pdf 2 ICPAC (2023), Cyclone Displacement Risk Profile (2023). Source: https://environmentalmigration.iom.int/sites/g/files/tmzbdl1411/files/documents/2023-04/IGAD_CY- CLONE_RISKASSESSMENT_SOMALIA.pdf 3 Over the past thirty years, East Africa has experienced a relative sea level rise of approximately 3.5 mm per year, slightly surpassing the global average. This trend is anticipated to continue in the coming decades, resulting in an increased risk of more frequent and severe coastal flooding and erosion, especially during high tides and storms, and particularly in low-lying areas and along sandy coastlines. The confidence in this projection is robust, highlighting the urgent need for enhanced data collection and risk assessment strategies. While the immediate risk to critical infrastructure assets, such as landing stations located in coastal regions, is not expected to escalate significantly in the near future, long-term projections—especially towards the end of the century—indicate that the combined impacts of erosion, storms, and rising sea levels may pose substantial challenges. Source of information: https://climateknowledgeportal.worldbank.org 19 2.4 Other hazards The following sections outline recommendations for resilience measures based on the available high-level In addition to acute hazards, prolonged exposure to high assessment of risks to digital infrastructure. Given the temperatures4, drought, dust, and humidity can have findings of this assessment and its inherent limitations, it is long-term impacts on digital infrastructure. This can lead essential to further integrate climate hazard risk analysis to accelerated degradation of electronic components and into infrastructure planning. To promote this, the following cooling systems, resulting in more frequent failures and recommendations are proposed: shorter operational lifespans. Extreme heat and drought further increase risks by boosting energy consumption • Mandate risk assessments: Establish policies, for cooling, reducing equipment efficiency, and causing procurement documents, and contracts that mandate infrastructure damage. Some data centers, which heavily the incorporation of climate risk assessments at every rely on water-based cooling systems, are being redesigned stage of digital infrastructure planning and development. to incorporate air cooling technology, which is becoming the • Improve data accessibility: Improve access to relevant standard for most new facilities.5 However, reducing water data to validate and clarify risk analyses. Enhanced consumption often necessitates increased electricity use for data availability will empower stakeholders to make cooling, which in turn encourages the development of on- informed decisions. site renewable energy sources. • Invest in geospatial data: Support the collection and analysis of geospatial hazard data. This investment will The risk of earthquakes impacting telecom infrastructure deepen the understanding of vulnerabilities in digital and digital connectivity varies across the Horn of Africa infrastructure and facilitate better decision-making (HoA) and Madagascar, influenced by geological factors, processes. historical seismic activity, and socio-economic conditions. Adhering to seismic building codes ensures that structures can withstand seismic events without collapsing.6 Installing bracing for equipment minimizes movement and prevents overturning during earthquakes. • The Horn of Africa is characterized by complex tectonic settings due to the Eastern Africa Rift System. This region experiences higher seismic risks compared to Madagascar, with active tectonic movements leading to occasional earthquakes, mostly along the Rift Valley regions of Djibouti, southern Eritrea, and northeastern Ethiopia. Despite significant earthquakes, they are relatively infrequent compared to other natural hazards like droughts and floods, which are more prevalent and damaging in the region. • Madagascar experiences moderate seismic activity, but the overall earthquake hazard risk is classified as low. Maps and models indicate that certain regions, particularly Toamasina, are more prone to losses from seismic events. The risk analysis has also identified tsunami risk zones, which adds another layer of concern for coastal areas.7 4 Annual mean near-surface temperatures are projected to increase by more than 1°C and 1.5° celcius over most of the Greater Horn of Africa. The highest temperature in- creases are expected in the northern region, including most of Somalia, Djibouti and northern parts of Ethiopia. The frequency and intensity of temperature extremes are expected to rise significantly. For instance, projections suggest that by 2100, average daily high temperatures will increase substantially under high-emission scenarios (RCP 8.5), Source: https://resilience.igad.int/resource/projected-climate-over-the-greater-horn-of-africa-under-1-5-c-and-2-c-global-warming/ 5 Data centers in Kenya and Ethiopia primarily use air-based cooling systems as the dominant cooling technique 6 such as the International Building Code (IBC) 7 https://downloads.openquake.org/countryprofiles/v2023.0.0/Africa/madagascar.pdf 20 Climate resilience of digital infrastructure 3 In digital infrastructure, resilience is the capacity of The resilience of a project design refers to how well its communication systems or assets to maintain or rapidly digital infrastructure assets incorporate climate and restore functionality after disruptions, whether natural disaster risks and risk mitigation measures. This includes or human induced. It involves avoiding shocks, managing structural robustness and adaptation measures, such as risks, adapting to change, and transforming systems that telecom towers designed to withstand local wind risks, limit current or future adaptability8. Since infrastructure and integrating these risks into the economic and financial exists to serve users, resilience is measured by its ability to analyses to ensure project viability. Resilient infrastructure continue meeting user needs during and after disruptive reduces life-cycle costs and increases confidence that events. While resilience is one of several factors defining investment outcomes will hold despite climate threats. high-quality digital infrastructure, embedding it in design Additionally, resilience can be reflected in project objectives and implementation enhances not only the management of that enhance the climate adaptability of sectors and natural shocks but also the overall cost-effectiveness and beneficiaries, for example, by improving early warning service quality of digital infrastructure investments. systems or emergency digital payments in flood- or drought- prone areas. Distinguishing between resilience in asset Disruptive events can affect any network segment, but design and resilience through project outcomes is key for there are some critical elements in which failure could effective monitoring of investment portfolios9. severely hamper communication services. These include elements in the backbone network, such as transmission links and switching nodes. Other core network components, including central offices and data centers that manage routing, traffic control, and end-user authentication, are critical because they handle massive data flows and ensure network reliability. Submarine cables and their landing stations, as well as cross-border terrestrial links, are crucial infrastructures for international connectivity. Box 3.1: Key advantages of resilient digital infrastructure • Reduced downtime costs: Enhanced resilience minimizes disruptions caused by natural disasters, ensuring continuous operation of digital services essential for businesses. This stability reduces economic losses associated with downtime, which can be substantial in sectors reliant on digital connectivity. • Mitigation of repair costs: By investing in resilience upfront, countries and operators can save on future repair costs associated with disaster recovery. For instance, resilient infrastructure can withstand extreme weather events, reducing the need for extensive repairs or replacements post-disaster. • Improved emergency response: Resilient digital infrastructure supports better emergency management systems. This capability enables quicker responses to disasters, ultimately saving lives and reducing recovery costs. Efficient communication networks ensure that critical information reaches the public and authorities swiftly. • Better access to services: With resilient infrastructure, governments can offer more reliable public services, such as healthcare and education, through digital platforms. This accessibility can lead to improved societal outcomes and economic productivity. • Attraction of investments: Countries with robust digital infrastructure are more attractive to investors. A resilient digital environment signals reliability and enhances confidence among potential investors, potentially leading to increased foreign direct investment and economic growth. • Sustainability and climate policy goals: Resilient digital infrastructure often aligns with broader global and country sustainability and climate policy goals. 8 This concept has been defined and refined in several World Bank reports, such as Hallegatte, Stephane; Rentschler, Jun; Rozenberg, Julie. 2019. Lifelines: The Resilient Infrastructure Opportunity. Source: https://hdl.handle.net/10986/31805 9 World Bank. 2024. Resilient Telecommunications Infrastructure - A Practitioner’s Guide. Source: https://hdl.handle.net/10986/42290 and World Bank Group. 2021. Resilience Rating System: A Methodology for Building and Tracking Resilience to Climate Change. Source: https://hdl.handle.net/10986/35039 21 3.1 Technical measures for network Reliable power supply is another critical factor, since power outages are a common cause of network failure14. Key resilience resilience measures include uninterruptible power supplies (UPS), backup generators, and diversified power lines, At a structural “system” level, digital infrastructure particularly at critical sites. Integrating renewable energy networks can be built with resilience in mind, using two key solutions, such as solar-powered base stations, offers principles: redundancy and diversity. Redundancy provides sustainable alternatives during power outages and fuel backup components, like duplicate switches and separate supply disruptions and enhances overall system reliability. communication links (physical paths), so if one fails, another can take over. Diversity, such as using different operators, The physical infrastructure, including buildings, towers, suppliers and technologies, reduces the risk of simultaneous and cell sites, must also withstand disruptions from climate failures caused by shared vulnerabilities. These principles and natural hazards. This requires strategic site selection, apply across all layers of the network. While diversity can structural reinforcement, and hazard-resistant design. be costly and complex, it significantly lowers the chance For instance, fiber cables can be buried at safe depths and of widespread outages. Advances in technology have waterproofed to prevent flood damage, while towers and further enabled the effective management of such complex facilities can be built to withstand earthquakes and strong network systems (see box 3.2). Overall, adopting a systemic winds. Although physical structures generally endure approach to resilience in digital infrastructure is preferable, tough weather conditions, sensitive electronic equipment even though it may not always be feasible in every context.10 needs additional protection through hardened enclosures, elevated foundations, and proactive monitoring and Layered architectures further enhance resilience by maintenance15. Detailed specifications for various assets combining terrestrial and non-terrestrial networks, such and hazards are provided in Table 4.3. as low-earth-orbit (LEO) satellites. These are especially valuable in disasters, providing emergency communication, Finally, road access to network sites is essential for backhaul, remote sensing, IoT, and broadband services when maintenance effort before or during disruptive incidents, terrestrial networks are damaged or congested11. Other and for recovery after disasters. Operators must proactively resilience strategies include network segmentation and the assess risks and coordinate logistics, sometimes partnering ability to isolate sections during disruptions until normal with government or military authorities to secure operations resume. Automated network management, emergency access. Strategically pre-positioning staff and including proactive monitoring and recovery, also helps spare parts near key sites also help ensure rapid restoration ensure uninterrupted service across all network layers.12 and recovery when disasters strike. Resilience of digital infrastructure in a specific country or region also depends on external infrastructure. Global managed services, including edge cloud and content delivery networks (CDNs), are increasingly essential, making redundancy and diversity in these services critical for robustness. Similarly, satellite connectivity, dominated by providers like Starlink, plays a growing role in market accessibility and control, underscoring the need for diversified external infrastructure to enhance overall resilience13. 10 World Bank. 2024. Resilient Telecommunications Infrastructure - A Practitioner’s Guide. Source: https://hdl.handle.net/10986/42290 11 Höyhtyä, M.; Anttonen, A.; Majanen, M.; Yastrebova-Castillo, A.; Varga, M.; Lodigiani, L.; Corici, M.; Zope, H. Multi-Layered Satellite Communications Systems for Ul- tra-High Availability and Resilience. Electronics 2024, 13, 1269. https://doi.org/10.3390/electronics13071269 12 Akinola, Oluwaseun Ibrahim, Oluwaseun Oladeji Olaniyi, Olumide Samuel Ogungbemi, Oluseun Babatunde Oladoyinbo, and Anthony Obulor Olisa. 2024. “Resilience and Recovery Mechanisms for Software-Defined Networking (SDN) and Cloud Networks”. Journal of Engineering Research and Reports 26 (8):112-34. https://doi. org/10.9734/jerr/2024/v26i81234. OECD (2023), “Enhancing the security of communication infrastructure”, OECD Digital Economy Papers, No. 358, OECD Publishing, Paris, https://doi.org/10.1787/bb608fe5-en. 13 OECD (2022), “Broadband networks of the future”, OECD Digital Economy Papers, No. 327, OECD Publishing, Paris, https://doi.org/10.1787/755e2d0c-en. 14 Cabrera-Tobar, A., Grimaccia, F., & Leva, S. (2023). Energy Resilience in Telecommunication Networks: A Comprehensive Review of Strategies and Challenges. Energies, 16(18), 6633. https://doi.org/10.3390/en16186633 15 World Bank. 2024. Resilient Telecommunications Infrastructure - A Practitioner’s Guide. Source: https://hdl.handle.net/10986/42290. See also OECD (2024), Infra- structure for a Climate-Resilient Future, OECD Publishing, Paris, https://doi.org/10.1787/a74a45b0-en. 22 Box 3.2. Recent technological advancements in resilience Several technological advances are enhancing network resilience, including increased network virtualization, cloud integration, network slicing, and the use of AI and machine learning.16  Virtualization, such as Software-Defined Networking (SDN), improves resilience by enabling dynamic traffic routing through software controllers or APIs, allowing networks to quickly adapt and reroute in case of failures.17  Network Function Virtualization (NFV) further boosts resilience by distributing network functions across multiple servers rather than relying on specific hardware.18  Cloud integration (cloud services) leverages SDN and NFV to enhance resilience by enabling workload relocation across different regions without data loss or service disruption.19  Network slicing enables the creation of reliable, virtual segments within mobile networks, allowing a single physical network to support diverse applications with varying capacity needs. While available in 4G, this feature is a key aspect of 5G standalone networks, enabling multiple “slices” tailored for different uses.20  AI and machine learning (ML) enhance network management by analyzing data from network elements and IoT sensors to detect issues early and automate network management. For example, AI can coordinate aerial platforms to deploy temporary base stations during terrestrial network failures. However, AI adoption requires skilled personnel and adds complexity.21  Mesh networks have increasingly become a foundational design principle in the industry, offering resilient, decentralized connectivity that dynamically self-heals and efficiently routes data. Operators may form commercial partnerships with competitors to develop robust mesh topologies that enhance network resilience and fault tolerance. The table below illustrates different categories of resilient infrastructure planning. While the table is not exhaustive, it underscores the need for a holistic approach to building resilient telecommunications systems. The different elements are further described in a World Bank resilience guidance note.22 Table 3.1: Different categories of resilient infrastructure planning Operations, Contingency planning Systems planning Engineering and design Institutional capacity maintenance Response/operational Network design Customized resilience Policy and regulation Resilient monitoring procedures to a and architecture measures incentivizing resilience and management particular hazard • Risk-informed • Site preparation, e.g. • Emergency call • Operational • National disaster design, routes, elevated foundation center. emergency response response plans sites, assets • Technology choice, • Coordination unit procedures (i.e. • ICT resiliency policies • Plan for e.g. underground for alarm handling, Disaster Recovery and standards redundancy8 fiber fault resolution, Plan) • Financial • Minimize dead • Better/stronger configuration • Procurement & incentivization, ends. materials. management, storage of spare parts subsidization • Spectrum • Asset protection & and proactive • Regular drills. • Critical infrastructure flexibility and reinforcement. performance • Backup power designation. Traffic overload • Asset relocation monitoring • Backup • Mandatory incident management • Modify surroundings • Network engineers communication reporting • Diversify to reduce risks standby to resolve • Backup storage of • Public disclosure over operators issues data hazard disruptions • Automated • Pre-positioning of • Collab across sectors detection of faults additional fuel and • Skills and capacity and threats. batteries and off- development • Weather/hazard shore cable repair monitoring ships 17 OECD (2022), “Broadband networks of the future”, OECD Digital Economy Papers, No. 327, OECD Publishing, Paris, https://doi.org/10.1787/755e2d0c-en. 17 Josbert, N. N., Ping, W., Wei, M., & Li, Y. (2021). Industrial Networks Driven by SDN Technology for Dynamic Fast Resilience. Information, 12(10), 420. https://doi. org/10.3390/info12100420 18 https://www.ericsson.com/en/nfv accessed on January 15th, 2025 19 Singh, R., Larsen, L. M. P., Ollora Zaballa, E., Berger, M. S., Kloch, C., & Dittmann, L. (2025). Enabling Green Cellular Networks: A Review and Proposal Leverag- ing Software-Defined Networking, Network Function Virtualization, and Cloud-Radio Access Network. Future Internet, 17(4), 161. https://doi.org/10.3390/ fi17040161 20 https://www.ericsson.com/en/network-slicing accessed on January 15th, 2025 21 El-Hajj, M. (2025). Enhancing Communication Networks in the New Era with Artificial Intelligence: Techniques, Applications, and Future Directions. Network, 5(1), 1. https://doi.org/10.3390/network5010001 22 World Bank. 2024. Resilient Telecommunications Infrastructure - A Practitioner’s Guide. © Washington, DC: World Bank. http://hdl.handle.net/10986/42290 23 3.2 Institutional capacity for resilience Scope of Policy Measures • Infrastructure enhancement: Upgrading technical Public policy and regulation play a crucial role in enhancing capabilities and building redundancy in communication network resilience to mitigate the impacts of disruptions. networks. This is recognized by the three World Bank projects, as • Disaster preparedness and response: Establishing illustrated by project objectives listed in table 1.1. Among the disaster plans, governance models, and business six Eastern Africa countries within scope, only Kenya and, to continuity protocols. a lesser extent, Somalia have developed policies promoting • Technological innovation: Promoting research and technical and organizational measures to manage risks and adoption of technologies that reduce vulnerabilities. network disruptions (see box 3.3). However, these policies • Emergency communications: Ensuring reliable have seen limited implementation and enforcement to date. communication services for responders and the public. • Reporting and transparency: Mandating timely Effective resilience demands coordinated efforts among disclosure of disruptions and recovery efforts. governments, the private sector, and civil society. Establishing comprehensive regulatory frameworks, Implementation Instruments guidelines, and standards is essential to guide all • Information and awareness: Raising stakeholder and stakeholders. Organizations must adopt proactive strategies public understanding of resilience, e.g. by promoting that prioritize prevention for both software and hardware, knowledge, best practices, and standards for robust alongside robust response and recovery plans. This dual communication infrastructure. focus ensures systems are safeguarded and can quickly • Facilitation: Encouraging collaboration among providers recover from disruptions. and utilities. Resilience also includes flexible spectrum and traffic • Financial incentives: Supporting infrastructure management, particularly during emergencies, by allocating upgrades and innovation through funding, grants, frequencies for public safety on land and satellite networks subsidies, or taxation schemes. where applicable.23 Training communication personnel, • Regulation: Enforcing standards and mandatory conducting emergency drills, and developing clear resilience requirements. These may include technical emergency communication and continuity plans are also requirements focusing on network design and vital.24 These efforts require identifying key stakeholders, infrastructure resilience, as well as outcome-based assessing risks, and establishing strong administrative and rules setting specific availability and recovery metrics. legislative frameworks. Building institutional capacity is fundamental to sustaining these resilience measures. To translate some of these considerations into practice, subsequent chapters will delve into specifying resilience There are multiple approaches to developing policy performance requirements that align with international frameworks that promote resilient digital infrastructure, standards and regional contexts. They will also explore as demonstrated by existing initiatives in Kenya and procurement strategies tailored to digital infrastructure, Somalia. While these frameworks must be tailored to emphasizing sustainable and resilient procurement each country’s unique context, several common elements practices that prioritize long-term value, adaptability, and are relevant across the Eastern Africa region and within risk mitigation. the scope of the three projects under consideration. A comprehensive policy framework can be organized around two principal dimensions: the scope of policy measures and the instruments employed for their implementation. 23 As also highlighted by GSMA, both regulators and operators should consider maintaining emergency telecommunications bandwidth to manage increased congestion and demand during emergency situations. Source: GSMA 2020, Building a Resilient Industry: How Mobile Network Operators Prepare for and Respond to Natural Disasters. 24 These plans should detail specific actions to be carried out before, during, and after a crisis, in order to maintain communication, ensure operational resilience, and facili- tate recovery. Aside from World Bank support, the International Telecommunication Union (ITU) also offers assistance in developing National Emergency Telecommuni- cations Plans (NETPs), which can greatly enhance a country's preparedness for emergencies. See https://digitalregulation.org/category/emergency-communications/ For details and guidelines on service continuity, there are several standards and guidelines targeted different types of organizations, including those developed by GSMA for telecom operators. See https://www.gsma.com/ 24 Box 3.3: Kenya and Somalia ICT policy frameworks with resilience objectives Kenya  Kenya National Digital Master Plan (2022–2032) emphasizes building a resilient and secure infrastructure that supports digital services and data management but does not provide specific directions how this will be implemented. The Kenya Cloud Policy 2024 complements this by mandating government agencies to prioritize cloud-based solutions that ensure scalability, centralized and redundant data storage, disaster recovery, and data backup, thus enhancing operational continuity.  Kenya’s ICT Authority (ICTA) enforces specific standards to strengthen infrastructure resilience and redundancy.25 These include Network Design and Management standards requiring resilient IT networks with redundancy and fail-over capabilities to maintain critical system operations; Data Centre Standards mandating full redundancy in network connections, power supply (including generators and UPS with at least 8-hour battery backup), and IT equipment. Information Security Standards obliging all Ministries, Counties, Departments, and Agencies (MCDA) to develop and maintain documented business continuity and disaster recovery plans that cover management structures, incident response teams, and procedures to sustain operations and information security during disruptions.  Additionally, the Communications Authority of Kenya (CA) has issued a guideline requiring network operators to submit incident reports and to implement redundancy, resilience, and diversity measures to prevent service downtime and ensure reliable ICT network availability. Somalia:  For Somalia, the National ICT Policy and Strategy (2019–2024) prioritizes extending national backbone infrastructure with redundant links connecting major urban centers and establishing a Government Data Center with offshore/cloud backups to protect critical government data and services. The ICT Regulatory Transformational Strategy and Roadmap (2023–2027) focuses on developing robust physical and network security protocols and rapid response capabilities to cyber and physical threats.  Somalia’s National EmergencyTelecommunications Plan (NETP) complements these policies by integrating telecommunications and ICT into the broader national disaster risk management (DRM) framework. The NETP provides a strategic blueprint (developed by ITU) that addresses all four phases of disaster management-mitigation, preparedness, response, and recovery- specifically for the telecom/ICT sector. 25 These measures are detailed primarily in ICTA standards ICTA.1.001:2016 (General Guiding Principles), ICTA.2.002:2019 (Data Centre Standard), ICTA.3.002:2019 (Information Security Standard) 25 Resilience standards and performance requirements 4 Climate resilience standards and performance requirements for digital infrastructure set minimum requirements for design, construction, and maintenance to protect against climate risks, for example, flood-proofing, material durability, or wind resistance. Standards can be adopted voluntarily by stakeholders without legal compulsion. However, regulators or contracting authorities may choose to mandate specific standards as legal requirements. A range of standards, guidelines and technical specifications have been established globally to enhance digital infrastructure resilience, including those developed by organizations such as the International Telecommunication Union (ITU), the International Organization for Standardization (ISO), as well as regional entities such as the International Electrotechnical Commission (IEC), and national ICT authorities and regulators. Table 4.1: Key industry standards on digital infrastructure resilience and disaster risk reduction Standard Description ISO • ISO 22301:2019: Security and resilience — Business continuity management systems — Requirements. Provides a framework to identify threats, mitigate risks, and maintain essential services during disruptions such as cyberattacks, natural disasters, or pandemics. It supports organizations in planning, implementing, monitoring, and improving business continuity to minimize impact and ensure rapid recovery. • ISO 22320:2018: Guidelines for emergency management and incident management to enhance security and resilience during emergencies. • ISO/IEC 27031:2011: Guidelines for ICT readiness for business continuity, focusing on ensuring technology systems support continuity objectives during disruptions. • ISO 21110:2019: Standards for emergency preparedness and response in information and documentation. • ISO 27001:2022: Global standard for Information Security Management Systems (ISMS), covering cybersecurity and privacy protection. Includes climate action amendment ISO/IEC 27001:2022/Amd 1:2024. • ISO/DIS 14091:2021: Adaptation to climate change — Guidelines on vulnerability, impacts and risk assessment. guidelines for assessing the risks related to the potential impacts of climate change • ISO/DIS 14097:2021: Greenhouse gas management and related activities — Framework including principles and requirements for assessing and reporting investments and financing activities related to climate change. • ISO 22372 (Security and Resilience – Resilient Infrastructure): This is a new and emerging ISO standard specifically focused on infrastructure resilience, aiming to provide comprehensive guidance on developing, implementing, monitoring, and improving infrastructure resilience. It addresses the interdependent nature of infrastructure systems and cascading disaster impacts, which is critical for digital infrastructure resilience. • ISO/IEC 20000-1:2018 (IT Service Management): This standard includes provisions for service continuity and availability management, key to ensuring IT service resilience and disaster recovery. IEC • IEC 60529, Degrees of protection provided by enclosures (IP Code) has been developed to rate and grade the resistance of enclosures of electric and electronic devices against the intrusion of dust and liquids. ETSI • ETSI ETS 300 019-1-3: Equipment Engineering (EE). Environmental conditions and environmental tests for telecommunications equipment standard. • ETSI TR 102 445 V1.2.1 (2023-04): Emergency Communications (EMTEL); Overview of Emergency Communications Network Resilience and Preparedness • ETSI TS 102 182 V1.5.1 (2020-07): Emergency Communications (EMTEL); Requirements for communications from authorities/organizations to individuals, groups or the general public during emergencies • ETSI TS 102 181 V1.3.1 (2020-06): Emergency Communications (EMTEL); Requirements for communication between authorities/organizations during emergencies 26 ITU ITU-T Study Group 5 on Environment, EMF and Circular Economy is responsible for developing methodologies to evaluate the impacts of ICT on climate change. This includes a long list of recommendations for construction, installation and protection of cables and other elements of outside plant (not listed here), and specific recommendations for climate adaptation and disaster management listed below: • Recommendation ITU L.1500: Framework for information and communication technologies (ICTs) and adaptation to the effects of climate change. • Recommendation ITU L.1502: Adapting information and communication technology infrastructure to the effects of climate change. Identifies direct and indirect threats of climate change on ICT services and provides options for adaptation and mitigation. • Recommendation ITU-T L.1506: Framework of climate change risk assessment for telecommunication and electrical facilities. It consists of risk assessment methodology and considerations for applying the defined methodology. ITU has released various guidelines and technical reports, including: • ITU-T Technical Report, 2022: Guide on the use of ITU-T L-series Recommendations related to optical technologies for outside plant • ITU-D, 2020: ITU Guidelines for national emergency telecommunications plan. The guidelines can be used for developing tailored contingency plans for emergencies caused by natural hazards, epidemics and pandemics. • ITU-T FG-DR&NRR, 2014: Requirements for Network Resilience and Recovery • ITU, 2014. Resilient pathways: the adaptation of the ICT sector to climate change. • ITU-T et al, 2020: Frontier Technologies to Protect the Environment and Tackle Climate Change. ANSI American National Standards Institute (ANSI) with telecom accredited standard organizations such as BICSI and ANSI has several relevant standards, including: • ANSI/BICSI 002-2024, a widely adopted standard for Data Center Design and Implementation Best Practices1 • ANSI/TIA-942, a widely adopted standard specifying the minimum requirements for data centers, covering all physical infrastructure. • ANSI/TIA-222-I-2023, a Structural Standard for Antenna Supporting Structures and Antennas, specifying required loads, material choice, anchoring etc. under different environmental settings. IEEE Institute of Electrical and Electronics Engineers (IEEE) offers a vast array of research papers, articles, and standards that delve into various aspects of network resilience, including methodologies for enhancing the robustness, reliability, and recovery capabilities of communication networks. ENISA ENISA (European Union Agency for Cybersecurity) offers many publications and tools designed to improve network resilience and cybersecurity within the EU. Its work includes research on best practices, methods, and standards for ensuring the resilience of communication networks. Uptime Uptime Institute Tier Standard provides certification data center resilience levels, including considerations for climate change impacts on data center operations. Codes Regulations governing design, deployment, and maintenance of telecom infrastructure, e.g., International Building Code (IBC) 2024 referencing TIA 222 for telecom towers, and national codes like Kenya's National Building Code 2024 mandating fiber optic installations in new constructions. The types of resilience standards can be categorized into several key areas, including those that focus on system performance, systems recovery, and service quality metrics. Based on the risk analysis findings, table 4.2 and table 4.3 provide a selection of resilience performance requirements and measures per asset type which can be applied across the three projects. It is advisable to include only the relevant resilience requirements in each contract, tailored to the project’s objectives, the network’s location, the nature of the requested services, and the climate and disaster risk assessments conducted during the project design phase. The contractor can then further refine and adapt these requirements based on a specific evaluation of the project site. This approach helps prevent overwhelming the private sector partner with unnecessary contractual obligations. 27 Table 4.2: Performance requirements Category Description Network • Redundant Connectivity: Establish resilient network connectivity through increased redundancy to eliminate reliability single points of failure, ensuring reliable communication and data transmission. • Power Backup Systems: Utilize uninterruptible power supplies (UPS), backup generators, or batteries to ensure continuous operation during outages or extreme weather disruptions. Climate • Temperature Tolerances: Equipment and infrastructure must be designed to operate effectively within specified and natural temperature ranges to maintain performance and durability. hazard • Water and Wind Resistance: Critical components should have water-resistant (or waterproof) or wind-resistant resilience designs to mitigate damage from flooding, heavy rainfall and strong winds. • Landslide Resistance: Implement engineering solutions such as retaining walls and slope stabilization techniques to protect infrastructure from landslides in areas with high or medium risk. • Earthquake Resistance: Design structures to meet seismic standards, incorporating flexible materials and shock absorbers to minimize damage during earthquakes. • Coastal Flooding Mitigation: Elevate critical infrastructure and employ flood barriers or levees in coastal areas to protect against rising sea levels and storm surges. Disaster • Operational Continuity Plans: Develop disaster preparedness and recovery strategies and operational plans preparedness to mitigate climate-related disruptions, this may include regular drills and data backups, emergency spectrum/ and recovery traffic management, stored and/or mobile equipment for quick repair or replacement, alternative operational sites • Monitoring and Early Warning Systems (EWS): Integrate systems for real-time monitoring of climate-related risks, including weather stations and sensor networks, to enable proactive risk management. When proposing standards and specifications at the national or regional level, it is essential to carefully review existing regulatory standards that govern specific infrastructure components or services, assessing their relevance and effectiveness. This review is only partially covered in this assignment and should be thoroughly addressed in country- or project-specific technical advisory services and feasibility studies, ideally conducted in collaboration with local ICT authorities and regulators. Compliance with local regulations and standards related to hazard risk management is critical to maintaining operational integrity. However, when binding national standards exist, imposing more stringent requirements may discourage potential bidders or increase the need for subsidies to meet these. An important exception to this principle occurs when additional requirements are necessary to meet particular project objectives and goals. Box 4.1: Country ICT standards - Kenya In Kenya, the deployment of public digital infrastructure, including fiber optic cables, telecom towers and data centers are governed by specific standards and requirements aimed at ensuring effective and reliable communication infrastructure. These regulations are outlined in various documents from government and regulatory bodies, including Kenya Communications Authority (CA)26 and ICT Authority (ICTA).27 • For instance, The Fiber Optic-Backbone, Metro and Last Mile Infrastructure Standard (ICTA 2.001:2021) sets minimum requirements for planning, design, deployment, operation, maintenance, and management of fiber optic infrastructure in Kenya. It mandates installation of fiber cables in approved ducting infrastructure to protect against physical damage, which supports resilience. The standard requires fiber cables to be buried at specific minimum depths (e.g., at least 1.5 meters below carriageway or road reserve surfaces) to reduce risk from environmental and human activities. It specifies that backbone fiber cables should have a minimum of 96 cores with two parallel cables (one as actual backbone and one as access cable), providing redundancy in the fiber network. The standard also emphasizes coordination with other utilities to avoid accidental damage during trenching and installation • The Data Centre Standard (ICTA 2.002:2019) includes requirements for physical infrastructure to enhance resilience and disaster preparedness. It specifies the use of watertight flexible metal conduits for branch circuits and solid metal conduits for feeder circuits to protect electrical wiring from water damage. Adequate separation and organization of power and telecommunications cabling are required to minimize risks of interference and damage. The standard mandates full network redundancy in data centers, including the use of two or more Internet Service Providers (ISPs) and pairs of IT equipment, ensuring continuous service availability during failures. Coordination with architects and engineers is required for cable tray planning to avoid conflicts with fire protection, lighting, and plumbing systems, reducing disaster risks 26 The Communications Authority of Kenya (CA) serves as the independent regulatory body for the Information and Communications Technology (ICT) sector in Kenya. Established in 1999, its primary responsibilities include: Licensing; Frequency Spectrum Management; Consumer Protection; E-commerce Development; Universal Service Fund Management; Compliance Monitoring: The CA monitors licensees to ensure adherence to regulations and standards. https://www.ca.go. ke/ 27 The ICT Authority (ICTA) was established in 2013 as a state corporation under the Ministry of Information, Communication, and Technology. Its key functions include: Management of Government ICT Functions; Setting ICT Standards; Promoting ICT Literacy; Facilitating E-Government Services; Commercialization of National ICT Infrastructure 28 • Environmental Considerations: Operators are required to conduct an Environmental Assessment for areas where communication infrastructure will be installed. This includes considerations for radio masts, towers, antennas, and other related equipment • Installation Guidelines: The CA Guidelines for Installation and Maintenance of External Communication Infrastructure specify that cables must conform to Kenyan specifications as per the Kenya Bureau of Standards (KEBS). This includes using approved materials and adhering to safety protocols during installation • Collocation and Sharing: The CA Code of Practice encourages operators to collocate or share facilities (e.g. towers) and ducts where feasible, promoting efficient use of resources and minimizing environmental impact. • Compliance and Enforcement: Both CA and ICTA conduct regular audits to ensure compliance with standards and guidelines. Entities found non-compliant will receive a report detailing deviations, with recommendations for remediation. • Licensing: Entities wishing to deploy fiber optic cables or telecom towers must obtain relevant licenses from the Communications Authority of Kenya (CA). When national standards are absent or lack sufficient detail, internationally recognized standards can serve as valuable references - see Table 4.1. If no such standards exist or are widely accepted, it is advisable to propose a set of functional specifications grounded in sector expertise. These specifications can then be incorporated into the procurement process, including the formulation of Terms of Reference (TORs), evaluation criteria for bidders, and key performance requirements within contracts. This approach helps ensure that infrastructure development remains robust and resilient, even in the absence of comprehensive local regulations. The following table summarizes asset specifications designed to reduce risks from specific hazards, focusing on common physical assets in connectivity and data infrastructure across East Africa. It is important to remember that the specific actions needed to enhance resilience will vary depending on the context, including: • The specific hazards present • The characteristics of existing infrastructure • The broader enabling environment 29 Table 4.3: Asset specifications for risk reduction Infrastructure Risk Resilience measure asset type Route Planning and Risk Assessment • Prioritize Risk Assessment: Before any installation, conduct a thorough risk assessment to identify areas prone to flooding or other environmental hazards. • Strategic Routing: Whenever possible, select cable routes that avoid high-risk flood zones. Cable Protection and Material Selection • Protective Conduits: Encase cables in durable conduits or protective tubing made from materials like PVC or polyethylene to provide a physical barrier against damage and water ingress. • Water-Resistant Materials: Utilize water-resistant cables and coatings specifically designed to withstand exposure to moisture and humidity. • Dedicated Conduits: Install cables in dedicated conduits to simplify the identification and repair of affected fiber lines in case of damage. • Joint Integrity: Ensure all cable joints are meticulously sealed using appropriate waterproofing techniques to prevent water from entering and causing corrosion or signal degradation. • In flood-prone locations, implement robust sealing measures such as concrete encasements when laying cables to create a watertight barrier. Burial Depth and Soil Considerations • Adequate Depth: Bury cables at a sufficient depth (typically 1 to 1.5 meters, but adjust based on local regulations and soil conditions) to protect them from erosion, exposure, and physical damage from surface activities. • Soil-Specific Techniques: Adapt trenching techniques to the specific soil type. Use appropriate shoring and stabilization methods to prevent cave-ins. Rocky soils may require specialized excavation equipment or additional protective layers around the cables. Deployment and Technology • Technology Choice: Carefully select cable types based on ease of deployment, damage resistance, Floods optical performance requirements, and compatibility with mid-span splicing techniques. • Manufacturer Data Integration: Combine manufacturer specifications with a detailed understanding of local environmental conditions to optimize material selection and deployment strategies. • Trenchless Technology: Whenever feasible, employ trenchless technologies like directional boring or air-blown fiber to minimize surface disruption, reduce environmental impact, and preserve soil integrity.12 Flood Mitigation Strategies Underground • Drainage Systems: Implement drainage systems strategically to divert water away from critical cables cable infrastructure. • Ground Stabilization: Use gabions (wire mesh baskets filled with rocks) to stabilize the ground in vulnerable areas and provide protection against erosion and washout. • Elevated Chambers: Construct fiber chambers slightly above ground level and ensure they are water-resistant to prevent water intrusion. Environmental Protection • Minimize Disruption: Use directional boring techniques to minimize surface disruption in sensitive areas, especially when crossing rivers and roads.13 • Soil and Vegetation Conservation: Preserve soil texture and existing vegetation to minimize soil erosion and maintain ecological balance. Post-Installation Monitoring and Maintenance • Regular Inspections: Conduct regular post-installation inspections to verify the integrity and functionality of the cable infrastructure, particularly after severe weather events or flooding. • Prompt Repairs: Address any identified issues promptly to prevent further damage and ensure the continued reliability of the network. In addition to the general principles outlined above: • Enhanced cable protection: Utilize reinforced/armored cables, rugged cable ducts, and secure anchoring systems to provide enhanced protection and durability for cables, especially in areas prone to landslides or other geological hazards. • Landslide risk management through activity control: Actively limit and monitor human activities 30 Landslide that increase landslide risks, such as deforestation or unsustainable construction practices. • Landslide risk management through nature based solutions: Promote tree planting and terracing initiatives to stabilize slopes and reduce the risk of landslides by improving soil structure and Post-Installation Monitoring and Maintenance • Regular Inspections: Conduct regular post-installation inspections to verify the integrity and functionality of the cable infrastructure, particularly after severe weather events or flooding. • Prompt Repairs: Address any identified issues promptly to prevent further damage and ensure the continued reliability of the network. In addition to the general principles outlined above: • Enhanced cable protection: Utilize reinforced/armored cables, rugged cable ducts, and secure anchoring systems to provide enhanced protection and durability for cables, especially in areas prone to landslides or other geological hazards. • Landslide risk management through activity control: Actively limit and monitor human activities Landslide that increase landslide risks, such as deforestation or unsustainable construction practices. • Landslide risk management through nature based solutions: Promote tree planting and terracing initiatives to stabilize slopes and reduce the risk of landslides by improving soil structure and moisture retention. This can be particularly effective in areas with steep slopes and vulnerable soil types. To protect buried cables from heat stress and fire, ensure adequate burial depth and consider these additional measures: Heat and • Thermal Insulation: Use cable sheaths or coatings that provide thermal insulation to enhance wildfire temperature resistance. • Soil Considerations: Be aware that soil properties, including thermal resistivity, affect heat transfer. Corrective backfill is important. In earthquake-prone areas, utilize flexible ducts and conduits for cable installation, along with seismic Earthquake bracing, to enhance the resilience of cables. For aerial cables in flood-prone areas, consider these measures: • Pole Reinforcement: Strengthen poles to enhance stability. • Tensioned Installation: Minimize cable sagging using proper tensioning during installation to Floods prevent signal loss or cable failure. • Water-Resistant Cables: Use cables with water-resistant insulation and protective coatings. • Weather-Resistant Components: Protect all components, including connectors and junction boxes, with weather-resistant materials. To effectively mitigate the risk of landslides: • Geotechnical Screening: Conduct thorough geotechnical screening on slopes to assess stability and identify potential landslide risks. This involves analyzing soil composition, slope angles, and drainage conditions to inform appropriate engineering solutions. • Human activity control: Actively limit and monitor human activities that increase landslide risks, such as deforestation or unsustainable construction and land development practices. Landslide • Nature based solutions: Promote tree planting and terracing initiatives to stabilize slopes and reduce the risk of landslides by improving soil structure and moisture retention. This can be particularly effective in areas with steep slopes and vulnerable soil types. • Monitoring Systems: Establish monitoring systems to detect early signs of slope movement or instability, such as ground cracks or unusual seepage. Early detection allows for timely intervention and mitigation efforts. For protecting aerial cables against wind-related hazards, consider implementing the following measures: • Pole Strengthening: Strengthen poles to enhance their stability. Install stays at the last end posts Aerial Cables to balance wind forces. • Tension and Sag: Install cables with appropriate tension and sag, adhering to local wind load Wind requirements. • Line Stays: Install line stays at every n-th pole along the route, as well as on poles either side of rivers and road crossings where normal span lengths are exceeded. • Regular Inspections: Conduct regular inspections of poles and cables to identify any signs of wear, damage, or instability. Prompt maintenance can prevent issues from escalating and ensure that the infrastructure remains secure against wind forces. To protect buried cables from heat stress and fire, consider these measures: • Thermal Insulation: Strengthen thermal insulation with cost-efficient solutions to improve the heat resistance of aerial cables. • Fire-Resistant Cables: In areas prone to fire, consider utilizing fire-resistant cables with flame- Heat and retardant materials. wildfire • Spacing: Maintain adequate spacing between aerial cables to facilitate proper heat dissipation and minimize the risk of fire propagation. • Vegetation Management: Maintain a clear space around the poles by removing vegetation, debris, and other combustible materials. • Non-Combustible Materials: Apply non-combustible pole materials like zinc or steel poles. In earthquake-prone areas, consider these measures in addition to those proposed for landslide risk. • Vibration Dampening: Employ vibration dampening mechanisms to mitigate the impact of 31 Earthquake vibrations on aerial cables. • Slack Loops: Incorporate slack loops in the aerial cables to provide flexibility, allowing for wildfire • Spacing: Maintain adequate spacing between aerial cables to facilitate proper heat dissipation and minimize the risk of fire propagation. • Vegetation Management: Maintain a clear space around the poles by removing vegetation, debris, and other combustible materials. • Non-Combustible Materials: Apply non-combustible pole materials like zinc or steel poles. In earthquake-prone areas, consider these measures in addition to those proposed for landslide risk. • Vibration Dampening: Employ vibration dampening mechanisms to mitigate the impact of Earthquake vibrations on aerial cables. • Slack Loops: Incorporate slack loops in the aerial cables to provide flexibility, allowing for movement and preventing excessive tension during earthquakes. To enhance the resilience of critical connectivity and data facilities against flooding, consider the following measures: • Elevated Foundations: Ensure that the foundations and floors of critical facilities, such as internet exchange points and data centers, are constructed at elevated levels above potential flood levels. If feasible, consider relocating new facilities to areas with lower flood risks to minimize exposure. • Flood-Resistant Design: Incorporate flood-resistant designs and materials throughout the facility. This includes: • Installing critical equipment (servers, switches, power distribution units) at higher elevations within the facility. • Implementing green roofs and permeable pavements to manage stormwater and reduce runoff. Floods • Creating vegetated buffer zones along rivers and coastlines to act as natural flood defences. • Constructing water management ponds and stormwater drainage systems near key digital infrastructure sites. • Data Redundancy: Secure backup data offsite or in the cloud to ensure data redundancy and availability. This reduces the risk of data loss or damage during flood events. • Regular Risk Assessments: Conduct regular assessments to evaluate existing and projected flood risks. Identify vulnerabilities within the infrastructure and prioritize necessary improvements to enhance resilience. • Actionable Disaster Recovery Plans: Develop comprehensive actionable plans that outline protection measures and strategies for quick recovery from flooding disasters. This ensures that critical facilities can maintain operations during and after an event. Landslide risk mitigation measures, in addition to the above: • Activity Monitoring: Limit or monitor human activities likely to trigger slope instability and increase landslide or flash flood risks. Consider slope stabilization activities and stormwater Landslide management systems diverting water away from slopes. Critical • Reinforced Construction: Reinforce buildings and construct retaining walls facing areas prone to connectivity landslides and flash floods. and data Wind risk mitigation measures facilities • Structural Reinforcement: Reinforce the structural integrity of walls, roofs, and foundations to withstand potential impacts from various hazards. Wind • Equipment Anchoring: Securely anchor outdoor equipment, such as generators, fuel tanks, solar panels, and air-conditioning units, to prevent damage caused by strong winds. • Vegetated Buffer Zones: Consider vegetated buffer zones as windbreakers. Heat and wildfire risk mitigation measures: • Heat Stress Monitoring: Monitor heat stress to identify potential issues related to excessive heat, enabling timely corrective actions to maintain optimal operating conditions. • Redundant Cooling Systems: Implement redundant and energy-efficient cooling systems to ensure resilience in case of cooling system failures or power outages/restrictions. • Vegetation Management: Consider vegetated surroundings and green roofs to reduce heat Heat and absorption, while ensuring a clear space is maintained by regularly removing dry vegetation, wildfires debris, and other combustible materials. • Fire Suppression Systems: Implement automatic fire suppression systems for rapid response and effective extinguishment in the event of a fire, safeguarding critical equipment and minimizing potential damage. • Data Redundancy: Ensure backup data offsite and/or in the cloud to ensure data redundancy and availability, reducing risks of data loss or damage in the event of a fire. Earthquake risk mitigation measures, in addition to those proposed for landslide risk. • Geological Assessments: Conduct thorough geological assessments to evaluate risks associated with fault lines and soil stability before site selection. • Structural Reinforcement: Reinforce the structural integrity of walls, roofs, and foundations to Earthquake withstand potential impacts from vibrations caused by earthquakes. • Data Redundancy: Ensure backup data offsite and/or in the cloud to ensure data redundancy and availability (applies especially for data centers). 32 Flood risk mitigation measures • Site Selection: Assess flood risks and locate sites at a safe distance from flood-prone areas. • Elevation: Select elevated locations and/or elevate electronic equipment and power backup supplies. • Reinforced Foundations: Construct tower foundations using reinforced concrete. Determine soil bearing capacity through local soil samples for each tower site. Floods • Drainage Systems: Implement proper drainage systems to prevent water accumulation around the tower base. • Corrosion-Resistant Materials: Utilize corrosion-resistant materials (e.g., galvanized steel) and protective coatings to enhance the longevity and durability of mobile towers. Smaller towers may be constructed using aluminium. • Regular Maintenance: Conduct regular maintenance to identify and address any potential issues or deterioration promptly, especially in coastal or high-humidity locations. Landslide risk mitigation measures • Site Selection: Assess landslide risks and locate sites at a safe distance from areas subject to landslides. • Structurally Sound Foundation: Ensure mobile towers have a structurally sound foundation, reinforced to withstand potential stresses. Landslide • Flexible Connectors: Consider incorporating flexible connectors to absorb vibrations and movements, enhancing the tower's resilience. • Activity Monitoring: Implement measures to limit or monitor human activities that may increase Mobile Towers the risk of landslides. • Retaining Walls: Consider constructing retaining walls in locations prone to landslides. Wind risk mitigation measures • Structural Integrity: Reinforce the structural integrity of mobile towers and review local wind load requirements to ensure they can withstand strong winds and severe weather conditions. For extreme wind conditions, a safety factor of 1.5 -2 or higher is advisable to accommodate potential future increases in wind intensity.14 In cyclone-prone regions, design wind speed/load capacity Wind should be > 150km per hour. • Guy Wire Maintenance: Conduct regular inspections and maintenance of guy wires and anchors to ensure they are properly tensioned. • Wind Pattern Monitoring: Monitor/control local geographical features that will influence wind patterns, prior to and following construction. Heat and wildfire risk mitigation measures • Thermal Insulation: Enhance thermal insulation beyond surface paint by implementing additional Heat and measures such as insulation materials or coatings. wildfire • Cooling Systems: Implement cooling systems for cell sites housing electronic components. • Vegetation Management: Maintain a clear space around the structures by removing vegetation, debris, and other combustible materials. Earthquake risk mitigation measures, in addition to those proposed for landslide risk. Earthquake • Structural Soundness: Ensure mobile towers are structurally sound, incorporating appropriate dampening mechanisms and potentially flexible connectors. In addition to the technical specifications mentioned above, a variety of additional details can be considered, but will require more information on sites, network specifications, national standards and regulatory environment, and other project-specific elements. These details may include specific criteria for material selection, power supply requirements, installation techniques, site selection and preparation methods, as well as various standards that apply to electronic components and equipment parts. 33 The use of low Earth Orbit (LEO) satellites is not reflected 4.1.1 Floods risk reduction in the table above but has become increasingly relevant for improved and resilient connectivity, especially in Ideally, infrastructure should be placed outside flood- underserved areas and climate hotspots. LEO satellite prone areas, but this is not always feasible due to logistical communication offers considerable advantages, notably or economic constraints. To mitigate flood risks, several in deployment costs and climate resilience. Compared strategies can be employed. Common best practice to traditional digital infrastructure, which demands measures are included in table 4.3, shown above, hence the substantial time and resources to build, satellite internet following highlights a couple of those. To protect fiber optic can be deployed rapidly, making it a particularly effective cables from flood risks, they can be installed underground solution for remote and underserved regions, as well as in within reinforced conduits and buried at depths greater than emergency situations where rapid communication need 1.5 meters (where feasible). To further enhance protection to be established. With the exception of ground stations, against water ingress and mechanical damage, moisture- LEO satellites themselves are largely immune to climate resistant jackets and metal reinforcements can be applied. and natural hazard risks. Furthermore, recent industry Aerial cables are less susceptible to flooding but are more investment in commercial LEO satellite networks has led to vulnerable to damage from extreme weather conditions, relatively competitive user connection prices.28 While LEO such as high winds and lightning strikes. Proper protection satellites hold promise for enhanced connectivity, the extent and maintenance of poles are crucial to ensure the integrity to which they can replace or complement existing fixed and of aerial installations. For cellular networks, careful mobile broadband technologies in sparsely populated rural selection of cell site locations is essential to avoid flood- areas remains to be seen in East Africa. It’s also important to prone areas. If such areas cannot be avoided, creating flood consider that, within the broader effort to make the digital barriers around the site can help mitigate risks. Also, raising sector greener, LEO broadband satellites are associated critical electronic equipment and power sources at least one with negative environmental and climate externalities, such meter above ground level can provide basic protection from as emissions from rocket launches.29 floodwaters, ensuring that essential components remain operational during flooding events. Implementing self- sufficient power systems, such as solar panels and batteries, 4.1 Resilience specifications can ensure continuous operation of cell sites during power outages caused by flooding. for different hazards The following subsections examine specific hazard Additionally, physical network (digital) inventory can aid scenarios and demonstrate how selected resilience in the efficient planning, management and maintenance measures from the preceding table can be effectively applied of fiber networks, especially during post-flood recovery to infrastructure assets. These examples also illustrate the efforts. Access to detailed maps and real-time updates and type of language suitable for inclusion in Terms of Reference access to inventory data, can facilitate precise repairs and (TORs) or Requests for Proposals (RFPs). In alignment with resource allocation. Utilizing advanced monitoring systems, the World Bank’s principles, procurement documents should such as fiber-optic sensors, can help detect potential avoid prescribing specific technologies, thereby enabling issues early, allowing for proactive measures to prevent or a diverse range of bidders to propose solutions grounded minimize damage from flooding. in their expertise and experience. However, depending on the country context and procurement modality, vendors may require clear guidance on project greening objectives, referencing relevant standards and recommended resilience measures to support these goals. 28 Starlink has recently launched a rental option for its internet kits in Kenya, making its services more accessible to customers. The rental model requires a one-time activation fee of KSh 2,700 (approximately USD 21) and a monthly hardware rental fee of KSh 1,950 (about USD 15). In addition to these fees, customers can choose service plans starting at KSh 1,300 (USD 10) per month for a 50GB data plan with speeds reaching up to 200 Mbps. By offering a rental option, Starlink aims to broaden its customer base and compete effectively with local internet providers like Safaricom and Airtel accessed on oct 3rd, 2024 29 Edward Oughton, Bonface Osoro, Andrew Wilson et al. Sustainability assessment of Low Earth Orbit (LEO) satellite broadband mega-constellations, 12 October 2023, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-3325730/v1] 34 4.1.2 Landslide risk reduction 4.1.4 Other hazards: Heat/humidity and earthquake risk reduction To protect digital infrastructure, such as fiber networks and telecom towers, from landslide impacts, several strategies Installing stress-resistant and heat-tolerant equipment, as can be implemented. These strategies focus on monitoring, well as sensors for real-time monitoring of temperature, infrastructure design, and emergency preparedness. humidity, and system performance allows for proactive Utilizing wireless technologies, sensors and remote sensing management of potential issues before they escalate allows for continuous monitoring of landslide-prone areas.30 into failures. Adhering to best practices and international Choosing appropriate locations for telecom infrastructure standards are crucial to ensure the reliability and is vital. This includes avoiding high-risk areas identified performance of telecommunications systems in harsh through GIS mapping of landslide-prone regions. In addition, environmental conditions. various reinforcing engineering works can help to secure cables, power installations, and structures better against Building earthquake-resistant cell towers and using landslides, just as landscape management (e.g. reforestation underground cables can protect against physical or drainage improvements) can reduce the risk of landslides. damage. Critical assets such as data centers are protected Developing comprehensive disaster management plans that against earthquakes through a combination of advanced include response strategies for landslides is essential. This construction techniques (seismic-resistant design), involves regular risk assessments and maintenance checks strategic location, and specialized equipment. It is also on existing infrastructure to identify vulnerabilities. worth mentioning that fiber optic cables have emerged as a promising tool for seismic early warning systems.32 4.1.3 Storm and coastal flood risk reduction To enhance resilience against storms and coastal floods, 4.2 Resilience objectives and indicators various strategies can be implemented across different To effectively achieve outcomes that align with resilience types of infrastructure. Critical facilities like cable landing standards, establishing resilience objectives and stations, as well as data centers and IXPs should be elevated performance indicators is essential. Performance indicators to mitigate flood risks and ideally redundant facilities should serve as a vital tool for the public sector, enabling the be established in different locations. Telecom towers and monitoring of project execution quality by private partners. other structures located in areas prone to strong winds When designed thoughtfully and measured consistently, can be designed with reinforced bases and wind-resistant performance indicatorss can significantly enhance the features to withstand high winds31. This includes using performance of resilience initiatives. Regular performance aerodynamic shapes, impact-resistant materials and guyed measurement not only reinforces successful activities but anchoring. Installing securely anchored and flood proof also identifies areas that require improvement. The primary backup power supply is crucial for maintaining operations challenge lies in developing performance indicators that are during power outages caused by storms and storm surges. As directly relevant to the overarching resilience initiative. previously indicated, burying fiber optic cables in protected ducts at appropriate depths can protect them from wind damage and storm surges. Having emergency plans and A practical approach to creating effective performance maintenance programs that involve conducting inspections indicators begins with identifying measurable resilience prior to storm seasons can help to identify weaknesses in criteria. It is important to consider both quantitative and infrastructure and ensure a prompt response to restore qualitative metrics in this process. Each performance lost connectivity in the event of a disaster. Furthermore, indicator should focus on a specific aspect of resilience, apart from sector-specific adaptations, different forms of set clear performance goals, and establish appropriate urban management and coastal protection are important, time frames for assessment. The number of performance including nature-based solutions such as the restoration of indicators to monitor should correspond to the project’s mangroves and coral reefs. size, value, and the number of assets involved. For smaller projects or those with short-term objectives, it may be more pragmatic to limit the performance indicators to broader goals that the private sector must achieve, allowing them the flexibility to devise their own strategies for meeting these objectives. 30 New technologies using a combination of satellite radar and artificial intelligence (AI) can monitor slope stability in near real-time, to predict early stages of landslide or even the risk of potential landslides in advance 31 Guyed towers are recommended to be used in spacious areas and when high towers (and coverage) are needed to cater to high anchor radius that helps in better tower resistance to winds and other stressors. Lattice towers are recommended when the available land is not big, while monopole tower can be used when an aesthetical aspect is of great importance to fit in the surrounding environment, e.g. in dense urban areas. 32 A method known as Distributed Acoustic Sensing (DAS) significantly improves earthquake detection by transforming fiber optic cables into ultra-dense arrays of seismic sensors. This technology uses laser pulses to measure the phase changes in the light scattered back from imperfections in the fiber, allowing for the detection of subtle changes in acoustic signals caused by seismic activity. Source: https://www.preventionweb.net/news/measuring-earthquakes-and-tsunamis-fibre-optic-networks 35 Table 4.4: Resilience objectives and indicators General objectives: • All contractors are to conduct site specific risk assessments in the area where infrastructure will be deployed. • The telecommunication infrastructure is designed/constructed to withstand climate-related risks. • The telecommunication assets adhere to relevant building codes, standards, and guidelines for climate resilience. • The telecommunication assets should have robust power supply systems and back-up systems, reflecting the risks of climate hazards and power outages. • Regular maintenance, inspections, and upgrades should be conducted to ensure the ongoing resilience of the assets aligned with evolving risks. • Establish clear recovery objectives, which define the maximum acceptable downtime per year or incident and/or specify the maximum acceptable amount of data loss. • User/customer satisfaction on accessibility and performance of the digital infrastructure during and after a disaster. Specific objectives Development of: • Climate and/or disaster risk assessments; Project design: • Climate adaptation resilience studies and plans; • Government standards and regulatory rules/guidelines; • Emergency response plan addressing climate events Ratio of change in or Existence of: • The kilometres of climate-resilient fibre; • Coverage of early warning systems; Implement • Backup systems or routes to primary systems. metrics: • Redundancy plans; • Emergency response plans; • Resilience standards and regulatory rules/guidelines. Number of, Size of or Changes in: • Climate-related incidents causing disruptions or requiring significant repair; • Users/customers, or critical facilities, or network elements affected by the network service disruption; • Size of the region affected by the disruption. Impact metrics: • Direct damage costs/loss (e.g. cost of repair); • Financial impact of the climate-related disruption, e.g. direct financial liabilities to contractor (cost of fines and/or loss of revenues). • Users affected: Number of users disrupted by the outage. • Increase in support calls or incident tickets indicating disruption severity. Incident response time or Percentage of time: Duration • To repair physical damages and/or restore service after a failure; metrics: • The service meets agreed performance metrics; • The network is (not) operational and accessible to users (due to climate events). Maximum acceptable or Number/ratio of: • Time to restore services after a disaster; data loss or data access (measured in time). • Digital infrastructure in a network protected against power outages (for example: by having a battery Disruption backup); tolerance: • Available capacity (e.g. bandwidth) in the digital infrastructure preventing traffic overload; • Redundancy level (e.g. separate fibre routes or connectivity options) built into the digital infrastructure preventing single point failure, ensuring minimal downtime, and/or preventing data loss during a disaster. Installation/operation of: • Regularly updated inventory for digital network assets with specs (essential for maintenance and repair) and info on exposure and risk, existing safeguards, incident history etc.; Monitoring • Digital monitoring system that includes fault or weather detection and/or forecasting modules (e.g. for system: seismic vibrations, extreme rainfall, extreme heat or humidity levels); • An Early Warning System (EWS) for operators and end users; • A feedback system on user experienced service quality and reliability. Note: The included performance indicators are adapted from existing World Bank projects and industry best practices. The specific performance indicator objectives can also be broken down and customized to the different types of digital infrastructure assets. There are specific standards in place for performance indicators, such as the guidelines created by ISO for data centers. https://www.iso.org/standard/83496.html 36 It is advisable for the private party to collect performance risk mitigation measures. A fair distribution of risks is indicators, while the public party should oversee the crucial for maintaining fiscal sustainability and ensuring monitoring process to ensure that these performance that both parties are incentivized to perform well. This indicators comply with the contractual terms. This involves recognizing that while the private sector may take oversight should be supplemented by external audits of on significant risks, the public sector must also retain some the performance indicator collection process. To facilitate responsibility to ensure accountability effective monitoring, the contract must explicitly define the indicators to be monitored, specify the phases during which The main idea guiding risk allocation is that all risks whose they will be evaluated, and include a detailed data collection impact can be avoided by implementing the adaptation protocol. This protocol should guide the private sector on measures listed in the contract should be borne by the how to accurately gather the required data. Additionally, private party. Conversely, the financial consequences of risks the contract should outline any variations in reporting that will impact the infrastructure despite the deployment frequency for individual KPIs. Information regarding project of these measures must be either shared or borne by the performance should be considered made publicly accessible. public party. Shared or public risks may be associated with situations of force majeure, where adaptation is no longer sufficient to guarantee infrastructure resilience. 4.3 Risk allocation management This means that the private partner should take on risks they can control effectively, such as design, construction, and Risk sharing management is a strategic approach that aims operational risks, while the public sector in part or entirely to distribute potential risks among various stakeholders retains risks that are more efficiently managed by them, involved in a project. This method not only mitigates the such as regulatory changes or force majeure events. This impact of unforeseen events but at best reduces the risk creates a strong incentive for the private player to respect of disputes and failures, all while enhancing collaboration the adaptation measures listed in the contract. and resource optimization. By allocating risks to those best equipped to manage them, organizations can create a more Developing annotated risk allocation matrices, shown in a resilient framework for project execution. Effective risk simple version below, can provide clarity on how risks are management encompasses several key activities, including shared between public and private partners.34 It also allows thorough risk assessment, development of risk mitigation for defining anticipated risks as opposed to unforeseen strategies and risk allocation, as well as regular review and ‘force majeure’ events relevant to the project. Experts evaluation of risk throughout the project’s lifecycle. with relevant digital infrastructure experience, as well as technical, environmental, insurance, and legal professionals, Fig 4.1: Risk Management are required to create the matrix and evaluate the risks. The public and private sectors often approach risk management from fundamentally different perspectives. Public agencies may prioritize service delivery, compliance, and transparency, while private entities tend to focus more Risk Determination Risk on financial implications and risk pricing. This divergence can Identification of risk mitigation Allocation lead to ineffective risk transfer and management, resulting in cost overruns, project delays, and ultimately, project failure. To mitigate these challenges, fostering a collaborative exchange of knowledge regarding all relevant risks associated with projects is essential. It is crucial to ensure that the risks assumed by both the public and private parties are balanced with the expected benefits. As previously mentioned, this should be done by regularly evaluating and A simplified linear representation of the risk management updating these risks in response to changing circumstances process can be helpful in illustrating the steps involved. and predictions, while keeping in mind the high uncertainty However, in reality, the process is more interlinked and involved in predicting future climate challenges. Moreover, cyclical as risks, strategies for managing them, and risk providing financial support, including public subsidies, tax allocation between parties must be continuously assessed breaks, or reductions, as well as other economic incentives, and possibly revised. The World Bank and PPIAF Risk can significantly enhance private sector motivation to Allocation Tool33 can serve as a reference guide for invest in and maintain resilient digital infrastructure. Such governments and other relevant stakeholders in deciding incentives can help alleviate the financial burdens associated on the appropriate allocation of project risks in a given with infrastructure investments and encourage private Public Private Partnerships project, as well as potential entities to adopt more robust risk management practices. 33 https://ppp-risk.gihub.org/ 34 Several examples of risk allocation matrix can be found in the PPP Risk Allocation Tool 2019 https://ppp-risk.gihub.org/, each tailored to a given project type. 37 Additionally, regulatory support and security for these investments can serve as a strategic approach to meet the additional requirements for project execution, particularly those related to building resilience. By aligning the interests of both sectors through these mechanisms, it becomes possible to achieve more effective risk management and successful project outcomes. Table 4.5: Risk allocation matrix template Risk category and Risk allocation Rationale and risk mitigation measure description Private Shared Public Increase in preparation and construction cost Public party typically responsible of - obtaining all environmental licenses and environmental authorizations - assessing environmental and geotechnical risks as part of feasibility Climate and studies natural hazard risks in design and Private party typically responsible of x construction - complying with all environmental licenses/law - environmental impacts caused during the project construction x - compliance of technical design with the climate resilient standards/ output specifications required by the contracting authority - buying insurance, although this could be eligible for support from the public party. Increase in operation and maintenance cost Private party typically responsible of - increased maintenance costs exacerbated by climate change and natural hazards as they must handle the risk of meeting the appropriate maintenance standards as set out in the performance specification. Ideally, contractual provisions should describe minimum requirements Climate and for day-to-day maintenance and life-cycle maintenance and replacement natural hazard of assets. risks in operation/ x - Increased maintenance costs from physical damage and disruptions maintenance caused by climate and natural hazards for which the project should have been adequately designed and appraised. Ideally, there is an established threshold/cap above which the contracting authority will bear/share risk, and the private partner will be entitled to compensation. Thus, the risk of financing additional adaptation beyond those planned should be shared among parties. Increase in costs, despite having implemented the resilience measures required by contract, due to unexpected events occurring, beyond the control of the parties, that will delay or prevent performance. Private and public party typically share the responsibility of - Climate risks caused by extreme and unpredictable events that overly exceed the design limits of adaptation work and are beyond any Force majeure risks insurable extreme. Ideally, define climate-related and other events that can be classified as force majeure and the benchmark intensity level x above which the climate event is considered force majeure. Other events could include political disruptions or changes in applicable legislation that can neither be foreseen nor managed, in which case the x public partner would typically bear risks. Other risks The allocation of risks may vary based on the nature of risks and the (financial, legal, x x x knowledge, projection or expectations existing at the time of contract demand, social etc.) formation. Note: The table includes information adopted from the GIF 2019 risk allocation tool https://ppp-risk.gihub.org/ 38 4.4 Enhancing Resilience: Costs and Benefits The costs and benefits of improving resilience in digital infrastructure can vary widely depending on several factors, such as the type of asset involved. Other important considerations include the specific resilience measures implemented, the degree of exposure to risks and vulnerabilities, and various contextual factors. While several studies have evaluated cost- benefit ratios for resilience investments in other infrastructure sectors, often finding benefit-cost ratios between 2 and 4, which suggests these investments are generally economically advantageous, there is a notable lack of similar data for the telecommunications sector.35 Despite this data shortage, it is reasonable to conclude that investing in resilience is economically justifiable, especially in regions where the risk of costly connectivity disruptions and asset damage is currently high and expected to remain so. However, to maximize cost-effectiveness, such investments should be informed by risk assessments, careful prioritization and close dialogue with key sector stakeholders. When evaluating the adoption of resilience measures, the costs associated with these resilience measures can vary widely. In many instances, the additional costs for implementing resilience strategies are relatively minor, constituting a small percentage of the total investment. For example, in data centres where significant capital is already allocated, these costs may represent only a few percent of overall expenditures. However, for linear infrastructure projects such as fibre optic cables, adhering to recommended standards for materials, installation and maintenance, as well as redundancy can lead to substantial increases in expenses. Moreover, investing in resilience presents a less compelling business case in sparsely populated areas, highlighting the critical need for subsidies or other economic incentives to bridge this gap. In these regions, both residents and businesses generally have a lower willingness and ability to pay for improved services compared to their counterparts in densely populated urban centres, where a larger number of users share the same infrastructure, leading to greater cumulative financial losses during disruptions, which makes investment in resilience more economically justifiable.36 Box 4.2: Enhancing redundancy cost-benefits Implementing redundancy: This typically involves duplicating essential components, such as servers, network pathways, power backups, and storage systems. While this strategy can significantly bolster system reliability and resilience, it often necessitates considerable initial investments. Depending on the complexity and scale of the infrastructure, these upfront costs can increase by 20-50% or more. Ongoing maintenance and operational costs: Redundant systems also require continuous maintenance, which can elevate operational expenses. To ensure that backup systems function effectively during failures, regular updates and testing are crucial. This ongoing commitment may add an estimated 10-20% to annual operational budgets. Furthermore, the introduction of redundancy increases system complexity, which in turn demands more skilled personnel for management and oversight. Consequently, organizations may face higher staffing costs or the necessity for specialized training programs. Long-term saving through reduced downtime: Despite these initial and ongoing costs, enhanced redundancy can yield significant long-term savings by minimizing downtime and the associated losses during system failures. Outages can be financially devastating; for instance, the 2023 Uptime Institute global data center survey revealed that each outage can easily cost over USD 1 million.37 Power issues consistently rank as the most common cause of severe data center outages, while network-related problems represent the largest single cause of IT service disruptions. Alarmingly, the time gap between the onset of major public outages and full recovery has widened significantly over the past five years. Additionally, extreme weather events exacerbated by climate change have increasingly contributed to data center and connectivity disruptions —a trend likely to intensify and elevate outage risks unless proactive measures are taken.38 Strategic implementation of redundancy: Rather than applying redundancy uniformly across all systems, project teams and organizations should consider a selective approach grounded in thorough risk assessments. Identifying which components of their infrastructure are most critical and vulnerable to failure is essential. Furthermore, defining acceptable levels of downtime for various systems can guide decisions on where redundancy is most necessary. For example, systems integral to critical operations may require complete redundancy, while less critical systems might be able to tolerate some level of risk. In addition to these considerations, ensuring equal access to connectivity should also factor into redundancy planning, particularly in climate hotspots and areas with especially vulnerable communities, such as refugee camps.39 35 Hallegatte, Stephane; Rentschler, Jun; Rozenberg, Julie. 2019. Lifelines: The Resilient Infrastructure Opportunity. Sustainable Infrastructure;. © Washington, DC: World Bank. http://hdl.handle.net/10986/31805 36 However, prioritizing resilience investments in these less populated areas is not just an economic issue; it is fundamentally a question of inclusion and equity. Ensuring that all communities, regardless of their population density, have access to reliable connectivity is essential for fostering equitable digital and economic development and safeguarding the well-being of all residents. 37 https://uptimeinstitute.com/resources/research-and-reports/uptime-institute-global-data-center-survey-results-2023 38 Uptime Institute 2024, Keynote Report 131: Annual outage analysis 2024. See also Rosie McDonald; Sara Ballan. Green Data Centers : Towards a Sustainable Digital Transformation - A Practitioner’s Guide (English). Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/099112923171023760 39 Climate change can exacerbate existing inequalities, and infrastructure systems can amplify these disparities when affected by climate hazards. Therefore, strategic redundancy investments may be justified to protect vulnerable populations and guarantee their access to essential services. More information is avia from other reports from the World Bank Global Digital Team which are referenced in the text. 39 Digital infrastructure contracts can be strategically Infrastructure contracts often extend over several decades, designed to encourage effective risk mitigation, integrating making it essential to anticipate how risks may evolve these incentives directly into the procurement process. One throughout their duration. The increasing frequency of effective approach is the implementation of performance- climate-related hazards is likely to drive up both investment based payments or bonuses for private partners. These and maintenance costs, particularly as performance incentives reward partners for successfully managing risks requirements are adjusted to address these emerging risks. and ensuring that projects are completed on time and Given the specialized nature of climate-related challenges, within budget. For instance, if the infrastructure remains engaging experts as mediators can significantly enhance operational and uninterrupted before, during, and after a the resolution of disputes arising from changing external hazard incident due to the proactive measures taken by the circumstances. This approach ensures that conflicts are private partner, they may receive a financial reward. addressed fairly and knowledgeably, thereby minimizing the potential for prolonged disputes and uncertainty. Conversely, contracts can also incorporate penalty clauses. These clauses impose financial penalties on private partners To mitigate disputes and uncertainties in infrastructure who fail to adequately manage risks or who do not meet contracts, it is important to incorporate comprehensive project deadlines and budgetary constraints. This approach climate assessments and resilience measures from the not only addresses the costs associated with repairing or project’s outset. Additionally, as described in the previous replacing damaged infrastructure but also emphasizes section, establishing clear provisions for the allocation the importance of accountability in risk management. By of risks and associated costs following climate events combining performance-based incentives with penalty can further reduce conflicts and enhance overall project clauses, digital infrastructure contracts can create a balanced stability. Contracts should include provisions for adjusting framework that promotes both proactive risk management risk allocations as circumstances change, particularly for and accountability among private partners. dynamic risks that may not be predictable at the outset. Such changes may arise due to technological advancements or the need for quick adaptation in the face of a climate event. Contracts may also allow private partners to share in cost Renegotiation mechanisms should be included to manage savings achieved through effective risk management to any change that could put the project at risk, for instance provide an additional incentive to prioritize innovative, incorporating updated climate risk analysis or scenarios cost-effective climate adaptation and mitigation measures. or new technologies, e.g., next-generation fiber optics and For example, if renewable energy solutions are deployed, satellite communication systems. resulting in cost savings, they could be entitled to a portion of those savings. Box 4.3: Risk allocation requirements in Kenya PPP Contract Public-Private Partnerships (PPPs) in Kenya are governed by a framework that aims to ensure transparency, accountability, and efficiency in the procurement and management of public projects. Among the key requirements for entering into a PPP contract, risks associated with projects need to be adequately managed and mitigated. The PPP framework emphasizes the importance of clearly identifying and allocating risks between public and private partners. Contracts must specify which party is responsible for each risk to avoid ambiguity and ensure accountability, especially when public funds are involved. Aside from this, the contract should also detail responsibilities, payment mechanisms, dispute resolution processes, and other essential elements. The private partner is typically responsible for managing risks related to construction and operational aspects of the project. This would therefore include ensuring that adequate measures are in place to mitigate potential disasters during the project’s lifecycle. The public partner (the government) retains responsibility for certain disaster risks that it can manage more effectively than private entities. This includes providing support mechanisms such as guarantees or financial assistance when necessary 40 Greening of energy for digital infrastructure 5 Greening digital infrastructure means both reducing its Electricity is the primary energy source for carbon footprint (climate mitigation) and enhancing its telecommunications, but operators also rely on fuel- resilience to climate risks (climate adaptation). While powered generators, especially in off-grid or unreliable earlier sections focused on resilience, this chapter addresses grid areas, leading to significant greenhouse gas emissions. improving energy efficiency in data centres, networks, While connecting to cleaner electricity grids remains the and devices, alongside increasing renewable energy use. ideal, deploying emissions-free on-site power solutions, like Although greening also involves broader issues like e-waste solar, offers a practical alternative. For example, Safaricom in management, this report concentrates on energy and Kenya has installed numerous solar-powered base stations. emissions aspects of digital infrastructure Given East Africa’s abundant solar resources, as shown by photovoltaic potential maps for Kenya and Madagascar Integrating reliable, efficient, and green energy systems is (Table 5.1), the region has strong potential to power its vital for East Africa’s digital infrastructure development. digital infrastructure sustainably. With supportive policies Rapid digitization growth and rising digital service demand and investments, East Africa can harness solar energy to means that the energy landscape must evolve to support drive a greener digital future. these changes sustainably. Many areas still lack stable electricity access, limiting connectivity, stability, and resilience. Regional efforts, such as those by the East African Community (EAC), promote cooperation to improve digital infrastructure alongside energy integration. Box 5.1: Kenya, Solar energy for base stations Safaricom, the largest mobile operator in Kenya, has recently invested heavily in solar power. By the end of 2023, it had installed over 1,400 solar powered base stations, i.e. 23 per cent of its 5300 base stations, with a long-term plan to install solar and battery storage for 5,000 sites by 2050. In Ethiopia, the company has 31 sites using solar with a target of 40 per cent sites on solar by 2030. Besides ambitious environmental goals, rising prices for grid electricity and diesel for generators, as well as the risk of power cuts, are important reasons for this shift. The use of solar has not only lowered carbon emissions, but also helped to reduce costs, with the company reporting that costs savings amounted to KSh23,000 at on-grid sites and KSh59,000 at off-grid sites.40 The new green solution erases the problem of electricity provision across the extended network and significantly reduces OPEX. 2023 2050 1,400 Solar powered base stations 5,300 Solar powered base stations 40 Safaricom 2023 Sustainable Business Report, https://www.safaricom.co.ke/images/Downloads/Safaricom-2023-Sustainable-Business-Report.pdf 41 Table 5.1: Solar resource and PV power potential maps for Kenya and Madagascar Photovoltaic power output: 4.03 - 5.15 kWh/kWp Photovoltaic power output 3.59 -5.25 kWh/kWp KENYA MADAGASCAR Note: The regional statistics of solar resource and PVOUT available from World Bank Solar Atlas are calculated from long-term averages Available data indicates that solar panels represent a highly valuable investment for digital infrastructure in East Africa, offering significant advantages both in operational cost savings and environmental impact. Analyses conducted using a specialized tool developed by Deloitte in partnership with the World Bank demonstrate that the upfront costs of installing solar panels are typically recovered within approximately seven years through reduced electricity and fuel expenses. Moreover, the adoption of solar energy leads to a substantial reduction in greenhouse gas emissions (exceeding 60%, with lowest reductions in Kenya due to greener power grid), with the exact figure varying based on the country’s energy mix and the proportion of renewable energy integrated into its national grid. Beyond cost efficiency and reduced climate footprint, solar energy enhances connectivity resilience, particularly in remote or hard-to-reach areas where the national power grid is either unreliable or insufficient. This exemplifies how renewable energy can simultaneously support climate mitigation and adaptation goals in digital infrastructure. 42 Table 5.2: Calculations of emission and cost reduction from solar installed at 120 tower sites in Madagascar In this hypothetical project scenario, it is assumed that a total of 120 tower sites in Madagascar will be equipped with solar power systems. The transition to solar energy is planned to occur progressively over a six-year period, starting from year 0. Specifically, 20 sites will be converted each year, beginning with 20 sites in year 0, followed by an additional 20 sites annually, until all 120 sites are fully transitioned by year 5. 1 BASELINE: General information about a site´s emissions and distribution of emissions between grid and generator energy. All calculations of GHG emissions due to grid utilization have been performed using country grid emissions factors provided by the UNFCCC and extrapolated to the total number of sites in the project. Generator emissions have been calculated using emissions factors based on UNFCCC. 2 GHG EMISSIONS: A calculation of saved emissions due to the solar panels in all project sites are provided below. It must be noted that, as the implementation of panels is done in different years, only batch 0 will register savings in year 0. The rest of batches will show savings in subsequent years. Additionally, in the case of the hybrid approach, depending on the selected scenario (solar + grid or solar + generator), savings will be 0 in one component or another respectively. If a panel´s useful life ends (set to 20 years for this calculation), it is assumed here that a site goes back to its original state, thus providing no savings. 3 FINANCIAL ANALYSIS: Calculation of CAPEX (Solar panel purchase and installation) is provided. Cash flows are computed, considering grid energy and generator fuel and maintenance as savings, as well as energy surplus that can be sold (this case assumed energy surplus being sold). NPV and IRR are also calculated. Three scenarios are presented, based on the Shadow Price of Carbon (SPC) value that is used or if SPC savings are omitted.41 41 SPC ANALYSIS: Shadow price of carbon is calculated using the emissions savings and multiplying them by the SPC prices, thus generating two additional scenarios to one where SPC is not taken into account. 43 Similarly, energy efficiency of digital infrastructure in East Governments are crucial in promoting energy efficiency Africa can be greatly improved by upgrading both new and and use of renewables within the ICT sector. High-level existing systems. Replacing traditional copper cables with recognition of ICT’s environmental impact is necessary fiber optic cables reduces energy loss and enhances data to drive relevant policies and regulations. Reducing ICT transmission efficiency. Modernizing mobile networks to 4G sector emissions depends on transparent data disclosure, and 5G technologies further lowers power consumption per aggregation, and analysis to support informed decision- data unit by enabling faster and more reliable connectivity. making. ICT regulators, with their sector expertise, are Incorporating AI and machine learning into energy well-positioned to lead this effort but will require capacity management enables predictive maintenance, optimized building to manage data collection and assessment energy distribution, and real-time monitoring, which effectively. Effective collaboration among ICT regulators, together reduce energy waste and improve grid reliability. energy authorities, and environmental agencies is vital for For instance, AI-driven technologies and IoT sensors can developing and enforcing these measures. For example, power down idle equipment during low usage, while machine the Green Digital Action Track at COP28 called for annual learning forecasts optimize renewable energy output. industry reporting on Scope 1, 2, and 3 emissions-a practice that could greatly enhance data availability if widely Data centers and cloud computing operators also play adopted. a crucial role for energy consumption. Designing data centers adapted to local climates-using natural cooling and Regional and national policies aimed at enhancing network modular designs-enhances efficiency. Locating facilities efficiency must be closely aligned with ongoing market near reliable renewable power sources and partnering with innovations while also drawing on lessons learned from utilities ensures stable, sustainable energy supply. New global experiences. The transition from 2G and 3G to more green hyperscale data centers, such as Kenya’s geothermal- advanced 4G and 5G technologies offers significant energy powered EcoCloud data center, offer superior cooling and efficiency improvements and supports a wider array of data- power efficiency, significantly reducing energy use and intensive applications-such as early-warning systems-that carbon emissions. The shift to virtualization and cloud deliver vital benefits to vulnerable communities. However, integration further optimizes resource use. the substantial increase in data traffic associated with these newer technologies can offset some of the energy savings, potentially maintaining overall energy consumption levels.46 Box 5.2: Technological advancement The successful adoption of 4G and beyond depends not only in energy savings for digital infrastructure on network upgrades but also on the widespread availability and affordability of compatible devices. To maximize both Equipment vendors and network operators are increasingly environmental and social benefits across East Africa, it is leveraging artificial intelligence (AI) and machine learning (ML) crucial to ensure access to affordable 4G-enabled devices. to enhance energy efficiency in digital infrastructure without compromising network performance. For instance, AI-driven An energy conscious approach can be adopted through wireless network technologies, such as massive multiple-input various measures of which examples are included in the and multiple-output (MIMO) sleeping modes, help optimize energy consumption in networks.42 Ericsson, a leading network table below. equipment vendor, reports that their AI-enabled RAN solutions have demonstrated potential energy savings of 10-12% for operators by dynamically adjusting network operations based on traffic patterns and demand.43 Similarly, Nokia and Orange announced a partnership to deploy AI and ML-powered software in Nokia’s radio units, introducing an “extreme deep sleep” power-saving mode to optimize energy use in radio access networks.44 GSMA provides a case study of the Finnish operator Elisa, who has developed a machine learning solution called Intelligent Energy Saver, which reduces RAN energy consumption and achieves operational expenditure (OPEX) savings of up to 14%, all while maintaining network quality.45 42 OECD 2025, The Environment Sustainability of Communication Networks. OECD Digital Economy Papers. 43 Ericsson (2023), Why AI-powered RAN is an energy efficiency breakthrough, https://www.ericsson.com/en/blog/2023/1/ai-powered-ran-energy-efficiency. 44 Nokia (2024), Nokia and Orange announce extreme deep sleep energy power saving mode, Press release, https://www.nokia.com/about-us/news/releases/2024/02/20/ nokia-and-orange-announce-extreme-deep-sleep-energy-power-saving-mode/. 45 https://www.gsma.com/solutions-and-impact/technologies/networks/gsma_resources/case-study-elisa-automate/ 46 The World Bank and ITU. 2024. Measuring the Emissions & Energy Footprint of the ICT Sector: Implications for Climate Action. 44 Table 5.3: Different Examples of climate mitigation strategies Section Strategi Energy Advanced Cooling Techniques: Implementing advanced cooling techniques in telecommunications networks and Conservation data centers can drastically reduce energy consumption. and Energy Efficiency Strategies Software Optimization: Employing software and technologies that enhance network performance while reducing energy usage. Data analytics can identify inefficiencies in network operations, leading to more efficient practices. Energy-Efficient Hardware: Investing in hardware specially designed for energy-efficiency, such as next-generation routers and switches. Prioritizing materials and components that lower operational energy consumption. Material Consumption Reduction: Strategies like infrastructure sharing can speed up digital network development and reduce duplication. Renewable On-Site Renewable Generation: In areas with unreliable or non-existing grid access, telecom operators can invest Energy in on-site solutions like solar panels and wind turbines. Strategies Other areas E-Waste Management: Implementing effective e-waste strategies based on recycling, reuse, and repair can save costs and reduce the wider environmental footprint (e.g., energy used for producing equipment and toxins related to e-waste disposal). 5.1 Greening standards As highlighted in the previous section on resilience, it is crucial to establish performance requirements and standards for mitigation strategies. These standards serve as minimum criteria for the design, construction, and maintenance of digital infrastructure. By developing regional and national standards and guidelines—such as those presented in the table above—we can promote consistency among producers, consumers, government agencies, and other stakeholders. Various international standardization organizations, including the International Telecommunication Union (ITU), the International Organization for Standardization (ISO), and the European Telecommunications Standards Institute (ETSI), have created standards and technical guidance documents that serve as valuable references. In the digital sector, these standards not only promote equipment interoperability but also help define best practices. The highlighted reports include overviews of green standards of relevance for the digital sector. Standards can be adopted voluntarily by stakeholders without legal compulsion. However, regulators or contracting authorities may choose to mandate specific standards as legal requirements. Integrating green considerations into procurement processes can encourage telecom operators to prioritize sustainability in their operations. For instance, public- private partnerships can set specific requirements for private contractors to: • Leverage solar in off-grid regions with suitable weather patterns. • Use energy efficient equipment meeting certain criteria.47 • Leverage network optimization systems to reduce energy consumption. • Utilize existing infrastructure e.g., infrastructure sharing and ‘dig-once’ priorities. It would be relevant to establish clear guidelines on both a regional and national level on how climate considerations are integrated into performance requirements, risk allocation, and financial support. These should feed into both the tendering processes and the contracts formed between public and private entities. 47 If a specific label or standard limits the supplier pool, or does not fully apply to telecom networks, consider specifying underlying criteria instead to allow broader partici- pation. 45 Table 5.4: International standards relevant to greening telecom Standard Description Organization ITU International Telecom Union (ITU) is a UN specialized agency for information and communication technologies. ITU’s Telecommunication Standardization Sector (ITU-T) assemble experts from around the world to develop international standards known as ITU-T Recommendations. ITU-T Study Group 5 is responsible for the development of standards on the ICT sector´s environmental aspects.29 Standards applicable to telecom networks: • Construction, installation. and protection of cables.: These standards cover a broad range of topics, such as the construction and deployment of telecom equipment (cables, poles, protections, security etc). They analyze aspects with limited application to greening telecom networks, such as the use of trenching techniques, soil analysis, or cable protection and installation methodologies. • E-waste and circular economy: These standards focus on helping telecom operators reduce electronic waste from networking equipment, as well as integrate circularity into their operations. These tackle critical operational aspects such as managing waste from power equipment, including user devices and batteries. Additionally, they provide guidance to operators for migrating toward circular ICT goods and networks and measuring their progress. • Power feeding and energy storage: These standards provide technical guidance regarding the use of power equipment and storage, although standard L.1210 “Sustainable power-feeding solutions for 5G networks” also considers the environmental implications of proposed solutions. • Energy efficiency, smart energy and green data centres: These standards provide best practices for greening data centers, including energy efficiency frameworks. However, they also provide information regarding telecom netwoks and their energy efficiency, as illustrated in standards L. 1325 “Green ICT solutions for telecom network facilities” or L. 1332 “Total network infrastructure energy efficiency metrics”. • Assessment methodologies of ICTs and CO2 trajectories: These standards provide a higher-level view of the interaction of ICTs with other sectors and the environment. • Circular and sustainable cities and communities: These standards address the specific topic of sustainable cities and communities, including how these can leverage ICTs to mitigate climate impact and key performatnce indicators (KPIs) that can help track development of such projects. • Low-cost sustainable infrastructure: This standard aims to identify general requirements and frameworks for achieving affordable and sustainable telecommunications infrastructures with a special focus on rural communications in developing countries. The focus is not specifically on green aspects. ISO International Standards Organisation (ISO) develops standards in many different domains. Some of these are relevant for green telecom networks; for example, standards related to circular practices. There are two main ISOs affecting • Energy management (50001) • Sustainable practices (14000 series) Telecom operators can certify their different businesses and assets with ISO to demonstrate performance, and most major providers are already certified (for example, Telefonica, Orange, Vodafone). 46 IEC Regional standards organisations also exist in some areas. The most prominent example is ETSI the European Standards organisation for ICT. Within some areas, ETSI has defined green telecom standards for areas not covered by ITU or ISO. ETSI supports European regulations and legislation through creation of harmonised European standards. To tackle sustainability issues and create standards aligned with greening the economy, ETSI has created different committees that address environmental issues associated with the sector: • Environmental Engineering Committee (EEC): The EEC develops standards for reducing the eco-environmental impact of ICT equipment. Some of the main aspects they address include: Life Cycle Assessment (LCA) of network equipment; Methodologies to assess energy efficiency; Circular economy applied to ICT products; Efficient power feeding solutions. The EEC and ITU-T Study Group 5 are working together to develop technically aligned standards on energy efficiency, power feeding solution, circular economy and network efficiency KPI and eco-design requirement for ICT, with the aim to build an international eco-environmental standardization. • Industry Specification Group (ISG) on Operational energy Efficiency for Users (OEU): The ISG OEU defines measures to minimize the power consumption and greenhouse gas emissions of infrastructure, utilities, equipment, and software within ICT sites and networks. They include: Measurement of energy consumption by IT servers, storage units, broadband fixed access and mobile access, with a view to developing global KPIs; Management of the end of life of ICT equipment; Defining global KPI modelling for green smart cities; • Access, Terminals, Transmission and Multiplexing committee (TC ATTM): The TC ATTM focuses on the green needs of operational networks and sites and broadband transmission, including: Developing global Key Performance Indicators (KPIs) to provide users of ICT with the tools to monitor their eco-efficiency and energy management; Defining the networks connecting digital multi-services in cities, producing KPIs for monitoring the sustainability of broadband solutions; Improving standards for transmission equipment to support the European Commission’s Ecodesign of Energy Related Products Directive; Supporting efficient ICT waste management (maintenance period and end of life). To effectively address these challenges, it is crucial for stakeholders from both the public and private sectors to collaborate in defining and prioritizing policy objectives, as well as a roadmap for achieving these. This partnership should aim to integrate climate strategies into the digital transformation agenda while ensuring that climate considerations are harmonized with other vital priorities, including those that promote digital inclusion in unconnected and underserved areas. This collaborative effort involves a thorough assessment of existing infrastructure, including supply and demand dynamics, as well as the associated climate risks and energy footprints. Governments in Eastern Africa can enhance investments in low-carbon digital infrastructure by offering regulatory and financial incentives. These incentives may include tax breaks and subsidies aimed at promoting renewable energy adoption.48 For instance, Kenya has implemented various financial incentives to support solar power adoption, including subsidies for solar panels and equipment. These subsidies come in different forms, including value added tax (VAT) exemption on solar and wind energy specialized equipment. Tax breaks to businesses that invest in solar energy help offset the cost of installing solar panels and make it more profitable for businesses to use solar energy. Similar incentives have recently been introduced in introduced in Ethiopia49 and Madagascar.50 Moreover, effective policies should encourage the co-deployment and sharing of infrastructure, which can yield regulatory and economic benefits for all stakeholders involved. Policies and strategies should also enable and promote co-deployment and sharing of infrastructure, resulting in regulatory and economic benefits for all parties involved. Governments need to establish transparent rules and incentives that encourage collaboration among operators while safeguarding competition and consumer interests. By creating a conducive regulatory environment, policy makers can facilitate infrastructure-sharing agreements and promote investment in shared networks. 48 Innovative financing models, such as viability gap financing, insurance products, service agreements should also be considered. For climate resilience, this is less common (compared to climate mitigation); thus, financial incentives can be used to ensure that the full social cost of infrastructure disruptions are accounted for, encouraging service providers to go beyond just meeting basic quality standards. 49 https://press.et/herald/?p=102385 50 https://taxsummaries.pwc.com/madagascar/corporate/tax-credits-and-incentives 47 Designing for green digital infrastructure requires balancing network resilience with energy efficiency and carbon-emission reductions to create systems that are both robust and environmentally sustainable. Enhancing network redundancy is an important reason to invest in additional network routes and active equipment. For operators, costs are the main limiting factor when considering additional routes and not sustainability considerations. Certain strategies, however, can cut both cost and climate footprint, while maintaining resilience. For instance, exploring collaborative options to leverage networks across different providers can enhance resilience while also cutting costs and lowering the climate footprint 5.2. Greening objectives and sustainability indicators An effective way to achieve the greening standards is to establish clear greening objectives and sustainability indicators. Sustainability indicators are an important tool. Performance indicators are an important tool for the public sector to monitor the quality of project execution by the private partner. Properly designed and regularly measured, performance indicators can make a big difference in how well greening initiative is performing. Regularly measuring performance reinforces activities that are meeting the goals and highlights those that are underperforming. A mix of performance indicators can help in measuring and monitoring the relative efficiency of networks. Here two main categories of performance indicators should be mentioned, one that is focused on the environment - e.g. CO2 equivalent emissions to assessing the company impact on climate change or the circular economy efforts including waste generation and recycling activities, as well as use of additional resources like water - and one that is focussed on energy consumption – e.g. total energy and electricity used by an operator, as well as the efficiency and intensity of its usage. The Global System for Mobile Communications Association (GSMA)51 and The Next Generation Mobile Networks Alliance (NGMN) has both released a set of performance indicators for green networks. Other efforts have been done to develop common indicators for measuring the environmental footprint of electronic communications networks, including by the EU Joint Research Centre (JRC).52 Based on that, the following summary table can be used for guidance on the selection of indicators. Tabel 5.5: Sustainability indicators Sustainability indicators Energy consumption Energy efficiency Energy indicator Use of renewable energy (rate) Must have GHG scope 1,2,3 emissions Climate indicators E-waste production Distribution or utilization of recycled/ refurbished/ reused products Recycled/refurbished/ reused components (also the excavate mass) used in products Recyclability Should have Reparability (optional) Expected lifetime Raw materials depletion (mineral) Environment Water usage/ consumption indicators Waste heat recovery/ reuse Land use Nice to have Eco toxicity (including incidence on biodiversity, water pollution…) (preferable Human toxicity (including air pollution) Eutrophication (terrestrial, freshwater, marine) Note: this table was adopted from the website https://joint-research-centre.ec.europa.eu/scientific-activities-z/green-and-sustainable-telecom- networks/sustainability-indicators-telecom-networks_en 51 GSMA ESG Metrics for Mobile: Realising value for society through common industry KPIs, June 2024 52 European Commission, Joint Research Centre, Baldini, G., Cerutti, I. and Chountala, C., Identifying common indicators for measuring the environmental footprint of elec- tronic communications networks (ECNs) for the provision of electronic communications services (ECSs), Publications Office of the European Union, Luxembourg, 2024, https://data.europa.eu/doi/10.2760/093662, JRC136475. 48 Each measure has its pros and cons, so the exercise of Compared to other parts of the digital value chain, telecom selection becomes a question of balance. In the case of operators have been better at reporting emissions. The a mobile network operator, energy efficiency can also be GHG Protocol currently classifies performance indicators interpreted at different levels. Different metrics can be for measuring GHGs in three different categories: scope 1, more suitable, depending on if the focus is on one piece of direct emissions; scope 2, purchased electricity; and scope 3, equipment, a site, the whole network or even the entire indirect emissions from the value chain.54 Operators widely operation of a mobile operator.53 Based on the standard of use this classification to publish data on emissions, though ITU and ETSI, mobile network data energy efficiency is the not all report on scope 3.55 ratio between the data volume and the energy consumption during the same period, expressed in bit/J (see relevant standards in annex A, table 2). 53 More details to find on this in GSMA A blueprint for green networks, October 2022. As an example, Orange operates is tasked with implementing an effective, con- text-specific action plan. To drive this programme, Orange has set up a high-level energy dashboard comprising four KPIs: 1) an economic KPI (ENOV), which is a ratio of IT and energy opex over revenue, where a decrease of one point in this KPI results in a corresponding one-point loss in EBITDA; 2) technical KPIs about RAN efficiency (RAN kWh/Gb and RAN/kWh/site); 3) the power usage effectiveness (PUE); 4) ecological efficiency, measured through the renewable energy ratio (RER). All these KPIs are driven by energy consumption. Energy type directly impacts emissions, while procurement determines energy cost. Hence, the energy sourcing strategy is vital. 54 Scope 1, 2, and 3 are commonly accepted designations to categorize the source of greenhouse emissions. Scope 1 emissions refer to emissions directly generated by firms in service delivery. Scope 2 emissions refer to emissions generated from consumption of electricity. Scope 3 emissions include all other emissions associated by a firm’s activity, including across its supply chain. Source: GHG Protocol. na. Standards and Guidance 55 World Bank; ITU. 2024. Measuring the Emissions and Energy Footprint of the ICT Sector: Implications for Climate Action. © World Bank and International Telecommuni- cation Union. http://hdl.handle.net/10986/41238 License: CC BY-NC-SA 3.0 IGO 49 Procurement 6 To effectively integrate greening standards and The public party should promote sustainable ‘green’ requirements into the design of digital infrastructure, procurement practices that prioritize climate-resilient it is crucial to incorporate tailored specifications and solution, the use of renewable energy sources, energy- performance criteria within procurement documents and efficient technologies, and environmentally conscious contracts. These specifications should be developed on a materials in the procurement process. This includes case-by-case basis, taking into account the specific exposure integrating green considerations in procurement as part of network routes and the vulnerabilities of assets to of technical specifications or award criteria. Much of the relevant hazards, as well as other greening considerations, existing telecom guidance on procurement has limited focus such as energy access. on greening, particularly when it comes to climate risks and resilience. The following table and checklist are adopted from procurement practices and guidance on resilience in Engaging the private sector offers a valuable opportunity other infrastructure sectors, as well as broader sustainable to attract additional investment while simultaneously procurement guidance.56 improving climate resilience, provided that the appropriate conditions are established. To fully leverage these advantages, it is essential to develop clear guidelines for incorporating climate considerations into performance requirements, risk allocation, and financial support. This clarity must be maintained throughout the tendering process and reflected in the contracts established between public and private entities. By doing so, we can ensure that greening standards are effectively integrated, fostering sustainable development in digital infrastructure. 56 Examples of guidance for sustainable procurement practices include ASCE 2020. Sustainable Procurement for Infrastructure; BS ISO 20400:2017; Global Center for Ad- aptation 2021. Climate-Resilient Infrastructure Officer Handbook; UNEP 2020. Global Review of Sustainable Public Procurement; and World Bank 2019. Lifelines: The Resilient Infrastructure Opportunity. The website https://infrastructure-pathways.org provides an excellent resource and easy-to-navigate guidance on climate-resilient infrastructure. 50 Table 6.1: Procurement approaches Traditional Effects of Climate Change New Tools and Approaches Approaches Traditional piecemeal, Procurement for isolated projects not enabling Long-term, systemic approach to procurement. Shift project-by-project consideration of climate changes (CC) systemic focus from short-term CAPEX savings to OPEX benefits approach, with a impacts. CC is a long-term issue, and resilience over the project’s full life. Network criticality and risk view to short-term and other greening measures a proactive assessments can aid in the prioritization. cost savings. Tend to adaptive process not achieved by status quo. maintain status quo, thus challenging long- term transformation and improvement. Requirements For uncertain climate futures, inputs can be Include definition of climate resilience and climate typically focus technically complicated and not always clear mitigation actions and required contractor minimum on prescriptive what ensures long-term continuity of service experience/ qualifications in procurement/contract inputs. And seek and delivery of value. documents. qualifications and experience in Contractors may have little or no prior Consider applying: 1) Appropriate rating systems delivering similar experience even if they have extensive and evaluation approaches to reflect resilience and projects, which rarely experience delivering traditional infrastructure. other greening value in addition to traditional costs; 2) include climate Outcome-based procurement approach to deal with resilience and Supplier evaluation based on costs may fail to uncertainty surrounding climate change; 3) Resilience climate mitigation recognize the long-term value in the face of and other greening objectives and performance-based considerations. climate change. measurements to help ensure parties are incentivized to deliver. Consider involving climate risk and climate mitigation experts in design and procurement teams. Sizeable pool The market in smaller or less developed New tools and approaches may help build the capacity of competent economies often has limited capacity to deliver of the market to deliver climate resilience projects and suppliers used to climate resilience and other greening objectives encourage cooperation between suppliers. ensure equitable leading to a limited pool of possible (and willing) and competitive suppliers, reducing competition. The use of transparent risk sharing strategies through procurement in well– procurement can help reassure involved parties that established markets. Uncertainties around climate change may lead climate risks can be managed. to climate resilience and climate mitigation projects being perceived as higher risk, further limiting the number of willing suppliers. Procurement is typically undertaken through a piecemeal, project-by-project approach. Typically, this is performed as required for the project and on relatively short timescales with a view to short-term cost savings. This can result in operators becoming “locked-in” to procurement that maintains the status quo, making long-term transformation and improvement challenging. Procurement is often seen as a necessary step to enable the delivery and use of infrastructure projects rather than an integral tool to deliver long-term strategies and goals. Adopting a long-term, systemic approach to procurement of infrastructure projects can help create the right environment for the delivery of climate mitigation and resilience objective. One of the primary goals of this is to shift focus from short-term capital expenditure savings to a view of holistic value and benefits over the project’s full life. To achieve resilience and other infrastructure greening objectives, it is important to have a risk analysis early in the project preparation and design phase, where the exposure and vulnerability of relevant climate hazards are evaluated for existing and planned infrastructure. However, as this analysis is often not very detailed, there is typically a need to repeat it in greater detail for specific sections of the infrastructure and the assets being implemented under each contract. Therefore, this should be carried out via a detailed technical feasibility study or alternatively by the responsible contract holder. Procurement documents should include a clear definition of greening objectives and criterion. To promote transparency in the evaluation process, the Request for Proposal (RFP) should include a definition or explanation for each evaluation criterion, outlining the factors that will be considered in the overall scoring. Proposals that do not plan any resilience and/or mitigation measure or that do not align with any applicable standards or specifications can be excluded from the tendering process. 51 Establishing relevant qualification standards for bidders is To mitigate these risks, employing procurement methods an efficient way to enhance climate resilience and other that facilitate price discovery is crucial, as suggested in the greening objectives. This can be achieved by implementing report “Leveraging Private Sector Investment in Digital bid criteria with weightings that consider factors such as Communications Infrastructure in Eastern Africa.”57 By the quality of approach, greening approach, monitoring promoting competition among private suppliers seeking plan, and personnel. For resilient and green infrastructure government subsidies for digital infrastructure investments, projects, it is advisable that technical criteria are evaluated these procurement methods can enhance value for money before financial criteria and scores are assigned to limit the through effective price discovery. This approach ensures that negative impact in the scoring undergone by green projects public funds are used optimally, aligning costs with benefits that occur to be more expensive than those with little or no and fostering a more efficient allocation of resources. green objectives. Some qualification standards to include Auctions are a well-established method for uncovering when evaluating bidders for a project in the digital sector valuations from both buyers (in forward auctions) and are bidders’ experience in green infrastructure projects; sellers (in reverse auctions), particularly in situations where technical expertise in climate change, energy management, these valuations are not transparent, and market prices are and digital infrastructure; proposed use of resilient and not readily available as benchmarks. Reverse auctions, if low-carbon (innovative) technologies. Another relevant properly designed, can be particularly effective in achieving qualification could be collaboration with relevant industry value for money. However, as previously stated, it is essential associations and government agencies to foster innovation that resilience and greening requirements are integrated and exchange of best practices. into the evaluation and comparison of value across bidders to ensure that these considerations are reflected in the Prospective private partners seeking public financing may procurement process. exaggerate deliverables and propose unrealistic terms in their tender responses to secure initial funding. This poses Performance monitoring is a crucial aspect of the contract a risk if private bidders anticipate the possibility of holding execution phase. The specific modalities of performance up the public party and renegotiating later. To mitigate this monitoring will typically be detailed in the contract and risk, it is crucial for projects to be well-defined, ensuring may vary from one project to another. The responsibility for that private entities submit sufficient bids and that the monitoring and reporting on KPIs can be shared between bids received are genuinely comparable, meaning they the public and private parties, depending on the specific are not based on disparate underlying assumptions. This terms of the contract and the agreements made. However, comparability allows for a more transparent and equitable as the private party is often responsible for the day-to-day evaluation process. implementation of the project (collecting data on operational performance, service availability, service quality, etc.), they Public finance encounters significant risks when seeking are usually better placed to regularly report this data in private investment in digital infrastructure projects, accordance with the contract provisions (i.e. KPI collection). primarily due to asymmetric information regarding funding While the public party often lacks expertise, time or budget gaps. These gaps refer to the additional funds required to to take care of this monitoring, it must review the reports deploy infrastructure in areas that are not commercially provided by the private party (i.e. KPI monitoring), and – viable, while also implementing standards for infrastructure possibly helped by consultants – may conduct audits every resilience and other greening measures. A major challenge 3 or 5 years depending on the duration of the PPP and the for public sector entities providing finance is their lack of lifespan of the asset. precise knowledge about the required gap funding, which exposes them to the risks of overpayment and adverse selection. This uncertainty can undermine efforts to achieve value for money. 57 World Bank and CEPA. 2024. Leveraging Private Sector Investment in Digital Communications Infrastructure in Eastern Africa. Washington, DC: World Bank. 52 Key recommendations 7 The analysis presented in this report highlights the urgent • Incorporate energy efficiency guidelines and renewable need to integrate climate considerations into the planning, energy requirements into all project tenders and design, and implementation of digital infrastructure technical specifications, and monitor compliance projects across Djibouti, Ethiopia, Kenya, Madagascar, throughout project lifecycles. Somalia, and South Sudan. To ensure the long-term viability and sustainability of digital investments, the following key Strengthening E-Waste Management and Circular Economy recommendations are put forward: Practices Systematic Integration of Greening Objectives into Policy • Develop and enforce policies for responsible e-waste and Implementation Frameworks management, including recycling, safe disposal, and circular economy practices for digital equipment and • Ensure that climate adaptation and mitigation measures infrastructure components1. are systematically embedded in national digital • Build capacity among public and private stakeholders strategies, sector policies, regulatory frameworks to implement e-waste protocols and adopt green ICT and procurement processes. This includes integrating standards. greening requirements into procurement documents such as TORs, RFPs, and contracts to guarantee that Enhanced Private Sector Engagement and Incentivization resilience and energy efficiency are prioritized from project inception through implementation. • Mobilize private sector investment and expertise by • Align digital infrastructure investments with national designing financial and contractual incentives that and regional climate action plans, leveraging frameworks offset higher upfront costs of green infrastructure (e.g., developed or under development by the EAC and AU to targeted subsidies, risk-sharing mechanisms, public- harmonize standards and share best practices. private partnerships). • Encourage private ownership models and operational Prioritization of Climate-Resilient Infrastructure partnerships to leverage efficiency and innovation in Investments project delivery and maintenance. • Direct funding and technical assistance toward climate- Capacity Building and Knowledge Sharing resilient infrastructure, focusing on the most vulnerable assets such as fiber optic cables, cell towers, and data • Invest in targeted training for government officials, centers. Use geospatial risk analysis to inform site regulators, and private sector actors on climate risk selection, design, and construction, ensuring that assessment, resilience-building, energy efficiency, and infrastructure is sited and built to withstand local green procurement practices. hazard profiles (e.g., floods, landslides, storms). • Foster regional knowledge exchange and collaboration • Adopt robust construction practices, including climate- to disseminate lessons learned, innovative solutions, proofing for underground and aerial cables, resilient and best practices across countries and projects. tower design, and disaster recovery planning for data centers1. Monitoring, Evaluation, and Adaptive Management Acceleration of Energy Efficiency and Renewable Energy • Establish robust monitoring and evaluation systems Adoption to track the effectiveness of greening interventions, measure progress toward climate and sustainability • Promote the use of energy-efficient technologies goals, and enable adaptive management in response to (e.g., fiber optics over copper, low-emission network emerging risks and lessons learned. equipment) and renewable energy solutions (solar, wind, • Use standardized metrics for resilience and mitigation battery storage) for powering digital infrastructure, outcomes to facilitate cross-country benchmarking and especially in rural and off-grid areas. reporting. 53 Emergency Preparedness and Disaster Response Strengthen Regional and International Collaboration Integration • Harmonize standards and guidelines: Support regional • Integrate digital infrastructure resilience planning and continental organizations (e.g., EAC, IGAD, AU) in into national and regional disaster risk management establishing cross-country frameworks for greening frameworks to ensure continuity of critical services digital infrastructure and harmonizing standards and during and after climate-related events. guidelines. • Develop contingency plans, redundant network designs, • Knowledge sharing: Facilitate the exchange of best and rapid response protocols for infrastructure recovery practices and lessons learned among countries in the following extreme weather or natural disasters. region and with international partners, e.g. regional working group or workshops. • Access climate finance: Explore the options to tap into international climate finance mechanisms to support investments in climate-resilient and green digital infrastructure. 54 Annexes 55 Annex A: Country case - Djibouti The Republic of Djibouti, strategically located in the Horn of Africa at the southern entrance to the Red Sea, is one of Africa’s smallest nations, sharing borders with Eritrea, Ethiopia, and Somalia. Covering over 23,000 km² with a 372 km coastline, Djibouti is predominantly arid, with nearly 90% of its land classified as desert. The country faces significant environmental challenges, including recurrent droughts and floods, which have impacted its population and economy over the years.58 Despite being a country characterized by its arid climate and limited water resources, the risk of terrestrial and coastal flooding poses a significant risk to physical infrastructure, including telecommunication networks and assets. Climate change will likely increase these risks, due to projected increase in intensity of heavy rainfall events, storms and sea level.59 Fig A1. Projected Change in Temperature and Precipitation in Djibouti Projected temperature Projected precipitation Source: World Bank, Climate Change Knowledge Portal (2025). URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: 14 April 2025 Despite these challenges, Djibouti is steadily advancing its digital infrastructure as part of its broader economic transformation goals, leveraging its strategic location to position itself as a regional digital hub. The country hosts nine submarine data cables and multiple cable landing stations, connecting it to Europe, the Middle East, East Africa, and South Asia. This robust connectivity enhances the reliability of internet and data services within Djibouti and across the Horn of Africa. A milestone achievement in this journey was the establishment of the Djibouti Data Center in 2013, the first Tier-3 data center in East Africa, which significantly bolstered regional ICT capacity. 58 https://climateknowledgeportal.worldbank.org 59 The World Bank Climate Change and Development Report (CCDR) for Djibouti reveals significant climate challenges. Over the past 50 years, average temperatures in the country have risen by approximately 1°C. Projections under the CCDR’s "hot" scenario indicate that by 2050, the number of high heat index days could increase from 66 to 123 annually. While precipitation predictions remain uncertain, models consistently show that extreme wet events will become more frequent. For example, rainfall events that historically occurred every decade are expected to happen every six years by mid-century. Additionally, droughts have become more frequent and severe, exacerbating water scarcity issues. Rising sea levels pose a serious threat to Djibouti’s coastal areas, which are vital for its economy. These changes could undermine live- lihoods, disrupt infrastructure, and damage critical economic hubs concentrated in low-lying regions like Djibouti City. Without urgent adaptation measures, climate-re- lated damages could cost the country up to 6% of its GDP annually by 2050. However, strategic investments in resilience, infrastructure development, and renewable energy can help mitigate these risks and support sustainable growth. Source: World Bank Group. 2024. Djibouti Country Climate and Development Report. CCDR Series. http://hdl.handle.net/10986/42439 56 Domestically, Djibouti has developed a strong terrestrial backbone network concentrated in the southern regions, linking urban centers and extending into Ethiopia. However, broadband penetration remains low due to high costs and limited infrastructure. While mobile internet usage is growing, it is constrained by affordability issues and inadequate rural coverage. Fixed internet services are underdeveloped, hindered by a lack of competition in the state-controlled telecommunications sector.60 Compared to neighboring Ethiopia and Kenya, Djibouti’s mobile penetration rate is lower, partly due to its monopoly on telecommunications services. In contrast, Kenya’s liberalized market fosters greater competition and higher penetration rates. The government has prioritized ICT development under its ambitious “Djibouti Vision 2035” strategy, focusing on inclusion, connectivity, and regional integration. Public institutions have made strides in digitizing services, though many platforms remain fragmented and lack interoperability. Regional integration efforts include participation in initiatives like the Horn of Africa program for connectivity and e-commerce markets. International support has also played a role in Djibouti’s digital transformation. For instance, the World Bank has funded projects such as the Djibouti Digital Foundations Project and the East Africa Regional Digital Integration Project (EARDIP). These initiatives allocate resources to ICT services, public administration, and infrastructure while promoting competition and private-sector investment in the digital economy. Despite challenges such as affordability and accessibility gaps, Djibouti continues to make progress toward becoming a key player in regional digital connectivity and economic growth. Fig A2-3: Mobile cell locations and fiber optic networks routes 60 The ICT sector in Djibouti is one of the final remaining monopolies in the global telecommunications sector, the other in neighbouring Eritrea. Djibouti Telecom is the sole provider of mobile and fixed services in the country, including mobile 2G, 3G and 4G technologies, fixed-line voice, internet and data services. 57 Risk analysis for Djibouti’s digital infrastructure Flood risks to digital infrastructure Below are the results from an exposure and vulnerability analysis, which shows flood-prone locations in Djibouti compared to site locations of fiber network routes and cell sites, as well as the estimated damage costs from a sever flood hazard incident on the two different types of digital infrastructure. The map displays areas at risk of flooding in shades of gray, showing vast areas in the central and northeastern regions of the country that are highly exposed. However, these areas do not yet have a significant presence of digital infrastructure. Compared to this, the digital infrastructure in the capital city and an area near the Sudan border are at risk of being affected by floods. Fig A4: Projected riverine flooding, RCP4.5 climate scenario, severe flooding event The figures below show the estimated direct damage costs from flood impacts under 2 different climate scenarios and flood intensities. As can be seen, over 37 percent of cell sites and more than 13 percent of the fiber network are estimated to be impacted. The most extensive damage occurs at cell sites, with damage costs of roughly $3 million, while the damage to the fiber network is more modest in economic terms. Fig A5-A8: Estimate riverine flooding damage and damage costs Flood risk: Mobile cells Flood risk: Fiber network 58 In the 0.1% probability category (representing high- In the 0.1% probability category (representing high- intensity, low-frequency flooding events) for estimated intensity, low-frequency flooding events) for estimated riverine flooding damage costs to mobile cells in Djibouti, riverine flooding damage costs to fiber optic routes in the results are as follows: Djibouti, the results are as follows: Direct Damage Costs: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is 3.3 • Under RCP4.5, the estimated direct damage cost is million USD. 0.04 million USD. • Under RCP8.5, the estimated direct damage cost is • Under RCP8.5, the estimated direct damage cost is slightly lower, at 2.9 million USD. slightly lower, at 0.04 million USD. Landslide risk to digital infrastructure On the map below, areas with land slide risk and the digital infrastructure, respectively cell sites and fiber network, have been overlayed. The image below contains two bar charts representing the estimated landslide exposure for mobile cell and fiber network assets in Djibouti. The data is broken down by landslide risk categories: No Risk, Low Risk, Medium Risk, and High Risk. The mobile cell infrastructure in Djibouti is predominantly located in safe zones, with a minimal proportion (5.9%) in areas exposed to landslide risk. This implies that landslides pose a relatively low threat to mobile cell infrastructure, although a small fraction of assets remains vulnerable, especially those in high-risk zones. Fig A9: Landslide risk exposure for medium and high-risk areas 59 The fiber network in Djibouti has considerable exposure to landslide risk, with over 16.8 percent of fiber assets in medium- risk and high-risk areas. This suggests a substantial vulnerability, as landslide events could significantly disrupt the fiber network. Fig A10: Estimated landslide exposure for mobile cells and fiber assets in Djibouti, reported by risk category • No Risk: The vast majority (94.1%) of mobile cell • No Risk and Low Risk: There majority of fiber network assets are assets are located in no-risk areas, indicating that in no-risk or low-risk areas. they are generally safe from landslide exposure. • Medium Risk: A significant portion, 11.4%, of fiber network • Low Risk: 1.6% of mobile cell assets are in low-risk assets are situated in medium-risk areas. areas. • High Risk: 5.4% of fiber assets are in high-risk zones. • Medium Risk: 1.2% of mobile cells fall within medium-risk zones. • High Risk: 3.1% of mobile cell assets are situated in high-risk areas. 60 Other hazard risks Based on the available geospatial data on hurricanes and coastal flooding, it is not possible to determine the level of risk these hazards may pose to digital infrastructure. From other sources, it can be deduced that hurricanes occurring further south off Somalia and Yemen have previously caused derived impacts in Djibouti, especially in the capital, where the associated low-pressure system have delivered heavy rainfall, while the wind has given rise to storm surges, which have caused flooding in coastal areas.61 Approximately 88% of Djibouti’s population lives in coastal areas, making them and the digital services they rely upon somewhat vulnerable to flooding. This applies in particular to submarine cables and landing stations, as well as coastal network routes and facilities. The country’s low-lying coastal regions are at risk from sea-level rise and storm surges associated with cyclones, which are expected to intensify due to climate change. Fig A11: Historical cyclone tracks for Djibouti Fig A12: Projected coastal flooding, RCP8.5 climate scenario, 2080 Lastly, Djibouti is located in an active seismic zone, which has implications for its infrastructure. Efforts to expand fiber optic networks and other digital infrastructure in Djibouti must account for geological risks. Djibouti is a crucial hub for submarine fiber optic cables that connect Africa to global internet networks. Seismic events, particularly those occurring on the seabed, can lead to cable faults that will not only affect Djibouti, but other countries in East Africa. 61 https://reliefweb.int/disaster/tc-2018-000059-som 61 Annex B: Country case - Ethiopia Ethiopia, located in the Horn of Africa, is the second most populous country on the continent, with over 128 million residents as of 2023. The country spans diverse climatic zones, ranging from tropical savannah in the west to desert and semi-arid regions in the north and east. Ethiopia’s varied topography and elevation profiles contribute to its complex climate conditions, which include warm, temperate, and humid climates in the highlands, to dry and desert-like conditions in the northeastern and eastern lowlands. Despite being one of the fastest-growing economies in sub-Saharan Africa, Ethiopia faces significant environmental and developmental challenges exacerbated by climate change. These include recurrent droughts, erratic precipitation patterns, flooding, and landslides. Over the past four decades, the country has experienced severe droughts—the most recent being the worst in 40 years—and devastating floods that have impacted infrastructure and livelihoods. Climate projections indicate further warming and increased rainfall variability, which are expected to intensify these hazards. The government has implemented strategies such as the Climate-Resilient Green Economy (CRGE) initiative to address these challenges. However, adaptation measures need further enhancement to protect infrastructure assets and vulnerable populations from worsening climate shocks.62 Fig B1: Projected change in temperature and precipitation, Ethiopia Projected temperature Projected precipitation Source: World Bank, Climate Change Knowledge Portal (2025).URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: 14 April, 2025 Ethiopia has made notable strides in developing its digital infrastructure, despite facing significant environmental and socioeconomic challenges. Central to this progress is the Digital Ethiopia 2025 Strategy, which aims to transform the country’s economy by integrating digital technologies across all sectors and ensuring equitable access to digital services for all citizens. Key initiatives under this strategy include the establishment of data centers, IT parks, and the digitization of public services. The country has achieved substantial improvements in digital connectivity, marked by the rollout of 4G networks and the introduction of Safaricom Ethiopia as a second telecommunications operator. Internet access has expanded significantly, with over 40 million users and 75 million mobile subscribers recorded by 2024. Ethiopia also hosts advanced data centers, such as Raxio’s Tier III-certified facility launched in 2022, which enhances ICT capacity and reliability. Additionally, an IT Park near Addis Ababa has been developed to attract ICT outsourcing firms and IT equipment manufacturers. These efforts aim to modernize infrastructure and boost private sector involvement. 62 World Bank Group. 2024. Ethiopia Country Climate and Development Report. CCDR Series. http://hdl.handle.net/10986/41114 62 However, challenges persist. Broadband penetration remains limited, particularly in rural areas where infrastructure is insufficiently tailored to local needs. International connectivity relies heavily on cross-border links to Djibouti and Kenya. Despite the liberalization of Ethiopia’s telecom sector in 2021 to foster competition, issues such as affordability and limited infrastructure continue to hinder widespread adoption. Although Safaricom Ethiopia’s entry has improved market competition slightly, extending connectivity beyond urban areas and strengthening both national and cross-border fiber networks remain critical priorities. In comparison to neighboring Kenya’s more mature liberalized telecom market, Ethiopia’s services are still evolving toward greater accessibility. Fig B 2-3: Mobile cell site and fiber optic network locations 63 Climate risks to digital infrastructure Flood risks to digital infrastructure Below are the results from an exposure and vulnerability analysis, which shows flood-prone locations in Ethiopia compared to site locations of fiber network routes and cell sites, as well as the estimated damage costs from a sever flood hazard incident on the two different types of digital infrastructure. The map shows areas at risk of flooding around major rivers that flow from the Ethiopian highlands. It reveals multiple areas where the risk of flooding overlaps with fiber network routes, demonstrating the considerable risk present in various parts of the country. This can have an impact on connectivity, affecting both urban business centers and isolated rural communities. Compared to fiber, the available data indicate that cell sites are significantly less exposed to direct flood impacts. Fig B4: Projected River flooding, RCP4.5 climate scenario, severe flood event 64 The figures below show the estimated direct damage costs from flood impacts under 2 different climate scenarios and flood intensities. As can be seen, around 1 percent of cell sites and 5 percent of the fiber network are estimated to be impacted. The most extensive damage relates to fiber, with estimated damage costs of between 3-4 million USD, while the damage to the cell sites network ranges from 2.1-2.4 USD. Fig B5-8: Estimate riverine flooding damage and damage costs Flood risk: Mobile cells Flood risk: Fiber network In the 0.1% probability category (representing high-intensity, low- In the 0.1% probability category (representing high-intensity, frequency flooding events) for estimated riverine flooding damage low-frequency flooding events) for estimated riverine flooding costs to mobile cells in Ethiopia, the results are as follows: damage costs to fiber optic routes in Ethiopia, the results are as follows: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is 2.1 million Direct Damage Costs: USD. • Under RCP4.5, the estimated direct damage cost is 3.29 • Under RCP8.5, the estimated direct damage cost is slightly million USD. lower, at 2.4 million USD. • Under RCP8.5, the estimated direct damage cost is slightly • lower, at 3.99 million USD. • 65 The data suggests that under the more extreme climate change scenario (RCP8.5), the costs associated with flooding damage to both mobile cells and fiber optic routes are higher compared to the moderate scenario (RCP4.5). This indicates a potential increase in flooding intensity or frequency as climate change progresses. Incorporating resilience into planning can mitigate future damage costs and enhance the sustainability of critical services. Landslide risk to digital infrastructure On the map below, areas with landslide risk and the digital infrastructure, respectively cell sites and fiber network, have been overlayed. The image below contains two bar charts representing the estimated landslide exposure for mobile cell and fiber network assets in Ethiopia. The data is broken down by landslide risk categories: No Risk, Low Risk, Medium Risk, and High Risk. The mobile cell infrastructure in Ethiopia is largely positioned in areas with minimal landslide risk, with almost 80% of assets classified in no-risk zones. However, a notable fraction is exposed to varying levels of landslide risk, including medium and high-risk areas. This indicates that while the majority of mobile cells are safe, there is a significant proportion that may be vulnerable in the event of landslides. Fig B9: Landslide risk exposure for medium and high-risk areas The fiber network in Ethiopia has a high degree of exposure to landslide risk, with 15.9 % assets situated in either medium or high-risk zones. This distribution highlights a considerable vulnerability, as landslide events could potentially disrupt a large portion of the fiber network infrastructure, impacting connectivity. 66 Fig B10: Estimated landslide exposure for mobile cells and fiber assets, reported by landslide category • No Risk and low risk: The majority of mobile • No Risk and Low Risk: The majority of fiber assets (60.7% and cell assets (79.1% and 12.2%) are in no-risk or 23.4%) are in no-risk or low-risk zones. low risk areas, suggesting they are generally • Medium Risk: A significant proportion of fiber network assets safe from landslide exposure. (8.9%) are located in medium-risk areas. • Medium Risk: 4.6% of mobile cells are in • High Risk: Another 7% of fiber assets are in high-risk areas. medium-risk areas. • High Risk: 4.2% of mobile cells are located in high-risk zones.. Other hazard risks Based on the hazard data on hurricanes, Ethiopia’s digital infrastructure has a low degree of exposure to cyclone winds. This does not imply that storms are not a concern, and therefore local data must be obtained, and wind loads must be considered, especially when constructing mobile towers in areas exposed to strong wind gusts. The main Ethiopian Rift is prone to seismic hazards, which can lead to ground motion that damages roads, bridges, and buildings. A probabilistic seismic hazard assessment indicates that the risk of earthquake-induced ground failures is high in areas with significant infrastructural development.63 Damage resulting from seismic hazards has not generally been recognized as a problem of national importance. However, careful analysis should be given before doing any projects in seismically prone areas. 63 Ayele, A., Woldearegay, K. & Meten, M. A review on the multi-criteria seismic hazard analysis of Ethiopia: with implications of infrastructural development. Geoenviron Disasters 8, 9 (2021). https://doi.org/10.1186/s40677-020-00175-7 67 Annex C: Country case - Kenya Kenya, the region’s largest economy, boasts a diverse climate and landscape, which ranges from high rainfall and humidity in the central highlands and coastal regions to arid conditions in the northern and eastern parts. Kenya faces recurrent climate-related challenges, including prolonged droughts and intense flooding. Climate variability is already a source of significant economic risk for Kenya, with estimates suggesting that more than 70 percent of disasters from natural hazards are attributable to extreme climatic events. The repeating patterns of floods and droughts in the country have had devastating socioeconomic impacts and high economic costs. A range of climate models forecast that Kenya’s climate future will entail the mean temperature increasing, with greater increases in temperature under a scenario with no global decarbonization effort. Precipitation will fluctuate significantly on an annual basis, with all but the most extreme GCMs indicating that Kenya will get wetter. Climate projections indicate that without adaptation measures, Kenya could experience real GDP losses of up to 7% by 2050 under pessimistic scenarios.64 Fig C1: Projected change in temperature and precipitation, Kenya Projected temperature Projected precipitation Source: World Bank, Climate Change Knowledge Portal (2025). URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: 14 April 2025 Kenya has established itself as a digital leader in East Africa, earning recognition for its innovative digital services and widespread connectivity. The country serves as a regional digital hub, supported by advanced broadband connectivity, pioneering solutions like mobile money, a robust tech ecosystem, and a digitally skilled workforce. It has become a testbed for emerging technologies like cloud computing and artificial intelligence. Kenya’s progress stems from early liberalization of the telecom sector, strategic public investments in fiber optic infrastructure, and an active private sector. This progress is rooted in strategic initiatives, including early liberalization of the telecom sector, substantial public investments in fiber optic networks, and active participation from the private sector. Kenya’s connectivity is supported by six undersea cables, mobile broadband coverage reaching over 96% of the population, and more than 20,000 kilometers of fiber optic network crisscrossing the country. However, parts of this infrastructure require upgrades due to prolonged outages in some segments. To address these gaps, the government launched the Digital Superhighway Project, aiming to expand fiber optic coverage nationwide by laying 100,000 kilometers of cable and establishing 25,000 public Wi-Fi hotspots and digital hubs 64 World Bank Group. 2023. Kenya Country Climate and Development Report (CCDR). Source: http://hdl.handle.net/10986/40572 68 Despite these advancements, challenges persist. Disparities in broadband access remain a concern, alongside limited availability of digital public services and gaps in critical digital skills needed for an increasingly digitized economy. The Kenya Digital Economy Acceleration Project (KDEAP), funded by $390 million from the World Bank, focuses on expanding high-speed internet access, digitizing government services, and building digital skills for economic growth. Fig C2-3: Mobile cell and Fiber-optic network locations 69 Climate risks to digital infrastructure Flood risks to digital infrastructure The results of an exposure and vulnerability analysis highlight flood-prone areas in Kenya and their intersection with critical digital infrastructure, including fiber network routes and cell sites. The analysis also estimates potential damage costs from a severe flood hazard affecting these two types of infrastructure. The accompanying map identifies regions at risk of flooding, primarily concentrated around major rivers that flow from the Kenyan highlands into lowland areas, extending toward border regions and the sea. The map reveals several locations where flood risks overlap with fiber network routes, underscoring the significant vulnerability of these networks in various parts of the country. Such disruptions could impact connectivity, affecting both urban economic hubs and remote rural communities. In contrast, the data suggest that cell sites are considerably less exposed to direct impacts from flooding compared to fiber networks. Fig C4: Projected river flooding RCP4.5 climate scenario, severe flood event, 2080 70 In analyzing the economic impacts of riverine flooding in Kenya, it is estimated that 7.9–8.2% of mobile cell sites and 5.6–6.5% of the fiber optic network could be affected by flooding. Among these, fiber optic infrastructure is projected to suffer the highest overall damage costs, ranging between $3–4 million USD. In comparison, damage to mobile cell sites is estimated at $2.1–2.4 million USD. These figures underscore the significant financial implications of climate-induced flooding on critical telecommunications infrastructure in Kenya. Damage costs for both types of assets show a slight increase under the RCP8.5 scenario compared to RCP4.5. The difference in damage costs between RCP4.5 and RCP8.5 highlights the potential impact of climate change on flooding frequency and intensity in Kenya. Also, this underscores the heightened vulnerability of these infrastructures under the climate scenario that is most probable given current emission trends. RCP4.5, which represents a moderate emissions scenario, still leads to substantial damage costs, indicating that even with some mitigation efforts, flooding remains a critical risk. Fig C5-C8: Estimate riverine flooding damage and damage costs, Kenya Flood risk: Mobile cells Flood risk: Fiber network In the 0.1% probability category (representing high-intensity, In the 0.1% probability category (representing high-intensity, low-frequency flooding events) for estimated riverine flooding low-frequency flooding events) for estimated riverine flooding damage costs to mobile cells in Kenya, the results are as damage costs to fiber optic routes in Kenya, the results are as follows: follows: Direct Damage Costs: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is 73.4 • Under RCP4.5, the estimated direct damage cost is 10.15 million USD. million USD. • Under RCP8.5, the estimated direct damage cost is • Under RCP8.5, the estimated direct damage cost is slightly slightly lower, at 79.2 million USD. lower, at 11.23 million USD. 71 Although the financial impact of these estimates is substantial, it is crucial to recognize that damage to infrastructure extends beyond direct repair costs, as it also disrupts vital services, transportation systems, and economic activities. For instance, the loss of mobile and fiber optic connectivity can severely impair communication and access to critical information during emergencies, further intensifying the difficulties faced by affected communities. Moreover, businesses that rely on data transactions and internet access will suffer substantial financial losses from extended shutdowns. Consequently, the reported damage figures reflect a highly conservative estimate. The true cost of an internet shutdown—whether localized or nationwide—would likely be significantly higher. These findings underscore the necessity for proactive measures in infrastructure planning and disaster management in Kenya. Developing climate-resilient infrastructure and implementing effective flood management strategies will be crucial to mitigate future risks and minimize economic losses associated with flooding events. Landslide risk to digital infrastructure The map below illustrates the overlay of landslide risk areas with digital infrastructure, specifically mobile cell sites and fiber networks. It is followed by an image, that includes two bar charts depicting the estimated exposure of mobile cell and fiber network assets in Kenya to landslide risks. The data is categorized into four levels of landslide risk: No Risk, Low Risk, Medium Risk, and High Risk. Fig C9: Landslide risk exposure for medium and high-risk areas 72 This distribution suggests that mobile cell infrastructure in Kenya is largely secure from landslide risks, with over 90% of assets located in no-risk areas. The small percentages in low, medium, and high-risk zones indicate minimal exposure to landslide vulnerability for mobile cells in Kenya. The fiber network in Kenya shows some exposure to landslide risk, with 4.7% of fiber assets in medium-and high-risk zones and highlights a vulnerability that could lead to significant service disruptions if landslides occur in these areas. Fig C10: Estimated landslide exposure for mobile cells and fiber assets in Kenya, reported by risk category • No Risk and Low Risk: A vast majority of mobile • No Risk and Low Risk: The majority of fiber assets (86.9% and 8.4%) cell assets (90.5% and 7.3%) are located in areas are located in areas with no risk or low risk for landslides. classified as having no or low landslide risk. • Medium Risk: A small portion of fiber network assets, 2.9%, are in • Medium Risk: 1.4% of mobile cells are in areas of medium-risk areas for landslides. medium risk. • High Risk: Another 1.8% of fiber network assets are in high-risk • High Risk: Only 0.8% of mobile cells are exposed to zones. high-risk zones. Other hazard risks Coastal flooding in Kenya, driven by storm surges and rising sea levels, poses a low to moderate risk to the country’s digital infrastructure. However, urban areas such as Mombasa face heightened vulnerability due to the concentration of critical facilities like submarine cable landing stations, some of which are situated near the shoreline. Although specific data on the combined flood exposure of these landing stations is unavailable, the combined effect of coastal flooding, storm surges and erosion could potentially disrupt their operations. While these facilities are generally designed to endure environmental challenges, the escalating impacts of climate change—such as rising sea levels and intensified storms—may increase future risks. In the short term, the probability of significant hazard impacts remains relatively low. Nevertheless, given the potentially severe consequences of disruptions or damage to these infrastructures, it is imperative to conduct comprehensive risk assessments and implement resilience measures to safeguard them against emerging climate threats 65 65 The recent damage to Seacom and EASSY cables led to significant losses for Kenyan businesses, impacting internet services, international voice calls, and overall business operations. See for example and news report responding to cable cuts in early 2024, e.g. https://businesstoday.co.ke/undersea-fibre-cable-cuts-cause-losses-for-ken- yan-businesses/ news report responding to cable cuts in early 2024, e.g. https://businesstoday.co.ke/undersea-fibre-cable-cuts-cause-losses-for-kenyan-businesses/. Further documentation of the economic importance of submarine cables is described here: RTI 2020 “Economic Impacts of Submarine Fiber Optic Cables and Broadband Connectivity in Kenya” https://www.rti.org/publication/economic-impacts-submarine-fiber-optic-cables-and-broadband-connectivity-kenya/fulltext.pdf and 73 Fig C11: Projected coastal flooding RCP4.5 climate scenario, severe flood event, 2080 Like neighboring countries affected by the Rift Valley’s tectonic movements, there is a recognized risk—ranging from small to moderate—that earthquakes and associated landslides could impact segments of the digital infrastructure, particularly those located in and around the Rift Valley. It is crucial to carefully assess this risk during the site selection, design, and construction phases of data centers, mobile towers, and other facilities. Moreover, technological advancements can be utilized to effectively monitor these risks. For example, integrating sensors with fiber optic cables can provide real-time seismic activity data, thereby enhancing the resilience of our infrastructure. 74 Annex D: Country case - Madagascar Madagascar, the fourth-largest island in the world, has a climate shaped by its rugged terrain and the influence of trade winds and ocean currents from the Indian Ocean. The island is highly vulnerable to cyclones, experiencing three to five storms annually—the highest exposure in Africa. These cyclones often cause severe damage to infrastructure and agriculture, particularly during the rainy season. On average, a typical cyclone reduces the country’s GDP by about 1% annually, while rare, high-impact cyclones can lead to losses of up to 8% of GDP. The effects of climate shocks are unevenly distributed across Madagascar. The southern and southwestern regions are most affected by severe droughts and water stress, while the northern and eastern areas face the highest exposure to tropical cyclones. This uneven distribution highlights the need for targeted disaster risk reduction strategies. Madagascar’s latest climate modeling predicts continued warming, increased rainfall variability, and more intense but less frequent cyclones, with widespread impacts across all sectors of society.66 Future precipitation patterns are more uncertain and vary by region and scenario. Madagascar’s vulnerability to both rapid and slow- onset climate shocks underscore the importance of climate resilience for its development agenda. Fig D1: Projected temperature and precipitation Projected temperature Projected precipitation Source: World Bank, Climate Change Knowledge Portal (2025). URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: 14 April 2025 Madagascar’s digital infrastructure has undergone significant development in recent years, aiming to bridge the digital divide, modernize governance, and expand access to technology. Madagascar is relatively well-connected with high mobile network population coverage and four submarine cable links. As of early 2025, Madagascar had 18.2 million mobile connections (56.2% of the population), with 84.6% classified as broadband connections. The telecoms market is competitive and comprises of four operators. However, rural areas are still largely unconnected, which is partly explained by the fact that only around a quarter of households are supplied with electricity. 66 World Bank Group. 2024. Madagascar Country Climate and Development Report (CCDR). http://hdl.handle.net/10986/42263 75 In April 2023, the World Bank approved a $400 million credit for the DECIM project, which aims to double energy access and add 3.4 million new internet users. The project targets both digital and energy sectors, focusing on underserved communities, and will connect around 2,000 health centers and schools to renewable energy and digital services. DECIM’s digital components include expanding broadband infrastructure, increasing internet penetration, and supporting the digitalization of public services. Through the DECIM project, the World Bank aims to increase electricity access from 33.7% to 67%, reaching at least 10 million people, including 2 million households and over 150 villages. Figure D2-3: Mobile cell site and fiber-optic network locations Climate risks to digital infrastructure Flood risks to digital infrastructure The results of an exposure and vulnerability analysis highlight the flood-prone areas in Madagascar, comparing them to the locations of fiber network routes and cell sites. The analysis also estimates the potential damage costs from severe flood hazards on these digital infrastructures. The map illustrates areas vulnerable to flooding, primarily along major rivers that flow from the central highlands to the coastal lowlands and the sea. Notably, several regions show an overlap between flood risk zones and fiber network routes, indicating significant exposure across various parts of the country. This poses a threat to connectivity, affecting both urban economic hubs and rural communities. In contrast, the available data suggests that cell sites are substantially less exposed to direct flood impacts compared to fiber networks. 76 Fig D4: Projected river flooding RCP4.5 climate scenario, severe flood event 77 The projected impacts of flooding on telecommunications infrastructure are significant. It is estimated that between 4% and 5.5% of cell sites, and 8.1% to 9.4% of the fiber network, could be affected by severe flood events under different climate scenarios. Cell sites are expected to sustain the greatest financial losses, with estimated damage ranging from $12 million to $21 million. In comparison, damages to the fiber network are projected to be between $1.1 million and $3.3 million. Under the higher emissions scenario (RCP8.5), the estimated costs for both mobile cell sites and fiber optic routes are slightly lower than those under the RCP4.5 scenario. This may be due to some climate models predicting a decrease in rainfall, which would reduce the overall exposure to flooding. However, it is important to acknowledge the uncertainties inherent in climate models, particularly regarding local rainfall variability and the frequency of extreme events. Given that storms are likely to become more intense as sea-surface temperatures rise and atmospheric instability increases, the risk of extreme rainfall events may also grow. Therefore, it is essential to take precautionary measures to protect telecommunications infrastructure, regardless of which climate scenario is used as a reference. Fig D5-D8: Estimate riverine flooding damage and damage costs, Madagascar Flood risk: Mobile cells Flood risk: Fiber network 78 In the 0.1% probability category (representing high-intensity, low- In the 0.1% probability category (representing high-intensity, frequency flooding events) for estimated riverine flooding damage low-frequency flooding events) for estimated riverine flooding costs to mobile cells in Madagascar, the results are as follows: damage costs to fiber optic routes in Madagascar, the results are as follows: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is 21.6 Direct Damage Costs: million USD. • Under RCP4.5, the estimated direct damage cost is 3.34 • Under RCP8.5, the estimated direct damage cost is slightly million USD. lower, at 20.5 million USD. • Under RCP8.5, the estimated direct damage cost is slightly lower, at 3.18 million USD. Landslide risk to digital infrastructure Fig D9: Landslide risk exposure for medium and high-risk areas The map below illustrates the overlay of landslide risk zones and digital infrastructure, specifically mobile cell sites and fiber networks. The accompanying image features two bar charts that depict the estimated exposure of mobile cell and fiber network assets to landslide risks in Madagascar. These risks are categorized into four levels: No Risk, Low Risk, Medium Risk, and High Risk. The analysis reveals that mobile cell infrastructure in Madagascar is predominantly situated in areas with no or low landslide risk. However, approximately 10.4% of mobile assets are located in medium and high-risk zones, making them potentially vulnerable during landslide events. This distribution suggests that while most mobile infrastructure is relatively secure, a significant portion remains exposed to elevated risks. In contrast, the fiber network exhibits a higher level of exposure to landslide risks. Nearly 14% of fiber assets are located in medium and high-risk zones, underscoring their susceptibility to damage during landslide occurrences. This pattern highlights the critical vulnerability of Madagascar’s fiber infrastructure to natural disasters, emphasizing the need for targeted risk mitigation strategies to ensure network resilience. 79 Fig D10: Estimated landslide exposure for mobile cells and fiber assets in Madagascar, reportet by risk category • No Risk: The majority of mobile cell assets (66.6%) are in no- • No Risk: There are 44.7% of fiber assets in areas risk areas, meaning they are generally safe from landslide categorized as no risk exposure. • Low Risk: There are 41.3% of fiber assets in low-risk areas. • Low Risk: 23% of mobile cell assets are classified as being in • Medium Risk: A significant portion, 9.6%, of fiber network low-risk areas. assets are located in medium-risk areas, indicating a • Medium Risk: 8.4% of mobile cells fall into the medium-risk moderate exposure to landslides. category. • High Risk: 4.4% of fiber network assets are in high-risk • High Risk: A small percentage, 2%, of mobile cell assets are areas, representing a minor proportion of assets with located in high-risk areas. substantial vulnerability. 80 Storm and cyclone hazards Madagascar’s digital infrastructure faces significant risks due to frequent tropical cyclones. These cyclones occur predominantly during the December-to-April season, with their intensity increasing over recent decades. Between 2000 and 2023, Madagascar was hit by 47 tropical storms and cyclones, resulting in extensive damage to infrastructure and economic losses.67 The rainy season, which overlaps with the cyclone season from January to April, exacerbates these challenges through strong winds, heavy rainfall, flooding, and landslides. Historical cyclone track analysis (Fig. D11) highlights that the entire island is highly vulnerable to these storms. However, this broad assessment lacks the detailed insights needed to support targeted initiatives. To address this gap, more granular analyses have been conducted by others (see Fig. D12), but further assessments would be needed. Fig D11: Historical cyclone tracks for Madagascar Fig D12: Cyclone exposure Source: ACAPS thematic report68 Cyclones generate strong winds, often exceeding 150 km/h, which can cause significant structural damage to telecom towers and aerial fiber cables, disrupting mobile communication services. Heavy rainfall during cyclones can cause flooding and trigger landslides that directly damage fiber optic cables and other infrastructure. Additionally, cyclones often lead to power outages that can incapacitate mobile networks even if the physical infrastructure remains intact. Recent cyclones, such as Cyclone Batsirai in February 2022 and Cyclone Freddy in February-March 2023, exemplify the destructive potential of these storms.69 As previously stated, climate change is exacerbating these challenges, with projections indicating higher average wind speeds, increased rainfall, and a greater proportion of Category 4 and 5 cyclones. Although fewer cyclones may make landfall compared to historical records, those that do will be increasingly severe. Madagascar’s eastern and northern coastal regions are particularly exposed to these impacts. 67 World Bank Group. 2024. Madagascar Country Climate and Development Report (CCDR) http://hdl.handle.net/10986/42263 68 The map is adopted from Acaps thematic report ”Madagascar. Cyclone Exposure and vulnerabilities”. Source: https://www.acaps.org/fileadmin/Data_Product/Main_me- dia/20240119_ACAPS_Thematic_Madagascar-cyclone_exposure_and_vulnerabilities.pdf 69 ACAPS, 2024 “Madagascar: Cyclone exposure and vulnerabilities” Thematic report, 19 January 2024. Source: https://www.acaps.org/fileadmin/Data_Product/Main_me- dia/20241019_ACAPS_Thematic_Madagascar-cyclone_exposure_and_vulnerabilities.pdf 81 The increasing frequency and intensity of tropical Fig D13: Projected coastal flooding, RCP8.5 climate scenario, cyclones demands comprehensive resilience and disaster severe flooding event preparedness strategies to safeguard critical infrastructure. To mitigate the impacts on digital connectivity, it is essential to adopt enhanced resilience and redundancy measures. For instance, mobile towers should be engineered to endure the high wind speeds typical of cyclones by utilizing durable materials and adhering to cyclone-resistant building codes. Strengthening tower foundations can further prevent structural failures during severe winds. Additionally, robust contingency plans must be developed to ensure the rapid restoration of connectivity and maintain emergency communication systems during such events. The combined threats of floods, landslides, and cyclones highlight the urgent need for detailed risk assessments and targeted resilience strategies in regions of Madagascar where infrastructure—both existing and planned—is located. Such measures are essential to reduce current vulnerabilities and protect critical assets and connectivity from future climate impacts. All stakeholders should be required to conduct comprehensive, risk-informed site selection and preparation, and to implement appropriate hardening measures for all types of digital infrastructure and their power supplies. These measures should address extreme events, including high winds, heavy rainfall, surface flooding, and landslides. Wherever possible, these requirements should be incorporated into regulations and guidelines governing the sector. Lastly, Madagascar does experience earthquakes, but the overall seismic risk is relatively low compared to other regions. Assessments classify Madagascar’s earthquake Other hazards hazard as low, with only a 2% chance of experiencing Madagascar, with the longest coastline in Africa, faces potentially damaging earthquake shaking in any given area significant challenges from coastal flooding and rising sea over the next 50 years.70 This suggests that while earthquakes levels. Although current data indicates relatively low risk can occur, they are not a primary concern compared to other levels for fiber networks and cell sites, submarine cable natural hazards like cyclones and floods. However, it is landing stations are at a notably higher risk. These critical prudent to consider earthquake risk alongside landslide risk, infrastructure points—two on the island’s east coast and two as both share similar resilience strategies. These strategies on the west—require immediate and long-term protective include designing telecom towers for enhanced stability, measures to mitigate threats from coastal floods, erosion, burying fiber optic cables to protect against landslides and and tropical storm surges. Ensuring their protection is vital surface damage, ensuring overhead cables have sufficient for maintaining the resilience of communication networks slack to absorb stress, and implementing redundant power and supporting regional stability. sources (such as generators) and network paths to maintain service during outages. Additionally, it is advisable to avoid placing critical infrastructure in high-risk zones for both earthquakes and landslides. 70 https://thinkhazard.org/en/report/150-madagascar/EQ 82 Annex E: Country case - Somalia Somalia, located in the Horn of Africa, experiences predominantly arid and semi-arid climates. The country has four distinct seasons: two rainy seasons (April-June, Oct-Dec) and two dry seasons (Dec-March, July-Sept). Rainfall is highly variable, with annual precipitation ranging from less than 100 mm in the northeast to 500-700 mm in the southwest. The mean annual temperature is among the highest globally, averaging close to 30°C, with very hot conditions prevailing throughout the year. Somalia is highly exposed and vulnerable to extreme climatic events. Droughts occur moderately every 3-4 years and severely every 7-9 years. Flash floods are common during intense rainfall periods, affecting riverine areas along the Juba and Shabelle Rivers, as well as mountainous regions in the north. Tropical storms and cyclones occasionally impact Somalia’s coastal regions, causing strong winds and storm surges. Heatwaves are increasingly frequent, with temperatures often exceeding 35°C. Climate change is expected to exacerbate Somalia’s existing climate hazards. Projections indicate a rise in mean annual temperatures by 1–1.75°C by mid-century (2040–2060) and up to 3.2–4.3°C by 2100. This will lead to more frequent heatwaves and higher heat-related mortality rates. While annual rainfall may slightly increase (1% by 2030, 3% by 2050), its distribution will become more erratic, intensifying droughts and floods. Climate models predict stronger tropical cyclones, increasing Somalia’s exposure to extreme storms along its coastline. Coastal communities face risks from rising sea levels, including erosion, saltwater intrusion into freshwater systems, and habitat degradation. Fig E1: Projected temperature and rainfall, Somalia, Somalia Projected temperature Projected rainfall Source: World Bank, Climate Change Knowledge Portal (2025). URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: Somalia’s digital infrastructure is experiencing notable growth, marked by advancements and ongoing challenges. Internet access is expanding, with mobile broadband penetration at 15% and 3G coverage reaching 76%. The country is connected to international fiber optic networks through systems like EASSy, DARE1, and the 2Africa cable. However, Somalia lacks a national fiber optic backbone, and existing regional fiber networks operate independently without interconnection. This fragmented setup complicates international connectivity, which relies on submarine cable landings in Mogadishu and Bosaso, as well as terrestrial links to Djibouti and microwave links to Kenya. Key obstacles include security issues, high deployment costs for fiber optics, and inadequate regulatory frameworks. 83 Despite these challenges, Somalia offers the lowest mobile internet costs in Africa, driven by a competitive private- sector telecom market. This liberalized environment fosters innovation, particularly in mobile money platforms and digital services. Strengthening fiber optic infrastructure and improving regulatory frameworks could significantly enhance access to affordable, high-quality internet services. International initiatives are supporting these efforts. Projects like the World Bank’s Somalia-Horn of Africa Infrastructure Integration Project and the Eastern Africa Regional Digital Integration Project (EARDIP) aim to improve cross-border terrestrial connectivity and expand both backbone and last-mile internet access, especially in rural and remote areas. Fig E2-3: Mobile cell site and fiber network locations Climate risks to digital infrastructure Flood risks to digital infrastructure The results of an exposure and vulnerability analysis highlight flood-prone areas in Somalia, comparing these locations with the sites and routes of fiber network infrastructure and cell sites. Additionally, the analysis estimates potential damage costs to these two types of digital infrastructure in the event of a severe flood hazard. The map below illustrates regions at risk of flooding, primarily around major rivers originating from Somalia’s northern highlands or Ethiopia’s highlands. These rivers flow into low-lying plateaus, coastal lowlands, or directly into the sea. The map also reveals significant overlap between flood-prone areas and current and planned fiber network routes, underscoring the substantial risks to connectivity across various parts of Somalia. This disruption could impact both urban economic hubs and remote rural communities. In contrast, the data indicates that cell sites are considerably less exposed to direct flood impacts compared to fiber networks. 84 Fig E4: Projected river flooding, RCP4.5 climate scenario, severe flood event In the analysis of riverine flooding damage costs in Somalia, the estimated direct damage costs present a concerning outlook under two different climate scenarios: RCP4.5 and RCP8.5. Results indicate that while both scenarios predict significant costs for both cells sites and fiber network, the RCP8.5 scenario, which assumes higher greenhouse gas emissions and more severe climate impacts, results in a marginally higher estimate for damage costs, although the absolute damage cost values for fiber network are significantly lower than those for mobile cells. 85 Fig E5-E8: Estimate riverine flooding damage and damage costs, Somalia Flood risk: Mobile cells Flood risk: Fiber network In the 0.1% probability category (representing high-intensity, In the 0.1% probability category (representing high-intensity, low-frequency flooding events) for estimated riverine flooding low-frequency flooding events) for estimated riverine flooding damage costs to mobile cells in Somalia, the results are as damage costs to fiber optic routes in Somalia, the results are as follows: follows: Direct Damage Costs: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is 86.6 • Under RCP4.5, the estimated direct damage cost is 6.52 million USD. million USD. • Under RCP8.5, the estimated direct damage cost is slightly • Under RCP8.5, the estimated direct damage cost is slightly lower, at 92.3 million USD. higher, at 7.33 million USD. 86 Landslide risk to digital infrastructure On the map below, areas with landslide risk and the digital infrastructure, respectively cell sites and fiber network, have been overlayed. This is followed by figures containing two bar charts representing the estimated landslide exposure for mobile cell and fiber network assets in Somalia. The data is broken down by landslide risk categories: No Risk, Low Risk, Medium Risk, and High Risk. The data indicates that mobile cell infrastructure in Somalia is largely protected from landslide risks, with almost all assets located in areas with no risk. Only a very small fraction of mobile cells is exposed to any level of landslide risk, suggesting that landslide vulnerability for mobile assets is not a significant concern. Compared to mobile cells, the fiber network in Somalia has slightly higher exposure to landslide risks, with 3.2% of assets in medium-risk and high-risk areas. This distribution indicates a small, but significant vulnerability of fiber infrastructure to landslides, which could lead to considerable damage and disruption if landslides occur in these regions, especially fiber networks located in the northern mountainous parts of the country. Fig E9: Landslide risk exposure for medium and high-risk areas 87 Fig E10: Estimated landslide exposure for mobile cells and fiber assets Somalia, reported by risk category No Risk: A substantial majority (98.5%) of mobile cell assets in No Risk and Low Risk: The majority of fiber network assets Somalia are in areas with no landslide risk. (91.7% and 5.2%) are located in areas with no risk or low risk Low Risk: Only 1% of mobile cell assets fall into the low-risk Medium Risk: 2.4% of fiber network assets are located in category. medium-risk landslide areas. Medium Risk and High Risk: Very minimal exposure, with 0.4% High Risk: Only 0.8% of fiber network assets are situated in of mobile cells in medium-risk areas and 0.1% in high-risk areas. high-risk zones. Cyclone hazards Cyclones are relatively rare in Somalia, with only a few making landfall since reliable records began. However, when they do occur, their impacts can be devastating. These storms generate powerful winds exceeding 150 km/h, often causing significant structural damage to telecom towers and disrupting mobile communication services. Heavy rainfall accompanying cyclones frequently leads to flooding, which damages fiber optic cables and other critical infrastructure. Additionally, power outages during cyclones can incapacitate mobile networks even if the physical infrastructure remains intact. Notable recent cyclones include Cyclone Gati in November 2020 and Cyclone Sagar in May 2018. Cyclone Gati was particularly severe, reaching peak winds of 165 km/h at landfall, making it equivalent to a Category 3 hurricane. The increasing intensity and frequency of cyclones in Somalia underscore the need for robust resilience and disaster preparedness strategies. Essential measures include designing mobile towers to withstand high wind speeds typical of cyclones, using cyclone-resistant materials and engineering practices, and reinforcing tower foundations to prevent tipping during strong winds. 88 Fig E11: Historical cyclone tracks Somalia 89 Other hazards Somalia’s extensive coastline also exposes it to challenges from coastal flooding and rising sea levels. Current assessments indicate that the risk level for fiber networks and cell sites is relatively low, whereas submarine cable landing points and stations in Berbera, Bosaso or Mogadishu would face a notably higher risk. To safeguard these critical infrastructure points, it is essential to implement protective measures against coastal floods, erosion, and tropical storm surges in both short- and long-term strategies. Prioritizing their protection will enhance the resilience of international connection. Fig E12: Projected coastal flooding RCP8.5 climate scenario, 2080 Lastly, while Somalia does experience earthquakes, the overall seismic risk is relatively low compared to other natural hazards like cyclones and floods, although the risk is considerably higher in the northern parts of the country and primarily concentrated along the Rift Valley areas. Therefore, in the mountainous northern region earthquake risk could to be considered alongside landslide risks since both require similar resilience strategies. 90 Annex F: Country case - South Sudan South Sudan, a landlocked nation in East-Central Africa, experiences a tropical climate marked by high temperatures and distinct wet and dry seasons. Average temperatures range between 26°C and 32°C, with peaks exceeding 35°C during the dry season (January to April). Rainfall varies significantly across regions, from as little as 200 mm annually in the southeastern areas to between 1,200 mm and 2,200 mm in the forested highlands of Western Equatoria. The country is highly exposed and vulnerable to extreme climate events. Seasonal flooding is common during the rainy season, particularly in August and September when the Nile River and its tributaries overflow. Flash floods, caused by intense localized rainfall, are also frequent. Conversely, droughts occur due to below-average rainfall, leading to water scarcity and heat stress, especially during the dry season. South Sudan has already observed rising temperatures, with increases of over 1°C since 1980. This trend has made “normal” years hotter and drier. Climate change is expected to exacerbate these challenges. While future rainfall patterns remain uncertain, historical trends suggest a decline in precipitation in some regions. Reduced rainfall combined with rising temperatures may result in prolonged dry spells and diminished water availability. Additionally, increased rainfall variability could intensify riverine and flash floods, further straining the country’s resilience to climate hazards. Fig F1: Projected temperature and precipitation Projected temperature Projected precipitation Source: World Bank, Climate Change Knowledge Portal (2025). URL: https://climateknowledgeportal.worldbank.org/. Date Accessed: 15 April 2025 South Sudan’s digital infrastructure is in a nascent stage but shows signs of gradual development. The key challenge is the lack of an operational domestic fiber backbone. South Sudan is served by a relatively recent route (2020) connecting Juba to Mombasa, Kenya, via Uganda, but this has been plagued by unreliability. However, this reduced internet costs and improved access. As of early 2024, internet penetration in South Sudan reached 12.1%, up from 7% in 2023, with approximately 1.36 million users.71 However, this remains one of the lowest rates globally. The government, through the Ministry of Information, Communication Technology, and Postal Services, has been working on enhancing digital services. The lack of digital skills among the population hinders widespread adoption of digital tools. 71 https://dig.watch/countries/south-sudan 91 The World Bank is enhancing digital infrastructure and connectivity in South Sudan primarily through the Eastern Africa Regional Digital Integration Project (EARDIP), which aims to promote an integrated digital market across Eastern Africa by increasing cross-border broadband connectivity, data flows, and digital trade. EARDIP focuses on deploying broadband infrastructure along priority cross-border routes to improve redundancy and resilience. The project plans to establish additional fiber rings, including routes connecting northern and southern parts of the country and links to neighboring countries like Kenya, Sudan, and potentially the Central African Republic. Discussions also include promoting green solar energy solutions to power mobile network infrastructure, reducing reliance on generators and improving sustainability. Fig F2-3: Mobile cell site and fiber optic network locations 92 Climate risks to digital infrastructure Flood risks to digital infrastructure The analysis of exposure and vulnerability highlights flood-prone areas in South Sudan and their intersection with critical digital infrastructure, including fiber network routes and cell sites. The study compares these locations to the estimated damage costs that could result from severe flood incidents. The map below illustrates regions at high risk of flooding, particularly around major rivers originating from the highlands in southern South Sudan and Ethiopia in the east. It identifies multiple zones where flood risks overlap with fiber network routes and cell sites, underscoring the significant threat posed to digital infrastructure across various parts of the country. Fig F4: Projected river flooding, RCP4.5 climate scenario, severe flood event The estimated direct damage costs from riverine flooding in South Sudan reveal a concerning trend under two climate scenarios. Under the moderate climate scenario RCP4.5, the direct damage cost is estimated at 16 million USD. However, under the more severe RCP8.5 scenario, this cost rises significantly to 24.6 million USD. This represents an approximate 54% increase in financial losses as the climate scenario worsens. Similarly, the estimated damage costs for fiber networks increase from 1.91 million USD under RCP4.5 to 2.75 million USD under RCP8.5. This corresponds to about a 44% increase in costs due to more extreme climate conditions. 93 Fig F5-F8: Estimate riverine flooding damage and damage costs, South Sudan Flood risk: Mobile cells Flood risk: Fiber network In the 0.1% probability category (representing high- In the 0.1% probability category (representing high- intensity, low-frequency flooding events) for estimated intensity, low-frequency flooding events) for estimated riverine flooding damage costs to mobile cells in South riverine flooding damage costs to fiber optic routes in Sudan, the results are as follows: South Sudan, the results are as follows: Direct Damage Costs: Direct Damage Costs: • Under RCP4.5, the estimated direct damage cost is • Under RCP4.5, the estimated direct damage cost is 16 million USD. 1.91 million USD. • Under RCP8.5, the estimated direct damage cost is • Under RCP8.5, the estimated direct damage cost is significantly higher, at 24.6 million USD. higher, at 2.75 million USD. The financial impact of flooding on mobile infrastructure is notably higher compared to fiber optic routes, reflecting the greater vulnerability and the higher direct damage costs of mobile systems. These findings highlight the growing economic vulnerability of South Sudans telecommunications infrastructure to worsening climate change impacts, emphasizing the need for adaptive strategies to mitigate future losses. Given the high costs associated with flooding, it is critical to prioritize investments in flood-resistant infrastructure, particularly in vulnerable areas. 94 Landslide risk to digital infrastructure On the map below, areas with landslide risk and the digital infrastructure, respectively cell sites and fiber network, have been overlayed. The following image contains two bar charts representing the estimated landslide exposure for mobile cell and fiber network assets in South Sudan. The data is broken down by landslide risk categories: No Risk, Low Risk, Medium Risk, and High Risk. Fig F9: Landslide risk exposure for medium and high-risk areas Fig F10: Estimated landslide exposure for mobile cells and fiber assets in South Sudan, reported by risk category 95 No Risk: The far majority (99.3%) of mobile cell No Risk: The majority of fiber network assets are located in areas towers are categorized as having no landslide risk. with no risk for landslides Low Risk: 0.6% of mobile cells fall under the low- Low Risk: 2.1% of fiber network assets are categorized as having risk category. low risk. Medium and High Risk: Only a minimal 0.1% of Medium Risk: A small proportion of fiber network assets (0.7%) is mobile cells are exposed to medium risk, and no classified under the medium risk category. mobile cells fall under high-risk exposure High Risk: Only 0.3% of fiber assets are in high-risk areas, making them particularly vulnerable to landslide events. In conclusion, the overwhelming majority of mobile cell towers are safe from landslide risks, with only a tiny fraction facing any exposure. Likewise, only a small proportion (1%) of the fiber network is classified under medium and high-risk categories, indicating a need for further assessment and potential risk mitigation strategies to protect these assets. Other hazards In South Sudan, the risk of significant hazards from storms or earthquakes impacting digital infrastructure, such as fiber networks and telecom towers, is minimal. The region’s geographical characteristics and climate patterns do not typically lead to severe storm events that could disrupt these systems. Furthermore, South Sudan experiences a low frequency of seismic activity, reducing the likelihood of earthquakes that could damage infrastructure. 96 Annex G: Technical note on data analysis and risk assessment For this report, a scenario-based method was developed CHEM from the Japanese Atmosphere and Ocean Research to assess telecom asset exposure and vulnerability for Institute (The University of Tokyo), National Institute for a variety of hazards. The approach consists of three Environmental Studies, and Japan Agency for Marine- main components, including the (i) selection of scenario Earth Science and Technology, and finally (v) NorESM1-M parameters, (ii) processing of global hazard data layers and from the Bjerknes Centre for Climate Research, at the affiliated data, and (iii) intersection processing of geolocated Norwegian Meteorological Institute. This analysis does not telecom assets with hazard layers to estimate asset exposure aim to preference a specific model, instead opting to use and vulnerability. The method steps will now be articulated these capabilities as an ensemble to obtain broad statistical in detail, beginning with the data sources utilized. estimates given each of these five global circulation model options. For landslide risks hazard data is taken from the Arup Global Hazard data Landslide Hazard Map. The data represent a combination of the median annual rainfall-triggered landslide hazards Firstly, for flood hazards the Global Aqueduct geospatial between 1980-2018, and earthquake-triggered landslide hazard datasets are obtained, selected for key scenario hazards. Available as a raster layer, each tile is graded into parameters of interest, and then processed for both riverine categories of High Risk, Medium Risk, Low Risk, or No Risk. and coastal flooding layers (World Resources Institute, 2022). Out of the >700 available data layers, two different climate scenarios are selected (as well as a 1980 historical For tropical cyclone risks hazard data is taken from baseline), based on the Intergovernmental Panel on Climate the United States National Oceanic and Atmospheric Change (IPCC) 5th assessment report. These include (i) Administration (NOAA) for historical hurricane landfalls. Representative Concentration Pathway 4.5 (steady carbon The available time-series spans the last 150 years for some emissions) (intermediate climate outcomes) (RCP 4.5) locations, with this dataset identifying events of a variety of and (ii) Representative Concentration Pathway 8.5 (rising sizes, all the way up to Category 5. carbon emissions) (limited climate outcomes) (RCP 8.5). The scenario names are possible values for radiative forcing in 2100, for example, either 4.5 or 8.5 Watts/m2. Infrastructure asset data The former scenario (RCP 4.5) is seen to be an optimistic future, reflecting considerable carbon mitigation by 2040 Crowdsourced data for December 2022 are gathered from onwards. Whereas the latter (RCP 8.5) is essentially a OpenCelliD for current mobile infrastructure networks, pessimistic business-as-usual case (rising carbon emissions) covering 46.7 million global cellular assets for 2G, 3G, with very little carbon mitigation taking place (thus, Earth’s 4G and 5G, and then separated by country. The Global current global trajectory is closest to RCP 8.5). Four main System for Mobile Communications (GSM) is regarded return periods are also selected for assessment, including more commonly as the second cellular generation or ‘2G’. events equating to 1% annual probability (1-in-100-year), following standardization in 1991, commercial deployments 0.4% annual probability (1-in-250-year), 0.2% annual were launched across the world in the 1990s onwards, with probability (1-in-500-year), and the most extreme 0.01% the technology offering revolutionary voice and text Short annual probability (1-in-1000-year). Messaging System (SMS) services. Although outdated in many high-income countries, 2G is still very much widely Riverine hazards in this evaluation represent overflow river used technology, especially in low- and middle-income flooding events, most likely to occur in river basins from countries. excessive water loads. There are five common models widely available from different institutions for riverine flooding, The third cellular generation known as ‘3G’ consists of including (i) GFDL-ESM2M from the US Geophysical Fluid multiple technologies, with the main standard known Dynamics Laboratory (NOAA), (ii) HadGEM2-es from the as the Universal Mobile Telecommunications System UK Met Office Hadley Centre, (iii) IPSL-CM5A-LR from the (UMTS). Standardized in 2001, commercial launches widely French Institut Pierre-Simon Laplace, (iv) MIROC-ESM- took place throughout the 2000s. The key benefit of 3G 97 deployment was the introduction of basic data capabilities, assets (Koks et al., 2022). Unfortunately, within the US enabling users to access a peak connection speed of ~3 HAZUS model, the development of fragility relationships Mbps. This enables web browsing and very basic low- specifically for telecommunications infrastructure assets resolution video functions. based on inundation has been stated as “deferred to a later date” (p7-10), with no current update on the status of these The fourth generation of cellular technology (‘4G’) was improvements (DHS and FEMA, 2011) (thus, identifying an standardized in 2009 with the name Long Term Evolution important area of future research). (LTE) and brought mass market mobile broadband to billions of consumers around the world (mainly deployed in Unfortunately, we have limited information relating to the the 2010s onwards). This was driven by the introduction of depth-damage relationship for these assets with regard to smartphones as a key general-purpose technology, largely flooding. For example, Figure G1 (A) illustrates the depth- resulting from Apple’s pioneering iPhone. With peak speeds damage curves adopted for mobile cells, with the baseline capable of up to ~100 Mbps, users have the capability to adopted from Kok et al. (2004). The low curve represents download high-definition video while mobile, as well as carry an approximate 50% decrease in vulnerability from the out a large range of online activities, most notably sharing baseline, whereas the high curve represents an approximate and consuming high-quality content across a range of social 100% increase in vulnerability. The baseline is utilized here media platforms. in this assessment. The introduction of the fifth generation of cellular Moreover, Figure G1 (B) illustrates the depth-damage technology known as 5G has only taken place in recent years, curves adopted from the literature for fiber optic cable. from 2020 onwards. The standardized radio technology Unfortunately, we do not know if the cable is deployed via is referred to as the 5G New Radio (NR) standard. Three aerial means (across telecommunication or electricity poles/ key use cases exist for users, including enhanced mobile pylons), or within underground ducts along the road network. broadband, ultra-reliable low-latency communications, Therefore, the low depth-damage curve represents the and massive machine type communications. Over the next more conservative shape for aerial electricity distribution decade, older cells will be decommissioned and replaced poles developed by Miyamoto (2019). This is akin to all fiber with 5G NR equipment. However, current availability is very available being aerially deployed across the existing telecom low. pole network. The baseline is adopted from Koks (2019), based on Espinet et al., (2018), representing the depth- For fixed fiber optic network data, spatial layers are utilized damage curve for road infrastructure. This is akin to all fiber from a recent assessment of the Horn of Africa, as well as being deployed in ducting along the existing road network. the popular Afterfibre dataset, and clipped to the boundaries Finally, the high scenario is adopted from Kok et al. (2004) of each country, ready for intersecting. and represents a more vulnerable depth-damage curve. Fig G1 Damage curves by hazard typeCell site depth- Flooding vulnerability analysis damage curves The flooding vulnerability analysis covers both riverine and coastal flooding hazards. Firstly, mobile cellular towers generally consist of the three main design types identified within the literature review, including monopoles, self- supported structures, or guyed structures. These assets usually rely on steel or aluminum for the structural frame, utilizing concrete at the base of the tower either for the foundation plinth or to secure guy lines. Vulnerability to hazards and the consequential damage cost is estimate using a depth-damage curve approach which is a standard way in the literature to relate the depth of flooding inundation Fiber depth-damage curves to a potential direct economic cost of damage. However, there are not readily available depth-damage curves for communication towers. Therefore, existing studies in the literature adopt a similar approach to assessing the vulnerability of electricity infrastructure assets, such as distribution poles and transmission pylons which share similar structural characteristics (Kok et al., 2004). Indeed, some studies utilize curves from the US HAZUS model (DHS and FEMA, 2011) for electricity infrastructure, to assess the vulnerability of communications infrastructure 98 The investment costs for rebuilding mobile cells are Hotspot analysis for landslides and tropical adapted from the existing literature (Oughton et al., cyclone 2022), to enable damage costs for different sized events A hotspot analysis is undertaken for landslide and tropical and climate scenarios to be estimated. Broadly, the cost of cyclone to provide insight into potential asset exposure to building a new three-sector 4G macro cell site is treated here these threats. Firstly, for landslide the mobile cells and fixed as approximately US$90,000, equating to approximately fiber assets exposed to the four categories are capture by US $30,000 total capital expenditure per cell. Thus, a cell intersecting with the underlying landslide hazard layers. The costing US $30,000 with a flooding depth of 0.6 meters quantity of assets exposed are reported both in absolute and a damage quantity of 0.5, leads to a direct damage terms (e.g., number of exposure assets by category), and in cost estimate of US $15,000. As identified in the literature relative terms (e.g., the percentage of overall assets falling review, the specifics of the damage depend on the cell tower within each category). Secondly, a set of maps are produced construction. At the lower end of the fragility curve, damage which illustrate the existing telecom assets for each country, to electronic equipment could take place. Whereas in more as well as the historical tropical cyclone tracks. These tracks severe outcomes, this could include the full destruction are obtained by intersecting each country boundary with of both active electronic radio equipment and passive civil the NOAA historical dataset. For reference, a buffer of engineering tower structures, requiring a total rebuild of a 33 kilometers is added to the track line to aid in exposure site. understanding for the highest risk areas. Finally, the direct damage cost to the ith cell asset is calculated using the estimated asset replacement cost and the damage value from the depth-damage curve, as per equation (1). References In contrast, the fiber construction cost is treated as US$20 DHS, FEMA, 2011. Hazus-MH MR3 flood technical manual. per meter enabling the damage cost to be estimated for the Version 2.1. Flood Model. ith asset, calculated as per equation (2) Espinet, X., Rozenberg, J., Rao, K.S., Ogita, S., 2018. Where the length of the ith asset represents the flooded Piloting the Use of Network Analysis and Decision-Making portion of the asset intersecting with the hazard layer, and under Uncertainty in Transport Operations. https://doi. the damage value represents the estimated damage state org/10.1596/1813-9450-8490 from the dept-damage curve. Kok, M., Huizinga, H.J., Vrouwenfelder, A., Barendregt, A., The intersection process is carried out practically using a 2004. Standard method 2004. Damage Casualties Caused range of python packages. For example, managing spatial Flooding Client Highw. Hydraul. Eng. Dep. point data is carried out via geopandas, and then raster- based queries of the hazard layers are undertaken using a Koks, E.E., Le Bars, D., Essenfelder, A.H., Nirandjan, S., Sayers, combination of rasterio and rasterstats. For each lower layer P., 2022. The impacts of coastal flooding and sea level rise on region, this information is exported to a comma-separated critical infrastructure: a novel storyline approach. Sustain. value file, and then later aggregated for reference in the Resilient Infrastruct. 0, 1–25. https://doi.org/10.1080/2378 following results section. 9689.2022.2142741 The final results are aggregated to the national level for Oughton, E.J., Comini, N., Foster, V., Hall, J.W., 2022. Policy reporting. choices can help keep 4G and 5G universal broadband affordable. Technol. Forecast. Soc. Change 176, 121409. https://doi.org/10.1016/j.techfore.2021.121409 World Resources Institute, 2022. Aqueduct Floods Hazard Maps [WWW Document]. World Resour. Inst. URL https:// www.wri.org/data/aqueduct-floods-hazard-maps (accessed 2.24.22). 99 Annex H: Project-specific climate adaptation and mitigation objectives and policy alignments Despite their differences in geographical scope and content, the three projects share a common objective: ensuring that investments in digital infrastructure align with the climate adaptation and mitigation standards outlined in the Paris Framework, relevant Sustainable Development Goals (SDGs) and each country’s specific climate commitments. This involves designing and implementing cellular and fiber networks, along with other critical components of the telecom network, in a resilient manner that protects them from typical climate-related and natural hazards. Additionally, a key priority is the adoption of energy-efficient solutions and methods to reduce carbon emissions, where technically and economically viable. Each of the four projects has distinct climate adaptation and mitigation goals that must be integrated into the overall investment strategy, as shown in Table 1 below. In summary, the projects aim to ensure that infrastructure is implemented according to international best practices and standards for resilience, energy efficiency, and renewable energy. This will be achieved by developing policies and guidelines at both regional and national levels, as well as through procurement documents for private companies responsible for deploying and maintaining various aspects of digital infrastructure and equipment. Specifically, the goal is to provide technical requirements and standards that can be incorporated into bidding documents for relevant infrastructure procurement. This includes: • New Cell-Tower Locations and Upgrades: Upgrading existing 2G cell towers to 4G+ (EARDIP and DECIM). • Backbone Fiber Networks: Establishing robust fiber networks (KDEAP and EARDIP). • Last-Mile Connectivity: Ensuring connectivity for schools, health centers, government offices in rural areas, refugee camps, and host communities (all projects). • Satellite Services: Deploying satellite dishes and ground stations for low-earth orbit satellite services (all projects except EARDIP-2). Table H.1. Summary of resilience and mitigation objectives included in project appraisal documents (PADs). Projects Climate adaptation / resilience Climate mitigation  Survey of climate risks in the area where  Investments in digital infrastructure will consider infrastructure will be deployed. energy efficiency measures, including applying use of  Risk informed recommendations for resilience energy-efficient fiber optic cables. measures based on public experiences and  Solar and wind used by the government for towers international best practice. in areas where conditions are suitable. This practice  ToRs for contractors will include requirements for will be sustained in KDEAP IDA investments. Good contractors to include site specific risk assessments. practices from the government and internationally will  TA to develop climate-informed policy and be documented and shared with contractors to inform regulation in the ICT sector, including: (i) a Kenya project implementation. specific strategy to enhance resilience to climate  Connectivity for Education, renewable energy sources related shocks linking to guidelines and frameworks will be used to power schools, TVETs and universities, KDEAP developed by the EAC; and (ii) review of network either as a main or back up power source Kenya and data infrastructure construction guidelines and  TA to develop climate-informed policy and regulation services regulations and support for planning and in the ICT sector and leveraging digital technologies preparedness to ensure the resilience and recovery for climate change mitigation, including: (i) a strategy of essential digital infrastructure and services in the to minimize the environmental impact of ICT event of climate-related shocks. infrastructure (referencing standards, best practices,  Digitizing services public service delivery will be country level targets etc.), devices (repair, reuse, safe enhanced and enable business continuity during disposal) and services, and leverage digital solutions climate induced or other shocks for climate mitigation across the economy.  Support to an e-waste management program including implementation strategy for e-waste act, technical training etc. 100  TA for detailed network design including technical  TA to refine technology and network design to specifications for prioritized routes and network be energy efficient (following international good architecture/configuration, leveraging parallel practices) and prioritizing the use of energy-efficient deployment of linear infrastructure wherever fiber-optic cables for cross-border connectivity and possible. backbone network deployment.  TA to refine location and technology design, which  Energy efficiency considerations in tenders (design, will consider site-specific climate risks, in relation equipment, and operational modes). Reducing the to cross-border connectivity and backbone climate footprint and emissions stemming from deployment new network infrastructure deployed and/or  Embedding climate-proof design in technical upgraded by  (a) using energy-efficient technology, specifications and tenders to ensure implementation, when developing technical specifications/bidding as part of the infrastructure deployment phase, in documents (for example, encouraging use of fiber) and relation to all network infrastructure investments including these in tenders launched under the project; EARDIP made, building in resilience and redundancy such and  (b) encouraging parallel deployment of linear Somalia and as use of buried fiber versus aerial fiber (that can be infrastructure as part of broadband network planning South Sudan prone to climatic changes, others) to reduce the emissions associated with deployment.  TA to develop both regional guidelines and national  Parallel deployment of digital infrastructure with protocols on embedding climate resilience for other linear infrastructure (such as roads or energy connectivity, and data infrastructure (EAC, IGAD) transmission lines), based on ‘dig-once’ principles, can  Regional level (EAC & IGAD) Situational analysis help reduce the emissions stemming from deployment/ and needs assessment on regional data centers for construction. identifying options for enhancing their resilience to  TA to develop both regional guidelines and national climate events. protocols for greening digital infrastructure, including  Financing for deploying resilient and agile hybrid leveraging renewable energy resources, and standards data hosting solutions, climate-proofing solutions to for energy or resource-use efficiency. prevent data loss.  Service contracts with commercial operators to pre- purchase capacity, connecting public institutions along fiber routes and preparation of specifications and tenders for energy efficient hardware/software,  Financing of new construction, repair, and upgrade  Replacing copper with [energy-efficient] fiber optics of cross-border terrestrial links and national cables backbone network infrastructure using climate  Replacing diesel generator cellular base stations with resilient technology such as fiber optic, others. solar powered  TA for feasibility studies, including drafting technical  All new cell towers will be supplied with solar power specifications for prioritized routes, identifying sites and battery storage to be connected along the priority routes, leveraging  New IT equipment and related digital infrastructure parallel deployment of linear infrastructure (for will follow the best international practices for energy example, roads, electricity distribution networks, efficiency and were possible, be certified.72 and so on) wherever possible. Survey work will take  The procurement of network equipment will leverage account of climate-related risks, such as flash floods, the use of energy-efficient technologies and measures coastal flooding etc. to support energy conservation and reduce GHG EARDIP II  The construction of passive elements of the network emissions, where possible, such as by using routers (e.g., cable ducts, shelters for equipment) will factor or other equipment with automatic switch-off Ethiopia and in network redundancy by minimizing single points mechanisms and those that conserve power when not Djibouti of failure in digital networks. This includes choices of in use. whether to deploy cables underground or overhead,  TA for the development of regional guidelines and embedding elevation or strengthening the base and national protocols for the greening of digital of cell towers, for example, to mitigate against the infrastructure, including the reduction of GHG risks of flooding emissions and adaptation to potential climate impacts,  Developing climate-oriented training content on for instance through the exploitation of renewable leveraging digital tools and services as an adaptation energy and through the development of e-Waste mechanism in the event of climate shocks. management (recycle, reuse) plans and guidelines on  Establishing climate emergency response facilities integrating climate resilience for connectivity and data (such as satellite terminals) infrastructure.  TA to assess regional data hosting and data  TA to assess regional data hosting, includind energy management needs (including adoption of a efficiency considerations guided by international best cloud-based approach), looking at demand and practices such as ITU-T recommendations for Green supply, regional demand aggregation, need for for ICT infrastructure. embedding automated backup and disaster recovery sites (in climate-safe locations). 72 Certification where feasible by Energy Star or equivalent standards surpassing the current energy efficiency standards, and in accordance with recommendations from The European Commission's Joint Research Centre (JRC) report Best Environmental Management Practice in the Telecommunications and ICT Services Sector 101  Feasibility study to analyze location-specific climate  Solar power and batteries would be used for cell change risks and exposure to natural disasters and towers and base stations, and these would be ‘’over- identify adequate mitigation measures. dimensioned’’ to allow for co-deployment of mini grids  A Commercial Transactions Manual (CTM) will to power both towers and local communities be developed to guide the allocation of the gap-  Wherever possible, green energy solutions (e.g., solar financing subsidies; (ii) the recruitment of an power and battery storage) will be used to power cell independent monitoring firm to ensure compliance towers and networks. in technical, environment and social and service  Tenders for digital connectivity infrastructure will level requirements during both construction and follow energy-efficiency guidelines taking local DECIM operations phases, including application of relevant conditions into account, including encouraging a shift Madagascar standards for climate-resilience away from high-energy-consuming legacy technologies  Technical and design specifications for the tenders toward more energy-efficient alternative network will be developed factoring in climate-induced technologies such as fiber optics. risks and resilience measures, e.g., related to site selection. Bidders will be required to comply with specific infrastructure robustness requirements to increase resilience to climate shocks (e.g., investments in flood barriers, more resilient ducts and towers, etc.). Note: World Bank Project Appraisal Documents (PADs) have several elements and proposed activities linked to climate, which are considered “Climate Co-Benefits” when they refer to the share of WBG lending commitment that contributes to climate change mitigation and/or adaptation. In digital development projects, the core infrastructure investments such as broadband connectivity, communication towers, data centers etc. are generally implemented to increase digital access and strengthen digital connectivity – so the infrastructure is built irrespective of climate change and is not specifically intended as an adaptation or mitigation measure. For these types of investments, climate change is a consideration, and measures are embedded, but climate is not the main driver behind the investment and therefore only the marginal or additional cost of such climate-resilient or green design considerations/features is counted as adaptation or mitigation finance. 102 Project alignment with national agenda on climate goals Kenya Digital Economic Acceleration Project The project aims to support mitigation goals through (KDEAP) investments in energy efficiency measures for digital infrastructure, including applying use of energy-efficient fiber optic cables, project design and operational practices that The project has among its objectives to contribute to take energy efficiency into account in line with international advancing Kenya’s climate objectives.73 Notably, Kenya is good practices and standards. It will also enhance the use among the few countries where the government explicitly of solar and wind for towers in areas where conditions are acknowledges the digital sector’s role in climate action. suitable through its investment. The project will provide TA The 2019 National Information, Communications, and to develop climate-informed policy and regulation in the Technology Policy includes an environmental component, ICT sector and leveraging digital technologies for climate positioning ICT as a crucial tool for both adaptation and change mitigation, including: (i) a strategy to minimize the mitigation. Specifically, it emphasizes the need for ICT environmental impact of ICT infrastructure (referencing players and consumers to minimize the environmental standards, best practices, country level targets etc.), devices impact of infrastructure, appliances, and devices. The (repair, reuse, safe disposal) and services, and leverage digital Digital Economy Blueprint further integrates Green ICT by solutions for climate mitigation across the economy. The promoting mechanisms to reduce e-waste and enhance the project will also provide support to an e-waste management efficiency of ICT equipment. program including implementation strategy for e-waste act, technical training etc. Although Kenya’s National Determined Contributions The project also aligns with and supports the ICT vision (NDC) does not mention digital technologies or digital outlined in the National Policy on Gender and Development. infrastructure, the National Climate Change Action Plan This policy aims to enhance women’s inclusion in the ICT (NCCAP) and National Adaptation Plan (NAP) both highlight sector by increasing their access to STEM education, them as enablers for adaptation and mitigation actions. eliminating social barriers, and creating opportunities for Kenya’s current National Adaptation Plan (NAP) highlights women in ICT. To achieve this, the project will focus on the importance of “climate-proofing” infrastructure to ensure several key initiatives: The project will work to increase that investments remain viable in the future. However, the women’s access to the internet, bridging the digital divide plan does not specify means or ways of implementation. and ensuring they can fully participate in the digital economy. It will offer customized training programs in The project aims to support adaptation goals by focusing schools to improve digital literacy and promote safe internet on the resilience of investments, taking into account use among women and girls. The project will also provide country and location-specific risks. It will also target use targeted support to female students pursuing STEM fields cases, such as education, that enhance population resilience by facilitating access to laptops and tablets, enabling them and consider relevant mitigation measures to reduce to engage more effectively with digital technologies. Lastly, greenhouse gas emissions from project investments. As it will support the collection of gender-disaggregated summarized in the table above, the project procurement, data on ICT usage, which is crucial for understanding and including ToRs for contractors, will include requirements addressing the specific challenges faced by women in the for contractors to include site specific risk assessments. ICT sector. This data will inform future policy decisions and The project will also provide technical assistance (TA) to interventions aimed at promoting gender equality in ICT. develop climate-informed policy and regulation in the ICT sector, including: (i) a Kenya specific strategy to enhance resilience to climate related shocks linking to guidelines and frameworks developed by the EAC and others; and (ii) review of network and data infrastructure construction guidelines and services regulations and support for planning and preparedness to ensure the resilience and recovery of essential digital infrastructure and services in the event of climate-related shocks. 73 Kenya has ratified Multilateral Environmental Agreements such as the UNFCC, Kyoto Protocol, the Paris Agreement, and others and formulated national policies such as the National Climate Change Response Strategy in 2010, the National Climate Change Action Plan (2013–2017) in 2012, and the current NAP (2015–2030). 103 Eastern Africa Regional Digital Integration Additionally, the project aligns with the World Bank’s Project (EARDIP) - Somalia and South Sudan Gender Strategy (2016–2023) by improving women’s access to broadband services, addressing barriers related to The project aims to support Somalia and South Sudan in low skills attainment and empowering women’s voices and achieving their Nationally Determined Contributions (NDCs) agency. and contribute to their efforts in mitigation and adaptation. Somalia has committed to reducing its greenhouse gas emissions by 30% by 2030 compared to a business-as- Eastern Africa Regional Digital Integration usual scenario. The Somalia NDC emphasizes adaptation Project (EARDIP SOP II) - Ethiopia and Djibouti measures across sectors, including infrastructure, although not specifically mentioning the digital sector, as well as The project aims to support Ethiopia and Djibouti in disaster preparedness and management. South Sudan achieving their Nationally Determined Contributions aims to reduce emissions by 109.87 million tonnes of (NDCs) and contribute to their efforts towards climate carbon dioxide equivalent and sequester an additional change mitigation and adaptation. In their latest NDC 45.06 million tonnes by 2030. South Sudan’s Nationally submissions to the United Nations Framework Convention Determined Contribution (NDC) targets several key sectors on Climate Change, Ethiopia and Djibouti have committed for mitigation efforts, including infrastructure, but similar to reducing their greenhouse gas emissions by 68.8% and to Somalia the digital sector is not specifically mentioned, 40%, respectively, by 2030. However, a significant portion of however there is a focus on developing mini-grids and off- these commitments—80% for Ethiopia and approximately grid solar energy projects to enhance electricity access in 70% for Djibouti—is contingent upon international support. rural areas. Both countries have identified renewable energy as a key mitigation strategy. On the adaptation front, the NDCs The EARDIP project aims to address country climate emphasize the need to enhance resilience in infrastructure commitments and priorities by focusing on climate and rural communities vulnerable to climate-related adaptation, mitigation, and enhancing digital inclusion. disasters. An initial climate and disaster risk assessment revealed the region’s high vulnerability to climate change and limited capacity to respond. As demonstrated by project The project aim to support adaptation/resilience goals by components summarized in the table above; to address these targeting climate hotspots (focusing on flood-prone areas challenges, the project will specifically tackle climate risks in Ethiopia and coastal regions in Djibouti) to ensure better affecting physical digital infrastructure and the regulatory protection and preparedness for these communities. The environment. project will establish emergency communication response systems in remote, vulnerable areas, including rural and The project aims to support adaptation goals through a refugee-hosting borderlands in Ethiopia. For infrastructure refinement of the technical network design for prioritized investments, the financing of new construction, repair, routes and network configuration, with consideration and upgrade of cross-border terrestrial links and of local climate risk to allow for climate-proofing when national backbone network infrastructure will be using planning connectivity infrastructure deployment. It plans to climate resilient technology such as underground fiber embed climate-proof design in technical specifications and optic. Special design provisions will ensure continuity of tenders, and for deploying connectivity infrastructure in all connectivity services at times of natural disasters e.g., by flood and drought prone areas. Moreover, the project will introducing network redundancy to improve resilience. provide TA to develop both regional guidelines and national The construction of passive elements of the network (e.g., protocols on embedding climate resilience for connectivity, cable ducts, shelters for equipment) will factor in network and data infrastructure (EAC, IGAD). It will also support redundancy by minimizing single points of failure in digital through finance the deployment of resilient and agile hybrid networks. This includes choices of whether to deploy cables data hosting solutions, climate-proofing solutions to prevent underground or overhead, for example, to mitigate against data loss. the risks of flooding. All new cell towers will be supplied with solar power and battery storage The project will also provide The project aims to support mitigation goals through TA TA to assess regional data hosting and data management to develop both regional guidelines and national protocols needs, looking at demand and supply, regional demand for greening digital infrastructure, including leveraging aggregation, need for disaster recovery sites (in climate-safe renewable energy resources, and standards for energy or locations). resource-use efficiency. As summarized in the table above, the project procurement process includes preparation of specifications and tenders for energy efficient hardware/ software. 104 The project aim to support mitigation goals by several The project will adopt climate-resilient and energy-efficient means, including by replacing copper with [energy-efficient] technologies. By providing energy access, it can mitigate fiber optics cables, by replacing diesel generator for cellular adverse climate impacts, such as powering cooling appliances base stations with solar power and battery storage capacity. and enhancing food security through solar irrigation and cold New IT equipment and related digital infrastructure will chain systems. On the digital side, the project will contribute follow best international practices for energy efficiency. to building back better by increasing connectivity reliability. The project will provide financing for ICT equipment This is crucial for mitigating economic slowdowns, sustaining (certified by Energy Star or equivalent standards) to up to well-being, and speeding up recovery during crises, as seen additional 50 national universities and a minimum of seven during the COVID-19 pandemic. TVET regional centers of excellence. The procurement of network equipment will leverage the use of energy-efficient The project will include a feasibility study to analyze technologies and measures to support energy conservation location-specific climate change risks and exposure to and reduce GHG emissions, where possible, such as by natural disasters and identify adequate mitigation measures. using routers or other equipment with automatic switch-off A Commercial Transactions Manual (CTM) will be developed mechanisms and those that conserve power when not in use. to guide the allocation of the gap-financing subsidies; (ii) the recruitment of an independent monitoring firm to Additionally, the project will provide technical assistance ensure compliance in technical, environment and social to Djibouti in developing a National Strategy for a Green and service level requirements during both construction Economy and to Ethiopia in creating a National Adaptation and operations phases, including application of relevant Plan (NAP) implementation roadmap for the ICT sector. standards for climate-resilience. Technical and design specifications for the tenders will be developed factoring in climate-induced risks and resilience measures, e.g., related Digital and Energy Connectivity for Inclusion in to site selection. Bidders will be required to comply with specific infrastructure robustness requirements to increase Madagascar project (DECIM) resilience to climate shocks (e.g., investments in flood barriers, more resilient ducts and towers, etc.). The project aims to support Madagascar in achieving its Nationally Determined Contributions (NDCs), in which The project will address mitigation goals by implementing the country has committed to reducing its greenhouse gas solar power and batteries for cell towers and base stations, emissions by at least 14% by 2030 compared to a business- and these would be ‘over-dimensioned’ to allow for co- as-usual scenario. Among other priorities, the country deployment of mini grids to power both towers and local NDC identified the following actions to contribute to the communities. Tenders for digital connectivity infrastructure reduction of GHG emissions: promoting rural electrification will follow energy-efficiency guidelines taking local and renewable and alternative energies, by reinforcing conditions into account, including encouraging a shift away renewable energy (hydraulic and solar) from the current from high-energy-consuming legacy technologies toward level of 35% to 79%. more energy-efficient alternative network technologies such as fiber optics. The project aligns with the priorities in the country NDC and World Bank Group Climate Change Action Plan 2021-2025 and the Paris Agreement by supporting the transitioning towards a digitally transformed economy with enhanced energy access. The project will also increase resilience and adaptation, e.g., by ensuring uninterrupted access to essential services during emergencies and supporting a shift away from resource-intensive and climate-vulnerable economic activities. Hence, renewable energy sources will be utilized to reduce emissions, and digital technologies will substitute for physical movements, further reducing emissions. 105