Digital Monitoring, Reporting, and Verification Systems and Their Application in Future Carbon Markets June 2022 Administered by THE NETWORKED CARBON MARKETS Climate INITIATIVE Warehouse Acknowledgements Disclaimer This report was prepared by Lucas Belenky, Keisuke This work is a product of the staff of The World Iyadomi, Susan David Carevic, and Harikumar Bank with external contributions. The findings, Gadde of the World Bank’s Carbon Markets and interpretations, and conclusions expressed in this Innovation Unit, under the leadership of Wendy work do not necessarily reflect the views of The Hughes, with support from the Partnership for World Bank, its Board of Executive Directors, or the Market Implementation Facility (PMIF). governments they represent. The following people provided valuable contributions The World Bank does not guarantee the accuracy in their personal capacity: of the data included in this work. The boundaries, colors, denominations, and other information shown Case study participants: Sven Braden (Wood on any map in this work do not imply any judgment Tracking Protocol); Michael Fabing (Wood Tracking on the part of The World Bank concerning the Protocol); Tom Baumann (Chile case study); Dmitry legal status of any territory or the endorsement or Halubouski (European Bank for Reconstruction and acceptance of such boundaries. All amounts in $ are Development case study); Polly Thompson (Sylvera United States dollars unless stated otherwise. case study); Sandeep Kanda (India case study); and EY India (India case study). RIGHTS AND PERMISSIONS The material in this work is subject to copyright. Land-use companies and non-profit organizations Because The World Bank encourages dissemination consulted: Global Mangrove Trust (Simon of its knowledge, this work may be reproduced, in Schillebeeckx); Kumi Analytics (Clinton Libbey); whole or in part, for noncommercial purposes as Regen Network (Gregory Landua and Ron Steinherz, long as full attribution to this work is given. now an independent contractor); and Pachama (Diego Saez Gil and Marcela Corradini). Attribution—Please cite the work as follows: World Bank. 2022. Digital Monitoring, Reporting, and Expert reviewers: Andres Espejo; Owen Hewlett; Verification Systems and Their Application in Future Dmitry Halubouski; Tom Baumann; Rachel Mok; Carbon Markets. World Bank, Washington, DC. Massamba Thioye; and Gemma Torras Vives. All queries on rights and licenses should be © 2022 International Bank for Reconstruction addressed to the Publishing and Knowledge and Development / The World Bank Division, The World Bank, 1818 H Street NW, 1818 H Street NW, Washington, DC 20433 Washington, DC 20433, USA; fax: 202-522-2625; Telephone: +1-202-473-1000 email: pubrights@worldbank.org. Internet: www.worldbank.org Editing and design by Clarity Global Strategic Communications (www.clarityglobal.net). i About this report T he purpose of this technical report is to illustrate the need for digital monitoring, reporting, and verification (D-MRV) systems to underpin future carbon markets under the goals of the Paris Agreement by discussing the available technologies and barriers to their adoption. It includes guidelines, tools, and lessons learned to promote the use of these systems and emerging technologies. Section 1 of the report makes the case for transitioning from a conventional monitoring, reporting, and verification (MRV) system to a D-MRV system. It examines the various types of D-MRV, the barriers to implementing a D-MRV system, and the benefits of overcoming these barriers. It also examines the resources needed to develop and implement a D-MRV system, and what an enabling policy and regulatory environment for D-MRV systems might look like. Finally, it suggests a tool for assessing whether a parameter can beneficially be monitored and reported under a D-MRV system. Section 2 offers case studies from across the world demonstrating how D-MRV systems can be used to monitor, report, and verify mitigation actions and greenhouse gas inventories linked to forestry and land-use projects, household and rural renewable energy projects, and even waste-to-energy projects. The case studies include lessons learned and best practices for developing, implementing, and managing a D-MRV system. Executive summary P ost-2020 markets under the Paris Agreement D-MRV systems are one component in the end-to- will be built through a bottom-up approach, as end digitalization of the generation, transfer, and each party to the Agreement is required to track reporting of carbon assets under post-2020 carbon the greenhouse gas (GHG) emission reductions (or markets. In addition to using emerging technologies removals) achieved—and has considerable leeway to to improve data collection and verification, D-MRV determine how this will be done. systems can be connected to national or global registries to automate fulfillment of reporting The bottom-up nature of future carbon markets requirements. Figure 1 shows how D-MRV systems comes with increased complexity and diversity fit into this ecosystem. of reporting and verification approaches for GHG emissions inventories and mitigation outcomes. The use of blockchain technology to create There is significant potential for digital monitoring, immutable and auditable data and transfer records, reporting, and verification (D-MRV) systems to including the creation of mitigation outcomes underpin and streamline the functioning of post- in digital form underpinned by smart contracts, 2020 carbon markets. is another important component of end-to-end digitalization of carbon markets that the industry is The current methods to report GHG emissions designing and implementing. and validate or verify emission reductions can be costly, error-prone, and time-consuming, often Many partially or fully digitalized MRV systems are relying on manual processes and in-person surveys. being piloted or implemented in different countries Increasingly, digital technologies are used to and across various sectors. This report includes case streamline data collection, processing, and quality studies on selected pilot D-MRV systems, providing control in monitoring, reporting, and verification a useful record of insights and lessons learned to (MRV) processes. Examples of such technologies promote the further use of digital technologies. The include smart sensors, satellites and drones, cloud systems examined in the case studies cover forestry, computing, artificial intelligence, the internet of waste, and energy projects in South America, the things, and blockchain encryption. Further, D-MRV Middle East, Asia, and Africa. systems can be applied across the commonly defined MRV types: MRVs of GHG emissions, Multiple barriers need to be removed to enable the mitigation actions, and support. widespread use of D-MRV systems. As some of the case studies show, the funds and expertise required Transitioning towards the use of credible and to implement these technologies for a new D-MRV compatible D-MRV systems to track GHG emissions, system, or to upgrade an existing MRV with these as well as the generation of mitigation outcomes, technologies, are often prohibitive. D-MRV systems will improve the functioning, enable scaling-up, and also sometimes capture sensitive data (for example, increase transparency of post-2020 carbon markets. a mitigation action MRV covering household appliances or decentralized energy generation may ii D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS FIGURE 1: END-TO-END DIGITALIZATION OF CARBON MARKET INFRASTRUCTURE 1 Security Digital MRV based tokens Reference Digital assets Forecasting tokens Native or Program/project level MRV permissioned Compliance reporting Direct API access Wallets token Exchange A ## Conflict resolution Level 1: Meter Cloud Level 2: Inverter, string, combiner box Data capture Level 3: Weather ## sensors Centralized Services Climate Benchmarking database/ Registry Warehouse Ratings Level 4: Data Switchgear, transformers, aggregation MRV system fuses ## Exchange X Due diligence checks Certifications Facility level MRV Verification of units Reporting for markets, UNFCCC store personal customer information), necessitating In addition to these measures, independent standard additional privacy controls to be put in place. organizations and regulatory bodies need to adapt Finally, there is little familiarity of emerging digital their MRV protocols to allow for digitally collected technologies, especially in developing countries, so data and establish rules for how these systems knowledge-sharing and capacity-building efforts are are validated and verified. For example, allowing crucial. The latter is especially important to ensure for one-off, on-site validation using a D-MRV equitable access to carbon markets, particularly in system certification, followed by recurring remote areas where capacity for innovative approaches is low. verifications, will increase the speed of validation and verification while reducing the overall cost of In addition to the barriers listed, the enabling generating mitigation outcomes and incentivizing environment for D-MRV systems is insufficient, the use of digital technologies. making the design, implementation, and operation of these systems burdensome and risky. To manage Capacity building, grants, and knowledge sharing this risk, governments or institutions need to for private companies and non-profit organizations develop policies or guidelines that clearly state how implementing D-MRV systems for carbon markets sensitive information may be collected, used, and are also needed to overcome initial gaps in stored. Governments also need to ensure that the financing and technological expertise. This extends required infrastructure, power, and data networks to third-party verifiers, who need to develop an are available for digital technologies to be effectively understanding of digital technologies so that they implemented. Strategic tax incentives might also can implement revised protocols from independent promote the use of desired emerging technologies. standard organizations and regulatory bodies. 1 Sourced from internal World Bank presentation on the Climate Warehouse Initiative. iii Building capacity in implementing organizations D-MRV systems have the potential to make post- and verifiers can be achieved by hosting information 2020 carbon markets more streamlined and cost- and training sessions, while “mock” review and effective in generating carbon assets and verifying certification of pilot D-MRV systems may provide emission reductions—provided there is impetus to valuable practical insights. overcome the upfront costs and increased technical complexity. However, once these systems are in This report proposes a tool for assessing the place, the effort to scale a D-MRV system is much suitability of monitored parameters for D-MRV less than a conventional monitoring system. systems that might help to create an enabling environment for D-MRV systems and digital Creating an enabling environment and further technologies. Among other considerations, the tool piloting of D-MRV systems will promote continued provides a step-by-step guide for evaluating the improvement of emerging technologies while need to digitize data collection for a given parameter, contributing to the knowledge base of D-MRV whether it is permissible to do so under the chosen systems. The full potential of carbon markets to monitoring methodology, and if it is cost-effective. combat climate change can be unlocked through Implementing organizations can use the tool to digital technologies and D-MRV systems, setting the quickly identify any barriers to the digital monitoring, stage for future innovations such as the tokenization reporting, and verifying of the monitored parameter. of carbon assets through blockchain technology, The report also sets out commonly monitored and real-time issuance of mitigation outcomes from parameters, with descriptions of the parameter and projects with system-wide certification of their D-MRV. appropriate digital technologies, or combinations of technologies, to track these parameters. iv D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Contents About this report i Executive summary ii Abbreviations and acronyms vii SECTION 1: THE CASE FOR DIGITAL MONITORING, REPORTING, AND VERIFICATION (D-MRV) 1 1. Introduction 1 2. Characteristics and types of D-MRV systems 4 3. Barriers to D-MRV systems 8 4. Benefits of transitioning to D-MRV 9 5. Resources needed to transition from MRV to D-MRV 11 6. Elements of an enabling environment for D-MRV systems 13 7. A parameter assessment tool for methodologies under D-MRV systems 18 8. Conclusion and way forward 22 SECTION 2: CASE STUDIES 24 1. Chile waste-to-energy landfill D-MRV 25 2. Grid-connected rooftop solar photovoltaic system in India 29 3. Uganda’s electronic database and information management system 32 4. Forestry project D-MRV systems 36 5. Third-party ratings of nature-based carbon credit projects 43 6. European Bank for Reconstruction and Development D-MRV system for renewable energy 48 APPENDIX 53 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS v List of tables Table 1: Summary of case studies 3 Table 2: Comparison of conventional and digital MRVs for GHG emissions 6 Table 3: Comparison of conventional and digital MRVs for mitigation actions 7 Table 4: Summary of differences between D-MRV and MRV at various process stages 11 Table 5: D-MRV checklist 12 Table 6: Commonly monitored parameters and corresponding digital technologies 21 Table 7: Parameters tracked by the EDIMS system 33 Table 8: Forestry D-MRV system case studies 36 Table 9: Commonly monitored parameters and corresponding digital technologies 54 List of figures Figure 1: End-to-end digitalization of carbon market infrastructure iii Figure 2: Decision tree for assessing if a parameter is suitable for a D-MRV system 20 Figure 3: Diagram of Chile’s waste-to-energy D-MRV system 26 Figure 4: Components and structure of the Chile waste-to-energy landfill D-MRV system 28 Figure 5: Arrangement of system components in D-MRV 30 Figure 6: The EDIMS web application 32 Figure 7: Sample of a certified emission reduction report produced by the EDIMS 34 Figure 8: Global Forest Watch data portal 37 Figure 9: Wood Tracking Protocol D-MRV process diagram 39 Figure 10: GROVE: Forestry Smart Ledger design 40 Figure 11: Venture capital funding for climate tech start-ups 42 Figure 12: Satellite imagery of forest cover (green) is compared to time series deforestation data (red), with other detected land classes, such as water, roads, and settlements (white) 43 Figure 13: Overview of how Sylvera’s machine learning outputs feed into the ratings of carbon score and additionality 45 Figure 14: Sylvera’s terrestrial laser scanner mapping vegetation in the Peruvian rainforest 46 Figure 15: Point clouds of individual trees inside an African tropical forest stand, captured using terrestrial LiDAR scanning 46 Figure 16: EBRD D-MRV structure 50 vi D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Abbreviations and acronyms AI Artificial intelligence API Application programming interface CDM Clean Development Mechanism CONAF Corporación Nacional Forestal CTF Clean Technology Fund DISCOMs Electricity distribution companies in India DLT Distributed ledger technology (blockchain) D-MRV Digital monitoring, reporting, and verification EBRD European Bank for Reconstruction and Development EDIMS Electronic database and information management system GHG Greenhouse gas GIS Geographic information system GRPV Grid-connected rooftop photovoltaic ha hectare IoT Internet of things kW kilowatt kWh kilowatt-hour MDB Multilateral development banks MRV Monitoring, reporting, and verification MWp Megawatt-peak NDC Nationally determined contributions NISE National Institute of Solar Energy (India) QA/QC Quality assurance/quality control REA Rural Electrification Agency (Uganda) tCO2e Tonnes of carbon dioxide equivalent VCS Verified Carbon Standard YDE Yellow Door Energy Jordan WTP Wood Tracking Protocol $ Dollar. All dollar amounts are US dollars unless otherwise indicated D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS vii Section 1: The case for digital monitoring, reporting, and verification (D-MRV) 1 Introduction I nternational carbon markets under the Paris The CDM greatly enhanced knowledge Agreement are significantly different from about emission reduction project cycles. those under the Kyoto Protocol. Under the The protocols, standards, and monitoring Kyoto Protocol, only developed countries had methodologies developed under the CDM are greenhouse gas (GHG) emission reduction likely to see continued use in future carbon targets, and the protocol defined how carbon markets. However, the CDM’s monitoring, units could be traded across countries under verification, and reporting processes—while international market mechanisms, such as providing a high degree of certainty that any the Clean Development Mechanism (CDM). certified emission reductions are additional, By contrast, under the Paris Agreement both real, and credible—were time-consuming and developed and developing countries are required expensive to implement and validate. Post- to submit GHG mitigation goals as part of their 2020 carbon markets will need a better way to nationally determined contributions (NDCs). monitor, report, and verify mitigation activities and resulting carbon emission reductions if Article 6.2 of the Paris Agreement allows these markets are to function properly. countries to partake in voluntary bilateral or plurilateral cooperative approaches to achieve Participants under Article 6 will be required their NDC targets through the transfer and use to regularly collect and report data on their of mitigation outcomes. NDCs are diverse in GHG emissions and the performance of nature, with some countries using business-as- mitigation activities in their countries, called usual emissions projections as their reference “collaborative approaches” under the article. point, while others use the emissions targets The Article 6 rulebook adopted at the 26th from a baseline year or emission intensity Conference of the Parties to the United Nations per unit of economic outputs as their point Framework Convention on Climate Change of reference. Furthermore, the bottom-up (COP26) describes the reporting requirements nature of market mechanisms could generate for participating parties. These requirements a variety of mitigation outcomes, which could cover an initial report on, among other things: make it difficult to compare and trade units across different mechanisms, especially when The description of the country’s NDC NDC targets are in non-GHG metrics. A quantification of the mitigation information in the NDC, in tonnes of carbon dioxide equivalent (tCO2e) 1 D MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS A quantification of the NDC, or the portion in under the Paris Agreement. Once D-MRV systems the relevant non-GHG indicator are in place, they can be connected to national or international registries. Tradeable mitigation A list of proposed cooperative approaches outcomes can also be tokenized to enable the use with their expected mitigation outcomes and of smart contracts. their alignment with the participating party’s sustainable development objectives This report highlights the potential of D-MRV systems, defines the types of D-MRV systems, Recurring annual information related to the explains the resources needed to move from a internationally transferred mitigation outcomes, conventional to a digital MRV system and the as well as regular information in an annex to its benefits of doing so, explores the necessary biennial transparency reports.2 enabling environment, and provides a tool for assessing monitoring methodologies and their Emerging digital technologies such as artificial suitability for digital technologies. intelligence (AI), smart sensors, and blockchain could help countries meet those reporting requirements The report’s findings are supported by case studies at a relatively lower cost and, potentially, further on D-MRV systems around the world, which are expand and operationalize market mechanisms summarized in Table 1 and set out in greater detail in such as the one referred to under Article 6 of the Section 2. These case studies cover different types Paris Agreement. The use of blockchain technology of mitigation activities, from renewable energy for digital monitoring, reporting, and verification generation to tracking forest growth. (D-MRV) systems is explored in the Climate Ledger Initiative’s recurring annual report, Navigating While there are barriers to the widespread use Blockchain and Climate Action. of D-MRV systems, such as the complexity and upfront costs of implementation, the benefits of D-MRV systems represent the first step in this these systems are expected to be significant. end-to-end digitalization of post-2020 carbon D-MRV systems will reduce the cost of generating markets. End-to-end digitalization includes using carbon assets, increase the transparency and digital technologies and processes to automate the security of carbon market transactions, and even collection of data, the reporting of GHG emissions, make it possible to tokenize carbon assets; conduct and the generation of mitigation outcomes to improve intermittent, system-wide verification of monitoring the functioning of future carbon markets and ease systems; and move towards real-time generation of the burden of satisfying reporting requirements carbon credits. 2 World Bank. 2021. Country Processes and Institutional Arrangements for Article 6 Transactions. Article 6 Approach Paper Series No. 2. World Bank: Washington, DC. [Online]. Available: https://openknowledge.worldbank.org/handle/10986/35392 2 TABLE 1: SUMMARY OF CASE STUDIES NAME COUNTRY STAGE DIGITAL TECHNOLOGY TECHNOLOGY APPLIED APPLICATION Rooftop Solar India Pilot Smart sensors Sensor technology measured energy usage, replacing paper processes Electronic Uganda Operation Smart electric meters Smart meters report on pre- Database and paid electricity usage used to Information calculate consumption for an Management adapted methodology System (EDIMS) Waste to Chile Pilot Smart sensors, distributed Sensor data secured on the Energy ledger technology blockchain is provided to Chile’s MRV system Open Surface, Peru, Chile, Various Blockchain, satellite and Digital imagery of the forests Mangrove Singapore drone data, AI is analyzed by AI algorithms Trust, Wood to identify changes to the Tracking forest and conduct remote Protocol monitoring Sylvera Global Operation Satellite, AI, drones, LiDAR3 Using satellite and LiDAR data in combination with AI to rate nature-based carbon credit projects European Kazakhstan, Pilot Smart sensors, cloud Near real-time digital data Bank for Jordan computing acquisition and processing in Reconstruction the cloud, immutable storage and of raw data, automation of Development verification process D-MRV System for Renewable Energy 3 LiDAR stands for “light detection and ranging” and involves pointing a laser at an object or a surface and measuring how long it takes for the reflected light to return to the receiver. It can be used to make digital 3D representations of areas on the Earth’s surface by varying the wavelength of light. 3 2 Characteristics and types of D-MRV systems M onitoring, reporting, and verification (MRV) An MRV system for GHG emissions estimates, systems have been a core component of the reports, and verifies actual emissions over regulatory framework of market mechanisms a period. In other words, it is an emissions under both the Kyoto Protocol and voluntary carbon inventory. This type of system can be applied markets. MRV is a term used to describe all measures at the national, sectoral, organizational, or taken to collect data on emissions, mitigation facility level. actions, and support. This includes information on methodologies, assumptions, and data parameters An MRV system for support tracks flows used. Typically, such systems were used both to and evaluates the impact of various means determine GHG emissions for cap-and-trade systems of implementation, such as financial support, in Annex I countries, and for mitigation actions in technology transfer, or capacity building. non-Annex I countries to quantify and verify emission This type of MRV can be performed at the sub- reductions under CDM projects. national, national, or multinational organization level.4 Due to the bottom-up nature of the Paris Agreement, participating parties are required to regularly submit Under the Kyoto Protocol, the monitoring, reporting, information on projects and programs related to their and verification of GHG emissions and emission Article 6 activities. The reporting requirements are reductions was costly and time-consuming, intended to ensure the integrity and transparency negatively impacting the liquidity of carbon markets. of Article 6.2 cooperative approaches. As a result, Project data was manually collected and recorded in post-2020 carbon markets will probably rely even hard copy or on unsecured electronic worksheets, more on MRV systems than under the Kyoto Protocol. and verification required on-site inspections by third-party auditors, which could take weeks to Robust MRV systems will be a core component of complete for projects that involved dispersed the infrastructure needed to support post-2020 technologies in rural areas, such as cookstoves carbon markets. The World Resources Institute or solar lamps. Private sector project developers defines three main types of MRV systems: typically had to wait two or three years before any emission reductions, or revenue from their sale, An MRV system for mitigation action could be realized. For smaller projects, the effort and assesses changes in GHG emissions and cost of MRV caused long delays in issuing carbon other sustainable development metrics before credits, affecting project cash flow. Many of these and after mitigation action has been taken, or MRV processes continue today. against specific reference levels. This type of MRV can be performed at the national, sub- To ensure well-functioning, liquid carbon markets national, sectoral, organizational, facility, under the Paris Agreement, there needs to be a or project level. concerted move towards reducing the cost and time needed to certify emission reductions. One way to achieve this is to use emerging D-MRV technologies to streamline MRV processes. 4 World Resources Institute. No date. 3 Types of Measurement, Reporting, and Verification (MRV). [Online]. Available: https://www.wri.org/data/3-types-measurement-reporting-and-verification-mrv [Accessed April 22, 2022]. 4 What is a D-MRV system? Table 2 and Table 3 highlight some of the key differences between conventional MRV and D-MRV Recent technological developments and innovations systems for GHG emissions and mitigation action. to reduce the cost of emerging technologies have While digital technologies are also being used to opened the door to the use of AI, machine learning, bring transparency and efficiency to support MRV satellite imagery, blockchain, smart sensors, the processes, this paper does not explore this type of internet of things (IoT), cloud computing, and drones MRV in depth. in MRV systems to fully or partially automate data collection, recording, and processing for reporting The key difference between a conventional and and verification, which are conducted manually in a digital GHG emission MRV system lies in the a conventional MRV. Further digital technologies methods used to collect, process, record, and that could be used in MRV systems, along with report on data. Conventional GHG MRVs manually potential applications, are presented in a recent record fuel and electricity usage, which is time- Asian Development Bank publication on digital consuming and leaves room for human error. D-MRV technologies for climate action.5 systems, on the other hand, collect data in near real time and automate the recording and reporting A well-functioning D-MRV system integrates these functions. These systems can be set to flag outliers, technologies into a single, overarching system, making it easier to detect changes in processes, using common standards of data exchange and errors in the system, or faulty sensors, providing application programming interfaces (APIs) to ensure project developers with the information needed to compatibility and interoperability across the different streamline maintenance. types of D-MRV systems. This interconnectedness allows for another important benefit of D-MRV For mitigation action MRVs, digital systems offer systems, which is the ability to conduct analysis and near real-time data collection and automated infer important insights from data covering multiple reporting plus faster and simpler verification of similar projects in multiple locations. claimed emission reductions. D-MRV systems use technology and data- Mitigation actions cover projects that can be management tools to quantify, communicate, and costly and difficult to monitor and verify, such as authenticate outcomes in near real time. They have decentralized, off-grid power generation; efficient the potential to improve the speed, consistency, household appliances; and forestry projects in and accuracy of reporting, while lowering reporting remote areas. Emerging digital technologies offer and verification costs and increasing the scalability many advantages when leveraged by a D-MRV and security of databases. These systems also have system. Forestry and land-use activities, for example, modular flexibility, allowing project implementers to can be monitored using satellite data, which can be develop the monitoring component of the D-MRV combined with AI and machine learning to better and then connect their system to the reporting and track project performance. Finally, D-MRV systems verification system of another entity mandated by can simplify the verification of mitigation actions, as the government or an international body to oversee the verifying party can remotely access the system reporting and verification. D-MRV systems can also and confirm the reported information, relying on be connected to registries at the sectoral, national, D-MRV system-defined algorithms to ascertain data or international level to streamline monitoring, quality, thereby reducing the frequency of on-site reporting, and verification of GHG inventories or a inspections, or removing the need completely for country’s mitigation activities. applicable project types. 5 Asian Development Bank. 2021. Digital Technologies for Climate Action, Disaster Resilience, and Environmental Sustainability. [Online]. Available: https://www.adb.org/publications/digital-technologies-climate-change [Accessed April 28, 2022]. 5 TABLE 2: COMPARISON OF CONVENTIONAL AND DIGITAL MRVs FOR GHG EMISSIONS CONVENTIONAL MRV D-MRV Monitoring Manual recording of data from different Near real-time digital monitoring of electricity entities within defined boundaries. and fossil fuel consumption/production through smart meters, linked billing systems, and Entities monitor fuel consumption/ equipment sensors. production or electricity consumption/ production through paper receipts, The resources dedicated to data collection can Excel files, or other manual systems. be reduced and redirected to QA/QC, reducing the time, travel expenses, and effort required to Processes typically involve multiple operate the MRV system. people, are prone to human error, and are time-consuming. Notifications/alerts could be built into the system to avoid data gaps or issues developing due to on-site disruptions. Reporting Recorded data is analyzed and Digital MRV systems can generate automatic compiled into a GHG emissions report. reports on GHG emissions using predefined This is labor intensive, may require templates. personnel to follow up on incomplete Emissions data from an automated or manual or incorrect reporting, and requires monitoring process can be seamlessly analyzed, supervisory review of the report. formatted, and reported. The system can flag errors or be programmed to highlight large deviations in reported values relative to historic reports or similar activities in the same year. Verification Step-by-step audit of GHG report to Validation/certification can be undertaken once ensure procedures are followed and at the digital MRV system level to ensure the no human error occurred. Facilitated workflow follows the GHG emissions reporting through manual review of paper or standards. Then GHG emission reports can be electronic records supporting the MRV verified remotely through dedicated verifier report. Hardcopy documents are easily user profiles, making the verification process lost or damaged over time, making faster and cheaper. A degree of automation in review costly and time-consuming. QA/QC could be achieved through screening of data based on predefined rules. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 6 TABLE 3: COMPARISON OF CONVENTIONAL AND DIGITAL MRVs FOR MITIGATION ACTIONS CONVENTIONAL MRV D-MRV Monitoring Manual collection of project data. This For most mitigation action projects, a digital could be paper receipts or end-user MRV can collect project performance data in data collected on a sampling basis for real time. For power projects this can be done decentralized or household technologies. using smart meters. Other activities can use remote sensors, satellite data, smartphone Collecting technology performance/ applications, and other tools to collect, upload, user data is resource intensive and often and analyze data on project performance. involves travel to remote locations. Reporting A monitoring report is prepared in Digital MRV systems generate automatic reports line with the project’s methodology, based on the mitigation action’s methodology. illustrating the steps to determine the The monitoring data is automatically analyzed, measured sustainability benefit of the formatted, and reported. mitigation action. This usually requires Depending on the type of data, triangulation a dedicated methodology or monitoring techniques can be built into the system to expert to gather evidence and project flag errors or outliers and quickly detect data from activity implementers. a hardware problem, such as a faulty sensor. For example, the electrical output of dispersed rooftop solar installations can be cross-checked against weather data on solar irradiance in the area to make sure the output is within an expected range. Verification Step-by-step audit of monitoring report, Verification can be undertaken once at the with manual review of paper or electronic digital MRV system level to ensure the workflow records supporting the report. follows the applied methodology. Field visits are often required to verify Mitigation action monitoring reports from evidence or samples taken during certified systems can be verified faster and monitoring. at less cost. Verification tools can be built into digitized MRV systems through dedicated verifier user profiles. Verification process can be further improved through D-MRV system flagging the monitored data according to predefined rules. 7 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 3 Barriers to D-MRV systems B arriers to the design, development, and Using D-MRV systems exposes operators to effective implementation of D-MRV systems potential data privacy risks because of the need include a lack of technical expertise when to store large amounts of operational data—which designing and implementing the digital solution; a could include information on private citizens—in a lack of availability of relevant technologies in certain digital format. Sufficient measures to protect this jurisdictions; the need to put in place security measures data need to be put in place and maintained. to protect operational data; and lack of familiarity with the emerging technologies used by D-MRV systems. The methodologies used to determine emission reductions from different types of mitigation A lack of technical expertise can affect every actions, as well as carbon standard-level validation stage of the development process, from identifying and verification rules, predate emerging digital appropriate technologies and designing the technologies. This presents a challenge as standard- hardware and software solutions to calibrating setters may need real-world experience of running smart sensors, maintaining distributed ledgers when such systems before allowing them into their carbon using blockchain encryption, and tweaking machine project cycle. learning algorithms. Relevant digital technologies are also generally more expensive and are not Significant investments of both time and money are available in every country. needed to overcome these obstacles to achieve all the functions that separate D-MRV systems from conventional MRVs. 4 Benefits of transitioning to D-MRV D -MRV systems offer improved monitoring, scaling up rural electrification use smart meters reporting, and verification performance to enable users to buy electricity through their relative to conventional MRVs for both GHG mobile phones. emissions and mitigation actions. Satellites and drone data are already being used to analyze forest cover and other land-use Easier, more reliable monitoring projects. In combination with AI and machine Data monitoring is where the greatest chance of learning, the D-MRV can be a powerful tool to incorrect or incomplete data recording occurs. quickly analyze large, remote areas. Conventional MRV systems still use hard copies of raw data from various meters, which are manually These are a few examples of how new technologies inputted to electronic worksheets. Hardcopy data is that are becoming cheaper and easier to implement easily lost, damaged, or destroyed, and the inputting can be used to streamline monitoring for GHG and process is time-consuming and prone to error. mitigation action MRV systems. The case studies in Section 2 provide examples of how these D-MRV systems can significantly increase the technologies are being used. In Uganda, grid- accuracy of recorded data by using digital connected residents use their mobile phones to technologies to collect, process, and store data. This purchase prepaid electricity via their smart meters. reduces the potential for human error and lowers In India, the Ministry of Power has committed to the risk of data loss. replacing existing meters with smart meters across the country by 2025.6 To transition to digital monitoring, systems need to be put in place to automatically record data on Automated reporting site, at the required interval, and to digitally transfer the information to a database. Multiple technology D-MRV systems can be set up to automatically solutions can be used to acquire and record this analyze, process, and report on monitored data in data, depending on the type of MRV: custom formats. Unit conversions, calculations, and statistical analyses required by the methodology Electronic billing systems can automatically that underlies a GHG or mitigation action MRV can report on the consumption of electricity derived be programmed into the system, eliminating the from fossil fuels, provided these purchases chance of human error. Further QA/QC can be built are logged on the unit. Renewable energy into the system, such as noting large deviations in supplied to the grid can similarly be tracked. reported parameters during the monitored period or This technology is used in combination with IoT relative to previous reporting periods. technologies (such as smart sensors) in many pay-as-you-go business models for solar home Standardized reporting formats are built into D-MRV or biogas systems. systems so the monitored results, outliers, or flagged errors are clearly presented. D-MRV systems can Smart sensors can be programmed to capture also allow for seamless integration of the monitored and report on monitored variables—such as and processed data, with narrative descriptions flow rates, electricity, and heat generation—at and comments by relevant stakeholders providing regular intervals. Many countries that are valuable detail on the project status or explaining potential issues with the data that the system might detect. 6 Smart Energy. 2021. India’s Smart Meter Rollout Timeline Released. [Online]. Available: https://www.smart-energy.com/industry-sectors/smart-meters/indias-smart-meter-rollout-timeline-released/ [Accessed: April 28, 2022]. 9 Streamlined verification Initial system-wide validation and remote verification Verification can be the most expensive and time- consuming aspect of MRV systems. Verifiers often D-MRV facilitates a move towards one-time need to sift through large volumes of hardcopy data certification/validation of the entire D-MRV system. to validate a sample of monitored information. Digital Once certified, the system is then able to remotely technologies can streamline the verification process monitor data, while robust system checks ensure in three ways: by allowing for remote verification, by the confidence and credibility of the GHG impacts allowing initial system-wide validation with remote claimed. Appropriate data flagging allows verifiers verification, and by improving accuracy and quality to only investigate flagged data points. of data. All three of these methods reduce the cost and time required to verify GHG inventories and Improved accuracy and quality of data mitigation actions. Real-time data collection from all systems under a GHG or mitigation action MRV reduces the need Remote verification for default parameters or sampling, and increases With remote verification, a third-party verifier can the accuracy of reporting. Mitigation outcomes that log in to the D-MRV system via the internet without are verified using a robust D-MRV system could traveling to an on-site location in order to gain access command higher market value than comparable to a complete set of relevant quantitative data (such outcomes using conventional MRV systems. as on-site measurements), as well as qualitative data (such as geotagged photos of project sites). In fact, the system they connect to does not need to be co-located with the project. This approach can reduce the frequency and length of site visits, especially for remote areas that are difficult to reach. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 10 5 Resources needed to transition from MRV to D-MRV T able 4 summarizes the differences between conventional and digital MRV systems at each stage of the MRV process to illustrate the additional resources needed to transition between these two types of MRV systems. Along with the checklist provided after, this table can be used to assess an existing MRV system and determine which processes or functionalities meet the conditions of a D-MRV system and which can be improved. TABLE 4: SUMMARY OF DIFFERENCES BETWEEN D-MRV AND MRV AT VARIOUS PROCESS STAGES MRV CONVENTIONAL MRV D-MRV RESOURCES NEEDED PROCESS Collection of Data is recorded in hard Data is automatically, digitally Smart sensors and software MRV data copies (forms or report recorded using smart sensors are necessary for collection templates) that are or satellite images at intervals and transfer of information manually collected on a or in real time recurring basis Transfer and Hardcopy data is Data is digitally transferred Digital transfer, storage, storage of stored as is or manually to a secure on-site or cloud security, and backup of data MRV data transferred to an server for storage and backup. electronic format (such as The data can be “hashed” a worksheet) (converted from plain text into cipher text) to a blockchain or recorded in cloud-based, immutable storage (for traditional ledgers) to ensure it hasn’t been tampered with Analysis and Analysis and reporting Algorithms for data conversion Data digitalization and reporting are manual and prone to and analysis have been computer code for analysis human error programmed into the system, and reporting enabling a repeatable process. Reports are produced in different formats Quality Any QA/QC requires Detection of outliers or A reporting system with assurance/ manual, expert review missing data can be pre- algorithms that send quality of hardcopy data or programmed and is done notifications when faulty control electronic worksheets continuously to detect faulty data has been collected and sensors or other outliers needs to be addressed Third-party Recurring sampling Remote access to monitored Dataset control source review and and manual review data. Automatic verification of (backup repository, immutable verification of hardcopy data, data by comparing data with cloud storage, cryptographic often requiring on-site a control copy (or blockchain/ proof or blockchain) inspection cryptographic hash) accessible via the internet 11 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Converting a conventional D-MRV systems can be partially digitalized. For MRV to a D-MRV system example, data collection, transfer, and reporting could be digital without remote access or securing of To further illustrate how to move from a conventional collected data. The case study on the EDIMS MRV in to a D-MRV system, Table 5 below provides a Uganda is an example of a partially digitalized MRV checklist for key D-MRV functionalities across that sees electricity consumption digitally recorded monitoring, reporting, and verification processes. and reported through smart meters. The EDIMS is Ideally, all the technical functionalities listed should also able to calculate emission reductions resulting be met to realize the benefits of a D-MRV system. from this data; however, digital technologies are not being used to ensure data integrity or for remote data validation or verification. TABLE 5: D-MRV CHECKLIST STAGE REQUIRED TECHNICAL FUNCTIONALITY TICK FOR “YES” 1. Monitoring Digital measurement of monitored parameter ☐ Data collection and storage technologies Secure transfer of data to digital storage location ☐ Storage of data on an accessible device ☐ Immutability of collected data ☐ Automatic backup of monitored data ☐ Conversion or transformation of data formats into a standard ☐ format for analysis 2. Reporting Data analytics that comply with the methodology and provide ☐ Data processing repeatable results have been programmed into the system and analytics technologies Generation and sharing of monitoring reports ☐ Automated QA/QC to detect data outliers or sensor malfunction ☐ 3. Verification Remote access to monitored data and automated verification ☐ Data processing through rule-based data flagging (for example, automated QA/QC) and analytics technologies Ability to compare data against the backup raw monitored data ☐ D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 12 6 Elements of an enabling environment for D-MRV systems M oving towards widespread use of D-MRV The domestic and systems, and realizing the benefits of these legal environment systems, will require overcoming multiple barriers. An enabling environment for emerging digital To promote the implementation and use of D-MRV technologies—and the systems using them—will systems, governments need to ensure that facilitate overcoming these barriers. regulations, policies, and guidance on how sensitive data is collected, secured, stored, and used are in For project entities to adopt digital technologies place. While both conventional MRV and D-MRV for MRV systems, it is crucial to have a clear picture systems may contain confidential information on on domestic policy requirements (or support to businesses, private citizens, or other stakeholders, adopt them); understand the benefits of adopting the digital format of information in a D-MRV system such systems (mainly to reduce transaction costs could increase the risk that the data could be and minimize operating and maintenance costs); accessed and distributed by unauthorized users. and fully understand the roles and responsibilities associated with better data management. The importance of data-management controls depends on the sector or type of mitigation activity An enabling environment seeks to provide the being monitored. Some MRV systems contain policies, institutions, regulations, and infrastructure customer contact information, while others work conducive to the desired outcome, in this case the only with weather data or satellite images that may widespread use of D-MRV systems. There are four already be publicly available. Guidelines on the key aspects to an enabling environment: privacy or security of sensitive information should be provided, and the D-MRV system operator should A conducive domestic and legal environment inform individuals providing data to the system about how their data may or may not be treated and A conducive institutional environment used. If the data provider is a private person, they should be given the option to provide consent for An enabling private sector and NGOs their personal information to be used. Social dialogue and public participation. 13 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Countries should explore how they can use as the regulation underpins what data the system technologies that are already in place for other can collect and how the data is managed.7 This services for D-MRV, or how they can use D-MRV highlights the importance of policy or regulations to promote or foster technologies for purposes around MRV systems and data collection to facilitate that the country has a strategic interest in. In the the use of D-MRV systems by both governmental power sector, for example, digital monitoring and non-governmental organizations in a country. and reporting of electricity generation and consumption is often already standard practice, The institutional environment with relevant guidelines and policies in place. These existing D-MRV systems and guidelines can In addition to national regulations, policies around be leveraged both to track emissions or mitigation data collection, storage, and use need to be in outcomes from the power sector, and to inform the place. An organization managing a D-MRV system, development of D-MRV policies for other sectors. or accessing data from one, should have internal guidance or protocols in place specific to the type of data it is handling. D-MRV systems for reforestation Where guidelines or policies for D-MRV systems or afforestation activities, for example, may monitor still need to be developed, the following should be land ownership by individuals to track the number considered: of trees on the property. This would require different data-protection protocols to a D-MRV that tracks power The reporting requirements issued by the production at large-scale renewable energy facilities. government with respect to a particular carbon pricing initiative or international agreements Regulatory bodies and international standards that such as the Paris Agreement have protocols and rules to assess compliance need to update these standards and protocols to Minimum security standards for data, for allow for use of D-MRV systems. For example, the example, encryption and backup requirements, international standards that certify carbon credits, how data should be secured, and which users such as the Verified Carbon Standard (VCS), the Gold have access Standard, the CDM, and the Article 6.4 mechanism, How consumer or end-user data should be will need to revise their monitoring methodologies collected, terms-of-use agreements, how long and verification protocols to allow for the use of data may be stored, user consent processes, digitally monitored and reported project data, as and anonymizing data already initiated by the Gold Standard in its clean cooking monitoring methodology. The European What type of personal data is eligible for Bank for Reconstruction and Development (EBRD), aggregation under the system. with support from the Spanish Office of Climate Change and inputs from the Joint Multilateral As an example, Turkey adopted regulations for Development Banks (MDB) Working Group MRV systems when the country developed a data- on Article 6, commissioned a protocol for D-MRV management system to comply with the reporting systems using renewable energy generation requirements of the Paris Agreement. In the lessons projects as an example. The protocol provides learned report, understanding the stipulations of the a blueprint for data collection, automating data regulation on MRV systems was the first step taken processing and reporting, remote verification, and towards developing the data-management system, system-wide certification of projects using D-MRV systems. The case study supporting this paper, titled EBRD D-MRV System for Renewable Energy, shows the application of the EBRD’s protocol. 7 Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ). 2017. MRV in Practice: Experience in Turkey with Designing and Implementing a System for Monitoring, Reporting and Verification of GHG Emissions. [Online]. Available: https://www.giz.de/en/downloads/MRV_in_Practice_Booklet_2017.pdf [Accessed April 28, 2022]. 14 Beyond data privacy and security issues, To support the implementation of the changes organizations may need to adjust policies and described, mechanisms should be in place for procedures to shift from paper-based to digital institutions, standards, regulators, and verifiers who processes. Monitoring protocols need to be revised are working with D-MRV systems (or are planning to allow digital data in place of self-reported, to), to share their experiences with one another. manually collected data. The use of digitally Examples of such mechanisms include the Joint collected data requires the development of MDB Working Group on Article 68 and the Gold repositories designed to capture information in the Standard’s D-MRV collaboration work.9 data format from the relevant system. Guidelines need to be developed on how to handle outlier An enabling private sector and NGOs or false data, and instructions on usage and interpretation of monitored data should be available. For D-MRV systems to support mitigation actions, The international standards and regulatory bodies digital technologies and skills need to be both should reassess their reliance on PDF-style reports available and affordable. Widespread internet access or provide options for automating the production of and the presence of cellphone towers, especially these reports, while moving towards dashboard-like in rural areas, can facilitate near real-time data systems that can be remotely accessed. collection from technologies like smart meters and digital sensors. Tax exemptions can also be applied In addition to revising monitoring and reporting to drones, smart meters, digital sensors, and other protocols, validation and verification rules need to technologies that make up D-MRV infrastructure to be updated or revised. One of the main anticipated reduce the overall systems cost. benefits of D-MRV systems is cheaper and faster verification of mitigation actions without detracting Verifiers of mitigation actions need to familiarize from accuracy or environmental integrity. This can themselves with D-MRV systems so that they can be achieved by initially validating the monitoring implement revised protocols from regulatory bodies system, after which system reviews can be limited and international standards. This includes validating to data anomalies. Verification can be supported systems using blockchain for unique identification by remote access to monitored data and eventually in place of hardcopy records, or employing AI become continuous. Verifiers will need to be able programs that use historical data to predict to ensure that reporting results are repeatable and project performance. Capacity in implementing prove the integrity of datasets. organizations and third-party verifiers can be developed by hosting information and training Successfully moving towards this type of system- sessions on D-MRV systems, by demonstrating how wide certification and remote, automated verification to harness the integration of emerging technologies will streamline the generation of carbon credits, into these systems, and by engaging auditors to potentially allowing for near real-time issuance as directly review and mock-certify pilot D-MRV systems. the underlying project(s) are monitored. Periodic audits of the D-MRV system may still be required, Sylvera, a company that provides independent but overall using system-wide certification in place ratings for carbon credits from nature-based of recurring, labor-intensive verification processes solutions, uses D-MRV systems and AI to develop will greatly reduce the time and cost of monitoring, models to assess different performance criteria, and real-time generation of carbon credit assets will such as above-ground biomass and changes in improve the liquidity of carbon markets. forest cover across project areas. The company then assigns a rating to the project so that carbon credit 8 World Bank Group. 2018. MDB Working Group on Article 6 of the Paris Agreement. [Online]. Available: https://www.worldbank.org/en/topic/climatechange/brief/mdb-working-group-on-article-6-of-the-paris-agreement [Accessed April 28, 2022]. 9 Gold Standard Foundation. 2021. Open Collaboration on Next Generation Digital Solutions for MRV. [Online]. Available: https://www.goldstandard.org/blog-item/open-collaboration-next-generation-digital-solutions-mrv [Accessed February 18, 2022]. 15 buyers can assess the quality of those projects. Taking a multifaceted approach to Sylvera raised $32 million in private capital, developing an enabling environment demonstrating the need for assessments of the quality of different types of carbon credits.10 Sylvera Creating an enabling environment for D-MRV plans to expand beyond nature-based solutions systems requires a multifaceted approach. National and provide ratings for all types of carbon credit regulations on data collection, management, projects. The company’s service and underlying and use are needed so that systems operators D-MRV systems are one of the case studies shown understand the framework within which they can in Section 2. operate. A comprehensive and clear policy on digital technologies and data-management systems Social dialogue and simplifies the implementation of complex D-MRV systems. Regulations—in combination with further public participation government support for D-MRV systems through Making D-MRV data publicly available has many tax exemption of digital technologies, a robust benefits. For example, some D-MRV systems telecommunications network, or promotion of have the potential to aggregate large amounts of internet connectivity in underserved regions—can information that can be useful beyond tracking GHG help reduce the cost of transitioning to D-MRV. At emissions or emission reductions from a mitigation the same time, international standards on emission action. Data can be publicized to promote the reporting and emission-reduction certification need technology or measure being tracked, and projects to update their protocols and guidelines so that that make their data available will benefit from implementers of D-MRV systems, which take on the crowdsourced QA/QC. increased complexity and cost of these systems, can also see the benefits. However, there is also the need to protect personal privacy and respect data ownership rights. Proper A national framework for the creation of an enabling safeguards need to be put in place. For example, environment for D-MRV systems would be a valuable the rooftop solar home system case study in India, tool for countries seeking to meet the reporting managed by the National Institute of Solar Energy requirements of the Paris Agreement, specifically (NISE) in India, tracked solar energy generated by Article 6. The development of this national rooftop solar systems. NISE considered making the framework would benefit greatly from capacity generation data publicly available, but the solar building and education of relevant stakeholders technology vendors providing the data objected on D-MRV systems, the underlying technologies, as these companies were charging consumers a and best practices. Lessons learned from piloting subscription fee to view their generation data. of digital technologies for GHG inventories or mitigation actions should be widely shared to avoid repetition of mistakes and ease the uptake of these technologies. 10 Verdantix. 2022. Sylvera Fund Raising of $32 Million Will Enhance the Integrity of Nature-Based Carbon Credits. [Online]. Available: https://www.verdantix.com/blog/sylvera-fund-raising-of-32-million-will-enhance-the-integrity-of-nature-based-carbon-credits [Accessed April 28, 2022]. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 16 An example of a framework for D-MRV systems is The independent standards have accumulated a paper by Microsoft and the InterWork Alliance extensive experience on suitable guidelines for titled Digital MRV Framework: Digital Measurement, development of mitigation actions and monitoring Reporting & Verification Framework.11 This white of those activities. This experience is reflected paper suggests terminology and processes that in the library of monitoring methodologies and should underpin any D-MRV system so that there project standards/protocols governing project is a common understanding and foundation from development, validation, monitoring, verification, which to build D-MRV systems and the required and issuance. These documents are regularly enabling environment. updated to streamline usability. As discussed in the section on an enabling environment, to promote the One of the main criticisms of the carbon use of D-MRV systems, monitoring methodologies credit generation governed by independent and project standards need to be adapted to reflect standards—whether for compliance purposes digital technologies and processes. under the CDM or voluntary markets, as with the Gold Standard and VCS—is the time and cost of Two case studies included in this report have monitoring and verifying projects. As discussed already taken steps to adapt available monitoring earlier, D-MRV systems could significantly reduce methodologies for D-MRV systems. The case these barriers, while maintaining the rigorous study on EDIMS in Uganda explains how a CDM accuracy and quality requirements of certification methodology on rural electrification activities was standards for post-2020 carbon markets. revised to allow for monitoring data to be reported by prepaid smart meters. This was a deviation from conventional electricity meters, which record power usage after the fact, since the prepaid meters report on electricity purchased before consumption, on the assumption that users would consume the electricity they purchased. The D-MRV system covering waste in Chile also adapted an existing monitoring methodology to accommodate digital technologies, and is aiming for system-wide verification. 11 Microsoft Corporation. 2021. Digital MRV Framework: Digital Measurement, Reporting & Verification Framework. [Online]. Available: https://interwork.org/wp-content/uploads/2021/11/Digital-MRV-Framework-1.0.pdf [Accessed April 28, 2022]. 17 7 A parameter assessment tool for methodologies under D-MRV systems M ethodologies will need to be revised to Can the parameter be substituted with an accommodate AI, smart meters, remote approved default value? sensors, and satellite or drone images. The If the parameter is expensive or time- Gold Standard is expanding its project cycle to include consuming to monitor, a default value may be use of D-MRV and in July 2021 revised its cookstove provided in the monitoring methodology. If this methodology to allow for “direct measurement and value is representative of the conditions of the remote monitoring methods”.12 To further support project, it can be applied and a D-MRV system this process, this chapter includes an assessment is not needed. Default values are provided for tool to determine if a parameter, as defined in each project parameters such as wood savings from methodology, can be incorporated in a D-MRV system. efficient cookstoves and emission reductions from solar lamp usage. To facilitate the transition to D-MRV systems, a high- level tool was developed to assess if a parameter Are digital technologies available to measure can be monitored and reported under a D-MRV and record the parameter? system. The tool consists of a decision tree that For parameters for which no usable aims to establish if a parameter should be tracked default value is provided in the monitoring using digital technologies, given that the use of methodology, the next question is whether it these systems and technologies can be costly. The is possible to use a D-MRV system to measure tool considers the following questions, which are that parameter. The digital technologies and not necessarily sequential and could be asked in a underlying systems that can monitor the different order: parameter need to be identified. It is possible that some parameters cannot be measured Is the monitoring, reporting, and verification under D-MRV systems. An example of this of the parameter costly or time-consuming? is the average household size in a project The decision tree begins by assessing the cost boundary or country. This information may be and effort to monitor the parameter in question. available from studies, but cannot easily be If the cost and/or effort is relatively low, then measured by a D-MRV system. a D-MRV system is probably not necessary. An example of such a parameter is the lifetime Is the use of these technologies for this purpose allowed by governing laws or of a technology or product, which is easily regulations? identified by technical specifications, contracts, or warranties, and is constant for the life of the If a suitable digital technology is identified to technology. measure and record the parameter in question, any applicable laws and/or regulations should be reviewed to ensure that using the identified technologies to collect the desired data is permitted. 12 The Gold Standard Foundation. 2021. Reduced Emissions from Cooking and Heating Version 4.0. p 31. [Online]. Available: https://globalgoals.goldstandard.org/standards/407_V4.0_EE_ICS_Reduced-Emissions-from-Cooking-and-Heating-TPDDTEC.pdf [Accessed April 28, 2022]. 18 Is the mitigation activity using a D-MRV If no modification is required, and the above system for other parameters? steps have been followed, the parameter is If steps 1-4 are satisfied—meaning that it is suitable for a D-MRV system. worthwhile to monitor the parameter digitally, If modification is required, determine if the no suitable default value is available, digital cost of doing so is unviable, too costly, or technologies are available to monitor the difficult. If not, the parameter is suitable for a parameter, and the identified technologies are D-MRV system. legally permitted—the next step is to identify if a D-MRV system is already in place for the Figure 2 provides a general framework for assessing mitigation activity for a different purpose whether a parameter specified under a monitoring (for example, a supervisory control and data methodology is suitable for a D-MRV system. The acquisition system in a power grid). If so, the affordability of digital technologies and the underlying monitoring of the new parameter might be system infrastructure is a key consideration when possible using the existing system. Otherwise, making the final decision. The D-MRV systems in the a D-MRV can be built from scratch. case studies all had external funding—usually from If yes, then can the monitoring and reporting development banks or donors—to develop the D-MRV of the parameter be incorporated into the systems. Monitoring methodologies may also need to existing system? be updated to include the use of digital technologies If a D-MRV system is already in place, for monitoring, reporting, and verification. determine if monitoring of the new parameter, and any new digital technologies, can be built When applying this tool, it is important to view into the existing system. digitalization of an individual parameter in the context of the wider scope of MRV of a particular If no, is developing a D-MRV for the parameter mitigation activity, since typically there are multiple cost prohibitive? parameters to be monitored. Even if a parameter is If no D-MRV system is in place, evaluate the not suitable for a D-MRV system, there could still be cost of developing the system. a benefit to using a D-MRV for the overall project. For instance, default values could still be built into D-MRV, Does the digital monitoring of the parameter while the other parameters are digitally monitored require a modification in the project and integrated into D-MRV to yield GHG results. technology or activity? The next step is to determine if the technology or service implemented under the mitigation activity requires modification to accommodate monitoring and reporting under a digital system. An example of this is installing digital heat sensors on efficient cookstoves to track usage and operating temperature. 19 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS FIGURE 2: DECISION TREE FOR ASSESSING IF A PARAMETER IS SUITABLE FOR A D-MRV SYSTEM Should a parameter be covered under a digital MRV system? STEP 1 Is monitoring, reporting, and verification of NO No digital MRV is needed the parameter costly or time-consuming? YES STEP 2 Can the parameter be substituted YES No digital MRV is needed with an approved default value? NO STEP 3 Are digital technologies available to NO measure and record the parameter? YES STEP 4 Is the use of these technologies NO for this purpose allowed by governing laws or regulations? YES STEP 5 Is the mitigation activity using a digital MRV system for other parameters? NO YES Can the monitoring and reporting NO Is developing a digital STEP 6 of the parameter be incorporated MRV for the parameter into the existing system? cost prohibitive? YES YES NO STEP 7 Does digital MRV of parameter require a modification in the project technology or activity? NO YES STEP 8 NO Is the required modification YES Parameter is suitable for cost prohibitive? digital MRV system Parameter is not suitable for digital MRV system D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 20 Common mitigation activity Blockchain encryption technology can be applied parameters under D-MRV systems alongside D-MRV systems, regardless of the digital technologies used to collect the project data, to Table 6 provides a summary of commonly monitored ensure the integrity of the information. Blockchain parameters for mitigation activities to help project technology and immutable data storage in the cloud developers assess if and how these parameters can support post-2020 carbon markets to transition can be measured under a D-MRV system (see the to one-time certification of D-MRV systems with Appendix for the full table including examples, recurring remote verification of project data. monitoring frequency, reporting measures, possible remote verification methods, and parameters for which D-MRV systems are not necessary or beneficial). TABLE 6: COMMONLY MONITORED PARAMETERS AND CORRESPONDING DIGITAL TECHNOLOGIES PARAMETER DESCRIPTION MONITORING DIGITAL QUALITY METHOD/S TECHNOLOGY CONTROL EGgrid13 Electricity generated Electrical meters Smart meters (or Power generation during a time period at interconnection digital meters data recorded by by a gird or mini-grid points can measure with data loggers) smart meters can connected plant power generation continuously record be uploaded to and report power D-MRV and cross- generation checked against other independent variables (such as solar irradiance) Number of Number of project Sales records, Sales or installation Cross-checked with units systems or users installation data using wholesale purchase under activity certificates, number smartphones or or import data of connected purchasing software, systems system monitoring software Carbon stock 14 Tonnes of carbon Satellite data or in- Use AI to analyze Periodic in-person dioxide per volume person inspection satellite or drone inspection to verify or mass of biomass of biomass data to measure reports from satellite deforestation data and AI software 13 This parameter is monitored using a D-MRV system in the India rooftop solar and EBRD solar photovoltaic case studies. 14 This parameter is monitored using a D-MRV system in the case study on forestry systems. 21 8 Conclusion and way forward T he transition to D-MRV systems, made Initial barriers to D-MRV systems can be more easily possible by emerging technologies, will have overcome with a strong enabling environment the potential to become a core component of for digital technologies. The support of key post-2020 carbon markets. Forestry and renewable stakeholders—whether governments, multinational energy projects are already relying on satellite data, AI, organizations, or international standards—for D-MRV smart meters, and digital sensors to record and report systems is a necessary first step. This includes the emission reductions (or removals). Examples of these following components: D-MRV systems already in pilot or implementation phases are shown in the case studies in Section 2. Governments developing policies and frameworks for data collection, management, In addition to cheaper and faster data collection, and use. securing the information through blockchain technology or immutable cloud storage is important Existing international voluntary standards to facilitate the verification of emission reductions adapting monitoring methodologies and from mitigation activities. The Wood Tracking standard-level requirements for validation and Protocol (WTP) project in Peru demonstrates how verification to allow, and accommodate, digital blockchain can be used to track a product through its technologies and D-MRV systems. development or supply chain. Beyond MRV systems, New country-level crediting frameworks and blockchain technology is also being considered to emerging international compliance mechanisms secure transaction registries to track the transfers such as CORSIA (the Carbon Offsetting and of mitigation outcomes between participants under Reduction Scheme for International Aviation) Article 6 of the Paris Agreement.15 and Article 6.4 designing their protocols and methodologies with D-MRV systems in mind. While D-MRV systems and the underlying technologies have the potential to streamline Governments and multinational organizations carbon asset generation and reduce the cost of providing financial support and technical verifying emission reductions, there is an upfront expertise to overcome the initial barriers to cost and increased complexity that needs to be implementing D-MRV systems, including through overcome. Once in place, though, it is likely that the practical demonstrations of pilot systems. cost of scaling or replicating a D-MRV system will be less than for a conventional MRV system. In addition, the potentially modular structure of D-MRV systems can reduce the cost to implementers of mitigation activities by providing access to digital reporting and verification systems at the sectoral or national level, and focusing their efforts on digitizing the monitoring of their activity. 15 Climate Ledger Initiative. 2021. Navigating Blockchain and Climate Action. [Online]. Available: https://www.climateledger.org/resources/CLI-Navigating-Report-December-20213.pdf [Accessed April 28, 2022]. 22 In addition to the enabling environment, further A strong enabling environment and concerted piloting of D-MRV systems and their functionalities efforts to pilot a range of emerging technologies will in different regions for a range of project types will support the widespread adoption of D-MRV systems. allow for further testing of emerging technologies This could unlock carbon markets’ full potential to and the development of a knowledge base to combat climate change while setting the stage for optimize, replicate, and scale D-MRV systems. In future innovations such as the tokenization of carbon addition to the case studies highlighted in this paper, assets through blockchain technology and real-time the World Bank’s Climate Warehouse initiative is issuance of mitigation outcomes from projects with simulating the interconnection of D-MRV systems system-wide certification of their D-MRV. and carbon market registries to demonstrate how these systems can share information. These lessons learned will have benefits across project types and regions. While the data used to train an AI to track forest growth in South America may not be applicable to Southeast Asia, the same machine-learning algorithm could be used with a different training dataset. Further, blockchain technology to secure the data of a D-MRV is mostly the same regardless of the technology used for data collection. 23 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Section 2: Case studies JORDAN INDIA European Bank for Reconstruction Grid-connected rooftop solar and Development D-MRV system photovoltaic system for renewable energy p29 p48 Global Mangrove Trust’s GROVE: Forestry Smart Ledger p39 PERU Wood Tracking Protocol D-MRV p38 Third-party ratings of nature-based carbon credit projects UGANDA p43 Electronic database and information management system p32 CHILE GABON Waste-to-energy landfill D-MRV Third-party ratings of nature- p25 based carbon credit projects p43 Open Surface reforestation/ afforestation D-MRV p37 1 Chile waste-to-energy landfill D-MRV Type of system: Mitigation action T he Chile waste-to-energy D-MRV system was Technical details piloted at a landfill capture-and-destruction The use of digital/IoT sensors system at the Copiulemu landfill in Chile between January 2020 and March 2022. It was a project The D-MRV was supported by various digital under the Reciclo Orgánicos Program, a partnership technologies. On-site digital sensors recorded data between the Canadian and Chilean governments to related to the volume and composition of biogas support the implementation of Chile’s NDC under the produced at the landfill, in accordance with the MRV Paris Agreement. The mitigation action did not pursue methodology. The volume of biogas was recorded registration under an international carbon credit in cubic meters and monitored every 15 minutes. standard because the emission reductions will be By measuring the composition of the gas, it was counted towards the country’s NDC. possible to record the percentage of methane in the biogas. The destruction of methane resulted in the The D-MRV system for tracking and reporting project’s emission reductions due to the high global emission reductions at Copiulemu was built on top warming factor of the methane gas (approximately of the facility’s newly commissioned, state-of-the-art 25 times that of carbon dioxide). measurement and management system. The D-MRV system tracked gas flow rates, gas composition, and The digital sensors recording gas flow and combustion efficiency, among other metrics. composition reported the data to a dedicated server at the landfill, from where it was uploaded to the An online, multistakeholder process was followed to cloud. A dedicated on-site computer was used for develop a landfill gas methodology for Chile based the D-MRV system to avoid the complexity (including on several established methodologies and protocols. security and compatibility issues) of installing the The D-MRV system itself was jointly developed D-MRV software on the landfill operator’s computer by the IOTA Foundation and ClimateCHECK, a system. Monitoring and data records that acted Canadian company that provides measurement, as supporting evidence—for example, photos of reporting, and verification for climate, cleantech, sensors and equipment, calibration records, sensor and sustainability solutions. manuals, and monitoring plans—were incorporated into the D-MRV system. A 3D digital twin of the This first phase of the project covered testing of project site and sensors was incorporated into the monitoring and data systems, such as direct the portal’s user interface to enable a virtual audit monitoring of system data, quality assurance, device user experience. The D-MRV portal was integrated connectivity, and information security. The main with ScribeHub, an online document collaboration parameter tracked by the D-MRV system was the platform, to enable customizable online project capture and burning of biogas, which was recorded reporting and verification reporting according to the at 15-minute intervals, covering about 100,000 MRV methodology and international standards. unique data points per year. 25 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS FIGURE 3: DIAGRAM OF CHILE’S WASTE-TO-ENERGY D-MRV SYSTEM Project type Biogas: Electricity Methodology Chile GHG Emission Reductions Quantification Protocol for Landfill Gas Capture and Destruction Project developer Duero Energía Copiulemu SpA Sensor D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 26 Information security through Full-cycle digital MRV distributed ledger technology This D-MRV system was unique as it was designed An innovative component of the D-MRV system to cover the full MRV process, from developing the was the use of distributed ledger technology (DLT, MRV methodology to the verification of emission popularly known as blockchain) to secure data reductions. Figure 4 shows the different components collected by the system. The D-MRV used IOTA’s of the D-MRV system. The project site covers all permissionless DLT, which uses a directed acyclic hardware involved in the D-MRV, including digital graph rather than a traditional blockchain. This sensors and the on-site computer. means that data is not stored in a single chain and can have multiple branches, enabling transactions The figure illustrates how monitoring was conducted to take place simultaneously. Unlike blockchains, on site and how the collected data was secured using IOTA’s DLT enables energy-efficient performance the IOTA DLT technology and stored in the D-MRV and high transaction throughput with no fees for portal. The other main component of the D-MRV transferring data. IOTA’s technology can also be system was the online verification reporting through used for data marketplaces (MRV data can be used ScribeHub, which streamlined processes that are for multiple purposes) and for the tokenization of expensive and time-consuming in conventional carbon credits generated by the mitigation action MRV systems. underlying the D-MRV system. Lessons learned and continued Adaptation of landfill use of digital technologies gas MRV methodology To ensure that the D-MRV system could support The use of DLT the MRV of emission-reduction certification from While Chile’s waste-to-energy D-MRV system has the underlying mitigation action, ClimateCHECK only recently completed its pilot stage, some notable managed an online, multistakeholder process to lessons were learned. develop a custom MRV methodology that was adapted from: At the conceptualization stage, there was concern that the DLT would increase the complexity of the The Quebec Ministry of Environment’s Cap and D-MRV system. In practice, the implementation of Trade Protocol: Landfill Sites—CH4 Treatment the DLT was simpler than expected as IOTA’s DLT or Destruction is specifically designed for these types of D-MRV systems. The Chilean government also had little Offset Initiative Protocols for Ontario’s Cap and difficulty in endorsing IOTA’s DLT and understanding Trade Program: Landfill Initiative Protocol how the DLT encryption ensures the integrity of collected data for the purpose of validation and Climate Action Reserve’s United States verification. Landfill Protocol. Productization of D-MRV system Most internationally recognized MRV methodologies were developed without specific consideration of The pilot phase was implemented at a single landfill. digital technologies. As a result, projects cannot As a next step, under the Reciclo Orgánicos Program, access the full benefits of digital technologies the system will be expanded to cover other landfills without customizing the MRV methodology. The with methane capture-and-destruction measures. customized Chile Landfill Gas protocol allowed for The system will aggregate emission reductions quantification of emission reductions using the data across Chile’s waste sector and transparently report collected by the on-site digital sensors, as well as the sector’s contribution to Chile’s NDC. Eventually, remote verification due to the security and integrity the D-MRV system can be modified to suit other of data offered by IOTA’s DLT technology. types of mitigation actions. 27 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS FIGURE 4: COMPONENTS AND STRUCTURE OF THE CHILE WASTE-TO-ENERGY LANDFILL D-MRV SYSTEM 16 Digital system at site Alvarium Manual Dell Edge data data entry server(IOTA IOTA Tangle Back end confidence Distributed Ledger (spreadsheet) Node + DCF) annotations IOTA cloud node + Alvarium author CLI Activity 1 Export Programmable Industrial Network logic controller router switch router Front end APIs • 360° viewable 3D digital twin with digital sensors for UI/UX • Data logs (real-time, historical) Digital sensors SCADA system Incoming Validated onsite report • QA/QC, calibration records, sensor data data manuals, photos Activity 2 • Integrated GHG calculations • Integrated data confidence score SCADA database Single sign on Exported integration content Online standards Climate and Standards and development system sustainability methodologies for Export programs and measurement, quantification, Online Online Specific MRV monitoring and Registries standards reporting, and disclosure standard for sector, report reporting template Markets technologies, organizations projects etc. Linked Finance (UN, governments, Reports voluntary) Standards for auditing Online verification Verification Certification and assurance reporting remplate report In addition to expanding the geographic and sectoral Systems such as the Chile waste-to-energy scope of the D-MRV system, further functionalities D-MRV are an important part of cost-effective and can be incorporated. From the IT side, future transparent reporting of GHG emission reductions improvements of the D-MRV system include using in sectors that fall under a country’s NDC and have digital solutions to: industrial processes where smart sensors can record data in real time. The innovative use of DLT Report data from the digital MRV system to ensure the integrity of data collected is equally directly to a national GHG inventory important for the integrity of future carbon credit markets. Verify emission reductions in real time as data is collected Link to carbon credit registries and/or market-places. 16 Dell Technologies. No date. Project Alvarium Accurately Tracks Carbon Footprints with Edge Solutions. [Online]. Available: https://www.dell.com/en-us/dt/video-collateral/project-alvarium-tracks-carbon-footprint-with-edge-solutions.htm [Accessed June 14, 2022]. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 28 2 Grid-connected rooftop solar photovoltaic system in India Type of system: GHG emissions T he government of India, in partnership with which provided access to integrated data loggers state governments and electricity regulators, to track generation data. For existing rooftop solar adopted and implemented policies and systems, the smart meters in the D-MRV system programs promoting the deployment of grid-connected were integrated with these data loggers. Greenfield rooftop photovoltaic (GRPV) systems. To determine the solar home systems use smart meters without the success of these programs, an MRV system to collect need for data loggers. data on the performance of these systems was needed. The data would populate a vast database that would: India’s National Institute of Solar Energy (NISE), which falls under the Ministry of New and Allow project developers and financiers to Renewable Energy, oversaw the implementation moni-tor project performance and identify and management of the D-MRV system with financial systemic inefficiencies support from the World Bank starting in June 2017. The aim of the system was to streamline certification Enable electricity distribution companies of emission reductions/mitigation outcomes from a (locally known as DISCOMs) to plan for supply World Bank project. To this end, the D-MRV system requirements monitors and reports data stipulated by CDM- approved monitoring methodology (ACM0002 Grid- Generate market data on the performance of connected electricity generation from renewable various rooftop systems sources) to determine the amount of emission Potentially support the generation of mitigation reductions resulting from the project. The main outcomes for India’s NDCs.17 parameter tracked by the D-MRV system is electricity generated by the rooftop solar photovoltaic system. The expected use of data by various stakeholders had implications for the type of data collected and The D-MRV reached pilot scale, covering about how the data was to be managed. 40 sites with rooftop solar photovoltaic installations with a total capacity of 9.4 megawatt-peak (MWp). The system was developed as a pilot mitigation Currently, there are no plans to expand the D-MRV action D-MRV to track the electricity generation system beyond the pilot stage. of rooftop solar photovoltaic systems. DISCOMs previously relied on data from inverter companies, 17 India announced at the Conference of the Parties (COP21) in Paris that it aims to increase to 40 percent the share of installed electric power capacity from non-fossil-fuel-based energy resources by 2030. 29 FIGURE 5: ARRANGEMENT OF SYSTEM COMPONENTS IN D-MRV Project developer Project owner/consumer DISCOM Inverter with integrated data logger Energy generation Bi-directional meter Solar PV panels meter (net metering) (solar meter) Meter under consideration Technical details the surplus energy is automatically exported (energy export) to the DISCOM distribution network (the A central data monitoring center was set up to grid) with the help of a two-way meter (bi-directional aggregate and analyze the data from the 40 sites at meter). If the solar energy produced is less than the NISE campus in Gurugram in northern India. The the total energy required in the building, then the pilot data monitoring initiative was conducted with shortfall energy is automatically supplied by the grid the support of the Chandigarh Renewable Energy (energy import). The overall amount of grid electricity and Science & Technology Promotion Society and displaced is the net electricity generated by the Chandigarh’s Nodal Renewable Energy Agency, GRPV system. Figure 5 illustrates the arrangements with DataGlen as an implementation partner. As part of the D-MRV and GRPV systems. of this pilot, NISE installed data monitoring devices, developed the data-acquisition methodology, and The smart meter continuously, and in real time, developed a web application for online data analysis measures net electricity in kilowatt-hours (kWh). and data visualizations. These measurements can be reported on a daily, weekly, or monthly basis. The meter integrates with Data collected from GRPV systems supervisory control and data acquisition software. At The D-MRV system mainly measures electricity in the time of piloting, the cost of a conventional meter order to accurately calculate the associated emission was about $10 versus $77 for a smart meter. The reductions. However, where the meter is to be placed resulting cost increase for using a smart meter was and what it should measure are equally important. roughly $67 per connection, increasing the cost of a 1 kilowatt (kW) GRPV by about 12 percent, to $570. For GRPV projects, the concept of net metering is a consideration. In GRPV systems, the direct current To determine emission reductions from the GRPV generated by the solar panels is converted into systems, the net electricity monitored by the alternating current by a solar inverter. This is then smart meter is the only parameter that needs to be fed to the building’s distribution board before it is tracked. This measurement is then used to calculate consumed by electrical appliances. If the electricity the emission reductions from the project, based on produced is more than what the building consumes, the regional or national grid emission factor. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 30 Using weather data for QA/QC Mitigating the cost of The D-MRV system draws on weather data to digital technologies determine local solar irradiance over a specified For smaller photovoltaic systems (1 kW), smart period, which is then used to estimate a GRPV meters increased the total system cost by 12 percent. system’s anticipated performance in a given The accounting consultancy EY recommended that, location. This approach can be used for brownfield given the fixed cost of using smart meters with a projects where conventional meters are used, or to GRPV system (an increase of $67 per connection complement smart meter data by ensuring that the compared with a conventional meter), the benefits reported electricity production system falls within of the technology made sense for photovoltaic an expected range, given local weather conditions. systems of 3 kW capacity or greater. This would ensure that the increased cost was small relative to Uses of D-MRV system data the cost of the system. Given that users are willing The reporting of the D-MRV system data was to pay a subscription fee to access the system’s streamlined through a web portal showing the real-time generation data, the cost increase could performance of GRPV systems across the locations. also initially be carried by the DISCOMs and then recovered through subscriptions. While the D-MRV system was initially only used to calculate emission reductions, the data aggregated Considering the needs of the in the portal was found to have other uses too. D-MRV system from the outset Project developers used the data to evaluate the The case study involved both existing and new performance of their systems and develop data- GRPV installations, so providing insight into the backed marketing material. differences between brownfield and greenfield D-MRV projects. The consensus was that it is better During the pilot, NISE and the project developers to use digital technologies for a greenfield project had differing views on the use of the D-MRV data. than to incorporate them into an existing activity, as NISE was in favor of making all system data publicly the use of digital technologies typically comes with available to promote solar technologies, while the increased cost and complexity. If a D-MRV system developers wanted the data to remain private so is to be used, its requirements should inform the that competitors could not benefit from it and it project design from the outset. could be offered to customers as a paid subscription service allowing them access to their system’s real- The mandated use of smart meters time power generation data. The D-MRV system was implemented to facilitate the monitoring required under the Program for Lessons learned and continued Results Financing structure of the project. However, use of digital technologies the use of smart meters showed benefits beyond There are no plans to scale the D-MRV system these requirements, yielding added value to the beyond the initial pilot of 40 locations providing DISCOMs and users. As a result, in August 2019 the 9.4 MWp of solar photovoltaic capacity. However, Ministry of New and Renewable Energy mandated the project yielded worthwhile insights into the use the use of smart meters by DISCOMs for all GRPV of D-MRV systems and digital technologies. systems. Compared to other digital technologies, smart meters are an inexpensive, easy-to-use, and proven technology. 31 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 3 Uganda’s electronic database and information management system Type of system: Mitigation action T he electronic database and information AMS-III.BL Version 1, to measure emission reductions management system (EDIMS) is a mitigation from electrification of communities. This methodology action D-MRV system that tracks new grid was revised to accommodate the use of prepaid smart connections and electricity consumption of those meters to determine the annual power consumption connections using a mobile application, prepaid of a connection. electricity meters, and a web-based online application. It was developed by RMSI, a global geospatial and REA is a semi-autonomous body established by engineering firm, for Uganda’s Rural Electrification the Ugandan government to operationalize its rural Agency (REA) to support the implementation of a rural electrification function by working with service electrification program funded by the World Bank. providers in the country’s 12 different service territories. The service providers are private companies that bid The goal of the EDIMS is to ensure an optimal, for the rights to provide electricity within a service reliable flow of information to allow for the central territory. Before the EDIMS, service providers used review, approval, and tracking of connection either paper or electronic records of their electric progress, while allowing for the validation of certified meters and meter consumption. REA received data emission reductions generated by the new grid on connections and power consumption in the form connections under the mitigation action. The D-MRV of Microsoft Excel worksheets. follows the CDM ex-post monitoring methodology, FIGURE 6: THE EDIMS WEB APPLICATION D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 32 Development of the EDIMS began in March 2017, Generation of certified with implementation taking place in early 2021. emission reductions The system currently holds customer and power In addition to streamlining data collection and consumption data on over a million connections reporting, the EDIMS also calculates the certified throughout Uganda and automatically generates emission reductions generated in line with the reports on certified emission reductions achieved chosen CDM methodology on electrification by service providers for a chosen year. of communities. In terms of this methodology, the annual power consumption of a customer is Technical details multiplied by a fixed emission factor, which varies Data collection and technologies depending on the customer’s power-usage bracket, to calculate the certified emission reductions. In terms of the chosen CDM methodology, projects The EDIMS can generate reports for each service must track the data points noted in Table 7 for each provider over a given time period for a verifier to new connection to the national electricity grid. Before review and certify the emission reductions achieved the EDIMS, power blackouts at service provider by the mitigation action. Figure 7 provides a sample offices could cause lost data. Duplicate meter of a certified emission reductions report produced numbers were found for different customers when by the system. reviewing service provider data. The EDIMS uses multiple digital technologies to allow the prepaid meter to automatically report electricity purchased by the customer, so reducing errors caused by manually recording data and then re-entering it into a database. TABLE 7: PARAMETERS TRACKED BY THE EDIMS SYSTEM NAME UNIT FREQUENCY MEASUREMENT METHOD Power purchased kWh/month Monthly Prepaid smart meters digitally report data to billing system Meter number – Once at connection Recorded through mobile application at time of connection Connection type Label (household Once at connection Recorded through mobile application at or business) time of connection Location GPS coordinates Once at connection Recorded through mobile application at time of connection Connection date DD/MM/YYYY Once at connection Recorded through mobile application at time of connection 33 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS FIGURE 7: SAMPLE OF A CERTIFIED EMISSION REDUCTION REPORT PRODUCED BY THE EDIMS D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 34 Data privacy and sharing Familiar technologies support adoption There is no concern about data privacy or sharing New grid connections in Uganda were already with the EDIMS because data collected by the transitioning to the use of prepaid smart meters. D-MRV system is identical to the data collected Combining this technology with mobile applications manually prior to the implementation of the EDIMS. and an online web application—both familiar technologies—allowed for the development of a Lessons learned and continued fully functioning D-MRV. As such, a large capacity- building effort to train users on these technologies use of digital technologies was not required. However, the implementation time The D-MRV system at REA was implemented as of four years (from the project initiation in Q1 2017 part of a results-based-financing program with to implementation in Q1 2021) required significant the World Bank, under which the World Bank com- commitment of time and resources. mitted to purchasing certified emission reductions resulting from the extension ofUganda’s national grid. The usage of the EDIMS has exceeded expectations. Initially envisioned to be used mainly by REA and The development of the EDIMS was financed by the 12 service providers covering Uganda’s service the World Bank to streamline data collection and territories, the D-MRV is now also being used by reporting of information required to generate the Umeme, Uganda’s main electricity provider, which certified emission reductions. provides power to the service providers and installs new connections. CDM methodology revision As already noted, the methodology covering Possible future applications electrification of communities was revised to The EDIMS has reduced the cost of collecting and accommodate the use of prepaid smart meters. Data monitoring data for REA’s results-based-finance from prepaid meters technically only shows power program with the World Bank. While the benefits are purchased, not necessarily power consumed. To only being realized after four years of development, allow the use of prepaid meter data, the methodology the system provides REA, service providers, was revised to stipulate that consumption for a given and verifiers with easier access to (and greater month is determined by purchases in the previous confidence in) performance reports and generated month. For example, when calculating power certified emission reductions. Human error in consumption for the 2020 calendar year, the EDIMS recording new connections and occurrences of data uses power purchases from December 2019 through loss due to power outages are less prevalent. November 2020. The methodology assumed that electricity purchased by consumers in Uganda The EDIMS currently covers more than a million is used within the following month. Updating or individual grid connections and will continue growing revising ex-post monitoring methodologies to adapt as Uganda works to achieve its electrification goals. to the use of emerging technologies and the data Expansion of the D-MRV to cover off-grid solar home they collect is a significant challenge that needs to systems is also being considered as the mobile be met to enable the use of D-MRV systems. applications can easily be modified to collect data on solar home systems at the time of installation, along with the location of the systems. 35 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 4 Forestry project D-MRV systems F orestry is a major sector in the work to mitigate Technical details the effects of climate change. Tree-planting This case study covers three different D-MRV and forest conservation are often presented systems for forestry-related mitigation actions and as important tools for capturing and storing carbon GHG inventory projects. The systems use digital dioxide emissions from the atmosphere. Projects technologies to track the growth or removal of in this sector have recently benefited from an influx forests. The three systems are: of investment as new digital technologies—such as satellites and drones, AI, and blockchain encryption— Open Surface MRV in Chile make surveying and tracking large areas of remote forest possible for the first time. Developing D-MRV Wood Tracking Protocol (WTP) D-MRV in Peru systems for the forestry sector is also crucial because some projects are criticized for attributing too many Global Mangrove Trust’s GROVE: Forestry emission reductions to their activity. A 2021 study Smart Ledger in India. estimated that 29 percent of forest projects under California’s cap-and-trade system overestimated their climate impact.18 Accurate and cost-effective D-MRV Table 8 summarizes the main characteristics of the systems for mitigation actions and GHG inventories for various D-MRV systems. forestry projects could solve this problem. TABLE 8: FORESTRY D-MRV SYSTEM CASE STUDIES D-MRV TYPE OPERATING IMPLEMENTING PARAMETER/S STAGE COUNTRY/ AGENCY TRACKED REGION (UNITS) Open Surface Mitigation action/ Chile Corporación Area of forest (m2) Pilot GHG inventory Nacional Forestal (CONAF) Wood Tracking GHG inventory Peru Swiss Agency for Volume or mass of Pilot Protocol Development and wood (m3 or ton) Cooperation GROVE: Forestry Mitigation action/ India Global Mangrove Area of forest (m2) Operation Smart Ledger support Trust 18 Reuters. 2021. California Program Overestimates Climate Benefits of Forest Offsets–study. [Online]. Available: https://www.reuters.com/business/environment/california-program-overestimates-climate-benefits-forest-offsets-study-2021-04-30/ [Accessed April 28, 2022]. 36 While the Open Surface and GROVE D-MRV systems The software underpinning the D-MRV is open support mitigation actions, the Wood Tracking source and used by the Corporación Nacional Protocol D-MRV is a GHG inventory system that Forestal (CONAF, Chile’s National Forest tracks the harvesting and processing of wood in Corporation) to monitor forests in Chile’s Valdivia Peru to combat illegal logging. The GROVE D-MRV region. The system tracks 100 m2 areas of land every uses a blockchain ledger to tokenize areas of forest three to five days—when the satellite passes over protected or regrown as part of a project. Each the area again. Over a one-year period, this equals D-MRV is discussed in more detail below. about 500 million data points in the pilot region. Open Surface D-MRV CONAF uses Open Surface’s AI software to comb The Open Surface D-MRV is a mitigation action vast amounts of satellite data to detect and report and GHG inventory MRV system. The application loss in forest cover. Prior to the D-MRV system, can be applied to track reforestation/afforestation CONAF would send a team to the site to evaluate efforts (mitigation actions) or to monitor protection deforestation reports. As a result, the AI had data of existing forests (GHG inventory). Open Surface on historical deforestation events that it could D-MRV is a greenfield MRV system that uses AI to match to changes in satellite images when the comb historical satellite data and compare it with event occurred. This allowed the system to learn reported deforestation events (such as fire and how different events affect the data, so enabling it clear-cutting) to learn which satellite images are to detect deforestation. Figure 8 shows a map of indicative of forest destruction. The system is not recent deforestation events in South America from pursuing carbon finance but could be applied under the Global Forest Watch data portal. a reforestation/afforestation project to track forest growth or maintenance. Open Surface is loosely based on Global Forest Watch’s tools to determine deforestation. The development of the system is 19 Global Forest Watch. No date. Explore Our Data. [Online]. funded by the Inter-American Development Bank Available: https://www.globalforestwatch.org/map/ and Climate-KIC. [Accessed April 28, 2022]. FIGURE 8: GLOBAL FOREST WATCH DATA PORTAL19 37 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Under the pilot, the educated AI analyzes satellite The WTP is not supported by carbon credit data and, if an event is detected, the location is finance because it does not directly track emission checked against the database of areas marked for reductions. The D-MRV system hopes to combat deforestation. If the location is not exempted, the illegal harvesting of wood in Peru by providing incident and location are flagged for inspection by assurance to buyers of wood products that their CONAF. purchases adhere to national guidelines. This will likely reduce deforestation and avoid GHG Before piloting Open Surface, forests were monitored emissions since deforestation is the main source of on a rotational basis. The COVID-19 pandemic GHG emissions in Peru. provided an extra push towards accepting remote, digital processes. Adopting digital processes had The WTP is owned and funded by the Swiss Agency the added benefit of enabling monitors to prioritize for Development and Cooperation through the flagged forest areas and conserve limited staff Climate Ledger Initiative. The pilot phase of the resources for on-site visits. D-MRV covers the Peruvian Amazon region and the main parameter tracked is cubic meters or tons of The main benefit of the D-MRV system is the ability wood. For conversion between mass and volume of to track a large area without requiring physical wood, only the density of the species needs to be inspection. The downside is the relatively high known. cost of AI expertise and access to the satellite data. The initial set-up also posed the challenge The WTP uses a smartphone to measure and of consolidating data from land-usage plans that, tag trees that are allowed to be harvested. The although digital, were stored in different formats. landowner uses an application to tag the tree to be harvested with a picture, its GPS location, and key After piloting, the system will be expanded measurements to estimate the tree’s volume, in line across Chile and to other Latin American countries with Peru’s national guidelines. In Peru, cellphone or regions. numbers are linked to a person’s national identity number so there is assurance that data is coming Wood Tracking Protocol D-MRV in Peru from the landowner. It is estimated that more than 80 percent of wood from Peru is illegally harvested. The WTP is built on The WTP then uses blockchain technology to top of an existing MRV system implemented by Peru securely track the wood from harvested trees as to enforce rules and regulations on harvesting wood it moves through the supply chain. After being and to combat illegal logging. The pre-existing designated for harvesting, the tree and its volume MRV relied on on-site inspections and hardcopy are logged in the blockchain ledger. The person documentation of the dimensions of trees and which cutting the tree follows a similar process, using the trees could be harvested. The D-MRV digitized application to record the volume of wood harvested established procedures required to track wood use, as well as where it will be transported for processing. per the logging guidelines of Peru. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 38 FIGURE 9: WOOD TRACKING PROTOCOL D-MRV PROCESS DIAGRAM20 1 Logging 2 Transport 3 Processing Real world activity Corresponding digitalization Logging relevant data Transport verifies logging data by Processing receives wood and WTP (location/time/size/species) is applying a coherence algorithm. verifies data for coherence. generated via smartphone and If check is positive—logger If check is positive—transport stored on distributed ledgers. receives automated payment. receives automated payment. This process repeats throughout the supply chain, ensuring that local communities benefit from as shown in Figure 9, allowing wood buyers to the projects. Global Mangrove Trust’s goal is to track where the wood comes from and be assured “plant as many trees, as quickly as possible”.22 To that the tree was legally harvested. In future, accomplish this, it is working to create an enabling the D-MRV could scale nationally and leverage environment for mangrove conservation projects satellite data and AI to complement current data. by supporting digital monitoring technologies and It is also envisioned that payment for the tracked methodologies, as well as innovative financing. wood will be automated through the application when predetermined milestones are reached. The GROVE: Forestry Smart Ledger D-MRV system uses satellite data and AI to measure and verify GROVE: Forestry Smart Ledger mangrove conservation and growth projects. Mangrove forests have the ability to capture more The system also has a fintech component, going carbon dioxide than rainforests, usually at a lower beyond monitoring, reporting, and verification to cost.21 GROVE: Forestry Smart Ledger is an open connect mangrove projects with funders through source, greenfield D-MRV being rolled out in India GRO-coin, a blockchain-backed, non-fungible that tracks both the flow of funds to mangrove digital token. conservation projects (support) and the impact of mitigation actions by such projects. Figure 10 shows how GROVE: Forestry Smart Ledger is made up of two components: GROVE, GROVE: Forestry Smart Ledger is championed the blockchain-backed platform that connects by Global Mangrove Trust, a non-profit based funders with projects and tracks the flow of in Singapore that aims to improve the financial money, and Forestry Smart Ledger, which runs viability of mangrove conservation projects by machine learning algorithms on satellite data of making access to carbon finance easier, reducing forest growth to measure the impact of project monitoring and verification costs, and using smart activities and create smart contracts underpinning contracts to ensure financial transparency while GRO-coin. 20 Wood Tracking Protocol. No date. Blockchain WTP. [Online]. Available: https://wtp-project.com/ [Accessed April 28, 2022]. 21 Climate Finance Lab. No date. GROVE: Forestry Smart Ledger— About. [Online]. Available: https://www.climatefinancelab.org/project/grove-forestry-smart-ledger/ [Accessed April 28, 2022]. 22 Global Mangrove Trust. No date. What is GMT? [Online]. Available: https://globalmangrove.org/ [Accessed April 28, 2022]. 39 FIGURE 10: GROVE: FORESTRY SMART LEDGER DESIGN23 GROVE (blockchain-backed platform) Forestry smart ledger (FSL) stack $ Donor/funder 1. GRO-coin GROVE Forest growth Impact (website/app) reporting measurement 2. Impact dividend $ Project listing Planting $ Escrow Smart contracts Conservation Conservation $ linked Community deposit $ to forest growth forest project Forest Conservation Conservation growth Rewards $ trust (DBS Bank) Measure forest growth using satellite + machine learning The D-MRV does not follow a specific monitoring application in order to estimate the total volume of methodology and can be adapted to work with most wood remains a challenging task. The WTP team available forestry and soil carbon methodologies. hopes that using new smartphone models with GROVE: Forestry Smart Ledger is being rolled out in multiple cameras and AI will help automatically India, but Global Mangrove Trust sees potential for it determine the tree’s volume more accurately while to be used across Southeast Asia for mangrove and requiring less input from the user. The combination other forestry or soil carbon projects. of improving the user experience and incorporating payment options in the application will further Lessons learned and continued promote adoption of the WTP by landowners, lumber companies, and buyers. use of digital technologies Training and automating complex tasks Drones and scaling AI to other climates All three D-MRV systems discussed in this section The Open Surface D-MRV team plans to complement have plans to improve on and expand their use satellite data with images from low-flying drones of technology. This could result in higher system and to improve their AI. Drones can provide higher- development costs and increase the need for resolution images of areas where satellite data has training. For example, the smartphone application prompted the AI system to report a deforestation developed by the WTP was initially difficult for event. The use of low-flying drones is currently landowners in rural Peru to understand and use costly, considering the small area a drone can cover effectively. Landowners were willing to use it when relative to satellite data. they were shown how the application simplified the process of designating trees for harvesting. However, entering the dimensions of a tree into the 23 Climate Finance Lab. No date. GROVE: Forestry Smart Ledger—Design. [Online]. Available: https://www.climatefinancelab.org/project/grove-forestry-smart-ledger/ [Accessed April 28, 2022]. 40 Scaling an AI-backed D-MRV to other regions and an impediment for scaling this technology. Open climates that have different colors and patterns Surface circumvented this problem because the of plants, soil, and weather presents a challenge implementing partner, CONAF, owns the data used because the AI needs to relearn the relationship by the machine learning model. between satellite images and deforestation events. A further challenge highlighted by the Open Forest projects need D-MRV the most Surface team is that while the D-MRV system can Regrowing and maintaining the Earth’s forests identify false positives—in other words, where a are important tools for reducing carbon dioxide deforestation event is reported but no deforestation emissions. It is estimated that forests and other is found on inspection—the system cannot easily ecosystems can make up a third of the required identify false negatives (unreported deforestation reduction in emissions to stay below 2oC.25 But events) because these are not inspected. mitigation activities on forestry have unique challenges that D-MRV systems can address. System-wide verification and data-sharing issues Emission reductions from conservation, afforestation, The cost of monitoring and verifying mitigation or reforestation are treated differently than those actions is often seen as limiting the viability of from other mitigation actions, such as renewable emission reduction tracking. Typically, each project energy, efficiency measures, or switching fuel. This needs to engage in the MRV process on an annual is because of the question of permanence. Carbon or biennial basis. D-MRV systems can reduce the dioxide captured in a tree can be released back into time and cost of these processes. the atmosphere in the event of a fire. This is unlike other mitigation actions, which permanently avoid To realize this benefit, the team behind GROVE: more carbon-intensive alternatives. In addition, Forestry Smart Ledger is working towards a one-off, power plants or factories already report their system-wide verification so that any projects using the generation, fuel usage, and other metrics, while verified system are automatically deemed verified, and forests are largely not tracked or monitored at the emission reductions can be certified as project data is macro level. recorded. Global Mangrove Trust is working with the Smith School of Enterprise and the Environment at The technologies used for D-MRV for forest projects the University of Oxford, Kumi Analytics, and Marex also bring positive spillover effects into complementary to develop a sequestration methodology that relies areas. CONAF found it valuable to have consolidated on remote verification, backed by machine learning, visual information that combined historical views of to make verification quick, accurate, transparent, and the forests that were not previously available. All of in real time.24 the use cases use the system to assist with optimizing in-person monitoring activities, and Mangrove Trust One of the hurdles faced by GROVE: Forestry uses the data to determine where new trees should Smart Ledger is a lack of transparent historical data be planted to achieve the greatest impact. showing reforestation progress and growth rates. GROVE uses this data to test and train its machine For these reasons, robust and transparent D-MRV learning models. Global Mangrove Trust described systems are highly important for forestry projects. a lack of willingness by organizations with this data Since 2018 there has been growing interest in to share it, citing legal or financial reasons. This investing in digital technologies to build cost- data can understandably be deemed sensitive as it effective D-MRV systems, with an increase in venture shows the numbers underlying project performance capital for start-ups working in the space (Figure 11). of forestry projects and could invite scrutiny on a The three case studies in this section are illustrative project’s performance or implementation. The lack of the technologies being used to monitor forests, of test data to use for training models could be forests’ carbon stocks, and the movement of funds. 24 Global Mangrove Trust. No date. What is GMT—Our Current Partnership. [Online]. Available: https://globalmangrove.org/ [Accessed April 28, 2022]. 25 Griscom, B. W., Adams, J., Ellis, P. W., et al. 2017. Natural Climate Solutions. Proceedings of the National Academy of Sciences of the United States of America. PNAS: Washington, DC. [Online]. Available: https://www.pnas.org/doi/10.1073/pnas.1710465114 [Accessed April 28, 2022]. 41 FIGURE 11: VENTURE CAPITAL FUNDING FOR CLIMATE TECH START-UPS Global venture capital deal flow in climate tech ($ billion) 18 16 14 12 10 8 6 4 2 0 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021* Angel/seed Early venture capital Later venture capital Global investors have closed as many climate-focused funds in 2021 as the previous five years combined. Seventeen US states, the District of Columbia, and Puerto Rico have adopted policies aiming to move to either all-renewable or zero-emission electricity supplies. The US officially rejoined the Paris Agreement in February 2021. The Biden administration aims to spend $2 trillion over four years on clean energy sources, with a goal of reaching net-zero emissions by 2050. Source: PitchBook *As of June 25, 2021 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 42 5 Third-party ratings of nature-based carbon credit projects P rivate companies are using D-MRV systems Legacy MRV methodologies for forest crediting to provide ratings of carbon credits and the projects rely on manual assessments of sample areas projects they are generated from. Sylvera is within a project, which are then extrapolated across an example of such a company and uses emerging the project, with monitoring undertaken every few digital technologies to produce independent ratings of years. Sylvera’s system accesses raw satellite earth nature-based carbon crediting projects. The company observation data from across entire project sites, as has developed machine learning models that interpret frequently as satellites pass over (usually every two multispectral satellite data to estimate key parameters to four weeks). This geospatial data is fed to machine such as above-ground biomass and changes in forest learning algorithms, which analyze and visualize this cover across project areas. These outputs come with a data, allowing for up-to-date assessments of forest level of uncertainty that can be significantly reduced by cover and condition. Historical satellite data of the integrating on the ground data, which will be gathered project site is also available, allowing for baseline using LiDAR scanning. assessments and better understanding of risks to permanence such as fire history. FIGURE 12: SATELLITE IMAGERY OF FOREST COVER (GREEN) IS COMPARED TO TIME SERIES DEFORESTATION DATA (RED), WITH OTHER DETECTED LAND CLASSES, SUCH AS WATER, ROADS, AND SETTLEMENTS (WHITE) 43 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Sylvera is also using new technology to gather more Organization, support, accurate on the ground data to train the machine and current status learning models. Current methods for estimating Sylvera is a United Kingdom-based start-up founded forest above-ground biomass, and therefore in March 2020. Revenue is generated through carbon stored, rely on allometric models. Estimates subscription to Sylvera’s ratings and analysis data, from these methods can have potentially large available to clients through a web application and uncertainties, often upwards of 40 percent at the via an API. stand-scale,26 particularly in the world’s tropical and subtropical forests. Sylvera’s second-generation Sylvera received seed funding from venture capital machine learning models are trained using proprietary firms, along with a research grant from Innovate UK’s datasets gathered using multi-level LiDAR scanning, Small Business Research and Innovation program. which significantly reduces the uncertainty of above- This grant funded proof-of-concept and initial field ground biomass assessments. work using LiDAR to gather on the ground above- ground biomass data in Gabon and Peru. In early This data is used to improve the accuracy of Sylvera’s 2022, Sylvera raised an additional $32.6 million in machine learning models. Sylvera combines quanti- venture capital and private equity funding. tative and qualitative data from carbon projects to rate the performance of carbon credits on a scale of Sylvera receives technical support from several AAA to D, similar to how financial bonds are rated. areas. The company partnered with Dr António This is calculated based on the project’s rating in Ferraz (the University of California in Los Angeles and three areas: the NASA Jet Propulsion Laboratory) and Professor Mathias Disney (University College London) on Carbon score (emissions reduction achieved the use of LiDAR. Through the SPace Research compared to the credits issued) and Innovation Network for Technology business support program, Sylvera also collaborated with Additionality (the extent that results would not the University of Leicester in the United Kingdom to have occurred without the project) verify its machine learning methodology. Permanence (how long the benefits will last). The project is also rated for its community and biodiversity co-benefits. The project’s deliberate contributions to the local community are assessed in terms of their alignment with the United Nations Sustainable Development Goals. Data from project developers and geographic information system (GIS) mapping, along with local research, is used to assess the project’s contributions to species richness and diversity and biodiversity protection in the context of local, regional, and national threats. 26 Picard, N., et. al. 2015. Reducing the Error in Biomass Estimates Strongly Depends on Model Selection. Annals of Forest Science. doi: 10.1007/s13595-014-0434-9 44 Technical details Quality assurance Data sources and pre-processing Quality assurance is performed at several stages. Internal machine learning QA validates predictions by Sylvera accesses open-source, multispectral satellite comparing datasets to global and local data on forest data from Google Earth Engine, including optical data cover. The machine learning outputs are assessed from satellites and topographical ranging data from by the GIS team before being fed into the ratings the Global Ecosystem Dynamics Investigation LiDAR framework. Visual QA is performed by comparing instrument on board the International Space Station. machine learning predictions with satellite images and time series plots. This data has global coverage, with spatial resolutions that vary from 10 meters to 30 meters. Improved biomass data to train Data is calibrated and pre-processed to reduce machine learning models artefacts and achieve a clear image for each region. Inferring reliable pixel-wise above-ground biomass Use of data values from satellite data requires accurate on the ground data to calibrate the machine learning Land class is labeled using the 2019 Global Forest models. Legacy methodologies to collect this data Canopy Height dataset, with additional segmentation use allometric models based on manual sampling labels for training and validation provided by of tree height and diameter at breast height to CloudFactory, which specializes in annotations for estimate forest above-ground biomass. Validation machine learning. This on the ground data is used to studies using destructive methods, where model train and validate deep-learning models to interpret predictions are compared to direct measurements optical satellite images and apply a binary “forest/ of the biomass of felled trees, have shown these not forest” mask, considering the local definition models to underestimate above-ground biomass of forest. These models can then be used to track (except for when measuring coniferous forests). forest cover in project areas, as well as reference When scaled across a one-hectare (ha) stand, these and leakage areas, over time. Data can be updated measurements also result in an uncertainty rating of as often as satellites pass over project sites. Forest/ up to 40 percent.27 not forest predictions are used to calculate both the carbon score and additionality ratings. FIGURE 13: OVERVIEW OF HOW SYLVERA’S MACHINE LEARNING OUTPUTS FEED INTO THE RATINGS OF CARBON SCORE AND ADDITIONALITY Quantify and verify deforestation emissions claims Rating methodology extract Carbon score Quantify and verify emissions reduction claims Satellite Prediction Machine of forest and imagery per learning non-forest year from model per year from project start project start Validate baseline claims about the reference area and leakage area Additionality score Validate deforestation drivers claims 27 Demol, M., et al. 2022. Estimating Forest Aboveground Biomass with Terrestrial Laser Scanning: Current Status and Future Directions. Methods in Ecology and Evolution. https://doi.org/10.1111/2041-210X.13906 45 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Sylvera’s approach to this problem is to gather FIGURE 14: SYLVERA’S TERRESTRIAL biomass data using multi-level LiDAR scanning and LASER SCANNER MAPPING VEGETATION volumetric tree modeling. Terrestrial LiDAR scanning IN THE PERUVIAN RAINFOREST maps vegetation across three single hectare sample plots to millimeter detail, collecting about 400,000 data points per square meter. This is scaled up using drone-mounted, unoccupied aerial vehicle-based LiDAR scanning 1,000 ha and upwards of 50,000 ha in slow- and fast-flying configurations. The 3D models of tree and forest structure derived from the data enable estimates of above-ground biomass with significantly improved accuracy. FIGURE 15: POINT CLOUDS OF INDIVIDUAL TREES INSIDE AN AFRICAN TROPICAL FOREST STAND, CAPTURED USING TERRESTRIAL LiDAR SCANNING D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 46 Sylvera’s research team, in partnership with the World When tracking large-scale activities in forests such Bank, University College London, the University of as clear cuts or forest fires, public data is suitable Edinburgh, and NASA’s Jet Propulsion Laboratory, as it has sufficient resolution, is extremely well have so far conducted research in the African calibrated, and is available with global coverage. tropics, the South American tropics, and temperate To track small-scale activities such as selective forest, funded by the Small Business Research and logging and degradation, high-resolution imagery is Innovation grant. Future research aims to cover all required, which could be obtained from commercial biomes relevant to their rated projects, with plans companies. Access to this higher-resolution data to collect measurements from 20 to 24 more sites could also help overcome some challenges faced by across 10 countries in the coming years. the machine learning models in expanding project cover, including seasonality, mountains, cloud cover, Lessons learned and continued and variability of vegetation. use of digital technologies New models are being developed to cover other D-MRV verification types of nature-based crediting projects that have The SPace Research and Innovation Network previously lacked accurate MRV. These include for Technology supported Sylvera in developing models tracking soil carbon and blue carbon and verifying its methodology and quantifying crediting projects. The robust ratings framework can uncertainty. Together with a team from the University be used to assess the additionality, permanence, and of Leicester, a transparent method of validating co-benefits of wider crediting projects included in the machine learning outputs and subsequent voluntary carbon markets, such as renewable energy interpretations has been identified in order to and cookstoves. Through partnerships with the World increase credibility and confidence in the credits Bank and national governments, Sylvera’s D-MRV rating framework. This work is being continued to systems will also be used as part of jurisdictional consider possible changes in data sources and how REDD+ programs and wider carbon accounting. they might impact estimates and uncertainty. Policy considerations Future developments Sylvera’s stated mission is to “deploy state-of- Initial ratings focused on avoided tropical the-art D-MRV systems to restore confidence in deforestation projects. To expand the range of nature-based solutions and help quickly scale VCMs biomes and crediting projects covered, the machine [voluntary carbon markets]”. Its rating framework learning models are being updated to go beyond aims to address a key criticism of avoided emissions binary classifications to include sparse forest, credits from nature-based projects, which is that the mangroves, plantations, and canopy cover and carbon benefits are poorly quantified and unreliably height. The next iteration of models will output reported. More accurate assessments of the carbon continuous definitions such as the percentage of tree stored in the world’s standing forests allows them to cover, allowing activities such as forest degradation be properly valued, and accurate, real-time monitoring and vegetation growth to be monitored, and for of their condition allows fair results-based payments afforestation, reforestation, and revegetation and to be made, incentivizing national governments and improved forest management projects to be rated. local communities to conserve forests. 47 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 6 European Bank for Reconstruction and Development D-MRV system for renewable energy Type of system: Mitigation action T he EBRD D-MRV system is being piloted in from the Spanish Climate Change Office, in effect support of two broad use cases: results-based monetizing its mitigation outcomes (up to an agreed climate finance provided by the Spanish Climate limit). The resulting GHG emission reductions paid Change Office for a grouped solar project in Jordan, for under this arrangement have to be monitored, and climate finance impact reporting to the Clean reported, and verified by the EBRD’s D-MRV system. Technology Fund (CTF) under the EBRD program The objective of the D-MRV pilot is to showcase how on Accelerating Innovations in Renewable Energy, digitalization and automation of MRV processes— co-financed by CTF. An initial set of solar plants in with associated benefits in terms of reduced costs Jordan has been integrated into the D-MRV as of the and time, and increased accuracy and transparency end of 2021. This case study is thus based on initial of data—could facilitate scaling up of mitigation experiences gained with these plants only. actions under the existing and emerging carbon market mechanisms and results-based climate The fully functional D-MRV system, which was finance instruments. developed in 2021, applies the key principles and requirements laid out in the D-MRV Protocol, 28 which The D-MRV system was developed by BowTie was released by EBRD at the end of 2020 and builds Technology BV, under contract to EBRD. It is a on the inputs from members of the Joint Multilateral cloud-based software solution that enables on-site, Development Banks Working Group on Article 6. measured data to be directly acquired from renewable energy projects and automatically sent for validation The D-MRV system serves a portfolio of eight and cross-checking. GHG emission reductions are solar photovoltaic plants with a total capacity of then automatically calculated according to the 48.3 MWp that is run by Yellow Door Energy Jordan applicable methodology, and monitoring reports (YDE).29 This is Jordan’s largest portfolio of private- on system-verified GHG emission reductions are to-private renewable energy plants supplying generated. The main parameter monitored is the net directly to the private sector. The YDE project has electricity generated and supplied to the grid, with been registered with the VCS30 as a grouped project solar irradiance data also tracked for the purpose of and, as such, benefits from results-based payment cross-checking the power output. 28 European Bank for Reconstruction and Development. Protocol for Digitalised Monitoring, Reporting and Verification (D-MRV Protocol). December 2020. [Online]. Available: https://www.ebrd.com/documents/climate-finance/digitised-mrv-protocol.pdf?blobnocache=true [Accessed April 28, 2022]. 29 Yellow Door Energy. No date. Projects. [Online]. Available: https://www.yellowdoorenergy.com/projects/ [Accessed April 28, 2022]. 30 Verra. No date. Grouped Solar Projects in Jordan. [Online]. Available: https://registry.verra.org/app/projectDetail/VCS/2016 [Accessed April 28, 2022]. 48 Implementation of the YDE solar photovoltaic The D-MRV software and data are hosted in the project in Jordan was carried out over several years, industry-class cloud, which ensures scalability, with the first plants completed in 2019 and the full flexibility, and high system availability. Information portfolio expected to be finalized in late 2022. The security is assured through data encryption, both D-MRV system launched in September 2021 was in transit and at rest. High-frequency data retrieval, connected to the plants that were operational at which sees measurements taken at five-minute the time. The D-MRV piloting phase is expected to intervals and data pulled into the D-MRV on an hourly continue through to 2023 to allow for the remaining basis, helps to prevent potential data tampering as plants to be commissioned and to verify delivery it limits the “non-supervised” time the data spends of the targeted mitigation outcomes under the at the project site before being transferred to the results-based arrangement. This phased rollout secure D-MRV environment. The system’s credibility will also allow for insights into real-time monitoring and transparency are further assured by the use of and connectivity issues to be identified and used to cloud-native, immutable storage of the raw data, enhance efficiency and robustness of data validation allowing full traceability of raw data for the purpose and cross-checks (including, potentially, exploring of verification. Access to the D-MRV system is options for applying machine learning). restricted to preauthorized users, with their rights to access and/or handle data and perform certain The initial results of the pilot, which are broadly functions defined broadly in line with the relevant summarized in this case study, indicate strong provisions of the D-MRV Protocol. For example, potential for D-MRV to enhance the efficiency, YDE, as the project owner, mostly has read-only timeliness, accuracy, and transparency of MRV access, particularly to critical elements that may processes. compromise the integrity of the results claimed. Technical details D-MRV system data verification In addition to automatically acquiring data in near real The use of digital technologies for time, the D-MRV system performs quality assurance data collection and processing and verification of the data received in several ways. The D-MRV system relies on a set of digital First, data provenance is assured by the use of technologies to monitor and collect the data on authenticated data sources, which are configured the relevant parameters. Electricity supplied to the in the D-MRV system from the outset, so preventing grid, and consumed from the grid to power internal potential tampering at the point of data import. loads during non-operating hours, is continuously Second, the data undergoes several numeric checks measured by revenue-grade, on-site, bi-directional for completeness and consistency on arrival in the digital electricity meters at five-minute intervals, with D-MRV to detect and fill data gaps or irregularities. time-stamped readings delivered via a server to the Finally, the D-MRV conducts a plausibility cross- D-MRV cloud environment. This enables the system check to ensure that the monitored energy produced to calculate the project’s net energy generation, does not exceed the theoretical maximum output which is the key input into the calculation of the based on the plant’s configuration and the solar mitigation results achieved. The other key input irradiance available over the period covered. into the GHG mitigation—grid emission factor—is preset ex-ante defined in the registered VCS project Depending on the outcome of the above checks, the documentation and is not monitored. In parallel, solar D-MRV system automatically labels each data point irradiance is measured by on-site pyranometers and as “auto-approved”, “pending”, or “rejected”. This acquired via the same process to enable cross- significantly streamlines periodic verification by an checks of the main output parameter. auditor, as only the “pending” data points will need to be manually approved or rejected (although all data points are available to the auditor for review, if needed). 49 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS In order to facilitate remote verification of the The GHG monitoring and verification report is project’s claimed mitigation results, the measured intended to enable the monetization of mitigation data directly acquired by the D-MRV system is outcomes claimed. The report automatically supported with additional digital records that draws information—including energy generation, are manually uploaded to the system, including calculated GHG emission reductions for the selected geotagged site photos, metering diagrams, meter monitoring period, and manually entered narrative calibration, and power plant commission certificates. data—from the D-MRV database and combines it in a familiar VCS-type report without requiring additional Data presentation and reporting manual intervention. Combined with system- The D-MRV system presents data in two ways: generated data quality and approval labels, as well via a dashboard and via a reporting interface. as verification feedback of an auditor (provided via The reporting interface supports data reports dedicated auditor access to the system), the report (downloadable worksheets for any parameter could be finalized in the system and immediately handled by the system) and GHG emission shared (for example, via an API) with the respective monitoring and verification reports. While the environmental attribute standard for an issuance dashboard provides a quick overview of the key decision. project performance indicators in numeric and graphic form, deeper insights can be gleaned through the data reports. FIGURE 16: EBRD D-MRV STRUCTURE Note: Some data has been hidden for confidentiality reasons. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 50 Methodological modifications Lessons learned and continued to D-MRV approach use of digital technologies The YDE project registered with the VCS relies The current D-MRV pilot started in September 2021. on the “standard” CDM methodology for grid- Some early insights gained from monitoring the connected electricity generation from renewable operational portfolio are summarized below. sources, ACM002, for the monitoring and accounting requirements to calculate its GHG mitigation Onboarding new projects results. While the methodology and the associated tools mandate that the key parameter (quantity of In order to fully realize the cost- and time-savings electricity supplied to or consumed from the grid) benefits that the D-MRV offers, it is essential that the should be measured by electricity meters installed process of connecting new projects to the D-MRV at the grid interface, these meters are inaccessible system is streamlined and automated to the extent to the project company—and hence, to the D-MRV feasible. The set-up of the pilot YDE project was system—for remote communication, as is the case handled manually by the developer, in coordination with most commercial utility-scale renewable energy with the project owner, which took a fair amount of projects. The D-MRV system therefore relies on time. Going forward, a (semi) automated on-boarding plant-site check meters that are under control and process will be trialed for new projects, whereby some used by the project company for plant-monitoring or most of the project’s elements will be configured in purposes. The check meters installed are of revenue the D-MRV based on inputs provided by the project grade and accuracy class, which makes them suitable owner. This is expected to significantly reduce the for the purpose of D-MRV system monitoring. duration and cost of the project set-up stage. In order to accurately account for the actual Advanced verification approaches amount of net electricity generation fed into the The pilot D-MRV system implemented some basic grid, as would be measured at the grid interface, verification checks to ensure that the energy the unavoidable cabling and transformer losses production numbers fall within the expected between two sets of meters—grid-side utility and performance range, based on a plant’s respective plant-side check meters—had to be included in the configuration, capacity, and efficiency, combined methodology-prescribed calculation approach to with solar irradiance, among other metrics. As avoid overstating the results. The corresponding currently designed, these checks have limitations. losses have been conservatively estimated, based For instance, they are “hard-coded” for each plant on the pre-D-MRV period electricity generation data and have, at times, triggered “false positive” flags (from both sets of meters), and embedded into the when plants were suspected of producing output D-MRV calculation algorithm. that exceeded their rated capacity (which could happen due to a combination of factors such as high irradiance, the module temperature, and the angle of the sun). The ongoing piloting phase is refining these cross-correlation checks to arrive at a more robust, flexible, and potentially more scalable approach. One alternative could be a machine learning model that can be “trained” on a set of trusted data in order to then provide continuous plausibility assessment for energy production and help better detect potential data anomalies. 51 D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS Interoperability with other MRV systems Alternative measurement arrangements The EBRD D-MRV system operates with near real- As evidenced by the ongoing D-MRV pilot, and as time, high-resolution raw data and should be able is likely the case for other commercial renewable to exchange this data with other stakeholders. energy projects, remote access to billing meters However, such high-resolution data (for example, at interconnection points may not be available to GHG emission reductions at five-minute intervals) project owners and/or D-MRV system operators. could prove excessive and are probably not As a result, alternative measurement hardware with necessary for other MRV systems, such as a appropriate metering and communication capabilities national-level GHG reporting and inventory system. may need to be put in place. Energy losses between Nevertheless, the two types of MRV systems can two measurement points need to be reflected when be integrated, for example, by having the project- calculating the net energy exported and GHG impacts level D-MRV system pass on aggregated data at claimed by the project. This will limit the ability for an agreed resolution (for example, monthly energy direct cross-checking against digital billing system generated and mitigation results achieved) to the records due to obvious discrepancies in readings national-level MRV system through an API. The between the two metering points. compatibility of the systems will need to be ensured by using the same emission factors, as an example. The ongoing pilot will provide further insights into The ongoing D-MRV pilot will explore these issues this issue as new projects are connected, different on the basis of the YDE project, as well as potential interconnection configurations are analyzed, and connectivity options with the national MRV system clarity is achieved on the requirements for the in Jordan. measurement and communication hardware. These learnings are expected to feed into updates of the respective provisions of the D-MRV Protocol. D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 52 Appendix TABLE 9: COMMONLY MONITORED PARAMETERS AND CORRESPONDING DIGITAL TECHNOLOGIES PARAMETER DESCRIPTION RELEVANT DATA UNIT MONITORING MONITORING DIGITAL QUALITY MITIGATION FREQUENCY METHOD/S TECHNOLOGY CONTROL ACTIONS EGgrid Electricity Grid-connected kWh, MWh Continuous Electrical meters Smart meters to Power generation data generated during renewable at interconnection continuously record recorded by smart a period of time energy points can measure and report power meters can be uploaded by a grid or mini- generation, fuel power generation generation to D-MRV to D-MRV linked to grid connected switching, or system digital billing system plant energy efficiency so generation data is measures in automatically checked power plants against invoices EGsinglesys Electricity Decentralized/ kWh, MWh Monthly Electrical Smart meters with SIM Sampling of technology generated by household (billing cycle) meters, technical card, AI, and weather functionality subsystem renewable specifications, and data energy weather data generation Emission CO2 content Fuel switching CO2 per – – – factor (fuel) of fossil fuel and energy energy or efficiency mass/ measures volume Emission CO2 content Fuel switching CO2/MWh – – – factor (plant) per MWh and energy efficiency measures Emission CO2 content Off-grid or CO2/MWh Yes Electrical meters Smart meters to Data can be cross- factor (grid) per MWh grid-connected to report power automatically report checked with meters at energy generation power generation interconnection points generation combined with from each plant with in the electrical grid and default efficiency linked CO2 factor at the plant level against or fuel consumption to automatically billing information values determine grid emission factor Net calorific Energy per mass Energy Megajoules No (default or – – – values or volume generation, fuel (MJ)/m3 or ton specific value) switching, energy savings D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 54 TABLE 9: COMMONLY MONITORED PARAMETERS AND CORRESPONDING DIGITAL TECHNOLOGIES (cont...) PARAMETER DESCRIPTION RELEVANT DATA UNIT MONITORING MONITORING DIGITAL QUALITY MITIGATION FREQUENCY METHOD/S TECHNOLOGY CONTROL ACTIONS Energy saved Thermal energy Energy savings, MJ, British Yes Purchases receipts Smart meters and Consumption data from (thermal) saved by activity fuel switching thermal units or flowrate digital purchasing/ sensors at pipes or counters on pipes billing systems similar can be cross- checked with purchases of fuel Energy saved Electrical energy Energy savings, kWh, MWh Yes Meters or invoices Smart meters and Usage data at meters (electrical) saved by the fuel switching digital purchasing/ can be cross-checked activity billing systems with electricity generation or purchases Number of Number of Household or – Yes Sales records Sales or installation Cross-checked with units project systems decentralized or installation data using wholesale purchase or or users under systems or certificates smartphones or importation data activity products purchasing software Operationality Fraction of Household or Percent Yes Telephonic or in- If performance is Can be cross- systems or users decentralized person surveys reported digitally, checked with in- in operation systems or then operationality person inspection or products is automatically telephonic interview/ confirmed mobile phone-based interaction with user Fuel use Fuel switch or Liters or tons Yes Purchases receipts Smart meters and Consumption data from project emissions or flowrate digital purchasing/ sensors at pipes or counters on pipes billing systems similar can be cross- checked with purchases of fuel Carbon stock Tons of carbon Forestry and tCO2/m3 or ton Yes Satellite data or in- Analyze satellite or Periodic in-person dioxide per agriculture person inspection drone data with AI to inspection to verify volume or mass projects of biomass learn how to estimate reports from satellite of biomass biomass and AI software D-MRV SYSTEMS AND THEIR APPLICATION IN FUTURE CARBON MARKETS 55 Administered by THE NETWORKED CARBON MARKETS Climate INITIATIVE Warehouse