A Cost-Effective Open-Source Synchrophasor for Power Grid Monitoring: A Case Study for Rwanda

A Cost-Effective Open-Source Synchrophasor for Power Grid Monitoring: A Case Study for Rwanda Opeyemi Ajayi, Student Member, IEEE,Barry Rawn, Senior Member, IEEE and Nathalie Uwamahoro, Student Member, IEEE Abstract—This paper aims to describe multiple analysis use cases of a future open-source grid situational awareness operational system for resilience and flexibility of the power grid This study presents a ReactJS, Django, and time series database, InfluxDBPostgreSQL Grafana used as open-source components that function as a dashboard for monitoring and storing realtime data from the OpenPMU. The platform is cloud-based and runs on Amazon Web Services(AWS). The platform aims to enable academics and small-scale energy utilities interested in cost savings to analyze events and track the quality and reliability of the electricity supply. Index Terms—Distributed energy resources (DER), Grid visualization platform, Intelligent energy networks, Modular opensource software, Situational awareness, Synchrophasor technology, Time-series databases I. INTRODUCTION Modern power grids increasingly depend on real-time monitoring and control mechanisms to maintain system reliability, enhance operational stability, and enable the efficient integration of renewable energy sources [1], [2]. Wide-area situational awareness is central to these objectives, allowing operators to detect, localize, and respond to disturbances promptly. However, achieving such awareness remains challenging in lowand middle-income countries (LMICs) due to persistent financial, infrastructural, and technological constraints. Phasor Measurement Units (PMUs) have emerged as a cornerstone technology for next-generation grid monitoring. These devices provide time-synchronized, high-resolution voltage and current phasor measurements that significantly enhance grid observability and situational awareness compared to conventional Supervisory Control and Data Acquisition (SCADA) systems, hence, essential for dynamic state estimation, and transient stability assessment [3]–[6]. Recent advancements have also introduced other variants such as optimal PMU placement help improve measurement redundancy, and resilience to component failures [7]–[9]. microPMUs (µPMUs), capable of sampling rates up to 120 Hz, facilitating detailed monitoring of fast-changing dynamics within distribution networks [3]. Field implementations by the Electric Reliability Council of Texas (ERCOT) and other utilities have also demonstrated the value of synchrophasor data in oscillation detection, topology verification, and dynamic model validation [10], [11]. Despite these achievements, academic bodies and small-scale utilities face limited access challenges to its high data volumes management, and effective visualization of its real-time data streams. To mitigate these challenges, open-source platforms like the Open Phasor Data Concentrator (OpenPDC) [12] remain one of the most mature tools for PMU data aggregation and management, supporting interoperability across multiple communication protocols. However, OpenPDC lacks an inherent cloud-native architecture and requires extensive expertise for configuration and testing for deployments. One other opensource alternative is the modular OpenPMU framework [13]– [15], which provides a foundation for synchrophasor data streaming and real-time communication but does not integrate essential database storage, analytics, or visualization components required for comprehensive situational analysis. As emphasized in [4], the absence of integrated data management and lightweight visualization capabilities hinders the full exploitation of PMU data, particularly in LMIC contexts where cost-effective and computationally efficient solutions are essential. To address these limitations, this study presents the development of a cost-effective, cloud-native, open-source, cloudfriendly synchrophasor monitoring system bringing together OpenPMU data ingestion via UDP, InfluxDB time-series database storage, and visualization within Grafana, so that utilities and research institutions can view, save, and examine PMU information in real time. The proposed system integrates data acquisition, processing, storage, and visualization within a unified architecture deployed on Amazon Web Services (AWS). Built upon the OpenPMU framework [16], the platform first incorporates database connectivity, lightweight analytics, and interactive dashboards that enable real-time visualization and performance tracking of synchrophasor data streams. This approach enhances scalability, reduces computational overhead, and provides an accessible pathway for LMIC utilities to leverage advanced monitoring tools, thereby promoting more resilient, flexible, and intelligent grid operations. The remainder of this paper is organized as follows. Section II describes the proposed framework, detailing the data acquisition process, middleware components, and cloud deployment strategy. Section III presents experimental validation, performance evaluation in terms of grid resilience and flexibility, and comparative analysis with existing solutions. Finally, Section IV concludes the paper and outlines directions for future research and platform enhancement. II. DEVELOPMENT OF A COST-EFFECTIVE CLOUD-NATIVE OPEN-SOURCE SYNCHROPHASOR PLATFORM The proposed platform provides a cloud-native framework for synchrophasor data acquisition, processing, and visualization to enhance real-time monitoring and decision support in smart grids. The architecture is composed of three major components: (1) a Data Acquisition Layer that communicates with Phasor Data Concentrators (PDCs) to collect timesynchronized phasor measurements; (2) a Data Processing Middleware responsible for stream management, preprocessing, and data aggregation; and (3) a Web-Based Visualization Layer for interactive analytics and real-time monitoring dashboards. The overall workflow of the system is illustrated in Figure 1, and data acquisition and processing are outlined in Algorithm 1, representing the end-to-end open-source technologies data pipeline from ingestion to storage, analysis, and live dashboard deployment. Fig. 1. OpenPMU Data Concentration and Management Architecture. The proposed platform integrates backend server that supports defined data exchange structures for synchrophasor measurements and a frontend module that visualizes these measurements through an exposed Application Programming Interface (API), enabling seamless interfacing between acquisition devices, middleware, and web dashboards. First, we illustrate data acquisition procedures. A. Data Acquisition and Measurement Rwanda’s national grid has a nominal system frequency of 50 Hz and operates with a transmission voltage of 110 kV, mainly controlled by the Rwanda Energy Group (REG). At the end of 2024, the national installed capacity is around 332 MW and is increasingly integrating distributed solar and micro-hydro generation. This diversification brings operational variations that spur the constant monitoring with synchrophasors to ensure stability monitoring and anomaly identification. Our OpenPMU simulation is an abstraction of an individual PMU node functioning under this kind of grid operations. The data acquisition process begins with the BeagleBone Black device, which serves as the local data acquisition module that produces voltage and current phasors from analogto-digital (ADC) conversions. In this study, a Python-based OpenPMU platform simulation tool with a graphical user interface (GUI) was used to emulate a single node PMU measurement data streams generated in real-time over a UDP communication protocol following the Extensible Markup Language (XML) schema of sampled values defined in [14]. The phasor estimator, following established algorithms in [17], [18], computes positive-sequence voltages and currents using the linear phase estimation method implemented in Algorithm 1. The general PMU phasor representation is given by: y(t) = Ymcos(ωt +θ)(1) where Ymis the amplitude, ωthe angular frequency, and θ the phase angle. The corresponding complex phasor form is expressed as: Y=Vs √2ejωt =Vs √2(cos(ωt)+jsin(ωt)) (2) where the real and imaginary components represent the inphase and quadrature signals, respectively. Figure 2 illustrates the functional settings of the PMU. Fig. 2. Functional settings and operational structure of the PMU. The statistical distribution and probability density function of the sample phasor data collected for a 10-minute duration are also displayed in Table ?? 3 4 as shown. TABLE I STATISTICAL DISTRIBUTION OF PMU MEASUREMENTS FOR CHANNEL 0 Metric Count Mean Std Min 5% Median 95% Max Angle 30000 0.00 0.05 -0.20 -0.08 0.00 0.08 0.21 Freq 30000 49.99 0.02 49.66 49.96 49.99 50.03 50.07 Mag 30000 5.00 0.03 4.92 4.95 5.00 5.04 5.07 ROCOF 30000 -0.00 0.73 -8.67 -1.20 -0.00 1.20 2.85 Before Potential Transformers and Current Transformers(PT/CT) scaling, the OpenPMU outputs are displayed as raw normalized Root Mean Square (RMS) values in secondary-level phasor magnitude (Mag0–Mag5). B. Phasor Data Concentration and Processing Multiple PMUs deployed across substations continuously stream synchronized voltage and current measurements to the Phasor Data Concentrator (PDC). The PDC aggregates, time-aligns, and pre-processes incoming data to minimize latency and ensure measurementintegrity [19], [20]. Within the proposed system, the OpenPMU data concentrator module Fig. 3. Timeseries of OpenPMU voltage and current data Fig. 4. Probability density distribution of simulated OpenPMU simulation data showing stable operation centered at 50 Hz ± 0.03 Hz. performs essential preprocessing operations, including data validation, timestamp correction, and synchronization verification. The OpenPMU modules (1) and (2) in 1 streams continuous data to the backend Django web framework, provide a RESTful API for data ingestion, processing for bad data detection, and retrieval of event logs stored in a PostgreSQL database for quality reports.Frontend visuals is implemented using D3.js and React, providing interactive real-time plots of voltage and current phasors, quality indicators of device status. Additional functionalities include frequency and rateof-change analysis,and real-time computation of key stability indicators, such as voltage stability indices (VSI).To ensure security and integrity, synchrophasor data transmitted over wide-area networks (via TCP/IP or UDP/IP) is encrypted to mitigate potential cyber threats, including man-in-the-middle and false data injection attacks. C. Data Archival and Decision Support The system incorporates long-term data archival and historian functionalities for trend analysis and performance evaluation. The processed and aggregated PMU data were then feed directly into the decision-support dashboard that helps operators visualize key performance indicators (KPIs) such as bus voltage magnitudes in per unit, frequency deviations, voltage stability indices through color-coded geospatial maps and real-time alarms when operational limits are violated e.g., VSI <0.2 or reactive reserve <15%. D. Data Fetching and Hosting Workflow The automated data fetching process is executed using Django management commands that record synchronized measurements into the PmuData model in real time. This ensures temporal consistency and supports continuous data streaming for visualization and analysis. The corresponding algorithm is presented in Algorithm 1. Algorithm 1 PMU Data Reception and Processing Algorithm 1: function XMLTYPECONVERT(tag) 2: if tag ∈ {“Frame”, “Channels”}then return integer(tag) 3: else if tag ∈ {“Freq”, “Angle”, “ROCOF”, “Mag”} then return float(tag) 4: elsereturn string(tag) 5: end if 6: end function 7: Initialize: localIP = “127.0.0.1”, localPort = 48011, bufferSize = 10240 8: Create UDP socket and bind to (localIP, localPort) 9: while true do 10: Receive XML message from PMU and decode using UTF-8 11: Parse XML fields: Date, Time, Frame, Channels 12: for each channel in Channels do 13: Extract Mag, Angle, Freq, ROCOF 14: Merge with general data and store record in database 15: end for 16: end while To ensure high data integrity and low latency, data validation and filtering operations are executed at each transmission stage [21]. The system supports data storage in open-source time-series databases such as InfluxDB and PostgreSQL, while visualizations are rendered dynamically in Grafana and React/D3.js. The Django backend follows the Model-TemplateView (MTV) design pattern, ensuring modularity, fault tolerance, and efficient synchronization with the cloud database. Cloud Deployment and Scalability: The complete application is deployed on Amazon Web Services (AWS) Elastic Beanstalk, leveraging containerized deployment for scalability and high availability. Secure Shell (SSH) connections enable remote access, system updates, and performance monitoring. The web-based dashboards are fully interactive, allowing zooming, filtering, and customization of data views. Phase voltage was selected as the default metric to demonstrate voltage drop behavior, as it provides immediate insight into the grid. TABLE II COST COMPARISON BETWEEN OPENPMU AND COMMERCIAL PMU SYSTEMS Parameter OpenPMU Commercial Cost (This Study) PMU Reduction Hardware Cost ˜$200 $15k–$40k >98% (RPi + ADC + GPS) Software License Open-source Proprietary 100% (Python/UDP) Sampling Rate 50 fps 60 fps Comparable Accuracy ±0.02 Hz ±0.01 Hz Acceptable (frequency) Data Interface UDP, JSON IEEE C37.118 Compatible Power Supply 12 V DC 120/240 V AC Flexible III. CASE STUDIES A. Result The study showcases several built UI components and screenshots of the voltage in phase 1, and its corresponding frequency. Dash graph component tools served as the prototype [22] for the initial version of the pages, as shown in figure 5; however, its open-source version lacks flexible data management and error checks where later versions are designed using ReactJS and D3.js thrive. The D3.js figure 6 binds the data from the data API to the chart display elements, such that the solution is completely responsive and compatible with all screen widths. The final version visualizes all phasor quantities using Grafana/InfluxDB version in figure 7. Fig. 5. First iteration of Phasor Visualization Dashboard. The performance analysis of this solution formulation would focus on its cost-effectiveness and its usefulness due to its speed and accuracy in calculating stability indices for enhancing grid resilience and flexibility. Table II shows the cost effectiveness of our system B. Performance Analysis Real-time computation of stability indicators like the Voltage Stability Index (VSI) was performed with comparable Fig. 6. Final Phasor Visualization Dashboard. Fig. 7. The monitoring dashboard. speed and accurate computations. The PMU system was timely and of comparable speed as data were measured and provided at a high rate 50 samples/sec. The system’s operational usability and visualization were another metric used. The visualization was made clear and quick for an operator under pressure to understand phasors’ complex information; hence, results on the dashboard are color-coded maps and time-series plots so operators can identify a stressed zone on the map, understand the severity from the VSI value. C. Relevance for future Power Grid Resilience and Flexibility By offering an open, low-latency of <50ms, in contrast to 100ms using alternatives in [6], and an affordable platform for PMU data streaming and analytics, the framework offers a probable observability solution for small utilities and research organizations for resilience studies, load forecasting, and distributed generation control.For flexibility, the system can allow operators to run the grid closer to its operational limits safely due to a high-fidelity view of the stability margin, e.g., VSI, to accommodate the variable and less predictable nature of renewable energy sources. IV. CONCLUSION This study demonstrates multiple iterations of the proposed solution and its potential to exemplify ideal use cases for both academia and utilities seeking an affordable venture visualization solution from the OpenPMU. The prototypes aimed to cover a breadth of features to showcase the potential for the integration of a database feature for the OpenPMU rather than delving deeply into the typical feature’s robustness and optimization of a synchrophasor analytics visualization dashboard. The dashboard is scalable in that it accommodates highrate PMU data as shown here[https://phasormeasurementdatacmuenergylab.link/]. To enhance the system resilience and flexibility, one could improve the dashboard observability by integrating early warning alarms, displaying damping ratios for oscillatory modes, implement conversion of raw measurements into derived quantities such as real and reactive powerdV/dQ sensitivities and interactive drill-downs for localized analysis of stressed zones PV/QV curve snapshot for dynamic stability monitoring, and inertia monitoring as in ERCOT [6], [10], [11], [23], and cross-validating VSI with the offline tools (Modal Analysis), showing how it correlates with actual grid behavior during both normal operation and stressed conditions. Thus, allow for experimentation to learn from past disturbances, perform decision-making post-event analysis, resilience assessment, and model validation, investigate and prevent cyberattacks in grid architecture for blackout prevention in mixed conventional plus renewable grids, load shedding, reactive power compensation, and generator voltage regulation remedial action schemes. REFERENCES [1] N. Zhou, Z. Huang, J. Nieplocha, and T. B. Nguyen, “Wide-area situational awareness of power grids with limited phasor measurements,” Pacific Northwest National Lab.(PNNL), Richland, WA (United States), Tech. Rep., 2006. [2] H. M. Khalid, F. Flitti, M. S. Mahmoud, M. M. Hamdan, S. Muyeen, and Z. Y. Dong, “Wide area monitoring system operations in modern power grids: A median regression function-based state estimation approach towards cyber attacks,” Sustainable Energy, Grids and Networks, vol. 34, p. 101009, 2023. [3] E. Dusabimana and S.-G. Yoon, “A survey on the micro-phasor measurement unit in distribution networks,” Electronics, vol. 9, no. 2, p. 305, 2020. [4] T. Menezes et al., “A survey on the application of phasor measurement units to the protection of transmission and smart distribution systems,” Journal Name, 2023. [5] S. S. Yu et al., “Comprehensive review of pmu applications in smart grid: Enhancing grid reliability and efficiency,” Journal Name, 2023. [6] K. Chatterjee et al., “Online monitoring applications enabled by phasor measurement units,” Pacific Northwest National Laboratory, Tech. Rep., 2023. [7] R. Babu, S. Raj, and S. Mahapatra, “Allocation of synchronized phasor measurement units for power grid observability using advanced binary accelerated particle swarm optimization approach,” Journal of Electrical Systems and Information Technology, vol. 10, no. 1, p. 42, 2023. [8] A. K. Singh et al., “A critical review on phasor measurement units installation planning and application in smart grid environment,” Renewable and Sustainable Energy Reviews, vol. 191, p. 114024, 2024. [9] J. Liu, F. Tang, F. Ponci, and A. Monti, “Optimal phasor measurement unit placement for power system observability using teaching–learning based optimization,” Electric Power Systems Research, vol. 201, p. 107960, 2021. [10] J. Chen et al., “Use of synchronized phasor measurements for dynamic stability monitoring and model validation in ercot,” in IEEE PES General Meeting, 2012, pp. 1–7. [11] K. M. Koellner et al., “Synchrophasors across texas: The deployment of phasor measurement technology in the ercot region,” IEEE Power and Energy Magazine, vol. 13, no. 5, pp. 36–40, 2015. [12] G. P. A. OpenPDC, “Openpdc: The open source phasor data concentrator,” 2024. [Online]. Available: https://github.com/GridProtectionAlliance/openPDC [13] D. Laverty, J. Hastings, D. Morrow, R. Khan, K. McLaughlin, and S. Sezer, “A modular phasor measurement unit design featuring open data exchange methods,” in Proceedings of the Power and Energy Society General Meeting (PESGM), 2017”,. Institute of Electrical and Electronics Engineers Inc., Feb 2018. [Online]. Available: http://pes-gm.org/2017/, [14] Laverty, “Open-source phasor measurement unit,,” 2024. [Online]. Available: https://sites.google.com/site/openpmu/ [15] D. M. Laverty, L. Vanfretti, R. J. Best, D. J. Morrow, L. Nordstrom, and M. Chenine, “Openpmu technology platform for synchrophasor research applications,” in 2012 IEEE Power and Energy Society General Meeting. IEEE, 2012, pp. 1–5. [16] H. e. a. Hachisu.H, Nakagawa.F, “Months-long real-time generation of a time scale based on an optical clock,” in Generation, 2018. [17] Y. Amirat, Z. Oubrahim, H. Ahmed, M. Benbouzid, and T. Wang, “Phasor estimation for grid power monitoring: Least square vs. linear kalman filter,” Energies, vol. 13, no. 10, p. 2456, 2020. [18] S. Wang, W. Gao, and A. P. S. Meliopoulos, “Power system real-time monitoring by using pmu-based robust state estimation method,” IEEE Transactions on Smart Grid, vol. 7, no. 1, pp. 300–309, 2015. [19] S. S. Yu, M. S. Rahman, G. Zhang, S. T. Meraj, and H. Trinh, “Comprehensive review of pmu applications in smart grid: Enhancing grid reliability and efficiency,” Chinese Journal of Electrical Engineering, 2025. [20] R. Kumar, B. Singh, and B. K. Panigrahi, “Wide area monitoring system using synchrophasors: Applications and challenges,” in 2023 IEEE Power & Energy Society General Meeting. IEEE, 2023, pp. 1–5. [21] F. Ahmad, A. Rasool, and E. Ozsoyoglu, “Blockchain-enabled secure data sharing for distributed pmu networks,” IEEE Transactions on Smart Grid, vol. 15, no. 2, pp. 1456–1467, 2024. [22] M. A. Hassan and M. A. Abido, “A pmu prototype for synchronized phasor and frequency measurements for smart grid applications,” in 2015 IEEE 8th GCC Conference & Exhibition. IEEE, 2016, pp. 1–6. [23] A. Aguto et al., “Project umuriro: University measurement units: Reliable integration by the rwandan operator,” 2023.