Loading...
Smart monitoring of low voltage networks to combat illegal power connections
Author(s)
Solani, Thulani
Date Issued
2026
Type
master thesis
Publisher
Cape Peninsula University of Technology
Abstract
The power system has continually been evolving with the integration of sophisticated software and hardware. This evolution has focused on medium- and high-voltage networks with a stateof-the-art level of automation, while automation for low-voltage (LV) distribution grids remained relatively unchanged. The current LV distribution infrastructure still contain many legacy devices and systems that require intervention to enhance efficiency and address issues such as non-technical losses (NTLs). NTLs are not technical shortcomings but are attributable to electricity theft, meter tampering, administrative inefficiencies, and non-payment of bills. The demand for an optimized, resilient and sustainable power supply necessitates the development of intelligent methods for the production, storage, and distribution of electricity. Currently, government strategies include the adoption of smart grids in place of conventional electrical grids. Smart grids allow for the distribution of electricity to consumers using digital communication networks, thus enabling continuous monitoring and analysis of the electrical supply. This research study addresses the global imperative to reduce losses in electricity networks by designing and implementing a smart low-voltage (LV) distribution network grounded in an Internet of Things (IoT) architecture. As the world shifts toward a green-energy future, optimizing power grids to operate efficiently becomes essential for sustainable energy resilience, with electricity theft identified as a significant threat to grid security and performance. Smart grid technologies, underpinned by Internet of Things (IoT) capabilities, offer continuous monitoring and analysis of electrical supply by delivering electricity through digital communication networks. The primary objective of is to design, develop, and implement a smart low-voltage (LV) distribution network monitoring and control system anchored in IoT architecture to mitigate electricity theft and enhance protection, automation, and control of LV networks. The study posits a gap in the literature: while numerous solutions focus on detection, they often rely on time-consuming utility-led inspections and disconnections, imposing substantial resource burdens. There is a need for scalable, proactive, and automated interventions that minimize direct utility involvement while effectively reducing non-technical losses and improving grid resilience. The research study proposes a framework that leverages real-time data and theory-driven algorithms to enable rapid detection of faults and illicit connections, thereby enhancing observability, fault detection, and proactive maintenance. The thesis findings and deliverables contribute to the expansion of the knowledge base in the field of smart grids in different ways: 1. The provision of a novel scalable, proactive and automated solution at the low-voltage distribution level, that minimizes direct involvement of the power utility while effectively reducing not-technical losses and improving grid resilience. 2. The development of an integrated IoT-based framework with current and voltage sensors deployed at strategic locations within the LV network in order to identify any anomalies. 3. The construction, design and development of a lab-scale prototype smart LV network solution with validation of operation by means of Proteus software simulations. 4. A pilot-ready solution with the capability for deployment within a LV network of the South African power utility, ESKOM. The thesis findings and deliverable contribute to extending the knowledge base within academic institutions and other research institutions.
Additional information
Thesis (MEng (Electrical Engineering))--Cape Peninsula University of Technology, 2026
File(s)![Thumbnail Image]()
Loading...
Name
Solani, T_212236946 (1).pdf
Size
11.13 MB
Format
Adobe PDF
Checksum
(MD5):f55fa217c9f8f2d78901b6ef8dc28a64