In the recent past, the electrical power system has been changing by leaps and bounds. In this new generation, consumers are becoming prosumers. There has been a quantum shift from conventional generation to renewable generation with the rise of distributed solar generators and the changing behaviour of load due to the increase in data centres, electric vehicle (EV) charging infrastructure and battery storage stations; and the growing use of variable frequency drives for motors and LED lights. The changing dynamics of the power system call for a change in the conventional system of network management. Power professionals are utilising various innovative solutions to meet the increasing demand from smart consumers and prosumers for improved power quality and reliability, and the latest power generation and storage technologies. The increased focus on environmental sustainability and reducing the carbon footprint has in turn increased the requirement for smart solutions such as smart grids, smart meters, advanced metering infrastructure, internet of things (IoT) devices and sensors-based monitoring.
Technological interventions for network monitoring
At Tata Power, we are using various new smart tools and software to develop an agile electrical network that can embrace any new technological solution. Our Mumbai distribution network has one of the lowest indices with a system average interruption duration index (SAIDI) at 8.5 minutes, and a system average interruption frequency index (SAIFI) at 0.5. This achievement was made possible by the strong electrical network and the use of smart solutions. To facilitate supply for new consumers and maintain the health of the network, it is necessary to carry out extensive studies before initiating planning, analysis and operation, with a holistic view of the existing and future network requirements. We carry out various network studies at regular intervals with the help of software simulations. Various software such as Cyme, MiPower and e-tap are used by our engineering, network and protection teams for various studies.
We have mapped all of our network with our geographic information system (GIS) software. Our systems are always updated such that any change in the network can be captured via the GIS software. We also integrate all network-related data from various applications such as GIS, enterprise resource planning software, Cyme, order management systems, and advanced distribution management systems (ADMS), with handshaking of all system data, thus enabling improved predictive analysis with a pan-picture view. All our electrical equipment is mapped via software for real-time network modelling and analysis, improving accuracy. With these tools, we could reduce distribution losses to less than 1.4 per cent.
These days, we are witnessing large-scale penetration of rooftop solar and EV charging infrastructure into the network. To assess the impact, we perform simulation studies on our network and observe changes in performance parameters. Special attention is required to keep the reactive power and harmonics in the network under control, as new-age devices and loads are non-linear. Based on our analysis, we have introduced a few dedicated harmonic meters in the network for deep-stick study, as well as filters in harmonics-prone areas.
To increase reliability, we have created a self-healing network through the automation of our ring man units at the 11 kV and 22 kV levels. This system automatically disconnects faulty sections, while supply is restored through a healthy network. The changeovers can be achieved in seconds with the use of voltage sensors and artificial intelligence (AI), and the motorisation of switchgears. Moreover, we are using our patented voice-guided auto instructor to ensure the safety of working personnel in the field. The coordinates of all our existing consumers are mapped in the land base layer of our GIS platform. When processing a new connection application, we first take the customer coordinates and mark them in the GIS land base. Then we check the feasibility of the network in the vicinity of the prospective consumer. We run various studies after the simulation to check the voltage drop and load flow in the network, and perform loading analysis. Before adding a new consumer substation to the network, a load flow analysis is done using simulation software. Accordingly, the network action plan is decided. We have rolled out around 2 million smart meters across the Tata Power network, which provide real-time details of outages and loading to consumers through a mobile application. The application empowers consumers to perform energy efficiency studies by providing data such as projected monthly consumption, past trend analysis and peer-to-peer comparison. As we get the meter data in mobile data management servers, we are using machine learning (ML) tools to detect theft based on consumption signature analysis.
We are also strengthening our network by using low voltage automation and IoT devices to improve the reliability of our last-mile asset monitoring. We are using smart IoT devices such as distribution transformer monitoring systems for monitoring of temperature, oil level and load. We have executed a few pilot projects for life extension of transformers, wherein the recorded database of temperature, weather conditions and seasonal changes was used, trend analysis was conducted, and the development of AI/ML tools was undertaken for mass-scale deployment. We are using QR code-based asset monitoring and maintenance history monitoring systems for ease of operation. We have equipped our network staff with smart devices for remote monitoring of network equipment and to enable them to address anomalies as quickly as possible. Field force automation, supervisory control and data acquisition, distribution management system (DMS) and ADMS applications enable real-time monitoring and control alongside data analysis. This helps in optimal network planning.
Periodic review of network through simulation study
We perform periodic reviews of our network through various studies to increase its efficiency, and reduce the disruption and safety issues caused by the variability and intermittency of loads and renewable generation.
- Network optimisation: This study is done at regular intervals, as it is critical for continuously growing networks. The normally open points (NOPs) are strategically placed in the network to meet the load in case a source or ring network fails. As the network grows, the NOP nodes need to be changed. The optimal placements are generally determined through simulation in order to minimise losses and optimise the section length and loading.
- Short-circuit analysis: This study is done to review fault levels at different locations, or to determine the fault levels of equipment used. Data extracted from the GIS and data regarding source fault levels are modelled in the simulation software. This helps in checking network reliability and equipment safety.
- Reactive power planning: This study is done to check the adequacy of the capacitor installed and the prevailing network parameters. This helps in reducing unscheduled interchange charges and optimally utilising capacitor banks, based on reactive drawl.
- Power quality analysis: We periodically perform power quality analysis of our network and monitor the sags, swells and fluctuations. Accordingly, we have deployed active filters, passive filters, dynamic voltage regulators, and ride-through devices at the nodes identified after analysis. Such studies are essential when generation, transmission and distribution networks are placed very close to each other.
- Contingency analysis: Ensuring the reliability of a network is of utmost importance, and this requires planning and operational enhancement. Contingency analysis helps in planning restoration after scheduled or forced outages, with switching sequences. This can improve the pace of restoration. It also facilitates planning for the installation of new equipment and the diversion of existing assets to ensure reliability for critical consumers.
- Reliability indices computation: Load-point indices indicate the interruption frequency and the outage duration experienced at a load point. These are computed for each “sub-zone” in the feeder. All customer loads within a sub-zone experience the same interruption frequency and outage duration. Once the interruption frequency and outage duration are known at every load point on the feeder, system indices such as SAIFI, SAIDI and the customer average interruption duration index are computed.
Smart grids and digital grids
In the metro city areas, we are facing a space crunch with regard to augmentation and addition of grid substations. Gas-insulated substations and digital grid stations have become viable options. We have conducted a pilot for a digital grid with a process bus at one of our grid stations in Delhi, and the results are encouraging. The increased digitisation also requires us to adapt the best cybersecurity systems to protect our network.
Conclusion
The spontaneously changing dynamics of electrical power generators and loads necessitate a smart, dynamic distribution network management system. There is large scope for innovative solutions with respect to network monitoring systems, to facilitate the integration of all network elements. Increasing digitisation and use of IoT devices necessitates a focus on cybersecurity to ensure grid resilience and power security. The next decade will be one of data analysis and ML, for optimal network management.