Smart metering deployment typically comprises field devices and solutions, as well as software and hardware. Field devices include meters, fibre optics, power line communication systems, radio frequency mesh networks, and communication technologies (like GPRS, code division multiple access, cloud computing) for last mile and backhaul connectivity. Enterprise applications, such as billing, outage management system (OMS), demand response and revenue assurance, also form a key part of the smart metering system.
However, the most critical building blocks of a smart metering structure are meter data acquisition systems (also known as head-end systems) and meter data management systems. Historically, utilities across the world have adopted multiple approaches for meter data management.
Approaches
Initially, individual head-end systems were used to send data to each utility system. A centralised meter data repository was missing and the interface between each vendor’s meter data collection and utility systems was developed individually. This resulted in duplication of work as each utility system had to perform integration of meter data. Further, any change in the automated meter reading (AMR) system required changes to core utility systems. This resulted in destabilisation risks for the customer information system (CIS) and other complex systems. Moreover, the integration process was difficult and long drawn out.
To overcome the issue of multiple utilities receiving the same data, another approach was developed, which involved building more processing capabilities on the utility system side. All of the new functions required to manage an AMR system were integrated into core business systems such as CIS, the OMS and energy management system. However, there were certain issues with this approach, since changes to core business systems would be expensive and complex. This also resulted in patchy utility systems that were less robust and more difficult to maintain. A new AMR system also required a repetition of the custom development efforts.
Over time, another approach was adopted, which entailed the development of a centralised meter data repository that collects and stores data from diverse AMR systems. A meter data repository makes the system more robust by isolating the utility systems from the details of various AMR systems. Utilities can, therefore, upgrade their AMR systems without putting the core business systems at risk. In this approach, some duplication remained, since similar data processing requirements were still handled individually by various utility systems. A single interface point for AMR data encouraged the integration of meter data repositories with more utility systems. However, diverse requirements from more systems made it critical for the repository to store information in an optimised manner.
This approach has been further modified to build vendor-neutral meter data warehouses with pre-processing capabilities. The method is currently being used globally and by some utilities in India. Data warehouses are geared towards storing transaction data specifically structured for query and analysis, and can therefore address diverse queries from various utility systems with greater ease. It is easier to keep track of the details of edits made to the data using this approach. It also helps in better regulatory compliance. A major issue, however, pertains to data sensitivity since a meter data warehouse with built-in pre-processing capabilities becomes a mission-critical system and requires high levels of security and robustness.
The most preferred approach for utilities is an enterprise service bus model, which has also been mentioned in the Restructured Accelerated Power Development and Reforms Programme specifications. In this approach, there should ideally be one head-end system communicating with the meter data management system. An enterprise bus can connect with other utility systems like geographic information system (GIS), OMS, customer care and billing.
Meter data analytics
Analytics as a concept is the discovery and communication of meaningful patterns in data, involving visualisation. It combines three disciplines – computer science, mathematics and statistics. The key drivers for analytics have been the growing role of data-driven decision-making, the need to discover relationships between groups and behaviour, and the huge amount of data that is now available.
Today, there are various solution providers for different layers of utility analytics applications. Analytics using meter data, with the combination of other applications, can be used for multiple purposes. First and foremost, it helps in theft prevention and revenue protection. Other benefits include the estimation of energy and revenue losses across the distribution grid – from feeder to meter, theft detection, and deciphering tamper patterns and prioritising actions based on them. For instance, a dashboard can be created by utilities for the detection of tampering or thefts in feeders, using analytics that can send real-time information for revenue protection alerts. Other useful information which can be generated by utilities using analytics pertains to area-wise energy consumption, substation revenues, technical and commercial losses and customer type.
It can also help improve power quality by quantifying system losses, evaluating voltage improvement, monitoring delivery voltages, ensuring optimal system performance on a real-time basis, replacing reactive processes with proactive management and monitoring outages, and reporting performance. With smart meters, smart nodes can be placed across the grid, which can greatly improve power quality. For instance, a mash-up of applications such as meter data management, GIS and OMS can help generate a real-time summary of consumers affected by outages and the sequence of automated metering infrastructure (AMI) events, on a GIS map.
The third area of application relates to condition monitoring. Intelligent systems can help utilities identify faults and take corrective steps by providing up-to-date load details, identifying unanticipated load increases, evaluating transformer sizing using loading history and peak seasonal loads, identifying overutilised, underutilised and at-risk transformers and other assets. For instance, using a pie chart, the overall health status of the transformers in the distribution area can be tracked. The user can also locate the transformer on a GIS map to get a graphical view of the distribution network. The user can click on any particular transformer on the map and go on to perform actions like creating a work order or notifying the designated authorities about failures, based on trend and dissolved gas analysis. Similarly, a colour-coded substation maintenance status sheet can be created for providing the status of all substations in that region.
A dashboard for OMS, which is updated on a real-time basis, can also be created using analytics. For instance, a utility can generate a tabular grid showing a summary of the outage incidents at the feeder level based on events and calls received by the call centre. Using an outage incident detail panel, the outage incident detail at the device level (circuit breaker/distribution transformer) at the lowest available granularity by network model can be obtained by the user. The user can also view the device on the GIS map at an appropriate zoom level. Also, an affected customer panel, showing the list of customers affected by the outage incident, can be created. It will include customers who either call the call centre or are detected via the AMI system event. With the outage incident dashboard updated every few seconds, the utility can track the incident details and affected consumers as the outages are fixed. The GIS map of the distribution area can provide a visual indication of the network areas experiencing outages. The user can zoom into the map to see a street-level view with lateral components and equipment of the distribution grid.
An upcoming area where analytics is being used is forecasting. This includes forecasting data such as total customer usage at the feeder or substation level, net usage reduced for distributed generation, demand response available at the facility level or the use of new technologies like plug-in electric vehicles, among other things.
In summary, meter analytics can significantly help utilities derive and manage data to gain useful insights for improving their business processes.
Based on a presentation by Mitul Thapliyal, Senior Principal and Head of Utilities Consulting, Infosys, at a recent Power Line conference