Over the last few years, power distribution companies in India have invested heavily in IT infrastructure and operational technology (OT) systems. They have also shifted almost entirely to electrostatic meters. In fact, the move to smart meters has already begun.
As a result, the discoms are able to generate a lot of data relating to electricity usage and consumption. Most of the discoms have also acquired meter data acquisition systems (MDAS), partly because it was mandated by the Restructured Accelerated Power Development and Reforms Programme (R-APDRP). However, with a few exceptions, they are not yet fully utilising these systems and their capabilities. This is partly because of inadequate in-house capacity and partly because of technical issues.
That said, the discoms that have begun to utilise these systems are seeing very positive results, not just in load management and revenue protection, but also in areas such as outage management and condition-based monitoring. As discoms become more familiar with these systems and make greater use of them, MDAS is likely to play a big role in improving operational efficiency.
This write-up attempts to describe the basic elements of MDAS, the key potential areas of application, and the experience so far.
MDAS/MDM/MDA
The three key steps in using metering data most effectively are acquisition, management and analytics. To meet these tasks, we have MDAS, meter data management systems (MDM) and meter data analytics. They have many overlapping areas and it is perhaps best to think of them as part of an overall meter data acquisition and analytics system.
The key components of these systems are field devices (such as meters), communication media (RF or PLC or cellular), head-end systems, meter data management systems (data warehouse or repository) and analytics. The objective of this extensive structure of multiple devices and systems is to measure, collect, transfer, store and analyse electricity consumption-related data.
In the initial stage, data is gathered and stored in the form of logs and meter events. The stored data is then transmitted from the meter to the collection device. Data is retrieved through an interoperable head-end system. The raw data is then transformed into an understandable format, which is validated and standardised. This data is integrated and stored in a proper format for further query and analysis. In the last stage, information is analysed and displayed through dashboards and reports.
MDAS enables the integration of metering data (from AMI or otherwise) with enterprise-wide systems and acts as an interface for various applications within the utility.
Deployment approaches
The approach to the deployment of MDAS has evolved over the years (see Figure 1). Initial deployments involved the installation of individual head-end systems to send data to each utility system such as GIS, OMS, customer care and billing systems, revenue protection systems, etc. This often resulted in duplication of work, which was overcome by customising core business systems to accommodate the metered data.
Utilities later started deploying centralised meter data repositories to collect and store data. A centralised repository isolated utility systems from the details of the associated AMR system, thereby allowing utilities to upgrade their AMR systems without changing the core business systems. Recently, the concept of vendor-neutral meter data warehouses has gained popularity. These warehouses are equipped with data processing capabilities and allow for interoperability at the device and head-end levels. A key advantage is that these warehouses are specifically designed and structured for undertaking analysis and answering queries. However, data confidentiality and ownership continue to remain a concern.
The “latest” and the most preferred approach these days, which Indian utilities can leapfrog to, is the enterprise service bus model (see Figure 2 on opposite page). A head-end system communicates with the MDAS. Data from the AMR and AMI systems is conveyed to the MDAS, which transfers it to an enterprise bus. The enterprise bus then makes the data available to utility systems. This model was also mentioned in the R-APDRP specifications.
Benefits of MDAS
MDAS enables efficient collection of large volumes of interval meter data and transfer of the same to multiple applications. This allows utilities to adopt new technologies and unify billing processes without completely re-engineering the existing meter reading processes. Importantly, since MDAS allows the framing of volumes of interval data into manageable information, it helps utilities move beyond billing information to focus on utility management. It enables the deployment of advanced analytics and the use of big data solutions.
With MDAS, utilities can better exploit systems that they may already have in the areas of billing, outage management, demand response, revenue assurance, SCADA, etc. Utilities are able to control theft, protect revenue, improve power quality, manage assets and forecast load/demand in a more effective manner. Some of the key applications are listed in the accompanying box. In addition, with two-way communication, MDAS provides customers with access to information on consumption patterns. It enables energy efficiency by allowing them to access demand-side management applications and control their consumption. It helps settle billing disputes. Customer service calls decline since the transactions can be viewed online. MDAS also helps utilities in complying with the regulatory and policy requirements related to power quality and conservation.
Deployment in India:
R-APDRP
Almost all of the power discoms have deployed MDAS in one form or the other. The pioneers in this area were private discoms such as Tata Power, BSES and CESC, which were able to significantly improve operational efficiencies and reduce AT&C losses with the deployment of these systems.
The public sector discoms began acquiring MDAS primarily because it was one of the 18 IT modules that were mandated under the central government’s R-APDRP, which was launched in 2008. The objective of the R-APDRP (now subsumed into the IPDS) was to reduce AT&C losses with the help of wide-scale deployment of IT systems. The role of MDAS, as stated under the programme, was to “remotely acquire meter data from select DTs, feeders, HT consumers and boundary meters; monitor key distribution parameters; use meter data for accurate billing purposes and for identifying exceptions; and generate MIS reports for efficient planning and decision-making”.
The next several paragraphs describe the experience in India so far….
MSEDCL
Maharashtra State Electricity Distribution Company Limited (MSEDCL) implemented an AMR system for its HT consumers, feeder meters and DTs under the R-APDRP. The MDAS was later developed and deployed in-house. “We have been using our home-grown MDAS for the past five to six years and as such, there have been no limitations from the software perspective,” says Yogesh Gadkari, chief general manager, IT, MSEDCL. The biggest advantage of in-house MDAS development and maintenance is that enhancements to the software are executed almost immediately.
All protocols for IR and RF meters have also been developed and standardised in-house. Further, the IR and RF meters, which have been sourced from around 13 vendors, have 100 per cent interoperability with the data acquisition system.
The utility frequently utilises the data generated and analysed by MDAS to identify exceptions to detect tampering and pilferage. In order to maximise the benefits, the utility has extended the use of MDAS to non-R-APDRP areas.
The utility has engaged multiple service providers depending on the network connectivity in each area so as to maximise the meter-modem read count. CDMA is the primary mode of communication used, failing which, the utility relies on CMRIs. Network connectivity remains a key concern for the utility. Nearly 25,000 SIM cards installed across the state for meter communications are currently not working. MSEDCL is working with Reliance Communications to resolve this issue.
CESC
CESC Limited deployed MDAS in 2007-08, following the implementation of AMR in 2005. The billing system software was developed in-house, while the MDAS platform was sourced from Secure Meters. For collection of meter data, the utility initially relied on GSM technology, but eventually migrated to a GPRS-based system.
Communication network availability has been a challenge, especially in a few pockets. However, CESC has developed a close-ended mechanism, in partnership with Vodafone, to resolve such issues. It has put in place a dynamic monitoring system, under which network health is monitored on an everyday basis. The system ensures that problems do not crop up on the final billing day. With this mechanism, CESC has been able to achieve a success rate of more than 99.6 per cent.
CESC has also been using home-grown analytics tools since 1995 for generating exceptions and alerting field crew to cases of pilferage or metering defects. In 2010-11, CESC deployed a DT analytics tool, which has helped in identifying DTs that are overloaded, operating on low/high voltages, low power factors, etc. “The result is that we hardly have any overloaded transformers. If a DT is 80 per cent loaded in winter, we know that it will get overloaded in summer. We have thus moved from a reactive regime to a proactive regime. A significant impact has been that we are now procuring around 50 transformers a year, compared to 350 to 400 earlier. The failure rate of transformers has declined to much below 0.5 per cent,” says Udayan Ganguly, deputy general manager, CESC.
A key success factor has been that, in order to deal with the issue of interoperability, CESC had stayed with the same vendor for the supply of all components including modems, meters and platforms. Further, the utility has been able to re-orient all systems and procedures within the organisation to reap the maximum benefits from MDAS.
BSES
Meter data in areas served by BSES is electronically downloaded every month from its 3.6 million meters, including single-phase meters. Multiple methods are used for collecting data, including AMR, optical fibre, CMRI and hand-held devices. Given the sheer volume of data generated by these systems, the utility decided to deploy MDAS.
The system divides the collected data into two parts, which are sent for billing and analytics. The data stored in the meter memory is analysed for clues to fraudulent abstraction by consumers using the Meter Data Analytics Module (MDAM). The module uses various techniques and maturity models to help detect theft. Energy input data from over 11,000 DTs is compared and correlated with the energy consumption at the consumer end to identify probable cases of theft. The energy usage pattern across various consumer categories is analysed for deviations, indicating theft or unauthorised use. In one instance, BSES detected 52,000 cases of ESD tampering with the help of analytics. Aside from theft detection, the MDAM is used to generate operational reports related to supply quality and network health.
Key issues
The deployment of MDAS, particularly by public discoms under the R-APDRP, has not been without issues. There have been problems with modem installations, interoperability and communication technologies. To begin with, there were delays in the installation of modems because of incorrect mapping of GIS base data. The technicians had difficulty in identifying the correct locations for modem installation. Asset mapping was not complete because many DTs and feeders were not metered. The existence of meters of different makes within the same area caused interoperability problems.
The biggest issue, which has still not been fully resolved, relates to communication technology. Many discoms opted for cellular (GSM/GPRS/CDMA) technology, but there were reliability and uptime issues. Discoms also faced significant difficulties in transferring data to the data centre using GPRS due to the huge volume of data generated. In many towns where connectivity was provided through MPLS links, meter data could not be downloaded properly through GPRS modems. There were also challenges relating to range and spectrum availability, operating frequency and network costs.
The discoms are beginning to find solutions to the communication problem. In some cases, they have reconfigured their MDAS to transfer the daily load survey data in a staggered manner during off-peak hours, while still using GPRS. In other cases, they are moving towards a “hybrid” approach using a mix of RF, PLC, cellular and even optic fibre.
Some of the discoms did not have adequate data storage capacity and some data was, therefore, lost. Irregular meter readings also meant data inconsistency and missing data. This limited the use of MDAS. However, issues like these are not insurmountable and are slowly being addressed and resolved.
The utilities need to make sure that they pick the right communication technology (cellular or otherwise), demand interoperability, and ensure data integrity and security. They also need to develop analytic capacity (preferably in-house) to use these systems. Most importantly, the management should proactively leverage these systems and capabilities for more effective decision-making. Private discoms have been able to, and public discoms can, use MDAS for detecting theft, reducing AT&C losses, managing outages and reducing equipment failure rate.