By Dr. Thierry Godart, Raymond Kelley and Bob Fesmire
Meter data is arguably the lifeblood of any utility. It represents the basis for the bulk of utility revenues and is the raw material for building customer relationships, particularly as consumers become more sophisticated and markets become more competitive. How well a given energy retailer handles meter data has a significant impact on the company’s overall performance. Automation is therefore an important part of the utility’s meter data strategy.
In terms of the metering function, however, “automation” typically refers only to automatic meter reading (AMR), the use of advanced metering and communication technologies to automate data collection. This is only part of the puzzle. A utility’s approach to metering actually runs from procurement, installation, data collection and maintenance to value-added information services based on the data collected. This article looks at the entire meter data process (Figure 1) and examines how utilities can make the most of their automation initiatives.
Down with Costs and Risk
The primary motivation for automation is cost savings, and indeed AMR has made great progress in reducing the cost (and time) associated with collecting meter data. However, metering automation can bring cost savings in a variety of areas, beyond data collection and beyond customer meters, and can convey other benefits to the utility and its customers.
Metering automation can drive cost savings in a number of ways. Lower procurement costs can be realized by automating the link between metering vendors and the utility’s inventory system. Remote reading and configuration lower the cost of installing and testing meters. Less human intervention means fewer errors associated with manual readings. Meter data is also a key component in T&D system planning, controlling losses and other areas that can benefit from the efficiency and reliability that comes with automation.
Metering automation also serves to reduce risk. Using meter notification as confirmation of restored power during an outage, for example, mitigates the physical risk of obtaining confirmation on-site. Planning with better input meter data reduces the financial risk of exposure to spot market price fluctuations. Metering automation also can lower risks associated with billing and settlement errors by increasing data integrity and reducing the number and magnitude of customer complaints.
Since true metering automation impacts all components from the utility’s enterprise systems to the meter, significant technology risks exist. Metering automation technology should provide investment protection against inevitable changes in business processes (e.g., from acquisitions, a new CIS, etc.), against changes in communication systems and media, and against increasing data volumes as the utility’s business grows (i.e., scalability).
Up with Customers
For commercial and industrial (C&I) customers, the benefits of metering automation are especially compelling. These organizations are constantly seeking to lower costs, and many have undertaken initiatives to optimize energy usage. Strategic application of meter data plays a vital role both in customer efficiency efforts, and new utility rate structures and special programs.
The need for more data and higher frequency of meter readings has already driven initiatives for metering automation at many utilities. Today, commercial energy rates can rely on actual interval data as well as register values computed by electronic meters (e.g., demand, consumption, time-of-use). A utility’s ability to capture, validate and apply vast amounts of meter data is therefore paramount.
Metering automation can have a direct impact on the cost associated with providing information to customers and account managers, aggregating data for billing purposes, and changing meter configuration. The utility also can realize higher revenues by using actual interval data instead of load profiles in the billing process.
Residential customers also stand to benefit from metering automation, especially with respect to new programs offered by utilities that are predicated on timely, accurate meter data.
So, metering automation is good for the utility, good for customers and good for creating new sources of revenue, but how should the utility go about implementing it? We suggest a strategy using two incremental phases:
- Surgical deployment for C&I customers
- Mass market deployment for residential customers
In Phase 1, the utility targets high-revenue customers for which a return on investment or competitive pressures make metering automation most attractive. The implementation also sets the foundation for Phase 2. For example, enterprise applications such as customer information systems, billing systems, call centers and outage management systems can already be integrated with the central meter data repository.
In Phase 2, the utility deploys a large number of meters with adapted communication technology. The implementation focuses on procurement, installation and testing of the meters, so in the back office, only the new residential data collection systems need to be integrated with the metering automation system. For Phase 2, it is also desirable to have inventory management systems and work order management systems well integrated with the metering automation system.
The risk in Phase 1 resides mostly in back office IT integration, while Phase 2’s risk is mostly in the meter and communication deployment project, assuming the IT solution in Phase 1 was designed to support a large number of meters.
C&I Metering Automation
The automation strategy should recognize a few important considerations:
- Billing for energy sales is becoming a complex process that must be automated with care in order to be profitable
- The utility must be prepared to support metering in multiple jurisdictions
- The utility may want to consider reading meters on the transmission and distribution grid for load survey, distribution planning or other applications
Within the scope of this phase, the utility may also want to consider residential meters in costly (e.g., hard-to-access, unsafe or high turnover) areas. This will require a residential solution that is cost effective for “surgical” deployments.
Figure 2 illustrates a proposed automation strategy including the following main characteristics:
- A layer of data collection systems that feed a central meter data repository and process requests from the metering automation system
- A metering automation system that automates metering processes and sends requests from business users to the data collection systems
- A central meter data repository shared among all business applications, but carefully integrated so queries and reports do not impact billing data transactions
- A file import/export mechanism to accommodate legacy meters and data collection systems in the near term
The installation is typically configured initially for the largest industrial customers (collectively representing fewer than 5,000 meters), but can progress toward small commercial customers (10,000 to 50,000 meters).
The data collection systems are communicating with meters using frequent two-way communication. They can be placed locally near the meters or centrally with the metering automation system depending on costs and the utility’s WAN capability. The systems also should be able to keep executing their scheduled reads with or without a network connection so that when a network failure is resolved, the data collection system will synchronize automatically with the central repository.
The metering automation and data collection systems are integrated so requests can be sent to the data collection system from the enterprise business applications. Requests such as billing scheduling changes, meter configuration changes or special reads can be processed from a uniform user interface independent of the technology supporting a given meter.
Traditionally, validation, editing and estimation (VEE) has been part of the billing process. In the proposed approach, VEE is part of the metering automation system. Data is validated and edited upon collection and stored in both raw and billing quality forms with the billing quality data shared with all systems.
Residential Metering Automation
As Phase 1 of the automation strategy moves into production, the utility should leverage the existing Phase 1 back office integration to accommodate the new Phase 2 residential data collection system(s).
Figure 3 illustrates the automation strategy for residential metering automation. We can expect different data collection systems will be used depending on the meter population density (rural vs. urban).
Typically, the data collected is one monthly consumption value. However, more data such as time-of-use consumption and outage notification may be collected as electronic residential meters become cost-effective. Similar to Phase 1, a two-way communication capability allows for better integration with business applications.
In a deregulated environment, business applications that deal with customer choice issues (e.g., customer enrollment, supplier registration and third party settlement processes) need to access meter data on a timely basis. In particular, the number of special reads may increase as customers switch suppliers and final reads are requested. The proposed strategy positions the utility for cost-effective support of customer choice.
Nuts & Bolts
The proposed solution for enterprise metering automation has three main components: communication server(s), enterprise software, and enterprise application integration (Figure 4).
The communication servers implement a communication strategy and manage the data collection process. The enterprise software includes a meter data repository and a workflow management system. The enterprise application integration (EAI) is the link between meter data and users, and is implemented using program interfaces as well as batch file exchange.
Communication Servers. Metering and communication technologies continue to change. As a result, significant features will remain non-standard, and proprietary communication servers will be required. For this reason, the solution needs a distinct layer between the metering automation and data collection systems. The data collection process is implemented using communication server(s), eliminating the need for the metering automation system to “know” meter and communications specifics.
This arrangement offers a few advantages:
- Flexibility–Diverse communication servers for diverse communication technologies and reading schedules independent of communication and meter technology
- Scalability–Add more communication servers to read more meters
- Reliability–Redundant communication servers and asynchronous parallel operation
Enterprise Software. The solution’s software architecture utilizes three technologies adapted for metering automation: distributed multi-tier architecture, a relational database and a workflow management system.
Multi-tiered architecture is now common in IT systems, where a browser-based client tier is connected through a Web tier to an application tier that manages the business logic and connections with a database tier. Only a Web browser is required to utilize the underlying functionality, eliminating the need for client software. Multi-tier structure also improves network efficiency and allows requests from multiple clients to be executed in parallel.
Relational database technology makes the meter data repository readily available to the enterprise and maintains a high level of reliability. The operational meter data repository is first dedicated to receiving and validating data from the various communications servers, then to estimating and preparing exports for billing purposes. The database is also designed to provide views for business applications without disrupting the ongoing billing process, and is the repository of record for billing and other auditing purposes.
The enterprise software utilizes a specialized workflow management system that is designed to handle large volumes of meter data. In this context, a workflow is an activity that starts from a business user’s request–or batch process–and involves diverse systems. The workflow management system has several key features that are important to note. First, little or no programming is required to implement changing business rules; these changes are managed through a user interface. Every workflow is associated with an owner and timestamp, and the complete state of each workflow is stored in a database to maintain operations across system shutdowns or failures.
EAI. Webopedia (www.webopedia.com) states that: “EAI is the unrestricted sharing of data and business processes throughout the networked applications or data sources in an organization.” For our purpose, EAI is enabled by four integration levels: an application interface level, a business method level, a data integration level and a presentation level.
The application interface level is provided by a messaging interface (API) for interactive use of metering automation workflows and rapid response of requests. In the solution, a library of API functions is designed to process requests from external business applications. Every request is sent to the workflow management system, which in turn initiates a workflow and returns acknowledgement. Similarly, a library of API functions houses data collection services, which are specific to the manner in which a communication server transmits data. For example, requesting a meter read or a meter modification is transparent to the user across various communication servers. The particular API interface to a data collection system performs any required translation into proprietary formats expected by the external systems.
The business method level of integration is provided by an interface that uses the contents of an XML file to control the operations performed by the system. The XML file may contain many operations that are independent and can be executed in parallel, and may also contain a sequence of operations that implement a specific business process.
The data integration level relies on file transfer of bulk data using an XML format. XML offers a unified view of meter data across the enterprise. It also allows automated information processing across different business applications without building intricate data mapping routines.
The implementation of metering automation enables billing specialists, customer service representatives, system planners, marketing analysts and executive management to access and act upon metered information. Metering automation also allows the traditional metering operation to focus on meter equipment issues and optimize its productivity.
In the future, the profitability and sustainability of new energy products will depend largely on the scalability of their associated business processes. Metering automation can enable these services cost-effectively. It allows business applications to interact with the metering system without requiring costly manual interventions of meter specialists.
With an approach like the one discussed here, the metering automation system can be designed for scalability, reliability and flexibility, and risk can be managed from a surgical deployment of meters to a mass-market residential implementation.
Thierry Godart is director of software product development and systems integration in ABB’s Electricity Metering unit.
Raymond Kelley is development manager for software and systems Integration in ABB’s Electricity Metering unit.
Bob Fesmire is a marketing writer in ABB’s Utilities division.