Transforming AMI data into business information


PPL Electric Utilities Corporation (PPLEU), a subsidiary of PPL Corporation that provides electricity delivery to approximately 1.4 million customers in Pennsylvania, recently implemented a Meter Data Management System (MDMS). The MDMS, developed and installed by Nexus Energy Software, is providing data validation, process management, storage, and analysis of daily, hourly and 15-minute interval meter readings collected throughout the day by signals sent over power lines from all customers’ meters.

This combination of AMI and MDM systems makes PPL the first to implement the largest operational system in North America in terms of hourly read data volume, and will be the most comprehensive in terms of applications that leverage the data.

The MDMS (Figure 1) is the centralised storage and control system for data operations to process meter data collected from PPL’s AMI systems – TWACS® by DCSI, Comvergeâ„¢ PowerCAMP and Itron’s MV-90. It also provides the platform to apply this data to integrated applications for enhanced revenue protection, complex billing, load forecasting, distribution planning and maintenance, load research, ISO settlement and customer energy management services, which will be implemented over the course of 2007.

Nexus’ Energy Vision® and ENERGYprism® products provide PPL with a system where data from the various metering and supporting systems are stored, managed and leveraged from a central location. Previously, PPL collected data from its different automated meter reading systems, and stored and used the data in various systems throughout the utility. Employees from diverse business areas can now combine this meter data with the other data stored in a variety of supporting systems, such as weather, SCADA data and GIS information, providing additional analysis and benefits.

Implementation of the central repository includes interfacing meter sources to collect daily readings of hourly consumption for all 1.4 million end points as well as 15-minute interval data for 2,500 end points. Data from the meter reading sources are imported via specific schedules generated by PPL’s AMI Operations group and placed into collections. T Configuration specialists establish the collection groups and create specific estimation, editing and validation rules using on-line screens within the central repository. As a result, this provides PPLEU with enough data to create 1.4 million unique profiles for future use.

Let’s review how PPL will transform AMI data into business information with MDMS applications.


The AMI Operations group administers the MDMS as well as the AMI systems. It is accountable for assuring that the data is collected at the highest success level achievable. The Operations group is able to achieve meter reading success between 96 and 99.9% from hourly to billing reads. Achieving these high levels is important to meet the business requirements for data availability and to support the analysis of data within the MDMS.

Once this data is collected from the AMI systems it is then scheduled to be loaded into the MDMS on a daily basis. The data files are transported through an enterprise architecture infrastructure using PPL and vendor-designed interfaces from PPLEU’s AMI systems, Customer Service System (CSS) and weather data to the MDMS data repository. Each interface generates a set of messages. The messages are categorised in several levels of severity from “˜Warning’ through “˜Fatal errors’. A member of the Operations group checks the logs (Figure 2) to make sure that the data is loaded as expected. If problems are found, the source data is corrected and the interface is rerun.


Once the data is in the repository, it is validated for various data conditions. If the data has gaps they are detected through the validation process, then the gaps are filled using automated estimation processes. Additional validations include sum and spike check procedures based on using available data, and validation, estimation and editing (VEE) routines. For example, when the hourly data is collected there can be gaps that exist in the data collected through the AMI systems. Using the daily data, the gaps are filled based on the consumption calculated, using the difference between today’s and the previous day’s daily readings.

Since the Nexus application provides for various levels of security, a Configuration Manager from the business side defines the collections (Figure 3) and sets the validation rules using seven validation tests in the Standard Validation Library found within the application. Data from the repository is stored in various scenarios. For example, the actual data retrieved from the meter source is considered to be “˜working data’ and data that is edited or estimated is considered “˜approved’ data (Figure 4). The various scenarios will be used in other Nexus applications such as Bill Vision® for complex billing, Bill Prismâ„¢ for customer service representatives and customers, and Load Vision® for forecasting and settlement.

Customer care applications

The ENERGYprism application deployed in the fall of 2006 has provided benefits of the new MDMS to PPLEU’s Call Centre customer service representatives. The user interface for the customer service representatives gives them access to TNS, MV-90 and Comverge data all through one single source. In addition, an analysis tool for helping customers to understand why their bills have changed help the customer service representatives provide a better customer experience with “˜First Time Every Time” call resolution (Figure 5).

In the summer of 2007, the features available to customers will be expanded via the Internet. Customers will have the ability to view detailed information about their account, perform analyses to better understand their energy usage, and answer questions about their bill and why it might have changed. They will be able to view and analyse their energy use and use the information to help them make decisions about what rate might be best for them. For example, a customer will be able to graph his usage against temperature to view how weather affects his energy usage (Figure 6). Commercial customers who receive this load information in a printed hard copy today will be able to log in online to view, download and print their usage at any time, allowing them to make better informed energy choices.

Revenue protection

In phase two of the MDM project, PPL plans to implement Revenue Vision®. This application will aid in identifying revenue loss from energy theft and equipment error. The PPL Revenue Protection team will enhance its already thorough and comprehensive revenue protection efforts by using this analytic tool to create a “˜hot list’ of potential revenue loss accounts for further investigation.

This tool will allow a revenue protection specialist to configure and maintain analytic logic to test the contents of the MDM database and isolate those data anomalies that suggest a high likelihood of current revenue loss to PPL. The specialist will use a set of new screens and reports to automate existing manual processes that are currently performed by PPL. It is believed that over time, as PPL gains more experience using the application, that new functionality will be built into the workflow to provide additional capabilities and streamline the process.

Complex billing

The implementation of complex billing at PPLEU will provide an automated scalable solution to replace the current manual processes in place for calculating complex bills for the largest industrial and commercial customers. This application will be used in conjunction with the CSS system to provide demand and usage determinants to bill these customers. PPL’s CSS will continue to be responsible for handling “˜simple’ calculations, where Bill Vision will be used for “˜complex’ calculations. Additionally, it will enable PPL to prepare for the lifting of its generation caps in 2010, by enabling billing to TOU customers using hourly meter data.

Distribution planning

PPLEU is currently using a legacy application to support distribution planning and capital budgeting decisions. This application was developed in-house with the purpose of facilitating feeder loading analysis over a 5-year planning horizon. The primary objective of applying the Nexus Wire Vision® implementation will be to replace the capabilities of this legacy system and serve as the single repository of data and results that support the distribution planning function.

Through aggregation and analysis of interval data, it will address the business needs of distribution planners and operators by enabling them to better understand and anticipate loading conditions, identify and prevent device overloads, and forecast short term and long term growth to help improve reliability and reduce costs. Regression modelling and forecasting capabilities are provided within the application to deliver accurate and versatile historical and forecast load reporting. This application is scheduled for 4th quarter 2007 implementation and is presently in the functional requirements phase.

Load research

Historically, the PPLEU’s Load Analysis group has managed all interval data using a legacy system designed in-house. This system will be retired and all its functionality replaced by the MDMS. Several of the tasks previously performed by the Load Analysis group will be replaced by other MDMS business areas. Validation of billing, rates and market research, and real time pricing interval data will be performed automatically by the VEE for the most part. Also, the printed plots of interval usage data currently mailed to some industrial and commercial customers will be replaced with a graphical online view of usage available through customer presentment via the Internet.

The new application will be used by the Load Analysis group directly in its daily analysis tasks. They will use Load Vision to perform analysis of load research samples to produce hourly rate class expansions. This involves identifying samples, stratifying samples to produce case weights, and expanding the sample to yield total hourly loads for each rate class along with an associated confidence interval. Additionally, the Load Analysis group will use Energy Vision for unbilled revenue estimation.

ISO settlement

Settlement will enable PPLEU to reconcile load estimations in a shorter period of time and more frequently than completed today through more accurate load forecasts. Also, it will be possible to use more current hourly load information to reconcile billing with PJM on a close to realtime basis.

Benefits and metrics

From a performance standpoint, PPL is working on metrics with Nexus to determine efficiency of data loads and validating that record counts are accurate on a daily basis, as well as the length of time taken to load the data from AMI and other business systems to the MDMS. In addition, AMI Operations is working with the application business owners to define requirements for data quality and reliability metrics.

From a benefits perspective, PPL will be leveraging the meter data to accomplish the following objectives:

  • Provide customers with tools to analyse their bill, view payment history and energy saving tips.
  • Streamline the billing process.
  • Validate, edit and estimate daily, hourly and interval data automatically without manual intervention.
  • Improve theft detection.
  • Increase reporting of unbilled revenue.
  • Enhance load profiling. 
  • Expand distribution operations and planning functions.

The value of leveraging the meter data will:

  • Help to further increase customer satisfaction.
  • Enable flexibility in introducing new rates.
  • Leverage profile data for enhanced load forecasting and ISO settlement.
  • Reduce theft and increase revenue. 
  • Increase operational efficiency.


When the MDMS is implemented with all the applications, it will be the largest system in North America to use AMI hourly interval data. PPLEU is very excited about how it will be able to leverage this data to further increase customer satisfaction, seek improvements in theft detection, enhance distribution planning efforts, and improve load forecasting and analysis – in essence, transforming AMI data into business information with MDMS applications.