With several major infrastructure initiatives on the horizon, including a large meter replacement project, PPL needed a better way to manage and analyze financial, operational and asset data.
A partnership with IBM is helping PPL integrate and analyze its data—for example, helping it predict which accounts are at risk of falling into arrears, and reach out to customers who need help.
99%success in predictive analysis of customer payment behavior
5,400additional at-risk customers identified, helping PPL reach out and offer support
Targetedinfrastructure investments optimize reliability, costs and customer care
Business challenge story
Transforming to a more intelligent gridPPL Corporation runs six regulated utilities: three in the US (Pennsylvania, and two in Kentucky) and three in the UK (Western Power Distribution in the Midlands, South West, and South Wales).Each of the six utilities has its own unique characteristics, but a common factor across all six is the need to maintain reliable electricity distribution networks. As a result, PPL has a strong focus on infrastructure and asset management: investing in maintaining and upgrading its networks, and introducing new technologies to create an efficient, smart, self-healing grid.Additionally, PPL must meet exacting regulatory standards that govern reliability, security, performance and costs, in a constantly changing business environment.Domenic Breininger, Supervisor – Business Intelligence, comments, “We need to transform our grid and replace aging infrastructure, with significant capital and operational expenditure programs. For example, in Pennsylvania we are investing about USD1 billion per year in our system.“It has become increasingly clear that the only way to build a sustainable, cost-effective grid for the future is to gain deeper insight into how the network operates. And that means we need to find better ways to capture, manage and analyze data from across the whole business.”
Data holds the key
Planning infrastructure investments effectively requires a detailed understanding of the load, usage and timings of electricity consumption. For example, for one of PPL’s utilities, consumption data is generated by 1.4 million advanced meters installed in customers’ homes and businesses, which capture accurate readings at regular intervals throughout the day. Other important data sources include the company’s asset management and maintenance systems and finance systems.<br><br>Over time, PPL had assembled multiple data warehouses to manage data from these various sources—but there was no overall framework to integrate these warehouses together. The volume of data that needed to be managed was also growing explosively—especially due to the streams of information generated by hourly meter readings.<br><br>Domenic Breininger explains, “We will soon be upgrading many of our existing advanced meters to second-generation hardware which sends readings every 15 minutes—effectively quadrupling the load on our metering database. In almost every area of the business, the rate of generating information is rising year-on-year.<br><br>“To turn this data into valuable insight, we needed to transform our siloed data warehouse landscape into a fully integrated data model across the whole business. This would give us better data governance and flexibility for incremental expansion of analytics. To make this possible, we wanted to implement pre-designed data warehouse schemas that would be appropriate for our business, instead of reinventing the wheel for ourselves.<br><br>“Adopting a standard industry data model would make it easier for us to perform analyses ourselves, rather than outsourcing to a third-party data provider. This in turn would increase our business agility and help us make better decisions based on accurate, timely information.”<br><br>He continues: “We knew IBM had created data models for other industries, and we created a business case for working with IBM to develop a model for energy and utilities too. Even though utilities are usually conservative about adopting new technology, we realized we needed to innovate. Our leadership team was impressed by the quality of IBM’s proposal: the President of the Pennsylvania utility became the executive sponsor, and our CIO sponsors the project across the wider group.”<br><br>PPL is working with IBM to implement the IBM® Data Model for Energy & Utilities, which provides an integrated methodology and approach for creating advanced industry-specific information solutions. The Data Model combines deep expertise and industry best practice for both business and IT stakeholders, helping to accelerate the creation of business conceptual models, the design and deployment of data warehouses, and the development of data extraction and analytics solutions.<br><br>The solution at PPL is based on IBM InfoSphere® Data Architect, which facilitates the discovery and modeling of the company’s data assets; IBM InfoSphere Information Server, which provides data integration, data quality management and governance; and IBM PureData® System for Analytics, a high-performance data warehousing appliance that dramatically accelerates analytics performance.<br><br>Domenic Breininger elaborates: “The first big use-case will be to support our second-generation meter rollout in Pennsylvania, which is set to produce around 32 TB of new data every year.<br><br>“This USD 0.5 billion investment aims to reduce the cost of service for customers by giving them insight into how their energy usage affects their bills. The data will enable us to diagnose which of their behaviors are increasing their costs, and show how using off-peak power can reduce costs.”<br><br>
Creating a standard way to manage data
The introduction of the Data Model for Energy & Utilities is helping PPL build a comprehensive information and data warehouse model, with integrated reporting and analytical capabilities based on standardized business terminology. This robust set of business and technical data models is fully extensible and scalable, and will help PPL manage the enormous increase in metering and other data.<br><br>While the Data Model is being deployed, PPL has also been working with the IBM Client Center for Advanced Analytics on a series of Rapid Analytics Results projects—brief eight- to ten-week engagements that provide “data science as a service” to address specific business problems quickly.<br><br>The IBM Client Center has helped PPL leverage IBM SPSS® Modeler and IBM PureData System for Analytics to build and run predictive models that identify customers who are likely to have difficulty paying their electricity bill. By being able to predict payment behavior, PPL will be able to focus on accounts that are most likely to fall into the collections process, and, if possible, intervene before that happens.<br><br>Domenic Breininger remarks, “The collections model aims to help us avoid bad debts by identifying people who are struggling, and trying to help them find better payment options. For example, we can make them aware of government programs that they would be eligible for, or we can work out a payment plan, before they go into collections.<br><br>“The initial results have been very, very positive. With two years of data, we were able to predict which accounts would go into collections with an accuracy rate of 99 percent. The IBM team helped us identify 5,400 accounts that were at risk, which gives us an opportunity to track down revenue that might otherwise be lost.<br><br>He adds: “For accounts that are already in the collections process, we are now building a payment propensity model, which enables us to predict which accounts we are most likely to be able to collect from successfully.<br><br>“This helps us focus our collection resources on the customers who are actually most likely to pay. In the future, we may factor some of this model into our credit scoring to help us decide whether new customers need to put down a deposit when they sign up.”<br><br>Domenic Breininger concludes, “The IBM Data Model for Energy & Utilities is helping us unify our data into a coherent architecture, while the IBM Client Center for Advanced Analytics is giving us a preview of what we will be able to achieve in terms of developing game-changing predictive models. On the longer term, our partnership with IBM is going to help reduce our operating expenses and ensure that the significant investment we’re making in our assets produces the best value and outcome for our customers.”<br><br>
About PPL Corporation
PPL Corporation, one of the largest investor-owned utilities in the US, provides electricity for 10.5 million customers in the US and UK. Employing around 13,000 people, the company achieved annual revenues of USD 11.5 billion in 2014, and has won 38 J.D. Power awards for customer satisfaction.
- E&U: Intelligent Utility Network (IUN)
- IBM Global Business Services
- InfoSphere Data Architect
- InfoSphere Information Server
- PureData System for Analytics (powered by Netezza technology)
- SPSS Modeler
- SPSS Modeler
Take the next step
IBM Analytics offers one of the world's deepest and broadest analytics platform, domain and industry solutions that deliver new value to businesses, governments and individuals. For more information about how IBM Analytics helps to transform industries and professions with data, visit ibm.com/analytics. Follow us on Twitter at @IBMAnalytics, on our blog at ibmbigdatahub.com and join the conversation #IBMAnalytics. View more client stories or learn more about IBM Analytics