Grid analytics: How much data do you really need?

Electric utilities are experiencing a technology transformation, which includes the deployment of distributed intelligent devices, two-way communications systems, and information technologies to enable greater monitoring and control of their distribution systems. These technologies will increase the volume of data flowing into a utility\´s Information Systems, which in turn will need to be stored and managed. The amount of data that can be generated from AMI/AMR, SCADA, and other intelligent devices can be significant. One strategy, which seems to be a popular option, is to collect all of the data possible and figure out what to do with the data at a later date. As the amount of data increases that needs to be stored, there is a corresponding increase in Information Technology infrastructure, skill sets, and cost. If collecting all data were possible, it is doubtful the data would be stored or structured in a way that would be useable in the future. An alternative to capturing and storing all data possible is to use Information Engineering methodologies to focus on what data is needed for a given task or application. Information Engineering is defined as a set if interrelated disciplines that are needed to build a computerized enterprise based on information systems. The focus of Information Engineering is the data that is stored and maintained by computers and the information that is distilled from this data. [1] It is foundational to understand that data and information are terms used interchangeably, but are distinctly different terms with different meanings. A kWh is a piece of data that is useful information when added to other data such as meter number, account name, or service location. The amount and kind of data required, as well as the necessary time frames, (i.e., historical, real-time, predicted) depend upon the applications, which may include: 1. Geospatial Information System 2. Meter/billing information 3. Asset management 4. Work and workforce manage- ent 5. Network modeling and analysis for planning (e.g., voltage drop, power flow, short circuit, arc flash, contingency studies, reliability metrics, loss analysis, protective device coordination) 6. Operations (outage management, fault location, Volt/VAR control, power quality, demand management, distributed generation and storage) 7. Real-time, active grid management, grid analytics. This paper will discuss how to apply Information Engineering principles to turn data into useful information for a utility as an alternative to the "Big Data" approach to capturing and storing data. Using kWh, and Voltage as examples, we will outline the Information Engineering process to turn these data elements into useful information. Once you have the information engineered, the next step is to use data management methodologies to manage the data that is being gathered. Data Management is a detailed topic on its own and will not be covered on this paper. What data has been captured needs to be managed across the enterprise.