EPRI: Electric Power Research Institute

Smart Grid Resource Center

Data Analytics Initiatives for Transmission and Distribution

EPRI Smart Grid Demonstration

Distribution Modernization Demonstration

The Distribution Modernization Demonstration (DMD) project is one of EPRI's two parallel Big Data initiatives focused on identifying, documenting, and valuing data oriented applications and relevant support infrastructure. The DMD target is distribution system sensors, datasets and use cases, while the sister initiative described on the opposite page targets transmission sensors, datasets and use cases.

Toward this focus, the member advisory teams for the two initiatives collaborated with EPRI to develop a data-to-value vision that generically describes "The Path to a Smarter Power Grid." The premise is that the smarter grid of the future will have five data driven attributes, with each enabled by information and communications. These five attributes describe a grid that is more intelligent, more automated, more flexible, more predictive, and more interactive as compared to today's power-delivery system. You can read more about this five-attributes approach on pages 6 and 7 of this brochure. To better demonstrate how each attribute will play a role in a more efficient, cost-effective, and environmentally responsible power grid, we have prioritized and completed a selection of value cases and demonstration activities to help support understanding and implementation. The distribution-related demonstrations and initiatives presently underway reside within the following focus areas:
  • Outage Awareness
  • Asset Optimization
  • Load and Distributed Energy Resource Awareness
  • System Performance Awareness
  • Emerging Practices and Technologies

Our member organizations all recognize that innovative uses of data and analytics are key enablers for the integrated grid of the future. The DMD initiative was by design very hands on with the member advisors, subject-matter experts, and real datasets from our member organizations. Overall outcomes were documented in a few key focus areas to include:

  1. Use Cases: That document valuable and innovative "data-driven applications" which leverage existing distribution sensors and other integrated data sources.
  2. Technical Insights: Which support members with shared learnings and leading practices as they develop plans and roadmaps for "data and analytics."
  3. Demonstrations: Which provide members with documented deployment and cost/benefit case histories where one or more members are leveraging sensors and data sources for more effective situational intelligence.
  4. Concept Documents and Surveys: To understand industry benchmarks with respect to philosophies and leading practices on emergent topics. For example, general preparedness for big data, leveraging the cloud to host big data, leveraging smart meters for better intelligence at the edge of the grid, and leading predictive and prescriptive use cases.
  5. The Data-Mining Initiative: Which has bridged a major industry barrier by enabling our research partners (vendors, universities, and other subject-matter experts) to leverage real power system and sensor data to solve clearly defined and valuable distribution system use cases.
  6. The Algorithm Sandbox: When we teamed the data science community with the power system experts, we found that the most impactful way to deliver data innovation is through documented and expandable data stories that include the raw code as well as the data scientist's insights regarding what worked best, what didn't work, and why certain approaches should be vetted and expanded with new contributions. More on this and how the concept is foundational for an analytics center of excellence is detailed on page 38.
Transformational is the one word I would use to describe the outcomes from the EPRI DMD and TMD Big Data initiatives, and I sincerely hope you enjoy the remainder of this document, which already has impacted and will continue to impact the electric power industry over the next decade.

Transmission Modernization Demonstration

EPRI's Transmission Modernization Demonstration (TMD) initiative started in 2013 to help electric power companies identify, prioritize, develop, and valuate advanced data-driven tools that enhance and optimize the way the transmission grid is operated, managed, and maintained. The first stages of the project focused on a) getting a solid understanding of the transmission landscape related to the needs and priorities of data-intensive applications and b) the capability of power companies to understand and use datasets. We surveyed the industry to identify areas of innovation and worked with our members to select high-priority areas for demonstrations. Ten demonstrations on data-driven technologies were then consolidated and developed. The scope of the projects included not only demonstration of existing applications but also the development of new tools, analytics algorithms, and data-management and data-integration techniques to address specific needs. The demonstrations include five focus areas:

  1. Situational Awareness
  2. Event Detection
  3. Data Integration
  4. Network Model Management
  5. Asset Management and Optimization

It All Starts With Data Awareness

The DMD and TMD initiatives began with a series of member immersions and workshops so that we could better understand the industry benchmarks in terms of data-driven aspirations and the challenges associated with turning that data into insights. As we began the dialogue with the earliest adopters, each had a unique perspective regarding sensors, data sources, and their journey toward analytics successes. Most had stories of disillusionment when they came to the realization that the ideal use cases that they endeavored to solve were much harder to operationalize than originally anticipated—primarily because the data was either siloed and difficult to access, or else the senor data was not available in the timeframe and in the structure needed for the use case. Others found that even when they began to warehouse the data, they learned after the fact that their data warehouses were generally designed for volume and storage, but that didn't necessarily translate to effective querying and mining within the new storage platforms. The vendor community was telling utilities that they had the data storage and analytics solutions to help them become the Googles of the power industry, but the reality was nowhere close to the claim. The key takeaways from the early workshops and member immersions were:

  1. Data Management – The industry needed to have a solid data-governance structure in place with ways to measure the quality and availability of the data.
  2. Data Integration – Tools and methodology to semantically integrate the data sources of the greatest value were necessary.
  3. Data to Value and Early Successes – The industry needed to gain consensus and define the most valuable and achievable use cases so that researchers and the vendor community could help. To support these takeaways, the EPRI team focused on creating an initial series of support documents that could describe the industry aspirations, the topical challenges, what the leaders were doing, and how early successes could be measured. The output of this early support focus included the following publications:

DATA MANAGEMENT

DATA INTEGRATION

DATA TO VALUE

DATA SCIENCE

Download the complete Data Analytics Initiative for Transmission and Distribution document.