Grid Optimization Through InGrid Data Analytics Mischa Steiner

Grid Optimization Through In-Grid Data Analytics Mischa Steiner CEO Awesense India-Canada Smart Grid Workshop

A Growing Global Issue +50% Growth in Energy Demand 1200+ Coal Fire Plants Proposed $202 B USD/yr is Lost Mostly in India and China 13 B tonnes of C 02 emissions Rising 2. 5% per year

The Challenge The Invisible Grid ✔ ✔ • Generation & transmission are well-monitored. • Consumers also well monitored (especially with smart meters). ✔ ✔ ? • MV distribution lines are poorly monitored. • Pure analytics solutions relying only on smart meters are blind to this part of the grid. Impossible to permanently monitor every part of the grid • The only realistic way to know what is happening on the grid is to intelligently sample. ? ? ? ? Estimated 70 million km of power lines worldwide.

Reduce Avoidable Losses Technical Losses Non-Technical Losses Billing errors Wiring errors MV Theft & LV Theft Transformer copper loss – phase imbalance Transformer copper loss - overload Transformer core loss Line Losses Capital costs ARE needed to reduce these losses Capital costs ARE NOT needed to reduce these losses Total Losses Metering errors

Reduce Avoidable Losses TGI Analytics + Io. T Metering errors Wiring errors MV Theft & LV Theft Transformer copper loss – phase imbalance Transformer copper loss - overload Transformer core loss Transformer copper loss - overload Line Losses Capital costs ARE needed to reduce these losses Capital costs ARE NOT needed to reduce these losses Total Losses Savings Billing errors

Maturity Model Capital Optimization Process Optimization Risk-based case management Meter Data Analytics Grid Segmentation Risk Modeling Local Knowledge Smart Meters AMR Data Integration Prescriptive Analytics Predictive Analytics Grid Segment Monitoring In-Grid Sensors Machine Learning

Our Unique Approach True Grid Intelligence (TGI) • Risk-based modeling for grid optimization • “In-grid data” enhances models • Grid Segmentation & Energy Balancing • Predictive, prescriptive grid data analytics • • • MV- and LV-capable current sensors Io. T: cellular and mesh backhaul Cloud software as a service Temporary/permanent in-grid data handling Wide area in-grid monitoring Loss and efficiency reporting Conventional Approach • Continuous monitoring of the entire grid edge and a few points in the grid • Extrapolate data from the edge of the grid and hope to be able to find problem areas Awesense Approach • Choose the grid segments that require attention • Get detailed data from within high risk areas and actually solve the root cause of the problems

5 Steps to the Optimized Grid 1 Grid Segmentation and Risk Modeling to determine high-risk grid segments 2 Case Selection and Management: plan the optimal investigation of target segments Machine Learning, 5 Feedback, KPIs: Assess priorities, recommend Next Best Actions Io. T Sensors to Collect In-Grid Data: TGI determines WHERE to put them. 3 Reporting and Analytics Energy Balances, Phase Balances, Transformer Load Studies, etc. 4

“THANK YOU” Questions? Please visit www. awesense. com General questions: tweet @awesense Questions for Mischa: tweet @mischa_steiner