150 likes | 266 Vues
Designing a Smarter and Greener Electric Grid -- IT Challenges. Objectives and Goals. Use IT to make the electric grid smarter, greener and more efficient - leverage sensing, computing and communication
E N D
Designing a Smarter and Greener Electric Grid -- IT Challenges
Objectives and Goals Use IT to make the electric grid smarter, greener and more efficient - leverage sensing, computing and communication Match the needs of suppliers and consumers of electricity 1. Continuous measurement of various parameters through judicious location of sensors 2. Collecting data from widespread sources, disseminating information/insights to service providers, consumers, aggregators 3. Demand-side energy management 2
Monitoring and Measurement • PMUs, and other (expensive) sensor devices, for voltage/current phasors • Wide Area Power Control • Distributed Generation Control -- especially with Renewable Energy Sources • Voltage Stability Detection • Post-event Analysis • Operator Displays 3
PMU Placement Challenges • PMUs expensive: high cost to place everywhere • Place PMUs on subset of nodes • observe non-PMU nodes using basic laws of electricity • Placement problem: maximize observed nodes while minimizing PMUs used (a la vertex cover) • Impact of PMU placement on observability and cross-validation • PMU placement for max observability and error detection • Approximation algorithms • PMU errors: place PMUs “near” each other to detect errors • place PMUs for max observability and to detect errors 4
Monitoring and Measurement In-building sensors to track usage and consumption • Predict energy needs • Better energy management • Identify opportunities for energy savings Sensors subject to outages and failures • cross-validation of measurements Sensors create privacy issues University of Massachusetts Amherst & Indian Institute of Technology Bombay 5
Data Dissemination and Processing • Handling myriad data, sources, consumers • Fast-sampled PMU phasor (and IED) measurements, control and event log data, alarm data, data from digital fault records, sequence-of-events recorders, …. • Network operators, balancing authorities (BA), regional coordinators (RC), data archivers, monitoring agencies, energy markets, demand response aggregators. University of Massachusetts Amherst & Indian Institute of Technology Bombay 6
Sources Data aggregators Clients Data dissemination and Processing Approaches for Data dissemination/aggregation Query processing in an in-network setting Execution of continuous queries using a network of data aggregators, generalizing a pub-sub dissemination mechanism. • Phasor Data Concentrator (PDC) collects data from other PMUs & PDCs • Hierarchical organization • Application can run at a PDC having all the data required for that application University of Massachusetts Amherst & Indian Institute of Technology Bombay 7
Example queries • Monitoring Transmission Faults • Difference between time aligned measurements of voltage angles between two ends of a bus > 10 deg., send trigger • Monitoring Power Factor • Difference between angles of voltage and current is greater than some threshold, send trigger • Monitoring renewable energy sources • Difference between predicted value and actual aggregated value of power generated at renewable energy sources is greater than a threshold, send trigger • Complex analytics queries • (Vm< 96) || (Vm> 120) followed within 2 secs by (PJD event) • determine probability of a voltage collapse and identify vulnerable buses (sub-stations) University of Massachusetts Amherst & Indian Institute of Technology Bombay 8
Scheduling and Optimization Demand-side energy management for smart home/ office buildings • Utilizing device usage data from homes and buildings to study techniques for reduction/capping of usage • Electricity Load Forecasting for Office Buildings using Online resources -- room scheduler, timetable, event-calendar, weather data Design of techniques for decision making • when to rely on the grid and • when to use local sources • Reduce energy bills (given different billing models) • Minimize peak usage (in specified time windows) • Minimize (total) usage above a limit University of Massachusetts Amherst & Indian Institute of Technology Bombay 9
Scheduling and Optimization Leverage appliance and device scheduling and optimization --use of batteries as energy storage devices Adapt algorithms from (real-time) scheduling, e.g., EDF Design appliance-specific approaches, e.g., coolest-AC-last Feasibility analysis, given price/peak energy University of Massachusetts Amherst & Indian Institute of Technology Bombay 10
Motivation for focusing on TCBM Devices University of Massachusetts Amherst & Indian Institute of Technology Bombay 11
Summary • Goal: Better sensor-driven resource management, to ensure smarter, greener and more efficient grid • By: Integrating communications, control and computing into the grid core and edge while reliably and efficiently harnessing resources • Consumer participation and contribution • -- Use of data obtained from measurements to drive actuation/control, optimization or resource management • Special-purpose low-cost fixed sensors & mobile-based data gathering • -- Providing instrumentation for several homes and office buildings to drive research on peak usage capping and demand response • Towards: Green power, smart homes, net-zero buildings... University of Massachusetts Amherst & Indian Institute of Technology Bombay 12
Lecture Timings & Venue Tuesday & Friday 3:30pm to 5:00pm SIC 205, KR building Credit/Audit Requirements All students: will be expected to • 1. make a class presentation (developed in consultation with one of the professors) • 2. actively participate in class discussions. • 3. write critiques of 5 papers (chosen for presentations) and their presentations. Credit students: will, in addition, define and complete a research project on a topic of interest, in consultation with the course professors/TAs