Making the Most out of Energy Technologies at Value

1. Making the Most out of Energy Technologies at Value Prof. Marija Ilić milic@ece.cmu.edu Electric Energy Systems Group (EESG) http://www.eesg.ece.cmu.edu/, Director SRC Smart Grid Research Center (SGRC) http://www.src.org/program/eri/, Director New Electricity Transmission Software Solutions (NETSS), Inc. Founder IEE-EPC Talk November 8 2012

2. Outline • Designing and operating low-cost green electric energy systems –possible problem statement (Smart Grid and Ostrom sustainable SES) • Making the most out of energy technologies at value [8] -industry operations and planning: the way it was (underlying assumptions and major hidden inefficiencies) - estimates of possible efficiency enhancements (short- and long-term) -possible changes in operations and planning to enhance efficiency (interactive planning; IT-enabled corrective resource management; the need for next generation software)

3. Possible problem statement - Coarse modeling of Socio-Ecological Systems(using SES interaction variables) [6]

4. “Smart Grid”  electric power grid and IT for sustainable energy SES [2,5]

5. Future Electric Energy Systems [1] Energy Sources Transmission Network Electro-mechanical Devices (Generators) Load (Converts Electricity into different forms of work) Photo-voltaic Device Electro-mechanical Device Demand Response PHEVs Energy Sources

6. An illustrative future electric grid [2]

7. Transformational change in metrics of future energy systems [2]

8. Future: Multi-layered smart balancing authorities [3]

9. DYMONDS-enabled Physical Grid [2,5]

10. Making the most out of given technologies-New technical problem • At present the physical energy system, including its communications and control, does not readily enable choice and multi-participant information exchange and processing for aligning often conflicting goals. • It is essential to design intelligence for T&D operations to align these goals and consequently to make the most out of available resources while simultaneously offering robust and affordable quality of service. • New flexible energy processing equipment will also be needed to handle increased variety and bandwidth of many participant requests.

11. Building a stronger and smarter electric energy infrastructure • IT-enabled methods for radical enhancements in reliability and efficiency • Possible new operating framework in future electric energy systems (Dynamic Monitoring and Decision Systems (DYMONDS) [2,5] • Proof-of-concept using Azores Islands electric power systems [4]

12. Difficult Questions: Systematic IT Design for Complex Energy Systems • Establish sufficiently accurate (but not too complex) modeling framework which captures inter-dependencies of energy Socio-Ecological Systems (SES), physical grid, ICT and governance system • The key objective: Match attributes of energy SES, physical grid, ICT and governance system by designing around a given energy SES • Interaction variables: A means of going from very coarse to granular and back • IT design to manage interaction variables (temporal, spatial and contextual) • Interaction variables-based unifying framework for relating engineering design, financial and environmental objectives

13. The challenge of multi-temporal and multi-spatial integration [2] • Today’s practice--- all information given to the control center; very complex decision making given many constraints • Future paradigm—distributed decision making to internalize many externalities (diverse inter-temporal dependencies, uncertainties, congestion management) • Therefore… DYMONDS

14. Major challenges and opportunities [5] • Making ’’the most” out of available technologies • Under-estimated role of IT • Performance metrics for the changing industry • Managing resources at value: - the key role of corrective actions (vs. preventive (N-1) ‘’reliability”); -the key role of automation (distributed power-electronically switched ‘’smart” devices) • Design of model-based quantifiable compliance standards; must be easy to implement).

15. Efficiency metrics for systematic use by all? [5] • Multiple values brought by different technologies • Physical efficiency of stand-alone components (generation, demand) • Physical efficiency due to loss delivery savings • Physical efficiency due to reduced reserve requirements (reliability reserves-1) relaxed steady state limits; 2) relaxed stability limits. • Physical efficiency due to temporal shifts during normal operations (peak load shaving) • Economic efficiency in electricity markets (spot and capacity) • These are not additive. System-dependent. • Tradeoffs defined by the system users, not by the hard constraints. • Need systematic framework for modeling, and IT for implementing.

16. The key role of corrective actions [6,7] • reduction of spilled wind power (380GWh spilled in ERCOT in 2002); • economic efficiency increase measured in terms of annual generation cost savings (25% , estimated by FERC; $600M-$900M NY only); • increased power transfer into large load centers (300MW to 1GWW into NYC; 1 GW into SW CT. • increased power transfers across key transmission corridors. • Much higher equipment utilization; should study the effect on reducing the estimated $200B

17. The key role of PE-enabled automation:Power system with flywheel control [8]

18. The key role of ‘’smart” automation [8]

19. Potential stabilization by means of fast storage –nonlinear control [8]

20. Challenges to IT • Software for complex bids (distributed optimization under uncertainties) • Software for corrective actions during non-time critical contingencies • Software for ensuring dynamic stability and observability of the system • Primary multi-modal nonlinear control design for fast automation, with minimal coordination • Integration of these functionalities in a single platform • Proof-of-concept for small islands [6]; need to pursue for continental power grids

21. Three qualitatively different paradigms for standardization of dynamics in future smart grids [15,16] • Plug-and-play standards for dynamics, with no requirements for on-line communications. Much stricter standards at the component level will be needed for this to work. • System-level technical standards based on minimal coordination of interactions of decentralized component-level standards. • Interactive protocols for ensuring technical performance according to choice and at value—dynamic monitoring and decision systems (DYMONDS) [3].

22. Major differences • Plug-and-play standards for dynamics –enhanced decentralized control for internalizing effects of interactions and canceling them. Lots of advanced local control. • Standards based on minimal coordinated control of interaction variables for given nested architecture of future electric energy systems. Technical specifications at the decentralized level, economic and technical specifications at the system level. Minimal exchange of technical signals. • Interactive protocols in terms of interaction variables evolving dynamically over time and space according to system users’ preferences. Both economic and technical specifications at all levels. Minimal exchange of technical and economic signals. (markets)

23. Market Load serving entity End-user IT-enabled markets to harvest hidden efficiencies [18-20] Key role of aggregators to account for the effects of DERs in distributions systems on whole-sale markets Adaptive Load Management (ALM) Long-term contract LSE LSE

24. Potential Use of Real-Time Measurements for Data-Driven Control and Decision-Making (new) • GPS synchronized measurements (synchrophasors ; power measurements at the customer side. • The key role of off-line and on-line computing. Too complex to manage relevant interactions using models and software currently used for planning and operations. • Our proposed design: Dynamic Monitoring and Decision Systems (DYMONDS)

25. Proof-of-Concept for Low-Cost Green Flores and Sao Miguel [4] • Collected data and used to derive dynamic models (linear and non-linear; with wind power dynamics, flywheels and power-electronics-control included) -equilibrium solutions (power flow); predictive models for wind power and demand power -demonstrate the use of DYMONDS decision-making algorithms (distributed, MPC-based) for enabling efficient integration of wind power; efficient integration of Adaptive Load Management (ALM); efficient integration of electric vehicles (EVs) - demonstrate new methods for automated load following, E-AGC and E-AVC for balancing hard-to-predict small wind power fluctuations

26. Azores Island: Electrical Characteristics[3], [4, Ch. 3] Flores IslandRadial 15 kV distribution network Total demand : ~2 MWDiesel generator with total capacity: 2.5 MWHydro power generator with total capacity: 1.3 MW (reservoir)Wind turbine with total capacity: 0.6 MW • Sao Miguel IslandRing 60 kV and 30 kV distribution network Total demand - ~70 MWTwo large diesel generators with total capacity: 97 MWTwo large geothermalplants with total capacity: 27 MW 7 small hydro power generator with total capacity: 5 MW • M. Honarvar Nazari and M. Ilić, “Electrical Networks of Azores Archipelago”, Chapter 3, Engineering IT-Enabled Electricity Services, Springer 2012.

27. Overall educational challenge and opportunity [8] • Important to educate ourselves how more efficient services can be provided w/o creating operating problems • The opportunity is presented itself as IT is deployed • The challenge: How to manage the infrastructure instead of simply deploying more hardware • Need for new software for interactive planning, corrective on-line resource management and automation; non-transmission solutions • States should take the lead in educating themselves and the industry (NARUC); design regulation to support this [10]

28. Educational Challenge to the US Universities [9] • An important educational challenge: How to pose the problem, and how to design sensing, communications, automated control and decision-making computer algorithms using well-understood concepts from basic disciplines? • The boundaries between electric energy processing and other types of energy processing (mechanical to electrical in generators; chemical/wind/hydro, diesel, nuclear into mechanical and/or electrical) becoming more gray than in the past as new energy resources are used • One possible unifying path– model the electric energy systems as dynamical systems and use systematic control design to pose the design objectives, and data-driven feedback and decision making for adaptation (18-618, Spring 2012, Carnegie Mellon University [9])

29. Conclusions and Next Steps • IT has key role to play in operating future electric energy systems with renewable resources. • Systematic modeling of electric energy systems as complex dynamic systems needed. • Not possible to manage all as a single problem. • The challenge is what IT to embed into hardware, what information needs to be exchanged and why. • Possible to be both reliable and efficient green with carefully designed IT. Many examples of this not being possible with today’s IT in electric power systems. • Need to communicate to NARUC/US Congress • Demonstrate using simulations for continental systems (``Smart Grid in a Room”?)

30. References [1] Ilic, M., Smart Grid and Future Electric Energy Systems, Lecture Notes, 18-618, Carnegie Mellon Univ, ECE, Spring 2012. [2] ] Ilic, M, et al, A Decision Making Framework and Simulator for Sustainable Electric Energy Systems, The IEEE Trans. On Sustainable Energy, TSTE-00011-2010, January 2011. [3] ] Ilic, M., Dynamic Monitoring and Decision Systems for Sustainable Electric Energy, Proc of the IEEE, Jan 2011. [4] Engineering IT-Enabled Sustainable Electricity Services: The Case of Low-Cost Azores Islands (co-editors, M. Ilic and Le Xie), Springer Monograph (2012, to appear) [5] Ilic, M., Dynamic Monitoring and Decision Systems for Sustainable Electric Energy Systems, Proc. of the IEEE, January 2011. [6] ElinorOstrom, et al, A General Framework for Analyzing Sustainability of social-Ecological Systems, Science 325, 419 (2009). [7] MIT Portugal, Universidade dos Açores, ”Characterization of the Azorean Residential Building Stock”, Report, 2010. [8] Ilic, M. , ``Making the Most out of Energy Technologies at Value”, Ilic, White paper, April 17, 2012; under revision with FERC comments, November 2012. [9] Ilic, M., ``Critical Needs for Multi-Disciplinary Approach to Teaching Electric Energy Systems”, IEEE Transactions on Education (Special Issue, invited), November 2012. [10] Ilic, Marija, “3Rs for Power and Demand”, Public Utilities Fortnightly Magazine, December 2009.

31. References [11 ] Ilic, Marija. “From Hierarchical to Open Access Electric Power Systems.” IEEE Special Issue on “Modeling, Identification, and Control of Large-Scale Dynamical Systems,” Simon Haykin and Eric Mouines, Guest Editors. Vol. 95, No. 5, May 2007. [12] M. Ilic. Toward reliable and efficient on-line resource management: A ramp rate-limited ac optimal power flow for integrating renewable resources and responsive demand. FERC Staff Technical Conference on Increasing Real-Time and Day-Ahead Market Efficiency through Improved Software, Docket No. AD10-12-003, Washington DC, June 25-27, 2012. [13] M. Ilic. Voltage dispatch and pricing in support of efficient real power dispatch. Final Report, NYSERDA Funded NETSS Project #10476, November 20, 2011. [14] M.Ilic, ‘’Toward Standards for Dynamics in Future Electric Energy Systems”, PSERC White Paper, June 2012. [15] M. Ilic,’’ Toward IT-enabled power systems: Large-scale distributed control for tomorrow's electricity grid”,Semi-plenary talk, American Control Conference, Montreal, CA, June 2012. [16] Ilic, M., Liu, Q., ``Toward standards for dynamics in future electric energy systems”, APSIPA, Dec. 2012. [17] ] Computing Research for Sustainability, NRC of the National Academies of Science, http://www.nap.edu/catalog.php?record_id=13415 , 2012. [18] J.-Y. Joo and M.D. Ilić, A multi-layered adaptive load management system: information exchange between market participants for efficient and reliable energy use, IEEE PES Transmission and Distribution Conference, Apr 2010 [19] J.-Y. Joo and M. Ilić, “Distributed Multi-Temporal Risk Management Approach To Designing Dynamic Pricing”, IEEE Power and Energy Society General Meeting, July 2012 [20] J.-Y. Joo and M. Ilić, Multi-Temporal Risk Minimization Of Adaptive Load Management In Electricity Spot Markets, IEEE PES Innovative Smart Grid Technologies, Europe, Dec 2011 [18]