Control and Coordination in a Transactive Energy Environment

1. Control and Coordination in a Transactive Energy Environment Jeffrey D. Taft, PhD Chief Architect for Electric Grid Transformation Jakob stoustrup, PhD Chief scientist / advanced controls program manager Pacific Northwest National Laboratory 28 March 2014

2. Our Grids Are Changing We are in the process of violating these principles! • Generation is dispatchable • No significant energy storage in the grid • Power must be kept in balance • Generation follows load • Distribution “floats” on transmission • Designed/operated for reliability, not economy Now we want to operate pervasively for joint economic/control optimization in this newer highly complex environment with potentially millions of interactive endpoints.

3. Emerging Trends for US Utilities • Integration of renewable sources at T and D levels • Reduction of grid rotational inertia • Interactive loads/prosumers/expanded markets • Time scale shift (to faster grid behaviors) • T&D level power electronics penetration • DG/DR penetration and local energy networks (microgrids) 30%

4. New(er) Grid Functions • Electric Vehicle (EV) charge management • Inverter control for fast VAr regulation • Local energy network and microgrid power balance, load sharing and flow control • Multi-tier virtual power plants • Energy/power market interactions for prosumers; Transactive Energy • Electronic grid stabilization (FACTS for transmission; DSTATCOM for distribution) • Load modulation of buildings, electric vehicle chargers, and data centers for local balancing • VER integration (wind, solar, etc.) • Wide area measurement, protection, and closed loop control • DER/DG integration (distribution level) • Energy storage integration • Responsive loads (command, price, and /or system frequency) • Integrated Volt/VAr control (LTC/cap) • •Advanced distribution fault isolation/service restoration • • Third party energy services integration • Grid stabilization (H reduction) No single use case predominates; the control approach must support ensembles of new functions; utilities are being driven to select their unique function sets.

5. Issue: Grid Coupling/Feedback • Electrical physics rules the grid – shaped by grid connectivity • Business models and software cannot change this • Must be taken into account in control design to avoid unintended consequences • IVVR/DR* • Becomes important as new rollouts of smart devices scale to full deployment • Implications for architecture, design, and control • Market/responsive loads • CVR/PV *Jose Medina, Nelson Muller, and IlyaRoytelman, Demand Response and Distribution Grid Operations: Opportunities and Challenges, IEEE Trans. On Smart Grid, vol. 1, pp. 193-198, Sept. 2010.

6. Issue: The Coordination Problem • Power grids do not have a strong multi-tier coordination framework • Power grids are inherently tiered now • QSE’s-> Bulk System-> DSO’s->End Users • Distribution “floats” on transmission • DER as a “threat” and an opportunity • A proper coordination mechanism is key to enabling new business models • Interop standards alone do not solve this issue

7. Systemic Issues/Requirements In fact, there are about 80 “-ity” type characteristics that everyone quotes* • The most obvious: • Reliability/scalability • Stability • Security • Less obvious: • Federation/constraint fusion • Disaggregation • Boundary deference • Local “selfish” optimization (organizational autonomy) • Communications compatibility scalar control *John Doyle and John G Brown, Caltech, Universal Laws and Architecture. Available online as 1_DoyleSageLec1_May7_2012.pdf

8. Wide Area Scalar Control Approaches • Solve the communications and disaggregation issues simultaneously and scalably (maybe!) • Send scalar signals to endpoints • Global common signal broadcast • Location-dependent values • Various Approaches • Distribution-Locational Marginal Pricing • Transactive Energy • “Prices to Devices”

9. Embedded Markets and Prices to Devices Scalar signal disaggregation models Scalability Directionally good but still lacking coordination capability, so needs a better foundation/framework Embedded market and prices-to-devices models Instability Roozbehani, M., et al, Volatility of Power Grids under Real-Time Pricing, MIT, 2011, available online

10. Transactive Energy Definition from GridWise Architecture Council “The term “transactive energy” is used here to refer to techniques for managing the generation, consumption or flow of electric power within an electric power system through the use of economic or market based constructs while considering grid reliability constraints. The term “transactive” comes from considering that decisions are made based on a value. These decisions may be analogous to or literally economic transactions.” Key elements: • Decisions (and control) based on value • Reliability constraints

11. Network Utility Maximization: Layering for Optimization Decomposition • Combine ideas from Control Engineering and Networking • Multi-tier control coordination • Benefits from layered architectural paradigm LAMINAR COORDINATION Mung Chiang, Steven Low, et. al., Layering as Optimization Decomposition: A Mathematical Theory of Network Architectures, Proceedings of the IEEE, Vol. 95, No. 1, January 2007.

12. Coordination: Essential Structure • Matches the “clean” layered model for the grid hierarchy • Provides scalable coordination while respecting organizational and system boundaries • Consistent with solid architectural principles • Laminar coordinators* can incorporate TE nodes • Replaces ad hoc TE node approach with uniform control architecture for joint economic-control optimization *J. Taft and P. De Martini, Scalability, Resilience, and Complexity Management in Laminar Control of Ultra-Large Scale Systems, available online.

13. Transactive Energy and Laminar Framework • By adding the Laminar Coordination concept to TE, can we address the issues listed earlier? • Decomposition allows mapping to real systems • Formulations provide constraint fusion and local goals

14. Make It a Platform and They Will Innovate • We cannot predict all the new business models and applications that may emerge: • Local energy balance via B2G • Third party ancillary services • Feeder stabilization/regulation services via smart inverters • Utility-owned, customer-sited DER • Customer-owned, utility-managed DER • Multi-microgrid management • DSO as DER coordinator • But, we can position the grid as platform and the DSO as an energy business hub if the control architecture is done properly

15. thank you Jeffrey D. Taft jeffrey.taft@pnnl.gov

16. resource slides

17. Issues with Prices to Devices • Hidden feedback, flash crashes, lack of tools to ensure stability leads to price and grid instability • Multiple prices for the same kW-hr to a single end used from separate parties/processes; price-overwriting at the endpoint • Differing prices to different endpoints for what is an apparently equivalent kW-hr • Time scale mismatch between necessary control actions and market functions • Presumed over-disaggregation (cannot really have individual components of a system, say in a factory, responding differently to prices, so cannot disaggregate to individual devices)

18. Embedded Markets/Prices to Devices Embedded market approaches may act as a control elements in a feedback control loops, whether intended or not. • Hidden feedback, flash crashes, lack of tools to ensure stability -> price and grid instability • Multiple prices for the same kW-hr to a single end used from separate parties/processes; price-overwriting at the endpoint • Differing prices to different endpoints for what is apparently equivalent kW-hr (due to D-LMP or non-system, i.e. very local, markets) • Time scale mismatch between necessary control actions and market functions • Presumed over-disaggregation (cannot really have individual components of a system, say in a factory, responding differently to prices, so cannot disaggregate to this level)

19. Value Representation in TE • Value might be represented in a traditional currency form (dollars and cents) • Advantage: easily understandable • More generally, signals do not have to be currency to be economic or value-based • Not all values are necessarily converted to $ • Economic signals do not have to be currency – behavior and math are determining factors • Dual decomposition uses abstract price signals for coordination, for example* • Want to enable value unlocking in general, regardless of business model or market structure * Alternative Distributed Algorithms for Network Utility Maximization: Framework and Applications , Daniel P. Palomar and Mung Chiang, IEEE Trans. Automatic Control, Vol. 52, No. 12, December 2007.

20. Issue: Emerging Lines of Grid Control Emerging Architectural Chaos!