Simulation of Power Systems with integrated time-dependent resources

1. Simulation of Power Systems with integrated time-dependent resources Yannick Degeilh and Pr. George Gross Friday, May 20th 2011 Power Affiliate Program

2. Outline • Context and Motivation • The challenges of modeling time dependent resources • Wind • Demand Response Resources (DRRs) • Utility-Scale Storage Device • Overview of the proposed simulation approach • Application study results

3. Context • There is a growing interest in integrating wind resources into the smart grid due to its benefits: • Renewable • emission free • 0 fuel cost • Meets energy independence goals • New policies currently being implemented to foster wind power development include: • tax credits and incentives • Federal stimulus package • Renewable Portfolio Standards (RPS)

4. Renewable Portfolio Standards www.dsireusa.org / April 2011 ME: 30% x 2000 New RE: 10% x 2017 VT: (1) RE meets any increase in retail sales x 2012; (2) 20% RE & CHP x 2017 WA: 15% x 2020* MN: 25% x 2025 (Xcel: 30% x 2020) MT: 15% x 2015 NH: 23.8% x 2025 MA: 22.1% x 2020 New RE: 15% x 2020(+1% annually thereafter) MI: 10% & 1,100 MW x 2015* ND: 10% x 2015 OR: 25% x 2025(large utilities)* 5% - 10% x 2025 (smaller utilities) SD: 10% x 2015 WI: Varies by utility; 10% x 2015 statewide RI: 16% x 2020 NY: 29% x 2015 CT: 23% x 2020 NV: 25% x 2025* IA: 105 MW OH: 25% x 2025† PA: ~18% x 2021† CO: 30% by 2020(IOUs) 10% by 2020 (co-ops & large munis)* IL: 25% x 2025 WV: 25% x 2025*† NJ: 22.5% x 2021 CA: 33% x 2020 KS: 20% x 2020 UT: 20% by 2025* VA: 15% x 2025* MD: 20% x 2022 MO: 15% x 2021 DE: 25% x 2026* AZ: 15% x 2025 DC OK: 15% x 2015 NC: 12.5% x 2021(IOUs) 10% x 2018 (co-ops & munis) DC: 20% x 2020 NM: 20% x 2020(IOUs) 10% x 2020 (co-ops) PR: 20% x 2035 TX: 5,880 MW x 2015 HI: 40% x 2030 29 states + DC and PR have an RPS (7 states have goals) Renewable portfolio standard Minimum solar or customer-sited requirement * Renewable portfolio goal Extra credit for solar or customer-sited renewables † Solar water heating eligible Includes non-renewable alternative resources

5. Complicating factors in the integration of wind Wind power salient features: • High variability/intermittency • Wind has a highly time-dependent nature • Uncertainty • Lack of predictability • Limited dispatchability

6. Wind time-varying nature

7. Wind power outputs are uncertain

8. DRRs Demand response resources (DRRs) = Market demand-side players who: • offer load curtailment servicesthatcompete with generators offers during peak-hours. • Potentially recover a fraction of the load that has not been consumed during peak hours in the low load hours: “Payback effect” DRR players effectively shift the demand from peak hours to low load hours => can reduce overall consumer payments, improve system reliability. Impact on emissions? Time - Dependent

9. Utility-Scale Storage MW Storage discharge during peak hours Modified load to be met by non-storage units Storage charge during low load hours day hours

10. V Wind/Storage Interactions energy discharged during peak hours MW unit i+3 higher cost unit i+2 daily load shape unit i+1 unit i unit i-1 daily load minus wind unit i-2 … Base-loaded units lower cost hours 0 12 24 energy charged during low load hours

11. Operational Planning conventional units load forecast Commitment of the conventional units for every subperiodh of period T wind power forecast unit commitment (UC) DRRs storage transmission resources

12. Hourly DAM realization load conventional units unit commitment wind power transmission-constrained DAM for hour h resource dispatch DRRs storage transmission resources • evaluation of : • economic indices • Reliability metrics • Environmental impacts

13. Nature and Scope of the study • The proposed research focuses on the assessment over longer term periods of wind, storage and demand response resources (DRRs) impacts on: • power system economics(LMPs, congestion rents…) • pollutant emissions (including greenhouse gas) • power system reliability (LOLP, EUE…) • Direct applications of the approach include: • resource planning • production costing • reliability analysis • investment risk analysis • policy analysis and study of their potential impacts

14. Main Contribution • A simulation approach that emulates the hourly transmission-constrained day-ahead markets(DAMs) for power systems integrating wind, demand response and storage resources • modeling of wind into daily wind patterns • modeling of DRR offers in the market and recovery effect • modeling of storage arbitrage via economic and cycle-efficiency considerations • The simulation approach is based upon the principles of a Monte Carlo simulation so that it can accommodate: • resource variability/intermittency • resource uncertainty. • Definition of representative simulation periods and implementation of Latin hypercube sampling (LCS) to ensure computational tractability used to approximate the market outcomeprobabilistic distribution for each snapshot

15. . . . . . . subperiod H subperiod 1 subperiod h Capture of both time-dependency and uncertainty • Capture of time-dependency… … … snapshotH snapshot1 snapshoth . . . . . . . . . period1 periodt period T time study period … … … Monte Carlo Simulation of eachperiod t • … and uncertainty.

16. Conceptual Approach

17. Insights in storage operation • In such conditions, how to emulate storage operations? • Premises: • Storage unit operation decided by the ISO. • Storage exploited in view of maximizing social welfare • Economic criterion: stored energy must be sold at higher price than that at which it has been bought. • Storage cycle efficiency consideration. Over a period of one week, what is charged must be discharged

18. Insights in storage operation The load is greater than 8000 MW 40% of the time “Time abstracted” domain Time domain

19. Discharged Energy MW p.u. 0 1 Insights in storage operation “Time Abstracted” domain Time domain MW charging energy unit i+1 unit i+1 unit i unit i … … base-loaded units base-loaded units hour

20. Insights in storage operation Maximum price at which energy is bought ($/MWh) Minimum price of displaced energy ($/MWh) Overall cycle efficiency

21. Insights in storage operation • Translating insights from the “time-abstracted” domain back into the chronological time domain • We have obtained the sets of units that can participate in charging the storage. Let be the marginal cost of the most expensive unit of such set. • We have obtained the sets of units that can be displaced by the storage. Let be the marginal cost of the cheapest unit of such set.

22. Insights in storage operation • In the hourly DAM: • if there is no congestion, scrupulously follow the nominal schedule yielded by the analysis in the “time- abstracted” domain. • if there is congestion, maximize the social welfare by both bidding and offering the storage unit: • Bid the storage as a load with willingness to pay • Offer the storage as a generator with willingness to sell • It can be shown that the optimal solution only allows the storage to be used either as a load or a generator

23. Application study results • IEEE 118-bus test system • Annual peak load of 8090.3 MW. • Reserve margin: 20% • Conventional generation: 9714 MW of installed capacity • 4 Midwestern wind farms for a total nameplate capacity of 2040 MW (25% of the peak load) • 300 MW utility-scale storage unitwith a 10,000 MWh reservoir. We illustrate the capabilities of the approach by examining the impacts of storage on power system economics in the following cases:

24. Hourly average storage utilization

25. Total Consumption Payments

27. Outcomes average values

28. Concluding remarks Findings • Wind resources may reduce electricity prices • Storage drives overall consumption payments down and improves system reliability • The impact of storage increases with deepening wind penetrations Future work • Improve the emulation of storage operation • Extend the simulation capabilities so as to take into account multiple storage units • Integrate a unit commitment to the simulation framework • Account for the intra-hourly variability of wind and its impacts on power system economics and reliability

29. Thank you! Any question?