Ramping and Demand Shifting: A Case Study Tim Mount Dyson School of Applied Economics and Management Cornell University Demand Response Workshop Cornell, January 17, 2011

Results from Previous Research Page 2 Rate Payments by a Wholesale Customer = Billing Cost + Wholesale Price x MWh + Capacity Price x MW Adding wind generation will result in: Wholesale Energy Prices going DOWN Capacity Prices going UP (MORE MISSING MONEY/MW) The financial viability of controllable demand and storage depends on getting paid correctly for providing services Buying low at night and selling high during the day (peak shifting) Paying lower capacity charges by reducing demand at the system peak Receiving payments for providing ramping services as well as regulation

Page 3 PART 1 Specifications for the Case Study

Criteria used for the Optimum Dispatch 2 OBJECTIVE FUNCTION FOR PLANNING Minimize the expected annual cost of operations over a set of credible contingencies, Including the costs of load-not-served, reserves and ramping (+incremental capital). subject to network constraints operating reliability system adequacy (financial adequacy)

Page 5 30-BUS TEST NETWORK Area 1 - Urban - High Load - High Cost - VOLL = $10,000/MWh Area 3 - Rural - Low Load - Low Cost - VOLL = $5,000/MWh Area 2 - Rural - Low Load - Low Cost - VOLL = $5,000/MWh Wind Farm

Specifications for a Windy Day System Conditions - Typical Demand Cycle - Wind is 25% of Demand - Three Big Cutouts Research Questions - How much potential wind is dispatched? - How much capacity is needed for reliability? Underlying Policy Question for Renewables - More transmission capacity v More efficiency using DER

Page 7 PART 2 Results of the Case Study

Effects of Including Ramping Costs Page 8 Case 2n: No Ramping CostsCase 2: With Ramping Costs Wind variability mitigated by GCT LESS wind dispatched Wind variability mitigated by Coal MORE wind dispatched RAMPING COSTS MATTER

Effects of Including Ramping Costs (Typical Day with 0MW/50MW of Wind Capacity) 9 NO Wind Capacity Case 1n: NO Ramping Costs Case 1: WITH Ramping Costs Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW MW Wind Capacity Case 2n: NO Ramping Costs Case 2: WITH Ramping Costs Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: %

Effects of Constant Wind (Typical Day with Ramping Costs) 10 50MW Wind Capacity Case 2: Normal Wind Case 4 Constant Wind Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Lower Operating Costs/ More Wind Dispatched Less Capacity Needed/ Cutouts Eliminated

Effects of Two Wind Sites (Typical Day with Ramping Costs) 11 50MW Wind Capacity Case 2: Normal Wind Case 7 Two Wind Sites Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Lower Operating Costs/ More Wind Dispatched - Not as low as constant wind Slightly More Capacity Needed

Effects of No Network Congestion (Typical Day with Ramping Costs) 12 50MW Wind Capacity Case 7: Two Wind Sites Case 3 No Congestion Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Lower Operating Costs/ Similar Wind Dispatched - Merit order dispatch BUT the cutouts are still there Similar Capacity Needed - The cutouts are still there

Effects of Demand Ramping (Typical Day with Ramping Costs) 13 50MW Wind Capacity Case 7: Two Wind Sites Case 8: Two Sites + DR Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Lower Operating Costs/ More Wind Dispatched - The gains are modest Less Capacity Needed - The cutouts are mitigated

Effects of Flat Demand + DR I (Typical Day with Ramping Costs) 14 50MW Wind Capacity Case 7: Two Wind Sites Case 9: Two + Flat + DR Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Lower Operating Costs/ More Wind Dispatched - The gains are substantial Much Less Capacity Needed - The cutouts are mitigated AND the peak load is reduced

Effects of Flat Demand + DR II (Typical Day with Ramping Costs) Page 15 Case 7: Two Wind SitesCase 9: Two + Flat + DR Lower Operating Costs/ More Wind Dispatched - The gains are substantial Much Less Capacity Needed - The cutouts are mitigated AND the peak load is reduced

No Congestion v Flat Demand + DR (Typical Day with Ramping Costs) 16 50MW Wind Capacity Case 3: No Congestion Case 9: Two + Flat + DR Percentage Change Operating Costs: $1000/day Conventional Capacity Committed: MW Potential Daily Wind Dispatched: % Similar Operating Costs/ More Wind Dispatched - Merit order dispatch v mitigated variability Much Less Capacity Needed - The cutouts are mitigated AND the peak load is reduced

Conclusions Ramping costs combined with the high probability of cutouts results in less wind dispatched, Eliminating network congestion does not eliminate the adverse effects of wind variability (more wind dispatched but the same capacity needed for reliability), The main benefit of using controllable demand to mitigate wind variability is to reduce the capacity needed, Using controllable demand (electric vehicles and thermal storage) to flatten the daily pattern of demand and mitigate wind variability is the big winner. More wind is dispatched and much less capacity is needed. Page 17