1. Transmission Bottlenecks Joe Eto Lawrence Berkeley National Laboratory January 27-29, 2004 Washington, D.C. Transmission Reliability Peer Review

2. “DOE believes that identifying and eliminating major transmission bottlenecks is vital to our national interest. National-interest transmission bottlenecks create congestion that significantly decreases reliability, restricts competition, enhances opportunities for suppliers to exploit their market power unfairly, increase prices to consumers, and increases infrastructure vulnerabilities.” http://www.ntgs.doe.gov DOE National Transmission Grid Study

3. Project Objectives/Accomplishments Conduct scoping and planning studies to support DOE implementation of NTGS recommendations on national-interest transmission bottlenecks CERTS has prepared 4 reports (listed below), 1 memo, and is studying MISO:  Survey of current transmission bottlenecks, as reported by ISOs – J. Dyer, EPG  Review of commercially available transmission bottleneck analysis techniques/models – P. Sigari, KEMA  Assessment of tools under development by national labs that might be available to support bottleneck assessment – S. Thomas, et. al, Sandia  Review of recent reports of congestion costs – B. Lesieutre/J. Eto, LBNL

4. ISO Survey of Transmission Bottlenecks

5. PriorityISOComment 1NYISOCongestion costs over a three year period are averaging in excess of $900 million per year. 2ISO-NELoad is at risk now 3CAISOCalifornia has two significant load pockets that are forecasted to be in violation of reliability criteria and a path that has inhibited transactions between the northern and southern portions of the state. 4PJMPJM’s congestion costs continue a four year trend of almost doubling each year, but the majority of 2002 increase is a result of adding PJM West to its market. 5MISOAt this time, the true congestion costs are unknown. The region will have difficulty operating an efficient market with the limited EVH infrastructure. 6ERCOTERCOT will need to expand its transfer capability to accommodate new generation and achieve market efficiency.

6.  Review of Available Analysis Tools Electricity Price Forecast Run Fuel Price Forecast Run load Forecast The electricity market simulation model selects resources available to meet the anticipated demand plus necessary ancillary services and determines the forward or ex ante Market Clearing Prices (MCP). Optimal Power Flow (real time dispatch) Simulates the actual system dispatch and determines the real and reactive power flow in each hour. Any energy imbalances, voltage quality or congestion problems are mitigated by the OPF re-dispatch algorithm. Data Maintenance Tools 1Network Reduction 2Database Maintenance 3One line Diagrams Data Model including network data, generation data, fuel data, contracts data, etc. Results: 1Identified/Verified Congestion 2Price Forecast at Zones or Nodes without a transmission line at congested path. 3Price Forecast at Zones or Nodes with a transmission line at congested path. Monte Carlo Simulation of un- certainties

7. Review of Available Analysis Tools Product National Regional State Electrical NetworkMarket Simulation TRACEN, R, SYN GridViewN, R, SYY AURORARNY TRACEN, R, SYN CAR --- MAPSN, R, SYN PROSYMN, R, SY (w/PowerWorld) Y UPLANN, R, SYY SCOPEN, R, SYN PSS/E, MUSTN, R, SYN PROMOD VIN, R, SYY

8. Survey of National Laboratory Models  Electricity Market Complex Adaptive System – Argonne  Generation and Transmission MAXimazation – Argonne  Transmission Entities with Learning Capabilities and Online Self-Healing – Argonne  Power Market Simulator – LANL/NISAC  Power System Analyzer – LANL/NISAC  Positive Sequence Load Flow and Positive Sequence Dynamic Simulation by GE; and PSS/E from PTI – PNNL  BUZZARD – Sandia

9. Review of Congestion Costs

10. Information about the operation of congestion revenue rights markets is needed to assess the impacts of congestion revenue charges on consumers. Information on generators’ offers is needed to assess system redispatch payments. Many studies presume that generator offers reflect competitive market conditions. Customer costs may rise as a result of reducing congestion. Minimizing consumer costs may not increase aggregate social wealth. There is no standardized conceptual framework for studies of congestion costs

11. MISO LMP/Congestion Modeling  Midwest ISO (MISO) plans to use locational prices to value wholesale power and manage congestion (market scheduled to open 12/04)  FERC (and market participants) want to know what to expect –Changes in power flows/line loadings –Changes in prices –Costs of congestion –Existence of bottlenecks –Opportunities for exercise of market power  Project Team B. Lesieutre, E. Bartholomew, J. Eto, LBNL T. Luong, FERC D. Hale, Tom Lecky, EIA

12. MISO LMP/Congestion Modeling Snapshot at 1600h July 7, 2003: 37 Control Areas (106 modeled) 1,280 Generators (3,680 modeled) 132,027 MW of capacity (506,969 MW modeled) 12,118 Buses (27,798 modeled) 14,656 Lines and Transformers (34,981 modeled)

13. MISO Preliminary Findings % Change in Control Area Generation Significant reduction in operating cost (~ - 20%) Winners and losers –Some regions have large changes in generation/prices –Some have high prices Intra-control region congestion dominates interregional congestion Important data issues still need to be resolved – expect to work closely with MISO staff

14. Next Steps  Support OETD planning and implementation of NTGS recommendations – public process on criteria for and federal role in addressing bottlenecks  Complete MISO market pre-assessment –improve data/verify findings with MISO; –identify potential bottlenecks/opportunities for exercise of market power  Work with EIA to improve quality of transmission data  Plan workshop on advanced modeling/simulation needs  Support DOE efforts to assist regional planning entities

15. Congestion Cost Backup Slides

16. Congestion Costs – Uplift Charges Congestion costs = dispatch payments out of merit order Congestion costs are equal to the increased dispatch payments by the market to generators out of merit order. The dispatch payments are calculated using a uniform market clearing price for most generation. However, generators dispatched out of merit order because of congestion are paid at their offer prices. The uplift charge is shared equally among the consumers. Area A 0 100 200 300 400 500 600 700 800 900 1000 1000 900 800 700 600 500 400 300 200 100 0 Area B Supply 20 22 24 26 28 30 32 $/MWh 20 22 24 26 28 30 32 $/MWhArea A LoadArea B Load Generation dispatched “out of merit order” Congestion uplift costs

17. Congestion Costs – System Redispatch Payments Congestion costs = change in dispatch costs Congestion costs are equal to the difference in dispatch payments by the market to generators in the congested case relative to costs for the uncongested case. The dispatch payments are calculated using LMPs. Area A 0 100 200 300 400 500 600 700 800 900 1000 1000 900 800 700 600 500 400 300 200 100 0 Area B Supply 20 22 24 26 28 30 32 $/MWh 20 22 24 26 28 30 32 $/MWhArea A LoadArea B Load Net change in total dispatch costs paid to generators

18. Congestion Costs – Congestion Revenues Congestion costs = congestion charges In a market that uses LMPs, congestion revenues are the valuation of transmission of energy across a congested interface. Neglecting losses, these revenues equal the product of the energy flow and the price. Congestion revenues are also equal to the difference between what consumers pay for energy and what generators are paid for supply Area A 0 100 200 300 400 500 600 700 800 900 1000 1000 900 800 700 600 500 400 300 200 100 0 Area B Supply 20 22 24 26 28 30 32 $/MWh 20 22 24 26 28 30 32 $/MWhArea A LoadArea B Load Congestion charges