Superior LMP Implementation with Minimal Changes to ERCOT Market

1
Superior LMP Implementation with Minimal Changes
to ERCOT Market

• Shams Siddiqi, PhD
• shams_at_crescentpower.net
• ERCOT Market Design Workshop
• January 14, 2003

2
What is ZEN (Zonal-ERCOT-Nodal)?

• ZEN is a unique implementation of LMP that
• Provides all features and benefits of other forms
  of LMP
• Resolves ERCOTs operational issues in managing
  Congestion
• Provides Nodal prices
• In addition, resolves issues Not Solved by other
  LMP
• Structurally enforces ex ante local market power
  mitigation
• Clearly distinguishes between high prices due to
  true scarcity (not mitigated) and high prices due
  to manipulation (mitigated)
• Allows for TCRs that are Options and mitigation
  against market manipulation arising from
  excessive TCR ownership
• Strong mitigation eliminates unnecessary risks
  leading to a Simpler Market - reduced barrier to
  entry for smaller players

3
What you dont get with ZEN

• What comes with other LMP, but not with ZEN
• Increased costs of managing risk
• Huge implementation costs for ERCOT and Market
• Estimated to be from 55 Million to 122 Million
• Difficulty in detecting and deterring energy
  price manipulation to impact FTR prices
• Currently, Mitigation methods that can suppress
  high prices due to true scarcity
• No protection against predatory behavior -
  generators have greater investment risk
• Uncertain impact on Retail Markets

4
How does ZEN work?

• The ZEN Method can be summarized as follows
• ERCOT and Market determine constraints that are
  competitive and those requiring mitigation
• Optimization algorithm with Competitive
  Constraints is run with absolutely no mitigation
  to determine ref. nodal prices used in creating
  mitigated bid curves
• Security constrained optimal dispatch with
  mitigated bids and all Constraints modeled is run
  to determine LMPs
• All nodal LMPs are published
• Resources are settled on nodal LMPs and Loads are
  settled on Zonal LMPs like today
• See Dr. Sam Zhous (MOD) paper or contact me

5
How is ZEN Superior?

• Locational Market Power Mitigation
• Clearly distinguishes between high prices due to
  scarcity and those that would result from the
  exercise of locational market power
• Does not allow ERCOT to Arbitrarily mitigate
  prices based on changes in bidding behavior
• Does not mitigate high prices due to true
  scarcity
• Ex ante structural locational market power
  mitigation with clearly defined Competitive
  (Unmitigated) Non-Competitive (Mitigated)
  Constraints

6
How is ZEN Superior?

• Transmission Congestion Rights (TCR)
• Point-to-Point and Flowgate Rights on Competitive
  Constraints are Options and Not Obligations -
  similar to current design
• Obligations expose entities to greater Risks
• Allows for Mitigation against Market Manipulation
  through limits on TCR Ownership
• Current 25 limit on TCR ownership highly
  effective in deterring manipulation and
  encouraging active congestion management by
  market participants
• Allows for easier TCR trading in Secondary Markets

7
How is ZEN Superior?

• More Beneficial for Resource Owners
• No need to engage in expensive Risk Management
  against unpredictable local constraints and
  outages
• Intrinsic Protection against Predatory Behavior
• More Beneficial for Consumers
• Aggressive and transparent mitigation against
  locational market power
• Mitigation against Market Manipulation arising
  from excessive TCR ownership
• Benefit for Both - Simpler Market Design

8
Why is Market Impact Minimal?

• ZEN is the theoretical basis of the current ERCOT
  market design
• Only changes required to current design are
• Unit specific bidding scheduling for Resources
• Nodal settlement for Resources
• Publishing all Nodal prices for transparency
• Slight changes in ERCOT solver algorithm
• No change in Risk Exposures TCR markets

9
Congestion Management Explained

• CASE 1 No Congestion (line rated at 1100 MW)
• G1000 MW _at_ 10/MWh 1100 MW
  L1000 MW
• Bid 20/MWh, Gd 1000 MW 1000 MW
  ----gt G? _at_ 50/MWh
• MCPEg 50/MWh SP
  0/MWh MCPEl 50/MWh
• CASE 2 Congestion (line derated to 900 MW, G
  owns 1100 MW of FTRs)
• G1000 MW _at_ 10/MWh 900 MW
  L1000 MW
• Bid 0/MWh, Gd 900 MW 900 MW
  ----gt G? _at_ 50/MWh
• MCPEg 0/MWh SP
  50/MWh MCPEl 50/MWh
• Congestion Cost -110050 90050 100(50-10)
  -6,000
• With no FTR ownership limit, G has incentive to
  create congestion and increase its FTR value
• True for other LMP - situation created with local
  line deration since PTP FTRs are not derated

10
Congestion Management Explained

• CASE 3A Congestion (line derated to 900 MW, G
  owns no FTRs)
• G1000 MW _at_ 10/MWh 900 MW
  L1000 MW
• Bid 10/MWh, Gd 900 MW 900 MW
  ----gt G? _at_ 50/MWh
• MCPEg 10/MWh SP
  40/MWh MCPEl 50/MWh
• Congestion Cost 90040 100(50-10) 40,000
• 1
• CASE 3B Congestion (line derated to 900 MW, G
  owns no FTRs)
• G1000 MW _at_ 10/MWh 900 MW
  L1000 MW
• Bid 50/MWh, Gd 900 MW 900 MW
  ----gt G? _at_ 50/MWh
• MCPEg 50/MWh SP
  0/MWh MCPEl 50/MWh
• Congestion Cost 100(50-10) 4,000
• G much better off strategically bidding
• Mistakes in bidding lead to inefficient dispatch
• True for MOD Proposal - Very Costly Risk
  Management - necessitates strategic bidding

11
Implications of Direct-Assigning Local Congestion

• For other forms of LMP
• Exposes market to manipulation, especially for
  non-competitive (Local) constraints
• Inability to impose FTR ownership limits exposes
  market to manipulation for all constraints
• Greatly increases Systems and Risk Management
  costs
• Due to incentives to strategically bid to
  maximize Profits, does not ensure efficient
  dispatch
• Bid to eliminate congestion if FTRs owned less
  than need
• Bid to increase congestion if FTRs owned exceed
  need

12
Implications of Direct-Assigning Local Congestion

• For MOD Proposal
• No tools (FTR) to hedge against unpredictable
  costs
• Forces Strategic bidding with dire costs for
  mistakes
• Must always bid close to Zonal MCPE but never
  exceed it
• Exceeding MCPE implies uneconomic dispatch down
• Bidding marginal cost implies huge costs (40k
  instead of 4k)
• Greatly increases Risk Management costs
• Encourages Collusion Dr. Orens Good Collusion
• Costs parties 40k if they compete and 4k if
  they collude
• Obviously, by not competing, then can save 36k

13
ZEN Design Philosophy

• Since Non-Competitive (Local) congestion
  encourages or forces Market Manipulation through
  Strategic bidding (locational market power),
  structurally mitigate any such possibility
• For marginally Competitive Constraints, allow
  Market to decide whether to compete and likely
  incur higher costs or not compete and accept
  structural mitigation
• For Competitive Constraints, offer TCRs as
  Options and allow full competition with no price
  mitigation however, limit TCR ownership to
  eliminate incentives to manipulate prices

14
Day Ahead Market Preferred Design

• Continuous, Bilateral (Pay-as-bid) market
• Independent 3rd party exchange, e.g. NYMEX
• A continuous market for hourly products for
  prompt hours and days, but also a forward market
  for months to years into the future
• Exchange costs recovered through commission on
  matched trades - Not Uplifted
• Greater value to market if exchange were to
  manage counterparty credit risk

15
Day Ahead Market Uniform Auction

• DA Uniform Price Auction market needs to address
  the following
• Uplift of cost running DA market (Max ISO)
• Protection against market power abuse since DA
  uniform price auction more susceptible to such
  abuse
• Centralized unit commitment is a problem
• arbitrary results and gaming opportunities
• Uplifted costs from 3-part bidding
• TCRs settled on DA prices may be a problem
• DA market can be manipulated to increase TCR
  value
• Exposes entities to risks from outages after DA
  market

16
Conclusion

• ZEN, in addition to the benefits of LMP
• Structurally enforces ex ante local market power
  mitigation
• Clearly distinguishes between high prices due to
  true scarcity (not mitigated) and high prices due
  to manipulation (mitigated)
• Allows for TCRs that are Options and mitigation
  against market manipulation arising from
  excessive TCR ownership
• Strong mitigation eliminates unnecessary risks
  leading to a Simpler Market - reduced barrier to
  entry for smaller player
• ZEN can be implemented in ERCOT at significantly
  lower cost and disruption to the Market than
  other forms of LMP

17
Conclusion

• Other methods of direct-assignment of Local
  Congestion (incl. MOD Proposal) provide incentive
  or compel parties to collude or strategically bid
• SMD implementation in West may resemble ZEN
• ZEN like Hybrid (LMP with FGR) favored by West
• NYMEX-style DA market would benefit Market
• Uniform Price DA market needs further
  investigation
• ZEN compatible with either type of DA market
• Benefits of ZEN need to be compared to Loss of
  Portfolio Bidding and Implementation Cost