Nodal LMP Markets: Modeling, Data and Application Software Challenges

1
Nodal LMP Markets Modeling, Data and Application
Software Challenges

• ERCOT TNT Presentation October 1, 2003

2
Presentation Overview

• Nodal LMP Markets High Level Overview
• Modeling and Data Maintenance
• Network Model Building and Maintenance Services
• Application Software Challenges

3
Nodal LMP Markets High Level Overview

• An integrated framework to facilitate forward
  decentralized bilateral markets, spot day-ahead
  and real-time markets for transmission services,
  energy, and ancillary services
• A centrally dispatched energy and ancillary
  services spot market based on locational marginal
  pricing (LMP), with standard financial hedging
  instruments (CRRs)
• The co-optimization of energy and ancillary
  services with cascading of all prices and
  quantities from higher quality products to lower
  quality products

4
Nodal LMP Markets High Level Overview

• A multi settlement based on binding day ahead
  pricing and real-time pricing for deviations from
  the day ahead/bilateral schedules
• A real time balancing market
• A marginal loss model explicitly in the dispatch
  engine
• An explicit modeling of security (contingency)
  constraints
• CRRs as both options and obligations
• Handling of ancillary services as lost
  opportunity costs and availability bids
• Standardized pricing, especially for unbalanced
  schedules and deviations from quantities cleared
  in the day-ahead spot market

5
Nodal LMP Markets High Level Overview
Transmission Reservation, Transmission
Entry, Bidding (Generation Load), Data Entry
and Report Viewing
Market Participants

• Portal Technology
• Market User Interface
• (One Stop Shopping)

Regional Market Planning, Security,
Monitoring, Optimization, Settlement and Bulk
Functions
ISO/RTO

• Communication Technology
• ICCP (real-time info)
• IP ( transaction entry)
• XML data transfer (bulk information)
  

Local Planning, Security, Control Monitoring and
Dispatch Functions

• EMS Technology
• Existing Services

Tie Line Flows, Generation Capabilities and
Outputs, EHV Facilities Status and Flows
Particular to Each Area Dispatch Instruction to
Participants
6
Nodal LMP Markets High Level Overview
MARKET PARTICIPANTS Generation/Load Entities
Energy Offers
Energy Offers
Quantities Prices
Payments Bills
Day Ahead Schedules
Hour Ahead Adjustments
Real Time Dispatches
Week Ahead Projections
Billing and Settlement
Real Time Values Schedules and Forecast Operating
Constraints
Schedules and Contracts Set Points and
Prices Net Tie Schedules
ENERGY MANAGEMENT SYSTEM Control Area Operations
Outage Scheduling
Load Management
Generation Control
Security Analysis
State Estimation
7
Nodal LMP markets Main functional components

• Security constrained market clearing configurable
  to support day-ahead SCUC and real-time SCED
• Network Analysis application (State Estimator -
  SE, Power Flow - PF, Security Analysis - SA)
• LMP engine configurable to support different
  pricing mechanisms (ex-ante, ex-post, pay as bid,
  ...)
• Congestion Revenue Rights (CRRs) auction and
  simultaneous feasibility study
• Communications (ICCP, Enterprise Integration Bus)
• Support systems and CIM compliant model and data
  maintenance

8
Day-Ahead Market (24-48 h ahead, 1h time interval)
Market User Interface
Awarded CRRs Energy and AS Schedules, Prices
Bids
CRR Feasibility
Market Database
Billing Settlement
Load Forecast
Pricing Engine
Study NA (TP, PF, SA, ATC)
SCUC/SCED
ISR
Outage Scheduling
Tagging Scheduling
OASIS
Schedules, Prices
CIM NERC DEWG Format
XML Over EAI
XML Over EAI
CIMInformation Model Manger
Control Area EMS
9
Real-Time Market (15 ahead, 5 time interval)
Market User Interface
Schedules, Prices
Bids
Billing Settlement
Market Database
Very Short Term Load Forecast
Alarming Trending
Pricing Engine
Real-Time NA (SE, TP, PF, SA)
ISR
SCED
ICCP
Tagging Scheduling
Block 1,2
Outage Scheduling
CIMInformation Model Manger
CIM NERC DEWG Format
Schedules, Prices
Control Area EMS
XML Over EAI
XML Over EAI
10
Potential scope of the FERC SMD software
standardization effort

• Common data model (static extensions of CIM)
• Market participant interfaces to support
  electronic upload/download of data to/from RTO
  OASIS
• Identification and standardization of functional
  blocks in support of intra-RTO business processes
• Standardization of data exchanges between the
  standardized functional blocks (dynamic
  extensions of CIM)
• Creation of benchmark test data for SCUC, SE,
  LMPs and CRR simultaneous feasibility
• Standardization of data and related interfaces
  required to support Market Monitoring Unit (MMU)
  operation
• Identification and standardization of data
  exchanges between RTOs

11
Nodal LMP Markets Key Success Factors

• System Modeling, Data Collection and Maintenance
• Knowledgeable Support personnel
• Robust Application Software
• Technology Platform
• Standardization
• Transparency
• Advanced Training of Market Participants and
  System Operators
• Continuous Improvements of RTO Systems and
  Processes

12
System Modeling, Data Collection and Maintenance

• Common Information Model (CIM)
• CIM Market Extensions (CME)
• Standardized interfaces for XML based model and
  data exchange (CIM)
• Common source for maintenance of data and
  displays (auto generation of displays)
• Transparent Web tools for transmission owners and
  market participants to actively participate with
  ERCOT in on-going data collection, validation and
  maintenance
• Regular frequent (at least daily) incremental
  on-line updates of the network model
• Single source of all static engineering data
• Flexible for customer specific model extensions
  without source code changes

13
Centralized CIM compliant model and data
maintenance

• Interactive graphic designer
• Topology derivation
• Automatic display construction
• Interactive Type- and Instance Editors

Exchange between different systems of
multiple suppliers
Bulk import/export
RemoteControl sites
Domain Object Repository
CIM-based data model
System processobjects
XML
Excel
Applications
Data Migration
SystemConfiguration
Archive
Field resources
14
Network Model Building

• Timely updates to the network model become
  essential to market calculations
• LMPs require accurate models that produce
  reliable, quality solutions
• Model accuracy in support of reliability analysis
  was less critical than accuracy requirements for
  market calculations
• Creating or updating a network model can distract
  transmission operation organizations from
  day-to-day operational priorities

15
Whos responsible?

• Conflicting interests
• TOs reliability of my system
• Most knowledgeable of the data
• More detailed modeling needs (lower voltage
  levels, substation configurations)
• RTOs reliability of overall system and market
  needs
• Dependent on help from TOs
• More data volume but maybe less detail required

16
Network Model Building Service

• Internet technology and industry standards are
  reducing the barriers to timely updates to
  network models
• It is possible to provide a network model
  building service
• Internet
• CIM
• Two tools are essential to support this service
• e-utilities Team Room
• Model Review Web site

17
e-utilities Team Room

• Team Room provides the mechanism for
  collaborative work
• File sharing
• Discussion forums
• Chat functionality

18
Model Review Web Site

• Review of results is performed via the Model
  Review Web site
• Results of model validation are posted for review
• Provide common access for review of problematic
  cases

19
Overall Network Model Building Process
Web Site
Proprietary
CIM
CIM
De-facto Standards (e.g. PTI PSS/E)
Model Validation Results
Model Merge
20
Four aspects to model support

• Model formats (e.g. CIM, PSS/E, proprietary)
• Use of CIM is a viable means for exchanging
  models from other systems
• CIM does not solve all the problems
• The devils in the details
• Equipment names duplication, consistent names,
  use of aliases (common names)
• Engineering model filtering (voltage level),
  available telemetry
• Modeling philosophy CBs vs. Switches
• Engineering
• Model reduction use only what is needed (less
  is better)
• Model size is not a software issue a
  maintenance issue
• Equivalents
• Difficult to merge from different sources and
  ensure a quality solution
• Masks the real system making problems harder to
  resolve
• Model ties where are models cut?
• Mechanics
• Automation incremental changes versus bulk
  changes
• Manual entry clean-up

21
Four aspects to model support

• Model validation
• Power Flow solution
• State Estimator solution based on true
  measurements
• Model Management system that allows validation
  prior to production
• Minimize impact on operational system
• Minimize down time dont miss State Estimation
  cycles
• Perform a soak test prior to putting into
  production
• Flexible enough to allow analysis of past,
  present and future models

22
State Estimation Needs

• Need timely quality solution to support market
  calculations dollars are at stake
• Measures of quality
• Voltage profile
• Injections (e.g. mitigate negative generation)
• Convergence criteria and resulting bus mismatch
• Good news and bad news
• Convergence history (number of solved cases over
  time)
• Size of observable area
• Metering
• ICCP short term loss of link should not result
  in missed solution
• Knowledgeable Support personnel
• Understanding of State Estimator theory
• State Estimation vs. Power Flow
• Understanding of the transmission system
  characteristics
• TOs are familiar with the model
• RTOs/ISOs will pick up knowledge over time

23
ISO/RTO Application Software Super Meters

• SCUC/SCED, SE, CRRs and Ex-Post LMPs are key
  ISO/RTO business enablers
• Super Meters with unparalleled complexity in
  data requirements and solution processes on a
  very large network models with no formal
  calibration in place
• Small errors and inconsistency in super meters
  or underlying data and models may have
  significant negative impacts on money flows and
  bottom line of market participants
• Auditing of the results of key ISO/RTO
  application software components is of very high
  importance
• Due to the underlying complexities in most cases
  the only way to audit the key software components
  is to run different tools in parallel
• Integration cost of achieving the proper auditing
  goals require a standardization of interfaces
  among the main components

24
Questions ?