State Estimation (SE) Tuning Guide

Introduction & Tuning Phases

SE tuning is an iterative process to bring up the State Estimation function with real-time data. The process is broken down into three main phases:

Phase 1: Initial Bring-up

Bring up SE with the latest database. Tune gross error detection. On-the-job training for customer personnel.

Phase 2: Systematic Tuning

Conducted by the Customer. Systematic tuning of all SE control parameters for all possible power system conditions. Allocation files are smoothed.

Phase 3: System-wide Checkout

Conducted by {OCO}. System-wide checkout of interfaces between SE and other power applications (DLF, CA, etc.).


Pre-Tuning Checklist

  • Source Data (SDBGEN) processed without errors.
  • One-line and application displays translated.
  • Sufficient RTUs are operational for observability.
  • System cutover complete: Data acquisition is scanning.

Refer to Sections 1.1 - 1.3 of the source document.

Global Tuning Parameters

These parameters control the core SE functions: pre-processing, solution convergence, and gross error detection.

Pre-Processing (RTDBIF, BDPROC) - Ref: APANMC, AMFLTM
Multiplied by AERROR to detect significant load changes. (Doc: 3.2.3)
Number of cycles without a load change to declare steady-state. (Doc: 3.2.5)
For out-of-service equip, measurement must be < AMSCT * AERROR to be considered zero. (Doc: 4.2)
For in-service equipment, to detect if flow is significantly zero. (Doc: 4.2)
Solution Convergence (SDSOLN) - Ref: APSESC
Gross Error Detection (SDGMDA) - Ref: APSESC

Troubleshooting Guide for SE Alarm Messages

Select an alarm category or search for a specific message.

  • SE NOT RUN: NO STEADY STATE

    Cause: RTDBIF could not find a steady-state set of A/D measurements within MXECYC cycles. Tune MEMPYS, MXSCYC, or individual AERROR.

  • RTDB FETCH FOR A/D IN ERR: n

    Cause: Error fetching data from RTDB. Check RTU, point definition, and communication links.

  • IN-SERVICE EQUIPMENT [name] HAS [value] MW FLOW

    Cause: Equipment is in-service but flow is near zero. Check status (SSP), measurement error (AERROR), or ALSCNT multiplier.

  • OUT-OF-SERVICE BUS [name] HAS [value] KV

    Cause: Bus is out-of-service but has voltage. Check breaker statuses or adjust AMSCT multiplier.

  • CIRCUIT [name] LOSS [value] MW > LIMIT [value]

    Cause: MW loss > PDELTA. Check terminal measurements or increase the reasonability limit PDELTA.

  • GROSS MEAS ERR DETECTED : [type]

    A measurement has been rejected. Check the measurement and its associated equipment. May require tuning measurement error (SIGMA).

  • SE EXCEEDED ITERATION LIM : [n] : [p]

    Solution did not converge. Check the bus with largest mismatch (B_SEBUSMIS). Consider adjusting PTOL, QTOL, or ITRLIM.

SE Tuning Displays (Appendix B)

These are the primary displays referenced in the guide for viewing and modifying tuning data.

Display ID Display Name Description / Key Parameters
B_RTDATAAnalog Measurement DataAERROR (Measurement Error), ALPHA, IMTYPE (Measurement Type)
B_SEPREPSE Tuning for Pre-processingMEMPYS, MXSCYC, AMSCT, ALSCNT, BRFAC, BDFAC
B_SESOLNSE Tuning for Solution ConvergencePTOL, QTOL, ITRLIM, DVLIM, TAPFLG, BIGDOR
B_SEGROSSE Tuning for Gross Error DetectionCHITE, TCHEK, NSIGMA, REDNTH, DORMAN
B_SEEBIGBus Injection Generation Point ParametersPMEAS, QMEAS, BGPGME (SIGMA P), BGQGME (SIGMA Q)
B_SECIRMCircuit Measurement SigmaFCPMER (From MW Sigma), FCQMER (From MVAR Sigma)
B_SELFBRLoad Flow Bus Measurement SigmaBKVMER (Voltage Sigma), BPGMER (Generation MW Sigma)
B_SEBUSMISBus Injection Mismatch SummaryDisplays buses with largest MW/MVAR mismatches for debugging divergence.

Step-by-Step Tuning Process

Based on Section 1.4 (State Estimation Tuning Steps) and Section 7 (Adjustment for Solution Convergence).

Step 1: Confidence Test
Initial

Execute SE with simulated data from a load flow solution to ensure programs and database are integrated correctly.

Step 2: Review Standard Deviations (SIGMA)
Database Check

Ensure reliable measurements (gen MW) have small SIGMA. Pseudo measurements should have larger SIGMA. (Ref: B_RTDATA)

Step 3: Isolate External Areas
Initial Execution

Disable use of external telemetered values and voltage regulation to prevent corruption from external model errors.

Step 4: Run SE & Check Preprocessor Alarms
Fix Obvious Errors

Execute SE and check alarms from BDPROC (e.g., energized equip with zero flow). Fix RTU/database issues. (Ref: Troubleshooting Tab)

Step 5: Increase Tolerances for Initial Solution
Get First Solution

Increase PTOL, QTOL, CHITE, and measurement error multipliers to help SE obtain an initial, albeit rough, solution. (Ref: Global Params Tab)

Step 6: Analyze Bad Data
Iterative Improvement

Check the Bad Data Summary. Starting with the largest residual, investigate and fix meter definitions, topology, and model parameters.

Step 7: Reduce Tolerances & Repeat
Iterative Improvement

Gradually reduce PTOL, QTOL, CHITE back to normal values, repeating Step 6 until all major errors are fixed.

Step 8-12: Final Integration
Completion

Restore load distribution factors, enable external areas, and inspect results for negative loads or "perfect bus" injection errors.