How Do You Tune a GE EX2100 Power System Stabilizer?

GE EX2100 stabilizer tuning starts at the measured generator signals and follows the stabilizing path through the power system stabilizer, automatic voltage regulator, static excitation system, generator field, rotor, and transmission network. A break, polarity error, limiter interaction, or model mismatch anywhere along that path can turn intended damping into weak damping or sustained oscillation. Treat tuning as a model-validation and commissioning task, not as parameter copying.

Where does the stabilizing signal travel?

The power system stabilizer derives an oscillatory component from one or more measured generator quantities. Its filters reject steady-state content, phase-compensation stages align the stabilizing response with the electromechanical mode, and gain determines response magnitude. The resulting supplementary signal enters the excitation regulator. The EX2100 changes generator field voltage, which changes electrical torque and damps rotor oscillation when magnitude and phase are correct.

The energized stabilizer on the other combined-cycle unit proves only that the site operates another PSS. Its brushless excitation path differs from the static-excitation path of this unit. Do not transfer its gain or time constants to the 325 MVA, four-pole, hydrogen-cooled generator.

Do the physical inputs pass a layer-one check?

Layer one first. Record every PSS input at a stable operating point and during an approved small disturbance. Verify signal continuity, polarity, engineering-unit scaling, channel assignment, noise, and saturation. A mathematically correct tuning study cannot compensate for a reversed or incorrectly scaled input.

1. Trace each configured stabilizer input from its measurement source to the EX2100 input channel.
2. Compare the displayed value with an independent plant indication at the same operating condition.
3. Confirm that increasing generator output produces the expected direction of change.
4. Trend the raw input, filtered signal, stabilizer output, regulator output, field response, terminal voltage, real power, and any active limiter.
5. If any channel clips, jumps, drifts, or carries excessive noise, repair that path and repeat the measurement before proceeding.

If the inputs are valid, move to the base control loops. If they are not, stop: changing PSS gain would hide the measurement fault without correcting the feedback path.

Are the excitation and governor settings final?

Tune the stabilizer after completing the excitation-control settings. The PSS acts through the voltage regulator, so regulator gain, response, limits, and field-system dynamics determine the phase and magnitude ultimately delivered as electrical torque. A later excitation change invalidates part of the stabilizer model.

The governor belongs in the same dependency check. One commissioning case reported that changed governor characteristics upset the PSS model and forced renewed review by the transmission authority. Freeze the governing configuration used for the study, or update the model and repeat the stability cases after any change.

Does the dynamic model reproduce the unit?

Build a transient-stability representation of the generator, turbine-governor, excitation system, PSS, and relevant network. The model must reproduce the unit both with the stabilizer disabled and enabled. Suitable studies require specialized software and engineers who remain proficient with it; the generator owner may use the OEM or a qualified power-system consultant.

Validate the non-PSS response before optimizing stabilizer parameters. Compare simulated and measured voltage, field, real-power, and speed or frequency responses to an approved small excitation test. Tune model parameters to match the plant response rather than forcing PSS settings to compensate for an inaccurate plant model.

1. Establish the tested operating point and network configuration.
2. Run the same disturbance in the model with the PSS disabled.
3. Overlay calculated and recorded response channels.
4. Correct generator, excitation, governor, and network data responsible for material response differences.
5. Repeat until the model reproduces oscillation frequency, response direction, initial magnitude, and decay behavior closely enough for the study purpose.

OEM experience can provide a useful starting region for gain and time constants, but final values still require study and on-unit confirmation.

What must the transmission operator specify?

The transmission operator normally defines required behavior or operating boundaries; the owner determines settings that meet them. One statement in the installation record asserted that the grid must furnish tuning parameters, while multiple commissioning accounts distinguished compliance requirements from actual settings. Resolve that responsibility from the interconnection agreement and the operator's current test procedure before scheduling work.

Request measurable acceptance criteria: required operating cases, permitted test disturbances, signals to record, model format, review stages, and approval sequence. Do not interpret a requirement that the PSS be active and properly tuned as a parameter set.

Does simulation show positive damping in every required case?

Run each required disturbance and operating condition twice: first with the PSS out of service, then with it enabled. Compare the oscillatory decay, not merely the first peak. A useful setting increases damping without causing growing oscillation, unacceptable voltage modulation, control saturation, or harmful interaction with excitation limits.

Vary gain and compensation time constants in the validated model. Phase compensation must make the excitation-induced torque oppose the rotor-speed deviation over the target oscillation range. Too little gain produces negligible improvement; excessive gain magnifies noise, excites other modes, drives output limits, or reduces stability margin.

Changing transmission impedance changes the modes seen by the generator, so include the credible network configurations required by the operator. Series-compensated networks can introduce subsynchronous interaction, but a conventional PSS tuning exercise is not automatically a subsynchronous-resonance solution. Treat that risk as a separate torsional and network study and apply only a mitigation demonstrated by that study.

How should the resolving settings be commissioned?

Use controlled on-line testing only under an approved procedure with plant and transmission coordination. A reported EX2000 service workflow used a remote connection, small step changes, transferred records, and off-site calculation. That account does not establish the same access method for the EX2100; obtain the applicable OEM procedure and cybersecurity approval for this controller.

1. Back up the installed EX2100 configuration and record the current PSS enable state, gain, time constants, output limits, excitation settings, governor configuration, and software identification displayed by the system.
2. Load only the settings produced by the accepted study and independently check every entered value against the approved setting sheet.
3. Place the unit at the modeled operating point and confirm that no unexpected limiter is active.
4. Apply the approved small disturbance while recording synchronized PSS input and output, regulator output, field response, terminal voltage, real power, and speed or frequency.
5. Abort according to the approved test limits if oscillations grow, controls saturate, voltage departs from its permitted band, or protection and limiter states change unexpectedly.
6. Compare the measured enabled response with the disabled baseline and the simulated response. Confirm response polarity, oscillation frequency, decay, settling behavior, and output headroom.
7. Repeat the operator-required operating cases and network conditions, then submit settings, models, plots, test configuration, and results for approval.

How is the final setting verified?

Acceptance requires more than an enabled status bit. Demonstrate that the installed parameters match the approved set, the input channels remain valid, and measured oscillations decay faster with the PSS active than with the comparable disabled baseline. Check that the field and terminal-voltage responses remain controlled and that the stabilizer output does not spend the test against its limits.

Preserve the final model, raw test records, configuration backup, plots, network condition, unit loading, governor state, excitation state, and operator approval. Any later change to excitation tuning, governor characteristics, measurement scaling, generator data, or material network assumptions triggers an impact review and may require renewed simulation and testing.

FAQ