7.3 Modal analysis and stability assessment

Modal analysis for system stability

Modal analysis is the primary technique for assessing small-signal stability in power systems. It reveals the natural oscillatory modes embedded in the system, characterizes each mode by its frequency and damping, and identifies which generators or states drive those modes. Engineers use this information to find weak points where poorly damped oscillations could threaten synchronism.

Identifying critical modes and their characteristics

The starting point is the state matrix (often called the $A$ matrix) from the linearized system model. The eigenvalues of this matrix each correspond to one oscillatory mode, and they carry two pieces of information:

• Real part ($\sigma$): Represents damping. A more negative value means the oscillation decays faster. A positive value means the oscillation grows.
• Imaginary part ($\omega$): Represents the oscillation frequency in radians per second.

Beyond eigenvalues, the eigenvectors tell you who is involved in each mode:

• Right eigenvectors (mode shapes) show the relative activity of each state variable in a given mode. They answer: "When this mode is excited, which states move the most?"
• Left eigenvectors show how much each mode contributes to the response of a particular state variable. They answer: "When a state is disturbed, which modes get excited?"
• Participation factors combine both. They're computed element-by-element from the product of left and right eigenvectors and identify the most influential generators or states (rotor angle, generator speed) for each mode. A high participation factor tells you where to focus control efforts.

Critical modes are those with low damping or high participation from key generators. These are the modes most likely to cause sustained oscillations after a disturbance.

Assessing small-signal stability

Small-signal stability is the system's ability to maintain synchronism when subjected to small perturbations, such as minor load fluctuations or small generation changes.

The stability criterion is straightforward:

• If all eigenvalues have negative real parts, the system is stable. Every oscillation eventually decays.
• If any eigenvalue has a positive real part, that mode is unstable. Oscillations in that mode will grow over time.

To quantify how well-damped each mode is, you calculate the damping ratio:

$\zeta = \frac{-\sigma}{\sqrt{\sigma^2 + \omega^2}}$

This gives a normalized measure between 0 and 1. A damping ratio of 1 is critically damped (no oscillation at all), while values between 0 and 1 are underdamped (oscillatory but decaying). In practice, modes with damping ratios below about 5% are considered poorly damped and flagged as stability risks.

Damping and frequency of modes

Importance of damping and frequency

Damping and frequency are the two parameters that define how each oscillatory mode behaves. Damping determines whether oscillations die out or persist. Frequency tells you how fast the oscillations cycle.

Modes with low damping and high participation from important generators are the most concerning, because they can produce sustained or growing oscillations that propagate through the system. Two common categories stand out:

• Inter-area modes (0.1–1 Hz): Groups of generators in one region oscillate against groups in another region. These are typically the hardest to damp because they involve large portions of the interconnected system.
• Local modes (1–2 Hz): A single generator (or a small group) oscillates against the rest of the system. These are usually easier to address with local controls like power system stabilizers.

Calculating damping and frequency

Here's how to extract damping and frequency from an eigenvalue step by step:

1. Start with a complex eigenvalue in the form $\sigma \pm j\omega$.
2. The damping is given directly by $\sigma$. More negative means better damping.
3. The oscillation frequency in Hertz is $f = \frac{\omega}{2\pi}$.
4. The damping ratio is $\zeta = \frac{-\sigma}{\sqrt{\sigma^2 + \omega^2}}$.

Worked example: Consider an eigenvalue of $-0.2 \pm j3.14$.

• $\sigma = -0.2$, so the mode is stable (negative real part).
• Frequency: $f = \frac{3.14}{2\pi} = 0.5$ Hz. This falls in the inter-area range.
• Damping ratio: $\zeta = \frac{0.2}{\sqrt{0.04 + 9.86}} = \frac{0.2}{3.146} \approx 0.0635$, or 6.35%.

At 6.35%, this mode is above the typical 5% threshold but still relatively lightly damped and worth monitoring.

System changes and modal behavior

Impact of system changes

Power system operators need to understand how modifications affect modal behavior. Changes that shift eigenvalues, mode shapes, or participation factors can either improve or degrade stability. Common changes include:

• Generator control settings: Tuning excitation systems or adding/adjusting power system stabilizers (PSSs).
• Transmission parameters: Changing line reactance, adding series compensation, or switching lines in/out of service.
• Load characteristics: Shifts in load composition (e.g., more motor load vs. resistive load).
• Topology changes: Adding or removing generators, lines, or other components.

After any modification, modal analysis is re-run on the updated linearized model. You then compare the new eigenvalues against the original ones:

• Did the real parts become more negative? That means improved damping.
• Did any mode shift closer to the imaginary axis (or cross it)? That's a warning sign.
• Did participation factors shift, indicating different generators now dominate a critical mode?

The goal is to push critical eigenvalues further into the left half of the complex plane (more negative $\sigma$) while keeping frequencies in acceptable ranges.

Sensitivity analysis and trade-offs

Sensitivity analysis identifies which parameters have the greatest influence on a particular mode's damping. For example, you might find that increasing the gain of a PSS on a specific generator significantly improves the damping of a critical inter-area mode, while adjusting other generators has little effect. This targeted approach avoids unnecessary changes across the system.

However, stability improvements don't come for free. Trade-offs arise because:

• Tuning a PSS for better damping may require the generator to deviate from its optimal economic dispatch point.
• Adding series compensation to improve inter-area damping can affect voltage profiles along the transmission corridor.
• Control actions that stabilize one mode can sometimes reduce the damping of another.

Modal analysis provides the quantitative basis for balancing these competing objectives. By tracking how eigenvalues, participation factors, and mode shapes respond to proposed changes, engineers can find operating points that maintain adequate stability margins without sacrificing too much economic efficiency or voltage quality.