8.3 Multi-machine rotor angle stability analysis

Swing Equation for Multi-Machine Systems

Extension of Single-Machine Swing Equation

In a single-machine system, the swing equation describes one generator swinging against an infinite bus. In a multi-machine system, every generator interacts with every other generator through the transmission network, so each machine's swing equation becomes coupled to the rest.

The result is a set of coupled nonlinear differential equations describing the rotor angles and speeds of all machines simultaneously. Each equation accounts for:

• The individual machine's inertia constant and damping coefficient
• Its mechanical power input from the prime mover
• Its electrical power output, which depends on terminal voltages and angles determined by the network and the states of all other machines

The coupling happens through the network admittance matrix: a change in one machine's rotor angle alters power flows across the network, which in turn affects the electrical power seen by every other machine.

Formulation of Multi-Machine Swing Equation

The multi-machine swing equation in matrix form is:

$M\ddot{\delta} + D\dot{\delta} = P_m - P_e(\delta, V)$

where:

• $M$ = diagonal matrix of inertia constants (one per machine)
• $D$ = diagonal matrix of damping coefficients
• $\delta$ = vector of rotor angles
• $P_m$ = vector of mechanical power inputs
• $P_e$ = vector of electrical power outputs

The electrical power output vector $P_e$ is a nonlinear function of rotor angles $\delta$ and bus voltages $V$, computed from the power flow equations:

$P_e = \operatorname{Re}(\operatorname{diag}(V)(Y_b V)^*)$

where $Y_b$ is the bus admittance matrix and $^*$ denotes the complex conjugate.

Together, the swing equation and the power flow equations form a complete electromechanical model of the power system. The nonlinearity of $P_e$ is what makes multi-machine stability analysis significantly harder than the single-machine case.

Multi-Machine Rotor Angle Stability

Stability Assessment of Equilibrium Points

An equilibrium point is a steady-state operating condition where each machine's mechanical power input equals its electrical power output and all rotor angle derivatives are zero. You find these by solving the power flow equations with $\ddot{\delta} = 0$ and $\dot{\delta} = 0$.

To assess stability at an equilibrium point:

1. Linearize the multi-machine swing equation around the equilibrium point.
2. Form the state matrix of the resulting linear system.
3. Compute the eigenvalues of the state matrix.
4. Interpret the results:
   • All eigenvalues have negative real parts → the equilibrium is stable (the system returns to synchronism after a small disturbance).
   • Any eigenvalue has a positive real part → the equilibrium is unstable (even a small disturbance can cause loss of synchronism).

This eigenvalue-based approach is the foundation of small-signal stability analysis.

Nonlinear Dynamical System Perspective

From a nonlinear dynamics viewpoint, the multi-machine system has multiple equilibrium points, some stable and some unstable.

Each stable equilibrium has a stability region: the set of initial conditions from which the system will converge back to that equilibrium. The boundaries between stability regions are formed by the stable manifolds of the unstable equilibrium points.

Why this matters practically:

• A small disturbance keeps the system within its current stability region, and it returns to the original operating point.
• A large disturbance (like a severe fault) can push the system across a stability boundary into a different region, leading to either a new steady-state condition or outright instability.

This is the conceptual basis for understanding why transient stability depends so heavily on fault severity and clearing time.

Factors Affecting Rotor Angle Stability

Network and Machine Parameters

Transmission network strength directly affects synchronizing torque between machines. Higher admittance values (lower impedance paths) reduce the electrical distance between generators, making it easier for them to exchange synchronizing power and stay in step.

Machine parameters also play a major role:

• Higher inertia constants mean more stored kinetic energy, so the rotor resists speed changes more effectively during disturbances.
• Lower machine reactances allow greater power transfer and stronger synchronizing torque, improving stability margins.

Operating Conditions and Disturbances

Systems operating closer to their limits are more vulnerable:

• High power transfers mean machines are already near their stability boundaries, so even a moderate disturbance can push them out of synchronism.
• Low voltage profiles indicate a weak transmission system with reduced power transfer capability, shrinking stability margins further.

The disturbance itself matters just as much:

• Location: Faults near critical generators or in weak parts of the network have the greatest destabilizing effect.
• Severity and duration: A longer fault duration or larger load loss reduces the remaining stability margin and makes recovery harder.

Control Systems and Stability Enhancement

Control systems are the primary tools for improving multi-machine stability beyond what the passive network provides.

• Excitation systems regulate generator terminal voltage and provide fast voltage support during disturbances. Modern high-gain, fast-response excitation systems significantly improve transient stability.
• Power System Stabilizers (PSS) are supplementary controllers that modulate generator excitation to damp low-frequency oscillations. They use input signals like rotor speed, frequency, or electrical power to produce a damping torque that improves small-signal stability.
• FACTS devices (SVC, STATCOM, TCSC, etc.) control power flows by injecting or absorbing reactive power, regulating voltage, or modulating line impedance. They provide fast, continuous adjustment of network parameters and can enhance both transient and dynamic stability.

Stability Assessment of Multi-Machine Systems

Numerical Simulation Techniques

Because multi-machine dynamics are nonlinear and coupled, numerical simulation is the primary tool for stability assessment. The general process is:

1. Solve the steady-state power flow (using Newton-Raphson or fast-decoupled methods) to establish initial conditions for all bus voltages, angles, and machine outputs.
2. Initialize the dynamic model with these steady-state values.
3. Apply the disturbance at a specified time (e.g., a three-phase fault, line trip, or load change).
4. Integrate the swing equations forward in time using numerical methods like Runge-Kutta or trapezoidal integration, updating the network solution at each time step.
5. Observe the transient response of rotor angles, speeds, voltages, and powers over the simulation period.

The integration methods discretize the continuous differential equations into algebraic updates at each time step, iteratively computing state variables from the previous step's values.

Stability Assessment Criteria and Techniques

After running a simulation, you assess stability by examining the time-domain response:

• Stable: Oscillations are damped and rotor angles converge to a new equilibrium.
• Unstable: Oscillations grow, angles diverge, or machines lose synchronism (angles separate unboundedly).

Key quantitative measures include:

• Critical clearing time (CCT): The maximum fault duration the system can tolerate without losing synchronism. This is a primary transient stability indicator.
• Maximum power transfer capacity: The highest power level the system can handle without violating stability limits, dependent on network topology, machine parameters, and control settings.

Modal analysis complements time-domain simulation for small-signal stability:

• Eigenvalue analysis reveals the frequency and damping ratio of each oscillatory mode.
• Participation factors identify which machines or state variables contribute most to each mode.
• These results guide the design and tuning of PSS and FACTS controllers to target the most poorly damped modes.

Probabilistic Stability Assessment

Deterministic analysis assumes fixed operating conditions, but real power systems face continuous uncertainty from load variations, renewable generation intermittency, and equipment failures. Monte Carlo simulation addresses this by:

1. Defining probability distributions for uncertain parameters (loads, wind/solar output, component availability).
2. Randomly sampling from these distributions across many trials.
3. Running a stability simulation for each sampled scenario.
4. Collecting statistics on the outcomes.

This produces probabilistic stability indices such as:

• Probability of instability
• Expected energy not served
• Expected cost of blackout events

These indices give a more realistic picture of system risk than any single deterministic scenario. They support planning decisions like where to reinforce the network, how much storage to add, or how to set reliability standards. Probabilistic assessment is especially valuable for evaluating the impact of increasing renewable penetration, aging infrastructure, or extreme weather events on system stability.