Question 7 of 7: Equal-area criterion and the critical clearing angle
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. National Exams, May 2015 — 07-Elec-B7 Power Systems
Engineering. Open-book, three hours, seven problems of equal value; any five constitute a
complete paper. Every problem is answered here, because the set is intended as a study
resource rather than an exam script.
Reference texts for this subject.
M. E. El-Hawary, Electrical Power Systems: Design and Analysis, rev. ed.,
IEEE Press / Wiley — the syllabus text for 07-Elec-B7. Chapters 4 (transmission lines),
5 (synchronous machines), 3 (transformers), 6 (load flow), 7 (fault analysis) and
8 (system stability) map one-to-one onto the seven problems below.
J. D. Glover, T. J. Overbye and M. S. Sarma, Power System Analysis and Design,
6th ed., Cengage — parallel treatment of the ABCD long-line model, symmetrical
components and the equal-area criterion.
S. J. Chapman, Electric Machinery Fundamentals, 5th ed., McGraw-Hill —
salient-pole two-reaction theory and transformer loss/frequency behaviour.
CSA C22.3 No. 1, Overhead Systems, and CAN/CSA C22.2 standards give the
Canadian equipment and clearance framework these calculations feed into.
Question 7: Equal-area criterion and the critical clearing angle
Given. A machine of internal voltage E behind a transient reactance,
feeding an infinite bus through line 1 and then two parallel lines, with a three-phase fault at
F on line 3 immediately adjacent to the receiving busbar.
Circuit data (per unit)
Element
Reactance
Quantity
Value
Machine (transient)
j0.20
Internal voltage E
1.2 pu
Line 1
j0.05
Infinite-bus voltage V
1.00 ∡0° pu
Line 2
j0.50
Mechanical input Pm
1.0 pu
Line 3 (faulted)
j0.40
Fault type
bolted three-phase at F
Receiving transformer
j0.05
Clearing action
open line 3
Find. The initial power angle at 1.0 pu load; the power-angle curve while
the fault is on; the power-angle curve after line 3 is opened; and the critical clearing
angle.
Figure 8 — Circuit of Figure (4). The fault
at F sits at the receiving end of line 3, so it holds the parallel-line receiving busbar at
zero voltage until line 3 is opened.
Approach. Each of the three network conditions — pre-fault,
during-fault and post-fault — reduces to a single transfer reactance between E and V,
and each therefore has its own sinusoidal power-angle curve $P = (EV/X)\sin\delta$. The initial
angle comes from the pre-fault curve at $P = P_m$, and the critical clearing angle follows from
equating the accelerating area under the during-fault curve to the maximum decelerating area
available under the post-fault curve.
Part (a), reduce the pre-fault network. With both parallel lines in
service they combine in parallel and add to the two series reactances:
$$X_{\text{pre}}=0.20+0.05+\frac{(0.50)(0.40)}{0.50+0.40}+0.05=0.20+0.05+0.22222+0.05=0.52222\ \text{pu}$$
$$P_{\max,1}=\frac{EV}{X_{\text{pre}}}=\frac{(1.2)(1.0)}{0.52222}=2.29787\ \text{pu}$$
Solve for the initial angle. In the steady state the electrical output
equals the mechanical input, $P_e = P_m = 1.0$ pu, so
$$\sin\delta_0=\frac{P_m}{P_{\max,1}}=\frac{1.0}{2.29787}=0.43519$$
$$\boxed{\delta_0 = 25.80^\circ = 0.45025\ \text{rad}}$$
Part (b), examine the network during the fault. The fault at F sits at the
receiving end of line 3, that is at the busbar where lines 2 and 3 come together. A bolted
three-phase fault there holds that busbar at zero voltage. Both parallel lines then terminate
on a short circuit, and the infinite bus behind the receiving transformer is also connected to
that same zero-voltage point, so there is no path from E to V that does not pass through the
fault. The transfer reactance is infinite and
$$\boxed{P_{e,\text{fault}} = 0\ \text{for all }\delta}$$
Physically the machine still delivers current, but into a purely reactive short circuit, so
the real power it exports is zero. This is the worst case for stability: the full 1.0 pu of
mechanical input goes into accelerating the rotor from the instant the fault appears.
Part (c), reduce the post-fault network. Clearing opens line 3 at both
ends, removing both the fault and that line. Only line 2 remains in the parallel section:
$$X_{\text{post}}=0.20+0.05+0.50+0.05=0.80\ \text{pu}$$
$$P_{\max,3}=\frac{EV}{X_{\text{post}}}=\frac{1.2}{0.80}=1.500\ \text{pu}$$
so the post-fault characteristic is
$$\boxed{P_{e,\text{post}}=1.500\sin\delta\ \text{pu}}$$
It still peaks above the 1.0 pu mechanical input, so a stable post-fault equilibrium exists at
$\delta = \sin^{-1}(1/1.5) = 41.81^\circ$; had $P_{\max,3}$ fallen below 1.0 pu the machine
could not have been stabilised at any clearing time.
Part (d), fix the limit of the decelerating swing. The largest angle from
which the rotor can still be pulled back is the unstable equilibrium on the post-fault curve,
$$\delta_{\max}=180^\circ-\sin^{-1}\frac{P_m}{P_{\max,3}}=180^\circ-41.81^\circ=138.19^\circ=2.41188\ \text{rad}$$
Beyond this the post-fault electrical power again falls below $P_m$ and the rotor accelerates
without limit.
Write the equal-area condition. The accelerating area between $\delta_0$
and the clearing angle, and the decelerating area between clearing and $\delta_{\max}$, must be
equal:
$$\int_{\delta_0}^{\delta_{cr}}\left(P_m-0\right)d\delta=\int_{\delta_{cr}}^{\delta_{\max}}\left(P_{\max,3}\sin\delta-P_m\right)d\delta$$
Evaluating both integrals and collecting the terms in $\delta_{cr}$ gives the closed-form
result
$$\cos\delta_{cr}=\frac{P_m\left(\delta_{\max}-\delta_0\right)+P_{\max,3}\cos\delta_{\max}}{P_{\max,3}}$$
where $\delta_{\max}$ and $\delta_0$ are in radians.
Evaluate the critical clearing angle. Substituting the numbers,
$$\cos\delta_{cr}=\frac{(1.0)(2.41188-0.45025)+(1.500)(-0.74536)}{1.500}=\frac{1.96163-1.11803}{1.500}=0.56239$$
$$\boxed{\delta_{cr} = 55.78^\circ}$$
Checking the answer by direct integration, the accelerating area is
$A_1 = P_m(\delta_{cr}-\delta_0) = 0.52328$ pu-rad and the decelerating area available between
$55.78^\circ$ and $138.19^\circ$ is also $0.52328$ pu-rad, so the two balance exactly.
Interpret the result. The rotor may swing a further
$55.78^\circ - 25.80^\circ = 29.98^\circ$ beyond its initial position before the breakers must
have opened. Converting that to a clearing time requires the swing equation,
$M\,d^{2}\delta/dt^{2}=P_m$ during a zero-output fault, which integrates to
$t_{cr}=\sqrt{4H(\delta_{cr}-\delta_0)/(\pi f_0P_m)}$; for a typical inertia constant
$H = 4$ MJ/MVA at 60 Hz this gives 0.211 s, or roughly 13 cycles — comfortably
within the reach of modern protection, which clears a transmission fault in three to five
cycles.
Figure 9 — Equal-area construction. The
during-fault curve lies on the δ axis, so the accelerating area A1 is simply the
rectangle Pm(δcr − δ0).
Check: Figure (4) draws F on line 3 at its receiving-end terminal, and the
question states the fault is cleared by opening that line — the standard idealisation of
a fault immediately outside the busbar, on the line side of the breaker. The during-fault
transfer reactance is therefore infinite and the during-fault power is zero. Had the fault been
located a distance along line 3 rather than at its end, a wye-delta transformation of the
(source-node, receiving-bus, fault) triangle would be needed; with the whole 0.40 pu of line 3
acting as the shunt arm this would give a transfer reactance of 1.14375 pu, a during-fault peak
of 1.049 pu and a critical clearing angle of 102.95°. The location as drawn gives the more
conservative answer, and it is the one reported above.