Balanced three-phase circuits

A three-phase motor, a transformer bank, or a power line is never really solved phase-by-phase from scratch — once you know the source is balanced and the load is balanced, the whole three-wire (or four-wire) mess collapses to one ordinary single-phase AC circuit. Solve that one circuit, and the other two phases come for free by symmetry: same magnitude, shifted $120^\circ$ . This is the single biggest labour-saving trick in three-phase analysis, and it rests on a fact worth remembering for life: a perfectly balanced star load draws zero neutral current, so the neutral wire can be ignored entirely while solving. This article takes the star/delta relations you already know and puts them to work: reducing a balanced star or delta load to its per-phase equivalent, solving for line and phase currents, and computing total three-phase power with both the per-phase formula $P=3V_{ph}I_{ph}\cos\theta$ and the universal line-quantity formula $P=\sqrt3\,V_LI_L\cos\theta$ . Two fully worked numeric examples — one star, one delta with a star cross-check — pin every step to real numbers.

The core insight

A balanced three-phase circuit reduces to one single-phase equivalent per phase (current $I_{ph}=V_{ph}/Z_{ph}$ , the other two phases identical but shifted $120^\circ$ ), and total power is always $P=\sqrt3\,V_LI_L\cos\theta$ with $\theta$ the load's own impedance angle.

The core idea

A circuit is balanced when the source EMFs are equal and $120^\circ$ apart, and all three load impedances are identical in magnitude and phase angle. Under balance, the three line currents come out equal in magnitude and $120^\circ$ apart too, and their phasor sum is exactly zero — which is why a balanced star load's neutral wire carries no current at all.

Reduction to a single phase. Because all three branches are identical, you never need to solve three coupled loops. The recipe is short: If the load is delta-connected, first convert it to an equivalent star using

Z_Y = \frac{Z_\Delta}{3}

where: $Z_Y$ is the equivalent star (per-phase) impedance and $Z_\Delta$ is the given delta-branch impedance — the conversion only scales magnitude, it does not change the impedance angle. Then draw one phase: source EMF $\vec E_{ph}$ , any line impedance, and the star-equivalent load $Z_{ph}$ , returning through a fictitious (zero-current) neutral. Solve this ordinary single-phase branch,

I_{ph} = \frac{V_{ph}}{Z_{ph}}

where: $V_{ph}$ is the phase voltage across that branch and $Z_{ph}=R+jX$ is the branch impedance, with $|Z_{ph}|=\sqrt{R^2+X^2}$ and impedance angle $\theta=\tan^{-1}(X/R)$ . The other two phases carry currents of the same magnitude, merely displaced $\pm120^\circ$ — no extra work. Finally convert back to line quantities using the star or delta relations of the actual connection, not the equivalent one.

Figure

Single line source phase impedance loop with two identical phases shown displaced 120 degrees needing no separate solution

Power in a balanced circuit. Total real power is three times the per-phase power,

P = 3\,V_{ph}I_{ph}\cos\theta

where: $\theta$ is the phase (branch) impedance angle — the angle between phase voltage and phase current in one branch, never an angle between line quantities. Substituting the star or delta line-phase relations converts this into a single line-quantity formula that works for both connections:

P = \sqrt3\,V_L I_L\cos\theta,\qquad Q=\sqrt3\,V_LI_L\sin\theta,\qquad S=\sqrt3\,V_LI_L

where: $V_L$ and $I_L$ are the line (measured, terminal) voltage and current, and $Q$ , $S$ are reactive and apparent power. The convenience of the line-quantity form is exactly why it is the one quoted on nameplates and in power calculations — you never need to know whether the hidden load is star or delta.

Derivation AWhy P = root-3 VL IL cos theta holds for both star and delta loads

The goal is to show that although the per-phase power formula $P=3V_{ph}I_{ph}\cos\theta$ looks connection-specific, substituting either the star or the delta line-phase relations collapses it to the same single line-quantity expression $P=\sqrt3\,V_LI_L\cos\theta$ — which is why this one formula is safe to use without first knowing how the load is wired.

Step 1. Start from the per-phase definition, true for any connection since it is just three identical single-phase branches added up: $P=3V_{ph}I_{ph}\cos\theta$ , with $\theta$ the impedance angle of one branch.

Step 2. Substitute the star relations. For a star load $V_{ph}=V_L/\sqrt3$ and $I_{ph}=I_L$ (line current equals phase current in star). Plugging in: $P=3\left(\dfrac{V_L}{\sqrt3}\right)I_L\cos\theta = \dfrac{3}{\sqrt3}V_LI_L\cos\theta=\sqrt3\,V_LI_L\cos\theta$ , using $3/\sqrt3=\sqrt3$ .

Step 3. Substitute the delta relations instead. For a delta load $V_{ph}=V_L$ and $I_{ph}=I_L/\sqrt3$ (line current is $\sqrt3$ times phase current in delta). Plugging in: $P=3V_L\left(\dfrac{I_L}{\sqrt3}\right)\cos\theta=\dfrac{3}{\sqrt3}V_LI_L\cos\theta=\sqrt3\,V_LI_L\cos\theta$ — the identical result.

Step 4. Read the result. Both connections funnel down to the same line-quantity formula because the $\sqrt3$ that star hides in the voltage relation is exactly the $\sqrt3$ that delta hides in the current relation. This is why $\theta$ in $P=\sqrt3V_LI_L\cos\theta$ must still be read off the phase (branch) impedance, even though $V_L$ and $I_L$ are line quantities — the angle never transferred, only the algebra rearranged.

Worked example

A balanced delta-connected load, each branch $Z_{ph}=30\ \Omega$ purely resistive ( $\theta=0$ ), is fed at $V_L=400\ \text{V}$ . Find the phase current, line current and total power, and cross-check using the equivalent-star method.

Step 1. Direct delta method — phase voltage equals line voltage in delta: $V_{ph}=V_L=400\ \text{V}$ .

Step 2. Phase current: $I_{ph}=\dfrac{400}{30}=13.33\ \text{A}$ .

Step 3. Line current uses the delta relation $I_L=\sqrt3\,I_{ph}$ : $I_L=1.732\times13.33=23.09\ \text{A}$ .

Step 4. Total power, purely resistive so $\cos\theta=1$ : $P=\sqrt3\,V_LI_L\cos\theta=1.732\times400\times23.09\times1=16{,}000\ \text{W}=16\ \text{kW}$ .

Step 5. Cross-check via equivalent star. Convert $Z_Y=Z_\Delta/3=30/3=10\ \Omega$ . Star phase voltage $V_{ph}'=V_L/\sqrt3=400/1.732=230.9\ \text{V}$ , star phase current $I_{ph}'=230.9/10=23.09\ \text{A}$ — matching the delta line current from Step 3, exactly as the reduction method promises. Power: $P=3V_{ph}'I_{ph}'\cos\theta=3\times230.9\times23.09\times1=16{,}000\ \text{W}$ , matching Step 4.

Final answer: $I_{ph}=13.33\ \text{A}$ , $I_L=23.09\ \text{A}$ , $P=16\ \text{kW}$ .

Competitive stretch

A companion balanced star-connected load with $Z_{ph}=8+j6\ \Omega$ on the same class of supply, $V_L=415\ \text{V}$ , gives $V_{ph}=415/1.732=239.6\ \text{V}$ , $|Z_{ph}|=\sqrt{8^2+6^2}=10\ \Omega$ , $\theta=\tan^{-1}(6/8)=36.87^\circ$ , so $I_{ph}=I_L=239.6/10=23.96\ \text{A}$ , $\cos\theta=0.8$ lagging, and $P=\sqrt3\times415\times23.96\times0.8=13{,}789\ \text{W}\approx13.79\ \text{kW}$ . From the same numbers, $Q=\sqrt3\,V_LI_L\sin\theta=1.732\times415\times23.96\times0.6=10{,}342\ \text{VAR}$ and $S=\sqrt3\,V_LI_L=17{,}236\ \text{VA}$ — check: $S^2=P^2+Q^2$ gives $17{,}236^2\approx13{,}789^2+10{,}342^2$ , confirming the power triangle closes.

Common traps

Using $\theta$ from an assumed line-quantity relation instead of the actual load impedance angle — $\theta$ in $P=\sqrt3V_LI_L\cos\theta$ is always the phase (branch) impedance angle.
Forgetting to convert delta impedance to star ( $Z_Y=Z_\Delta/3$ ) before drawing the single-phase equivalent, or applying the conversion backwards ( $Z_\Delta=3Z_Y$ , not divide).
Assuming a neutral wire is required to solve a balanced star circuit — for a balanced load neutral current is zero, so the neutral (if present) can be omitted from the per-phase equivalent without changing the answer.
Plugging line voltage directly into $P=3V_{ph}I_{ph}\cos\theta$ without converting to phase voltage first — that shortcut only works for delta, where $V_{ph}=V_L$ , and is wrong for star.

Ask the tutor: want a walkthrough of a mixed problem where the source is star but the load is delta, or more practice computing reactive and apparent power for a lagging balanced load?

Why is the line-quantity formula P = root-3 VL IL cos theta so much more useful than the per-phase one?

Because line voltage and line current are exactly what you can measure at the terminals with an ordinary voltmeter and ammeter, without needing to know or open up whether the hidden load is star or delta connected. The derivation shows both connections collapse to this same formula, so it is safe to apply directly from terminal readings — that is why it is the one printed on equipment nameplates and used in almost all practical power calculations.

If the neutral current is zero for a balanced star load, why do real distribution systems still run a neutral wire?

Because real loads are rarely perfectly balanced in practice — different numbers of lights or appliances on each phase mean the three phase currents are not exactly equal, so some current does flow in the neutral to carry away that imbalance. The zero-neutral-current result you compute here is the idealised balanced case used for analysis; it tells you the theoretical minimum, not a reason to omit the wire in a real installation.

How do I quickly tell whether to use Vph = VL or Vph = VL/root3 in a problem?

Look at how the load is connected, not the source. In a delta-connected load, each branch sits directly across two line terminals, so phase voltage equals line voltage, $V_{ph}=V_L$ . In a star-connected load, each branch sits between one line terminal and the neutral point, so phase voltage is smaller by $\sqrt3$ , $V_{ph}=V_L/\sqrt3$ . When you convert delta to an equivalent star for solving, remember you are now using the star relation on the equivalent circuit even though the real load is delta.

Balanced three-phase circuits

The core idea

Worked example

Competitive stretch

Common traps

Before this

Where this leads