Kirchhoff's Rules

Series and parallel tricks are wonderful — for about two batteries. Add a second cell in an unexpected branch, or lay a resistor across the middle of a network, and suddenly nothing is in series or parallel with anything: the shortcuts from Resistors in Series & Parallel simply have nothing to grab. Yet the circuit itself is not confused. Switch it on and the currents settle instantly into one definite pattern, the same pattern every single time. Something decides those currents — and it is not cleverness. Two conservation laws decide, and this module is about wielding them. Charge cannot pile up at a junction, so the currents into any meeting point of wires must balance: that is Kirchhoff's junction rule, and it is the whole of it. Energy cannot appear from nowhere, so a charge carried around any closed loop must come back with exactly the energy it started with — every volt gained in a battery spent somewhere along the way: that is the loop rule. Gustav Kirchhoff wrote both down in 1845, as a 21-year-old student, and they have been the constitution of circuit analysis ever since. Every network, however tangled, surrenders to them by patient bookkeeping. The price of that power is discipline. The rules only work if you handle signs the same way every time — and sign slips are precisely where exam marks go to die. This article states the two rules, builds the one sign algorithm you need, solves a genuine two-battery, two-loop circuit end to end, and shows you how to audit your own answer so you know it is right before the examiner does.

Key takeaways

Junction rule: $\sum I_{\text{in}} = \sum I_{\text{out}}$ at every junction — conservation of charge; current cannot pile up at a node.
Loop rule: $\sum \Delta V = 0$ around any closed loop — conservation of energy; the potential returns to its starting value after a full tour.
The method: name every branch current with an arrow you choose freely, write junction equations, walk loops with disciplined signs, solve. A negative answer means the real current opposes your arrow — nothing is wrong.
Sign algorithm: crossing a resistor with its current arrow is $-IR$ , against it is $+IR$ ; crossing a battery from $-$ terminal to $+$ is $+\varepsilon$ , the reverse is $-\varepsilon$ — decided by the terminals, never by the current.
Audit every solve: currents must balance at each junction, and power delivered must equal power absorbed.
Key idea: two conservation laws, applied with disciplined signs, solve every circuit that exists — including all the ones series/parallel cannot touch.

The core insight

Kirchhoff's rules are charge conservation and energy conservation written as circuit bookkeeping — currents balance at every junction, and potential changes sum to zero around every closed loop — and together they determine every current in any network of batteries and resistors.

Where this lives today

Every integrated circuit designed this century was signed off by Kirchhoff's rules at industrial scale: SPICE and its descendants assemble exactly this article's junction and loop equations for networks of billions of transistors and solve them millions of times over as signals evolve — the two-loop worked example above, executed by machine, with not one idea added. The same bookkeeping runs at continental scale: national grid load-flow studies are Kirchhoff's equations on 400 kV branches hundreds of kilometres long. And the rules' most elegant offspring is just ahead — the Wheatstone Bridge, a diamond of four resistors that turns the loop rule into a precision instrument by arranging a galvanometer to read exactly zero.

The physical picture

Picture the circuit as a road network for charge. Junctions are the roundabouts: traffic flows through them, but no roundabout stores cars — whatever drives in per second drives out per second. That is the junction rule, and it is not an approximation: if charge ever piled up at a node, the pile would repel every newcomer and choke itself off within a nanosecond. A steady circuit has finished that argument — in equals out, at every node, always. The loop rule is an altitude statement. In a steady circuit every point has one definite potential — a single number, like the elevation of a village on a hillside. You met this landscape in Electric Potential & Potential Difference, and EMF, Terminal Voltage & Internal Resistance told you what batteries do to it: they are the ski lifts, raising charge by $\varepsilon$ ; resistors are the downhill runs, dropping it by $IR$ . Walk any closed tour of the landscape and your net change of altitude is exactly zero — climbs and descents must cancel, whatever route you take. That cancellation, written as an equation per loop, is the loop rule.

Figure

The circuit that defeats series and parallel thinking — two batteries, three branches, and no shortcut works

Look at that circuit honestly. Is the $2\ \Omega$ in series with the $6\ \Omega$ ? No — junction A sits between them, stealing current into the right branch. Are the $2\ \Omega$ and $3\ \Omega$ in parallel? No — their far ends connect through different batteries. The reduction vocabulary of Resistors in Series & Parallel cannot even begin this network. But count what the rules provide: three unknown currents, one junction equation ( $I_1 + I_2 = I_3$ at A — junction B repeats it), two independent loops. Three equations, three unknowns, one guaranteed answer — and this exact circuit is solved end to end below.

The two rules, stated for work

The junction rule is pure accounting at a point. At any junction, at every instant,

\sum I_{\text{in}} = \sum I_{\text{out}} \qquad (1)

where: each $I$ is a branch current, counted the way its arrow points. Physically this is charge conservation plus a steady circuit's refusal to store charge at a node. Notice what the rule does not need: no resistor or battery values. If 3 A and 5 A flow in, 8 A flows out — end of discussion. That indifference makes junction statements the fastest marks in any multi-branch problem. The loop rule is accounting around a tour. Choose any closed loop in the circuit, choose a direction to walk it, and add up the potential changes across everything you cross:

\sum_{\text{loop}} \Delta V = 0 \qquad (2)

where: each $\Delta V$ carries its own sign, set by how you cross the element — the algorithm below. The physical content is that potential is a single number at each point: complete a lap and you are back at your starting altitude, so gains and drops cancel exactly. A charge that lapped the loop and came back with extra energy would be a free-energy machine; one that came back short would have leaked energy to nowhere. Neither exists, and equation (2) is just that sentence with subscripts. One honest note for A-Level and beyond: the loop rule as stated holds for the steady battery-and-resistor world of this chapter. In Electromagnetic Induction you will meet loops threaded by changing magnetic fields, which can hand a lap energy — the rule then gains a term rather than breaking. Here, zero means zero.

The sign algorithm

Every mark lost on Kirchhoff problems is lost here, so fix the procedure once. Arrow every branch current — any directions you like. Choose a walking direction per loop. Then, as you walk:

Crossing a resistor. Walking with the current arrow: the potential drops, write $-IR$ . Walking against the arrow: you are climbing the drop, write $+IR$ .

Crossing a battery. Entering at the $-$ terminal and leaving at the $+$ : you rode the lift up, write $+\varepsilon$ . Entering at $+$ and leaving at $-$ : write $-\varepsilon$ . The battery's sign is decided by which terminal you enter — never by which way the current flows through it. (A battery being charged still lifts the potential from its $-$ plate to its $+$ plate; that is exactly how the rules detect charging, as the worked example shows.) If the battery has internal resistance $r$ , it also contributes a resistor term $\mp Ir$ by the resistor rule, exactly as EMF, Terminal Voltage & Internal Resistance promised.

Figure

The sign algorithm — four crossings, four signs, decided by direction of walk alone

How many equations, and which ones

A network needs as many equations as it has unknown currents, and the rules supply exactly that many — provided you do not double-count. Junctions: of $J$ junctions, only $J - 1$ give independent equations; the last repeats the others (in the hero circuit, junction B is A's statement read upside down). Loops: keep choosing loops until every branch has been walked, each new loop crossing at least one branch no earlier loop used. For the two-junction, three-branch circuits that dominate board and JEE papers the recipe is fixed: one junction equation, two loops, three unknowns. Done.

Derivation AThe two rules from two conservation laws

We want both rules as consequences of physics you already own — not as new postulates to memorise. The junction rule should fall out of conservation of charge, and the loop rule out of conservation of energy applied per coulomb, using only the meanings of current and potential difference from earlier modules. The strategy is to follow first the charge, then the energy, through one second of the circuit's steady life. Watch how "steady" does the real work in both arguments: it is the steadiness that forbids storage at a node and guarantees a well-defined altitude at each point.

Step 1. Take any junction and any one-second interval. Currents $I_1, I_2$ flow in; currents $I_3, I_4$ flow out. In that second, charge $I_1 + I_2$ (in coulombs) arrives and charge $I_3 + I_4$ departs.

Step 2. Suppose they were unequal — say more arrives than leaves. The junction accumulates positive charge, coulomb upon coulomb, every second. That growing charge repels the positive current flowing toward it, throttling the arrivals; in any real conductor this self-correction completes in about a nanosecond. A steady circuit is by definition one where all such transients are over. Why this matters for the rule: steady means no accumulation anywhere, so arrivals equal departures exactly — which is equation (1). The rule is not approximately true; it is what "steady" means at a node.

Step 3. Now follow the energy. Take one coulomb of charge and escort it around any closed loop, back to its starting point. Potential difference is energy per coulomb — Electric Potential & Potential Difference — so across each element the coulomb's energy changes by exactly the $\Delta V$ of that element: up $\varepsilon$ joules through a battery's chemistry, down $IR$ joules through each resistor (delivered to the lattice as heat).

Step 4. The coulomb returns to its starting point, and the starting point has one definite potential — the same number as when the tour began, because a steady circuit's potential landscape does not shift under your feet. Same point, same potential, same potential energy: the coulomb's net energy change over the lap is zero. Therefore the sum of the signed $\Delta V$ terms around the loop is zero — equation (2). Why energy conservation is the right name: if the signed sum came out positive, each lap would mint energy from nothing; negative, and energy would vanish untraceably. The loop rule is the accounting identity that forbids both. *Verified: applying Step 3–4 to a single loop — battery ε with internal resistance r driving resistor R — gives ε − Ir − IR = 0, i.e. I = ε/(R+r), exactly the circuit equation established in EMF, Terminal Voltage & Internal Resistance.*

Limiting cases

case	result	meaning
circuit reducible by series/parallel	rules reproduce the shortcuts	the rules of Resistors in Series & Parallel are equations (1) and (2) pre-solved for simple wiring
one loop, one battery	$I = \varepsilon/(R + r)$	the loop rule alone recovers the basic circuit equation
a guessed arrow is wrong	that current solves negative	flip the interpretation, never the algebra
two junctions	only 1 independent junction equation	the second junction repeats the first
all resistors doubled, all emfs doubled	every current unchanged	the equations are linear — scaling both sides cancels
bridge-shaped network	no series/parallel reduction exists	full Kirchhoff machinery required — the Wheatstone Bridge module turns this case into an instrument

Common trapA negative current means the analysis went wrong

It means precisely the opposite — the analysis worked and is handing you information. The arrows you draw before solving are bookkeeping labels, not physical claims; the algebra never cared which way you guessed. A current that solves to $-0.33$ A is a complete, correct answer: a current of $0.33$ A flowing opposite to your arrow. Do not go back and re-solve with the arrow flipped (you will get $+0.33$ A and learn nothing new), and never drop the sign silently. The real errors live elsewhere: changing your walking direction halfway around a loop, writing junction equations at every junction (one is always redundant), or setting a battery's sign from the current direction instead of from the terminal you enter. Sign discipline in, correct physics out.

Worked examples

Board level — one loop, two batteries facing off (CBSE/ISC pattern, 3 marks)

A battery of emf $6.0\ \mathrm{V}$ and internal resistance $0.5\ \Omega$ is connected in a single loop with a second battery of emf $2.0\ \mathrm{V}$ and internal resistance $0.5\ \Omega$ , so that the two emfs oppose each other, along with an external resistor $R = 3.0\ \Omega$ . Find (a) the current, (b) the terminal voltage of each battery, and (c) which battery is being charged.

Step 1. Draw the current arrow the way the stronger emf pushes — out of the $+$ terminal of the 6.0 V battery — and walk the loop the same way. Crossing the 6.0 V battery from $-$ to $+$ : $+6.0$ , then its internal resistance: $-0.5I$ . Crossing the 2.0 V battery from $+$ to $-$ (its emf opposes the walk): $-2.0$ , and its internal resistance: $-0.5I$ . Crossing $R$ with the current: $-3.0I$ .

Step 2. The loop rule: $6.0 - 0.5I - 2.0 - 0.5I - 3.0I = 0$ , so $4.0 = 4.0I$ and $\boxed{I = 1.0\ \mathrm{A}}$ .

Step 3. Terminal voltages from $V = \varepsilon - Ir$ for the discharging battery: $V_1 = 6.0 - 1.0 \times 0.5 = \boxed{5.5\ \mathrm{V}}$ . For the 2.0 V battery the current is driven into its $+$ terminal, so it gains: $V_2 = 2.0 + 1.0 \times 0.5 = \boxed{2.5\ \mathrm{V}}$ .

Step 4. Part (c): current enters the 2.0 V battery at its $+$ terminal — charge is being pushed backward through it, so it is absorbing energy: $\boxed{\text{the 2.0 V battery is being charged}}$ . This single loop is every phone charger and car alternator in miniature: a stronger source deliberately overpowering a weaker one through a controlled resistance.

Step 5. The sentence that earns the method mark: "Applying Kirchhoff's loop rule around the circuit and taking emf signs from the terminals crossed, the net emf $4.0$ V drives the current through the total resistance $4.0\ \Omega$ ."

Competitive level — the two-loop, two-battery network (JEE-Main pattern, ~4 minutes)

In the circuit of the hero figure, a $12\ \mathrm{V}$ battery (negligible internal resistance) in the left branch is in series with $2\ \Omega$ ; a $6\ \mathrm{V}$ battery in the right branch is in series with $3\ \Omega$ ; both feed junction A at the top, and a $6\ \Omega$ resistor connects A down to B. Find all three currents and verify the energy account.

Step 1. Name and guess. Let $I_1$ flow toward A through the left branch, $I_2$ toward A through the right branch, and $I_3$ down the middle. Junction A: $I_3 = I_1 + I_2$ — the only independent junction statement.

Step 2. Left loop, walked clockwise from B: through the 12 V battery from $-$ to $+$ ( $+12$ ), through $2\ \Omega$ with $I_1$ ( $-2I_1$ ), down the middle with $I_3$ ( $-6I_3$ ). Loop rule: $12 - 2I_1 - 6(I_1 + I_2) = 0$ , i.e. $8I_1 + 6I_2 = 12$ .

Step 3. Right loop, walked anticlockwise from B: through the 6 V battery from $-$ to $+$ ( $+6$ ), through $3\ \Omega$ with $I_2$ ( $-3I_2$ ), down the middle with $I_3$ ( $-6I_3$ ). Loop rule: $6 - 3I_2 - 6(I_1 + I_2) = 0$ , i.e. $6I_1 + 9I_2 = 6$ .

Step 4. Solve the pair. From the first equation $I_1 = (12 - 6I_2)/8 = 1.5 - 0.75I_2$ ; substituting: $6(1.5 - 0.75I_2) + 9I_2 = 6$ gives $9 + 4.5I_2 = 6$ , so $I_2 = -\tfrac{2}{3}\ \mathrm{A}$ and then $I_1 = 2.0\ \mathrm{A}$ , $I_3 = I_1 + I_2 = \tfrac{4}{3} \approx 1.33\ \mathrm{A}$ .

\boxed{I_1 = 2.0\ \mathrm{A}, \qquad I_2 = -0.67\ \mathrm{A}, \qquad I_3 = 1.33\ \mathrm{A}}

Step 5. Read the minus sign: $I_2$ really flows away from A — the 12 V battery is so dominant that it drives $0.67$ A backward through the 6 V battery, charging it. No re-solving needed; the sign is the answer to "what is really happening in the right branch."

Step 6. Audit with the power ledger. The 12 V battery delivers $12 \times 2.0 = 24.0$ W. The 6 V battery absorbs $6 \times 0.67 = 4.0$ W (being charged). The resistors dissipate $2(2.0)^2 + 3(0.67)^2 + 6(1.33)^2 = 8.0 + 1.33 + 10.67 = 20.0$ W. Ledger: $24.0 = 4.0 + 20.0$ — exact. When the ledger balances, every sign in your solution is right. Timing note: about 4 minutes at full writing; under 3 with practice. The junction substitution $I_3 = I_1 + I_2$ made at the start keeps it a 2×2 system — never carry three unknowns into the algebra when one junction equation kills the third immediately. In MCQ format, compute only the asked current, then spend 15 seconds on a junction check.

Exam traps

Renegotiating the walk. Choosing clockwise and then flipping direction halfway around the loop, usually at a battery "because the current goes the other way." The walk direction is yours; the circuit does not vote. Decide once, cross everything in that direction, and let the signs fall where the algorithm says.

Battery sign from the current. The emf term's sign comes from which terminal you enter, full stop. A battery being charged still contributes $+\varepsilon$ when you enter at its $-$ terminal. Mixing this up reverses the answer's sign pattern and is the single most common Kirchhoff error on CBSE and JEE papers alike.

Using all the junction equations. Writing balance statements at every junction produces a redundant system — one equation is always the sum of the others, and you will "discover" $0 = 0$ mid-solve. Use $J - 1$ junctions, then move to loops.

Forgetting internal resistance. Real batteries carry their own $-Ir$ term inside the loop walk. A question that quotes $\varepsilon$ and $r$ expects both terms; dropping $r$ shifts every current. (The full doctrine lives in EMF, Terminal Voltage & Internal Resistance.)

Dropping the minus sign in the final answer. If $I_2 = -0.67$ A, the exam answer is "0.67 A, directed from A toward the 6 V battery" (or equivalent words). Writing " $0.67$ A" with no direction — or worse, quietly writing $+0.67$ — loses the interpretation mark that the question was built to test.

Fraction fright. Two-loop systems love answers like $\tfrac{4}{3}$ A. Keep values as fractions until the final line; premature decimals ( $0.66$ , $1.3$ ) drift the power ledger and can land you between MCQ options.

Board scope notes. IGCSE (Extended) examines the junction rule only — currents at a node — not loop analysis; if that is your board, master equation (1) and the parallel-circuit logic it powers. AP Physics 2 tests both rules conceptually and in two-loop numericals. JEE Advanced expects fluent multi-loop algebra with internal resistances included. No special constants are needed anywhere in this module.

Origins · 1 minGustav Kirchhoff and how this was discovered

Gustav Kirchhoff

1824–1887 · Königsberg, Heidelberg, Berlin

Kirchhoff wrote these rules in 1845 as a 21-year-old student in Franz Neumann's seminar at Königsberg — set the exercise of extending Ohm's law to branched networks, he returned with the two laws that would govern every circuit built since. Two years later he proved how many independent equations any network supplies, seeding what became graph theory. The rules were his student work; his fame came at Heidelberg, where with Bunsen he founded spectroscopy, read the Sun's chemistry from its light, and posed the blackbody problem whose answer started quantum theory. Nothing aged better than the student-day sentences: every chip simulation today is those sentences, applied a billion nodes at a time.

History

1827

— Ohm publishes the law of the single conductor; branched networks remain unsolvable

1845

— Kirchhoff, aged 21, states the junction and loop rules in Neumann's Königsberg seminar

1847

— Kirchhoff's network paper counts the independent loops of any circuit, seeding graph theory

1873

— Maxwell's Treatise systematises loop analysis — solve whole meshes at once

1973

— SPICE at Berkeley automates Kirchhoff's bookkeeping; circuit simulation becomes universal engineering practice

Going further

Texts. NCERT Physics Class 12, Ch. 3 (Current Electricity — Kirchhoff's rules and the two-cell networks) · HC Verma, Concepts of Physics Vol. 2, Ch. 32 · Halliday, Resnick & Walker, Ch. 27 (multiloop circuits) · CIE 9702 Topic 10 (Kirchhoff's laws).

Provenance. All results are standard. A verifier should recompute: the board example $I = (6.0 - 2.0)/(0.5 + 0.5 + 3.0) = 1.0$ A, $V_1 = 5.5$ V, $V_2 = 2.5$ V; the flagship network $8I_1 + 6I_2 = 12$ , $6I_1 + 9I_2 = 6$ giving $I_1 = 2.0$ A, $I_2 = -2/3$ A, $I_3 = 4/3$ A; and the ledger $24.0 = 4.0 + 20.0$ W with $2(2)^2 + 3(2/3)^2 + 6(4/3)^2 = 20.0$ W exactly.

Prerequisites. Resistors in Series & Parallel · EMF, Terminal Voltage & Internal Resistance · Electric Potential & Potential Difference. Leads to. The Wheatstone Bridge (the rules' precision payoff) · Multi-Loop & Two-Battery Circuits (the drill gym, coming up in E3) · Sign Discipline in V, U & W (the E2 technique this module's algorithm extends) · Kirchhoff's Rules (UG deep dive).

How do I know which direction to draw the current arrows before solving?

You do not, and it genuinely does not matter. The arrows only define what "positive" means for each branch — they are bookkeeping labels, not predictions. Draw anything, apply the sign algorithm consistently, and the algebra reports back: a positive answer means your arrow was right, a negative one means the real flow is opposite. Experienced solvers guess plausibly (out of the biggest battery's positive terminal) just to keep minus signs rare, but their final physics is identical to yours. What does matter is never changing an arrow after the equations are written.

Why is one junction equation always redundant?

Because the circuit as a whole conserves charge. Every branch current flows out of one junction and into another, so if you add up the balance statements at all the junctions, each current appears once with a plus sign and once with a minus — the grand total is $0 = 0$ . Any one junction's statement is therefore already implied by all the others. That is why a circuit with $J$ junctions yields only $J - 1$ useful equations, and why writing them all gives you a system that mysteriously refuses to solve.

Is the loop rule just energy conservation, or something more?

For this chapter, it is exactly energy conservation, per coulomb. A coulomb making a full lap returns to a point whose potential has not changed, so its net energy change is zero — every joule the battery gave it was spent in the resistors. The subtlety arrives only later: a changing magnetic field threading the loop can hand the coulomb energy during the lap (that is electromagnetic induction, chapter E6), and the rule then gains a term rather than failing. In the steady, battery-and-resistor world of this chapter, the sum is zero, always and exactly.

When should I use Kirchhoff's rules instead of series and parallel formulas?

Use the shortcuts whenever they apply — they are faster and less error-prone. They apply when the network reduces step by step: every pair of elements you merge is genuinely in series (same current, no junction between) or in parallel (same two nodes). The moment a circuit has batteries in different loops, or a resistor bridging across branches, no such reduction exists, and Kirchhoff is not the fancy option but the only option. A good habit: spend ten seconds attempting a reduction; if any element resists classification, go straight to arrows and loops without regret.

My two loop equations turned out to be multiples of each other. What went wrong?

You walked the same information twice. Two loops are independent only if the second crosses at least one branch the first never touched. In a two-window circuit the safe choices are the two windows themselves (or one window plus the outer boundary — but never all three, since the outer loop is the two windows added together). If your system collapses to one equation, pick a replacement loop through the unused branch, and check you have not used both junction statements of a two-junction circuit — that redundancy produces the same symptom.

Does a battery being charged break the loop rule somehow?

Not at all — the rules are how you detect charging. A battery absorbs energy whenever the solved current enters at its positive terminal: the network is pushing charge backward through its chemistry. Its emf still enters your loop equation by the terminal rule ( $+\varepsilon$ entering at $-$ , $-\varepsilon$ entering at $+$ ), unchanged. In the power ledger it appears on the absorbing side, taking $\varepsilon I$ watts out of the circuit and into stored chemical energy. Every charger you own is a Kirchhoff problem whose "battery" term absorbs — the worked examples above contain two of them.

Is there a fast way to check my currents without re-solving everything?

Two audits, thirty seconds total. First, the junction check: your solved currents must balance at every junction, including the ones you never used — if they do not, an algebra slip exists. Second, the power ledger: total power delivered by discharging batteries ( $\varepsilon I$ each) must exactly equal total power dissipated in resistors ( $I^2R$ each) plus power absorbed by any charging batteries. The ledger is merciless — it balances only when every current and every sign is simultaneously right, which is precisely what makes it the perfect final line of a long-answer solution.