DEV Community

GridCircuits
GridCircuits

Posted on

Reflections on a Transmission Line: Why an Open Far End Sees Double Voltage

Close a breaker onto a long transmission line and the voltage does not appear everywhere at once. It travels. A step wave runs down the line, reaches the far end, and then — depending on what it finds there — either stops, turns back, or doubles up on itself. That last case is the one that catches people out.

Surge impedance is not series impedance

A line has resistance and inductance in series, and conductance and capacitance in parallel, spread along its whole length. For a travelling wave, though, what matters is the ratio of voltage to current in the wave itself. That ratio is the characteristic impedance, also called the surge impedance for the lossless case.

For a lossless line it is simply:

Z0 = sqrt(L' / C')

where L' and C' are the inductance and capacitance per unit length. It is not the series impedance of the line, and it does not grow with length. It is a property of the line's construction: how much inductance and capacitance each kilometre holds, and how the wave shares its energy between them.

The wave speed follows from the same two quantities, velocity = 1 / sqrt(L' C'), so the one-way travel time is tau = length * sqrt(L' C'). Until that time has passed, the far end has no idea the breaker closed.

A worked example with real numbers

Take the example from the GridCircuits calculator: L' = 1 mH/km, C' = 0.01 uF/km, and a line length of 100 km.

The lossless surge impedance works out to 316.23 ohms, and the one-way travel time is 0.31623 ms. That is a third of a millisecond for the wave to cross the line — short in everyday terms, long enough to matter when a breaker closes or a fault clears.

Now put a 500 ohm resistive load on the far end of that same line. The load reflection coefficient is:

Gamma_L = (R_L - Z0) / (R_L + Z0)

With R_L = 500 ohms and Z0 = 316.23 ohms, Gamma_L = 0.22515. Part of the wave is absorbed, part comes back. With a 1,000 V source step and a 100 ohm source resistance, the calculator gives a first load step of 930.8026 V when the wave first arrives at t = tau, and a further increment of -108.8696 V at t = 3 tau, after the reflected wave has been back to the source and returned. The source reflection coefficient in that case is -0.5195, so the wave keeps bouncing, smaller each time, until it settles.

You can run the same numbers yourself with the GridCircuits transmission line calculator, changing the load or the source resistance to see how the steps move.

Three endings, three very different answers

The reflection coefficient is what decides the far end's fate. It runs from -1 to +1, and the three ends of that range are worth remembering.

An open end: Gamma = +1. No current can flow out of an open terminal, so the reflected wave has to carry a current that cancels the incident current. To do that while keeping the same surge impedance, its voltage adds to the incident voltage. The far-end voltage doubles: incident plus reflected, twice the arriving step. Nothing has gone wrong. The line is doing exactly what the boundary condition asks of it.

A shorted end: Gamma = -1. Now the voltage at the terminal has to stay at zero. The reflected wave cancels the incident voltage, so the far-end voltage collapses to nothing. The current, on the other hand, doubles. Same physics, opposite constraint.

A matched load: Gamma = 0. When the load resistance equals the surge impedance, R_L = Z0, the wave sees no change at the boundary. All of its energy goes into the load. There is no reflected wave, no second step, and no ringing between the ends. That is the quiet case — and the reason matching matters wherever a clean step is wanted.

The 500 ohm load above sits between the matched and open cases: Gamma_L = 0.22515, so the far end rises above the incident wave, but well short of doubling.

Why this matters for insulation and switching

Switching a line in or out is not a steady-state event. Energizing an unloaded line, clearing a fault, or reclosing onto a line that is still carrying trapped charge all launch travelling waves. If the far end is open — a line energized from one end with the far breaker open, for example — the voltage there can approach twice the incident step.

Insulation does not care that the doubling lasts only until the reflected wave returns to the source. A brief overvoltage can still stress bushings, transformers, and cable terminations, and it can still flash over a weak point. That is why insulation coordination looks at transient peaks, not just the normal operating voltage, and why switching surges get their own place in planning alongside lightning.

The same bouncing explains why the details at the sending end matter too. A reflected wave that reaches a source with a low resistance reflects again, inverted or not, and heads back down the line. Each round trip takes 2 tau — 0.63246 ms on the 100 km line above — and each pass adds another step at the load. Pre-insertion resistors, controlled switching, and surge arresters are all ways of taming that sequence rather than hoping the peaks stay low.

What to take away

Think of a transmission line as a wave guide first and a pair of terminals second. The surge impedance sets the voltage-to-current ratio of the travelling wave. The load sets the reflection coefficient. The open end doubles the voltage, the shorted end doubles the current, and the matched load absorbs the wave cleanly.

Run the numbers for your own line before you guess at them. A 100 km line with 1 mH/km and 0.01 uF/km gives 316.23 ohms and 0.31623 ms — and from there, one formula for Gamma tells you most of what the far end is about to see.

GridCircuits builds free engineering calculators for preliminary design, estimating, and education. Results like the ones above are a starting point: verify them against the governing standards and the equipment ratings before using them in a design.

Top comments (0)