Electromagnetism

The stride that spans a lightning strike

When lightning strikes open ground, thirty thousand amperes flow out through the soil, and the ground for tens of metres around rises to tens of thousands of volts. Standing in that potential is harmless. Standing across it is not: two feet a stride apart touch two different voltages, and the difference drives a current up one leg and down the other. A cow touches the ground at two points a metre and a half apart, with its heart between them, which is why a single strike can kill a herd that was nowhere near the tree it hit. In dry ground the voltage rises and the current does not.

Assumes: One number for every point, and nothing at all is lost · The resistance that is a length

On 26 August 2016, in a single thunderstorm on the Hardangervidda plateau in Norway, 323 wild reindeer were killed together. The photographs show them lying where they stood, packed close in a herd, and no single animal looks as if it was hit. That is because almost none of them was. Lightning struck the ground nearby, and the current spreading out from the point it struck passed through their legs.

The mechanism has a name in electrical engineering, step voltage, and it is a consequence of the simplest thing a potential can do. One number for every point introduced the electric potential as a single number attached to each place, with the field as its slope. Ground carrying a current has a potential that varies from place to place, and anything touching the ground at two places is connected across the difference. The figures here take one lightning stroke, thirty thousand amperes, into uniform soil, and follow the current out, up a leg and down the other.

Where the current goes

How a cloud comes to hold that charge is a separate story, and the water drops that charge themselves to a spark built a small version of it; what matters here is only where the charge goes when it arrives. A stroke of lightning delivers its charge to the ground in tens of microseconds, with a peak current that for a typical first stroke is around thirty kiloamperes. Once that current reaches the soil it has to go somewhere, and in uniform ground with nothing nearby to prefer, it goes out in every downward direction equally. At a distance rr from the point of entry the current is crossing a hemisphere of area 2πr22\pi r^2, so the current density is I/2πr2I/2\pi r^2, and by Ohm’s law in a material of resistivity ρ\rho the field is ρ\rho times that. The potential that has this field as its slope is

V(r)=ρI2πr.V(r) = \frac{\rho I}{2\pi r}.

This is the potential of a point charge, written in amperes and ohm-metres instead of coulombs and farads, and the analogy is exact: current spreading through a uniform conductor obeys the same equation as an electric field spreading through empty space. Steady current in a uniform conductor has no sources or sinks except at its electrodes, so its potential satisfies Laplace’s equation, and the potential is where the wanderers stop gave that equation a reading that fits the ground well: the potential at any point is the average of the boundary values that random walkers starting there would reach. A walker released ten metres from the strike is far more likely to wander off to the distant, zero-potential earth than to find its way back to the point of entry, and the probability that it does find its way back falls as one over the distance. It is the same mathematics the inside of a conductor described from the other side, where a conductor carries no current and its potential is uniform. Here the conductor is poor, and it is carrying an enormous current, and the potential falls steeply across it.

Where a lightning current goes when it reaches the ground. A section through uniform ground of resistivity 100 Ω·m below a 30 kA strike. The current spreads radially (grey lines, equal current between each pair), so the potential falls as 1/r and the surfaces of equal potential are hemispheres: 200 kV at 2.4 m, 100 kV at 4.8 m, 50 kV at 9.5 m, 25 kV at 19.1 m. Anything touching the surface at two points spans whatever potential difference lies between them: 3.1 kV across a person's stride and 6.2 kV under a cow, both ten metres from the strike.
Fig. 1 A section through uniform ground of resistivity 100 Ω·m below a 30 kA strike. The current goes out radially, equal amounts between each pair of grey lines, and the equipotentials are hemispheres: 200 kV at 2.4 m, 100 kV at 4.8 m, 50 kV at 9.5 m, 25 kV at 19 m. A person’s stride at ten metres and a cow’s length at the same distance each touch two different hemispheres.

The numbers are large. Soil of 100 ohm-metres is ordinary moist ground, neither marsh nor rock. With thirty kiloamperes going in, the ground a metre from the strike rises to nearly half a million volts and the ground ten metres away to 48 kilovolts, relative to the distant earth. Those potentials last only as long as the current, a few tens of microseconds for the main part of the stroke, but while they last the whole neighbourhood is live.

A body standing on one foot in that neighbourhood rises to the local potential and nothing flows through it. A bird on a high-voltage line is in the same position and comes to no harm. So is a person inside a car or a metal-roofed building, for a stronger reason: the field a ring of wires lets through found that a conducting shell keeps the potential inside it nearly uniform, and a closed metal body gives the current an outside path that never crosses the occupants. The danger comes only from bridging: from touching the ground, or anything else, at two points whose potentials differ.

What a stride spans

The difference between two points a distance ss apart on a line through the strike, at distances rr and r+sr+s, is

ΔV=ρI2π(1r−1r+s)=ρI2π sr(r+s).\Delta V = \frac{\rho I}{2\pi}\left(\frac{1}{r} - \frac{1}{r+s}\right) = \frac{\rho I}{2\pi}\,\frac{s}{r(r+s)}.

Everything about the stance is in ss, and everything about the strike and the ground is in the factor in front.

The potential the ground takes on, and what a stride spans. The potential of the ground's surface against distance from a 30 kA strike into soil of 100 Ω·m, ρI/2πr: 477 kV at a metre, 47.7 kV at ten. The bands mark what two contacts span: a person's 0.7 m stride at 10 m spans 3.1 kV, and a cow's 1.5 m between fore and hind hooves spans more, 3.3 kV, four metres further away. The potential itself is harmless to stand in; only its difference across the body drives a current.
Fig. 2 The ground’s potential against distance from the strike, ρI/2πr\rho I/2\pi r, for 30 kA into 100 Ω·m. The shaded bands are what two contacts span: a person’s 0.7 m stride at 10 m spans 3.1 kV; a cow’s 1.5 m from fore to hind hooves, at 14 m, spans 3.3 kV.

The potential curve is steep near the strike and flattens with distance, and what matters at any point is how much it drops across the width of the stance. A person ten metres away, standing in a normal stride of seventy centimetres along the line to the strike, has 3.1 kilovolts between their feet. A cow at the same distance spans 6.2 kilovolts, and a cow fourteen metres away still spans more than the person at ten. The animal’s advantage in four legs, its stability, is its disadvantage here: it is a long object, permanently standing across the ground, and it cannot narrow its stance.

Step voltage by the width of the stance. The voltage across two contacts a given distance apart, on a line through a 30 kA strike in 100 Ω·m soil, against distance from it. Far from the strike it is proportional to the separation and falls as 1/r²; close in, a wider stance gains less, because its far contact already stands lower. At 10 m feet held together (0.15 m) span 706 V, a stride 3124 V, a cow 6228 V and a person lying along the radius 7283 V. Standing with the feet together is not a superstition; it divides the voltage by the ratio of the separations.
Fig. 3 The voltage between two contacts a given distance apart, on a line through a 30 kA strike in 100 Ω·m soil, against distance. At 10 m: feet together (0.15 m) span 706 V, a stride (0.7 m) 3,124 V, a cow (1.5 m) 6,228 V, a person lying along the radius (1.8 m) 7,283 V. Far from the strike every curve falls as 1/r21/r^2.

Plotted for four stances against distance, the same formula gives the whole of the standard advice about lightning in open country. Far from the strike, where rr is much larger than ss, the step voltage is very nearly ρIs/2πr2\rho I s / 2\pi r^2: proportional to the separation of the contacts and falling as the inverse square of the distance, twice as fast as the potential itself. So the separation is the lever. Feet together, about fifteen centimetres between the centres of the soles, take a fifth of the voltage a stride does. A person lying flat along the line to the strike spans the most of all, and the advice not to lie down in a thunderstorm, which can seem odd given that the point is to be low, comes from exactly this term. Being low reduces the chance of being struck directly; lying down increases what a nearby strike can do.

The inverse-square falloff also explains why a whole herd dies at once. A herd standing together in an open field occupies a disc tens of metres across. Every animal in it is within a few times the danger radius, every animal spans a metre and a half, and the step voltage across each one is set by where it stood rather than by anything it did. Reindeer and cattle huddle in storms, which puts them all within range of one strike.

The resistance of a foot

The voltage across the feet is not the voltage across the body. Between the two lies the ground’s own resistance to letting current into and out of each foot, and this turns out to be the most interesting part of the circuit.

A foot standing on soil touches it over an area roughly like a disc of radius bb, about eight centimetres for a shoe or a bare sole. Current leaving the body through that disc has to spread out into the soil below it, and the spreading is concentrated within a few disc radii of the contact. The resistance that is a length worked out exactly this geometry for two metals touching at a microscopic spot: the resistance of a contact of radius bb is not resistivity times length over area, but resistivity over a length, ρ/2b\rho/2b for a spot with metal on both sides. A foot on the ground has soil on one side only, which doubles the spreading area and halves the resistance:

Rfoot=ρ4b.R_{\text{foot}} = \frac{\rho}{4b}.

For 100 ohm-metre soil and an eight-centimetre foot that is about 310 ohms, and the two feet are in series with the body. A body’s internal resistance from one leg to the other, at the voltages involved, is of order a thousand ohms. So for a person ten metres from the strike, the 3.1 kilovolts across the stride drives a current of

Ilegs=ΔVRbody+2ρ/4b≈31001000+620 A≈1.9 AI_{\text{legs}} = \frac{\Delta V}{R_{\text{body}} + 2\rho/4b} \approx \frac{3100}{1000 + 620}\ \text{A} \approx 1.9\ \text{A}

for the duration of the stroke.

The surprising connection is that the contact resistance and the step voltage are both proportional to ρ\rho. Make the ground drier and more resistive and the step voltage across a stride rises in proportion, but so does the resistance of each foot.

Drier ground raises the voltage and not the current. The current through the legs of a person (0.7 m stride, foot contacts of 8 cm radius) and of a cow (1.5 m, hooves of 5 cm) standing 10 m from a 30 kA strike, against the soil's resistivity, with a body resistance of 1 kΩ. In wet ground the current rises in proportion to the resistivity. In dry ground each contact's own resistance, ρ/4b, rises just as fast as the step voltage does, and the current settles at Isb/πr(r + s): 5.0 A for the person and 6.2 A for the cow, whatever the soil.
Fig. 4 The current through the legs at 10 m from a 30 kA strike against soil resistivity, for a person (0.7 m stride, 8 cm feet) and a cow (1.5 m, 5 cm hooves), with 1 kΩ of body between the contacts. In wet ground it grows in proportion to the resistivity; in dry ground it levels off at Isb/πr(r+s)Isb/\pi r(r+s), 5.0 A and 6.2 A, whatever the soil.

In wet ground, where the feet’s resistance is small next to the body’s, the current through the legs rises in proportion to the resistivity. In dry ground, where the feet dominate, the resistivity cancels from the ratio and the current settles at

Ilegs→ρIs/2πr(r+s)ρ/2b=I s bπ r (r+s),I_{\text{legs}} \to \frac{\rho I s/2\pi r(r+s)}{\rho/2b} = \frac{I\,s\,b}{\pi\,r\,(r+s)},

which contains the stroke current and the geometry and no property of the ground at all. For the person at ten metres it is five amperes. The intuition that dry, poorly conducting ground protects a body is half right: it protects in the sense that it stops the current rising further, and it does nothing to lower it once the feet dominate. Rock, sand and dry turf all leave the legs carrying about the same current, because the body is a small resistor between two contacts whose resistance is set by the same soil that sets the voltage.

A cow’s hooves are smaller than a person’s feet, which helps it a little by raising the contact resistance. Its larger span helps it much less than its hooves help, and in the limit it carries six amperes against the person’s five. In wet ground, where the body’s own resistance dominates, the cow is worse off by the full ratio of the spans.

Where the current goes inside the body

The figures treat a body as a resistor between its contacts, which is right for the current and silent about what the current does. For a person standing on two feet, the current enters one leg, crosses the pelvis and leaves through the other. It does not pass through the chest. That path can burn and can stun the leg muscles into collapse, and people near strikes are often knocked down, but it is not the path that stops a heart.

For a four-legged animal the path is fore to hind, and between the front and rear legs lies the chest. Part of the current crosses the heart. This, rather than any difference in the step voltage, is probably the larger reason livestock die in such numbers: an animal spanning twice the stride also has the one organ that matters directly in the current’s way. Grazing animals huddled under a tree are a special case of a special case, standing in the place most likely to be struck and in a ring at exactly the distances where the ground’s potential gradient is steepest.

The danger radius, for whatever current is dangerous. The distance from a 30 kA strike in 100 Ω·m soil inside which the current through the legs exceeds a chosen value, for a person with a 0.7 m stride and a cow spanning 1.5 m. The threshold is left as the axis because the current a short lightning pulse needs to stop a heart is not well known and depends on the path. Whatever it is, the cow's radius is larger: at 1 A, 14.0 m for the person and 18.2 m for the cow, an area 1.7 times as large. And the cow's current crosses its chest, while a person's goes in at one foot and out at the other.
Fig. 5 The distance from a 30 kA strike in 100 Ω·m soil inside which the current through the legs exceeds a chosen value, for a person and for a cow. At 1 A the radii are 14.0 m and 18.2 m, an area 1.7 times larger for the cow; the ratio holds at every threshold.

How much current, for how long, is enough to stop a heart is the one number this calculation cannot supply. Thresholds for ventricular fibrillation are well established for currents lasting from milliseconds to seconds, and they fall as the duration rises. For pulses of tens of microseconds, the duration of a lightning stroke, the data are thin and come largely from animal experiments, and the effect depends on when in the heartbeat the pulse arrives. So the last figure leaves the threshold as the horizontal axis. Whatever the dangerous current turns out to be, the cow’s danger zone has the larger radius, and because the danger zone is a disc, the cow’s is the larger area by the square of that ratio. For a herd in a field the area is what decides how many animals are inside it.

The same arithmetic at a substation

The step-voltage formula was not worked out for lightning. It was worked out for electrical substations, where a fault on a high-voltage line can drive thousands of amperes into the earthing system for a fraction of a second, raising the ground of the whole site by thousands of volts. Engineers designing earthing grids calculate the step voltage across a one-metre stride and the touch voltage between a hand on a grounded structure and the feet a metre from it, and they are required to keep both below limits set by the fault’s duration. The remedies are the ones the formula suggests. Bury a mesh of conductors so that the whole site rises together and the gradient across the surface is shallow. Spread a layer of crushed rock over the surface, whose high resistivity raises ρ/4b\rho/4b for every foot while the soil beneath sets the voltage.

That second remedy is the saturation argument used on purpose. A layer of gravel of a few thousand ohm-metres on top of ordinary soil leaves the step voltage almost unchanged, because the voltage is set by the current flowing through the deeper, more conductive ground; but it multiplies the contact resistance of every foot standing on it. A layer that makes the feet the dominant resistance in the circuit turns a step voltage of thousands of volts into a leg current of milliamperes. Nature’s dry ground does the same, less reliably, which is why the current a person carries near a strike levels off rather than growing without limit.

What a uniform half-space leaves out

The figures assume ground that is uniform to great depth, a single point of entry, and a steady current. All three are approximations, and each fails in a direction worth knowing.

Real ground is layered. A wet surface over dry rock confines the current to the upper layer, which spreads it in two dimensions rather than three; the potential then falls more slowly than 1/r1/r and the danger extends further. Dry surface over wet subsoil does the opposite. Two layers can still be solved exactly, by the device the charge that has to be somewhere else used for a charge above a metal sheet: the boundary between the layers is replaced by an image current, the image by an image of itself in the surface, and so on, an infinite row of point sources at depths that are multiples of twice the layer’s thickness, each weaker than the last by the reflection factor (ρ2−ρ1)/(ρ2+ρ1)(\rho_2-\rho_1)/(\rho_2+\rho_1). Over rock of ten times the soil’s resistivity that factor is 0.82, the row of images decays slowly, and at distances several times the layer’s thickness the current has nowhere to go but sideways: the surface potential falls logarithmically, as in two dimensions, and the step voltage falls as 1/r1/r rather than 1/r21/r^2, so a thin wet layer over rock carries danger much further than uniform ground of the same surface resistivity. Lightning also often reaches the ground through a tree, a post or a wire fence, and the current then enters along a line or at several points rather than at one. A wire fence is especially dangerous: it carries current a long way from the strike and delivers it at every post, so animals along a fence line can be killed hundreds of metres from where the lightning hit. And the current of a lightning stroke is not steady. It rises in a few microseconds and decays over tens, and at those speeds the ground’s response is not purely resistive. A changing current induces fields that oppose it, and how far a field gets into metal measured the depth at which that opposition takes over: 2ρ/ωμ0\sqrt{2\rho/\omega\mu_0}. For a metal it is millimetres. For soil of 100 ohm-metres at the hundred kilohertz that describes a stroke’s rise, it is sixteen metres, comparable to the distances in every figure here. The fast front of the stroke therefore crowds towards the surface and spreads less deeply than the steady formula assumes, which steepens the gradient near the surface during the first microseconds; the slower tail, which carries most of the charge, is described by the resistive picture. The resistivity of soil falls at high field strengths as the soil breaks down electrically. Near the point of entry the field exceeds the breakdown strength of soil, of order a few hundred kilovolts per metre, and the current reaches out through ionised channels and branching arcs visible as burn marks on a lawn, the ground’s own version of the glow and the leaders that the glow a thin wire wears before it sparks found where a field is concentrated in air. The field at radius rr is ρI/2πr2\rho I/2\pi r^2; for 30 kA in 100 Ω·m it falls to 300 kV per metre at about 1.3 metres, so the arcing zone is a few metres across. Close to the strike, that makes the effective point of entry larger and the near-field gradient smaller than the point-source formula gives.

The body is also not a fixed resistor. Skin resistance falls sharply once a few hundred volts break through it, wet skin and bare feet conduct far better than dry shoes, and a current of amperes heats the tissue it crosses. The thousand-ohm figure is a conventional value, roughly what the internal resistance of the body settles to once skin has broken down, and it should be read as an order of magnitude.

What survives all of that is the shape of the argument: the potential of ground carrying a current falls as one over the distance, a body touching it at two points is connected across the difference, the difference scales with the separation, and the ground’s resistivity enters the current through the legs twice, once in the voltage and once in the contacts.

Still open: what a short pulse does to a heart

The fibrillation thresholds used in electrical-safety standards rest on experiments with currents lasting from ten milliseconds to several seconds, and they have been extrapolated to shorter pulses by assuming that what matters is the charge delivered, or the energy, or some mixture. For lightning-length pulses the extrapolation spans two or three orders of magnitude of duration, and the measurements that would anchor it are few. Whether the dangerous current for a fifty-microsecond pulse through a pair of legs is one ampere or twenty, and how much of it the heart sees in a four-legged animal whose chest lies between the contacts, is not settled. Nor is the fraction of lightning casualties among people that are due to step voltage rather than to a direct strike or a side flash from a nearby object, since most of the evidence comes from survivors’ accounts and from marks on the ground afterwards.

The habit worth carrying away is to ask what a body spans, not what it stands in. A 30 kA strike into 100 Ω·m soil raises the ground ten metres away to 48 kV, which is harmless to stand in, while a 0.7 m stride there spans 3.1 kV and a cow’s 1.5 m spans 6.2 kV; and because each foot’s contact resistance, ρ/4b\rho/4b, grows with the soil’s resistivity exactly as the step voltage does, the current through the legs in dry ground settles at Isb/πr(r+s)Isb/\pi r(r+s), about five amperes for a person, whatever the soil. A potential has no danger in it. Only differences do, and the difference is chosen by the width of the stance.

Part 9 of 9

This essay is one argument about Potential. The others:

What links here

Essays that reach for this one mid-argument — the half of a link its own author cannot write down.

The objects named here

The third axis, after the field and the reading path: the things themselves, and every essay that touches each one.

Contact resistanceCurrent densityElectric potentialEquipotentialLightningResistivity