Risk zones

Risk is assessed for each risk zone and compared with the tolerable risk in that zone; protection is required unless every zone complies.

Risk component breakdown — governing zone

Recommended protection measures

• Install a class I LPS on the structure, including SPDs designed for LPL I for equipotential bonding of the connected lines (Tables B.3 and B.13)

With the recommended measures every risk zone satisfies R <= R_T (projected governing-zone risk 4.51e-06/yr).

Detailed calculation steps

The derivation chain shared by every risk zone, with the substituted values and the clause each step implements — check any step against the standard.

Resolved factors — zone Production hall

Resolved factors — zone Storage

Projected outcome with the recommended measures

The same assessment re-run with the recommended protection measures applied. The projected figures are reproducible from the post-measure factors below, exactly as the unprotected result is from the tables above.

Resolved factors with measures applied — zone Production hall

Resolved factors with measures applied — zone Storage

Air termination

§5.2.1: for all types of air terminals only the real physical dimensions of the metal air-termination systems shall be used for the determination of the volume protected. Every protected volume in this design is computed from the entered physical height alone — no enhanced or extended radius is claimed for any device. §5.2.1 also states that radioactive air terminals are not allowed.

§5.2.1: the individual air-termination rods should be connected together at roof level to ensure current division.

The rolling sphere always touches the roof edge. That is geometry — it holds for every building and nothing placed mid-roof changes it. D.5.2.2.2 makes a positioning adequate when the sphere touches only the ground and the air-termination system, so a conductor on the edge is what makes the contact there a contact with the air termination rather than with the structure.

It protects the edge line only. A conductor lying flush in the roof plane obstructs nothing above it: a sphere coming straight down touches it at one tangent point, and the roof inward of the edge is still the rods' job. Nor is it protected space — an air termination is where the strike is meant to land, not volume that is safe — so the protected volume drawn for this design is the same with the conductor as without it.

Minimum cross-sectional area, Table 7 (§5.6.2) for Copper / tin-plated copper: 50 mm² solid tape or 50 mm² solid round. The Materials section resolves the same table for every LPS conductor, including the stranded option where the table gives one.

The length is the perimeter of the roof outline this design was given, and nothing more: no allowance for laps, joints, fixings, down-conductor connections, upstands or roof plant, and no route around anything on the roof, because no design here knows what is on the roof. It is the geometry to take off from, not a cut length to order.

§5.2.3: lateral impacts of flashes to the side are considered negligible for structures up to 60 m measured from ground level, and this structure is 20 m. However, the same clause states that elements significantly protruding the facade can be endangered (e.g. balconies, cameras, antennas), and NOTE 2 adds that such elements on structures up to 60 m are rarely endangered. No air termination is required on the vertical surfaces at this height — the requirement in §5.2.3 is scoped to structures taller than 60 m — so this is a judgement for the designer, about protruding elements this design does not know about.

D.5.2.3.3 proposes a simple approach for those elements: the positioning is adequate if all parts of the element to be protected are below a surface generated by a straight line at α = 15° to vertical (Figure D.6), with the horizontal width of the protected area limited to w = r/10 = 2 m at class I. α is independent of the class of LPS; only w moves with it. This design states those two quantities and positions nothing on the facade. w is a width limit rather than a facade solution — D.5.2.3.3 offers the method as a supplementary part of protected area, "especially useful just below the roof".

The air termination here stands 23 m above ground, at or above the class I rolling-sphere radius of 20 m — which is where D.5.2.3.3 says this method earns its place. At an air-termination height equal to the radius, the tangent to the arc of the sphere becomes vertical and the width of the area protected by the sphere becomes zero (Figure D.6); the lateral protection angle method then supplies a supplementary part of protected area, w = 2 m wide, which D.5.2.3.3 calls especially useful just below the roof. Above the radius the method stays suitable up to 60 m (Figure D.7).

This design does not lay out the lateral air termination. It carries no facade elevation, no balcony, camera or antenna position, and no facade conductor route — the quantities above size that work, they do not specify it.

The roof at 20 m.

An air-termination conductor runs around the roof edge, following the outline: 44.0 m. §5.2.2.1 requires air-termination components to be located at roof level on corners, exposed points and edges, and §5.2.1 requires the individual rods to be connected together at roof level to ensure current division — the roof edge is where both are satisfied at once, and where practice routes the connection. This is not additional material: it specifies a conductor the design already implied.

At 20 m this structure is at or above the class I rolling-sphere radius of 20 m, so there is no rod height with a roof-edge spacing: a ground-resting sphere reaches the edge from beyond any rod tip standing on the roof. The edge is the perimeter conductor's here, and §5.2.3's flashes to the side are sized separately below.

These rods carry the roof's interior and nothing else. §5.2.2.1 requires air-termination components on corners, exposed points and edges, and the perimeter conductor above passes through every corner and along every edge — so that requirement is met by the conductor, and the rods answer the one question left: bring every roof point within 10.5 m of a rod. Without the conductor the rods would have to hold the edge as well, and there would be 4 of them here instead of 2.

2 of the 4 lattice positions were kept. The selection is computed the same way for the same inputs, so re-running this design gives this layout again — but it is one covering layout, not the only one, and a rod moved to suit the roof is the engineer's call to make against the same criterion.

The layout is proved to cover rather than sampled for coverage. Any uncovered piece of roof would be an open region whose border is made of walls and of arcs of the rods' reach, and such a border can only change from one to the other at a corner of the outline, where a rod's reach crosses a wall, or where two rods' reaches cross. A piece with none of those would have to be ringed by a single unbroken circle, which cannot happen inside a closed outline. So that finite set of places decides the whole roof, and every one of them was tested — no grid is involved and nothing between the tested places is assumed. Two limits, stated because the claim is only worth what they leave: a point exactly at a rod's reach counts as outside it throughout, so a point on the boundary of the protected volume is treated as unprotected rather than protected; and where a tested place is a knife-edge contact rather than a gap, its surroundings are probed at a finite number of points to decide whether real area opens there, so a gap narrower than roughly a quarter-turn at that contact would not be reported. Both are properties of the layout as drawn: move or shorten any rod and neither this proof nor the coverage percentage applies until the design is recomputed. That percentage — the "Roof within a rod's protected radius" row above — is not a second proof. It answers the same question on a grid, by code that shares none with this argument, so where this proof closes it can only agree — read it as a check on the tool, not as further evidence about this roof. It is reported with its spacing because it is the number that says how far short a layout falls in the case this proof does not close.

The rod positions above are a proposed layout, computed from the LPS class, the plan outline and the rod height by the positioning rule of §5.2.2.1 — rods on corners, then edges, then the field, on a square grid no coarser than the spacing one rod's protected radius supports. They are not a survey and not an instruction: the engineer may move them, and the coverage figure is what any alternative should be judged against — but the coverage figure and the covering proof are both about the layout as drawn, and a rod moved, shortened or removed voids them until the design is recomputed against the positions actually adopted.

Natural air terminations are not counted. Metal parapets, copings, handrails, ladders and existing masts commonly protect a roof already, and §5.2.5 allows them; none of them is an input here, so the layout may put a rod where something on the roof already does the job. Check the layout against what is actually up there before pricing it.

The coverage figure is measured and not assumed: 120 roof points on a 1.00 m grid, each tested against the rolling-sphere criterion of §5.2.2 — a point is protected when the sphere resting on it cannot avoid touching the air termination, and every one of them is protected — so it reads 100% with no "at least" in front of it. Elsewhere the figure is a lower bound, because the only obstacles counted are the rods above and anything else on the roof can add protection but never remove it; here there is nothing left above it for a bound to allow for. What it remains is the grid's answer rather than the roof's: whether every point of the roof is covered is a different question, and the "Rod coverage proof" row above is what answers it.

The roof at 12 m.

An air-termination conductor runs around the roof edge, following the outline: 120.0 m. §5.2.2.1 requires air-termination components to be located at roof level on corners, exposed points and edges, and §5.2.1 requires the individual rods to be connected together at roof level to ensure current division — the roof edge is where both are satisfied at once, and where practice routes the connection. This is not additional material: it specifies a conductor the design already implied.

The roof edge is governed by a different figure from the interior. A sphere resting on the ground beside the building reaches the roof edge from 18.3 m away in plan; a rod tip at 15.0 m above ground is reached from 19.4 m. Rods on the edge keep that sphere off it while they are no more than d_edge = 2·√(a₁² − a₀²) = 12.5 m apart — against the 10.5 m that governs the interior. The two are not the same number and neither substitutes for the other.

d_edge is largest at a rod height of exactly R − H = 8.0 m for this class and building, and falls away on both sides of it. Taller is not better at the roof edge: past 16.0 m a rod of that height protects the edge at no spacing at all, because the sphere that clears its tip clears the roof with it.

This roof stands beside a taller part of the same structure — the roof at 20 m — and that part obstructs the rolling sphere over at least 99.9% of it: the sphere rests on that roof's edge, where this design runs its edge conductor (§5.2.2.1), so what it touches there is air termination and not the structure (§5.2.2.2). The obstruction reaches 16.0 m out from the taller plan at this level — the same √(d(2R − d)) that gives a rod its protected radius, with d the height the taller part stands above this roof. Those metres need no rod, and the layout above does not put one there. Nothing about it is an allowance: the covering proof is run against the obstruction and the rods together, and it is the same exact test the rods alone would have faced.

These rods carry the roof's interior and nothing else. §5.2.2.1 requires air-termination components on corners, exposed points and edges, and the perimeter conductor above passes through every corner and along every edge — so that requirement is met by the conductor, and the rods answer the one question left: bring every roof point within 10.5 m of a rod. Without the conductor the rods would have to hold the edge as well, and there would be 12 of them here instead of 2.

2 of the 12 lattice positions were kept. The selection is computed the same way for the same inputs, so re-running this design gives this layout again — but it is one covering layout, not the only one, and a rod moved to suit the roof is the engineer's call to make against the same criterion.

The layout is proved to cover rather than sampled for coverage. Any uncovered piece of roof would be an open region whose border is made of walls and of arcs of the rods' reach, and such a border can only change from one to the other at a corner of the outline, where a rod's reach crosses a wall, or where two rods' reaches cross. A piece with none of those would have to be ringed by a single unbroken circle, which cannot happen inside a closed outline. So that finite set of places decides the whole roof, and every one of them was tested — no grid is involved and nothing between the tested places is assumed. Two limits, stated because the claim is only worth what they leave: a point exactly at a rod's reach counts as outside it throughout, so a point on the boundary of the protected volume is treated as unprotected rather than protected; and where a tested place is a knife-edge contact rather than a gap, its surroundings are probed at a finite number of points to decide whether real area opens there, so a gap narrower than roughly a quarter-turn at that contact would not be reported. Both are properties of the layout as drawn: move or shorten any rod and neither this proof nor the coverage percentage applies until the design is recomputed. That percentage — the "Roof within a rod's protected radius" row above — is not a second proof. It answers the same question on a grid, by code that shares none with this argument, so where this proof closes it can only agree — read it as a check on the tool, not as further evidence about this roof. It is reported with its spacing because it is the number that says how far short a layout falls in the case this proof does not close.

The rod positions above are a proposed layout, computed from the LPS class, the plan outline and the rod height by the positioning rule of §5.2.2.1 — rods on corners, then edges, then the field, on a square grid no coarser than the spacing one rod's protected radius supports. They are not a survey and not an instruction: the engineer may move them, and the coverage figure is what any alternative should be judged against — but the coverage figure and the covering proof are both about the layout as drawn, and a rod moved, shortened or removed voids them until the design is recomputed against the positions actually adopted.

Natural air terminations are not counted. Metal parapets, copings, handrails, ladders and existing masts commonly protect a roof already, and §5.2.5 allows them; none of them is an input here, so the layout may put a rod where something on the roof already does the job. Check the layout against what is actually up there before pricing it.

The coverage figure is measured and not assumed: 680 roof points on a 1.00 m grid, each tested against the rolling-sphere criterion of §5.2.2 — a point is protected when the sphere resting on it cannot avoid touching the air termination, and every one of them is protected — so it reads 100% with no "at least" in front of it. Elsewhere the figure is a lower bound, because the only obstacles counted are the rods above and anything else on the roof can add protection but never remove it; here there is nothing left above it for a bound to allow for. What it remains is the grid's answer rather than the roof's: whether every point of the roof is covered is a different question, and the "Rod coverage proof" row above is what answers it.

The roof at 8 m.

An air-termination conductor runs around the roof edge, following the outline: 56.0 m. §5.2.2.1 requires air-termination components to be located at roof level on corners, exposed points and edges, and §5.2.1 requires the individual rods to be connected together at roof level to ensure current division — the roof edge is where both are satisfied at once, and where practice routes the connection. This is not additional material: it specifies a conductor the design already implied.

The roof edge is governed by a different figure from the interior. A sphere resting on the ground beside the building reaches the roof edge from 16.0 m away in plan; a rod tip at 11.0 m above ground is reached from 17.9 m. Rods on the edge keep that sphere off it while they are no more than d_edge = 2·√(a₁² − a₀²) = 15.9 m apart — against the 10.5 m that governs the interior. The two are not the same number and neither substitutes for the other.

d_edge is largest at a rod height of exactly R − H = 12.0 m for this class and building, and falls away on both sides of it. Taller is not better at the roof edge: past 24.0 m a rod of that height protects the edge at no spacing at all, because the sphere that clears its tip clears the roof with it.

This roof stands beside a taller part of the same structure — the roofs at 20 and 12 m — and that part obstructs the rolling sphere over at least 75.0% of it: the sphere rests on that roof's edge, where this design runs its edge conductor (§5.2.2.1), so what it touches there is air termination and not the structure (§5.2.2.2). The obstruction reaches 18.3 m out from the taller plan at this level — the same √(d(2R − d)) that gives a rod its protected radius, with d the height the taller part stands above this roof. Those metres need no rod, and the layout above does not put one there. Nothing about it is an allowance: the covering proof is run against the obstruction and the rods together, and it is the same exact test the rods alone would have faced.

These rods carry the roof's interior and nothing else. §5.2.2.1 requires air-termination components on corners, exposed points and edges, and the perimeter conductor above passes through every corner and along every edge — so that requirement is met by the conductor, and the rods answer the one question left: bring every roof point within 10.5 m of a rod. Without the conductor the rods would have to hold the edge as well, and there would be 6 of them here instead of 2.

2 of the 6 lattice positions were kept. The selection is computed the same way for the same inputs, so re-running this design gives this layout again — but it is one covering layout, not the only one, and a rod moved to suit the roof is the engineer's call to make against the same criterion.

The layout is proved to cover rather than sampled for coverage. Any uncovered piece of roof would be an open region whose border is made of walls and of arcs of the rods' reach, and such a border can only change from one to the other at a corner of the outline, where a rod's reach crosses a wall, or where two rods' reaches cross. A piece with none of those would have to be ringed by a single unbroken circle, which cannot happen inside a closed outline. So that finite set of places decides the whole roof, and every one of them was tested — no grid is involved and nothing between the tested places is assumed. Two limits, stated because the claim is only worth what they leave: a point exactly at a rod's reach counts as outside it throughout, so a point on the boundary of the protected volume is treated as unprotected rather than protected; and where a tested place is a knife-edge contact rather than a gap, its surroundings are probed at a finite number of points to decide whether real area opens there, so a gap narrower than roughly a quarter-turn at that contact would not be reported. Both are properties of the layout as drawn: move or shorten any rod and neither this proof nor the coverage percentage applies until the design is recomputed. That percentage — the "Roof within a rod's protected radius" row above — is not a second proof. It answers the same question on a grid, by code that shares none with this argument, so where this proof closes it can only agree — read it as a check on the tool, not as further evidence about this roof. It is reported with its spacing because it is the number that says how far short a layout falls in the case this proof does not close.

The rod positions above are a proposed layout, computed from the LPS class, the plan outline and the rod height by the positioning rule of §5.2.2.1 — rods on corners, then edges, then the field, on a square grid no coarser than the spacing one rod's protected radius supports. They are not a survey and not an instruction: the engineer may move them, and the coverage figure is what any alternative should be judged against — but the coverage figure and the covering proof are both about the layout as drawn, and a rod moved, shortened or removed voids them until the design is recomputed against the positions actually adopted.

Natural air terminations are not counted. Metal parapets, copings, handrails, ladders and existing masts commonly protect a roof already, and §5.2.5 allows them; none of them is an input here, so the layout may put a rod where something on the roof already does the job. Check the layout against what is actually up there before pricing it.

The coverage figure is measured and not assumed: 192 roof points on a 1.00 m grid, each tested against the rolling-sphere criterion of §5.2.2 — a point is protected when the sphere resting on it cannot avoid touching the air termination, and every one of them is protected — so it reads 100% with no "at least" in front of it. Elsewhere the figure is a lower bound, because the only obstacles counted are the rods above and anything else on the roof can add protection but never remove it; here there is nothing left above it for a bound to allow for. What it remains is the grid's answer rather than the roof's: whether every point of the roof is covered is a different question, and the "Rod coverage proof" row above is what answers it.

The roof at 8 m.

An air-termination conductor runs around the roof edge, following the outline: 45.7 m. §5.2.2.1 requires air-termination components to be located at roof level on corners, exposed points and edges, and §5.2.1 requires the individual rods to be connected together at roof level to ensure current division — the roof edge is where both are satisfied at once, and where practice routes the connection. This is not additional material: it specifies a conductor the design already implied.

The roof edge is governed by a different figure from the interior. A sphere resting on the ground beside the building reaches the roof edge from 16.0 m away in plan; a rod tip at 11.0 m above ground is reached from 17.9 m. Rods on the edge keep that sphere off it while they are no more than d_edge = 2·√(a₁² − a₀²) = 15.9 m apart — against the 10.5 m that governs the interior. The two are not the same number and neither substitutes for the other.

d_edge is largest at a rod height of exactly R − H = 12.0 m for this class and building, and falls away on both sides of it. Taller is not better at the roof edge: past 24.0 m a rod of that height protects the edge at no spacing at all, because the sphere that clears its tip clears the roof with it.

This roof needs no air-termination rods. A taller part of the same structure — the roofs at 20 and 12 m — obstructs the rolling sphere over the whole of it: the sphere rests on that roof's edge, where this design runs its edge conductor (§5.2.2.1), which the sphere method treats exactly as a rod would — proved, not sampled, by the same test every other layout in this design is held to. The perimeter air-termination conductor and the down conductors are unaffected: §5.2.2.1 puts air-termination components on the roof's corners and edges whatever stands over it, and §5.3 brings the current down regardless.

This roof stands beside a taller part of the same structure — the roofs at 20 and 12 m — and that part obstructs the rolling sphere over 100% of it: the sphere rests on that roof's edge, where this design runs its edge conductor (§5.2.2.1), so what it touches there is air termination and not the structure (§5.2.2.2). The obstruction reaches 18.3 m out from the taller plan at this level — the same √(d(2R − d)) that gives a rod its protected radius, with d the height the taller part stands above this roof. Those metres need no rod, and the layout above does not put one there. Nothing about it is an allowance: the covering proof is run against the obstruction and the rods together, and it is the same exact test the rods alone would have faced.

The roof at 8 m.

An air-termination conductor runs around the roof edge, following the outline: 64.0 m. §5.2.2.1 requires air-termination components to be located at roof level on corners, exposed points and edges, and §5.2.1 requires the individual rods to be connected together at roof level to ensure current division — the roof edge is where both are satisfied at once, and where practice routes the connection. This is not additional material: it specifies a conductor the design already implied.

The roof edge is governed by a different figure from the interior. A sphere resting on the ground beside the building reaches the roof edge from 16.0 m away in plan; a rod tip at 11.0 m above ground is reached from 17.9 m. Rods on the edge keep that sphere off it while they are no more than d_edge = 2·√(a₁² − a₀²) = 15.9 m apart — against the 10.5 m that governs the interior. The two are not the same number and neither substitutes for the other.

d_edge is largest at a rod height of exactly R − H = 12.0 m for this class and building, and falls away on both sides of it. Taller is not better at the roof edge: past 24.0 m a rod of that height protects the edge at no spacing at all, because the sphere that clears its tip clears the roof with it.

This roof stands beside a taller part of the same structure — the roofs at 20 and 12 m — and that part obstructs the rolling sphere over at least 76.2% of it: the sphere rests on that roof's edge, where this design runs its edge conductor (§5.2.2.1), so what it touches there is air termination and not the structure (§5.2.2.2). The obstruction reaches 18.3 m out from the taller plan at this level — the same √(d(2R − d)) that gives a rod its protected radius, with d the height the taller part stands above this roof. Those metres need no rod, and the layout above does not put one there. Nothing about it is an allowance: the covering proof is run against the obstruction and the rods together, and it is the same exact test the rods alone would have faced.

These rods carry the roof's interior and nothing else. §5.2.2.1 requires air-termination components on corners, exposed points and edges, and the perimeter conductor above passes through every corner and along every edge — so that requirement is met by the conductor, and the rods answer the one question left: bring every roof point within 10.5 m of a rod. Without the conductor the rods would have to hold the edge as well, and there would be 6 of them here instead of 2.

2 of the 6 lattice positions were kept. The selection is computed the same way for the same inputs, so re-running this design gives this layout again — but it is one covering layout, not the only one, and a rod moved to suit the roof is the engineer's call to make against the same criterion.

The layout is proved to cover rather than sampled for coverage. Any uncovered piece of roof would be an open region whose border is made of walls and of arcs of the rods' reach, and such a border can only change from one to the other at a corner of the outline, where a rod's reach crosses a wall, or where two rods' reaches cross. A piece with none of those would have to be ringed by a single unbroken circle, which cannot happen inside a closed outline. So that finite set of places decides the whole roof, and every one of them was tested — no grid is involved and nothing between the tested places is assumed. Two limits, stated because the claim is only worth what they leave: a point exactly at a rod's reach counts as outside it throughout, so a point on the boundary of the protected volume is treated as unprotected rather than protected; and where a tested place is a knife-edge contact rather than a gap, its surroundings are probed at a finite number of points to decide whether real area opens there, so a gap narrower than roughly a quarter-turn at that contact would not be reported. Both are properties of the layout as drawn: move or shorten any rod and neither this proof nor the coverage percentage applies until the design is recomputed. That percentage — the "Roof within a rod's protected radius" row above — is not a second proof. It answers the same question on a grid, by code that shares none with this argument, so where this proof closes it can only agree — read it as a check on the tool, not as further evidence about this roof. It is reported with its spacing because it is the number that says how far short a layout falls in the case this proof does not close.

The rod positions above are a proposed layout, computed from the LPS class, the plan outline and the rod height by the positioning rule of §5.2.2.1 — rods on corners, then edges, then the field, on a square grid no coarser than the spacing one rod's protected radius supports. They are not a survey and not an instruction: the engineer may move them, and the coverage figure is what any alternative should be judged against — but the coverage figure and the covering proof are both about the layout as drawn, and a rod moved, shortened or removed voids them until the design is recomputed against the positions actually adopted.

Natural air terminations are not counted. Metal parapets, copings, handrails, ladders and existing masts commonly protect a roof already, and §5.2.5 allows them; none of them is an input here, so the layout may put a rod where something on the roof already does the job. Check the layout against what is actually up there before pricing it.

The coverage figure is measured and not assumed: 252 roof points on a 1.00 m grid, each tested against the rolling-sphere criterion of §5.2.2 — a point is protected when the sphere resting on it cannot avoid touching the air termination, and every one of them is protected — so it reads 100% with no "at least" in front of it. Elsewhere the figure is a lower bound, because the only obstacles counted are the rods above and anything else on the roof can add protection but never remove it; here there is nothing left above it for a bound to allow for. What it remains is the grid's answer rather than the roof's: whether every point of the roof is covered is a different question, and the "Rod coverage proof" row above is what answers it.

Down conductors

IEC 62305-3:2024 §5.3.3 requires only that an attached LPS have not fewer than two down conductors. Distributing them around the perimeter is stated as a recommendation, explicitly subject to architectural and practical constraints, and both equal spacing and the Table 5 distances are preferred values rather than limits. The count above is derived from the preferred distance for this class, which is the conservative reading; a layout that departs from it is a departure from a preference, and the design records it as one.

Fit a test joint on each of the 21 down conductors at its connection to the earth termination, openable for measurement and closed in normal use (§5.3.6).

Separation distance

s is a distance at a point, and this design has two worth quoting. §6.3.1 measures l from the point where the separation distance is to be considered, along the air termination and the down conductor. The figure above is for the down-conductor run; at the tip of the rod the path is longer and the current has not yet divided among the down conductors, so k_c = 1 over the rod (Annex B, Figure B.4 a) — the values of k_c are considered from the point of strike). Separate any metal installation by the figure for its own elevation; this design carries no position for any metal installation, so it cannot say which applies where — a proposed rod layout locates the air termination and nothing else. The governing figure below is the larger of the two, because a single distance applied everywhere has to be the worst case.

Every internal metal installation closer than s to a down conductor must be bonded or the distance increased.

This separation distance is not altitude-corrected. §6.3.2 requires a correction for an LPS installed at high elevations, following IEC 60071-2. Site elevation is not an input to this tool and no correction has been applied, so where the site is at elevation, read §6.3.2 against it and adjust s accordingly.

s is the one figure in this design that the finished drawing can correct. Everything else above is fixed by the class and the building; s is computed from the conductor run length and, on the general approach, from the down-conductor and ring spacings — and the values used here for those are stated defaults, not measurements of a layout. Two of them can move s the wrong way: k_c1 grows with the down-conductor spacing c, and rings further apart than assumed make the computed s non-conservative. Once the layout is drawn, re-run this design with the measured figures.

Earth termination

The minimum lengths may be disregarded where an earthing resistance of the Type A arrangement below 10 Ω is achieved — measured at a frequency different from the power frequency and its multiples, to avoid interference (§5.4.2.2).

Ring in contact with the soil for at least 80 % of its length, buried ≥ 0,5 m deep, about 1 m from the external walls (§5.4.3).

Materials & minimum dimensions

All LPS components (conductors, rods, electrodes, clamps) shall meet IEC 62561-2 mechanical/electrical/corrosion requirements (Table 7/8 note a).

Air-termination rods where mechanical stress such as wind loading is not critical may be reduced to 9,5 mm diameter, 1 m long (Table 7 note c).

Copper solid round 50 mm² (8 mm Ø) may be reduced to 28 mm² (6 mm Ø) where mechanical strength is not essential; reduce fastener spacing accordingly (Table 7 note b).

A buried copper earth electrode shall not be used in contact with the structure's steel members, pipes or piles — corrosion (Table 8 note g).

Equipotential bonding (internal LPS)

Install the main bonding bar in the basement or at ground level, accessible for inspection, and connect it to the earth-termination system (§6.2.2 a).

Structure length 56 m exceeds 20 m — use a ring bonding bar or several interconnected bonding bars (§6.2.2 a).

Bond additionally at every point where the separation distance s of the section above cannot be kept (§6.2.2 b). Keep bonding connections as direct and straight as possible.

Live conductors of internal systems are bonded via SPDs — coordinate per IEC 62305-4.

Touch & step voltages

The insulation withstand voltage should be considered in wet conditions (§8.1). Signage shall conform to ISO 3864-1.

Alternative exemption: a measured contact resistance within 3 m of the down conductors of at least 100 kΩ also brings the hazard to a tolerable level. §8.1 c) sets that over the area of a typical footprint (0,02 m²) for touch voltage; §8.2 c) states the 100 kΩ for the soil surface layer without giving an area. A 5 cm layer of insulating material, e.g. asphalt, generally achieves it (§8.1 NOTE 2 / §8.2 NOTE 1).

Inspection & maintenance

Also inspect: during construction (components that become concealed), after LPS installation, after alterations or repairs, and after a known or suspected lightning strike to the structure (§7.4).

Periodic checks: deterioration and corrosion of air-termination elements, conductors and connections; earthing resistance of the earth-termination system; condition of connections, equipotential bonding and fixings (§7.4).

Testing on a 14–15-month cycle is acceptable where earth-resistance readings across different seasons are wanted (D.7.2.2). Consider improving the earthing system when measured resistance rises over successive inspections.

Record all inspection and test results and keep them with the LPS design documentation for the property manager (§7.4).

Calculation trace

Installation notes

• Bond every internal metal installation that cannot keep the separation distance s from LPS conductors.
• Connect the earth termination to the equipotential bonding bar; coordinate SPDs per IEC 62305-4.

This document specifies the surge protection measures for the structure: the lightning protection zones, the earthing and bonding network, and a coordinated SPD system for every incoming service. The SPD specification is manufacturer-neutral — it states the required test class, discharge current and voltage protection level, and can be issued to any supplier.

Lightning protection zones

An LPZ 1 created by a normal external LPS in accordance with IEC 62305-3 has mesh widths and typical distances greater than 5 m, so its spatial shielding effect is negligible (IEC 62305-4:2024, B.14.5). Effective spatial shielding requires a mesh width typically below 5 m.

Equipotential bonding SPDs are always required at the entrance of an LPZ to bond incoming lines connected to internal systems within that zone. Using an interconnected or extended LPZ can reduce the number of SPDs required (IEC 62305-4:2024, 5.5).

IEC 62305-4:2024, 4.3.2 divides the region outside the structure into LPZ 0A and LPZ 0B, and which applies at a given point is decided by the rolling sphere against the real geometry (IEC 62305-3:2024, 5.2.2) — for the roof, by the LPS design; for a service's route outside the structure, by nothing modelled here. This design therefore assumes LPZ 0A at every line entrance, which is the more severe of the two: the threat is the direct flash and the full surge current. Where an air termination is shown to cover a line's approach, LPZ 0B applies instead and this design does not determine that. The assumption changes no SPD specified here — the test class required at the entrance is resolved from C.2.2 and does not depend on it — but the threat the installation is stated to be exposed to does.

This structure has more than one roof, and the protected height above a roof is a property of that roof rather than of the building — so the LPS design states it roof by roof, and there is no single figure for the structure. Whether each externally installed item's exposure was answered at its own position against the air-termination layout, or taken as declared, is stated with that item under B.11. Treat roof-mounted equipment as LPZ 0A wherever its exposure was not answered — one that carries no position, or stands where no rod gives a protected volume to answer it against (IEC 62305-3:2024, 5.2.2).

LPZ 0A — IEC 62305-4:2024, 4.3.2 defines this zone as: "Zone where the threat is due to the direct lightning flash and the full lightning electromagnetic field. The internal systems can be subjected to full lightning surge current."

LPZ 0B — IEC 62305-4:2024, 4.3.2 defines this zone as: "Zone protected against direct lightning flashes but where the threat is the full lightning electromagnetic field. The internal systems can be subjected to partial lightning surge currents."

LPZ 1 — IEC 62305-4:2024, 4.3.3 defines this zone as: "Zone where the surge current is limited by current sharing and isolating interfaces and/or by SPDs at the boundary. Spatial shielding can attenuate the lightning electromagnetic field."

LPZ 2 — IEC 62305-4:2024, 4.3.3 defines LPZ 2…n as: "Zone where the surge current may be further limited by current sharing, isolating interfaces and/or by additional SPDs at the boundary. Additional spatial shielding may be used to further attenuate the lightning electromagnetic field."

Whether the separation distance s of IEC 62305-3 is maintained has not been verified for this installation. Where it is not, electrical lines carrying partial lightning currents shall be treated like down conductors (IEC 62305-4:2024, B.11.3), and where a ring conductor is bonded to each down conductor for this reason, class I tested SPDs are required (B.3). State it on the basis of design once it has been checked.

Earthing and bonding network

A Type B (ring or foundation) arrangement is used, as 5.2 recommends.

Meshed, integrated by multiple bonding points; metal components are not isolated from the earthing system. Preferred where internal systems extend over wide zones or many lines enter at several points.

Do not isolate the internal systems' metal parts from the earthing system; tie them into it at several points instead.

One earthing system serves the whole site. Giving the LPS, the SPDs, the power installation and the telecommunications installation an earth each is not an acceptable arrangement for lightning and surge protection; they must form a single integrated system (IEC 62305-4:2024, 5.2).

Bonding bars shall be installed for bonding of: all conductive services entering an LPZ, directly or via suitable SPDs; the protective earth conductor PE; metal components of the internal systems (cabinets, enclosures, racks); the magnetic shields of the LPZ, at the periphery and inside the structure (IEC 62305-4:2024, 5.4).

Where possible, incoming services should enter the LPZ at the same location and connect to the same bonding bar. Where services enter at different locations, each shall connect to a bonding bar and those bonding bars shall be interconnected — bonding to a ring bonding bar is recommended (IEC 62305-4:2024, 5.5).

Treat the bonding network as a protective path only: it should carry no power or signal return current. On that basis the PE conductor is tied into the bonding network and the PEN conductor is kept out of it, along with any metalwork connected to the PEN. A functional earthing conductor may be bonded straight to the low-impedance network. Keeping the PEN separate is what stops power-frequency current finding a way into the electronic systems (IEC 62305-4:2024, B.7 — Annex B is informative, so these are conditions of following its method rather than requirements standing on their own).

For efficient bonding the following installation rules are important: the basis for all bonding measures is a low impedance bonding network; run each bonding bar back to the earthing system by the most direct route the building allows; SPDs should be installed with the shortest possible connections to the bonding bar and to the live conductors, minimising inductive voltage drops; downstream of an SPD, mutual induction should be minimised by reducing the loop area or by using shielded cables or ducts earthed at both ends (IEC 62305-4:2024, 5.4).

Smaller cross-sections are permitted in some countries where the thermal and mechanical requirements of IEC 62305-1:2024, Annex D are satisfied (IEC 62305-4:2024, Table 1, footnote b). That verification is outside this tool's scope, so the tabulated minimum is applied.

For SPDs in power applications the connecting conductors shall also be dimensioned according to IEC 60364-5-53 and IEC 61643-12 (IEC 62305-4:2024, Table 1, footnote c).

Earthing and bonding are required in every design — there is no case in which they can be left out. Every conductive service must be tied to earth where it enters the structure: directly where that is possible, and otherwise through an equipotential bonding SPD (IEC 62305-4:2024, Clause 4.4).

Coordinated SPD schedule

Cascaded SPDs shall be energy coordinated in accordance with IEC 61643-12 and IEC 61643-22; the SPD manufacturer should provide the information needed to achieve energy and voltage protection level coordination between the devices (IEC 62305-4:2024, C.3.4). This tool states the requirement but does not compute the coordination.

This is a telecommunication or signal line, so its discharge currents are not taken from Annex C: Tables C.3, C.4 and C.5 are keyed on the supply system and its modes of protection, and describe power lines only. IEC 62305-4:2024, C.2.2 provides the required current for a signal line in IEC 62305-2:2024, Annex B (Table B.8), and classifies the device under IEC 61643-21 — class D1 where a power SPD would be class I tested, and class C2 where it would be class II. Each source of damage carries its own waveform, so a device at this boundary has both a 10/350 and an 8/20 duty.

These currents are per line conductor, not the total through the cable (IEC 62305-1:2024, Table E.2: "All values refer to each line conductor"). A multi-core cable carries the sum, so a device protecting n cores must carry n times the figure above in total, and manufacturers' data sheets state both a per-line and a total rating. Check which one you are reading before comparing.

The figures above are those IEC 62305-2:2024, Table B.8 gives for overhead unshielded lines, which is its worst case. Its NOTE 3 permits the currents to be halved for a buried line, and halved for a shielded line. This design applies neither reduction, so the requirement stated here is conservative for a line that is buried, shielded, or both.

No voltage protection level is stated for this signal line. Every U_p ceiling in this schedule is a fraction of the equipment's rated impulse withstand U_W, and U_W is resolved from Table C.1, whose rows are line-to-neutral voltages of the power supply. The withstand of a telecommunication or signal port is a property of that interface and is declared by its manufacturer; select U_p against it, and against IEC 61643-21, rather than against the value used for the power lines here.

The voltage drop in the SPD connecting leads is computed from the approximation of IEC 62305-4:2024, eq. C.4 — 0,1 x I_SPD kV per metre. The standard notes that the true value depends on the current waveform, the conductor arrangement and the loop geometry; published manufacturer figures for the same case run up to about 20 % higher. The margin this consumes is small only while the leads are kept within 0,5 m, which is why that limit is a requirement and not a recommendation.

SPD selection and installation requirements

U_c, the maximum continuous operating voltage of each SPD, shall be equal to or higher than the value IEC 60364-5-53:2019+AMD1:2020, Table 2 gives for its mode of protection and the earthing arrangement, and the device shall withstand the temporary overvoltages of IEC 61643-11 for the same combination. Both are checked against a candidate's data sheet in the verification section, which needs the line-to-line voltage U of the supply — the Table C.1 band above does not fix it, since its 300 V row covers 230/400 V and 277/480 V alike.

Where SPDs are cascaded, a decoupling element or a minimum line length between them is required for correct energy coordination (IEC 62305-4:2024, Figure D.1). The required value is given in IEC 61643-12 and by the SPD manufacturer; a combined device may be used instead.

The back-up overcurrent protective device shall be the maximum permissible type specified by the SPD manufacturer, and shall be coordinated with the upstream system OCPD; where those two requirements conflict, the lower rating is generally the safer choice (IEC 62305-4:2024, C.2.8). That coordination is not computed here. The SPD's own short-circuit current rating I_SCCR shall not be lower than the maximum prospective short-circuit current at its connection points (IEC 60364-5-53:2019+AMD1:2020, 534.4.4.6), and the same holds for the follow current interrupting rating I_fi where the manufacturer declares one (534.4.4.7). Neither requirement applies to an SPD connected between the neutral conductor and PE in a TN or TT system — both clauses exempt that case, as the product standard IEC 61643-11 already covers it. Both are checked in the verification section against a prospective current you state, since it is a property of the installation rather than of the lightning protection design.

The behaviour of the back-up overcurrent protective device under partial lightning current or high impulse current should be considered as one of three cases (IEC 62305-4:2024, C.2.8.1). Case 1 — No melting and no tripping. The OCPD withstands the expected surge current through the SPD without tripping or melting. Case 2 — Melting and tripping permissible. The lightning impulse energy is enough to operate the OCPD. Case 3 — Mechanical destruction permissible. The impulse energy exceeds the OCPD's withstand level.

In cases 2 and 3 the installation is no longer protected against subsequent strokes within the same flash; additional SPDs on specific downstream circuits can maintain protection. Take care before replacing a blown OCPD or re-closing a tripped MCB, since the SPD may have failed into a short circuit. Because lightning behaves as an ideal current source, operation or destruction of the back-up OCPD does not reduce the impulse current the SPD must carry (IEC 62305-4:2024, C.2.8).

Where the circuit between the SPD and the equipment is long, propagation can lead to oscillation raising the overvoltage at an open circuit to as much as 2 x U_p/f, even where U_p/f <= U_W is satisfied (IEC 62305-4:2024, C.2.3.4). The effect is negligible where the SPD is installed at the equipment input.

Services connected to different ports of the same equipment shall not be connected to different points on the earthing system, except where those points are equipotential bonded (a densely meshed foundation electrode of the order of 1 m x 1 m, or a ring electrode welded or bolted to the steel reinforcement), or the internal services are shielded or in a metallic duct bonded at both ends, or a multiservice SPD able to withstand S1/S3 stress is provided between the ports (IEC 62305-4:2024, C.2.6).

Externally installed equipment

Externally installed equipment — aerials, meteorological sensors, cameras and exposed process sensors — is assessed to IEC 62305-4:2024, B.11. Positions are given here, but no item's exposure could be answered against this building's air-termination layout (IEC 62305-3), so every exposure is taken as declared and nothing here has checked it — each item says on its own row why. Complete what it names, or confirm each declared exposure against the installation.

Calculation trace

Every requirement above, with the clause that governs it — check any step against the standard.