What a cordon costs, in perimeter and in people

Work it out in CORDON, which gives the perimeter, the area, the number of posts at your spacing and the multiple for holding it past a shift change, holds the published stand-off distances with both figures, and turns a centre and a radius into eight positions a unit can actually be sent to. Nothing is uploaded, including the position.

The arithmetic nobody does

The perimeter of a circle is 2πr, which everybody knows and almost nobody applies at the moment it would change a decision.

Two things in that table are worth dwelling on.

The first is that perimeter is linear in the radius. Doubling the radius doubles the people. That is intuitive once stated and is not what anybody's instinct says when a distance is being argued about on a radio.

The second is that area is quadratic. Doubling the radius quadruples the ground enclosed, which is what determines how many premises are inside, how many people have to be moved, and how long the evacuation takes. A cordon decision is really two decisions with different cost curves, and the one that sounds cheap — a bit further back — is the expensive one.

The number everybody quotes is the single-shift number

Twenty-six posts is twenty-six people standing there now. It is not twenty-six people to hold the cordon, because they will need relieving, and a cordon of any consequence outlasts a shift.

Three reliefs is the honest planning figure for anything expected to run past a few hours, and it turns twenty-six into seventy-eight. That number is the one that determines whether you are asking for mutual aid, and it is generally arrived at around hour four rather than minute one.

Spacing is the other lever and it is not free. A hundred metres between posts on an open perimeter means each person is watching fifty metres in each direction, which is fine in daylight across open ground and useless at night in a built-up area with side streets. Urban cordons are held at junctions rather than at intervals, and the count is then the number of ways in, which has to be counted rather than calculated.

Why published stand-off distances carry two numbers

Responder stand-off guidance for a suspected device gives two distances, and the difference between them is the whole point.

The first is the distance inside which a person is at direct risk. It is the smaller figure, and it is the one people quote.

The second, which is typically several times larger, is the distance at which flying and falling glass stops being the dominant hazard. Blast breaks windows over a far wider radius than it injures people directly, and in a built-up area the injuries at distance are overwhelmingly from glass rather than from the blast itself.

Quoting the first figure as though it were the second is the mistake that produces the characteristic failure: an orderly, well-managed evacuation that moves a crowd out of the direct-risk zone and assembles it on a pavement beneath a glass frontage. The cordon was correct and the assembly point was not.

Which is why an evacuation has two questions, not one: how far back, and to where. Standing people in the open at an adequate distance is fine. Standing them at the same distance under glass is not.

Positions, not a radius

“Four hundred metres” is not an instruction anybody can act on. A unit told to hold four hundred metres to the north-east has to decide what that means on the ground, and two units will decide differently.

Eight positions — a bearing and a distance from a known centre, expressed as coordinates — can be read out, typed in and driven to. The conversion is a geodesic calculation, and it is worth doing on the ellipsoid rather than on a sphere: a spherical approximation is out by up to a fifth of a per cent, which over a two-kilometre cordon is metres in the wrong place for somebody who has been told to stand there.

The log, which is the part that gets asked about

Who was on the north-east junction at eleven minutes past, and when were they relieved?

Three weeks later, that answer exists only in somebody's memory of a radio call, unless it was written down at the time. It is the single most common gap in the record of a cordon, and it is entirely avoidable: a line per post, a name, a time posted and a time relieved.

The value is not defensive. A commander who can see which posts have been held for four hours and which were relieved an hour ago is making a better decision about where to put the next arriving unit.

None of this is a decision

The distances published for responders exist so that somebody arriving first has a defensible number to work to in the first two minutes, before anybody qualified is there. They are planning figures and they are not advice about the thing in front of you.

The person who owns the scene sets the cordon, and where there is ordnance involved that is the ordnance officer, on the ground, looking at it. Arithmetic can tell you what a decision costs. It cannot tell you what the decision should be.