Everybody thinks the hard part of D-Day was the beaches. The planners did not.
The hard part was the second week, and the fifth week, and the tenth. An army of a million men consumes an enormous tonnage every single day: ammunition, food, fuel, spare parts, medical supplies, and replacement everything.
Delivering that requires a port. A proper port with deep water, cranes, quays and rail connections, where a ship can tie up alongside and be unloaded in hours rather than days.
To put a number on it: the Allied armies in France were consuming, at their peak, somewhere over twenty thousand tons a day. That is a very large number of ships arriving, forever, without a pause.
The Germans understood this perfectly. Every major port on the French coast was fortified into a fortress, on the correct assumption that any invasion would have to go for one.
In August nineteen forty two, a raid was made on the port of Dieppe, largely with Canadian troops.
It was a disaster. Of roughly six thousand men who went, more than half were killed, wounded or captured, most of them in a few hours. The tanks could not get off the shingle beach. The naval bombardment was inadequate. Nearly everything that could go wrong did.
Its value was entirely in what it proved, which is that a defended port could not be taken from the sea by direct assault, at least not in a state where it would be any use afterwards, since the defenders would demolish it before surrendering.
That left a question that appeared to have no answer. You cannot supply an invasion without a port, and you cannot capture a port.
Somebody, and there is a good deal of argument about exactly who, wrote a memorandum containing the sentence that solved it. If we cannot capture a port, we must take one with us.
What was built is still, eighty years on, one of the largest civil engineering projects ever undertaken, and it was done in about eight months, in secret, by firms all over Britain who were not told what they were making.
Each harbour, and there were two, had three parts.
The outer breakwater was made of old ships, sailed over under their own power and then sunk in position by opening the sea cocks. Inside that came the real breakwater: enormous hollow concrete boxes called caissons, the largest of them two hundred feet long and sixty feet high, weighing six thousand tons each. They were built in dry docks and in holes dug in riverbanks all round the coast, floated out, towed across the Channel, and sunk in place.
Then came the piers. A quay that is fixed cannot work where the tide rises twenty feet, so the pierheads were platforms on legs that rested on the seabed and slid up and down the legs with the tide.
And connecting the pierheads to the shore were floating roadways, made of steel spans on pontoons, flexible enough to ride the swell, up to a mile long. Vehicles drove off ships onto a road that was floating on the sea.
The whole thing had to be moved across the Channel after the invasion had already started, in pieces, at a walking pace.
There were around four hundred separate units to be towed, and towing a six-thousand-ton concrete box across the Channel takes tugs and time and reasonable weather. Everything available was used, and a good many of the tugs were American and had been brought over for the purpose.
Assembly began on the seventh of June, the day after the landings. Within about twelve days both harbours were substantially working.
And then, on the nineteenth of June, the Channel produced the worst summer storm it had seen in forty years. It blew for three days.
The American harbour at Omaha was wrecked beyond repair. The British one at Arromanches, which had been built in slightly more sheltered water and had more of its breakwater complete, survived with damage. It was repaired using parts salvaged from the other one.
The one that survived stayed in use for around ten months, far longer than the ninety days it had been designed for, and something like two and a half million men, half a million vehicles and four million tons of supplies came through it.
The other half of the problem was fuel, and fuel is the worst thing to move by ship, because a tanker is slow, vulnerable and takes up a berth for a long time.
So they built a pipeline along the bottom of the Channel.
The project was called PLUTO, standing for Pipe Line Under The Ocean, which is a piece of naming that everybody involved seems to have enjoyed.
There were two designs, and both were peculiar. One was a flexible lead pipe, essentially an enormous version of an undersea telegraph cable with nothing in the middle. The other was a steel pipe wound onto a floating drum ninety feet across, nicknamed a Conundrum, which was towed across the Channel unwinding pipe behind it as it went.
The first lines ran from the Isle of Wight to Cherbourg once that port was taken, and later ones ran the shorter route from Dungeness to Boulogne. The system eventually delivered somewhere over a million gallons a day.
Not all of it worked. Several lines failed, the early ones were laid late, and for the first months of the campaign most fuel still arrived by tanker and was pumped ashore. PLUTO's reputation has probably always run slightly ahead of its actual delivery figures, which is worth knowing.
It is worth stepping back and noticing what this episode is really about.
None of this is exciting. There is no combat in it. Nobody makes a speech. The heroes are civil engineers, dock workers, tug crews and lorry drivers, and the achievements are measured in tons per day.
And it is the reason the invasion held.
The German assumption had been that the Allies must go for a port, and that until they had one, their build-up would be slow enough to be contained. Every calculation about how quickly the beachhead could be sealed off rested on that.
Instead, supplies came over open beaches at a rate nobody had thought possible, and the build-up in Normandy outpaced the German build-up by road, because the roads and railways were being taken apart from the air.
Within three months of the landings, more than two million men, half a million vehicles and four million tons of stores had crossed to France, most of it over open sand.
The campaign was decided by the arithmetic of arrival. It is not the part anybody makes a film about.
You can still see it. At Arromanches in Normandy, the remains of the caissons are sitting out in the bay, in a long curve, and at low tide sections of the roadway are visible. Photographs do not really convey the size of them.
The secrecy was remarkable, given the number of people involved. Thousands of workers built components without being told what they were, and the standard assumption on the sites was that they were making floating grain silos, or breakwaters for after the war.
The storm of the nineteenth of June is a useful reminder about how much was luck. If it had come thirteen days earlier it would have caught the invasion fleet at sea, and if the invasion had been postponed to the next tidal window later in June, it would have sailed straight into it.
Here is something to think about tonight. The Germans built a wall along the entire coast and fortified every port, and the Allies simply brought their own port and put it on a beach where there was nothing to defend. When a problem looks impossible, how do you tell whether you are solving it or just accepting somebody else's version of it?