Every carrier in the tube system is tracked: sending station, destination, time, and the transaction number on the label. One a week arrives at the laboratory from station 9, and station 9 was taken out when the old block came down.
I am the estates supervisor responsible for the tube system, among other things. Twenty three years on this site.
How it works, since most people only ever see the hatch.
Carriers are plastic tubes with a lid. You open a station, put your sample or your notes or your blood unit in, close it, and key the destination on the keypad. The system routes it through diverters and blowers to the station you asked for, and it lands in a basket at the other end.
The controller logs every transaction: origin station, destination, time out, time in, and a number. Fault conditions are logged too, and there are a great many of those, because a tube system in a hospital is a thing that jams.
Our system has thirty one stations. Station numbers are physical: each station has an address in the controller and a number on the front of it, and when a station is removed the address is deleted so the system does not route to a wall.
Station 9 was in the old maternity block, which was demolished in 2011.
The station was taken out, the pipework was cut and capped at the plant room, and the address was deleted from the controller by the contractor who did the works. I have the commissioning certificate.
The laboratory receives a carrier from station 9 about once a week.
The transaction shows in the log. Origin 9, destination 14, which is the lab, time out, time in, and a number in the right sequence for that day.
The carrier is in the basket. Somebody picks it up.
What is in it.
It is empty. Every one of them, and there have been about six hundred since 2011.
Empty and clean and dry, with the lid closed, which is a thing a carrier does not do by itself because the lid is a screw lid with an O ring.
What has been done about it, in order.
The controller was reprogrammed in 2013, from scratch, by the manufacturer's engineer, from a station schedule I wrote out on paper myself. Thirty one stations, no station 9.
It logged one from station 9 eleven days later.
The controller was replaced in 2018 as part of a wider upgrade. New hardware, new software, new station table.
Same.
The pipework. I have been round the whole run with the drawings and a torch. The line to the old maternity block is capped at the plant room, and it is capped with a welded plate, and I have had my hand on it. Beyond the plate the pipe goes into the ground under what is now a car park.
The pipe is there. Nobody digs a tube system out. You cut it, cap it and leave it.
What the log says that I cannot make sense of.
A transaction has a time out and a time in, and the difference is the transit time, and transit times in this system run between forty seconds and about three minutes depending on how far it goes and how busy it is.
The station 9 transactions have a transit time of two minutes eleven seconds. All of them. Six hundred, to the second.
Our system does not do anything to the second. Two carriers sent from the same station to the same destination a minute apart will differ by ten or fifteen seconds, because the diverters and the blowers are mechanical and the traffic changes.
The one figure that never varies in that log is the one belonging to the station that is not there.
What happens about it.
It is on the fault log as a spurious transaction, and the manufacturer's position, in writing, twice, is that the transaction is a software artefact and that the carriers are being sent from another station by staff keying the wrong destination, arriving at the lab, and being logged incorrectly.
That would be a decent explanation if the carriers were not empty, and if the transit time varied, and if the origin were anything other than an address that has been deleted from the table three times.
The lab staff empty the basket and put the carriers back in the rack. It costs nothing and delays nobody.
What I do.
The tube system is on my planned maintenance and it gets the same attention as everything else.
The one thing I do differently is the plant room.
The plant room has the blowers, the diverters and the capped stub of the old maternity line, and it is loud in there when the system is running, which is most of the day.
I do that room in the morning now, and if I need to be in there in the evening I take a colleague.
In November 2019 I was in there at about seven with a torch, checking a diverter that had been jamming, and the system was quiet because it was between transactions.
You can hear a carrier coming a long way off. It is a rising rush in the pipe and then the bang of the diverter.
I heard one coming up the capped line, from under the car park, and it stopped at the plate, and the plate is welded steel and about eight inches from where I was standing, and it did not sound like something hitting a wall. It sounded like something arriving.