Return to Archive -by date - by topic - Archive 2002.

INFORMED SOURCES May 2002

Buffer stops breached driving techniques

My joke that HMRI were trying to kill me by forcing drivers to power up to buffer stops in terminal platforms protected by TPWS was a more serious issue than I thought.

As you will recall, certain types of train running at certain speeds could cause buffer stop TPWS to make a spurious intervention. A train could be running at less than the nominal 10mile/h approach speed, only for the TPWS to slam on the brakes when it passed the over speed sensor ( OSS ) halfway down the platform.

As usual it was an interface problem. Normally, you might expect the metal of the bogie to attenuate the signal radiated from the OSS ‘grids'. But in some case it can have the opposite effect. This allows the aerial to pick up the secondary radiation fields, colloquially ‘side lobes', which are generated around the ends of the grids.

TPWS works on the basis that when the aerial passes the first OSS grid a timer starts. Passing the second grid stops the timer. Grids are spaced so that this process always takes 1 second for a train running at the maximum safe speed. If the time is less than a second, the train is running too fast and TPWS intervenes with a brake application.

Combination

Technicians emphasised that the spurious buffer stop trips were due to a combination of heightened aerial sensitivity, the length and spacing of the grids and the train speed. The system had not been tested at low speeds before installation.

Given the ‘right' speed (5.5-6.5mile/h) and enhanced aerial sensitivity, detection of the ‘side lobe' at the end of the first could restart the counter, so that it registered a shorter time than 1 sec, tripping the TPWS.

In the new design, the grids are one third shorter which prevents such spurious timer readings. The smaller grids, are now being produced by Thales and will replace standard OSS fitted at buffer stop platforms.

So that's alright then? Well, no, because despite the growing number of cases of spurious tripping, leading to the imposition of 5mile/h approach speeds at terminal platforms, Railtrack continued to install buffer stop TPWS without informing drivers.

As recently as mid March, three trains entering a London terminus on the morning after buffer stop TPWS had been commissioned, suffered spurious trips in rapid succession. Mid-March, note, when this column drew attention to the safety risks in the February Modern Railways.

Risks ignored

What safety risks? Standing commuters falling down under the abrupt deceleration. Commuters in the rear coaches of slam door stock, on early morning autopilot, thinking that the train has stopped stepping out onto the track. Commuters leaping out onto the platform as the train crawls along at 5mile/h or less. And, last, but more important, drivers having to power up to the buffer stops. At several stations. Kings Cross, Paddington and Euston, for example, you cannot coast to the buffers if you are doing under 5mile/h halfway down the platform.

But drivers are specifically instructed not to take power when approaching buffers stops. At one London terminus the firm policy is that drivers cannot be allowed to enter a platform if there is the chance that they will then have to apply power to reach the buffer stops.

And yet drivers have been powering up to buffer stops all over the place. How can Railway Safety, let alone Railtrack, let alone the train operators, have been happy for drivers to disobey a tenet of professional driving? Powering up after passing the OSS negates the concept of buffer stop TPWS anyway.

Why, when the problem was well reported (well, in this column) did not Railtrack remove the fuses from buffer stop TPWS installations until the modified grids could be replaced?

Above all, what was Her Majesty's Railway Inspectorate playing at? Here was a basket of risks and, as far as I can tell, HMRI did nothing about it. Could it be that they were embarrassed because the whole potty idea of buffer-stop TPWS emanated from – err the HMRI?.