Bill West wrote:Your description brings out that it is not enough for the dispatcher to go along with the train crew inspecting and seeing if the mitre rails are down. Which is a process that might be acceptable at a track switch. For our case, if the wedges are not in, the end of the bridge will flex down an inch or two when the train rolls on it and a flange may pick at the rail joint, to say nothing of damaging the pivot bearing. Hence having to wait for a maintainer to find if the trouble really is major or just a minor adjustment. This is where newer electronic overlays could help, they can at least tell you which interlock is unhappy without having to spend time probing around with your voltmeter.
We had our own separate switches sensing that the lift rails were fully lowered. We would have rather used a spare contact off of the actual signal department switches, but they were four-pole switches with all four poles used for signal department "vital circuits".
During the renovation of our bridge, the derailment at Portal that was eventually linked to a lift rail problem occurred. As the South Channel bridge is the only moveable bridge on the New York City Subway system, NYCTA engineering asked us to review and report on the switches and interlocks associated with the lift rails. It was decided that a redundant set of switches and indicators be installed to sense lift rail position.
We installed an additional 8 switches and some relays wired into the bridge control relay logic and to existing lamps on the new console. The two existing lamps were "North Rails Down" and "South Rails Down"; there just wasn't room to add another 8 lamps. We interlocked all of this as follows: If all 4 rail lift switches on one end indicated "up", the light was extinquished. If any one switch indicated "down", the light flashed and only when all 4 rails were sensed down did the console lamp illuminate steadily. This showed the bridge operator whether any switches failed to close or were stuck in one position.
But we weren't finished yet... Each of the relays had an internal indicator lamp, so the operator only had to walk a few steps to the equipment room and open a cabinet door to see which switches reported "open" or "closed".
We didn't stop there, either... Other modifications used all of the spare inputs left in the PLC rack, so we were asked to supply some additional I/O cards. With plenty of inputs now available, we wired the second set of contacts from our eight new rail lift switches into the PLC, and programmed the PLC as follows: When the rail lift machinery was stopped after lifting the rails, the PLC took a snapshot of all 8 switch contacts. If all 8 were open, an "All Rails Fully Raised" message was stored and printed to the log. If any contact was still sensed as closed, a message was stored and printed to the log "Track F3 (or F4) North (or south end) East (or west side) rail lift switch fail closed" for any switch that failed. The same thing occurred when the lift machinery stopped after lowering the rails; either an all rails lowered message was logged or an individual message "(track#) (end) (side) rail lift switch fail open" was logged.
While the "bridge" switches did not replicate the "signal department" switches, there was at least some redundancy! The signal department did accept and allow us to install a set of indicator lamps on the bridge operator's console that wired back to their relays operated by their rail lift switches. These "repeater lamps" were wired in parallel with 4 signal department vital relay coils using very low current, high brightness LED's so lamp current would not affect the vital circuit. The bridge operators finally had indicator lamps that repeated the status of the signal department circuits, and could immediately identify if a signal department switch was the reason for failing to get a "bridge lock" signal. In the end, there were 3 independent sets of switches sensing lift rail position on the bridge, which did speed troubleshooting.
The PLC system was designed years before the project was bid and constructed; and ruggedized flat panel displays for industrial use just did not exist at that time. Toward the end of the project when modifications and additions were being considered, a flat panel display showing an "aerial view" of the bridge was suggested. Every limit switch could have been displayed as a red or green indicator shown exactly where it was physically located on the display; bridge rotation, position and speed could also have been displayed along with critical messages. The operators would have had a real time graphical display of the status of every limit switch on the bridge. Unfortunately, it was too costly to add this technology near the end of the project; it sure would have been nice!
