What's Up With RTD and Caltrain's Grade Crossings?

Ever since the Denver RTD A Line opened in 2016, it's been plagued by strange grade crossing issues that eventually resulted in human crossing watchmen being posted for a number of years followed by a regulatory battle with the FRA.  Since then Caltrain has reported similar issues with a new grade crossing activation system citing them as one reason for the delays in their new electrification project. This begs the question, WTF is going on with RTD and Caltrain that would involve trouble with grade crossing activation, a problem that has been solved for well over a century. 

RTD A-Line Crossing Guards in 2017

A Twitter threat got me to Google this issue and I found a communication filed with the FRA from Denver Transit Partners that appeared to answer most of the questions if one read between the lines a bit. If you've tuned into my PTC reporting before you might remember how the entire concept was lifted from a grade crossing safety technology and while that has had a lot of implications on the signaling side it also means that project planners and political leaders have a ready made tool they can use to "solve problems", chief among these being NINBY complaints about grade crossing related crossing delays. 

RTD A-Line WiMAX PTC/WCAS Antennas

To make some more technology connections the RTD commuter rail network uses a traditional pulse code cab signal based ATC system and as it is an isolated network one might assume it would look to use ACSES as its PTC solution as that's the best way to leverage the cab signal system and avoid issues with wireless data.  Well for some reason they decided not to go in this direction and the reason is likely tied to the desire to implement a Wireless Crossing Activation System (WCAS).

At this point you might be wondering why an isolated 80mph commuter rail line needs a Wireless Crossing Activation System.  Well I was also confused until I saw that Caltrain had gotten itself into the same trap. Anyone that has ridden Caltrain knows that it's crossings tie up a lot of congested downtowns, especially around station stops and while Caltrain is in the process of a pretty extensive grade crossing elimination project, it's pretty clear if one reading the WCAS capabilities outlined in the previously mentioned DTP document that local motorists have a pretty strong voice when it comes to a frequent (4tph+ all day) commuter rail line snarling downtown traffic. 

Typically the example to justify WCAS is one of differential train speeds.  A "snow" freight train will activate the gates longer than a "fast" commuter trains will. While that is technically true there are scores of rail mileage in the US where 80mph passenger trains are mixed with 50mph or slower freight trains wit no ill effect.  On systems like Caltrain or RTD the number of freight is so tiny as to render this a non-issue.  The real "problem" is stopping trains.  Trains stop adjacent to a crossing and then keep the gates down for the duration of the station stop.  In addition to the longer delay there is likely a psychological element with drivers being delayed by a train that is just sitting there.  I have been told that some commuter rail operators like SEPTA and the LIRR have made use of timeouts to more aggressively raise the gates during long station stops, but I have no first hand experience with these. What WCAS looks to do is to have trains approaching the crossing negotiate if they are stopping or non-stopping  and then the crossing system delays the crossing activation and uses the attached PTC system to protect the crossing using a temporary speed restriction of some kind. 

This sounds great, but remember we have wireless involved as well as general complexity so, in the case of RTD, if this process fails, the backup is to use the full length signal block to trigger the crossing, resulting in a 90 second activation time.  There are additional issues of stopping trains (or trains that need to stop) getting non-stop activations that again result in a longer than planned crossing activation.  This brings us back to the DTP submission.  They are upset because CFR Part 49 Section 225 specifies a minimum 20 second activation time and at no point has this ever been a problem, in their eyes the FRA is baselessly trying to prevent motorists suffering from a few seconds of extra delay here and there.  Unfortunately Section 225 also says that a crossing system has to function AS DESIGNED and any deviation is considered a crossing malfunction.  This means that if your crossing protection uses neon lights and an air raid siren they had better work even if those aren't generally used elsewhere. 

From the FRA's point of view, the WCAS systems (at least used by RTD) was designed to have a certain activation time, the implementation isn't meeting the design spec so regardless of the 20 second minimum activation time the WCAS is not operating as designed and is therefore in violation of the regulations, thus the watchmen.  Sure, there are safety considerations involved with inconsistent crossing times, but the real issue is the meeting a design spec that was driven by the political considerations of crossing delays. 

TLDR NIMBYs demand the lowest possible crossing latency, planners see their shiny new PTC driven WCAS tools as a way to placate the NIMBY's, the technology fails to live up to the promise and finally the FRA gets mad.  There are multiple points to "fix" this issue  The FRA can be more flexible, the politicians can stand up to the NIMBY's or maybe the technology can be simplified.  Whatever the ultimate solution hopefully this sheds some light on the problem.