The Other Rule G

The evolution from "Stop and Proceed" to "Restricted Proceed" in North America is long and something I hope to tackle at a later date. A holdover from the timetable and train order era where waiting for a train running ahead increase its lead made some sense to reduce collisions, the fuel and time savings of allowing following movements to keep moving is hard to deny.  However even where Stop and Proceed are the norm, exceptions are necessary where something like a stiff grade might make restarting a heavy freight train difficult. To implement these exceptions, the (G) or Grade marker was used on automatic signals to relieve some or all trains of the need to stop when paired with a Stop and Proceed indication.

Stop and Proceed was especially popular in the Northeast United States, but after Conrail was picked apart by NS and CSX, most of the major northeastern freight territory fell under the spell of Restricted Proceed.  At this point the G marker appeared to be on its way out as the remaining NORAC roads were either passenger focused or too small to have signals or heavy freight trains. However, as the limited speed triangle, Amtrak's Harrisburg line turned into an unwitting refuge. The PRR's Main Line between Philadelphia and Harrisburg was surprisingly hilly and saw more than its fair share of automatic signals with (G) markers applied.

One might assume that the spread of Rule 562 cab signals without intermediate wayside signals and complete lack of through freight would put the Main Line's (G) markers into a terminal decline, however at the Milepost 21 intermediate signal, a minor capacity project may have what could be the last new main line (G) marker to be installed in the United States. 

When Amtrak wanted to upgrade the single direction tracks tracks between the SEPTA Yard at Frazer and the start of 4-track operation at Paoli, they modified Milepost 21 intermediate signal location for bi-directional operation on both tracks with new Safetran colorized position lights and (G) markers in the eastbound direction.  Moreover, the (G) even uses the same old school blocky font as opposed to the smoother letter used on things like (D)istant and (A)pproach markers.

It's anyone's guess how much longer these (G) markers will be around for as the track segment they govern will soon be subjected to the Rule 562 treatment. Another part of Amtrak territory south of Wilmington also features both long freight trains and (G) markers, but it too going the 562 route so its a good chance that these will be the last new (G)'s installed in the United States. 

Small Changes at Western Ave

While out visiting METRA TOWER A-2 at Chicago's Western Ave for the first time in many years, I noticed that the US&S H-2 dwarf searchlight signals at the adjacent TOWER A-3 interlocking had been replaced by L&W LED searchlights with a small backing target. It is unclear if this took place in conjunction with the 2019 removal of legacy PRR position lights at TOWER A-2. TOWER A-3 is remote from a panel inside the staffed TOWER A-2 and the interlocking's other distinctive features including gantry mounted main line searchlights and A-5 pneumatic point machines had been left untouched. You can see the difference between the old H-2 searchlights and newer LED dwarfs below.

  


An interesting feature is that the 6L dwarf appears to illuminate only with a route is lines (or I assume upon the occupation of adjacent track circuits). This signal is placed after a train exits the main line, but before it is routed to a yard track. I managed to catch the 6L LED dwarf displaying a flashing red "Restricted Proceed" indication for a yard move, in conjunction with a lunar white Restricting aspect on the 10L gantry signal. (Note the 10L signal displays "3rd head Restricting" with a phantom center head.)



I assume that the H-2 yard dwarfs always displayed R stop or L Restricting and the LED units changed this to *R* Restricted Proceed as is the style today. I guess its cheaper to have an LED unit with just a single color of LED's, but I will point out that Union Pacific did choose to use lunar LED searchlights at the nearby KEDZIE interlocking.

Tight Jeans: Position Lights As The PRR Intended

We all know the PRR Position Light signal looks like this with the round upper head and the (usually) squared off lower head. Both backing targets are technically interchangeable with each other (Although I have never seen a lower head type target used on an upper head.). 

However what would you think if I told you that this concept of the PRR PL did not reflect the original vision of the author. A while ago I wrote up a post discussing the compact position light and how the lamp arms of a US&S pattern position light can support a PL lamp module at any point along their length.  At the time I thought that there were two standard PL setups, full size and compact. Well it turns out there is a third method of lamp arrangement which was actually the standard setup used by the PRR in its final few decades. 

While the PRR was an industrial behemoth that didn't hesitate to spend on quality, there were a few areas where its famed standardization turned more towards penny pinching. One of these was the configuration for the lower heads of its position light signals. To make the signals stand out in bright sun, a dark backing target would be necessary. However there would be little need for this contrast if an approaching train was moving at slow speed below 30mph. Therefore the PRR only applied a backing to the lower | position as the lower / and \ positions would be approached at something more like 15mph. 

If you look closely at one of these original pattern lower heads you'll also see that the modules for the / and \ positions are not positioned at the ends of the mounting arms. Instead they are scrunched in couple inches. Although not to the level seen in compact PL arrangements, this scrunch may have been to provide some contrast without the "expense" of a backing plate. 

See the difference between the placement of the lamps on the | vs the \ position.

I am sure the rivet counting modelers are yelling at their screens right now about my obliviousness, but somehow I had never noticed this before. I guess this is partly because unmodified PRR position lights have become so uncommon, but also because later lower backing plate thoroughly set the expectation that the spacing would be the same on both heads. Anyway now I know and now you know as well.

Hoboken Terminal's Suspended Signals

Stations with large terminal fans often exhibit a problem at one or both of edge tracks. Because rail vehicles can't generally make hard right turns, one needs a longer series of switches to reach the highest and/or lowest track. Unfortunately terminals often abut a fixed location like a river or street or public square and to avoid having those extreme tracks things can get a little squished. In the early 1980s, New Jersey Transit engaged in a near total reconstruction of the former Delaware, Lackawanna and Western electrified suburban network. Not only was the 3kv DC traction power system under powered and on its last legs, but the signaling system was firmly rooted in the 1920's. In the re-signaling project that followed, NJT built the last great North American interlocking tower, TERMINAL TOWER and replaced most vestiges of the former DL&W plant.

New (frame left) and old HOBOKEN TERMINAL TOWERs shortly before cutover.

In the late DL&W era Hoboken Terminal signals consisted largely of two head, small target searchlight signals displaying the DL&W's unique mix of color light signal indications. Long platforms in the center of the terminal had these mounted on signal gantries in the normal way, while tracks towards the edge of the low 1908 train shed saw signals mounted below the gantries under the shed itself. Moreover, while center tracks could get a "straight" clear out of town, tracks towards the edges had to make do with a Restricting indication as the DL&W lacked Slow Clear and Slow Approach options, even on dwarfs. 

When NJT rebuilt the signaling there appears to be a time when the new signals were wired into the old HOBOKEN TERMINAL tower's old US&S 1908 Electro-Pneumatic interlocking machine. To make the cutover process easier, signal locations were largely kept the same, but NJT had no desire to keep the old handicap of trains of edge tracks departing under a Restricting signal indication. This meant that the replacement "high" signals had to display R/R/G Slow Clear as opposed to the G and G/G on a dwarf. In most locations this wasn't a big deal as the signal gantries could easily support three tri-light heads, however on tracks 2, 3, 12 and 13, the need to suspend signals under the trainshed due to the closeness of the first switches presented a bit of a problem.

TERMINAL TOWER 36W2 and 36W1 suspended signals governing tracks 3 and 2 respectively.

TERMINAL TOWER 88W and 92W suspended signals governing tracks 13 and 12 respectively.

The solution was somewhat akin to a Penn Station signal, just with three "heads" instead of two. To display R/R/G, R/R/*Y* and R/R/Y, two fixed Safetran modular dwarf cubes were mounted close above an additional three-unit cluster of modular dwarf cubes. These were then mounted close under the roof at the westernmost end of the trainshed.

36W2 signal governing track #3.

 

Although a bit inelegant compared to a dwarf or today's compact LED searchlights, they are a unique solution in today's largely homogeneous signaling environment and are likely to persist for many years to come. 

The Last N&W CPL Intermediates Fall

 For many years a lightly used NS line in southern Virginia has been been carrying the torch for N&W position light signaling. Well it appears that the torch has gone out not just for its own N&W colorized position lights, but, in all likelihood, the last example of N&W intermediate signals anywhere on the NS system. The NS Altavista district had become an island of classic signaling due to the typical reason of decline. Like the former Conrail Buffalo Line, C&O Virginia Central route and the B&O Mountain Sub, the Altavista district was involved with moving coal from the Appalachian interior to various ports. 

Comprised of the Virginian Railway's main line that was absorbed into the N&W in 1959, the gentler alignment makes it the preferred route for loaded eastbound coal trains in a form of directional running. Converted by the N&W to bi-directional CTC signaling with N&W style position lights between 1960 and 1963, the relatively young age of the signaling combined traffic that made the route PTC exempt put the Altavista District low down on the capital improvement totem pole.  Unfortunately with the great PTC push of the 2010's now in the rearview mirror, NS started to pick away at the Altavista district PLs in the early 2020's.

In the particular case of automatic N&W Pl locations, the N&W switched to installing new color lights before the end of the "right hand rule" in 1985. This made "split" automatics ripe for replacement during signal logic refreshes in the 90's and 2000's, even if PL signals at interlockings were left in place. If the assessment of the community is accurate, the Altavista district was the last place one could find active N&W PL signals with an automatic number plate, although like with the dwarf signals, its entirely possible that one or two might be hanging on deep in the forests of coal country. 

H-Line CPL's, not Altavista, but you get the point.

Note the qualification of "active" as the former N&W route eastward out of Cincinnati had its N&W CTC signaling, including automatic locations, taken out of service, but not removed. Now leased to a shortline it is the most practical place to get photos of automatic N&W PLs.

 

 

 

 

 

Amtrak's Q and R Towers Close For Good

 Signaling control for the Pennsylvania Railroad's massive 1910 Penn Station project consisted of 10 block and interlocking stations. Towers N and S (later HUDSON) managed traffic at Manhattan Transfer, W (later PORTAL) handled the Hackensack River swing bridge, A, KN (B?), C and JO (D?) controlled Penn Station proper, F managed the eastern approaches and, R and Q controlled either end of Sunnyside Yard. N went first as Manhattan Transfer was consolidated into HUDSON interlocking in the early 1930's. PORTAL went next in 1988 followed by the four Penn Station towers and F being  replaced by the Penn Station Control Center in 1994. When HUDSON closed in 2003 it brought an end to the era of main line towers on the Pennsylvania Tunnel and Terminal railroad, but R and Q soldiered on at either end of Sunnyside Yard. 

In the early 2000's the 1910 R was re-signaled and the 1910 tower was demolished to make room for an Acela maintenance facility. Still, the complexity of the Sunnyside Yard operations still required the use of local operators, who were relocated into the yardmasters office in the Acela building to run the interlocking through a video interface. Around this time both Q and R were reclassified as "switching centers" to alleviate Amtrak of various regulatory requirements associated with main line interlockings. 

 

While the operator at R was playing video games in an office, Q kept going with its original US&S Electro-Pneumatic interlocking machine from 1910. Similar to the machine in ALTO and NORTH PHILADELPHIA, the Q machine was likely to be the last pre-Model 14 in active service in North America. Unfortunately, as early as 2015 one began to see the replacement of pneumatic A-5 point machines with electric M3's and by the early 2020's new relay cabinets were popping up all over. In 2024 I got word that the 1910 Q machine had been retired, the operator switched over to a video display like R, and just now I was sent these this Amtrak bulletin order taking both R and Q out of service and replacing them with the new Section S dispatcher in the PSCC.
 

Why this is happening now when long term closures of each of the four East River Tunnels are likely to complicate Sunnyside Yard operations is unknown. It could be that the Q re-signaling project was completed and found to be in working order and moving to PSCC was always part of the plan. Or it could be that Amtrak was having issues managing the two remote yard towers when everyone else except said towers were in the same room. If new productivity tools allow a single dispatcher to do the work of two or more operators (especially if they lack a clear view of the yard) remains to be seen.

A Quick Look at WALBRIDGE Tower

WALBRIDGE was one of several active towers that ringed Toledo into the first decade of the 21st Century. Technically a PRR tower, WALBRIDGE stood at the triangular junction of the C&O Northern Branch, PRR Toledo Branch and the Toledo Terminal ring railroad. Like many PRR towers west of Pittsburgh, it lacked the fancy architecture and bay windows, and was instead a pretty basic brick box with a US&S Model 14 interlocking machine controlling a pneumatic interlocking plant with position lights in all the typical places. 
 

 

Like many of the survivor towers, it stayed in service due to the personal service needed to manage the trains of conflicting railroads, namely CSX and Conrail. After Conrail was broken up, it was pretty straightforward to rationalize the lines around Toledo, eliminate a lot of diamonds and close the towers. Anyway, Railroad Media Archives has posted another fantastic video from 1993 showing operations around WALBRIDGE including interior video as trains moved through the plant. You can cut to the exact time of the interior video or watch the whole thing. 

For some additional reference here is how the interlocking machine and model board looked in 2001.

 

Unfortunately the tower was demolished in 2015. If I ever find more than the handful of interior photos I'll write up a longer piece on it, but today just sit back and enjoy the video.  

Dispatcher Signal Stacking

The story of railway signaling, like many others, has been one of technology increasing human productivity. Power operation allowed "levers" to be thrown faster, N-X machines allowed routes to be set with fewer button presses and remote operation combined with CTC allowed one person to control many miles of track. However remote control and ease of operation were not the only types of automation added to towers and dispatch offices. Fleeting allowed infrequently manipulated interlockings to do the same thing over and over and in cases where routing decisions needed to be made, automatic route setting combined with train labeling could keep track of train movements and automatically request routes as they approached junction points.

Recently another method of automation has crept into dispatch software, route stacking. While similar to automatic route setting, it is different in a significant way. Consider the following pair of photos showing an Amtrak Springfield shuttle train departing SPRING interlocking at Springfield Union Station.  

 

Within seconds of clearing the interlocking the 2N dwarf signal (upper right of the photo) changes from Stop to Restricting for an opposing movement. In the past this would signify a very attentive dispatcher, hovering over their CTC panel or video display, ready to line the new route as soon as possible. If this was an auto-routing system we would have had to wait for the opposing movement  to enter the track segment to trigger the route request.  With stacking the dispatcher simply uses their computerized dispatching interface to queue up a sequence of routes which will be executed as soon as the last established route is taken. Then they can do and pay attention to something else.  

This works well at places like Springfield where reverse movements are common. For a northbound train the dispatcher could set up a route from the 2S at SWEENEY, 12 switch reverse. Then immediately have a second route set for the 1N 13 switch reverse to the CT River Main Line.  It's not super complicated, but it is something that, due to the time penalties associated with bad routes, took a little while to catch on.

Amtrak's Secaucus High Density Cab Signaling in Action

When the Secaucus transfer was being designed and built around the year 2000, Amtrak implemented a new "High Density Cab Signaling" concept with shorter blocks and additional cab signal codes in order to increase the capacity of the approach into Penn Station especially with all the NJT Corridor and Midtown Direct trains stopping at Secaucus. This would replace the PRR era Rule 261 signaling with mile long blocks. Below we see the 2E signal at ERIE interlocking cycle through its indications after an eastbound NJT push-pull takes a Clear signal upon departing the track 2 station platform. 2E was fleeted for a normal route and immediately went to Stop and Proceed, then Slow Approach, Approach, Approach Limited and  finally Cab Speed (which could represent both 60 and 80mph speed codes). The horn at the end was an approaching Amtrak Regional which ultimately took 2E at Clear (not captured on video).

Unfortunately this was just in time for NJT to lean into Push-Pull service which lacked the acceleration to and braking systems to take advantage of the rapidly changing speed codes. On push-pull sets with traditional air brakes, crews would often target one speed below the one being displayed to avoid jerky train handling and the risk of penalty brake applications in the face of frequent cab signal drops.This negated much of the intended speed benefits from the high density cab signals. For whatever reason, similar High Density Cab Signals could be found on the 2005 era western Harrisburg Line re-signaling, however due to the low traffic density, Amtrak has taken to disabling some of these extra code change points.

NAN Drawbridge: A Crazy Thing We Lost

While this is ostensibly a blog about railroad signaling, I'll sometimes cover related bits of historic technology like railroad telecom systems, electrification systems and drawbridges. Today I get to cover two of those topics at once. Located on the former New Haven Railroad Shore Line, the Niantic River Drawbridge sat at the mouth of the Niantic River in the town of, you guessed it, Niantic, CT. A former New Haven Railroad signaling station, the 1907 bridge would later take the name NAN in the Penn Central and would serve as a staffed interlocking and block station up until the Boston electrification project of the late 1990's. Due to the New Haven being in perpetual bankruptcy NAN saw little in the way of upgrades and was documented by the Library of Congress before the changes of the NEC Improvement Project. While a chain-driven rolling lift bascule bridge is pretty cool in and of itself, it was the electrification project that provided "Old Nan" with one heck of a curtain call.

Railroad movable bridges have to employ complex mechanisms to separate the running rails and unlock the bridge, however when overhead electrification is involved a separate yet equally important system to isolate the contact wire across the movable span must also be employed. Because the 1907 NAN bridge was not electrified for its first 93 years of existence, this system had to be retrofitted onto the rolling bascule span in situ.  

On the contemporary Pellham Bay bridge, which was electrified within a decade of construction, a mechanism isolates the dynamic tension contact wires, while an alternate routing takes the 12kv lines up and over the waterway. The span then rolls back causing the wires across it to sag down to touch the track structure.  (I have seen a photo of this, but cannot currently locate it.) However when NAN was electrified in the 1990's, Amtrak had to use a fancy constant tension solution and in this case that meant a solid aluminum rail type conductor running the length of the span. Now, if you're sharp you might already realize that when the bascule span rolls back, a solid rail can't bend and flex with it. Well, this video shows the solution.  

That's right, a traveling telescoping carriage would move back and forth to extend and withdraw the contact rails from below the rolling structure of the bridge. What's even crazier was that NAN was the busiest drawbridge on the NEC due to a low clearance of 11 feet and a steady stream of boat traffic. Due to complaints from boaters, an agreement with the Coast Guard in 2003 limited closings to "only" 48 per day, so in its final decade of service, Old Nan was putting on quite the show. Unfortunately the 1907 bridge was finally replaced in 2012 by a new bascule span with greater clearance to reduce the need for openings. Fortunately its presence in plain view right downtown allowed for many to personally experience the show and record it for posterity. 

 

MBTA's Everything's Alright Signal

Those who find themselves looking down the MBTA Red Line's distinctive open span subway tunnels might be forgiven if they sometimes think their eyes are playing tricks on them. One moment a lunar white wayside marker signal is visible, the next moment it has seemingly vanished. Those with above average powers of observation might also notice that these signals are facing against the standard current of traffic and also vanish in the minutes before a train's arrival. These are worker protection or "TAK" lights for "Train Approach IndiKator" (also called "snitch lights") and their job is to illuminate when everything is alright, and extinguish when a train is approaching.  

Light on, all clear.

Light out, train approaching.

 

As one might have guessed these use track circuit state to indicate when a train is approaching in a failsafe manner. Placement of the TAK lights appears to have been around curves or other situations that limit the sight lines of track workers. They seem to have been designed for the older single direction ABS system and may only still be in service in the tunnels as the few I have seen on the above ground Braintree Branch appear to be bagged. 

TAK lights are one of those things that you never notice until someone points them out, then you can't help but notice. Back before the real time departure information they were a life hack to know if a Red Line train was approaching, but today, unless you are a track worker, they are mostly a neat bit of trivia. 

LINCOLN Lefty: Story of the PRR's Left Handed Dwarfs

 Anyone looking at the NORAC signal aspect charts may have noticed that the familiar PRR dwarf position lights have a curve to them (or yaw) that can go both ways. For all four dwarf indications, Slow Clear, Slow Approach, Restricting and Stop, one will see two entries on the diagram chart, one with a left hand curve and one with a right hand curve. To the uninitiated this can be dismissed as having to do with clearances or which side of the track a dwarf signal is placed, but it actually represents a fairly early change by the PRR to make a more practical signal vs a more conceptually accurate signal. 

The first PRR position light signals were designed as a replacement for semaphores. Version 1 or "tombstone" position lights (due to the shape of the backing) consisted of 4-lamp lines with an off-center pivot point that directly mirrored the shape and function of a semaphore blade. The position light dwarf signal was no different, but used two lamps to mimic the right to left sweep of a dwarf semaphore. Originally only three positions were provided, stop, low speed prepared to stop and low speed, again matching the function of the dwarf semaphore. 

Between 1918 and 1925 the three lamp system with circular mounting became standard for high signals, but the dwarfs, unable to be simplified further, retained the design aesthetic of a semaphore blade sweeping out a curve. The one major change was that someone noticed that the 4-lamp PL dwarf could display the reverse \ for free and a distinction between "proceed expecting stop" and "proceed expecting obstruction" was able to be made. The result was the left-handed dwarf position light.

Made of cast iron by Union Switch and Signal, the left hand dwarf had bulb access hatches on both the straight and curved sides. The rest was a sealed unit with a small vent on the back. 
 

This model of PL dwarf would go on to be installed at all the great PRR signaling achievements of the 1930's such as ZOO and HARRIS, however the march of progress is never satisfied and the desire to simplify the design merged with the desire to reduce confusion. The new dwarf would be reversed such that the curve would proceed in a high hand direction with the Restricting \ position being reduced in size compared to the Slow Approach / Position. The design was also changed to one with a single large rear access door, possibly enabled through advances in large gasket technology. 

The new style position light proceeded to almost entirely replace the old, especially as the staffed interlocking towers and their electro-mechanical interlocking machines were replaced by CTC starting in the 1970's. By the 21st century left handed PL dwarfs were a rare sight with most appearing at those parts of Amtrak and Conrail territory that had escaped re-signaling since the 1930's. However while out at LINCOLN interlocking on Amtrak's Northeast Corridor in Metuchen, NJ, I spotted a left handed survivor  governing movements out of the MoW yard.

Somehow escaping replacement in the early 90's when LINCOLN tower was closed and re-signaled, LINCIOLN's 5W dwarf is now one of the more accessible left-handed PL dwarfs, mounted on the side girder of the Lake St bridge for all to see.

Hopefully it will escape scrutiny and continue to perform its job for years to come. 

Amtrak's New Position Light Configuration

Following from Pennsylvania railroad practices, Amtrak used the following two configurations for all of its "complete" position color light mast signals at interlockings. They either have a lunar white Restricting \, or they don't. In both cases the single central lunar white Stop and Proceed marker has been retained.  

While most other North American railroads have shifted to using Restricted Proceed in place of Stop and Proceed, Amtrak, and the NORAC rulebook for whom Amtrak is the biggest member, have both retained the Stop and Proceed practice.  

So consider my surprise when saw the recently installed colorized position light signals at Amtrak's new LEGGETT interlocking on the Hellgate Line. The restricting \ is present, but, please correct me if I am wrong, I believe this is the first time I have seen Amtrak PL interlocking signals without a Stop and Proceed marker. 
 

Part of  the Penn Station Access Project, these colorized position lights are the first of many that will wipe out the last stretch of amber position lights on the NEC (PHIL is a point, not a stretch) and they appear to indicate a shift at Amtrak towards that all important stop before entering an occupied block.

In case your wondering the typical Amtrak signal progression is to display a Stop and Proceed if there is an obstruction within the interlocking limits and a Restricting (if available) if the interlocking is clear, but the block beyond is not. The stop certainly seems like a worthwhile safety precaution, but the distinction between it and Restricted proceed is minimal, mostly just the salience of entering a stretch of track with an obstruction or broken rail. 
 

As you can see the new signals on the replacement Portal Bridge have retained the Stop and Proceed marker, but that project was designed and specified some years before Penn Station Access. So it remains to be seen if this is the shape of things to come, or a one off trial.