can anyone give a good definition between "dual control points" and a "non dual control points"???

my understanding is the dual control has a hand throw lever ... and a non dual control has a hand crank

thanks in advance..

can anyone give a good definition between "dual control points" and a "non dual control points"???

my understanding is the dual control has a hand throw lever ... and a non dual control has a hand crank

thanks in advance..

In NSW, electric point machines are wound over from N to R and vice versa using hand crank handle stored in an Electric Switch Machine Lock (ESML) box, which in turn is locked by a SL key. The ESML Handle has a Kellogg Key welded on one end, and has switch machine specific key on the other end.

The ESML box is located a minimum of say 40m from the switch machine, so that the operator cannot act too hastily. (See head-on collision on Trans line (Zanthus)).

I cannot say whether such switch machines are described in NSW as "dual control" or "non dual control". Clearly a switch machine with an ESML are kind of dual control, but without using a throw over lever incorporated in the machine itself.

Maybe better terms would be:

Direct Dual Control: Hand throw handle incorporated in Switch Machine.

Indirect Dual Control: ESML handle stored say 40m from switch.

Non-dual control: No ability to operate points in an emergency.

BTW, Switch Machines operated pneumatically have emergency push buttons in a locked box, probably secured with a Lock.

See: https://en.wikipedia.org/wiki/Zanthus_train_collision

In Victoria they are termed DUAL CONTROL POINT  MACHINES .    They operate normally in motororized mode, and can be operated if required in MANUAL mode .

The smaller of the two visible levers selects either  MOTOR or MANUAL mode .
The larger throw lever when unlocked actually operates the points  Normal (N) or Reverse (R) .

When the smaller lever is moved to the MANUAL position it places all appropriate signals controlling movements over those points at STOP.

can anyone give a good definition between "dual control points" and a "non dual control points"???

my understanding is the dual control has a hand throw lever ... and a non dual control has a hand crank

thanks in advance..

This datasheet (http://www.rail-automation.com/downloads/1271vmQKwVaDLhR8.pdf) from Inversys describes the M3A and M23A point machines - these are identical except that the M3A can be manually operated by an emergency hand crank, while the M23A can be manually operated by a hand lever and is known as a 'dual control' point machine. Other manufacturers offer similar point machines.

The original intention - in the US around eighty years ago - was that dual control point machines were provided at remote locations where the points would have to be operated by the train crew during a failure. Ordinary point machines were provide at larger interlockings where they would be operated by a signaller or technician during failures.

In both types of point machines the risk is that they will be hand operated while a train is approaching them. This risk is fairly small - taking the hand crank from its lock or placing a dual control machine in hand mode will restore any signal reading over the machine to danger. The risk is slightly higher in the case of an ordinary point machine due to the time required to crank over the points (the lever on the dual control machine is faster). But this can be more than offset by mounting the hand crank at a distance from the point machine - by the time the hand crank has been brought to a machine any approaching train will have passed over the points or come to a stand at the signal.

The ESML box is located a minimum of say 40m from the switch machine, so that the operator cannot act too hastily. (See head-on collision on Trans line (Zanthus)).

See: https://en.wikipedia.org/wiki/Zanthus_train_collision

The ATSB report on Zanthus can be found at http://www.atsb.gov.au/publications/investigation_reports/1999/rair/rair1999001.aspx

The accident occurred because the point controls at Zanthus were a very poor design and completely ignored roughly 100 years of experience:
* There was no approach locking on the points (probably because they didn't want to install the necessary track circuits or detector).
* The point machines took a substantial time to throw the points (probably because they were DC battery driven using power from solar cells - this wasn't a poor design, but increased the risk which should have been managed by other aspects of the design)
* Once the points had been called, they could not be reversed until the points had completely thrown and locked (the operator realised immediately he had made a mistake, but couldn't do anything but wait for the points to ***slowly*** throw, and then he would have had to wait for them to ***slowly*** throw back to prevent the head-on collision).
* The operating controls were mounted on the side wall of a relay hut where the operator could not see approaching trains (the operator knew there was an approaching train, that was why his train was in the loop, but there was no visual cue reinforcing this knowledge).

The resulting design alterations were not really about preventing the operator from acting hastily. They were about emulating approach locking. As mentioned, the 'solution' was to put massive time delays in. Opening the control box caused the point indicator to show 'unlocked'. Then a ***long*** delay before the controls became active meant that any approaching train would have either come to a stand or be on the points.

Incidentally, throwing the switch in the face of an oncoming train and sending into an opposing train in the loop was a recognised problem in the US. The usual solution was to require the brakeman operating the switch to stand on the opposite side of the line to the switchstand until the approaching train had passed.

I cannot say whether such switch machines are described in NSW as "dual control" or "non dual control". Clearly a switch machine with an ESML are kind of dual control, but without using a throw over lever incorporated in the machine itself.

Maybe better terms would be:

Direct Dual Control: Hand throw handle incorporated in Switch Machine.

Indirect Dual Control: ESML handle stored say 40m from switch.

Non-dual control: No ability to operate points in an emergency.

I'd be surprised if there were any modern point machines that lacked any ability to hand operate when necessary.

Even ignoring operation during failure conditions, it is necessary to hand operate point machines during cleaning and maintenance work on both the points and point machines.

I forget how old EP (Electro-Pneumatic) could be controlled "by hand".

New installations have a "Fortress Lock", which allows push buttons to control the electric solenoids that operate the air cylinders. The main safety feature is a time release, so that after the necessary signal are put to stop, only after a wait of say 60 seconds can the points be moved.

The ESML Handle has a Kellogg Key welded on one end

Isn't that an Annett key? That's what they look like to me, anyway.

Kellogg keys (Milo G. Kellogg, founder of the Kellogg Switchboard & Supply Co., Chicago IL USA) are toggle switches found on telephone switchboards / concentrators and also on some early electric signalling panels for the control of signals, points and releases etc.

Isn't that an Annett key? That's what they look like to me, anyway.

Kellogg keys (Milo G. Kellogg, founder of the Kellogg Switchboard & Supply Co., Chicago IL USA) are toggle switches found on telephone switchboards / concentrators and also on some early electric signalling panels for the control of signals, points and releases etc.

You are quite right. The ESML has an ANNETT key welded on.

Must have got confused by all those double letters.

A KELLOGG key is quite small, with a plate about 40mm high * 25mm wide, with the handle sticking out about 30mm.

I forget how old EP (Electro-Pneumatic) could be controlled "by hand".

New installations have a "Fortress Lock", which allows push buttons to control the electric solenoids that operate the air cylinders. The main safety feature is a time release, so that after the necessary signal are put to stop, only after a wait of say 60 seconds can the points be moved.

The older Style 'C' EP points valves could be manipulated by maintenance staff to "blow" over a set of EP points.
The plunger lock valve was opened, making the reverse contact in the plunger lock, energising the valve lock.
At the same time the normal or reverse valve was opened to blow the points, depending on which way you were going.
This was usually a 2 man job unless you "jammed up" the P.L. valve first in the open position, commonly done with a fiber stick or small screwdriver.

The Style 'E' valve had a handle/key device that, when the micro-switch block was removed (failing the detection), could be inserted into the front of the valve unit and, once pressed home, toggled from side to side to operate the points, ie: normal or reverse. This also required the air to be still available.

Newer Style 'S' (and others) have press buttons.

In almost every case, there was no practical way to move the points if the air supply itself was not available, other than disconnecting the drive and barring them over.

When did the different kinds of power points and signals (air or electric) become available?

I think they started using air in the 1930s with the advent of pistol grip levers

As far as I know, the first power signalling scheme in NSW was for Sydney Station, brought into use 13 March 1910, replacing a large mechanical box of over 100 levers. Photos of this original "Station Box" in action, the adjacent "Tunnel Box" and the mechanical boxes they replaced are not often seen.


Equipment was provided mainly by the McKenzie, Holland and Westinghouse Power Signal Co. Ltd. of Worcester, England, including electro-pneumatic two position lower quadrant semaphore signals, points and some track circuiting but no track circuit control of signals or points. There was no track circuit block or automatic signals; the absolute block system remained in use. The first automatic signalling (in Australia) was introduced 22 June 1913 between Eveleigh Loco Jctn and Sydenham, consisting of two position lower quadrant home and distant semaphore arms. "Station Box" had a large McK, H & W miniature lever frame that was broken up and reused at (IIRC) Illawarra Junction, Sydenham and Petersham after its closure in 1924 (when Sydney Station West Signal Box - the one most people remember - was brought into use). Sydney Station East Signal Box was brought into use in 1916, also with a miniature lever frame. The large brick Station Box building was situated next to the Sydney Yard Controller's office and survived longer than the newer Sydney Station West.

I wonder how similar the original EP point mechanisms from Sydney Station were to the 1920s era NSWR Signal Branch designs which seem to remain in use throughout the old EP area.

When did the different kinds of power points and signals (air or electric) become available?

The first pneumatic interlocking was installed as a trial in 1876 on the PRR in the US. George Westinghouse (of the brake frame) became interested in power interlocking and formed the Union Switch & Signal Co in 1881. Their first products were hydraulic type and were installed in 1883. After 1884 the US&S developed the EP system, however it was not considered safe to use air to operate points. In 1891 the US&S developed the EP switch valve - already mentioned here at Sydney - and EP point mechanisms became safe and feasible. McKenzie, Holland & Westinghouse company was a joint venture to push this technology into the UK (and empire) market. Eventually it was re-organised (with most of the other UK signalling companies) into Westinghouse.

Electric point motors were first introduced in 1888 by Ramsey and Weir (in the US). J.D. Taylor - again in the US - developed the 'all electric' system which was first installed in 1889. Taylor's company eventually formed a foundation part of the General Railway Signal Company.

The UK was considerably behind. The first main line use of power operation of points wasn't until 1899 - EP at Granary Junction (using US&S technology) and Crewe (the LNWR's own system of electric points operation).

As far as I know, the first power signalling scheme in NSW was for Sydney Station, brought into use 13 March 1910, replacing a large mechanical box of over 100 levers. Photos of this original "Station Box" in action, the adjacent "Tunnel Box" and the mechanical boxes they replaced are not often seen.


Equipment was provided mainly by the McKenzie, Holland and Westinghouse Power Signal Co. Ltd. of Worcester, England, including electro-pneumatic two position lower quadrant semaphore signals, points and some track circuiting but no track circuit control of signals or points. There was no track circuit block or automatic signals; the absolute block system remained in use. The first automatic signalling (in Australia) was introduced in 1914 between Eveleigh Loco Jctn and Sydenham, consisting of two position lower quadrant home and distant semaphore arms. "Station Box" had a large McK, H & W miniature lever frame that was broken up and reused at (IIRC) Illawarra Junction, Sydenham and Petersham after its closure in 1924 (when Sydney Station West Signal Box - the one most people remember - was brought into use). Sydney Station East Signal Box was brought into use in 1916, also with a miniature lever frame. The large brick Station Box building was situated next to the Sydney Yard Controller's office and survived longer than the newer Sydney Station West.

I wonder how similar the original EP point mechanisms from Sydney Station were to the 1920s era NSWR Signal Branch designs which seem to remain in use throughout the old EP area.

The six-platform station at Brisbane Central was controlled from aremarkably small signal cabin. There was a reason for this. This box opened in1904, replacing two mechanical boxes controlling either end of the station dating from 1891.

The 36-lever "B" battern electro-pneumaticframe, built by McKenzie & Holland, was in fact the first power installation in the whole of Australia.This frame was of just 24 levers when new, but was enlarged to 36 in July 1917

http://www.signalbox.org/overseas/australia/brisbanecentral.htm

Just shows that Queensland is not as backward as southerns like to think.

The UK was considerably behind. The first main line use of power operation of points wasn't until 1899 - EP at Granary Junction (using US&S technology) and Crewe (the LNWR's own system of electric points operation).

The UK was behind when it came to implementing power signalling technology, but the story doesn't really end there, because the British were miles ahead when it came to actually using proper interlocking and the absolute block system, a situation which essentially continues to this day.

The early power systems in the US were at the forefront of technology but only ever covered a small proportion of route mileage, versus the busier and mostly double line UK network which was already almost fully signalled and interlocked using mechanical equipment by the time power signalling was becoming common in the US. Having just spent large sums of money on mechanical equipment, it appears the private British railway companies (whose fortunes were in decline in the early 20th century) mostly just elected to keep it except where the upgrade to power signalling was economically justified (in addition to the tales of prejudice commonly cited). The American railroads who have in more recent times simply switched off life expired power signalling systems without replacing them adopted a similar attitude. The availability of cheap labour to work the equipment was probably also a factor, with many American lines running through sparsely settled areas and virgin bush compared to the more densely populated UK.

Also of relevance is the fact that George Westinghouse sought technical assistance from the UK when he set up US&S, primarily from Saxby & Farmer of Kilburn, pioneers of large lever mechanical interlocking frames and supplier to US railroads; there were a number of British engineers working for US&S including a general manager at one point. They may not have invented the power signalling equipment, but they were apparently used to ensure British signalling safety principles were broadly applied. Given the continuing lack of proper interlocking to British standards in many parts of the US, it seems money always came first... as it does everywhere to a greater or lesser degree.

The UK was behind when it came to implementing power signalling technology, but the story doesn't really end there, because the British were miles ahead when it came to actually using interlocking and the block system, a situation which essentially continues to this day.

There are a couple of things being discussed here, but the invention versus application comparison is an interesting one.

The American example demonstrates that power signalling technology does not necessarily mean higher safety integrity of the system overall.

Despite the great innovation of American signalling firms like GRS with NX route control, APB, CTC and ATP technology, the American railroad companies themselves were behind in terms of enforcing tight safety regimens bolstered by strong engineered controls. If the rules are rubbery or there is a heavy reliance on purely administrative controls to guard against human error, power equipment is no better than barbaric mechanical levers, rods and wires, and in fact it could be argued that it is worse due to its "deadly convenience".

It is rather ironic that the relatively primitive British manual block system had a better safety record than pre-APB/CTC American lines with automatic block signalling, but that is what happens when signals are relegated to the role of a guide, as an overlay on the pre-existing timetable and train order system that still forms the basis of American main line working.

The American railroads who have in more recent times simply switched off life expired power signalling systems without replacing them adopted a similar attitude.

The Australians have done a similar thing to various country lines which were "designalled" from the 1990s onwards, although it was more a case of ripping out old mechanical signalling and replacing it with el cheapo "alternative" systems heavily reliant on administrative controls, as the traditional American system always was.

The UK was behind when it came to implementing power signalling technology, but the story doesn't really end there, because the British were miles ahead when it came to actually using proper interlocking and the absolute block system, a situation which essentially continues to this day.

I would agree with you c1900. I wouldn't agree c1950, and I wouldn't agree that the UK is leading today.

The UK achieved a remarkable level of signalling by the mid 1890s. As you say this was far and away in advance of any other country at the time and was based on mechanical interlocking and manual block working. However, signalling in the UK then largely stagnated for the next 50 years. Signalling innovation was largely restricted to minor tweaks, and classic mechanical signalling and block working largely saw out steam. Even with traditional signalling, I would suggest that other countries surpassed the UK in the twentieth century - in Germany, for example, the standard was lock and block, not plain block.

The baton for signalling innovation passed to the Americans in the mid 1890s. The signalling technology of the twentieth century was largely developed in the US - track circuits (both AC and DC), automatic signalling, light signals, power operation of points, power interlocking, relay locking, CTC, train control... Such modern installations that were provided in the UK almost exclusively used technology that was developed in the US - to the extent that even today the standards for naming relays in the UK is recognisably based on the US scheme.

The result of this stagnation in the UK and rapid development in the US was an evening out of the safety race. Shaw, in his book on US accidents, asserts that by WWII the accident rate in the US *per train mile* was equivalent to the UK. There were a large number of accidents, but this was because the US railway system was far bigger than any other system and had more trains. Any casual reading of the UK accident reports in the '50s will be struck by the number of serious accidents caused by block failures or misreading signals. This didn't really improve until the massive resignalling schemes really started to take their toll of traditional signalling - schemes that were largely based on US technology.

Today? IMO the UK has largely stagnated again. This time the signalling advances are largely coming out Europe - the UK, for example, is way behind the rest of Europe in the installation of any form of ATP. The development of SSI was the one bright spot...

I would agree with you c1900. I wouldn't agree c1950, and I wouldn't agree that the UK is leading today.

There is a big difference between inventing technology and using it, as Headshunt said. The Yanks were great at inventing signal technology that could greatly improve safety if it was used specifically for that purpose, but unfortunately USS and GRS were not the people running trains - they had to sell their equipment to penny pinching railroad operators whose safety philosophy was somewhat different to that which had prevailed in the UK from 1889. There were requirements like the ICC ruling for ATP on passenger lines operating at 80 mph and over, but many railroads circumvented such expensive provisions where possible (e.g. by imposing a 79 mph limit, only equipping the fast passenger trains or one insignificant route ), just as many British railway companies did not want to spend money on upgrades. That said, up to 10,000 miles of American railroads had continuous or intermittent ATP by 1960, but that declined as the airlines killed off the passenger rail business to the extent that there isn't much left outside the Northeast Corridor (and commuter networks). The US was also very good at designing crashworthy rolling stock that outperformed European carriages, but given their flimsy safe working procedures and the frequency of "cornfield meets" it was just as well.

the invention versus application comparison is an interesting one.

The American example demonstrates that power signalling technology does not necessarily mean higher safety integrity of the system overall.

Despite the great innovation of American signalling firms like GRS with NX route control, APB, CTC and ATP technology, the American railroad companies themselves were behind in terms of enforcing tight safety regimens bolstered by strong engineered controls. If the rules are rubbery or there is a heavy reliance on purely administrative controls to guard against human error, power equipment is no better than barbaric mechanical levers, rods and wires, and in fact it could be argued that it is worse due to its "deadly convenience".

That's what I was getting at - the "integrity" of the safety system - a product of the operating philosophy, rules and technology combined.

The technology was there, but most railroads still used a system (the Standard Code of Operating Rules or SCOR) where accidentally leaving a few words off the end of a train order could cause a fatal head on collision. That simply wouldn't have happened on the other side of the pond in ca 1900, but it continued well beyond 1900 in the US despite new fangled technology that could have prevented it.

When I said "a situation that essentially continues to this day" I meant that safety standards for interlocking and the block system in the US still fell behind the UK until recent times, despite the use of more advanced technology; I'll provide a few examples below as to why I believe that to be the case.

It is rather ironic that the relatively primitive British manual block system had a better safety record than pre-APB/CTC American lines with automatic block signalling, but that is what happens when signals are relegated to the role of a guide, as an overlay on the pre-existing timetable and train order system that still forms the basis of American main line working.

1. The block system with track circuit controlled automatic signals has been used in the US since the 1870s (after they invented it) but, for reasons of economy, flexibility and the established operating philosophy, it is mostly stop and proceed (permissive) working outside "interlockings", which reduces safety margins and undermines the safety benefit of the technology. As Headshunt said, automatic signals were more of a guide to the presence of other trains on the line ahead than the primary authority for train movements which was still the timetable and any train orders that may have been issued. Meanwhile in the UK it was the opposite - the signals (and token) were the authority for trains to move and the timetable was essentially only a guide; absolute block working was already the norm before 1900 and mandated on passenger lines; permissive working was highly restricted; British train controllers could not arrange a head on collision by mistake in drawing up their graph as an American train dispatcher could because they merely passed their instructions to signallers who worked fully interlocked and strictly controlled absolute block sections. See 2 regarding conditions at "interlockings".

2. Main lines running through shunting yards that were normally fully interlocked in the UK were not always given the same level of protection in the US. Well into the second half of the 20th century you could find yards full of non-interlocked facing points connecting to main lines on which passenger trains (the few that actually still ran outside metropolitan commuter networks by that time) routinely ran at 60 mph or more. Sometimes the only proper warning of the position of the points provided to drivers was pieces of green and red reflective sticky tape on the switch stand. No electric release from the signaller required, no approach locking, no Annett key... nothing - just unlock the frame and throw the points. Did it lead to crashes? Hell yes! Would it have been allowed in the UK (or most of Australia) in the 1970s/80s? Probably not. That's basically 1830s stuff still going on in the 1980s and it probably still happens now. Positive Train Control might finally put an end to it, if it is ever fully implemented.

3. Compare single line crossing stations, passing loops etc of the same era on both sides of the pond and again you'll find the US typically falls behind UK practice due to allowing simultaneous reception with short overlaps and no catch/trap points and sometimes automatic signalling on the main line but no proper signal protection or interlocking for the loop. It speeds things up and cuts costs but also cuts safety margins.

Most of this makes sense when it is realised that on classic US main lines without CTC or ATP the timetable was the primary authority for trains to move, not the signals.

4. Single line working during engineering work - by pilotman in the UK vs a form of telephone block in the US with no pilotman and no token for the single line section. Still happening in the 1980s and no doubt beyond.

The result of this stagnation in the UK and rapid development in the US was an evening out of the safety race. Shaw, in his book on US accidents, asserts that by WWII the accident rate in the US *per train mile* was equivalent to the UK. There were a large number of accidents, but this was because the US railway system was far bigger than any other system and had more trains. Any casual reading of the UK accident reports in the '50s will be struck by the number of serious accidents caused by block failures or misreading signals. This didn't really improve until the massive resignalling schemes really started to take their toll of traditional signalling - schemes that were largely based on US technology.

I think to say that the large number of accidents in the US was simply due to the size of the network overlooks the relatively flimsy safe working procedures and standards applied, even in the presence of power signal technology like automatic block signalling. Yes, the network was huge, but so were the holes in the system. I think Robert Shaw paints rather a rosy picture with his selective crash stats. Such stats can also be skewed the other way very easily, e.g. according to one source on accidents involving trains carrying US Mail alone, there were as many as 9,355 crashes with over 200 US Mail employees killed and 1,500 seriously injured between 1876 and 1905 - that's one crash every 27 hours. Many of those crashes would have been minor, but that's still quite a staggering figure. Another source claims it was only one crash every 39 hours for mail trains between 1875 and 1917 for a total of about 9,400. Either way, American trains operating under the classic SCOR crashed often.

It's true that the 1950s was not a good decade in the UK, where the holes in the manual block cheese aligned with catastrophic results on several occasions. The holes certainly were there to be exploited by an unfortunate convergence of circumstances and technology invented in the US helped to plug them. However, that Shaw appears to have used the UK as a benchmark for safety in his comparison is perhaps revealing.

Looking at the modern rail safety stats, it seems the UK is still among the top European performers despite the troubles of privatisation and a widespread reliance on pre WW2, if not pre 1900 technology. If the stats in the ORR Health and Safety Report 2013 can be trusted, it just goes to show there is more to rail safety than high tech equipment.

Such modern installations that were provided in the UK almost exclusively used technology that was developed in the US - to the extent that even today the standards for naming relays in the UK is recognisably based on the US scheme.

Could Historian supply a list of US abbreviations for signalling relays, please?

Could Historian supply a list of US abbreviations for signalling relays, please?

For NSW codes (based on UK) see: For NSW codes, see: http://www.railpage.com.au/f-p1910900.htm#1910900

There is a big difference between inventing technology and using it, as Headshunt said...

I don't think that you've really understood my argument.

Essentially it is economic. The provision of signalling equipment provides two benefits - increased line capacity and reduced risk. (To a certain extent these benefits can be traded - the goal of signalling is often to increase capacity without changing the risk.) But there is an opportunity cost to the provision of signalling. The money spent on a particular signalling installation (both in providing and operating) cannot be spent on other projects.

In the UK the BoT essentially mandated in the 1890s a uniform level of signalling infrastructure over the entire country. There was *no* provision for varying the infrastructure depending on traffic. (There was subsequently some backing away from this position - light railways had lower infrastructure requirements. But these were never numerous in the UK.) This did provide a very safe railway, but at a price. Essentially the railway companies must have been wasting resources over signalling their branch and minor lines. These resources was then not available to be spent in other areas - such as improved signalling on their main lines.

This waste was probably managable before WWI, but after that the strain showed. It is clear from accident reports that the companies were under investing in safety (new technology) on main lines - as shown by accidents caused by lack of ATP, by lack of track circuiting, and by lack of electric signals. I would consider this came to a head in the '50s. It wasn't just a matter of having a bad decade with the holes lines up. It was a consequence of lack of investment, and part of the problem must have been the deadweight of money spent on the less important lines.

The US during the same period is a useful counterpoint. While you criticize the railways for the lack of uniformity and the manual processes, you fail to note that it was completely impossible for anything approaching a uniform signalling system to be used in the US. And certainly impossible for any technology to be uniformly adopted. The US had lines with far greater traffic than anything in the UK than anything in the UK. Equally, they had huge mileages of line with, at best, one or two trains a day.

I would agree - as I said in the previous posting - that the US managed safety poorly in the 19th century. But safety management improved greatly in the 20th. And this was done more efficiently than in the UK. Safety technology was applied where the risks were highest. Important lines got full automatic signalling with interlocking and (eventually) cab signalling. Minor branch lines had no signalling and the main line points were secured by switchstands. In the middle were vast mileages protected by automatic signalling superimposed over TT&TO working - because the major risk in this form of safeworking was rear end collisions.

The result was that by the '50s, Shaw notes that the US railways were as safe as the UK railways per train mile (i.e. using a statistic reflecting the size of the network, not an absolute measure of number of accidents).

It is interesting at this point that the UK began a crash program of modernising their signalling - using largely US technology. This investment, however, was focussed on the most important lines - in essence they abandoned the uniformity of the signalling. If you like, they didn't just adopt US technology, they also adopted the US approach to signalling. The main lines got MAS. Secondary lines remained mechanical with little investment. Branch lines were ruthlessly de-signalled back to the 'basic railway'. Safety went up and costs went down.

Incidentally, I think that Australian railways were far better than the UK at adopting signalling systems that suited traffic levels.

In the UK the BoT essentially mandated in the 1890s a uniform level of signalling infrastructure over the entire country. There was *no* provision for varying the infrastructure depending on traffic. (There was subsequently some backing away from this position - light railways had lower infrastructure requirements. But these were never numerous in the UK.) This did provide a very safe railway, but at a price. Essentially the railway companies must have been wasting resources over signalling their branch and minor lines. These resources was then not available to be spent in other areas - such as improved signalling on their main lines.

Did Beeching ever think of introducing low level signalling in the UK to save lines from complete closure?

I don't think that you've really understood my argument.

Essentially it is economic.

I fully appreciate that it is a case of economics and I thought I mentioned factors relating to the differing circumstances, fortunes and motives of private railway companies in both the UK and US, governmental regulation for safety in both countries etc. However, economics is not a strict science.... it's not science at all. There is room for many views when it comes to assessing what is acceptable or justified and what balance of safety vs other factors is to be struck. Your point that most of the power equipment was invented in the US stands as a simple fact.

The US during the same period is a useful counterpoint. While you criticize the railways for the lack of uniformity and the manual processes, you fail to note that it was completely impossible for anything approaching a uniform signalling system to be used in the US. And certainly impossible for any technology to be uniformly adopted. The US had lines with far greater traffic than anything in the UK. Equally, they had huge mileages of line with, at best, one or two trains a day.

Yeah, as I said, I am critical of things like:

• running passenger trains at 60 mph through yards with non-interlocked points connecting to the main lines, the operation of which was mostly governed by written rules rather than engineered safety equipment to protect against human error and "open switch" accidents, well known in the US but much rarer in the UK since the 1880s. I don't think it's a very good idea. The suits do their sums and make a rational business decision, but they aren't the ones who can get killed when something goes wrong.


• a lack of full interlocking to UK/Australian standards on mostly freight railroads even if the economics says it is right ("right" meaning the odd payout of compensation to families of dead train crew and potentially passengers is justified because the cost of mitigating against them is undesirable or prohibitive)


• telephone block for opposing movements during single line working, because it is cheaper and faster than working by pilotman

• generally, allowing the market to determine what safety precautions are necessary, justified etc - something even the US authorities were not satisfied with, although their views on safety vs economic efficiency were better for business than in the UK


• simultaneous reception at crossing stations with short overlaps/no ATP

Why don't I like simultaneous reception?
Here's an example of what can go wrong:


http://www.youtube.com/watch?v=oXUSUAYMwWQ

Luckily, no-one was killed at Kismet CA on 14 June 2006, but one crew baled out of the offending train and it's a terrifying video clip; the after shots of the locos are also pretty scary. The repair bill was over $5,000,000.

A lack of uniform signalling systems is not, in itself, a major concern to me, especially since a uniform standard might not be adequate. In fact, I like the variety. I'm mainly concerned with safety integrity, what can go wrong and what has gone wrong.

In the UK the BoT essentially mandated in the 1890s a uniform level of signalling infrastructure over the entire country. There was *no* provision for varying the infrastructure depending on traffic.

The requirements were more or less the same for passenger lines irrespective of traffic density (not goods lines). Despite the economic argument against such a position I'm not convinced that it would have been advisable to have lower safety standards for quieter lines or allowed to private companies to self-regulate what standards were required. A similar cost vs safety argument is also seen in the airline safety, which is dominated by for-profit entities with a revolving door between manufacturers and regulators that has led to dubious safety decisions and coverups.

The impact of spending money on less important lines is not merely a case of a burden or deadweight on the rest of the business, but I agree it would have been a cost that for-profit companies would have preferred to avoid and worse still made it harder to adopt power signalling in the following decades. By the 1950s when a number of serious accidents occurred, most of those less important lines were already decades old, the signalling was still relatively basic (e.g "one engine in steam" on quiet single lines), largely unchanged since it was first installed and amortised.

Shifting the focus from money to safety, what about the safety benefit that expensive equipment conferred on those less important lines? The 1889 UK lock, block and brake laws resulted in a high degree of mechanical signalling and a high degree of safety. It's fairly safe to assume there was a safety benefit even on lesser lines, however unjustified or inefficient the beancounters may have considered it.

How many lives did the 1889 laws save over the following century? I'd like to believe they saved more than they cost, but the problem in defending the cost burden on operators is that because the accidents didn't happen, it's not known how many were prevented by the expensive equipment.
How many times did signalmen reach for locked levers - levers that they would have thrown if they were not interlocked?
How many times did a signalman offer a train to the box in advance only to be reminded that there was already a train in the section by the block instrument or cop an earful of abuse on the phone after sending the "Is Line Clear?" bell signal?
Block working failure could be disastrous, but the humble block instrument worked far more often as a reminder than it didn't, and could be enhanced by the various lock and block systems that were in use (also used in NSW) or replaced with primitive manual track circuit block arrangements (also popular in NSW where a number are still in use). An automated, power operated system is not necessarily safer than an enhanced manual one.

Despite their lower traffic density, would the potential for head on collisions still have been there if token working had not been adopted on lesser single lines? I think so. The fact is, token working was on the whole very effective at minimising the risk of head on collisions on single lines despite its cost in terms of time and money and the occasional crash when the wrong staff was given to a driver who failed to check it etc.

Would the potential for open switch accidents have existed if points were not interlocked on lesser lines? I think so. We can see from early UK reports the sorts of accidents that used to happen before the block system and interlocking, systems that were adopted for (apparently arguably) good reasons.

It wasn't just a matter of having a bad decade with the holes lines up. It was a consequence of lack of investment, and part of the problem must have been the deadweight of money spent on the less important lines.

The holes in the manual block safety system were present in 1890, and in some cases still in 2014, waiting to be exploited by a convergence of circumstances. Sometimes it takes decades and increased operational pressure for a disaster to occur. Sure, many of them could have been and eventually were plugged by investment in American invented equipment tailored to suit local conditions.

Important lines got full automatic signalling with interlocking and (eventually) cab signalling. Minor branch lines had no signalling and the main line points were secured by switchstands. In the middle were vast mileages protected by automatic signalling superimposed over TT&TO working - because the major risk in this form of safeworking was rear end collisions.

Important lines in the US got full automatic signalling, but not necessarily full interlocking by British and Australian standards, as I covered in my last post.

I find it hard to agree that the major risk on single lines is rear end collision, but I guess it depends how you define "major risk"; that definition could be tailored to suit different locations. The obvious major generic risk to me is that of a head-on collision in between stations - a "cornfield meet" as they are known the US and which happened often under timetable based working.

Basic installations of automatic block signalling certainly reduce the risk of rear end collisions on single lines but they don't prevent cornfield meets. Returning to the horrendous safety data for US Mail trains (over 9,000 crashes from the mid 1870s to 1910s), bear in mind that automatic signalling and full interlocking was already available for most of that period. "Efficient" safety management is easy when the cost of human life is low and government regulations loose or non existent.

Incidentally, I think that Australian railways were far better than the UK at adopting signalling systems that suited traffic levels.

I'm sure C.B. Byles, NSWR Signal Engineer formerly of the Lancashire & Yorkshire Railway in the UK enjoyed the government railway budget that allowed him to instal loads of track circuiting and other American-made or derived power equipment in Sydney and busier country lines in NSW at the expense (and (arguably - if you want to talk efficiency) to the benefit of) taxpayers, many years ahead of most of the UK. Even before his arrival in Australia, Byles was a strong proponent of power signalling and important member of the IRSE along with several of his contemporaries in the UK such as A.F. Bound. The result at least in NSW was still a railway run in accordance with British practice, but a safety integrity often better than England and higher than most of the US outside best practice commuter networks and main lines with continuous ATP and CSS. NSW was also fond of the Westinghouse air brake and automatic couplers, both American inventions.

An example of adjusting signalling to suit traffic is the Molong to Dubbo Absolute Permissive Block installation, known locally as Single Line Automatic, brought into use in 1925. This Anglicised application of American technology was a technical success but failed from a cost-benefit perspective due to lower than anticipated levels of traffic on the line. It was replaced with divisible miniature electric staff in 1933 - older, more primitive and manual British technology.

I would agree - as I said in the previous posting - that the US managed safety poorly in the 19th century. But safety management improved greatly in the 20th. And this was done more efficiently than in the UK. Safety technology was applied where the risks were highest.

And they removed that technology when they could. The economic rationale for lowering standards in the absence of passenger traffic is understood (non passenger lines in the UK also had lower standards), but when the odd passenger train still runs on a mostly freight line, the risk to passenger safety with low safety standards is still there; the risks to employee safety are ever present even if the economic stakes are not particularly high.

by the '50s, Shaw notes that the US railways were as safe as the UK railways per train mile

Robert Shaw's calculation is interesting. Shouldn't safety in the technologically superior US have been better than the inferior UK, not just equal? For me this only reinforces the notion that there is more to rail safety than technology. As I said in the last post, if Shaw used the technologically inferior UK as a safety benchmark for comparison, that is perhaps revealing. Anyway, by the 1950s US intercity passenger rail was already in decline as the golden age of the automobile and airliners was taking hold. The reality is a long distance train could crash with minimal casualties because there were sometimes only a handful of passengers on board. The reluctance of railroad companies outside a select few to comply with the 1947 ICC requirement for ATP for operating at over 79 mph must have had an impact on the attractiveness of passenger rail services in the face of competition from cars, buses and airlines, but then again had they installed it and maintained high speeds it's likely that it would merely have delayed the inevitable.

High tech American ATO outfits can suffer wrong-side failures and crashes like this one on the Washington Metro on 22 June 2009 with GRS-Alstom equipment. The death toll was not as bad as Clapham Junction, the landmark wrong-side crash, but scary because it was a you beaut high tech system. The rolling stock was also uncrashworthy.


http://www.youtube.com/watch?v=KHMosix9bQ0

Incidentally, I think that Australian railways were far better than the UK at adopting signalling systems that suited traffic levels.
- A user

Before CTC and TOW, single lines in NSW had basically two levels of signalling:

* Annett Key locked loops, with all points operated by ground frames.
* Centrally controlled loops with loop points controlled from the signal boxes, and ground frames for other sidings.
These two forms might intermix on lines.
Unanderra - central
Dapto - key locked (now CTC with simultaneous arrivals)
Albion Park - key locked (now CTC with SA)
Shellharbour - central (now CTC with SA)
Bombo - central (now CTC)
Nowra - central (Up and Down Working) (now CTC with Main and Loop Working)
Berry - key locked (though an early 1890 diagram shows it being Central) (now reduced to a crossing siding)
Nowra - key locked.