Metric Railway

[snowball]

Mars Climate Orbiter.

 

[Up_Tilt_390]

First rule of Engineering if it is not broken don't fix it.

Unless I've misunderstood what you meant to say with that, then that kind of uses the argument I specifically debunked at the start of the thread saying "if the imperial system wasn't broken, the simpler metric system wouldn't need to exist. Correct me if I have misunderstood your argument of course, because I am very interested to hear people's thoughts on metrication of the railway.

 

[Murph]

Unless I've misunderstood what you meant to say with that, then that kind of uses the argument I specifically debunked at the start of the thread saying "if the imperial system wasn't broken, the simpler metric system wouldn't need to exist. Correct me if I have misunderstood your argument of course, because I am very interested to hear people's thoughts on metrication of the railway.

The imperial measurement system is not broken, and never has been broken. The problems are nothing to do with imperial vs. metric, but either bad records keeping, records not being adequately kept up to date, human error at the point of measurement, or outright incompetence of someone involved. All of those problems exist equally in both measurement systems, and change for the sake of change will do nothing to improve that. Frankly, if the people responsible for the measurement can't cope with reliably converting between imperial and metric, they should not be doing the job.

The people who really need to deal with the measurement, such as the surveyors and engineers, they can cope equally well with both systems, as they have been doing that successfully for a very long time. New entrants into those fields should really not have any difficulty in learning the systems from more experienced people; and if they do have problems learning that then they probably should not be doing a job of that nature.

 

[Maurice3000]

What system the network is measured in is irrelevant, except that doing it in imperial is probably easier than metric, as you would just be extending existing and updating the existing data sets.

The problem is that the existing data sets already use three different measurement systems and are considered to be a mess. NR have been receiving a lot of flak about it in recent years. Not only will the current mess be expensive to maintain, it's also a nightmare to manage and, quite possibly, even dangerous.

As for 99.99% of all engineering being done in metric, the USA alone probably makes that a lie.

I just did a quick calculation and Americans make up approx. 0.00000005 percent of the world’s population. If you then consider that large parts of American engineering is already metric (because otherwise nobody would buy their products, because most their drawings are in metric, because most of their tools are in metric, because most of the parts they need are in metric, because most of the materials they use are in metric, because the factories in Asia they outsource to only understand metric etc. etc.) then precious little engineering in the world is done in anything other than metric.

The conversion of current records to metric, would cost millions for no real benefit as multiplying a record by 0.914 does nothing to improve its accuracy. It would also add an extra level of disconnect in that older paper records would be in one system, with newer in another.

Who is talking about multiplying anything? That wouldn’t help anything as you say. You just draw a line in the sand to fix things from now on:

1. Don’t trust any old records.
2. Any new measurement is done exclusively in metric.
3. Any new track is laid in metric.
4. Any upgrade is planned in metric.
5. All new signalling systems work in metric. Practically all trains we’re build in metric anyway so all you'd need is to change the speedometer in the cab when they get the new signalling system retrofitted.

First rule of Engineering if it is not broken don't fix it.

Agreed. And because the current mess is definitely broken NR has decided to move to metric because they want to fix it. What's the benefit of prolonging the mess when the world has moved on?

 

[TDK]

It's about time that time was easier to work out, 100 minutes in an hour, 20 hours in a day maybe but that will never change.

For me personally KM/h MPH miles, kilometres or any type of measurement wouldn't confuse me driving a train in the slightest. If the speed is signed at 70mph or 110km/h that what you travel at as long as the speedometer is in the correct format.

 

[Up_Tilt_390]

The imperial measurement system is not broken, and never has been broken.

I suppose it depends on how you define broken. When I say broken, I mean that the system has no scientific basis and as such is very tricky to remember and learn things. For example...

5280ft = 1760 yards = 1 mile
10 chains = 1 furlong
80 chains = 8 furlongs = 1 mile
12 inches = 1ft
36 inches = 3ft = 1 yard
16 ounces = 1 pound
224 ounces = 14 pound = 1 stone

As evident from what I've listed above, the imperial system has no particular scientific basis, and this means it can be difficult for people to learn and remember the system. Compare this to the metric system...

1000 millimetres = 1 metre
1000 metres = 1 kilometre
1000 milligrams = 1 gram
1000 grams = 1 kilogram
1000 kilograms = 1 metric tonne

This is when I consider the imperial system broken, because the metric system has a scientific basis, with everything being based on 10 (I think) and this makes it easier to remember and learn, along with the fact that even with different units of weight, the prefix is the same, so you can remember what is heavier than what.

 

[Trog]

How many minutes do you think it would take someone suitable to be trusted with the safety of the line to learn your little list. As for throwing out all existing records and starting again, that is a crazy idea, as some things are just not available to measure again. Also remember that the staff are used to these records and their eccentricities we know which are good and which are a bit iffy. If I look at a list of alleged track components most of the dodgy records would not stick out more if they were printed in fluorescent ink. As I know the year ranges the different rail sections and sleeper types were used and which combinations are possible, and which are not.

As for the scientific basis of the metric system the metre was originally defined as being one ten millionth of the distance between the North Pole and the Equator through Paris. Which as it would relate directly relate to the size of the earth would make metric units easy to use for navigation. The only trouble was that the base measurement on which the metric system is based was wrong, which is one of the reasons the nautical mile and knot remain in use.

 

[Emblematic]

You obviously know little about the railway and its systems, distance measurements in imperial are common, and many current systems use imperial or sometimes both imperial and metric depending on what is most suitable.

The conversion of current records to metric, would cost millions for no real benefit as multiplying a record by 0.914 does nothing to improve its accuracy. It would also add an extra level of disconnect in that older paper records would be in one system, with newer in another.

First rule of Engineering if it is not broken don't fix it.

It's the railway, at least the Network Rail part of it, that has decided to metricate speeds and distances during transition to ETCS and ERTMS. The system and regulations would have permitted imperial measurements to be used, if desired. They must have their reasons.
"If it's not broken don't fix it" - I thought that was the first law of Bangernomics (the second being "if it IS broken don't fix it")

 

[jimm]

It's the railway, at least the Network Rail part of it, that has decided to metricate speeds and distances during transition to ETCS and ERTMS. The system and regulations would have permitted imperial measurements to be used, if desired. They must have their reasons.
"If it's not broken don't fix it" - I thought that was the first law of Bangernomics (the second being "if it IS broken don't fix it")

I should the think the reason is very simple. That all the signalling suppliers work in metric and Network Rail is not going to waste money it doesn't have getting the software rewritten in miles and chains just to satisfy Trog...

 

[Up_Tilt_390]

As for throwing out all existing records and starting again, that is a crazy idea, as some things are just not available to measure again.

I don't recall ever saying anything about that, nor do I recall implying it, maybe I did and I have a bad memory, but I'm pretty sure I didn't mention what you said there. The idea I put forward is that the railway starts switching to the metric system. Old staff can stick with their primitive measurements, but as people start to learn metric in school, we need to keep up with the times and do the same for the railway.

Older generation workers probably won't be able to get out of the habit of imperial measurements, but young people are being taught metric as the years go on, so it's a step backwards to give them old measurements they weren't taught should they go on the railway as a career path.

 

[Trog]

I just did a quick calculation and Americans make up approx. 0.00000005 percent of the world’s population. If you then consider that large parts of American engineering is already metric (because otherwise nobody would buy their products, because most their drawings are in metric, because most of their tools are in metric, because most of the parts they need are in metric, because most of the materials they use are in metric, because the factories in Asia they outsource to only understand metric etc. etc.) then precious little engineering in the world is done in anything other than metric.

I think given the above a simpler system than metric is required for the younger generation as if you divide the population of the world 7 billion by the population of the US 320 million it turns out that about 4.6% of the worlds population live in the USA. Also as they have the largest economy they probably build more per head than their simple numbers would suggest.

But what would an old dinosaur like me know compared to the up and coming, exponents of metrification, after all I only have a few decades of doing the job to go on.

 

[Up_Tilt_390]

Also as they have the largest economy they probably build more per head than their simple numbers would suggest.

Actually, I believe that China has overtaken the United States in terms of the largest economy if I'm not mistaken...

 

[Nym]

*cough*

Manchester Airport Branch - 1993
Windsor Link - 1989
Selby Diversion - 1983

East London Line Extensions - c.2007-2009
Jubilee Line Extension - 1999
Docklands Light Railway - 1980s
Fleet Line / Jubilee Line - 1977

 

[DerekC]

To answer the specific question - in ERTMS (ETCS) the train displays to the driver in the units set by the infrastructure manager (NR) and transmitted to the train in either a radio or balise message (I forget which). So the driver can't choose.

I wouldn't comment on the Imperial system being "broken" or "out of date" because these are subjective terms. It undoubtedly is complex (contains lots of different units of length, weight etc which have different names that you have to remember) and illogical (16 oz = 1 lb; 14 lb = 1 stone etc etc) and engineers have been trained to think in metric since at least the 1960s (can vouch for this personally).

It is correct to say that all modern control systems, rolling stock, track etc etc are specified and designed in metric units. Converting them to imperial would be a waste of money and would result in GB railways costing even more than they do. The bit which keeps slipping away from us is the infrastructure data and the driver-facing stuff which is all still in miles, chains and yards. NR has a programme called "ORBIS" which is intended to rationalise all the different measurements and reference points and will allow everything to be metricated. It will happen one day - just a question of when!

The Technical Specifications for Interoperability (TSIs) are a mandatory requirement of the Interoperability Directive, but they contain negotiated exceptions and special case rules for individual member states which have been negotiated with the expenditure of lots of committee time and Eurostar fares to Lille. Hence the special provision for imperial or metric units (nobody else in Europe or anywhere else ERTMS is used would want this).

ERTMS Level 2 does have benefits in terms of cost and capacity but there is a big problem in terms of changing over to it. Level 3 has greater benefits but doesn't exist yet. This is worth a whole thread on its own which I would be happy to contribute to.

 

[Joseph_Locke]

NR has a programme called "ORBIS" which is intended to rationalise all the different measurements and reference points and will allow everything to be metricated. It will happen one day - just a question of when!

Sadly, ORBIS is doomed to fail in this respect - the earth isn't flat, so any orthogonal survey grid will accumulate errors, quite big ones on long railway routes covering 600 miles or more. An answer is snakegrids, but these cannot be tied together at junctions because the snakegrid principle knowingly distorts the survey data onto a plane (making all surveys "wrong" deliberately).

A practical upshot of all this is that linear metreages don't work properly ...

It has been pointed out to them, who knows if they will listen?

Anyway, no easy solution.

 

[SouthStand]

Why we need mileposts any way?

 

[DerekC]

I know of two or three reasons for needing mileposts (kilometre posts would be just as good):

a) When placing temporary speed boards, defining possession boundaries etc and referring to them in drivers' instructions

b) So that drivers have something to refer to when contacting the signaller when stationary and not at a signal.

c) Drivers also seem to like them as a reference in their mental models of the route

As I recall, the original reason for them was so that passengers could check the fares, which were set by the mile.

 

[QueensCurve]

I'm afraid that's simply not true.

There are hundreds and hundreds of SIs that derive their authority only from the 1972 Act. If you have access to Westlaw etc, just go to the 1972 Act, look at the "SIs made under this Act" function and read the preambles to the SIs. For many of these SIs there is no corresponding domestic legislation.

If you really want to disagree with me, perhaps you could tell me what the non-1972 Act legislative basis was for the Utilities Contracts Regulations 2006?

Section 18 of the 2011 Act does not say anything about how directives are implemented. It is a reminder that directly applicable EU law is applicable only because the 1972 Act (or other Acts) says that it is and thus at least technically parliamentary supremacy is maintained. It was inserted for largely political reasons.

The general principle is that an SI requires an "Enabling Act". One Such is the Health and Safety at Work Etc Act 1974.

Once an enabling act is in place, it makes it possible to put in place SIs to implement regulations under the act. Thus if the EU makes a Directive in relation to Ionising Radiations, an SI can be put in place under HSW74 to make regulations to implement the Directive.

Schedule 2,1(1)(c) of the 1972 Act gives a power "to confer any power to legislate by means of orders, rules, regulations or other subordinate instrument, other than rules of procedure for any court or tribunal":-

http://www.legislation.gov.uk/ukpga/1972/68/schedule/2
--- old post above --- --- new post below ---

*cough*

Manchester Airport Branch - 1993
Windsor Link - 1989
Selby Diversion - 1983

Dare one mention Norton Bridge?

 

[SouthStand]

I know of two or three reasons for needing mileposts (kilometre posts would be just as good):

a) When placing temporary speed boards, defining possession boundaries etc and referring to them in drivers' instructions

b) So that drivers have something to refer to when contacting the signaller when stationary and not at a signal.

c) Drivers also seem to like them as a reference in their mental models of the route

As I recall, the original reason for them was so that passengers could check the fares, which were set by the mile.

Would be a lot cheaper just to use the GPS that most people carry around with them nowadays.

 

[najaB]

Would be a lot cheaper just to use the GPS that most people carry around with them nowadays.

It would be cheaper, but would only be helpful for half of one of DerekC's reasons. It would help with defining the limits of possessions and TSR's but since drivers don't generally watch a GPS while driving it wouldn't help them understand where those boundaries are.

I can't see it being useful for communications with the signaller (they would need to get a map) nor with the driver's mental model.

 

[Ploughman]

The problem is that the existing data sets already use three different measurement systems and are considered to be a mess.

Who is talking about multiplying anything? That wouldn’t help anything as you say. You just draw a line in the sand to fix things from now on:

1. Don’t trust any old records. - Don't trust any records, always check.
2. Any new measurement is done exclusively in metric. - Done already since before 1994
3. Any new track is laid in metric. - Ditto
4. Any upgrade is planned in metric. - Ditto
5. All new signalling systems work in metric. Practically all trains we’re build in metric anyway so all you'd need is to change the speedometer in the cab when they get the new signalling system retrofitted. - All OHLE and Signalling on the ground is planned already in Metric again since pre 1994

Just thought I would a couple of comments on your list.
In 20 years on track renewals all jobs planned and executed in Metric but because old historic records are in Imperial then this must be taken into account so therefore a 10 second calculation as to where the metric measurement of an object is in relation to the records must take place to confirm that the item you have found is that particular record.
No problem whatsoever.
As someone else has pointed out the historic records will always remain due to never being able to fully remeasure all locations on the railway, primarily under track structures which are buried under 60ft / 18.288 metres of ballast.

 

[w1bbl3]

Imperial measurements in the US are really only used by the general public and construction. Manufacturing industry and engineering in US are basically metricated.

In terms of not being able to resurvey fortunately with modern technology if you can see it you can resurvey it with mm level precision, if you can't see it but can get near it may well also be possible to survey with mm level precision. In most cases far more accurately than traditionally produced engineers surveys.

Buried structures being most obvious element that couldn't still be re-surveyed easily but there are admittedly manual processes that can be used to try and remediate existing records in these situations. The records themselves are probably best guess representations anyway of such structures unless as built surveying was carried out.

Asset records that contain errors or omissions are not really records at all and should be corrected to properly reflect the asset. A bad set of records will in many cases be more dangerous that having no records, hence the adage never trust records.

The system of distance and weight measurement be it metric in SI units, British imperial pre 1959, US imperial pre 1959, British imperial post 1962/US standard units (imperial) or hilariously decimalised imperial metric (don't ask), really isn't that important.

However imperial distances and weights as we understand them today are defined relative to metric measurements 1 yard being 0.9144 metres.

Digital record map data should surely be stored with ETRS89 datapoints so that whatever flat earth system is being used locally can be appropriately derived, most decent surveying kit can do this for you. Storing data as linear metric or imperial is a recipe find stuff doesn't fit as world isn't flat.

I think given the above a simpler system than metric is required for the younger generation as if you divide the population of the world 7 billion by the population of the US 320 million it turns out that about 4.6% of the worlds population live in the USA.

The OP's error raises a valid if unintended point about differing measurement systems and errors that can result from interpretation. Remember US and (old) British units are different in a number of ways, if the world's population was 7 English billions then they'd be correct but it is instead only 7 US billions or to express it simply 7,000 million.

 

[alxndr]

Old staff can stick with their primitive measurements, but as people start to learn metric in school, we need to keep up with the times and do the same for the railway.



Older generation workers probably won't be able to get out of the habit of imperial measurements, but young people are being taught metric as the years go on, so it's a step backwards to give them old measurements they weren't taught should they go on the railway as a career path.

I was taught metric mostly at school, never heard of a chain before, and don't have a problem using imperial where necessary at work.

 

[Trog]

The OP's error raises a valid if unintended point about differing measurement systems and errors that can result from interpretation. Remember US and (old) British units are different in a number of ways, if the world's population was 7 English billions then they'd be correct but it is instead only 7 US billions or to express it simply 7,000 million.

No they would still be wrong by several decimal places. :roll:

 

[Murph]

Would be a lot cheaper just to use the GPS that most people carry around with them nowadays.

There's a few technical issues with that. While some consumer GPS units will be capable of good precision and additionally allow the operator to properly examine the satellite status and HDOP, some may not (which means they are not suitable for anything where precision is important). In the railway environment, inside a metal vehicle body, most portable GPS units will have degraded accuracy and frequent loss of fix from periods of running in cuttings and tunnels. If 100-500m accuracy is sufficient, a consumer GPS with less than ideal satellite constellation will normally provide that. Getting the full 1m accuracy from GPS requires a combination of the correct equipment, suitable satellite constellation, and operator skill.

Beyond that, even with a professional differential GPS with near perfect satellite constellation and good HDOP, there's the issue of geodetic systems, with many to choose from and significant differences in the coordinates if there's a mismatch between GPS and cartography. The native coordinate system for GPS is WGS84, which is now commonly the system used for nautical and aeronautic charts (but all professional navigators should know very well to verify the system used by each chart). For terrestrial surveying in the UK (Ordnance Survey, but _NOT_ the Admiralty), however, OSGB36 or ETRS89 are more likely to be used than WGS84, and the same numerical coordinates refer to a different physical location depending on the system used.

Not all GPS receivers are created equal; GPS operation does actually have some details which can easily trip up an unskilled operator.

 

[steverailer]

There's a few technical issues with that. While some consumer GPS units will be capable of good precision and additionally allow the operator to properly examine the satellite status and HDOP, some may not (which means they are not suitable for anything where precision is important). In the railway environment, inside a metal vehicle body, most portable GPS units will have degraded accuracy and frequent loss of fix from periods of running in cuttings and tunnels. If 100-500m accuracy is sufficient, a consumer GPS with less than ideal satellite constellation will normally provide that. Getting the full 1m accuracy from GPS requires a combination of the correct equipment, suitable satellite constellation, and operator skill.

Beyond that, even with a professional differential GPS with near perfect satellite constellation and good HDOP, there's the issue of geodetic systems, with many to choose from and significant differences in the coordinates if there's a mismatch between GPS and cartography. The native coordinate system for GPS is WGS84, which is now commonly the system used for nautical and aeronautic charts (but all professional navigators should know very well to verify the system used by each chart). For terrestrial surveying in the UK (Ordnance Survey, but _NOT_ the Admiralty), however, OSGB36 or ETRS89 are more likely to be used than WGS84, and the same numerical coordinates refer to a different physical location depending on the system used.

Not all GPS receivers are created equal; GPS operation does actually have some details which can easily trip up an unskilled operator.

We found this the other night, using Omnicom track locator app on 3 different smart phones, all stood next to each other on track and there was a 4 chain difference between the highest and lowest. We only use it as a reference to find the correct location which is already marked on the tracks.

 

[cjmillsnun]

So was I !!
The only use is in a pub.

Or when driving.
--- old post above --- --- new post below ---

The imperial measurement system is not broken, and never has been broken. The problems are nothing to do with imperial vs. metric, but either bad records keeping, records not being adequately kept up to date, human error at the point of measurement, or outright incompetence of someone involved. All of those problems exist equally in both measurement systems, and change for the sake of change will do nothing to improve that. Frankly, if the people responsible for the measurement can't cope with reliably converting between imperial and metric, they should not be doing the job.

The people who really need to deal with the measurement, such as the surveyors and engineers, they can cope equally well with both systems, as they have been doing that successfully for a very long time. New entrants into those fields should really not have any difficulty in learning the systems from more experienced people; and if they do have problems learning that then they probably should not be doing a job of that nature.

If the records keeping is bad, then it makes sense to go out and check the records and digitise them accurately. You may say that is a waste of money, but in reality having a mapped digital asset records system will save an absolute shedload of money in the long term. To do this easily requires one measuring standard. It makes sense for that to be metric, as any new railway will be in that standard.

 

[The Planner]

ERTMS Level 2 does have benefits in terms of cost and capacity but there is a big problem in terms of changing over to it. Level 3 has greater benefits but doesn't exist yet. This is worth a whole thread on its own which I would be happy to contribute to.

Indeed it is, am I am one of those that is pretty sceptical as to what level 2 can do in terms of capacity, level 3 I agree with, 2 no...

 

[cjmillsnun]

It's the railway, at least the Network Rail part of it, that has decided to metricate speeds and distances during transition to ETCS and ERTMS. The system and regulations would have permitted imperial measurements to be used, if desired. They must have their reasons.
"If it's not broken don't fix it" - I thought that was the first law of Bangernomics (the second being "if it IS broken don't fix it")

The reason to convert is cost. They can use off the shelf products as used throughout the EU on other ETCS/ERTMS projects and not have to customise everything.

 

[Joseph_Locke]

Dare one mention Norton Bridge?

If one must, and it's actually open ...