Showing posts with label reliability. Show all posts
Showing posts with label reliability. Show all posts

Monday, August 02, 2010

Real-time information: need for, reliability and management

Real-time information has been a growth area for Melbourne’s trains, trams and major bus routes. Although less important than the service basics of coverage, span and frequency, it is still appreciated by passengers.

Available and accurate real-time information can mitigate the effects of service disruptions, particuarly on the rail network. Some passengers may choose to change or defer their trip, reducing loadings on scarce train replacement buses. And where services are delayed, reliable information allows passengers to make better use of their time, especially if the station or stop is near a shopping area. Advising of services just ahead eases crowding by spreading loadings across several trains (especially if just after a cancellation) or improving the efficiency of transfers at major stations. Hence good real-time information potentially offers substantial operational and customer service benefits.

Real-time information can vary from manual staff announcements to automated position reporting systems that transmit their data to information displays on station platforms or mobile phones. As staff rely heavily on automated systems, this item will concentrate on these.


Printed timetables are simple and reliable

Need

The need for real-time information varies across the network. I believe there are three factors (reliability, urgency and trust) that determine whether real-time information is justified or not.

Consider a suburban bus route that caters mainly for local shoppers. Passengers may not need to connect with another service or be particularly time-sensitive. Buses are almost always on time and cancellations are rare. Printed timetables fairly reflect actual times and are trusted by passengers. Low urgency, high reliability and high trust make real-time displays a frill rather than a necessity, and in any case the per passenger cost is high if installed at quiet stops.

Suburban railways are quite different. More passengers are time-sensitive, expecially those who need to change to a bus or have fixed work times. Even a ten minute train delay can increase end-to-end travel time by 30 or 40 minutes if a bus is missed.

Even allowing for measurement changes, rail delays approximately trebled since 2003. Punctuality declined from 96-97% to around 85% network-wide. It is lower again during peak periods and on long lines served by the troubled Siemens trains (around 75%). Customer satisfaction also fell as patronage rose faster than service levels, causing crowding and delays. High urgency, lower reliability and falling trust all make real-time information a high priority on the train system, second only to network strenghtening measures that increase capacity, frequency and reliability.

In between are the major bus and tram routes. Here heavy traffic and long routes can increase variability. The city-bound routes have a large commuter function, while the orbital routes feed passengers to railway stations and major trip generators. Again real-time information is desirable, providing extra assurance, especially on orbital routes where the bus might be starting its run 60 kilometres away.

The installation of real-time information in Melbourne has generally followed the above priorities. About the only exception is the busier bus corridors that offer above-SmartBus span and frequencies but provide only fixed timetables at stops.


Unreliable information can reflect poorly on a network

Reliability The extent to which an information system is worthwhile depends on whether its content is useful, provided when the passenger needs it, and can be relied on.

Required standards for the latter are high; if 99% of trains run (as is frequently attained in Melbourne) and information displays are 99% accurate then there is just as much chance of the information being wrong as the train being cancelled.

To be fit for purpose, measuring and reporting equipment (such as passenger information systems) should be sufficiently accurate to show variations in the item being measured without introducing substantial errors or uncertainties of their own. Hence in a system where 99% of services run, information systems must be operational and accurate (in this case faithfully indicating which services have been cancelled) for 99.9 to 99.99% of the time.

Information system reliability may fall during service disruptions, but should always be at least an order of magnitude better than the reliability of the underlying service. This increases trust that information is accurate (despite problems the service itself is experiencing).

In Melbourne some SmartBus displays at stops, ‘next station’ displays on the new Siemens trains and ‘next train’ displays at stations are particularly at risk of not meeting availability and reliability standards (in contrast the 1980s AVM system used for trams appears more reliable).

An inconsistent quality of information can make a transit service appear less reliable than it actually is (and undeservedly lower its reputation). And customer complaints are often as much about missing or inaccurate information as the service interruptions themselves. To lessen these risks electronic information displays must be much more reliable than the system itself, and if this cannot be attained they are best removed.

Upgraded equipment is being installed across the rail network

Management

The simplest information display systems (eg blackboard and chalk) require no special expertise to maintain. Lapses in availability (eg running out of chalk) or poor readability (due to bad writing) can be diagnosed and resolved at the local station level. The key issue here is likely to be the extent to which customer-facing staff are kept informed of short-notice service changes and disruptions.

Arrangements for managing automated real-time information are more complicated. Day to day ‘babysitting’ may be done at control desks at bus depots, operations centres or major railway stations. Faults that are beyond the control of stations and can only be fixed by centrally-deployed technically-trained personnel.

Responsibility and accountability are divided (a likelihood in any technically specialised area but accentuated through franchising) and recurring problems may be unresolved defects from two or three operators ago.

A difference between the treatment of service delivery and passenger information is that operator contracts (at least for train and tram) impose penalties for non-performance of the former each month. There is also a requirement for public reporting through ‘Track Record’ and these topics are frequently matters for media comment.

In contrast contracts specify no hard performance standards for real-time information system reliability nor the public reporting of same. While the latter is in line with other cities it does mean that (unlike service reliability) discussions of PID reliability are more anecdotal than factual.

The reliabilty of passenger information systems has a lower profile than either of the above, so is unlikely to the the topic of such special action. Nevertheless it sometimes makes the media when tested and found wanting, for instance during last Tuesday’s disruptions when it was found that the SMS alert system lacked the message sending capacity to promptly advise all subscribers.

Even if public standards existed, the extent to which a new incoming operator could or should be held liable for systems inherited from and unfixed by previous government and private operators is debatable. But in the end it boils down to contractual obligations (which the government drafts, signs, enforces and should be accountable for). Additionally the government can negotiate with the operator enhancements beyond what’s in the contract at an additional public cost (examples being the strengthening of the Comeng train air conditioners or late night services).

Conclusion

The fragmentation of responsibility, only limited accountability, the dependence on IT, the interconnection of various systems at multiple sites (of various origin and age) and use of wireless communication all increase the risk of failure.

Real-time passenger information systems are of only limited value unless they are much more reliable than the services they are intended to monitor and indicate.

If such systems are to be a beneficial part of the transit network (and passengers increasinly appear to want and expect it), increased attention will need to be paid to the management, procedures and technology that make the difference between information reliability and unreliability.

Sunday, October 25, 2009

The search for Melbourne's best railway line

Supposing you were moving to Melbourne and planned to commute to the CBD area by train. You had no pre-conceived suburb preferences and based your location choice soley on train service quality. Which train-served suburb or area would you pick?

The following are some key measures of service quality:

* Service frequency, both peak and off-peak. Can you 'turn up and go'?
* Service delivery. How common is it for trains to be cancelled?
* Service reliability. Do trains usually run on time?
* Comfort of travel. Are trains free of overcrowding?

Statistics on all the above are available from the following sources:

Timetables reveal service frequency. Delivery and reliability statistics are reported each month. The twelve month average is most useful here. Train loading figures are collected by the Department of Transport.

A look through these indicate significant variations between lines for frequency, reliability and comfort/crowding. Variations were fewer for service delivery, with almost all lines having around 99 percent of trains running. For this reason only frequency, reliability and comfort are needed for this survey.

An A to E scale is used to allow easier comparison. These are as follows:

Frequency

A = 3 to 10 minutes
B = 11 to 15 minutes
C = 16 to 20 minutes
D = 21 to 30 minutes
E = >31 minutes

Reliability

A = 96 - 100%*
B = 92 - 96 %*
C = 88 - 92%
D = 84 - 88%
E = less than 84%

(*) Exceeds compensation threshold.

Comfort

A = not indicated in DoT load chart
C = feature moderately in DoT load chart
E = feature prominently in DoT load chart

Line by line comparisons are tabulated below. Peak frequency figures are for am services from the terminus unless otherwise indicated. This is based on maximum intervals and service is usually higher at stations identified (+). Ratings are given in order of peak frequency/ offpeak frequency/ punctuality/ comfort.

Werribee (Werribee): C+/B+/C/C
Werribee (Altona): C/C/C/C
Williamstown (Williamstown): C/C/B/B
Sydenham (Watergardens): B+/C/B/E
Craigieburn (Craigieburn): C+/C/C/C
Upfield (Upfield): C/C/B/A

Epping (Epping): B+/C/B/C
Hurstbridge (Hurstbridge): E+/E/C/C
Hurstbridge (Greensborough): B+/C/C/C

Belgrave (Belgrave): C+/D/C/B
Lilydale (Lilydale): C+/D/B/B
Belgrave/Lilydale (Ringwood): A+/B/B/B
Belgrave/Lilydale (Auburn): A+/B/B/B
Alamein (Alamein): D+/B/A/A
Glen Waverley (Glen Waverley): B+/B/B/A

Pakenham (Pakenham): D+/D/E/E
Cranbourne (Cranbourne): C+/D/D/C
Pakenham/Cranbourne (Clayton): A+/B/C/D
Frankston (Frankston): B+/B/E/C
Frankston (Glenhuntly): C+/B/E/C
Sandringham (Sandringham): B+/B/B/A

The winner is:

The Glen Waverley line has good peak and interpeak service frequencies and high reliability without the crowding of some other lines. Parts of other lines score higher on some individual items but Glen Waverley presents the best overall package for the commuter.

Sunday, October 18, 2009

Does system complexity lessen reliability?

It's been a rough week on parts of Melbourne's train network. Power failures, equipment faults, errant prams and unruly passengers all contributed to delays and line suspensions. And during peak periods replacement buses aren't necessarily available either.

This leads one to ask whether there is a relationship between a transit system's complexity and its reliability, as measured by cancellation statistics.

First let's look at the relative complexity of buses, trams and trains.

Buses are the simplest. The driver must be rostered and present, the bus must be available and serviceable and there must be a road with flowing traffic. True, buses do get delayed when roads are blocked due to an accident. But replacement buses can be sent out from the depot (during off-peak periods, subject to driver availability) are can overtake defective buses. Also buses are self-powered, and being trackless, do not rely on points and signals to run.

Trams are a bit more complex. Rather than being self-powered they require power from an overhead. They can also derail and a stuck tram can block the line. However they do not have the sorts of complex centralised signalling systems we have with rail.

Trains are the most complex of the lot. A train is useless by itself. Not only does it need overhead power, but it also needs a track and a signal. The route a train takes is determined by the way the points are set, and this is interlocked with the signals to provide seperation between trains through a safeworking system. Signals can be manual (controlled by the city-based Metrol or suburban signal boxes) or automatic, controlled by the presence of trains, as detected by track circuitry.

Unlike trams, trains can be a mixture of stopping and express, with some lines shared with country freight and passenger trains. Similar to trams, delays can cause knock-on delays to later trains, but with magnified effects.

For a train to proceed so many more things need to be 'just right' than with trams or buses. And because any spare trains/drivers are often not where needed, one train taken out of service can result in two services not running.

What do the reliability statistics from Track Record say?

These indeed confirm that trains are cancelled more often than trams which in turn are cancelled more often than buses.

For the 2008-2009 year, 0.1 percent of buses did not run. The figure for trams is 0.3 percent. 1.3 percent metropolitan trains did not run. Regional trains were cancelled 1.7 percent of the time, though replacement coaches are more commonly run for country services. Note that these are overall system-wide figures; performance varies by time of day and line group.

Longer-term, trams always ran more of their timetabled services than trains, with the difference widening from small to large as train performance declined since its 2000-2003 peak.

Trains however come into their own when it comes to on-time running. Once a train runs you are more likely to arrive on time than if you took a tram. This is because they have their own right of way, unlike trams and buses, which mostly operate in mixed traffic.

The cancellation (but not punctuality) figures do seem to confirm that the more complex systems have signifantly more difficulties than the simpler modes. Metro trams versus buses is a 3:1 difference, while metro trains verus buses is a 13:1 difference.

However the above by itself does not constitute evidence that there is a causal relationship between complexity and (un)reliability.

It is possible to conceive of a situation where a more complex system comprising of three reliable processes produces better outcomes than a simpler one with one less reliable process.

As an example, supposing the simple process required only one thing to work properly, and this happened 90 percent of the time. A more complex system might require three sequential processes to happen. However the probability of each working was 99 percent.

In this case despite its greater complexity the second system was more reliable; its failure rate was near enough to 3% versus 10% for the simpler one-stage system. Its higher reliability appears to be because each process was more reliable and this contributed to overall better system performance. But add a lot of extra steps (each at 99 percent reliability) to the second system and the overall result could deteriorate to be worse than 90 percent. The solution is to either to simplify the process (fewer steps to go wrong) or improve the reliability of each process to say 99.9 percent (ie reducing failure rates to one-tenth the previous level).

Nevertheless research in other fields eg Alexandrov's paper do find such increasing system complexity does risk reliability. While he writes about computer systems, his conclusion could well hold true for transit systems (noting that computer systems are involved here too).

This seems to support the idea that instead of a massive interconnected system, high-reliabity railways should operate as several independent lines or sectors, each with their own tracks and points. Problems are thus isolated rather than risk bringing down the whole network. Crucially, connectedness must still be provided in the eyes of the passenger, and this requires very careful station and interchange design due to the possible increased need to transfer.

Tuesday, July 21, 2009

Legislative Council Select Committee on Train Services

After a period where public submissions were sought, the Legislative Council Select Committee on Train Services held its first hearing today.

Those called before the hearing were from the Department of Transport in the morning and Connex /Mainco/United in the afternoon. They gave short opening addresses followed by answers to questions from committee members. Media and about 20 members of the public witnessed the proceedings.

The Department of Transport emphasised the massive patronage growth on the state's passenger railways. Our 50% increase (metropolitan) and 70% increase (country) in the last 4 years outstripped all other states. On some measures at least Melbourne has overtaken Sydney's CityRail network, with us now having more peak trains in the busiest hour than Sydney.

This growth lengthened station dwell times and resulted in reduced punctuality (dropping from 97% obtained in 2002-3 to over 90% most months/most lines today). One way of looking at is is that trains are 2 or 3 times more likely to be late now than in 2002. However 90% is still regarded as an achievement in the context of unprecedented patronage growth and was described as the 'real story about Melbourne trains'. Platform dwell times (rising from 30 seconds 5 years ago to up to 75 seconds now) were a major contributor to late running.

Main drivers of patronage growth (which ocurred after decades of decline, stagnation or slow growth) included high metropolitan population growth, a strong economy with rising CBD employment and high petrol prices. More people were using the railway by choice, there was no sign of slowing demand and the '20% by 2030' mode share target was likely to be met. Once people changed to rail they remained with rail, which was a vote of confidence in the service.

The system of governance, based on government funding and ownership of infrastructure but private operation through a franchise agreement was defended. It was pointed out that Yarra Trams operated under similar arrangements.

The implication was that the tram system (though significantly delayed by car traffic) has generally operated smoothly without political or media attention (compared to metro trains). Also the train system had high reliability in 2002-3, two or three years after franchising started.

Hence, the department stated, the 2008 summer failures could not be seen as being caused by governance arrangements, contrary to the claims of un-named but well-known 'transport academics'. Nevertheless it was acknowledged that train reliability during this period was below legitimate passenger expectations and that the government was ultimately responsible for the performance of the system.

Operators had put in 'agressive' bids in 1999 (in order to win business). This model, based on diminishing public subsidy and (as it turned out premature) projections of patronage growth, proved unsustainable, even after a cash injection in 2002. When National Express asked for no the government refused and it quit, giving a week's notice. Refranchising in 2004 led to the reunification of both the train and tram networks, the creation of Metlink and a more secure financial footing for the operators. The savings (compared to government operation) proved to be 'illusory' but the Auditor Generals Office reported that the new franchises delivered fair value for money. Services ran through the NX receivership period without disruption to passengers, while driver training resumed (this had stopped and caused a subsequent shortage).

While punctuality had declined from the 97% obtained in 2002/3, it still normally exceeded 90%. This was considered a creditable result given the high patronage growth. Patronage is both an expression of success but also a challenge for service reliability due to the previously mentioned difficulties of crowding and longer dwell times.

What was being done to accommodate growth? Some extra services have been added and maintenance has been increased, but this was 'sweating' the infrastructure (ie being worked harder). More recently governments have announced significant upgrade programs (including South Morang and Sunbury extensions, which also include upgrades elsewhere along those lines) with a lead time applying. Upgrades also aim to simplify the variety of sinalling equipment to improve response time and there is a program to replace timber sleepers with concrete.

In hindsight more could have been done to address overcrowding. However at that time governments had other priorities and it was considered that the case for major rail improvements on the strength of a year's high patronage growth was less strong then than now. The high investment now, while it has some lag time, is regarded as appropriate and one 'can't complain'.

Many questions were asked about Comeng train air conditioning since their failure reduces summer train reliability. The problem here is that these often cut out when temperatures exceed 34.5 degrees C. When specified in the 1970s the debate then was whether to air condition the new trains or not, rather than what type of air conditioning.

Summer 2008/9 was regarded as exceptional, due to the number of days over 40 degrees C, which is more now than when they were specified. This is one reason why the problems this year were more significant than previous summers. It was acknowledged that beyond basic maintenance, the operator had no contractural obligation to make them more suitable for our temperatures. Trials were being run (involving at least 2 trains) and the result of these will influence whether any decision is made to retrofit. With regard to this the Department specified a number of reasons why this might not be a sensible use of funds, including (i) hot days are rare, (ii) the trains are old with limited future life, (iii) there are service consequences of trains being out, and (iv) it's not cheap.

Air conditioning wasn't the only cause of service failures. Widespread carelessness around our 180 suburban level crossings cause incidents that can disrupt trains for an entire peak period (or longer). 10-15% of delays are due to infrastructure, and of this 70 to 90% is due to signalling faults. Point failures can also interrupt services.

Another persistent line of questioning was the extent to which the Department takes an interest in all incidents that cause service failure. Parliamentarians wanted the Department to receive regular detailed reports, and possibly prepare prevention plans to prevent reoccurrence.

In reply the department said it received briefings and system performance information in agregate form. However the operator and department do not go through the reasons for all 14000 delays each month. This appeared not to satisfy some parliamentary members given the follow-up questions about specific incidents.

I suspect that this because of aspects of the franchise agreements and the operator/department relationship did not get fully fleshed out. In the absence of direct questions, on this, I suspect that the department holds the operator to account on broad figures derived from monitoring of operational performance (ie cancellations/late running). This includes penalties for cancellations or lateness.

The outsourcing aspect of franchising means that how good performance is achieved is up to the operator. If they can find a cost-effective way to reduce cancellations (for example) then the statistics the department receives improves and the penalties are less. The operator is rewarded for their innovation. Besides it's presumed that the operator knows more about the system than the department, so are in a better position to specify which improvements deliver most bang for the buck (eg the revised fleet maintenance plan later cited by the operator).

On the department's part, the mechanics of how the operator does it seems to be less important to the department than whether a specified performance is achieved or not. This appears to be a design feature of franchising, with the potential advantages described above (provided incentives/penalties are set to make improved performance worthwhile). It also prevents the department trying to 'micromanage' train operations, and proposing fixes that might be less effective than an alternative proposed by the operator (who has a financial incentive to achieve the results).

This could make some non-government politicians impatient with the Department, who might not be able to instantly answer questions about specific incidents (eg how can delays at 5am be due to 'passenger loading').

Some politicians seemed to want the department to 'supervise more' and 'do more' at the operational level (eg upgrading Comeng air conditioners and platform staff were suggested) rather than leave to to the franchisee (who is taking other steps to improve performance, eg its revised Fleet Maintenance Plan and employment agreements).

Direct Department intervention in operational areas (while it has occurred in some cases eg additional late night services being funded or driver training) would go against franchising's philosophy at its purest, where the franchisee controls resource allocation, subject only to meeting performance and safety standards. The current 'franchising model' may have a lesser role for the department compared to alternatives such as a 'fee for service' contract, or direct government running.

It seems to me that franchising for the Department could be described as it being a contract manager who sets specifications, but leaves the solving of day to day problems with the operator. However time was too short to be articulated in any answers to the questions. Nevertheless such an explanation could have benefited some parliamentarians, who, I suspect may be unfamiliar with our operational arrangements and the extent to which this would shape the Department's responses on questions such as detailed performance information, Comeng air conditioners and the desirability of platform staff to reduce dwell time (currently being trialed at Parliament Station).

A question was asked about travel origins and destinations. The Department was doing work on this through the VISTA study. Many trips were local or to the next municipality and for these buses and trams met these needs. As to the effect of bus improvements on rail patronage, it was stated that the two may well be related. However modal interchange was not currently high and most trips would be seperate bus or train, rather than bus+train. However the current bus reviews were cited as a means of improving service co-ordination and ensuring buses met trains at stations. As to the future, the Victorian Transport Plan was prepared in parallel with 'Melbourne @ 5 million', with the emphasis on early provision of transport services.

A number of rail infrastructure questions were asked. These included the $650m costings for South Morang (amount is a 'best professional estimate' and includes a package of 3 grade seperations, stabling at Eltham and other improvement for the Clifton Hill group), the desirability of standard gauge (not a high priority for passenger, and seperating freight and passenger traffic was more important), and expansion gaps in rail (continuous welded was world-best), Regional Fast Rail costs (higher costs went to renew infrastructure found to be in poor condition), most stressed section (Sunshine - Footscray, with Regional Rail Link and Sunbury electrification assisting).

The Department's submission

Connex appreciated the opportunity to appear at the committee so that it could hear the views of those professionally concerned with running the system. It also sought to articulate the view of rail workers and dispel myths from 'transport academics'.

Our system was described as unlike a dedicated metro. This is due to it sharing track with freight and country trains and the large number of level crossings. There has also been less recent investment to the network. It is however large by world standards, with about double the track km of the Paris Metro.

Financial performance and efficiency was described as good - with sources for this being the Auditor General's report on train franchising and comparisons with the Sydney and Brisbane networks. Our costs per passenger were less than both of these while patronage growth was higher, reaching 214 million trips pa and up by 80% since 1999, on a train fleet only 9% bigger, increasing the number of load breaches. An extra 1500 services per week have been added in this time.

Other major achievements Connex cited include reunifying the network after the exit of National Express and the growth of two incompatible networks. Train services were maintained during the refranchising, but a driver shortage due to the previous operator's lack of recruitment prevented service increases during this time.

Customer service initiatives included 'Meet the Managers' and 'We Hear You', but this was not a major part of the presentation.

Generally Connex enjoyed industrial peace (unlike the Met in the 1980s/early 1990s). A Memorandum of Understanding was signed in 2008. However disagreements over this led to industrial problems which exacerbated heat-related cancellations during the summer of 2008/9 as many were considered preventable. The matter went before the Industrial Relations Commission and both sides have accepted its finding.

Much discussion time was spent on the improvements that will ensue due to the acceptance of the revised Fleet Management Plan in conjunction with more maintenance staff. The former would allow trains with certain minor faults to continue until the end of the day's peak period (rather than be taken out of service immediately). The falling rate of cancellations during 2009 has been attributed to the application of this plan. As an example, just 0.4% of trains were cancelled in June 2009, which is comparable to the good results obtained in the 2000-2003 period. The FMP is included in the industrial agreement, so will continue when MTM takes over.

Also agreed was the ability to run trains where one air conditioning unit had failed. Such trains are often still usable with no safety implicatons. These could be run for the remainder of that peak period instead of being cancelled.

There was agreement with the Department that replacement was probably not cost-effective, but the operator was confident that the above measures would result in a more reliable service this summer than last.

Improved maintenance practices have increased the availabilty of more trains, with 94% (or 149) being scheduled to run in the am peak in the current timetable. This represents high utilisation and pressure on the fleet. However relief will come when the new trains start arriving next year.

Apart from air conditioner faults during the heat, other sources of failures include tracks buckling, signal failures (which due to electrical equipment double in hot weather), infrastructure faults, vandalism and loss of power. Like everyone else, trains are susceptible to 'load shedding', ie power cuts where consumption exceeds generating capacity. Connex is negotiating to get priority treatment with the power authorities so that trains can retain power. These discussions are continuing.

A question was asked about platform staff, particuarly to assist wheelchair passengers and otherwise speed boarding. No commitments were made regarding this. However the member was referred to the DDA Action plan and said that the operator was on track to achieve this. In addition a modified 'demonstration train' provided more space for wheelchairs with the decision to modify more trains resting with the Department.

Related to previous comments about the franchise and what it can drive operators to do, was a question about whether it was better to cop a fine or pay for a fix. While paying fines for poor performance was a problem, this was a part of the contract and was accepted as such. In addition because the operators receive a portion of the farebox revenue the growing patronage increased this part of their revenue, hence offsetting the fine (it was not mentioned whether this was a partial or full offset). In addition the operator receives income from joint ventures with United in infrastructure projects such as North Melbourne Station, Cranbourne stabling and Laverton upgrade.

The sleeper replacement program involves replacing 1 in 4 sleepers. In practice this involves 1 in 1 around curves and 1 in 5 on the straights. Ideally all would be replaced but there are constraints on funding and on the capacity of the factory to produce (100 000 pa available to Connex).

Connex was also asked whether the patronage growth was projected and if preparation was adequate. In hindsight the answer was no. At the time everyone concerned was involved in picking up the pieces from the NX walkout, refranchising, the driver shortage and reunifying the network. The driver shortage remained throughout 2005, while the first double digit patronage rise was in 2006.

Progress however has acelerated in the last few years, bearing in mind that some improvements (eg timetables) might have a 2 year lead time and major infrastructure longer. Connex drew up a plan, with its first major example being the November 2008 timetable. This started to 'untangle the loop' by taking peak period Werribee trains out of the loop and modified the Clifton Hill group pattern. Later improvements involve a more metro-style operation, seperating lines and reducing the crossing of paths. After this the extra trains should allow more services to be run.

Other questions were asked on such topics such as drivers calling in sick (2 hour notice requested), driver numbers (now very close to establishment levels) and who ensures train is operable before start (driver during preparation).

Connex submission

The questions members asked pretty much stuck rigidly to topic. Related matters such as the handling of disruptions and customer service (eg adequacy of passenger information/advice of disruptions/website crashes, possible extra staff training to better handle disruptions and better 'alternative transport' information at stations) were not raised in either the introductory addresses or subsequent questions from committee members. Disruptions can't be prevented, but communication may have helped mitigated their effects.

Saturday, July 11, 2009

Wednesday, May 14, 2008

Crowds, delays and late-running - Part 2

It's only slightly harder to find a parking spot in a car park that's 91% full than one that's 90% full. It might take 33 seconds instead of 30 seconds, for example. However it's twice as hard to find a spot where it's 99% full instead of 98% full. Finding one would take twice as long on average. And when it gets really full, from 99 to 99.9%, you'd expect parking would take 10 times longer, since there's only 1/10 the number of free spaces.

This is the rule that even a tiny increase in usage of a system nearing capacity causes service faults (or more accurately, their impact) to rise exponentially.

In Melbourne we are currently experiencing the same on our peak-hour trains. Operators struggle to meet performance targets and rising patronage is cited as the key reason why. Today's stats compare to five years ago, where the likelihood of your train being late or cancelled was 50 to 70% less, but patronage was also lower, and if your train was cancelled you were more likely to be able to board the next one.

Studying lateness and cancellation figures is useful but don't always tell the full story, especially as seen by the passenger. In some cases a cancelled train may be effectively replaced by the previous running late, and the overall impact might be negligible. But in others, even a 2 minute delay could cause buses or other trains to be missed, resulting in 15 - 60 minute delays. Similarly the cancellation of a busy service might lengthen waits as passengers are unable to board the next three or four crush-loaded trains.

For simplicity, let's assume that trains can carry 1000 people crush-loaded. An on-time morning peak train approaching Station A already has 800 on board. As 80 people board at this station, all can board.

The same is also true for the next station (Station B). By the one after that the train will be full, but assuming some alight (eg a nearby school is a popular destination) then Station C's 80 people should also just be able to fit on.

The situation changes if the train departs late.

For a start there will be more on the platform at previous stations, as some intending to catch the following train will have started walking on. The total number of people walking through the validators at all stations before Station A could be 50 persons per minute (but may be more).

If all these people board, and the train departs the terminus 3 minutes late, then that's at least 150 extra people on board by Station A. With 950 people already, and at least 80 wishing to board, then at least 30 will be left behind. With the crush load train presenting at Stations B and C, most of their passengers will be unable to board.

Why did I say 'at least'? The graphs in Part 1 show that as a train approaches capacity, boarding time rises exponentially. So the train will likely be 5 or more minutes late at Station A, and even later down the line. And in that time more people (say 5 or 10 more per minute on average) will have passed through the validators at Stations A, B and C, all expecting to catch the train. So the train could well be full even before it reaches Station A.

Hence we have three station platforms of passengers, increasing every minute, unable to board the train. The next train, assuming it left on time might be emptier than usual on leaving the terminus, as some of its passengers will have boarded the previous late-runner. And for a while it might even be only a couple of minutes behind that train, its progress retarded by the slower-loading train in front. However as it travels more of its passengers will be comprised of those left behind than the increasingly delayed earlier train, and early arrivals for the train after that. So the second train that started on time could end up also being late and crowded.

On the busier lines in Melbourne, this and the following several trains will also be leaving passengers behind. And at the more inner stations (especially those served by one line only and bypassed by expresses) passengers may need to wait for several trains to pass before being able to board. This causes the actual delay experienced to be higher than what might be apparent from train running data alone (as some trains were unboardable).

To summarise, a line with frequent trains that are not crush-loaded (eg Belgrave/Lilydale) will outperform ones that are (Pakenham/Cranbourne and Sydenham lines) in reliability. Belgrave/Lilydale are nearer the 90% carpark full level, while the others are like 99%. And it's much easier to find 100 passener places in a hurry to clear a platform on the former than the latter.

Tuesday, May 13, 2008

Crowding, delays and late-running - Part 1

A points failure, lightning strike or level crossing accident delays thousands of passengers on a busy line. Some get frustrated and blame the train operator. Others take refuge in a good book. But for me such incidents are beneficial in that they force me to think about them and their management. I confess that blog postings here would be fewer if I lived on a shorter, quieter and less delayed line!

The following graphs illustrate some links between passenger loading, delays and their propagation (click for a larger image).

Friday, December 15, 2006

How reliable was the Met Part II

Back in August I wrote about reliability of the pre-privatisation Met train services.

The main pre-1999 trends were vastly reduced cancellations from the early 1990s. Reliability also improved and has been sustained at a higher level than during the strike-torn 1980s.

What has happened since?

To find out, I compiled a table, drawing on PTC annual reports for the pre-1999 figures. Afterwards I used statistics from Track Record 1, 5, 9, 13, 17, 21, 25, 28. Because results were listed by operator, these were averaged for the pre-reunification years (1999 - 2003).

Metropolitan Trains: reliability and cancellations 1983 - 2006

Period | % OT(i) | % Canc(ii)
1983 ------ 93 ----- ?
1984 n/a
1985-6 ----- 87.8 ----- 3.7
1987 -------- 92 ----- 2.8
1988 -------- 92.1 ----- 2.5
1989 -------- 92.1 ----- 2.4
1990 -------- 88.3 ----- 6.3
1991 n/a
1992 n/a
1993 n/a
1994 -------- 92.3 ----- 0.3
1995 -------- 92.3 ----- 0.5
1996 -------- 93.3 ----- 0.5
1997 -------- 94.8 ----- 0.9
1998 -------- 93.7 ----- 1.4
1999 Q1 ---- 93.5 ----- 1.1
1999 Q2 ---- 93.6 ----- 0.7
1999 Q3 ---- 96.0 ----- 0.4
1999 Q4 ---- 97.2 ----- 0.5
2000 Q1 ---- 95.2 ----- 0.5
2000 Q2 ---- 94.8 ----- 1.3
2000 Q3 ---- 96.5 ----- 0.6
2000 Q4 ---- 96.6 ----- 0.6
2001 Q1 ---- 96.5 ----- 0.6
2001 Q2 ---- 96.5 ----- 0.6
2001 Q3 ---- 97.0 ----- 0.3
2001 Q4 ---- 96.9 ----- 0.5
2002 Q1 ---- 96.6 ----- 0.5
2002 Q2 ---- 96.1 ----- 0.6
2002 Q3 ---- 96.5 ----- 0.4
2002 Q4 ---- 96.9 ----- 0.4
2003 Q1 ---- 96.9 ----- 0.5
2003 Q2 ---- 96.9 ----- 0.4
2003 Q3 ---- 96.2 ----- 0.7
2003 Q4 ---- 96.2 ----- 1.2
2004 Q1 ---- 95.8 ----- 1.0
2004 Q2 ---- 95.0 ----- 1.9
2004 Q3 ---- 94.2 ----- 1.4
2004 Q4 ---- 94.0 ----- 1.3
2005 Q1 ---- 93.9 ----- 1.4
2005 Q2 ---- 93.1 ----- 1.2
2005 Q3 ---- 93.1 ----- 0.8
2005 Q4 ---- 93.2 ----- 1.0
2006 Q1 ---- 92.6 ----- 1.3
2006 Q2 ---- 93.2 ----- 0.6
2006 Q3 ---- 93.9 ----- 0.5

(i) Percent of trains on time to 5:59 min (all day)
(ii) Percent of trains cancelled (all day)

Analysis

The post-1998 figures can be divided into four stages. The first occurred during 1998 and 2000 when the network was split (Bayside and Hillside) and then franchised. Apart from a poor Q2 of 2000, the rise in service cancellations experienced in 1997 and 1998 was arrested and then reversed.

The third quarter of 2000 ushered in a 'golden age' for trains in Melbourne, and not just for those who were taking advantage of the new increased Sunday services either. Reliability was consistently high (96-97%) and cancellations were few (0.5% approx). Good performance continued for nearly three years.

The (time?)tables turned sometime in 2003. The organisation of the system was falling apart with National Express pulling out and half the network reverting to government operation until new contracts could be negotiated with the remaining operator. Chronic driver shortages (due to a previous lack of recruitment and training) increased cancellations on the 'Bayside' part of the network. Major rail projects reduced operational flexibility while the consequences of delays increased due to patronage pressures (thought to be due to rising fuel prices and urban growth). These factors (plus the odd Metrol breakdown and storm) combined to double lateness and cancellations between 2002 and 2005. This put on-time running back to 1995 levels and wiped out the improvements of the early 2000s. However, cancellation performance, though poorer than in the recent past, was still at approximately half the level experienced during the Met days.

It's too early to say much about the last year. However there are some encouraging signs of a turn-around, notably sharply fewer cancellations (similar to those experienced during the 'golden age'). On-time running has also improved, with comparable figures to late 2004/early 2005 and the last few (best) years of the Met.

These figures are of value in several ways. Firstly they help to dispel myths one hears said about the network. The first is that 'things were better during the Met years'. Such an assertion is not borne out in the figures, even during less reliable years such as 2004 and 2005. Secondly they demonstrate what our network can achieve, ie a reliability exceeding 96% and cancellations around 0.5%. Thirdly, under a short-term management style, though current performance may be good, the seeds of later problems may be sown (eg not training drivers during the 2000-2003 period).

Friday, August 11, 2006

How reliable was the Met?

Train Reliability Statistics 1983 - 1998

Source: MTA/PTC Annual Reports

Rpt Yr | % OT(i) | % OT(ii) | % Canc(iii) | % Canc (iv)
1983 ------ 93 ---------- ? ----------- ? ------------ ?
1984
1985-6 --- 87.8 ------ 84.1 -------- 3.7 -------- 6.0
1987 ------ 92 -------- 86 ---------- 2.8 -------- 3.8
1988 ----- 92.1 ------ 90.2 --------- 2.5 -------- 4.1
1989 ----- 92.1 ------ 86.4 --------- 2.4 -------- 5.2
1990 ----- 88.3 ------ 85.2 --------- 6.3 -------- 7.6
1991
1992
1993
1994 ----- 92.3 -------- ? ---------- 0.3 --------- ?
1995 ----- 92.3 ------ 86.7 --------- 0.5 --------- ?
1996 ----- 93.3 ------ 89.5 --------- 0.5 --------- ?
1997 ----- 94.8 ------ 92.4 --------- 0.9 --------- ?
1998 ----- 93.7 ------ 90.8 --------- 1.4 --------- ?

(i) Percent of trains on time to five minutes (all day)
(ii)Percent of trains on time to five minutes (peak periods)
(iii) Percent of trains cancelled (all day)
(iv) Percent of trains cancelled (peak periods)

General trends

- The long-term average for on-time running is around 93% with few variations.
- Cancellations and service disruptions declined during the late 1980s but increased greatly around 1990.
- There were unusually low cancellation rates in the 1994 - 1996 period. Subsequent years showed increases, but they remained below their late 1980s levels.
- Current stats are here. These indicate on-time is running at the long-term average and retention of the early 1990s gains in service reliability.