Summary of Selected NTSB Accident Reportsonlinepubs.trb.org/Onlinepubs/tcrp/tcrp_rpt_52-k.pdf · Page Ia-16 Table Ia-1: Summary of Selected NTSB Railroad Accident Reports (cont'd)

Page Ia-1

APPENDIX Ia

Summary of Selected NTSB Accident Reports

Page Ia-2

Table Ia-1: Summary of Selected NTSB Railroad Accident Reports

Page Ia-3

Table Ia-1: Summary of Selected NTSB Railroad Accident Reports (cont'd)

Page Ia-4


Page Ia-5


Page Ia-6


Page Ia-7


Page Ia-8


Page Ia-9


Page Ia-10


Page Ia-11


Page Ia-12


Page Ia-13


Page Ia-14


Page Ia-15


Page Ia-16


1Could also be classified as "SEPARATION" failure; note NTSB unable to determine probable cause.

Source: NTSB Railroad Accident Reports.

Page Ib-1

APPENDIX Ib

Complementary Analysis of FRA Data

Page Ib-2

Ib-1: Analysis of Accident Data for Passenger Rail Service

To begin to identify potential incrementalrisks, an analysis of passenger railaccidents was done. Table Ib-1 summarizesthe accident data by railroad. The data aregrouped into the following categories:

! Passenger Railroads;! Amtrak (separated out from other

"Freight Railroads on whichPassenger Trains Operate" becauseof its unique role as the mainprovider of long-distance railpassenger service);

! Freight Rai lroads on whichPassenger Trains Operate; and

! Other (which essentially representsfreight-only railroads).

It is important to note that there is asubstantial amount of variability in theaccident rates (per million train miles andper million passenger miles) for individualrailroads. For example, for passengerrailroads, the range in total accident rate isfrom 0.59 to 7.2, i.e., more than one orderof magnitude. Also, there is not enoughdata in a single year to produce statisticallysignificant estimates. In particular, there isvery little data on collisions and onpassenger railroads (few collisions).However, the following conclusions can betaken from Table Ib-1:

! The accident rate for all railroads is3.85 per million train miles, whichbreaks down by accident type asfollows:collisions: 0.31 per million train miles;derailments: 2.71 per million train miles;other: 0.84 per million train miles.

! Passenger Railroads have the lowestoverall accident rate, i.e., 2.86 permillion train miles, which breaksdown by accident type as follows:collisions: 0.23 per million train miles;derailments: 1.01 per million train miles;other: 1.62 per million train miles.

! The accident rate per million trainmiles for Freight Railroads on whichpassenger trains operate is 3.67,which breaks down by accident typeas follows:collisions: 0.30 per million train miles;derailments: 2.65 per million train miles;other: 0.72 per million train miles.

The high proportion of derailmentsrelative to passenger railroads is tobe expected as longer and heavierfreight trains deteriorate track andinfrastructure at a faster rate thanwould passenger operations. Thissub-category had the most data andtherefore the most stable frequencyestimates.

! Amtrak's accident rate per milliontrain miles was 3.07, which breaksdown by accident type as follows:collisions: 0.12 per million train miles;derailments: 1.68 per million train miles;other: 1.27 per million train miles.

Amtrak's record for 1996 is similar tothat of passenger railroads with themain difference being that Amtrakhad a higher proportion ofd e r a i l m e n t s . T h e i n c r e a s e dderailments is consistent with thehigher number of derailments forFreight Railroads on whichpassenger trains operate.

! The accident rate for the Othercategory is 6.26 per million trainmiles, which breaks down byaccident type as follows:collisions: 0.48 per million train miles;derailments: 4.64 per million train miles;other: 1.15 per million train miles.

Page Ib-3

Table Ib-1: Summary of Accident Data By Railroad, 1996 (Source: FRA, 1997)

Page Ib-4

The high rate of derailments for the"Other" category (i.e., 4.64) relativeto passenger railroads (i.e., 1.01)and Freight Railroads on whichpassenger trains operate (i.e., 2.65)is to be expected, as the "other"category comprises mostly freightoperations which tend to deterioratetrack quality at a faster rate thanwould passenger or a mix ofpassenger and freight operations,assuming an equal volume of trains.Passenger trains using freightrailroad track tend to use the highestquality track on the freight railsystem.

Ib-2: Analysis of Accident Causes by TypeofAccident (All Data, 1996)

To identify the causes of accidents thatwould likely have increased frequencyunder the joint running situation, ananalysis was done of the breakdown ofaccidents in 1996 by cause and type. TableIb-2 shows that "human factors" is theleading cause of "collisions" and "other(excluding highway-rail crossing)"accidents and a major source of all types of

accidents, accounting for 78% of allcollisions, 23% of derailments, and 49% ofother accidents (excluding highway-railcrossing accidents). The main exception isthe case of derailments, for which the leadingcause was "signal and track." More detailedbreakdowns of the causes of collision andderailment accidents are needed to identifypotential target areas for the development ofmitigation measures to reduce incrementalrisk. Figure Ib.1 illustrates the breakdown ofdetailed causes for collisions due to "humanfactors." The 159 collisions attributed to"human factors" are represented in the figure.Figure Ib-1 provides some preliminaryinsight as to where to improve onprocedures, training etc.

Figure Ib-2 shows the breakdown ofdetailed causes for derailments due to"track & signal" for the 1996 data. The 870derailments attributed to "track & signal"are represented in the figure. Figure Ib.4provides some insight as to where toimprove on infrastructure, inspection,maintenance, etc. to mitigate incrementalrisk due to derailments arising from a joint-use situation.

Table Ib-2Accident Type by Cause, 1996

Source: (FRA, 1997).

Page Ib-5

Figure Ib-1:

Collisions Due to Human Factors by Detailed Cause, 1996

Source: FRA, 1997; Table 20

Page Ib-6

Figure Ib-2:

Derailments Due to Track & Signal by Detailed Cause, 1996

Source: FRA, 1997; Table 22

Page Ib-7

Ib-3: Analysis of FRA Accident Causes of Serious Collisionsand Derailments (Passenger Train Data, 1992-1996).

An analysis of FRA data for seriouspassenger train collisions and derailmentswas done by FRA cause categories. TableIb-3 shows the breakdown by cause class(i.e., "Operating Practices", "Mechanical &Electrical Failures", "Track, Roadbed, andStructures", and "Miscellaneous"). Themain cause of collisions is "OperatingPractices", whereas the main cause ofderailments (excluding "Miscellaneous") is"Track, Roadbed, and Structures". Thedetailed causes are shown in Table Ib-4.

Figure Ib-3 shows the breakdown of FRAcauses for the main cause class ofCollisions, i.e., "Operating Practices." As

shown in Figure Ib-4, of the 14 collisionsinvolvements attributed to operatingpractices, 50% are due to "Main TrackAuthority", 36% are due to "Flagging,Fixed, Hand, and Radio Signals", and 14%are due to "Speed".

Figure Ib-5 shows the breakdown of FRAcauses for the main cause class ofDerailments (excluding "Miscellaneous"),i.e., "Track, Roadbed, and Structures." Asshown in Figure IB-6, of the 14 collisioninvolvements attributed to operatingpractices, 56% are due to "Track GeometryDefects," 11% are due to "Rail and JointBar Defects," 11% are due to "RoadbedDefects," 11% are due to "Frogs, Switchesand Track Appliances," and 11% are due to"Other Way and Structures."

Table Ib-3Summary of Serious Passenger Train Collisions and Derailments

(mainline track, 1992-1996).

Page Ib-8

Figure Ib-3Serious Passenger Collisions on Mainline Track by FRA Cause

Category (1992-1996)

Source: Derived by IRR from FRA Data

Page Ib-9

Figure Ib-4Breakdown of Serious Passenger Collisions on Mainline Track Due

to "Operating Practices" Cause (1992-1996)


Page Ib-10

Figure Ib-5Serious Passenger Derailments on Mainline Track by FRA

Cause Category (1992-1996)


Page Ib-11

Figure Ib-6Breakdown of Serious Passenger Derailments on Mainline Track

due to "Operating Practices" by FRA Cause (1992-1996)


Page Ib-12

Table Ib-4Serious Passenger Train Accidents (1992-1996) by FRA Cause

Collisions DerailmentsOPERATING PRACTICES

Flagging Fixed, Hand And Radio SignalsH204 Fixed signal, failure to comply 2 1H216 Interlocking signal, failure to comply 1 1H299 Other signal causes 2 0

Subtotal 5 2General Switching Rules

H399 Other general switching rules 0 1Subtotal 0 1

Main Track AuthorityH401 Failure to stop train in clear 2 0H402 Motor car or on-track equipment rules, failure to

comply1 0

H403 Movement of engine(s) or car(s) withoutauthority (railroad employee)

1 0

H404 Train order, track warrant, track bull., ortimetable auth., failure to comply

0 1

H499 Other main track authority causes 3 1Subtotal 7 2

Train Handling/Train Make-upH501 Improper train make-up at initial terminal 0 1

Subtotal 0 1Speed

H605 Failure to comply with restricted speed 2 0Subtotal 2 0

Switches, Use OfH705 Moveable point track from improperly lined

switches0 1

H799 Use of switches, other 0 1Subtotal 0 2

MECHANICAL AND ELECTRICAL FAILURESBrakes

E04C Other brake components damaged, worn, broken,or disconnected

1 0

Subtotal 1 0Truck Components

E49C Other truck component defects, (CAR) 1 0Subtotal 1 0

WheelsE60L Broken flange, (LOCOMOTIVE) 0 1

Subtotal 0 1General Mechanical and Electrical Failures

E99L Other mechanical and electrical failures,(LOCOMOTIVE)

1 0

Subtotal 1 0

Page Ib-13

Table Ib-4 Serious Passenger Train Accidents (1992-1996) by FRA Cause (Cont'd)

Collisions DerailmentsTRACK, ROADBED, AND STRUCTURES

Roadbed DefectsT002 Washout/rain/slide/flood/snow/ice damage to

track0 1

Subtotal 0 1Track Geometry Defects

T102 Cross level of track irregular (not at joints) 0 1T103 Deviation from uniform top of rail profile 0 1T109 Track alignment irregular (buckled/sunkink) 0 1T110 Wide gage (due to defective or missing crossties) 0 1T199 Other track geometry defects 0 1

Subtotal 0 5Rail and Joint Bar Defects

T299 Other rail and joint bar defects 0 1Subtotal 0 1

Frogs, Switches and Track AppliancesT314 Switch point worn or broken 0 1

Subtotal 0 1Other Way and Structures

T499 Other way and structure defect 0 1Subtotal 0 1

Signal and Communications FailuresS013 Other communication equipment failure 1 0

Subtotal 1 0MISCELLANEOUS

Loading ProceduresM201 Load shifted 1 0

Subtotal 1 0Unusual Operational Situations

M402 Object or equipment on or fouling track (motorvehicle-not at highway-rail Xing)

1 0

M404 Object or equipment on or fouling track, other 2 0Subtotal 3 0

Other MiscellaneousM501 Interference (other than vandalism) with RR

operations by nonrailroad employee0 1

M503 Vandalism of track or track appliances 0 8M505 Cause under investigation 2 3

Subtotal 2 12GRAND TOTAL 24 30

APPENDIX JOverseas Joint Use Exhibits

Railroad/LRT, Overseas/N. American - Clearances and Remedies ContrastedExhibit J1

Category 2 and 3 DMU Railcar ModularityExhibit J2

Sanriku Railway Shared Track PromotionExhibit J3

Karlsruhe Rail Passenger System DiagramExhibit J4

Page J-1

TWO DIFFERENT SOLUTIONS TO SAME PROBLEM IN JOINT USE - FREIGHT RAILROAD CLEARANCE AT LRT STATIONS:Kassel, Germany (left) uses gauntlet track (note offset freight car(EBO) profile) for its low platform/low floor LRVs.Salt Lake City uses bridge plates and state granted variances for high bloc/high floor SD-100 LRVs

Railroad/LR

T, O

verseas/N. A

merican - C

learances and Rem

edies Contrasted

Exhibit J1

Page J-2

Category 2 and 3 DMU Railcar Modularity

Exhibit J2

courtesy ABB Daimler-Benz Transportation (Schweiz)

Page J-3

Sanriku Railway Shared Track Promotion

Exhibit J3

Page J-4

Karlsruhe Rail Passenger System DiagramExhibit J4

Page Ka-1

APPENDIX KaINVENTORY OF JAPANESE RAILWAYS FOR JOINT USE

This appendix contains a matrix of thoseJapanese railways that participate in jointuse of track/reciprocal running or havepotential for such. It is provided as aresource for subsequent researchers. This isthe starting point for a process employedby the research team to reduce thecandidate railways to the six case studiesthat are described in detail. The followingparagraphs explain the columns in thetable:

! First column, "Commutersheds" arecomparable to U.S. Urbanized Areas(UZAs) but are defined by typicalweekday commuter patterns focusedon a core city (or cities). Includedwithout separation are the nation'sthree major conurbations – the KantoRegion (focused upon the core citiesof Tokyo, Yokohama, Kawasaki, andChiba), the Kansai Region(containing the core cities of Osaka,Kyoto, Kobe, and Wakamatsu), andthe Nagoya Region (containing thecore cities of Nagoya, Gifu, andToyohashi). Because of theinterwoven intricacy of rail lines andtheir participation in reciprocalrunning, it is not possible to separatethese regions into smaller units.

! Second column, "Operator"Corporate Names are l istedalphabetically for each urban place.Translations of common suffixesappear at the end of the table as theymight appear in direct Japanesetranslations.

! Third column, "Track Gauge" isshown in millimeters. 1435mm =4'8½"; 1372mm = 4'6"; 1067mm =3'6"; 762mm = 2'6"; and 609 = 2'0".Inasmuch as the conventionalintercity Japan Rail track gauge is

1067mm,1 only railways of thatgauge can interchange freight carswith the national railway network.Nevertheless, several railways ofother gauges – particularly"standard" - 1435 mm – participate inreciprocal running with railways oflike gauge. There are a few instancesof shared trackbed with three-railmulti-gauge track. There are alsorailways of 1372mm (4'6") gaugewhich participate together in jointuse of track. The railways of 762 mmand 609 mm gauge are isolated anddo not participate in joint use.

! Fourth column, "Sector" identifieswhether the operator is nationalrailway (i.e., Japan Rail), private, orThird Sector. Municipal operationsare generally regarded as "private"(i.e., not national) in Japan, but inthis table "municipal" systems areidentified as such. A separateexplanation of "Third Sector"(public/private joint ventures)appears earlier in Chapter 8.

! Fifth column, Rail "Mode": DCR =diesel commuter rail; ECR = electriccommuter rail; HRT = heavy rapidtransit; HSR = high-speed rail; ICR =intercity rail; IND = industrial (butcarries passengers); INT =interurban; LRT = light-rail transit;StR = street railway; and TRy =tourist railway (or vintage trolley).Japan Rail-Hokkaido, Japan Rail-Nishi, and Oigawa Railway alsooperate steam on regular mainlines,and Hakodate City Transport,Hiroshima Electric Railway, andTosa Electric Railway operatevintage trolleys on regular streetrailway trackage.

Page Ka-2

! Sixth column, "Status" has beenobserved in order to convey an imageof the operator. Some of theserailways have been purpose-built inrecent decades for such reasons as toopen new land for residentialdevelopment; their facilities andequipment are generally quitemodern. Some are older railways thathave more recently been rebuilt orupgraded as – for example – theirterritory has changed from rural tosuburban. Others are going-concernrailways that have broughtthemselves up to date by purchase ofnew rolling stock or acquisition ofsecond-hand rolling stock withremaining economic life.2 "Intransition" indicates that they are inthe process of upgrading, or in a fewinstances, diminishing.

! Seventh column, "Propulsion":Electric or diesel operation is noted."Transition" refers to either thetemporary operation of diesel rollingstock until electrification isimplemented, or the process isunderway to discontinueelectrification and use diesel rollingstock. Cable propulsion systems(funiculars) are not in the list, butrack-assistance railways are included.

! Eighth column, "Setting" or TypeTerritory Served: Japanese land-useand development is more akin toU.S. and Canadian practice than toE u r o p e a n , s o t h e s i m p l ecategorization as urban, rural, and

intercity as railway settings caneasily be understood by NorthAmerican readers.

This tabulation represents the first siftingof candidate operators for furtherexamination, the goal having been to selectno more than six for which additional datawould be arrayed. The final sifting wassubjective, inasmuch as the intention wasnot to find six similar candidates, but ratherto find six candidates demonstrating avariety of joint-use circumstances thatmight be applicable in North America.

The following tabulation is also useful asan inventory of all major Japanese railpassenger transportation services.

ENDNOTES - APPENDIX Ka1. The high-speed intercity Shinkansen

network is 1435mm (4' 8 1/2") gauge.

2. The 'cascading' of Japanese railwayrolling stock is a separate andinteresting topic. The purchase of newrolling stock by the larger, moreprosperous railways generally meansthat the older rolling stock will behanded down to lesser railways.'Standardization' is an unplannedresult. But it often involves some ofthe factors that enter into anyintroduction of joint use or reciprocalrunning, such as changing the catenaryvoltage from 600vDC to 750vDC or1,500vDC, adding threshold extendersto fill the gap between second-handrapid transit cars, and station platformsset back to accommodate standardfreight cars (yes, secondhand heavyrapid transit cars are often thoroughlyrefurbished as modern interurban carswith comfortable seating, on-boardfare-collection, etc.).

Page Ka-3

SELECTION MATRIX OF JAPANESE RAILWAYS WITH REGARD TO THEIR POTENTIALPARTICIPATION IN JOINT-USE OF TRACK OR RECIPROCAL RUNNING

Page Ka-4

Page Ka-5

Page Ka-6

Page Ka-7

Page Ka-8

Page Ka-9

Chiho = District; Electric = Electric; Dentetsu = Electric Railway; Kaisha = Company; K.K. =Kabushikigaisha [company]; Kanko = Sightseeing; Kido = Tramway; Kosoku = Rapid; Kotsu= Transport; Kuko = Airport; Kyoku = Department or Bureau; Kyuko = Express; Renraku =Connecting; Passenger = Passenger; -sen = Line; -shi = City; and Tetsudo = Railway

Page Kb-1

APPENDIX Kb

This tabulation provides more detailed information on those systems screened for furtherexamination. The goal was to select no more than six case studies for which additional datawas obtained through a survey. The ones that were carried forward for further considerationare in bold in the two left columns:

FINAL MATRIX OF JAPANESE RAILWAYS WITH REGARD TO THEIR POTENTIALPARTICIPATION IN JOINT-USE OF TRACK OR RECIPROCAL RUNNING

Page Kb-2

* = including municipal

This does not fully represent the final selection. The Abukuma Railway, for example, wasdropped from consideration in order to include the Nagoya Railway's lines at Gifu.

Page Kc-1

APPENDIX KcMATRIX OF PACIFIC RIM CONURBATIONS AND THEIR POTENTIAL

PARTICIPATION IN JOINT USE OF TRACK OR RECIPROCAL RUNNING

The possibilities for Pacific Rim examples of joint-use were reduced from a review of nationsand urban places to the 36 places that have fixed-guideway urban transportation systems andare identified on the following table. The cities selected for further examination arehighlighted.

ECR = electrified commuter rail; DCR = diesel commuter rail; HRT = heavy rapid transit; INT =interurban; LRT = light-rail transit; StR = street railway; FUN = funicular; P-M = people-mover; AC =aerial cableway; FY = ferry; ! = existing; # = under construction or planned; note: all places areserved by motor bus services.

Page Kc-2

1 Hanoi had streetcars until replaced by electric trolley buses in 19872 Harbin had streetcars until 19873 Kunming diesel commuter rail includes two track gauges (meter and standard)4 Manila did have streetcars and electric trolley buses5 Pusan did have streetcars6 Seoul did have streetcars and interurbans7 Wuhan has two separate electric trolley bus networks

Documents

Summary of Selected NTSB Accident Reportsonlinepubs.trb.org/Onlinepubs/tcrp/tcrp_rpt_52-k.pdf · Page Ia-16 Table Ia-1: Summary of Selected NTSB Railroad Accident Reports (cont'd)