Route Selection and Design of a Coal Shuttle Railroad: Experiences of a Recent Railroad Engineering Graduate

C. Tyler Dick, P.E.HDR Engineering Inc.

University of Illinois W.W. Hay Railroad Engineering SeminarFebruary 1, 2008

2

Outline

Project Background and OverviewRoute and Alignment Selection

Comparison techniques

Detailed Alignment DesignSnags and solutions

ConstructionExpecting the unexpected

3

Coal and Powerin Texas

Texas Lignite mined locally for electricitalgenerationSeveral plants use short-haul railroads to transport ligniteOak Grove project currently under construction will use an 11-mile railroad

Dallas-Ft.Worth

Oak GroveProject

4

Oak Grove Project

Twin Oak ReservoirOak Grove Steam Electric Station (SES)

Two lignite-fired generation units, 1600 MWUnit 1 online late 2009, Unit 2 in 2010

Kosse Lignite MineOak Grove RailroadOwner: Luminant Energy (formerly TXU)

5

Oak Grove Railroad

Deliver lignite from Kosse Mine to Oak Grove SES on dedicated rail lineApproximately 9 million tons per yearTwo 40-car trains, push-pull operationLimited lignite storage at plant

needs to be highly reliable system

Self contained with maintenance facility

6

Project Area Kosse Mine Loader

Oak Grove SES

Kosse

Existing spur to SES

UPRR EnnisSubdivision

SH7

1 MILE

7

The Challenge – HDR Tasks

Oak Grove 11-Mile RailroadRailroad route/alignment studyEnvironmental permittingDetailed design of preferred alignment and mine site yard facilitiesBidding and general construction support

Oak Grove SES Plant SiteDesign reconfigured plant site trackage

lignite unloadingFuture bottom ash and flyash opertions

8

Project Timeline

Route/alignment study: June ‘05 - Jan ‘06

Environmental permitting: Oct ‘05 – May ‘06

Detailed design: May ‘06 – May ‘07

Earthwork construction: Jan ‘07 – Jan ‘08

Bridge construction: Jul ‘07 – Dec ’07

Trackwork construction: Jan ’08 – Fall ’08

In-Service for “first-fire”: Early 2009

9

Route/Alignment StudyDevelop three possible routes meeting design criteria

40mph operation1 percent grades desirable12 MGT annual traffic

Evaluate routes and select preferred alternativeDevelop feasible alignment along preferred route alternative

10

Possible Routes Kosse Mine Loader

Oak Grove SESExisting spur to SES

SH7

1 MILE

A

B

C

11

Which is preferred?“Minimize length, curves and grades” –Wellington & AREMAHow to compare “curviness”?

Tightest degree curveTotal degrees central angle = S delta“Curve Index” = S delta x degree

How to compare “hilliness”?Ruling gradeRise and Fall = S length x |grade|“Grade Index” = S length x grade x grade

12

Need for “Curve Index”

D = 3D = 90 D = 3

D = 90

D = 3D = 90

D = 3D = 90

D = 3D = 90 D = 1

D = 90

D = 1D = 90

D = 1D = 90

Tightest Curve = 3 degrees Total Central Angle = 360 degrees “Curve Index” = 1080

Tightest Curve = 3 degrees Total Central Angle = 360 degrees “Curve Index” = 540

13

Need for “Grade Index”G

= 1%L =

1000

’

G = 0%L = 2000’

Ruling Grade = 1.0% Rise and Fall = 30 feet “Grade Index” = 30G

= 1%L =

1000

’

G = -1%

L = 1000’

G = 1%

L = 10

00’

Ruling Grade = 1.0% Rise and Fall = 30 feet “Grade Index” = 20

G = 0.5%

L = 2000’

G = -0.5%L = 2000’

14

Route Characteristics

265733183188Culverts (lf)

10911893Grade Index

0 / 07 / 1169 / 148Parcels Crossed / ROW (ac)Property Owners

888Grade Crossings

111Grade SeparationsRoad Crossings

999Stream/Creek Crossings

000River Crossings

Bridges and Drainage

148146127Rise and Fall

1.001.001.00Ruling Grade

Grade

598357285Curve Index

258158144Total DegreesCurvature

13.0812.4812.55Total Route Length (mi)

9.588.989.05New Construction (mi)Length

CBAParameter

15

Matrix Analysis ApproachAllows for comparison of alignments relative to multiple decision criteriaAlignments judged in five evaluation categoriesCategories subdivided into several criterionDouble weight scoring system

Criterion scores weighted to determine category scoreCategory scores weighted to determine overall score for the alignment

16

Evaluation Categories

Operational Efficiency,

Mobility and Safety Effects

25%

Cost Effectiveness

30%

Public and Agency Support

10%

Environmental Effects15%

Social and Economic Effects

20%

17

Sample Criterion

15%

10%10%

15%

50%

15%

10%

25%

50%

Operational Efficiency, Mobility and Safety Effects (25%)

OperationalEfficiency

Public Safety

Railroad FreightSafety

Grade CrossingsAdded

Grade CrossingLevel of Service

Cost Effectiveness (30%)

TotalCost

Railroad Ops & Maintenance Cost

RoadwayMaint.

Ease ofImplementation

18

Scoring Results

758888Efficiency, Mobility and Safety

362325316Total

372727Public and Agency Support

515749Environmental Effects

905656Social and Economic Effects

1109898Cost-Effectiveness

CBACategory

19

Preferred Alignment

Alternative C selected as preferred alignment for detailed designDeciding factors:

Minimizes right-of-way impactsEase of implementationPublic supportCost of constructionShorter alignments had environmental and/or access and constructability issues

20

Route Study Learning Points

Beyond Wellington: need to consider range of environmental and social factors, not just engineering geometry and costCurve Index and Grade Index are tools to quickly compare efficiency of alternativesMatrix analysis approach can weight various criteria relative to their importance and allow for easy comparisons between alternatives

21

Detailed Design

Develop detailed design package for preferred alignment

Planset covering all construction tasksStandard detailsSpecifications for earthwork, track and bridge

Designed with MicroStation and Bentley InRail SelectCAD

Yard Tracks (2.3mi)

SH7 Grade Separation

Two Creek Bridges

8 Grade CrossingsLoader

Main Track (8.9mi)

25

Detailed Design WorkflowAerial survey – 1000’ strip – 1’ contoursGeotechnical field investigationTrack design

Horizontal and vertical geometrySubgrade, earthwork and basic drainageCulverts and special ditchesRoadway crossings and signage

Bridge design (by HDR bridge group)

26

Snag:Aerial Survey and Vegetation

Aerial survey conducted in January 2006 to generate design topoVegetation obscured many stream channelsCulverts designed to topodon’t match GPS stream traces used for Corps permit!!!

27

Solution:Aerial Survey and Vegetation

Tried to incorporate GPS traces into aerial topo date but some differences too greatConduct ground survey at key jurisdictional water crossingsReplaced aerial topo with ground survey data then redesigned earthwork and culverts

28

Snag:Horizontal Obstructions

Unmarked pioneer graves and abandoned well near working alignmentFortunately outside limits of cut/fill constructionAlternative solutions if needed:

Steepen side slopesRetaining wallsrealignment

29

Crossing Streams & Floodplains

Bridges at Duck Creek and Pool Branch, both in FEMA designated floodplainsRailroad must create “no rise” in floodplain – HEC-RAS analysisHow to set profile across bridge?

Too high = costly fill and tall bridge bentsToo low = interferes with flood flows and unnecessary grades

30

Top of Rail Elevation over Bridges

Baseline: 100-year water surface elevation from FEMA flood mapsAdd 2’ of freeboard to structure low chord to clear debris during floodAdd structure depth approx = span/12Add allowance for deck, ballast, track

Total is 7 feet above flood elevationUse to set low point of sag curve over Duck Creek

31

Duck Creek Bridge – 170’

32

Duck Creek 5-10’x10’ Relief Culvert

33

Snag:Mine Loader Elevation

Confused top of subgrade with top of rail elevation at mine loaderLoader designed and parts fabricatedNeed to lower railroad 3 feet at loaderGrades allowed redesign for clearanceAdded 3 feet of climb for loaded train starting from stop up the ruling grade!

34

Solution:Train Performance Calculation

Change in elevation at loader negated previous operations analysis by CanacVolunteered to run analysisWent back to U of I grad school spreadsheet for train acceleration

Speed TE R F’a F’a bar L step T step L tot (ft) T tot (s)

1 100880 62581 7.442 100880 62818 7.39 7.42 28 13 28 133 100880 63059 7.35 7.37 47 13 76 26

etc…

35

Train Performance Calculation:SD50 and Loaded 40-car Train

05

1015202530

0 2500 5000 7500 10000 12500Distance from Loading Station (ft)

Spee

d (m

ph)

Accelerating from rest at loading station

0.0%+0.5%

+0.9%-0.4%

3 min2 min

1 min

4 min 5 min6 min

7 min 8 min12mph

60’

36

Used Locomotive Shopping

“We’re up here at NRE in Mt. Vernon looking at used SD50s and wanted your opinion…”Advised client on features to look for such as traction control, de-rating etc.Use ALL of your railroad knowledge to assist client: value-added service!

37

Detailed Design Learning Points

Ground survey is still importantLeave room for expansion in yards and industrial track designsYou can’t miss everything, but miss the expensive things Bridge basics help set efficient profilesUse full rail knowledge of operations to optimize design and aid client

38

Bidding and Construction

Detailed design yielded 230-sheet plansetBidding also required:

Specifications (track, earthwork, bridge)Material breakdowns and bid tabulations

Construction phase also entailed advising owner on suitability of materials procured by the contractor

39

Material Lead Times

Time to procure materials can hamper fast-track projects

Timber ties – 9 to 12 monthsRail – 6 monthsSpecial trackwork – 6 to 8 monthsBallast – wait for break in Class 1 poductionschedule

Price volatility and fuel charges hamper getting a fixed price from suppliers

40

Snag: UPRR Connection and CWR Train

Facing point turnout at UPRR removed

Contractors to use 1600’ CWR stringsCWR train: ramp & threader cars, two locosCan’t bring 1850’ train into plant!

To plant

1750’ clearUPRR to DFW

41

Solution:On-Site Rail Welding Plant

42

Construction Progress:Earthwork and Subgrade

43

Construction Progress:Duck Creek Bridge

44

Construction Progress:Subballast

45

Trackwork Progress:Tie Layout

46

Trackwork Progress:Plating

47

Trackwork Progress:Threading Rail and Spiking

48

Trackwork Progress:Skeleton Track

49

Summary

Can find oneself designing a “whole new railroad” a few years out of schoolModern alignment study considers many more factors than just rail geometry and costRail design is more than just trackExpect the unexpected and be creative!