1. Wayside Detection The Detroit Edison Experience
Dennis J. Buda
Detroit Edison
Railroad Engineering Program
University of Illinois
Urbana-Champaign, ILL
December 2006

2. Discussion Topics The Detroit Edison Strategy WILD Issues
Financial Issues
Conclusions and Recommendations

3. Introduction 36” Wheels Heavy Haul High Mileage Service
Comparison of DTE and AAR CRB Data
2002 and 2003 Data - CRB Limitations
DTE - Michigan Publicly Traded Utility
2.1 Million Customers - $17.0 Billion Assets
7,000 Aluminum Rotary Railcars
Average Car Mileage 100,000/Year
Fuel Source from PRB and Appalachian Mines

4. Wheel Failures and WM Codes
WM 66, Flange Cracked or Broken
WM 68, Cracked Rim
WM 71, Shattered Rim
WM 72, Spread Rim
WM 83, Cracked Plate
WM 75, Shelling, Spalling, Thermal-Mechanical
WM 65 and 67, Out of Round (WILD or Gage)
WM 76, Built-Up Tread
WM 78, Slid Flat

5. Recent AAR Rule Changes
WM 65 Wheel Impact Load Detector Removal at Impact Greater Than 90 Kips
97% of the Benefit Falls to Class I RR’s
Rim Thickness Credit – Removed
Window of Opportunity - Stress State

6. DTE vs. CRB Comparison Why Made Codes 66, 72, 74, 83
DTE = 0, AAR CRB = 4,197 total
DTE = 0, AAR CRB = 4,841 total
DTE = 1 shattered rim in 2002, 0 in 2003
AAR CRB = 174 in 2002, 143 in 2003
Chi-square was then used to determine if there is a statistically significant difference for various Why Made Codes…..

7. Chi-square Comparisons
There is a statistically significant difference in wheel removal performance for DTE vs. AAR CRB - DTE has fewer than expected removals for the following Why Made Codes:
65, 67, 75, 76, 78,
This is true despite heavy haul, high mileage coal train operations….

8. The Detroit Edison Strategy
The Theory
The WILD Program
50+ Kip Maintenance Strategy

9. The Theory
The private car owner can effectively manage the life and performance of heavy haul-high mileage wheel sets, by developing a proactive maintenance strategy based upon way-side detection methodologies in combination with Class I Railroads, private railcar maintenance providers, and railway suppliers.

10. The WILD Program Data Purchased from TTCI By Train By Railcar
By Wheel Location
By Graph
Proactive Maintenance Strategy

11. WILD Data

15. 90 70 50

16. 50+ Kip Maintenance Strategy
Wheels Replaced Above 85 kips (3Pt. MA)
50+ Kips - Complete Visual Inspection
Install Cobra Tread Guard Brake Shoes
Computerized SCAT and Date Stencil
Constant Contact Measurement

17. 50+ Kip Maintenance Strategy
Brake Beam & Polymer Slides
Springs & Truck Mechanisms
Re-torque Bolts - Air Valve & Structural
Inspect Relief on Bearing Adaptors
AAR Condemnable Criteria (1/8 WM 64)
Modified NCTA Wheel Specification

18. Wheel Life Trend Analysis
50+ Kip Strategy Began Nov. 2003
1,500 Mile Inspection on Western Class I
35 Trains Completed
33,600 Wheels YTD

19. Wheel Life Trend Analysis
Without 50+ Kip Strategy
01/01/03…11/14/03
80 Randomly Selected
Wheel Sets
With 50+ Kip
Strategy
11/15/03…12/31/03
102 Randomly Selected
Wheel Sets

23. days (avg.) 90 70 50

24. Targeted Considerations
Rail Maintenance Management Forum
Continue Investigative Wheel Failure Analysis
Equipment Health Monitoring TAG
Component Performance Within AAR Rulings
Railcar Capacity and Loading Criteria
Break Even Point: Railroad vs. Private Shop
WILD Site and Software Performance
The PUR Owner Requires a Rule Making Vote

25. DTE Wheel Removals With Brake System Repairs
DTE wheel changes that also had a brake system repair within previous year
Data for 1/03 to 7/04, Job Codes Analyzed
21,035 wheel changes - 1,738 also w/brake repair
Air hose (1,182)
Emergency portion (380)
Service portion (125)
Hand brake related (51)

26. Wheel Removals With Brake System Repairs
By Why Made Code (Data 1/03-7/04)

27. Wheel Removals With Brake System Repairs
By Why Made Code - 75, 65, 78, 67 (1/03-7/04)

28. 2003 Removals with Brake Repairs Chi-Square Analysis
DTE 21,035 removals - 1,185 also with air brake system repair within previous year
WM ,156 removals w/brake repair
WM removals w/brake repair
WM removals - 30 w/brake repair
Compared WM 65, 75, 78 with brake repairs to all 2003 DTE removals with brake repairs - statistically significant difference - “Worse”

29. WILD Issues WILD Calibration: Frequency & Uniformity
Effects of Train Speed
Effects of Track Structure & Sub-grade
Effects of Concrete Ties
WILD Reading Variation
Etc.

30. Conclusions DTE Wheel Removals Wayside Detectors are Useful Tools
Better vs. AAR CRB for WM 65, 67, 75, 76, 78
Also Better Failure Performance Than AAR CRB
Wayside Detectors are Useful Tools
Variation is a Concern
DTE 50+ Kip Strategy Lengthens Wheel Life
Brake System Repairs Linked to Wheel Life
WM 65, 67, 75 Appear to be Affected

31. Wild sites

32. Pareto of whymade since 2003

33. Life span since 2002

34. Whymade 65 life span

35. 3 pt mavg & kips

36. Scatter plot of kips variation vs. train date

37. Grahical summary of sample pool

38. Distribution of kips variation

39. Hypothesis Kips variation vs. Speed change?
Kips variation vs. Car weight?
Kips variation vs. Site change?

40. Kips vs. spd correlation

41. Kips vs. weight correlation

42. Scatter plot of kips var by site-comb

43. Observations Kips variation vs. Speed variation – small correlation.
Kips variation vs. Weight variation – none.
Kips variation vs. different sites – small correlation.

44. MSA – precise & accurate

45. Scatter plot of kips var vs. train date

46. Graphical variation vs. upper and lower control limits

47. Measuring system variation vs. process tolerance

48. Measuring gage tolerance vs. Process tolerance example
Process Tolerance = +/- 7.5 kips, 70 – 85 kips.
Measurement Tolerance is 1/10th of Process Tolerance, = +/-.75 kips.

49. Scatter plot of kips variation vs. train date

50. Financial Impact WM 65, 67 and WM 75 Wheel Changes 2002, 2003 and 2004
,279 Total - 1,769
,179 Total - 2,179
,013 Total - 2,005
AAR Labor Rates per wheel switch $1,000.00
Total Cost $1,769,000
Total Cost $2,179,000
Total Cost $2,005,000

51. Financial Impact Train days lost equals number of wheels pulled.
One train cycle equals seven days (52 cycles per train per year
Total number of cycles lost
(5 sets)
2003 – 311 (6 sets)
2004 – 286 (5.5 sets)
At current lease rates ($500 pcpm), cost to Detroit Edison is approximately $3,600,000.00

52. Summary The 50+ Kip Strategy Extends Wheel Life.
WM65 Exhibits Shortest Life Span.
What’s the contribution of variation by the measuring system? - What are the variation contributions from the Repeatability & Reproducibility tests of the measuring devices?

53. Recommendations Root Cause of Wheel Defects WM 65 & WM 75
Computerized SCAT at All Repair Tracks
Rail Maintenance Management Forum
WILD Calibration Should be Standardized and Meet AAR M-1003 Quality System
Improved Efforts to Share Costs and Benefits Within the Railroad Industry
Private Railcar Owner - Voting AAR Member

54. The Final Thought
Our greatest challenge as an industry is to get the departments responsible for the vehicle performance and the track maintenance to work together and define a wheel rail interaction strategy to obtain synergy between the vehicles, track and the long term profitability of the heavy haul operation.
Source: International Heavy Haul Association (2001)