National Academies Press: OpenBook
« Previous: Part 1: Guidebook
Page 37
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 37
Page 38
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 38
Page 39
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 39
Page 40
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 40
Page 41
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 41
Page 42
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 42
Page 43
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 43
Page 44
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 44
Page 45
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 45
Page 46
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 46
Page 47
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 47
Page 48
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 48
Page 49
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 49
Page 50
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 50
Page 51
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 51
Page 52
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 52
Page 53
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 53
Page 54
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 54
Page 55
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 55
Page 56
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 56
Page 57
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 57
Page 58
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 58
Page 59
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 59
Page 60
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 60
Page 61
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 61
Page 62
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 62
Page 63
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 63
Page 64
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 64
Page 65
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 65
Page 66
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 66
Page 67
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 67
Page 68
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 68
Page 69
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 69
Page 70
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 70
Page 71
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 71
Page 72
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 72
Page 73
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 73
Page 74
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 74
Page 75
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 75
Page 76
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 76
Page 77
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 77
Page 78
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 78
Page 79
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 79
Page 80
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 80
Page 81
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 81
Page 82
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 82
Page 83
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 83
Page 84
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 84
Page 85
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 85
Page 86
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 86
Page 87
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 87
Page 88
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 88
Page 89
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 89
Page 90
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 90
Page 91
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 91
Page 92
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 92
Page 93
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 93
Page 94
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 94
Page 95
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 95
Page 96
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 96
Page 97
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 97
Page 98
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 98
Page 99
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 99
Page 100
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 100
Page 101
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 101
Page 102
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 102
Page 103
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 103
Page 104
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 104
Page 105
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 105
Page 106
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 106
Page 107
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 107
Page 108
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 108
Page 109
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 109
Page 110
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 110
Page 111
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 111
Page 112
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 112
Page 113
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 113
Page 114
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 114
Page 115
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 115
Page 116
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 116
Page 117
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 117
Page 118
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 118
Page 119
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 119
Page 120
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 120
Page 121
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 121
Page 122
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 122
Page 123
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 123
Page 124
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 124
Page 125
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 125
Page 126
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 126
Page 127
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 127
Page 128
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 128
Page 129
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 129
Page 130
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 130
Page 131
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 131
Page 132
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 132
Page 133
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 133
Page 134
Suggested Citation:"Part 2: Final Report." National Academies of Sciences, Engineering, and Medicine. 2010. Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects. Washington, DC: The National Academies Press. doi: 10.17226/14369.
×
Page 134

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Final Report P A R T 2

5 Summary 5 S.1. Definition of Soft Costs 6 S.2. Soft Cost Estimation: State of the Practice 8 S.3. Soft Cost Expenditures: As-Built Analysis 8 S.4. A New Approach to Estimate Soft Costs 9 S.5. Future Research Direction 10 Chapter 1 Introduction 10 1.1. Purpose of This Report 10 1.2. Background 11 1.3. Definition of Soft Costs 13 1.4. Organization of This Report 14 Chapter 2 Literature Review on Soft Cost Definition and Components 14 2.1. Papers and Websites 15 2.2. Indirect Costs 15 2.3. Textbooks and Technical Books 16 2.4. U.S. Army Corps of Engineers Publications 16 2.5. European Sources 17 2.6. Summary and Conclusion 18 Chapter 3 Soft Cost Estimation: State of the Practice 18 3.1. In-Depth Interviews with Professional Cost Estimators 20 3.2. Questionnaire of Transit Cost Estimators 20 3.3. Questionnaire Results: Magnitude of Estimated Soft Costs 25 3.4. Questionnaire Results: Drivers Identified 27 3.5. Questionnaire Results: Impact of Drivers 31 Chapter 4 As-Built Soft Cost Analysis 31 4.1. Approach 31 4.2. Data Source: FTA Capital Cost Database 33 4.3. Potential Issues in Soft Cost Categorization 35 4.4. Historical Soft Costs 42 4.5. Relationships between Cost Drivers and Historical Soft Costs 53 Chapter 5 Conclusion 53 5.1. Literature Review 53 5.2. Soft Cost Estimation: State of the Practice 54 5.3. As-Built Cost Analysis 55 5.4. Future Research Directions 56 Bibliography C O N T E N T S

58 Appendix A Cost Estimators Interviewed 59 Appendix B Project Names and Descriptions in As-Built Analysis 59 B.1. Data Sources for Project Descriptions 59 B.2. Project Descriptions 73 Appendix C Supplementary As-Built Cost Analysis 73 C.1. Data Preparation and Standardization 73 C.2. Adjustments Addressing Different Cost Categorization 74 C.3. Adjustment for Inflation and Nationalization 74 C.4. Outliers Omitted 74 C.5. Vehicle Soft Costs 75 C.6. Soft Costs by Mode and Year 78 C.7. Soft Costs by Complexity: Overall Project Size 82 C.8. Soft Costs by Complexity: New versus Extension 84 C.9. Soft Costs by Complexity: Percentage of Guideway Not at grade 86 C.10. Soft Costs by Complexity: Percentage of Guideway Below Grade 88 C.11. Relationships Among Other Category Unit Costs 90 C.12. Soft Costs by Complexity: Right-of-Way Costs 90 C.13. Soft Costs and Project Development Budget 92 C.14. Soft Costs and Project Development Schedule 95 C.15. Vertical Profile and Soft Cost Measurement 96 C.16. Isolating Agency-Specific Effects

This report presents the research, data sources, and analysis underlying Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects, Part 1: Guidebook, which came out of TCRP Project G-10. This Final Report is one of two final products from the proj- ect and is intended to support the information summarized in the Guidebook in Part 1. Please refer to the Guidebook for a summary of how the results of the research presented here can be applied to practice, including an introduction to “soft costs” and a new method- ology to estimate these soft costs based on historical projects. To support the development of a guidebook for agencies on soft costs, this report: • Identifies a working definition of soft costs, • Describes the current industry practice of estimating soft costs through a questionnaire of the transit industry and interviews with industry professionals, • Statistically analyzes the as-built costs of 59 past transit projects to determine how proj- ect characteristics have driven soft costs historically, and • Introduces a new methodology for estimating soft costs based on actual past expenditures, presented in the Guidebook. S.1. Definition of Soft Costs Generally, soft costs (or indirect costs) are the capital expenditures that are required to complete an operational transit project but that are not spent directly on activities related to brick-and-mortar construction, vehicle and equipment procurement, or land acquisition. Instead, these expenses are incurred on ancillary professional services that are necessary to complete the project. After reviewing a variety of financial, engineering, academic, and other literature, this study concludes that the Federal Transit Administration’s (FTA) definition of Standard Cost Category (SCC) 80, Professional Services, is an equivalent operational definition of soft costs for the purposes of this project. FTA (U.S. FTA, 2008) defines SCC 80 as follows: [Soft costs include] all professional, technical and management services (and related professional liabil- ity insurance costs) related to the design and construction of fixed infrastructure during the preliminary engineering, final design, and construction phases of the project. This includes environmental work, design, engineering and architectural services; specialty services such as safety or security analyses; and value engineering, risk assessment, cost estimating, scheduling, before and after studies, ridership modeling and analyses, auditing, legal services, administration and management, etc. by agency staff or outside consultants. 5 S U M M A R Y Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

The kinds of soft costs encountered in rail transit construction projects in the United States can vary widely depending on project characteristics, local regulations, and the administra- tion practices of the sponsor agency. Most new rail construction projects will incur certain “typical” soft cost expenditures, while other soft cost components can be unique to the proj- ect, as shown in Table 1. Evidence suggests that European cost estimators are also trying to standardize a defini- tion of soft costs, but the term is rarely comparable to how it is used in U.S. practice. The definition of soft costs to a transit agency can differ from a construction contractor or other stakeholders, depending on institutional context. The point of view of a U.S. transit agency is taken in this report. S.2. Soft Cost Estimation: State of the Practice Interviews with and a questionnaire administered to estimators revealed that cost esti- mators for transit construction projects follow different approaches to estimating soft costs depending on the phase of the project. During early phases of planning [alternatives analysis (AA) or preliminary engineer- ing (PE)], a transit project is only conceptually defined, and the soft costs are as well. At these early stages, transportation planners usually identify a single corridor for construc- tion but develop a range of options for more specific details such as mode, alignment, station locations, and, as a result, construction costs. Most attention is on construction costs at this phase since soft costs are difficult to predict given the conceptual nature of the project. Estimators apply default costs to approximate construction quantities, reme- diation, and other “hard” costs, and then simply add a set of percentages of hard costs (e.g., 30%) to approximate an initial soft cost estimate, as shown in Figure 1. At this phase, the central question is what percentages to apply. Based on interviews and an industry questionnaire, most estimators report that they choose percentages from within a range for each soft cost component based on historical experience and project characteristics. Figure 2 shows the midpoint percentages used by 10 cost estimators rep- resentative of the transit industry, broken down by the soft cost component. Typically, these “add-ons” represent an additional cost to the project of around 25–35% of con- struction costs. However, these midpoints are not applied blindly. Estimators may begin with these aver- ages but choose higher or lower percentages from within a range based on their knowledge 6 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects TYPICAL SOFT COSTS INCURRED IN MOST PROJECTS Design and engineering services for preliminary engineering and final design Transit agency staff managing project, development, construction, and customer information Reimbursement to external entities such as police, utilities, and other costs of local and state government Insurance LESS TYPICAL SOFT COSTS INCURRED IN SOME PROJECTS, DEPENDING ON CHARACTERISTICS Professional services to support acquiring real estate for right-of-way Third-party contractor managing construction Design and engineering services to re-design a project, due to unforseen circumstances Table 1. Types of soft costs encountered in rail transit construction. Construction Cost Soft Costs x Percentage = $ Guideway $ Vehicle Cost Vehicles $ Vehicle Soft Costs $ Stations $ Maintenance Yard $ Etc. $ TOTAL $ TOTAL $ Real Estate Cost Acquisitions $ RE Soft Costs $ TOTAL $ TOTAL PROJECT COST $ Figure 1. Cost esti- mation in early project phases.

Summary 7 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1* 2* 3 6 7 4 8 9 5 10 M id -R an ge E st im at ed S of t C os t (% of C on str uc tio n) Questionnaire Respondents Other Insurance + Legal Project Mgmt. and Construction Admin. FD PE *Respondents estimate PE + FD as combined amount; PE displayed here using average split Figure 2. Midpoint soft cost estimates for all components during project planning phases. LOWER % SOFT COSTS MODE PROJECT DELIVERY Bus Rapid Transit Design–Build MIXED/MID-RANGE % SOFT COSTS Commuter Rail Light Rail Design–Bid–Build Elevated Alignment New Right-of-Way HIGHER % SOFT COSTS Heavy Rail Tunnel Alignment Differing Subsurface Conditions Design–Build–Operate–Maintain Full Turnkey ALIGNMENT OTHER CONDITIONS Table 2. Project characteristics guiding soft cost percent estimates within a range. of the project, the sponsor, and their experience with similar past projects. Table 2 lists some of the project characteristics that estimators generally use to guide their choice of a percent- age within a range during planning phases. Estimators report that they may choose figures up to 10% higher or lower than their starting points based on judgment and the character- istics of the project. During the final design (FD) and construction phases, estimates of soft costs based on a percentage of construction cost are replaced with more closely tailored, bottom-up estimates relying heavily on past experience with similar projects. For instance, adminis- tration costs may be estimated based on headcount multiplied by the duration of the project, as determined from construction schedules. Also at this stage, more costs are known: preliminary engineering work is largely complete, and the sponsor’s contracts for construction management and any remaining design work may already be executed for an agreed-upon cost.

S.3. Soft Cost Expenditures: As-Built Analysis This report analyzes a database assembled by the Federal Transit Administration of as-built costs for 59 rail transit construction projects in the United States over the past four decades (summarized in Figure 3) and concludes that: • The current industry practice of using percentage add-ons for soft costs appears to be a valid approach to estimating soft costs. Project characteristics such as complexity, mag- nitude, mode, context, and others identified by industry estimators are correlated with soft costs in dollar and percentage terms. • Soft cost expenditures have averaged around 30% of construction costs, with a range across all projects of between 11% and 54%, depending on the characteristics of the proj- ect, as Figure 3 indicates (outliers excluded). • Cost estimators typically begin estimating soft costs with average percentages that corre- spond closely to historical averages for each soft cost component, as Table 3 shows. • However, actual soft costs in past projects have shown a wider range of variability than estimators currently use. While estimators report choosing from within a range of around 20 percentage points, past projects have varied within a range of around 40 percentage points (outliers excluded). • Some variability in soft costs cannot be explained solely with information available to the estimator prior to construction. The statistical analysis applied in this research was able to explain around 60% of the changing relationship between hard and soft costs with data available during planning phases. This suggests that the remaining variability in soft costs must be estimated with a blend of science, judgment, and art. S.4. A New Approach to Estimate Soft Costs This report is accompanied by a guidebook that presents a new method to estimate soft costs for a planned transit project that is firmly rooted in historical experience. The Guide- book also serves as a primer on soft costs and takes the reader through a step-by-step process to estimate the relationship between a given transit project’s hard costs and its likely soft costs, given certain characteristics about the project and its sponsor. 8 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 0% 10% 20% 30% 40% 50% 60% H ud -B er g II H ia w at ha Sa lt La ke Sa cr am . I H ud -B er g I Po rtl an d So Sa cr am . F ol Po rtl an d In t Ch ar lo tte LA G ol d Ea st LA G ol d Pa sa Sa cr am . S o St . L ou is Po rtl an d W So ut h NJ Pi tts bu rg h N Po rtl an d Se g1 Pi tts bu rg h I LA B lu e Sa n Di eg o VT A Ta s E Pi tts bu rg h II D en ve r S W VT A Ca pi to l Ph oe ni x CT A O 'H ar e D C U St . D C An ac os t D C Ad di so n D C L'E nf an t D C Ne w Ca D C Sh ad y G D C Hu nt gt n D C G le nm t 1 CT A Do ug la s N YC T 63 rd CT A O ra ng e Ba lti m or e D C Vi en na M BT A O ra ng CT A Br ow n M AR TA N -S M ia m i Sa n Ju an D C An ac os t O D C Sp rin gf ld D C G le nm t 2 D C G re en bl t BA R T SF O Ph il Fr an kf . N YC T St illw So ft Co st s (% of C on str uc tio n) Light Rail Heavy Rail Figure 3. Historical soft costs by project and mode (outliers excluded).

S.5. Future Research Direction This report and the accompanying Guidebook give transit agencies and other project sponsors a better understanding of what soft costs are, how they are estimated, and what has caused changes in soft costs in past projects. The Guidebook synthesizes the research and analysis from this technical report into a straightforward primer on soft costs and introduces a new methodology to estimate soft costs based on a review of historical drivers and costs. More in-depth research into the documentation of one or more recent construction proj- ects will enhance the understanding of the exact composition of soft costs and cost drivers. Future research might further examine the more-detailed elements of soft costs below the Standard Cost Category component level and document more of the estimation techniques used in later project phases. Given the specificity of this work, the research may need to be more closely tailored to a specific mode or operating environment (e.g., streetcar versus light rail on exclusive right-of-way). Moreover, a comprehensive industry outreach program will provide further insight on context-specific soft-cost estimation practices. Finally, the methodology to estimate soft costs for public transportation infrastructure projects developed here is based on past heavy and light rail construction projects and is therefore not entirely applicable to other prevalent public transportation capital infrastruc- ture projects such as bus rapid transit (BRT), commuter rail, streetcar, or state-of-good- repair projects to repair or replace aging infrastructure. Additional data and research would help estimate soft costs for these kinds of projects. Summary 9 Table 3. Comparing industry practice to historical actuals: soft costs as a percentage of hard costs.

1.1. Purpose of This Report This Final Report presents the research, data sources, and analysis underlying Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects, Part 1: Guidebook, which came out of TCRP Project G-10. The purpose of TCRP Project G-10 was to research soft costs in major public transportation capital infrastructure projects, with the goal of producing a guide for proj- ect sponsors to learn more about these costs and better estimate them in the future. This Final Report is one of two final products from the project and is intended to support the information summarized in the Guidebook in Part 1. Please refer to the Guidebook for a summary of how the results of the research presented here can be applied to practice, including an introduc- tion to soft costs and a new methodology to estimate these soft costs based on historical projects. 1.2. Background When a new rail transit project is proposed in the United States, its capital cost is one of the most visible and important characteristics in the public deliberation over whether to build the project. The capital cost of a rail project factors prominently when deciding alignment and mode during the alternatives analysis and preliminary engineering phases. It is reported in the press, debated by stakeholders, influences the public’s perception of a sponsoring transit agency, and ultimately helps determine whether the project is ever constructed. The capital cost is also a cru- cial input to the cost-effectiveness indicator that helps determine the project’s eligibility for fed- eral funds. In addition, the transit industry has recently come under scrutiny for perceived per- sistent underestimating of capital costs and its ability to contain such costs. While research on the transit industry has focused primarily on hard construction costs and estimation techniques, rel- atively little literature exists on the composition and estimation of soft costs for transit projects. Historically, soft costs have accounted for a significant portion of a capital project’s total expenditures, yet many agencies know little about soft costs. As this report discusses, most rail transit projects’ soft costs have ranged from as low as 11% to as high as 54% of hard construc- tion costs. Given the importance and public scrutiny of transit capital costs and the relative inattention to a cost category that makes up a significant portion of expenditures, the transit industry may benefit from improved information on soft costs. Therefore, the research team for TCRP Project G-10 hopes to help the transit industry better understand: • The definition, importance, composition, and timing of soft costs; • How the industry currently estimates soft costs, depending on project phase; 10 C H A P T E R 1 Introduction

Introduction 11 • What has driven soft cost expenditures in the past; and • How soft costs can be estimated in the future. Increasing the integrity, accuracy, and reliability of soft cost estimates will improve the indus- try’s public perception and deliver public transportation infrastructure more cost-effectively. The ultimate objective of the researchers was a guidebook for estimating soft costs for major transit capital projects that walks a project sponsor through each step in building up a soft cost estimate. 1.3. Definition of Soft Costs Generally, soft costs are the capital expenditures that are required to complete an operational transit project but which are not spent directly on activities related to brick-and-mortar con- struction, vehicle and equipment procurement, or land acquisition. Instead, these expenses are incurred on ancillary professional services that are necessary to complete the project. Soft costs are the expenditures necessary to develop and manage the project, whereas hard costs are the expenditures required for construction. Soft costs are a necessary part of a construction project because building or rehabilitating transit infrastructure requires more than the direct payments made to a general construction contractor or a vehicle vendor. The Federal Transit Administration requires that all candidate and recipient projects of New Starts funds organize and report their project cost estimates in the same way, using the Standard Cost Category structure. This structure consists of ten major categories (as shown in Table 4), each of which is further broken down into components. For example, the SCC 50 Systems cost category includes separate components for Train Control, Traction Power, Communications, and Fare Collection. Standard Cost Category 80, Professional Services, consists of eight separate components that together encompass all services and activities commonly associated with project soft costs (although some exceptions are discussed below). In addition, a literature review on soft costs concludes that the existing engineering, technical and international professional literature on the definition of soft costs is consistent with the FTA’s description of SCC 80, Professional Services, in the Standard Cost Category Workbook (U.S. FTA, 10 Guideway & Track Elements (route miles) 20 Stations, Stops, Terminals, Intermodal (number) 30 Support Facilities: Yards, Shops, Admin. Bldgs 40 Sitework & Special Conditions 50 Systems 60 ROW, Land, Existing Improvements 70 Vehicles (number) 80 Professional Services 90 Unallocated Contingency 100 Finance Charges Total Project Cost (10–100) 80.01 Preliminary Engineering 80.02 Final Design 80.03 Project Management for Design and Construction 80.04 Construction Administration and Management 80.05 Professional Liability and Other Non-Construction Insurance 80.06 Legal; Permits; Review Fees by Other Agencies, Cities, etc. 80.07 Surveys, Testing, Investigation, Inspection 80.08 Start Up Table 4. FTA Standard Cost Categories and Category 80 components.

12 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 2008). Furthermore, using the SCC structure and the definition of SCC 80 is consistent with the historical analysis that underpins the new soft-cost estimation methodology discussed later. For this reason, the researchers for this project adopted the definition and structure of FTA SCC 80, Professional Services, as being equivalent to the definition of soft costs. The FTA’s char- acterization (U.S. FTA, 2008), restated below, is therefore a reasonable and consistent definition and has been used throughout the project: [Soft costs include] all professional, technical and management services (and related professional liability in- surance costs) related to the design and construction of fixed infrastructure during the preliminary engineering, final design, and construction phases of the project. This includes environmental work, design, engineering and architectural services; specialty services such as safety or security analyses; and value engineering, risk assessment, cost estimating, scheduling, before and after studies, ridership modeling and analyses, auditing, legal services, administration and management, etc. by agency staff or outside consultants. It is important to keep in mind institutional or contractual perspective when referring to soft costs. Although this research views soft costs from the perspective of the project sponsor or FTA, the classification of soft costs within the construction industry can take on somewhat different meanings, depending on institutional context. As Figure 4 illustrates, the project sponsor will likely view soft costs as the non-construction professional services costs identified in Standard Cost Category 80. Expenditures in other cost categories reflect the sponsor’s expenditures on direct activities, perhaps primarily composed of payments to the vehicle vendor or construction contractor. The construction contractor, in turn, may view some portion of their total construction con- tract as indirect or soft costs for their organization, such as the cost of contract administration, home office overhead, and related expenses that are built into the contract amount. These indi- rect costs represent real costs of doing business to the construction contractor, but since they cannot be clearly attributable to a specific project, the construction contractor is likely to charge various projects in some proportional manner. In addition, some costs that are clearly attributable a specific project cannot be attributed to physical components of the project (such as concrete or steel), and these may be referred to as “general conditions.” While these activities sound similar to the types of services identified in SCC 80, they are the contractor’s (not the sponsor’s) costs and are therefore considered hard costs outside of SCC 80. These multiple perspectives on indirect or soft costs are illustrated schematically in Figure 4. To keep matters clear, this research assumes the perspective of a transit agency sponsoring major construction where at least some design and all construction work is to be performed by Veh/R.E. Costs Total Project Capital Costs “Hard” Construction Costs “Soft” Costs “Hard” Construction Contract “Indirect” Costs Sponsor Agency and FTA’s Perspective Construction Contractor’s Perspective “Direct” Costs Focus of this Report “Direct” Construction ProjectProjectPerspective Direct “General Conditions” Figure 4. Capital costs from sponsor, contractor, and project perspectives.

outside contractor(s). For example, while a construction contractor might build their expected overhead costs into their bid, their total bid price from the transit agency’s perspective, and according to the definition above, is a hard construction cost. 1.4. Organization of This Report This report consists of five broad sections: • A literature review on the definition and components of soft costs; • Results from an industry questionnaire and interviews about how soft costs are estimated; • Analysis of the relationship between project characteristics and actual as-built soft costs from 59 past rail projects, including univariate and multivariate analyses; • A summary of the analysis underlying the Guidebook’s new soft-cost estimation technique; and • Concluding remarks and directions for future research. Introduction 13

The objective of this literature review is to find a consistent definition for the term “soft costs” in the context of capital construction projects and to decide what cost items fall under this term. The following general sources of information were reviewed: • Technical and scholarly articles published in archival U.S. journals; • Articles presented in various U.S. professional conferences and published as part of the proceedings; • Internet sources; • Books, and specifically, engineering textbooks; • U.S. Army Corps of Engineers publications; and • European cost estimation sources. 2.1. Papers and Websites Many organizations and industry groups publish definitions of soft costs on the Internet or in readily available literature. It is important to note that the term “soft cost” is not used commonly in the technical literature. However, many of the cost items associated with soft costs are cov- ered under the definition of indirect costs. • The Association of Physical Plant Administrators (APPA, 2005) has the following definition: Soft costs include such items as architecture, design, engineering, permits, inspections, consultants, environ- mental studies and regulatory demands needing approval before construction begins. Soft costs do not include construction, telecommunications, furnishings, fixed equipment and expenditures for any other permanent components of the project. . . . . These costs are related to those items in a project that are necessary to prepare and complete the non-construction needs of the project. While the main components are the same, there is a distinction in this terminology that limits the definition of soft costs to costs incurred prior to construction. In the FTA SCC defi- nition, soft costs include professional and managerial services during the construction phase as well. • KRG Insurance Group (2002) defines soft costs for building and entrepreneurial projects as follows: “Soft Costs” may be defined as those indirect additional expenses that form part of the construction or repair of property. They not only impact on building cost but on business revenues. . . . In a typical ac- counting summary of construction costs on a new project it is normal for soft costs to comprise up to 30% of total expenses. 14 C H A P T E R 2 Literature Review on Soft Cost Definition and Components

Literature Review on Soft Cost Definition and Components 15 Here is a partial listing of soft costs incurred in building construction: Architect and engineer fees, audit and bookkeeping charges, realty taxes/assessments, advertising and promotional expenses, real estate commissions, tenant inducement expense, premiums—Insurance/bonds, license and permit fees, increased mortgage costs, additional loan expenses, legal expenses, cost of vacancy, increased cost of labor, security expenses, and penalties. • Constructionplace.com (as of 10/5/2007) has the following definition for soft costs: Soft Costs are cost items in addition to the direct Construction Cost. Soft Costs generally include archi- tectural and engineering, legal, permits and fees, financing fees, construction interest and operating expenses, leasing and real estate commissions, advertising and promotion, and supervision. Further, this source contends that the terms indirect costs and soft costs are synonymous. 2.2. Indirect Costs Few professional publications have used the term “soft costs,” instead discussing many ele- ments of soft costs at some length under indirect costs. However, extending the search to include the term “indirect costs” yielded more information regarding various elements of soft costs. In considering indirect costs it is important to identify the relevant perspective. As an example, a cost item that is considered an indirect cost for a project (general conditions) may be categorized as a direct cost from a contractor’s perspective. In the same way, a cost that can be considered a direct cost for a provider of professional services (such as labor cost) could be considered an indirect cost from the owner’s or sponsor’s perspective. • For example, the Association for the Advancement of Cost Engineering (AACE, 2007) offers the following definition for indirect costs: Costs [that] are not directly attributable to the completion of an activity. Indirect costs are typically allocated or spread across all activities on a predetermined basis. In construction, all costs which do not become a final part of the installation, but which are required for the orderly completion of the installation and may include, but are not limited to, field administration, direct supervision, capital tools, startup costs, contractor’s fees, insurance, taxes, etc. This definition is from the perspective of the contractor. For the sponsor of a capital project, such as a transit agency dealing with multiple contracts in the same project, the contractor’s gen- eral conditions and home office overhead could be considered direct costs because they are clearly attributable to that specific contract. It is therefore conceivable that the whole construction con- tract could be considered a hard cost. Since the purpose of this guide is to help project sponsors estimate project soft costs with greater accuracy, this analysis takes the perspective of the proj- ect sponsor and treats the general conditions and home office overhead as construction costs. 2.3. Textbooks and Technical Books Ten construction management textbooks were also selected for review because they have been commonly used in various universities and other academic settings for years. The term “soft costs” was only used in one of these textbooks (Bartholomew, 2000, p.252) as follows: In development, as distinct from actual construction, direct costs are the hard costs, the total construction costs that include what we call in estimating both direct and indirect construction costs. The indirect, or soft, costs in development include the costs of financing, advertising and sales, fees, insurance, ground rent and taxes during construction, and the costs of land rights.

16 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 2.4. U.S. Army Corps of Engineers Publications U.S. Army Corps of Engineers publications on the Internet addressing the subject of construc- tion costs included no reference to the term “soft costs.” However, the Corps’ estimating sources deal with non-construction costs, and these sources can be used in the current research. As an example, the U.S. Army Corps of Engineers document Engineering Instructions— Construction Cost Estimates (1997) describes the process of cost estimating as prescribed by the Corps. This document does not use the term “soft costs”; however, costs are divided into con- struction costs and the non-construction activities costs for real estate; planning, engineering, and design; and construction management. This document also provides a work breakdown structure (WBS) for organizing the cost estimate; several categories of this WBS include soft cost items as defined in the FTA SCC description of Category 80. The document can be used for obtaining information on the Corps approach for estimating non-construction costs even though the main emphasis is on construction costs. 2.5. European Sources Research was conducted to evaluate the European approach to cost classification and to see how European countries keep track of project soft costs. Sources were reviewed in places includ- ing Germany, the United Kingdom, and Switzerland. It appears that the term “soft costs” is not used to identify non-construction costs. However, one useful source was the European Committee for Construction Economists (CEEC) Code of Measurement for Cost Planning (CEEC, 2004). The CEEC was established over 20 years ago as a European organization in the field of real estate economy. A working group of this committee focused on the problem of differences between national codes for the measure- ment of quantities and classification of construction costs (Stoy and Wright, 2006). This work- ing group created the CEEC Code of Measurement for Cost Planning as a high-level standard sum- mary for the classification of costs in construction and real estate. Table 5 provides general categories of costs (cost groups) according to the CEEC. CEEC’s Code of Measurement for Cost Planning applies the codes of Belgium, Switzerland, Germany, Netherlands, Ireland, and the United Kingdom to arrive at a uniform approach for categorizing construction costs. In this cost breakdown, general conditions costs (project over- head) can be found under cost group A (in Table 5). Items that may be classified as soft costs are found mainly under cost groups L, M, O, and X. As an example, cost group M, “ancillary costs and charges,” is described as follows: General incidental costs to the client including the costs of models, documentation, copies of drawings, etc., laying of foundation stone, topping out, inauguration, competitions, permits, planning charges, connection charges for utilities, insurances, third party compensation, client’s involvement, legal fees in association with construction, compensation payments due to statuary requirements. The titles of cost groups can sometimes be misleading. As an example, project management and project administration costs are listed under cost group L, “design team fees.” Insurance costs can be captured under items “779—general incidental building costs, other items” and “790—other incidental building costs” (cross-referenced from German DIN 276/1993). Legal fees for land acquisition are under category U. Categories J and N are reserved for con- tingencies, and category V is reserved for finance. There is a major effort in Europe to standardize construction cost categories across European nations. The general approach is not unlike the U.S. approach where the costs are divided into

construction and non-construction costs, although the term “soft costs” is not used even in English speaking nations (the United Kingdom and Ireland). 2.6. Summary and Conclusion In general, despite minor differences, the various definitions of soft costs in the professional publications are generally consistent. Methods of estimating or allocating these costs vary and change from organization to organization. From this literature review the researchers conclude that the definition provided by the FTA in Standard Cost Category 80 is a comprehensive defi- nition consistent with most of the sources that were reviewed. Literature Review on Soft Cost Definition and Components 17 Table 5. Breakdown of cost categories according to the CEEC. CONSTRUCTION COSTS A Preliminaries B Substructure C External superstructure/envelope structure D Internal superstructure E Internal finishings F Services installations G Special equipment H Furniture and fittings I Site and external works J Construction contingencies K Taxes on construction DESIGN AND INCIDENTAL COSTS L Design team fees M Ancillary costs and charges N Project budget contingencies O Taxes on design and incidental costs COSTS IN USE P Maintenance Q Operation R Disposal S Decommissioning T Taxes LAND AND FINANCE U Land costs V Finance W Grants and subsidies X Taxes on land

Industry practices for developing soft cost budgets were assessed using a questionnaire com- pleted by construction cost estimators at a variety of transit agencies, and in-depth interviews were conducted with experienced professional cost estimators in public transportation. 3.1. In-Depth Interviews with Professional Cost Estimators To develop an initial picture of how the transit industry estimates soft costs, in-depth inter- views were conducted with professional cost estimators. The following sections describe the find- ings of these interviews. 3.1.1. General Approach From the interviews, it is clear that sponsors of major new transit projects approach estimat- ing soft costs differently depending on how far along the project is in the planning process. Over time, as a project becomes better defined, the soft-cost estimate process increases in sophistica- tion from a proportionate approximation to a more detailed or “bottom-up” estimation for each functional aspect of soft costs. 3.1.2. Soft Cost Estimation during Early Planning Phases Early in the project development phase, such as during alternatives analysis or preliminary engineering, a transit project is only conceptually defined, as are the soft costs. At these early stages, transportation planners may identify a single corridor for construction but develop a range of options for more specific details such as mode, alignment, station locations, and, as a result, construction costs. At this stage, capital cost estimates are very important, especially because they are a crucial input to the project’s cost effectiveness, which can help determine eligibility for federal funding. However, despite the early importance of capital cost estimates, soft cost estimates are approxi- mations at best in such early phases. Soft costs are generally approached as a percentage add-on to capital costs during alternatives analysis and are an approximation only. As a result, most atten- tion focuses on hard costs, not soft costs, at this stage. Because of the conceptual nature of the project and the emphasis on hard costs at this stage, soft costs are usually treated as percentage add-ons to estimates of hard construction costs. Estimators begin by estimating each soft cost component as a percentage of construction costs, choosing a percentage for each component within a range depending on a variety of factors. For instance, 18 C H A P T E R 3 Soft Cost Estimation: State of the Practice

Soft Cost Estimation: State of the Practice 19 during conceptual design a sponsor might begin by estimating final design costs as 9% of con- struction costs, but then increase that estimate to 11% if they know the project is likely to require a more complex design due to special circumstances. Cost estimators interviewed for this study identified the following project characteristics as cost drivers: • Mode (generally, soft cost percentages for highway projects are lower than for transit projects, which tend to be more complex with more unknowns); • Vertical alignment (underground segments usually add to soft cost percentages); • Traffic impacts and relocations around the construction site; • Level of public support and acceptance of the project; and • Local and regional politics that can complicate the project development process, including alignments, delays, local funding share and methods, and other concerns. The project characteristics that were identified in the interviews as cost drivers, listed above, are very similar to those identified in the questionnaire (as shown in Table 8). 3.1.3. Soft Cost Estimation during Later Design and Construction Phases If a project proceeds into preliminary engineering and final design and becomes better defined, the soft-cost estimation approach changes, and percentages are rarely used. Instead, percentage estimates are replaced with more closely tailored, bottom-up estimates relying on a more detailed understanding of the project than was available in earlier stages and relying on past experience with similar projects. For example, an estimator might forecast design costs using a standard number of drawings per station and drawings per linear foot of guideway and apply a standard per-drawing cost. Agency administration and management costs might be based on headcount, staff salaries, and project duration, in combination with the project’s operational requirements. Third-party reimbursement and other costs in SCC 80.06 might be estimated based on construc- tion duration per station as well as headcount. Right-of-way soft costs might apply assessed actual property values rather than a gross estimate of acquisition and real estate costs. Importantly, the project faces external pressure to adhere to whatever soft cost estimate is assigned to the project during final design. The public, agency staff, FTA, and other oversight bodies tend to expect that the SCC budget line items as defined at final design will not change. In particular, FTA wants to avoid major budget revisions after final design and highly scrutinizes soft cost esti- mates at this stage. As a result, the cost estimator will typically approach each soft cost compo- nent with a conservative estimate. During the construction phase, project management has little influence on the incurrence of soft costs. Due to the prior attention to the major SCC budget, the project sponsor might be reluctant to change the major SCC line items at the category level, although budget revisions within components are less difficult. To a great degree, the sponsor may be “stuck with the number” once construction begins. Some redesign may be necessary for differing or unexpected site conditions. Once construction is underway, the management interface between agency and contractor is the most important determinant of soft cost expenditures; other potential factors have relatively little influence on soft costs at this point. The FTA’s oversight, local regulations and building codes, and other potential complexities will have only minimal effect on soft cost expenditures. The effect of project delay (for whatever reason) can be mixed: some soft costs, such as manager salaries, are calendar-based and will continue regardless of progress, while other soft costs can be slowed or halted altogether as the project demands, such as when the design contractor tem- porarily reduces ongoing work on a project.

20 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects The market for construction management and design professional services itself can have an impact on construction costs and thereby the relative magnitude of soft costs. Contractors may bid lower if the market is weak, and vice versa. 3.2. Questionnaire of Transit Cost Estimators To supplement the interviews, a questionnaire on soft cost estimation was completed by tran- sit professionals and cost estimators at several consulting and engineering firms and transit agen- cies. The questionnaire was intended to build on the qualitative information gathered from the interviews by adding more quantitative information. The questionnaire had three objectives: to summarize the spectrum of soft cost percentages used in the industry by soft cost components, to identify the characteristics (or cost drivers) of a project that would change those percentages, and to measure how much the percentages might change within the range based on project char- acteristics. The questionnaire was transmitted to nine transit industry members of various sizes, from which 7 data points were collected—5 from transit agencies and 2 from agencies’ planning consultants working on a specific capital project. Several respondents reported different estima- tion techniques and percentages at different project phases; this yielded a total of 10 data points for analysis. 3.3. Questionnaire Results: Magnitude of Estimated Soft Costs Results from the first section of the questionnaire revealed that most agencies and contractors estimate soft costs as a percent of construction costs roughly consistent with the SCC structure; however, they use a fairly wide range of percentages, depending on context. These results are pre- sented in Table 6. The questionnaire asked respondents to report a midpoint as well as a high and low percent- age for each cost component; however, some respondents supplied only a range or only an approx- imate midpoint. Where only a midpoint was noted, ranges are omitted, and where only high and low ranges were given, the mathematical average is shown. Some agencies provided percentage estimates that varied depending on project phase, resulting in multiple data points in the results presented in this document. Several respondents noted other soft costs that are estimated on some basis other than a fixed percentage of construction costs. For example: • Respondent 7 usually reserves around $1 million for a before-and-after study, regardless of relative project magnitude; • Respondent 9 estimates resource needs for agency force account and flagging work on a project- specific basis, without using a percentage; and • Similarly, respondent 10 estimates startup costs not as a percentage of construction costs but on a project-specific basis. While most questionnaire respondents roughly followed the FTA SCC structure when esti- mating costs, there were some exceptions. For example, respondents 3, 4, and 5 use a single value to address both SCC 80.03, Project Management for Design and Construction, and 80.04, Con- struction Administration and Management. Respondents 1 and 2 estimate preliminary engineer- ing and final design with a single value as well. Some respondents noted a percentage multiplier to estimate planning efforts in the early phases of project development, such as alternatives analy- sis, whereas many did not. This may be because these costs are already largely spent by the time

1 98765432 10 --- --- --- --- --- --- --- 2% for Environmental --- --- 80.01 1-2% (<1 to 2+%) (complete, same as at left) (complete, same as at left) 3% (2.3 to 3.8%) 3% 4% (3 to 6%) (3 to 5%) 7.5% (7 to 8.5%) 80.02 10% (6-15%) 11% (7-16%) 9-11% (<8 to 11+%) 9% (7% to 10%) (complete, same as at left) 10% (7.5 to 12.5%) 7% 8% (7 to 12%) (6 to 10%) 7.5% (7 to 8.5%) 80.03 8% (5-12%) 9% (6-12%) 8-10% (<8 to 10- 12%) 18-20% (15 to 20+%) 17-19% (15 to 20%) 10% (7.5 to 12.5%) 12% of PE, then 12% of FD for PMC 8% (4 to 8%) (4 to 6%) 9% (8.5 to 9.5%) 80.04 --- --- 5% (3.8 to 6.3%) 12% 5% (2 to 5%) (6 to 10%) 14.5% (14 to 15%) 80.05 4% (2-6%) 4% (2-6%) 1.5-2% (<1 to 2.5+%) 1.5-2.5% (<1 to 2.5+%) 2% (1 to 3%) 0.1% (0.0 to 0.1%) 4% (3 to 6%) (3 to 7%) 1% 80.06 --- --- 0-2% (0 to 2+%) --- 0-2% (0 to 2+%) 0.7% (0.5 to 0.9%) 0.25% (2 to 4%) 3% 80.07 2% (1-4%) 2% (1-4%) 0-2% (0 to 2+%) --- 0-2% (0 to 2+%) 1% (0.8 to 1.3%) --- 2.5% (2 to 3%) (2 to 3%) 0.5% 80.08 --- --- 2% (0 to 2+%) 2% (0 to 2+%) 3% (2 to 4%) 0.6% (0.5 to 0.8%) 6% 3% for Start up and Artwork --- Not estimated with % --- --- --- --- --- --- $1m for Before/After Study 4% of SCC 60 for ROW Engineering; 7% of SCC 60 for Agency ROW Costs; 12% of SCC 70 for Vehicle Design and Agency Costs Agency Force Account Work - Flagging Costs: As Needed Note: Midpoint reported first, figures in parentheses indicate upper and lower bound of range Other (included in Project Management above) Insurance 3% for Insurance and LegalLegal; Permits; Review Fees by other agencies, cities, etc. Project Management for Design and Construction Construction Administration & Management Surveys, Testing, Investigation, Inspection Start up Planning and Feasibility Preliminary Engineering Final Design Questionnaire Respondent: (included in Final Design below) Table 6. Summary of soft cost percentages reported in questionnaire.

22 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects FTA SCC Category Shown Here 80.01 Preliminary Engineering Preliminary Engineering 80.02 Final Design Final Design 80.03 Project Management for Design and Construction 80.04 Construction Administration and Management Project Management and Construction Administration 80.05 Professional Liability and other Non- Construction Insurance 80.06 Legal; Permits; Review Fees by other agencies, cities, etc. Insurance and Legal 80.07 Surveys, Testing, Investigation, Inspection Surveys, etc. 80.08 Start Up Start Up Table 7. FTA Standard Cost Categories combined to report questionnaire results. 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 1 2 3 4 5 6 7 8 9 10 PE C os t E st im at e (% of Co ns tru ct io n) Questionnaire Respondent Upper Bound Midpoint Lower Bound Es tim at ed w ith F D Es tim at ed w ith F D Figure 5. Preliminary engineering soft cost estimates. an estimate is made or because the FTA directs agencies to exclude these costs in the SCC work- sheet instructions (U.S. FTA, 2008). This report combines several cost categories, as shown in Table 7 above: The following section compares the questionnaire responses for each cost component, again with some FTA SCC components combined for reporting purposes. Figure 5 shows the estimates for preliminary engineering provided by questionnaire respon- dents. Most agencies report a range of approximately 2–4%. Questionnaire respondents reported using a fairly consistent range of between 7 and 11% of construction costs to estimate final design costs, as shown in Figure 6. However, these estimates go as high as 16%. Note that the percentages for respondents 1 and 2 include an estimate of pre- liminary engineering soft costs as well. Responses were more varied as to the percentage of construction costs estimated for project management, construction management, and administration, as Figure 7 shows. Most estimates were in a range of around 7–19%, but some were as low as 5% and some were as high as 23%.

The relatively wider variance between respondents here may be due to the definition of manage- ment costs in major infrastructure projects involving a sponsoring public entity and multiple contractors, and the demarcation of where agency oversight ends and contractor oversight begins. As the literature review indicated, the definition of soft costs can often depend on institutional perspective or a project sponsors’ decision regarding how much oversight and management to retain for agency staff and how much to contract out. If an agency expects a construction con- tractor to assume more management responsibility, these costs might appear to the agency as a higher construction bid. Alternatively, a transit agency might segment a large construction proj- ect into multiple contracts and hire a third-party construction manager to be responsible for their coordination and integration. The division of management labor between agency staff, management contractor, and construction contractor can differ depending on the sponsor agency. Figure 8 and Figure 9 show that sponsors typically estimate around 2–4% of construction costs for insurance and legal soft costs, and another 1–2% for the cost of surveys, testing, and other costs. Similar to administration and management costs, however, these types of costs, particularly Soft Cost Estimation: State of the Practice 23 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 1* 2* 3 4 6 5 7 8 9 10 Questionnaire Respondent FD C os t E st im at e (% of Co ns tru ct io n) Upper Bound Midpoint Lower Bound * Includes PE Figure 6. Final design soft cost estimates. 0% 5% 10% 15% 20% 25% 30% 1 2 3 4 5 6 7 8 9 10 Questionnaire Respondent Pj t M gm t + C on st. A dm in. C os t Es tim at e (% of C on str uc tio n) Upper Bound Midpoint Lower Bound Figure 7. Project management and construction administration soft cost estimates.

insurance, can depend on the practice of the agency and local circumstances, and it can be diffi- cult to characterize industry-wide estimation patterns for these cost categories. Sponsors appear to estimate startup costs quite differently, with estimates ranging from 0% to 7% for this category, as Figure 10 shows. Note the wide range given by respondents 1 and 3, further supporting this uncertainty. When viewed as individual components or groups of components, as Figures 5 through 10 show, some estimators use fairly consistent soft cost percentages, while others vary more widely. However, some of the differences at the component level may be somewhat offset at the aggre- gate level. Figure 11, therefore, shows the sum of all soft cost components for each questionnaire response. The stacked bars represent midpoint estimates, while the error bars show the sum of the range of all elements. The midpoints of each soft cost component sum to approximately 25–35% of construction costs fairly consistently, even though the individual soft cost compo- nents may differ somewhat from respondent to respondent. 24 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 0% 1% 2% 3% 4% 5% 6% 7% 8% 1 2 3 4 5 6 7 8 9 10 Questionnaire Respondent In su ra nc e + Le ga l C os t E st im at e (% of C on str uc tio n) Upper Bound Midpoint Lower Bound Figure 8. Insurance and legal soft cost estimates. 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 4.5% Questionnaire Respondent Su rv ey s, e tc . C os t E st im at e (% of Co ns tru ct io n) Upper Bound Midpoint Lower Bound 1 2 3 4 5 6 7 8 9 10 Figure 9. Surveys and other soft cost estimates.

3.4. Questionnaire Results: Drivers Identified Cost estimators were asked in an open-ended format to identify the kinds of project charac- teristics or circumstances that would ultimately impact their choice of percentages and that have impacted soft costs for past projects. The questionnaire suggested several attributes, but estima- tors were free to make their own responses as well. For each soft cost component (following the FTA SCC structure), estimators were requested to identify “cost drivers” that would have high, moderate, or minimal/no impact on soft costs in percentage terms. The results of this part of the questionnaire are presented in Table 8. Respondents generally identified a wide variety of soft cost drivers, and this research uses these as a starting point for its historical analysis presented later. Some drivers relate to the physical Soft Cost Estimation: State of the Practice 25 0% 1% 2% 3% 4% 5% 6% 7% Questionnaire Respondent St ar tu p Co st E st im at e (% of Co ns tru ct io n) Upper Bound Midpoint Lower Bound 1 2 3 4 5 6 7 8 9 10 Figure 10. Startup soft cost estimates. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1* 2* 3 6 7 4 8 9 5 10 M id -R an ge E st im at ed S of t C os t (% of C on str uc tio n) Questionnaire Respondents Other Insurance + Legal Project Mgmt. and Construction Admin. FD PE * Respondents estimate PE + FD as combined amount; PE displayed here using average split Figure 11. Midpoint soft cost estimates for all components.

Cost QUESTIONNAIRE RESPONDENT Impact 1 and 2 3, 4, and 5 6 10 High Alignment Grade; City v. Rural Alignment Grade; City v. Rural Alignment Profile Alignment Grade, City v. Rural Moderate Vehicle Quantity; Design Speed Project Delivery; Mode Quantity and Type of Stations Vehicle Quantity, Design Speed None/ Minimal Mode Peak Throughput Procurement Strategy Mode High Tunnel and Aerial Guideway; Quantity of Stations; Mode Completeness of P.E.; City v. Rural Community Outreach Alignment Grade, City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation Deviation from P.E. Decisions; Alignment Grade Value Engineering Vehicle Quantity, Design Speed None/ Minimal Design Speed; Grade; Peak Period Throughput Peak Throughput Budget Mode High Tunnel and Aerial Guideway; Quantity of Stations; Mode Ability and Experience of Contractor City v. Rural Vehicle Qty., Design Speed, Stations per LF; City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation Alignment Grade Special Design Skills Project Delivery Method None/ Minimal Design Speed; Grade; Peak Period Throughput Peak Throughput Available Engineering Pool --- High Tunnel and Aerial Guideway; Quantity of Stations; Mode Ability and Experience of Contractor Available Resources City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation Alignment Grade Available Skills Alignment Grade Design Speed; Grade; Peak Period Throughput Peak Throughput Avoid Owner / Contractor Duplication --- High Tunnel and Aerial Guideway; Quantity of Stations; Mode Market Forces Risk Assessment City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation Owner's experience; Brownfield v. Greenfield; City v. Rural Risk Assessment Alignment Grade None/ Minimal None/ Minimal None/ Minimal Design Speed; Grade; Peak Period Throughput Project Delivery Method Safety Record --- High Tunnel and Aerial Guideway; Quantity of Stations; Mode Brownfield v. Greenfield Requirements Identification City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation City v. Rural Schedule Station Density Design Speed; Grade; Peak Period Throughput Vehicles; Design Speed Agency Coordination Brownfield v. Greenfield High Tunnel and Aerial Guideway; Quantity of Stations; Mode Elevated or Tunnel Necessary Balance of Requirements City v. Rural Moderate City v. Rural; Project Delivery Method; Mitigation Vehicles; Design Speed; Mode Avoid Duplication Brownfield v. Greenfield None/ Minimal Design Speed; Grade; Peak Period Throughput --- Share Historical Information Alignment Grade High Tunnel and Aerial Guideway; Quantity of Stations; Mode New Line v. Extension Operation Coordination Vehicle Quantity, Design Speed Moderate City v. Rural; Project Delivery Method; Mitigation Elevated or Tunnel; Design Speed Skill Level Available Station Density None/ Minimal Design Speed; Grade; Peak Period Throughput Vehicles Schedule and Warranty Issues --- 80.03 80.04 Construction Administration & Management Preliminary Engineering80.01 80.02 Final Design SCC 80.07 Surveys, Testing, Investigation, Inspection Start up80.08 Insurance 80.05 Legal; Permits; Review Fees by other agencies, cities, etc. 80.06 Project Management for Design and Construction Table 8. Soft cost drivers identified by questionnaire respondents.

characteristics of the project, the setting and circumstances in which the project is built, the skills and experience of the sponsor and its contractors, and mitigation and unexpected issues. Look- ing ahead to how these drivers might be used to estimate future soft costs, some of these drivers are relatively straightforward to predict (e.g., alignment grade), while others are much more dif- ficult to foresee (e.g., agency coordination). 3.5. Questionnaire Results: Impact of Drivers Finally, cost estimators were asked to quantify the impact of 11 project characteristics on soft costs within the following scenario: • First, consider 7 project attributes that were designed to reflect increasing technical complexity; • Second, consider 4 additional attributes highlighting different institutional arrangements between the public sponsor and private contractor; • Third, consider a hypothetical base-case project: a simple light rail construction project, fully at grade, using an existing right-of-way, and delivered with a traditional design–bid–build method; and • Fourth, consider changes from the base case and report whether the soft cost estimate for each soft cost element would go up or down in percentage terms, using a scale of from 1 to 5, 1 mean- ing “significant reduction,” 3 meaning “no impact,” and 5 meaning “significant increase.” To help visualize patterns in the data, the color scheme presented in Figure 12 was applied to the responses. Table 9 shows the impact of mode on soft cost estimates, using light rail as the base case. Many respondents did not give information here or the response was not complete, perhaps because they lacked historical experience to respond. However, the table shows that, relative to light rail, estimators generally estimate higher soft costs for heavy rail projects, and only moderately higher for commuter rail projects. The results for BRT are mixed; one respondent predicted higher costs in some areas but lower in others, while another respondent predicted lower costs generally. However, these two questionnaire respondents should be interpreted within the context of their sample size. Cost estimators generally reported that higher project complexity, as measured by a number of indicators in Table 10 below, will tend to increase soft cost expenditures. Most respondents noted that an elevated alignment increases soft costs only moderately compared to at grade, but that tunneling tends to increase soft costs more significantly. Respondent 10, however, noted that soft costs might decline in some categories when tunneling. Estimators at all agencies sur- veyed predicted rising costs, especially in design and construction management, when subsur- face conditions differ from original plans. Results were mixed on the creation of a new right-of- way (versus the base-case existing ROW): some respondents foresaw no change, others predicted uneven increases, and others predicted significant increases. The final three project attributes included in the questionnaire describe alternative project delivery methods, which generally intend to shift risk from the public agency to the private con- tractor. Table 11 shows that cost estimators generally estimate that soft costs to the transit agency will go down as more risk is borne by the constructor. However, it is unclear whether this pat- tern describes a real reduction in costs or merely a shifting of soft costs out of the transit agency’s view and into a different cost category. Contractors bidding on a design–build contract, for exam- ple, might build soft costs into their bid. Soft Cost Estimation: State of the Practice 27

28 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Project Characteristic Change from Base Case SCC SCC Description 1, 2 3, 4, 5 6 10 80.01 Preliminary Engineering 5 N/A 4 N/A 80.02 Final Design 5 N/A 4 N/A 80.03 Project Management for Design and Constructio 5 N/A 4 N/A 80.04 Construction Administration & Management 5 N/A 4 N/A 80.05 Insurance 5 N/A 5 N/A 80.06 Legal; Permits; Review Fees by other agencies, 5 N/A 5 N/A 80.07 Surveys, Testing, Investigation, Inspection 5 N/A 5 N/A 80.08 Start up 5 N/A 3 N/A 80.01 Preliminary Engineering 4 N/A 4 N/A 80.02 Final Design 4 N/A 3 N/A 80.03 Project Management for Design and Constructio 4 N/A 3 N/A 80.04 Construction Administration & Management 4 N/A 3 N/A 80.05 Insurance 4 N/A 3 N/A 80.06 Legal; Permits; Review Fees by other agencies, 4 N/A 3 N/A 80.07 Surveys, Testing, Investigation, Inspection 4 N/A 3 N/A 80.08 Start up 4 N/A 3 N/A 80.01 Preliminary Engineering 2 N/A 4 N/A 80.02 Final Design 2 N/A 5 N/A 80.03 Project Management for Design and Constructio 2 N/A 4 N/A 80.04 Construction Administration & Management 2 N/A 4 N/A 80.05 Insurance 2 N/A 3 N/A 80.06 Legal; Permits; Review Fees by other agencies, 2 N/A 5 N/A 80.07 Surveys, Testing, Investigation, Inspection 2 N/A 4 N/A 80.08 Start up 2 N/A 3 N/A Questionnaire Respondent 1, 2 Mode: Heavy Rail Mode: Commuter Rail Mode: Bus Rapid Transit Notes: Base case is light rail. 1 Respondents 3, 4, 5, and 10 provided partial responses due to lack of experience; lack of response is noted as “N/A.” 2 Respondents 7, 8, and 9 did not provide responses and are omitted. Table 9. Impact of mode on soft cost estimate. No No Historical Impact Experience 1 2 3 4 5 or N/A LOWER COSTS HIGHER COSTS Significant Negative Moderate Negative Moderate Positive Significant Positive Figure 12. Questionnaire measurement system to quantify impact of cost drivers.

Soft Cost Estimation: State of the Practice 29 Project Characteristic Change from Base Case SCC SCC Description 1, 2 3, 4, 5 6 10 80.01 Preliminary Engineering 4 4 4 4 80.02 Final Design 4 4 5 4 80.03 Project Management for Design and Construction 4 3 4 3 80.04 Construction Administration & Management 4 3 4 4 80.05 Insurance 4 4 3 3 80.06 Legal; Permits; Review Fees by other agencies, 4 3 5 4 80.07 Surveys, Testing, Investigation, Inspection 4 4 3 4 80.08 Start up 4 3 3 3 80.01 Preliminary Engineering 5 3 5 3 80.02 Final Design 5 4 5 4 80.03 Project Management for Design and Construction 5 4 4 N/A 80.04 Construction Administration & Management 5 4 4 N/A 80.05 Insurance 5 4 5 4 80.06 Legal; Permits; Review Fees by other agencies, 5 3 5 2 80.07 Surveys, Testing, Investigation, Inspection 5 4 5 2 80.08 Start up 5 4 5 3 80.01 Preliminary Engineering 5 3 4 4 80.02 Final Design 5 5 5 4 80.03 Project Management for Design and Construction 5 4 4 5 80.04 Construction Administration & Management 5 4 4 5 80.05 Insurance 5 4 4 3 80.06 Legal; Permits; Review Fees by other agencies, 5 4 4 5 80.07 Surveys, Testing, Investigation, Inspection 5 4 5 5 80.08 Start up 5 3 3 3 80.01 Preliminary Engineering 5 3 5 5 80.02 Final Design 5 3 5 5 80.03 Project Management for Design and Construction 5 3 3 5 80.04 Construction Administration & Management 5 3 3 5 80.05 Insurance 5 3 3 4 80.06 Legal; Permits; Review Fees by other agencies, 5 3 5 5 80.07 Surveys, Testing, Investigation, Inspection 5 3 5 5 80.08 Start up 5 3 3 5 Questionnaire Respondent1, 2 New Right-of-Way Alignment: Elevated Alignment: Tunnel Differing Subsurface Conditions Notes: 1 Respondents 3, 4, 5, and 10 provided partial responses due to lack of experience; lack of response is noted as “N/A.” 2 Respondents 7, 8, and 9 did not provide responses and are omitted. Table 10. Impact of project complexity on soft cost estimate.

30 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Project Characteristic Change from Base Case SCC SCC Description 1, 2 3, 4, 5 6 10 80.01 Preliminary Engineering 3 3 4 3 80.02 Final Design 3 3 5 4 80.03 Project Management for Design and Construction 3 3 4 4 80.04 Construction Administration & Management 3 3 5 4 80.05 Insurance 3 3 3 3 80.06 Legal; Permits; Review Fees by other agencies, 3 3 4 4 80.07 Surveys, Testing, Investigation, Inspection 3 3 5 4 80.08 Start up 3 3 3 3 80.01 Preliminary Engineering 3 3 5 3 80.02 Final Design 3 2 3 4 80.03 Project Management for Design and Construction 1 2 5 2 80.04 Construction Administration & Management 2 2 3 2 80.05 Insurance 2 3 3 3 80.06 Legal; Permits; Review Fees by other agencies, 2 3 4 2 80.07 Surveys, Testing, Investigation, Inspection 2 2 4 2 80.08 Start up 3 3 4 3 80.01 Preliminary Engineering 3 N/A N/A 3 80.02 Final Design 3 N/A N/A 4 80.03 Project Management for Design and Construction 1 N/A N/A 2 80.04 Construction Administration & Management 2 N/A N/A 2 80.05 Insurance 2 N/A N/A 3 80.06 Legal; Permits; Review Fees by other agencies, 2 N/A N/A 2 80.07 Surveys, Testing, Investigation, Inspection 2 N/A N/A 2 80.08 Start up 2 N/A N/A 2 80.01 Preliminary Engineering 3 N/A N/A N/A 80.02 Final Design 3 N/A N/A N/A 80.03 Project Management for Design and Construction 1 N/A N/A N/A 80.04 Construction Administration & Management 2 N/A N/A N/A 80.05 Insurance 2 N/A N/A N/A 80.06 Legal; Permits; Review Fees by other agencies, 2 N/A N/A N/A 80.07 Surveys, Testing, Investigation, Inspection 2 N/A N/A N/A 80.08 Start up 2 N/A N/A N/A Questionnaire Respondent1, 2 Procurement: Full Turnkey Procurement: Design- Bid-Build (DBB) Procurement: Design- Build (DB) Procurement: Design- Build-Operate-Maintain (DBOM) Notes: 1 Respondents 3, 4, 5, and 10 provided partial responses due to lack of experience; lack of response is noted as “N/A.” 2 Respondents 7, 8, and 9 did not provide responses and are omitted. Table 11. Impact of project delivery method on soft cost estimate.

31 This report has thus far summarized efforts to assess the practice of soft cost estimation, as revealed through interviews and a questionnaire of cost estimators. To complement this research, this section examines actual soft cost expenditures from past construction projects. This as-built analysis also assesses the relationship between characteristics of transit infrastructure projects and actual soft cost expenditures for as-built projects. 4.1. Approach This analysis has three major objectives: • Describe the magnitude and range of soft cost expenditures in previous projects; • Analyze the relationship between these soft costs and other project characteristics as cost drivers, such as project complexity, mode, year, size, delivery method, and economic conditions; and • Form the ultimate basis of a new historically based methodology to estimate soft costs for future rail transit construction. 4.2. Data Source: FTA Capital Cost Database To examine historical costs, this analysis used as-built cost data and characteristics on 59 urban rail transit projects constructed over the past four decades in the United States. This cost data has been adapted from the capital cost databases developed for the FTA. In addition to this dataset, this study also relied on project schedule data adapted from the final report of TCRP Project G-07, Managing Capital Costs of Major Federally Funded Public Transportation Projects (Booz Allen Hamilton Inc., 2005), and developed some additional data on project char- acteristics such as public involvement, installation conditions, and sponsor agency capitalization policies. 4.2.1. About the Projects Included The projects included in this database were constructed by transit agencies in major urban centers and distributed throughout the various geographic regions across the United States. Over the period of 1984 through 2008, 29 light rail projects were constructed, and 30 heavy rail projects date from 1974 through 2005. This project cost database includes the costs of 59 proj- ects of various sizes, ranging from $100 million to over $2 billion, and represents new rail line segments, extensions of existing networks, and several rehabilitation and replacement proj- ects. This wide range of rail projects provides a good distribution of projects to examine the soft cost requirements needed in their development and offers a reasonable representation of C H A P T E R 4 As-Built Soft Cost Analysis

32 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects the requirements for professional services and soft costs for passenger rail construction in the United States. 4.2.2. About the Cost Data Format All project expenditures are reported in standardized formats for individual light and heavy rail project segments. The light rail database reports as-built costs in the same format as the Fed- eral Transit Administration’s current SCCs, while the heavy rail data is reported using a prior SCC format. Both formats use common element definitions and consistent structures to docu- ment the as-built costs of these passenger rail projects. Costs were adjusted to an average of the 38 largest U.S. metropolitan areas and then escalated to a common base year of 2008 using Means Construction Cost Index (Murphy, 2008) for this consistent dollar value. The cost categories for these two datasets are listed below in Table 12 using the present FTA SCC category format in order at left and the corresponding heavy rail categories at right. Most capital cost categories examined in this section are comparable between the two data structures, with minor exceptions. In addition, this analysis took several steps to prepare and standardize the cost data: 1. All dollar costs were inflated to constant 2008 dollars; 2. All dollar costs were adjusted for local–national cost variations using the Means Construction Cost Index (Murphy, 2008); and 3. Outlier data points were eliminated. The details of these adjustments can be found in Appendix C. 4.2.3. Project Development Schedule Database The project development schedules used in this analysis have been adapted from the results of the contractor’s final report from TCRP Project G-07 entitled Managing Capital Costs of Major Federally Funded Public Transportation Projects (Booz Allen Hamilton Inc., 2005). The G-07 report examined the various strategies, tools, and techniques available to better manage major transit capital projects and developed another separate project development schedule database to examine project schedule delays and their impacts on project costs. The evaluation of soft costs relies on Project G-07’s schedule database to measure the relationship between project schedule and soft costs incurred. 4.2.4. Drivers Tested Table 13 presents the non-financial data items that are tested here as potential cost drivers for actual soft cost expenditures. Some of these results are shown in Appendix C. Light Rail Heavy Rail 10 Guideway and Track Elements 1.00 Guideway Elements 20 Stations, Stops, Terminals, Intermodal 4.00 Stations 30 Support Facilities: Yards, Shops, Admin. 2.00 Yards and Shops 40 Sitework and Special Conditions 6.00 Special Conditions 50 Systems (Signals, Power, Communications) 3.00 Systems 60 ROW, Land, Existing Improvements 7.00 Right-olf-Way 70 Vehicles 5.00 Vehicles 80 Professional Services (Soft Costs) 8.00 Soft Costs 90 Unallocated Contingency 100 Finance Charges Table 12. Light and heavy rail capital cost categories correspondence table.

As-Built Soft Cost Analysis 33 4.3. Potential Issues in Soft Cost Categorization As described in Chapter 2, this project considers soft costs to be equivalent to professional ser- vices as defined in FTA’s Standard Cost Category 80, Professional Services, in the Standard Cost Category Workbook (U.S. FTA, 2008). Refer to Section 1.3 for a definition of soft costs. According to the FTA definitions, however, other SCC categories besides Category 80 may contain expenditures that may be very similar to soft costs. This analysis has addressed these cases as follows: • Construction costs (Categories 10 through 50) contain some indirect costs that could be con- sidered soft costs, such as project and construction supervision, general conditions, contrac- tor’s general liability, insurance, overhead, and profit, plus comparable subcontractors’ costs. Because these soft costs are more associated with direct construction functions, they are treated as hard construction costs. • ROW, Land, Existing Improvements (Category 60) may include professional services associated with the real estate component of the project such as agency staff oversight and administration, real estate and relocation consultants, assessors, legal counsel, court expenses, and insurance. These costs have been considered separately in this analysis. • The Vehicles category (Category 70) includes supporting services associated with the vehicle procurement aspect of the project. These costs may include agency staff oversight and admin- istration, vehicle consultants, design and manufacturing contractors, legal counsel, and warranty and insurance costs that, like real estate soft costs, have been considered separately in this analysis. • Unallocated Contingency (Category 90) includes some costs that could depend on other costs. These costs are essential to cost estimates in earlier project phases, but by the completion of the project, these costs are zero in the as-built cost. This cost category was therefore excluded from this soft cost analysis. • Finance Charges (Category 100) contains costs that could be considered soft costs. These financing charges have been excluded from this analysis of soft costs because these costs are more project specific and depend on the availability of funding. They have more in common with the financing plan than the overall project development process. Mode Guideway length (linear feet) Percentage of guideway below grade Percentage of guideway not at grade Percentage of guideway at grade Percentage of guideway at grade (incl. built-up fill and retained cut) New line/extension of existing line/rehabilitation Procurement or delivery method (design–build, etc.) Midpoint of expenditures (year) Planning/draft environmental impact statement (DEIS) midpoint (year) Preliminary engineering/final environmental impact statement (FEIS) midpoint (year) Final design midpoint (year) Construction midpoint (year) Revenue service begins (year) Number of stations Total project cost estimated at preliminary engineering Total project cost – actual Economic conditions at estimated bid date (U.S. GDP growth) Experience level of sponsor Installation conditions (active service, no active service, etc.) Public or political involvement Use of contractors in management or development Unusual delays in project planning phases Agency tendency to minimize capital charges Table 13. Project characteristics tested as cost drivers.

34 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects This analysis uses terminology that implies certain groupings of cost categories from the two datasets for light and heavy rail. The numerical definition of these groupings is presented in Table 14. Project year in this analysis means the midpoint of expenditures, derived as the average year of expenditure for each individual cost element, weighted by expenditure amount. Other schedule years used in the analysis to denote project phases (e.g., preliminary engineering, design, construction and operations) mean the midpoint of that phase within the project schedule. This analysis relies on FTA’s prior categorization of costs for projects constructed prior to the current SCC structure and clarifying guidance. Therefore, users of this analysis must be mindful of potentially inconsistent classification of costs within the data. This is because of a number of possible reasons, including: • Inconsistent reporting across agencies – At a basic level, some judgment is required to classify specific expenditures within the SCC structure, even with the available guidance from FTA. Broad categories such as the demar- cation between vehicle and systems costs are likely to be more consistently comparable among the reporting agencies, while detailed cost items such as the difference between “Project Management for Design and Construction” and “Construction Administration and Management” are likely more susceptible to inconsistencies in reporting definitions. Since this dataset includes projects from across the country and across decades of construc- tion, the data may be susceptible to some level of inconsistent definitions of cost categories. – Of particular relevance to this study is the reporting of professional service soft costs for vehi- cles and rights-of-way. These were initially reported into a database structure that was unclear about some of these related vehicle and right-of-way soft costs. More recent data- base structure and instructional guidance expressly defines the cost elements for vehicle and right-of-way soft costs. These more category-specific soft costs have been segmented from this analysis of construction-related soft cost. • Refinements to the cost structure – The structure of FTA’s SCC capital cost database has evolved over the past 20 years. This evolving framework for the cost data and varying levels of detail directly and indirectly affect some of the more detailed reporting and thereby the resulting relationships. – The FTA Standard Cost Categories have clarified the right-of-way and vehicle cost cate- gories to include those categories related to soft costs. However, prior structures may not have been as clear and vehicle and right-of-way associated soft costs could be mixed into the general soft cost category. • Agency capital program policies – The financial and administrative policies of the sponsoring agency can affect how soft costs are reported for a capital project, which could affect the amount and proportion of soft costs Term Used Here Light Rail Cost Categories Applied from Table 12 Heavy Rail Cost Categories Applied from Table 12 Soft costs as % of total costs [80] ÷ ([10] + [20] + [30] + [40] + [50] + [60] + [70]) [8] ÷ ([1] + [2] + [3] + [4] + [5] + [6] + [7]) Soft costs as % of construction costs [80] ÷ ([10] + [20] + [30] + [40] + [50]) [8] ÷ ([1] + [2] + [3] + [4] + [6]) Vehicle costs [70] [5] ROW costs [60] [7] Engineering soft costs [80.010] + [80.020] [8.02] + [8.03] Management soft costs [80.030] + [80.040] [8.03] + [8.04] + [8.05] + [8.06] Table 14. Capital cost definitions of soft cost analysis terms.

when comparing projects across agencies. For example, staff and contractor soft cost charges can be funded through separate grants and are not always reported into the project budget. – The salaries of some agency staff who support engineering, design, and/or construction may be treated as an operating expense rather than charged to the capital project. – Early planning and preliminary engineering costs may be charged to a general planning grant rather than attributed directly to the capital project. – Insurance may be carried by the construction contractor or the sponsor agency, and/or it may be embedded into individual cost elements as an overhead cost. • Project delivery mechanism – The varying methods of project development and procurement present unique challenges to the breakdown and classification of project costs because cost classification can depend on institutional perspective. Sections 3.5 and 4.5.3 discuss this issue more thoroughly. 4.4. Historical Soft Costs This first portion of the soft cost analysis presents the general breakdown of project soft cost attributes within the as-built project cost database. Total project costs are described using the follow- ing categories: Soft, Vehicle, and Construction costs. Soft costs are then examined as a proportion of the Construction Costs category and then further examined by individual soft cost components. 4.4.1. Describing the Data As shown in Figure 13, construction costs made up the largest share of expenses for most proj- ects, vehicle costs range from 0 to 25% of total project cost, ROW costs 0 to 20%, and soft costs 10 to 35%. While all projects incurred construction and soft costs, some projects had no ROW or vehicle procurement costs. For example, the extension of Bay Area Rapid Transit (BART) to San Francisco International Airport (SFO) required the purchase of no additional vehicles, while the extension of the CTA’s Blue Line to O’Hare Airport did not entail right-of-way costs. Pro- fessional services for the many varied rail transit capital projects in this database usually accounted for around 10–35% of total project costs. This pattern forms the focus of the more detailed segmentation of these costs, presented briefly here and in more detail in Appendix C. Figure 13 illustrates soft costs, with light bars at the top, expressed as a percentage of total costs. To measure soft costs in a more commonly used format, Figure 14 shows soft costs as a percentage As-Built Soft Cost Analysis 35 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% % o f T ot al P ro jec t C os ts Construction Costs ROW Costs Vehicle Costs Soft Costs Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 13. Project costs by category.

of construction costs for these same projects. Construction costs include all of the guideway, trackwork, facility, station, systems, sitework, and special conditions costs (refer to Table 14). When expressed as a percentage of construction costs, soft costs vary considerably more across these same projects than when expressed as a percentage of the total cost—from 11% to a high of 54% of construction costs. Expressing soft costs as a percentage of construction costs is pertinent to this analysis since soft costs associated with the vehicle and right-of-way costs are expressly defined as a separate cost element in each of those associated cost categories. This relatively wide range in soft costs as a percentage of construction costs merits further examination. Note that in Figures 14 through 18, 20, and 22, the historical projects are ordered in terms of increasing soft costs as a percentage of construction costs, with separate ordering for light rail and heavy rail projects. To explore the wide range in this soft cost measure, the individual cost components that com- pose total soft costs were analyzed. Total soft costs can be segmented into six major components, as defined in the FTA SCC structure: • Preliminary Engineering, • Final Design, • Project Management for Design and Construction, • Construction Administration and Management, • Insurance, and • All Other Soft Costs in SCC 80. These six soft cost components are shown as a percentage of construction costs in the bar chart in Figure 15. The total percentages are consistent with those presented above in Figure 14. The six components are expressed as a percentage of overall soft costs in Figure 16, where the bar chart for each project totals 100%. The components of soft costs appear to vary considerably across projects, especially as a pro- portion of overall soft costs. For example, preliminary engineering costs (bottom measure and dark aqua in Figure 15 and Figure 16) are a very small or near-zero proportion of soft costs for some projects, while for others (e.g., Hudson-Bergen Phase 1, Phoenix) these costs are signifi- cant expenditures. In projects with little or no reported preliminary engineering costs, there was likely either a missing expenditure or it was rolled into a combined grant with another soft cost component. Insurance can account for almost 10% of construction costs (e.g., CTA Douglas Branch) for some projects, or none at all for others. This may be due to different agencies’ 36 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 0% 10% 20% 30% 40% 50% 60% So ft Co st s (% of C on str uc tio n) Light Rail Heavy Rail Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 14. Soft costs percent of construction costs by project and mode.

approaches to project development, where one agency may provide project-wide wrap-up insur- ance and others may require each contractor to provide their own insurance, or some combination of these. In general, individual variances may be due to real differences in expenses incurred as a result of project characteristics, while some variation is probably due to the way in which costs are reported or categorized. The more consistent soft cost components were final design, project management, and construction management. Some projects appear to have inconsistencies in the reported soft cost experience that may indicate questionable data. For example, some projects show zero engineering or design costs, which is unlikely given the complexity of constructing major transit capital projects. In these cases, expenditures may have been classified elsewhere in the SCC structure or charged to a sep- arate, off-project funding source and not reported into the project budget. In subsequent analysis in this report, certain outliers were omitted from the more detailed analyses to eliminate the effect of these uncertain data. The decision to remove an outlier was based on analyzing the distribution of projects’ soft costs, and is more fully described in Section C.4 in Appendix C. Figure 17 shows the average soft cost percentages by component for all projects in the dataset (outliers excluded) and the range of percentages encountered. The bars represent average soft-cost As-Built Soft Cost Analysis 37 0% 10% 20% 30% 40% 50% 60% So ft Co st s (% of C on str uc tio n C os ts) PE FD Project Mgmt. for D&C Construction Admin. & Mgmt. Insurance All Other Soft Costs Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 15. Soft cost components as a percentage of construction costs. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% So ft Co st s (% of To tal S of t C os ts) PE FD Project Mgmt. for D&C Construction Admin. & Mgmt. Insurance All Other Soft Costs Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 16. Soft cost components as percentage of total soft costs.

component expenditures, and the lines indicate the maximum and minimum values in the dataset. For instance, the average project incurred final design expenses of 9.7% of construction costs, but this percentage ranged as low as 2.6% for one project and as high as 31.0% for another. Most cat- egories contained projects with zero expenditures for that category, resulting in the minimum of the range being zero. Figure 17 also shows that when all components are combined, projects show average soft costs of around 31% of construction costs. However, the range of total soft costs has been as low as 11.4% for one project and as high as 53.6% for another project, after excluding outliers. To test the hypothesis that soft-cost component costs may have been inadvertently assigned and reported to a related soft cost component, the analysis grouped some related soft cost com- ponents and subtotaled them into the following three soft-cost component categories: • Pre-construction costs (design and engineering), • Construction expenditures (construction management, administration, etc.), and • Other costs (insurance, others). Although an approximation of these project development phases, this broad categorization produces the results displayed in Figure 18 (as a percentage of construction costs) and Figure 20 (as a percentage of total soft costs). A more consistent soft cost basis appears to emerge from the analysis when soft cost compo- nents are grouped by these categories, which approximates the project development phase in which the expenditures were incurred. Figure 19 shows the averages and ranges of these three groups of soft cost components, expressed as a percentage of construction costs. This figure indicates that a typical project incurs preliminary engineering and final design costs of 12.4% of construction, and construction management and project administration soft costs of 15.1% of construction, but that these percentages can range from around 3% to 33% for some projects. When expressed as a percentage of total soft costs as shown in Figure 20, the resulting cost proportions are more consistent. About 40–50% of soft costs are generally related to engineer- ing and final design, another 40–50% of soft costs are related to construction management and administration, and about 10% are other costs. The first two of these three categories (engineer- ing and final design, and construction management and administration) are sometimes used in subsequent analysis in this report. 38 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Soft Costs as % of Construction Average 2.7% 9.7% 8.8% 6.3% 1.6% 2.2% Minimum 0.0% 2.6% 0.0% 0.0% 0.0% 0.0% Maximum 8.3% 31.0% 23.2% 19.4% 9.4% 10.5% PE Proj. Mgmt. for D&C FD All OtherSoft CostsInsurance Construction Admin. & Mgmt. 31.3% 11.4% 53.6% 0% 5% 10% 15% 20% 25% 30% 35% So ft Co st s (% of C on str uc tio n) Maximum Minimum 0% 10% 20% 30% 40% 50% 60% All Components So ft Co st s (% of C on str uc tio n) Figure 17. Average and range of soft cost components as percent of construction.

As-Built Soft Cost Analysis 39 Soft Costs as % of Construction Average 12.4% 15.1% 3.8% 31.3% Minimum 3.6% 3.2% 0.0% 11.4% Maximum 31.0% 33.2% 16.0% 53.6% 0% 5% 10% 15% 20% 25% 30% 35% Design (PE + FD) Admin. & Management Other Soft Costs So ft Co st s (% of C on str uc tio n) Maximum Minimum 0% 10% 20% 30% 40% 50% 60% All Components So ft Co st s (% of C on str uc tio n) Figure 19. Average and range of subtotaled soft cost components as a percentage of construction. 0% 20% 40% 60% 80% 100% So ft Co st s (% of To tal S of t C os ts) Preliminary Engineering & Final Design Project and Construction Management All Other Soft Costs Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 20. Subtotaled soft cost components as a percentage of total soft costs. 0% 10% 20% 30% 40% 50% 60% So ft Co st s (% of C on str uc tio n) Preliminary Engineering + Final Design Project and Construction Management All Other Soft Costs Hu d- Be rg II Hi aw ath a Sa lt L ak e Sa cra m. I Hu d- Be rg I Po rtla nd S o Sa cra m. Fo l Po rtla nd In t Ch arl ott e LA G old E as t LA G old P as a Sa cra m. S o St . L ou is Po rtla nd W So uth N J Pi tts bu rgh N Po rtla nd S eg 1 Pi tts bu rgh I LA B lue Sa n D ieg o VT A Ta s E Pi tts bu rgh II De nv er S W VT A Ca pit ol Ph oe nix CT A O' Ha re DC U S t. DC A dd iso n DC L' En fan t DC N ew C a DC S ha dy G DC H un tgt n DC G len mt 1 CT A Do ug las NY CT 63 rd CT A Or an ge Ba ltim or e DC V ien na MB TA O ran g CT A Br ow n MA RT A N- S Mi am i Sa n J ua n DC A na co st O DC A na co st DC S pri ng fld DC G len mt 2 DC G ree nb lt BA RT S FO Ph il F ra nk f. NY CT S till w Figure 18. Subtotaled soft cost components as a percentage of construction costs.

4.4.2. Measuring Soft Costs Developing a guidebook on the estimation of soft costs requires the identification of specific measures. This section tests a number of different ways to measure soft costs and explores how each may be used in a guidebook context. Soft costs of as-built projects can be measured in the following ways: • As a percentage of total project cost; • As a percentage of all other costs, excluding only soft costs; • As a percentage of construction costs; • In constant dollar value terms; or • In constant dollars per linear foot of constructed guideway. This analysis does not rely on the first and fourth measurements on this list. Figure 13 above showed soft costs as a percentage of total project cost, and this measurement is sometimes used to describe soft costs. However, measuring soft costs as a percentage of total project cost is not an appropriate metric for a cost estimator since the estimator does not know total project cost until the soft cost estimate is complete. Soft costs may also be expressed in dollar value terms, but this measure would fail to account for differences in project size across the dataset. Therefore this analysis focuses on measuring soft costs as a percentage of all other costs, as a percentage of construction costs, and in dollars per linear foot of guideway. Figure 21 compares measuring soft costs as a percentage of all other total costs (i.e., all other costs besides soft costs themselves) and as a percentage of construction costs (i.e., excluding vehicle and right-of-way costs) and shows that these two percentage-based methods of measure- ment are highly correlated. This applies to both light and heavy rail modes and the combined analysis of projects of both modes. These results suggest that ROW and Vehicle category costs (those that are excluded when measuring construction costs only) have a relatively small effect on soft costs. This may indicate that their related soft costs (ROW and Vehicle category costs) have been accurately accounted for within each of these categories. Measuring soft costs per linear foot is another way to measure soft costs. To test the quality of this measure, all project costs were normalized by applying the national average metropolitan area Means Construction Cost Index (Murphy, 2008) and then inflating to 2008 dollars. Signifi- cant project outliers were excluded from this analysis to focus on the more consistent results. Figure 22 shows this measurement for all included projects. 40 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.79 t-Stat = 9.4 R2 = 0.836 t-Stat = 11.07 t-Stat = 13.2 R2 = 0.84 0% 20% Soft Costs as % of All Other Costs R2 = 0.79 0% 10% 20% 30% 40% 50% 60% 70% 0% 10% 20% 30% 40% 50% 60% 70% 0% 20% 40% 60% 40% 60% 40% 60% Soft Costs as % of All Other Costs So ft Co st s (% of C on str uc tio n) So ft Co st s (% of C on str uc tio n) R2 = 0.781 0% 10% 20% 30% 40% 50% 60% 70% 0% 20% Soft Costs as % of All Other Costs So ft Co st s (% of C on str uc tio n) Figure 21. Soft cost percentage of construction costs versus soft cost percentage of total other costs.

Soft costs on a per-linear-foot basis vary considerably, even with the removal of outliers, from less than $1,000 to nearly $10,000 per linear foot (all costs in 2008 dollars). Specifically, light rail projects averaged $2,572 per linear foot, heavy rail $5,726, and all projects combined $4,044 per linear foot, as shown in Figure 23. The range for soft costs in light rail is somewhat less than for heavy rail projects. In general, soft costs tend to be higher for heavy rail, consistent with the generally higher cost of heavy rail overall. The soft cost per linear foot measure appeared to offer some consistency with the range estimates noted above. The next step in the analysis was to see if there was any relationship with the soft cost percentage of construction. Figure 24 compares the measurement of soft costs as a percentage of construction cost and as a dollar value cost per linear foot. As Figure 24 indicates, measuring soft costs as a dollar-value cost per linear foot versus a per- centage of construction cost would not yield similar results. The heavy rail projects have a some- what better relationship that may indicate a greater relationship of increasing complexity of the heavy rail projects with greater soft cost requirements. As-Built Soft Cost Analysis 41 $- $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 Sa cra m. I Sa lt L ak e Sa cra m. Fo l St . L ou is De nv er S W So uth N J Sa cra m. S o Hi aw ath a Po rtla nd S eg 1 Ch arl ott e Po rtla nd S o LA G old P as a Po rtla nd In t VT A Ca pit ol VT A Ta s E Ph oe nix LA B lue Pi tts bu rgh I Po rtla nd W Hu d- Be rg I Pi tts bu rgh II Sa n D ieg o Hu d- Be rg II LA G old E as t CT A O' Ha re CT A Or an ge DC G len mt 2 CT A Do ug las DC S pri ng fld DC G ree nb lt Mi am i DC A na co st DC A na co st O DC V ien na Ba ltim or e DC G len mt 1 MA RT A N- S DC U S t. DC A dd iso n DC H un tgt n DC S ha dy G DC N ew C a DC L' En fan t BA RT S FO Ph il F ra nk f. So ft Co st s pe r L in ea r F oo t o f G ui de w ay Light Rail Heavy Rail Figure 22. Soft costs per linear foot of constructed guideway by project and mode. Soft Costs per Linear Foot Average $ 2,572 5,726$ 4,044$ Minimum $ 335 1,191$ 335$ Maximum $ 6,201 9,728$ 9,728$ $- $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 Light Rail Heavy Rail Light & Heavy Rail So ft Co st s pe r L in ea r F oo t Figure 23. Average and range of soft costs per linear foot of constructed guideway.

It is unclear from these initial analyses which basic measurement of soft costs (percentage or dollar value terms) is most appropriate. Therefore this analysis shows results with both unless one measure appears more appropriate given the circumstances. 4.5. Relationships between Cost Drivers and Historical Soft Costs This section tests the relationship between various project characteristics such as mode, align- ment, and year (detailed above in Table 13) and actual soft cost expenditures. This research took two approaches to measuring how soft cost drivers have impacted actual soft costs: • Univariate testing of soft cost drivers suggested in interviews and the questionnaire. First, this research began by creating a series of scatter diagrams comparing soft costs with the kinds of project characteristics that estimators currently use to choose higher or lower soft cost per- centages. This kind of analysis tests only whether one project characteristic alone influences soft costs. As the results below demonstrate, some of these tests showed that soft costs are cor- related with certain project characteristics, while other tests yielded less conclusive results. Many of the less conclusive results are presented in Appendix C. These single-variable results served to guide the research into the next phase described in Section 4.5.7. • Multivariate testing of combinations of soft cost drivers. Second, this research tested a multitude of combinations of soft cost drivers and their effect on soft costs in a multivariate regression. Project characteristics were the independent variables, and soft costs as percent of construction costs acted as the dependent variable. After several hundred tests, a single multi- variate regression was developed that can explain approximately 60% of the differences in soft cost percentages by variations in project characteristics (R2 = 0.58), as will be described later. This kind of analysis tests the cumulative effect of how changes in a variety of project attri- butes have affected resulting soft costs. 4.5.1. Assembling Data on Soft Cost Drivers A set of characteristics was gathered for the projects to help identify cost relationships, including the following: • Physical attributes, such as alignment length, profile (e.g., below grade, at grade, aerial), number of stations, or whether the project initiated new service or extended an existing line. 42 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL $- $5 $10 $15 Soft Costs per LF (000) (2008$) R2 = 0.05 R2 = 0.00 R2 = 0.02 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $- $2 $4 $6 $8 Soft Costs per LF (000) (2008$) So ft Co st s (% of C on str uc tio n) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% So ft Co st s (% of C on str uc tio n) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% So ft Co st s (% of C on str uc tio n) $- $5 $10 $15 Soft Costs per LF (000) (2008$) HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.05 t-Stat = 1.07 R2 = 0.00 t-Stat = 0.09 R2 = 0.02 t-Stat = -0.81 Figure 24. Soft costs as a percentage of construction versus soft cost per linear foot of constructed guideway.

• Installation conditions, such as whether the project interacted with other active rail transit lines. • Schedule information, including major milestones in the project lifecycle for a subset of projects in the dataset. While each project had a midyear of expenditure, only some projects had full schedule data available. • Characteristics of the project sponsor, such as experience level, internal policies on capital costs, and use of outside contractors. • The context of the project development process, such as the level of public involvement, delivery method, or whether a significant redesign was necessary. For these last two types of characteristics, the definition and determination of values required some judgment based on knowledge of those projects’ development process. Many measures were derived from this primary dataset that were intended to act as a proxy to capture other project characteristics, such as project magnitude (e.g., construction costs per linear foot), complexity (e.g., percent of alignment below grade), unique circumstances (e.g., real estate acquisition costs, project occurred prior to certain federal requirements), and many others. 4.5.2. Soft Costs by Mode and Year Figure 25 shows the average soft costs as percentage of construction costs across modes and by decade. The amount spent on soft costs appears to vary little depending on mode, as indicated in the left pane. Light rail projects averaged 33.8%, heavy rail projects averaged 28.0%, and the combined database projects averaged 30.9% of soft cost percentage of construction. Soft costs have been rising over time since the 1970s. The right pane of Figure 25 shows that on average, soft costs for both heavy and light rail have recently amounted to approximately 34.6% of construction costs, and this figure is an increase from about 21.4% three decades ago. 4.5.3. Soft Costs by Project Delivery Method Project delivery method or procurement strategy also appears to affect expenditures on soft costs. Although most projects in the dataset were delivered via a DBB methodology, evidence for light rail projects indicates that DB projects have lower soft costs, as shown in Figure 26. With only nine design–build projects and one construction management/general contractor (CM/GC) project out of all database projects, these findings need to be considered within the limitations caused by the small sample size. As-Built Soft Cost Analysis 43 Sample Size: 25 26 51 6 14 10 21 33.0% 34.6% 21.4% 27.7% 0% 5% 10% 15% 20% 25% 30% 35% 40% 1970s 1990s 2000s So ft Co st s (% of C on str uc tio n C os ts) 33.8% 28.0% 30.9% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Light Rail Heavy Rail All Modes So ft Co st s (% of C on str uc tio n C os ts) 1980s Figure 25. Average soft costs by mode and by decade.

Projects selected for design–build delivery method may be chosen for their simplicity, how- ever, so care should be exercised when considering the above chart. An agency may choose to advertise for design–build projects that would incur low soft costs regardless of delivery method. The Hudson-Bergen project, for example, was delivered with a design–build contract, which may have contributed to lower soft costs. Alternatively, design–build contractors may classify soft costs in different ways than a public agency (e.g., in the construction line item), which might make soft costs appear lower. One of the problems with these delivery methods is that they are not yet very common in the United States, and transit agencies may not fully understand them. Some transit agencies may award a design–build or other alternative delivery contract but then continue to perform engi- neering work in a more traditional project delivery mode, unknowingly duplicating soft costs. 4.5.4. Soft Costs by Project Development Schedule Figure 27 shows the effect of pre-construction duration (from planning/DEIS to construction phases) on soft costs in dollar terms. Total soft costs are presented in the left pane, and engineer- 44 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL Sample Size: 27.3%29.9% 25.6% 36.3% 0% 5% 10% 15% 20% 25% 30% 35% 40% DBB DB CM/GC DBB DB CM/GC So ft Co st s (% of C on str uc tio n) 0% 5% 10% 15% 20% 25% 30% 35% 40% So ft Co st s (% of C on str uc tio n) 0% 5% 10% 15% 20% 25% 30% 35% 40% So ft Co st s (% of C on str uc tio n) 29.9% 25.6% 31.1% DBB DB CM/GC 16 8 1 22 0 0 38 8 1 Figure 26. Soft costs as a percentage of construction versus project delivery method. LIGHT + HEAVY RAIL: ALL SOFT COSTS Sample Size: 11 t-Stat =1.69 Sample Size: 11 R2 = 0.24 R2 = 0.24 t-Stat =1.58R2 = 0.22 $- $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $14,000 $16,000 $18,000 - 2 4 6 8 10 12 - 2 4 6 8 10 12 Years Elapsed between Planning/DEIS and Construction So ft Co st s pe r L F (20 08 $) R2 = 0.22 $- $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000 Years Elapsed between Planning/DEIS and Construction PE + F D C os ts p er L F (20 08 $) L IGHT + HEAVY RAIL: PE + FD COSTS ONLY Figure 27. Soft costs per linear foot versus years elapsed between completion of the draft environmental impact statement and construction.

ing costs (preliminary engineering and final design) are presented in the right pane. In the left pane the results are pronounced, from zero soft cost at 4 years to a maximum of about $16,000 per linear foot at about 15 years between the DEIS completion and construction. This relation- ship holds for engineering soft costs as well, as shown in the right pane. This finding seems to suggest that the duration of pre-construction phases should be consid- ered within the estimate of soft costs. However, the findings in Figure 27 may simply show that costly projects take longer to plan and design. The relatively small sample size (11) and the role of one relatively costly project in this chart should be recognized in a careful consideration of these findings. 4.5.5. Soft Costs by Project Complexity The remainder of the univariate analysis focuses on project characteristics that address com- plexity (such as percentage of guideway not at grade), number of stations, and other factors and the impact of these characteristics on soft costs. In general, indicators of complexity tend to cor- relate well with soft costs when measured in dollar terms per linear foot. Many of these relation- ships where soft costs are measured as a percentage of construction costs are presented in the appendices. The following figures compare soft cost percentages to the project’s alignment profile and typify many of the other results addressing project complexity. The alignment profile of new rail construction can substantially influence the technical com- plexity of the project. In the proposed hypothesis, as the proportion of guideway that is not at grade (in tunnels, on aerial structures, etc.) increases, complexity increases, and soft costs may increase likewise. In the first part of this analysis, “not at grade” is defined as an aerial structure, built-up fill, underground cut and cover, underground tunnel, or retained cut or fill guideway. Figure 28 shows little correlation between the proportion of alignment not at grade and soft costs as a percentage of construction costs. The light rail soft cost percentage is flat at about 40%, while heavy rail shows an increasing trend in the soft cost percentage from 25% to about 35%. The combined project database is flat at about 38%. The statistical trend line for all three relation- ships shows a very weak correlation (R2 less than 0.04), and all relationships are statistically insignificant. The issue of project complexity can be examined in another way, by measuring soft costs in terms of dollar per linear foot. Figure 29 expresses soft costs in dollars per linear foot and shows that soft costs indeed rise as greater portions of the alignment are not at grade. As-Built Soft Cost Analysis 45 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.04 t-Stat = 0.94 R2 = 0.04 t-Stat = 0.941 R2 = 0.01 t-Stat: -0.51 R2 = 0.04 0% 10% 20% 30% 40% 50% 60% 0% 25% 75% 100% 100%100% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) R2 = 0.04 0% 10% 20% 30% 40% 50% 60% 0% 25% 50% 50% 50%75% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 0% 25% 75% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) Figure 28. Soft costs as a percentage of construction versus percentage of guideway not at grade.

As Figure 29 shows, the relationship between the percent of guideway not at grade and soft costs per linear foot is statistically significant for light rail, and for both modes combined, but not for heavy rail alone. Indeed, the R2 value for light rail indicates that the proportion of guide- way not at grade can explain about half of the variation in soft costs per linear foot for heavy rail projects. As the not-at-grade percentage of the projects increase, the soft costs as measured in dollar value terms per linear foot increase. So while the dollar value of soft costs does measurably increase with project complexity as shown in Figure 29, the pattern is not significant enough to increase soft costs in percentage terms, as demonstrated in Figure 28. Although it is tempting to measure soft costs in dollar value terms because this measure pro- duces more correlation with complexity variables, it is worth exploring the measure further. One benefit of measuring soft costs in percentage terms is that the measure controls for variations in unit costs. Soft cost requirements of more expensive projects can be consistently compared to inexpensive projects in percentage terms. Measuring soft costs in dollars-per-linear-foot terms risks autocorrelation between unit costs—high soft costs could be correlated with higher other costs. In general, Figure 30 tends to confirm this hypothesis: in dollar terms, soft costs increase proportionately to construction costs. The correlations shown are strong and statistically signif- 46 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.35 t-Stat = 3.45 R2 = 0.01 t-Stat = 0.52 R2 = 0.36 t-Stat: 4.88 R2 = 0.35 $- $2 $4 $6 $8 $10 $12 0% 25% 75% % Guideway Not At-Grade So ft Co st s (00 0) pe r L ine ar Fo ot R2 = 0.01 $- $2 $4 $6 $8 $10 $12 0% 25% 75%50%50% 50%100%100% 100% % Guideway Not At-Grade So ft Co st s (00 0) pe r L ine ar Fo ot R2 = 0.36 $- $2 $4 $6 $8 $10 $12 0% 25% 75% % Guideway Not At-Grade So ft Co st s (00 0) pe r L ine ar Fo ot Figure 29. Soft costs per linear foot versus percentage of guideway not at grade. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.56 t-Stat = 5.31 R2 = 0.51 t-Stat = 4.487 R2 = 0.68 t-Stat: 9.48 R2 = 0.56 $100 $1,000 $10,000 $100,000 $1,000 $10,000 $100,000 Construction Cost (2008$) per Linear Foot So ft Co st s pe r L in ea r F oo t R2 = 0.51 $100 $1,000 $10,000 $100,000 $1,000 $10,000 $100,000 Construction Cost (2008$) per Linear Foot So ft Co st s pe r L in ea r F oo t R2 = 0.68 $100 $1,000 $10,000 $100,000 $1,000 $10,000 $100,000 Construction Cost (2008$) per Linear Foot So ft Co st s pe r L in ea r F oo t Figure 30. Soft costs per linear foot versus construction costs per linear foot on a logarithmic scale.

icant for both modes and the combined database. This trend may help explain why soft costs measured in percentage terms appear unrelated to many other variables like alignment profile— these other variables may simply drive up construction costs at the same rate. 4.5.6. Soft Costs by Other Characteristics As the questionnaire responses and interviews with cost estimators indicated, other impor- tant determinants of soft costs are the characteristics of the sponsor agency, and the political, operational, or other circumstances under which the project is being developed. Figure 31 shows the correlation between soft costs and the experience level of the sponsor agency (in the left pane), and the installation conditions of the project (in the right pane). The left pane shows a rough spectrum of experience levels across the x-axis, from inexperienced on the left to fairly experienced with both mode and delivery/procurement method at right. The experience level of the project sponsor has a mixed correlation with soft costs. The right pane of figure 31 shows how the level of a project’s interaction with existing transit service can affect soft costs. Specifically, the more a project must coordinate with and work around other services, the more soft costs tend to increase in percentage terms. A project to con- struct a new, stand-alone transit line that is not adjacent to any previous service seems to require less design costs than projects to extend or expand an existing rail line. When a construction proj- ect interacts with existing transit service in any way, more engineering and design work has typ- ically been required in the final design phase. Working on or near an active rail right-of-way poses additional logistical challenges that must be planned for, and may also trigger additional safety requirements. Extending a rail line will mean integrating the new track and station(s) into the older infrastructure, and additional work is usually required to ensure that signal, power, safety, and other systems operate compatibly. Figure 32 summarizes the relationship between soft costs and three other project characteris- tics: whether the project required a direct interface with existing service, whether political or pub- lic influence was unusually high, and whether public involvement or opposition was significant. As Figure 32 shows, a project that requires a direct connection or interface with existing revenue service, such as a line extension, a new branch intersecting an existing line, or the rehabilitation of an existing line, tends to show somewhat higher soft costs. Projects where political influence As-Built Soft Cost Analysis 47 Experience Level of Sponsor Agency at the Time Installation Under Active Revenue Service Sample Size: 14 4 39 21 17 31 3 28.1% 31.3% 41.5% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% No Active Service Adjacent Active Service Rebuild Under Operation So ft Co st s (% of C on str uc tio n)34.1% 34.5% 25.8% 27.3% 32.2% 0% 5% 10% 15% 20% 25% 30% 35% 40% No experience with Mode or Procurement No recent experience with mode or tunnel, other rail or past experience Experience with mode, new procurement Recent experience with mode, not with procurement Recent experience with mode and procurement So ft Co st s (% of C on str uc tio n) Figure 31. Soft costs versus sponsor experience level and installation conditions.

is unusually high or where public involvement or opposition is significant also tend to be correlated with higher soft cost percentages. Figure 33 compares soft cost percentages to the sponsor agency’s tendency to use outside con- tractors to varying degrees, to whether the project was ever required to be redesigned for any rea- son, and to whether the project’s planning phase was unusually long. As the left pane shows, sponsors that make more extensive use of outside contractors in early project development phases to design and plan tend to incur somewhat higher soft cost expenditures. However, spon- sors who use contractors in both the development and construction phases do not typically see significant differences in soft cost percentages. The middle pane of Figure 33 shows that the two projects in the dataset that had to undergo significant redesign do not show significantly different soft cost percentages. The right pane of Figure 33 demonstrates that when projects remain in development stages for an unusually long period of time (beyond approximately five to seven years), their soft cost per- centages tend to increase. A significant component of engineering and design soft cost is simply the salary and benefit costs of planners working on the project. When the planning phases for a project take an unusually long time, these costs tend to continue to be charged to the project, increasing 48 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Direct Interface w/Existing Service Unusually High Political or Public Influence Significant Public Involvement or Opposition Sample Size: 32 19 20 10 4131 33.4% 29.2% 0% 5% 10% 15% 20% 25% 30% 35% 40% True FalseFalse So ft Co st s (% of C on str uc tio n) 29.0%31.9% 0% 5% 10% 15% 20% 25% 30% 35% True So ft Co st s (% of C on str uc tio n) 29.8% 35.1% 0% 5% 10% 15% 20% 25% 30% 35% 40% True False So ft Co st s (% of C on str uc tio n) Figure 32. Soft costs versus installation conditions, political influence, and public involvement. Significant Use of Contractors Significant Redesign Required Abnormally Lengthy Project Development Sample Size: 4 9 2 7 444938 30.6% 30.9% 0% 5% 10% 15% 20% 25% 30% 35% 40% True False So ft Co st s (% of C on str uc tio n) 29.5%30.1% 41.3% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% Little roles outside of traditional More extensive use in project development Extensive use in project development and construction So ft Co st s (% of C on str uc tio n) 29.4% 39.7% 0% 5% 10% 15% 20% 25% 30% 35% 40% True False So ft Co st s (% of C on str uc tio n) Figure 33. Soft costs versus use of contractors, redesign required, and lengthy project development phase.

overall soft costs. When a significant amount of time elapses between entering preliminary engi- neering and the beginning of construction, projects incur higher soft cost percentages. 4.5.7. Soft Costs by Multiple Project Characteristics So far, this analysis has focused on testing the cost relationship between soft cost percentages and a project’s characteristics one variable at a time. The next step of this analysis tests the ability of a number of variables in combination to predict the variability in soft costs between projects using multivariate regression techniques. To do this, various combinations of variables were tested, including those variables that did not show particularly strong correlations in the univariate analysis. Soft costs as a percentage of construction costs was the dependent variable, and different combinations of project characteris- tics were the independent variables. Variables that described broadly similar project characteristics were grouped, and the relative contribution of each variable to the overall predictive power (R2) of the regression was measured. In an iterative fashion, one or several variables for each broad facet of the project were retained while many other indicators were left out. The following describes the variables tested and the resulting decision. Project Magnitude The variables with the best ability to predict soft cost percentages were alignment length (in linear feet) and construction costs, adjusted to 2008 dollars. It may seem counterintuitive that alignment length and construction cost in combination pro- duced opposite signs since both measures broadly describe the magnitude of the project. However, these two measures in tandem are good predictors of soft costs and produce better statistical re- sults together than either of them alone, one divided by another, or other measures of project magnitude such as number of stations or station density. The two variables together capture the special cases where short, expensive projects (such as a tunnel project) or long, less- expensive projects (such as service on existing right-of-way or in less developed areas) may tend to demonstrate differing soft costs. Several other variables describing the magnitude of a project were tested but were eliminated since they contributed relatively less to the regression analysis: • Construction costs per linear foot, • ROW costs as a percentage of construction, • Vehicle costs as a percentage of construction, and • Number of stations. Project Complexity Of many measures of project complexity, its mode, an indicator of installation conditions (i.e., whether the project is a new standalone line with no active adjacent service or not), and an indicator of an unusually lengthy project development phase were the best predictors of soft cost percentages. Heavy rail projects tend to incur somewhat higher soft costs than light rail, other things being equal, perhaps due to their relative complexity. This finding contrasts somewhat with that of Figure 25 because this multivariate regression controls for other factors influencing soft costs. Heavy rail projects can typically involve constructing guideway and systems that have been designed to more rigorous engineering standards that support more complex systems, move higher passenger volumes, and operate at higher speeds relative to light rail. This finding in the multivariate analysis confirms the results of the industry questionnaire and the interviews with cost estimators. As-Built Soft Cost Analysis 49

A project to construct a new, stand-alone transit line that is not adjacent to any previous ser- vice will usually require less design costs than projects to extend or expand an existing rail line. When a construction project interacts with existing transit service in any way, more engineering and design work has typically been required in the final design phase. Working on or near an active rail right-of-way poses additional logistical challenges that must be planned for and may also trigger additional safety requirements. Extending a rail line will mean integrating the new track and station(s) into the older infrastructure, and additional work is usually required to ensure that signal, power, safety, and other systems operate compatibly. Note that this variable is not statistically significant to a high degree of certainty (t-statistic of −1.25). A significant component of engineering and design cost is simply the salary and benefit costs of planners working on the project. When the planning phases for a project take an unusually long time, beyond approximately five to seven years, these costs tend to continue to be charged to the project, increasing overall soft costs. Other variables were eliminated due to their relatively low contribution to the regression analysis’ predictive power: • Station density (number of stations per mile of guideway constructed), • Percentage of guideway below grade, • Percentage of guideway not at grade, • Rebuild or rehabilitation under operation (dummy variable), • Project type (new service, extension of existing service, or rehabilitation of existing service), and • Direct interface with existing revenue service required. Delivery Method A dummy variable indicating whether the project sponsor chose an alternative project delivery method (i.e., a method that is not the traditional design–bid–build) contributed the most to the regression analysis in a statistically significant way. Including the specific effects of a certain kind of alternative delivery method did not strengthen the regression analysis, primarily due to the small sample size of such projects. When sponsors choose to procure their project through an alternative delivery mechanism such as design–build, design–build–own–maintain, or construction manager/general contractor, these projects have historically incurred lower soft costs. In addition, these alternative delivery methods tend to frontload more design and planning costs in preliminary engineering. However, the lower soft costs of projects implemented with alternative delivery methods may be partially the result of differences in measurement rather than a real reduction in cost. Con- tractors may simply categorize their costs in different ways than transit agencies (in the construc- tion line item, for example), which makes that project’s soft costs as a percent of construction appear low. Sponsor Agency Characteristics An indicator of whether the sponsor agency tended to minimize capital charges contributed the most to the regression. A dummy variable indicating if the sponsor agency tended to rely heavily on outside contractors during project development phases did not demonstrate significant power to predict soft cost percentages, and was excluded. When a transit agency sponsors a construction project, it usually contributes some of its own labor and even materials. Agency employees often inspect construction activities, monitor safety, administer the contract, acquire property, manage the project, and perform many other tasks. As opening day approaches, agency staff contribute time coordinating testing, training, safety inspec- tions, and shared tasks with other agencies. The agency chooses whether to charge these expendi- 50 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

tures to the capital project (either directly or as an overhead-type allocation) or to absorb them into the operating budget, and project sponsors each have different internal policies for this. External Factors Of many indicators of the broader circumstances in which a project is developed, two variables stood out: economic conditions and unusual political influence. The overall health of the economy, as well as the level of construction activity, can affect the construction bids a transit project sponsor can expect to receive. If the construction sector or economy at large is in a downturn when a project sponsor accepts bids, contractors may reduce their bids due to economic forces. In this case, soft costs computed as a percentage of the engineered construction cost estimate might look relatively higher simply because the bid construction cost is lower. Historically, some change in soft costs can be attributed to the rate of gross domestic product (GDP) growth when construction contracts are bid, after accounting for other variables. Although GDP growth rises and falls with the economy, it has historically risen an average of 2.5% to 3.0% per year. However, it is difficult to use this driver to estimate soft costs for a project years away from construction since future GDP growth is difficult to predict. The Guidebook therefore recommends using this cost relationship only when a cost estimator can be reasonably sure the project is to be bid within one year. When public involvement or political pressures are high, such as in a contentious design and planning process, soft costs tend to rise relative to construction costs. When, for example, mul- tiple planning boards, citizen advisory councils, and officials must approve the design and could even call for a redesign, these external factors were shown to increase soft costs. Other measures of project context and external circumstances contributed less to the regression analysis and were excluded: • Unusually high public involvement and/or opposition (dummy variable), • Major project redesign required (dummy variable), and • Decade. The multivariate regression also used midyear of expenditures as an independent variable. As Figure 25 showed earlier, soft costs have been rising over time, so including this variable controls for the effect of the historic rise in soft costs. However, in estimating soft costs for a given project, the Guidebook does not recommend increasing soft cost percentages for future projects. Extension regression analysis yielded a 10-variable equation that can explain approximately 60% of the difference in soft cost percentages by variations in the projects’ characteristics (R2 = 0.58). Table 15 shows the resulting coefficients from this regression, whose dependent variable is total soft costs as percent of construction costs. As-Built Soft Cost Analysis 51 Variable Name Unit Coefficient t-Stat Guideway alignment length 10,000 linear feet 1.4% 2.69 Construction costs Billions, 2008$ -5.9% -2.49 Mode Dummy, heavy rail = 1 6.0% 1.64 Installation conditions Dummy, no active service = 1 -3.8% -1.25 Delivery method Dummy, non-DBB = 1 -7.2% -2.10 Economic conditions GDP % annual growth -1.4% -2.34 Unusually long project development phase Dummy, yes = 1 7.1% 2.08 Unusual political influence Dummy, yes = 1 6.6% 2.22 Agency tendency to minimize capital charges Dummy, yes = 1 -6.0% -1.65 Years from 2008 Years -0.4% 2.22 Table 15. Multivariate regression results on soft costs as a percentage of construction costs.

Using the projects contained in this FTA capital cost database, the strongest correlation that could be produced is the regression described above. After testing many combinations of explanatory independent variables, these 10 could best predict the relationship between soft and hard costs. Although the strength of this correlation is not ideal (the R2 and t-statistics are rela- tively small), the relationship does highlight the importance of judgment in cost estimation. In addition, as more projects are included in this cost database, it may be possible to perform analysis with stronger cost relationships. 4.5.8. Preparing Multivariate Results for Use in Guidebook Alternative multivariate regressions were examined using different actual soft cost components (rather than total soft costs) as the dependent variable. The coefficient from the overall soft cost analysis was distributed to the soft cost components that correlated to the project characteristics in a statistically significant way. For example, alignment length showed an overall coefficient of around 1.4% per 10,000 linear feet regressed against overall soft costs, and this relationship was strongest when regressed against project management and other soft costs, so the Guidebook rec- ommends adjusting the percentage estimate for those two components to a total of 1.4% per 10,000 linear feet. Finally, the starting points and recommended percentage adjustments were validated against the original projects to gauge the potential error in the Guidebook methodology. Some minor adjustments to the coefficients were made to minimize the sum of each component’s root mean square error for all projects. 52 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

This Final Report presents the research, data sources, and analysis underlying Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects, Part 1: Guidebook, which came out of TCRP Project G-10. This Final Report is intended to support the information summa- rized in the Guidebook in Part 1. Please refer to the Guidebook in Part 1 for a summary of how the results of the research presented here can be applied to practice, including an introduction to soft costs and a new methodology to estimate these soft costs based on historical projects. This conclusion section summarizes the key points from previous sections, and presents objectives for future research. 5.1. Literature Review While the term “soft costs” is often similar to “indirect costs” or other terminology, the FTA’s definition of Standard Cost Category 80, Professional Services, is an operational definition con- sidered equivalent to soft costs for this report and consistent with the financial, construction, and related literature. 5.2. Soft Cost Estimation: State of the Practice Cost estimators for transit construction projects follow different approaches to estimating soft costs depending on the phase of the project. During alternative analysis through preliminary engineering, soft costs are estimated for each cost component as a percentage of hard construction costs. Estimators begin with a range of per- centages for each soft cost component and apply a value within that range to a specific project based on knowledge about the project and its sponsor. Figure 34 shows the percentages used for each cost component by each cost estimator questioned for this research. For each cost component, estimators choose one percentage from within that range based on historical experience and their knowledge of the specific project characteristics. During the final design and construction phases, estimates of soft costs based on a percentage of construction cost are replaced with more closely tailored, bottom-up estimates relying heav- ily on past experience with similar projects, as indicated earlier. Estimators usually perform a resource-driven analysis for each cost element. For instance, administration costs may be esti- mated based on headcount and construction schedules. 53 C H A P T E R 5 Conclusion

5.3. As-Built Cost Analysis Analyzing the database of actual as-built soft cost expenditures provided the following insights into soft costs: • Soft costs have historically averaged 31% of construction costs, a value that is consistent with how the industry currently estimates soft costs both in total and at the component level. • However, the range of variability in past projects has been wider than the range estimators report. While estimators report an uncertainty range of ±10%, actual soft costs have been as low as 11% of hard costs and as high as 54% of hard costs, or an uncertainty range of around ±20%. • Soft costs have averaged around $2,600 per linear foot for light rail, and around $5,700 per linear foot for heavy rail, with a range between $300 and $10,000 per linear foot of guideway for both modes (2008$, outliers removed). The as-built analysis also revealed relationships between project characteristics and soft costs: • Soft costs have been increasing over the past four decades, particularly for heavy rail projects. • Project complexity, mode, delivery method, magnitude, and context all appear to drive soft costs. Univariate analysis reveals some relationships between these considerations and soft costs, but a more complete and consistent picture emerges through a multivariate regression analysis. A multivariate analysis of 10 variables captured the cumulative effect of a number of variables on soft cost percentages and was able to explain approximately 60% of variability in soft costs. • Projects where alignments stretch longer distances tend to incur somewhat higher soft costs as a percentage of construction cost. • More expensive construction projects tend to display somewhat smaller soft cost percentages, other things being equal. • Heavy rail projects tend to incur somewhat higher soft costs than light rail, perhaps due to their relative complexity and higher engineering standards. • A project to construct a new stand-alone transit line will usually require less design costs than a project to extend, expand, or interface with existing transit services. 54 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1* 2* 3 98476 5 10 M id -R an ge E st im at ed S of t C os t (% of C on str uc tio n) Questionnaire Respondents Other Insurance + Legal Project Mgmt. and Construction Admin. FD PE *Respondents estimate PE + FD as combined amount; PE displayed here using average split Figure 34. Midpoint soft cost estimates for all components reported by surveyed cost estimators.

Conclusion 55 • Projects procured with alternative delivery methods such as design–build appear to have incurred less soft costs. • The health of the national economy and the level of construction activity can affect the relation- ships between soft and hard costs. • Longer project planning phases, unusual political influence, and a sponsor agency’s capital- ization policies may increase soft cost requirements. • In the end, cost relationships based in historical evidence cannot explain 100% of the variability in soft costs. Therefore, soft cost estimation must blend the art of human judgment with the science of cost relationships. 5.4. Future Research Directions More in-depth research into the documentation of one or several recent construction proj- ects will enhance the understanding of soft cost drivers. Moreover, a comprehensive industry outreach will provide further insight on context-specific soft cost estimation practices. Finally, the methodology to estimate soft costs for public transportation infrastructure projects devel- oped here is based on past heavy and light rail construction projects and is therefore not entirely applicable to other prevalent public transportation capital infrastructure projects such as BRT, commuter rail, streetcar, or other state-of-good-repair projects to repair or replace aging infra- structure. Additional data and research would help estimate soft costs for these kinds of projects.

56 Adrian, J.J. (1982). Construction Estimating. Reston Publishing Company, Reston, VA. American Public Transportation Association (2005). RE: DOT DMS Docket Number FTA-2005-20585. http://www. apta.com/government_affairs/regulations/docket050404.cfm (As of October 5, 2007). Association for the Advancement of Cost Engineering (AACE) (2007). Cost Engineering Terminology. AACE International Recommended Practice No. 10S-90. Association of Physical Plant Administrators (APPA) (2005). Asset Lifecycle Model for Total Cost of Ownership Management; Framework, Glossary & Definitions. http://www.ifma.org/daily_articles/2005/july/07_19.cfm (As of October 5, 2007). Association of Physical Plant Administrators (APPA) (2006). 2004–05 Facilities Performance Indicators Report. ISBN: 1-890956-33-3. Ball, J.F. and D.T. Petrowski (2007). Projecting Construction Costs with Tempered Cost Indices. Association for the Advancement of Cost Engineering (AACE), International Transaction, EST. 09. Bartholomew, S.H. (2000). Estimating and Bidding for Heavy Construction. Prentice Hall, Upper Saddle River, NJ. Booz Allen Hamilton Inc. (2005). TCRP Web-Only Document 31: Managing Capital Costs of Major Federally Funded Public Transportation Projects: Contractor’s Final Report. Transportation Research Board of the National Academies, Washington, DC. http://onlinepubs.trb.org/onlinepubs/tcrp/tcrp_w31.pdf (As of November 9, 2009). CEEC (2004). Code of Measurement for Cost Planning. Comité Européen des Economistes de la Construction, January. Collier, K. (1984). Estimating Construction Costs. Reston Publishing Company, Reston, VA. Commercial Construction Loans and Soft Costs. Loans.com Free Commercial Lender Databank, http://www. cloans.com/commercialconstructionloans.html (As of October 5, 2007). Construction Management Terms. Constructionplace.com Incorporated, http://www.constructionplace.com (As of October 5, 2007). Czarnecky, J. and D. Schmitt (2007). How to Conduct a Cost-Effective Transit Alternatives Analysis. 11th TRB National Transportation Planning Applications Conference, Session 15: Transit Matters. Dagostino, F.R. and L. Feigenbaum (1996). Estimating in Building Construction, 5th ed. Prentice Hall, Upper Saddle River, NJ. Estimated Soft Cost. http://www.angelarmsworks.com/karen/project_cost.htm (As of October 5, 2007). Flyvbjerg, B., M.K.S. Holm, and S.L. Buhl (2005). “How (In)accurate Are Demand Forecasts in Public Works Projects?” Journal of the American Planning Association, Vol. 71, No. 2, 131–146. Gould, F.E. (2005). Managing the Construction Process, 3rd ed. Pearson Prentice Hall, Upper Saddle River, NJ. Halpin, D.W. and R.W. Woodhead (1998). Construction Management, 2nd ed. John Wiley & Sons, New York, NY. Hendrickson, C. and T. Au (1989). Project Management for Construction. Prentice Hall, Englewood Cliffs, NJ. Holland, N. and D. Hobson (1999). “Indirect Cost Categorization and Allocation by Construction Contractor.” Journal of Architectural Engineering, American Society of Civil Engineers, Vol. 5. Hurley, M.W. and A. Touran (2002). “Cost Structure and Profitability of Design Services Industry.” Journal of Management in Engineering, ASCE, Vol. 18, No. 4, 167–172. KRG Insurance Group (2002). Soft Cost—Building Insurance Valuation, Volume 2, Issue#2, http://www. krg.com/News.asp?A=9 (As of October 5, 2007). Mason, J.J. (1993). “The Cost Approach.” Appraisal Journal, Vol. 61, Issue 1, 116. Murphy, Jeannene D. ed. (2008). Means Construction Cost Index, Vol. 34. No. 1, R. S. Means Company: Kingston, MA. Norfleet, A.D. (2007). The Theory of Indirect Costs, AACE International Transactions, 12.1–12.6. Bibliography

Paulson, B.C. and D.S. Barrie (1992). Professional Construction Management, 3rd ed., McGraw-Hill, New York, NY. Pellicer, E. (2005). “Cost Control in Consulting Engineering Firms.” Journal of Management in Engineering, ASCE, Vol. 21, No. 4, 189–192. Schexnayder, C.J. and R.E. Mayo (2004). Construction Management Fundamentals. McGraw-Hill, New York, NY. Schuette, S.D. and R.W. Liska (1994). Building Construction Estimating, McGraw-Hill, New York, NY. Soft Cost Definition. BusinessDictionary.com, http://www.businessdictionary.com/definition/soft-cost.html (As of October 24, 2007). Standard Form of Cost Analysis (2003). Building Cost Information Service (BCIS), the Royal Institution of Chartered Surveyers, UK. Stoy, C. and M. Wright (2006). “CEEC Code of Measurement for Cost Planning,” Zeitschrift fur Immobilienokonomie—German Journal of Property Research, No. 1, 47–60. Tah, J., A. Thorpe, and R. McCaffer (1994). “A Survey of Indirect Cost Estimating in Practice.” Journal of Construction Management and Economics, 12, 31–36. Ubeck, J. (2006). “Insight into Indirect Rates.” Contract Management, Vol. 46, Issue 5, 26–35. U.S. Army Corps of Engineers (1997). Engineering Instructions: Construction Cost Estimates. EI 01D010, Department of Army, Washington, DC. http://www.saj.usace.army.mil/Divisions/Engineering/Branches/CostEngineering/ DOCS/EI01D010CostEstimates.pdf (As of January 19, 2010). U.S. Federal Transit Administration (1995). Transit Capital Cost Price Index Study, Chapter 2: Definitions: Proj- ect Elements, Cost Categories and Fixed Guideway Technologies. http://www.fta.dot.gov/publications/ reports/other_reports/publications_4855.html (As of October 5, 2007). U.S. Federal Transit Administration (2003). Project & Construction—Management Guidelines, Appendix A: Cost Estimation Methodology. http://www.fta.dot.gov/publications/reports/other_reports/planning_environment_ 1344.html (As of October 5, 2007). U.S. Federal Transit Administration (2008). Standard Cost Category Workbook. Rev. 10, June 4, 2008. http:// www.fta.dot.gov/documents/SCC_Workbook.xls (As of July 21, 2008). What Are Hard Costs & Soft Costs, Build Max Learning Center, http://www.buildmax.com/help/budgeting_ what_are_hard_costs_and_soft_costs.asp (As of October 5, 2007). Bibliography 57

58 The following section presents the experience and qualifications of the professional cost esti- mators interviewed for this research. Cesare DeRose, a vice president with AECOM Transportation, has over 25 years of experience in the field of heavy construction. Projects Mr. DeRose has worked on have been in the $1 to $200 million range. Typical duties have included cost estimating, scheduling, engineering, design, constructability, project management, site supervision, and project planning in all types of involved engineering tasks. Mr. DeRose has worked on many large projects, including the Lincoln Center Development Corporation, preliminary and final design of the Second Avenue Subway, the New York City Water Tunnel project, the Queensboro Bridge, the Charles River Bridge crossing in Boston, the Amawalk and Titicus dams, rehabilitation of the Brooklyn Bat- tery Tunnel, the Tappan Zee fender replacement, and the Hillview Reservoir wall extension and sediment removal. In addition to this, his involvement has been with bridge, highway, and other heavy construction; sewer and utility work; foundation supports; marine construction; and com- mercial rehabilitation. James T. Czarnecky, AICP, a senior project manager with AECOM Transportation, is a nation- ally certified professional planner (AICP) and Master of Community Planning (MCP), and has 19 years of transportation planning and civil engineering experience. He has managed and par- ticipated in all phases of project development, including systems planning, major investment studies (MIS), alternatives analysis, environmental impact statements, preliminary engineering, and final design and construction support. He specializes in the development and analysis of multimodal transportation networks, inclusive of transit and highway components designed to complement the foreseeable socioeconomic conditions and related planning initiatives unique to each community. Mr. Czarnecky is focused on a realistic approach to solving transportation problems with respect to the balancing of costs and benefits. Raul V. Bravo, president of Raul V. Bravo & Associates Inc., has over 40 years of experience in the design, development, construction, and implementation of transportation vehicles and sys- tems. Since 1974 Mr. Bravo has been primarily involved with guided transportation; first as engineering manager for Rohr Industries and later as Amtrak’s director of equipment design and operations planning; since 1979, Mr. Bravo has been managing director of Raul V. Bravo & Associates Inc., transportation planners and engineers, located in the Washington, DC, metro- politan area. Mr. Bravo is a member of the Railroad Safety Advisory Committee (RSAC) assist- ing the Federal Railroad Administrator in developing new rules and regulations. Mr. Bravo is also a member of TRB committees examining the future of intercity passenger rail in the United States and management structures, and standardization of rail systems and vehicles. A P P E N D I X A Cost Estimators Interviewed

This appendix provides a key of the abbreviated names for the projects identified in this analy- sis and offers a short description of each project. The project descriptions and graphics below provide a brief snapshot of the variety of projects contained in the capital cost databases used in this analysis. B.1. Data Sources for Project Descriptions While the detailed capital costs are from FTA cost databases, the following descriptions were developed from the following data sources: • FTA’s Light Rail Transit Capital Cost Study Update, 2003 (3–6) • FTA’s Annual Report on New Starts, various years (2006–2009), Alphabetical List of Projects by Development Phase and State, Full Funding Grant Agreements, Appendix A: New Starts Project Profiles • Project information and fact sheets from project sponsors • Transit agency/project sponsor websites • Internet sources 59 A P P E N D I X B Project Names and Descriptions in As-Built Analysis Abbreviated Name Full Project Name Approx. Length (mi) Midyear of Expend. Mode Delivery Method Sacram. I Sacramento Stage I 20.6 1985 Light DBB Pittsburgh I Pittsburgh Light Rail Stage I 24.5 1984 Light DBB Portland Seg1 Portland MAX Segment I 15.0 1984 Light DBB LA Blue Los Angeles – Long Beach Blue Line 22.6 1987 Light DBB San Jose N San Jose North Corridor 20.8 1985 Light DBB Hud-Berg I Hudson-Bergen MOS-I 8.7 1999 Light DB Hud-Berg II Hudson-Bergen MOS-II 6.1 2000 Light DB Hiawatha Hiawatha Corridor 11.6 2001 Light DB Portland Int Portland Interstate MAX 5.8 2002 Light DB San Diego San Diego Mission Valley East 5.5 2003 Light DBB St. Louis St. Louis St. Clair County Extension 17.4 1999 Light DBB Salt Lake Salt Lake North-South Corridor 15.0 1998 Light DBB South NJ Southern New Jersey Light Rail Transit System 34.0 2002 Light DB Portland W Portland Westside/Hillsboro MAX 18.0 1996 Light DBB Sacram. So Sacramento South Corridor 6.3 2002 Light DBB Sacram. Fol Sacramento Folsom Corridor 11.4 2002 Light DBB LA Gold Pasa Pasadena Gold Line 13.7 2002 Light DB Denver SW Denver Southwest Corridor 8.5 1999 Light DBB Pittsburgh II Pittsburgh Light Rail Stage II 5.5 2002 Light DBB LA Gold East Los Angeles Eastside Gold Line 5.9 2006 Light DB Phoenix Phoenix Central/East Valley Light Rail Line 19.6 2008 Light DB Portland So Portland South Corridor 6.5 2005 Light CM/GC Seattle Cen Seattle Central Link Light Rail Project 13.9 2006 Light DBB Pittsburgh N Pittsburgh Northshore Light Rail Connector 1.2 2008 Light DBB Table 16. Data on projects included in as-built cost analysis.

60 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Abbreviated Name Full Project Name Approx. Length (mi) Midyear of Expend. Mode Delivery Method Charlotte Charlotte South Corridor 9.6 2005 Light DBB VTA Tas W VTA Tasman West 7.6 1999 Light DBB VTA Tas E VTA Tasman East 4.9 2004 Light DBB VTA Capitol VTA Capitol Segment – Connected to Tasman East 3.3 2004 Light DBB VTA Vasona VTA Vasona Segment 5.3 2005 Light DBB MARTA N-S Atlanta MARTA North-South Line 22.2 1984 Heavy DBB MARTA Dun Atlanta MARTA North Line Dunwoody Extension 7.0 1998 Heavy DBB MBTA Orang Boston MBTA Orange Line 4.7 1983 Heavy DBB Baltimore Baltimore MDMTA Metro Sections A and B 15.0 1982 Heavy DBB CTA Orange Chicago CTA – Southwest Orange Line 9.0 1990 Heavy DBB CTA O'Hare Chicago CTA – O'Hare Extension Blue Line 7.1 1981 Heavy DBB CTA Brown Chicago CTA Brown Line (Ravenswood) Rehabilitation 9.1 2006 Heavy DBB CTA Douglas Chicago CTA Blue Line (Douglas) Rehabilitation 5.6 2002 Heavy DBB LA Red 1 Los Angeles Red Line Segment I 3.4 1988 Heavy DBB LA Red 2 Los Angeles Red Line Segments 2A & 2B 6.7 1994 Heavy DBB LA Red 3 Los Angeles Red Line Segment III 6.5 1998 Heavy DBB Miami Miami Dade Metrorail 21.0 1982 Heavy DBB San Juan San Juan Tren Urbano 10.7 2002 Heavy DBB BART SFO San Francisco, CA BART SFO Extension 8.7 2002 Heavy DBB DC Shady G Washington, DC – Shady Grove (A Route) 18.0 1977 Heavy DBB DC Glenmt 1 Washington, DC – Glenmont (B Route) 5.7 1980 Heavy N/A DC Glenmt 2 Washington, DC – Glenmont Outer (B Route) 6.2 1996 Heavy N/A DC Huntgtn Washington, DC – Huntington (C Route) 12.1 1977 Heavy DBB DC New Ca Washington, DC – New Carrollton (D Route) 11.8 1974 Heavy DBB DC U St. Washington, DC – U Street (E Route) 1.7 1988 Heavy DBB DC Greenblt Washington, DC – Greenbelt Mid (E Route) 2.3 1997 Heavy DBB DC Anacost Washington, DC – Anacostia (F Route) 4.3 1988 Heavy DBB DC Anacost O Washington, DC – Anacostia Outer (F Route) 6.7 1999 Heavy DBB DC Addison Washington, DC – Addison (G Route) 3.5 1978 Heavy DBB DC Springfld Washington, DC – Springfield (J,H Route) 3.5 1988 Heavy DBB DC Vienna Washington, DC – Vienna (K Route) 12.0 1980 Heavy DBB DC L'Enfant Washington, DC – L'Enfant (L Route) 1.7 1974 Heavy DBB Phil Frankf. Philadelphia SEPTA Frankford Rehabilitation 5.3 1997 Heavy DBB NYCT 63rd New York NYCT 63rd Street Tunnel 0.4 1977 Heavy N/A NYCT Stillw New York NYCT Stillwell Terminal Rehabilitation 0.3 2000 Heavy N/A Table 16. (Continued). B.2. Project Descriptions Sacramento Stage I Sacramento, CA Label: Sacram. 1 The Sacramento Stage I project included a 20.6-mile light rail system with two lines, the Northeast (Blue) and Folsom (Gold) lines, which connect the eastern and northeastern suburbs to downtown Sacramento. The segment is mostly single-track, with double-tracking in passing sections for about 40% of its length. The alignment is largely at grade and is located on existing rights-of-way in freeway medians and abandoned railroad corridors. Pittsburgh Light Rail Stage I Pittsburgh, PA Label: Pittsburgh 1 The Stage I Light Rail Transit Program in 1980 to restore light rail transit service on the old trolley routes connecting the South Hills suburbs with downtown Pittsburgh. Stage I consisted of 12.5 miles of new alignment construction and 12 miles of right-of-way rehabilitation. In 1985 the first segment started operating for 1.6 miles underground in the downtown business district and at grade south of the Monongahela River to the South Hills Village.

Portland MAX Segment I Portland, OR Label: Portland Seg1 The Portland MAX Segment I construction project resulted in the opening of the first mod- ern light rail line in Portland in 1986. A 15-mile east-west alignment, named the Banfield Cor- ridor, was built mostly at grade with some elevated portions along joint highway alignments. It extended from the Cleveland Avenue station in Gresham to downtown Portland. The Segment I alignment permits trains to operate in reserved rights-of-way in city streets, arterials, and high- way medians. Of the 30 stations built, 25 are at grade, less than a mile apart, and have easy access for pedestrians. Stations generally lack park-and-ride facilities but have bus transfer facilities with good intermodal coordination. MAX Segment I was the first segment to open on the present day Hillsboro-Gresham (Blue) line that was extended in 1998 with the opening of the Portland Westside/Hillsboro MAX segment. Los Angeles—Long Beach Blue Line Los Angeles, CA Label: LA Blue The Blue Line is a modern light rail transit line in Los Angeles and primarily uses the original Pacific Electric right-of-way. It provides riders from the communities of Vernon, Huntington Park, South Gate, Watts, Compton, Carson, and Long Beach with access to downtown Los Angeles and the greater Metro system. The 22.6-mile line required 22 stations and connects at its down- town terminus to the Metro heavy rail lines at the 7th Street/Metro Center station. Its southern ter- mini stations are in the 4-station loop in downtown Long Beach. Approximately 80% of the line is a dedicated alignment, mostly at grade or elevated with an underground portion. Construction began in October 1987 and revenue service commenced in July 1990. This project encountered some complications in planning and design due to unexpected environmental review, state environmental laws, and an active political and stakeholder environment. San Jose North Corridor San Jose, CA Label: San Jose N Revenue service commenced in December 1987 in a small segment of the San Jose North Cor- ridor that would become the first section built of a longer San Jose Guadalupe Corridor that would require two phases to reach completion. The full 20.8-mile North Corridor was completed and servicing passengers in April 1991. This project’s alignment is mainly located along the median area of major roadways and a transitway through downtown San Jose. The alignment is at grade for nearly the full length and required only one bridge, two overpasses, and a short underpass to be built in the new guideway. The guideway is double-tracked for its entirety except for two small sections of single-track operation. Hudson-Bergen MOS-I Newark, NJ Label: Hud-Berg I The first two lines of the Hudson-Bergen Light Rail began full revenue operations in 2002. (Ser- vice had opened in three phases between 2000 and 2002.) The project included 8.7 miles of double- tracking and 14 stations (including intermodal transfer stations). The fully built segment starts at the Hoboken Terminal and runs south towards the Liberty State Park station after which the 22nd Street-Hoboken (Blue) and West Side Avenue-Tonnelle Avenue (Orange) lines separate with the latter running a 3-station spur line to western Jersey City. The alignment of the 22nd Street- Hoboken Terminal line continues south from the Liberty State Park junction parallel with the I-78 and Garfield Avenue corridors and then along Avenue E, terminating at the 22nd Street station. Project Names and Descriptions in As-Built Analysis 61

This project encountered some soft cost complexities when the project underwent an engi- neering redesign after the sponsor had executed design–build contracts and had begun util- ity relocation. Hudson-Bergen MOS-II Newark, NJ Label: Hud-Berg II This project included a new light rail line from the Hoboken Terminal station to North Bergen County (the green-colored Tonnelle Avenue-Hoboken Terminal Line). In addition, a station was added to the Blue Line, moving the southern terminus from the 34th Street station to the 22nd Street station. The completed project required 6.1 miles of track and 7 new stations. The MOS-2 segment opened for revenue service in increments from 2003 to 2006. Hiawatha Corridor Minneapolis, MN Label: Hiawatha The Hiawatha Corridor LRT project included an 11.6-mile light rail transit line with 17 stations that operates primarily in the Hiawatha Avenue/Trunk Highway 55 Corridor linking downtown Minneapolis to the Mall of America in Bloomington and also servicing the Minneapolis-St. Paul International Airport. The alignment includes a 1.5-mile tunnel under the airport runways. Revenue operations began in December 2004. Portland Interstate MAX Portland, OR Label: Portland Int The Portland Interstate MAX Light Rail Project included a 5.8-mile, 10-station light rail tran- sit line (Yellow) that extends north from downtown Portland parallel to the I-5 Corridor. The line branches from the existing Blue Line in the Rose Quarter District, follows the median of Interstate Avenue for 4.5 miles, between the Albina and Overlook Park stations, to Kenton, and then is on a separate alignment to the Portland Exposition Center terminus, which is just south of the Columbia River. The original design called for the line to extend across the river to Van- couver, Washington, but Tri-Met scaled back alignment options after Portland voters rejected a bond measure. This project’s alignment near an active highway also raised design complexities. The project opened to revenue service in May 2004. San Diego Mission Valley East San Diego, CA Label: San Diego The Mission Valley East (MVE) project included in a new double-track light rail line that runs from the Mission San Diego Trolley station east of I-15 to the Grossmont Center Trolley station. The new line provides important connectivity between the pre-existing Blue and Orange Lines, as well as San Diego State University, which was an active stakeholder in the design process. The 5.9-mile project required 4 new stations and the renovation of an existing station. The project opened for revenue service in July 2005. St. Louis St. Clair County Extension St. Louis, MO Label: St. Louis The St. Clair County Metrolink Extension Project is a three-phase light rail construction proj- ect that will eventually extend service over 26 miles from East St. Louis, IL, to the MidAmerica 62 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

Airport in St. Clair County. The Phase 1 segment opened for revenue service in May 2001. It is a 17.4-mile Minimum Operable Segment (MOS) light rail extension of the existing Red Line from the prior terminus at the 5th & Missouri station in East St. Louis to the College Sta- tion. The project required 8 new stations, 7 park-and-ride lots, 20 new LRT vehicles, and a new vehicle maintenance facility. Salt Lake North-South Corridor Salt Lake City, UT Label: Salt Lake The North-South Corridor included construction of the SLC-Sandy line, which opened for revenue service in 1999 from the Arena Station to the Sandy Civic Center Station. The original 15-mile light rail alignment starts on South Temple, turns right onto Main Street, right at 700 South, left at 200 West, and then follows the Union Pacific (UP) corridor. The remainder of the alignment goes south within the UP corridor to the 10000 South (Sandy Civic Center) station. The original line was mainly built with double-tracking, with two single track sections at the I-215 overpass and the State Street Bridge (U.S. Highway 89), and had 16 stations. Southern New Jersey Light Rail Transit System Trenton, NJ Label: South NJ The Southern New Jersey Light Rail Transit System, known as the “River Line,” was built for intercity travel in the southwestern part of the state. The line has 20 stations between Tren- ton and Camden, near Philadelphia. The 34-mile light rail system runs roughly parallel to New Jersey Highway Route 130 in the former Conrail right-of-way adjacent to the Delaware River. The line’s construction required upgrading 50 at-grade crossings on local streets and the reconstruction of 20 bridges. Stations connect to other public transport services offered by NJ TRANSIT, PATCO, SEPTA, and Amtrak to provide passengers with easy connections to New York City, Philadelphia, Trenton, and Atlantic City. Construction began in May 2000, and rev- enue operations began in 2004. Portland Westside/Hillsboro MAX Portland, OR Label: Portland W The Westside/Hillsboro extension is an 18-mile light rail extension to the TriMet MAX Blue line from downtown Portland to Beaverton and Hillsboro. While TriMet, the sponsor agency, initially considered designing an alignment that runs at 6% grade to cross the West Hills, which rise 700 feet higher than the downtown area, this plan was eventually changed in favor of a 3-mile twin tube tunnel. The alignment emerges from the twin tube tunnel, which includes the Washington Park Station at 260 feet below ground, to follow Highway 26 to the Sunset Transit Center before turning onto Highway 217. The alignment approach at the Beaverton Transit Center required newly constructed right-of-way. The line eventually ends on 12th Avenue in Hillsboro before terminating on Washington Street. Revenue operation began in 1998. Sacramento South Corridor Sacramento, CA Label: Sacram. So The Sacramento South Corridor includes a 6.3-mile light rail line with 7 stations that spurs southward at the 16th Street station from the original Sacramento Light Rail alignment. The Project Names and Descriptions in As-Built Analysis 63

constructed section originates in downtown Sacramento at the intersection of 16th and Q streets and follows the Union Pacific freight corridor until it terminates at the Meadowview Road sta- tion. The extension opened for revenue operation in 2003. Sacramento Folsom Corridor Sacramento, CA Label: Sacram. Fol The Sacramento Folsom Corridor light rail project was built to extend transit service within a corridor following Highway 50 to downtown Folsom. The 10.7-mile suburban extension required 9 stations between the Mather Field/Mills station and downtown Folsom. In addition, the Sacramento Valley station (adjacent to the Amtrak station) was built and connected via a new 0.7-mile double-track extension to the existing 8th & K station. The con- nection to Amtrak service required additional boarding platforms to be constructed at exist- ing stations. Pasadena Gold Line Los Angeles, CA Label: LA Gold Pasa The Pasadena Gold Line runs 13.7 miles, stopping at 13 stations, to connect Chinatown, High- land Park, South Pasadena, and Pasadena to downtown Los Angeles via Union Station (its west- ern terminus). At Union Station this light rail line provides walking connections to the Red and Purple heavy rail lines. Construction commenced in 1994 and revenue operations were sched- uled to begin in May 2001. Unfortunately, a lack of funding and other complications resulted in construction stoppage. The state of California authorized the creation of the Metro Gold Line Construction Authority in 1998 with the sole purpose of immediately instituting tighter cost con- trols and resuming design, contracting, and construction of the Los Angeles to Pasadena Metro Gold Line. The newly formed construction authority completed construction in three years and the line opened for revenue service in 2003. Denver Southwest Corridor Denver, CO Label: Denver SW The Southwest Corridor line was built to connect the southern portion of Denver with its downtown via the already operational Central Corridor at the I-25 & Broadway station. The extension added 8.5 miles and 5 stations of service to the growing Denver light rail system. The extension is entirely grade-separated from the I-25/Broadway station to the Mineral Avenue sta- tion in Littleton, Colorado. This project planning phase spent some time addressing complexi- ties arising from the need to accommodate through-routing of trains. Revenue service on the extension began in 2000. Pittsburgh Light Rail Stage II Pittsburgh, PA Label: Pittsburgh II The Stage II LRT Priority Project included the reconstruction of the Overbrook line, a 5.5-mile existing rail line, which had closed in 1993 because of the deterioration of old bridges. This included rebuilding the existing light rail track bed, new bridges, and retaining walls through its entire length. The first segment connected with the existing operating light rail system at the South Hills Junction on its northern end and with the Castle Shannon Junction at its south- ern end. These operational challenges resulted in some design complexities. The service opened in June 2004. 64 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

Los Angeles Eastside Gold Line Los Angeles, CA Label: LA Gold East The eastside extension will provide transit access from the east side of Los Angeles to the regional Metro system. The 5.9-mile eastside extension of the Gold Line will be primarily at grade, with a 1.8-mile mid-section tunnel. It will originate at Union Station in downtown Los Angeles, where it connects to the Pasadena extension of the line and the heavy rail lines. The proj- ect alignment runs eastward along Alameda Street, 1st Street, and 3rd Street before terminating just before the intersection of Pomona and Atlantic Boulevards. This project was originally designed as a heavy rail line, but was altered to light rail because of funding constraints. Con- struction began in 2004 and revenue operation is scheduled to begin in late 2009. Phoenix Central/East Valley Light Rail Line Phoenix, AZ Label: Phoenix After some initial complications in the early planning phases, the City of Phoenix and Valley Metro Rail, Inc., a nonprofit public corporation in charge of the design, construction, and operation of the regional light rail system, partnered to construct a 19.6-mile, 27 station light rail system. The system’s alignment, located primarily in the street median from 19th Avenue and Bethany Home Road, starts in north central Phoenix and runs through the City of Tempe to the intersection of Main Street and Longmore in Mesa. The City of Phoenix entered into a Full Fund- ing Grant Agreement (FFGA) in January 2005, construction started the same month, and rev- enue operations began in December 2008. Portland South Corridor Portland, OR Label: Portland So The Tri-County Metropolitan Transportation District (TriMet) and Portland Metro, the region’s metropolitan planning organization, are constructing 8.3 miles of new light rail transit consisting of two segments connecting to the existing “MAX” LRT system along Interstate 84. The South Corridor Extension will provide a new rail line, “the Green Line,” from Clackamas Town Center to Portland State University (PSU). A portion of the Green Line will merge with and share 6.2 miles of the existing Blue Line along I-84 before continuing in the right-of-way of I-205 from the Gateway/NE 99th Avenue Transit Center to a new rail transit center at the Clacka- mas Town Center. The I-205 alignment is 6.5 miles of double-tracked and at-grade line with sev- eral grade-separated roadway crossings. The alignment in downtown Portland will run along the North-South Transit Mall Portland Union Station to the PSU campus while providing connec- tivity to the Red Line. The project includes 8 bi-directional stations for the I-205 segment and 14 unidirectional stations along the downtown Portland Mall alignment, with 7 on each leg of the one-way loop. Revenue operation is scheduled to begin in September 2009. Seattle Central Link Light Rail Project Seattle, WA Label: Seattle Cen Central Puget Sound Regional Transit Authority (Sound Transit) is constructing a 13.9-mile double-track light rail system for the initial segment of the Central Link Light Rail transit proj- ect. This segment is scheduled to open for revenue operations in July 2009. Its alignment runs from Westlake Center station through downtown Seattle to the Tukwila International Boulevard station. The system will use the existing 1.3-mile Downtown Seattle Transit Tunnel (DSTT), a new 1-mile long Beacon Hill tunnel, and a new 0.1-mile tunnel used for crossover and turnback Project Names and Descriptions in As-Built Analysis 65

operations. The scope of work includes 7 new stations, the renovation of 4 stations in the DSTT, a maintenance and operations facility, and a park-and-ride lot at the Tukwila International Blvd. station. A 1.7-mile extension to the Seattle-Tacoma Airport is scheduled to open in late 2009. Pittsburgh Northshore Light Rail Connector Pittsburgh, PA Label: Pittsburgh N The Port Authority of Allegheny County (Port Authority) is constructing a 1.2-mile double- tracked light rail transit extension from the existing Gateway terminus station in the Golden Triangle area of downtown Pittsburgh across the Allegheny River to the rapidly developing North Shore area. While remaining underground along the North Shore, the alignment travels adjacent to Bill Mazeroski Way accessing a station near the PNC Park stadium. The alignment con- tinues below grade adjacent to Reedsdale Street and transitions to an elevated alignment near Art Rooney Avenue to a station along Allegheny Avenue, near the Heinz Field stadium, before termi- nating near the West End Bridge. The project includes two bored tunnels below the Allegheny River and 3 newly constructed stations, and includes a new Gateway Station that will be constructed adjacent to the current Gateway Station to facilitate the tie-in to the existing system. The first North Shore station (North Side Station) will be located underground, and the terminus at Allegheny Station will be aerial. Charlotte South Corridor Charlotte, NC Label: Charlotte The Charlotte Area Transit System (CATS) and the City of Charlotte managed the construc- tion of a 9.6-mile and 15-station light rail transit line from the city’s central business district (CBD) to I-485 in south Mecklenburg County. A 3.7-mile portion of the system—between the CBD and the Scaleybark Road station—operates in an abandoned Norfolk Southern Railroad right-of-way owned by the City of Charlotte. The remainder of the operating service (5.9 miles) runs on separate tracks parallel to this right-of-way. The single-line system opened for revenue service in 2007. This project’s planning process encountered some difficulties when a redesign was required to meet FTA’s cost-effectiveness threshold and other requirements. Construction was stalled because CATS had to remove from the railroad right-of-way a species of flower listed as endan- gered under the provisions of the Endangered Species Act. VTA Tasman West San Jose, CA Label: VTA Tas W The VTA Tasman West construction project was the first leg of the Tasman Light Rail Proj- ect. The entire project was originally planned as a 12.4-mile expansion of an existing line; how- ever, funding constraints forced the VTA to scale back immediate construction to a 7.6-mile Tasman West segment that opened for revenue service in December 1999. This project had an extensive public outreach and involvement process. VTA Tasman East San Jose, CA Label: VTA Tas E The Tasman East Project was a 4.9-mile light rail extension from the existing San Jose Guadalupe corridor Baypointe station to the Hostetter station. The alignment runs along Tasman Drive from North First Street to I-880 and then follows the Great Mall Parkway and 66 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

Capitol Avenue. Phase I construction from the Baypointe Transfer station to the I-880/Milpitas station aligned the track for 1.9 miles in the median of Tasman Drive between the Baypointe Parkway and Alder Drive to the I-880 in Malpitas. It includes 3 new stations and opened for revenue service in May 2001. The second segment was a 3-mile extension in the median of Capitol Avenue between Alder Drive to just south of Hostetter Road. Both 4 new stations and a 7,200-ft bridge for grade separation were completed in June 2004. This project had an extensive public outreach and involvement process. VTA Capitol Segment San Jose, CA Label: VTA Capitol The Capitol Light Rail Project was a 3.3-mile light rail extension of the Tasman East Project that continued the alignment in the median of Capitol Avenue to extend service to the present terminus just south of Alum Rock Avenue. It opened simultaneously with Tasman East II for revenue operations in June 2004. This project had an extensive public outreach and involvement process. VTA Vasona Segment San Jose, CA Label: VTA Vasona The Vasona Light Rail Project is a 5.3-mile light rail extension from downtown San Jose to the Winchester Transit Center. The project added 8 new stations between Woz Way in downtown San Jose and Winchester Station in Campbell. The Vasona Light Rail operates primarily on the existing Union Pacific Railroad right-of-way between the San Jose Diridon Station and Winchester Station. Additionally, the segment between the San Fernando and San Jose Diridon Stations is in a tunnel, and the segment between Bascom Avenue and Route 17 bridges over Hamilton Avenue. This project had an extensive public outreach and involvement process and opened for revenue operations in October 2005. Atlanta MARTA North-South Line Atlanta, GA Label: MARTA N-S The MARTA North-South Line project included a 22.2-mile heavy rail line from the Hartsfield- Jackson Atlanta International Airport to the Doraville station south of the I-285 Beltway in northeast Atlanta. The alignment runs up Main Street to the Arthur Langford Parkway where it continues on Lee Street SW. The alignment veers east onto W. Whitehall Street SW just south of downtown Atlanta. In downtown the line runs underneath Peachtree Street and follows a railroad right-of-way after the Arts Center station to its northeastern terminus. The complete 18-station heavy rail line became operational in 1992. Atlanta MARTA North Line Dunwoody Extension Atlanta, GA Label: MARTA Dun The Dunwoody extension project created a spur line off the North-South line’s alignment. It opened for revenue service in 1996. This line and the North-South line are co-aligned from the Airport to the Lindbergh Center station. Its alignment is a 7-mile spur line off of the North-South alignment splitting off north of the Lindbergh Center station and runs to the Dunwoody station north of the I-285 Beltway. The alignment parallels Georgia State Route 400 between the Buckhead and Medical Center stations and ends at Dunwoody between Route 400 and I-295. Project Names and Descriptions in As-Built Analysis 67

Boston MBTA Orange Line Boston, MA Label: MBTA Orang After anti-highway protests stalled the construction of a freeway into downtown Boston through the Southwest Corridor, the Massachusetts Bay Transportation Authority (MBTA) con- structed a heavy rail line through the corridor. This double-tracked 4.7-mile line extended and rerouted the Orange Line south of the Chinatown station from the former Washington Street Elevated to the Southwest Corridor right-of-way. The Southwest Corridor alignment runs pri- marily below grade, with some portions in open-cut and other portions in subway, and primar- ily serves Boston’s South End, Roxbury, and Jamaica Plain neighborhoods. Baltimore MTA Metro Sections A and B Baltimore, MD Label: Baltimore The Maryland Transit Administration (MTA) built a 15-mile, 12-station heavy rail line in two phases. The first phase of construction built the line from the Charles Center station in down- town Baltimore to the Reisterstown Plaza station in the northwest section of the city along Eutaw Street, Pennsylvania Avenue and briefly on Reisterstown Road before re-emerging at grade in the Western Maryland Railroad (WMR) right-of-way adjacent to Wabash Avenue. Revenue service began in 1983 along this 9-station line. A 3-station extension, which continues in the WMR right-of-way and the I-795 median to the current western terminus at Owings Mills in Baltimore County, opened for revenue service in 1987. Chicago CTA—Southwest Orange Line Chicago, IL Label: CTA Orange The CTA Orange Line, the first rapid transit line to operate in southwest Chicago, runs 9.0 miles (double-tracked) from the downtown loop to its terminus at Midway Airport (eight stations) along freight rights-of-way. Approximately 2.7 miles of the fixed guideway is aerial structure, and the remaining 6.3 miles is on embankment. It connects the neighborhoods of Burbank, Bedford Park, Bridgeview, Hometown, Justice, Merrionette Park, Oak Lawn, and Summit to the downtown Chicago loop and connections with the other five heavy rail lines. The line opened for revenue service in 1993. Chicago CTA—O’Hare Extension Blue Line Chicago, IL Label: CTA O’Hare The O’Hare project extended the Blue Line (formerly called the Milwaukee Line) within the median of the Kennedy Expressway in northwest Chicago. Construction began in the early 1980s to extend the line 7.1 miles with 4 new stations from the previous terminus at the Jefferson Park station to the present terminus at the O’Hare Airport station. Revenue service to the Rosemont station began in 1983 and to O’Hare in September 1984. Chicago CTA—Ravenswood Brown Line Rehabilitation Chicago, IL Label: CTA Brown Persistent crowding on the Brown Line platforms prompted the Chicago Transit Authority (CTA) to begin reconstructing existing platforms and stations to accommodate eight-car trains, along with other related capital improvements The Ravenswood (Brown) Line extends approximately 9.1 miles with 18 stations from the Kimball Terminal on the north side of Chicago through the “Loop Elevated” section in down- 68 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

town Chicago. The majority of the heavy rail line operates on an elevated structure (8.0 miles), except for a portion near the northern end of the line that operates at grade (1.1 miles). The proj- ect began in late 2004 and is under construction. As of March 2009, 16 of 18 station project ren- ovations have been completed. Chicago CTA—Douglas Blue Line Rehabilitation Chicago, IL Label: CTA Douglas The Chicago Transit Authority reconstructed 5.6 miles of the Douglas Branch, then a portion of the Blue Line (now operated as the Pink Line). The heavy rail line extends from the Clinton station, to the west of downtown Chicago, to its terminus at the 54th St./Cermak Avenue sta- tion. The project required the reconstruction and rehabilitation of 11 stations, aerial structures, upgrading power distribution and signal systems, and the reconstruction of the 54th Street main- tenance yard. The rehabilitation project was completed on schedule and the line opened to rev- enue operation in January 2005. Los Angeles Red Line Los Angeles, CA Labels: LA Red 1, LA Red 2, LA Red 3 The Red Line is a heavy rail line in Los Angeles between Union Station and North Hollywood. This line opened for revenue service in three phases between 1993 and 2000. The line includes a 3.4-mile segment of underground guideway from Union station to Westlake/MacArthur Park station. Miami-Dade Transit Metrorail Miami, FL Label: Miami Miami-Dade Metrorail built a 21-mile elevated rapid transit line with 21 stations in the early 1980s. Most of the heavy rail line operates on an aerial structure. This rapid transit line opened for revenue service in May 1984. San Juan Tren Urbano San Juan, PR Label: San Juan The Puerto Rico Highway and Transportation Authority, a division of the Puerto Rico Depart- ment of Transportation and Public Works, constructed a 10.7-mile (17.2-km) double-track heavy rail system between Bayamón Centro and the Sagrado Corazon area of Santurce in San Juan. The entire project includes 5.7 miles (9.3 km) of aerial structures and a 0.8-mile (1.4-km) tunnel. When the existing publico service was incorporated into the project during planning phases, ridership requirements increased and the design sequence changed, which impacted the project’s budget. Approximately 40% of the alignment is at grade or near at grade. Aside from a short below-grade segment in the Centro Medico area, and an underground segment through Rio Piedras, the remainder is elevated track. The project includes 16 stations, 74 vehicles, and a maintenance/storage facility. The project opened for revenue service in June 2005. Bay Area Rapid Transit San Francisco Airport Extension San Francisco, CA Label: BART SFO After an extended planning process (the project’s original EIS occurred in 1985), BART and San Mateo County Transit District (SamTrans) completed a rail extension in 2003. BART and Project Names and Descriptions in As-Built Analysis 69

SamTrans completed this 8.7-mile double track, 4-station, heavy rail extension that runs from the Colma station through the cities of Colma, South San Francisco, and San Bruno along the Cal- train right-of-way to Millbrae. Approximately 1.5 miles north of the Millbrae Avenue intermodal terminal, an east-west aerial “Y” stub branches to the east to service the San Francisco International Airport (SFO). Because this project extended BART service beyond the existing five counties in BART’s service area, the project involved coordination with San Mateo County, including the ex- ecution of an agreement for the county to fund East Bay projects and to share the operating sub- sidy. With the support of the airport, the project sponsor was BART and the principal funding sources were San Mateo County and FTA. The extension opened for service in June 2003. WMATA—Shady Grove Extension (A Route) Washington, DC Label: DC Shady G The Shady Grove (A Route) construction project added 15 stations along 18 miles of heavy rail alignment in the District of Columbia and Montgomery County, MD. This segment of the Red Line extends from the Farragut North station to the present terminus at the Shady Grove station. Revenue service on the Shady Grove extension began in January 1977 with the opening of the Dupont Circle station. Revenue operations to the Van Ness-UDC station began in December 1981 and the full extension opened in December 1984. WMATA—Glenmont Extension (B Route) Washington, DC Labels: DC Glenmt 1, DC Glenmt 2 The Glenmont and Glenmont Outer (B Route) project was an 11.9-mile extension of WMATA’s heavy rail Red Line in northeastern Washington, DC, and eastern Montgomery County, MD. This extension starts in the B&O Railroad right-of-way with an above-grade cross- ing of U.S. Route 50 (Rhode Island Ave.) right after the Rhode Island Ave-Brentwood station. It continues at grade in the railroad right-of-way through Washington, DC, with grade-elevated crossings through the downtown of the Silver Spring, MD, suburb. It submerges south of the intersection of 16th and Georgia Avenue and continues underground beneath Georgia Avenue to the terminus at the Glenmont station. The extension first opened for revenue service to the Silver Spring station in February 1978 followed by the opening of service to Wheaton in 1990. The full extension began revenue operations in January 1998. WMATA—Huntington (C Route) Washington, DC Label: DC Huntgtn WMATA’s Huntington project included a 12.1-mile new heavy rail line (present-day Yellow Line) that opened for revenue in two phases. In the first phase, which opened in 1983, Yellow Line trains began operating across the Fenwick bridge over the Potomac River, and the Archives- Navy Memorial-Penn Quarter station opened. Later in 1983, the Yellow Line was extended south of Washington National Airport to its current terminus at Huntington. WMATA—New Carrollton (D Route) and Washington, DC Vienna (K Route) Labels: DC New Ca, DC Vienna The New Carrollton project included 11.8 miles of heavy rail and 14 stations, including a 5-station extension from the Stadium-Armory station in Southeast Washington to New Carrollton, 70 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

MD. When it opened for revenue service in November 1978, the line originated in suburban Maryland, ran through downtown Washington via Pennsylvania Avenue, D Street, 12th Street and I Street before passing under the Potomac River to the Rosslyn station in Arlington County, Virginia. The original service alignment terminated on the Virginia side at Washington National Airport. The Maryland portion of the alignment proceeds underground from the Stadium- Armory station. It continues above ground after crossing the Anacostia River and follows the Anacostia Freeway (DC 295) and US 50 corridors at grade and on elevated structure before terminating at the New Carrollton station, a major intermodal transfer center. The Vienna project opened eight new stations on the Orange Line in two phases. In 1979, underground stations on the Wilson Boulevard corridor in Arlington County opened between Rosslyn and Ballston stations. An extension to the Vienna station, which runs primarily at grade in the median of I-66, opened for revenue service in 1986. WMATA—Green Line (E, F Routes) Washington, DC Labels: DC U St., DC Greenblt, DC Anacost, DC Anacost O The Green Line opened for revenue service in several phases between May 1991 and January 2001. The initial Anacostia alignment to be built ran north and south of downtown Washington. The DC U St. Project included a 3-station, 1.65-mile northern section (the “Mid-City line”) that runs north underneath 7th Street NW from the Gallery Pl.-Chinatown station before turning west to the U Street/African-American Civil War Memorial/Cardozo station at 13th and U Streets NW. The 3-station, 4.3-mile southern section runs from L’Enfant Plaza along M Street before crossing under- neath the Anacostia River to reach the Anacostia station adjacent to Suitland Parkway. The north- ern section opened for revenue service in May 1991 with the full 6-station line opening for service in December 1991. In September 1999, with two additional stations, the full line was operational. The Outer Anacostia project extended the line 5 stations and 6.7 miles into southeast Wash- ington, DC, and Prince George’s County, MD. The alignment runs underground equidistant between Martin Luther King Jr. Avenue SE and Suitland Parkway to Southern Avenue. It runs at grade parallel with Southern Avenue in a northeastward direction, briefly submerges, and reappears above-grade at Branch Avenue and Naylor Road in Temple Hills, MD. It continues parallel to the Suitland Parkway before terminating east of Branch Avenue in Suitland, MD. Rev- enue operations on this extension commenced in January 2001. The 2-station, 2.3-mile Greenbelt extension from the Prince George’s Plaza station opened for revenue service in December 1993. WMATA—Addison (G Route) and Washington, DC Springfield Extensions (J, H Routes) Labels: DC Addison, DC Springfld The Addison project extended the Blue Line for 3.5 miles, adding 3 stations, from the previous terminus at the Stadium-Armory station to the Addison Road-Seat Pleasant station in Prince George’s County, MD. It continues east under E. Capitol St. NE and follows that major thorough- fare underground until that corridor becomes Central Avenue in Capitol Heights, MD. Revenue service commenced on this extension in November 1980. The 2-station extension to the present terminus at Largo Town Center opened up for revenue service in December 2004. The database costs do not reflect the latest extension to Largo Town Center. The Springfield project extended Blue Line service 3.5 miles from the King Street station to the present terminus at the Franconia-Springfield station. Service with the Yellow Line south of Project Names and Descriptions in As-Built Analysis 71

the National Airport station is shared to the King Street station. Blue Line revenue operations south of the National Airport began in June 1991 with the opening of the Van Dorn Street sta- tion. The full extension was opened for revenue service in June 1997. WMATA—L’Enfant Plaza (L Route) Washington, DC Label: DC L’Enfant The L’Enfant Plaza project included 1.71 miles connecting the L’Enfant Plaza station and the Pentagon Station via the 14th Street Bridge. This addition enabled service underground in Washington, DC, in what is today the Yellow Line via 7th Street NW. Philadelphia SEPTA Frankford Rehabilitation Philadelphia, PA Label: Phil Frankf. SEPTA began rebuilding the entire Frankford Elevated Line in 1986 with new track, signal sys- tems, and stations along a 5.25-mile span between Girard Avenue and Bridge Street. In addition to renovating 10 smaller stations, the project transformed the prior terminus into the larger modern intermodal Frankford Transportation Center (FTC) in northeast Philadelphia. The new FTC terminal building was opened on August 4, 2003. This project had an extensive public out- reach process. New York NYCT 63rd Street Tunnel New York, NY Label: NYCT 63rd The project included a two-level tunnel. The NYCT F rail service uses the upper level, connect- ing the IND Queens Boulevard Line in Queens to the IND Sixth Avenue Line in Manhattan via the IND 63rd Street Line. The lower level will be used by the Long Island Rail Road East Side Access project, which will bring LIRR commuter trains to Grand Central Terminal. The tunnel is con- structed with immersed tubes in trenches at the bottom of the East River bed. Beyond the river, the tunnel was built using cut-and-cover construction. The tunnel opened in October 1989. New York NYCT Stillwell Terminal Rehabilitation New York, NY Label: NYCT Stillw The New York Metropolitan Transportation Authority (MTA) completed the rehabilitation of its eight-track Stillwell Avenue Terminal station in Brooklyn, NY, in May of 2004. In addition to rehabilitating 90-year-old platforms, the project included a new triple-vaulted glass and steel structure with solar panels on the roof. This project had an extensive public outreach process, including the existing ridership on NYCT service as a significant stakeholder. 72 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

73 This appendix summarizes additional analysis of historical capital costs performed to support TCRP Project G-10. This appendix describes how the historical data was prepared and analyzed for cost relationships between soft costs and other project characteristics. C.1. Data Preparation and Standardization This analysis used actual historical capital cost data from two FTA Capital Cost Databases for light and heavy rail, respectively. This analysis took several steps to standardize and prepare the data in both databases for an accurate comparison. Most capital cost categories in the two data structures are similar, with minor exceptions. For example, vehicle costs are separated as their own category in both systems, although presented in a different numbering category sequence. Otherwise, the full capital costs to complete each project are represented in each dataset and these results are reflected in the analysis. C.2. Adjustments Addressing Different Cost Categorization This analysis combines light and heavy rail transit project capital cost databases using slightly different categorization structures for each mode. To correct for small variations in reporting protocols, the following modifications were made. • Project Initiation: Cost category 8.07 in the heavy rail database, Project Initiation, contains two sub-items for Mobilization and Maintenance of Traffic which are reported under 8.00 Soft Costs. The light rail dataset includes these items as SCC 40.073 and 40.074 under 40.000 Site- work and Special Conditions. To ensure comparability, the two heavy rail cost components were reclassified as an element of 40.073 and 40.074 of the Special Conditions category. • Planning and Feasibility Costs: Only a few projects reported these costs. This is for work that is typically carried out early in the initial phase of a transit project’s development lifecycle. These efforts are conducted prior to entry into the FTA New Starts Program and have been inconsis- tently documented at the project level. Therefore, FTA has eliminated these early efforts from the SCC structure. Transit agencies might assign these costs to general planning activities or other grants rather than a specific project budget, and FTA’s current SCC worksheet excludes planning costs incurred prior to FTA approval to enter preliminary engineering. To ensure com- parability, this cost category (8.01 for heavy rail and 80.090 for light rail) was omitted entirely. • Unallocated Contingency: The light rail dataset reports Category 90, Unallocated Contin- gency. However, since costs are final as-built expenditures, unallocated contingency is zero A P P E N D I X C Supplementary As-Built Cost Analysis

for all projects. The heavy rail dataset does not report contingencies. Therefore, this cost cat- egory has no impact on the analysis and was omitted. • Finance Charges: A small number of light rail projects report finance charges. However, these costs are largely a function of the financial structure and policies of the sponsor agency, and do not affect the relationship between project characteristics and construction-related soft costs. To ensure comparability, finance charges (8.08 for heavy rail and 100.00 for light rail) were omitted entirely from the analysis. These steps help to ensure that this technical analysis is based on a uniformly reported dataset for both light and heavy rail construction projects. C.3. Adjustment for Inflation and Nationalization The soft cost analysis adjusted all costs for inflation and local price differentials, and expresses nominal costs in U.S. 2008 dollars. The historical cost index from Means Construction Cost Index (Murphy, 2008) was applied to inflate all costs to the study base year of 2008. Differences in local metropolitan area labor, equipment, and material costs were adjusted to U.S. average 2008 dol- lars based on the Means Construction Cost Index (Murphy, 2008) for the 38 largest U.S. metro- politan areas. For example, cost of labor was less expensive in Charlotte than New York, so this analysis factors base-year dollars up in Charlotte and down in New York to the average nation- wide value. Each cost amount is also associated with a year of expenditure corresponding to the midpoint of the individual element expenditure. C.4. Outliers Omitted Some inconsistencies in the data appear to be a result of conflicting cost reporting or inter- pretation of the cost element definitions. These projects were omitted because they were consid- ered as non-representative outliers or as reflecting incomplete data. For example, the Chicago Transit Authority Brown Line/Ravenswood Rehabilitation project overhauled an existing rapid transit line and built only minimal new guideway; therefore this project does not offer a consis- tent cost basis to express the project costs on a per-linear-foot basis and compare that with the other projects in the database. In other project cases, while the overall soft costs total was in the reasonable range and could be used, the breakdown by individual soft cost element was not and that project was withdrawn from the more detailed analyses. For example, some projects reported zero costs for an individ- ual soft cost component such as preliminary engineering or final design, but the overall soft cost value was in the reasonable range. Therefore, the total soft cost was used, but the cost analysis at the component level was not used. Finally, not all detailed information on project schedule was always available. Wherever data was considered incomplete, questionable, or incomparable to other projects, these projects were omitted from the analysis in situations where appropriate. Table 17 below shows the resulting sample size from removing outliers or incomplete data points. C.5. Vehicle Soft Costs This analysis sought to determine if any soft costs were reported in a category outside of SCC 80. FTA instructions for reporting project costs within the Standard Cost Categories guide grantees to report professional services related to vehicle procurement under SCC 70 Vehicles, not the general soft cost category (SCC 80 Professional Services). However, the strict separation 74 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects

Supplementary As-Built Cost Analysis 75 of soft costs for vehicles from other soft costs may not hold consistently across the dataset. (Section C.1 discusses this potential shortfall.) Therefore, several figures below test for the pos- sibility that vehicle soft costs are included in the directed vehicle-specific category and not the overall soft costs category. Establishing the clear use of these related terms (soft costs generally and vehicle soft costs) is an important step in evaluating soft costs and developing a soft cost guidebook. Figure 35 shows the effect of vehicle costs on soft costs as percent of construction. If vehicle soft costs are included mistakenly in overall soft costs, one would expect to see that bigger vehicle purchases cause soft costs as percentage of construction to rise if the underlying guideway construction remains the same. Many of these project cost summaries were collected before there was federal guidance for classifying capital costs into a consistent set of cost cate- gories. Indeed, the data included 59 projects sponsored by numerous different agencies across nearly 35 years of experience. Instead, however, Figure 35 shows that soft costs appeared mostly immune to changing levels of vehicle procurements—the trend was slightly downward in light rail, upward in heavy rail, and zero for both modes, and all correlations were statistically insignif- icant. This is a good indication that vehicle soft costs are not included or reflected within the gen- eral soft costs category (SCC 80). C.6. Soft Costs by Mode and Year Figure 36 expands on the analysis of soft costs by decade in Figure 25 by analyzing average soft costs by mode and decade. The pattern shown in Figure 25 of increasing soft costs over time may in part be the result of no light rail projects from the 1970s being included in the dataset. Data Analysis Type Sample Size All projects in dataset 59 All projects used for analysis 51 Soft costs per linear foot 45 Soft cost subcomponents (engineering, management, etc.) 48 Duration from planning/DEIS to construction 13 Duration from preliminary engineering to construction 13 Duration from construction to operations 12 Duration from preliminary engineering to operations 13 Project delay 15 Table 17. Resulting sample sizes for each project characteristic. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.04 t-Stat = -1.02 R2 = 0.00 t-Stat = 0.10 R2 =0 .00 t-Stat: 0.17 R2 = 0.00 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 30%20%10% Vehicle Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) R2 = 0.040% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 30%20%10% Vehicle Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) R2 = 0.000% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 30%20%10% Vehicle Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) Figure 35. Soft costs as a percentage of construction versus vehicle costs as a percentage of total other costs.

76 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Figure 36 confirms that heavy rail projects are primarily responsible for the pattern of rising soft costs over time. Soft costs for light rail projects have been stable over this same period. However, the higher soft cost percentages are related to light rail projects constructed in the 1980s, possi- bly by agencies developing their initial segments. Figure 37 disaggregates the data in Figure 36 further from decade to actual year of construc- tion. This analysis confirms that the overall correlation for all modes combined is statistically significant, but that heavy rail projects are primarily responsible for the pattern of rising soft costs over time. Although light rail projects show a limited correlation in the increasing relationship, heavy rail projects exhibit a stronger relationship in increasing soft costs over time. Note that midyear of expenditure represents the midpoint of all project expenditures, which is similar to, but not neces- sarily the midpoint of, physical construction. Figure 38, Figure 39, and Figure 40 present the same analysis as the two previous figures but fur- ther disaggregate the soft cost category into several groups of components: PE+FD, FD alone, and construction management and administration. The same overall relationship of rising soft cost per- centage of construction costs holds true, but the relationship is weak. The final design soft costs show a stronger relationship and the same increasing relationship over time for both modes com- bined. The soft costs incurred in construction phases (measured as a percentage of construction LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL Sample Size: 0 4 5 16 10 5 56 6 14 10 21 25.1% 35.2% 34.5% 21.4% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 1970s 1990s So ft Co st s (% of C on str uc tio n) 34.2% 34.6% 30.9% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 1970s 1980s 1980s 1980s1990s 2000s 2000s 2000s So ft Co st s (% of C on str uc tio n) 34.6%33.0% 27.7% 21.4% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 1970s 1990s So ft Co st s (% of C on str uc tio n) Figure 36. Soft costs as a percentage of construction by decade and mode. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.02 t-Stat = 0.71 R2 = 0.35 t-Stat = 3.628 R2 = 0.217 t-Stat: 3.69 R2 = 0.35 0% 10% 20% 30% 40% 50% 60% 1970 1980 1990 2000 2010 Midyear of Expenditure So ft Co st s (% of C on str uc tio n)R2 = 0.02 0% 10% 20% 30% 40% 50% 60% 1980 1990 20102000 Midyear of Expenditure So ft Co st s (% of C on str uc tio n) R2 = 0.22 0% 10% 20% 30% 40% 50% 60% 1970 1980 1990 2000 2010 Midyear of Expenditure So ft Co st s (% of C on str uc tio n) Figure 37. Soft costs as a percentage of construction versus midyear of expenditure.

Supplementary As-Built Cost Analysis 77 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.02 t-Stat = -0.68 R2 = 0.00 t-Stat = 0.273 R2 = 0.04 t-Stat: 1.44 R2 = 0.02 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1980 1990 2000 2010 1980 1990 2000 2010 Midyear of Expenditure PE +F D Co st s (% of C on str uc tio n) R2 = 0.00 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1970 Midyear of Expenditure PE +F D Co st s (% of C on str uc tio n) R2 = 0.04 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1970 1980 1990 2000 2010 Midyear of Expenditure PE +F D Co st s (% of C on str uc tio n) Figure 38. Preliminary engineering and final design costs as a percentage of construction versus midyear of expenditure. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.09 t-Stat = -1.42 R2 = 0.14 t-Stat = 1.917 R2 = 0.08 t-Stat: 1.93 R2 = 0.09 0% 5% 10% 15% 20% 25% 30% 1980 1990 2000 2010 Midyear of Expenditure Fi na l D es ig n (% of C on str uc tio n) R2 = 0.14 0% 5% 10% 15% 20% 25% 30% 1970 1980 1990 2000 20002010 2010 Midyear of Expenditure Fi na l D es ig n (% of C on str uc tio n) R2 = 0.08 0% 5% 10% 15% 20% 25% 30% 1970 1980 1990 Midyear of Expenditure Fi na l D es ig n (% of C on str uc tio n) Figure 39. Final design costs as a percentage of construction versus midyear of expenditure. Figure 40. Management and administration costs as a percentage of construction versus midyear of expenditure. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.04 t-Stat = 0.94 R2 = 0.32 t-Stat = 3.31 R2 = 0.20 t-Stat: 3.38 R2 = 0.04 0% 5% 10% 15% 20% 25% 30% 35% 1980 1990 2000 2010 Midyear of Expenditure M an ag em en t (% of C on str uc tio n) R2 = 0.32 0% 5% 10% 15% 20% 25% 30% 35% 1970 1980 1990 2000 2010 Midyear of Expenditure M an ag em en t (% of C on str uc tio n) R2 = 0.20 0% 5% 10% 15% 20% 25% 30% 35% 1970 1980 1990 2000 2010 Midyear of Expenditure M an ag em en t (% of C on str uc tio n)

78 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects costs) illustrates a higher increasing trend over this time period. This pattern is more prevalent and statistically significant for the heavy rail projects than for the light rail projects. These soft cost percentages of construction costs by project development phase figures are consistent with the findings of Figure 37. Soft costs of all kinds have risen since the 1970s, but the pattern is strongest in heavy rail projects. Causes of this trend may include increasingly stringent environmental or mitigation requirements, the trend from new construction toward extending existing rail lines, or changing institutional roles or construction management techniques. The opposite logic is also likely true and may have a greater impact on these results, although in the same direction. Many of the heavy rail projects started in the 1970s were extension proj- ects along already well-established networks and constructed by sponsoring organizations with significant engineering and design capability. Light rail projects, by contrast, were constructed at emerging agencies that had to contract and develop their engineering and design capabilities. The project development demands may have increased for all of the projects; the actual percent- age increase was relatively larger for the heavy rail agencies since they started from a lower soft cost percentage due to more limited learning curve effects. C.7. Soft Costs by Complexity: Overall Project Size Soft costs can generally be expected to rise with the technical complexity of the project. However, there are myriad ways to quantify complexity, and the choice of soft cost measure- ment may be important since construction costs can also generally be expected to rise with technical complexity. Figure 41 shows that soft cost percentage is not dependent on the total cost of the overall proj- ect. There is virtually no relationship or correlation of the soft cost percent of construction to the total project expense. This is consistent for each of the light and heavy rail modes and the total project database. In a similar vein, Figure 42 shows that soft costs do not depend on the total cost for the con- struction portion of the project either. As noted above, there was no soft cost percentage relation- ship with total project cost and also here with project construction costs. If anything, soft costs appear to decline as construction costs decline, suggesting some economies of scale in engineer- ing and management. The correlations, however, are not statistically significant. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 =0 .01 t-Stat = -0.43 R2 = 0.02 t-Stat = -0.66 R2 = 0.04 t-Stat: -1.47 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $0.0 $0.5 $1.0 $1.5 $2.0 Billions Total Proj. Cost (National 2008$) So ft Co st s (% of C on str uc tio n) R2 = 0.02 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $0 $2 $4 $6 Billions Total Proj. Cost (National 2008$) So ft Co st s (% of C on str uc tio n) R2 = 0.04 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $- $2 $4 $6 Billions Total Proj. Cost (National 2008$) So ft Co st s (% of C on str uc tio n) Figure 41. Soft costs as a percentage of construction versus overall project cost.

Supplementary As-Built Cost Analysis 79 Figure 43, Figure 44, and Figure 45 disaggregate the analysis in Figure 42 summarizing the soft costs by project development phase as defined earlier: preliminary engineering and final design, construction administration and management, and all other soft costs. Figure 43 presents the combined engineering and design phase costs as a percentage of total construction costs. These subsets combine the project development aspects of the engineering and design phases, the various development functions during the construction phase, and then all of the other supporting project development efforts. The light rail, heavy rail, and combined analysis show no relationship. Heavy rail projects are more complex, especially those with higher project costs. This greater complexity would predict a flat or slightly increasing soft cost percent- age of construction costs, yet the combined project database mixes these contrasting relation- ships with a slightly declining relationship with little statistical reliability. These results confirm that engineering and design costs as a percentage of construction cost do not consistently depend on the total cost of the overall construction project, other things being equal. Figure 44 presents construction phase soft costs as a percent of construction costs against the dollar value construction cost of a project. In light rail, these project administration and man- agement costs fall in percentage terms as the magnitude of the project grows; however, no sta- tistically significant pattern holds for heavy rail or the combined project database. This finding for light rail is consistent with the same pattern for final design costs and further supports the LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 =0 .05 t-Stat = -1.04 R2 = 0.04 t-Stat = -0.98 R2 = 0.07 t-Stat: -1.94 R2 = 0.05 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $0.0 $0.5 $1.0 $1.5 Billions Construction Cost (2008$) So ft Co st s (% of C on str uc tio n) R2 = 0.04 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $0 $1 $2 $3 Billions Construction Cost (2008$) So ft Co st s (% of C on str uc tio n) R2 = 0.07 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% $- $1 $2 $3 Billions Construction Cost (2008$) So ft Co st s (% of C on str uc tio n) Figure 42. Soft costs as a percentage of construction versus construction cost. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat =- 0.16 R2 = 0.02 t-Stat =0 .65 R2 = 0.01 t-Stat: = -0.78 R2 = 0.00 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% $0 $500 $1,000 $1,500 $2,000 Millions Construction Cost (2008$) PE + F D Co st s (% of Co ns tru ct io n) R 2 = 0.02 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% $0 $2,000 $4,000 $6,000 Millions Construction Cost (2008$) PE + F D Co st s (% of Co ns tru ct io n) R2 = 0.01 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% $- $2,000 $4,000 $6,000 Millions Construction Cost (2008$) PE + F D Co st s (% of Co ns tru ct io n) Figure 43. Preliminary engineering and final design costs as a percentage of construction versus construction cost.

80 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects hypothesis that light rail is less complex and therefore its soft costs do not scale up with construc- tion costs. Lastly, Figure 45 completes the analysis by measuring the relationship between dollar value construction cost and all other soft costs not explicitly accounted for in the engineering and con- struction phases. No relationship is shown, which indicates the relatively inconsistent makeup of other soft costs. The next refinement of soft costs is to examine the phase breakdown for the engineering and design phases into the preliminary engineering and final design phases. Figure 46 presents the preliminary engineering phase soft costs compared to overall construction costs. The prelimi- nary engineering phase suggests an increase in the soft cost percentage of construction cost with increasing construction costs for both modes, but since the relationship is not significant in sta- tistical terms, it is not clear that the relationship is not zero. Figure 47 presents the same analy- sis structure for the final design phase. The light rail analysis shows a more (but not profoundly) statistically significant decline in soft cost percentage with the increasing project construction cost. The heavy rail mode results are flat for the full range of construction costs, indicating that the increasing complexity of more expensive heavy rail projects requires greater soft cost resources through a consistent percentage of construction costs. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.18 t-Stat =- 2.13 R2 = 0.01 t-Stat =- 0.45 R2 = 0.04 t-Stat: = -1.46 R2 = 0.18 0% 5% 10% 15% 20% 25% $0 $500 $1,000 $1,500 Millions Construction Cost (2008$) Ad m in . C os ts (% of C on str uc tio n) R2 = 0.01 0% 5% 10% 15% 20% 25% $0 $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) Ad m in . C os ts (% of C on str uc tio n) R2 = 0.04 0% 5% 10% 15% 20% 25% $- $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) Ad m in . C os ts (% of C on str uc tio n) Figure 44. Project administration and construction management costs as a percentage of construction versus construction cost. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat =0 .09 R2 = 0.10 t-Stat =- 1.60 R2 = 0.06 t-Stat: = -1.67 R2 = 0.00 0% 5% 10% 15% 20% 25% $0 $500 $1,000 $1,500 Millions Construction Cost (2008$) O th er C os ts (% of C on str uc tio n) R2 = 0.10 0% 5% 10% 15% 20% 25% $0 $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) O th er C os ts (% of C on str uc tio n) R2 = 0.06 0% 5% 10% 15% 20% 25% $- $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) O th er C os ts (% of C on str uc tio n) Figure 45. Other soft costs as a percentage of construction versus construction cost.

Supplementary As-Built Cost Analysis 81 This analysis suggests economies of scale in light rail construction, primarily through the reduction of final design expenses, but the results are inconclusive. Heavy rail shows no such trend, nor does the pattern appear for preliminary engineering. An alternative explanation for the notion of economies of scale in light rail is that certain con- struction conditions, such as tunneling and bridging, cause overall construction costs to rise much faster than the design and engineering of these conditions. This would cause engineering costs as percent age of construction to decline, not because of economies of scale but because of the way soft costs are measured. The heavy rail analysis in Figure 47 may not show this pattern because of the complexity of heavy rail. This possibility is explored further in sections below. The preceding exhibits focused on project magnitude as a proxy for complexity, and have magnitude as overall costs and construction costs. Two alternative ways to measure project mag- nitude may be alignment length and number of stations, as the following figures explore. Figure 48 measures project magnitude by alignment length (linear feet of guideway) and shows only a weak and statistically insignificant correlation with percentage soft costs. No con- clusion can be drawn here. Number of stations also indicates overall project size. Locating and designing stations can present challenges to the project development process and could be factors influencing soft costs LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.08 t-Stat = 1.33 R2 = 0.03 t-Stat =0 .88 R2 = 0.06 t-Stat: = 1.73 R2 = 0.08 0% 5% 10% 15% 20% 25% $0 $500 $1,000 $1,500 Millions Construction Cost (2008$) PE C os ts (% of C on str uc tio n) R2 = 0.03 0% 5% 10% 15% 20% 25% $0 $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) PE C os ts (% of C on str uc tio n) R2 = 0.06 0% 5% 10% 15% 20% 25% $- $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) PE C os ts (% of C on str uc tio n) Figure 46. Preliminary engineering costs as a percentage of construction versus construction cost. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.06 t-Stat = -1.11 R2 = 0.00 t-Stat = -0.30 R2 = 0.07 t-Stat: = -1.91 R2 = 0.06 0% 5% 10% 15% 20% 25% $0 $500 $1,000 $1,500 Millions Construction Cost (2008$) FD C os ts (% of C on str uc tio n) R2 = 0.00 0% 5% 10% 15% 20% 25% $0 $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) FD C os ts (% of C on str uc tio n) R2 = 0.07 0% 5% 10% 15% 20% 25% $- $1,000 $2,000 $3,000 $4,000 Millions Construction Cost (2008$) FD C os ts (% of C on str uc tio n) Figure 47. Final design costs as a percentage of construction versus construction cost.

82 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects as the design and construction of stations require many professional services functions. Yet, as Figure 49 shows, soft costs do not appear to depend on the number of stations. Soft costs as a percentage of construction appear to decline somewhat weakly with a greater number of stations, but the relationship is not statistically significant. Stations do not appear to have any effect upon soft costs for either mode. Beyond the simple number of stations, their frequency may also drive technical complexity. Since stations and ancillary facilities (e.g., train control rooms) may require more engineering and design than non-station components, the hypothesis is that a higher mix of stations along the guideway may increase soft costs in percentage terms. Figure 50 compares the number of stations per 10,000 linear feet of guideway to soft costs but finds minimal correlation and no statistical significance. C.8. Soft Costs by Complexity: New versus Extension Whether a rail construction project consists of a new line or extends an existing line may influ- ence its soft costs. On the one hand, more professional services or agency staff time may be required, for example, to integrate a guideway extension with existing train control or traction LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = -0.08 R2 = 0.00 t-Stat = 0.09 R2 = 0.00 t-Stat: 0.42 R2 = 0.00 0% 10% 20% 30% 40% 50% 60% - 50 150100 Alignment Length (000 LF) So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% - 50 150100 Alignment Length (000 LF) So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% 0 50 100 150 Alignment Length (000 LF) So ft Co st s (% of C on str uc tio n) Figure 48. Soft costs as a percentage of construction versus constructed alignment length. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.01 t-Stat = -0.38 R2 = 0.13 t-Stat = 1.86 R2 = 0.03 t-Stat: = 1.18 R2 = 0.13 0% 10% 20% 30% 40% 50% 60% - 10 20 30 Stations PE + F D So ft Co st s (% of Co ns tru ct io n) R2 = 0.01 0% 10% 20% 30% 40% 50% 60% - 10 20 30 40 Stations PE + F D So ft Co st s (% of Co ns tru ct io n) R2 = 0.03 0% 10% 20% 30% 40% 50% 60% 0 10 20 30 40 Stations PE + F D So ft Co st s (% of Co ns tru ct io n) Figure 49. Soft costs as a percentage of construction versus station quantity.

Supplementary As-Built Cost Analysis 83 power systems. On the other hand, a transit agency undertaking an extension project may suggest that relatively experienced agency staff with the necessary expertise be involved, which could result in lower soft costs. Figure 51 shows average soft costs by mode and by project type (new/extension/rehabilitation). The four rehabilitation projects included in this dataset include: SEPTA Frankford, CTA Brown Line (Ravenswood), CTA Blue Line (Douglas), and NYCT Stillwell Terminal. For both modes, and for both measures, average soft costs do not seem to change whether the project is a new line or an extension. Unexpectedly, extensions, not new rail lines, incurred slightly higher average soft cost percentages. Rehabilitation projects had somewhat higher soft cost percentages, but this sample is limited to four heavy rail projects. These findings indicate that the provision of professional services may be slightly lower for the initiation of new lines than for the extension or rehabilitation of existing segments. Figure 52 breaks Figure 51 down into the soft cost components of engineering and design, and project administration and management. Engineering costs appear higher for new and exten- sion light rail projects, while heavy rail engineering costs are fairly consistent between new and extension projects. Heavy rail engineering and design costs are lower for rehabilitation projects. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = 0.25 R2 = 0.01 t-Stat = -0.34 R2 = 0.00 t-Stat: 0.44 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% Stations per 10,000 Linear Feet of Guideway So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% Stations per 10,000 Linear Feet of Guideway So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% - 2 4 6- 4 62 8 0 642 Stations per 10,000 Linear Feet of Guideway So ft Co st s (% of C on str uc tio n) Figure 50. Soft costs as a percentage of construction versus station density. Sample Size: 19 28 4 29% 36%31% 35% 26%27% 35%31% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% New Line Extension of Line Rehabilitation of Line So ft Co st s (% of C on str uc tio n) Light Rail Heavy Rail All Modes Figure 51. Soft costs as a percentage of construction by project type.

Soft costs for project administration and construction management are higher than engineering- related activities, as shown in the difference between the left and right panes of Figure 52. However, the difference attributable to projects being extensions or new construction appears negligible. C.9. Soft Costs by Complexity: Percentage of Guideway Not at Grade Figure 53 extends the examination of project complexity by focusing on preliminary engineer- ing and final design costs, and suggests a similar conclusion. Engineering costs in percentage terms do not appear to be influenced by the extent to which the alignment is not at grade. Light rail projects have a fairly consistent 15% soft cost percent of construction costs. Heavy rail proj- ects are about 10% to 15%, and the combined database is about 13%. Figure 54 examines the effect of alignment complexity on the construction management and project administration soft costs of a project. Similar to the above findings, the proportion of guideway not at grade does not appear to affect the soft costs as a percentage of construc- tion costs. 84 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects ENGINEERING (PE + FD) ADMINISTRATION & MANAGEMENT Sample Size: 19 28 4 19 428 15% 17% 14% 13% 19% 19% 15% 15% 0% 5% 10% 15% 20% 25% 30% New Line Extension of Line Rehabilitation of Line A dm in . & M gm t. (% of C on str uc tio n) Light Rail Heavy Rail All Modes 15% 9% 10% 12% 10% 12% 10% 11% 0% 5% 10% 15% 20% 25% 30% New Line Extension of Line Rehabilitation of Line PE +F D (% of C on str uc tio n) Figure 52. Subtotaled soft cost components as a percentage of construction by project type. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.02 t-Stat = 0.58 R2 = 0.02 t-Stat = -0.73 R2 = 0.067 t-Stat: -1.81 R2 = 0.02 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% 25% 50% 75% 100% % of Guideway Not At-Grade PE + F D Co st s (% of Co ns tru ct io n) R2 = 0.02 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% % of Guideway Not At-Grade PE + F D Co st s (% of Co ns tru ct io n) R2 = 0.07 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% % of Guideway Not At-Grade PE + F D Co st s (% of Co ns tru ct io n) 25% 50% 75% 100% 25% 50% 75% 100% Figure 53. Engineering soft costs as a percentage of construction versus percentage of guideway not at grade.

The analysis so far has defined “not at grade” to include aerial structures, underground cut and cover, underground tunnel, retained cut or fill, and built-up fill guideway. Vertical align- ment has been applied as proxy for project complexity. However, these three last alignment types (retained cut or fill, and built-up fill) can be designed and constructed with fairly standardized engineering and design requirements that are similar to at-grade alignments. Therefore, Fig- ure 55, Figure 56, and Figure 57 designate these alignment types as “at grade,” and re-examine the relationship between soft costs and project complexity. These three figures, then, include only aerial structure, underground cut and cover, and underground tunneling alignments as “not at grade.” Figure 55 is comparable to Figure 28 and produces similarly statistically insignificant findings. Light rail projects are nearly flat at about 39% soft costs as a percentage of construction costs. Heavy rail projects range from about 28% to about 33%. The combined project database is nearly flat at about 35% to 38% soft costs as a percent of construction costs. Figure 56 and Figure 57 are comparable to the analysis presented in Figure 53 and Figure 54 and are mostly inconclusive. In Figure 56, both rail modes and the combined project database result in a slightly decreasing trend in engineering and design soft costs as a percentage of con- struction with increasing alignment complexity. The results are mixed for Figure 57, where light Supplementary As-Built Cost Analysis 85 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = 0.25 R2 = 0.03 t-Stat = 0.87 R2 = 0.00 t-Stat: -0.38 R2 = 0.03 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% % of Guideway Not At-Grade Ad m in . C os ts (% of C on str uc tio n) R2 = 0.00 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% 25% 50% 75% 100% 25% 50% 75% 100% % of Guideway Not At-Grade Ad m in . C os ts (% of C on str uc tio n) R2 = 0.00 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 0% % of Guideway Not At-Grade Ad m in . C os ts (% of C on str uc tio n) 25% 50% 75% 100% Figure 54. Administration soft costs as a percentage of construction versus percentage of guideway not at grade. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = -0.03 R2 = 0.01 t-Stat = 0.60 R2 = 0.03 t-Stat: = -1.18 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 50% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 50% 100% 100% 100% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) R2 = 0.030% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 50% % of Guideway Not At-Grade So ft Co st s (% of C on str uc tio n) Figure 55. Soft costs as a percentage of construction versus percentage of guideway not at grade (retained cut and built-up fill designated as “at grade”).

rail shows a slight downward trend and heavy rail shows a slight upward trend, but the combined project database is flat and all relationships are not statistically significant. While the relationships are weak, there may be some decline in engineering soft cost percentage with increasing project complexity. The greater capital costs of these more complex alignments results in higher soft costs, even with a slight decline in the soft cost percentage. Combining the two modes produces a weak negative correlation, surprisingly suggesting that soft costs decline as more aerial and tunnel segments are built. C.10. Soft Costs by Complexity: Percentage of Guideway Below Grade Underground alignment segments introduce several unique costs that other alignment grades do not, particularly excavation and complex structures. This report so far has used per- centage of guideway not at grade as a proxy for complexity; however, the portion of guideway below grade may be a useful indicator of complexity as well. Tunneling and excavating may pro- 86 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.01 t-Stat = -0.35 R2 = 0.12 t-Stat = -1.74 R2 = 0.16 t-Stat: = -2.96 R2 = 0.12 0% 10% 20% 30% 40% 50% 60% 0% % of Guideway Not At-Grade PE + F D (% of C on str uc tio n)R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 0% 50% 100% 50% 100% 50% 100% % of Guideway Not At-Grade PE + F D (% of C on str uc tio n) R2 = 0.16 0% 10% 20% 30% 40% 50% 60% 0% % of Guideway Not At-Grade PE + F D (% of C on str uc tio n) Figure 56. Engineering soft costs as a percentage of construction versus percentage of guideway not at grade (retained cut and built-up fill designated as “at grade”). LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = -0.01 R2 = 0.03 t-Stat = 0.80 R2 = 0.00 t-Stat: = -0.45 R2 = 0.03 0% 10% 20% 30% 40% 50% 60% 0% 50% % of Guideway Not At-Grade Ad m in . (% of C on str uc tio n)R2 = 0.00 0% 10% 20% 30% 40% 50% 60% 0% 50% 100% 100% 100% % of Guideway Not At-Grade Ad m in . (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% 0% 50% % of Guideway Not At-Grade Ad m in . (% of C on str uc tio n) Figure 57. Administration soft costs as a percentage of construction versus percentage of guideway not at grade (retained cut and built-up fill designated as “at grade”).

duce a unique set of engineering and management requirements, separate from aerial or built- up fill structures, which might influence project soft costs. Figure 58 shows that the proportion of the alignment in tunnels (cut and cover or deep-bore) has a mixed effect on soft costs as a percentage of construction. Light rail projects showed a slight increase in soft cost percentages as percentage below grade increased, whereas heavy rail projects showed a slight decrease from 30% to 24% with higher proportions of below-grade guideway. The combined project database shows a decreasing trend as well. Figure 59 and Figure 60 present this same analysis, but focus solely on engineering and admin- istration soft costs, respectively. Figure 59 shows that engineering and design soft costs (preliminary engineering and final design) tend to be only slightly negatively correlated to the percentage of guideway below grade, but the pattern is only statistically significant among heavy rail projects. Figure 60 finds a simi- lar general trend for administrative soft costs, but the trend is statistically less significant. Finally, another view into project complexity and soft costs is presented in Figure 61, which examines the effect of guideway grade on soft costs per linear foot and finds a positive correlation that is statistically significant for light rail and both modes combined. This figure is presented on a logarithmic y-axis scale to more clearly illustrate the relationship. Figure 61 shows that more Supplementary As-Built Cost Analysis 87 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.01 t-Stat = 0.39 R2 = 0.10 t-Stat = -1.62 R2 = 0.07 t-Stat: -1.98 R2 = 0.10 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 25% 75% % of Guideway Below Grade So ft Co st s (% of C on str uc tio n) R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 25% 75%50% 100% 50% 100% 50% 100% % of Guideway Below Grade So ft Co st s (% of C on str uc tio n) R2 = 0.07 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 25% 75% % of Guideway Below Grade So ft Co st s (% of C on str uc tio n) Figure 58. Soft costs as a percentage of construction versus percentage of guideway below grade. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.03 t-Stat = 0.83 R2 = 0.25 t-Stat = -2.74 R2 = 0.10 t-Stat: -2.21 R2 = 0.25 0% 10% 20% 30% 40% 50% 60% 0% 25% 75% % of Guideway Below Grade PE + F D (% of C on str uc tio n) R2 = 0.03 0% 10% 20% 30% 40% 50% 60% 0% 25% 50% 75% 100% 50% 100% 50% 100% % of Guideway Below Grade PE + F D (% of C on str uc tio n) R2 = 0.10 0% 10% 20% 30% 40% 50% 60% 0% 25% 75% % of Guideway Not At-Grade PE + F D (% of C on str uc tio n) Figure 59. Engineering soft costs as a percentage of construction versus percentage of guideway below grade.

complex alignment profiles are consistently tied to higher soft costs per linear foot and that the relationship is statistically significant for light rail and the combined project database. The percentage of guideway not at grade or below grade therefore appears weakly related to soft costs when measured as a percentage of construction costs. When soft costs are measured in dollar terms per linear foot of guideway, however, a stronger relationship appears: more com- plex alignment profiles are tied to higher soft costs per linear foot. This finding suggests that more alignment below grade may be driving capital costs in all categories, so that soft costs will rise in dollar value terms but remain unchanged in percentage terms. C.11. Relationships Among Other Category Unit Costs Although it is tempting to measure soft costs in dollar value terms because this measure pro- duces more correlation with expected complexity variables, it is worth exploring the measure further. One benefit of measuring soft costs in percentage terms is that the measure controls for variations in unit costs. Soft cost requirements of more expensive projects can be more consistently compared to inexpensive projects in percentage terms. Measuring soft costs in per-linear-foot 88 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.02 t-Stat = -0.72 R2 = 0.06 t-Stat = -1.23 R2 = 0.07 t-Stat: -1.87 R2 = 0.06 0% 10% 20% 30% 40% 50% 60% 0% 25% 75%75% % of Guideway Below Grade Ad m in . C os ts (% of C on str uc tio n) R2 = 0.02 0% 10% 20% 30% 40% 50% 60% 0% 25% 50% 50% 50%100% 100% 100% % of Guideway Below Grade Ad m in . C os ts (% of C on str uc tio n) R2 = 0.07 0% 10% 20% 30% 40% 50% 60% 0% 25% 75% % of Guideway Not At-Grade Ad m in . C os ts (% of C on str uc tio n) Figure 60. Administration soft costs as a percentage of construction versus percentage of guideway below grade. R² = 0.41 R² = 0.03 R² = 0.27 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.41 t-Stat = 3.87 R2 = 0.03 t-Stat = 0.76 R2 = 0.27 t-Stat: = 3.94 $100 $1,000 $10,000 $100,000 0% 25% 50% 100% So ft Co st s pe r L in ea r F oo t ( 20 08 $) Percent Below Grade $100 $1,000 $10,000 $100,000 0% 25% 75%75% 100% 100% So ft Co st s pe r L in ea r F oo t ( 20 08 $) Percent Below Grade $100 $1,000 $10,000 $100,000 0% 25% 50%50% 75% So ft Co st s pe r L in ea r F oo t ( 20 08 $) Percent Below Grade Figure 61. Soft costs per linear foot versus percentage of guideway below grade.

terms risks autocorrelation between unit costs—high soft costs could be correlated with higher other costs. In general, the analyses below tend to confirm this hypothesis: in dollar terms, soft costs and most cost categories tend to increase proportionately to construction costs. Figure 62 shows that right-of-way costs grow along with sitework and special conditions costs. The relationship is weak, but this finding mildly supports the hypothesis that all categories of capital costs may be growing together, which may help explain the previous results showing that soft costs grow in dollar value, but not percentage terms in relation to complexity (i.e., in terms of percent of alignment not at grade or below grade). Another perspective on the relationships between these soft cost categories is the relationship of guideway costs to right-of-way costs. As shown in Figure 63, these two cost categories appear to be correlated, similar to Figure 30 and Figure 62. The statistical significance is not as pronounced, but the relationship is clear: as right-of-way costs increase, guideway construction costs are also shown to increase. This correlation is best demonstrated for light rail, and the near-zero intercept makes intuitive sense. The heavy rail correlation is statistically insignificant but directionally consistent with light rail. The combined project database also shows a statistical relationship. Supplementary As-Built Cost Analysis 89 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.12 t-Stat = 1.75 R2 = 0.14 t-Stat =1.759 R2 = 0.16 t-Stat: 2.88 R2 = 0.12 $- $1 $2 $3 $4 $5 $6 $7 $8 $- $1 $2 $3 $4 ROW Cost (2008$) (000) per LF Si te w or k & Sp ec ia l C on di tio ns Co st s (00 0) pe r L ine ar Fo ot (2 00 8$ ) R2 = 0.14 $- $1 $2 $3 $4 $5 $6 $7 $0 $2 $4 $6 ROW Cost (2008$) (000) per LF Si te w or k & Sp ec ia l C on di tio ns Co st s (00 0) pe r L ine ar Fo ot (2 00 8$ ) R2 = 0.16 $- $1 $2 $3 $4 $5 $6 $7 $8 $0 $2 $4 $6 ROW Cost (2008$) (000) per LF Si te w or k & Sp ec ia l C on di tio ns Co st s (00 0) pe r L ine ar Fo ot (2 00 8$ ) Figure 62. Sitework and special conditions costs per linear foot versus right-of-way costs per linear foot. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.33 t-Stat = 3.27 R2 = 0.04 t-Stat = 0.90 R2 = 0.14 t-Stat: 2.66 R2 = 0.33 $- $5 $10 $15 $20 $25 $30 $- $1 $2 $3 $4 ROW Cost (2008$) (000) per LF G ui de w ay C os ts (0 00 ) ( 20 08 $) pe r L F R2 = 0.04 $- $5 $10 $15 $20 $25 $30 $0 $2 $4 $6 ROW Cost (2008$) (000) per LF G ui de w ay C os ts (0 00 ) ( 20 08 $) pe r L F R2 = 0.14 $- $5 $10 $15 $20 $25 $30 $0 $2 $4 $6 ROW Cost (2008$) (000) per LF G ui de w ay C os ts (0 00 ) ( 20 08 $) pe r L F Figure 63. Guideway construction costs per linear foot versus right of way cost per linear foot.

In general, Figure 62 and Figure 63 (and Figure 30 in Section 4.5.5) show that all categories of capital costs tend to grow together. These three figures help explain why the data show that soft costs rise in dollar terms but not percentage terms when plotted against expected complexity vari- ables such as alignment profile. To further support this point, Figure 64 can be compared to Fig- ure 30. Both display the same variables on the x- and y-axes; however, Figure 64 measures soft costs as a percentage of construction costs whereas Figure 30 measures these in dollar value terms. When one variable is expressed in percentage terms, as in Figure 64, the correlation is non-existent. The preceding figures demonstrate that despite the relatively stronger cost relationships pro- duced by measuring soft costs in dollar terms, such a measurement may not provide an accurate understanding of the changing relationship between soft costs and other project characteristics. Indicators of project complexity are correlated with higher soft costs in dollar terms, and with higher costs in all categories. C.12. Soft Costs by Complexity: Right-of-Way Costs Right-of-way costs, which are primarily the cost to acquire real estate and relocate existing res- idences and businesses, appear to be mildly related to soft costs as a percentage of construction costs. High expenditures to acquire real estate and relocate land uses may be correlated with proj- ects in more dense, urban areas where soft costs might be relatively high in proportion to the construction budget. Figure 65 compares soft costs as a percentage of construction cost to right- of-way costs and shows that right-of-way costs as a percentage of total costs appear to explain a small amount of soft cost variation. Figure 66, however, shows that ROW costs per linear foot are not correlated with soft cost per- centages. These relationships indicate that soft cost percentages do not change significantly as right-of-way costs increase per linear foot. C.13. Soft Costs and Project Development Budget As a project is developed through the planning and design phases, its budgeted cost is likely to change as the project is further defined. Similarly, a project can face cost overruns during con- struction phases due to a variety of factors such as unforeseen subsurface conditions, inaccurate 90 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.12 t-Stat = -1.72 R2 = 0.37 t-Stat = -3.32 R2 = 0.30 t-Stat: = -4.27 R2 = 0.12 10% 100% $1,000 $10,000 $100,000 Construction Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n) R2 = 0.37 10% 100% $1,000 $10,000 Construction Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n) R2 = 0.30 10% 100% $1,000 $10,000 $100,000 Construction Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n)$100,000 Figure 64. Soft costs as a percentage of construction versus construction costs per linear foot on a logarithmic scale.

preliminary estimates, and unexpectedly high bids from contractors. To explore the potential effects of early budget estimates on actual soft cost expenditures, this report used data from the report from TCRP Project G-07 (Booz Allen Hamilton Inc., 2005). This data, provided for 22 projects in the original database, is summarized in Table 18. Figure 67 graphs the data above (outliers removed) and show that budget overruns have lit- tle impact on a project’s final proportion of soft costs. Cost overruns were measured by dividing the actual as-built cost by the total project cost as it was estimated during the preliminary engi- neering phase. The outlier shown with significant cost overruns is the Tren Urbano project in San Juan, whose project requirements and design sequence changed substantially during project development, impacting the budget of the project. Soft costs as a percentage of construction decline slightly as the projects increase in cost esca- lation, but this trend is not statistically significant. This slight decline was not evident for con- struction phase project administration costs. This pattern is consistent with the previous figures: since soft costs may tend to grow in relation to other project cost categories, cost overruns have Supplementary As-Built Cost Analysis 91 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.01 t-Stat = 0.40 R2 = 0.10 t-Stat = 1.645 R2 = 0.07 t-Stat: 1.87 R2 = 0.10 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% ROW Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% ROW Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) R2 = 0.07 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% ROW Costs as % of Total Cost So ft Co st s (% of C on str uc tio n) Figure 65. Soft costs as a percentage of construction with right of way costs as a percentage of total cost. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL R2 = 0.00 t-Stat = -0.18 R2 = 0.00 t-Stat = 0.25 R2 = 0.00 t-Stat: -0.45 R2 = 0.00 0% 10% 20% 30% 40% 50% 60% $- $1 $2 $3 $4 Thousands ROW Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% $0 $2 $4 $6 Thousands ROW Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% $0 $2 $4 $6 Thousands ROW Cost (2008$) per Linear Foot So ft Co st s (% of C on str uc tio n) Figure 66. Soft costs as a percentage of construction with right of way costs per linear foot.

little impact on the relative proportion of soft costs. In short, dollar value costs tend to increase together for a given project, regardless of the characteristics of the project. C.14. Soft Costs and Project Development Schedule The length of time it takes to plan, design, and construct a rail transit project may impact soft cost expenditures, as may schedule delay during the project development process. As pre-construction project development phases extend, design costs and project management costs may tend to increase. In addition, delay from the original schedule may also increase soft costs, as certain soft costs continue to be incurred steadily through these schedule delays. 92 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Project Cost Estimated at PE As-Built Cost Actual Cost as % of PE Estimate Portland MAX Segment 1 $214.0 $246.8 115% Hudson-Bergen MOS-I $775.0 $1,113.0 144% Hiawatha Corridor $548.6 $715.3 130% Portland MAX Interstate $301.8 $349.4 116% St. Louis Clair County Extension $359.1 $336.5 94% Salt Lake North-South $261.3 $311.8 119% Portland MAX Westside/Hillsboro $913.0 $963.5 106% Pasadena Gold Line $803.8 $677.6 84% Denver Southwest Corridor $142.5 $175.0 123% Portland South Corridor $125.0 $127.0 102% VTA Tasman West $327.8 $280.6 86% VTA Tasman East $275.9 $276.2 100% VTA Capitol Segment $147.1 $162.5 110% VTA Vasona Segment $269.1 $316.8 118% MARTA Dunwoody Extension $438.9 $472.7 108% CTA Orange Line $496.0 $474.6 96% LA Red Line Segment 1 $914.4 $1,417.8 155% LA Red Line Segment 2 $1,446.4 $1,921.7 133% LA Red Line Segment 3 $1,310.8 $1,313.2 100% San Juan Tren Urbano $950.6 $2,250.0 237% BART SFO Extension $1,070.0 $1,550.2 145% NYCT 63rd Street Tunnel $537.9 $632.3 118% Note: all dollar amounts in year-of-expenditure dollars. Table 18. Project development budgetary database used. LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: ADMIN COSTS ONLY Sample Size: 16 R2 = 0.01 t-Stat = -0.31 Sample Size: 16 R2 = 0.01 t-Stat = 0.32 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 50% 100% 200%150% 250% 100% 200%150% 250% Actual Project Cost as % of Predicted Cost at PE So ft Co st s (% of C on str uc tio n) R2 = 0.01 0% 10% 20% 30% 40% 50% 60% 50% Actual Project Cost as % of Predicted Cost at PE M an ag em en t ( % of C on str uc tio n) Figure 67. Soft costs as a percentage of construction versus cost overruns.

To explore this potential, this report again turned to data provided from the report from TCRP Project G-07, Managing Capital Costs of Major Federally Funded Public Transportation Projects (Booz Allen Hamilton Inc., 2005). Table 19 shows the project schedule data used in this analysis. This data represents the year in which a project phase began, which is somewhat differ- ent from the midyear of expenditure used in other sections of this analysis. When data was not available for all project phases, or when phases appeared to be unreasonable, projects were omit- ted where appropriate. Resulting sample sizes are noted in the figures, as well as in Table 17. Figure 68 shows the effect of pre-construction duration (from Planning/DEIS to construction phases) on soft costs as a percentage of construction. Total soft costs are presented in the left pane, and engineering costs (preliminary engineering and final design) costs are presented in the right pane. Note that it may be difficult to identify a single year for the “Planning and DEIS” Supplementary As-Built Cost Analysis 93 Project Planning/ DEIS PE/FEIS Final Design Constr- uction Operation Portland MAX Segment 1 1980 1983 1984 1986 Hudson-Bergen MOS-I 1993 1996 1997 1999 2002 Hiawatha Corridor 1993 1999 2000 2001 2004 Portland MAX Interstate 1999 2000 2002 2004 2004 St. Louis Clair County Extension 1995 1998 1999 2001 2001 Salt Lake North-South 1994 1995 1998 1998 1999 Portland MAX Westside/Hillsboro 1990 1991 1994 1996 1998 Pasadena Gold Line 1993 1996 2000 2003 Denver Southwest Corridor 1992 1996 1997 1999 2000 Portland South Corridor 1995 1997 2001 VTA Tasman West 1992 1993 1996 1999 VTA Tasman East 1992 1995 1999 2001 VTA Capitol Segment 1999 2000 2004 VTA Vasona Segment 1999 2000 2005 MARTA Dunwoody Extension 1990 1991 1994 1998 2000 CTA Orange Line 1982 1984 1986 1990 1993 LA Red Line Segment 1 1983 1988 1989 LA Red Line Segment 2 1983 1990 1994 LA Red Line Segment 3 1983 1993 1998 San Juan Tren Urbano 1992 1995 1996 2002 2004 BART SFO Extension 1992 1996 1997 1998 2002 NYCT 63rd Street Tunnel 1989 1992 1994 1998 2001 Table 19. Project development schedule data used. LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: PE + FD COSTS ONLY Sample Size: 13 R2 = 0.01 t-Stat = 0.32 Sample Size: 13 R2 = 0.02 t-Stat = 0.46 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% - 2 5 7 9 10 11 Years Elapsed between Planning/DEIS and Construction So ft Co st s (% of C on str uc tio n) R2 = 0.02 0% 10% 20% 30% 40% 50% 60% - 21 1 5 73 6 8 3 6 84 4 9 10 11 Years Elapsed between Planning/DEIS and Construction PE + F D C os ts (% of C on str uc tio n) Figure 68. Soft costs as a percentage of construction versus years elapsed between completion of the draft environmental impact statement and construction.

phase for a project since the long-range planning process may be very different for each metro- politan area or agency. The correlation is positive, as expected, but the relationship is statistically insignificant. Measuring soft costs on a per-linear-foot basis, however, produces a stronger relationship, as shown in Figure 69. Total soft costs are presented in the left pane, and engineering costs (pre- liminary engineering and final design) costs are presented in the right pane. In the right pane, the results are pronounced from an x-axis intercept at four years toward a maximum range of about $20,000 per linear foot at about 15 years between the DEIS completion and construction. This relationship holds for engineering soft costs as well, as shown in the right pane. These findings seem to suggest that the duration of pre-construction phases should be con- sidered within the estimate of soft costs. However, the findings in Figure 69 may simply show that costly projects take longer to plan and design. Caution should be given due to the relatively small sample size (15) and the role of four relatively costly projects in this chart. Figure 70 measures the effect of a more narrowly defined pre-construction phase (PE/FEIS to construction) on soft costs, and shows insignificant findings. The relationship shows the correct direction of increasing soft cost percentage with increasing schedule duration but is statistically insignificant. The relative magnitude of soft costs, including engineering costs only, appears to be unaffected by the years elapsed between the preliminary engineering and construction phases. Figure 71 extends the above analysis to include the duration through construction all the way to operations, and finds similarly inconclusive results. Total soft costs are presented in the left pane, and construction management and administration costs are isolated in the right pane. Administration costs are shown here to test the hypothesis that construction and other admin- istration costs may be more likely to be affected by the duration of the construction phase. Although soft costs do tend to go up for lengthier projects in Figure 71, the relationship is not statistically significant. The duration of a project may not cause soft costs to increase as much as delay or deviation from a prior schedule. During a delay, if construction costs and project scope remain stable, but administration activities continue steadily, soft costs in relation to construction costs might increase. 94 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: PE + FD COSTS ONLY Sample Size: 11 R2 = 0.24 t-Stat = 1.69 Sample Size: 11 R2 = 0.22 t-Stat = 1.58 R2 = 0.24 $- $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $14,000 $16,000 $18,000 - 4 8 10 12 Years Elapsed between Planning/DEIS and Construction So ft Co st s pe r L F (20 08 $) R2 = 0.22 $- $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000 - 42 6 2 86 10 12 Years Elapsed between Planning/DEIS and Construction PE + F D C os ts p er L F (20 08 $) Figure 69. Soft costs per linear foot versus years elapsed between completion of the draft environmental impact statement and construction.

Figure 72 presents the variance between the project opening date projected during the prelim- inary engineering phase and the actual project opening date and compares this to soft costs as a percentage of construction. Presumably, a deviation from the opening date predicted during engineering phases represents a delay. Note that many projects in this dataset were not delayed at all (zero years), while two actually opened ahead of schedule. Figure 72 shows no strong relation- ship with years of delay and the proportion of soft costs. C.15. Vertical Profile and Soft Cost Measurement Somewhat surprisingly, this soft cost analysis found a relatively weak correlation between vertical profile (and by extension, project complexity) and a variety of soft costs measured as a percentage of construction costs. One possible explanation for this finding is that tunneling and aerial structures increase construction costs so rapidly that soft costs as a share of the project do not change measurably beyond the construction costs and increase the soft cost proportions. Supplementary As-Built Cost Analysis 95 LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: PE + FD COSTS ONLY Sample Size: 13 R2 = 0.06 t-Stat = 0.86 Sample Size: 13 R2 = 0.00 t-Stat = 0.15 R2 = 0.06 0% 10% 20% 30% 40% 50% 60% - 4 8 Years Elapsed between PE and Construction So ft Co st s (% of C on str uc tio n) R2 = 0.00 0% 10% 20% 30% 40% 50% 60% - 4 82 6 2 6 Years Elapsed between PE and Construction PE + F D C os ts (% of C on str uc tio n) Figure 70. Soft costs as a percentage of construction versus years elapsed between preliminary engineering and construction. LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: ADMIN COSTS ONLY Sample Size: 13 R2 = 0.00 t-Stat = -0.06 Sample Size: 13 R2 = 0.01 t-Stat = -0.39 R2 = 0.00 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% - 4 8 10 Years Elapsed between PE and Operations So ft Co st s (% of C on str uc tio n) R2 = 0.01 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% - 4 82 6 2 6 10 Years Elapsed between PE and Operations M an ag em en t ( % of C on str uc tio n) Figure 71. Soft costs as a percentage of construction versus years elapsed between preliminary engineering and operations.

Figure 73 and Figure 74 examine this potential explanation by comparing the rate of growth of soft and hard costs as the vertical profile becomes more complex. Soft and construction costs are on a per-linear-foot basis, and in every pane construction costs are shown as green diamonds, with a green dashed trend line. Figure 73 shows that for light and heavy rail and both modes combined, as more of the align- ment is in cut and cover and tunnels, construction costs rise faster than soft costs. Figure 74 shows a similar trend when alignment is simply not at grade, although the pattern is less strong. These trends affirm that when the alignment moves from at grade to more complex tunnel, bridge, or aerial structures, construction costs expand rapidly, sometimes faster than soft costs. C.16. Isolating Agency-Specific Effects One potential source of variance within the dataset used here is that financial and construc- tion management practices differ from agency to agency. Where one agency maintains construc- tion inspectors and managers on staff through the operating budget, another agency might 96 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects LIGHT + HEAVY RAIL: ALL SOFT COSTS LIGHT + HEAVY RAIL: ADMIN COSTS ONLY Sample Size: 15 R2 = 0.01 t-Stat = -0.43 Sample Size: 15 R2 = 0.09 t-Stat = -1.14 R2 = 0.01 0% 10% 20% 30% 40% 50% 60% (2) - 2 64 8 Years of Delay So ft Co st s (% of Co ns tru ct io n) R2 = 0.09 0% 5% 10% 15% 20% 25% 30% 35% (2) - 4 82 6 Years of Delay M an ag em en t ( % of C on str uc tio n) Figure 72. Soft costs as a percentage of construction versus years of delay in opening. LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL Soft Costs trend: R2 = 0.41 t-Stat = 3.87 R2 = 0.03 t-Stat = 0.76 R2 = 0.27 t-Stat: = 3.94 R2 = 0.41 R2 = 0.18 $- $5 $10 $15 $20 $25 $30 $35 $40 0% 25% 50% 100% 50% 100% 50% 100% % Guideway Below Grade Co st s pe r L F (00 0) R2 = 0.03 R2 = 0.19 $- $10 $20 $30 $40 $50 $60 0% 25% 75%75% % Guideway Below Grade Co st s pe r L F (00 0) R2 = 0.27 R2 = 0.38 $- $10 $20 $30 $40 $50 $60 0% 25% 75% % Guideway Below Grade Co st s pe r L F (00 0) Construction Costs Construction Costs Construction Costs Soft Costs Soft Costs Soft Costs Figure 73. Soft costs and construction costs per linear foot with percent of guideway below grade.

charge these employees to the project’s capital budget, for example. In another case, certain agen- cies collect internal staff, force account, and contractor staff into separate operating accounts and then allocate them back to specific projects. These differing approaches may have some impact on the soft cost amount used in this analysis. The philosophy or style of agency management might have just as much impact on soft costs as the alignment profile or number of stations. However, it is particularly difficult to control for agency-specific effects given the range of potential impacts and the generally small number of new rail construction projects per agency. The dataset contains twelve distinct projects for Washington, DC, presenting an opportunity to try to isolate agency-specific effects. Examining only projects constructed by WMATA means analyzing projects with very similar project development processes and cost allocation practices. WMATA has expanded its Metrorail system incrementally over the past four decades, with each extension or new line treated as a discrete project in the database. This section of the report restates some of the previous analysis for WMATA projects only. Figure 75 shows that soft costs for WMATA projects have been increasing in percentage terms over time. From the initial Metrorail segments completed in the 1970s through the projects com- pleted in the late 1990s, WMATA has seen an increasing trend in soft costs. The initial segments of Washington DC’s rail system had soft cost percentages of construction at about 20%. Through the rail extensions in the 1980s, soft cost percentages were mixed, with projects higher and lower than 20%, with a range from as low as 11% to as high as 38%. The three projects completed in the 1990s, however, had more consistent soft cost values of about 38%. Figure 76 examines the soft cost percentage of construction costs with the percent of project alignment not at grade. In contrast to the full database for heavy rail, WMATA projects suggest a declining trend in soft costs with more complex alignments. The heavy rail project database showed an increasing trend from 27% to 36% with a statistical relationship, but little correla- tion. The WMATA statistical relationship is also insignificant, but does result in a declining trend more similar to the combined project database. Analysis of the full project database suggests that this pattern may be caused by other project categories growing faster than soft costs for these alignment types. Figure 77 presents a similar analysis but is focused on only the proportion of alignment that is below grade for WMATA projects. The results of this WMATA analysis are similar to those Supplementary As-Built Cost Analysis 97 LIGHT RAIL HEAVY RAIL LIGHT + HEAVY RAIL Soft Costs trend: R2 = 0.35 t-Stat = 3.45 R2 = 0.01 t-Stat = 0.52 R2 = 0.36 t-Stat: = 4.88 R2 = 0.35 R2 = 0.12 $- $5 $10 $15 $20 $25 $30 $35 $40 $45 0% 25% 50% 75% 100% % Guideway Not At-Grade Co st s pe r L F (00 0) R2 = 0.01 R2 = 0.02 $- $10 $20 $30 $40 $50 $60 0% 25% 50% 50%75% % Guideway Not At-Grade Co st s pe r L F (00 0) R2 = 0.36 R2 = 0.28 $- $10 $20 $30 $40 $50 $60 0% 25% 75% 100%100% % Guideway Not At-Grade Co st s pe r L F (00 0)ConstructionCosts Construction Costs Construction Costs Soft Costs Soft Costs Soft Costs Figure 74. Soft costs and construction costs per linear foot with percentage of guideway not at grade.

98 Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects HEAVY RAIL R2 = 0.40 t-Stat = 2.69 R2 = 0.40 0% 10% 20% 30% 40% 50% 60% 1970 1975 19851980 1990 1995 2000 Midyear of Expenditure So ft Co st s (% of C on str uc tio n) Figure 75. Soft costs as a percentage of construction versus midyear of expenditure, WMATA only. HEAVY RAIL R2 = 0.05 t-Stat = -0.79 R2 = 0.05 0% 10% 20% 30% 40% 50% 60% 0% 20% 40% 60% 80% 100% Percent of Guideway not At Grade So ft Co st s (% of C on str uc tio n) Figure 76. Soft costs as a percentage of construction with percentage of guideway not at grade, WMATA only. HEAVY RAIL R2 = 0.08 t-Stat = -0.98 R2 = 0.08 0% 10% 20% 30% 40% 50% 60% 0% 20% 40% 60% 80% 100% Percent of Guideway Below Grade So ft Co st s (% of C on str uc tio n) Figure 77. Soft costs as a percentage of construction with percentage of guideway below grade, WMATA only.

from the full database of heavy rail projects. The trend line shows a declining trend from a high of 30% to a low of about 22%, but the relationship is not statistically significant. Note that some WMATA projects are 100% below grade. These results suggest that below-grade alignment has no effect on soft costs as a percentage of construction costs. These WMATA Metrorail results do not support the full heavy rail database, nor do they demonstrate consistent relationships that may be expected for projects from the same agency. The results from the preceding WMATA-only data demonstrate the difficulty in identifying project characteristics that can be used to help estimate construction soft costs of major public transportation capital projects. Supplementary As-Built Cost Analysis 99

Next: Abbreviations used without definitions in TRB publications »
Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Transit Cooperative Research Program (TCRP) Report 138: Estimating Soft Costs for Major Public Transportation Fixed Guideway Projects defines and describes soft costs and provides a new suggested methodology to estimate soft costs based on historical projects. The report also examines detailed technical information about the data collection, methodology, and statistical analysis that was used to develop the suggested methodology.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!