Forecasting Asset Conditions with Decay Curves April 16, 2012 Keith Gates, PE Senior Analyst, Strategic Planning & Analysis 9 th National Conference on Transportation Asset Management San Diego, California
FTA Capital Investment Needs Analysis Asset Data Asset types and quantities In-service dates Cost to replace Investment Policy Available Funding When to rehab, replace & expand Priorities Forecast Asset conditions Reinvestment backlog & ongoing needs $90 $80 $70 $60 $50 $40 $30 $20 $10 $0 SGR Backlog & Annual Outlays SGR 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 $Billions $160 $140 $120 $100 $80 $60 $40 $20 $0 Distribution of Asset Physical Conditions by Asset Type Guideway Elements Facilities Systems Stations Vehicles 5. Excellent 4. Good 3. Adequate 2. Marginal 1. Poor 2
Asset Conditions Decay over Time 4.8-5.0 Excellent 4.0-4.7 Good 3.0-3.9 Adequate 2.0-2.9 Marginal 1.0-1.9 Poor Physical Condition Rating 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 Observed Physical Condition Versus Age: 40 Foot Buses SGR Spline - Bus (High PM) Spline - Bus (Avg PM) Spline - Bus (Low PM) Bus Inspection 1.0 0 2 4 6 8 10 12 14 16 18 20 Vehicle Age (Years) 3
Original Source of TERM Decay Curves Chicago Transit Authority s (CTA) 1992 Engineering Condition Assessment This $20 million study assessed the physical condition of all of CTA s fixed assets Used a reverse logit functional form to achieve the best statistical fit with the following three variables: Age Utilization Rate Maintenance Rate 4
FTA Research on Decay Curves Between 1998 and 2006, FTA invested in developing nationally representative decay curves New curves based on detailed, on-site asset condition inspections at 43 US transit agencies Inspections covered more than 1000 buses, 300 rail vehicles, 150 maintenance facilities, 100 rail stations and samples for train control, electrification and communications systems These new curves use spline regression models, which rely on the same 3 factors 5
Typical Decay Curves 6
$160 $140 Distribution of Asset Physical Conditions by Asset Type Asset Categories Guideway Elements Track Tunnels Elevated Structure Roadway $Billions $120 $100 $80 $60 $40 $20 5. Excellent 4. Good 3. Adequate 2. Marginal 1. Poor Facilities Admin Maintenance Systems Train Control Traction Power Communications ITS Fare Collection Stations Structures Parking Elevators / Escalators Bus / Pedestrian Access $0 Guideway Elements Facilities Systems Stations Vehicles Vehicles Revenue Non-Revenue (2007 NTD and TERM data, excluding generated assets) 7
Why the Condition 2.5 Threshold? Condition 2.5 is still serviceable but looks bad, provides poor quality service, and is starting to have an unacceptable risk of failure 2.5 gives backlogs for major agencies that are close to what they report as their unfunded needs 8
Useful Service Life Consumed for all Transit Assets Expected Life using 2.5 Condition Threshold SGR Backlog 9
Reinvestment Policy Determined by Asset Conditions 100% 5 Investment (% of replacement) 80% 60% 40% 20% New maintenance ¼ Rehab Investment and Condition vs. age Replace ½ Rehab ¾ Rehab 4 3 2.5 2 Asset Condition 0% 0 5 10 15 20 Asset Age (years) 1 10
TERM 40-foot Bus Investment Strategy 100% 5 Investment (% of Replacement) 75% 50% 25% Investment Condition 4.5 4 3.5 3 2.5 Condition 0% 2 0 2 4 6 8 10 12 14 Vehicle Age (years) no annual capital maintenance no ¼ or ¾ rehab 11
Example: New 100 Bus Agency Started in 2004 12
Recap TERM produces asset condition distributions and investment need forecasts Decay curves based on extensive asset condition assessments represent national averages Investment strategies use expected life from decay curves simplified models of typical industry practice 13
Going Beyond TERM Goal is to minimize overall cost of providing safe, comfortable, reliable service Costs not condition should be the primary drivers of reinvestment decisions Cost of maintenance Safety priority Cost of in-service failures Cost of customer time Cost of money / financing Not explicit in TERM analysis 14
Most cost-effective service life? 1000 windshield wiper blades Mean life expectancy is 30 months In-shop cost to replace is $10 In-service cost to replace is $50 Hypothetical Example Failed Units Interest rate is 1.5% per year 80 70 60 50 40 30 20 10 0 Time to Failure 0 10 20 30 40 50 Months Failed Units 1000 900 800 700 600 500 400 300 200 100 0 Total Failures 0 10 20 30 40 50 Months 15
Proactive or Reactive? Windshield wiper cost factors Cost to replace units that fail in service (reactive) Cost to replace all units still in service that have not yet failed (proactive) Adjustment for number of replacements needed over time Minimizing costs 5-year cost curve has a minimum of $35,000 at 20 months Waiting until all units fail in service costs $75,000 $1000s $60 $50 $40 $30 $20 $10 $0 Cost Factors Cumulative cost of in-service failures Cost to replace all remaining 0 10 20 Months 30 40 50 $1000s 120 100 80 60 40 20 0 5-year cost 6 10 14 18 22 26 30 34 38 Service Life (months) 16
Questions? Keith Gates, PE Senior Analyst Office of Budget and Policy Federal Transit Administration keith.gates@dot.gov 17