Amtrak Cascades Performance as of November 2016

Amtrak has released their monthly reports for October and November 2016.  Examining the Cascades performance for this time period shows ridership has continued to be above 2015, on time performance was relatively poor, and revenue was up.


  • October
    • 2015 – 54,190 
    • 2016 – 62,930
  • November
    • 2015 – 57,844
    • 2016 – 63,177

Ridership is on pace for a 9% improvement over 2015 and if December is similar to November and October the total for the year may be 810,000, barely surpassing the 807,000 riders from 2013.  In 2012 with 836,000 riders was the last time ridership was higher.

On Time Performance

  • October
    • 2015 – 78.5%
    • 2016 – 72%
  • November
    • 2015 – 77.9%
    • 2016 – 63.7%

The on time performance dramatically dropped.  For most of the year it was tracking close or better than 2015.  The poor performance in October and November made it extremely unlikely for on time performance to exceed 2015.

Though on time performance was much worse, the amount of time spent delayed is very similar.  In September 2016 the on time performance was 82% with 11,148 minutes of delay.  October had 10,942 minutes of delay with on time performance of 72% and November 11,134 minutes of delay with on time performance of 63.7%.  There is not a correlation between minutes of delay and on time performance for this period.  With the current Amtrak reports I am unable to narrow the cause.  I presume there was an increase in the number of trains delayed, but the delays themselves were shorter.


  • October
    • 2015 – $2,108,965
    • 2016 -$2,230,003
  • November
    • 2015 – $2,236,715
    • 2016 – $2,297,269

Average revenue per ride has varied so far this year, similar or better than 2015 in the first half of 2016 and below 2015 in the second half of 2016.  Due to October and November, the revenue per ride is now slightly lower overall, down 4%, but total revenue from fares is up 5% over the course of the year.  This points to a decrease in ticket prices leading to a boost in ridership.  If the costs are similar to last year this could be a smart change as overall revenue is up.  I now expect the end of year recovery rate to be around 61%.

A few months ago and before the October and November data, almost all metrics were better than 2015.  Now that October and November data is in, the metrics look a little less rosy, but still very good.  810,000 riders for the year, 61% recovery rate, overall revenue up, but revenue per ride slightly lower and on time performance similar to 2015.


Amtrak Cascades Performance Improves

In 2016 Cascades performance numbers are getting better.  Ridership, on time performance, recovery rate, and other metrics, as of August, are better than last year and likely the past two years.

Cascades 2016 ridership data looks very promising.  2016 is currently a better year than both 2015 and 2014 when comparing ridership totals up to August.  Total ridership this year is up 7% compared to the same point in 2015.  It is possible 2016 could be an even better year for ridership than 2013.  This reverses four years of declines.

On time performance also improved.  Trains are arriving on time more often and there have been no dismal months.  In 2015 there were two months where trains arrived less than 70% of the time.  This year each month has been above 70%.

Average income per rider is nearly the same compared to 2015.  In 2015, up to August, the average was $37.89.  This year the average is $37.19.  The increased numbers of riders should greatly improve the recovery rate as well.  Briefly looking at Amtrak’s report for per mile seat loss and overall loss, year over year, makes me believe the recovery rate at the end of this year will easily exceed 60%.  The highest recovery rate on record for Washington trains is from 2011 at 66.2%, however a different cost structure was in place where Amtrak picked up some costs.  Last year, the second year with the new cost structure, ended with a recovery rate of 58.6% and 2014 was 58.1%.

Expect at the end of the year total ridership to be at 800,000 or greater, 80% or higher on time performance, and possibly even a 62% recovery rate.


Cascades Seattle to Portland Travel Time Improvements

Once the current projects are completed, Seattle to Portland travel time via Cascades trains will drop.  The current published expectation is a decrease of 20 minutes from 3 hours and 40 minutes to 3 hours and 20 minutes.  10 minutes due to construction delays and 10 minutes from improved speed and on time performance.  The actual decrease will be even greater.

The Washington State Long-Range Plan for Amtrak Cascades from 2006 indicates three hour travel time is possible after the in progress improvements are complete.  Specifically page 3-4:

Current travel times from Seattle to either Vancouver, BC or Portland, OR will decrease by approximately thirty minutes each way.

The projects required to complete that kind of travel time are listed in the document and those projects are very similar to the ones already completed or in progress as part of the current WSDOT High-Speed Rail Program.  The 20 minute delta between estimates in the Long-Range Plan and the current planned reduction is significant and meaningful.  I have not yet found the cause of the delta.  One possibility for 5 minutes of the delta is the Tukwila stop, but this is only a part of the delta.  I surmise planners want to be cautious at first and see how the completed projects impact on time performance in practice versus their modeling.  Near 3 hour travel times are possible in the next couple of years.

The long term goal is a travel time of 2 hours and 30 minutes.  This has been re-iterated in multiple of the Washington rail plans.  For example in the 2014 rail plan:

Thirteen round trips between Seattle and Portland (1-hour frequency during peak travel times) with a travel time of two hours and 30 minutes (2:30).

The current projects will decrease delays and improve the average speed.  Currently, the maximum speed is 79 mph indicating a maximum track class of 4.  The long range planning documents indicate 110 mph speeds and class 6 track will be needed to reach the 2 hours and 30 minutes goal.

Many of the current projects, once completed, will allow for the track to become class 5.  Class 5 track allows for up to a maximum speed of 90 mph.  On page 4-14 of the Amtrak Cascades Mid-Range Plan on the WSDOT website:

This project will upgrade and maintain all existing main line tracks to FRA “Class V” standards. However, trains would still be limited to 79 mph maximum due to signal limitations.

The signal limitations, I believe, were addressed in some of the current projects.  The quote is referencing track between Blaine and Vancouver WA.  An associated cost increase in maintenance would occur if the track became class 5, from page 4-15 of the Amtrack Cascades Mid-Range Plan on the WSDOT website:

WSDOT estimates it will cost more than $200,000 per track mile. This equates to $97.4 million with delivery in 2014 (Exhibit 5A-8, Appendix 5). In addition, the cost of maintaining the tracks to the higher standard will be higher than today. This will take about four years to implement without severely disrupting existing service. BNSF estimates it will cost between $10,000 and $13,000 (2008 estimates) per track mile annually for ongoing maintenance at the higher track standard.

There are roughly 300 miles of track for the Cascades in Washington.  If being unkind with inflation and staying on the higher side of the BNSF maintenance estimate that brings the total extra cost in maintenance to $4.5 million annually.  Not all of the track needs to be of higher classification and if focused on the Seattle to Portland route the annual cost impact could be lowered.  It sounds feasible for at least some of the track to receive a class 5 rating.  I believe, based time estimates in the WSDOT plans, this would reduce travel time by another 10 minutes.

Basic internet searching for class 6 maintenance costs versus class 4 maintenance costs suggests the increase would be double the class 5 versus class 4 cost.  It seems class 6 track is an unreasonable expectation at this time.  The past couple of years WSDOT has named operating cost reduction as a goal.  It is also unclear how much if any track at this time is built for class 6 or if the existing engines are economical to run at class 6 speeds.

Aside from track improvements, the new Charger locomotives should also have an effect on travel times.  They are more powerful and more efficient.  Better acceleration will improve the average speed and the greater efficiency should allow for more use of the greater power due to lower operational costs.  There are hints in the various WSDOT plans and articles online that the new engines should help reduce travel time slightly, but I have not found any numbers to indicate how much of an effect they will have.

Over the next four years we should see a gradual decrease in travel time.  In 2017 travel time should be 3 hours and 20 minutes.  By 2019 I expect to see 3 hour travel times.  With a class 5 track rating, by 2020 I would hope for a 2 hours 50 minute travel time between Seattle and Portland.