Metro’s last quarterly report showed us the agency was struggling to have enough trains for service and that those running were frequently late and overcrowded. According to the latest one, customers faced even more single-tracking and delays in July, August, and September than they did last quarter, largely because Metro doesn’t have enough train cars.

Metrorail’s recent on-time performance. Graphic from WMATA.

On-time numbers have been hurt especially by the Stadium-Armory power problems, caused by a major fire at a power substation a few months ago. Additionally, the agency still doesn’t have enough railcars available to run normal service, a situation exacerbated by the opening of the Silver Line and delays in getting the new 7000 series cars. The data backs up how current customers feel about the system.

Keeping rail cars in service is still a challenge, three months later

According to this quarter’s Vital Signs report, WMATA was only able to supply the 954 rail cars it needs on 10 of the 64 weekdays in the quarter from April to June, or 16% of the time. That’s down from 19% in the second quarter.

When fewer than 954 rail cars are available for use during service, WMATA must either cut the number of eight-car trains it runs, or cancel some trains altogether. A report from Martin di Caro of WAMU suggested that there in fact are fewer eight-car trains running, possibly related to the phenomenon of having too few cars available.

One of the reasons WMATA says it doesn’t have enough cars is that 62 of its approximately 1130-car fleet are “indefinitely parked” due to a lack of parts. The agency attempts to keep a spare rail car ratio of around 20% so there would always extras when some need to be fixed or sent in for maintenance. That spare ratio currently sits at around 11%.

Having too few spares available means cars won’t get as much necessary maintenance time, and that there will not be extra cars to replace ones that break during service. As the WMATA data suggests, this lack of cars is significantly impacting service and means fewer eight-car trains and more cancelled trains.

On-time performance sank to new lows, and there were more cancelled trains

Most notable from the Vital Signs report, Metro’s on-time performance fell in September to less than 80%. That’s the lowest since Metro started putting out the report. And it’s the lowest in at least six years.

On-time performance since 2009. Data from WMATA, graphic by the author

The number of cancelled trains also increased compared to the previous quarter. All 100 4000-series rail cars were taken out of service in mid-June 2015 due to two instances of doors opening when they weren’t supposed to. Those went back into service in July 2015. The number of reported train cancellations during these two months and August are noticeably higher than the rest of the year, and the 4000-series is one likely reason why.

When a train doesn’t operate, it creates a gap in service averaging just over six minutes. So instead of waiting, say, six minutes for a train, customers have to wait up to 12 minutes or more, depending on the line. During that period the platform gets more crowded, and when the next train shows up, it has to carry a larger load.

The more crowded a train is, the longer it dwells in stations, which exacerbates the delay, and can cause bunching. Crowded trains can be more likely to be offloaded themselves as passengers hold the doors trying to get on and off.

Monthly train cancelations. Data from WMATA, graphic by the author.

The number of trains that did not operate (because not enough cars were available) spiked in July 2015. And while the number has been declining, the number of trains that did not operate was still above 150 in September, a major increase over pre-Silver Line numbers.

This quarter’s on-time performance was unusually rough

From the report:

“On-Time Performance (OTP) fell below 80 percent this quarter, the lowest rail OTP published in Vital Signs. Causes included: railcar shortages, speed restrictions and service reductions.”

Metrorail notched an average on-time performance of 82% for the quarter ending in September. August and September’s numbers (82.8% and 78.9%) are the lowest recorded in the public data dating back to 2009. July’s OTP of 84.4% is only beaten once by February 2015’s recorded on-time performance of 83.9%, which was directly attributable to the effects of winter.

The target goal for Metrorail on-time performance is 91% and was last met in July 2014 when averaged across all six lines.

On-time performance by line. Data from WMATA, graphic by the author.

Except on the Yellow Line, on-time performance fell across the system during the period from October 2014 to September 2015 compared to the year before.

The 7000 series is still being broken in, but the fleet is becoming more reliable

Four 7000-series trains have been placed into passenger service. Of the 64 cars that WMATA ordered in the first batch, 52 had been “conditionally accepted” when the Vital Signs report was put together. This means that WMATA has accepted ownership of the cars, but they have not been placed into passenger service. However, the acceptance process itself is held up waiting for software changes needed to fix various bugs discovered in the cars. Until Kawasaki can update, verify, check, and deliver this software, there aren’t likely to be many more of the cars in the system soon.

The reliability of the 32 7000-series cars that have already entered passenger service is still increasing. The first month of data (June 2015) showed a Mean Distance Between Delay (MDBD) of 14,902 miles for the series, while September’s data shows a MDBD of 70,202. The 7000-series contract requires a minimum MDBD of 150,000 miles fleet-wide, so we expect to see the MDBD increase to around that number.

This reliability requirement would make the new cars the most reliable in the fleet, followed by the 6000-series. Given the amount of time the 7000-series cars are in use currently (rush hours only) and how few there are in service, the data itself is pulling from a small sample set compared to the other series.

Waiting a couple months to get more and better data is the best thing to do in this situation in order to properly judge the new rail cars. In the best scenario, the reliability number keeps going up as WMATA learns the intricacies of the cars and can manage them better, leading to fewer breakdowns and train issues.

The Metrorail data released in this version of the Vital Signs report does not indicate any short-term solutions to the problems that having too few rail cars causes. As long as no others are crashed or damaged and the 7000-series cars start being introduced to the system again, it would take around eight months before WMATA has all the rail cars it needs for service plus spares. Only after this point in time is when 1000-series cars would start to be replaced, and the oldest cars start leaving for good. That day can’t come soon enough.

Stephen Repetski is a Virginia native and has lived in the Fairfax area for over 20 years. He has a BS in Applied Networking and Systems Administration from Rochester Institute of Technology and works in Information Technology. Learning about, discussing, and analyzing transit (especially planes and trains) is a hobby he enjoys.