North Carolina NCDOT-Amtrak Piedmont Service

My personal observation is that 506 (1215pm) and 507 (220pm) are the first to sell out and that there is almost always a seat available on 501 (730am). My conclusions are somewhat backed up by data in the WSDOT Gray Notebook, see the section on Farebox Recovery in the Amtrak Cascades chapter. Data for 11, 500 and 509 are not included because those trains are funded by Amtrak, not Washington state. So I realize my statement is more of a casual observation than a fact, but I think North Carolina's midday trains will prove to be very popular.

BTW, if Oregon publishes anything similar to the Gray Notebook, I'd appreciate a link, thanks.

villager wrote:There is supposed to be another pair of midday runs added in 2008. The tentative schedule is a 12:00 PM departure from CLT and a 12:30 departure from RGH.

This would make much better use of the equipment and I imagine we would see passengers per train climb after about a year or so as people found out they had some choices in regards to train travel.

Not to mention the crew cost savings. The current scheme has the 79/80 crew originate in Raleigh on 79 at about 4 PM or later, spend the night in Charlotte in a motel, then back to RGH on 80 the next morning. 73/74 crews do the same thing, being based in Charlotte. The mid day train will allow one day crew turns reducing motel and food costs.

AUGUST 2007 PIEDMONT OTP REPORT

Good evening. Here is your monthly Piedmont OTP report. As usual, the current stats for August are first, with last month’s (July 07) results in parentheses.


Here are the numbers:

Train 73: RGH - CLT
Scheduled Running Time: 3:09
Average Running Time: 3:12 (3:17)
Median Running Time: 3:09 (3:08)
Minimum Running Time: 3:03 (3:02)
Maximum Running Time: 4:34 (6:11)
St Deviation (in minutes): 15 (33)
90th Percentile Running Time: 3:19 (3:25)
Amtrak On-Time Standard Percentage: 90% (83%)


Train 74: CLT - RGH
Scheduled Running Time: 3:10
Average Running Time: 3:27 (3:19)
Median Running Time: 3:16 (3:13)
Minimum Running Time: 3:05 (3:01)
Maximum Running Time: 5:02 (4:14)
St Deviation (in minutes): 28 (18)
90th Percentile Running Time: 3:58 (3:36)
Amtrak On-Time Standard Percentage: 61% (72%)

DISCUSSION:

I’m pressed for time today, so here’s a short summary. Train 73: Great! Train 74: Tied worst performance of the last 12 months.

I have heard that some delays are due to double tracking project south of Greensboro, but that can't excuse ALL of them. NS normally does a better job dispatching here.

Semi-annual train host meeting with NCDOT is Saturday week (9/15). I will be interested in hearing what Allen Paul (NCDOT) has to say about that and updates on equipment and mid day service. I'll post something when I hear.

SEPTEMBER 2007 PIEDMONT OTP REPORT

Happy Friday. Here is your monthly Piedmont OTP report. As usual, the current stats for September are first, with last month’s (August 07) results in parentheses.


Here are the numbers:

Train 73: RGH - CLT
Scheduled Running Time: 3:09
Average Running Time: 3:14 (3:12)
Median Running Time: 3:08 (3:09)
Minimum Running Time: 3:02 (3:03)
Maximum Running Time: 4:23 (4:34)
St Deviation (in minutes): 17 (15)
90th Percentile Running Time: 3:33 (3:19)
Amtrak On-Time Standard Percentage: 80% (90%)


Train 74: CLT - RGH
Scheduled Running Time: 3:10
Average Running Time: 3:22 (3:27)
Median Running Time: 3:16 (3:16)
Minimum Running Time: 3:01 (3:05)
Maximum Running Time: 4:59 (5:02)
St Deviation (in minutes): 26 (28)
90th Percentile Running Time: 3:34 (3:58)
Amtrak On-Time Standard Percentage: 70% (61%)

DISCUSSION:

74 improves over a dismal last month but 73 falls off the high perch of 90% OTP. Here’s a more interesting wrinkle in the data:

Look at Train 74’s 90th percentile run times over the last several months.

JAN 3:28
FEB 3:56
MAR 3:30
APR 3:57
MAY 3:29
JUN 3:39
JUL 3:36
AUG 3:58
SEP 3:34

Average the odd numbered months and even-numbered months, and you get this:

Odds: 3:31
Evens: 3:52

Coincidence? Or is it that NS dispatchers alternate the night shift every month and the guy who is more sympathetic to keeping Amtrak moving works the odd numbered months? Maybe I’m in tinfoil hat territory here, but there’s definitely a pattern.

**Update**

I just checked the OTP 90th percentile of Train 79, which often meets Train 74 in single track territory between Raleigh and Greensboro when 79 is running late.

It has a similar pattern for 90th percentile:

Odd months: 15:32
Even months: 16:03

Maybe it's the CSX dispatcher between DC and Raleigh. I'd be curious to hear if any of the folks in rail operations note differences in how individual dispatchers handle passenger trains...

1. Reports that I get from hosts indicates that the track work between Greensboro and High Point is causing a number of delays on both trains, especially the morning ones (73/80). Most of the grading is done for the first segment, but there still a lot of work needed on signal relocations, etc before any track is put down.

2. Crews will clearly tell you that some dispatchers are good and some are not. People throw down on CSX dispatchers, but they have a terribly hand to play with too many trains on too little track, especially in the Richmond to Rocky Mount segment. NS dispatchers are generally good on the Selma to Charlotte segment, in my experience, but there are a few "off" ones there, too.

At the train host meeting last month it was reported that the track work (GRO-HPT) continues very slowly, that the mid-day train was at least 2 years off, and that the last of the cars for the Piedmont should complete renovations this year. The new siding between Clayton NC and Selma NC should be completed soon.

-RSM

matthewsaggie wrote:1. Reports that I get from hosts indicates that the track work between Greensboro and High Point is causing a number of delays on both trains, especially the morning ones (73/80). Most of the grading is done for the first segment, but there still a lot of work needed on signal relocations, etc before any track is put down.

I assume you are talking about the Cox/Hoskins double-tracking project, yes?

At the train host meeting last month it was reported that the track work (GRO-HPT) continues very slowly, that the mid-day train was at least 2 years off, and that the last of the cars for the Piedmont should complete renovations this year. The new siding between Clayton NC and Selma NC should be completed soon.
-RSM

A few other questions- once the last cars for the Piedmont are renovated, NCDOT will then have 6 fully renovated cars, correct? This would allow for the two F59-PHIs to pull renovated cars rather than any Heritage equipment, I assume?

At the train host meeting, was there any discussion about which projects are most critical to the introduction of the midday train, and why they are critical?

I assume that since the track speeds are pretty good at this point throughout most of the corridor, at least for conventional rail, that all the pre-midday frequency improvements that are deemed necessary prior to startup must be reliability-related.

The GRO-HPT double tracking is the hold-up (I use the station names as out of state readers won't know where the other "actual" locations are). All the passengers pass currently at the HPT station or within a mile or two south, but they often hold each other up, since the single track starts just north of the HPT station, if there are and delays. Throw in a lot of mid-day NS trains, and you have a problem that only completing the double track will solve.

To me, what's obscene is that its only 9 miles, but the NS, with the states money, is only rebuilding 3 miles a year. What ever happend to 10 miles of track in a day on the UP in 1869! ( I know different times and different track, but a real "can do" attitude)

Just guessing, but I think you're answer might be in the terms of the contract which may or may not specify a construction timeframe, just that those improvements will be made.

Thanks for the responses. That 9 mile gap is tragic, and the 3 miles per year pace is ridiculous.

I hate to ask again, but I'm super-curious. Will we have 6 refurbished cars for two Piedmont consists, each pulled by a F59-PHI, by the end of December?

Or am I overestimating the refurbished fleet we will have?

gt7348b wrote:Just guessing, but I think you're answer might be in the terms of the contract which may or may not specify a construction timeframe, just that those improvements will be made.

NS set the 3 year timetable, we just get to write the checks.

I think that the Piedmont's will remain 3 car trains, with the remaining 3 cars rotated into the set on occassion, and held in reserve for special moves and the future mid-day train.

matthew-Thanks for your response.

I assume that if we cut bigger checks, NS would develop faster timetables?

OCTOBER 2007 PIEDMONT OTP REPORT

Here is your monthly Piedmont OTP report. As usual, the current stats for October are first, with last month’s (Sept 07) results in parentheses.


Here are the numbers:

Train 73: RGH - CLT
Scheduled Running Time: 3:09
Average Running Time: 3:11 (3:14)
Median Running Time: 3:08 (3:08)
Minimum Running Time: 3:02 (3:02)
Maximum Running Time: 3:45 (4:23)
St Deviation (in minutes): 09 (17)
90th Percentile Running Time: 3:20 (3:33)
Amtrak On-Time Standard Percentage: 84% (80%)


Train 74: CLT - RGH
Scheduled Running Time: 3:10
Average Running Time: 3:15 (3:22)
Median Running Time: 3:13 (3:16)
Minimum Running Time: 3:04 (3:01)
Maximum Running Time: 3:58 (4:59)
St Deviation (in minutes): 11 (26)
90th Percentile Running Time: 3:25 (3:34)
Amtrak On-Time Standard Percentage: 77% (70%)

DISCUSSION:

What’s a more important goal? Very high On-Time Performance according to the Amtrak standard, or minimizing delays on trips that finish outside of the OTP window?

The longer I track the data, the more I think the latter goal is the most important for building and sustaining ridership. October 2007 was a terrific month for this goal. If you look at Train 73 and 74, both trains had a 90th percentile performance of 3:25 or below. This is a first since I began tracking performance. I wonder if the East Durham siding, which was completed in September, is helping with meets in the evenings for 74.

Also, the standard deviation in travel times (9 and 11 minutes, respectively) is terrific for trains with a 189 or 190 minute running time. As someone who travels on tight schedules with limited vacation, and cannot afford the uncertainty by a lot of Amtrak’s long distance services, I’m very willing to be 10, 20 or even 30 minutes late on a 3 hour trip. No big deal. It’s the low but reasonable chance of being 1-6 hours late with no other options that deters me from using a service.

While the Carolinian over CSX still suffers from this problem, the Piedmont along NCRR seems to be making the multi-hour delay an increasingly rare outcome. Kudos to NCDOT, NCRR, and NS for making it so.

MODERATOR'S NOTE: removed duplicate post - thread was created twice, removed the first post. 11/01/07 12:38 PM

NOVEMBER 2007 PIEDMONT OTP REPORT

Here is your monthly Piedmont OTP report. As usual, the current stats for November are first, with last month’s (Oct 07) results in parentheses.


Here are the numbers:

Train 73: RGH - CLT
Scheduled Running Time: 3:09
Average Running Time: 3:14 (3:14)
Median Running Time: 3:11 (3:08)
Minimum Running Time: 3:02 (3:02)
Maximum Running Time: 4:03 (3:45)
St Deviation (in minutes): 13 (09)
90th Percentile Running Time: 3:25 (3:20)
Amtrak On-Time Standard Percentage: 83% (84%)


Train 74: CLT - RGH
Scheduled Running Time: 3:10
Average Running Time: 3:17 (3:15)
Median Running Time: 3:10 (3:13)
Minimum Running Time: 3:05 (3:04)
Maximum Running Time: 3:55 (3:58)
St Deviation (in minutes): 13 (11)
90th Percentile Running Time: 3:36 (3:25)
Amtrak On-Time Standard Percentage: 73% (77%)

DISCUSSION:

This month was pretty similar to last month. This being the case, it’s hard to spot adjustments in the trains’ performances, or to speculate why things are the way they are. So for fun, I started running moving averages of 90th percentile performance, my favorite stat to track. I was looking for the overall pattern I see in my graphs in Excel, and I found it at a 5-month moving average of 90th percentile times. Starting in Nov 2006 and coming forward to Nov 2007, you see the following:

Train 73 (90th percentiles, Nov 06 – Nov 07)

3:35
3:34
3:35
3:32
3:32
3:29
3:25
3:21
3:22
3:21
3:22
3:22
3:24

Train 74 (90th percentiles, Nov 06 – Nov 07)

3:43
3:43
3:41
3:45
3:39
3:42
3:40
3:42
3:38
3:43
3:39
3:38
3:37

One train is moving from the 3:30s into the 3:20s. The other is moving from the 3:40s into the 3:30s. This is good.