|
Post by Gokhan on Mar 21, 2012 13:48:12 GMT -8
I dedicate this thread to technical matters regarding Expo Line operations! I start the thread with the following question, for which I will provide an answer myself: How many trains are needed on Expo Line Phase 1?My answer: Expected running time from 7th/Metro to Culver City is about 25 minutes. Metro plans to run three-car trains with 12-minute headways. Consider the whole system as a loop or circle, where the trains run continuously on the loop. The length of the loop is about 50 minutes. If we had five trains on the loop, there would be five loop segments between the five trains. That would correspond to 5 x 12 = 60 minutes. Therefore, if the trains spend about five minutes at both ends (such as turning around, dwelling, or the operator going to the restroom), 25 + 25 + 5 + 5 = 60 = 5 x 12! The math works out! Therefore, my answer is 5 three-car trains, requiring 15 cars total, running with 12-minute headways, and spending about 5 minutes at each end of the line.Comments are welcome!
|
|
|
Post by Gokhan on Mar 21, 2012 14:20:24 GMT -8
In addition to any comments for the question and answer above, provide your answers to this new question:
Does Metro have the 15 cars plus the spare cars to run the Expo Line Phase 1?
|
|
|
Post by transitfan on Mar 22, 2012 7:20:14 GMT -8
Hmm, well we would need to know the total number of LRVs assigned to Division 11 (the Blue Line division, and for now the Expo Line division as well). There are 69 Nippon-Sharyo P865 (P685?) cars (100-168). I'm not sure how many of the Siemens P2000 cars are assigned to Div. 11 now (there are a total of 52 (201-250, 301 & 302). I think 301 and 302 are still at the Gold Line (Division 21). I don't know how many cars the Green Line (Div. 22 has, earlier today I saw a pic of car 228 running on the Green Line). So assume that Div. 22 has 201-228 (28 cars), that leaves 22 cars (229-250) at Div. 11, for a total of 91 cars. I don't know how many 3-car trains the Blue Line needs, let's say 20 trains, that's 60 cars, which leaves 31 for the Expo Line, so 6 3-car trains is definitely doable, leaves 13 spares.
|
|
|
Post by Gokhan on Mar 22, 2012 9:41:55 GMT -8
Thanks transitfan! Great info! It looks like they have enough cars and spares. I would think not all cars can run, as some of them need to be repaired or maintained.
According to my calculations, 5 3-car trains is what is needed if Expo can make it between 7th/Metro and Culver City in 25 minutes and 5 minutes of dead time at each end is sufficient.
If the line turns out to be slower and/or they need more dead time (such as more time for the operators to rest), then they might need 6 3-car trains but this could lead to the bundling of the trains.
They certainly cannot run more than 6 trains at a time with 12-minute headways unless the line is really slow because it would lead to the bundling of the trains.
|
|
|
Post by James Fujita on Mar 22, 2012 17:54:14 GMT -8
I suck at math but I get the feeling that they are going to want those new Kinki-Sharyo LRVs that the Metro Board put on hold at today's meeting.
|
|
|
Post by bluelineshawn on Mar 22, 2012 20:20:44 GMT -8
Gokhan that sounds right to me although Expo trains probably won't dwell for 5 minutes at 7th/metro because a blue line train will likely need to leave in 6 minutes or so and sooner during rush hour. They'll probably switch train operators for each run. And from what I saw, they have a person designated to take the arriving trains into the tail tracks and to the opposite platform for departure. Like they used to do. Oh and the crew break room is plenty crowded now. They may not have thought that through. They need more space and another restroom.
I think that 301 is in the blue line yard now because I think that I saw it earlier this week either testing on Expo or running on the blue line. And I think that the blue line yard has maybe 6 or7 3-car P2000 trains. Whatever it is, that's all that it looks like it can fit. They are probably down some trains because of the accidents and various mechanical issues. Like when I was at 7th/metro on Monday they had a stalled train on the tail tracks with a propulsion issue, but that was probably temporary.
|
|
|
Post by crzwdjk on Mar 22, 2012 22:52:46 GMT -8
I did the calculations at some point, and came to the conclusion that Metro has enough rolling stock to run Phase I at 12 minute headways and, if they shuffle things around and really cut their spare ratio way down, they might have just barely enough trains for either Phase II at 12 minute headways or Phase I at 6 minute headways, but not the full line at 6 minute headways. But that doesn't take into account things like the Foothill extension, which should be done somewhere around the same time as Expo Phase II, and there are obvious reasons why it might be a bad idea to open to Santa Monica with only enough trains to run 12 minute headways, in case ridership really does grow that quickly.
|
|
|
Post by carter on Mar 23, 2012 13:11:44 GMT -8
Gokhan that sounds right to me although Expo trains probably won't dwell for 5 minutes at 7th/metro because a blue line train will likely need to leave in 6 minutes or so and sooner during rush hour. They'll probably switch train operators for each run. And from what I saw, they have a person designated to take the arriving trains into the tail tracks and to the opposite platform for departure. Like they used to do. Oh and the crew break room is plenty crowded now. They may not have thought that through. They need more space and another restroom. I think that 301 is in the blue line yard now because I think that I saw it earlier this week either testing on Expo or running on the blue line. And I think that the blue line yard has maybe 6 or7 3-car P2000 trains. Whatever it is, that's all that it looks like it can fit. They are probably down some trains because of the accidents and various mechanical issues. Like when I was at 7th/metro on Monday they had a stalled train on the tail tracks with a propulsion issue, but that was probably temporary. Well, here is the math: Round-trip time + dead time at the origin + dead time at the terminus = (Number of trains - 1) x headwayLet's assume that the round-trip time is 2 x 27 = 54 minutes Dead time at 7th/Metro is 3 minutes For six trains, this gives 54 + 3 + dead time at Culver City = 5 x 12, which results in dead time at Culver City = 3 minutes So, this would work out with 6 trains. If they had 7 trains, it wouldn't work out because you can't have the dead time at either end greater than the headway, as it would lead to bundling of the trains. Obviously, you can't have 5 trains, as it would give negative dead time, which means the train wouldn't be on time for departure for the return trip. Nowadays, I am only seeing three-car Nippon Sharyo's on the Expo Line. Could it be that they gave up on Siemens' because of the problems at the junction? Perhaps running Nippon Sharyo's only on the Expo Line was the solution to the problems at the junction. So, there are two tail tracks at 7th/Metro only 300-ft-long north of the rail switch there, correct? The Siemens and Nippon-Sharyo can be used interchangeably on the Expo and Blue Lines, because a train may need to be dispatched to Long Beach or Culver at a moments notice if there are any service interruptions. Today I only saw Siemens cars on Expo and that was what was used for the press event.
|
|
|
Post by Gokhan on Mar 23, 2012 13:29:27 GMT -8
Well, here is the math: Round-trip time + dead time at the origin + dead time at the terminus = (Number of trains - 1) x headwayLet's assume that the round-trip time is 2 x 27 = 54 minutes Dead time at 7th/Metro is 3 minutes For six trains, this gives 54 + 3 + dead time at Culver City = 5 x 12, which results in dead time at Culver City = 3 minutes So, this would work out with 6 trains. If they had 7 trains, it wouldn't work out because you can't have the dead time at either end greater than the headway, as it would lead to bundling of the trains. Obviously, you can't have 5 trains, as it would give negative dead time, which means the train wouldn't be on time for departure for the return trip. The Source is reporting the travel time to Culver City as about 30 minutes. If this is the case, we would probably need 6 trains when the line opens to Culver City: (2 x 30) + 3 min dead time at 7th/Metro + 9 min dead time at Culver City = 6 x 12
|
|
|
Post by Gokhan on Mar 23, 2012 16:55:26 GMT -8
Note that I made a mistake in my original formula, which I have now corrected in the above posts. I for some reason assumed that if there are n trains in a circle (or a loop), there are n-1 segments between them on the circle. There are obviously n segments. The corrected formula is actually simpler:
round-trip time + dead time at the origin + dead time at the terminus = number of trains x one-way headway
round-trip time: round-trip travel time from the origin to the terminus and then back dead time at the origin: time an arriving train spends at the origin before its scheduled departure dead time at the terminus: time an arriving train spends at the terminus before its scheduled departure one-way headway: time before the next train in the same direction on the schedule
Let's assume that
the round-trip time is 2 x 27 = 54 minutes Dead time at 7th/Metro is 3 minutes
For 5 trains, this gives 54 + 3 + dead time at Culver City = 5 x 12,
which results in
dead time at Culver City = 3 minutes
So, this would work out with 5 trains. They may need 6 trains if the one-way travel time is 30 minutes instead. For La Cienega as the terminus, 5 trains are definitely sufficient.
With 3-car trains, they would need 15 cars or 18 cars, depending on whether they need to run 5 trains or 6 trains at a time, respectively. Chances are that they have enough cars for Phase 1.
|
|
|
Post by spokker on Mar 24, 2012 17:14:13 GMT -8
At 30 minutes, average speed is 15 MPH, similar the Gold Line Eastside Extension.
At 27 minutes, average speed is 16.6 MPH.
At 25 minutes, average speed is 18 MPH.
I'm using 7.5 miles as the distance between Metro Center and Culver City. I calculated an average speed of 20 MPH back in April 2012 for the entire line (15.2 miles in 46 minutes per BuildExpo).
It will be slower than the Blue Line and the Gold Line to Pasadena. Is it the cost of safety? There are other trade offs, of course. The Green Line has a high average but its stations are relatively inaccessible to pedestrians compared to Expo.
LA average speed for automobiles is 26 MPH apparently, but I have not vetted that figure much. My own door-to-door average speed during my monster work commute was 20 MPH. Apples to oranges, however, because the average passenger rail speeds above are not door-to-door.
Not complaining. Just interesting to see where everything falls. Many more reasons to use transit exist, such as comfort, environment and stress. It's not my line and I have no need to use it. Its success or failure will not rest on Internet pundits but the community.
|
|
|
Post by Gokhan on Mar 24, 2012 18:16:16 GMT -8
These are the official travel times by Metro:
Expo westbound to Culver City: 29 minutes Expo eastbound from Culver City: 28 minutes
The distance from Culver City is 8.7 miles.
Therefore, the current average speed of the Expo Line is 8.7 / 28.5 * 60 = 18.3 MPH.
|
|
|
Post by masonite on Mar 24, 2012 18:47:41 GMT -8
These are the official travel times by Metro: Expo westbound to Culver City: 29 minutes Expo eastbound from Culver City: 28 minutes The distance from Culver City is 8.7 miles. Therefore, the current average speed of the Expo Line is 8.7 / 28.5 * 60 = 18.3 MPH. Phase 2 will be faster, but isn't this a lot slower than some of the estimates on this board. It is going to be tough to make the 46 minutes to the end if this can't be sped up. With better signal timing.
|
|
|
Post by Gokhan on Mar 24, 2012 19:25:26 GMT -8
These are the official travel times by Metro: Expo westbound to Culver City: 29 minutes Expo eastbound from Culver City: 28 minutes The distance from Culver City is 8.7 miles. Therefore, the current average speed of the Expo Line is 8.7 / 28.5 * 60 = 18.3 MPH. Phase 2 will be faster, but isn't this a lot slower than some of the estimates on this board. It is going to be tough to make the 46 minutes to the end if this can't be sped up. With better signal timing. The trains lose 3+ minutes at the junction.
|
|
|
Post by bluelineshawn on Mar 25, 2012 13:54:15 GMT -8
The trains lose 3+ minutes at the junction. Based on what I've seen I don't know that I would go that far, but if we combine the time lost from both the junction and the loss on Flower due to increasingly poor signal synchronization then it may be close to that on average. That will need to be incorporated into a new blue line schedule as well.
|
|
|
Post by bluelineshawn on Mar 25, 2012 14:21:58 GMT -8
These are the official travel times by Metro: Expo westbound to Culver City: 29 minutes Expo eastbound from Culver City: 28 minutes The distance from Culver City is 8.7 miles. Therefore, the current average speed of the Expo Line is 8.7 / 28.5 * 60 = 18.3 MPH. Do these include the full stopping time at Farmdale?
|
|
|
Post by Gokhan on Mar 25, 2012 15:23:46 GMT -8
Yes, the times are from the official Metro schedule. Trains do stop at Farmdale, even though the station won't open for a while.
Looking at the schedule again, I guess the time lost at the junction is more like 2 minutes. I will analyze it more.
|
|
|
Post by Gokhan on Mar 25, 2012 15:34:10 GMT -8
|
|
|
Post by Gokhan on Mar 25, 2012 15:48:41 GMT -8
Quickly analyzing the schedule and comparing it to the optimal running times on Wikipedia, the problem is the street-running segment, as we all expected.
The line is fairly slow from Pico to Western, which is the street running segment.
The time lost at the junction is not that bad actually. It seems to be about a minute or more.
The private-right-of-way segment is also not up to full speed yet.
I expect them to speed up the line by a few minutes next year, with better signal sync in the street-running section and a little faster operation in the private-right-of-way section.
Once again, the current running times are 29 and 28 minutes westbound to and eastbound from Culver City, respectively.
|
|
|
Post by Gokhan on Mar 25, 2012 16:16:18 GMT -8
Note that according to the schedule, dead time at Culver City is 12 minutes and at 7th/Metro it is 6 minutes.
Doing the math using the formula above:
(round-trip time + dead times) / one-way headway = number of trains
(29 + 28 + 6 + 12)/12 = 6.25 trains at a given time run on the Expo Line on the average.
The result is fractional because they switch trains from the Expo Line to the Blue Line and vice versa.
I was surprised that the dead time at Culver City is so long but this must be because they want to depart the trains there at the same moment as the other train arrives so that the trains don't run into each other on the same track.
Note that all Expo and Blue Line trains are three-car trains.
|
|
|
Post by bluelineshawn on Mar 25, 2012 18:34:28 GMT -8
Thanks! Although honestly the blue line schedule has gone out the window since testing started. Usually most of the trains are within a minute or two of the timetable, but lately there have been so many missed intervals that I don't know if the train is 2 minutes early or 10 minutes late. I waited at Del Amo for 30 minutes during rush hour last week and there should have been 5 trains and there were 3.
|
|
|
Post by joemagruder on Mar 25, 2012 19:42:58 GMT -8
If I read the schedule correctly, the 4 "sweep" trains are stored overnight at the USC station. Is that correct?
|
|
|
Post by Justin Walker on Mar 25, 2012 20:26:07 GMT -8
If I read the schedule correctly, the 4 "sweep" trains are stored overnight at the USC station. Is that correct? That does seem to be what the schedule says. Specifically: Two westbound trains (deadheading from 7th St./Metro Center) arrive at USC/Expo Park station and "Secure Cars at Passenger Platform" for the night. Two eastbound trains (deadheading from Venice/Robertson) arrive west of the USC/Expo Park station and "Secure Cars Between Watt Way and USC Crosswalk" for the night. The four trains spend the night there and each serves as a sweep train in the morning. Sweep trains are the first trains "to operate over any section of mainline track each day or after any interruption of service exceeding one (1) hour." Trains performing sweeps cannot exceed 25 mph, hence the longer run times given in the timetable.
|
|
|
Post by jamesinclair on Mar 25, 2012 21:06:16 GMT -8
That does seem to be what the schedule says. Specifically: Two westbound trains ( deadheading from 7th St./Metro Center) arrive at USC/Expo Park station and "Secure Cars at Passenger Platform" for the night. Two eastbound trains ( deadheading from Venice/Robertson) arrive west of the USC/Expo Park station and "Secure Cars Between Watt Way and USC Crosswalk" for the night. The four trains spend the night there and each serves as a sweep train in the morning. Sweep trains are the first trains "to operate over any section of mainline track each day or after any interruption of service exceeding one (1) hour." Trains performing sweeps cannot exceed 25 mph, hence the longer run times given in the timetable. Why the deadheading? Such a waste. USC will be the biggest rider generator, the trains that terminate there can and should carry riders.
|
|
|
Post by masonite on Mar 26, 2012 9:48:36 GMT -8
Quickly analyzing the schedule and comparing it to the optimal running times on Wikipedia, the problem is the street-running segment, as we all expected.The line is fairly slow from Pico to Western, which is the street running segment. The time lost at the junction is not that bad actually. It seems to be about a minute or more. The private-right-of-way segment is also not up to full speed yet. I expect them to speed up the line by a few minutes next year, with better signal sync in the street-running section and a little faster operation in the private-right-of-way section. Once again, the current running times are 29 and 28 minutes westbound to and eastbound from Culver City, respectively. 29 minutes to Culver City is pretty slow. Unfortunately, I don't know how much they will be able to speed up the line if the GLEE is any indication as it has only gone from 25 minutes to 24 minutes in its several years of operations. As a comparison, the Purple Line if built to Santa Monica would be well past Bundy and into Santa Monica from 7th/Metro in 29 minutes.
|
|
|
Post by bzcat on Mar 26, 2012 11:58:35 GMT -8
That does seem to be what the schedule says. Specifically: Two westbound trains ( deadheading from 7th St./Metro Center) arrive at USC/Expo Park station and "Secure Cars at Passenger Platform" for the night. Two eastbound trains ( deadheading from Venice/Robertson) arrive west of the USC/Expo Park station and "Secure Cars Between Watt Way and USC Crosswalk" for the night. The four trains spend the night there and each serves as a sweep train in the morning. Sweep trains are the first trains "to operate over any section of mainline track each day or after any interruption of service exceeding one (1) hour." Trains performing sweeps cannot exceed 25 mph, hence the longer run times given in the timetable. Why the deadheading? Such a waste. USC will be the biggest rider generator, the trains that terminate there can and should carry riders. Your posts are getting more reactionary... ;D I think Sweep trains do carry passenger, they just can't go very fast. Besides, the sweep train starts service at 4 AM... I can assure you not many people are going to USC at that hour USC is also the first place south of Metro Center that can reasonably be used as overnight rest area. You can't use Metro Center as overnight rest area because I assume that is where the Blue line train rests now. That Blue line train will sweep the track down to Washington. The Flower segment Expo track is using shared road space so not the best for overnight rest area as there are high vehicle traffic volume due to 110 freeway on and off ramp. Exposition by USC has dedicated right of way in the wide medium, and ample security presence by USC DPS. It's a good place to store the train overnight.
|
|
|
Post by bluelineshawn on Mar 26, 2012 16:03:09 GMT -8
Why the deadheading? Such a waste. USC will be the biggest rider generator, the trains that terminate there can and should carry riders. Your posts are getting more reactionary... ;D You misread something. He's saying that the last trains from 7th/metro apparently deadhead to the USC station when they could be used to carry riders. He's not saying that the sweep trains deadhead.
|
|
|
Post by bzcat on Mar 26, 2012 17:35:42 GMT -8
You misread something. He's saying that the last trains from 7th/metro apparently deadhead to the USC station when they could be used to carry riders. He's not saying that the sweep trains deadhead. Yes I see... sorry James
|
|
|
Post by calwatch on Apr 20, 2012 21:57:15 GMT -8
They have updated the schedule and it now shows only going to La Cienega: www.scribd.com/doc/90462480/d11-SummariesThis implies that they are not planning on going to Culver City any time soon (or until they issue a schedule showing trips to Culver City).
|
|
|
Post by Gokhan on Apr 21, 2012 0:30:08 GMT -8
Note that the time from 7th/Metro to La Cienega didn't change at 26 minutes but the time from La Cienega to 7th/Metro increased by 5 minutes from 25 minutes to 30 minutes. This is because the northbound Expo Line trains wait for the (northbound and southbound) Blue Line trains at the junction.
They have stopped testing the Culver City section after Day One, as they started installing emergency lighting on the two bridge ramps there. The work seems to have stopped but there is still no testing. I don't expect an opening there before late June or even July.
|
|