I've been going on about the Rail Rapid Transit Jan 2019 report and how I don't like how I perceive it as "slamming the tram" by certain suggestions that don't seem right (you can see my comments from before?). I was honestly trying to make a "what is the best for Vancouver" decision. Whether you believe that or not, or if you question my motives, so be it. I question the motives of the Jan 2019 document, but who cares what one person thinks?
I dug deeper into Jan 2019 Rail rapid transit to UBC studies assumptions based on gut feeling. I decided to write about what I found here. It's not complex, but the proof uses probability theory 101 from university (proof pulled from documents linked below), but I explain it in plain text and no math is required. The interesting statistical proof of it is in a link near the end where you have to re-read what I have written here and then proceed to some nice tables where the flawed assumption about LRT capacity is at once revealed and questioned.
I'm undecided on what technology choice is best.
BTW: I also like the namings of the station. They are simple.
------------------------------------------------------------------------------
The Rail Rapid Transit Study to UBC from Jan 2019 is fundamentally flawed. It states the maximum sustainable headway is 4 minutes instead of 2 minutes for the LRT, where 2 minutes was the initial headway given to the LRT by previous studies/experts on the same subject and then quashed as not practical in this report and this effectively gave the LRT half the capacity and a no-go for a technology option right from the start of the assessment. I am not convinced that this is true and that for 1 billion dollars or more in savings, more study should be done on the matter. Based on the selection criteria and final conclusions of the Jan 2019 Rail Rapid Transit Study from Arbutus to UBC, it is safe to say that this capacity decrease is the primary reason why LRT is not even on the table. Who would build a system that goes out of capacity in 15 years? No one would. And that is effectively the case against LRT. It is all based on what appears to be an assumption that should be challenged and questioned. In the next sections I will give about 10 reasons why I don’t think this is a fair assumption at all (with links to documents). Basically, by using some statistical math (Probability 101 in uni) and using worse dwell times, worse traffic conditions, worse traffic signal management that is being used in Jan 2019 report and the same infrastructure allowance (one bus berth), I show that 2 minute headway can still be maintained and is regularly maintained in developing countries at higher rates. In addition, the 99 B-line from commercial drive to arbutus is already maintaining uni-directional bus headway of 2 minutes currently during peak times of day, so the 4 minute headway assumption is not valid.
The Jan 2019 study was the first document that made a skytrain technology choice (and the most recent) and it was fed to the Mayor’s council of transit where they maintained their preferred option to UBC was skytrain, but mentioned it “is not a regional priority”.
The fundamental flaw of the Jan 2019 study is where most of the conclusions are drawn from a model of the transit system. The transit max capacity of the LRT was downgraded in this Jan 2019 report from 14,400 to 7200, where a previous study on the same subject suggested 14,400 was a valid assumption. I am challenging this downgrade.
This one downgrade and this one assumption (with no deep dive scientific/factual based observations from other documents or even rudimentary study of the existing traffic from Arbutus to UBC which would already exist) is the reason why the skytrain surpasses the LRT option. The LRT option costs half as much.
I don’t know if the LRT is the best option. However I surmise that it won’t be travelling very fast down the corridor with frequent intersections (can’t get up to speed on 7km stretch) and the peak headway will be 2 minutes during rush hour (in the morning), when there are few people crossing the street and this headway will only be in one direction. I also surmise that this is sustainable and for someone as competent as Translink: easy to operate. The business along Broadway will likely like it better (especially underused sections) and there will be opportunity for medium density all along it’s length. A tram coming every 2 minutes (worst case) and travelling slow..can easily be crossed?
Previous documents/experts maintained 14,400 as the max capacity of the LRT. The difference is that the headway of the system was questioned. In the previous report a headway of 2 minutes was assumed. In this report, there was cold water and doubt splashed over this headway of 2 minutes and 4 minutes was used instead. This effectively changed the whole dynamics of the report as the capacity of the LRT was cut in half. Keep in mind that BRT systems can get 10-15000 pphpd. There will be some evidence throughout that indicates this, but my main goal is to throw serious doubt on the one thing that changed all the outcomes for the LRT.
I believed the report the first 10 times I read it. I don’t know anything about headways, bus bunching, Traffic Signal Priority (TSP) or semi-exclusive ROW (ROW B), active vs passive or conditional vs unconditional TSP, bus boarding time, dwell time, or how capacity is measured or configured or calculated for BRT or LRT systems in ROWs. However the report has some very questionable assumptions as I have already pointed out in this forum and this report seems to “slam the tram”. So where there is fire…. I decided to investigate myself. From here to the end are paragraphs describing why I believe 2 minute headway during peak time in one direction is sustainable. I read about 100 documents on BRT and LRT planning in order to come to this conclusion: This is an assumption I’m making and I'm not an expert, but it is not a complicated science. There could be a traffic engineer or COV capacity engineer that reads everything I write and says...well...there is this one assumption which he is making that makes everything he (santak003) is saying invalid, but I reiterate that I’m not an expert. However, operationally there is a significant wealth of knowledge for this old industry and 2 minute headways for Bus BRT is quite normal (30 seconds for Delhi) and even when in semi-exclusive ROW’s. So what follows is just quotes from various on-line links that show why I think there was an error in judgement when downgrading the capacity of the LRT:
The Jan 2019 report states:
“The Phase 2 Report operating plan indicates that the LRT 1 option would operate on a 4-minute headway with ‘full signal priority.’ However, the type of priority given to transit is not clearly defined within the operating plan. “
However, For LRT, it is common practice that unconditional TSP is used, which is exactly what the above is implying. The previous authors are experts in the field as well.
We can see below that operating full block lengths and unconditional TSP is expected. In fact, in every rail capacity book, there is an assumption that unconditional TSP is used for at grade crossings. For Buses, the type of TSP becomes important, but not for the Tram. It says the the rule of thumb is one transit signal cycle time can be used as headway, but as a rule of thumb, 2x the traffic signal cycle time can be used and sustained.
Transit Capacity and Quality of Service Manual
http://onlinepubs.trb.org/onlinepubs/tcrp/tcrp_webdoc_6-c.pdf
“Where, as is often the
case, light rail train lengths approach the downtown block lengths, then the throughput is
simply one train per traffic signal cycle, provided the track area is restricted from other
traffic. When other traffic, such as queuing left-turning vehicles, prevents a train from
occupying a full block, throughput drops as not every train can proceed upon receiving a
green signal. A common rule of thumb is that the minimum sustainable headway is double
the longest traffic signal cycle on the at-grade portions of the line. “
I am not talking about full block trains and only 2 car trains. Certainly when building the ROW B in the median, the left turn lanes can be accommodated? I don't go into this much, but either does the Jan 2019 document and I can find nothing online about it. The rule of thumb above is clear: 2x the traffic signal cycle time. From the Jan 2019 document of Rail Rapid Transit to UBC it is clear the traffic signal cycle time in Vancouver is 1 minute.
From the Jan 2019 Rail to UBC Rapid Transit Study, it also talks about the rule of thumb, but then goes on to change the fate of the LRT:
“With these above factors in mind, the shortest possible headway for every transit unit to be able to take advantage of TSP is 2x the traffic signal cycle time. However, one must also consider that transit units travel in two directions along the Broadway corridor. As such, the bi-directional headway must be 2x the traffic signal cycle time, with the one directional headway no less than 4x the traffic signal cycle time, in order for every transit unit to be able to request TSP activation. “
This is one very large assumption being made. That it is 4x the headway. It disagrees with every document out there.
When looking at the 99 bline buses the wiki states:
“Today the 99 B-Line operates on a 2-minute headway in the morning peak direction, with a 4.5 minute day base headway.”
So we can see that it is already using directional headway not bi-directional.
The Jan 2019 report also assumed BRT TSP will be used (not active/unconditional, but active/conditional). For LRT’s it is normal to do unconditional/active TSP for at-grade crossings. TSP is not expensive and is usually not centrally done. Most TSP studies find a 20% improvement in through traffic (for Bus BRT) and negligible (2%) impact to other traffic. But the documents out there suggest for a tram, that unconditional is used. In fact in a bus survey done for North American systems with 30 different cities, 50% were using First In First Serve, active/unconditional TSP for Bus BRT. So it’s quite popular for Bus systems. It is quite normal to have direction of travel in the Traffic Signal Priority. So again this 4 minute headway is a questionable assumption with bi-directional 2 minute headways which are not required at any time of the day.. More work should be done on this. It appears to be very feasible to have a 2 minute headway as the original document suggests.
Now, there are definitely issues when approaching 60% theoretical capacity of the line. It is a fundamental concept of transit capacity that from 60-100% theoretical capacity there are slow downs for a number of reasons, one of which is that the on-boarding time slows down and this impacts the dwell time. However LRT (just like skytrain) has multiple doors to help this out. And more card readers. In a 40m train on the gold coast in Australia, there are 8 doors for example. Also, the Jan 2019 Rail Rapid Transit Study to UBC is careful to use practical capacity and not maximum capacity. However, first step is to determine maximum capacity.
The best document I’ve been able to find that explains these concepts with minimal mathematics (only probabability theory 101 from university is required):
Public Transport Capacity Analysis Procedures for Developing Cities
Jack Reilly and Herbert Levinson
September 2011
http://documents1.worldbank.org/curated/.../697190ESW0P1170d0Quality0of0Service.doc
For the technical details from that report with tables and further proof that 2 minute headway is normal in practice etc just click this link, but you have to reread everything I said and then continue to the further proofs with nicely formatted tables.
https://docs.google.com/document/d/1_-aSdehWkkGMiw3LrlcwnU0FVW5Go_MAQKx6yjKvHUg/edit?usp=sharing
Prev
. 
Linear Mode
