Wednesday, July 23, 2008

What can the endowment do for us? The Berea college example.

UBC Student Senator Alfie Lee posted this New York Times article on facebook a few days ago. I think it's worth thinking pretty hard about. Berea College, which outwardly looks like a typical New England private school, uses it's 1.1 billion dollar endowment for students. From the Berea website:/

Berea continues to build upon a distinctive history of 150 years of learning, labor and service, and find new ways to apply our mission (the Great Commitments) to contemporary times by promoting kinship among all people, serving communities in Appalachia and beyond and living sustainably to conserve limited natural resources....
Berea continues to build upon a distinctive history of 150 years of learning, labor and service, and find new ways to apply our mission (the Great Commitments) to contemporary times by promoting kinship among all people, serving communities in Appalachia and beyond and living sustainably to conserve limited natural resources

Now true, this college is different from UBC in a lot of ways. It's much smaller, not a research institution, has a bigger endowment (UBC's is about 700 million) and only accepts low-income students. Students work 10 hours a week on campus and pay no tuition. Food comes from the on-campus farm, furniture in the workshops, and crafts are produced for sale. Still though, this school is an example of what it looks like to actually live up to the high aspirations of lofty mission statements (like UBC's Trek 2010), and using an endowment fund for this purpose. UBC's endowment fund definitely has potential benefits to students and research. But the debates about how much to use now, how much to save, and to what lengths to go to enrich the endowment (by leasing out our land for development, for instance) are hugely important.


Monday, July 21, 2008

The Underground Bus Loop - analyzed to death

This is a guest post by Dr. Darren Peets, who just completed both his Phd in physics, and his term as a student represetitive to the UBC Board of Governors.

One issue at UBC that's particularly bothered me, and which wasn't resolved to my satisfaction during my Board term, is the proposed underground bus terminal at University Boulevard and East Mall. The University Boulevard Neighbourhood Plan, from which the terminal sprung, was the reason I got involved on campus -- up until then, I'd assumed that common sense would prevail. The Plan disproved that resoundingly for me. While there has been no shortage of opposition to the terminal, there hasn't been a great deal of in-depth analysis of its shortcomings, how they might be fixed, or what the alternatives might be.

Now that I'm on the verge of graduating, I feel a need to share what I still see as problematic about the bus terminal as currently designed. This will not be short.

The concerns can be categorized as financial, technical, aesthetic, convenience, safety, and process-related.


Having one central bus terminal emerged as the preferred option from the 2003 Campus Transit Plan (based on finances that do not resemble the current model). It was believed at the time that the terminal would cost $17M to build, and my recollection is that the above-ground housing and commercial were going to pay for it. It's now closer to $50M, and the above-ground program (aside from the new SUB) has been struggling to find a way to break even.

I personally like the idea of a central transit terminal. The ideal place for it would be the centre of campus, University at Main Mall (notwithstanding that there's a pedestrian zone there), but it needs quite a lot of land, and the nearest location that might be large enough is the proposed site, bounded by University Boulevard, East Mall, the grid of trees, SUB, and the Aquatic centre.

However, the decision to build entirely below grade and the plans for the development atop it were made prior to consultation, and the public consultation followed UBC's then-standard model of design-display-defend. They drew consistent, vociferous criticism, to little avail. It wasn't until 2007 that these concerns were listened to, and the area atop the terminal has now been basically fixed. The bus terminal was framed as being off the table in this discussion, and in May 2007, Board approved $15.5M to move utilities out of the way and construct a tunnel, which would become quite useful if we were to build a bus terminal that happened to connect to it. That terminal would be a completely separate and unrelated project, however -- possibly because the terminal would exceed $20M and would therefore be a public-private partnership (P3) by default. Without the terminal, the tunnel would make for a handy forklift-accessible storage area. The utilites have been moved, but the tunnel is not under construction.


Many people simply will not feel safe in a largely closed-in below-ground waiting area with a single exit stairway, even if that stairway is 40' (12m) wide. There will be good lighting and internal sightlines and security will be present at all times, but actual and perceived safety are very different matters. There are ways of improving this somewhat, for instance by taking part of the lid off so the terminal is open to the square or the atrium of a building, but there's only so much you can do with an enclosed space below ground. At present, the terminal is being compared to Burrard SkyTrain station, but the part that does not have a concrete lid on top of it is quite small, and will itself be covered, likely about 15' (5m) above ground level (this cover could potentially involve some limited use of glass). However, I doubt there's any way that such a terminal could be built underground and feel safe for all patrons.

The image shows the approximate outline of the bus terminal (blue) and its waiting area (blue with cross-hatching), the platforms where people get onto and off of buses (red), and the part of the waiting area that's open to the outside world (dark green), which is basically all stairway. (The shaded areas represent one scheme under consideration for buildings that could go in the area, including a new SUB.)


The point of having a central terminal is convenience, and the current design has shortcomings in that regard. The electric trolley buses get a separate loop on the street half a block away, because the terminal is not large enough for them. If the trolleys shared the terminal, riders could decide on the fly which bus to take, based on lineups and timing. If the trolleys instead looped around campus, we'd have a convenient and frequent campus shuttle network. Neither opportunity is being taken.

A wide variety of explanations have been given for why trolleys can't be accommodated within the terminal, and some have been quite imaginative. The current line is that they require additional depth for the trolley poles and would cause delays. I understand that both of these reasons are incorrect (and note that they'd be fairly easily solved). It's actually quite simple -- the terminal as currently designed isn't large enough to hold all the buses, and making it larger would cost money. Particularly if it were expanded under the East Mall sidewalk, where the main campus IT spine is, or downward to allow a second floor. While cost increases from $17M to $50M aren't enough to trigger a rethink, tossing an extra $5-10M into the hole to accommodate the rest of the buses would apparently be outrageous. The trolleys' overhead wiring and the possibility of their poles coming off got them kicked out first (note that the new trolleys automatically retract their poles if they come off the wires).

Putting campus shuttle buses in the terminal has never been seriously considered -- a more sensible place for them is on the street near the top of the stairs, particularly if they're on loops that involve East Mall. This means that we'll have a shuttle bus loop, a bus terminal, and a trolley loop, all hopefully within a block of each other.


More than half of all students get around by bus, and transit is the single most popular means for getting to campus. The grand entrance to campus will be a diesel-soot-blackened tunnel, followed by a well-lit and possibly well-finished crowded concrete basement. The original reasoning was to free up the surface for shopping, parking and a plaza.

The only groups who want cars in this area are the Aquatic Centre, who have a legitimate need for drop-off access, the Alumni Association, who would like drop-off access and short-term (15 minute) parking if possible, and the Bookstore, who believe that their ability to stay solvent relies upon people being able to park immediately outside their door. If the Bookstore's business model is inconsistent with their location on campus, the solution is to change their business model or location, not campus.

In my opinion, this is not a suitable main entrance for the campus.

Technical Issues: Capacity

The biggest problem with the terminal is capacity.

If you ask UBC's planning department "who looked into whether the capacity would be adequate to meet the demand?", they'll tell you TransLink spent $250 000 on studies looking to this and other issues. If you ask TransLink's planning department where they got the ridership demand and bus numbers required for these studies, they'll tell you the bus numbers were based on 2005 service levels plus 2% annual inflation, and the ridership demand was based on UBC's projections.

This circular logic means that ridership estimates are, to this day, based on pre-U-Pass estimates that badly misjudged the effect of U-Pass, while TransLink's bus numbers assume that the level of bus service in 2005 was adequate, and service levels should increase by 2% per year thereafter with inflation.

In UBC's 2003 Campus Transit Plan, projections indicated that the student U-Pass would increase ridership by 28% initially, settling out at 38% by 2011. I recall a modified projection that assumed the first four years of U-Pass to have 130%, 140%, 150%, and 160% of the baseline ridership respectively, as drivers graduated and transit riders came in, but I can't find a reference to this. Regardless, the increase in transit ridership in just the first year of U-Pass was 53% -- significantly larger than projected. This past fall's ridership counts put the increase over 2002/2003 (the year before U-Pass) at 82%.

Our ridership projections have been updated to reflect what has happened so far, by assuming that ridership will hold dead flat for half a decade then resume its previously-calculated inflationary increases (calculations based on faculty/staff U-Passes before 2011). While I do not have access to a ridership projection graph, I do have bus projections calculated by the same method. Pay close attention to the horizontal axis:

So TransLink believes UBC's demand projections because they don't have anything better (UBC knows the ridership and can give them surveys), and UBC believes their own projections because TransLink presumably spent a quarter of a million dollars confirming them. UBC trusts TransLink to look into this sort of thing because they'd need the terminal to work, and TransLink assumes UBC must have, because UBC is paying for the terminal and any required fixes to it, and would have a strong interest in it working.

Based on my experience with the UBC routes, I'd say that there's a significant latent demand. There are a lot of people who have to get on a bus very early or can't get on a bus because the buses are too crowded, have switched to driving in frustration, or have adjusted their schedules to avoid rush hour. If we had adequate bus service, there would be more trips on transit and inbound trips would be concentrated about 10 minutes before classes start. The distribution of buses around the hour is important, because the design can only handle so many buses -- it fails based on peak demand, not the demand averaged over two hours.

A further driver of demand will be fuel prices. As gasoline prices continue to rise, more people can be expected to consider transit. The gas price assumptions used to date have not proven particularly realistic.

A reasonable question might be whether TransLink anticipates improving service to UBC beyond the 2% per year that the terminal can accommodate through to 2018 or 2021. Well, the service improvements for 2008 are outlined in TransLink's 2008 Transportation and Financial Plan

  • Improvements to U-Pass service, primarily on the 17, 25, 41, 49, 84, and 145 routes.
  • A new #33 route along 16th Avenue to UBC (this was originally planned for 16 buses per two-hour peak period, but the numbers for this year suggest to me about 12).
  • A new community shuttle route linking UBC with Broadway at Alma via Spanish Banks (this shouldn't require space in the terminal).
On top of this, TransLink plans to upgrade the #43 to a #91 B-Line in the fall of 2009 when the Canada Line opens, and I've heard rumours that the 480 will likely be replaced at the same time by routes from Bridgeport Station and Steveston. I don't have enough information to calculate how many buses per two-hour peak period all these improvements would come to, and I don't even know how many we had this spring, but we're exceeding 2% inflation by a pretty substantial margin regardless.

A few words must also be said on what "capacity" means, because the report from TransLink's consultants suggested the terminal would not be at capacity if the trolleys were removed. Amongst other metrics, the terminal is not considered to be at capacity if the line-up to get into it in peak periods does not extend into the intersection at Wesbrook, 19 bus-lengths back, more than 5% of the time. The simulations assumed that buses arriving Not in Service would be held at Blanca Loop or on 16th Avenue to mitigate a lack of storage capacity in the terminal, and several other parameters had to be fine-tuned for efficiency, such as stoplight cycles at two separate lights. "Capacity" is the absolute breaking point for the terminal under the best possible conditions. Capacity was found to be 218-221 buses per 2-hour PM peak period (this might be a good time to refer back to UBC's bus estimates bar graph). It's not clear to me whether the simulations took account of such rare occurences as wheelchairs, bikes, school groups, or people asking for directions, but it has been indicated to me that bikes would likely be barred from the terminal during peak times to save space and time.

On the passenger side, capacity is defined more arbitrarily, based on crowding (and there is an assumption that many people will wait upstairs in the plaza). With the trolleys removed, the crowding is projected to fall to more-or-less acceptable levels. I'm fairly certain bikes weren't included in these simulations, and I'm not sure about wheelchairs.

Even if the likely-incorrect demand numbers and bus projections are to be believed, the terminal is designed to run near the breaking point basically from day one, and must be bailed out by SkyTrain after about a decade. Its design does not otherwise anticipate rail, and leaves no room on University Boulevard between Wesbrook and East Malls for trenches or foundations if grade-separated rail (e.g. SkyTrain) were to be added.

Solving this would need to involve a ridership demand study, preferably including a user survey. If my suspicions are correct and the terminal as designed is too small, new design options would need to be looked at. I'd much rather risk being proven wrong by a $20 000 study than proven right and having a good "I told you so" a few years from now.

Technical Issues: Driving on the Left

Another amusing aspect of the design is that, for the sake of space-efficiency, there is one central pedestrian platform, which the buses drive around. This means the buses are effectively driving on the left, and they have to cross over. I've tossed in an image tracing out the path of a bus that does a loop through the terminal without stopping, and you'll note that this path crosses itself where the ramp meets the terminal. This is handled with a stoplight, and one of its stop lines is at the front of a pick-up bay. Several interesting problems can arise from this, including blocking access to that bus bay, delays, crashes, and light cycles wasted because a bus stopped too far forward and blocked traffic in the other direction. I can even imagine a scenario where one bus tries to go around another bus that's too far forward, but can't make it, requiring several other buses to back up. Regardless, I'd expect this part of the design to be a serious problem. TransLink's consultants' simulation suggested other things would limit capacity first. In any case, it doesn't need to be there.

The ramp in the tunnel is designed to have a fairly shallow slope, giving outbound buses a good run at Wesbrook. Coming into the terminal, I see no reason for a gentle grade. The buses should be able to pass over each other instead of through, and this should be cheaper and more space-efficient to build. Nobody has thought about this problem in three dimensions.

It's not just the ramp -- so far, nobody has thought about the terminal design itself in three dimensions. The pedestrian area could be on the surface with storage below, the storage could be multi-level, or loading and unloading of passengers could happen at different levels (loading on top, for safety and to give the buses a better run at Wesbrook). If the capacity turned out to be an issue, I'd hope that options such as these would be considered in any rethink.

Other Technical Issues

I should point out that the entire terminal relies upon a computer system that manages all buses to the second, requires an operator at peak times to fine-tune it on the fly, and which is so specialized that it will take two years to design and purchase. Buses are notoriously bad at exact timing around here.

For a fun exercise, try coming up with a list of everything involved that runs on electricity (e.g. stoplights, electric doors, the massive ventilation fans feeding an exhaust flue with a 200 square foot cross-section, etc.). Now, what on that list has to be on emergency backup power so the terminal can operate in a blackout? I haven't seen an emergency generator in the plans so far, but they're not very detailed yet. The generator would be the size of a small moving van, would need to vent fumes, and would require a fuel tank buried somewhere where it can be excavated.

Finally, if there is a bus terminal resembling this anywhere in the world, none of the army of consultants has found it. This design has 1/3 the bus bays and 3-4 times the passenger flow of any comparator. While I like being leading-edge, getting this far out in front of the rest of the world means we need to pay a great deal of attention to every possible detail -- everything's new and untested. Very few systems work exactly as planned from the very first prototype, and this is definitely the first prototype.


UBC is paying 80% of the capital costs for a TransLink bus terminal, mainly from infrastructure charges on market housing developments. This piece is $31M that would otherwise either be in the endowment, building childcare, beautifying campus, or creating social space. A bus terminal may be nice to have, but $31M nice? Is a student more likely to drop out of school from lack of childcare, or from an extra block's walk in the rain? So far as I can tell, nobody has looked into the possibility of getting government money for this project. Oh, and who's on the hook for cost overruns or repairs due to shortcomings in the design? UBC. If the terminal's too small and an extra third or fourth bus loop has to be built elsewhere for the rest of the demand, who pays? UBC. If it has to be replaced within 40 years, who pays? UBC. I find this mind-boggling.

The bus terminal must be done right. If it feels unsafe, doesn't function, isn't what students want, or breaks the bank, it will be a very expensive mistake. Given that we're putting buildings on top of it, it can't be expanded or repaired. We only have one shot at this.