Updated April 13, 2015:
The TTC has issued a press release regarding the management of the Spadina subway extension project:
The Toronto Transit Commission has entered into an agreement with Bechtel Canada Co. for project management of the Toronto-York Spadina Subway Extension (TYSSE) for up to $80 million.
The contract value to Bechtel is based on staffing costs, management fees and incentives to open the subway extension by Dec. 31, 2017. Bechtel staff begin work today and will form an integrated team with existing TTC personnel. The Bechtel contract will expire March 31, 2018. Bechtel’s project director will report directly to TTC CEO Andy Byford.
On March 26, the TTC board approved a report from staff that recommended TTC enter into a sole source agreement with a project manager with a proven track record of delivering similar-sized projects on time, and with experience working with multiple contractors, in order to have the TYSSE in service by Dec. 31, 2017.
Toronto City Council subsequently authorized the expenditure of $90 million, while the Regional Municipality of York authorized the expenditure of $60 million, for a total of $150 million (third party contractor, plus in-house project costs), to fully deliver TYSSE by the end of 2017.
The release is silent on the issue of what might be done with the remaining $70m of Toronto/York’s $160m authorization.
Original article of March 29, 2015:
In a previous article, I reviewed information from a media briefing by Andy Byford on the status of the Toronto York Spadina Subway Extension (TYSSE) project. At the TTC Board meeting on March 26, 2015, further information was made public both in Byford’s presentation, and in additional material appended to his report.
Updated March 30, 2015 at 1:30 pm: The slides from Byford’s presentation are now available starting at page 58 of the linked pdf.
Updated March 30, 2015 at 11:30 pm: A new report from the Toronto City Manager to Council advises that the interest earnings on the “Move Ontario Trust” (the repository for provincial contributions to the TYSSE project) have not achieved the target rate of 4% resulting in an $85m shortfall. Oliver Moore reports in the Globe that Ontario has refused to make up this amount as per the original agreement between the funding partners. Toronto and York Region are on the hook for this additional cost estimated at $51m for Toronto and $34m for York Region. This expense is over and above the cost overruns on various contracts, but at least Council cannot blame the TTC because the trust fund is not under TTC control.
Appendix F (beginning at page 33 of the linked PDF), is a presentation given to the Executive Task Force who oversee the project on behalf of the sponsoring governments on July 28, 2014. The presentation was given by Parsons Brinkerhoff who had been retained by the TTC to review the project.
Appendix G (beginning at page 56) is a two-page summary of Bechtel’s work reviewing PB’s original study and a subsequent APTA (American Public Transit Association) peer review. APTA concluded that an earlier completion date would be possible than PB had projected, but only with major changes to the project management structure. Bechtel concurred in these findings.
It is abundantly clear from this material that the TYSSE’s problems were known at the top level of the project in mid-2014 at the latest. At the time, their severity was so great that the project would still be incomplete by the time of the next municipal and provincial election cycles, and that considerable additional cost could be facing the funding partners. This very serious issue did not arise in public discussion until six months later, notably after Toronto’s 2015 budget cycle was complete.
Also included as Appendix E (beginning at page 19) is an October 24, 2012 TTC Board report advising of delays to the TYSSE project and adjusting the opening date to “fall 2016”. The report includes a detailed chronology of the project beginning in December 2005.
As of October 2012, construction work had been underway for barely two years with contract award dates ranging from November 2010 to July 2012. Construction work at York University Station was scheduled to begin in January 2013. It is important to note that the work fatality at the site occurred in October 2011 during the first phase of construction. This was part of the Northern Tunnels contract that would eventually leave a hole into which the station would be built under a separate contract. This is the same hole that filled with water (“Lake York” as described in the National Post) because of miscommunication about the responsibility for pumping out groundwater between the tunnelling and station contractors.
Problems with all of the construction sites were apparent in 2012, and the risk analysis concluded:
By early 2012 a review of the schedule impacts as noted strongly suggested that completing the project by December 2015 was unlikely. A number of workarounds, alternate work methods and acceleration achieved limited schedule recovery but were not sufficient to maintain the original schedule.
In mid-2012 schedule risk assessments were facilitated by independent transit scheduling experts. The conclusion was that maintaining the original schedule was no longer viable even by extensive acceleration measures and corresponding additional expenditures.
The risk analysis has been followed up by intensive schedule recovery workshops held on a weekly basis. These workshops explored potential initiatives to recover schedule that allow TYSSE staff to initiate action to achieve schedule recovery.
Many initiatives have been adopted and others are in discussion with the contractors. Some, such as major acceleration have some risk of failure and will require the outlay of significant funds as contract changes beyond the contracted amount and will require extensive analysis, scrutiny and successful negotiations with contractors. Efforts in this regard are ongoing, but the level of success in negotiating major accelerations with contractors is unknown at this time.
Initiatives coming from schedule analysis, observed performance and better contractual situations and risk analysis are expected to achieve some schedule improvements.
The potential for contract slippage and for “significant” additional expenditure are quite clear. However, at this point the assumption was that the project could still be completed within its original budget.
Pages 30-32 show the evolution of the project schedule from an original view in 2006 to a then-current view in 2012, and even that proved optimistic. Note that by 2012, the effect of the delay from the York University fatality should already have been incorporated in the schedule.
What is missing is an updated chart showing newly projected dates for the project components or a discussion of the factors since October 2012 that would lead to even further delay.
By July 2014, the transition from tunnel building to stations was well underway, but the PB report notes that five of six sites were “behind the spending curve”, i.e. the contractors were not building at the expected rate. Without stations, installation of many systems cannot occur including track and power, not to mention equipment and finishes in the stations themself. Delays to contracts for such work would lead to claims and the whole process of managing the project would be extended. The most telling goal of the PB review is to “Define the magnitude” of the station construction problem. This suggests that the TTC’s project managers did not actually know the status of their own project.
A detailed review took place with a series of meetings in June and July 2014 leading to a conclusion that a May 2018 opening would be possible subject to various caveats notably that the then-current rate of progress by contractors continued and that no contractor would go into receivership or abandon their work.
The new schedule was subject to formal risk analysis to examine the linkages between contracts and activities, the risk of various events, and the combined effect on a likely opening date for the line. This revealed that at May 2018, there was a very high risk (see chart, page 47) that many of the potential delays would occur, and that far too much faith would be required in ideal circumstances to actually hit that date. The project would not get “over the hump” of the risk chart until late 2018, with mid-2019 as a worst case scenario. Of particular note were the late completion of the trainways at Steeles West (now Pioneer Village) and York University Stations that would slow down many dependent contracts. This was not a question of fixing many small problems (although there were plenty), but of addressing two very big ones.
Three schemes for interim operation of an incomplete line were considered:
- Service to Vaughan bypassing Steeles West Station. This option would bring the opening date to November 2018, but would be subject to added cost of $17-25m, and could be subject to safety concerns (operating through a construction site, emergency exit provisions).
- Degraded service from Finch West to York University. This option would see an April 2018 opening, but only one in three trains would run through to York U via single track, two-way operation north of Finch West Station, the nearest crossover location. The added cost would be $28-40m, and this option too had safety concerns.
- Service to York University with a turnback at Steeles West. This option too would yield an April 2018 opening date with Steeles West being the turnback location for revenue service as far as York U. The added cost would be comparable to option 2, but with more technical and safety concerns.
All three of these depended on completion of the two much-delayed stations at least to a point where some train operation was possible.
The PB recommendations include detailed review of the options and implementation of very close monitoring and planning to track the project against any revised targets.
The Bechtel review (building on the APTA review) includes recommendations for strengthened project management and improved relationships between the TTC and its contractors. That so much is needed to bring the project back on track only hints at the underlying scope and details of problems, some of which were alluded to by Andy Byford in his verbal presentation at the Board meeting. Project management may have existed on paper, but the widespread lack of co-ordination shows that this was more show than substance.
The essential difference in the approach recommended by APTA and Bechtel requires a new project management structure and improved relationships with contractors. Although some of their claims for extra costs may be challenged, some have legitimacy and at least partial payment should add to good will between the parties. Whether it will result in better completion rates and quality is quite another matter.
At the end of a long discussion, much of which was held in camera, and with a section from which even most management were excluded, the Board decided to accept Byford’s recommendation that a project management consultant be retained to take over control of the project. Two Board members objected to the sole sourcing of this work (almost certainly to Bechtel because of their previous involvement, although this detail was withheld from the public discussion), but in the end the proposal passed.
From here forward, we can only hope that a detailed project plan and tracking will be published at every TTC Board meeting so that, if there are problems, the issues get a public airing early in the process rather than hiding conveniently out of sight.
And in the real world, that means dismissal for cause – if they did not know what was going on, it was incompetence.
I know a couple of managers have been dismissed, but this needs to occur more often.
Steve to me the idea of the need for continual expansion, and experience in construction and its management is another reason to try and break up projects into smaller more manageable chunks. It would permit detailed study of the area in question a complete project review of requirements for say 2 kilometers of track, and say a station or 2. Review the project as you proceed, and learn your lessons. Proceed to the next section required with a new proposal, and lessons learned.
It is also a reason to generally where possible build lighter smaller projects on an ongoing and dispersed manner. Ideally the TTC would not be directly for this kind of project management, as I believe that even for LRT of BRT projects they are quite involved, and a distraction, also as the final operator, the temptation to tweak will be very strong. I would prefer to see an ongoing process of expansion, with a process that would support an ongoing annual more modest investment in transit (although given how long nothing has been done that would need to be substantial now).
The problem with this is that some parts, like tunneling, need to be done before other parts can proceed. This could result in one part holding up all the rest, which appears to be what happened, resulting in contractors being unable to proceed when they expected to. I believe that it is better to have one main contractor responsible for the whole project so that you only have one to blame. It is then their responsibility to get all the parts working together.
LikeLiked by 1 person
@Robert Wightman – I was actually thinking in terms of short extensions, where only 1 or 2 stations and tunnel would be added at a time. However, yes, I think it would be better to have a single point of contact that had overall responsibility. The TTC should be expected to know how to run a transit system, not construction management. However, even at this, why get into bed with a 6 or 8 km extension and 5, 6 or 7 stations with a single contractor in a single contract. If you could build say 2 stations and 3 km of line, and review after each project. The next extension not starting until this one was complete, buttoned up and in service all issues settled. The need to extend the system in massive segments, seeming to be a little bit of an artifice.
Steve: Equally artificial is the “just build one km a year” line. That does not mean start-build-stop-accept-operate for each segment as a separate tunnel for the simple reason that this would be an incredibly inefficient way to do things. For starters it would mean that every project would be cut-and-cover simply given the overhead of creating an access shaft for each piece of work individually.
It is the zillions of individual contracts that can get us into a mess because there’s just that many more contractors to manage and projects to treat as separate entities. We don’t buy a few subway trains every year because it is uneconomical to manufacture them that way. Similar considerations apply to many other aspects of subway construction.
The other problem, of course, is that by our choice of subways, subways, subways, we make projects more complex, expensive and long-lasting than they need to be in all cases.
Steve, yes clearly building a single km of subway is impossibly costly way of eventually getting 6 . There are basic costs of setting contracts and mobilizing people and equipment. Also yes the massive nature of what is required to build subway makes this much worse than lrt or brt. However we seem to build subway in massive ill considered waves, well beyond the minimum reasonable extension. A greenfield build of a km or 2 of brt or lrt a year would be imaginable with a very small amount of equipment and crew.
Doing a go slow approach in an existing roadway would be infuriating, and best to be getting it over with. However a 200 million dollar extension of Lrt would be meaningful, and could be reasonably undertaken, not so much subway. This would also move an LRT project across neighborhoods quickly enough to not be devastating, even if built 200 million at a throw. Unfortunately now is no longer the time for small or greenfield builds, but rather 15 km LRTs are now needed, but moderate future extensions should be in our minds.
Steve – oh and to your point of subways, subways, subways, why is it that we obsessed with subways for every point, except where there is the most compelling argument for a need to have very large capacity, and no choice but to be underground? While it would appear we are slowly actually acknowledging a need, how is it that Vaughan VCC) or Richmond HIll (RCC) make a more compelling case than say Thorncliffe Park, or Pape & Cosburn, as reasonable new station locations?
Malcolm’s suggestion of breaking up a big subway project into smaller chunks, and building only a portion at a time in order to review & control costs has its appeal.
How feasible would it have been if, say, the TYSSE were tunneled to completion, just as it was, but then, instead of constructing all of the stations and all of the infrastructure all at once, we started at the Downsview end and built a couple of stations at a time under separate contracts?
I can imagine some problems.
The TTC is looking at building only that portion of the Relief Line from Danforth southwards. Though, in my submission, I said that planning for north of Danforth was absolutely necessary even if it isn’t built at the same time, in order to protect the route and to eventually have an integrated whole.
Isn’t the Crosstown LRT being built under separate contracts tunneling vs stations?
Steve: The tunneling strategy will always be unique to each line depending on which locations are appropriate for tunnel, which for cut-and-cover, where the TBM access shafts can go, etc. Depending on how the Don River crossing is handled, the section downtown could be a completely separate build. On the east side of the river, it certainly would make sense to bore all the way north to the Don Valley beyond O’Connor (whatever route they might take) just to get that out of the way in a single go. The river crossing at or south of Queen is going to be tricky for many reasons including various obstacles, the river itself, and the lack of planning for a route west of the river.
As for the ECLRT, yes, the tunnels are a separate contract from the stations. At this point in construction, the future station locations are having “headwalls” build where the outer ends of the box structures will be, and the TBMs are just chewing through them and building a tunnel that will eventually be exposed and at least partly disassembled for the station construction. The stations will be built under a separate contract, but awarded as one big block to a single consortium. Such an approach would not be as simple with an incremental build because the project would be stretched over a longer time.
Steve, I appreciate your answer very much regarding phased subway construction, Malcolm’s idea.
Though this thread is supposed to be about the Downsview subway extension, you’re bringing up an issue with the DRL, that is, how to cross the Don River. I’ve only seen this mentioned just once, and in not too much depth [pun not intended].
I agree with you, Steve, that tunneling would be the sensible way to build. Tunneling allows existing buildings to remain; the streetscape doesn’t get ripped up. And if any part of the route follows the CNR right-of-way, apparently it is too narrow to allow side-by-side train & subway tracks, so it has to be tunneled here, as well.
I would imagine, if the DRL route runs (most sensibly) along Wellington Street, and we want to service the redevelopment of the lower Don lands, that the route would run below King and Queen and somewhere there would cross the Don River.
I would imagine that tunneling across the river would have to be very deep in order to bore through bedrock and not soft silt.
On the other hand, a bridge over the Don means that the line has to rise to meet the bridge’s elevation, and then descend again into the tunnel.
Steve: I am looking forward to whatever alignment studies come out of the RER/ST/DRL work now underway. I can guess at the permutations, but it would be good to see something more definitive if only to eliminate certain options from consideration.
I am really suggesting that we stop building is such massive extensions. When I said break up into different projects, I am really pushing the idea, that if we had wanted to be in Vaughan eventually, there was no reason to build it all at once. Build it to the University, and stop, operate it, make sure you got it right, and ask yourself the question does the next section make sense, is there something different you would do.
The only reason we should be building massive segments in my mind, is that we are so far behind what we should have built, that we are now required to go big. However, this is hard to see in terms of extensions of the existing lines – 3 or 4 kms is a lot of subway.
Steve: The original idea was to build only to York U, but that didn’t sit well with the powers that be in the “City Above Toronto” and at Queen’s Park. Also, frankly, I don’t think 8km is a particularly long extension. After all, the tunneled portion of the Eglinton Crosstown line is over 10km and nobody seems to be claiming it should be smaller.
To the first, I think that the first plan for York extension is what should have been done, but well … To the Eglinton, issue, I had been tempted to talk about the DRL for similar reasons. Neither of these (Crosstown or DRL) are extensions, they are entirely new lines. The Eglinton line at its current length is required to fill its basic function (actually needs to be several km longer), and well also falls into the area of something that should have been done a long time ago (just not as subway). The DRL also would not be an extension but an entirely new line required for linkage reasons, and I cannot imagine building less than 7 km and would like to start at 11 or 12 (linkage to both Crosstown and Danforth through core), and a corridor secured that would allow it to eventually get to Parkside and the Queensway at a minimum, likely back to Bloor.
I really do think that these need to be considered in different lights, hence I was trying to be clear with regards to extensions of existing lines, which in the case of Toronto, are already quite long. I would point out that 5 km on the Bloor line, would get us a Dixie stop, 5 km on Yonge puts us in Thornhill (Thornhill Ave) – extensions that long on these lines at this point should be carefully considered, and challenged.
It’s very off topic for this thread but I think it’s all probable that the south Don crossing for the DRL will likely be tunneled thanks to the city’s lack of foresight in protecting lands for potential routes. The same thing is happening at the possible western terminus of DRL phase 1 where the city is about to approve more developments without provisions for future transit.
In the past I would have thought closing Sunlight Park Rd (the appendage sticking out from Eastern Avenue @ Broadview) for a new portal and bridge across the Don would have been perfect leaving open the option for any number of downtown alignments. The sudden appearance of a new neighbourhood and giant park on the west bank of the Don eliminates most of the above grade possibilities.
How much of this fiasco would have been evident while Webster was GM? Could his realization that things were out of control lead him to say we shouldn’t build subways subways subways — and thus got him fired?
Steve: I suspect that Webster left the engineering side of the house to that branch of the TTC, and coupled with the “see no evil” attitude of Karen Stintz as chair, things got out of hand for want of proper monitoring. His objection to subways was based on the simple fact that we don’t need one in Scarborough or on Finch West or on Sheppard West. Webster also opposed the DRL and espoused the “we can fit all of the passengers on Yonge” outlook.
Daylighting south of Queen is a very bad idea from a disaster management perspective. From the Don Flood Plain Rehabilition Project, a Hurricane Hazel event would flood the whole western bank between Queen and the rail corridor under more than 1m of water. With the DVP, Queen, Eastern, and Lakeshore closed, the DRL would be one of the only options of travel that might remain open for the area. In this area, bedrock is around a depth of 67m to 70m and surface are 79m to 82m, so a tunnel would only be in the 12m to 20m depth range.
As the build 1 km per year doesn’t seem to be an option, the only other way would be to complete the entire tunnel and with signals and crossovers now and rough in intermediate stations for construction and opening up in future. However this only makes sense if there are trip generators beyond the unfinished stations, and if the incomplete stations can be built without major disruption to the line itself.
From Toronto’s experience we have two examples of the approach. North York Centre was built along an active subway in between Sheppard and Finch stations. In contrast Bayview, Bessarion (and to a lesser extent Leslie) on the Sheppard Line were built with the knowledge that these would not be heavily used stations for many years.
I recall that at one time it was thought/suggested that we build the line from Sheppard-Yonge to Don Mills without having interim stations built in order to save money. I suppose it was found that it would cost less over time to just build the stations and open them with the line than to try and build the line with the ability to complete the interim stations in the future.
I suppose it may be because of that decision to build and hope that Willowdale station was sacrificed.
I assume by depth that you actually mean height above sea level.
Not sure where else to put this comment, but anyway – crazypants idea time.
For a comparatively low cost, imagine a cablecar gondola going from Don Mills/Eg–Thorncliffe Park–Donlands Stn–Unilever site–West Don Lands (at, say, the new Cherry Loop)–East Bayfront–the new office development slated to replace the GO bus terminal and expand the streetcar loop.
Pros: inexpensive, fast to build, easier-to-maintain infrastructure, a “DRL” that serves many of they key areas a subway would for a fraction of the cost
Cons: limited capacity (3000 pph, from my research), need for support structures through residential areas (possibility of locating them at intersections reconfigured as roundabouts?), legality/liability of having suspended cars travel over private property
Is this even a remotely feasible idea for the DRL route in Toronto?
Actually, there has been a private study of such a system from the Don Lands to Exhibition Place, but as a premium private service, apart from the TTC, and mostly focusing on tourists. Even this wasn’t feasible. As the peak hour, peak point ridership is between 12K and 13K and peak hour boardings are 24K to 27K, you’d need an awful lot of parallel gondalas.
Steve: Yes, and the structures, particularly the stations, would be very intrusive in an area intended for a modestly-sized roadway with buildings out to the lot line. I am getting quite tired of the gondola advocates who show up from time to time as if they are “the answer” to transit problems. Gondolas have specific applications, typically point to point where there is a major problem with terrain. They are not a replacement for surface transit.
Bad Ross: When rebuilding intersections, like Spadina-College, crews assemble track, then dissassemble & transport to the site
Why then does it take a month?
Steve: The elapsed time for the trackwork is about three weeks, and if the weather co-operates, as it did last year at Dundas & Bathurst, they finish a bit early. It takes two days to demolish and remove the existing intersection. That’s with four rams breaking up concrete and round-the clock work clearing things away. By the third day, they can level the site and start prepping for the foundation slab. Some advance work is required for new drains, and other utilities may be involved. By day five, the track starts to go in, and it takes until well into day seven to get it all assembled. This does not include the approach sections on the four sides that are excavated but not yet rebuilt.
(As a side note, when we get to the point of rebuilding intersections that have had this treatment, they will not have to build a new foundation slab, and this will shorten things a bit.)
By the start of the second week (tomorrow, April 13), the concrete pour around the new track should start. This is typically done in two “lifts” one for the layer including the ties and another for the layer including the track. This makes it easier to remove only the top layer, if necessary, for later repairs. My guess is that toward the end of the week, the concrete will be done (you cannot pour it all at once as some curing is required).
Then there are all of the auxiliary bits like the intersection approaches, new safety islands (2 in this case, expanded for accessibility eastbound and westbound on College). Depending on what might be planned for the curb lanes and sidewalks, that’s extra work, but it can be done in parallel with restored service once the concrete has cured. There is no point in putting streetcars back on track before this happens because the concrete will not set properly, and there will likely be gaps left between the rails and the concrete. This is not good because it provides a place for water to enter and freeze thereby undoing the effort to make an intersection last 20-30 years depending on the level of demand it serves.
What does Bechtel bring to the table for $80m that TTC and its Consultants couldn’t for all this time? I guess time will tell.
Inspired asked about Gondolas:
There are three systems in the US:
They are good for very specific functions hauling between two points, especially ones that need a quick change of elevation.
Robert, I find it interesting that the Star would also discuss the issue [of gondolas].
Steve: For what it’s worth, the article is three years old, and dates from a period when there was a gondola advocate active in Toronto. It’s a technology looking for an implementation.
I am not a huge fan of the idea of having something like this moving over the streets etc. I could see it in a very special application to link in something in the valley, however, it would be a very exceptional application. I would think that its construction would as Steve notes be very intrusive on a streetscape, and where it was in a business area with any density would limit development below it. It is hard to imagine it being able to manage a single car capacity of more than 60, and being able to run more than a car every 90 seconds or so. While the article discusses being able to match the ridership of the Queen car, the Queen car is far from the densest streetcar service in Toronto. If only Toronto, had left an allowance in the railway lands for a ROW, and would actually build LRT in the core the other way. Elsewhere it is hard to see application now in Toronto.
I’m late in offering a 2007 perspective from TheStrand.ca, the Victoria University student newspaper in an article by Andrea Lynn Palmer titled Toronto’s transit plan gets off track. It shows how perverted the “planning” has become, including a quote from then-chair Giambrone “There are better ways to service more people and get more riders….than the York subway”, and the article writer referred to a 2005 report by the TTC where the extension “ranks among the lowest options”, due to density lack and overall sprawl.
So we know the Sheppard line isn’t paying for itself; and now it’s becoming ever-clearer that beyond the York University area it will also be a white elephant, and now the fiscal conservatives are promoting a further extension of the subway in through Scarborough???
Before we go too much further down the road to yet another drain, where’s the business case analysis for the Scarborough extension? If it doesn’t exist yet, when will it? Why not halt it all until we see it, and it gets digested.
A smarter option I think would be to use the Gatineau hydro corridor for a busway: it’s on the diagonal to advantage some trips; touches on a few areas of density and demand; and it’s also on the surface, plus it’s owned by the province already through Hydro. I guess we don’t want to squeeze the billions though.