VIA Rail

[crs1026]

It isn't so much how long the passing tracks are but what percentage of the ROW still has passing tracks. According to this article, they removed
"60 miles of double track on the 104.9-mile Winchester Subdivision," so that works out to 57% (between a 1/2 and 2/3). VIA could either build new track parallel to the passing tracks or they could use the passing tracks for their dedicated track in exchange for CP being able to use the VIA's track as double track when VIA isn't using it (or a combination of the two).

There are now 4 longish (5 mile or so) sidings (passing segments, really) spaced about 10 miles apart, and double track still in place east of St-Telesphore. The turnouts at the ends of the passing segments are reportedly good for 45 mph which is pretty good.

The question is, how many VIA trains and how long do they occupy the territory. Supposing the remaining track is as good as it was in 1966, it would likely be fine-tunable to 90 mph except on curves, which aren't that frequent or restrictive. So those six hypothetical bypass trains, running in alternate directions, would each consume the single track for maybe an hour and change each. We don't need to make any assumptions that VIA would seek to upgrade that track to the standard of the full HFR line.... ie no 177 km/hour speeds needed.

At current freight volume there will seldom if ever be more than two CP trains in that segment, at least for the next 5 years. Even if there were four, they would all fit into the passing zones...and the amount of delay they would encounter from a single VIA is pretty minimal.

My point being, the operability of VIA (at a low frequency) on that route is pretty darn good. And the capital cost of getting to that 90 mph is probably pretty low.

That doesn't imply that CP is all that thrilled about having VIA on that section.... but it might not look that bad a proposition if the price were right. I do wonder if there might be a quid pro quo somewhere that kept CP from laughing Ottawa out of the room when this idea was proposed.

It's still a poor choice - especially since there is little upward growth potential without putting some of the former track back. But on paper it might not look like a terrible investment.

- Paul

 

[Bordercollie]

Business class isn't quite as nice as it was pre-pandemic. The liquor service has been reduced to beer, wine, or soft drinks only; previously they served broad choice of drinks prior to meal service and liqueurs afterwards. I understand the need for the change - it limits their interaction with passengers - but I miss it.

I won't miss the HEP-II coaches used on the corridor though.

The rebuilt HEP II care are almost as comfortable as an LRC. You would be surprised at the difference. They are much quieter than before they where rebuilt.
Just have to make sure that the doors are closed on either end.

What is the track speed on the Winchester sub?

 

[Urban Sky]

What is the track speed on the Winchester sub?

From a Railway Investigation Report in 2010:

https://bst.gc.ca/ENG/rapports-reports/rail/2010/r10d0077/r10d0077.pdf

 

[nfitz]

As far as I know, CP has since decommissioned one of tracks, to save money on maintenance.

 

[Urban Sky]

As far as I know, CP has since decommissioned one of tracks, to save money on maintenance.

Refer back to just a few posts ago:

It isn't so much how long the passing tracks are but what percentage of the ROW still has passing tracks. According to this article, they removed
"60 miles of double track on the 104.9-mile Winchester Subdivision," so that works out to 57% (between a 1/2 and 2/3).

As far as track speeds (and especially: track Classes) are concerned, I doubt that they have improved since then...

 

[crs1026]

^Class 4 track is rated by TC regs at 80 mph for passenger,

Some high priority intermodal trains used to be allowed 65mph on the Winchester - I don’t know if that’s still the case.

- Paul

 

[Urban Sky]

Thanks for the great analysis! Certainly if the capital cost to build the bypass per km is the same as the entire TOM HFR project, your calculations are correct. The problem is I don't think that is a valid assumption. CP removed 2/3 of the double track (leaving 1/3 as passing tracks) less than 1.5 years ago, meaning that the base should still be in good condition. Also, most of the ROW is very straight, so little if any adjustments to the curves will be necessary.

I don't know what a good cost per km would be, but (reverse engineering your calculations) if we used the same as OTTW-MTRL, you would get the following results. As I said, I don't know if this would be correct either, so if anyone wants to try anything other values, send me a PM, and I will send you the link to the Google Sheets sheet I made.

Metric	HFR TRTO-OTTW	HFR OTTW-MTRL	HFR TRTO-OTTW-MTRL (Base Case)	Ottawa Bypass	vs. HFR without Ottawa Bypass
Capital Cost	$2.1 billion	$91.5 million	$2.2 billion	$74.2 million	+3%
Segment length (km)	400	180	580	146	+25%
Frequencies/day	15	15	15	6	+40%
Trains/year	10,950	10,950	10,950	4,380	+40%
Train-km/year	4,380,000	1,971,000	6,351,000	639,480	+10% (ignoring everything outside the Ottawa Bypass)
Capital cost per km	$5.3 million	$508.3 thousand	$3.8 million	$508.3 thousand
Capital cost per year (assuming 40-year asset life)	$52.5 million	$2.29 million	$54.75 million	$1.86 million	+3%
Capital cost per train movement	$4,795	$208.90	$5,000	$424	+8% (ignoring everything outside the Ottawa Bypass)
Capital cost per train-km	$11.99	$1.16	$8.62	$2.90	+34%

Thank you for volunteering a modification to my table, which is highly appreciated!

When looking at that $508.3 thousand per km capital cost, I struggle to believe that you could obtain any guaranteed slots out of the Winchester Sub at such a low-ball capital cost (recall that CN charged VIA $4.5 million for every km of triple-tracking and that was about a decade ago).

Maybe let's try to first break down the travel time between Montreal and Ottawa:

• MTRL-DORV (19 km): If we go to the last pre-Covid schedule, the fastest scheduled travel time from MTRL to DORV (departure time in both cases, i.e. including dwell time in DORV) was 24 minutes.
• De Beaujeu-OTTW (114 km): If we now look at the fastest scheduled travel time between OTTW and ALEX, we get 44 minutes pre-Covid and 37 minutes as a historic low (achieved by train 39 in October 2005). If we extrapolate this travel time from 87 to 114 km, we obtain a travel time of 55 or 48 minutes (depending on whether we use the pre-covid fastest or the historic fastest).
• DORV-De Beaujeu (47 km): This means that in order to achieve the travel time of 93 minutes in that Globe&Mail article I keep referring to, you would need to achieve a travel time of 14 or 21 minutes. 47 km in 14 minutes would equal an average speed of 202 km/h, which exceeds the maximum speed of VIA's new trainsets, but the same distance in 21 minutes equals an average speed of 137 km/h, which would be slightly less than what 87 km in 37 minutes represented back in 2005 (i.e. 141 km/h).

The above means that HFR needs Track Class 5 and tracks separate from CP's much slower freight trains for most of the distance. Happily, both seems to be achievable at relatively low cost except for a relatively short segment from Sainte-Anne-de-Bellevue and across the Ile-Perrot to Dorion. If we assume that the entirety of the $91.5 million is spent on the Winchester Sub alone, we now have a per-mile capital cost of $1.95 million, which is basically half of the per-mile cost estimate for HFR.

Consequently, the construction cost of the Ottawa Bypass would be $284.2 million (representing a cost premium of 13% rather than 3% on the capital cost of the HFR Base Case) and still 29% higher per train-km than the HFR trunk line chosen as the base case. Put differently: total capital costs of the Ottawa Bypass would be three times as high as in the Base Case for the entire OTTW-MTRL segment, despite covering a shorter distance (146 vs. 180 km), and per-train-mile they would be eight times as high:

Metric	HFR TRTO-OTTW	HFR OTTW-MTRL	HFR TRTO-OTTW-MTRL (Base Case)	Ottawa Bypass	vs. HFR without Ottawa Bypass
Capital Cost	$2.1 billion	$91.5 million	$2.2 billion	$284.2 million	+13%
Segment length (km)	400	180	580	146	+25%
Frequencies/day	15	15	15	6	+40%
Trains/year	10,950	10,950	10,950	4,380	+40%
Train-km/year	4,380,000	1,971,000	6,351,000	639,480	+10% (ignoring everything outside the Ottawa Bypass)
Capital cost per km	$5.3 million	$508.3 thousand	$3.8 million	$1.95 million	52% of T-O-M Base Case
Capital cost per year (assuming 40-year asset life)	$52.5 million	$2.29 million	$54.75 million	$7.11 million	+13%
Capital cost per train movement	$4,795	$208.90	$5,003	$1,622	32% of T-O-M Base Case (ignoring everything outside the Ottawa Bypass)
Capital cost per train-km	$11.99	$1.16	$8.62	$11.11	129% of T-O-M Base Case

 

[kEiThZ]

The next obvious question: can ~$300 spent elsewhere, save close to the 15 mins in rolling time that the bypass would save?

 

[Urban Sky]

The next obvious question: can ~$300 spent elsewhere, save close to the 15 mins in rolling time that the bypass would save?

Not to mention that every minute shaved off the T-O-M trunk (rather than an Ottawa Bypass) would not just be saved on MTRL-TRTO, but also on either MTRL-OTTW or OTTW-TRTO...

 

[roger1818]

Thank you for volunteering a modification to my table!, which is highly appreciated!

When looking at that $508.3 thousand per km capital cost, I struggle to believe that you could obtain any guaranteed slots out of the Winchester Sub at such a low-ball capital cost (recall that CN charged VIA $4.5 million for every km of triple-tracking and that was about a decade ago).

Maybe let's try to first break down the travel time between Montreal and Ottawa:

• MTRL-DORV (19 km): If we go to the last pre-Covid schedule, the fastest scheduled travel time from MTRL to DORV (departure time in both cases, i.e. including dwell time in DORV) was 24 minutes.
• De Beaujeu-OTTW (114 km): If we now look at the fastest scheduled travel time between OTTW and ALEX, we get 44 minutes pre-Covid and 37 minutes as a historic low (achieved by train 39 in October 2005). If we extrapolate this travel time from 87 to 114 km, we obtain a travel time of 55 or 48 minutes (depending on whether we use the pre-covid fastest or the historic fastest).
• DORV-De Beaujeu (47 km): This means that in order to achieve the travel time of 93 minutes in that Globe&Mail article I keep referring to, you would need to achieve a travel time of 14 or 21 minutes. 47 km in 14 minutes would equal an average speed of 202 km/h, which exceeds the maximum speed of VIA's new trainsets, but the same distance in 21 minutes equals an average speed of 137 km/h, which would be slightly less than what 87 km in 37 minutes represented back in 2005 (i.e. 141 km/h).

The above means that HFR needs Track Class 5 and tracks separate from CP's much slower freight trains for most of the distance. Happily, both seems to be achievable at relatively low cost except for a relatively short segment from Sainte-Anne-de-Bellevue and across the Ile-Perrot to Dorion. If we assume that the entirety of the $91.5 million is spent on the Winchester Sub alone, we now have a per-mile capital cost of $1.95 million, which is basically half of the per-mile cost estimate for HFR.

Consequently, the construction cost of the Ottawa Bypass would be $284.2 million (representing a cost premium of 13% rather than 3% on the capital cost of the HFR Base Case) and still 29% higher per train-km than the HFR trunk line chosen as the base case. Put differently: total capital costs of the Ottawa Bypass would be three times as high as in the Base Case for the entire OTTW-MTRL segment, despite covering a shorter distance (146 vs. 180 km), and per-train-mile they would be eight times as high:

Metric	HFR TRTO-OTTW	HFR OTTW-MTRL	HFR TRTO-OTTW-MTRL (Base Case)	Ottawa Bypass	vs. HFR without Ottawa Bypass
Capital Cost	$2.1 billion	$91.5 million	$2.2 billion	$284.2 million	+13%
Segment length (km)	400	180	580	146	+25%
Frequencies/day	15	15	15	6	+40%
Trains/year	10,950	10,950	10,950	4,380	+40%
Train-km/year	4,380,000	1,971,000	6,351,000	639,480	+10% (ignoring everything outside the Ottawa Bypass)
Capital cost per km	$5.3 million	$508.3 thousand	$3.8 million	$1.95 million	52% of T-O-M Base Case
Capital cost per year (assuming 40-year asset life)	$52.5 million	$2.29 million	$54.75 million	$7.11 million	+13%
Capital cost per train movement	$4,795	$208.90	$5,003	$1,622	32% of T-O-M Base Case (ignoring everything outside the Ottawa Bypass)
Capital cost per train-km	$11.99	$1.16	$8.62	$11.11	129% of T-O-M Base Case

You do bring up some valid points. A couple thoughts though:

First of all, while for HFR OTTW-MTRL, a significant portion of the cost would directed towards DORV-De Beaujeu, I don't think it would be 100%. For one, in order to achieve that historic low travel time of 37 minutes between OTTW and ALEX, some upgrades would be necessary, especially if they are wanting to run 15 trains a day (my thinking is more/longer sidings). I also think some upgrades will be needed to CN's Montreal Sub (especially near where it crosses the Lachine Canal) to manage the extra traffic and allow for a faster approach into Central Station.

Secondly, how much the Winchester sub needs to be upgraded really depends how many trains they want to run along it. If they are only running a couple trains a day, they might (with some quid pro quo as mentioned by crs1026) be able to negotiate a deal with CP to give them priority on the existing track with minimal upgrades. While class 4 certainly isn't optimal, if they don't have to slow down, the top speed of 80 mph (129 km/h) should allow trains to travel from De Beaujeu to Smiths Falls in less than 90 minutes (an average speed of 97 km/h). If they want to run 4-6 trains a day, they would could likely get away with only rebuilding the double track that CP removed (connecting the passing tracks). Only if they wanted full HFR service on the bypass they would likely require fully dedicated track that is independent from CP's track. I agree that last option would be very expensive and the money would be better spent elsewhere..

 

[crs1026]

Secondly, how much the Winchester sub needs to be upgraded really depends how many trains they want to run along it. If they are only running a couple trains a day, they might (with some quid pro quo as mentioned by crs1026) be able to negotiate a deal with CP to give them priority on the existing track with minimal upgrades. While class 4 certainly isn't optimal, if they don't have to slow down, the top speed of 80 mph (129 km/h) should allow trains to travel from De Beaujeu to Smiths Falls in less than 90 minutes (an average speed of 97 km/h). If they want to run 4-6 trains a day, they would could likely get away with only rebuilding the double track that CP removed (connecting the passing tracks). Only if they wanted full HFR service on the bypass they would likely require fully dedicated track that is independent from CP's track. I agree that last option would be very expensive and the money would be better spent elsewhere..

Getting back to Class 5 on the Winchester can’t be that expensive, because there are no major civil projects required…. no new separations, no curves to be eased, no new rail. Perhaps some tie replacement , surfacing, or ballasting… and maybe renewing crossings or fixing low spots. It’s hard to say what CP would ask to maintain that for a decade or more… it certainly implies more maintenance done more often, but not a new investment in bigger or better rails or track materials.

I’m not even sure they need to relay much of the former double track, unless they are building in an option for growth. Once VIA needed to have its trains meet and pass each other, more and longer passing sections might be needed, but a morning, midday, and afternoon service plan leaves CP lots of capacity. There might have to be a one-hour curfew before each VIa slot at each end, to remove any risk of a VIA overtaking a slower freight. Again, there is no need for new grading or culverts so the nominal figure based on Kingston Sub triple tracking is high.

- Paul

 

[Urban Sky]

You do bring up some valid points. A couple thoughts though:

First of all, while for HFR OTTW-MTRL, a significant portion of the cost would directed towards DORV-De Beaujeu, I don't think it would be 100%. For one, in order to achieve that historic low travel time of 37 minutes between OTTW and ALEX, some upgrades would be necessary, especially if they are wanting to run 15 trains a day (my thinking is more/longer sidings). I also think some upgrades will be needed to CN's Montreal Sub (especially near where it crosses the Lachine Canal) to manage the extra traffic and allow for a faster approach into Central Station.

Should it? I thought the money was being spent on the segments that weren't already VIA owned and operated.

VIA has been working for years to incrementally upgrade this track. If you look back at 1988 the best time to Casselman was 31 minutes, compared to 25 minutes today.

I'm with @nfitz on this one: I don't see how the tight budgets communicated so far ($2.1 billion for TRTO-OTTW and $91.5 million for OTTW-MTRL) leave any room for improvements of VIA's existing frequency and whereas complying with the new Track Class 6 in order to achieve speeds of 110 mph will for sure require higher track maintenance levels, these improvements would also benefit VIA's legacy Corridor services (even if HFR doesn't get built at all), as - HFR or not - all Corridor trains will be able to achieve 110 mph. I'm not even sure these improvements would count as capital expenditure, but that depends on how profound the new requirements will be. Even the construction of new sidings (which undoubtedly counts as CapEx) would benefit Corridor services in a non-HFR future, as increased capacity reduces delays and thus increases OTP, especially when the number of frequencies can't increase. Therefore, I wouldn't count any improvements between Smiths Falls and De Beaujeu (excluding the connectors to the Belleville and Winchester Subdivisions as HFR-related...

Fixing the math ... 15 minutes extra travel time, add 5 minutes to stop ... so 20 extra minutes instead of 30.

15 extra minutes. The Express train could just roll through Ottawa Station without incurring the time penalty from stopping there...

It isn't so much how long the passing tracks are but what percentage of the ROW still has passing tracks. According to this article, they removed
"60 miles of double track on the 104.9-mile Winchester Subdivision," so that works out to 57% (between a 1/2 and 2/3). VIA could either build new track parallel to the passing tracks or they could use the passing tracks for their dedicated track in exchange for CP being able to use the VIA's track as double track when VIA isn't using it (or a combination of the two).

According to my Canadian Trackside Guide 2021, the second track has been removed on the following sections of the Winchester Sub:

Location (at start of section)	(MP)	Length	# of tracks
Dorion	18.9	20.01	2
St. Telesphore	38.91	13.04	1
Glenroy	51.95	5.96	2
Apple Hill	57.91	12.13	1
Avonmore	70.04	4.49	2
Finch	74.53	12.12	1
Winchester	86.65	4.59	2
Inkerman	91.24	11.96	1
Bedell	103.2	7.05	2
Burritts	110.25	9.72	1
Rosedale	119.97	3.83	2
Smiths Falls

This suggests that 58.97 (or 56.2%) out of 104.9 miles of second track have been removed and all (!) of that West of De Beaujeu (MP 35.5).

Secondly, how much the Winchester sub needs to be upgraded really depends how many trains they want to run along it. If they are only running a couple trains a day, they might (with some quid pro quo as mentioned by crs1026) be able to negotiate a deal with CP to give them priority on the existing track with minimal upgrades. While class 4 certainly isn't optimal, if they don't have to slow down, the top speed of 80 mph (129 km/h) should allow trains to travel from De Beaujeu to Smiths Falls in less than 90 minutes (an average speed of 97 km/h). If they want to run 4-6 trains a day, they would could likely get away with only rebuilding the double track that CP removed (connecting the passing tracks). Only if they wanted full HFR service on the bypass they would likely require fully dedicated track that is independent from CP's track. I agree that last option would be very expensive and the money would be better spent elsewhere..

Getting back to Class 5 on the Winchester can’t be that expensive, because there are no major civil projects required…. no new separations, no curves to be eased, no new rail. Perhaps some tie replacement , surfacing, or ballasting… and maybe renewing crossings or fixing low spots. It’s hard to say what CP would ask to maintain that for a decade or more… it certainly implies more maintenance done more often, but not a new investment in bigger or better rails or track materials.

I’m not even sure they need to relay much of the former double track, unless they are building in an option for growth. Once VIA needed to have its trains meet and pass each other, more and longer passing sections might be needed, but a morning, midday, and afternoon service plan leaves CP lots of capacity. There might have to be a one-hour curfew before each VIa slot at each end, to remove any risk of a VIA overtaking a slower freight. Again, there is no need for new grading or culverts so the nominal figure based on Kingston Sub triple tracking is high.

- Paul

Let's do yet another back-of-the-envelope calcualtion:

Suppose that a VIA train can obtain an average speed of 90% of the maximum allowed speed at Track Class 4, 80% at Track Class 5 and a freight train can obtain 70% at Track Class 4 and 60% at Track Class 5. In this case, the travel times would be as follows between Dorval and Smiths Falls:

Segment	TT (VIA, Track Class 5)	TT (VIA, Track Class 4)	TT (Freight, Track Class 5)	TT (Freight, Track Class 4)
Assumed Track Speed	100 mph (161 km/h)	80 mph (129 km/h)	80 mph (129 km/h)	60 mph (97 km/h)
Assumed Average Speed	128.7 km/h	115.9 km/h	77.2 km/h	67.6 km/h
(in % of track speed)	80%	90%	60%	70%
Dorval-De Beaujeu (47 km)	21.9 minutes	24.3 minutes	36.5 minutes	41.7 minutes
De Beaujeu-Smiths Falls (142.1 km)	66.2 minutes	73.6 minutes	110.4 minutes	126.1 minutes
Combined (189.1 km)	88.1 minutes	97.9 minutes	146.9 minutes	167.9 minutes

That means that every VIA train operating on Track Class 4 will consume a departure window of 2:20 hours ([168-98]*2=140 minutes) and at Track Class 5 1:58 hours ([147-88]*2=118 minutes). Even without adding any headway buffers (determined by the frequency of intermediary signals), you can easily see that even operating 3 trains per day would wipe out a large proportion of daylight operating hours. Conversely, for Dorion-De Beaujeu, the corresponding figure would be 35 minutes for Track Class 4 ([42-24.5]*2) and 29 minutes for Track Class 5 ([36.5-22]*2) - thus enough to let CP's freight trains comfortably swim in-between hourly departures...

 

[crs1026]

That means that every VIA train operating on Track Class 4 will consume a departure window of 2:20 hours ([168-98]*2=140 minutes) and at Track Class 5 1:58 hours ([147-88]*2=118 minutes). Even without adding any headway buffers (determined by the frequency of intermediary signals), you can easily see that even operating 3 trains per day would wipe out a large proportion of daylight operating hours. Conversely, for Dorion-De Beaujeu, the corresponding figure would be 35 minutes for Track Class 4 ([42-24.5]*2) and 29 minutes for Track Class 5 ([36.5-22]*2) - thus enough to let CP's freight trains comfortably swim in-between hourly departures...

That's if you assume that VIA has exclusive use of the entire main line end-to-end from the moment a VIA train is due (signals lined up ahead of time) to the time it departs the other end. In reality that wouldn't be the case.

Say the day begins with the first express VIA westbound leaving Dorion-ish at 07:30. Eastbounds out of Smiths Falls may proceed, so long as they are able to reach one of the unoccupied passing zones ahead of the VIA. They might have to clear, and hold in the passing zone, for 20-30 minutes ahead of the VIA. CP would start holding westbounds out of Montreal around 05:30ish, so that the VIA westbound can clear Smiths Falls without catching up to and being delayed by a westbound. (And even that can be shaved, if there aren't any eastbounds due, as the passing tracks are available and a westbound freight can slide into one to be overtaken). Once the VIA reaches Dorion, westbound freights can be fleeted behind it, making whatever meets with eastbound freights are required.

Assuming a fairly symmetrical schedule, the eastbound VIA out of Toronto won't be met until the westbound VIA is west of Smiths Falls, so that becomes VIA's problem. When the eastbound shows up, the hold/proceed pattern reverses, but the delays to freight remain fairly brief.

And in fact, CP is down to 3-4 freights in each direction per 24 hour period, so even doubling their freight business doesn't overload this operating pattern. The only problem comes when VIA wants to run a second train in each of those windows.... dispatching becomes much more complicated, and freight delays are magnified, if passing capacity has to be reserved for a meet between two VIA's.

The biggest impact to CP is actually the shortening of track availability for maintenance of way. With only a few freights out there, MOW crews get longer stretches of track and time to fix things. If they have to clear every couple hours for a VIA, their productivity falls.

The absurdity is, in spite of everything we agree on about the need to get passenger away from freight - here we are applying brain power to find a clever solution to help VIA run on a freight line, and mitigate the risks and likelihood of problems in freight operations cascading to VIA. Maybe @nfitz has a point.... if we applied the same brain power to the CN line, VIA wouldn't need to move at all? I'm more inclined to think the Winchester thing is just a bad idea and we shouldn't try to polish a cannonball.

- Paul

 

[kEiThZ]

I'm more inclined to think the Winchester thing is just a bad idea and we shouldn't try to polish a cannonball

Just the brain power required on this forum too figure this out, should tell you how bad this will be in real life.

I don't buy for a second, that faster service to Montreal is about the business case. Something tells me, the government is just frustrated with the political blowback of spending $8B and getting 4:45 hr for TO-MTL. Understandable. But building a bypass is not the solution here. I think the public interest is better served and business case improved by improvements along the whole line, than one specific segment. Saving everybody 15 mins is much preferable to saving TO-MTL pax half an hour.

 

[roger1818]

If we were willing to sacrifice all local service on the Lakeshore, VIA certainly could run a few express trains along the Kingston Sub. The problem is, you can't have it both ways, and sacrificing service to the communities along the lakeshore is not an option.

I do agree with @kEiThZ that this may be more political than anything. Having said that, I do think by having a couple (i.e. 2) of high yielding express trains a day during the most popular travel times could be lucrative (a third, mid-day train, not so much).

While it is true that VIA could run the express trains through Ottawa on existing track without stopping, the problem is the track geometry isn't conducive for maintaining higher speeds. Straightening curves in urban areas would be expensive and, similar to costs of using a bypass, doing so wouldn't benefit trains that are slowing to a stop in Ottawa anyway. Using the Winchester sub as a bypass only becomes a reasonable option if the costs can be kept to a minimum.

VIA's track in Ottawa