This photo taken from the rear platform of an observation car sometime between 1942 and 1959 shows
the ventilation building at the west portal of the Connaught Tunnel - Andy Morin.
When the Last Spike was driven in Canadian Pacific's main line on 7 Nov 1885 at Craigellachie the tracks climbed over
the Selkirk mountains through Rogers Pass. Snow conditions were so bad that the first winter there were no trains moving over
that portion of the line. CPR solved the problem by constructing 31 snowsheds in the area of
the pass over the next two years. Then on 4 Mar 1910 a huge avalanche came roaring down Mount Avalanche in Rogers Pass near
shed 17 and killed 58 men who were in the process of clearing a previous avalanche from Mount Cheops. In February 1913 they
announced the construction of a tunnel beneath the pass.
Such a tunnel would eliminate the need for maintenance of all those snowsheds, reduce the grade over the pass, and shorten
the length of their main line.
Construction of this tunnel was commenced on 2 Apr 1914 by the firm of Foley Brothers, Welch, and Stewart. It went
operational ahead of schedule on 16 Dec 1916.
Trackage abandoned: 14.5 miles (23.335 kilometers)
Snowsheds abandoned: 31
Snowsheds length: 4 miles (6.437 kilometers)
The newly built double-tracked tunnel was plagued with falling rock, wet rails causing locomotive slippage, and
insufficient ventilation. A concrete lining solved the loose rock problem and controlled water seepage. To solve the
ventilation problem a large fan building was constructed at the western portal of the tunnel.
Over the years, and still so today, railway rolling stock has gradually increased in size. In 1959 the double-track through
the tunnel was removed and a single track laid down the center of the tunnel to accommodate larger and higher equipment.
Just as the Rogers Pass route with its snowsheds and looping track was a bottleneck for the railway in the early years,
over time the Connaught Tunnel became a bottleneck too. Operation westbound through the tunnel required pusher locomotives to
assist heavy tonnage up the hill to the eastern portal on a grade of 2.2 percent. An operational nightmare. So to solve the
problem once more Canadian Pacific turned to tunneling. In 1988 a second route with two tunnels and a viaduct were put into
operation. The short tunnel was named Mount Shaughnessy Tunnel (1.1 mile 1.77 kilometers long). The main tunnel was named the
Mount Macdonald Tunnel. Lying 360 feet (109.728 metres) lower than the Connaught Tunnel and 9.11 miles (14.645 kilometers)
long, the Mount Macdonald Tunnel reduced the grade over the Selkirks to only 1 percent on this new
Currently, loaded and heavier westbound trains use the lesser grade of the Mount MacDonald Tunnel route while the lighter
eastbounds use the Connaught. However, this can change. For example, in 2002, a new tunnel lining was needed on the Mount
Shaughnessy Tunnel which required the tunnel's closure to install it. All trains were therefore forced to use only the
Connaught Tunnel. This necessitated reinstating pusher service from Rogers up to Stoney Creek siding near the east portal of
the Connaught Tunnel. (See the Canadian Pacific employee news article "Push Comes to Shove in Rogers Pass"
Operational conditions vary, so beware, you may find trains travelling in any direction on either route. Beware, indeed,
that applies to any track anywhere.
CP 9159 East exits the east portal of the Connaught Tunnel - Date/Photographer
Here is more information that has come to light regarding the Connaught tunnel:
For many years, there was a row of CPR houses behind the west portal of the Connaught tunnel. These housed the fan
operators and their families. Contrary to popular legend, the fans were never steam operated. Ancient one lung diesel engines
connected to a large flywheel and enclosed squirrel cage fans forced air into the mouth of the tunnel via a short wooden false
There was some skill involved in starting the one lung diesels. Once in full operation, the machines made a loud wooshing
sound. Because of the noise generated by the fans, CP had a gong attached to the main line signal eastbound at the east end of
Three quarters of a mile inside the tunnel was a lunar searchlight signal, connected to a short ABS block. When a westbound
train entered this block, the bell on the signal outside the tunnel would start to ring. The control (home) signal for Glacier
East was for many years located just inside the tunnel portal.
Sometime in the late 1960's, the operation of the fans was transferred to the train dispatchers in Revelstoke. For
westbound trains entering the short block between the lunar and the east switch meant the fans were turned on automatically.
The train dispatcher also had a start-stop switch as well as a status light on his/her panel.
Failures of the fans to start and stop were not uncommon, and following automation, there was no signal maintainer on duty
at night to fix or monitor the fans.
Sometime circa 1974, the Mountain Subdivision dispatcher had a steady fail light on his panel, and asked the crew of a
westbound train going through the tunnel if there was any indication the fans were working.
When they exited the tunnel, they noted that the fan house was on fire. The original one lung engines were badly damaged
and had to be replaced. The new engines to power the fans didn't require the tall steel smokestacks used by the original
engines. The replacement engines didn't have a long service life, and in turn, they caught fire circa 1980, this time
destroying the two tall fan houses.
The fan complex had to be entirely rebuilt as a lower structure. I think this is when the original engines were dumped
beside the rail yard at Glacier. (A photo of the engine cylinders lying outside the fan
building shows the date to be July 1972.)
The fans still work at Glacier, but see much less use.
CP's first EMD 710 engines were installed in the fan houses on the MacDonald tunnel. Again, the fan houses are automated,
but have a much more sophisticated monitoring system for the Mountain Subdivision RTC's in Calgary to keep an eye on.
Phil Mason - 20 Apr 2013.
Still More Information
Don Thomas has provided still more information regarding the Connaught tunnel:
It has commonly been suggested that the 1910 avalanche was the last straw for CP and the Rogers Pass line, and led directly
to the decision to construct the Connaught Tunnel. This has been refuted by Gary Backler in his 1981 graduate thesis at the
University of British Columbia, "The C.P.R.'s Capacity and Investment Strategy in Rogers Pass, B.C., 1882-1916."
Backler's research demonstrated that the avalanche was in fact not directly responsible for CP's subsequent decision to build
the Connaught Tunnel. CP correspondence and files made it clear that the 1910 avalanche was considered a freak event at the
time, as it came from a direction completely different from the usual avalanche paths in Rogers Pass. The cost of constructing
a tunnel was found not to be worth the expense under conditions existing at that time. What changed CP's mind soon after was
the need to increase capacity of the line, and specifically the 1912 decision to double track the main line between Calgary
and Vancouver. This was an extension of the double tracking then under way between the Lakehead and Calgary, as well as
elsewhere on CP. Traffic was increasing so rapidly that CP quickly began planning major changes to the mountain lines to
accommodate a second track. This may seem ironic in view of the fact that the Canadian Northern and Grand Trunk Pacific were
nearing completion and would compete for the same traffic the second track was intended to carry. Perhaps CP believed that a
greater track capacity would help it compete for this traffic and help mitigate the steeper grades and higher summits it
traversed. CP may have anticipated that business would increase so much that even with competing railways there would still be
an overload of traffic without the planned second track.
However that may be, the need for a second track through Rogers Pass, with heavy expenditures to build double track
snowsheds, and to duplicate the extensive loop viaducts west of the pass and the large bridges on the east slope, made a
tunnel more economical for the first time. Even then, however, a long tunnel roughly comparable to the present Macdonald
Tunnel, or even one intermediate in length and elevation between Macdonald and Connaught, and with lower grades than the
ruling 2.2 percent, was not found to be economic. After several iterations a five mile tunnel was found to be economic, and
the present Connaught Tunnel was chosen. However it was initially planned to have a different east approach. This would have
left the original line near Rogers, roughly where the Connaught and Macdonald tracks divide today, and rise at 1 percent much
as today's Macdonald track does. About three and a half miles east of the Connaught Tunnel (roughly where Wakely siding is
today) the grade would have increased to 2.2 percent, then eased off to just under 1 percent close to the east tunnel mouth.
The purpose of this was to limit the length of grade greater than 1 percent to the minimum necessary to reach the elevation
required for a five mile tunnel. The intention was to electrify from the base of this grade, through the tunnel to the present
Glacier station. The chosen profile allowed for the shortest mileage of electrification, and consequently the smallest number
of locomotives. Apart from the steep eastern approach grade, the electrification would have been similar in concept to those
recently established for the Hoosac Tunnel or the first Cascade Tunnel, which were restricted to the tunnels and their
immediate approaches. And if electrification had not proceeded, the pusher territory would have been comparable to, but
shorter than, the existing line west of Rogers.
Since tunnel construction would take the greatest time of all the work required for this capacity project, it was started
first. Subsequently the First World War caused most other phases of the double tracking project to be suspended or cancelled.
So much had already been invested in the tunnel that it was now economic to continue building it. However the cost of the
proposed new eastern approach for only one track was not justified by the small operating saving it would produce. The most
economical way to approach the east end of the tunnel was to use as much of the existing east slope as possible. The route
chosen diverged from the old line just around the curve west of Stoney Creek bridge, at a level grade until about half a mile
from the east portal. From this point the grade gradually increased until it reached the tunnel's own grade of just under one
percent, just east of the tunnel itself. This new alignment had the advantage of retaining the substantial investments in
Stoney Creek bridge and all the other structures east of it including Surprise Creek and Mountain Creek, which would have been
written off under the proposed lower double track alignment. By the time traffic increased again, larger locomotives allowed
this line to handle it without the need for the lower approach, for a period of over fifty years. When a new tunnel was
needed, the economics ruled against a new double track line (since Connaught would serve for eastbound traffic), but they did
allow for a maximum 1 percent grade and longer tunnel, that had not been found economic in 1913.
Obviously a double track between Calgary and Vancouver would have affected other noteworthy sections of the line. In 1912,
when the Spiral Tunnels were only three years old, alignments were proposed for second tunnels beside them. Between Revelstoke
and Clanwilliam in Eagle Pass, there was a proposal for a viaduct beginning at the west end of Revelstoke yard, at the point
where the track today begins to descend toward the Columbia River bridge. This viaduct would have begun to rise at 1 percent,
heading to the right (more northerly) of the present bridge, then curving west to parallel the old line at a higher level and
on an easier grade. The new line would return to the old alignment close to where the present old and new lines rejoin going
west, which is where the grade on the old track eases out to about 1 percent. This line would have been a little shorter than
the present new (westbound) track which similarly has a 1 percent maximum, but the cost of the viaduct would probably have
been greater (adjusted for inflation) than the cost of building the new track on embankments in 1980.
Various alternatives were proposed to reduce the grade on Notch Hill as part of double tracking, including a tunnel and a
new line across the shallow water offshore from Salmon Arm. None of them resembled the loop built for the present westward
The grade between Lake Louise and Field would have been lowered using a different route than the one adopted in the 1970s,
rather closer to the original track and to Bath Creek. The principle would have been the same as with the present westbound
track, with the increased grade beginning close to the station instead of starting to rise gradually before becoming much
steeper at the top, as the original (eastbound) track does.
All of these works would have provided a maximum 1 percent grade against westbound freight, the dominant direction of
traffic between Calgary and Vancouver. This goal was not achieved until the present Lake Louise, Clanwilliam, Notch Hill
grades, and Macdonald Tunnel were all completed by the 1980s. Elsewhere on this route, double track would have served to
increase capacity but not, for the most part, to reduce grades. In some areas, particularly the Thompson and Fraser Canyons,
both finding places for the second track, and constructing it, would certainly have been challenging.
Don Thomas - 16 Apr 2014.
Editor's Note: The current co-production track through the Thompson and Fraser canyons certainly solves the
problem of constructing a second track.
And Still More Information
Doug Mayer has provided even more information, a drawing, plus two new photos regarding the Connaught tunnel. The photos
and drawing may be found with the thumbnail photos on the left side of this page.
As you may know, the Connaught Tunnel will be 100-years-old in December 2016. Here's a sampling of what I know so far. Some
of these "known" facts may turn out to be false:
The first official train through the tunnel was on 9 Dec 1916. Contrary to popular legend, the tunnel fans were never
powered by steam. They were powered by large McIntosh and Seymour 4-cylinder oil engines of an early design. The engines
weighed 156,000 pounds and were over 20 feet long and over 14 feet high. They operated at a speed of 190 rpm and burned Bunker
C fuel oil. The engines were started with diesel fuel and then switched over to bunker C. The bunker C fuel required heating
to keep it viscous, so a boiler room was built below the engine room on the south side. This may be where the steam legend
came from. The fans were 12 feet 3 inches in diameter and 8 feet 3 inches wide. They were capable of delivering 700,000 cubic
feet of air per minute. (Thumbnail photo number five shows a tank car on the building's siding which probably delivered the
Bunker C fuel oil.)
Between 1921 and 1925 the tunnel was relined with concrete throughout it's length. Originally, only 7,800 feet of it were
lined at the ends. The relining project was done while the tunnel was in use.
In 1931 the diverted Illecillewaet River broke through it's banks and flooded the west portal of the tunnel and it's
approach with debris. (see the thumbnail photo).
In 1958 the tunnel was converted to single track.
In 1972 one of the fans houses had a fire that destroyed it's engine. At this time the old engines were discarded and newer
type engines were installed. In 1976 another fire destroyed both fan houses requiring another rebuild with the result that the
fan houses are now much lower. Some time after that the diesels were replaced by electric motors. It should be noted however
that the original fans still exist.
Doug Mayer - 25 Apr 2014
We're Not Done Yet
Research by Doug Mayer has come up with even more data. The photos have been added to the list on the left side of this
My last entry in the article states that both diesel engines were replaced by electric motors. This is incorrect. Only the
engine in the south tower was switched to electric operation. The north tower is powered by a modern diesel unit. I think a
Detroit Diesel, but it could be a Cummins. I've attached photos of the electrical and diesel power units and one of the north
I've also attached a 1916 drawing of the fuel system for the original diesels. It is from a drawing I acquired and had
scanned by a local business. The original drawing is over 6 feet long.
If you study the drawing carefully you will see that the boilers in the drawing are different than the ones in the Tom
Parkin photo. It's possible that the boilers were changed at some point, but I think the drawing is not an
"as built" as they installed different boilers when they built the fan house.
You'll also notice that there is a large fuel tank shown inside the structure. It's still there. I've attached a photo.
There has been speculation that the tank was from an old tank car, but I don't think this is correct. It is much too large. It
was probably assembled on site from sections or would have had to be a special over-width movement to get it there. It just
barely fits in the room.
I've found someone in Revelstoke that tended the boilers and engines. They were started on kerosene and then switched to
bunker C fuel. To start them was quite a process. The engine valves were opened to eliminate the compression and then
compressed air was applied to the cylinders. Once the engine was up to speed on air, you closed the valves on one cylinder and
applied the kerosene. Once you got that cylinder working you did the same thing on the next cylinder, and then the next, and
the next. Once you got all four cylinders running on kerosene you switched over to bunker C.
The original engines were provided with air compressors as part of the power units. At some point the railway provided a
single stage steam powered air compressor to augment the engine powered compressors. The steam powered unit is still there.
See the photo. It is hoped we can acquire the compressor from Canadian Pacific and place it on display at the Revelstoke
If the engines failed to start you could run out of air. As the boilers and diesels were separate systems, if you ran out
of air you could recharge the air tanks from this standby compressor.
Lastly, the first official train through the tunnel was on 9 Dec 1916. The Revelstoke Railway Museum and I will be doing
something to celebrate that on 9 Dec 2016. Watch the museum web site for information about this.
Railway Contractors, The
Taylor, Geoffrey W.
Morriss Publishing 1745 Blanshard St. Victoria BC
1988 - Hard cover
6.25 x 9.25 inches - 16 x 23.5 centimetres
Information about the company that constructed the Connaught Tunnel.