Mitchell Point Tunnel and Viaduct
Carried the Historic Columbia River Highway through Mitchell Point, Hood River County, Oregon, east of milepost 58.
Date of Construction
John Arthur Elliott, locating engineer and designer
Standifer-Clarkson Co., contractor
Charles Nelson and Co., subcontractor
Oregon State Highway Department
None; closed, 1953; destroyed, 1966
One of four tunnels constructed on the Historic Columbia River Highway, and one of the earliest examples of highway tunnels in the United States constructed with adits used by motorists for viewing a scenic landscape.
Mitchell Point Tunnel and Viaduct
In 1914 and 1915, the Hood River and Wasco County courts contracted with John Arthur Elliott, a locating engineer, and his crews to prepare a plan and profile of the proposed Historic Columbia River Highway through their counties. Many voters in both counties were reluctant to spend the money on constructing the HCRH, including employing a locating engineer to survey a route. Hood River County only approved its bond issue to cover construction on the condition that local businessman and road supporter Simon Benson would guarantee to make up the difference if costs overran the bond. Indeed, Benson paid at least $13,000 for costs in Hood River County. Many citizens of Wasco County were equally reluctant to spend their own funds on a new highway. They were satisfied with the present county road system which included grades of up to 18 percent on routes between Hood River and The Dalles.
Travel along the Oregon side of the Columbia River from Portland to The Dalles was extremely difficult until construction of the Historic Columbia River Highway from 1913 to 1922. The first road passing along the river between these two points was begun in 1872, when the state of Oregon appropriated $50,000 to begin construction of the Troutdale to The Dalles Road. An additional $50,000 in 1876 furthered construction, but the completed road was narrow, crooked, and steep. In 1883, the Oregon Navigation Company constructed the first long-distance railroad through this section and in the process destroyed and detached many sections of the Troutdale to The Dalles Road. Only a few sections remained as primary local routes for inhabitants and their wagons.
When the HCRH was surveyed through Hood River County in 1914, Elliott continued the philosophy used on the Multnomah County section of the route – maintaining a 5 percent maximum grade and 100′ minimum turning radius. He encountered a host of obstacles in designing the Mitchell Point section in Hood River County. There, the old wagon road passed through the saddle between the 400′ Little Mitchell Point and the 1,100′ Big Mitchell Point at an elevation of 250′. The route included grades between 10 and 23 percent to bring it up to the saddle from near water level. Elliott feared that to carry the HCRH over the same pass he needed to “develop” distance to maintain maximum grade. An alternate water-level route, skirting around Lower Mitchell Point, was impossible because the CRN’s successor, the Oregon-Washington Railroad and Navigation Company (OWRN) main line from Portland to points east was located there. Elliott believed that a highway built through the saddle would be much longer, requiring 1 1/2 miles through “a rough and broken country” to reach the saddle, and then 1 1/2 miles down the other side across a shellrock slide. The alignment was poor and came with heavy maintenance expenses. Elliott chose instead to take a shorter, more direct route, but it required finding a location “which would not endanger the railroad and at the same time would not cost excessively.” He eventually found his route by cutting a ledge into a cliff, building a viaduct, and tunneling through Lower Mitchell Point.
One news report from July 1915 illustrates just how impassable the old road over the saddle on Mitchell Point was. According to a Portland Oregon Journal reporter, several parties of Portlanders and their motorcars ventured out along the Historic Columbia River Highway for a drive to Hood River, even though the old wagon road was still the only passage through the Mitchell Point section. The reporter chronicled most drivers’ failure to go further upon reaching Mitchell Point:
Some machines refused to climb the hill because the oil would settle back in the tank beyond reach of the motors, others had brakes the driver would not trust, but a great many machines were turned back when the man at the wheel took a look at the narrow, winding and rocky path with a wall of rock and gravel on one side and a death dealing abyss on the other.
Elliott saw his plan of a cliff-hugging road, viaduct, and tunnel as the practical solution for ending hair-raising motor travel in the Columbia River Gorge.
Elliott began his field survey of Hood River County on October 11, 1913 and completed it the following spring. He brought with him a locating party of 15 men, consisting of a chief of the party, draftsman, a seven-man transit party, levelman, level rodman, topographer, topog rodman, tapeman, and a cook. Throughout the county, including the Mitchell Point section, they began by running a baseline, or preliminary line, following closely the desired highway location. From this, the party ran levels and noted topography for about 100′ on either side of the line. They staked the projected center line, drew maps, and made profiles.
According to Elliott, the party used two methods of location, contour and cross-section. First, the men located lines of equal elevation at 5′ intervals. On cliffs they used a second method; they located ground breaks by elevation and distance from the base line.
Design and Description
Construction began on the Mitchell Point section of the Historic Columbia River Highway on March 23, 1915. At the western end, the highway’s alignment left the wagon road’s route. The first challenge from this point was to round a cliff that was too high and too expensive to take out as an open cut. Elliott found that he could hold a line out as far as possible, undercutting the narrowest possible ledge from the cliff for the roadbed (essentially a half-tunnel) and constructing masonry retaining walls to gain width. From there he built a 192′ reinforced-concrete slab viaduct over a shellrock talus slope, before then cutting a 390′ windowed tunnel through Lower Mitchell Point. From the east portal, the road continued on to rejoin the wagon road’s alignment. The total distance of the Mitchell Point section was .84 miles.
At the west end of the Mitchell Point section lay what Elliott believed was the most difficult portion of the project. This involved undercutting the cliff for half the width of the roadbed without jeopardizing the OWRN mainline below. The contractors took great care in preventing damage to the track. Before dynamite blasting ever began, the Western Union telegraph and OWRN communication lines were taken off their poles and buried. Prior to each dynamite blast, the railroad semaphore signals were removed, and the track was protected by covering the rails with ties. Then, the contractor blasted away only small sections of the cliff and hurriedly removed it from the track below between train schedules.
Undercutting the cliff was a long and tedious process because of the need to use only small dynamite shots or only black powder to loosen the rock so that it would fall only to the grade, and carting it away in ore cars. Nevertheless, the contractor spent much time protecting the tracks and removing the small amounts of rock that fell on them so that the OWRN could maintain its schedule. On May 3, 1915, one blast showered the track with so much rock that it took over two hours to clear and also delayed a train. The remaining rock left after the blast was unstable and prone to fall at any moment. The contractor, Standifer-Clarkson, finally convinced the OWRN to permit one large shot to clear the remaining 300 to 500 yards of material and with the help of railroad crews and machinery, cleared the track of the rubble with a minimum of delay for train service along the gorge.
On May 10th, at 12:25 p.m., the big shot was fired. Cut holes loaded with 40 percent dynamite defined the slope. Following closely behind them were lift holes, loaded with 17 kegs of black powder, which lifted much of the cut and dropped it over the cliff on to the OWRN main line, 90′ below. The blast covered the tracks with 20′ of debris. After an hour of cleaning up it was evident to all involved that the rock would not be cleared by midnight as intended. Work continued all afternoon and into the night, and appeared to be finished at 10:30 the following morning, 22 hours after the blast. Yet, additional rock loosened by the blast continued falling, and a second shot was made to attempt to bring all of it down at once. By then, there were at least 12 passenger trains delayed by the cleanup efforts. After bringing in additional steam shovels, the OWRN eventually cleared the tracks at 11:55 p.m. Wednesday, May 12th. The OWRN and Standifer-Clarkson engaged in a verbal battle for the next month over responsibility for the blast, and in the meantime, other explosions caused more rock fall on the company’s main line and delayed more trains.
Standifer-Clarkson’s crews evened up the cut, constructing cement rubble masonry walls, and filling, where needed to even up the road bed. In the meantime, work was progressing on the tunnel, and crews were preparing the talus slope for the viaduct that connected the excavated cliff side with the tunnel. The 192′-0″ reinforced-concrete viaduct was a slab and girder type, supported on sets of columns 15′-6″ apart center-to-center, and 32′-0″ longitudinally. The roadway was 20′-0″ curb-to-curb. Crews found it difficult to locate footings because the talus slope was unstable. Excavations were made by hand and proved very time consuming. Cribs were placed and men worked as deep as 65′, dumping excavated material into buckets hoisted by hand to the surface by a windlass. Even so, several columns were founded on shellrock, requiring that footings be enlarged from 4′-0″ to 5′-0″ square. Further complicating matters was the OWRN’s easement providing for a 25′ width to the toe of the slope at this point. This required sinking footings below the slope of repose of material extending upward from the 25′ line.
Because the road line was so close to the OWRN right-of-way at the foot of the slope, there was no level area there for a concreting plant, nor was there enough room at grade level. Standifer-Clarkson solved the problem by building a platform along the toe of the slope, next to the railroad right-of-way, for its cement mixing plant. There, it mixed concrete using aggregate from the tunnel excavations and water pumped from the Columbia. Crews hoisted the concrete in % yard buckets about 100′ to the road level, where they dumped it into cars that ran along a track to the viaduct site. They also used this method of transport for forms, falsework, and reinforcing steel. It worked smoothly, but the cramped quarters near the railroad main line meant that operations ceased whenever trains passed by.
Standifer-Clarkson’s crews were experts at concrete work. On the Mitchell Point Viaduct, they poured the footings, then built up the forms and poured the columns and struts. They took special precautions to ensure that the long column forms were filled without leaving any voids. Similarly, because an expansion joint was included in the design, 128′ from the east end, the deck was divided into two pours. The eastern part was the longer and took 3 8 hours to pour. There was minimal settling of the structure considering the different types of foundation materials that crews encountered with the footing. They used wedges and shims between the falsework and forms to maintain true lines in the deck.
As an adjunct to the viaduct construction, Standifer- Clarkson and John Elliott field-designed a shoart half bridge to span a portion of the cliff that fell away during the big blast in May 1915. It was not feasible to pour this structure until after they had built the viaduct and completed some nearby grading. The bridge was not in the original plans for this section of the HCRH, and Elliott pulled it together, “using odd steel sizes and lengths on hand.”
Meanwhile, Standifer-Clarkson progressed on tunnel excavations. The bore ran nearly 4 00′ west-east through a portion of the point projecting east of the shell rock slide. As completed it was 18′ wide and 19′ high at the midpoint of the roadway. The alignment ran a tangent from the west portal to about midway, where it took a 10° curve with a 11%® central angle. Five windows with the same cross-section as the tunnel bore were cut along the cliff face, arranged in a one-three-one formation, with a 50′ space between the first and second windows, a 10′ space between the second and third, and third and fourth windows, and a 46′ space between the fourth and fifth windows. The curvature in alignment was designed for practicality and to increase the light effect in the tunnel. The curve was so slight it was still possible to see through the tunnel from end to end. At the fifth window a trail was constructed leading out 25′ along the cliff face for pedestrians to view both the exterior of the tunnel and the river gorge.
The idea for a tunnel with windows, or the “Tunnel of Many Vistas,” is attributed to Samuel C. Lancaster, who on a trip with Samuel Hill to the First International Road Congress in Paris in 1908 visited roads throughout western and southern Europe, Of these, the parkways along the Rhine, in Germany, and the Auxenstrasse, on the banks of Lake Lucerne, in Switzerland, most interested him. In particular, he was taken with the Auxenstrasse’s long tunnel with its series of windows mostly hewn out of the natural rock to bring light into the tunnel’s interior. Shortly after Lancaster returned from Europe, during the 1908-09 academic year, Elliot was one of his highway engineering students and saw his photographs of the Auxenstrasse and likely heard him speak of this tunnel.
In surveying the Historic Columbia River Highway in the Mitchell Point section, Elliott picked the Lower Mitchell Point for a similar design. He hoped to improve upon the Auxenstrasse tunnel, which had pillars between windows built up from masonry, by creating one on the HCRH that had no artificial construction. The natural columns, though, could not be too thick, for Elliot feared the windows might take on the appearance of side tunnels, “If the openings were to convey the idea of windows,” he wrote, “they must be seen as such.” He also chose a curved alignment rather than a straight bore, because he believed the “the light effect would be lost.” The adits would admit a continuous glow during daylight hours, one for which the motorists would not know the source. It was the easiest and most economical alternative.
In his reconnaissance of the tunnel site, Elliott noted indentations in the cliff wall that he believed were “cheap window locations,”and with some testing, he pinpointed the five window openings where he thought they best illuminated the bore. In addition, the bore’s curvature was such that drivers approaching the tunnel from either end had a head-on view of the central three windows and the rock columns that separated them. To insure that the firm awarded the excavation contract used care in boring the tunnel and in cutting the adits that Elliot had located, the highway department contract provided a premium for “close work.” It allowed a variation of 5 percent from the section that Elliott specified without any price adjustment, while overbreak in excess of 10 percent was not tolerated. So while the tunnel was designed with economics as the first concern, aesthetics and an incentive for accuracy in cutting followed closely behind.
Standifer-Clarkson’s crews began the bore by excavating the west portal. Men dangled from ropes attached to the cliff above to support themselves while they cut a working bench, squaring up the cliff and creating a ledge. The material was Columbia River Basalt, with frequent cleavage places, and commonly known as “dice” rock, because it broke up unto small fragments when it was blasted. With the tunnel section 18′ wide and with 10′ vertical walls, the crown was 19′ given an arch radius of 9′. The heading, or top portion of the bore was taken out first, followed by the bench. Blasting crews used extreme care when working near the outside wall and especially in the vicinity of the windows so that they did not cut through the wall or create oversize adits. It was a tedious process that involved the sure skill of an experienced explosives expert. The arrangement of blast holes, and the sequence of explosions, using 40 percent dynamite and black powder was the key to precise boring. According to Elliott, it worked this way:
At the beginning the entire heading was pulled at ones, but this was abandoned, as the working face neared the windows and the thin outside wall. Here the lower outside lift hole was dropped one round back, leaving each time about 3* of the heading next to the thin wall, thus increasing the effective thickness of the outer wall. This lift hole was always one round behind until all danger of breaking out had been passed. . , . Bad fuse gave some trouble, but danger from this source was avoided by inserting two fuses in the center and cut holes. . . . The firing order was controlled by varying the length of the fuses. Shorter fuses were used in the center, and cut holes and longer fuses in the lift holes. In this way, the center section was broken before the heavier charges in the lift holes exploded, which made the work of the lift holes easier. By holding the outside lift hole back one round, the necessity of breaking to a wall on both sides was eliminated, and the explosives broke out along diagonal lines converging towards the powder charges, which also made the heading easier to break.
Standifer-Clarkson limited hole depth in the heading to 4′. The crews used 1% sticks of 40 percent dynamite to spring each hole, followed by a per hole charge of 16 to 20 sticks of the dynamite. Each explosive round moved the bore along about 4 1/2′.
After detonating the charges, crews loaded the heading material into horse-drawn muck wagons on tracks and dumped the material out the portal. They also disposed of only a limited amount through the adits because of the OWRN main line’s close proximity below. Crews trimmed the roof with picks and hammers after each shot, and then pilled or stripped off the bench. They used two sticks of 40 percent dynamite to spring each hole and followed this with loads of 18 to 22 sticks. Crews loaded the bench material in the same manner as the heading, dumping much of it out the west portal. Once crews broke through from the east portal, they dumped all bench material into a large fill outside of the east portal.
While Standifer-Clarkson crews were cutting the tunnel from west, others were excavating the east portal of 6,10G cubic yards earth, shell rock, and loose rock to reach ledge rock that it bored through in much the same manner as crew did from the west portal. Their procedure was to first cut a trap tunnel along the centerline of the projected bore until they reached the ledge rock. Then, as Elliott explained:
The trap tunnel was 6′ x 6′ in cross-section, and was timbered through the mixed material, the roof timbers being placed without fastening. A track was built into the tunnel, on which a one-yard car was run. When the trap tunnel was ready for use, a few of the roof timbers were pulled and others placed longitudinal, forming a small hole through which the material fell into the car below. One man with a wide board guarded the hole and stopped the flow of material when the car was loaded. The car ran out onto the dump by gravity, and a horse was used to pull it back. Two men worked up on the slope loosening the material around the hole in the shape of a half funnel. Even the horse learned to perform his act without direction. After pulling one car back with sufficient momentum to carry it into the trap tunnel, he would walk back to the dump, turn around and wait to be hooked onto the second car to come back.
Five men and the horse handled as much as 200 cubic yards a day. Once the men had mucked out the debris, the tunnel crew began driving the heading from the east in the same manner was as the other crew had from the west.
Elliott hoped to take advantage of the natural rock in creating sills for the adits, but after completing the western most window he found that the rail was rough and uneven. As an alternative, he carried on the rubble masonry guard wall theme that was used throughout the Historic Columbia River Highway along the exposed roadway. Stone masons built the rubble walls with semi-circular arched drainage cutouts and a screeded concrete cap in the window openings. According to Elliott, they were “of proportions to give a feeling of security to a driver.
The Mitchell Point Tunnel and Viaduct section of the HCRH opened for traffic in early September 1915. Standifer-Clarkson completed the project on the .84 miles of road on November 25, 1915. The original estimate for this work was $60,000 and the Oregon state Highway Commission appropriated $50,000. To better match the state allocation the alignment was revised and costs projected at $53,104.00. Yet, Standifer-Clarkson’s bid was the lowest at $40,343.50. Adding to this $2,500 for engineering and $3,600 for railroad inspection, total costs were about $47,000, This was $3,000 less than the state appropriation.
Mitchell Point Tunnel became known as the “Tunnel of Many Vistas.” Samuel Lancaster believed that it was “among the most wonderful pieces of highway construction in the civilized world.” Lancaster continued, “It is fully equal to the famous ‘Auxenstrasse’ of Switzerland and one the great features of the Highway.” Indeed, while the Auxenstrasse’s tunnel had three windows, the Mitchell Point Tunnel had five.
Repairs and Maintenance
The Mitchell Point Tunnel, like the Oneonta Tunnel and the Mosier Twin Tunnels, all received widespread praise for their construction, and more significantly, for their artistic contribution to the Historic Columbia River Highway. Nevertheless, all three were financial drains on the Oregon State Highway Department because they required almost continuous maintenance. Surrounding unstable basalt formations continually covered tunnel approaches with large rocks, endangering the lives of motorists on what became a heavily traveled road Traffic density almost from the day Mitchell Point Tunnel opened made it, and the other bores on the Historic Columbia River Highway, dangerous portions of the road.
The Historic Columbia River Highway became increasingly popular with the more widespread use of automobiles in the 1920s, By the late 1920s and the 1930s, the development of larger, more powerful cars with improved suspensions and brakes brought even greater numbers of fast-paced drivers to the HCRH. Finally, the advent of motor trucking by the 193 0s, and the adoption of the Historic Columbia River Highway as the major trunk route connecting Portland with points east, brought an additional group of drivers to the road. From almost the beginning, the Mitchell Point Tunnel’s adits became a traffic hazard as motorists stopped their automobiles along the tunnel to peer out the windows or walk out on the cliff.
Combining the inherent traffic dangers with the real possibilities of pedestrians being struck by falling rock while out on the balcony, the highway department saw the need to barricade the windows. The Union Pacific Railroad became increasingly concerned in the late 193 0s about the unstable nature of the rock formations in the tunnel’s vicinity. It was especially alarmed at the disintegration of tunnel’s window pilasters and feared a complete collapse of the tunnel with rock raining down on its OWRN main line below. The highway department strengthened one of the pilasters with concrete, and it closed one of the windows.
At the same time the department received numerous complaints about the narrow roadway of this tunnel and the others on the HCRH. The highway department considered widening the Mitchell Point Tunnel, but its engineers believed that the already weakened window pilasters were too unstable for any additional excavations. Similar situations were seen at Oneonta Tunnel and at the Mosier Twin Tunnels. At all three, the alternative solution was to install traffic-actuated one-way signals. These signals consisted of customary red-green lights mounted at the portals of the tunnel and tripped by vehicles passing over actuation pads set in the pavement near the portals.
By the late 1940s and early 1950s, the highway department constructed a new water-level route for U.S. 30, bypassing the Historic Columbia River Highway from Troutdale to The Dalles. Much of it was constructed on fill taken from dredgings of the Columbia River. The new route was not a tangent route, but consisted of long, sweeping curves that better accommodated faster traffic, yet still afforded splendid views of the Columbia River Gorge. The section from Troutdale to Dodson completely bypassed the HCRH, leaving this portion of the old road intact for pleasure drivers seeking close-up views of the many waterfalls. Likewise, the portion from Hosier to The Dalles was left intact, but the department sacrificed much of the old road between these two sections for the new alignment.
Even though Mitchell Point Tunnel remained intact after the new route was complete, it was selected for abandonment as part of a discontinuous portion of the Historic Columbia River Highway. The Union Pacific worried about rock fall damage to its main line and finally relocated its tracks out into the river on fill. Likewise, the highway department designed the new waterlevel route through this section to follow the railroad alignment at the base of Mitchell Point. Before the state could complete the segment of the new U.S. 3 0 near the tunnel, a portion of the cliff supporting the roadbed near the tunnel’s west end collapsed, spelling an early demise for the Mitchell Point Tunnel.
The Oneonta Tunnel had been mothballed in the late 1940s when the railroad and the highway departments both moved their alignments around Oneonta Point. Subsequent to the opening of the new highway, state crews mothballed the Mitchell Point Tunnel and the Mosier Twin Tunnels also by backfilling with rubble. This backfilling was done by dump trucks and caterpillar tractors, with crushed basalt. State Highway Engineer Robert H. Baldock believed that because of safety concerns all of the tunnels on the old highway were a liability to the state and that the best plan was to close them forever.
In December 1954, eighteen months after Mitchell Point Tunnel’s closure, highway department officials reported that continual rock fall had littered the closed roadway approaches making it nearly impossible to walk on bare pavement in the area. For the next decade the tunnel sat mothballed and unused, to be reclaimed by the landscape. Yet in the mid-1960s, the state highway department constructed an additional two traffic lanes along the water-level route of U.S. 30 to make it a four-lane divided highway (it was renamed Interstate SON, then again renamed Interstate 84). In the course of widening the section at the base of Mitchell Point, the department needed to cut back the cliff and scale it of loose debris. This meant, in part, the destruction of the Mitchell Point Tunnel and Viaduct. By 1966, it was gone; only a short section of retaining wall to the west of the viaduct remains.