Construction of the Stockwood Fill, 1906–1909

Construction of the Stockwood Fill in Clay County between 1906 and 1909 taught Northern Pacific Railway engineers a bitter lesson about building big on northwestern Minnesota’s former-lake-bottom soil.

The Red River Valley straddles the border between North Dakota and northwestern Minnesota, a flatland that was once the floor of Lake Agassiz. The flat surface of the Red River Valley and wave-cut strandlines record the ancient lake’s presence. Below the valley’s surface lie weak lacustrine and fluvial sediments on top of glacial drift. Below this sediment package is a granitic bedrock floor.

In 1872, the Northern Pacific Railway (NP) reached the Red River and crossed into North Dakota. Railroad tracks laid by the NP ran from east to west and created a steep grade between the towns of Glyndon and Hawley. The grade slowed railway traffic and required assistance from extra locomotives that could pull or push loaded trains from the bottom of the valley to the top. In 1906, NP engineers decided to construct an earthen embankment gradually rising eastward from Glyndon.

Construction on this 7.3-mile stretch of track started in April 1906. The project had a budget of $2,133,000 and a deadline of two years. Railway engineers estimated that the trestle would require over 2 million board-feet of lumber and millions of cubic yards of dirt fill.

At the beginning of August, it was discovered that the Stockwood Fill was sinking. Between August 7 and 13, the structure continued to settle. With each new settlement came compression ridges that appeared on each side of the fill. On August 8, a baffled S. A. McCoy (the project’s manager) ordered a geologic investigation of the soils beneath the fill. The investigation revealed unstable glacial sediment, including quicksand, gravel, and clay. The underlying geology of the Red River Valley had compromised the structure.

Construction of the fill continued despite the problems induced by the quicksand below. Soon, the sag in the track became too steep for work trains to navigate. McCoy ordered a second trestle to be built over the sagging region. With new tracks in place, NP crews continued to dump dirt onto the trestle, only to watch it continue to sag.

In December 1907, NP construction superintendent F. L. Birdsall took over the project. In the spring of 1908, he ordered the construction of a 3,000-foot-long permanent bridge over the same problem region, hoping to put an end to the sinking. Following construction of the new bridge, work progressed rapidly. The bridge allowed work trains to access the embankment once more, and the crew continued to bury the trestle.

In May 1908, two more zones of the fill started to show signs of settlement. Eight hundred fifty feet of track sank nearly fifty feet in one week. Hoping that a wider base would support the heavy embankment, Birdsall proposed extending the width of the fill. The strategy worked well throughout the summer of 1908, and the project was on track to be completed by November 1.

On October 10, 1907, a fourth zone of settlement appeared adjacent to a mile-long stretch of embankment opposite the Stockwood Depot and along the 3,000-foot-long permanent bridge. NP officials were devastated to hear that this new filling technique had failed, since the next part of the trestle to be filled included an $11,000 concrete archway. The 125-foot-long, 16-foot-high archway would allow local road access through the fill.

Filling progressed eastward toward the archway. The ground continued to crack along the embankment and the ground stresses eventually reached the concrete arch.

In July 1909, the fill of the trestle above the archway was almost complete when the embankment and arch began to sink. Under stress from the soil above and the compression ridges below, the archway broke at its center. Engineers saw no hope for the arch and abandoned it. Over the years, the arch continued to sink until it was completely swallowed by quicksand.

In September 1909, the NP admitted defeat, and the company began to use the embankment as it stood. The project was two years behind schedule and $700,000 over budget. Additional locomotives were still required to assist eastbound trains out of the valley until the arrival of more powerful engines in the 1930s.

Editor’s note: This article adapts material originally published in Donald P. Schwert and Mark E. Peihl's Toils on Weak Soils: A Photo Essay on the Construction of the Stockwood Fill (1906–1909).