Society of American Foresters

Northwest Regions

Evolution of Forest Road Design and Construction

By Julie Dieu, as told to Andrea Watts

Julie Dieu headshot

In the forestry profession, engineering makes it possible for a safe and steady flow of wood from forests to sawmills or ports. How engineers accomplish this has changed over the years as the methods of moving wood from the forests have changed and as available technology both for engineering and for construction has evolved. All of this has been influenced by regulations directing us to reduce our environmental impact on the landscape and comply with regulations.

The railroad era

From the 1920s to the 1940s, timber harvesting in western Washington was quite sophisticated in terms of pure engineering. As many of the pictures from this era depict, logs were moved by trains, and engineers were tasked with determining where the rail lines would be built.

Early in my career, I met a couple of retired engineers who worked at Dickey Camp, north of Forks, in the late 1940s. They made their own topographic sheets by walking out across long distances of ground and running survey lines to create contours at a fairly fine scale. (My employer, Rayonier, still has and uses a number of these topo sheets, which were made on oilcloth.) It was a massive effort to create these topo sheets, but that is what was needed to design rail line. Standard gauge rail line, with very rare exceptions, can’t exceed a 3 percent grade, which limits where it can be built. The details of design are also complicated because the rail line must support the weight of the trains across all soils including swampy ground.

As rail line was both expensive to engineer and expensive to build, swing landings were used to limit the length of rail line necessary to harvest an area. First, a steam donkey (a small, steam-powered engine) was rigged to a strong tree that became the “spar” pole. All the wood reachable within the pole’s circle was hauled to the pole. The size of the circle depended upon how tall a tree could be found. Out in the coastal zone, which is severely wind-impacted, it was very rare to find old growth trees that were taller than 150 feet. This kept the spar poles necessarily short, and the circles of yarding smaller, about 25 acres as viewed on aerial photos.

Since the tracts being logged were expansive, swing yarding connected these spar poles across the landscape. The logs would be swung from one spar pole to the next. Twenty-five acres of timber could go down one track between a series of spar poles until the logs reached the rail line.

The remnants of this type of yarding are still visible on the landscape. Occasionally, you will find a funny little stream valley, except there’s no stream in it. That is where old-growth timber, logged to the first spar pole, got dragged across the ground to the next spar pole and then the next spar pole, finally reaching the rail line.

Truck hauling and landslides

Around the late 1940s, there were a number of changes in how companies such as Rayonier practiced forestry. The first change: We figured out that we could replant trees and manage our forests for additional rotations. The second change: Companies were moving into steeper ground because the gentler hillslopes on the Olympic Peninsula had been harvested. This necessitated the building of truck roads, which could be built at much higher gradients—up to 18 percent grade—with short pitches steeper than that. (This is the same era when logging trucks had double clutches. How log truck drivers managed this is another story!)

The earliest truck roads were designed and built by the engineers who had built the railroad grades. As a result, they had the mindset of creating smooth curves, constant gradients, and needing to balance the cuts and fills.

This is how I can date the age of a road: If it has big cuts in the hillside with accompanying big fills and nice smooth curves (despite the fact that the road is running at 10-12 percent, which no railroad would ever run at), I know the road was designed by railroad engineers. How can someone who isn’t an engineer date a road? If you’re on a logging road that has a constant gradient with wide smooth curves and you can go flaming along at 45 miles an hour on gravel, it is an old railroad grade or at least designed by an old railroad engineer.

Landslide photo
This section of road experienced a landslide as a result of sidecast
failure. The faint black zone represents the original topsoil
and material above is sidecast. Photo courtesy of Julie Dieu

Today we know this mindset had environmental consequences in steeper ground. Particularly in the 1940s, ’50s, and ’60s, roads were cut into steep hillslopes with bulldozers. The excess material, known as “sidecast,” was cast over the side. Big cuts led to collapses, and big fills with small culverts led to storm-driven failures as the pipes plugged. But the biggest source of landslides were the failures from the sidecast along the edge of the road. This is why there was a tremendous landslide signature from these decades. In some watersheds, 75 percent of the historic failures are from roads, and that road prism is maybe two percent of the landscape. As you get into steeper and steeper ground, the footprint gets smaller and smaller because you want to build as little road as possible. This resulted in a very high landslide rate from a very small footprint.

Through the years, many engineers have thought that if the sidecast has been there for 40 years, it’s stable. That’s not the case: the failure potential of that material perched on steep ground still exists. It’s safe to build in this manner when you’re on 20-30 percent of ground, even 40 percent, but when you’re on 60 plus percent hillslope, that’s where you’re incurring landslide risk by having that material perched on the edge.

Reducing landslides through design

A current forest practices guideline for road construction is that you design the road on the topography, so you have sharper curves with smaller cuts and fills. As I have said to my engineers, “The less dirt you move, the less environmental hazard you incur.” We still need detailed topographic information to accomplish this, but today we utilize digital elevation models generated from Lidar, with minimal topographic data collected by field effort. And we use powerful engineering software to nuance the design elements and create diagrams that facilitate both the regulatory permitting process and the actual construction.

Today, when we design a new road on steep ground, we specify places where material cannot be pushed over the side; it has to be put in a truck and hauled out of the area to someplace where it’s safe to dump. When we reconstruct old roads, we pull back the old sidecast, put it in a truck, and haul it to a waste area. The development of hydraulics in the 1970s made these efforts possible— bulldozers just pushed material, but modern excavators can scoop soil and place it into a truck. Evidence indicates that this care with sidecast is bringing down landslide rates, which is what we want to see.

There are two other practices that are lowering road-caused landslides. One, we up-size pipes and place them on higher gradients in stream crossings, even if fish can’t reach that stretch of stream. These large, steep pipes have a better chance to pass debris and water during a big storm without plugging and failing. Two, a lot of cross drain culverts, pipes that pass ditch water under the road, are being added. In earlier decades, few cross drain culverts were installed. Today, in Rayonier’s coastal environment, our standard practice on steep ground is to place cross drains about every 300 feet.

Seeing the engineer’s signature

Although regulations and topography dictate how roads are constructed, there’s a very human individual aspect to road construction and road design. Road engineers tend to have 30-year careers. If you look at a landscape that a timber company has owned for many decades, you can see the imprint of the different thought processes, such as where the engineers decided to put cross drain culverts that relieve the ditch line onto the hillslope.

As for the signature I am leaving behind: There will be a series of roads on very steep ground that will have a limited landslide signature from this last era of harvest. Future Rayonier engineers will see side cast placement limited to safe locations; evidence of the removal of old side cast; the installation of lots of cross drains; and in the presence of larger stream-crossing culverts at high gradients. My job is to make sure this road is going to be here for 40 years when we come back for the next rotation. 

Julie Dieu is a geomorphologist with Rayonier and is based in the office at Hoquiam,Washington. She can be reached at (360) 580-0088 or

Back to Table of Contents                    Next article