Site studies show that log length is the most important factor influencing accumulation width. Site studies and the descriptive statistics of larger sets of drift-laden bridges also show that drift-accumulation width on single piers and the width of blocked spans are related to upstream channel width (figures 18, 19, and 21). In this report, a design log length for use in estimating the potential for drift accumulation is inferred from the width of the largest single-pier accumulations and the longest blocked spans.

Because single-pier drift accumulations are based on logs extending the full width of the accumulation, and spans are blocked by logs extending from pier to pier, the maximum width of these types of accumulation is about equal to the maximum length of sturdy logs delivered to bridges. This design log length does not represent the absolute maximum length of drift pieces; longer pieces were observed at several sites. It represents a length above which logs are insufficiently abundant, or insufficiently strong throughout their full length, to produce drift accumulations equal to their length.

Design log length is defined at a given site by the smallest of three values:

• the width of the channel upstream from the site,

• the maximum length of sturdy logs, and

• in much of the United States, 9 m (30 ft) plus one quarter of the width of the channel upstream from the site.

The minimum width of relatively narrow channel reaches immediately upstream from the bridge can be used as an estimate of the length of the longest logs arriving at the bridge (Lagasse and others, 1991). Channels less than 12 m (40 ft) wide with forested banks receive an abundant influx of logs longer than one channel width when generation processes are active. Drift transported over a long distance is limited to pieces too short to jam between banks or between trees on opposite banks. Longer drift either accumulates or is broken until it can fit crosswise in the channel. As a result, few logs arriving at a bridge are able to span the distance between piers farther apart than the width of the channel upstream from the bridge.

The height and diameter of mature trees on the banks determine the maximum length of the logs that are delivered to the bridge as drift and are capable of withstanding hydraulic forces when forced against piers. This maximum sturdy-log length seems to reach about 24 m (about 80 ft) in much of the eastern United States, and may be as long as about 45 m (about 150 ft) in parts of northern California and the Pacific Northwest.

Typical mature heights of tree species common on river banks are not identical to maximum sturdy-log length, but give a rough guide to the maximum length of transported logs that can be expected in wide rivers. A comparison of mature tree heights among forest regions can be used to illustrate possible regional differences in the design log length and corresponding differences in maximum widths of accumulation and span blockage (table 2). The maximum sturdy-log length can be modified to fit regional and local conditions as more data on actual dimensions of transported logs are gathered.

Table 2. Height and diameter of mature large trees, by region and species.

Throughout much of the United States, the maximum sturdy-log length is 24 m (80 ft). In the Southern Forest Region and Central Forest Region, several tree species common on moist sites reach mature heights between 24 and 30 m (80 and 100 ft) (Preston, 1976). These species provide the large logs in accumulations observed in Indiana, Georgia, Tennessee, and Texas, and those that have occurred in most of the Eastern United States. On the Great Plains, some typical eastern species of large hardwoods extend westward along river valleys to the limits of their ranges. Probably the most extensively distributed large tree along rivers of the Great Plains is the cottonwood. The cottonwood can produce large logs, as shown by the blockage of multiple 24-m (80-ft) spans by cottonwood logs in Idaho.

In those parts of the United States where the maximum sturdy-log length is 24 m (80 ft), design log length is less than either the upstream channel width or the maximum sturdy-log length over an intermediate range of channel width from 12 m to 60 m (40 to 200 ft). Based on the width of drift accumulations and blocked spans, the design log length over this range of channel width is 9 m (30 ft) plus one quarter of the channel width (figure 24). This third constraint on design log length reflects the rarity of long logs and their breakage during transport.

Local knowledge should be used to select a regional maximum sturdy-log length in Washington, Oregon, coastal Alaska, and northwestern California. Many trees of the Pacific Coast Forest Region are more than 30 m (100 ft) high at maturity. Sturdy logs from 30 m (100 ft) to 45 m (150 ft) long were observed in Washington, and sturdy logs more than 24 m (80 ft) long are common. Single-pier accumulations as wide as about 50 m (about 170 ft), and blocked spans as wide as about 45 m (about 150 ft) were observed in this study (figure 25). Other comparable accumulations have been reported (I. Nagai, California Department of Transportation, written commun., 1992; Phillip D. Martin, Quinault Tribe, oral commun., 1995).

Where many logs exceed 24 m (80 ft), design log length is equal to the lesser of either upstream channel width or the regional maximum sturdy-log length (figure 24). In the Pacific Northwest, there seems to be no intermediate range of channel width over which design log length is less than both channel width and the maximum sturdy-log length.

The Alaskan interior may have a maximum sturdy-log length comparable to that in the Eastern United States. McFadden and Stallion (1976) measured the length of spruce, birch, cottonwood, larch, and aspen logs in the Chena River in central Alaska. The average log was 12 m (39 ft) long, two-thirds of the logs were between 7 and 16 m (24 and 54 ft) long, and the longest log was 26 m (85 ft) long.

Across the northern tier of the eastern United States, only a few species reach mature heights of more than 24 m (80 ft). The largest common species in this size class are white oak, yellow poplar, cottonwood, white pine, and American elm. The first three of these species are rare or absent in a region stretching along the Canadian border across northern Minnesota, Wisconsin, Michigan, and New York, and across most of Vermont, New Hampshire, and Maine. Large white pines are rare, and American elm is increasingly rare due to the spread of Dutch elm disease. Thus, along the Canadian border, trees reaching typical mature heights greater than 24 m (80 ft) are relatively rare. The maximum sturdy-log length and corresponding drift-accumulation width in this region may be less than in areas farther south. Data presently available on the dimensions of transported logs and the maximum width of drift accumulations are insufficient to define this shorter maximum sturdy-log length precisely; based on two examples of blocked spans in Manitoba, it is at least 18 m (60 ft) (James Lukashenko, Penner and Keeler Partners, written commun., 1994).

If drift removal and streambank clearing cease, the maximum sturdy-log length will likely increase. Before extensive snag removal (mostly from 1870 to 1920), logs in North American rivers were much larger (Sedell and Frogatt, 1984; Triska, 1984; Sedell and others, 1988). The largest logs in the Red River (Louisiana) jam were 30 to 36 m (100 to 120 ft) long and as much as 1.75 m (6 ft) in diameter. Snags, primarily sycamore and cottonwood, were historically abundant in the lower Mississippi River. On the average, these logs were 1.7 m (6 ft) in diameter at the base, 0.7 m (2 ft) in diameter at the top, and 35 m (115 ft) long. On the Williamette River of Oregon, the very abundant snags were, on the average, 0.5 to 2 m (1.6 to 7 ft) in diameter and 30 to 60 m (100 to 200 ft) long. These historical sizes indicate the potential maximum sturdy-log length that could eventually result from reduced bank clearing and snag removal.