Home Archived January 13, 2017

Potential Drift Accumulation at Bridges

Previous Studies

Published accounts of drift accumulation at bridges represent only a small fraction of the cases that have occurred. Most reports do not include details such as the composition of the accumulation, its size and shape, or its position with respect to the channel and the bridge piers. Unpublished written accounts and photographs are typically not catalogued or segregated from other material in bridge files, and important details may be absent from these accounts. The memories of maintenance engineers are the largest repository of information on drift accumulation. As Brice and others (1978a) found in their study of scour problems:

"No systematic record of bridge losses and hydraulic problems, kept separately from the files on individual bridges, was found in any state or agency; therefore, the estimates given depend on the memory and experience of the persons interviewed. Problems or losses that occurred within the last year are more likely to be recalled than those that occurred ten years ago."

Several published studies describe instances in which drift contributed to bridge failure and damage caused by scour. Brice and others (1978a, 1978b), in a study of countermeasures for hydraulic problems at 283 bridges, identified drift as contributing to scour problems at 22 bridges, and as a major cause of scour problems and bridge failures in California, Louisiana, Massachusetts, Nevada, Oklahoma, Oregon, Pennsylvania, Texas, and Wisconsin. Chang and Shen (1979) published data on drift problems at 61 bridges in the U.S. and Canada. These two studies include some of the same sites; the total number of drift-damage sites covered by the two studies is 67. Harik and others (1990) reported on 114 bridge failures in the United States, blaming one failure on drift without giving details. Smith (1976) reported on the causes of 143 bridge failures worldwide. One of these failures was caused by drift. Dongol (1989) gathered responses from local authorities in New Zealand that identified 12 bridges where drift caused failure.

Many other instances of drift accumulation have been reported in engineering literature, but few reports contain much detail about drift itself. Pangallo and others (1992) report the case of the U.S. Highway 40 bridge over the Wabash River, where a 23.6-m (77.5-ft) span was blocked, and provide photographs of six other sites of drift accumulation. Klingeman (1971) studied scour resulting from drift accumulation at the Deerhorn bridge in western Oregon. The National Transportation Safety Board (1990) reported on the failure of a temporary bridge in Ohio due to drift accumulation. Wright and Harrison (1990) reported on the details of one bridge failure due to scour and lateral loads caused by a drift raft in New Zealand. The Engineering News-Record reported a bridge failure in Tennessee caused by drift-related scour, but detailed information on the drift accumulation was unavailable (Engineering News-Record, 1980; Harik and others, 1990; James Schall, written commun., 1993). Foster (1988) reported on a modeling study of the failure of a temporary construction trestle on the Mississippi River due to scour under a drift raft. The Corps of Engineers published several photographs of bridges that failed due to drift in Hurricane Agnes (Gannett Fleming Corddry and Carpenter, Engineers, 1974).

Several unpublished accounts of drift accumulation and associated damage were provided by bridge engineers (I. Nagai, California Highway Department, written commun.., 1992; Luis Ybanez, 1992, Texas Highway Department, written commun.; James Schall, written commun., 1993; Martin Fisher, Washington State Department of Transportation, written commun., 1994; James Lukashenko, Penner and Keeler Partners, written commun., 1994; Mark Miles, Alaska Department of Transportation, oral commun., 1995). Like most published reports of drift accumulation, these accounts lack specific information on the size of drift accumulations and logs. Because information on the cross-sectional area transverse to the direction of flow is not available, indirect methods must be used to estimate drift forces and the degree of constriction of the bridge opening. The lack of information on log dimensions makes the estimation of log-impact forces more difficult.

Bridge inspection programs provide qualitative information on drift (Strautman and others, 1987; Avent and Whitmer, 1990; Huizinga and Waite, 1994; Bryan and others, 1995). Information from such programs does not include the shape, location, and porosity of drift accumulations, or the size of the component pieces. Lambeston and others (1981) recognize that drift-related forces and scour threaten bridges, and recommend inspection of the "nature and location of debris" during underwater inspections of bridges, but do not specify which drift properties should be recorded.

Potential Accumulation Size and Scour

The potential scour depth (and potential lateral forces on bridges) associated with drift depend on the maximum size that drift accumulations can reach. Methods for estimating a maximum drift-accumulation size for use in bridge design have been recommended for Australia and New Zealand, but not for the United States(American Association of State Highway and Transportation Officials, 1989).

Australian design practice assumes that the potential width of drift at a pier is equal to the average of the adjacent span lengths, up to a maximum of 20 m (66 ft), and the minimum assumed vertical depth is 1.2 m (4 ft) (National Association of Australian State Road Authorities, 1976; Wellwood and Fenwick, 1990). The potential width of drift on a submerged bridge superstructure is assumed to be the length of the superstructure. In developed river basins, the assumed minimum potential vertical depth of a drift accumulation is 1.2 m (4 ft) greater than the vertical extent of the submerged superstructure (typically, from low steel to the top of the parapet). The assumed maximum potential vertical depth is 3 m (10 ft), unless local information indicates that it should be greater (figure 6).

New Zealand's design practice is similar to Australian design practice (Dr. Thomas Fenske, University of Louisville, oral commun., 1992). A draft design specification states that the potential drift accumulation at a pier can be assumed to be triangular in cross section perpendicular to the approaching flow (Dr. Arthur Parola, University of Louisville, written commun., 1992 ). The triangle's greatest width (at the water surface) is half the sum of the adjacent span lengths up to a maximum of 15 m (49 ft). The triangle extends vertically downward along the pier nose to a depth equal to half the total water depth or 3 m (10 ft), whichever is less (figure 6). The maximum width and depth of drift accumulations observed in this study exceed the values used in design in Australia and New Zealand.

7K GIF version of this diagram

Figure 6. Vertical cross sections of assumed maximum drift accumulations on single piers.

The width of drift used in suggested Australian and New Zealand design practice may be related to the estimated length of key structural logs, and the suggested depth of drift may be based on observations of actual drift accumulations. However, no explicit basis is given for the design drift accumulations. The authors of these recommendations do not cite published or unpublished studies of actual drift accumulations, flume studies, or theoretical studies. A literature review of hydrodynamic forces on bridges, conducted at the University of Queensland in 1984, failed to uncover published accounts of the dimensions or characteristics of drift accumulations (Apelt, 1986a). As Apelt (1986b) points out:

"The estimation of the flood loads applied to bridges by debris mats is bedevilled by lack of reliable information about the size, shape, and compactness of accumulations of debris against bridges during floods."

Scour associated with drift accumulations can be estimated based on the size of the accumulation. Pier scour depths around drift accumulations are estimated by substituting an effective pier width into a conventional pier-scour equation ( Melville and Sutherland, 1988; Dongol, 1989; Melville and Dongol, 1992; Richardson and Davis, 1995). Contraction scour is calculated by reducing the area of the opening by the cross-sectional area of the drift accumulation perpendicular to the flow (Richardson and Davis, 1995). Scour underneath a drift raft can be calculated by assuming it is analogous to pressure flow under a submerged superstructure (Richardson and Davis, 1995).

Drift Countermeasures

Solid, round-nosed piers aligned to flood flow are recommended where drift is abundant (Neill, 1973; Brice and others, 1978a; Chang and Shen, 1979; Lagasse and others, 1991; Richardson and others, 1991; Richardson and Davis, 1995). Pangallo and others (1992) suggested not placing hammerhead piers in the water or pile caps above the streambed. To reduce drift accumulation, some piers have been designed with inclined upstream noses, or have been fitted with inclined drift deflectors (Brice and others, 1978a; Brice and others, 1978b; Martin, 1989).

Spaces between piles can clog with drift, increasing flow contraction and local scour depth (Brice and others, 1978a; Lagasse and others, 1991; Richardson and others, 1991; Richardson and Davis, 1995). Where pile bents or multiple-column piers are used, placement of a web wall between the columns is recommended (Brice and others, 1978a; Lagasse and others, 1991).

Long spans are less prone to drift blockage, but none of the publications that were reviewed endorsed a specific span length as adequate for North America (Bowser and Tsai, 1973; Neill, 1973; Brice and others, 1978b; Chang and Shen, 1979; Pangallo and others, 1992). In Australia and New Zealand, however, current design practices imply that spans will not be completely blocked if longer than 20 m (66 ft) or 15 m (49 ft), respectively.

Various other countermeasures have been suggested. Bridges that have adequate freeboard during the design flood are less prone to drift accumulation (Neill, 1973; Brice and others, 1978a; Chang and Shen, 1979). Several authors have recommended structures to deflect drift from piers and guide it through openings, or booms and trash racks to collect it (Kennedy, 1962; Kennedy and Lazier, 1965; McFadden and Stallion, 1976; Brice and others, 1978a; Chang and Shen, 1979; Perham, 1988; Lagasse and others, 1991; Richardson and others, 1991; Saunders and Oppenheimer, 1993). Pangallo and others (1992) suggested that piers should not be placed in the outside of bends because of the likelihood that drift will be concentrated there.

previous || table of contents || next

U.S. Department of the Interior, U.S. Geological Survey
Last update: Tuesday, 08-Jan-2013 14:46:22 EST
URL: http://tn.water.usgs.gov /publications/FHWA-RD-97-028/prevstud.htm
USGS Privacy Policy || Disclaimer