• Federal Highway Administration >
• Publications >
• Research Publications >
• Hydraulics Publications >
• 05072 >
• 04.Cfm >
• Assessing Stream Channel Stability At Bridges in Physiographic Regions

Assessing Stream Channel Stability At Bridges in Physiographic Regions

4. ASSESSING CHANNEL STABILITY

Based on the studies described in a previous section as well as on the observations made at bridges across numerous physiographic regions, a group of parameters that indicate channel stability can be selected. First, however, it is necessary to redefine stream channel stability in light of bridge engineering issues. For this purpose, a stable channel is defined as follows, based on Knox and modified for use at bridges: (15)

A stable channel in the vicinity of a bridge is one in which the relationship between geomorphic process and form is stationary and the morphology of the system remains relatively constant over the short-term (one to two years), over a short distance upstream and downstream from bridge, and with minimal lateral movement.

Although lateral migration of a stream channel can be considered normal and stable within a geomorphic definition of channel stability, it is detrimental to bridge safety and is, therefore, considered in the stability definition used here. The distance upstream and downstream of the bridge that should be considered in a stability assessment depends on the problem, the channel, and the bridge. However, it is acknowledged that a bridge inspector will not typically walk more than a few hundred feet in either direction. That stated, it should be noted that without walking well upstream and downstream of the bridge, channel instabilities, such as knickpoints, that are migrating toward the bridge area may be overlooked. Remember that the objective here is only to assess stream stability in the short term, as inspections of bridges over water are required every 2 years. Thus, it is not necessary to develop a complex method to examine the history or future of channel adjustments over a long time period. It is necessary, however, for each inspector to review previous stability analyses at the bridge of interest to determine whether any unstable trends are developing.

A stability assessment program for bridge inspections should be: (1) brief so that it can be completed rapidly; (2) simple in that extensive training is not required (although some training will be required); (3) based on sound indicators as discussed in the literature review; and (4) based on the needs of the bridge engineering community.

One way to insure that all aspects of channel stability are included is to start at the watershed or regional level and focus in on vertical and lateral aspects of the channel, following the concepts of Thorne et al. (7) and Montgomery and MacDonald. (20) Thus, at the broader level, watershed and flood plain activities as well as characteristics, flow habit, channel pattern and type, and entrenchment are selected as appropriate indicators. At the channel level, indicators such as bed material consolidation and armoring, bar development, and obstructions are used. Indicators of bank stability include bank material, angle, bank and riparian vegetation, bank (fluvial) cutting, and mass wasting (geotechnical failure). Finally, the position of the bridge relative to the channel can be indicated by meander impact point and alignment. In the previous method, the ratio of the average boundary shear stress to the critical shear stress for sediment movement had been found to be important; however, average shear conditions do not necessarily indicate processes that are occurring. (1) Also, critical shear stress is not a reliable number. In addition, it is difficult to measure and quantify as part of a rapid assessment. Therefore, the shear stress ratio is not used as a stability indicator in this current assessment method. In its place, bed material and percent of sand are used. These results are based on the Wilcock and Kenworthy study of bed material movement as a function of sand fractions. (86)

The 13 indicators identified for this study are listed in table 8. For each indicator, a rating of poor, fair, good, or excellent can be assigned based on descriptors listed in the table. After a rating is assigned for each of the indicators, an overall rank is obtained by summing the 13 ratings. Several assumptions are implicit in this method of obtaining an overall rank. First, all indicators are weighted equally. This assumption was tested by assigning weights to each of the indicators and creating a weighted score for every bridge where observations were made. The results showed that the weighted indicators yielded the same results as the equally weighted indicators. Thus, there was no advantage in using weights. Second, this method implies that each indicator is independent of all others. While it is possible that some correlation exists between several of the indicators, an attempt was made to select indicators that independently describe various aspects of channel stability; thus, correlation effects were judged to be insignificant. Third, the summing of the ratings implies a linear scheme. The impact of this is not precisely known; however, given that weighted ratings provided no change in the overall results, it can be assumed that the linearity will also not affect the results significantly.

Table 9 provides the rating results for each of the 13 stability indicators at all of the bridges where observations were made. Using the same 13 indicators for streams in all physiographic regions adequately described channel conditions at each of these sites. The sums of the ratings also are given in table 9. These overall rankings were then rated as excellent, good, fair, or poor. The division of the overall rankings among a single set of category divisions provided limited sensitivity to streams in some stream channel classifications and physiographic regions. Thus, it was desirable to rank the stability based on stream type. Given that the Montgomery-Buffington classification method is based on processes as well as physical characteristics, this scheme was used to provide additional sensitivity to the method. Since cascade and step-pool streams are both transport streams and are not sensitive to changes in sediment or water discharge, these streams were given a separate category of rankings. Plane-bed, pool-riffle, and dune-ripple streams, along with engineered channels, were given a second category as primarily response-type streams. Finally, braided streams were placed in a third category, as these represent a type of stream that is very sensitive to changes in sediment and water discharge and are primarily located in the western and southwestern regions of the United States. These divisions also agree loosely with the stability assessment method that Rosgen developed. (Rosgen has divisions according to stream type, resulting in 42 divisions. This implies a level of sensitivity for which there is no explanation given. It also provides an unwieldy and cumbersome accounting of rankings and tables.) Tables 10-12 provide the range of rankings for excellent, good, fair, and poor ratings of stability for each of the three divisions of stream channels. The final rankings, in terms of excellent, good, fair, and poor, are given in table 9.

*Range of values in ratings columns provide possible rating values for each factor

H = horizontal, V = vertical, Fs = fraction of sand, S = slope, w/y = width-to-depth ratio

H = horizontal, V = vertical, Fs = fraction of sand, S = slope, w/y = width-to-depth ratio

H = horizontal, V = vertical, Fs = fraction of sand, S = slope, w/y = width-to-depth ratio

H = horizontal, V = vertical, Fs = fraction of sand, S = slope, w/y = width-to-depth ratio

R. = River, Cr. = Creek

R. = River, Cr. = Creek

R. = River, Cr. = Creek

R. = River, Cr. = Creek

HEC-20 suggests that the lateral and vertical stability be examined as well as the overall stability. The indicators in table 8 can be divided into those that indicate vertical stability and those that indicate lateral stability. Results are given in table 13 in which vertical stability is described by indicators 4-6, while lateral stability is described by indicators 8-13. Each of the lateral and vertical stability ratings were normalized by the total number of points possible in each category so that they could be represented as a fraction and more readily compared. Thus, the lateral score was divided by 72 and the vertical score by 36. If the lateral score fraction is greater than the vertical score fraction, then it can be expected that the channel instability is primarily in the lateral direction. As an example, the Route 66 Wash is rated as "poor." However, the lateral score fraction is significantly higher than the vertical score fraction (0.93 versus 0.67), indicating that lateral instability is dominant. If, on the other hand, the vertical score fraction is greater than the lateral, then bed degradation is the dominant source of instability. An example of this type of scores is given by Wolf Run, for which the vertical score fraction is about double the lateral score fraction, indicating primarily vertical instability. If both scores are high, then the channel is unstable due to both lateral and vertical processes. For example, Beaver Creek has lateral and vertical fractions of 0.86 and 0.92, respectively. This indicates that the channel is both degrading and widening. The processes may be ongoing simultaneously or they may be occurring differentially. This is frequently the case-as a knickpoint moves upstream, the channel banks respond by collapsing and widening, then another knickpoint moves through, and the process repeats. If both scores are low, this indicates minimal instability in either direction. For example, Alligator Creek has similar scores in both lateral and vertical categories, indicating healthy adjustments in both directions.

Occasionally, rating each of the 13 factors for a particular bridge will result in one factor which stands out as being much higher (worse) than the others. For example, Little Elk Creek received an excellent as the overall rating. All of the assessment factors received scores between 2 and 5, except for the alignment factor (#13). This factor was given a rating of 8 due to the fact that the right abutment of the bridge was located just downstream of the outside of a gentle meander bend. The meander bend appears to be migrating at a very slow rate; this is based on observations that there is undercutting of tree roots on the right bank, but all trees are oriented vertically. Although the rate of lateral migration appears to be slow, it is worth noting and making additional observations during future inspections.

In collecting the data and observations for this method, the engineer or other inspector should walk some distance upstream and downstream from the bridge, rather than just observe from the bridge itself. The appropriate distance, however, depends on several factors, such as uniformity of stream conditions, magnitude of disturbances along the banks, in the flood plain, or in the watershed, time available, and accessibility. Ideally, the observer should walk at least 10 channel widths upstream and downstream of the bridge. Although it is possible to establish stability conditions in less distance, the more of the stream that is observed, the better understanding the observer will have of causes, processes, and rates of change.

Bridges often divide property and sometimes divide geomorphic features or regions. Thus, conditions upstream and downstream of the bridge may be significantly different. In this case, it may be necessary to conduct separate analyses upstream and downstream. Unless the disturbance downstream of the bridge is traveling upstream, as in knickpoint migration or lateral migration of an adjacent meander, then the conditions downstream will be unlikely to affect the bridge, and more emphasis should be placed on the upstream conditions.

Previous | Contents | Next