Effects of Debris on Bridge Pier Scour (2010)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

107 4.1 Applicability of Results to Highway Practice Approximately 83% of the 583,000 bridges in the National Bridge Inventory are built over waterways. Many, especially those on more active streams, will experience problems with scour, bank erosion, and channel instability during their use- ful life (Lagasse et al. 2001). The magnitude of these problems is demonstrated by the estimated average annual flood dam- age repair costs of approximately $50 million for bridges on the federal aid system. Highway bridge failures caused by scour and stream insta- bility account for most of the bridge failures in this country. A 1973 study for the FHWA (Chang 1973) indicated that about $75 million were expended annually up to 1973 to repair roads and bridges that were damaged by floods. Extrapolating the cost to the present makes this annual expenditure to roads and bridges on the order of $300 to $500 million. This cost does not include the additional indirect costs to highway users for fuel and operating costs resulting from temporary closure and detours and to the public for costs associated with higher tar- iffs, freight rates, additional labor costs and time. The indirect costs associated with a bridge failure have been estimated to exceed the direct cost of bridge repair by a factor of five (Rhodes and Trent 1993). Rhodes and Trent (1993) document that $1.2 billion was expended for the restoration of flood-damaged highway facilities during the 1980s. Although it is difficult to be precise regarding the actual cost to repair damage to the nation’s highway system from problems related to pier scour as a result of debris accumula- tion, the number is obviously very large. In addition, the costs cited in the preceding paragraphs do not include the extra costs that result from over design of bridge foundations that result from the inability to predict where and how debris will accumulate on bridge piers and calculate the resulting increase in pier scour. This lack of knowledge often results in overly conservative design. The guidelines and methods that resulted from this research provide guidance to bridge owners for predicting and calcu- lating the effects of debris on scour at piers. The end result will be a more efficient use of highway resources and a reduc- tion in costs associated with the impacts of debris on highway facilities. 4.2 Conclusions and Recommendations 4.2.1 Overview This research accomplished its basic objectives of develop- ing guidelines for predicting the size and geometry of debris accumulations at bridge piers and methods for quantifying scour at bridge piers resulting from debris accumulations. The project produced results on two related problems: (1) predict- ing the accumulation characteristics of debris from potentially widely varying source areas, in rivers with different geomor- phic characteristics, and on bridges with a variety of substruc- ture geometries and (2) developing improved methods for quantifying the depth and extent of scour at bridge piers con- sidering both the accumulation variables and the range of hydraulic factors involved. Waterborne debris (or drift), composed primarily of tree trunks and limbs, often accumulates on bridges during flood events. Debris accumulations can obstruct, constrict, or redi- rect flow through bridge openings resulting in flooding, dam- aging loads, or excessive scour at bridge foundations. The size and shape of debris accumulations vary widely, ranging from a small cluster of debris on a bridge pier to a near complete blockage of a bridge waterway opening. Debris accumulation geometry is dependent on the characteristics and supply of debris transported to bridges, on flow conditions, and on bridge and channel geometry. The effects of debris accumu- lation can vary from minor flow constrictions to severe flow contraction resulting in significant bridge foundation scour. C H A P T E R 4 Conclusions and Suggested Research

At the outset of this study, only limited guidance was avail- able on which to base critical public safety decisions during flooding on debris-prone rivers. There was a pressing need for accurate methods of quantifying the effects of debris on scour at bridge pier foundations for use by DOTs and other agencies in the design, operation, and maintenance of highway bridges. 4.2.2 Advances in the State of Practice Guidelines for Debris Production and Delivery As an extension of the original work by Diehl (1997) for the FHWA, expanded guidelines and detailed flowcharts are now available for estimating (1) the potential for debris production and delivery from the contributing watershed of a selected bridge and (2) the potential for accumulation on individual bridge elements. To facilitate the application of the guidelines, a case study of a debris-prone bridge on the South Platte River in Colorado is summarized in Chapter 3 and presented in detail in Appendix D. The case study also introduces and illus- trates the use of Field Data Sheets for evaluating the potential for debris production and delivery from a given watershed. These sheets were developed specifically for the debris loading problem and provide a rapid and efficient approach to identi- fying the hydrologic, hydraulic, geomorphic, and vegetative factors relevant to the problem. When used as intended dur- ing a site reconnaissance, the data sheets will: • Supply a methodological basis for field studies of the debris hazard • Present a format for the collection of qualitative informa- tion and quantitative data on the stream system and its riparian corridor • Supply the data and input information necessary to imple- ment the engineering analysis of the debris hazard, and estimate the depth and extent of scour expected at a spe- cific bridge pier under debris loading. Although the sheets appear complex, they were designed to produce a comprehensive record of the morphology of the stream and its surroundings and to be applicable to a wide range of river types and sizes in diverse settings. With this in mind, one should resist the temptation to omit filling out part of the sheets for the purposes of expediency or because of perceived irrele- vance, because the data may be used for other applications in the future. However, the sheets can be customized to a particu- lar region, basin, or river through the removal of extraneous material rather than the omission of entire topics or sections. Photographic Archive As a basis for laboratory testing, the photographic archive introduced in Chapter 2 (see also Appendix A), the field pilot study of debris sites in Kansas (see Appendix C), and the South Platte River case study (see Appendix D) were exam- ined to develop a limited number of debris shapes that repre- sent the maximum number of configurations found in the field. These simplified, yet realistic, shapes that could be con- structed and replicated with a reasonable range of geometric variables provided the physical characteristics of debris clus- ters needed for laboratory testing. Rectangular and triangu- lar shapes with varying planform and profile dimensions were selected to represent prototype debris accumulations. To account for additional variables thought to be relevant to debris clusters in the field, a method to simulate both the porosity and roughness of the clusters was developed. While the photographic archive in Appendix A provided the key to developing an efficient, comprehensive laboratory test program, the value of the archive will extend beyond the needs of this study. The archive provides a well-documented database on debris generation, movement, accumulation, and scour at bridges that can be used to inform and train design and maintenance personnel on debris-related hazards. As illustrated by example in the archive, supplemental data on a specific bridge can be acquired with relative ease using Inter- net and programmatic resources available to all DOTs (e.g., Google Earth™, Terraserver, the National Bridge Inventory). Observations from Laboratory Testing Laboratory-scale physical modeling of scour at debris-laden bridge piers was conducted using a range of pier types and widths, combined with different sizes and shapes of debris attached to the upstream pier face. In most (but not all) of the cases investigated, the presence of debris resulted in greater scour at the pier than the baseline (no-debris) condition. Rectangular, blocky debris masses tended to produce the greatest scour at the pier when the extent (“length” dimen- sion) of the debris upstream of the pier was on the order of one flow depth. This condition produced plunging flow that was directed toward the channel bed in the immediate vicinity of the pier face, resulting in a worst-case scour condition. Total scour at the pier was also significantly increased when the total frontal area of flow blockage (as a percentage of the cross- sectional area of the approach channel) was large. In that case, the debris-induced scour appeared to be similar to that created by pressure flow and contraction effects, for example, pressure flow beneath bridge decks that are submerged during floods. Triangular debris clusters (conical in profile) were also investigated, because the debris photographic archive revealed that a triangle is another very common shape that can be pro- duced in the field as drift accumulates at a pier. In a triangu- lar configuration, the thickness of the debris is greater at the pier face, tapering upward and thinning toward the leading (upstream) point. This shape tended to produce more scour 108

at the pier face compared to the baseline (no-debris) condi- tion. In addition, triangular debris shapes produced more pronounced scour laterally outward from the sides of the pier, apparently because much of the debris-blocked flow tends to be shed around this shape compared to a rectangu- lar, blunt-shaped blockage. Although the effect of lateral scour extent on adjacent piers was not investigated in detail, data collected from the labora- tory tests yield valuable information in this regard. Both rec- tangular and triangular (conical) debris shapes resulted in lateral scour that was directly related to the width of debris blockage. Interestingly, the lateral side slopes of the debris- induced scour holes were relatively mild. Pier width normal to the flow direction is also important in determining the total scour depth at the pier face, even when the pier is loaded with a large amount of debris. Given the same size and shape of debris, a slender pier with debris will experience less total scour than a wider pier with the same amount of debris, for the same hydraulic conditions of the approach flow. Scour Prediction at Bridge Piers with Debris Loading The laboratory testing program was designed and con- ducted to develop information on a variety of factors related to debris accumulations at piers that can potentially affect the depth of scour at the pier. The factors examined in this study included: • Shape: Rectangular or triangular • Size: Width, length, and thickness • Location: Surface (floating), mid-depth, or bed (partially buried) • Roughness: Smooth or roughened • Porosity: Impermeable or 25% porosity • Approach velocity: V/Vc ratios of 0.70 and 1.0 Selected combinations of the above factors were also tested; for example, a particular debris shape might be tested as (1) a smooth, impermeable body; (2) a smooth, porous body; (3) a rough, impermeable body; and (4) a rough, porous body. All tests were conducted in the 8 ft (2.4 m) wide flume at CSU. Most of the tests were conducted using 4 in. (10.2 cm) square piers. Tests using slender piers (pile bent and wall piers) were also conducted, but were necessarily limited in number, considering the resources available for this study. The tests were not designed to represent any particular scale ratio. However, considering typical pier sizes and dimensions of debris accumulations found in the field and the photo- graphic archive, a model:prototype scale ratio of approxi- mately 1:10 to 1:30 can be considered a reasonable range for the tests conducted in the flume. Factors not considered in the test program include the effect of bed material grain size, flow depth, live-bed conditions, and contraction scour. In addition, tests at different scales, including near-prototype scale ratios of approximately 1:2 to 1:4 were originally con- sidered but were ultimately dropped from the program so that other factors could be investigated in more detail. All pier scour prediction equations use pier width as a factor that contributes to the estimated scour depth. Intuitively, the accumulation of debris on a pier causes the pier to appear larger in the flow field, thereby increasing the total area blocked by obstruction. HEC-18 (Richardson and Davis 2001) uses the width of the debris perpendicular to the flow direction to estimate the additional obstruction. The concept of equivalent pier width has been widely accepted as a way to estimate scour at complex piers and assess the extent to which debris affects scour at piers. Using the data collected from the laboratory program, this concept was val- idated as the best approach for predicting the effect of debris on pier scour. Dongol (1989) and Melville and Dongol (1992) provide an equation to calculate the “equivalent width,” be, of a bridge pier that is loaded with debris. The equation uses both the width and thickness of the debris and is based on scour data from a limited number of tests in a laboratory flume. Only floating (surface) debris at cylindrical piers was tested, with the debris wrapped around the pier in all directions. The effects of a debris mass with variable length, width, and thick- ness upstream of a bridge pier and the effect of the vertical location of the debris mass within the water column could not be predicted. Under this NCHRP study, these effects were investigated in detail and can now be considered when esti- mating the impact of debris on bridge pier scour. Building on the algorithm originally proposed by Melville and Dongol and using an equivalent pier width, a*d, an improved predictive equation is now available. Considering the most common shapes of debris clusters found in the archive (rec- tangular in planform and profile, and triangular in planform but conical in profile), length, width, and thickness of the debris accumulation upstream of a bridge pier can now be considered. Different coefficients and exponents based on more extensive laboratory testing are recommended, but the basic form of the effective width equation is retained. The recommended equation is stable, can be adapted to most conditions found at bridge piers in the field, and complements the approach to estimating pier scour currently recommended in FHWA’s HEC-18. For the equivalent pier width equation, optimizing the coefficient Kd1 and exponent Kd2 to the observed laboratory data reveals that the shape and upstream extent of the debris do affect the resulting scour at the pier face. The coefficient Kd1 in the effective pier width equation is thus seen to be a shape factor, while the exponent Kd2 is a factor that describes 109

the intensity of the plunging flow created by the debris block- age. In addition, the methodology as formulated accounts for the occurrence of the upstream trough when a rectangular debris cluster extends upstream a distance, L, greater than the approach flow depth, y, as well as the observation of maxi- mum scour when L equals y for a rectangular accumulation of debris. The relationship derived using an “equivalent width” con- cept uses optimized coefficients and exponents based on lab- oratory data, but it is essentially a best-fit relationship that underestimates the observed scour as often as it overestimates. A relationship better suited to design should tend towards conservatism; that is, underestimation of the observed (i.e., actual) scour should be relatively rare. Consequently, design equations for estimating an equivalent pier width for use with the CSU pier scour equation using an “envelope” concept were derived. These equations for a*d are recommended for design. Roughness and porosity of the debris mass have long been assumed to significantly affect the depth of scour at a bridge pier. This assumption was based largely on anecdotal data. The data from this laboratory research program indicate that roughness and porosity of the debris mass do not significantly affect the observed scour. At most, roughness and porosity can be considered second-order variables that are not signif- icant compared to the size and shape of the debris mass. The data also indicate that the location of the debris in the water column affects the total depth of scour at the pier face. The effect of debris location in the water column on pier scour was investigated using two different rectangular debris shapes and was quantified by the ratio of scour depth caused by the debris to the baseline (no-debris) condition. In gen- eral, rectangular debris cluster placed as a surface raft caused greater scour at the model square piers compared to baseline conditions. In contrast, when this same shape was located at mid-depth in the flow, significantly less scour at the pier face was observed. This was presumably due to the relative distri- bution of flow over the debris compared to the plunging flow occurring beneath it. Similarly, when the debris was placed on the bed, less scour was observed at the pier face compared to baseline conditions. When a debris mass accumulates at a pier, it typically initi- ates and grows from floating drift material. The laboratory tests conducted under this study program indicate that the lateral extent of the scour caused by floating debris rafts is directly proportional to the width of the raft. The impact of the lateral scour extent on adjacent piers or abutments was not directly investigated; however, inferences in this regard can be drawn from the laboratory data collected: • Rectangular debris: The lateral extent of scour created by rectangular floating debris extends outward from the edge of the debris at a slope ranging from about 4H:1V to 6H:1V. • Triangular debris: The lateral extent of scour created by tri- angular floating debris extends outward from the edge of the debris at a slope ranging from 2H:1V to 3H:1V. Both 1-D and 2-D hydraulic models have some capability to incorporate debris at a bridge pier. In a hydraulic model, the actual geometry of the debris cluster should be used, rather than incorporating any formulation of an effective pier width. The effective pier width is only used for scour calculations. It is recommended that hydraulic modeling be used to deter- mine hydraulic variables in the bridge reach and that the effec- tive pier width approach be used directly in the HEC-18 pier scour equation to estimate debris impacts. Inspection, Monitoring, and Maintenance General guidelines and considerations for inspection, mon- itoring, and maintenance of debris-prone bridges are dis- cussed in Chapter 3. The major points include the following: • Field inspectors and bridge maintenance personnel are uniquely positioned to detect and report potential hazards relating to debris buildup on bridge foundation elements. These individuals are aware of those bridges that tend to accumulate debris more frequently than other bridges in their district. Records of biennial bridge inspections, as well as maintenance records associated with debris removal, can reveal trends that will help identify debris-prone bridges. • Obviously, removing debris from a pier before it becomes a large, fully developed mass is desirable, but this approach is often impractical from a management and operations standpoint. • However, when single- or multiple-log hangups are observed, a notation should be made in the inspection report and preventive maintenance should be requested. • A maintenance plan that clearly defines the activities and responsibilities of inspectors and maintenance personnel should be developed for any structure that is susceptible to debris problems. 4.2.3 Deliverables As a result of this research, bridge owners now have docu- mentation, guidelines, and analytical procedures to quantify the effects of debris-induced scour on bridge piers. These include the following: • A fully documented database on debris and case studies, photographs, and data related to debris generation, move- ment, accumulation, and scour at bridges that can be used to inform and train design and maintenance personnel on debris-related hazards. 110

• Necessary guidelines for predicting the size, location, and geometry of debris accumulations at bridge piers. • Methods for quantifying scour at bridge piers resulting from debris accumulations. • Guidance for incorporating debris effects in 1-D and 2-D hydraulic modeling. • Worked example problems and a case study illustrating the application of the guidelines and analytical methods. • Suggestions for implementing the results of this research. The end results of this research are practical, implementable guidelines for bridge owners that enhance their ability to pre- dict debris-related hazards at bridges and design, operate, inspect, and maintain bridges considering those hazards. 4.3 Suggested Research The observations of many researchers including Lyn et al. (2003b) would suggest that the recruitment, transport, and accumulation of debris appears to be a somewhat “random process,” while the results of Manners et al. (2007) show that “the relationship between individual logs and complete debris jams is complex and nonlinear.” Lyn also noted that the deliv- ery of debris to a given site “seems to occur in bursts, rather than continuously, even during a flow event of extended dura- tion” and that a possible explanation for this is that “the debris is not generated in the vicinity of the site, and the bursts result from different travel times from different contributing areas.” They also concluded that “the transport of debris occurs rather intermittently with long periods of comparative inactivity punctuated by short periods of intense activity, generally on the rising limb of the hydrograph.” Both the field work and the debris photographic archive compiled for this study validate these findings. Any attempt to develop more definitive guidelines for predicting the site- specific geometry of a debris cluster at a particular bridge is not likely to yield meaningful results. However, this research had limitations in both time and budget that precluded conduct- ing many of the laboratory tests originally planned. Tests of debris effects at a pile bent or long wall pier substructure with and without skew were not tested under this research. Also, time and budget did not permit fully exploiting the photo- graphic archive. Additional research is highly likely to yield important results on these topics. In addition, an expanded laboratory testing program would permit refining and expanding the applica- bility of the algorithm for predicting the depth and extent of scour at debris-prone bridges. Consequently, the following research is suggested: • Using the examples in Appendix A as a model, expand the photographic archive of debris at bridges into a fully doc- umented, searchable database for all sites currently in the archive. If resources are available, the archive could be expanded to include more sites. • No laboratory tests were conducted for a debris length to flow depth ratio less than 1.0. For deep rivers, this ratio is likely to be less than 1.0, and additional tests are warranted. • Conduct selected laboratory tests at a larger scale and include multiple-column (pile bents) and wall-type piers skewed to the flow direction. The prevalence of these pier types in the survey responses warrants these additional tests. • Conduct laboratory tests at greater flow depths, because flow depth was not varied in this study. In addition, con- duct higher velocity tests in the live-bed regime. • Investigate in the laboratory the additional case represent- ing a “true debris raft.” The raft would extend the full width of the flume and far enough upstream to result in a uniform flow field at the pier. • Further evaluate data from tests already completed under this study in order to develop guidance on potential impacts at adjacent piers or abutments or at a downstream bridge. • Construct a realistic debris mass from natural materials (branches, rootwads, etc.) and test it in the laboratory to val- idate the prediction equations, because the equations were developed from tests that utilized “idealized” debris masses. • Identify a debris-prone bridge on the South Platte River (sand bed channel); instrument it with fixed, telemetered scour-monitoring devices; and record the debris character- istics during and after each scour-producing event. • Although debris mass dimensions and shape cannot be predicted for a specific bridge, measure individual debris masses in the field to evaluate whether specific debris mass dimensions can be correlated to the key log length and diameter, and rootwad size. 111