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.

41 3.1 Introduction This chapter presents guidelines for assessing the poten- tial for debris production and accumulation at a given bridge site, an overview of laboratory testing of a variety of debris configurations on model bridge piers, and an appraisal of the results. The summary of the current state of practice in Chapter 2 is combined with the testing results to provide an improved methodology for quantifying the depth and extent of scour at bridge piers resulting from the accumulation of organic debris. As an extension of the original work by Diehl (1997) for the FHWA, guidelines and flowcharts are developed for esti- mating the potential for debris production and delivery from the contributing watershed of a selected bridge, and the potential for accumulation on individual bridge ele- ments. The application of the guidelines is illustrated by a case study of a debris-prone bridge on the South Platte River in Colorado. The case study is summarized in this chapter and presented in detail in Appendix D. The case study intro- duces and illustrates the use of field data sheets for evaluat- ing the potential for debris production and delivery from a given watershed. 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) are examined to develop a limited number of debris shapes that would rep- resent the maximum number of configurations found in the field. Simplified, yet realistic, shapes that can be constructed and replicated with a reasonable range of geometric variables were 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. The laboratory testing program is described in detail, including the flume and model piers, instrumentation for data acquisition, and the materials used to fabricate the debris clus- ters. Baseline tests are described and results are compared with several pier scour prediction equations. Tests with the various debris shapes are summarized and results are illustrated with tabular data, photographs, and post-test contour plots. An appraisal of testing results supports the development of an algorithm for predicting the scour anticipated at bridge piers from debris accumulations with rectangular and triangu- lar planforms and varying length, width, and depth geometries. An application methodology is presented to integrate the debris accumulation guidelines with the equation for predict- ing scour at bridge piers with debris loading using the South Platte River case study as an example. Finally, guidelines for inspection, monitoring, and mainte- nance for debris-prone bridges are discussed. An implementa- tion plan for the results of this research completes this chapter. Conclusions from this research and recommendations for future research are presented in Chapter 4. 3.2 Guidelines for Assessing Debris Production, Transport Delivery, and Accumulation Potential This section provides guidelines for assessing the potential for debris production and accumulation at a given bridge site (Tasks 3 and 6 of the research work plan). These guidelines expand on those originally presented by Diehl (1997). The following guidelines should be reviewed prior to a site evaluation and can be used to assist in the completion of the Field Data Sheets developed for this project. The Field Data Sheets (Appendix D, Part 1) have been developed to assist bridge inspectors in collecting and recording data and informa- tion that can be used in assessing the woody debris production, transport, delivery, and accumulation potential for a bridge site. The Field Data Sheets also contain brief descriptions for each of C H A P T E R 3 Guidelines, Testing, Appraisal, and Results

the data sheet entries and the flowcharts that are included within the guidelines. A considerable amount of information can be acquired prior to a site evaluation. Much of this information can be obtained from existing maps, aerial photographs, surveys, and DOT bridge records. Sections A and B of the Field Data Sheets (Lines 2-35) are used to record or identify the availability of this information. Section C of the Field Data Sheets is used to doc- ument specific information and data for a bridge site and the reach upstream that is acquired during a site visit and field examination for use in assessing the potential for debris pro- duction and accumulation at the bridge site. As originally suggested by Diehl (1997) and modified here, there are three major phases for assessing the poten- tial for debris production and accumulation at a bridge site: (1) estimate the potential for debris production and delivery; (2) estimate the potential for debris accumulation on individ- ual bridge elements; and, as modified from Diehl, (3) delineate bridge segments or zones that have the same accumulation potential ratings. These major phases and the associated tasks and subtasks that are required to assess the potential for debris production and accumulation at a bridge site provide the basis for the following guidelines and are described in Table 3.1. There may be direct or indirect evidence for the degree of debris production and delivery potential at any given bridge site; however, direct evidence should be evaluated first and given greater weight than indirect evidence. 3.2.1 Phase 1: Evaluate Potential for Debris Production and Delivery Under this phase, the potential for a stream to produce and deliver debris to a bridge site will be evaluated, the likely max- imum dimensions of individual logs will be estimated, and the site will be divided among location categories that define the local potential for debris accumulation. The location categories reflect local conditions and are not dependent on bridge design. Potential for Debris Production (Task 1a) Observations of the channel upstream of a bridge site as well as observations and knowledge of the physical conditions of the watershed upstream of the site and nearby watersheds can assist in determining the potential for debris production and delivery at a given site. A lack of debris at a bridge site does not indicate that there is a low potential for the production and delivery of debris at a site. Even if debris is relatively sparse at a particular site, infrequent or catastrophic events may pro- duce significant debris available for transport to the site. Thus, data and information collected under Items B1a, B1b, C1b, and C1c of the Field Data Sheets (Appendix D, Part 1) can be used to assist in estimating the potential for debris production upstream of a bridge site. Figure 3.1 provides a flowchart for use in evaluating the potential for debris production upstream of a bridge site. 42 Phase Task Subtask a. Evaluate potential for debris production upstream of the bridge site. Direct evidence Indirect evidence b. Evaluate the potential for debris delivery to the site. Direct evidence Indirect evidence c. Estimate the size of the largest debris that could be delivered to the site. Channel width and depth Maximum design log length 1. Evaluate potential for debris production and delivery. d. Assign location categories to all parts of the bridge crossing. Sheltered Bank/flood plain In the channel (bend or straight reach) In the path (bend or straight reach) a. Assign bridge structure characteristics to all submerged parts of the bridge. Horizontal and vertical gaps between fixed bridge elements Pier and substructure in flow field 2. Evaluate the potential for debris accumulation on individual elements. b. Determine accumulation potential for each part of the bridge. Assume maximum size of potential accumulations 3. Delineate bridge segments that have the same accumulation potential ratings. a. Identify elements with accumulation potential ratings of low, medium, high, and chronic. Delineate zones of low, medium, high, and chronic potential where adjacent elements have the same rating Table 3.1. Major phases, tasks, and subtasks for assessing the potential for debris production and accumulation at a bridge.

Direct Evidence. The primary method of debris produc- tion is through bank erosion that results in woody vegetation being introduced into the channel. Therefore, existing bank erosion along forested streams provides the most direct evi- dence of the potential for debris production. Bank erosion may be extensive and severe or localized and minor. Extensive and severe bank erosion may be evident along straight and meandering reaches of streams that are currently unstable and undergoing incision or downcutting and/or widening. Moderate bank erosion may occur along the outer banks of actively migrating meander bends and in reaches where bars are well developed such that the bars, and any trapped debris, cause flow to impinge directly on the bank. Localized, minor bank erosion may occur at any location and is generally insuf- ficient in magnitude to contribute a significant amount of debris. The most direct evidence for a high potential for debris pro- duction is the presence or absence and density of a riparian forest or corridor along a stream channel upstream of a bridge site. For a river or stream corridor to produce debris that may be available for transport, the channel must contain a forested area or trees along its banks. Streams with cleared banks or sparse or intermittent riparian zones will provide little debris to a channel unless the channel has become unstable because of land use practices. Of course, the upstream channel must have trees in proxim- ity to the channel banks [generally within 100 ft (30 m)] for their introduction through bank erosion to occur. Trees that are leaning over the water at the bankline, incorporated in failed bank sediments at the bank toe, or lying in the water are direct indicators of ongoing bank failure and potentially high debris production. Debris may not be present at a bridge site or in view from the site, but debris may be stored in significant quantities at sites in the upstream channel. These storage areas may not have had sufficient flows to mobilize the stored debris. Sites of debris storage include at the heads of flow splits, on bars and islands, and along the banks, especially along the outer banks of actively migrating meander bends. Rare, large magnitude or catastrophic events such as ice storms, hurricanes, tornados, microbursts or wind shear, and extreme floods can produce considerable amounts of deadfall available for introduction into a channel. If overbank flooding occurs regularly, deadfall in proximity to the channel may be introduced to the channel as large rafts if the path of movement from the flood plain to the channel is relatively unobstructed. Indirect Evidence. Indirect evidence for a high potential for debris production includes historic or ongoing channel changes that affect channel stability and ultimately bank 43 EVALUATE POTENTIAL FOR DEBRIS PRODUCTION Figure 3.1. Flowchart for evaluating debris production potential.

erosion, which is the primary method of debris introduction to a channel. These channel changes include the following: • Downcutting or incision • Lateral migration • Channel widening • Channelization • Dams and diversions • Widespread timber harvesting in the basin • Existing or potential changes in land use practices Some of the general upstream watershed characteristics that may have an indirect impact on debris production are documented under Item B1a, Lines 8–14, of the Field Data Sheets (Appendix D, Part 1). Indirect evidence for a low potential for debris production includes the following: • Woody vegetation growing along the channel and on the slopes leading down to the stream is absent or scarce. • Channel may be fully stabilized (both vertically and later- ally) and is unlikely to undergo significant change. Potential for Debris Transport and Delivery (Task 1b) The potential for debris transport and delivery is obviously dependent on the availability of debris and the channel geom- etry. If debris is readily available, the potential for transport is dependent on the size of the debris relative to the channel width, depth, and planform. Potential for transport is high if the channel width and depth exceed the maximum design log length and diameter. If there is a possibility that the width and depth of the upstream channel could increase in the future, the potential dimensions should be accounted for as well. Yet, even if the channel can accommodate the transport of debris, the channel planform may restrict delivery of the debris to the bridge site. Upstream channel geometry, including average width and depth, and flood plain characteristics are recorded under Item B1b, Lines 15–26, of the Field Data Sheets (Appendix D, Part 1). Figure 3.2 provides a flowchart for determining the potential for debris transport and deliv- ery to a bridge site. Direct Evidence. Observations of existing debris in the channel and at a site provide the most direct evidence for assessing the potential for debris transport and delivery to a site. Direct evidence for a high potential for debris delivery to a bridge site may include the following observations: • Documented chronic or frequent debris accumulations at one or more bridges in the area • Abundant debris stored in the channel and along the banks • Ongoing or prior need for debris removal from the channel 44 • • • • • • • • • • • Direct evidence of potential for debris transport and delivery Indirect evidence of potential for debris transport and delivery Documented chronic or frequent accumulations at one or more bridge sites along this channel or upstream tributaries Abundant debris stored in channel and/or along banks Ongoing or prior need for debris removal from channel Upstream channel geometry too small to transport debris Debris is absent after floods in typical debris accumulation sites other than bridges Negligible debris delivered to a site following major events Debris in the channel remains in place following floods because of low flow velocities Long straight or slightly sinuous reach upstream Direct or unobstructed transport path to bridge site HIGH Potential for Debris Transport & Delivery LOW Potential for Debris Transport & Delivery Estimate Size of the Largest Debris (Key Log Length) Potentially Delivered to Site EVALUATE POTENTIAL FOR DEBRIS TRANSPORT AND DELIVERY Highly sinuous reach upstream Obstructed transport path to bridge site Is it likely that debris will be delivered to the bridge site during subsequent floods? YES NO Figure 3.2. Flowchart for determining the potential for debris transport and delivery.

Direct evidence of a low potential for debris delivery includes the following: • The upstream channel is narrower and/or shallower dur- ing flood flows than most of the debris produced. • Debris is absent after floods in typical debris accumulation sites other than bridges (e.g., on bars and along the outer bank of meander bends). • Negligible debris is delivered to a site following large floods or other catastrophic events. • Debris in the channel remains in place following floods because of low flow velocities. Indirect Evidence. Long, straight reaches upstream of a bridge site will provide the greatest potential for debris delivery. The thalweg of a straight channel is generally near the center- line and debris transported in the channel will generally follow that path. Some debris may become lodged along the banks or on bars as it moves downstream. Streams or rivers with low sinuosity or long, high radius bends can also provide a high potential for debris delivery to a bridge site because the flood path, and consequently the debris path, will generally follow a relatively straight down valley path (Figure 3.3). On forest-lined streams or rivers, as channel sinuosity increases and bend radius decreases, the potential for debris transport and delivery during any given flood decreases because debris is generally transported along the thalweg. Consequently, debris in highly sinuous, well-forested rivers or streams will not be transported far from its source area during any given event and is often deposited along or on top of the outer bank or on the point bar of the next downstream bend. However, some of this debris may eventually be moved downstream to a bridge site during subsequent flows. Streams with actively migrating meander bends may be fairly sinuous, but can produce substantially more debris than less sinuous or relatively straight streams because of the bank erosion associated with lateral migration processes. Yet because of the planform geometry of forested meandering streams, delivery of debris on highly sinuous streams may take longer to reach a bridge in comparison to a low sinuos- ity or straight channel. Therefore, fairly sinuous streams with evident bend migra- tion or bank erosion and significant debris production can be considered to have a moderate to high delivery poten- tial, depending on existing conditions and sound engineer- ing judgment. Bankline characteristics, including evidence of vertical and lateral instability, active bank erosion, and active meander migration are documented under Item C1b, Lines 64–87 of the Field Data Sheets (Appendix D, Part 1). The general upstream riparian corridor characteristics are documented under Item C1c, Lines 88–95, of the Field Data Sheets (Appendix D, Part 1). This information can be used in conjunction with the flowcharts shown in Figures 3.1 and 3.2 to estimate the poten- tial for debris production, transport, and delivery (Lines 96 and 97, Field Data Sheets, Appendix D, Part 1). Estimated Size of the Largest Debris Potentially Delivered to Site (Task 1c) The potential for a channel to deliver debris to a site will be controlled by the ability of the stream to transport it. Existing or future channel dimensions, particularly width, upstream of a site determine the size of debris that can be transported and influences the potential size of accumulations. The maximum design (or key) log length (Figure 3.4) is esti- mated either by examining the largest existing logs in the chan- nel or on the basis of the channel width upstream of the site as measured at inflection points between bends (see Figure 2.6). Diehl (1997) states that the maximum log length on wide channels for much of the United States is about 80 ft (24 m) and that channels less than 40 ft (12 m) wide transport logs with lengths equal to or less than the upstream channel width. 45 Debris path during flooding Thalweg = Debris path during in-bank flows Flood flow path Figure 3.3. Hypothetical debris and flood flow paths during in-bank and out-of-bank flood flows for a low sinuosity channel.

In the eastern United States, channels that are 40 to 200 ft (12 to 60 m) wide transport logs with an estimated design (or key) log length of 30 ft (9 m) plus one quarter of the channel width. Depth is important as well; the depth sufficient to float logs (Figure 3.4) is the diameter of the butt of the tree plus the distance the root mass extends out from the butt, or approxi- mately 3% to 5% of the estimated log length (Diehl 1997). Diehl also indicates that the length of transported logs with attached rootwads rarely exceeds about 30 times the maximum flow depth. The sizes of existing debris and their key log dimen- sions for a given site are recorded under Item C1d, Lines 101–119 of the Field Data Sheets (Appendix D, Part 1). Location Categories for Debris Accumulation at a Bridge Crossing (Task 1d) The delivery and accumulation of debris at a bridge crossing is generally localized. Some areas of the bridge crossing may be free of debris while other areas may receive the bulk of the debris transported to a site. Therefore, each span and pier on a bridge should be evaluated for potential debris accumulation relative to the debris delivery path(s). Figure 3.5 provides the general location categories for a meander bend and a straight reach relative to the local debris delivery path. Figure 3.6 pro- vides a flowchart for determining the location categories. Sheltered by Upstream Flood Plain Forest. A location that can be considered “sheltered” may lie just downstream of a flood plain forest, heavily wooded area, or other major obstruction that can trap transported debris and prevent its delivery to the site. This category may be used where the spac- ing between trees is much narrower than the average tree height and the tree-lined corridor or zone width perpendicu- lar to flow is more than a single or double row of trees. If a forested corridor or zone is likely to be cleared, select a loca- tion category that assumes the location is unsheltered. Flood Plain and Top of Bank. The flood plain and the area along the top of bank are grouped together in a single loca- tion category for the purposes of estimating the local potential for debris delivery. This location category is assigned to areas that are currently clear of trees or are currently forested, but are likely to be cleared in the future. This location category includes any area outside the channel that can be inundated in a design flood event to a depth sufficient to transport debris. Channel. Debris can be transported anywhere in the channel as long as flow depth and channel width are sufficient to transport the debris. Piers on the bank slope or at the bank toe are just as likely to collect debris as those in the channel, especially if they are located at the outer bank of a meander bend. In many cases, piers located on the slope of the bank or at the bank toe can collect sufficient debris to cause severe scour and erosion of the bank. Diehl (1997) suggests that the “channel” location category contain the area between the bases of the banks except where judgment dictates that the bank slopes should be included in the “channel” category. However, most debris will be transported during events that have flow levels well above the base of the bank, especially in more tem- perate regions where flows that fill much of the channel are common. Even in arid regions where flood events are flashy and are generally of substantial magnitude, most flows that are able to transport debris often fill much of the channel well above the base of the bank. In many cases, debris can collect on the lower portions of meander bend point bars well below the top of the bank. Therefore, it is recommended that the “chan- nel” location category include the slopes as well as the basal area of the banks and that the “flood plain” category only extend to the top of the bank. Path of Concentrated Debris Transport. As indicated by Diehl (1997), secondary circulation currents converge at the surface, causing floating debris to be transported along a rela- 46 Butt Rootwad or BoleTrunk Min. Flow Depth Design Log Length Root Mass Extension Butt Diam. Figure 3.4. Schematic of design (or key) log length, butt diameter, and root mass extension.

tively confined path within the channel. This path is closely related to the thalweg. On straight channels this path is gener- ally in the middle of the channel, whereas in meander bends the path is close to the outer bank. However, the transport path during flooding can be significantly different from the path under less than flood flow conditions (see Figure 3.3). Therefore, it is recommended that debris transport during a range of flow conditions should be observed in order to evaluate the debris paths at the bridge crossing. If direct observations of the transport path of debris are not practi- cal or are impossible to obtain, then the location category should be based on the channel characteristics and the probable flow paths under high and low flow conditions as described above. In some cases, where flood flows are out- of-bank, the top of bank and flood plain may be in the debris path, especially along the outside bank of a meander bend. In these instances, these areas may have a “debris path” location category as well. The shape and location of existing debris accumulations on particular elements of an existing bridge are recorded under Item C1d, Lines 101–119 of the Field Data Sheets (Appen- dix D, Part 1). This information along with previously collected information can be used to estimate the debris accu- mulation potential for various elements of an existing bridge (Item C1d, Lines 120–122, Field Data Sheets, Appendix D, Part 1). For new bridges, it may be necessary to estimate the potential for debris accumulation on individual bridge ele- ments based on their location categories as described under 47 Seasonal High Flow Level Design Flood Flow Level Sheltered Top Bank/ Flood Plain Top Bank/ Flood Plain Channel Debris Path (approx.) Meander Bend Seasonal High Flow Level Design Flood Flow Level Sheltered Top Bank/ Flood Plain Top Bank/ Flood PlainChannel DebrisPath (approx.) Straight Reach Channel Sh el te re d Figure 3.5. Location categories relative to local debris delivery in bends and straight reaches.

Phase 2. Also, debris accumulation problems for an existing bridge can be tied to the location categories of those bridge elements identified under Phase 2 where debris accumulates, and this information could then be used to evaluate potential countermeasures and/or develop appropriate maintenance programs. 3.2.2 Phase 2: Estimate Potential for Debris Accumulation on Individual Bridge Elements Bridge characteristics can have a significant influence on the potential accumulation of debris. The design of the substruc- ture, piers, abutments, span widths, and bridge and pier skew all influence the location and degree of debris accumulation at the bridge. The following information along with information previously collected for the Field Data Sheets (Appendix D, Part 1) can be used to estimate the debris accumulation poten- tial for various elements and their location categories for new and existing bridges (Item C1d, Lines 120–122, Field Data Sheets, Appendix D, Part 1). Assign Bridge Structure Characteristics to All Submerged Parts of Bridge (Task 2a) For existing bridges, the current location and configuration of individual components should guide the selection of values for these characteristics. For new bridges, alternative locations and designs should be developed during the design process to assess how they affect the relative potential for debris accumu- lation. Diehl (1997) recommends the following procedure for each design under evaluation: 1. Assign each of the following to one of the location cate- gories as described previously: • Pier • Gap between fixed elements of the bridge crossing • Abutment base • Section of substructure where low steel is wetted by the design flood 2. Determine whether the effective width of each gap exceeds the design log length for the site. 3. Determine whether each pier or substructure section immersed in the design flood includes apertures that carry flow. Pier and Bridge Substructure Characteristics. The char- acteristics of the bridge piers and substructure where they are exposed to floating debris will determine the debris-trapping potential of these features. Piers and substructure elements with narrow apertures that can pass flow during high flows are considerably more likely to trap debris than single piers or solid wall piers. These elements can include pile bents with aligned or battered piles, piers founded on footings or caps over mul- tiple exposed pilings at or near the water surface, or even piers protected by dolphins or fenders. The skew of the flow align- 48 In channel between tops of banks? Sheltered by upstream forest? SHELTERED Location Category DETERMINE LOCATION CATEGORY YES NO TOP BANK/FLOOD PLAIN Location Category CHANNEL Location Category DEBRIS PATH Location Category In path of debris transport? NO NO YES LIKELY Figure 3.6. Flowchart for determining location category along bridge crossing.

ment to these types of piers under various flow conditions can significantly influence their debris-trapping potential. Where flow is skewed, the apertures may trap debris that would nor- mally pass between the piers if they were aligned to flow. In these cases, initiation of the debris accumulation may occur beneath the bridge well back from the upstream end of the pier or pile bent. A skewed flow alignment can produce debris accu- mulations that may be very large and may extend well under the bridge where access may be limited. Open trusses and open parapets with pillars and rails are also likely to trap debris if they are located at or below the water sur- face. Structural arrangements that include pier caps, beams, and deck create openings that can pass flow and trap debris. Pile bents that include bracing also contain openings that can trap debris. Wide Gaps Between Fixed Elements of the Bridge Crossing. New or existing bridges will include one or more gaps through which debris must pass. Gaps include the horizontal openings between support elements of the bridge, the banks, and/or the abutments. If submergence of the substructure is likely to occur under the design discharge, the height of vertical open- ings between the substructure and the streambed or flood plain should also be evaluated. The effective width and height of these gaps need to be estimated with regard to debris trapping potential. Span or Horizontal Gaps. Horizontal gaps or spans are created by openings between the vertical support elements of the bridge as well as the stream banks and the abutments. These areas can become potential sites for debris entrapment and accumulation. Once a log or raft has become lodged across the gap at the surface, additional debris can become trapped across the gap and the rate of accumulation can increase. A horizontal gap or span should be assigned to the most debris-prone location category occupied by fixed elements that define the gap (Diehl 1997). Therefore, a pier-to-bank or pier-to-abutment gap should have the same location category as the pier. If one of the fixed elements is sheltered and the other is not, the gap should be considered unsheltered. If the horizontal gaps or spans are narrower than the longest debris transported by the stream, then the potential for debris accumulation at the bridge site can be high. Under some cir- cumstances, the potential for a span blockage may be high even though the span may exceed the maximum log length. For existing bridges, this situation may be evident where large iden- tifiable debris accumulations on adjacent piers have induced a span accumulation (Figures 3.7 and 3.8). For new bridges where one or more piers will be located in the channel and the potential for debris accumulation on the piers is determined to be high then the potential for gap spanning debris accumula- tions can also be high, especially where large debris accumula- tions on piers can induce span accumulations. This holds true for pier-to-bank and pier-to-abutment gaps as well. Vertical Gaps. Vertical gaps are created when high flows reach the bridge substructure and debris becomes trapped between the structural elements below the bridge deck and the flood plain or streambed. Because most debris is transported at the surface, when pieces hit the substructure, most rotate to the side and are either trapped against the substructure ele- ments at the surface or swept under the substructure. Where debris intersects the bridge structure endwise, the upstream end of the debris, if it is sufficiently long enough, may rotate downward and become wedged in the streambed or flood plain. Hammerhead or similar type piers may also act to force 49 Pi er Pi er Figure 3.7. Schematic of span debris accumulation induced by adjacent pier accumulations. Source: Mike Collier, Debris Free, Inc. Figure 3.8. View of gap-spanning debris accumulation induced by adjacent pier on the Harpeth River in Tennessee.

one end of the debris downward toward the streambed or flood plain, which can cause the other end to project upward into the substructure as shown in Figure 3.9. For this reason, the height of the vertical gap (i.e., vertical gap width) along the bridge should be evaluated relative to the typical log length and location category. Debris that is significantly shorter than the design log length can become trapped between the bridge substructure and the ground, so a range of typical log lengths should be evaluated with regard to the vertical gap width, especially when evaluating the vertical gaps in flood plain areas. Because the width of the vertical gap is defined by the height of the gap between the low steel of the bridge substruc- ture and the streambed or the flood plain and may vary along the gap as well as along the bridge, a vertical gap should, therefore, be assigned to the most debris-prone location cat- egory occupied by the fixed elements that define the gap (Diehl 1997). Determine Potential for Accumulation by Location and Type (Task 2b) The potential for debris accumulation is determined separately for each pier, each section of the submerged sub- structure, and each horizontal/vertical gap/span between fixed elements of the bridge following determination of the poten- tial for delivery of debris to the site, assignment of location cat- egories, and identification of other bridge characteristics. It should be noted that the estimation of the potential for debris delivery and accumulation does not address the potential for or likely size of debris accumulations at a given site. Depending on log dimensions, flow depth, and the number and proxim- ity of gaps/spans and piers affected, debris accumulations can grow to maximum sizes both vertically and horizontally (Diehl 1997). Diehl suggests that estimates of debris accumulation sizes should be conservative. Therefore, he recommends assum- ing that the debris accumulation will extend from the water sur- face to the streambed, that an accumulation on a single pier will have a width equal to the design (or key) log length over its full depth, and that an accumulation across two or more piers will extend laterally half the design log length beyond the piers. Because the maximum size of debris accumulations on substructure cannot be evaluated, Diehl (1997) recom- mends assuming that the apertures in the bridge substructure will be completely closed by debris and that the entire vertical gap between the substructure and the streambed or flood plain could ultimately be closed by debris. Potential for Debris Accumulation at Each Pier and Substructure Section. The potential for debris to accumu- late on each pier and section of the substructure is determined using the flowchart shown in Figure 3.10. Once the potential for delivery of debris is estimated for the entire bridge site, the location category and the presence or absence of narrow aper- tures that can pass flow are evaluated for each pier and section of the substructure. Potential for Debris Accumulation Across a Span or Vertical Gap. The potential for debris to span an individ- ual gap between fixed elements of the bridge opening is deter- mined using the flowchart shown in Figure 3.11. Once the potential for delivery of debris is estimated for the entire bridge site, the effective length of the span/gap between fixed elements relative to the design log length and the location cat- egory of the gap are evaluated for both horizontal and verti- cal spans/gaps. Effects of Accumulations That Modify Analysis. Because hydraulics and debris-trapping characteristics of a bridge can change with debris accumulation, the potential for additional trapping can increase as well. Diehl (1997) recommends that, after all possible debris accumulations at a bridge have been assigned a potential for occurrence, the analysis should be run again to determine whether some individual accumulations increase in potential or may affect the potential for adjacent accumulations elsewhere along the bridge. 3.2.3 Phase 3: Determine the Overall Debris Accumulation Potential for the Bridge Diehl (1997) suggests that the potential for debris accu- mulation at a bridge is the maximum or worst case of the potentials determined for each pier, substructure section, or gap/span between fixed elements. However, this method could result in a relatively long bridge having an overall high 50 Source: Mike Collier, Debris Free, Inc. Figure 3.9. View of large debris wedged between bridge substructure and channel bed on Carson River in Nevada.

High LOW Potential LOCATION CATEGORY LOCATION CATEGORY MEDIUM Potential HIGH Potential HIGH, CHRONIC Potential YES NO PIER OR SUBSTRUCTURE DESIGN SOLID WITH APERTURES Potential for Delivery of Debris Low High Low Low High DEBRIS PATH DEBRIS PATHCHANNEL FLOOD PLAIN/ TOP BANK CHANNEL FLOOD PLAIN/ TOP BANK Potential for Delivery of Debris Potential for Delivery of Debris PIER OR SUBSTRUCTURE SECTION IN SHELTERED LOCATION CATEGORY? Figure 3.10. Flowchart for determining the potential for debris accumulation on a single pier or section of the bridge substructure. Horizontal/vertical gap/span wider than design log length? LOW Gap/Span Blockage Potential GAP/SPAN IN SHELTERED LOCATION CATEGORY? Potential for gap/span blockage due to large adjacent pier accumulations? MEDIUM Gap/Span Blockage Potential HIGH Gap/Span Blockage Potential HIGH, CHRONIC Gap/Span Blockage Potential Channel Is pier-to-pier (or pier- to-bank) gap wider than design log length? YES NO Debris Path Flood Plain / Top of Bank Channel Debris Path Location Category NO Location Category Potential for Delivery of Debris NOYES YESNO YES HIGHLOW Flood Plain / Top of Bank Figure 3.11. Flowchart for determining the potential for debris accumulation across a horizontal or vertical span or gap.

or chronic rating for debris accumulation potential that is based on a high or chronic rating for only one or two bridge elements. Therefore, it is recommended here that each indi- vidual element and gap/span should be treated as a separate entity in an effort to reduce design and construction costs, while maintaining structural integrity. Major bridge segments can be compartmentalized for design or maintenance purposes by identifying and delineating those zones where adjacent structural elements and gap/spans have the same debris accu- mulation potential rating. Construction or maintenance costs could be significantly reduced through the use of this compartmentalization or zonation of potential debris accu- mulation ratings. 3.3 Application of the Guidelines: South Platte River Case Study Based on recommendations from the NCHRP 24-26 panel, a field investigation and case study were conducted as part of the revised research work plan. The intent of the field inves- tigation was to identify potential case study locations on the South Platte River in Colorado in order to evaluate the appli- cability of the guidelines (see Section 3.2) and Field Data Sheets (see Part 1 of Appendix D) developed for this project. The Field Data Sheets were used to document site character- istics such as channel type and size, channel instability, bank erosion and retreat, and bank vegetation characteristics in detail. The purpose of the case study is to provide an example of how the practitioner should apply the guidelines for assess- ing debris production and the potential for debris accumula- tion at a bridge site. 3.3.1 Field Reconnaissance Potential case study bridge sites were examined on June 7, 2006, along a stretch of the South Platte River between the cities of Greeley and Sterling, Colorado. Upstream of the potential study sites, the South Platte River Basin has a drainage area of approximately 14,600 mi2. Headwaters of the South Platte River are located in the central Colorado mountains, where the mean annual precipitation is about 30 in. including approxi- mately 300 in. of snowfall (USGS 2006). On the northern Colorado plains between Greely and Sterling, Colorado, where the potential study sites were located, the mean annual precip- itation is about 12 in., primarily in the form of rain that typi- cally falls April through July. The average June flow in the South Platte River in the corri- dor containing the potential study sites is 1,260 ft3/s. During the field reconnaissance, the flow was approximately 90 ft3/s at the USGS gauge station in Fort Morgan. Unusually warm tem- peratures in May 2006 caused snowpack to melt faster than normal, resulting in an earlier runoff peak, which occurred on the South Platte River at Fort Morgan in early May 2006 at 182 ft3/s, just prior to the field reconnaissance. Land use in the corridor containing the potential study sites is primarily agriculture and some rangeland. Continuing mod- erate to severe drought conditions in Colorado have resulted in the need to shut off irrigation wells that draw from a shallow aquifer in the area in order to ensure minimum flow condi- tions in the South Platte River are maintained. During the field visit, flow depths were observed to decrease moving down- stream despite the confluence with several tributaries due to what appeared to be irrigation diversion. Sixteen bridge sites were examined during the field reconnais- sance; two of the bridge sites within Weld County, Colorado, were identified as potential candidates for case study locations. These sites were the bridges carrying Weld County Road 37 (WEL 37) and Weld County Road 50/67A (WEL 50/67A) over the South Platte River. The other fourteen sites were eliminated for a number of reasons: • Stable banks did not contribute debris to channel. • Several bridges had relief bridges that may reduce exposure to flood flows. • Flow diversion drastically reduced channel flow. • Older riparian vegetation was set back from the banks min- imizing large woody debris in channel. Bridge WEL 50/67A was selected for the case study. The bridge is located on the apex of a meander bend of the South Platte River (Figure 3.12) at Hardin, Colorado. Large woody debris, composed primarily of large mature cottonwoods, was observed in the channel and on the banks upstream of the bridge. Active bank erosion and channel migration were evi- 52 Source: Google Earth Pro Figure 3.12. Aerial view of the Bridge WEL 50/67A case study site.

dent on both banks upstream of the bridge and are the primary mode of debris delivery to the channel. Bridge WEL 50/67A has three sharp-nosed piers about 18 in. (0.5 m) wide and skewed approximately 5° to high flows. Two piers had debris buildup on the nose and sides. The left and right abutments had about 12 in. (0.3 m) diameter stone riprap protection, with the riprap protection on the left bank extending about 300 ft (91 m) upstream. Pier 1 is located on a mid-channel bar that has a moderately dense cover of grass. A triangular debris pile consisting of several logs with a small [less than 1 ft (0.3 m) deep] scour hole was located on the nose of Pier 1 (Figure 3.13). Debris on Pier 2 consisted of a single log approximately 14 in. (0.36 m) in diameter with a 4 to 6 ft (1.2 to 1.8 m) diameter rootwad positioned at the upstream end. A debris pile and a scour hole, approximately 2 to 3 ft (0.6 to 0.9 m) deep and 8 to 10 ft (2.4 to 3.0 m) wide, were observed along the upstream end of the right abutment. The debris that had accumulated on the right abutment had no discernable key log, but was composed of several logs approximately 18 in. (0.5 m) in diameter with one log having a 6 to 8 ft (1.8 to 2.4 m) diameter rootwad positioned at the upstream end. 3.3.2 Development of the Case Study Based on field conditions identified during the field recon- naissance, it was determined that Bridge WEL 50/67A was the best candidate for the case study. The case study was con- ducted by completing the Field Data Sheets for the bridge site both in the field and in the office following additional data collection efforts (see Part 2 of Appendix D). The completed Field Data Sheets for the bridge are provided in Part 3 of Appendix D. Once the Field Data Sheets were completed, the data and information were used with the guidelines to assess the potential for debris production, transport, delivery, and accumulation at the bridge site. Based on the data and infor- mation collected in the field and application of the guidelines, the following potentials were determined: • Estimated potential for debris production: HIGH • Estimated potential for debris transport and delivery: HIGH • Estimated debris accumulation potential – Left Bank Abutment: LOW – Pier 1: MEDIUM – Pier 2: HIGH – Pier 3: HIGH – Right Bank Abutment: LOW • Estimated span blockage potential: – Span 1: LOW – Span 2: LOW – Span 3: LOW – Span 4: LOW 3.4 Development of Debris Characteristics for Laboratory Testing 3.4.1 Debris Accumulation Characteristics Typical debris accumulation characteristics were identified based on the literature review (Section 2.1), an examination of the debris photographic archive (Section 2.2), the survey (Section 2.4), the field pilot study in Kansas (Section 2.5), and the field reconnaissance for the case study (Section 3.3.1). Identification of typical debris shapes and geometry was a necessary preliminary to developing a laboratory testing pro- gram. Considering laboratory testing budget constraints, the objective was to develop a limited number of debris shapes that would represent the maximum number of debris config- urations found in the field. It was also necessary that the shapes be simplified so that they could be constructed and replicated with a reasonable range of geometric variables for laboratory testing. At bridge piers, debris characteristics at a minimum might include a single-pier floating cluster, a floating raft bridging two or more piers, and submerged or sunken variations of these configurations. The debris accumulations shown in the photographs compiled in the debris photographic archive (Appendix A) were evaluated for specific geometric charac- teristics and guided identification of the specific geometries described in the following paragraphs. A summary of the observed planform and accumulation types—organized by region, photograph source, and location—is provided in Table 3.2. The photographs were also used to identify typical geometries of the debris piles and to provide rough estimates 53 Figure 3.13. View looking down at debris cluster and scour hole on the upstream end of Pier 1 of Bridge WEL 50/67A.

54 Rectangular Triangular Profile Geometry Source State Stream Number of Photos Physio- graphic Region Eco- region Single Log Multi Mass Multi Mass Conical Rectangular Inverted Cone Pacific Coast Arroyo Grande 11 24 260 X X Hopper Creek 3 24 M260 California (Central & South Coast) Salsipuedes Creek 5 24 M260 X California (Mountain) N. Fork Deer Creek 4 23 260 Mad River 4 24 260 X X Trinity River 1 24 260 X California North Coast Yager Creek 6 24 260 X X Sacramento River 2 24 260 X X Stony Creek 6 24 260 X X Kevin Flora Caltrans (Survey) California (Central Valley) Thomes Creek 4 24 260 X Malibu Laguna 1 21 260 Harris Creek 1 25 M262 California Stony Creek 1 24 260 X Ayres Assoc. Oregon Shitike Creek 1 24a 242 Debris Butte Creek 9 23 260 Free, California (North) Navarro River 13 24 260 X Inc. California Callegas Creek 18 24 260 X X (Mike (South Santa Clara River 26 24 260 X X Collier) San Antonio Creek 9 24 260 X Adams Creek 3 24 M260 Ventura River 29 24 260 X X Oregon Calapooia River 1 23 M240 North Santiam River 3 23 M240 X Washington Cowitz River 7 24 240 X N. Fork Skagit River 6 23 M240 X N. Fork Skagit River 2 23 M240 S. Fork Skagit River 1 23 M240 X Skykomish River 4 24 240 X Timothy Diehl (1997) Washington Queets River 1 24 M240 West Nevada Carson River 6 22 340 X Colorado River 4 21 340 X Debris Free, Inc. (Mike Collier) Utah San Rafael River 7 21 340 X X Jefferson River 14 19 M330 X Boulder River 7 19 M331 Russel Brewer MtDOT (Survey) Montana St. Regis River 3 19 M332 X C. Miller NDOT (Survey) Nevada Carson River 15 22 340 X Tim Ularich UDOT (Survey) Utah Santa Clara and Virgin Rivers 11 21 340 X X New Mexico Rio Grande 1 22 310 X X Colorado Bijou Creek 2 13 330 Ayres Assoc. Idaho Teton River 1 18 M330 Santa Clara River 8 22 340 X Beaver Dam Creek 4 22 340 X Various Web- sites (KSL Ch. 5 TV, St. George) Utah Unknown 11 22 340 X Midwest Debris Illinois East Skokie Ditch 3 12 220 X X Free, Iroquois River 4 12 250 Inc. Mississippi River 3 12 250 (Mike Mackinaw River 3 12 250 Collier) Unknown 6 Indiana Eel River 8 12 220 X Eel River 4 12 220 X Vermillion River 11 12 220 X Wabash River 3 12 220 X Kansas Republican River 6 12 250 X Smoky Hill River 34 13 330 X Minnesota Minnesota River 15 15 250 Missouri Unknown 1 X Table 3.2. Inventory of photographs of debris at bridge piers classified by shape.

Rectangular Triangular Profile Geometry Source State Stream Number of Photos Physio- graphic Region Eco- region Single Log Multi Mass Multi Mass Conical Rectangular Inverted Cone Kentucky Unknown 2 X X Texas Clear Fork Brazos Rv. 6 12 310 Debris Tennessee Harpeth River 8 11 220 Free, Inc. Oklahoma Wild Horse River 3 12 250 (Mike Collier) North Canadian River 12 12 310 X X Brad Kansas Verdigris River 12 12 250 X X Rognlie Smoky Hill River 73 13 330 X KDOT Chikaskia River 1 13 330 (Survey) Elm Creek 25 13 330 Neosho River 19 12 250 Elk River 25 12 250 Unknown 6 X Ayres Kansas Neosho River 38 12 250 X Assoc. Neosho River (US 400) 28 12 250 X Verdigris River 14 12 250 X X Elk River 32 12 250 Missouri Florida Creek 1 12 250 Nebraska South Platte River 1 12 250 Indiana Indian Creek 1 12 220 MNDOT District 1 (website photos) Minnesota Black River 2 1 212 X USGS (Robinson 2003) Indiana White River 6 12 220 X Tim Iowa Cedar Creek 3 12 220 Dunlay Iowa East Nishnabotna River 6 12 250 X Iowa DOT (Survey) Iowa West Nodoway River 3 12 250 X John Holmes MoDOT (Survey) Missouri Chariton River 9 12 250 X X Ken Foster MoDOT Missouri Big Creek 2 12 250 X (Survey) West Fork of Grand Rv. 8 12 250 X Larry Cooper Minnesota Minnesota River 2 12 220 MNDot (Survey) Minnesota Mississippi River 2 12 220 Duane Hill MNDOT (Survey) Minnesota S. Branch of Wild Rice River 2 12 250 Brandon Callett ODOT (Survey) Ohio Great Miami River 14 12 220 X Allan Bjorklund Wisconsin Pikes Creek 1 1 210 X WisDOT (Survey) Bad River 1 1 210 X Jerry Conner TxDOT (Survey) Texas Clear Fork of Brazos Rv. 6 12 310 Tim Hertel TxDOT (Survey) Texas Red River 12 12 250 X Jon Zirkle TDOT (Survey) Tennessee Harpeth River 13 11 220 Timothy Diehl Tennessee Harpeth River 2 11 220 X (1997) Indiana White River 1 12 220 East N. Carolina Deep River 4 4 230 X Debris Virginia Appomattox River 4 4 230 X X Free, James River 2 3 230 Inc. Meherrin River 3 4 230 (Mike Dan River 2 4 230 Collier) Nottoway River 4 4 230 Midwest (cont.) Ohio Unknown 5 X Table 3.2. (Continued). (continued on next page)

of debris accumulation length upstream of the pier and debris accumulation width relative to pier width. Table 3.3 provides a list of sites from the photographic archive that were used to acquire measurements of relative debris accumulation width and length. Debris was observed accumulating at bridge piers as single logs, multiple logs, or a mass of logs. The single-log accumula- tion was composed of one or two logs that had become trapped on a pier or between spans. The multiple-log accumulation consisted of several logs that were loosely intertwined and had no filling of the interstices or matrix with finer debris, detritus, and sediment. The mass of logs accumulation was composed of a cluster of logs and other debris tightly interlocked with almost all of the matrix or interstitial openings filled with smaller debris, detritus, and sediment. Unlike the mass of logs accumulation, in almost all cases, the single-log and multiple- log types of accumulations did not extend upstream for any significant distance. Figures 3.14 through 3.16 present schematics of the three observed accumulation types as well as example photographs for each accumulation type. Although Dongol (1989) used conical, cylindrical, and elliptical shapes in his modeling, most debris accumulations observed in the photographic archive could be considered either triangular or rectangular in planform. Triangular debris accumulations tend to have a conical shape in profile, while rectangular accumulations tend to have a rectangular profile. Figures 3.17 and 3.18 present plan view schematics of the tri- angular and rectangular planforms, respectively. Figures 3.19 and 3.20 illustrate the conical and rectangular profile geome- tries. Figure 3.21 provides a photograph of a triangular debris pile with a conical geometry after the water has receded and the pile has collapsed upon itself. Figure 3.22 presents a photo- graph of a rectangular debris pile with a rectangular geometry after the water has receded and the pile has collapsed upon itself. Both types of debris accumulation profiles can grow from being a surface accumulation to being partially or fully 56 Rectangular Triangular Profile Geometry Source State Stream Number of Photos Physio- graphic Region Eco- region Single Log Multi Mass Multi Mass Conical Rectangular Inverted Cone Photos Website Photos Tennessee Coal Creek 4 8 220 South Debris S. Carolina Little River 40 4 230 Free, Florida Escambia River 2 3 230 X Inc. Tennessee Wolf River 3 3 230 (Mike Jackson Dist. Var. 193 11 230 Collier) Arkansas St. Francis River 34 3 230 X Louisiana Amite River 10 3 230 X Ayres Louisiana Bayou Boeuf 1 3 230 Assoc. Mississippi Abiaca Creek 1 3 230 X Jack Creek 2 3 230 Sykes Creek 1 3 230 X Website Photos Louisiana Red River Raft 2 3 230 MDOT, Br. Des (Survey) Mississippi Coles Creek 1 3 230 X J. Zirkle TxDOT (Survey) Texas Wolf River 3 3 230 J. Howell TxDOT (Survey) Texas San Marcos River 1 3 250 J. Kilgore TxDOT Texas Cocklebur Creek 7 3 310 (Survey) Texas San Antonio River 13 3 250 B. Laywell TxDOT Texas Little River 7 3 250 X (Survey) Texas Brushy Creek 6 3 250 X H. Schroeder TxDOT (Survey) Texas Rocky Creek 6 3 250 X Timothy Texas Brazos River 1 12 310 X Diehl (1997) Unknown Unknown 5 X East (cont.) W. Virginia Mud River 2 8 220 X Tug River 2 8 220 X X NCDOT Website N. Carolina Deep River 5 4 230 X Table 3.2. (Continued).

submerged, depending on flow depth. Figures 3.23 and 3.24 display schematics of the various submergence possibilities. A third type of profile geometry is the inverted conical pro- file, which generally has a triangular planform. This type of accumulation is very common and usually occurs following one or more floods when an accumulation with a triangular- conical geometry settles onto the bed of the channel. The lower portion of the accumulation then becomes embedded in the bed. When the next flood occurs, the debris accumu- lation remains trapped on the bed, but can grow in size because of trapping of additional debris. As more debris is trapped by the existing debris pile during subsequent flows, 57 Geographic Region State Stream W/a L/a L/W Pacific Coast California (Central Valley) Stony Creek 10.0 6.0 0.6 Pacific Coast California (North Coast) Butte Creek 31.3 8.8 0.3 Pacific Coast California (South Coast ) Callegas Creek 16.3 13.3 0.8 Pacific Coast California (South Coast ) Callegas Creek 15.3 13.3 0.9 Pacific Coast California (North Coast) Navarro River 43.0 48.0 1.1 Pacific Coast California (South Coast ) San Antonio Creek 8.0 5.0 0.6 Pacific Coast California (Central Valley) Sacramento River 7.1 7.1 1.0 Pacific Coast California Stony Creek 15.0 10.0 0.7 Pacific Coast California (Central Valley) Thomes Creek 7.6 6.3 0.8 Pacific Coast Washington Skagit River 8.3 11.7 1.4 Pacific Coast California (Central and South Coast) Arroyo Grande 30.0 15.0 0.5 Pacific Coast California (North Coast) Yager Creek 5.0 6.7 1.3 West Utah Santa Clara River 13.3 8.0 0.6 West Utah Virgin River 28.0 36.0 1.3 West Utah Colorado River 10.0 4.0 0.4 West Utah San Rafael River 5.6 4.0 0.7 West New Mexico Rio Grande 8.5 3.2 0.4 Midwest Kansas Verdigris River 13.3 11.7 0.9 Midwest Iowa East Nishnabotna River 17.0 16.0 0.9 Midwest Indiana Eel River 22.0 24.7 1.1 Midwest Kansas Smoky Hill River 16.0 21.7 1.4 Midwest Ohio Unknown 28.0 10.0 0.4 Midwest Tennessee Harpeth River 23.3 19.0 0.8 Midwest Tennessee Harpeth River 14.0 20.0 1.4 Midwest Tennessee Harpeth River 18.0 9.0 0.5 Midwest Indiana White River 12.5 20.0 1.6 Midwest Kansas Neosho River 17.8 8.0 0.4 Midwest Kansas Verdigris River 12.0 17.5 1.5 Midwest Red River Texas 15.7 13.8 0.9 Midwest Red River Texas 6.9 6.7 1.0 East West Virginia Tug River 6.4 17.0 2.7 East Tennessee Coal Creek 17.5 3.0 0.2 East Virginia Nottoway River 10.0 7.0 0.7 South Louisiana Amite River 11.0 6.0 0.5 South Tennessee Wolf River 15.0 6.0 0.4 South Arkansas St. Francis River 12.5 15.0 1.2 South South Carolina Little River 10.0 7.0 0.7 Unknown Unknown Unknown 14.3 4.9 0.3 Average 15.1 12.4 0.9 Range 5.2 - 43 3 - 48 0.2-2.7 NOTE: a = Pier Width, L = Length of Debris Pile, W = Width of Debris Pile Table 3.3. Measurements of debris pile width and length.

factor to the overall equation (such as the Kw factor for wide piers) or an adjustment to the pier width used as an input variable to the equation (similar to the HEC-18 complex pier approach). The goal of the laboratory plan was to develop a series of tests for a wide range of debris configurations that can be run quickly and efficiently. The tests would be performed for single debris clusters at individual piers, which was the primary type of debris accumulation identified by all regions in the survey. These tests would then be supple- mented to address specific issues related to other factors that would not be incorporated into the initial runs. The majority of the testing would be performed for clear-water 58 Figure 3.15. Multiple-log debris accumulation. Figure 3.14. Single-log debris accumulation. a rectangular-rectangular geometry may develop. Figure 3.25 depicts a schematic of the inverted cone scenario. Figure 3.26 presents a photograph of a debris pile with an inverted cone geometry after the water has receded and the pile has col- lapsed upon an existing pile embedded in the bed. 3.4.2 Laboratory Testing of Debris Testing Requirements. The goal of the laboratory test- ing was to provide sufficient data for a range of debris accu- mulations to develop adjustment factors to the HEC-18 pier scour equation. The adjustment factors could be a correction

59 Figure 3.16. Mass of logs debris accumulation. Figure 3.17. Triangular debris pile planform. sediment transport conditions (approach flow velocity less than the critical velocity to initiate sediment transport) for durations much less than would be required to achieve ulti- mate scour. The duration will, however, be sufficient to achieve at least 60% of ultimate scour. This approach to the laboratory testing maximized the number and range of debris configurations that would be tested within the labo- ratory budget. The testing should include a range of debris characteristics including debris accumulation shape, thickness, width, and length. The range of debris accumulation size that would be tested in the laboratory was related to actual debris accumu- lations observed by the research team in the field or from the survey sources and the photographic archive (Section 3.4.1). Figures 3.27, 3.28, and 3.29 illustrate the debris shapes (rec- tangular, conical, and collapsed in profile and either rectan- gular or triangular in planform) that were modeled and define the dimensions for the various shapes. The dimensions were varied in order to model the range of conditions typi- cally seen in the field. Debris Dimensions. All of the physical modeling was conducted in the 8 ft (2.4 m) wide flume at Colorado State University under clear-water flow conditions using 4 in. (10.2 cm) square piers. This scale and flow condition were selected to maximize the number of debris conditions that can be modeled because scour should develop rapidly at this scale and clear-water runs are also less time consuming.

Table 3.4 shows a summary of the observed debris dimensions contained in Table 3.3 and the range of debris dimensions for the laboratory physical modeling. All of the dimensions were normalized by the pier width so the field conditions could be used to develop a realistic range of laboratory runs. The range of debris dimensions was selected to encompass the range observed in the field +/− one standard deviation around the mean. It should be noted that W/a of 24 requires use of the 8 ft (2.5 m) wide flume using a 4 in. (10.2 cm) pier. Additional testing was conducted using 1 in. × 8 in. (2.5 cm × 20 cm) wall piers, and a pile bent having four 0.5 in. (1.3 cm) diameter columns (see Section 3.5). 3.5 Laboratory Testing Program 3.5.1 Testing Facilities and Protocols Laboratory testing of effects of debris on bridge pier scour was conducted at the Hydraulics Laboratory at Colorado State University (CSU). An indoor recirculating flume measuring 8 ft (2.4 m) wide by 200 ft (61 m) long was utilized for all laboratory tests conducted during this research program. 60 Figure 3.18. Rectangular debris pile planform. Figure 3.19. Conical profile geometry. Figure 3.20. Rectangular profile geometry.

Flume Description All laboratory tests were conducted under clear-water con- ditions with a sand bed 1.5 ft (0.46 m) thick. The sand had a median grain size d50 of 0.7 mm. Water was supplied by two 125-horsepower pumps, which could operate separately or in tandem to achieve a desired discharge, up to a maximum flow capacity of 55 ft3/s (1.6 m3/s). For all tests, the slope of the flume was kept constant at 0.1%. A flow straightening assem- bly was placed at the entrance to the flume to eliminate large- scale circulation and condition the flow field prior to entering the test section. Four pier locations (A through D) were established in the flume along the centerline. All piers were designed to be constructed to a height of 3 ft (0.9 m) above the floor of the flume. Fabrication of the piers was performed by CSU shop personnel. 61 Figure 3.21. Collapsed triangular debris pile with conical profile geometry. Figure 3.22. Collapsed rectangular debris pile with rectangular profile geometry. Figure 3.23. Partially submerged debris accumulation. Figure 3.25. Inverted cone profile geometry. Figure 3.24. Fully submerged debris accumulation.

62 Figure 3.26. Collapsed debris pile with an inverted cone geometry. LOOKING DOWNSTREAM H d Pier width = a PROFILE H d Pier length = a L Flow W Figure 3.27. Rectangular shape definition sketch.

63 H d L LOOKING DOWNSTREAM H d Pier width = a W PROFILE Pier length = a Flow Figure 3.28. Conical shape definition sketch. PROFILE H d Pier length = a L Flow LOOKING DOWNSTREAM H d Pier width = a W Figure 3.29. Collapsed shape (inverted cone) definition sketch.

All piers were designed and constructed in two sections. The lower piece remained secured to the floor of the flume while the upper section could be removed at the pre-test sand level in order to easily re-level the bed after each test. A level bed sur- face was produced by a screed board attached to a cart and drawn across the surface of the bed. A photograph of the flume before installation of the piers is presented in Figure 3.30. A schematic of the flume showing the pier layout and ancillary features is presented in Figure 3.31. Piers Three types of piers were fabricated for the testing program: 4 in. (10.2 cm) square piers, 1 in. wide by 8 in. long (2.5 cm by 20 cm) wall piers, and pile bent piers having four 0.5 in. (1.3 cm) diameter columns. Before each run, four piers of predeter- mined shape were installed along the centerline of the flume. The piers were secured to the flume floor and extended approx- imately 1.5 ft (0.46 m) above the sand bed. Each pier has a removable upper section to allow automated mapping of scour holes using an array of 16 ultrasonic transducers on the data cart (further explained in the following section). Figure 3.32a shows a fully assembled 4 in. (10.2 cm) square pier, and Figure 3.32b shows the upper and lower halves of the same pier separated to allow for data collection. Data Acquisition Hydraulic Data Acquisition. Prior to each test, the tail- gate was closed and the flume slowly filled with water until the target flow depth of 1 ft (30.5 cm) was established. Flow was introduced very slowly to ensure no local scour occurred dur- ing startup. During the slow filling process, air was allowed to escape from the sand bed. Figure 3.33 is a photograph of a 4 in. (10.2 cm) square pier as flow was introduced to the test- ing flume. With the flume full of water, discharge was slowly increased to the target discharge, while simultaneously the tailgate was opened until steady flow at the target depth of 1 ft (30.5 cm) relative to the initial bed surface was obtained. This process ensured a very gradual acceleration of flow until the target velocity was achieved and maintained. This process took about 1.5 to 2 hours to accomplish. Each run then proceeded for a duration of 8 hours while velocity and water surface data were collected at each pier and at designated locations between piers. For long duration tests, the duration was increased to 72 hours per run. After each test, the discharge was gradually decreased and the tailgate adjusted to ensure that no additional scour occurred during the drain-out period. Typically, the flume was allowed to drain out overnight, and the sand bed around each pier was mapped the next day. Velocity and water surface elevation were monitored peri- odically at predetermined locations during each test to ensure target hydraulic conditions were maintained. A motorized cart traversed the flume along a track attached to the top of the flume and served as a platform to mount data acquisition equipment. Water surface elevations and velocity profiles were documented at designated locations along the flume. Water surface elevations were measured utilizing a point gauge assembly mounted to the mobile data acquisition cart. Accuracy of the point gauge was 0.01 ft (3 mm). Velocity acquisition equipment was mounted to the cart via the same point gauge. Measurements were recorded adjacent to each 64 W/a L/a L/W Field and Photograph Measurements Average 15.1 12.4 0.9 Range 5.2 – 43 3 – 48 0.2 – 2.7 St. Dev. 8.2 9.2 0.5 -/+ St. Dev. 6.9 – 23.3 3.2 – 21.5 0.4 – 1.3 Recommended Laboratory Tests Range 6 – 24 3 – 24 0.5 – 1.5 Table 3.4. Field and laboratory debris dimensions. Figure 3.30. Eight-foot (2.4 m) flume before pier installation.

65 60 m 13 m Mobile data acquisition cartConcrete cap PLAN VIEW 10 m10 m 10 m17 m Pier APier BPier CPier D Point gauge assembly with velocity probe Tailgate PROFILE flow 0.3 m 0.5 msand bed 2.4 m Figure 3.31. Schematic plan and profile of testing flume. a. Four-inch square pier fully assembled. b. Four-inch square pier separated for data collection purposes. Figure 3.32. Four-inch square pier.

pier and between each pier to quantify the water surface ele- vation and the velocity profile. Velocity data were collected and recorded with a SonTek Acoustic Doppler Velocimeter (ADV). Three main compo- nents made up the ADV: the probe head, the conditioning module, and the data recorder. The probe head, a three- pronged apparatus that was submerged to a predetermined depth, was attached to the point gauge assembly. Velocities were measured in a three-space coordinate system within a sampling volume located approximately 1.2 in. (5 cm) below the probe head. The data-conditioning module served as the link between the probe head and data recorder. Digital pro- cessing, necessary to interpret the Doppler signal from the probe head, was performed by the conditioning module. A personal computer was used as a data recorder. A schematic of the point gauge assembly is shown in Figure 3.34. Hydraulic data were collected at a fixed set of 52 locations for each of the configurations considered for the specified set of piers. Flow-depth data were collected at all 52 locations. Twelve of these locations consisted of 4-point velocity profile measurements (20%, 40%, 60%, and 80% depths) in addi- tion to the flow-depth data. The remaining 40 locations con- sisted of a velocity measurement taken at the 60% depth elevation in addition to the depth measurement, resulting in a total of 88 velocity measurements per data collection set. There were 13 different data collection locations at each pier. Data were generally collected at positions of 6 ft (1.8 m), 4 ft (1.2 m) (centerline), and 2 ft (0.6 m) from the right flume wall at stations 10 ft (3 m) and 4 ft (1.2 m) from each upstream pier face and 8 ft from each downstream pier face; 66 Figure 3.33. Flow initiation at the start of a test. Data Recorder Velocity Probe Head Data Acquisition Cart Point Gauge Assembly Point Gauge Track Elevation Reading Sand Bed Flume Wall Flume Wall Conditioning Module Figure 3.34. Velocity data acquisition setup.

and at positions of 7 ft (2 m), 6 ft (1.8 m), 2 ft (0.6 m), and 1 ft (0.3 m) from the right flume wall at locations aligned with each downstream pier face. To accurately collect the velocity data for the desired loca- tions, the ADV probe tip was set at an X-direction offset of 0.27 ft (0.08 m). This offset adjusted the X-location of where the point gauge data were collected. Table 3.5 identifies the location and description of hydraulic data collected during each test. A schematic of the hydraulic data collection loca- tions taken in the flume is shown in Figure 3.35. A higher res- olution schematic of the hydraulic data collection location taken in the vicinity of a pier is shown in Figure 3.36. Time-Dependent Scour Measurement. Four depth trans- ducers were mounted around a pier and/or debris cluster to track the depth of scour in real time throughout the test. Depth 67 Data Acquisition Point Identification # X (ft) Y (ft) Data Collected 1 31.94 6 Velocity @ 60% Depth 2 through 5 31.94 4 Velocity @ 20%, 40%, 60%, 80% Depth 6 31.94 2 Velocity @ 60% Depth 7 37.94 6 Velocity @ 60% Depth 8 37.94 4 Velocity @ 60% Depth 9 37.94 2 Velocity @ 60% Depth 10 42.60 7 Velocity @ 60% Depth 11 through 14 42.60 6 Velocity @ 20%, 40%, 60%, 80% Depth 15 through 18 42.60 2 Velocity @ 20%, 40%, 60%, 80% Depth 19 42.60 1 Velocity @ 60% Depth 20 50.60 6 Velocity @ 60% Depth 21 50.60 4 Velocity @ 60% Depth Pi er A 22 50.60 2 Velocity @ 60% Depth 23 63.94 6 Velocity @ 60% Depth 24 through 27 63.94 4 Velocity @ 20%, 40%, 60%, 80% Depth 28 63.94 2 Velocity @ 60% Depth 29 69.94 6 Velocity @ 60% Depth 30 69.94 4 Velocity @ 60% Depth 31 69.94 2 Velocity @ 60% Depth 32 74.60 7 Velocity @ 60% Depth 33 through 36 74.60 6 Velocity @ 20%, 40%, 60%, 80% Depth 37 through 40 74.60 2 Velocity @ 20%, 40%, 60%, 80% Depth 41 74.60 1 Velocity @ 60% Depth 42 82.60 6 Velocity @ 60% Depth 43 82.60 4 Velocity @ 60% Depth Pi er B 44 82.60 2 Velocity @ 60% Depth 45 95.94 6 Velocity @ 60% Depth 46 through 49 95.94 4 Velocity @ 20%, 40%, 60%, 80% Depth 50 95.94 2 Velocity @ 60% Depth 51 101.94 6 Velocity @ 60% Depth 52 101.94 4 Velocity @ 60% Depth 53 101.94 2 Velocity @ 60% Depth 54 106.60 7 Velocity @ 60% Depth 55 through 58 106.60 6 Velocity @ 20%, 40%, 60%, 80% Depth 59 through 62 106.60 2 Velocity @ 20%, 40%, 60%, 80% Depth 63 106.60 1 Velocity @ 60% Depth 64 114.60 6 Velocity @ 60% Depth 65 114.60 4 Velocity @ 60% Depth Pi er C 66 114.60 2 Velocity @ 60% Depth 67 127.94 6 Velocity @ 60% Depth 68 through 71 127.94 4 Velocity @ 20%, 40%, 60%, 80% Depth 72 127.94 2 Velocity @ 60% Depth 73 133.94 6 Velocity @ 60% Depth 74 133.94 4 Velocity @ 60% Depth 75 133.94 2 Velocity @ 60% Depth 76 138.60 7 Velocity @ 60% Depth 77 through 80 138.60 6 Velocity @ 20%, 40%, 60%, 80% Depth 81 through 84 138.60 2 Velocity @ 20%, 40%, 60%, 80% Depth 85 138.60 1 Velocity @ 60% Depth 86 146.60 6 Velocity @ 60% Depth 87 146.60 4 Velocity @ 60% Depth Pi er D 88 146.60 2 Velocity @ 60% Depth Table 3.5. Hydraulic data acquisition locations and descriptions.

P6 1 2 3 4 5 9 8 7 15 16 17 18 11 12 13 14 22 21 20 19 10 37 38 39 40 33 34 35 36 28 23 31 30 29 44 43 42 41 32 50 46 47 48 49 45 53 52 51 59 60 61 62 55 56 57 5854 66 65 64 63 72 67 75 74 73 81 82 83 84 77 78 79 80 88 87 86 P P FLOW Pier DP Pier C Pier B Pier A P P P P P P P P P P Water Surface and Velocity at 20%, 40%, 60%, and 80% of Depth Water Surface and Velocity at 60% of Depth Pier 24 25 26 27 68 69 70 71 85 76 Figure 3.35. Schematic of data collection locations in 8 ft (2.4 m) flume.

transducers, Figure 3.37, collected a depth measurement through the transmission of a sonar pulse from the bottom of the transducer to the surface of the sand bed every 30 sec- onds. The data were sent to a data logger and then saved to a computer data file. For baseline tests, where no debris cluster was mounted at the pier, the first depth transducer was mounted at the nose of the pier on the upstream face, the sec- ond depth transducer was mounted on the left upstream cor- ner of the pier, the third depth transducer was mounted at the tail of the pier on the downstream face, and the fourth depth transducer was mounted on the right upstream corner of the pier. A schematic of the baseline depth transducer layout is shown below in Figure 3.38. When a debris cluster was mounted such that it was sus- pended above the bed, the first transducer was placed upstream of the pier at a distance half the length (L/2) of the debris clus- ter, the second transducer was placed 1 in. (2.5 cm) from the upstream face of the pier, the third transducer was placed approximately 1 in. (2.5 cm) away from the right side of the pier and approximately flush to the upstream face, and the fourth transducer was placed directly downstream of the right edge of 69 P Water Surface and Velocity at 20%, 40%, 60%, and 80% of Depth Water Surface and Velocity at 60% of Depth Pier P P P 8 ft 6 ft4 ft Aligned with back of th e pie r Pier FLOW 2 ft 4 ft 1 ft #1 #2: 20% #3: 40% #4: 60% #5: 80% #6 #7 #15: 20% #16: 40% #17: 60% #18: 80% #8 #9 #11: 20% #12: 40% #13: 60% #14: 80% #10 #19 #20 #21 #22 Figure 3.36. Data collection location in 8 ft (2.4 m) flume. Figure 3.37. Transducers used for automated scour hole mapping. Transducers, located 1 inch from pier face Pier FLOW 1st of 4 2nd of 4 3rd of 4 4th of 4 Figure 3.38. Schematic of depth transducers around baseline piers during testing.

the debris cluster. A schematic of the transducer orientation around a suspended debris cluster at a pier is presented in Figure 3.39a. For debris clusters that were resting on the bed, the first transducer was placed approximately 1 in. (2.5 cm) away from the left side of the pier and approximately flush to the upstream face, the second transducer was placed approximately 1 in. (2.5 cm) away from the right side of the pier and approximately flush to the upstream face, the third transducer was placed directly downstream of the right edge of the debris cluster, and the fourth transducer was placed half the distance between the right corner of the debris cluster and the right flume wall. A schematic diagram of the transducer locations when the debris cluster was resting on the bed is presented in Figure 3.39b. Post-test Bed Survey and Mapping. Post-test bed sur- veys were completed using Seatek® 2 MHz Ultrasonic depth transducers, a potentiometer, and a motorized cart. When necessary, survey measurements were verified manually with a point gauge. Depth transducers were mounted on a 4 ft (1.2 m) wide frame, spaced 3 in. (7.5 cm) apart, and fixed to a rail on the data acquisition cart. A voltage-regulated potentiometer was also mounted on the data collection cart to measure the distance (X) along the flume of each depth measurement. Measurements from the depth transducers and the voltage- regulated potentiometer were recorded electronically by a data logger that was connected to a laptop computer that saved the raw data to a data file. Figure 3.40a shows the potentiome- ter cabled system, and Figure 3.40b shows the transducers mounted on the frame and cart. Around each pier, depth transducers collected post-test bed data at a small change in station distance and the poten- tiometer would then measure the station distance for the location of each set of data. Because the depth transducer frame was only 4 ft (1.2 m) wide, two passes along the flume were necessary to collect a flume-wide post-test bed survey. Figure 3.41 presents a schematic of the ultrasonic depth transducers, potentiometer, and frame setup. 3.5.2 Debris Cluster Test Materials All debris clusters were constructed with a 0.25 in. rolled steel frame; a 0.25 in. wire mesh was then placed around the 70 Flume wall Pier Debris Cluster L/2 L/2 Flow a. Debris suspended above the bed. Flume wall Pier Debris Cluster y/2 y/2 Flow b. Debris located on the bed. Figure 3.39. Schematic diagram showing locations of depth transducers around piers and debris during testing.

71 b. Transducer array used for automated scour hole mapping. a. Data collection cable system for establishing x coordinate. Figure 3.40. Post-test bed and scour hole mapping system. Left Flume Wall Right Flume Wall Potentiometer Data Logger 0.24 ft Sand BedPier Scour Hole Cart Rail Transducer Bracket 16 Transducers Figure 3.41. Depth transducers, potentiometer, and data logger setup for post-test mapping. steel frame. Figure 3.42 shows the basic framework and wire mesh common to all the debris clusters used in the testing program. For tests that modeled impermeable debris, the debris clusters were completely packed with 1 ft square sheets of woven geotextile (Figure 3.43). Because the intensity of flow vortices, separation, and tur- bulence may depend on the surface roughness of the debris pile, variations of the surface roughness were included to test the sensitivity of scour to this variable. For tests that simu- lated debris with increased surface roughness, the debris clus- ters were artificially roughened by placing 0.25 in. (0.6 cm) diameter wood dowels around the surface of the cluster. Each dowel projected outward from the surface of the cluster a dis- tance of 2 in. (5 cm) and was oriented orthogonally to the sur- face of the cluster. The average density of the dowels was one per every 4 in.2 (25.8 cm2) of surface area. An example of a roughened debris cluster is shown in Figure 3.44. To determine if porosity affects debris scour, selected debris clusters were filled with wooden dowels instead of the tightly packed geotextile. The dowels ranged in diameter from 0.25 to

Figure 3.45 illustrates the tank and an example debris cluster being tested to determine the porosity. 3.5.3 Baseline Tests Bed Material and Incipient Motion The sand comprising the bed material in the test flume was characterized by a d50 grain size of approximately 0.70 mm. The coefficient of uniformity, Cu, defined as d60/d10, was 3.14. A representative grain size distribution graph is shown in Figure 3.46. The critical velocity, Vc, for incipient motion of the bed material was estimated using the method described in HEC-18 (Richardson and Davis 2001): where: Vc = Critical velocity for the initiation of motion, m/s Ks = Shields parameter (dimensionless) Ss = Specific gravity of sediment (dimensionless) d50 = Median particle diameter, ft (m) y = Flow depth, ft (m) n = Manning’s resistance coefficient Assuming that Manning’s n value can be estimated from Strickler’s equation as n = 0.041(d50)1/6 where d50 is in meters, and using a range of the Shields parameter from 0.039 to V K S d y n c s s = −( )1 2 1 2 501 2 1 61 3 1( . ) 72 a. Woven geotextile used in filing debris clusters. b. Wedge-shaped, impermeable debris cluster after being filled with geotextile. Figure 3.43. Impermeable debris clusters. Figure 3.42. Wire framework and mesh for debris cluster construction. 2 in. (0.6 to 5 cm) and lengths ranging from 2 to 12 in. (5 to 30 cm). The dowels were placed into each debris cluster in a random orientation to achieve the overall desired porosity of 25% of the gross volume. To verify that this specification was achieved, Archimedes’s principle was applied. A water tank large enough to contain all the various debris cluster sizes and shapes was filled with water, and each debris cluster was then submerged. The displaced water was measured and divided by the gross volume of the cluster to confirm the porosity.

73 Figure 3.44. Typical photograph showing 2 in. (5 cm) wood dowels used to simulate roughness. Figure 3.45. Determining porosity of debris cluster. #60200140100201610 40 70840.5-in1-in2-in 0 10 20 30 40 50 60 70 80 90 100 0.00010.0010.010.1110100 Grain Size in millimeters Pe rc en t F in er b y w ei gh t - % Sieve Size GRAVEL Coarse Fine SAND Coarse Medium Fine SILT CLAY (ASTM) Figure 3.46. Grain size distribution of bed material.

0.047 for the initiation of movement, the critical velocity is estimated to range from 1.5 to 1.6 ft/s (0.45 to 0.49 m/s). From this analysis, a conservative value of 1.4 ft/s (0.43 m/s) was selected for establishing the target approach velocities. The intent was to create a condition for the initial run that resulted in true clear-water conditions, with no movement of the bed material except for local scour in the immediate vicin- ity of the piers. Tests confirmed that an approach velocity of 1.4 ft/s (0.43 m/s) resulted in no bed material movement except for local scour. Predicted Scour CSU Equation. The ultimate (equilibrium) depth of local scour at a pier for clear-water conditions can be estimated using the CSU equation as presented in HEC-18 (Richardson and Davis 2001): where: ys = Depth of scour, ft (m) y1 = Flow depth directly upstream of the pier, ft (m) K1 = Correction factor for pier nose shape K2 = Correction factor for angle of attack of flow K3 = Correction factor for bed condition K4 = Correction factor for armoring by bed material size a = Pier width, ft (m) Fr1 = Froude number directly upstream of the pier For a square nose pier, K1 = 1.1, and for a group of cylinders, K1 = 1.0; for an angle of attack equal to zero, K2 = 1. For plane bed clear-water scour, K3 = 1.1, and for no armoring of the bed material, K4 = 1.0. Pier widths were 4 in. (10.2 cm) for the square piers, 1 in. (2.5 cm) for the rectangular wall piers, and 0.5 in. (1.3 cm) for the multiple-column cylindrical piers. The Froude number of the approach flow was deter- mined by: where: V1 = Mean velocity of flow directly upstream of the pier, ft/s (m/s) g = Acceleration of gravity, ft/s2 (m/s2) Fr V gy 1 1 1 3 3= ( . ) y a K K K K y a Frs = ⎛⎝⎜ ⎞⎠⎟2 0 3 21 2 3 4 1 0 35 1 0 43. ( . ) . . The approach flow depth directly upstream of the pier was established at a depth of 1 ft (0.305 m) for all tests. The approach velocity, V1, was found to be approximately 1.2 times the cross-sectional average velocity due to the location of the piers along the centerline of the flume. Therefore, using a tar- get cross-sectional average velocity of 1.4 ft/s (0.43 m/s), the computed Froude number was 0.30. Sheppard Equation. Sheppard et al. (2004) present an alternative method for predicting the ultimate depth of clear- water scour at a pier: The functions f1, f2, and f3 are defined as: where: ds = Equilibrium scour depth, ft (m) D = Effective pier width, ft (m) where D = 1.0 times the diameter of a circular pier or 1.23 times the width of a square pier y0 = Depth of approach flow, ft (m) V = Velocity of approach flow, ft/s (m/s) Vc = Critical velocity of bed material, ft/s (m/s) D50 = Median particle size, ft (m) The computed values of local scour at a baseline pier (free from debris) using the CSU and Sheppard equations under the conditions described above are presented in Table 3.6. Contraction Scour. A modified version of Laursen’s equa- tion, as presented in HEC-18 (Richardson and Davis 2001), was used to determine depth of scour in a contracted section under clear-water conditions. f D D D D D D D D 3 50 50 50 1 2 50 4 10 6 ∗⎛ ⎝⎜ ⎞ ⎠⎟ = ∗( ) ∗( ) + ∗. .. 0 0 13 3 7( )− . ( . ) f V V V Vc c 2 2 1 1 75 3 6 ⎛ ⎝⎜ ⎞ ⎠⎟ = − ⎛ ⎝⎜ ⎞ ⎠⎟ ⎡ ⎣⎢ ⎤ ⎦⎥. ln ( . ) f y D y D 1 0 0 0 4 3 5 ∗ ⎛⎝⎜ ⎞⎠⎟ = ∗⎛⎝⎜ ⎞⎠⎟ ⎡ ⎣⎢ ⎤ ⎦⎥ tanh ( . ) . d D f y D f V V f D D s c = ∗ ∗ ⎛⎝⎜ ⎞⎠⎟ ⎛ ⎝⎜ ⎞ ⎠⎟ ∗⎛ ⎝⎜ ⎞ ⎠2 5 1 0 2 3 50. ⎟ ( . )3 4 74 Pier Size and Type CSU Equation Sheppard Equation 4" square pier 0.71 ft (0.22 m) 0.79 ft (0.24 m) 1" rectangular wall pier 0.29 ft (0.09 m) 0.25 ft (0.08 m) 0.5" cylindrical columns 0.17 ft (0.05 m) 0.09 ft (0.03 m) Table 3.6. Computed local scour depths for baseline conditions.

of baseline scour for all pier shapes. Figure 3.47 shows the results for square piers after 72 hours of testing; the upper segment of the pier has been removed for data collection purposes and ambient bed elevation is represented by the top of the lower segment of the pier. Figure 3.48 shows the results for multiple-column and wall piers. Arrows indicate direction of flow on the photographs. Table 3.7 provides the results (maximum scour at the pier only), details on the run number, hydraulic conditions, dura- tion, and location of the baseline tests. Note that all scour depth measurements in Table 3.7 were taken at the nose of the pier. 75 Figure 3.47. Square pier after 72-hour test with no debris (flow from right to left). a. Multiple 0.5" column pier after 72-hour test with no debris b. 1" wall pier after 72-hour test with no debris Figure 3.48. Baseline tests of multiple-column and wall piers. ys = y2 − yo = average contraction scour depth, ft (m) where: y2 = Average equilibrium depth in the contracted section after contraction scour, ft (m) Q = Flow discharge through the contracted section, ft3/s (m3/s) dm = Diameter of the smallest non-transportable particle in the bed material in the contracted section computed from dm (m) = (1.25  d50) W = Bottom width of the contracted section less pier widths, ft (m) Ku = Equal to 0.025 for SI units (0.0077 for English units) yo = Existing depth in the contracted section before scour, ft (m) For the square piers, using a total discharge Q = 11.2 ft3/s (0.32 m3/s) from continuity, W = width of the flume [8 ft (2.4 m)] less the width of the pier, and d50 = 0.7 mm (0.0023 ft), Equation 3.8 predicts that no contraction scour was anticipated for any of the piers under baseline (no-debris) conditions. Testing Baseline tests were designed as control tests to provide scour data for piers under clear-water conditions. No debris was affixed to the piers, allowing for determination y K Q d w u m 2 2 2 3 7 2 3 3 8= ⎡ ⎣⎢ ⎤ ⎦⎥ ( . )

3.5.4 Tests with Debris The goal of the laboratory plan was to develop a series of tests for a wide range of debris configurations that could be run quickly and efficiently. Tests were performed for single debris clusters at individual piers. The literature review, survey results, photographic archive, and site reconnaissance were used to determine the number of alternatives that should be included to investigate a representative range of bridge-specific debris characteristics. Initial runs considered various debris geomet- ric characteristics (see Section 3.4.2). These tests were then sup- plemented to address issues of porosity and roughness. All test runs were performed for clear-water sediment trans- port conditions (approach flow velocity less than the critical velocity required to initiate sediment transport). Typically, the test duration was 8 hours after steady flow conditions were established. Selected pier shapes and debris cluster types were run for 72 hours. Rectangular Debris Nine geometrically unique rectangular (in planform and pro- file) debris shapes were tested. Variations in testing included location within the water column, pier type, roughness, and porosity for a total of 39 tests. Five of these tests were con- ducted at a nominal velocity of 0.7 Vc; the remaining 34 tests were conducted at 1.0 Vc. Figure 3.49 provides a definition sketch of the rectangular debris geometry and associated variables. Table 3.8 shows the tests performed at 4 in. by 4 in. (10.2 cm by 10.2 cm) square piers, while Table 3.9 shows the tests per- formed at 1 in. (2.5 cm) rectangular wall piers (no skew) and 0.5 in. (1.3 cm) cylindrical column piers (no skew). Each letter in the table denotes a specific test and the variable examined. Figure 3.50 shows a 1 ft wide by 1 ft long by 1 ft high rec- tangular debris configuration incorporating roughness and porosity prior to Test 009_01C. Figure 3.51 shows the results from the debris configuration presented in the previous fig- ure after 72 hours of testing at 1 Vc. The upper segment of the pier has been removed for data collection purposes, and ambient bed elevation is represented by the top of the lower segment of the pier. Figure 3.52 shows a 4 ft wide by 3 ft long by 8 in. high debris configuration incorporating roughness and porosity prior to Test 004_03C. Figure 3.53 shows the results from the debris configuration presented in the previous figure after 8 hours of testing at 0.7 Vc. The upper segment of the pier has been removed for data collection purposes, and ambient bed eleva- tion is represented by the top of the lower segment of the pier. Figure 3.54 shows a 2 ft wide by 1 ft high, wedge-shaped debris cluster before Test 007_02B, incorporating roughness and porosity. Figure 3.55 shows the results from the debris configuration presented in the previous figure after 8 hours of testing at 1 Vc; the upper segment of the pier has been removed for data collection purposes and ambient bed eleva- tion is represented by the top of the lower segment of the pier. 76 Run No. Pier Nominal V/Vc Duration (h) Maximum Measured Scour Depth (ft) Pier Shape 003_01 A 0.7 72 0.67 4" square 003_01 D 0.7 72 0.41 4" square 004_01 A 0.7 8 0.49 4" square 004_01 B 0.7 8 0.38 4" square 004_01 C 0.7 8 0.34 4" square 004_01 D 0.7 8 0.34 4" square 003_02 A 1.0 72 1.02 4" square 003_02 D 1.0 72 0.86 4" square 004_02 A 1.0 8 0.85 4" square 004_02 B 1.0 8 0.70 4" square 004_02 C 1.0 8 0.68 4" square 004_02 D 1.0 8 0.70 4" square 009_01 B 1.0 72 0.80 4" square 003_01 C 0.7 72 0.16 1" rectangular wall 003_02 C 1.0 72 0.26 1" rectangular wall 008_01 A 1.0 8 0.17 1" rectangular wall 003_01 B 0.7 72 0.10 0.5" cylindrical columns 003_02 B 1.0 72 0.10 0.5" cylindrical columns 008_01 B 1.0 8 0.11 0.5" cylindrical columns Table 3.7. Maximum measured scour, baseline tests.

77 LOOKING DOWNSTREAM H d Pier width = a W PROFILE H d Pier length = a L Flow Figure 3.49. Rectangular debris geometry and associated variables. PositionDebris Width (ft) Debris Length (ft) Debris Height (ft) W/a L/a H/d Surface Mid-depth Bed 1.0 1.0 1.0 3 3 3/3 I, P, R, C, C(72) 2.0 1.0 0.33 6 3 1/3 I, P, R, C(72) 2.0 1.0 0.67 6 3 2/3 I, P, R 2.0 2.0 0.33 6 6 1/3 I, P, C(72), P(72) I C I I I 2.0 2.0 0.67 6 6 2/3 I I, P, R 2.0 3.0 0.67 6 9 2/3 C 4.0 3.0 0.67 12 9 2/3 I, C 6.0 4.0 0.67 18 12 2/3 I, I(72) Wedge Shape Variant, Rectangular in Plan, Triangular in Section 2.0 1.0 0.67 6 3 2/3 C, C(72) W = Debris width I = Impermeable fill, no roughness L = Debris length P = Porosity tested independently, no roughness H = Debris height R = Roughness tested independently, no porosity a = Pier dimension C = Porosity and roughness tested together (combined) d = Depth of water (72) = Test ran for 72 hours Table 3.8. Square piers with rectangular debris. PositionDebris Width (ft) Debris Length (ft) Debris Height (ft) W/a L/a H/d Surface Mid-depth Bed 1.0 1.0 1.0 12 12 3/3 C (1" wall pier) 1.0 1.0 0.67 24 24 2/3 C (0.5" columns) 4.0 3.0 0.67 96 72 2/3 C (0.5" columns) Wedge Shape Variant, Rectangular in Plan, Triangular in Section 2.0 1.0 0.67 48 24 2/3 C (0.5" columns) W = Debris width a = Pier dimension C = Porosity and roughness tested together (combined) L = Debris length d = Depth of water H = Debris height Table 3.9. Wall and multiple-column piers with rectangular debris.

Table 3.10 provides the results (maximum scour at the pier only) of tests performed with rectangular debris config- urations. Most scour depth measurements in Table 3.10 were taken at the pier nose. When nose scour measurements were not possible due to location of the debris in the vicinity of the bed, data was collected at the upstream corner of the pier; this is denoted by the use of a “c” after the test number. Triangular/Conical Debris Seven geometrically unique conical debris shapes (triangular in planform) were tested. Variations in testing included loca- tion within the water column, pier type, roughness, and poros- ity for a total of 14 tests. Three of these tests were conducted at a nominal velocity of 0.7 Vc; the remaining 11 tests were con- ducted at 1.0 Vc. Figure 3.56 provides a definition sketch of the triangular/conical debris geometry and associated variables. Table 3.11 shows the tests performed at 4 in. by 4 in. (10.2 cm by 10.2 cm) square piers. Table 3.12 shows the tests performed at multiple-column and wall piers. Each letter in the table denotes a specific test and the variable examined. Figure 3.57 shows a 4 ft wide by 3 ft long by 1 ft high triangular/conical debris configuration incorporating rough- ness and porosity before testing. Figure 3.58 shows the results 78 Figure 3.50. Rectangular debris cluster before Test 009_01C. Ambient bed elevation is represented by the top of the pier. Upper segment of the pier and the debris configuration have been removed for data collection purposes. Figure 3.51. Scour hole resulting from Test 009_01C after 72 hours of testing at 1 Vc. Figure 3.52. 4‘W  3‘L  8‘‘H debris cluster located at the water surface during Test 004_03C. Ambient bed elevation is represented by the top of the pier. Upper segment of the pier and the debris configuration have been removed for data collection purposes. Note that the deepest scour occurred upstream and away from the pier face. Figure 3.53. Scour hole resulting from Test 004_01C after 8 hours of testing at 0.7 Vc.

from the debris configuration presented in the previous fig- ure after 8 hours of testing at 1 Vc; the upper segment of the pier has been removed for data collection purposes and ambi- ent bed elevation is represented by the top of the lower seg- ment of the pier. Table 3.13 provides the results of tests performed with tri- angular/conical debris configurations at a pier. Most scour depth measurements in Table 3.13 were taken at the pier nose. When nose scour measurements were not possible due to location of the debris in the vicinity of the bed, data was collected at the upstream pier corner; this is denoted by a “c” after the test number. Figure 3.59 shows a 2 ft wide ×3 ft long ×8 in. high, impermeable, buried wedge debris configuration before Test 003_03D. Figure 3.60 shows the results from the debris configuration presented in the previous figure after 8 hours of testing at 0.7 Vc and an additional 8 hours of testing at 1 Vc; the upper segment of the pier has been removed for data col- lection purposes, and ambient bed elevation is represented by the top of the lower segment of the pier. 3.6 Appraisal of Testing Results 3.6.1 Baseline (No-Debris) Tests Predicted vs. Measured Scour A total of 19 baseline tests were run under clear-water scour conditions. Thirteen baseline tests were conducted using 4 in. (10.2 cm) square piers. Six of the square pier tests were con- ducted at a nominal (target) approach velocity of 0.7 Vcrit, where Vcrit was estimated to be 1.4 ft/s (0.43 m/s) for the initiation of motion of the bed material. The remaining seven square pier tests were conducted at a nominal velocity of 1.0 Vcrit. Three baseline tests were conducted using 1 in. by 8 in. (2.5 cm by 20 cm) rectangular wall piers with no skew angle. One of those tests was conducted at a nominal velocity of 0.7 Vcrit, and the other two tests were run at 1.0 Vcrit. Three baseline tests were conducted using piers having multiple 0.5 in. (1.25 cm) cylindrical columns with no skew angle. One of those tests was conducted at a nominal velocity of 0.7 Vcrit, and the other two tests were run at 1.0 Vcrit. 79 Figure 3.54. Wedge-shaped debris cluster before Test 007_02B. Ambient bed elevation is represented by the top of the pier. Upper segment of the pier and the debris configuration have been removed for data collection purposes. Note that the deepest scour is coincident with the pier face. Figure 3.55. Scour hole resulting from Test 007_02B after 8 hours of testing at 1 Vc.

80 Run No. Pier V/Vc Duration (h) Max Measured Scour Depth ds(ft) Transducer Location Debris Shape Width (ft) Length (ft) Height (ft) Debris Location Pier Type Debris Characteristic 005_02C-c C 1.0 8 0.42 at corner cube 1 1 1 full depth square impermeable 006_01C-c C 1.0 8 0.47 at corner cube 1 1 1 full depth square roughness 006_02C-c C 1.0 8 0.47 at corner cube 1 1 1 full depth square porosity 008_01C-c C 1.0 8 0.51 at corner cube 1 1 1 cube wall roughness and porosity 009_01D-c D 1.0 72 0.75 at corner cube 1 1 1 cube square roughness and porosity 005_02B B 1.0 8 0.92 at nose rectangle 2 1 0.33 surface square impermeable 006_01B B 1.0 8 0.91 at nose rectangle 2 1 0.33 surface square roughness 006_02A B 1.0 8 1.11 at nose rectangle 2 1 0.67 surface square porosity 009_02C C 1.0 72 1.06 at nose rectangle 2 1 0.33 surface square roughness and porosity 005_01A-c A 1.0 8 0.81 at corner rectangle 2 1 0.67 bed square impermeable 005_02A A 1.0 8 1.21 at nose rectangle 2 1 0.67 surface square impermeable 006_01A A 1.0 8 1.16 at nose rectangle 2 1 0.67 surface square roughness 006_02B A 1.0 8 1.00 at nose rectangle 2 1 0.33 surface square porosity 003_03A-c A 0.7 8 0.46 at corner rectangle 2 2 0.33 surface square impermeable 003_04A-c A 1.0 8 0.75 at corner rectangle 2 2 0.33 surface square impermeable 004_03A A 0.7 8 0.42 at nose rectangle 2 2 0.33 mid-depth square impermeable 004_04A A 1.0 8 0.70 at nose rectangle 2 2 0.33 mid-depth square impermeable 005_01B-c B 1.0 8 0.35 at corner rectangle 2 2 0.33 bed square impermeable 009_01A A 1.0 72 1.14 at nose rectangle 2 2 0.33 surface square roughness and porosity 009_02A A 1.0 72 1.13 at nose rectangle 2 2 0.33 surface square porosity 004_03B B 0.7 8 0.40 at nose rectangle 2 2 0.67 surface square impermeable 004_04B B 1.0 8 0.66 at nose rectangle 2 2 0.67 surface square impermeable 005_01D-c D 1.0 8 0.35 at corner rectangle 2 2 0.67 bed square impermeable 005_02D-c D 1.0 8 0.55 at corner rectangle 2 2 0.67 mid-depth square impermeable 006_01D-c D 1.0 8 0.42 at corner rectangle 2 2 0.67 mid-depth square roughness 008_01D D 1.0 8 0.30 at nose rectangle 2 2 0.67 surface multiple column roughness and porosity 007_01B B 1.0 8 0.59 at nose rectangle 2 3 0.67 surface square roughness and porosity 004_03C C 0.7 8 0.49 at nose rectangle 4 3 0.67 surface square impermeable 004_04C C 1.0 8 0.74 at nose rectangle 4 3 0.67 surface square impermeable 007_01D D 1.0 8 0.85 at nose rectangle 4 3 0.67 surface square roughness and porosity 007_02bD D 1.0 8 0.91 at nose rectangle 4 3 0.67 mid-depth square roughness and porosity 008_02D D 1.0 8 0.72 at nose rectangle 4 3 0.67 mid-depth multiple column roughness and porosity 004_03D D 0.7 8 0.58 at nose rectangle 6 4 0.67 surface square impermeable 004_04D D 1.0 8 0.95 at nose rectangle 6 4 0.67 surface square impermeable 009_02D D 1.0 72 1.23 at nose rectangle 6 4 0.67 surface square impermeable 007_02bB B 1.0 8 0.92 at nose wedge 4 6 0.67 surface square roughness and porosity 008_02B B 1.0 8 0.45 at nose wedge 2 1 0.67 surface multiple column roughness and porosity 000_01C C 1.0 72 1.08 at nose wedge 2 1 0.67 surface square roughness and porosity Table 3.10. Results of Rectangular Debris Tests.

81 LOOKING DOWNSTREAM H d Pier width = a W PROFILE H d Pier length = a L Flow Figure 3.56. Triangular/conical debris geometry and associated variables. Position and Shape Debris Width (ft) Debris Length (ft) Debris Height (ft) W/a L/a H/d Surface Bed Bed (partially buried) H2/H1 = 1 Bed (partially buried) H2/H1 = 0.5 2.0 2.0 0.67 6 6 2/3 C 2.0 3.0 0.33 6 9 1/3 I 2.0 3.0 0.67 6 9 2/3 I I 2.0 3.0 1.0 6 9 3/3 C 4.0 3.0 0.67 12 9 2/3 I 4.0 3.0 1.0 12 9 3/3 C, C(72) 4.0 6.0 0.67 12 18 2/3 C W = Debris width a = Pier dimension I = Impermeable fill, no roughness L = Debris length d = Depth of water C = Porosity and roughness tested together (combined) H = Debris height (72) = Test ran for 72 h Table 3.11. Square piers with triangular/conical debris. Position and Shape Debris Width (ft) Debris Length (ft) Debris Height (ft) W/a L/a H/d Surface Bed Bed (partially buried) H2/H1 = 1 Bed (partially buried) H2/H1 = 0.5 2.0 3.0 0.67 24 36 3/3 C (1" wall pier) 4.0 3.0 0.33 48 36 3/3 C (1" wall pier) W = Debris width a = Pier dimension C = Porosity and roughness tested together (combined)L = Debris length d = Depth of water H = Debris height Table 3.12. Wall and multiple-column piers with triangular/conical debris.

82 Figure 3.57. Triangular/conical debris cluster before Test 007_02A, mounted such that the top surface of the debris was located at the water surface. Ambient bed elevation is represented by the top of the pier. Upper segment of the pier and the debris configuration have been removed for data collection purposes. Figure 3.58. Scour hole resulting from Test 007_02A after 8 hours of testing at 1.0 Vc. Run No. Pier V/Vc Duration (h) Max Measured Scour Depth ds(ft) Transducer Location Debris Shape Width (ft) Length (ft) Height (ft) Debris Location Pier Type Debris Characteristic 007_01A A 1.0 8 0.86 at nose conical 2 2 0.67 surface square roughness and porosity 003_03B-c B 0.7 8 0.41 at right corner conical 2 3 0.33 surface square impermeable 003_04B-c B 1.0 8 0.72 at right corner conical 2 3 0.33 surface square impermeable 003_03D-c D 0.7 8 0.41 at right corner conical 2 3 0.67 buried square impermeable 003_04D-c D 1.0 8 0.69 at right corner conical 2 3 0.67 buried square impermeable 005_01C-c C 1.0 8 0.19 at right corner conical 2 3 0.67 bed square impermeable 007_02bC C 1.0 8 0.80 at nose conical 2 3 1 surface square roughness and porosity 008_02C C 1.0 8 0.37 at nose conical 2 3 1 surface wall roughness and porosity 007_02bA A 1.0 8 1.15 at nose conical 4 3 1 surface square roughness and porosity 008_02A A 1.0 8 0.54 at nose conical 4 3 1 surface wall roughness and porosity 009_02B B 1.0 72 1.19 at nose conical 4 3 1 surface square roughness and porosity 007_01C C 1.0 8 0.77 at nose conical 4 6 0.67 surface square roughness and porosity 003_03C-c C 0.7 8 0.09 at right corner conical 2 3 0.67 surface square H2/H1-0.5 003_04C-c C 1.0 8 0.21 at right corner conical 2 3 0.67 surface square H2/H1-0.6 Table 3.13. Results of triangular/conical debris tests.

where: Kt = Fraction of ultimate scour reached at time t (dimen- sionless) t = Elapsed time from start of scour, days te = Time to ultimate (equilibrium) scour, days Vc = Critical velocity for the initiation of motion of bed sediment, ft/s (m/s) V = Approach velocity upstream of pier, ft/s (m/s) and the time to ultimate scour is given by: where: a = Pier width or diameter, ft (m) y = Depth of approach flow, ft (m) te = Time to ultimate (equilibrium) scour, days 72-Hour Tests. The analysis of the time dependency of scour in the laboratory flume indicated that, for all tests, the 72-hour duration runs resulted in 98% to 100% of the expected ultimate scour depth at the pier face. Therefore, observed scour was adjusted minimally for all 72-hour tests, regardless of pier type. 8-Hour Tests. In contrast to the 72-hour tests, the 8-hour tests resulted in measured scour depths that were significantly less than the estimated ultimate scour. For the 4 in. square piers, the 8-hour tests at 0.7 Vcrit ranged from 77% to 81% of ultimate scour, while the tests at 1.0 Vcrit ranged from 84% to 87% of ulti- mate scour. The slimmer wall piers and multiple-column piers reached about 93% to 96% of ultimate scour after 8 hours. Table 3.14 summarizes the results of the baseline (no-debris) tests. In this table, the actual approach velocity as measured by ADV and the measured approach flow depth are shown, because both velocity and depth during any particular test were typically slightly higher or lower than the target conditions. The table also shows the measured scour depths at the end of each test, the ultimate scour depth estimated by the Melville method, and the ultimate scour depth predicted by the CSU equation. Figure 3.61 presents the results of the scour predictions vs. equilibrium scour in graphical form. For the runs conducted at a target velocity of 1.0 Vcrit, the ratio of clear-water equilibrium scour depth, Yse, to pier width, a, was slightly different depending on pier type, as follows: 4 in. square piers (7 tests): 2.4 < < Y a se 2 6. t a V V V y a e c = − ⎛ ⎝⎜ ⎞ ⎠⎟ ⎛⎝⎜ ⎞⎠⎟30 89 0 4 3 10 0 25 . . ( . ) . K V V t t t c e = − ⎛ ⎝⎜ ⎞ ⎠⎟ ⎧⎨⎪⎩⎪ ⎫⎬⎪⎭⎪ exp . ln ( . ) . 0 03 3 9 1 6 83 Figure 3.59. Buried wedge debris cluster placed on the bed showing initiation of flow during test. Ambient bed elevation is represented by the top of the pier. Upper segment of the pier and the debris configuration have been removed for data collection purposes. Figure 3.60. Scour hole resulting from Test 007_02A after 8 hours of testing at 0.7 Vc and an additional 8 hours of testing at 1.0 Vc. Most tests were run for 8 hours. However, selected tests were conducted with a duration of 72 hours. Because equa- tions for assessing pier scour are intended to reflect ultimate (equilibrium) conditions, the observed scour depths in the laboratory flume were adjusted to reflect estimated equilib- rium scour. The method provided in Melville and Coleman (2000) for estimating the time dependency of clear-water pier scour was used to develop a best estimate of the equilibrium condition, as follows:

84 Velocity ratio V/Vc Run No. Pier Target Meas. Meas. Approach Depth (ft) Test Duration (h) Measured Scour1 (ft) Estimated Ultimate Scour (ft) Predicted Scour, CSU Equation (ft) 4 in. (10 cm) Square Piers 003_01 A 0.7 0.81 1.00 72 0.54 0.55 0.59 003_01 D 0.7 0.65 1.05 72 0.41 0.41 0.54 003_02 A 1.0 1.14 1.03 72 0.83 0.84 0.69 003_02 D 1.0 0.91 1.08 72 0.86 0.87 0.63 004_01 A 0.7 0.78 1.02 8 0.40 0.49 0.58 004_01 B 0.7 0.69 1.03 8 0.38 0.48 0.55 004_01 C 0.7 0.65 1.09 8 0.34 0.44 0.55 004_01 D 0.7 0.63 1.08 8 0.34 0.44 0.54 004_02 A 1.0 1.12 1.01 8 0.70 0.80 0.68 004_02 B 1.0 1.00 1.04 8 0.70 0.82 0.65 004_02 C 1.0 0.94 1.10 8 0.68 0.81 0.64 004_02 D 1.0 0.94 1.06 8 0.70 0.83 0.64 009_01 B 1.0 0.99 1.05 72 0.80 0.81 0.65 1 in. (2.5 cm) Wall Piers (no skew) 003_01 C 0.7 0.64 1.08 72 0.16 0.16 0.22 003_02 C 1.0 0.90 1.09 72 0.26 0.26 0.25 008_01 A 1.0 1.02 1.04 8 0.17 0.18 0.27 Multiple 0.5 in. (1.25 cm) Cylindrical Columns (no skew) 003_01 B 0.7 0.72 1.02 72 0.11 0.11 0.13 003_02 B 1.0 0.97 1.05 72 0.10 0.10 0.15 008_01 B 1.0 1.04 1.06 8 0.11 0.12 0.16 1 Measured scour at Pier A was reduced by a factor of 1.23 to account for the observation that baseline scour was consistently larger at this pier. Table 3.14. Pier scour for baseline tests (no debris). 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.20 0.40 0.60 0.80 1.00 Estimated Equilibrium Scour, ft Pr ed ic te d Sc ou r ( CS U Eq ua tio n) , ft Figure 3.61. Predicted scour vs. equilibrium scour for all baseline tests.

Equilibrium scour depths for baseline (no-debris) condi- tions must be identified for each of these pier types in order to assess the effect of debris on scour. The 1.0 Vcrit condition was used to define baseline scour, as this condition represents the upper limit of clear-water scour, and because almost all of the laboratory tests with debris clusters used this condition. The average of the equilibrium scour depths from tests on square, wall, and multiple-column piers used in this study were 0.83 ft, 0.22 ft, and 0.11 ft, respectively. These values were used as baseline values in the assessment of debris effects on pier scour for the remainder of the laboratory testing program. Baseline tests typically resulted in a very symmetric ellip- tical scour hole (in plan view) for all pier shapes investigated. Figure 3.62 is a photograph of a typical scour hole for a base- line run at a 4 in. square pier. The resulting contour map of the scour from that test is provided in Figure 3.63. The lat- eral extent of the scour hole from the sides of the piers typi- cally ranged from 2.0 to 2.5 times the depth of scour for all pier shapes. 3.6.2 Tests with Debris Rectangular Debris Clusters A total of 39 tests were run under clear-water scour condi- tions with rectangular (in plan and profile) debris clusters affixed to the upstream face of the piers. Five of those tests were 0.5 in. multiple columns (2 tests): 2.4 < Y a se < 2 7. 1 in. wall piers (2 tests): 2.2 < < Y a se 3 1. conducted at a nominal (target) approach velocity of 0.7 Vcrit, where Vcrit was estimated to be 1.4 ft/s (0.43 m/s) for the initi- ation of motion of the bed material. The remaining 34 tests were conducted at a nominal velocity of 1.0 Vcrit. All tests were conducted with a nominal approach flow depth of 1.0 ft. Rectangular debris clusters were fabricated with a range of dimensions as follows: • Width: 1.0 to 6.0 ft (30 to 180 cm) (perpendicular to flow) • Length: 1.0 to 4.0 ft (30 to 120 cm) (aligned with flow direction) • Height: 0.33 to 1.0 ft (10 to 30 cm) (placed at various locations within the water column) Debris clusters were fabricated both with and without roughness elements and filled with either wooden dowels to achieve 70% porosity or wadded geotextile swatches to achieve an essentially impermeable mass. Figure 3.64 is a photograph of a porous, permeable rectangular debris cluster that incor- porates roughness elements. Dye tests performed at rectan- gular debris clusters showed that at the face of the debris mass, flow tended to plunge directly beneath the debris. Very little flow is shed around the sides of a rectangular debris cluster, as seen in the figure. Table 3.15 summarizes the results of the rectangular debris tests. In this table, the actual approach velocity as measured by ADV and the measured approach flow depth are shown, because both velocity and depth during any particular test were typically slightly higher or lower than target conditions. The 85 (Test 009-01, Pier B) Figure 3.62. Typical scour pattern at square pier with no debris. 72 73 74 75 76 77 78 79 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Pier B Flow (Test 009-01, Pier B) -0.05 -0.45 -0.15 -0.15 -0.05 - 0.05 -0.25 -0.35 Figure 3.63. Typical scour contour map at square pier with no debris.

table also shows the measured scour depths at the end of each test and the ultimate scour depth estimated by the Melville method. Scour created by rectangular debris clusters typically resulted in a trough upstream of the pier created by the plunging flow, as shown in Figures 3.65 and 3.66. An important observation made during the course of the testing was that when the upstream extent of the debris was approximately equal to the depth of flow, the deepest part of the trough was coincident with the front face of the pier, significantly increasing the total scour at the pier itself. When the debris extended further than one flow depth in the upstream direction, the trough was moved further upstream as well, and scour at the pier face was less severe. The depth of the upstream trough was found to be depend- ent on the frontal area of the debris mass (debris width times height). In many cases, the depth of the upstream trough was greater than the scour at the pier face. When the scour trough was located some distance upstream of the pier, the depth of the local scour at the pier face was sometimes greater and sometimes less than the baseline (no-debris) scour. The lateral extent of scour created by rectangular debris clus- ters was directly related to the width of the cluster. The impact of the lateral scour extent on adjacent piers was not investigated during this study; however, the lateral scour caused by rectan- gular debris was directly related to the lateral extent of the debris. The side slopes of the scour trough (perpendicular to the flow direction) consistently ranged from 4H:1V to 6H:1V. For many of the tests with rectangular debris, the slope of the lateral scour trough was noted to intersect the flume walls on both sides of the pier. In general, the scour processes described in the preceding paragraphs can be visualized by comparing idealized flow lines at a pier with no debris to those at a pier with a rectangular debris cluster. In Figure 3.67, the flow lines at an unobstructed pier are essentially uniform in the approach section. At the pier, the flow dives down the front face and spirals past the pier in the classic horseshoe vortex pattern. In contrast, flow at a pier with a rectangular debris cluster is significantly obstructed and forced to plunge beneath the upstream face of the debris as shown in Figure 3.68. The plunging flow creates the upstream scour trough that was observed consistently during the laboratory testing program. Because of the blockage created by the debris, some flow is forced around the sides as well. As the flow beneath the debris approaches the pier, the diving and spiral horseshoe patterns are still observed. Depending on the degree of blockage com- pared to the entire channel (flume) cross section, the relative strengths of the diving flow and horseshoe vortex may be greater or less than the unobstructed case. Triangular/Conical Debris Clusters A total of 14 tests were run under clear-water scour condi- tions with debris clusters having a triangular shape in plan view with a conical shape in profile. Three of those tests were con- ducted at a nominal (target) approach velocity of 0.7 Vcrit, where Vcrit was estimated to be 1.4 ft/s (0.43 m/s) for the initiation of 86 a. Upstream b. Downstream (note upwelling flow and dye dispersal) Figure 3.64. Typical plunging flow pattern (shown by dye) at a rectangular debris cluster.

motion of the bed material. The remaining 11 tests were con- ducted at a nominal velocity of 1.0 Vcrit. Two of the tests (the first at 0.7 Vcrit and the second at 1.0 Vcrit) were conducted with the debris mass partially buried beneath the ambient bed level prior to the start of the test. Triangular debris clusters were fabricated with a range of dimensions as follows: • Width: 2.0 to 4.0 ft (60 to 120 cm) (perpendicular to flow) • Length: 2.0 to 6.0 ft (60 to 180 cm) (aligned with flow direction) • Height: 0.33 to 1.0 ft (10 to 30 cm) (placed at the surface or at the bed) Debris clusters were fabricated both with and without rough- ness elements and filled with either wooden dowels to achieve 70% porosity, or wadded geotextile swatches to achieve an essentially impermeable mass. Figure 3.69 is a photograph of a porous, permeable triangular debris cluster that includes 87 Velocity Ratio V/Vc Run No. Pier Debris Shape Debris Location Target Meas. Meas. Approach Depth (ft) Test Duration (ft) Measured Scour1 (ft) Estimated Ultimate Scour (ft) 4 in. (10 cm) Square Piers 003_03 0.4 Rect Surface 0.7 0.82 0.20 A 0.37 1.1 004_03 0.3 Rect Surface 0.7 0.81 0.18 A 0.34 1.1 004_03 0.4 Rect Surface 0.7 0.61 0.15 B 0.40 0.8 004_03 0.5 Rect Surface 0.7 0.61 0.13 C 0.49 0.8 004_03 0.6 Rect Surface 0.7 0.58 0.12 D 0.58 0.7 005_02 0.9 Rect Surface 1.0 1.00 0.23 B 0.92 1.4 006_01 1 Rect Surface 1.0 1.05 0.25 B 1.04 1.4 006_02 0.9 Rect Surface 1.0 1.15 0.28 A 0.90 1.6 009_02 1.1 Rect Surface 1.0 1.00 0.21 C 1.06 1.3 003_04 0.6 Rect Surface 1.0 1.11 0.27 A 0.61 1.5 009_01 0.9 Rect Surface 1.0 1.14 0.26 A 0.93 1.5 009_02 0.9 Rect Surface 1.0 1.07 0.25 A 0.92 1.4 005_02 1 Rect Surface 1.0 1.17 0.27 A 0.98 1.6 006_01 0.9 Rect Surface 1.0 1.13 0.26 A 0.94 1.5 006_02 1 Rect Surface 1.0 1.01 0.23 B 1.02 1.3 004_04 0.7 Rect Surface 1.0 0.96 0.21 B 0.66 1.3 007_01 0.6 Rect Surface 1.0 0.97 0.22 B 0.59 1.3 004_04 0.7 Rect Surface 1.0 1.31 0.28 C 0.74 1.7 007_01 0.9 Rect Surface 1.0 0.94 0.23 D 0.85 1.3 004_04 1 Rect Surface 1.0 1.05 0.23 D 0.96 1.4 009_02 1.2 Rect Surface 1.0 0.92 0.20 D 1.23 1.2 003_03 0.4 Wedge Surface 1.0 0.67 0.15 D 0.41 0.9 007_02 0.9 Wedge Surface 1.0 1.07 0.25 B 0.92 1.5 009_01 1.1 Wedge Surface 1.0 0.99 0.21 C 1.08 1.3 007_02 0.8 Rect Mid-Depth 1.0 1.40 0.24 C 0.8 1.3 005_02 0.9 Rect Mid-Depth 1.0 1.39 0.25 A 0.9 1.4 006_01 1.2 Rect Mid-Depth 1.0 1.42 0.22 B 1.2 1.3 006_02 0.8 Rect Mid-Depth 1.0 1.33 0.23 C 0.8 1.4 004_04 0.4 Rect Mid-Depth 1.0 1.66 0.23 C 0.4 1.4 005_02 0.4 Rect Full Depth 1.0 1.39 0.22 C 0.4 1.3 006_01 0.5 Rect Full Depth 1.0 1.41 0.22 C 0.5 1.3 006_02 0.5 Rect Full Depth 1.0 1.37 0.21 C 0.5 1.3 009_01 0.8 Rect Full Depth 1.0 1.35 0.22 D 0.8 1.3 005_01 0.7 Rect Bed 1.0 1.66 0.28 A 0.7 1.6 1 in. (2.5 cm) Wall Piers (no skew) 008_01 0.5 Rect Full Depth 1.48 0.23 C 0.5 1.4 Multiple 0.5 in. (1.25 cm) Cylindrical Columns (no skew) 008_01 0.4 Rect Surface 1.0 1.02 0.23 D 0.35 1.4 008_02 0.5 Wedge Surface 1.0 1.12 0.26 B 0.53 1.5 008_02 0.5 Rect Mid-Depth 1.0 1.49 0.26 A 0.5 1.5 1 Measured scour at Pier A was reduced by a factor of 1.23 to account for the observation that baseline scour was consistently larger at this pier. Table 3.15. Pier scour for rectangular debris tests.

roughness elements. Dye tests performed at triangular clusters showed that at the centerline of the debris mass, flow tended to plunge beneath the debris but was also shed very readily around the sides of the debris, as seen in the figure. Table 3.16 summarizes the results of the triangular debris tests. In this table, the actual approach velocity as measured by ADV and the measured approach flow depth are shown, because both velocity and depth during any particular test were typically slightly higher or lower than target conditions. The table also shows the measured scour depths at the end of each test and the ultimate scour depth estimated by the Melville method. The scour pattern created by triangular debris clusters was markedly different from that exhibited by the rectangular clus- ters. No scour troughs upstream of the pier were observed with any of the triangular debris clusters. The portion of the flow that plunges beneath a triangular/ conical blockage is seen to be funneled towards the pier face, creating additional scour at the pier compared to the baseline condition. The scour at the pier face was found to be related to the thickness of the debris blockage at the pier face; i.e., a greater thickness of debris lodged directly against the pier created more scour at the pier face, with the triangular debris shapes. 88 Figure 3.65. Typical scour pattern at square pier with rectangular debris cluster. (Test 004-03, Pier C) Figure 3.66. Typical scour contour map at square pier with rectangular debris cluster. Flow 102 103 104 105 106 107 108 109 110 111 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Pier C -0.1 -0.1 - 0. 1 (Test 004-03, Pier C) -0.2 - 0.2 -0.1 0 - 0.3 0 0 Figure 3.67. Idealized flow pattern at an unobstructed pier.

As with the rectangular debris tests, lateral extent of scour created by triangular debris clusters was directly related to the width of the cluster. However, the lateral extent of scour caused by a triangular debris cluster was shown to be greater than that of a rectangular one. This greater lateral extent appears to be caused by the shedding of flow around the triangular debris and has implications regarding adjacent piers or abutments. Figures 3.70 and 3.71 illustrate typical scour patterns and contours created by triangular debris clusters. The slopes of the scour trough (oriented radially from the pier face) caused by triangular debris clusters consistently ranged from 2H:1V to 2.5H:1V. An idealized flow pattern around a triangular debris clus- ter is shown in Figure 3.72. Note that the flow both plunges beneath the debris and is shed to the sides, as discussed previously. Effect of Debris Roughness and Porosity on Scour The laboratory studies revealed that the roughness and porosity of a debris mass do not significantly affect the pat- tern of scour or the magnitude of the scour depth at the pier face. For the range of these properties examined during this investigation, debris roughness and porosity can at most be considered second-order factors affecting pier scour and are much less important than (1) the size and shape of the pier and (2) the size, shape, and location (floating or buried) of the debris that collects on the pier. The effects of roughness and porosity are discussed in more detail in Section 3.7.4. Summary 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” dimension) 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 89 Figure 3.68. Idealized flow pattern at a rectangular debris cluster. Rectangular debris Figure 3.69. Typical plunging flow with shedding around the sides (shown by dye and ripples) at a triangular debris cluster. a. Upstream b. Downstream (note upwelling flow and dye dispersal)

Figure 3.70. Typical scour pattern at square pier with triangular debris cluster. (Test 007-02, Pier A) ness 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 at the pier face compared to the baseline (no-debris) condition. 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 com- pared to a rectangular, blunt-shaped blockage. Although the effect of lateral scour extent on adjacent piers was not investigated in detail, data collected from the laboratory tests yield valuable information in this regard. Both rectangular and triangular 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 rela- tively mild (ranging from about 6H:1V to 4H:1V) for rectangu- lar debris compared to lateral side slopes produced by triangular debris (minimum 2.5H:1V). 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 experi- ence less total scour than a wider pier with the same amount of debris, for the same hydraulic conditions of the approach flow. 3.7 Scour Prediction at Bridge Piers with Debris Loading 3.7.1 Introduction 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 90 Velocity Ratio V/Vc Run No. Pier Debris Shape Debris Location Target Meas. Meas. Approach Depth (ft) Test Duration (ft) Measured Scour1 (ft) Estimated Ultimate Scour (ft) 4 in. (10 cm) Square Piers 003_03 0.4 Triangle Surface 0.7 0.66 0.16 B 0.41 0.9 007_01 0.7 Triangle Surface 1.0 1.08 0.26 A 0.71 1.5 003_04 0.7 Triangle Surface 1.0 1.01 0.24 B 0.72 1.4 003_04 0.7 Triangle Surface 1.0 0.99 0.22 D 0.69 1.3 007_02 0.8 Triangle Surface 1.0 0.85 0.24 C 0.80 1.3 007_02 0.9 Triangle Surface 1.0 1.07 0.25 A 0.93 1.4 009_02 1.2 Triangle Surface 1.0 0.99 0.22 B 1.19 1.3 007_01 0.8 Triangle Surface 1.0 1.04 0.23 C 0.77 1.4 003_03 0.1 Bur wedge Surface 0.7 0.66 0.14 C 0.09 0.9 003_04 0.2 Bur wedge Surface 2.0 0.99 0.21 C 0.21 1.3 005_01 0.2 Inv Cone on bed Surface 3.0 0.94 0.20 C 0.19 1.2 1 in. (2.5 cm) Wall Piers (no skew) 008_02 0.4 Triangle Surface 1.0 1.07 0.23 C 0.37 1.4 008_02 0.5 Triangle Surface 1.0 1.15 0.26 A 0.53 1.5 1 Measured scour at Pier A was reduced by a factor of 1.23 to account for the observation that baseline scour was consistently larger at this pier. Table 3.16. Pier scour for triangular debris tests. the pier face, resulting in a worst-case scour condition. Total scour at the pier also tended to increase 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 were also investigated, because the debris photographic archive revealed that this is another very common shape that can be produced in the field as drift accumulates at a pier. In a triangular configuration, the thick-

depth of scour at the pier. The factors examined in this study included the following: • 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 wide (2.4 m) flume at CSU. Most of the tests were conducted using 4 in. (10.2 cm) square piers. Tests of selected debris shapes were also conducted using slender piers, consisting of 1 in. wide (2.5 cm) square- nosed wall piers and multiple 0.5 in. wide (1.3 cm) cylindri- cal column piers. However, tests using the slender piers were necessarily limited in number, and only a small subset of debris shapes could be examined with the resources avail- able 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 to prototype scale ratio of approx- imately 1:10 to 1:30 can be considered a reasonable range for the tests conducted in the 8 ft (2.4 m) 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, includ- ing near-prototype scale ratios of approximately 1:2 to 1:4 were originally considered but were ultimately dropped from the program so that other factors could be investigated in more detail. A total of 53 tests of debris-laden piers was run under clear- water scour conditions. These are generally categorized by debris shape and target approach velocity as follows: • Rectangular debris shapes: – 0.7 Vcrit: 5 tests – Vcrit: 34 tests 91 Figure 3.71. Typical scour contour map at square pier with triangular debris cluster. - 0.1 5 -0.15 -0.35 -0.55 -0.65 -0.75 -0.45 -0.35 -0.25 -0.15 -0.05 0.05 -0.25 - 0.2 5 - 0.1 5 39 40 41 42 43 44 45 46 47 2 4 6 Pier A Flow (Test 007-02, Pier A) Figure 3.72. Idealized flow pattern at a triangular debris cluster. Triangular debris

• Triangular debris shapes: – 0.7 Vcrit: 3 tests – Vcrit: 11 tests Most of the tests (35 tests) were conducted with the top surface of the debris at the water surface, forming a “raft.” Selected tests were also performed with the debris located in the center of the water column, resting on the bed, or buried into the bed. 3.7.2 Equivalent Pier Width The concept of equivalent pier width has been widely accepted as a way to quantitatively assess the extent to which debris affects scour at piers. Using the data collected from the labora- tory program, this concept has been explored in great detail and appears to have the best promise for predicting the effect of debris on pier scour. 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, W, of the debris perpendicular to the flow direction to estimate the additional obstruction. 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, W, and thickness, T, of the debris and is based on scour data from a limited number of tests (17 tests) in a laboratory flume. Only floating (surface) debris at cylindrical piers was tested, with the debris wrapped around the pier in all direc- tions. The effect of the vertical location of the debris mass within the water column was not investigated. The equation to calculate equivalent pier width is: where: be = Effective width of the pier, ft (m) Kd1 = Dimensionless coefficient equal to 0.52 from labora- tory tests (Dongol 1989) T = Thickness of debris, ft (m) W = Width of debris normal to flow, ft (m) a = Pier width (without debris) normal to flow, ft (m) y = Depth of approach flow, ft (m) The effective width equation was compared to the results of the laboratory tests conducted at CSU under this study pro- gram. The observed effective width (denoted ad ) for all tests at 1.0 Vcrit with debris at the water surface was determined from the CSU pier scour equation, using the equilibrium scour depth Yse obtained from each test. The calculated value of be obtained from the Melville and Dongol equation was then plotted against ad to determine how well the effective width equation predicts the actual effective width. Figure 3.73 presents the results of that comparison for both rectangular and triangular shapes. Figure 3.73 indicates that the Melville–Dongol equation tends to overestimate the effective width of the pier when debris is present, particularly for triangular shapes. The Melville– Dongol equation does not take into account the shape of the debris mass (e.g., rectangular vs. triangular), nor does it con- sider the length, L, of the debris extending upstream from the pier. As discussed in Section 3.6, these aspects were observed to have an effect on the scour pattern as well as the total depth of scour at the pier face. b K TW y K T a y e d d = ( )+ −( )1 1 3 11( . ) 92 Figure 3.73. Comparison of the Melville and Dongol effective width to the observed effective width. Debris on Surface; V/Vc = 1.0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5 Observed a*d, ft Ca lc ul at ed b e, ft Rectangular debris Triangular debris

A modification to the equivalent width equation was therefore proposed and tested against the laboratory data. The proposed modification is denoted as “ad ” to distinguish it from the Melville and Dongol “be” and is given as: where: Kd1 = Dimensionless coefficient optimized from laboratory test data Kd2 = Dimensionless exponent optimized from laboratory test data L = Length of debris upstream from pier face, ft (m) Other terms as are as defined previously. a K TW L y y K T a y d d K d d ∗ = ( )( ) + −( )1 12 3 12( . ) 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. For rectangular debris shapes, Kd1 and Kd2 were found to be 0.39 and −0.79, respectively, whereas for triangular shapes, Kd1 and Kd2 were 0.14 and −0.17. The coefficient Kd1 is thus seen to be a shape factor, while the exponent Kd2 is a factor that describes the intensity of the plunging flow created by the debris blockage. The result of this comparison is presented in Figure 3.74 for both rectangular and triangular shapes. Figure 3.74 shows that accounting for debris shape and length significantly improves the ability to predict the equiv- alent pier width (compare with Figure 3.73). Predicted and observed equilibrium scour depths are shown in Figure 3.75 for all runs with debris at the water surface and an approach velocity of 1.0 Vcrit. In this figure, the CSU equation is used to 93 Figure 3.74. Comparison of the modified effective width to the observed effective width. Debris on surface; V/Vc = 1.0 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5 Observed a*d, ft Ca lc ul at ed a * d, ft Rectangular debris Triangular debris Figure 3.75. Comparison of observed scour to predicted scour using the modified equation for equivalent pier width. Debris on Surface; V/Vc = 1.0 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 Observed Scour, ft Pr ed ic te d Sc ou r, ft Rectangular debris Triangular debris Baseline (no debris)

predict ultimate clear-water scour at the pier face, with the equivalent pier width calculated using the modified version of the Melville–Dongol equation (Equation 3.12). 3.7.3 Recommended Design Equation The previous section developed a predictive relationship to estimate local scour at a debris-laden bridge pier. The rela- tionship was derived using an “equivalent width” concept by modifying the approach developed by Melville and Dongol. Figure 3.75 shows that the predictive equation, using opti- mized coefficients and exponents based on laboratory data, 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. Based on the labora- tory data developed for an approach velocity of 1.0 Vcrit, the shape coefficient Kd1 that provides overestimation 90% of the time (underestimating 10% of the observations) is 0.79 for rectangular debris shapes, and 0.21 for triangular shapes. The recommended design equations for estimating an equivalent pier width for use with the CSU pier scour equa- tion are, therefore: and where: Kd1 = 0.79 for rectangular debris, 0.21 for triangular debris Kd2 = −0.79 for rectangular debris, −0.17 for triangular debris L = Length of debris upstream from pier face, ft (m) y = Depth of approach flow, ft (m) a K TW y K T a y L yd d d∗ = ( )+ −( ) ≤1 1 1 0 3 14for . ( . ) a K TW L y y K T a y L yd d K d d ∗ = ( )( ) + −( ) >1 1 2 1 0 3 1for . ( . 3) Other terms are as defined previously. The design or “envelope” values using the recommended equations are shown in Figure 3.76 for all runs with debris at the water surface and an approach velocity of 1.0 Vcrit. In this figure, the CSU equation is used to predict ultimate clear- water scour at the pier face, using the equivalent pier width calculated by Equations 3.13 and 3.14 and the recommended Kd1 and Kd2 values presented above. 3.7.4 Effect of Debris Roughness and Porosity The data from the laboratory research program indicate that roughness and porosity of the debris mass do not significantly affect the observed scour. The effects of roughness and poros- ity were investigated using a rectangular debris shape 2 ft wide by 1 ft long (0.6 m by 0.3 m). Nine tests were conducted with this shape mounted on the front of 4 in. (10.2 cm) square piers, with an approach velocity of 1.0 Vcrit. An example of a smooth, impermeable debris shape is provided as Figure 3.77, and a rough, porous shape is shown in Figure 3.78. The effect of roughness and porosity characteristics on scour was quantified using the ratio of scour depth caused by the debris to the baseline (no-debris) condition. The average scour ratio for all tests was 1.35, with a range of 1.28 to 1.48 and a standard deviation of only 5% about the mean. Figure 3.79 shows the results of these tests. At most, roughness and porosity can be considered second-order variables that are not significant compared to the size and shape of the debris mass. 3.7.5 Effect of Debris Location in the Water Column The data from the laboratory research program 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 was 94 Figure 3.76. Comparison of observed scour to the recommended design equation using 90% envelope values. Debris on Surface; V/Vc = 1.0 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 Observed Scour, ft En ve lo pe S co ur , f t Rectangular debris Triangular debris Baseline (no debris)

95 Figure 3.77. Example of a smooth, impermeable debris shape. Figure 3.78. Example of a rough, porous debris shape. Figure 3.79. Effect of debris roughness and porosity on observed pier scour. Rectangular Debris on Surface, V/Vc = 1.0 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 Smooth and Impermeable (2 tests) Rough and Impermeable (2 tests) Smooth and Porous (2 tests) Rough and Porous (3 tests) Ys e/ Y se b as el in e All tests utilized a rectangular debris shape, 2 ft wide by 1 ft long. investigated using two different rectangular debris shapes. The first shape was 2 ft wide by 2 ft long (0.6 m by 0.6 m), and the second was 1 ft wide by 1 ft long (0.3 m by 0.3 m). Fourteen tests were conducted with these shapes mounted on the front of 4 in. (10.2 cm) square piers, with an approach velocity of 1.0 Vcrit. Figure 3.80 shows a 1 ft by 1 ft rectangular debris shape on the bed in front of Pier D prior to testing. Figures 3.81 and 3.82 show the scour resulting from Test 009_01, a 72-hour test conducted with an approach velocity of 1.0 Vcrit. The effect of debris location in the water column on pier scour was quantified by the ratio of scour depth caused by the debris to the baseline (no-debris) condition. In general, the 2 ft wide by 2 ft long (0.6 m by 0.6 m) debris placed as a sur- face raft caused slightly greater scour at the 4 in. (10.2 cm) square piers compared to baseline conditions.

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 in this study program. However, inferences in this regard can be drawn from the lab- oratory data collected: • Rectangular debris: The lateral extent of scour created by rectangular floating debris extends outward from the edge of the debris (as measured from the pier face) at a slope ranging from about 4H:1V to 6H:1V. • Triangular debris: The lateral extent of scour created by triangular floating debris extends outward from the edge of the debris (as measured from the pier face) at a slope ranging from 2H:1V to 3H:1V. To estimate the impact on adjacent bridge elements (piers or abutments) caused by debris loading on a single pier, the recommended guidance is as follows: 1. Estimate the total scour at the debris-laden pier. 2. Extending from the edge of the debris, compute a lateral slope of 6H:1V for rectangular debris or 3H:1V for trian- gular debris. 3. Determine the scour prism using these values and determine the effect on adjacent bridge foundation elements. 96 Figure 3.81. Pier D after Test 009_01, 72 hours at 1.0 Vcrit. Figure 3.82. Maximum scour depth at the Pier D face after Test 009_01 is less than that obtained from the baseline (no-debris) condition. - 0. 05 - 0.0 5 0 - 0.1 -0.3 - 0.6 - 0. 2 -0.15 -0.15 -0.1 -0.1 -0.15 -0.2 -0.2 -0.25 - 0.25 -0.45 -0.45 -0.4 -0.35 -0.3- 0.4 5 -0 .5-0.55 -0.65 -0.7 135 136 137 138 139 140 141 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 Pier D Flow Figure 3.80. Pier D prior to Test 009_01. In contrast, when this same shape was located at mid-depth in the flow, significantly less scour at the pier face was observed. This difference was presumably due to the relative distribution of flow over the debris compared to the plunging flow occur- ring beneath it. Similarly, when the debris was placed on the bed, less scour was observed at the pier face compared to base- line conditions. Figure 3.83 shows the results of the laboratory tests as a function of debris location in the water column. 3.7.6 Lateral Extent of Scour at Piers with Debris When a debris mass accumulates at a pier, it typically initiates and grows from floating (usually organic) drift material. The laboratory tests conducted under this study

3.8 Incorporating Debris in Hydraulic Models HEC-RAS (Brunner 2008) is a one-dimensional model that is the primary tool for simulating hydraulic conditions at bridges. The Finite-Element Surface Water Modeling System (FESWMS) (Froehlich 2002, rev. 2003) and RMA2 (Donnell et al. 2006) are used to simulate hydraulic conditions at bridges for more complex situations that require two-dimensional hydraulic analysis. The hydraulic effects of debris can be incor- porated into each of these models. HEC-9 (Bradley et al. 2005) provides guidance on incorporating debris in hydraulic mod- els. This section supplements the information in HEC-9. 3.8.1 HEC-RAS In HEC-RAS, floating debris clusters can be added to bridge piers. Figure 3.84 shows a bridge crossing with a 6 ft high by 18 ft wide (1.8 by 5.5 m) debris cluster only at the center pier. Consistently sized debris clusters can be included at all piers or varying sizes can be input at individual piers. The debris cluster is centered on the pier and moves up and down with the water surface. It becomes fixed in place when it comes in con- tact with the low chord of the deck at the centerline of the pier. A debris raft can be simulated by setting the width dimensions of the debris clusters to form a continuous blockage. When floating debris is included at piers, HEC-RAS only includes the debris at the upstream face of the internal bridge sec- tions. Debris that has accumulated on the bottom of the deck can be simulated by changing the low chord of the bridge deck. If the debris is only at the upstream face of the bridge, only the low chord of the upstream internal bridge section should be adjusted. The HEC-RAS model includes four bridge hydraulic compu- tation methods for low flow and two methods for high flow. Low flow is defined as flow through the bridge opening without the occurrence of either pressure flow (submerged deck) or roadway overtopping. The four low flow methods are energy (standard step method), momentum, WSPRO, and Yarnell. The two high flow methods are energy and pressure/weir. Each bridge hydraulic method incorporates the debris effects based on the hydraulic computation assumptions inherent to that method. For example, the energy method removes flow area at a bridge cross section based on the areas blocked by the embankments, piers, and abutments and includes wetted perimeter for each of these obstructions. In the energy method, the area and wetted perimeter of the debris cluster are included as part of the bridge pier obstruction. In the momentum method, the debris area is included in the drag force of piers in the force balance of the momentum solution at the bridge. Just as the results of the HEC-RAS analysis differ based on the user-selected bridge hydraulic method, the effects of the debris also differ between the methods. For the example model shown in Figure 3.84, the debris increased the energy 97 Figure 3.83. Effect of debris location in the water column on observed pier scour. Rectangular Debris, V/Vc = 1.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 Surface raft, 2'x2' On bed, 2'x2' Ys e/ Y se b as el in e (Average, 4 tests) (Average, 4 tests) (Average, 2 tests) (Average, 4 tests) Mid-depth, 2'x2' On bed, 1'x1'

grade upstream of the bridge by 0.06 ft (18.3 mm) for WSPRO, 0.10 ft (30.5 mm) for Yarnell, 0.14 ft (42.7 mm) for energy, and 0.18 ft (54.9 mm) for momentum. When the tailwater in the example model was increased to create a pressure and over- topping condition, the debris increased the upstream energy grade by 0.03 ft (9.1 mm) for pressure/weir and 0.04 (12.2 mm) for the energy method. The bridge hydraulic method (energy, momentum, etc.) should be selected based on the suitability of that approach to the particular bridge crossing without considering debris. When debris is included, the same method should be used. If the momentum equation is used, then NCHRP Report 445 (Parola et al. 2000) can be consulted for selection of the drag coefficient. Although the Parola report is directed at calculating debris forces on bridges, the guidance can be used in selecting the appropriate drag coefficients for debris in hydraulic model- ing. If debris blocks a large portion of a bridge, the upstream and downstream cross sections adjacent to the bridge may need to include ineffective flow areas to represent areas that are not actively conveying flow. HEC-RAS also includes scour computations in the Hydraulic Design menu. Debris only affects the scour results to the extent that it impacts the hydraulic results, such as a different hydraulic depth or flow distribution in the contraction scour calculation and a local depth or flow velocity in the pier scour calculation. The default pier width in the pier scour calculation is the pier width entered in the HEC-RAS bridge data editor. In the Hydraulic Design menu, the pier width value can be overridden so the value of ad (the effective full-depth pier width computed as in Section 3.7) based on the debris dimensions can be used. As with any automated calculation, the results should be checked for accuracy. While directly using ad as the pier width in the HEC-RAS model may be tempting, this is not recommended. The hydraulic effect of debris is to block area. Therefore, the actual blocked area should be used in bridge hydraulic computa- tions and ad should be used only for scour calculations. The total area of blockage (debris plus the area of the remaining pier below the debris cluster) would not be expected to equal the area of the “effective width” pier. 3.8.2 Two-Dimensional Models Two-dimensional finite-element models commonly used for bridge hydraulics (FESWMS and RMA2) use the momen- tum equation for hydraulic calculations. Actual pier dimen- sions and locations can be included directly in the FESWMS model. The model then computes the additional drag force 98 Figure 3.84. Debris cluster in the HEC-RAS hydraulic model. 0 100 200 300 400 500 600 700 800 80 85 90 95 100 105 Bridge - Debris - Energy Station (ft) El ev at io n (ft) Legend EG PF 1 WS PF 1 Ground Ineff Bank Sta Pier Debris

caused by the obstruction. To model debris in FESWMS, the user should calculate the total obstructed area of the pier and debris, divide by the flow depth, and use the resulting width as the pier width. This width is not the same as the effective width, ad , used in the scour calculation. Even though the debris may be at the surface, the computed drag force does not account for the vertical location, so the com- puted hydraulic effect is the same regardless of the vertical location. NCHRP Report 445 (Parola et al. 2000) can be con- sulted for selection of the drag coefficient. To include the additional drag force in an RMA2 model is not as easy. In RMA2 (and FESWMS), the shear stress on the bed can be estimated from το = γyS and substituting the Manning equation for energy slope (assuming gradually var- ied flow, which is the assumption in these models). The resulting equation for shear stress is: where: γ = Unit weight of water n = Manning flow resistance coefficient V = Velocity Y = Flow depth (wide channel assumption) Ku = 1.486 for U.S. customary units and 1.0 for SI The resulting force on an element in the 2-D model net- work is approximately equal to the area of that element times the shear stress computed from the average velocity and depth for that element. To include the additional drag force caused by a pier and debris, one can calculate the drag force (Fd = 0.5CdρApV2), add it to the bed force, and then back calculate an effective Manning n for that element that includes both forces. The resulting effective Manning n (ne) can be computed without knowing velocity, because veloc- ity squared is included in both forces and the equation for effective Manning n (ne) is: where: n = Manning n of the bed Cd = Drag coefficient Ap = Sum of the pier and debris cross-sectional areas for all the piers that are located within the element AE = Area of the element (in plan), y = Average flow depth g = Gravitational constant Ku = 1.486 for U.S. customary units and 1.0 for SI n n C A K y gA e d p u E 2 2 2 1 3 2 3 16= + ( . ) τ γ 0 2 2 2 1 3 3 15= n V K yu ( . ) For example, if V is 5.0 ft/s (1.52 m/s), n is 0.030, Cd is 1.2, Ap is 45 ft2 (4.18 m2), y is 10 ft (3.0 m), and AE is 400 ft2 (37.2 m2), then ne would be 0.1043. As a check, the force on this element from bed stress alone is 118 lb (525 N), the drag force is 1310 lb (5827 N), for a total of 1428 lb (6352 N), and the force from the adjusted Manning n (ne = 0.1043) is 1427 lb (6347 N) where the difference is due to rounding. An approximate overall hydraulic impact of piers and debris on a bridge can be included by using Ap equal to the sum of areas for all piers and debris at the bridge, y equal to the hydraulic depth of the bridge, and AE equal to the total area (in plan) of the elements representing the bridge in the finite-element network. Another way that piers can be included in 2-D models is to represent the pier with disabled elements in the finite-element network. This method is only used if the pier is large and the flow is significantly altered around the pier. If this method is used, then the dimensions of the disabled elements can be increased to account for debris blockage. This method would be most applicable to the situation where the debris predom- inantly blocks the water column. The width of the disabled element could also be estimated as the total area of the debris and pier divided by the flow depth. To calculate pier scour from a 2-D model, the user would use the hydraulic results of the model combined with ad deter- mined as in Section 3.7. The value of ad should not be used as the pier width for hydraulic calculations within the 2-D model. 3.9 Application Methodology and Examples 3.9.1 Methodology The methodology for determining scour at a bridge with debris at the piers involves all of the steps that would be per- formed for any other bridge (see Richardson and Davis 2001) plus additional steps that relate to debris scour. The steps are summarized below: 1. Conduct a field reconnaissance to determine site condi- tions for hydraulic modeling, including the potential for debris production, debris delivery, key log dimensions, and existing debris accumulations (see the guidelines and flowcharts in Section 3.2). 2. Review bridge inspection reports for information on debris, including size and shape. 3. Contact bridge maintenance staff for information on debris removal, including size, frequency, and shape. 4. Based on the information obtained in the previous steps determine debris dimensions and shape that can be expected during extreme events. The existing bridge infor- mation can be used for analysis of replacement bridges, 99

but care must be taken to assess whether changes in span length, deck height, pier type, pier orientation, or pier placement would affect potential debris characteristics. For a new bridge crossing, the assessment should include other nearby bridges and differences in reach conditions for debris production and delivery as well as differences in structure type when determining the debris characteris- tics. The photographic archive (Appendix B) can also be used as a resource by comparing the bridge characteristics with photographs in the archive. 5. Perform hydraulic modeling (typically using HEC-RAS, see Section 3.8) with and without debris at the piers. 6. Compute scour for the pier without debris. If the pier is skewed, compute the projected width of the pier using a maximum pier length of 12 times the pier width per HEC-18 guidance (Richardson and Davis 2001). 7. Compute the effective pier width, ad , for the selected debris dimensions using the equations in Section 3.7 and the hydraulic conditions from the with-debris model. If the pier is skewed, use the projected width as the pier width in the ad calculation. (Note: This guidance is based on the judgment of the research team and not on labora- tory investigation.) 8. If the length, L, of the debris cluster upstream of the pier exceeds the flow depth, y, also calculate ad for L equal to y. This step is necessary because the plunging flow scour is greatest for L/y = 1.0, so this may be the controlling case for scour (see Section 3.6). When reducing the length, L, also adjust the width, W, and thickness, T, based on the expected debris conditions. In some cases, the width and thickness would not be decreased. The width should not be decreased to a value less than the key log length, and the thickness should not be decreased less than the expected rootwad diameter. If there is a large difference in the debris size, a third calculation for an intermediate debris cluster size may also be warranted. 9. Calculate pier scour with debris for the largest value of ad obtained in Steps 7 and 8. Do not include the pier shape coefficient, K1, or the skewed pier coefficient, K2, in the HEC-18 equation for pier scour when making this calculation. 10. Check the scour estimates for reasonableness. For exam- ple, if the initial calculation of debris scour is for a rectan- gular cluster, a comparison could be made for a triangular cluster with the same dimensions. 3.9.2 Example Debris Scour Calculations Steps 1 through 5 This example uses the South Platte River case study pre- sented in Appendix D (Part 3) as the information for deter- mining bridge conditions, probable debris characteristics, and hydraulic conditions. The case study results indicate that there is a high potential for debris production, high potential for debris transport and delivery, and a high potential for accumulation of debris at Pier 2, which is located in the cen- ter of the bridge near the middle of the channel. The span length is 112.9 ft (34.4 m), which is much longer than the key log length of 20 ft (6 m). Therefore, the debris is extremely unlikely to bridge between piers to form a raft. The key log diameter is approximately 1.5 ft (0.46 m) and rootwad sizes can exceed 6 ft (1.8 m). Inspection records of the existing bridge, a previous bridge at this location and nearby bridges indicate frequent debris accumulations. The debris accumulation at Pier 2 for an extreme event is assumed to be 30 ft (9.1 m) wide, 6 ft (1.8 m) thick, and 20 ft (6.1 m) long. This size accumulation is based on the high debris accumulation potential, key log length, key log diam- eter, and rootwad size. The shape is assumed to be rectangu- lar. The selection of the debris accumulation dimensions is based on engineering judgment using the key log dimensions and the assumption that the accumulation would be significant during a design event. These assumptions should be confirmed for reasonableness with bridge maintenance and inspection personnel (and/or by evaluating debris size and shape in the photographic archive in Appendix A). The hydraulic conditions are calculated for a 100-year flood with and without debris loading. Without debris load- ing, the maximum channel velocity is 6.25 ft/s (1.91 m/s) and flow depth is 15.3 ft (4.66 m). When the hydraulic model is run to simulate debris loading, the maximum channel veloc- ity is 6.17 ft/s (1.88 m/s) and the flow depth is 15.5 ft (4.72 m). Because the length of the debris cluster exceeds the flow depth, the debris scour will also need to be computed for a length equal to the flow depth. It is assumed that this shorter debris pile is 26 ft (7.9 m) wide but remains 6 ft (1.8 m) thick and that the hydraulic conditions are the same as for the larger debris cluster. The wall pier at this bridge has a width, a, of 1.5 ft (0.46 m); a length, L, of 43 ft (13.1 m); a 5° angle of attack; and a sharp nose (actually a debris deflector). Because the pier is more than 12 times the pier width, a maximum length of 12 × 1.5 ft = 18 ft (12 × 0.46 m = 5.5 m) is used per HEC-18 guid- ance. In the pier scour equation, the K1 pier shape factor is 1.0 (because of the skew) rather than 0.9 for a sharp nose. The pier scour equation K2 factor can be calculated based on HEC-18 guidance, or the projected width of the pier can be used in lieu of using K2. K3 is 1.1 based on an assumption of plane bed or small dunes expected on the South Platte River during extreme floods, and K4 is 1.0 because armoring is not expected. Guidance on obtaining the above information is contained in Steps 1 through 5 of the methodology outlined in the previous section. 100

Step 6 Compute pier scour without debris: Alternatively, the pier scour can be computed using the projected width and excluding K2 from the pier scour equation. The projected width of the pier without debris is: Step 7 Determine the effective pier width with debris for the max- imum debris dimensions. Maximum debris dimensions are W = 30 ft (9.1 m), L = 20 ft (6.1 m), and T = 6 ft (1.8 m) and the projected width of the pier should be used. For a rectan- gular debris cluster Kd1 = 0.79 and Kd2 = −0.79. a TW L y y T a y a d d ∗ = ⎛ ⎝⎜ ⎞ ⎠⎟ + −( ) ∗ = − 0 79 0 79 0 79 . . . proj 0 79 6 30 20 15 5 15 5 0 79 6 3 0 79 . . . . . . × × ⎛⎝⎜ ⎞⎠⎟ + − ×( ) − 1 15 5 9 7 3 0 . . .= ( )ft m a a Lproj = ( )+ ( ) = ( )+ ( ) = Cos Sin Cos Sinθ θ 1 5 5 18 5 3 . .1 0 93 2 0 1 3 4 0 3 ft m proj proj . . . ( ) = ⎛ ⎝⎜ ⎞ ⎠⎟y a K K K y a s 5 0 43 2 0 3 1 1 0 1 1 1 0 15 3 3 1 Fr ys . . . . . . . . = × × × × ⎛⎝⎜ ⎞⎠⎟ 0 35 0 43 0 28 6 9 2 1 . . . . . ( ) = ( )ft m . . .y aK K K K y a Fr K s = ⎛⎝⎜ ⎞⎠⎟ = ( ) 2 0 1 2 3 4 0 35 0 43 2 Cos θ + ⎛⎝⎜ ⎞⎠⎟ ( )⎡⎣⎢ ⎤ ⎦⎥ = ( )+ L a Sin Cos θ 0 65 5 18 1 . . . . . . 5 5 1 6 6 25 32 0 65⎛⎝⎜ ⎞⎠⎟ ( )⎡⎣⎢ ⎤ ⎦⎥ = = = Sin Fr V gy 2 15 3 0 28 2 0 1 5 1 0 1 6 1 1 1 0 15 3 1 × = = × × × × × . . . . . . . . . ys . . . . . . 5 0 28 6 9 2 1 0 35 0 43⎛⎝⎜ ⎞⎠⎟ ( ) = ( )ft m Step 8 Determine the effective pier width for the debris length equal to the flow depth. For a debris length equal to the flow depth, the debris dimensions are W = 26 ft (7.9 m), L = 15.5 ft (4.7 m), and T = 6 ft (1.8 m), where W and T are assumed based on the guidance in the previous section. Step 9 Calculate scour for ad equal to the largest computed value of 10.1 ft (3.1 m) excluding K1 and K2 from the pier scour equation. Step 10 For comparison, compute the scour for a triangular debris accumulation with the same dimensions. Maximum debris dimensions are W = 30 ft (9.1 m), L = 20 ft (6.1 m), and T = 6 ft (1.8 m). For a triangular debris cluster, Kd1 = 0.21 and Kd2 = −0.17. For a debris length equal to the flow depth, the debris dimensions are W = 26 ft (7.9 m), L = 15.5 ft (4.7 m), and T = 6 ft (1.8 m). a TW y T a y a d d ∗ = + −( ) ∗ = × × + 0 21 0 21 0 21 6 26 15 . . . . proj 5 0 21 6 3 1 15 5 5 0 1 58 − ×( ) = ( ). . . . .ft m a TW L y y T a y a d d ∗ = ⎛ ⎝⎜ ⎞ ⎠⎟ + −( ) ∗ = − 0 21 0 21 0 17 . . . proj 0 21 6 30 20 15 5 15 5 0 21 6 3 0 17 . . . . . . × × ⎛⎝⎜ ⎞⎠⎟ + − ×( ) − 1 15 5 5 2 1 58 . . .= ( )ft m y a K K y a Fr Fr V gy s d d = ∗ ∗ ⎛ ⎝⎜ ⎞ ⎠⎟ = = 2 0 6 1 3 4 0 35 0 43. . . . 7 32 2 15 5 0 28 2 0 10 1 1 1 1 0 15 5 10 1 . . . . . . . . . × = = × × ×ys ⎛⎝⎜ ⎞⎠⎟ ( ) = ( ) 0 35 0 43 0 28 14 9 4 54 . . . . .ft m a TW y T a y a d d ∗ = + −( ) ∗ = × × + 0 79 0 79 0 79 6 26 15 . . . . proj 5 0 79 6 3 1 15 5 10 1 3 1 − ×( ) = ( ). . . . .ft m 101

Calculate scour for ad equal to the largest computed value of 5.2 ft (1.58 m) excluding K1 and K2 from the pier scour equation. Summary In summary, the pier scour would be 6.9 ft (2.1 m) with- out debris, 14.9 ft (4.5 m) with a rectangular debris cluster, and 9.7 ft (3.0 m) with a triangular debris cluster. The con- trolling condition for the rectangular cluster is when L/y = 1.0 (plunging flow coincident with the pier face) and, for the tri- angular cluster, the controlling condition is when the debris accumulation is at the maximum size. 3.10 Guidelines for Inspection, Monitoring, and Maintenance 3.10.1 Inspection At bridges, debris characteristics can include single or multiple logs at a single pier, floating clusters at a pier, a floating raft spanning two or more piers, and submerged or y a K K y a Fr Fr V gy s d d = ∗ ∗ ⎛ ⎝⎜ ⎞ ⎠⎟ = = 2 0 6 1 3 4 0 35 0 43. . . . 7 32 2 15 5 0 28 2 0 5 2 1 1 1 0 15 5 5 2 . . . . . . . . . × = = × × × ⎛⎝ys ⎜ ⎞⎠⎟ ( ) = ( ) 0 35 0 43 0 28 9 7 2 96 . . . . .ft m sunken variations of these configurations. The debris jams compiled in the photographic archive (Appendix A) were compared to detailed measurements developed from field sites in Kansas and Colorado. This information, combined with a comprehensive review of the literature, has been used to develop guidelines for field inspectors in the assess- ment of debris at bridges. Debris accumulates episodically at bridge piers, beginning as one or more logs, as shown in Figure 3.85. Single and multiple individual logs typically do not extend upstream for any signif- icant distance and typically do not present a severe potential for creating additional scour at the pier. However, if left unattended, more logs and branches will be caught and the jam will grow to become a mass of logs. The mass of logs then traps other floating debris, includ- ing relatively small branches, shrubs, twigs, etc. that would otherwise not hang up on a pier. A fully formed debris mass is shown in Figure 3.86 and represents a condition where a significant depth of additional scour at the pier can be expected. 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 during the early stages of accu- mulation will minimize subsequent trapping of additional 102 Figure 3.85. Single- and multiple-log debris accumulations (photographs taken at low flow).

debris that will eventually create a scour hazard; however, this approach is often not practical given limited maintenance resources. Guidance for inspectors includes the following elements: • Inspection of debris conditions will usually consist of sim- ple observations. However, during biennial bridge inspec- tions, whenever possible, a thorough documentation of debris conditions should be made, including photographs and scour measurements. Underwater inspection near or beneath debris masses is extremely hazardous and is not advised, as discussed in the next section. • When single- or multiple-log hangups are observed, a notation should be made in the inspection report and preventive maintenance should be requested. Although this condition does not represent a scour threat, it indi- cates that there is a potential for growth of the debris mass. • The potential for debris recruitment from the stream banks or flood plain areas should be noted. Evidence of debris potential includes tree-lined banks that are under- cut, leaning trees, and deadfall that may be swept into the stream at high flows. Areas previously affected by forest fires, beetle kill, or other circumstances that leave dead woody debris along or near the channel banks should be considered as having high potential for contributing float- ing material to the stream. • If an initial hangup occurs at a bridge with a known history of debris problems, more frequent observations may be warranted, especially during or immediately after storm events or high flows. High flows recruit more floating debris that is likely to become trapped by the initial material. Keep in mind that debris arrives at bridges in bursts. Monitoring a debris mass is discussed in Section 3.10.2. • When a debris mass is observed to have become more fully developed, expedited maintenance for removal should be scheduled. Maintenance is discussed in Section 3.10.3. • Inspectors should also look for evidence of scour not only at the pier or piers where debris is accumulating, but also at adjacent foundation elements. Flow that sheds around a debris mass can exacerbate scour at adjacent piers or abut- ments and can cause erosion and instability of stream banks in the vicinity of the pier. 3.10.2 Monitoring Monitoring a buildup of debris on bridge piers will typi- cally begin with simple visual observations during the initial stages of debris accumulation. Existing debris masses at piers, particularly at bridges with a known history of debris prob- lems, should be monitored more frequently than the biennial cycle, as mentioned in the previous section. During high flows, observation of the entire extent of a debris buildup may be difficult, because the sides and leading edges may be submerged. However, during periods of low or “average” stream flows, the width and length of debris masses can be easily measured from the bridge deck with standard surveying equipment and methods, as described in the fol- lowing paragraphs. A theodolite or clinometer can be used to quickly deter- mine the extent of a debris cluster from a bridge deck by measuring the instrument height above the water and declina- tion angle (θ) to the leading edge of the debris cluster. The horizontal distance from the observer is L = H/tan(θ). The distance should be measured at several radials (recording the azimuthal angle, α) from a position directly above the pier to determine the shape and lateral extent of the debris pile. For a height above the water surface measured to ±1 ft and declination measured to ±1°, the calculated lengths should be accurate to within about 10%. This method can be used by DOTs to determine the extent of debris accumula- tions and is illustrated in Figure 3.87. The thickness of debris is difficult to estimate when the water depth is greater than about 10 ft. During low flows, however, the debris-laden pier may be in shallow enough flow (or even dry, in many cases) such that a direct measurement can be made with a survey rod. Diving around or under debris masses is extremely dangerous because of turbulence and unexpected currents in the vicinity of the debris, poor vis- ibility, and the potential for becoming snagged or other- wise trapped underwater or being struck by a log. For these reasons, underwater inspection of a debris-laden pier is not advised. 103 Figure 3.86. A fully developed debris mass creates significant scour potential.

3.10.3 Maintenance Obviously, removing debris from a pier before it becomes a large, fully developed mass is desirable, but such removal is often impractical from a management and operations stand- point. As discussed in HEC-9 (Bradley et al. 2005), no spe- cific guidelines for maintenance debris removal currently exist, instead “general maintenance practices . . . should involve regular inspections and cleaning, coupled with emer- gency removal of debris.” HEC-9 also suggests that priority should be given to bridges carrying interstate or other primary highways. The frequency of maintenance may be greater at these bridges compared to those carrying secondary highways or other roadways with lower average daily traffic. Obviously, a high priority should be given to bridges that are known to be prone to debris prob- lems. Increased frequency of maintenance at these sites should be considered as well. 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. Considerations for maintenance activities include the following: • Access: It may be necessary to provide an area on one or both banks for mechanical equipment to reach the debris and remove it from the structure, ideally without having to disrupt traffic. Tracked vehicles can often be used after a flood event when the flow is very shallow or the affected pier is dry after flood waters recede. At large bridges with perennial flow, equipment may have to operate from a barge moored near the debris mass. • Debris disposal: It is not acceptable to simply dislodge debris from a bridge and let it float downstream. Nor should debris be moved from the upstream side of a bridge to the downstream side. Debris may be temporarily placed on the banks or overbank flood plain areas, but it must be removed in a timely fashion so that it is not reintroduced to the stream during the next flood. Potential disposal options include using it as firewood or chipped wood or, if it is of high grade, using it for structural purposes. These disposal options are preferable to burying or burning it, as they may provide an opportunity for some financial return. • Countermeasures: Countermeasures to control debris at bridge piers function in one of two ways: either trapping the debris upstream of the bridge or deflecting the debris away from the piers. Countermeasures have met with vary- ing degrees of success and must themselves be inspected and maintained. However, they can serve to minimize the debris loading on the pier itself. Design guidance for debris countermeasures is described in detail in HEC-9. Types of countermeasures include: – Deflectors constructed of steel rails or steel piles filled with concrete placed upstream of the pier (Figure 3.88). – Debris fins consisting of a thin wall made of concrete, steel rails, or timber installed upstream of the pier and aligned with the flow (Figure 3.89). – In-channel debris basins, which are storage basins exca- vated in the stream bed upstream of the bridge. A row of posts to catch and hold floating debris must be included. After flood events, the debris stored in the basin must be removed. – Debris sweepers consisting of a buoyant cylinder mounted on the leading edge of a pier. The sweeper can slide up and down on a vertical metal pole and spins in the cur- rent. When debris encounters the spinning cylinder, it is kicked off to one side (Figure 3.90). 104 Figure 3.87. Survey method for measuring the upstream and lateral extent of debris. H L Profile Li αi θ Plan Flow Unscoured bed Scoured bed

3.11 Implementation Plan 3.11.1 The Product As described in more detail in the preceding sections, the products of this research include practical guidelines for pre- dicting debris hazards at bridges and methods for predicting the depth, shape, and extent of scour at bridge piers resulting from debris accumulations. 3.11.2 The Market The market or audience for the results of this research will be hydraulic engineers and maintenance and inspection per- sonnel in state, federal, and local agencies with a bridge-related responsibility. These would include the following: • State highway agencies • Federal Highway Administration • City/county bridge engineers • Railroad bridge engineers • U.S. Army Corps of Engineers • Bureau of Land Management • National Park Service • Forest Service • Bureau of Indian Affairs • Any other governmental agency with bridges under their jurisdiction • Consultants to the agencies above 3.11.3 Impediments to Implementation A serious impediment to successful implementation of results of this research will be difficulties involved in reaching a diverse audience scattered among numerous agencies and institutions; however, this can be countered by a well-planned technology transfer program. Because of the complexity and geographic scope of the debris-related bridge scour problem and the diversity of bridge foundation geometries, a major challenge was to present the results in a format that can be applied by agencies with varying levels of engineering design capabilities and maintenance resources. Presenting the guidelines and methods in a format familiar to bridge own- ers, who are the target audience, will facilitate their use of the results of this research. 3.11.4 Leadership in Action Through the National Highway Institute (NHI) and its train- ing courses, FHWA has the program in place to reach a diverse 105 Figure 3.88. Debris deflectors made of steel pipes filled with concrete. Source: Bradley et al. (2005) Figure 3.89. Timber debris fins with sloped leading edge. Source: Bradley et al. (2005) Figure 3.90. Spinning-type debris sweeper. Source: Bradley et al. (2005)

and decentralized target audience. For example, recommen- dations from this study could be considered for the next edition of HEC-18, “Evaluating Scour at Bridges,” and NHI Course No. 135046, “Stream Stability and Scour at Highway Bridges.” TRB—through its annual meetings and committee activities, publications such as the Transportation Research Record, and periodic bridge conferences—can also play a leading role in disseminating the results of this research to the tar- get audience. AASHTO is the developer and sanctioning agency for stan- dards, methods, and specifications. Thus, research results can be formally adopted through the AASHTO process. As a col- lective representation of individual state DOTs, AASHTO can also suggest any needed training to be developed by FHWA or others. The AASHTO Subcommittee on Bridges and Structures could provide centralized leadership through the involvement of all state DOT bridge engineers. Regional bridge conferences, such as the Western Bridge Engineer Conference or the International Bridge Engineer- ing Conferences, reach a wide audience of bridge engineers, manufacturers, consultants, and contractors. The groups would have an obvious interest in the effects of debris on bridges and their acceptance of the results of this research will be key to implementation by bridge owners. 3.11.5 Activities for Implementation The activities necessary for successful implementation of the results of this research relate to technology transfer activ- ities, as discussed in the previous section, and the activities of appropriate AASHTO committees. 3.11.6 Criteria for Success The best criteria for judging the success of this implemen- tation plan will be acceptance and use of the guidelines and methodologies that result from this research by state highway agency engineers and others with responsibility for design, maintenance, rehabilitation, or inspection of highway facili- ties. Progress can be gauged by peer reviews of technical pre- sentations and publications and by the reaction of state DOT personnel during presentation of results at NHI courses. A supplemental critique sheet could be used during NHI courses to provide feedback on the applicability of the guidelines and suggestions for improvement. The desirable consequences of this project, when imple- mented, will be more efficient planning, design, maintenance, and inspection of highway facilities considering the threat from debris. The ultimate result will be a reduction in the number of bridge failures and reduction in damage to highway facilities attributable to accumulation of debris on bridge piers. 106