Minimizing Roadway Embankment Damage from Flooding (2016)

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.

39 CHAPTER FIVE GEOTECHNICAL AND GEOLOGICAL FACTORS INTRODUCTION The modes of failure identified in chapter two deal with the interaction between the water and the soil. The water repre- sents the “load” and the soil the “resistance.” The water effect is characterized by the water velocity v (m/s) and the corre- sponding hydraulic shear stress τ (N/m2). The soil resistance to erosion is characterized by the relationship between the erosion rate (mm/h) on one hand and the water velocity v or shear stress τ on the other (Figure 65). This relation- ship is called the erosion function of the soil; it represents the fundamental behavior of the soil much like the stress strain curve represents the fundamental behavior of the soil when subjected to mechanical loading. The erosion function is nonlinear and can be measured by erosion testing in the laboratory or in the field. Some of the erosion tests available are described later in this chapter. FIGURE 65 Erosion functions for a sand and for a clay. GEOTECHNICAL CONSIDERATIONS Critical Velocity As discussed previously, the erodibility of a soil is described by the erosion function. The erosion process does not start until the velocity is large enough to initiate erosion of the soil particles. This threshold of erodibility is called the critical velocity vc and plays a very important role in erosion engi- neering. Indeed, if this velocity can be precisely determined, any lower velocity will not create erosion. Figure 66 shows a relationship between the critical velocity and the mean grain size D50, which is the soil grain size for which 50% by weight of the soil grains is larger than D50. As Figure 66 illustrates, for sands and gravels there a good relationship between vc and D50 (Briaud 2013). The following relations are derived in SI units. If needed, the values in metric units can be cal- culated using the conversions table. FIGURE 66 Critical velocity and mean grain size. For sand and gravel 0.45 50( / ) 0.35 ( )cv m s D mm (5.1) However, this relationship breaks down for silts and clays because the behavior of the particles of sand and gravel is controlled by their weight, which is not the case for silts and clays. For silts and clays, the electrostatic and electromag- netic forces that exist between such fine particles become more important than the gravity force or weight of the par- ticle. As a result, the relationship with grain size is scattered and the critical velocity can only be measured on a site-spe- cific basis at this time. Upper and lower bounds can be used to bracket the critical velocity, as shown in Figure 66. Upper bound for silt and clay 1 50( / ) 0.03 ( )cv m s D mm (5.2) Lower bound for silt an clay 0.2 50( / ) 0.1 ( )cv m s D mm (5.3)

40 Erosion Function and Soil Classification The most effective way to obtain the erosion function is to measure it on a site-specific basis by testing samples or by in situ testing. If this preferred approach is not pos- sible, the erosion function can be estimated on the basis of the Unified Soil Classification System by using the erosion category chart shown in Figure 67 (Briaud 2013). To use this chart, the soil that will be eroded is classified accord- ing to the Unified Soil Classification System to obtain the dual symbol classification indicated in the chart, and the straight line that splits the erosion category zone is selected as an average for that soil. The lower bound or upper bound of that category can be used to be conservative, depending on the erosion problem. FIGURE 67 Erosion function and soil classification (Briaud 2013). Example Process for Calculating Erosion Depth The following process shows how the erosion function can be used for simple calculations by hand to estimate the ero- sion depth. The steps are listed as follows: 1. Obtain the erosion function for the soil that may be eroded by the flowing water. This can be done by ero- sion testing (Figure 65) or by using the erosion cat- egory chart based on soil classification. 2. Obtain the velocity hydrograph for the flood to be con- sidered. This is the relationship between the velocity of the water and time. Riverine floods usually last a few days. However, a hurricane that makes landfall may create a surge and associated overtopping that lasts only a couple of hours. 3. Decompose the velocity hydrograph into a series of constant velocity steps. 4. For each constant velocity step, find the erosion rate i (mm/h) that corresponds to the velocity vi on the erosion function. 5. Using that erosion rate, find out how much erosion will take place during the time increment correspond- ing to the velocity step. i i iz z t (5.4) 6. Repeat steps 4 and 5 for all subsequent velocity time steps and calculate all the erosion depth increments corresponding to all velocity time steps. 7. The final erosion depth Z is: i i iZ z z t (5.5) The following is an example of this simple step-by-step procedure applied to embankment overtopping. Figure 68a shows an embankment being overtopped during a major flood. The erosion function measured in an erosion test performed on a sample from the embankment is given in Figure 68b; it shows that the soil is very erosion resistant as the critical velocity is 4 m/s (13.12 ft/s). The velocity hydrograph at the bottom of the downstream side of the embankment where the water goes the fastest is given in Figure 68c. It shows that the flood is lasting about 2.5 days, with velocities reaching 12 m/s (39.37 ft/s) at the flood’s peak. The calculations proceed according to the steps described previously. The smooth velocity hydrograph is discretized into time increments of 3 hours each. For each increment, the veloc- ity is constant and as long as the velocity is lower than the critical velocity of 4 m/s (13.12 ft/s), no erosion occurs. For each velocity step, the calculations are simple. For example, during the fifth increment beyond the start of the erosion process, the velocity is 10 m/s (32.81 ft/s) for 3 hours. The erosion function gives an erosion rate of 55 mm/h (2.57 in./h) for a velocity of 10 m/s (32.81 ft/s). Therefore, the ero- sion depth during these 3 hours becomes larger by 55 mm/h x 3 h = 165 mm (6.5 in.). By following this reasoning from the beginning of the hydrograph all the way to the end, the erosion depth can be calculated and is shown in Figure 68d as 1.8 m (5.91 ft). Erosion Tests A soil’s erodibility can be tested using many different labo- ratory and in situ tests, which have been developed at an increasing rate mostly over the past 25 years. The laboratory erosion tests include the following: • Jet Erosion Test (JET) • Hole Erosion Test (HET) and the previous version of it, the Pin Hole test

41 • Rotating Cylinder Test (RCT) • Erosion Function Apparatus (EFA) tests and similar devices (e.g., Sedflume). FIGURE 68 Example of the erosion depth calculation. Jet Erosion Test (JET) The JET is a laboratory test that can be credited to Greg Hanson of the U.S. Department of Agriculture (Hanson and Cook 2004; USSD 2011). The JET (Figure 69) can be per- formed in the laboratory on a soil sample or in the field on the ground surface. It consists of directing a stationary water jet at a given velocity perpendicular to the soil surface and recording the depth of the hole made by the jet as a func- tion of time to obtain an erosion rate. The result of the test consists of a curve linking the depth of the hole to the time of jetting. Hanson used a linear relationship to describe the data linking the erosion rate to the net shear stress above critical and called the slope of that line KD, which he named the erosion coefficient: D cz K (5.6) He went on to classify soils according to their KD value, as shown in Figure 70. FIGURE 69 Jet erosion test (Hanson and Cook 2004). Hole Erosion Test (HET) The HET is a laboratory erosion test that evolved from the older Pin Hole test and can be credited to Robin Fell in Australia (Lefebvre et al. 1984; Wan and Fell 2004; Wahl 2009; Benahmed and Bonelli 2012). The test (Fig- ure 71) consists of drilling a 6-mm (0.24-in.) diameter hole through a soil sample and forcing water to f low through the hole at a chosen velocity while recording the increase in diameter of the hole as a function of time to obtain an erosion rate. The test results link the mass ero- sion rate to the net shear stress above critical. The equa- tion used is linear: e cm C (5.7) The parameter Ce is called the erosion coefficient. The erosion rate index is then defined as 10log ( / )HET eI C s m (5.8) Wan and Fell went on to propose some erosion categories based on IHET (Figure 72).

42 FIGURE 71 Hole erosion test. FIGURE 72 Hole erosion test (continued). Rotating Cylinder Test The Rotating Cylinder Test is a laboratory erosion test that has been in existence for a longer time than other tests (Moore and Masch 1962; Chapuis and Gatien 1986; Hender- son 1999; Kerr 2001; Sheppard et al. 2005; Bloomquist et al. 2012). It consists of placing a soil sample in a chamber filled with water and rotating the chamber at a speed sufficient to entrain the water up to a chosen velocity. The water erodes the sample, and the decrease in the weight of the sample ver- sus time gives the average erosion rate. EFA Test and Similar Devices The EFA test developed in 1991 (Figure 73) and the devel- opment of associated bridge scour design guidelines can be credited to Briaud (Briaud et al. 2001; Briaud 2013). Others have also worked on this type of device (McNeil, Taylor, and Lick 1996; Roberts et al. 2003; Crowley et al. 2012). The EFA test is a laboratory erosion test that consists of pushing a soil sample through the bottom of a conduit only as fast as the water flowing over it is eroding it. The erosion rate cor- responds to the rate at which the piston is pushing the sample upward. It is recorded for each velocity and gives the erosion function point by point (erosion rate versus velocity or shear stress curve). The equations used for the erosion function in this case are b cz a v v (5.9) cz (5.10) FIGURE 70 Jet erosion test: Hanson’s classification according to the erosion coefficient.

43 Based on many tests performed over the past 25 years, Briaud proposed an erosion classification and related the soil classification to the erosion classification (Figure 67). FIGURE 73 EFA test (Briaud 2001). The field erosion tests are • In Situ Scour Testing Device (ISTD) from FHWA • In Situ Scour Evaluation Probe (ISEP) from North Carolina State University • Borehole Erosion Test (BET) from Texas A&M University • Pocket Erodometer Test (PET) from Texas A&M University. In Situ Scour Testing Device The ISTD is an in situ erosion device being developed by FHWA at the Turner–Fairbanks Highway Research Center. It is a reverse jet test where the water is sucked upward from the outside to the inside of a vertical pipe placed in a borehole. In this process, the tool erodes the bottom of the borehole. In Situ Scour Evaluation Probe The ISEP test uses a jetting probe that penetrates into the soil under its own weight at a recorded erosion rate for a given jet velocity. It is credited to Gabr (Gabr et al. 2013; Caruso and Gabr 2010). Gabr follows the work of Hanson and uses the equation D cz K (5.11) Pocket Erodometer Test The PET (Figure 74) was proposed by Briaud, Bernhardt, and Leclair (2012) as a very simple portable device to test the sur- face of a sample in the field as the sample is extruded or in the lab before more advanced erosion testing is performed. It con- sists of using a repeated water jet impulse at 8 m/s (26.25 ft/s) aimed perpendicular to the sample surface, placed 50 mm from that surface and applying 20 repetitions of that jet. Once the test is completed, the depth of the hole generated in the sample surface is measured. Briaud et al. (2013) merged the results of the PET to their erosion classification by adding the depth of the PET hole on the erosion classification chart (Figure 75). FIGURE 74 Pocket erodometer test (Briaud et al. 2012). FIGURE 75 Pocket erodometer test chart (Briaud 2013). Borehole Erosion Test Briaud (Briaud et al. 2014) developed the BET. This in situ erosion test (Figure 76) consists of recording the increase in a borehole’s diameter as the water circulates in it. The test is performed by drilling the borehole to a chosen depth, removing the drilling tool and rods, lowering the diameter measuring tool (borehole caliper) to get a zero diameter reading profile, lower- ing the rods and bit into the open hole, circulating the water at a chosen velocity for 10 minutes, removing the rods, and lower- ing the borehole caliper again to take a second diameter mea- surement. The difference in radius divided by the time of flow gives the erosion rate of all the layers within the borehole depth for the chosen velocity (and, therefore, shear stress). The test is repeated at different velocities and point by point the erosion function of each layer is obtained in one test. FIGURE 76 Borehole erosion test (Briaud et al. 2014). GEOLOGICAL CONSIDERATIONS Geological considerations are an important part of a roadway embankment design to minimize flooding damage. Engineer-

44 ing geology brings a big-picture understanding of the sur- roundings of an embankment and river setting. Within the engineering geology framework, it is necessary to understand the geomorphology of rivers and the impact it has on nearby embankments. Throughout the estimation of the hydrologic input parameters, factors related to drainage basin, stream channel, and flood plain characteristics become essential. The discharge that determines the overtopping height and duration is controlled by the river cross section in addition to meteoro- logical characteristics. The flow velocity is an important design parameter that is controlled by the slope and basin roughness. Other processes affect the damage during flooding. Some of these factors include meandering potential in riverine environ- ments, and erosion and deposition processes in the coastal envi- ronments. Those two processes are important considerations in deciding on the location of a roadway embankment, or the recommended measures to mitigate damage. Meandering Potential Rivers are active systems; meanders can move laterally sev- eral meters per year. This lateral migration of the main channel affects bridges, embankments, and other structures that strad- dle the river. It is important to predict future meander move- ments and to design remedial measures or move the structure. An example of meandering is shown in Figure 77 where the average meander rate is 4m/year. Many have contributed to the advancement of knowledge in this field, including Brice (1974), Hickin and Nanson (1984), Hooke (1984), Lagasse et al. (2001), and de Moor et al. (2007). FIGURE 77 Measured migration of the meander over a 25-year period, the Brazos Meander Case History (after Briaud 2013). Briaud et al. (2007) developed the MEANDER method to predict the movement of a meander over time. First, the initial geometry of the river is described by fitting circles to the meander bends and placing straight-line tangents to the circles between circles. Second, the erosion function of the river banks is input. This can be done by using the results of EFA tests or by using the erosion classification charts of Fig- ure 67 adjusted for the presence of vegetation, trees, or other erosion-retarding layers. Third, the velocity hydrograph is input from measurements at a nearby gage station. Fourth, the circles describing the meanders are moved according to erosion rules developed through a series of very large-scale laboratory meander experiments (in sand and then in clay) as well as numerical simulations (Briaud 2013). This leads to a prediction of the location of the river after the period of time corresponding to the hydrograph. Erosion and Deposition As a result of over-wash, erosion and deposition environments can form on barrier islands. As shown in Figure 78, erosion naturally occurs from the ocean side of an island and deposi- tion occurs on the land side. The frontal dunes generally have the highest elevations on the island. In a storm event, these dunes are eroded or possibly overtopped, and the eroded sand can be pushed back through the island into the bay area. The coastal embankment could lie within the erosion zone or the deposition zone. If it is located in the erosion zone, it would be subjected to extreme wave attack; con- sequently, it could suffer more damage, depending on the available protection means. On the other hand, if a coastal embankment is located within the deposition environments, it would be found undamaged under a layer of sand that could be scraped off. DESIGN CONSIDERATIONS FOR FAILURE MODES The design of flood-prone roadway embankments includes the consideration of the different failure modes explained in FIGURE 78 Sketch for sand erosion and deposition processes on a barrier island resulting from overwash (Douglas and Krolak 2008).

45 chapter two. Design considerations for failure modes result- ing from overtopping and wave action were considered in previous chapters—within the limits of existing knowledge and practice. This section covers geotechnical considerations to minimize damage from through-seepage, underseepage, softening by saturation (rapid drawdown condition), over- topping, and lateral sliding. Through-seepage and under- seepage may cause internal erosion and piping, which would lead to failure. Internal erosion and piping depend on the material characteristics. Softening by saturation results from seepage through the embankment. Lateral sliding could be caused by softening of the embankment-soil interface and consequent loss of friction. This could also result from dam- age caused by other processes on the downslope and toe of the embankment, which would decrease the embankment stability. The discussion of the geotechnical aspect of these failure modes is carried out in light of knowledge pertinent to dams and levees. Internal Erosion Phenomenon Internal erosion is the reason behind an estimated 46% of earth dam failures, half of which occur during the first filling of the reservoir (Fell and Fry 2005). In general, through-seep- age becomes a concern when the embankment is made up of soils subject to internal erosion. The handling of the internal erosion phenomenon is still based primarily on engineering judgment and experience. Although guidelines and publica- tions exist, much remains to be studied in this field. For internal erosion to occur, the following conditions are required: 1. A seepage path and a source of water. 2. Erodible material that can be carried by the seepage flow within the flow path. 3. An unprotected exit from which the eroded material may escape. 4. For a pipe to form, the material must be able to form and support the roof of the pipe. Four different phenomena can lead to internal erosion of an embankment, as shown in Figure 79: 1. Backward erosion: initiated at the exit point of the seep- age path when the hydraulic gradient is too high and the erosion gradually progresses backward, forming a pipe. 2. Concentrated leak: internal to the soil mass; it initi- ates a crack or a soft zone emanating from the source of water and may or may not progress to an exit point. Erosion gradually continues and can create a pipe or a sinkhole. 3. Suffusion: develops when the fine particles of the soil wash out or erode through the voids formed by the coarser particles. This occurs when the amount of fine particles is smaller than the void space between the coarse particles. If, in contrast, the soil has a well- graded particle size distribution with sufficiently small voids, suffusion is unlikely. Soils are called internally unstable if suffusion takes place and internally stable if particles are not eroding under seepage flow. 4. Soil contact erosion: sheet flow at interfaces between soil types. This may occur, for example, when water seeps down the back face of the core at the interface with the filter and then the stabilizing mass. FIGURE 79 Mechanisms of internal erosion failures (after Perzlmaier 2005). Soils Mostly Susceptible to Internal Erosion Coarse silt and fine sand are among the most erodible soils. Therefore, embankments that contain significant amounts of such materials will be more prone to internal erosion. Clays in general, and high-plasticity clays in particular, are more resistant to erosion as long as the electrical bonds between particles are not destroyed by chemicals. It appears that some core materials of glacial origin, such as glacial tills, can be particularly susceptible to internal erosion. Sherard (1979) gives a range of gradation of soils that can lead to internal erosion problems (Figure 80). FIGURE 80 Range of problem soils for internal erosion (after Sherard 1979). The soils most susceptible to suffusion are those where the volume of fines is less than the volume of the voids

46 between coarse particles. In this case, the fines can move easily between the coarse particles and erode away to an exit face. After suffusion, such soils are devoid of fines and become very pervious clean gravel, for example. Fell and Fry (2005) indicate that gap-graded soils and coarsely graded soils with a flat tail of fines (Figure 81) are most sus- ceptible to suffusion. FIGURE 81 Range of problem soils for suffusion (after Fell and Fry 2005). Criterion to Evaluate Internal Erosion Potential One of the important criteria for evaluating erosion is to cal- culate the hydraulic gradient and compare it to the critical gradient. The critical gradient is given by sat w cr w i (5.12) Values of icr typically vary in the range of 0.85 to 1.2. In dam applications, the hydraulic gradient depends on many factors including the difference in water level between the upstream and the downstream, the length of the drainage path, and the relative hydraulic conductivity of the various zones. To avoid internal erosion, the target maximum gradient in the flow is kept much lower than the critical value, especially in areas where internal erosion is possible. For unfiltered exit faces, figure 82 shows ranges of hydraulic gradient values that are associated with the ini- tiation of internal erosion on one hand, and the full devel- opment of piping on the other. Generally speaking, there is a trend toward higher-porosity soils beginning to erode at lower hydraulic gradients, even lower than 0.3. Yet soils with plastic fines erode at higher gradients, and gap-graded soils begin to erode at lower gradients than nongap-graded soils with the same fine content. FIGURE 82 Range of hydraulic gradient values associated with internal erosion (after Perzlmaier 2005). For levee applications, USACE uses a lower-bound value of the critical hydraulic gradient equal to 0.8 and allows a hydraulic gradient of up to 0.5 at the toe of levees, provided a number of conditions are met (USACE 2003a). Another way to address the incipient motion of soil particles in internal erosion problems is to use the concept of critical velocity and charts such as Figure 67. However, these critical velocities were developed from sheet flow tests, and the critical veloc- ity may differ from those initiating internal erosion. Several methods, based in part on the analysis of the grain size curve, have been developed to evaluate the instability of soils in dams and their sensitivity to the suffusion phenom- enon. They include Sherard (1979), Kenney and Lau (1986), Burenkova (1993), and Fell and Fry (2005). Through-Seepage Seepage through the embankment, termed through-seepage, leads to a wet downstream slope, as shown in Figure 83. This would weaken the embankment and compromise its stabil- ity. Other schemes include the development of internal ero- sion mechanisms as described in the previous section. To alleviate the problems imposed by the hydrostatic forces,

47 drainage elements can be introduced into the embankment (USACE 2000). FIGURE 83 Homogeneous embankment section on impervious foundations with emerging seepage on the landside slope (after USACE 2000). For example, a pervious toe can be integrated into the system, as shown in Figure 84. If both through-seepage and underseepage are anticipated, a combined solution of a pervious toe and a partially penetrating toe trench can be adopted (Figure 85). Other horizontal and inclined drainage options are presented in Figure 86. FIGURE 84 Homogeneous embankment section with pervious toe (after USACE 2000). FIGURE 85 Homogeneous embankment section with pervious toe combined with partially penetrating toe trench (after USACE 2000). Underseepage Underseepage is a problem generally associated with pervi- ous foundation material. It would impose a severe problem in the following two cases (USACE 2000): 1. A connected pervious substratum underlies an embankment and stretches both upstream and downstream. Erosion could develop within the foundation materials until a void forms under the embankment. 2. A relatively impervious thin-top stratum is present on the downstream side. This could lead to excessive hydrostatic pressures on the downstream side in the underlying, more pervious stratum. Seepage control solutions in earth foundations are pro- vided in USACE EM 1110-2-1913, EM 1110-2-1901, and EM 1110-2-1914. The methods are namely cutoff walls placed beneath the embankment, riverside blankets, downstream seepage berms (Figure 87), pervious toe trenches (Figure 88), and pressure relief wells. Such measures would alleviate the severity of the seepage conditions. FIGURE 87 Installation of a berm while considering foundation material conditions (USACE 2000). FIGURE 88 Typical partially penetrating pervious toe trench (USACE 2000). Softening by Saturation (Rapid Drawdown Condition) Softening by saturation generally occurs after the embank- ment becomes saturated owing to prolonged exposure to water. This water could come from precipitation, the ris- ing headwater or riverside water level, or inundation dur- ing overtopping. The extent of saturation depends on the embankment’s permeability and the duration of precipita- tion, rise in flood level, or inundation. This is accompa- nied by the embankment’s loss of strength and is generally characterized by slope failures and pavement deterioration resulting from the saturation of the subgrade. FIGURE 86 Horizontal and inclined drainage layer options (after USACE 2000).

48 Rapid drawdown is associated with a prolonged flood stage that saturates—by seepage—a major part of the upstream portion of the embankment. The floodwaters then recede faster than the embankment soil can drain, thus resulting in potential instability. For analysis of rapid draw- down, the effective shear parameters are used as recom- mended by USACE (2000). Two preferred procedures for rapid drawdown analysis are presented in USACE’s Slope Stability Manual (2003b). To increase stability and prevent slope failure, flatter slopes could be adopted, as well as sta- bility berms. Stability berms also work as a good option in case of emergency to stop any further movement. Lateral Sliding on Foundations Figure 89 shows a simplified model for lateral sliding cal- culations. For sliding stability, the applied force from the upstream water is kept less than the maximum resisting force developed at the embankment-foundation interface by a certain safety factor. The driving force or push P in kilone- wtons/meter (kN/m) of embankment length is 21 2 P H (5.13) FIGURE 89 Sketch for lateral sliding calculations. Where γ is the soil unit weight and H is the water depth or, more precisely, the difference in water depth between the two sides of the embankment. The maximum resistance R per unit length of embankment is ' tanR W (5.14) Where W is the embankment effective weight per unit length of embankment and φ is the effective angle of friction of the interface at the bottom of the embankment. The factor of safety FS against sliding is RFS P (5.15) Overtopping of Embankments This section presents erosion concepts that are applicable to levees or dikes, and could be applied to roadway embank- ments. Levees are small dams built along a river or an ocean to prevent the water from inundating the land during floods. The top of the levee is set at a predetermined height corre- sponding to the water level for a chosen design flood. This flood corresponds to a certain return period, such as a 100- year flood. Similar to the case of the roadway embankments, if the flood exceeds the design return period, water is likely to flow over the levee and generate potential erosion. One of the first observations is that if the water flows above a levee of height h (also applicable to an embankment), by the time the water reaches the bottom of the dry side of the levee it will have a velocity v, which can be very high. One simple way to evaluate that velocity is to write conservation of energy: 21 2 2 mgh mv or v gh (5.16) Where g is the acceleration resulting from gravity. For example, if the levee is 5 m high (16.4 ft), the velocity v will be approximately 10 m/s (32.8 ft/s). Of course, Equation 5.15 does not take into account the energy lost in friction between the water and the levee surface, but it does indicate that the velocity range is much higher than typically encountered in rivers, where peak velocities range from 3 to 4 m/s (9.84 to 13.12 ft/s). Furthermore, a distinction is made between events such as hurricanes and river floods—the major dis- tinction being that hurricanes may overtop a levee for about 2 hours, while river floods may overtop a levee for 2 days. A levee-overtopping erosion chart has been developed for these two types of events and is presented in Figure 90, which could be useful for embankments. This chart is based on extensive work by Briaud and his co-workers in the aftermaths of both Hurricane Katrina in New Orleans and the Midwest floods of 2008. By combining Figures 90 and 91 (also Figure 67), one can get a sense of which soils, soil categories, and associated erosion functions are likely to resist overtopping during a 2-hour or 2-day overtopping event. Recall that Categories I to IV on the erosion chart are soils and Categories V and VI are rocks. As shown in the charts, only the most erosion-resistant soils can resist 2 hours of overtopping without protection (Category IV), and no soil can sustain 2 days of overtopping without being totally eroded away. Armoring or vegetation that satisfy strict criteria need to be used to ensure that overtopping can be sustained for longer than 2 hours. FIGURE 90 Overtopping erosion chart (Briaud 2013).

49 FIGURE 91 Erosion function and soil classification (Briaud 2013). Vegetation can help significantly to slow down erosion. To be effective, though, this vegetation must satisfy the fol- lowing minimum requirements: 1. have a mat-like appearance; 2. have a sod-forming root system; 3. be made of perennial grasses; 4. have a dense, consistent coverage; and 5. have a minimum height of 0.3 m during flood season. Figure 92 shows an embankment that, according to the farmer, was overtopped for 2 days during the June 2008 Mid- west Mississippi River flood; it resisted very well because the grass cover met the criteria discussed earlier. Figure 93 shows a weak grass cover on a levee that was overtopped and breached. Figure 94 shows an earth embankment that survived well after 2 hours of overtopping during Hurricane Katrina because the soil was quite erosion resistant (Cat- egory IV, Figures 90 and 91 in August 2005) while Figure 95 shows an earth embankment that was totally eroded during the same event (Category I, Figures 90 and 91). FIGURE 92 Good-quality grass cover to delay erosion during overtopping (Briaud 2013). FIGURE 93 Poor-quality grass cover unable to delay erosion during overtopping (Briaud 2013). FIGURE 94 Overtopped earth embankment that survived well after 2 hours of hurricane surge overtopping (Briaud 2008). FIGURE 95 Overtopped earth embankment totally eroded by 2 hours of hurricane surge overtopping (Briaud 2008).

50 Tree roots can be considered as reinforcement for the slope of a levee if the tree is on the levee and is alive and healthy. However, if the tree on a levee is uprooted and top- ples over during a flood, it will create a major hole in the levee. Also, if the tree dies, the disappearance of the roots will leave channels for the water to seep through the levee. On the whole, it is best to not let trees grow on or near levees. Culvert-Related Problems Based on the case examples presented and available infor- mation, problems associated with culverts are common in flooding events. Erosion is often concentrated at points of discontinuity in a soil mass. This is the case of the interface between a soil embankment and a culvert where internal erosion can be expected. Aggravated damage from erosion at the locations of culverts and overtopping that might have occurred as a result of blocked or undersized culverts are common as revealed by the case examples in chapter three. Based on Iowa’s experience with Western Iowa Missouri River Flooding (IHRP TR-638 by Vennapusa et al. 2013), culvert-related problems during flooding include erosion of culvert backfill, separation of culverts, and blockage of water outflow. An example is presented in Figure 96 on Highway 18 (North Dakota). The embankment was over- topped and breached, pavement was damaged, and the cul- vert was washed out. FIGURE 96 Pavement damage, washed-out culvert, and embankment breach in 2009 overtopping during spring thaw, Highway 18, North Dakota. Most DOTs have unique installation methodologies for culverts. The methodologies can include drawings or sec- tions that describe the bedding materials and the placement of the culverts. A number of procedures are included in rel- evant literature and are currently adopted by some DOTs. Such procedures are related to limiting the scour at culvert inlet and outlet, the allowable headwater behind the culvert, and the culvert bedding. Common scour problems include scour holes and scouring away of the embankment slope materials in the vicinity of the inlet as well as scour at the outlet. This usually occurs as the water velocity increases upon entering the inlet or as it carves its path outside the cul- vert. Structural problems are related to culvert integrity and proper selection and placement of bedding and fill materials. Available practical solutions are outlined in chapter nine. According to Bonelli (2013), the presence of a culvert can facilitate the initiation of internal erosion in many ways. Sherard et al. (1972a) and Charles (1997) discuss how a stiff conduit would cause unusual stress distributions when embedded in less-stiff surrounding soil. Also, the drying of surrounding soil could open some cracks that in turn would lead to initiation of piping during flooding. Another factor is that the compaction of the soil in the immediate vicinity of a culvert is often more difficult and less effective. Guidance on relevant most effective practice can be found in FEMA (2006) and guidance on how to assess the likelihood of ini- tiation of erosion around these features is included in Fell et al. (2008). Pavement Degradation and Failure Pavement degradation and failures (loss or erosion) are com- mon after and during flooding. Pavement failure comes in different modes, including rafting, edge failure, and post- flood damage (damage that develops after the flood) result- ing from the saturation of the pavement and the subgrade. The methodologies to minimize pavement damage during flooding vary among DOTs. Whereas some states specify some measures for this purpose, others do not. This section summarized the failure modes addressed throughout this report. Practical measures adopted to minimize the damage are included in chapter nine. Rafting During flooding, pavement failure can occur from rafting, as explained in chapter two. The floodwater can seep into the subgrade and when the uplift pressure becomes large enough, the pavement is carried away by the flow. An exam- ple of this phenomenon is presented in Figure 97. Edge Failure Another damage mechanism is partial or complete loss of pavement from erosion or loss of the embankment soil sup- port when the upstream slope (seaward slope) or downstream slope (landward slope) is subjected to erosion. Such modes would usually occur in three cases, as explained in chapter two: (1) erosion of the upstream slope (seaward slope) from wave action or toe erosion of the embankment stretching in the vicinity of a stream, (2) erosion of the seaward slope as the water recedes (generally common in coastal environ- ments), and (3) overtopping by surge or waves or the com- bination of both, which undermines the pavement on the

51 downstream (landward) slope. This mode of failure is com- mon among case examples mentioned in chapter two as well as damage reported by IOWA TR-368. Figure 98 shows a graph of embankment pavement losses prepared by Schnei- der and Wilson (1980). FIGURE 97 Pavement rafting damage after Hurricane Isabel in 2003, Kimsey Run Project, West Virginia. FIGURE 98 Embankment pavement losses (after Schneider and Wilson 1980). Post-Flood Damage After a major flood, there may be no visible pavement damage, but the engineer may wonder if some hidden damage exists. Three case examples are presented to address this issue. The discussion in this section is limited to hot mix asphalt (HMA) pavement generally used in roadway highways. Additional information relevant to saturation of the pavement and recov- ery of strength as a function of time is included in IOWA TR-638 (Vennapusa et al. 2013). Based on the latter, voids were obtained at shallow and deep depths in the aftermath of Western Iowa Missouri River Flooding in 2012. At shallow depths [<150 mm (6 in.)], voids occurred due to erosion of underlying base materials. At deep depths [>150 mm (6 in.)], voids formed due to erosion of subsurface materials. Case Example 1: TS-1 (Old Mormon Bridge Road), Western Iowa Missouri River Flooding A research team was assembled right after the Western Iowa River flooding to assess the geo-infrastructure damage, repair, and mitigation strategies. Vennapusa et al. (2013) document their experience with one secondary highway pavement stretch of TS-1 (Old Mormon Bridge Road) that they tested in the course of assessing the pavement condition after flooding. The pavement was 360-mm (14-in.) thick HMA under- lain by 300 mm (12 in.) of thick base over natural subgrade. This stretch was partially submerged for about 2 months during the 2011 flooding event. The goal was to evaluate the pavement performance two times: after the flooding receded and about 8 months later. For this purpose, in situ testing using falling weight deflectometer (FWD), dynamic cone penetrometer, and ground-penetrating radar was carried out in addition to hand-augured soil borings. The key findings based on the test results are as follows: • The pavements did not show structural failure, yet granular shoulder erosion occurred close to high water level lines. • EFWD and ESG were 1.3 to 1.4 times lower in flooded areas than in non-flooded zones. FWD results obtained 6 months after the flood were higher, on average, than the initial values, and those taken about 9 months after the flooding almost matched those of non-flooding areas. • California Bearing Ratio results for the base layer were almost matching (greater than 50) for flooding and non-flooding sites, unlike the subgrade, which was, on average, about 10 times lower in flooded zones. Measurements taken shortly after and 9 months after these initial measurement were about the same. Damage basically occurred in the form of voids from the erosion of base and subsurface materials.

52 Case Example 2: SH-24 North of Washington, Oklahoma/McClain County Christopher (2007) carried out FWD at this site after the flood. It was concluded that flood durations of 8 to 14 hours were not long enough to cause significant pavement damage. A “slight weakness” was initially observed in the subgrade that was later recovered. The importance of the site geology was further noted in understanding the potential damage to highways. For instance, subgrade made of sandy soils would become severely weak in comparison with clayey soils. Case Example 3: Pavement Structures Damage/ Hurricane Katrina Flooding Zhang et al. (2008) reported a study carried out by Louisiana Transportation Research Center to assess the impact of Hur- ricane Katrina on pavements in the region. This study basi- cally concluded that HMA and HMA subgrade layers were significantly weakened by the flooding. It was further noted that more damage was experienced by HMA pavements at lower elevations and by those with less thickness. Relation Between Overtopping Depth and Pavement Loss In 1980, FHWA collected data from highway agencies including Schneider and Wilson’s work. The data were mainly based on observations of pavement and embank- ment damage resulting from overtopping. The goal was to develop a relationship between overtopping depth and loss in pavement and embankment. Figure 98 shows the cumulative effects of overtopping over time based on these data (Chen and Anderson 1987). The embankment test data collected by Chen and Ander- son (1987) is also shown on Figure 98 for Type I and Type II embankment soil to be compared with the 20-hour curve. Type I (Clay with low plasticity-CL/USC Unified Soil Clas- sification) shows general agreement with the curve while that of Type II (SM-SC/USC) with freefall condition shows higher erosion rates than the curve. SUMMARY This chapter presented geotechnical and geological con- siderations in minimizing roadway embankment damage from flooding. Geotechnical concepts including critical velocity, erosion function, and the erodibility as a func- tion of soil classification were explained. Available in situ and lab erosion tests were briefly described. Related geo- logical factors include the meandering potential in river- ine environments and erosion and deposition in coastal environments. Geotechnical design considerations were then presented for each of the failure modes highlighted in chapter two. Aside from the hydraulic and hydrological aspects presented in chapter four, and the geotechnical and geological aspects presented in this chapter, nontechnical aspects are presented in the next chapter that include legal, regulatory, and funding aspects.