Aquatic Organism Passage Design Guidelines for Culverts


by Roger Kilgore, P.E., D.WRE1

1Kilgore Consulting and Management, 2963 Ash Street, Denver, CO 80207;             (303) 333-1408      ;

Waterway crossings, including bridges and culverts, represent a key element in our overall transportation system. Traditionally, the design of crossing structures has focused on hydraulic conveyance and flood capacity as the main design objectives. This emphasis has resulted in culvert designs able to perform well during floods, ensuring the efficient conveyance of large volumes of water, but this approach often did not include provisions for the passage of fish and other aquatic organisms through the culvert.

Figure 1. Changes in Fish Habitat Use Over Time after Roadway Fragmentation.

Culvert installations can be a barrier to aquatic organism movement between habitat areas in a number of ways including by presenting high flow velocities, shallow flow depths, or high jump requirements. Figure 1 depicts the possible results of barrier culverts on fish populations: (a) undisturbed habitat, with fill representing habitat in use; (b) habitat with ineffective culverts causing fragmentation with fill colors representing disconnected habitats; (c) fragmented system after a few years, areas with no fill represent population extirpation; (d) fragmented system after many years. In the undisturbed case, fish are free to use the entire stream system as habitat. After a road interrupts stream continuity, fragmented populations are forced to survive independently. In a short time frame, this interruption in continuity increases the susceptibility of smaller populations to elimination by chance events (Farhig and Merriam, 1985). Over the long-term, genetic homogeneity and natural disturbances are also likely to destroy larger populations (Jackson 2003).

The Federal Highway Administration (FHWA) has developed a procedure to design culverts to facilitate aquatic organism passage (AOP) (Kilgore, et al., 2010). The procedure is multidisciplinary making use of skills in aquatic biology, geomorphology, hydrology, sediment transport, hydraulic engineering, and geotechnical engineering. The degree to which any of these disciplines are required at a particular site will depend on the specific ecological and engineering conditions, constraints, and objectives.

Given the diverse behavior and capabilities of fish and other aquatic organisms it is highly desirable for a design procedure to be independent of those diverse behaviors. This has given rise to class of procedures known as stream simulation. The United States Forest Service (USFS) documented their stream simulation approach in 2008 (FSSSWG, 2008). The FHWA approach is also independent of aquatic organism behavior, but uses an alternative surrogate parameter.

To design for passage without focusing attention on specific organism behaviors, design procedures necessarily rely on surrogate parameters and indicators as measures for successful passage. Many of the existing AOP design procedures, most notably the USFS procedure, rely on dimensional characteristics of the stream such as bankfull width. There are several limitations to the use of dimensional stream characteristics as a surrogate: 1) bankfull width can be difficult to identify in the field, 2) bankfull width can be highly variable within a stream reach, 3) field measurement of bankfull width assumes the stream is in dynamic equilibrium, and 4) bankfull width has no known relationship to passage requirements.

The FHWA procedure uses streambed sediment behavior as its surrogate parameter. The hypothesis of using sediment behavior as a surrogate parameter is that aquatic organisms in the stream are exposed to similar forces and stresses experienced by the streambed material. The design goal is to provide a stream crossing that has an equivalent effect, over a range of stream flows, on the streambed material within the culvert compared with the streambed material upstream and downstream of the culvert. When this is achieved and the velocities and depths are comparable to those occurring in the stream, the conditions through the crossing should present no more of an obstacle to aquatic organisms than conditions in the adjacent natural channel.

FHWA Design Procedure:
The FHWA design procedure emphasizes the design of closed conduit culverts with a native bed material embedment that provides a natural bottom throughout the culvert. It also applies to open bottom culverts, which also provide a native bed material throughout the culvert, but additional foundation considerations are required.

The basis of the FHWA procedure is to create conditions in the culvert that are no more difficult to pass than those found in the upstream and downstream channel. Streambed sediment behavior is used as a surrogate by examining the permissible shear stress of the native bed, that is, quantifying how resistant the bed material is to movement, and comparing that stress to the stresses experienced by the bed during the periods that passage is needed.

Passage does not occur at very low flows nor at very high flows. The procedure provides a range of flow between the low passage flow and the high passage flow during which passage is anticipated to occur. The passage flows do not require determination of target species and life stages, though if they are known for a site should be used in defining the passage flows.

In addition to looking at sediment behavior, the FHWA procedure compares velocity conditions at the high passage flow with those in the upstream and downstream channel to insure that velocities are not higher in the culvert than the aquatic organisms would already have to negotiate elsewhere. Similarly, the procedure compares flow depths at the low passage flow with those in the upstream and downstream channel to insure that flow depths are not lower in the culvert creating a barrier to passage.

The procedure retains the requirement that the culvert must satisfy traditional criteria for addressing flood flows. The passage requirements, including the culvert embedment, generally result in a larger culvert than required by traditional criteria. However, the larger size may often generate lifecycle benefits of lower maintenance costs and longer service life because of the reduction of stresses in and around the culvert during flood flows and a reduction in the probability of retaining debris at the inlet or in the culvert barrel.

A final component of the FHWA procedure is a test for stability of the bed at flood flows. Although natural stream processes will replenish the bed material within the culvert if it is washed out during a flood event, it is not always advisable to rely on these processes. The procedure includes guidance for designing an armored layer underneath the native bed material to prevent a total washout and enhance the replenishment process.

Example Application:
The procedure is illustrated briefly here. This example, and two others, is provided in detail in the FHWA guidance document (Kilgore, et al., 2010).

Figure 2. Culvert Inlet at North Thompson Creek.

Site Description. The design procedure is applied to a road crossing of the North Thompson Creek, which is approximately 15 miles south and 3 miles east of Glenwood Springs, Colorado. The drainage area to the crossing is 2.33 mi2. There is an existing 36-in culvert at the stream-road crossing (see Figure 2), which was identified for replacement because it was considered a passage barrier, possibly because of high velocities in the barrel.

The high passage and low passage flows are determined by site-specific guidelines, if they exist. In the absence of site-specific guidelines, the high passage flow may be defined as the 10 percent exceedance quantile on the annual flow duration curve. If a flow duration curve is not available, the high passage flow is estimated as 0.25 times the 2-year discharge.

In the absence of site-specific guidelines, the low passage flow may be defined as the 90 percent exceedance quantile on the annual flow duration curve or the 7-day, 2-year low flow (7Q2). The minimum low passage flow is 1 ft3/s. By evaluating available data and methods for the site, the low passage and high passage flows were determined to be 1 and 8.8 ft3/s, respectively.

An initial CMP culvert diameter of 6.5 ft on a 0.0267 ft/ft slope is estimated considering that the existing 3-ft diameter CMP is inadequate and a 2 ft minimum embedment is required. The embedment criterion for a circular culvert is 30 percent of the culvert rise giving an embedded depth of 0.3 x 6.5 ft = 1.95 ft. However, the minimum embedment depth of 2.0 ft will be used for this design. The native bed material is selected for the embedment.

To check the bed stability, the modified permissible shear stress approach is applied (Kilgore et al., 2010). The shear stress in the culvert is compared with the permissible shear stress for the bed material. The resulting computations for the depth, velocity, experienced shear stress, and permissible shear stress are summarized in Table 1. Comparing inlet and the outlet conditions, the highest shear stress is at the outlet estimated as 1.3 lb/ft2. Since this is more than the permissible shear stress of 1.0 lb/ft2, the culvert bed is not stable. After performing additional evaluations of the shear stresses upstream and downstream, a larger (8.5 ft) culvert is tried. This culvert does satisfy the shear stress requirements.

Parameter Inlet Outlet
y (ft) 0.55 0.46
V (ft/s) 2.58 3.10
td (lbs/ft2) 0.8 1.3
tp (lbs/ft2) 1.0 1.0
Table 1. 6.5 ft Culvert
Inlet and Outlet Parameters
at High Passage Flow.

Once the previous steps are completed, a check is conducted to verify that the culvert velocity is less than or equal to at least part of the upstream or downstream channel. The check is satisfied if the culvert inlet and outlet velocities are within the range of the cross-section velocities. The results for this case are shown in Table 2. The velocity in the 46 ft culvert varies from 2.4 to 2.7 ft/s. Upstream of the culvert in the reaches indicated by cross-sections 342 and 307 there are higher velocities of 2.4 to 2.9 ft/s through a distance of 61 ft. Therefore, the velocity in the culvert does not present conditions more severe than are found elsewhere in the project reach.

Cross-section Applicable Reach Length (ft) Normal Depth/HY-8 (ft/s)
567 95 2.48
472 73 2.38
399 57 1.57
342 35 2.86
307 26 2.40
Culvert Inlet 23 2.68
Culvert Outlet 23 2.42
215 43 2.37
172 47 2.02
125 68 2.12
57 57 2.43
Table 2. Velocity Estimates at QH.

Next, a check is conducted to verify that the culvert depth is greater than or equal to at least part of the upstream or downstream channel. Table 3 summarizes the maximum depths estimated at each cross-section and within the culvert. The depths in the culvert bed are shallower than those in the upstream and downstream channel meaning that the culvert is the limiting location in terms of depth and, therefore, the depth check is not satisfied, requiring the creation of a low flow channel. A triangular low-flow channel with side slopes of 1:8 (V:H) is added. This will provide a thalweg 0.5 ft deeper in the center of the culvert.

125 0.30

Cross-section Normal Depth/HY-8 (ft)
567 0.30
472 0.30
399 0.58
342 0.28
307 0.24
Culvert Inlet 0.09
Culvert Outlet 0.17
215 0.18
172 0.29
57 0.25

An 8.5-ft embedded CMP on a 2.67 percent slope is proposed to replace the 3.0-ft CMP culvert on a 3.9 percent slope. A low-flow channel is to be created to maintain depths in the culvert at QL. No change to the road profile is needed. Alternative culvert shapes and materials may also be considered. A concrete box or pipe arch may offer an option to maintain a sufficiently wide span to meet the stability, velocity, and depth criteria with a lower rise.

The FHWA procedure offers a design procedure for culverts that provide for aquatic organism passage using bed stability as a surrogate measure. Species-specific data on passage requirements is not needed because the procedure focuses on a culvert installation that does not create conditions more challenging to passage than those found in the upstream and downstream channel.


Farhig, L., and Merriam, G.
1985. “Habitat Patch Connectivity and Population Survival.” Ecology, 66(6), 1762-1768.
Forest Service Stream Simulation Working Group (FSSWG)
2008. “Stream Simulation: An Ecological Approach to Road-Stream Crossings”, May.
Jackson, S.
2003. “Design and Construction of Aquatic Organism Passage at Road-Stream Crossings: Ecological Considerations in the Design of River and Stream Crossings.” 20-29 International Conference of Ecology and Transportation, Lake Placid, New York
Kilgore, Roger T., Bart S. Bergendahl, and Rollin H. Hotchkiss
2010. “Aquatic Organism Passage Design Guidelines for Culverts,” Hydraulic Engineering Circular, Number 26 (HEC 26), Federal Highway Administration, FHWA-HIF-11-0008.