Bridge Conceptual Design Guidelines
Bridge Conceptual Design Guidelines
Bridge Conceptual Design Guidelines
Version 3.0
The Crown in right of the Province of Alberta, as represented by the Minister of Transportation
Version 3.0
Volume 3.0
May 2020
These guidelines cover all aspects of bridge conceptual design, also referred to as bridge planning,
including bridge location, sizing, geometrics, hydrotechnical design, and river protection works.
These guidelines apply to all Alberta Transportation projects involving bridge size structures,
including all tasks identified as Bridge Planning as per the current version of the Engineering
Consultant Guidelines, Vol. 1.
Although this document is intended to be thorough, certain cases may arise where specific
guidance is not provided or not applicable. Consultants working for Alberta Transportation must
exercise technically sound and well-justified engineering judgment in the application of these
guidelines. It is not the intent of the document to limit progress or discourage innovation.
Consultants are encouraged to explore all engineering options they deem appropriate for a
specific site.
For situations where engineering analysis reveals that standards in this guideline are not
appropriate for a specific project, the design exception process shall be followed, as per the
Department’s Design Exceptions Guideline. Subject Matter Experts shall be informed for any
proposed deviations from current standards or guidelines and will make a determination of
whether a formal design exception submission is required. Documentation, either through a
formal design exception submission or through content within the Conceptual Design Report,
should include an appropriate level of engineering analysis, evaluation of alternatives (for
example an option to meet the standard in comparison to an option to not meet the standard),
risk assessments, mitigation strategies, and recommendations. The Subject Matter Expert shall
determine if the design exception is approved, with final sign off required by the Executive
Director of the Technical Standards Branch.
Any project specific questions relating to these guidelines should be directed to the Project
Manager. Any feedback or technical clarification requests relating to this document should be
directed to the Bridge Planning Specialist, Bridge Engineering Section, Technical Standards
Branch, Alberta Transportation.
Approved:
The following page is reserved for documenting changes to this version of the Bridge Conceptual
Design Guidelines. When changes are completed to the document, the following actions will be
completed:
The version of the document will be updated;
A basic description and the date of the change will be summarized below.
1 Introduction .................................................................................................. 8
1.1 What is Bridge Conceptual Design? ..................................................................... 9
1.2 Why do Bridge Conceptual Design? ..................................................................... 9
1.3 Process Overview ............................................................................................... 10
1.4 Technical Considerations Overview ................................................................... 10
8 Options Analysis.......................................................................................... 50
8.1 Evaluate Alternatives ......................................................................................... 50
8.2 Design Exceptions............................................................................................... 51
In the event that discrepancies exist between this document and other references, the Bridge
Conceptual Design Guidelines shall take precedence.
The purpose of the bridge conceptual design phase is to determine and document the most
suitable solution for a roadway to cross a stream, road, or other facility while considering relevant
issues, risks, and constraints, and exploring all potential options. The results should:
Document data compiled, project constraints, design parameters, alternatives
considered, and decisions made
Provide preliminary design information on the recommended concept to proceed to the
Detailed Design phase.
The main difference between a design project and a functional planning study is the level of detail
of data collection, analysis, and reporting. Refinement of bridge openings identified during high-
level planning studies shall occur during the bridge design process, as additional information
(survey, geotechnical, etc.) is gathered. Proceeding directly from functional planning level
concepts to detailed design is strongly discouraged.
During the functional planning and conceptual design phases of a project, significant savings can
result in comparison to the effort expended. Investing upfront effort into identification of
constraints and exploration of options oftentimes results in savings during future life cycle phases.
Additional benefits include better project scope definition, reduced project schedules, and
simplified issue resolution. Bridges are typically replaced due to structural condition rather than
functionality due to their high capital costs. With typical design lives of 75 years (50 years for
culverts), bridges are the least flexible infrastructure component of the roadway network. Failure
to consider future functional improvements can negatively affect safety, operations and
economics of the roadway network. Therefore, it is essential to consider scenarios that may occur
during the life span of a structure in order to develop an optimal lifecycle solution.
1) Data collection
2) Site inspection
3) Arrange for technical input
4) Hydrotechnical assessment (as required):
5) Geometric assessment
6) Review technical inputs
7) Develop feasible options
8) Prepare draft Conceptual Design Report
9) Submit final Conceptual Design Report
10) Follow up in future project stages (as required)
Technical considerations for any given project will vary depending on the specific project. In
general, the list below shall be considered in developing an optimal bridge concept.
Future Plans:
Life cycle analysis on bridge rehabilitation, culvert lining, traffic accommodation
Highway widening, twinning, minimize throw-away costs
Phased construction options
Net Present Value
Hydrotechnical (stream crossings):
Design parameters for stream at crossing site
Structure impact on hydraulics – constriction, drift/ice handling
River issues – scour/erosion, bank stability, flow alignment, protection works
Bridge/Highway Geometries:
Highway alignment – radius, superelevation, spiral, skew, safety, accesses, land severance
Gradeline – grade, K values, length of curve, bridge height, freeboard, cut/fill balance
Bridge geometrics - width, cross slope, sight distances, clear zone requirements
Roadside/median barrier requirements vs barrier free
Structural:
Span arrangements (single vs. multiple spans), , pier location, skew
Deck drainage
Structure type, girder depths
Culvert vs. bridge, major bridge vs. standard bridge
Retaining walls vs open headslopes
Geotechnical:
Slides, headslope ratios, remediation works
Pile depths, settlements
Retaining walls
Environmental:
Regulatory requirements (Fisheries Act, Water Act, Navigation Protection Act, First Nations
Consultation, etc.)
Oftentimes, data will provided or identified as available in a project’s terms of reference. Requests
for historical or corporate data (Alberta Transportation, 2020) from Alberta Transportation can
be made through a project’s Project Manager. Note that historical data contained in the Regional
offices may not be the same as date contained at the head office (Twin Atria Building).
The Consultant is responsible for obtaining sufficient data and survey for each particular project.
All sites are unique, however, a sample inspection checklist is provided below:
Action Description
Determine channel Identify portions of the channel that appear typical of the
dimensions reach and represent the natural channel
Estimate bed width, top width, and bank height
Note highwater data Note type, location, elevation of any highwater/ice marks
Talk to landowners, local officials about history
Note any backwater impacts e.g. farmlands, buildings
Characterize stream channel pattern – meandering or braided, incised channel or
and geomorphology floodplain, differences between natural channel and in the
vicinity of the structure
bank stability – slope, vegetation, material, height, erosion,
rock outcrops, slides, springs
bed material – gravel, sand, silt, D50
bed forms – bars, islands
flow alignment – bends, skew
dimensions – water level and velocity
Notes signs of degradation/aggradation
Assess operation of general scour – bed lowering through the opening
existing crossing local scour – holes near piers, protection works
condition of slope protection – cracks, loss of material
abrasion on pier nose plates
water or soil side corrosion
environmental sensitivities
drift accumulated at opening
high load strikes, deck drainage performance at overpasses
Identify hydraulic rock ledges
controls changes in cross section geometry or slope
hydraulic structures like weirs, nearby bridges
lakes, beaver dams
Assess basin terrain – flat, rolling, foothill etc.
characteristics land use – farming, forest, development
surface storage – lakes, slews
upstream controlling factors – outlets, weirs
Towards the end of the conceptual design phase it is useful to conduct a subsequent site visit,
particularly for complex sites or major bridge projects, to envision or lay out options in the field.
This helps to assess the feasibility of options in the field and allows an opportunity to identify any
additional unforeseen issues or constraints, before finalizing the recommended conceptual design
option.
The Consultant is responsible for determining the number and scope of site visits required
depending on the project and specific site needs, unless otherwise specified within a project’s
terms of reference.
Once preliminary conceptual options are developed, obtaining additional technical input is
recommended, particularly for major bridges or complex sites, including:
Checking the field survey and other data sets for completeness and accuracy
Obtaining opinions from others specialties on the feasibility of the option (roadway,
geotechnical, environmental, etc.), and
Noting any new site constraints (e.g. ROW, site access, detour, environmental concerns,
restricted activity periods, utility relocation, land purchase, budgets, schedule,
stakeholders) or projects risks.
Design of stream crossings requires the flow depth (Y), mean channel velocity (V) and resultant
flow (Q). Key principles in determination design parameters are:
Representative of the capacity of the channel to deliver flow from the upstream basin
Consistent with the historic high water observations
Consistent with existing hydraulic performance
Flood frequency analysis has proven to be incapable of meeting these principles at most sites, as
documented in Context of Extreme Floods in Alberta (Alberta Transportation, 2007)
The overall process to determine hydrotechnical design parameters for a site is:
Estimate typical natural channel parameters
o B (bed width), T (top width), h (bank height), S (slope)
Calculate Channel Capacity (CC), using channel capacity calculator tool (CCCT) or other
methods
o Determine Y, V, Q
Assemble Historic Highwater Data
Check if HW exceeds CC: If YHW > YCC, set Y = YHW in CCCT
o Determine V,Q using CCCT
Calculate Basin Runoff Potential (BRP)
o If drainage area < 100km2 , look up ‘q’, calculate QBRP
o If drainage area > 100km2 , method does not apply
Check if BRP governs: If QBRP < Q, set Y = YBRP
o Determine V,Q using CCCT
Recommend Q,Y,V values for Design
Drainage Area (DA) – topographic area potentially contributing flows and used in Basin
Runoff Potential analysis. DA can be determined using DEM data and GIS tools, with some
values noted in HIS.
Airphotos – can show changes in flow alignment over time, indicating lateral mobility of
a stream. Bank erosion can be tracked using georeferenced airphotos with some
banktracking summaries are available in HIS.
Scour surveys – can show vertical changes in streambed over time, including local and
general scour and bed form movement. Some scour surveys are available in HIS.
AEP Flood Hazard Mapping Tool – areas may be subject to additional constraints.
AEP Fish and Wildlife Mapping Tool – provides information related to fish and wildlife
data
AEP Code of Practice maps – these maps classify streams in terms of their importance in
fisheries management, and note restricted activity period dates.
Local site hydraulic influences – such as other structures (weir, bridge, culvert, dam),
sudden channel changes (slope, width), and confluences with other channels.
Site inspection observations – channel features, flow concentration and alignment, active
bank erosion, and ice scars on trees.
Historical Reports (TRANS, AEP, Universities, Municipalities)
This technique estimates the capacity of the channel to deliver flows to a site at a defined depth
above the bank height. The typical channel is a trapezoidal representation of the stream reach
that the crossing is located on, as shown below:
Sources of information for ‘B’ and ‘T’ include georeferenced airphotos, digital elevation models
including LiDAR, site measurements. Many cross sections within the river’s reach should be used
to determine an average natural channel section. The sections used should be on a natural, stable,
and straight portion of the river within proximity of the project site. In many cases, channels in
proximity to an existing crossing may have been modified during construction or influenced by
the existing structure or adjacent land use, and may not represent the natural channel. Surveyed
cross sections are usually too limited in number to enable estimation of the typical values. ‘B’
values can be estimated to the nearest meter from surface water width on airphotos at low water
levels, with adjustments based on observations and survey data as appropriate.
The bank height is the height at which the channel transitions to the floodplain, and there should
be a sudden change (decrease) in the slope of the terrain perpendicular to the channel. Sources
of information include surveyed cross sections, high resolution DEMs (e.g. LiDAR), photos with
scalable objects, and site measurements. Other bank height definitions exist such as based on the
line of permanent vegetation. This will typically be below the geometric definition of bank height
used in bridge hydraulic analysis, and should not be used for this purpose. For high-level
These values are based on the Context of Extreme Floods in Alberta (Alberta Transportation, 2007)
analysis. Values that exceed the channel capacity are accounted for in historic highwater analyses.
The equation for B>=10m is based on the Evaluation of Open Channel Flow Equations (Alberta
Transportation, 2005) study. The Channel Capacity Calculator tool built by AT has all of these
calculations built in.
B Vcc n
>= 10 14R0.6S0.4 NA
7–9 R0.67S0.5/n 0.040
4–6 R0.67S0.5/n 0.045
<= 3 R0.67S0.5/n 0.050
Where:
R = hydraulic radius = A/P
A = typical cross section area of flow at YCC (m2)
P = wetted perimeter of typical cross section at YCC
n = Manning coefficient, with adjustment for Slope (S) as follows:
S ‘n’ adjust
< 0.0005 (B > 8) n = n – 0.005
0.005 – 0.015 n = n + 0.005
> 0.015 n = n + 0.01
The channel slope ‘S’ is typically a small number (<<1) so a significant length of channel profile
(kilometers) is required to estimate this parameter. Variations in bed levels measured in a survey
usually exceed the total drop observed over the length of the survey. Natural channels are a
difficult environment for surveying, with challenges in locating the thalweg, limited line of sight,
and hazards from flowing water, soft silt deposits, and slippery banks. Extending the length of
surveys is generally not a practical solution. Channel slopes are best estimated from DEM data
and have been extracted for most streams in the province into HIS. In some cases, the crossing
may be located close to a sudden break in the channel profile, and additional investigation may
be required to confirm which slope would apply to the crossing site.
When evaluating highwater data, the flow depth should be for the channel. Any records caused
by a constricted opening, superstructure in the water, or blockage due to drift should be
accounted for. Data for structures located on the same stream and in proximity should be
considered and judgment should be applied when considering data from sites with substantially
different drainage areas or channel parameters. Measurements downstream of a bridge are more
representative of the natural channel response under flood conditions. Measurements upstream
and downstream will enable assessment of the bridge hydraulic performance. Timing of
At some sites, there may not be enough supply of water from the basin to fill the channel to its
physical capacity. This can occur when the drainage area (DA) is small (< 100km2) relative to the
capacity of the channel. One case would be a small drainage basin (e.g. 5km2) that drains into a
ravine that cuts through the valley wall of a larger river. The ravine may be steep and have high
banks but the surrounding basin will not supply enough water to fill it. The Basin Runoff Potential
Map (Figure 2) assigns the largest observed unit discharges (‘q’) to various hydrologic regions
within Alberta. To estimate this upper bound QBRP from the basin, the selected unit discharge q
and defined ‘DA’ are multiplied. Details are found in the Development of Runoff Depth Map for
Alberta (Alberta Transportation, 2006).
Known exceptions to the Basin Runoff Potential Map are the Cypress Hills and Swan Hills area,
where higher unit discharges have been recorded due to the higher basin gradients and ‘q’ =
0.4cms/km2 is recommended.
The hydrotechnical design values (along with fish passage and Q2, as required) will form the
boundary conditions for calculations. For culverts and constrictive bridges, calculations involving
gradually varied flow (such as backwater curves) and rapidly varied flow (abrupt energy losses
over a short distance) are necessary. These calculations can be done by simple models using
prismatic channels and one dimensional (section averaged) techniques, such as those facilitated
by Alberta Transportations Flow Profile (Alberta Transportation, 2015). Advanced techniques,
such as multi-section (e.g. HEC-RAS), two-dimensional, and unsteady flow calculations are not
necessary and offer little value in bridge design. Some of the reasons to avoid using models that
are more complex include:
Boundary conditions are one dimensional anyway
Natural rivers have mobile boundaries (scour, bed forms, lateral erosion)
Many natural factors cannot be modeled accurately – drift, ice, sediment transport
Data-sets don’t exist to support true calibration of complex models
Complicated outputs are difficult to interpret and assess
These models are expensive and require significant resources
Most of the output, accurate or not, is not needed to design a bridge
Channel capacity method calculations do not account for flow adjacent to the channel in the
floodplain. Hydraulic calculations suggest that the down-slope component of flow on the
floodplain is a small portion of the channel flow (typically < 10%). Additional reasons include:
Relatively shallow Y and low V (high relative roughness)
Lack of a defined and continuous channel in the floodplain
Presence of many natural and man-made obstructions (trees, roads, development)
Most flow interacts laterally with the channel as levels change
Sizing a standard bridge, major bridge, or major buried structure involves placing the bridge fills
and setting the roadway gradeline to provide the desired freeboard. A starting point is to place
the fills parallel to the channel banks at the crossing location. From this point, a range of options
can be considered in the optimization process. As the bridge opening decreases, the degree of
constriction increases. This will result in increased velocities (V) and headloss (V2/2g) through the
bridge opening. The hydraulic impacts of a constriction and increased velocity will result in
increased size and quantity of riprap protection, bank erosion, increased water level and flood
risk on adjacent developments, and reduced freeboard (possibly requiring a gradeline raise).
Bridge openings wider than the typical channel can provide advantages such as less protection
works, if sufficient buffer is provided. In some cases, these benefits may counteract the expense
of additional bridge length. Openings much larger than the typical channel can result in adverse
impacts, such as sediment deposition and local flow realignment issues that can lead to increased
bank erosion, and are generally more costly.
Freeboard should be determined through optimization. The starting point for freeboard shall be
1.0m between design highwater elevation and the lowest point of the structure. Higher values
are seldom justified hydraulically, but may result from gradeline optimization. Lower values
should be considered if any the following conditions are met:
Reducing freeboard could result in a significant cost reduction (>15%)
There is a high degree of confidence in the design highwater level
There is limited potential and/or history for drift or ice accumulation at the site
The bridge is on a longitudinal grade, where most of the bridge has more than 1.0m
freeboard
A single span bridge (no piers) is proposed, with less risk of blockage
The volume of traffic is low and detour length is short
A minimum freeboard amount of 0.3m is achieved
A shorter design life is desirable, such as for a temporary structure
Culverts are available in a range of materials, shapes and sizes. Historically, the most cost effective
and common solution is a single round culvert. In general, culvert shapes do not match the shape
of a natural channel, resulting in flow contraction and expansion when water is entering and
exiting a culvert. A useful starting point for sizing a culvert is the flow depth plus the burial depth
(diameter/4) that results in an opening that would have close to no freeboard. From here, the
culvert opening should be optimized for the site considering site-specific objectives and risks
including AADT, height of fill, detour length, and structure design life.
Hydraulic calculations are typically more complicated for culverts than for bridges, due to factors
such as the different shape, burial depth, and the potential for full flow. Various combinations of
rapidly varying flow, gradually varied flow, normal flow, and full flow are also possible.
Supercritical flow may result in some cases, with the potential for hydraulic jumps. As such,
hydraulic modeling is recommended to assess each option being considered. For the majority of
highway sites, ponding or pooling of water above the culvert is not desirable, resulting in most
culverts operating under subcritical flow conditions (flow depths controlled by downstream
tailwater conditions).
Transport Canada (TC) assesses navigation impact of a crossing structure based on the mean
annual flood (Q2 or 1 in 2-year flood). Vessel clearance is measured from the Q2 elevation flood
to the underside of the bridge. This applies to watercourses declared navigable under the
Navigation Protection Act, along with sites determined by AT to be navigable. Further guidance is
found on the Environmental Regulation webpage, including the Navigation Assessment form.
For sites with nearby WSC gauges with long records, Q2 can be calculated as the average of the
reported annual maximum mean daily flows. This analysis is documented in Estimation of
Navigation Clearance Box Reference Water Level (Alberta Transportation, 2011) in Appendix D.
For sites without data, the following method is proposed to estimate the equivalent to the Q2
water level:
1. Determine design flow (Q) as per Section 2.1
2. Calculate Q2 = Q/(4 + 600*S), where S is the channel slope in m/m
3. Calculate Y2 using channel capacity method
4. Add Y2 to streambed elevation to get the Q2 reference water level
Alberta Transportation projects require adherence to Provincial and Federal legislation related to
accommodation of fish passage through structures. The main principle of fish passage through
culverts is to ensure the mean velocity throughout the structure is less than or equal to the mean
velocity in the channel at QFPD. The reasoning behind the velocity comparison approach is that if
the fish can adapt to comparable velocities in the stream, the culvert itself should not be a velocity
barrier to them. This approach does not involve the use of fish swimming performance curves as
these curves have often resulted in mean velocities that are a small fraction of the mean velocity
in the channel, and cannot be met with a culvert or bridge crossing. Many of the studies used to
develop such curves were completed in laboratories where natural stream variations such as
pools and riffles, or vegetation were not present.
A fish passage design flow, QFPD, is required for culverts on fish bearing streams as below:
1. Calculate YFPD = 0.8 – 34.3*S , where S is the channel slope in m/m
2. Minimum YFPD = 0.2
3. Calculate QFPD at YFPD
When comparing mean velocities, the precision should be extended to 0.01m/s due to the
relatively low magnitude. This approach was developed with support from Alberta Environment
and Fisheries and Oceans Canada and is documented in the Discussion on the Selection of the
Recommended Fish Passage Design Discharge (Alberta Transportation, 2012), in Appendix C. This
method ensures that fish passage is evaluated at a relatively high flow, while providing more
consistent results than statistical estimates such as the 3Q10 flow.
In general, increasing pipe diameter or burial depth are ineffective methods of reducing mean
velocities at QFPD, as most of the additional area will be above the flow depth. Increasing the burial
depth can also lead to sedimentation/maintenance issues, and construction challenges due to
increased excavation depth and a more difficult (steep/long) upstream transition from the culvert
to the channel for the fish to traverse. If a feasible configuration cannot be found, installation of
substrate and holders inside the culvert should be considered. Substrate holders assist in
retaining substrate material thereby increasing the effective roughness of the culvert and
decreasing the mean velocity.
For sites where substrate and holders are proposed, the following parameters are recommended:
Substrate holder should be made of steel and conform to the shape of the pipe up to the
desired height.
Height of substrate holder should be 0.3m (0.2m if culvert diameter < 3m)
Spacing of holders should be based on height divided by culvert slope (minimum = 7m)
Substrate should be Class 1M or Class 1 rock, with an average thickness matching the height
of the holder.
Substrate and holders are only required for portions of the pipe where the mean velocity
exceeds the mean channel velocity (typically upstream 1/2 or 1/3).
Materials shall be as per the Standard Specifications for Bridge Construction (Alberta
Transportation, 2020)
The hydraulic effect of substrate is assessed by blocking off the flow area filled by the substrate
and increasing the effective Manning roughness coefficient “n”. The relative roughness depends
on the substrate type and flow depth, as shown in Table 1 below. AT’s Flow Profile tool will block
the substrate flow area and adjust the roughness parameters, if a substrate value is entered.
In February 2019, AT and AEP signed a Memorandum of Understanding stating that these Bridge
Conceptual Design Guidelines shall be used for bridge and culvert design for highway crossings
and that the Alberta Roadway Crossing Inspection Manual (Alberta Environment and Parks, 2015)
shall be used to assess fish passage inspection compliance. For further guidance on legislative
requirements, contact AT’s Environmental Regulation group.
The presence of barriers, curbs and raised medians impedes the ability of rainfall runoff to drain
off bridges. Rainfall collects and is channeled along these barriers until it reaches a drainage point
of sufficient capacity, or until the point of overtopping. Encroachment of water into driving lanes
can result in a road safety hazard due to hydroplaning, driver avoidance (swerving to avoid
ponding), and visibility (splashing on windshields). Local pooling of water for extended durations
on the bridge deck can also result in an increased rate of deck deterioration due to sub-surface
drainage. Historically, deck drains combined with optimized geometry is used to minimize lane
encroachment and local pooling. Use of below deck drainage systems is generally avoided due to
capital and maintenance costs, low reliability (durability, clogging, segments becoming
separated), and safety concerns. Drainage issues should receive early attention at the planning
stage, when there is opportunity to optimize bridge geometry. Optimization should include
considerations to longitudinal grade, shoulder width, number of deck drains, amount of driving
lane encroachment, roadway classification, safety concerns, risks, and costs. Detailed design of
components, including deck drain and trough design, is further discussed in the Bridge Structures
Design Criteria.
The minimum desirable longitudinal gradient for bridges of 1% is specified in the Bridge Structures
Design Criteria with deck drains as normal practice for river crossings. Bridge deck drainage
analysis shall combine the Rational Method equation for runoff flow rate estimation and the
For safety reasons, encroachment should be minimized with a desirable maximum encroachment
of 0m into the driving lane for divided highways and 1.0m for undivided highways. For all cases,
a minimum lane width of 2.5m shall be maintained and the maximum water depth within a travel
lane shall be 25mm. The following design parameters shall be used: i = 75 mm/hr, C = 0.9, n =
0.016, as further discussed below.
Manning Equation:
The Manning equation relates the depth of flow to the runoff rate as follows:
Q = 1000AfR2/3 S1/2 / n
Where:
Q= runoff rate (L/s)
Af= flow area (m2)
P= wetted perimeter (m)
R= hydraulic radius (m)
S= longitudinal slope of deck (m/m) at point of analysis
n= roughness coefficient (use n = 0.016 for bridge decks)
Where:
Y= depth of flow (m)
e= superelevation or crown rate
T= top width of flow (m)
Therefore:
Solution:
• For specified T (shoulder width for no encroachment), calculate longitudinal flow capacity (Q,
Manning eqn.).
• Use Q with the Rational Method equation to calculate length to first deck drain
• Calculate drain spacing using deck drain flow (at specified T) with Rational Method equation
• Use spacing as approximate guide to optimally locate deck drains on structure.
• Iterative solution may be required for variable grade/width bridges decks.
For detailed deck drainage design, including analysis and sizing of deck drains and trough drains,
Hydraulic Engineering Circular 21: Design of Bridge Deck Drainage (Federal Highways
Administration, 1993) and Hydraulic Engineering Circular 22: Urban Drainage Design Manual
(Federal Highway Administration, 2009) (Federal Highway Administration, 2009) shall be referred
to, in conjunction with the latest version of the Bridge Structure’s Design Criteria. Specific
considerations at the Conceptual Design Phase include:
Minimizing the number of deck joints and deck drains
Deck drains shall not discharge onto underpassing traffic lanes or pedestrian facilities
Drainage shall not be discharged onto any exposed substructure concrete surfaces
Shoulder use by bicyclists; eliminate snag hazards and minimize dips/elevation changes
with deck drains in the travel paths if bicycle traffic is expected
Locating and designing future drainage considerations for projects where deck widening
will occur in the future
Accommodation of hazardous materials or deleterious substances for environmentally
sensitive sites such as streams with critical fish habitat or adjacent water intake facilities
Directing drainage away from MSE wall structures and major buried bridge structures,
through grading and the use of membranes (refer to Bridge Structures Design Criteria)
Erosion control at discharge locations
Natural channels have mobile boundaries, both vertically and horizontally. Lateral movement of
stream banks is referred to as erosion (typically addressed with river protection works), while
vertical changes are referred to as scour. Scour and erosion susceptible structures have features
located within the active floodplain. Examples includes culvert soil back-fill envelopes, roadway
embankments, open bottomed culverts, and shallow bridge foundations as loss of soil support for
these structures could result in sudden and complete failure of the structure.
If such structures are proposed, mitigation measures are required to reduce the risk of
vulnerability and increase resiliency. Potential mitigation solutions may include
Increasing structure size or shape
placing foundations below scour depths
placing rock riprap or river protection works (see Section 5.6.1)
the use of cutoff walls
the use of clay seals
low permeability end treatments
encapsulating backfill with geotextile
the use of sheet piles to protect a footing foundation
establishing a monitoring program.
Historically, due to costs required for mitigation measures, open bottom structures and MSE walls
located within floodplains have had limited use for Provincial highway projects. Some situations
where they may prove to be cost effective include uses as temporary or detour structures, on
Local Road projects where additional risk may be acceptable, or at sites where erosion and scour
are not concerns such as wildlife passage structures or railways. For more guidance on culverts
and buried structures design, refer to Section 7.4.
5.6.1 Scour
Figure 3: Scour Definitions (Adapted from Guide to Bridge Hydraulics (TAC, 2008))
As part of a research project, data collected during this program was compared to the industry
standard modified Melville approach recommended in the Guide to Bridge Hydraulics
(Transportation Association of Canada, 2004). In general, it was found that the modified Melville
approach over predicted the local scour when compared to design conditions. It is worth noting
that all of AT’s field data was collected after flood conditions have passed due to safety concerns,
and that some infilling may have occurred. For some sites, it is likely that scour deeper than that
measured occurred during high flow events. AT intends to update these results over time to
account for the collection of additional data for this program.
Figure 5 below shows the results of this research. The modified Melville approach predicted local
pier scour depths between 1.8m to 13.2m, with an average of 4.8m. The observed scour depth
measured through the Scour Survey program ranged from -0.3m (increase in streambed
elevation) to 4.0m, with an average of 1.5m. The difference between predicted and measured
scour ranged from -0.2m (underestimation) to 10.8m, with an average over estimation of 3.3m.
There also appeared to be no noticeable trend that related observed scour depth to flow, average
sediment size, or pier width, although these factors are used in the modified Melville approach to
determine the theoretical scour depth.
Although this research project focused on a small sample of bridge sites across the Province, there
was an observed data trend towards a maximum prolonged scour value of less than 3m. Measured
scour values at two bridge sites exceeded this value: BF78104 and BF73949. For BF78104
(Highway 32 over the Macleod River), the 4.0m measured depth of scour is thought to be due to
the 45 degree flow alignment with the center pier and thalweg location. For BF73949 (Highway 2
over the Peace River), the 3.8m measured depth of scour is thought to be due to the accumulation
of drift and thalweg location.
Modern bridge construction equipment and materials has made piled in-stream foundations cost-
effective. In general, as long as the foundation penetrates >5m into the streambed, pier scour
should not be an issue. In the rare case of spread footings or short pile foundations, the bottom
of the foundation should be lower than the estimated pier scour depth. The design pier scour
depth shall be estimated as 2 times the effective pier width to a maximum of 3.0m below
streambed unless site-specific data is available to suggest deeper parameters are warranted. To
minimize impacts to navigation, the top of pile cap shall be beneath streambed elevation, to a
maximum practical limit of 2m below streambed. Foundation design should consider the impact
of loss of material up to the design scour depth. Geotechnical recommendations, structural design
implications, and constructability should also be considered in the foundation design, as further
described in the Bridge Structures Design Criteria.
Fills placed in the active waterway generally require protection to prevent erosion. The major
types of protection works systems include headslope protection, guidebanks, and spurs. Rock
riprap is the preferred material for protection of bridge headslopes, culvert ends, and river
protection works. Reasons include over 60 years of proven performance history; systems that
resist drift, abrasion and ice forces with the flexibility to accommodate settlement and launching;
proven velocity based criteria for selection of rock protection systems, within many publications
and studies; relatively low cost and generally readily available sources of rock riprap; relatively
easy monitoring, maintenance and repair procedures; and laterally mobile streams require a
“hard” solution to maintain flow alignment.
Bioengineering options, such as willow staking within a rock riprap protection system, may
compromise the function of the geotextile and impact hydraulic capacity of the bridge opening
on smaller channels. These options may be considered for projects beyond the extent of the
crossing, such as fisheries compensation in a channel, and erosion protection within ditches.
Selection of the appropriate class of rock riprap is based on mean velocity (V) at the design flow
with materials as per the Standard Specifications for Bridge Construction. Rock class shall be:
Class 1M riprap is seldom used on bridge projects, with either Class 1 or no protection options
typically assessed for very low velocity sites. Rock gradation is important to ensure interlocking of
the rock. In some very high velocity cases, a modified gradation has been used for aprons, with
smaller sizes excluded from the mix. The angularity of the rock (less rounded) becomes more
important at high V sites as the rock is more likely to interlock when it is angular. For sites where
V exceeds 4.0m/s, addition of H or sheet piles in the apron may be considered to enhance
protection. Rock larger than Class 3 is usually not considered due to limited availability, cost, and
difficulty in transportation and placement.
The protection works system should extend well into a stable bank or naturally protected feature.
This will minimize the risk of the system being outflanked/eroded from behind. On streams with high
lateral mobility, historical river banktracking will help to determine the extent of protection works
required. Stable natural features, such as rock outcrops, should be utilized as appropriate. Guidebank
and spur configurations are often best developed in plan view on top of airphotos. Alberta
Transportation has developed a Spur Planning Geometry tool to support this.
Berm EL
Top of Rock EL
(Des. HW)
t
Bed EL
Berm
(Theor.)
Width
1
t RPW Definition Sketch
m
Apron
Bottom of Length
Headslope
Rock EL
Ratio
Guidebanks are protected fills built parallel to the flow that extend beyond the bridge. They
improve and maintain flow alignment through the openings on laterally mobile streams. Parallel
flow alignment is preferred to skew to reduce the structure’s size and minimize erosive forces on
banks/protection works. Guidebanks typically extend from the headslope, and flare towards the
bank in an elliptical shape with a 2:1 ratio of distance along the stream to perpendicular to the
stream. The transition from the channel into the bridge opening should be smooth.
Spurs are fills that project perpendicular into the river, with protection works on the ends. They
deflect flow from a bank or align flow. They are typically used in groups with other spurs or in
addition to guidebanks, and can be more cost-effective that continuous protection. Spurs with
significant projection may cause a contraction of flow, be difficult to construct/maintain, and may
require extensive environmental approvals.
Additional references for river protection design include Hydraulic Engineering Circular 23: Bridge
Scour and Stream Instability Countermeasures Experience, Selection, and Design Guidance
5.6.3 Degradation
Degradation is the long term lowering of a channel elevation over a significant distance, in
comparison to localized streambed scour. It can occur naturally or because of manmade activities,
such as channel straightening. Degradation can result in unstable banks and exposed structural
elements. It is important to differentiate degradation from scour as solutions and mitigation
measures may be different. Signs that degradation may be an issue at a site include:
changes in historical streambed surveys
changes in historic airphotos, showing progressive slumping, channel deepening over a
significant length, or vertical banks
history of hydraulic structures and channel modifications/shortening on the stream
ongoing maintenance concerns
ravine like section approaching a confluence
If degradation has occurred, some judgment will be required to determine if further degradation
may occur. This can be based on changes in rates of progress over time, and whether the
degradation was caused by man-made (reaction of one-time intervention) or natural (potentially
ongoing) activities. Additional information is contained in the Steam Degradation Technical Note
(Alberta Transportation, 2013).
Ice impacts include design forces on piers, vertical clearance for ice jams, and structure blockage
due to icing (aufeis). Where historic observations of ice impacts are available, these shall be
considered in determining design parameters. Ice jams form when pieces of broken ice form a
partial blockage of the channel. The constricted opening may result in headloss, and the
accumulation of broken ice upstream of the toe may result in sustained high water levels for long
distances. Highwater levels are the result of the increased wetted perimeter of the floating ice
combined with the high effective roughness of the broken ice, and the submerged thickness of
the ice itself. Broken pieces of ice may also project above highwater. Jams can form during freeze-
up or break-up. Break-up jams form due to a weakened ice cover or increased runoff flows
physically breaking competent ice. In general, the more competent the ice, the more severe the
ice jam. Ice jam elevations can be several meters higher than highwater from summer flood events
and can form/release very quickly.
When ice breaks up or ice jams release, floating ice moves downstream and potentially impacts
piers. This can result in significant loading on piers. As defined in the Canadian Highway Bridge
Design Code, the three components of loads are ice strength (classified into situations, a – d),
elevation, and ice thickness. Historic records should be reviewed for information that may help
quantify these three parameters such as abrasion on pier nose plates. When historic observations
do not exist, the following parameters are recommended, based on a system wide review of
historical observations across Alberta. In Table 2 below, ‘Y’ is the design flow depth and ‘t’ is the
design ice thickness. In some cases, two combinations of parameters may be appropriate, such as
weaker load at higher elevation and stronger load at lower elevation. Structural analysis can
determine which set of parameters may govern pier design.
Damage History Small Stream (B < 50m) Large Stream (B > 50m)
Icing (aufeis) occurs when a structure is blocked by solid ice. The ice can form due to repeated
freezing of water supplied to the site, such as an upstream spring. During spring runoff, a buried
structure may be blocked/iced due to not being exposed to the sun. Icing reports are often noted
in maintenance records. Remediation options include a larger opening or installation of an
additional culvert above the main structure. Maintenance such as removal of the icing before
spring runoff by thermal or physical means may be required for some sites. Deicing lines have
been used historically but are not recommended due to maintenance and environmental
concerns.
Flood events are frequently accompanied by drift that can impede flow, or change flow
alignments. Factors indicating potential drift at a site include significant amount of trees near the
channel and its tributaries in the drainage basin, laterally mobile streams with active bank erosion,
historic records noting issues, accumulations at piers/on point bars/or at a culvert opening, and
presence of beaver dams. Debris is often a controlling factor in design for mountainous streams
where the sediment capacity need is greater than the hydraulic capacity need. Culverts are more
prone to drift/debris concerns than bridges, as the surface width provided by a culvert decreases
at higher stages. Some mitigation options for culverts include:
Increase in size/change in shape - consider cost-effectiveness.
Flared inlets with raised crown elevations to maintain flow in case of a drift blockage.
Channel realignments can result in cost-effective, sustainable, and optimal solutions. Many
projects, such as twinned highway structures, high fill culverts, buried structure bridges, and
bridges on highly mobile streams require some form of channel work. The main benefit channel
realignment is reduced skews, resulting in simpler designs. Flow alignment and fish passage may
be improved with realignments. The main principle in designing channel realignments is to mimic
a stable section of natural channel in plan-form, cross section, and profile. A man-made channel
should be designed with similar B, h, T, and S values as the natural channel. This should result in
a stable, low-maintenance, and low environmental impact solution. A larger opening or milder
slope has the potential to result in aggradation (potential sediment accumulation) while a smaller
opening or steeper slope has the potential to result in degradation. It is important to
communicate this to regulators in the approval process.
Existing data is required to assess the functionality and safety of the existing highway and bridge.
A new design should address and remediate existing concerns at a bridge site including high
collision rates, substandard geometrics, poor sight distances, access management, insufficient
clear zones, insufficient freeboard, high structure skew, or bridge grade. Future parameters are
required to ensure lifecycle performance.
Horizontal clearances for bridge structures shall be as per the Roadside Design Guide. If lower
values are proposed, the design exception process shall be followed, with evaluations based on
level of risk, length of impact, economics, and past precedence. Bridges shall be on tangent
horizontal alignments as curved bridges require extra design and detailing, and cost more for
construction and maintenance. Curve effects on design include extra width for sight distances,
impact on operations, safety (icing, braking on bridges), deck drainage, maintenance, etc. Tapers
and spirals on bridges introduce similar challenges. Where practical, skews on bridges should be
minimized as skewed bridges require extra design and detailing efforts along with additional costs
for construction and maintenance. The presence of bridge rails can affect horizontal sight distance
by obstructing a driver’s field of vision. Solutions can include modifying the tangent alignment of
the bridge, increasing curve radii, using a wider shoulder at the inside of the curve, and using a
longer offset to intersection from bridge end.
Bridge rails can affect vertical sight distance for ramps and intersections at grade separations. The
impact can be calculated by reducing both height of eye and height of object by the effective
bridge rail height. Solutions include increased K value for vertical curves, longer offset to
ramps/intersection from bridge end, and wider shoulders. Guidance on calculations can be found
in the HGDG. River crossings can have significant impacts on gradelines due to freeboard and
structure depth requirements combined with minimum desirable grade. For these crossings, it is
essential for hydrotechnical, roadway, and structural expertise be involved in developing roadway
profiles.
For bridges over a roadway, the design vertical clearance shall be 5.4 m at the most critical
location. The design vertical clearance for pedestrian structures shall be a minimum of 5.7m.
Increased clearance may be required where there is an increase level of risk or reduced structural
redundancy, as further discussed in the Bridge Structures Design Criteria. The Department has a
commitment to the trucking industry to not change the posted vertical clearance of a bridge
structure over a structure’s lifespan. One rare exception is widening of a grade separation bridge,
for which allowance should be made during the initial design if widening is expected within 20
years. Allowance for future pavement overlays are not considered beneath a bridge structure
and the under passing pavement designs should reflect this. Mill and fill practices are assumed
for pavement rehabilitation beneath a bridge structure. Consultation with highway network
planners should occur to determine if the under passing road is (or will be) designated as part of
the Over Sized Over Weight (OSOW) Corridor (Alberta Transportation, 2018).
Bridge decks, including abutments, are susceptible to preferential icing (bridge is icy when the
approach road is not). As such, braking and steering adjustments on the bridge deck should be
avoided. Solutions include reducing the longitudinal gradient, reducing the superelevation
(tangent alignment or large radius curve), providing a constant road cross section across the
bridge (no spirals, no tapers), avoiding intersections on or in close proximity to the bridge (avoid
braking on structure) or considering a large culvert structure. Anti-icing measures may be required
if a suitable solution cannot be developed.
Available span lengths and structure type may affect the configuration of complex grade
separations structures. Gradelines may be affected by structure depth requirements, and
alignments may be affected by location of piers. Bridge costs and complexity of design can
A bridge width that matches the road width provides continuity for drivers. Functional aspects of
bridge width include achieving shy line offsets to barriers, storage of stalled vehicles,
consideration to cyclists/pedestrians, snow storage, deck drainage, and traffic accommodation
during maintenance. The optimal bridge width will balance functionality, capital construction
cost, and lifecycle performance.
Approval from the railway operator and road authority must be obtained for one to cross another.
These approvals and associated conditions form the basis of a contractual agreement, called a
Crossing Agreement. These agreements typically contain information about who is senior at the
crossing location, apportionment of costs, maintenance and operation responsibilities, and any
future obligations. Historical agreements at a site set the precedent for future work at that site.
At the conceptual design phase, railway design criteria (horizontal and vertical curvature, sight
lines, span lengths, design loads, etc.) differ from those used on roadway design projects.
Standard railway clearance boxes are included as part of the Roadside Design Guide (Alberta
Transportation, 2020) and were developed in consultation with Railway operators and the CTA.
Design criteria shall be confirmed, in writing, with the railway for each site during the conceptual
design phase.
The basic grade separation concept (defined as “that portion of work that is required to provide
adequate facilities for present-day needs at the time of construction or reconstruction of the
grade separation”) forms the basis for apportionment of construction cost with typically an
85/15% split between the two parties (proponent/affected party) for grade separations on new
routes. Any deviations from the basic grade separation concept (such as overbuilding for future
highway lanes or an additional rail track) are typically considered at the full cost of the requesting
party.
Agreement details, including cost apportionment, are typically initiated during the conceptual
design phase of a project, as they can impact the crossing configuration, and are finalized during
latter project stages. It is prudent to seek legal advice when developing agreements. In the case
where the two parties involved cannot come to a consensus on crossing agreements or cost
apportionment, the CTA is consulted. Disputes may be handled by the CTA through facilitation,
mediation, adjudication, or arbitration with varying timeframes for each option. The CTA website
should be consulted for more information and updated references.
Additional Resources:
Alberta Transportation - Rail Safety Website (Alberta Transportation, 2020)
Railway Grade Separations: Application Guidelines Overview (historical reference guide)
(Alberta Transportation, 2004)
Canadian Transportation Agency – Rail Crossing Agreements (Canadian Transportation
Agency, 2020)
Canada Transportation Act (Government of Canada, 1996)
Railway Safety Act (Government of Canada, 2019)
Traffic Safety Act (Government of Alberta, 2018)
Railway (Alberta) Act (Government of Alberta, 2010)
There may be a need to step back in the process to refine elements to derive a feasible option,
particularly after consultation with other technical areas. Multiple options often result due to the
combination of alignments, profiles, and bridge openings. On combined road/bridge projects,
bridge geometric constraints must be integrated into the overall project design and optimization.
The ability for a given bridge option to work with a potential future roadway improvement should
be documented in the conceptual design report. Once a horizontal alignment, road profile, and
bridge opening have been established for a given option, an estimate of road grading impact
should be determined. This will facilitate further optimization of the horizontal alignment and
gradeline, and allow comparison with other options. In general, an approximate balance of cuts
and fills will be preferable, to minimize the amount of borrow/waste material.
Major differences between bridges and roads that can affect horizontal alignment options are:
the presence of bridge barriers (shy offset, sight lines)
Once a horizontal alignment is established, a vertical profile can be added. Key criteria include:
For streams, set bridge elevation to allow for flow, freeboard, and structure depth.
For grade separations, set bridge elevation below to allow for required vertical clearance and
structure depth.
Initial structure depth can be estimated as the maximum span length/20. This accounts for
all elements including girder depth, haunch, deck, wearing surface and crossfall.
Initial pier shaft width estimates can be estimated as the maximum span/30 (used to
determine impact of cross sectional flow area reductions, and estimate fisheries habitat
losses).
Proposed future widening of a structure needs to be considered in the structure depth
allowance (i.e. will widening to the outside reduce clearance?)
Sag/crest K values should meet requirements as per the HGDG.
Bridge elevation can influence the maximum grades on approach roads.
A 1% minimum longitudinal gradient shall be evaluated for all options.
A minimum of 0.5% longitudinal gradient shall be used on a structure to ensure positive
gravity drainage. For grade separations, this is typically accomplished by locating the VPI off
of the structure.
Proposals to use gradients less than 0.5% shall follow the Design Exception process
The bottom of a sag curve shall not be on a bridge.
For bridge replacements with minimal road changes, minor variances in elevation between
the bridge and the road could be accommodated with a variable thickness wearing surface.
This approach should be confirmed with the structural and pavement designers.
The resultant gradient (on diagonal) shall be less than 4% due to safety concerns related to
preferential bridge deck icing.
The minimum length of curve shall be 2 times the design speed to minimize visual effects, as
per the HGDG.
Intersection, stopping sight, and decision sight distance shall be considered as per the HGDG.
The bridge opening is defined by the bridge out-to-out fills and the bottom of the superstructure
(typically reported to the nearest meter). Laterally, the bridge headslope fills shall exceed the
structure width by 2m at the top of fill (bridge clear roadway width + 3m at conceptual design
stage). This provides a base for constructing the abutments, provides lateral support for
abutments and wingwalls, and allows for the installation of barrier systems with appropriate
deflections. Headslope ratios shall be 2:1 (H: V) and sideslope ratios shall be 3:1. An elliptical
transition is provided at the corners between the headslope and the sideslope. Open headslopes
are preferred to allow for improved sight lines, increased flexibility for future improvements, and
better performance from a drainage, operations, and maintenance perspective. The use of
retaining walls or MSE walls is discouraged unless major constraints are noted (e.g. urban setting)
or in the case of railway crossings. A sample is shown below, with more details contained in the
Fills in an active waterway shall be aligned with the expected flow direction at high flows. River
protection works are linked to the location of the headslopes and the skew of the crossing. Fills
that project into the waterway will require protection works. For normal bridge fills with elliptical
transitions, riprap is placed on the fill, with excavation at the ends to ensure the rock is tied in to
the banks upstream and downstream (HS Definition Sketch). An alternative is to use a guidebank,
which maintains a 2:1 slope through the protection works, with the elliptical bridge fill transition
located above and behind the guidebank (GB Definition Sketch). If the bridge fills are outside the
waterway, and a buffer distance is provided, protection works may not be required. In this
scenario, a river engineering assessment including historical bank erosion and lateral mobility
potential is required.
CL Road
CL Road (increasing station in direction of arrow)
(increasing station in direction of arrow)
Left Right
Left Right
GB Extent
Angle
GB Berm
Skew Width GB Nose
HS Berm Wrap Angle
Width
GB Radius
Skew
Conceptual design considerations for structures specifically designed for wildlife passage include:
Assessing overpass vs underpass, considering pros, cons, risks and costs
Determining wildlife specie(s) and associated design parameters (horizontal/vertical
clearance boxes, species and vegetation preferences)
Consider use of natural topography such as natural high areas for overpasses and natural
low areas for underpasses, and use of existing wildlife paths
Consider light and noise protection needs (berms/sound walls)
Consider fencing needs (minimum 3km on either side recommended)
Consider minimizing traffic sightlines and structure noise such as deck joints
Avoid accommodation of pedestrians and wildlife within the same structure (unsafe for
pedestrians and wildlife will be less likely to use the structure)
Further guidance on wildlife requirements can be found within the Environmental Regulation
Section (Alberta Transportation , 2020) of the website.
Structural options should be considered at a high level when developing bridge opening options.
Expected span lengths will confirm assumptions made in setting the roadway profile to
accommodate desired freeboard. Potential pier locations should be assessed for impact on drift
and ice, in-stream berm needs during construction, and proximity to fills. Pier location/type can
have an impact on roadway design including shy line offsets, clear zones, and barrier requirements
along with future interchange plans (directional ramps, bridge widening, and lane expansions).
For stream crossings, bathymetry (ex: thalweg location) and environmental sensitives (ex:
construction berms impacts) can affect pier location. For long span structures, it is beneficial to
consider structure type (steel I-girders, steel box girders, concrete segmental, cable stayed,
suspension, etc.), and associated limitations (maximum spans, span depths, pier connections)
during the conceptual design phase. For these types of projects, preliminary structural alternative
reports often form part of the conceptual design submission.
Stormwater drainage from highway ditches shall be considered in the design of bridge fills and
associated structures such as MSE walls. For rural sites, specific stormwater management
facilities, such as ponds, are not required with more details provide in Best Practice Guideline 11:
Stormwater Management at Rural Bridges (Alberta Transportation, 2020). Rural highways are
designed with open channel ditches, as per the Highway Geometric Design Guide, which have
sufficient capacity to attenuate peak flows originating from roadway surfaces, and the ability to
filter runoff. Ditch drainage is typically directed into low-lying areas adjacent to the channel, and
erosion and sedimentation guidelines apply. Within a grade separated bridge opening, swales are
used between the headslope and roadway shoulder, as per the Roadside Design Guide.
Stormwater ponds may be required at large interchanges in urban areas that significantly increase
impervious area, when regulated release rates are required, or when more stringent quality
control is required (such as upstream of a water intake facility). Depending on the size and risk of
a pond, AEP Dam Safety Regulations may also be triggered. Additional information regarding
stormwater management for urban areas is found in Design Bulletin 16 (Alberta Transportation,
2007) and the Stormwater Management Guidelines (Alberta Environmental Protection, 1999).
Drawing S-1418 (Alberta Transportation , 2017), Installation of CSP and SPCSP Structures, provides
guidance for the design and installation of culvert structures. This drawing is applicable to bridge
sized culverts with a diameter of 3m or less, with site-specific details required for culverts with
greater diameters, or classified as a major bridge such as an open bottomed structure. Conceptual
design considerations for culverts include:
Culvert inverts should be buried one quarter of the rise (D/4) below the average natural
streambed, up to a maximum depth of 1 m, to allow for future streambed changes and ensure
biological connectivity. Exceptions to the recommended burial depth may be considered
when site-specific features require special attention such as shallow competent bedrock or
historical stream degradation.
Culverts have fixed opening dimensions with transition zones located at both ends of a
culvert, as the channel shape moves from the bevel end to a trapezoidal shape that can fit the
natural stream. Transition zones also connect the culvert invert elevations to the natural
streambed elevations. In some cases, berms perpendicular to the roadway may be required
to support the protection works. Channel realignments may also be beneficial in some cases,
to reduce culvert skew (length), improve flow alignment, or reduce property impacts.
Sideslopes shall be designed as per the Roadside Design Guide (Alberta Transportation, 2020),
with a minimum 3:1 slope. Shallower, recoverable 4:1 sideslopes are preferred where not
limited by other sites constraints such as Right of Way. A 4:1 slope through the clear zone,
followed by 3:1 outside the clear zone (with no guardrail) is often a cost effective solution.
Minimum cover for steel culverts shall be in accordance with the current version of the CHBDC
(CSA, 2019) or 300mm, whichever is the greater. The minimum cover shall be taken as the
least dimension between the crown of the culvert and the edge of the roadway shoulders.
Multiple culverts shall only be used if they fit the natural channel (i.e. bed width created
should be similar to the natural channel). Flow expansion and contraction associated with
structures being too small or too big in comparison to the natural channel should be avoided
to minimize erosion and scour risks.
Open bottomed structures shall be larger than the natural channel bedwidth and shall have
mitigation measures to prevent failure due to scour or erosion, as further described in Section
5.9.
Horizontal spacing between adjacent culverts shall be at least 1.0 m or span/10 of the larger
span, whichever is the greater, to allow for the soil envelope to be established and properly
constructed. Site specific backfill design details are required for sites with multiple pipes.
Overflow culverts outside of the natural channel placed at a higher elevation or lateral
distance away shall be considered when aufeis/icing concerns are noted at existing or
adjacent culvert sites.
Guidance on available culvert types and materials can be found in the AT Products List
(Alberta Transportation, 2020) and Section 18 of the Standard Specifications for Bridge
Construction (Alberta Transportation, 2020). Selection of the optimal culvert configuration
that meets the required service life should be based on life cycle analysis and site specific
constraints.
Culvert upsizing to allow for future lining may be a cost effective strategy at sites where open-
cut replacement would be expensive or require extensive traffic accommodation (high fills,
high traffic volume).
Cathodic protection systems are not permitted as service life extension strategies, as they are
difficult and costly to maintain/operate.
Tunnelling shall be consider if the structure diameter is <3m, fill height is >5m, or traffic
volumes are >2000 AADT. A moderate increase in initial cost over an open-cut option may be
worth savings in user costs. Undertake a cost analysis of options and consider risks such as
geotechnical suitability. Costs may be affected by availability of experienced contractors and
equipment.
For larger culvert sizes, fill removal by a bobcat within the advanced WSP may be feasible or
tunnel-boring machines may be required. If tunnelling is proposed, contacting potential
contractors regarding equipment availability and limitations is recommended during the
conceptual design phase.
For cattlepass and pedestrian culvert structures, the minimum vertical clearance should be
2200 mm, or as required by TAC based on the expected users. A flat floor, with a minimum
depth of 150 mm at the midspan, shall be considered for circular culverts. Length should be
determined such that sideslopes terminate at the top of the floor level when no bevel is used.
Surface water should not pond inside the structure. This could be achieved by setting the
inverts slightly above adjacent ground, by longitudinally sloping or crowning the inverts, or by
using ditch drainage when necessary.
Further guidance regarding bridge-sized culverts and buried structures can be found in the Design
Guidelines for Bridge Sized Culverts (Alberta Transportation, 2020), the Corrugated Steel Pipe
Institute (Corrugated Steel Pipe Institute, 2020), the Canadian Concrete Pipe and Precast
Association (Canadian Concrete Pipe & Precast Association, 2020), and local suppliers.
Alternatives shall be assessed during the conceptual design phase to identify the optimal solution
for the site. The comparison shall consider costs, pros/cons, and risks. The comparison will
typically only involve replacement options, as rehabilitation and maintenance options are
generally considered in the bridge assessment phase. Life cycle cost analysis is required for
options with different expected structure lives.
Options may cover a range of alignments, profiles, and bridge openings. They may also cover a
range of structures including steel culverts, concrete culverts, standard bridges, and major
bridges. Costs shall include all project costs at a type ‘A’ level, including:
Structure costs - unit cost times deck area, in plan view
River protection works – unit cost times rock volume
Road Grading - unit cost for balanced grading, unit cost for borrow/waste
Road construction – unit cost times length of road
Geotechnical remediation – lump sum costs based on similar projects
Land/ROW purchase costs – unit cost times area
Utility costs – lump sum for major items paid for by the Department
User costs – construction speed/lane reduction impacts, if these differ for the options or in
urban/high traffic areas
Construction/detour costs
Environmental costs – lump sum mitigation costs
Operations/maintenance costs – if these differ for the options
Throw-a-way costs
Pros and Cons shall address issues not accounted for by the costs, such as:
Highway geometrics – functionality, performance, safety
Hydrotechnical – backwater impacts, erosion, debris, ice
Future plans - flexibility to upgrade in the future, ROW impacts
Route length – temporary, permanent
Construction issues – timing, risks, difficulty, staging, user impacts
Environmental – impacts, approvals, mitigations, sustainability
Geotechnical – unknowns, mitigations, risks
Land/development impacts – sensitive areas, purchase delays
Operations/Maintenance – icing mitigations, future rehabilitation
Safety – potential collision reduction/increase
Adherence to standards – non-conformance, or design exceptions
Risks – for the Owner, stakeholders, travelling public
Appropriate technical experts (road designers, environmental, geotechnical, structural, etc.) shall
be engaged during this phase to ensure considerations are fully assessed. For complex projects,
discussion with the Department prior to completing the option analysis phase is recommended
to minimize rework. This will confirm any additional factors to consider in the comparison, or
other directions from the Department such as schedule or budget limitations.
A recommendation on optimal bridge concept is required within the draft conceptual design
report. After discussion with the Department, direction may be given to proceed with the
recommended option, a different option, or evaluation of an additional option. Documentation
of the rationale for the final recommended plan shall be included in the final report.
For situations where engineering analysis reveals that standards in this guideline are not
appropriate for a specific project, the design exception process shall be followed, as per the
Department’s Design Exceptions Guideline (Alberta Transportation, 2018). Subject Matter Experts
shall be informed for any proposed deviations from current standards or guidelines and will make
a determination of whether a formal design exception submission is required. Documentation,
either through a formal design exception submission or through content within the Conceptual
Design Report, should include an appropriate level of engineering analysis, evaluation of
alternatives (for example an option to meet the standard in comparison to an option to not meet
the standard), risk assessments, mitigation strategies, and recommendations. The Subject Matter
Expert shall determine if the design exception is approved, with final sign off required by the
Executive Director of the Technical Standards Branch.
The bridge conceptual design process addresses the principals involved in the design exception
process, and a formal design exception request should not be required in most cases. It is the
responsibility of the project manager to ensure that appropriate effort has been undertaken and
sufficient documentation has been made from a liability perspective.
Conceptual design items that shall be clearly communicated and justified include:
Freeboard at bridges < 0.3m
Freeboard at culverts <0m
Non-standard river protection works details
Variable cross section on bridge deck
Roadway spiral on a bridge deck
Gradient on bridge deck < 0.5%
Resultant gradient on bridge deck > 4%
Bridge width less or more than standard
Bridge vertical clearance less or more than standard
The report shall summarize acquired data, analyses, options considered, and justification of the
recommended option. While the report content will vary depending on the project scope and
specific site constraints, all reports shall identify the following as a minimum:
Sketches shall include a site plan, elevation view, road profile view, and streambed profile (as
required). In the case of a culvert, the elevation view shall run along the culvert. Additional
sketches and details may be required for new road alignments, channel realignments, and river
protection works. Superimposing design elements, such as headslope protection works or bridge
fills, on airphotos and 3D views may be of value to communicate the proposed option.
The typical contents are as follows, unless otherwise specified in a project terms of reference:
Site Plan:
Extents – include fill transitions to roadway along road and tie-ins to bank along streams
Centerline of road with at least two station points shown*
Top and bottom of stream banks, as required
Bridge fill extents*
Headslope protection works
Utilities (with labels)
Existing and required right-of-way extents*
Land ownership
Existing bridge, as required*
Skew angle of opening relative to roadway*
Developments or significant natural features
Bridge and Roadway Elevations (cross section along bridge and road alignment):
Extents – same as site plan along the roadway
Elevation on vertical axis, station along road on horizontal axis
Gradeline of existing road centerline (if applicable)*
Gradeline of proposed road centerline*
Profile of existing natural ground along proposed road centerline (if applicable)*
Profile of existing natural ground offset on both sides of proposed roadway centerline (should
be clear of influence from existing fills, e.g. 20 to 30m offset)
Bridge headslopes from fill to natural ground (slope, station and elevation for top of fill points)
Design highwater elevation, as required*
Protection works (if applicable), detailing thickness, top/bottom elevations, apron length
Outline of existing bridge, including deck, abutments, piers and foundation, if applicable*
Geotechnical test holes
Road gradients at estimated bridge ends
For culverts – show invert elevations, burial, channel transition lengths, and all elements that
affect culvert length, including sideslopes, berms, road cross section and clearances.
Samples of functional planning sketches and bridge conceptual design sketches, are found in the
Appendix A. Historically, Design Data drawings were using by the Department but these were
found to be costly and provided limited additional value in comparison to well developed sketches
and general layout drawings generated in the detailed design phase.
Channel Capacity:
S = 0.014
B = 1.5 m, h = 1.2 m, Th = 6 m
n = 0.055
HW was noted in 1908 (extent unknown), 1924 – causing both road approaches to wash
out, 1989 – backwall scour noted, estimated highwater mark of 1.7m, and 2005 – no
damage noted.
There are no WSC gauges on Cripple Ck. There is a nearby gauge on Trout Creek
(Gauge 05AB005, DA~440 sqkm) with records from 1908 – present. The largest event
recorded were in June 2005 (Q~60cms), June 1923 (Q~40cms), and June 1995
(Q~30cms). The largest storms in the storm database for this area are June 1953, May
1923, and June 1963 with about 110mm of rainfall.
The basin is located within areas of 100mm (rainfall) and 15mm (snowmelt) zones.
Qp = Qp1 + Qp2
Qp = 14cms
Conclusion :
Consultant: CE Agreement #:
APPROVALS
Required? Status Reference Number
AEP – Water Act
AEP – Public Lands Act
Transport Canada
Fisheries and Oceans Canada
Right of Way
Other
Other:
Cost Estimate:
Note that some references within these documents are outdated, including, but not
limited to:
Navigable Waters Protection Act
Currently Canadian Navigable Waters Act
Hydrotechnical Design Guidelines (AT)
Superseded by the Bridge Conceptual Design Guidelines
Fish Habitat Protection Guidelines for Stream Crossings (AT)
Removed from circulation, contact AT’s Environmental Regulation group
for more information
Alberta Infrastructure and Transportation (AIT)
Currently Alberta Transportation (AT)
Introduction
Alberta Transportation (AT) has proposed a fish passage assessment methodology that
compares the mean flow velocity in the pipe to that of the typical channel. Comparable
mean velocities would suggest that a given culvert configuration would not be a velocity
barrier to fish passage. In addition, culverts providing similar mean flow velocities to the
channel at the most frequent flows should minimize the impact on overall river processes.
However, concerns have been expressed about the validity of the mean velocity being
used to compare fish passage in channels and culverts. Culverts provide a length of
constant section, profile, and alignment that in many cases will exceed that of a natural
channel, and this may affect the availability of low velocity zones in the flow. In order to
quantify this impact, a study has been undertaken to compare the distribution of flow
velocity over a cross section between culverts and channels. Knowledge of the velocity
distribution at culverts may also provide useful guidance in the application of fish
swimming performance curves.
Methodology
Data on the distribution of flow velocity over a cross section for both culverts and open
channels is available from several sources. Data on velocities within round corrugated
metal pipes has been published based on field studies such as Katopodis et al (1978) and
Behlke et al (1991), and based on laboratory studies such as Barber and Downs (1996)
and Ead et al (2000). Laboratory studies have the advantage of more control over flow
conditions, but come with unavoidable scale effects. Additional data for embedded
culverts has been published by House et al (2005) and Magura (2007).
A convenient source of open channel velocity distribution data covering a wide range of
flow and channel geometry conditions is the gauging data collected by the Water Survey
of Canada (WSC). Data from about 40 gaugings at sites across Alberta have been
analyzed. Some recent data sets, collected using Acoustic Doppler Current Profiler (ADCP)
equipment at the Hwy 63 over Athabasca River crossing near Fort McMurray in Alberta
provide an additional source of data.
Analysis
Figure 3 shows a plot of %A versus vmx/V for all culvert and WSC open channel data. Each
data-set shows some scatter, but significant trends are visible. Best-fit regression lines
have been added to the plot to clarify the differences between the two sets of data. Little
difference was observed between culvert data derived from field and laboratory
investigations, so all culvert data were combined into one set. The ADCP channel data set
compared well with the WSC channel data, but was not included on the plot due to the
influence of a channel constriction.
Two areas of difference between the two data-sets are readily apparent from Figure 3.
The first difference is that culverts provide a somewhat smaller %A for vmx/V ratios less
than 1.0. The magnitude of this difference can be seen in Table 1. Values of vmx/V are
reduced by about 10% of V for %A values in the 10 to 30% range. This range is likely of
most interest to fish passage assessment. The second difference is that channels tend to
have much higher vmx/V ratios at higher values of %A. This difference in the high velocity
regions is about 5 times larger (~ 50% of V), but is of less interest to fish passage
assessment.
The WSC channel data shows an increase in scatter when compared to the assembled
culvert data. As the WSC data-set covers a wide range of site, a brief investigation of
potential factors that may affect the velocity distribution was undertaken. Separate plots
of data within ranges of parameters such as channel width, depth, mean velocity (see
Figure 4a), and cross section aspect ratio (width to depth, Figure 4b) were prepared.
However, no significant trend or reduction in scatter with any of these parameters was
observed.
Conclusion
A comparison between flow velocity distribution in culverts and open channels was
undertaken, based on published data-sets. Best fit lines to the two data-sets on %A vs.
vmx/V plots show differences in regions both above and below the mean velocity. The
difference in the region of interest to fish passage assessment is about 10% of the mean
velocity, with the velocities in these regions being a higher percentage of the mean
velocity at culverts than at natural channels. This difference is within the range of
accuracy of determination of design parameters and hydraulic modelling results.
Therefore, it is concluded that the practice of assessing fish passage based on comparison
of mean velocity between the culvert and the channel appears to be valid. Culvert
embedment with granular substrate may reduce the difference in velocity distributions
between culverts and channels.
Barber, M.E., and Downs, R.C., 1996. “Investigation of Culvert Hydraulics Related to
Juvenile Fish Passage”. Washington State DOT Technical Report No. WA-RD 388.2.
Behlke, C., Kane, D.L., McLean, R.F., and Travis, M.D., 1991. “Fundamentals of Culvert
Design for Passage of Weak-Swimming Fish”. FHWA Research Report No. FHWA-AK-RD-
90-10.
Ead, S.A., Rajaratnam, N., Katopodis, C., and Ade, F., 2000. “Turbulent Open-Channel Flow
in Circular Corrugated Culverts”. Journal of Hydraulic Engineering, Vol. 126, No. 10, p750.
House, M.R., Marvin, R.P., White, D., 2005. “Velocity Distributions in Streambed
Simulation Culverts used for Fish Passage”. Journal of the American Water Resources
Association, 41(1):209-217.
Katopodis, C., Robinson, P.R., and Sutherland, B.G., 1978. "A study of model and
prototype culvert baffling." Canadian Fisheries and Marine Service, Technical Report 828
WSC Data – gauging records, Alberta Region, Water Survey of Canada, Environment
Canada
100
90 Natural Channels
(WSC)
80 Round Corrugated
Culverts
70
60
%A
50
40
30
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
vmx/V
Bridge Conceptual Design Guidelines 79
Classification: Public
Figure 4a – Velocity Distribution - Effect of Mean Velocity
100
90 V < 1.5m/s
80 V > 2m/s
70
60
%A
50
40
30
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
vmx/V
Bridge Conceptual Design Guidelines 80
Classification: Public
Figure 4b – Velocity Distribution - Effect of Channel Shape (T/Y)
100
90 T/Y < 10
80 T/W > 10
70
60
%A
50
40
30
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
vmx/V
Bridge Conceptual Design Guidelines 81
Classification: Public
Figure 5 – Velocity Distribution - Effect of Roughness
100
60
%A
50
40
30
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
v/V
Bridge Conceptual Design Guidelines 82
Classification: Public
Estimation of Navigation Clearance Box Reference Water Level (Originially
Published 2011)
Introduction
Stream crossings over streams that have been deemed navigable by Transport Canada require
approval under the Navigable Waters Protection Act (NWPA). Part of the submission to Transport
Canada involves identification of a clearance box that shows that navigation is provided for. This
clearance box sits on top of a reference water level. Current TC practice appears to be to use the
water level associated with 1 in 2 year flood (Q2) as this reference water level (TSH, 2006).
Alberta Transportation (AT) has identified many issues with the use of statistical techniques
involving return periods to determine design flow values for stream crossings (AT, 2007). Due to
these issues, AT have established Hydrotechnical Design Guidelines that are based on physical
parameters and historic observations (AT, 2006). Although the Q2 avoids many of the
extrapolation issues involved with structure design flows, it is still not possible to determine Q2
accurately and with confidence at ungauged sites using statistical basin transfer techniques. In
order to comply with the TC practice for clearance box determination, a practice for determining
Q2 and its associated water level is required. This document describes such an approach.
The use of a flood for the navigation reference level appears questionable to Alberta
Transportation. Flow conditions at this level may be unsafe for navigation, and the amount of
time that such a flow is exceeded is very small relative to the navigation season. In order to
quantify this impact, a study of the duration of flow exceeding portions of Q2 has been
undertaken.
Approach
In order to relate Q2 to structure design flows, WSC gauges in Alberta with at least 20 years of
record and in close proximity to stream crossings were identified. The selection of gauges was
limited to a drainage area of less than 5000km2, as most sites that exceed this value have gauge
records sufficient to determine Q2. Some additional sites were eliminated due to known influence
from water management structures that may affect runoff response and low flow conditions. For
each of the selected sites, the design flow was determined by applying the AT Hydrotechnical
Design Guidelines to the nearby structure.
The 1 in 2 year flood is equivalent to the mean annual flood. It can be estimated as the average
of annual flood peaks published for each gauge. However, the peak flow record is usually highly
skewed due to the presence of large flows that may be an order of magnitude higher than typical
annual peak flows. Therefore, the median of the annual peak flows has been selected in this
study, as it is more representative of the typical mean annual flood at each site.
Values for Q2 and structure design flow (Qdes) have been compiled for all identified WSC gauges,
and are shown in Table 1. Physical properties of each site, such as typical channel geometry, were
also compiled, in order to facilitate analysis of the results.
The daily flow gauging records for each site were also examined to quantify the duration of flows
exceeding certain values. A range of flows up to and exceeding Q2 were used as benchmarks. The
Figure 1 shows the ratios plotted against drainage area. No significant trend is visible. For larger
drainage areas (i.e. > 5000km2), it would be expected that the ratios would tend to decrease with
increasing drainage area, as the storm inputs are typically confined to a portion of the basin and
flood peaks will attenuate as they go downstream.
Figure 2 shows the ratios plotted against channel slope. Channel slope is a strong indicator of
hydraulic capacity. It would be expected that there would be less routing of extreme events with
higher slopes, yielding higher ratios. As with Figure 1, there is significant scatter, as there are
other significant factors that affect these ratios. However, a trend is apparent at the lower
envelope of the data.
Approximately 5 sites appear as outliers at the low end of the range of ratios. A common
characteristic of these sites is that they all have significant portions of their drainage basins at
very high altitudes (> 1500m). The high altitude areas are less likely to see high intensity rainfalls,
and these areas may provide snowmelt/glacier-melt flows to the basin throughout the navigation
season, resulting in lower ratios.
Approximately 3 sites have ratios > 20. Two of these sites are high capacity channels that have
little basin storage and infrequent runoff responses. The other is a basin that requires a high
volume input (e.g. snowmelt) to generate significant runoff, but this occurs infrequently.
Due to the significant scatter, a best fit line would have little relevance for prediction at ungauged
sites. However, a lower envelope line could be of use as a conservative estimate (i.e. high flow)
for the purposes of defining the navigation clearance box reference level. Using Figure 2 as a
reference, a lower envelope line can be constructed between a ratio of 4 and 10 as the channel
slope increases from 0 to 0.01. This line results in the equation:
where S is the channel slope. Although there are likely to be very few navigable crossings with S
> 0.01, a ratio of 10 can be applied for these sites, if necessary.
These results show that flows are lower than Q2 98% of the time, and are lower than 0.5Q2 ~ 90%
of the time, during the navigation season. This suggests that using Q2 for assessment of navigation
is a very conservative approach, and that using a smaller value (even as low as ½ of Q 2) would
have little impact on time available for navigation.
For sites with basins with significant drainage area higher than 1500m in elevation, lower ratios
may be applicable. For sites with channel slopes > 0.01, Q2 = Qdes / 10 can be used.
Once Q2 has been determined, the reference water level can be determined using the typical
channel parameters as identified in the calculation of Qdes in an open channel flow hydraulics tool,
such as the AT Channel Capacity Calculator tool. The reference water level can be taken as the
design water level minus the difference in depth between Ydes and Y2.
As the ratio is based on a lower envelope curve, the resulting reference water level should be a
very conservative approach to meeting the NWPA Q2 practice. In addition, an analysis of flow
durations indicates that navigation availability is not sensitive to this value. In cases where
navigability requirements appear to govern the minimum gradeline across a bridge, further
refinement of the reference water level and clarification of the navigation clearance criteria
should be considered.
References :
AT, “Hydrotechnical Design Guidelines for Stream Crossings”, Alberta Transportation WWW site
(http://www.transportation.alberta.ca/2646.htm), September 2006.
TSH. “Determining the Minimum Navigational Clearances Required for Un-Powered (Paddle)
Vessels”, Transport Canada Internal Use Document, May 2006
05BM008 CROWFOOT CREEK NEAR CLUNY 1374 21 6985 50.83 112.76 7 0.0008 15 3.5 30 4.5 1.9 200 28.6
16
DA < 5000sq. km
14
High Elevation
12
10
Qdes / Q2
0
0 1000 2000 3000 4000 5000
Drainage Area (sq. km)
16
14
12
10
Qdes / Q2
6 DA < 5000sq. km
High Elevation
4 Envelope Line
0
0 0.002 0.004 0.006 0.008 0.01
Channel Slope
1.2
0.8
Q/Q2
0.6
0.4
0.2
0
0 10 20 30 40 50 60 70 80 90 100
% Time Not Exceeded
Introduction
The provision of fish passage is a requirement for most proposed culvert and bridge installations in
Alberta, depending on the recommendations from a qualified environmental professional. A fish passage
design discharge is required to assess a structure’s performance and evaluate fish passage through the
structure. The difficulty lies in selecting an appropriate maximum fish passage discharge for culverts
where insufficient historical hydrological information exists.
In 2006, Alberta Transportation undertook a statistical analysis in an attempt to establish a depth based
approach for the provision of fish passage in consultation with the Department of Fisheries and Oceans.
This study used physical channel characteristics (bed width, slope, etc) to calculate discharge based on
the Manning’s Equation. Hydraulic performance was then evaluated against a range of discharges based
on flow depth, typically between 0.5 m and 1.0 m. The culvert was assumed to be passable by fish if the
section averaged velocities at the prescribed depth of flow within the culvert were the same, as or lower,
than the averaged velocity within the typical natural channel.
The work from 2006 has been continued here in an attempt to develop an alternate fish passage design
discharge (FPDD) that could be calculated based on physical channel characteristics and applied to all sites
across the Province.
Background
Various methodologies have been used in the past to determine the Fish Passage Design Discharge
(FPDD). Most recently (~2006 to 2014), comparing velocities at a range of flow depths from 0.5 to 1.0m
was used. Previous to this, the 1 in 10 year 3 day delay discharge (3Q10) was used. This method was
related to fish swimming performance and, oftentimes, resulted in very low velocities in the culvert
relative to the stream.
There are a couple of significant drawbacks in the application of the 3Q10 discharge. The first drawback
is the assumption about the applicability of its use. The use of a maximum delay of three days appears to
have first been recommended in a consultant report (Dillon, 1979) for construction of the Liard Highway
in Alaska which led to further studies in the region. In particular, a population of arctic grayling in Fish
Creek in Alaska was studied with the recommendation that arctic grayling not be subjected to an artificial
delay of more than 3 days (Fleming, 1991). The recommendation did not apply a return period to this
delay. It is not clear where the return period was originally recommended but it is assumed that it is
related to the consecutive risk of failure of the spawning migration. The 1:10 year return period was used
in calculation of fish passage discharges in the Fish Habitat Protection Guidelines for Stream Crossings
(Alberta Government, 1992).
The application of 3Q10 discharge assumes that this same delay period should be applied at all locations
for all fish species; there have been no additional studies completed to suggest that this is the case.
Anecdotal reports from staff at Alberta Environment and Park’s Fisheries Management Branch (FMB) have
suggested that some species of fish may wait longer to move upstream. No studies have been completed
in Alberta to suggest what length of delay would be appropriate.
The 3Q10 methodology overall generally did not work unless baffle systems were installed within culverts
to reduce velocities. These baffle systems were difficult to install, created maintenance issues including
acting as debris catchers, reduced the lifespan of the culvert, and caused safety issues during inspections.
The majority of baffles installed in the past have been abraded due to drift and bedload to the point of
being ineffective, or have washed away completely from within the structure.
Methodology
The Water Survey of Canada has hundreds of active gauging staging across the provinces which were
relied upon for the current study. Gauges selected for evaluation were required to be on a stream where
it would be feasible to install a culvert. Based on this requirement, streams with a bed width of less than
8 m were used. The second criterion was a minimum gauge record of 30 years in order to evaluate the
statistical record with some statistical reliability.
For the purpose of this review, mean daily discharges occurring during the spring spawning migration
were considered, as this is the timeframe when flows are the highest for an extended period of time. The
period selected by DFO staff for evaluation was March 1 to May 15 (76 days), noting that this should cover
the migration period for all spring spawners along with annual variations in the run timing.
Several different discharges were evaluated for this exercise to aid in assessing an appropriate maximum
FPDD. During the evaluation process five discharges were selected for further evaluation: 3Q10, 7Q10,
Q90, Q95 and Q98. The methodology for determining these discharges is summarized below and their
advantages and disadvantages are included in the Analysis and Discussion section.
The 3Q10 is based on the assumption that fish will only sustain a delay of up to 3 days before giving up on
their migration, and the females begin reabsorbing their eggs (Fleming, 1991). The 3Q10 is evaluated by
selecting the fourth highest consecutive mean daily discharge during the spawning period for each year
of record. These discharges are then used to complete a frequency analysis, typically using a Pearson III
distribution. For this evaluation both the Pearson III distribution and a rank and percentile evaluation
were employed with no significant variation. (Alberta, 1992)
The 7Q10 is based on similar assumptions to the 3Q10 only it allows for a delay of 7 days. The 7Q10
discharge is evaluated in a similar manner to the 3Q10 with the exception that the eighth highest
consecutive flow during the spring spawning migration period is selected instead of the fourth. This
assumption was used based on anecdotal reports from some staff at Alberta Environment and Sustainable
Resource Development (AESRD) Fish and Wildlife that some species of fish can withstand a delay of 7 days
or possibly more.
The Q90, Q95 and Q98 discharges are equivalent to the 10% exceedence, 5% exceedence and 2%
exceedence discharge respectively. Q90, Q95 and Q98 were used as the terminology to indicate that fish
passage would be achievable 90, 95 and 98% of the time in a given year.
The flow depth for each discharge, whether delay or exceedence discharge, was calculated based on the
physical channel characteristics to provide a consistent basis for comparing the discharges to one another.
As fish require adequate depth of flow for migration, the minimum depth of flow applied for determining
the FPDD was 0.2 m. Below this level, there may be insufficient water depth for fish migration to occur.
The migration window selected for this analysis is 76 days long. Given that not all fish move at the same
time, the probability of each species encountering the delay discharge is quite low. For example, the
probability of fish encountering the 3Q10 is 4/76 for the given year multiplied by the frequency of 1/10
years. This results in a probability of 1/190 or 0.5%. Similarly, the probability of fish encountering the
7Q10 is 1% (8/76 x 1/10), double the probability of encountering the 3Q10. The 7Q10 is more lenient
than 3Q10 with respect to the allowable delay but is still conservative when considering the probability
of encountering the discharge.
Due to the low probability of encountering the delay design discharge and because most species of fish in
Alberta move to spawn after the peak discharge (and many of them are less sensitive than arctic grayling),
the Department is suggesting that a less conservative discharge than the 3Q10 be applied for assessing
fish passage.
Two streams of analysis were taken for this discussion. The relationship between discharge and the
physical parameters of the stream was explored to determine if there is a strong enough correlation that
physical parameters can be relied upon to develop design criteria. Based on observations made during
the first portion of the analysis, the relationship between the exceedance discharges and the delay
discharges was also evaluated.
Slope was selected as the independent variable for the analysis as it is easily measured and it showed a
reasonably good relationship to the depth. The data sets for each of the depths at the defined discharge
were plotted against the corresponding slope. A linear regression was completed on each data set with
poor results (Figures 1 and 2).
Due to the poor correlations of all the data sets, a more conservative approach was taken and a best fit
envelope was established for each data set (Figures 3 and 4). The two outliers appear to be ice affected
water levels and were ignored.
With further examination of Figure 2, approximately 70 percent of the 3Q10 data points are below the
7Q10 envelope. This means that for the majority of sites, the 7Q10 envelope is more conservative than
using the 3Q10 method.
The recommendation of an envelope to use for determining the FPDD depth must consider the risk to the
fish populations. The Q98 has a 2% probability of exceedance during the spring spawning migration. Also,
earlier in the discussion, it was shown that during the 3Q10, now Q98, there was a 0.5% probability of fish
encountering the discharge. Likewise, for the Q95 there is a 5% probability of exceedance during the
spring spawning migration with a 1% probability of fish encountering the discharge. The likelihood of
experiencing the Q90 discharge is 10% on an annual basis. The probability of fish encountering this
discharge was not determined. The probability of encountering the delay is independent of the
exceedance probability. If a delay discharge were determined to correlate with the Q90, the delay would
be greater than 7 days. Increases in annual probability of occurrence, and associated delay probabilities,
may increase the risk of year upon year failures in the spawning and therefore could have significant
negative impacts on the fish populations. Without significant, and costly, research this cannot be
determined; as such a moderate approach to calculating the FPDD is suggested.
Recommendations
All of the exceedances evaluated in this study occur during high flow events with low probability where it
is possible that fish may choose to wait rather migrate. Ideally, risk to fish populations and fishery
management objectives during these types of events, should be considered. Without studying the
movements of fish populations at specific sites in the field, this application of risk is somewhat subjective.
Based on the discussion, it is clear that the 3Q10 (Q98) discharge is the most conservative approach, the
Q95 (7Q10) is somewhat conservative, and the Q90 is the least conservative of the discharges evaluated,
although all are still considered to be flood events.
Due to the uncertainty with risks to the fish populations it is recommended to use the moderate approach
to determination of the FPDD, the Q95 envelope. The envelope curve, Figure 6, is equated to be:
YFPDD = 0.8 – 34.3 * S, for S < 0.017
YFPDD = 0.2, for S ≥ 0.017
While this is a somewhat conservative approach it will help to ensure that fish passage is achievable in
years when larger runoff events occur. These larger events are considered to be essential for high
recruitment for juvenile fish by staff at AEP Fish and Wildlife Division. This approach will also allow for
consideration of site specific channel characteristics in a consistent manner.
1.40
1.20
1.00 Y@3Q10
Y@7Q10
Linear (Y@3Q10)
0.80
Linear
R3Q10²(Y@7Q10)
= 0.4166
Depth (m)
R7Q10² = 0.3219
0.60
0.40
0.20
0.00
0.0000 0.0050 0.0100 0.0150 0.0200
Slope (m/m)
Bridge Conceptual Design Guidelines 95
Classification: Public
Figure 2: Correlations for Depth as a Function of Slope at Exceedence Discharges
1.40
Y90
Y95
1.20
Y98
Linear (Y90)
Linear (Y95)
1.00 Linear (Y98)
RY90² = 0.1033
RY95² = 0.2186
0.80
R² = 0.3966
Depth (m)
0.60
0.40
0.20
0.00
0.0000 0.0050 0.0100 0.0150 0.0200
Slope (m/m)
1.40
Y@3Q10
Y@3Q10 envelope
1.20
7Q10
Y@7Q10 envelope
1.00
0.80
Depth (m)
0.60
0.40
0.20
0.00
0.0000 0.0050 0.0100 0.0150 0.0200
Slope (m/m)
1.40
1.20
Y90
1.00 Y90 envelope
Y95
Y95 envelope
0.80 Y98
Q98 envelope
Depth (m)
0.60
0.40
0.20
0.00
0.0000 0.0050 0.0100 0.0150 0.0200
Slope (m/m)
1.20
1.00
Delay Discharge (cms)
0.80
0.60
3Q10 vs Q98
0.40
7Q10 vs Q95
Linear (3Q10 vs Q98)
Linear (7Q10 vs Q95)
0.20
R² = 0.7141
R² = 0.3969
0.00
0 0.2 0.4 0.6 0.8 1 1.2
1.4
1.2
0.8
Depth (m)
0.6
YFPDD = 0.8-34.3 * S
0.4
0.2
0
0 0.005 0.01 0.015 0.02
Slope (m/m)
Classification: Public
Comparison of 3Q10 to Depth-Based Approach for Fish Passage Evaluation (Originally
Published 2008)
Introduction
Poor culvert installations can result in barriers to fish passage. Major factors that present
a barrier are hanging outlets which result in a break in connectivity, and constrictive
openings which can result in flow velocities that exceed the ability of the typical fish to
traverse. Current design standards for burial and end protection works, combined with
appropriate construction, have eliminated most of the concern with hanging outlets.
However, issues relating to the ability of fish to pass upstream through culverts under
certain hydraulic conditions remains.
One of these issues is the magnitude of the fish passage design flow. The concept of the
1 in 10 year, 3 day delay flow (3Q10) has been used in Alberta for hydraulic assessment
of fish passage dating back to 1992. The 3 day delay portion of the flow was based on the
observation that some fish types would not tolerate a delay of more than 3 days before
abandoning their attempts to pass a blockage. The 1 in 10 year portion was based on the
intention that it was permissible for fish passage to be blocked for no more than 1 year
out of every 10, on average.
Tools have been developed to facilitate the calculation of 3Q10 values at gauged sites.
However, due to the relatively small number of gauges and the great variance in
hydrologic response within and between basins, a great range of flow estimates can result
for any given culvert site. The two biggest factors in this variance are the selection of
nearby gauges and the drainage area transfer exponent. As this approach is statistical in
nature, there is no ability to make use of physical indicators of hydraulic response at the
site, such as channel geometry and slope.
Due to the difficulties in transferring 3Q10 values and the lack of assigned return period
with the AIT approach, direct comparison at any given site is difficult. However,
comparing the results of the two approaches at actual gauge sites with long records
should eliminate most of these issues. A comparison between the results of the statistical
approach and the physical approach at these sites may provide some context on the
application of the physical approach at the many un-gauged potential culvert sites across
the province.
For the purposes of this study, selected sites should have relatively long records to
minimize sampling issues, be in the range of stream where culverts are a feasible option,
and be on streams for which channel data is available. Therefore, the following criteria
have been established :
These criteria result in a list of 41 sites. The geographic distribution of these sites can be
seen can be seen to cover most of southern Alberta (Figure 1).
3Q10 flow values and associated flow depths have been developed for each site. Daily
flow data have been extracted from the WSC Hydat software, and used within the AIT
QFish tool to identify the 3 day delay flows for each year within the specified fish passage
period. The tool identifies and reports the highest value during the period for each year
that is the 4th highest out of any continuous 7 day group of daily flows. The Pearson III
distribution is then applied to the data to calculate the 1 in 10 year value (exceedance
probability of 0.1).
Fish passage periods have been provided by Rick Orr of DFO, specified as either spring or
fall. Spring dates have been set to the period of March 1 to May 15, and fall dates have
been set to between September 1 and October 30. The spring dates were assigned to the
few sites for which no dates were specified, as spring flows are typically higher than fall
flows.
The 3Q10 flow values have been converted to flow depths using hydraulic calculations.
This process requires estimation of channel geometry data at each site. Channel slopes
have been extracted from the Digital Terrain Model (DTM) based stream profiles in the
AIT HIS tool. Typical channel cross section data, in the form of bed width, top width, and
bank height have been estimated using a variety of sources, including :
The magnitude of flow depths associated with 3Q10 flows have been compared to the
AIT approach, which recommends evaluation of fish passage over a range of flow depths
between 0.5 and 1.0m. In addition, trends between these 3Q10 flow depths and some of
the channel parameters have been investigated. It is hoped that these observations will
help explain the results and provide additional confidence in the application of the AIT
approach.
Results
The results of this study are summarized in Table 1. The gauge column presents the WSC
gauge number, and the description column presents the stream name and approximate
location. The columns in the “Channel Parameters” group include the channel slope (S),
the channel bed width (B), the channel top width (T) at the channel bank height (h), and
the Manning roughness coefficient (n). The “3Q10” group of columns includes the fish
passage period (Period), the 3Q10 flow (Q), the calculated flow depth at 3Q10 (Y), and
the mean channel velocity at 3Q10 (V). For interest, results of the AIT approach are
presented in the “Q at Flow Depth” group, with the flow corresponding to flow depths of
0.5m, 0.8m, and 1.0m presented.
The average flow depth corresponding to 3Q10 is 0.65m with a standard deviation of
0.27m (meaning 2/3 of the data points would fall between 0.38m and 0.92m). The
minimum flow depth is 0.08m and the maximum is 1.28m. The 3Q10 flow depth exceeds
Y = 1.0m at 2 sites (gauges 05EC002 and 05GA003).
Investigation of trends with channel parameters reveals correlation between the flow
depth at 3Q10 and channel slope and width. Figure 2 shows a plot of 3Q10 flow depth
vs. channel slope, and although there is significant scatter, flow depth tends to decrease
with increasing channel slope. Figure 3 shows a plot of 3Q10 flow depth vs. mean channel
width. Again there is significant scatter, but it can be seen that 3Q10 flow depth tends to
increase with increased channel width.
The magnitude and range of equivalent 3Q10 flow depths appears to match the proposed
AIT criteria of Y = 0.5 to 1.0m quite well. The only 2 significant exceptions are at sites with
fairly wide channels where round culverts may not be competitive. In addition, the
channel slopes are relatively flat at these sites, resulting in velocities that are likely to be
less problematic for fish passage even when the flow depth is higher than 1.0m.
The observed trend between 3Q10 flow depth and channel slope is consistent with
expected hydraulic performance. At steep channels, runoff response is usually rapid and
large events generally occur in the summer, resulting in relatively lower values of 3Q10.
This combined with the higher flow capacity at shallower depths due to the increased
slope would lead to expectations for typically shallower depths. This observation suggests
that the proposed 0.5 to 1.0m range is likely somewhat conservative for evaluation of fish
passage at culverts on steeper channels.
Similarly, the observed trend between 3Q10 flow depth and channel width appears
reasonable. Wider channels will typically have larger basins with more storage potential,
resulting in relatively higher 3Q10 values. In addition, larger channels tend to have higher
banks to convey the higher runoff volume, with Y = 1.0m being a smaller fraction of bank-
full flow than at smaller channels. Many of the gauge sites examined have channel widths
that exceed the typical channel parameters of sites were culverts are considered a viable
solution, as there are very few gauges in the province on streams with mean widths less
than 5m. The observed trend between 3Q10 and channel width suggests that the
proposed 0.5 to 1.0m range is somewhat conservative for evaluation of fish passage at
culverts on relatively narrow channels.
Based on these results, it appears that the AIT flow depth approach is consistent with the
principles behind the 3Q10 flow. The value in the AIT approach is that it can be applied
at most sites based on derivable physical parameters. As the 3Q10 approach is statistical,
it loses consistency and physical relevance for sites that are not in the proximity of a long-
term runoff gauge. This represents most bridge-size culvert sites in Alberta.
Conclusion
Flow depths at 3Q10 flows at all long-record WSC gauges in Alberta over the size of
channel where culverts may be a feasible option have been calculated in this study. These
depths are consistent with those proposed for use with the depth-based AIT fish passage
flow assessment technique. For culverts on relatively steep or narrow streams, the AIT
approach appears to be somewhat more conservative than the 3Q10 approach. It
therefore appears that the AIT approach meets the goals of the 3Q10 technique. The AIT
approach has the advantage of being based on physical parameters at a site, resulting in
more consistent and defendable design parameters at sites lacking WSC gauge data,
which represents the vast majority of culvert installations in the AIT system.
Classification: Public
(m) (m) (m) (cms) (m) (m/s) (cms) (cms) (cms)
05AB005 TROUT CREEK NEAR GRANUM 0.0030 8 15 1.2 0.040 Spring 3.3 0.47 0.75 3.7 8.4 12.6
05AB013 BEAVER CREEK NEAR BROCKET 0.0075 5 12 1.0 0.050 Spring/Fall 2.0 0.39 0.80 3.1 7.5 11.5
05AB028 WILLOW CREEK ABOVE CHAIN LAKES 0.0085 8 15 1.5 0.045 Spring/Fall 7.1 0.59 1.30 5.4 12.2 18.1
05AB029 MEADOW CREEK NEAR THE MOUTH 0.0030 4 8 1.0 0.045 Spring 1.2 0.42 0.60 1.7 3.8 5.8
05AC023 LITTLE BOW RIVER NEAR THE MOUTH 0.0007 8 15 1.5 0.040 Spring 3.5 0.75 0.48 1.7 3.8 5.6
05AD016 DRYWOOD CREEK NEAR TWIN BUTTE 0.0110 5 10 1.5 0.045 Spring/Fall 3.1 0.44 1.21 3.8 8.6 12.8
05AD035 PRAIRIE BLOOD COULEE NEAR LETHBRIDGE0.0045 6 12 1.0 0.045 Spring 2.8 0.47 0.79 3.1 7.2 10.9
05AE005 ROLPH CREEK NEAR KIMBALL 0.0035 5 10 1.0 0.045 Spring 5.0 0.78 0.92 2.3 5.2 7.9
05AG003 EXPANSE COULEE NEAR THE MOUTH 0.0032 4 8 1.0 0.045 Spring 2.1 0.56 0.73 1.7 4.0 6.0
05AH037 GROS VENTRE CREEK NEAR DUNMORE 0.0020 7 14 1.5 0.040 Spring 5.9 0.80 0.83 2.6 5.9 8.8
05AH041 PEIGAN CREEK NEAR PAKOWKI ROAD 0.0022 8 18 1.0 0.040 Spring 9.0 0.85 0.86 3.3 8.0 12.2
05BF016 MARMOT CREEK MAIN STEM NEAR SEEBE 0.1200 3 5 1.0 0.060 0.3 0.08 1.04 5.5 12.2 18.0
05BM007 PARFLESH CREEK NEAR CHANCELLOR 0.0010 6 12 1.5 0.045 Spring 4.9 1.02 0.59 1.4 3.2 4.7
05BM009 TWELVE MILE COULEE SPILLWAY NEAR CARSELAND
0.0026 4 8 1.5 0.045 1.3 0.46 0.60 1.5 3.3 4.9
05GA008 SOUNDING CREEK NEAR OYEN 0.0005 9 18 2.0 0.040 Spring 3.8 0.81 0.43 1.8 4.0 5.9
06AC001 JACKFISH CREEK NEAR LA COREY 0.0007 7 12 1.0 0.040 Spring 4.3 0.89 0.52 1.5 3.5 5.3
07AA007 SUNWAPTA RIVER AT ATHABASCA GLACIER 0.0150 4 8 1.0 0.055 Spring/Fall 0.7 0.22 0.75 3.0 7.0 10.6
07AF003 WAMPUS CREEK NEAR HINTON 0.0150 4 8 1.0 0.055 Fall 0.8 0.23 0.77 3.0 7.0 10.6
07AF010 SUNDANCE CREEK NEAR BICKERDIKE 0.0015 7 12 2.0 0.040 Fall 2.4 0.54 0.59 2.1 4.7 6.9
07AH002 CHRISTMAS CREEK NEAR BLUE RIDGE 0.0050 8 14 1.0 0.045 Spring 11.9 0.90 1.24 4.4 10.2 15.2
07CA003 FLAT CREEK NEAR BOYLE 0.0027 6 12 1.0 0.045 Spring 3.7 0.64 0.73 2.4 5.6 8.4
11AA027 SAGE CREEK NEAR INTERNATIONAL BOUNDARY0.0006 7 14 1.0 0.040 Spring 1.0 0.40 0.30 1.1 2.6 4.0
11AA028 BEAR CREEK NEAR INTERNATIONAL BOUNDARY0.0050 3 6 1.0 0.050 Spring 1.3 0.48 0.74 1.4 3.3 4.9
107
11AA029 MINERS COULEE NEAR INTERNATIONAL BOUNDARY
0.0050 4 8 1.0 0.050 Spring 1.1 0.36 0.64 1.9 4.5 6.7
11AA032 NORTH FORK MILK RIVER ABOVE ST. MARY CANAL
0.0023 6 10 1.0 0.045 3.7 0.69 0.72 2.1 4.8 7.1
Figure 2 – Plot of Y at 3Q10 vs. Channel Slope
1.4
1.2
3Q10 Equivalent Flow Depth (m)
1.0
0.8
0.6
0.4
0.2
0.0
0.000 0.004 0.008 0.012 0.016
Channel Slope (m)
1.4
1.2
3Q10 Equivalent Flow Depth (m)
1.0
0.8
0.6
0.4
0.2
0.0
0 2 4 6 8 10 12 14 16
Mean Channel Width (m)