When obstructions are placed in a river or floodplain, the flow
characteristics of the river are affected. Flow is of particular
concern to bridge designers because it affects scour, which is the
scooping out of soil around bridge foundations. The faster the currents,
the worse scour can be, and, therefore, care must be taken to avoid
creating fast-flow conditions when choosing the span arrangement
of a bridge. Hydraulic analysis determines the flow velocities associated
with various span arrangements in order to determine the optimum
span placement. The flow velocities are then used to calculate the
severity of anticipated scour and the depth needed for sizing the
bridge foundations.
PB was asked to perform a hydraulic analysis and scour evaluation
of the 6.4-kilometer-wide floodplain of the Ohio River in the area
of the Owensboro Bridge. This evaluation included the main river
crossing and the relief structures in the overbank areas on the
Kentucky floodplain side.
Hydraulic Analysis
A state-of-the-art 2-dimensional finite element hydraulic analysis
of the floodplain was prepared. The 3-D flood analysis was not considered
because of the large ratio of width-to-depth, and because the 3-D
computer program was still under development. We used Finite Element
Surface Water Modelling System (FESWMS) software to calculate flow
veloticites and discharge at every nodal point of the finite element
model.
After careful evaluation of the site and its surface features, we
did a topographic mapping of the study area using stereophotogrammetric
plotting instruments. Our mapping had the following characteristics:
- Contour intervals of 1.5 meters with supplementary contours
at .75-meter intervals were included in the plot.
- The 122-meter level was the highest contour since it was close
to the mean level of a 500-year flood.
- The flood-stage width of the river at the project site was
approximately 6.4 kilometers.
- The length of the study area extended from river mile 750 downstream
to river mile 741 upstream, a distance of 14.5 kilometers, giving
an area of about 89.6 square kilometers.
We prepared a finite element mesh that consisted of trapezoidal
and triangular elements. The model was calibrated using high-water
profiles of the Ohio River prepared by the U.S. Army Corps of Engineers,
Louisville District. The maximum discharge associated with each
of the high-water profiles was extracted from values reported at
the Owensboro gauge located at river mile 756.5, maintained by the
U.S. Geological Survey. The model was then validated using other
flood data for both a 100-year and a 500-year occurrence.
Several Alternatives Were Studied
After the FESWMS model was calibrated and validated, 14 alternatives
for the Kentucky-approach relief structures were run. These alternatives
ranged from carrying the approach roadway on a single, long viaduct
structure across the full floodplain, to carrying it on a series
of short structures separated by embankments. Each alternative had
at least ten iterations or permutations and combinations of structure
lengths.
Two of the 14 alternatives studied best met the following criteria:
- Flow rate not to exceed 1.2 meters/second immediately downstream
of the crossing
- Backwater elevation not to exceed 7.6 centimeters.
One alternative consisted of five floodplain structures with a
total length of 1,210 meters, separated from each other by embankment
sections. The other alternative consisted of a single viaduct structure
for the full width of the floodplain.
To further compare the two alternatives, we looked at cost, hydraulics,
operations and maintenance, agricultural impact, etc. Both alternatives
had their advantages and disadvantages. The final determination
was based on cost considerations and the alternative consisting
of five structures was selected by the Kentucky Transportation Cabinet.
State-Of-The-Art Scour Analysis
The FESWMS hydraulic analyses provided us flow velocity and volume
of water passing each pier location as vectors. This information,
along with Federal Highway Manual information on scour calculation
(HEC18), was used to determine the scour at each pier. This information
was critical to our design of the foundations. Additionally, the
velocity vector information was extremely useful for determining
the ship impact forces on the piers.
PB was the first company to apply state-of-the-art analyses tools
to such a large area of river body. Doing so increased our technical
strength in the areas of hydraulics and scour prediction. It also
gave us an important tool to better evaluate ship impact forces
on piers in navigable waterways. (For another article on scour,
see PB Network Spring ‘94, page 30.) |
