Morphodynamic Modeling of Gravel Bar Formation At A Bridge Replacement

State Highway 14 spans the Cache la Poudre River on the east side of Fort Collins, CO. The Poudre River at this location underwent a substantial realignment in 1949 to accommodate the SH 14 bridge structure, shifting the channel via an engineered meandering section. Design of the replacement bridge began in 2013 and culminated in construction of the replacement bridge in 2015-16, along with a substantial change to the vertical alignment of SH 14 and associated river grading. A tall, largely stable vegetated bank was present on the right overbank area prior to the recent bridge replacement. Only one year after construction, a significant gravel bar with cobble top layer began rapidly developing upstream of, and under, the new bridge footprint. This bar continues to expand into the active channel and bridge opening.

A morphodynamic model of the fluvial system upstream and downstream of the bridge has been constructed using SRH-2D. Taking advantage of topographic data and employing SRH-2D's sediment transport capabilities, we have experimented with the predictability of the observed gravel bar formation using this hydraulic model. Towards that end, we have utilized a nearby USGS gaging station to run actual flow data through the simulated fluvial system.

Our preliminary findings suggest that the initial stages of the gravel bar formation may be expected within a single year of post-construction runoff, which matches observation of the site. Further, we experiment with multiple sediment transport equations in attempts to predict the lateral and vertical location of the bar's formation. In addition to these findings, we propose to also address the difficulties inherent in such a model during the pre-construction design phase for a DOT engineer: calibration of the model and sufficient understanding of the geomorphic response of the larger system to the disruption caused by the new bridge and associated grading.

An important take-away for hydraulic engineering practitioners, as demonstrated by this study, is that design using only large magnitude (small recurrence) peak flows, and assuming a static non-moveable bed, may not provide adequate understanding of the fluvial system when arriving at a proposed condition grading plan for infrastructure improvements. This could lead to diminished hydraulic capacity and performance at a bridge crossing, as well as larger floodplain encroachments than initially expected.