Author(s): Kevin Reiterer; Thomas Gold; Christoph Hauer; Helmut Habersack; Christine Sindelar
Keywords: Gravel Replenishment; Sediment Management; Dispersion and Advection; Physical Modelling
Abstract: Gravel replenishment is a common approach to counteract problems caused by a disturbed sediment continuity in many large rivers in Europe. At the Rhine River as well as at the Danube hundred thousand cubic meters of gravel are added each year to ensure river-bed stabilization, secure water-fairways and protect infrastructure downstream of run-of-river hydropower plants (RoR-HPP). Beside technical aspects, gravel augmentation is used as a common approach for ecological habitat improvement. Aim of this work is the introduction of a modelling approach to quantify possible effects and the effectiveness of such measures. Therefore, the 1D-advection-dispersion model (1D-ADM) was tested for its suitability to describe the longitudinal propagation of local-short-term (LST) gravel inputs based on laboratory experiments and field data. The 1D-ADM is commonly used when it comes to suspended and solute material transport like containments and tracers. The model-function has the form of a downstream propagating Gaussian distribution with an increasing measure of dispersion over time. The advection term is described by the mean value (µ) of the longitudinal propagation, while the dispersion parameter is related to the corresponding standard deviation (σ). Laboratory test runs were performed to derive the advection and dispersion parameters which are further used as an input for the 1D-ADM. The experiments were conducted in a 0.3 m wide and 10 m long laboratory flume with mobile bed conditions. A frequency-controlled conveyor-belt sediment feeder located at the upstream end of the flume was used to supply a heterogenous sediment mixture (0.7-6 mm). The constant-head water supply was provided by an internal water recirculation system including a frequency-controlled pump. The experiments started with a short period of high sediment input (SIH), representing the LST gravel augmentation, followed by a constant rate of fed sediment (SIC). The unit-width discharge (q) was constant for both periods (SIH and SIC), with SIC corresponding to the preliminary measured equilibrium bed-load transport rate for q. By analysing the volumetric changes of the experimental bed-surface the parameters for the 1D-ADM were determined. The 1D-ADM showed its capability for estimating the longitudinal propagation of the LST gravel input. The model was further applied to field data collected downstream of the Iffezheim Rhine RoR-HPP (Goelz et al., 2002). Based on the longitudinal mass-distribution of petrographic tracers added in October 1996 and subsequently tracked for three years, the 1D-ADM parameters were determined. Further, the mass distribution predicted by the model was compared with the field measurements at different cross sections. The 1D-ADM is capable to describe the longitudinal LST gravel propagation for both field and laboratory data with good to very good accuracy. In conclusion, the 1D-ADM is a useful tool for the planning, evaluation, and optimization of gravel replenishment.