Dam break simulation using the landslide module in MIKE21 HD.
The interesting thing in this study is the approach chosen to model the dam break. Instead of the normal MIKE 11 Dam Break method, the MIKE21 HD landslide module has been used describing the dam breach via the bathymetry. Hence a full 2D model.
The study was made in corporation with the University of Alicante, and the objective was to determine the flood risk for a particular plot of land in the event of a breach of the Crevillente dam. Additionally to determine, if the area was in a flood risk zone, what types of preventives actions should be put in place in order to avoid loss of life and property.
The client is a real estate developer, who in order to get the necessary building permission was required to prepare a flood study of the area. The area in question is shown on figure 1.
Figure 1: The study area. The figure to the left shows the area without the planned development, and the figure to the right the area with the development (Black contours).
Modelling Approach:
In the present study the propagation and transformation of the generated waves have been modelled using DHI´s MIKE 21 hydrodynamic module, which is based on the non linear shallow water equations. The effect of the Dam Break is modelled by forcing terms representing the dynamic vertical deformation of the bathymetry plus additional terms to represent the effect of the dam break due to viscous and inertia forces
Regarding the high velocities that can occur during a Dam break, the MIKE 21 model engine has incorporated a calculation scheme which allows it to model both sub - and super-critical flows.
The present approach has been validated against results obtained from a classic MIKE 11 model using both the Energy and the Dam Break equation to model the outflow hydrogram.
Definition of the breach and data received :
The Dam Break was defined based on the general guidelines set out in the “Guia tecnica para la elaboracion de los planes de emergencia de presas”, (Ministerio de medio Ambiente). The following parameters were used:
- Time of Breach=19.8 min
- Shape of breach = Trapeze
- Depth of breach = 93m
- Longitude: 101.5 m
- Pendiente: 1:1
- Crest level =146 m
- Initial water level in reservoir =145.5 m
- Volume of reservoir =12.7 Hm3
The client provided digitised bathymetric information for the area down stream of the dam. A horizontal resolution of 5 m x 5 m was used for the calculation grid. The geometric description of the Dam Breach was incorporated in a time varying bathymetry as illustrated in the figure below.
Figure 2: The dam before (t=0 min) and after the breach (t=19.8 min)
As no topographic data was available the reservoir was incorporated directly into the model bathymetry using an approximated shape, based on the given total volume. In cases where topographic data are available it will naturally be a great advantage.
Results:
As mentioned the objective of the study was to investigate the consequences of a dam break on particular plot of land. As it can be seen the Breach will cause flooding of the area, and could probably have serious consequences for the inhabitants.
Figure 3: The maximum water levels for the present situation and with the Floodwall..
The solution proposed be the client, was to construct a flood protection wall as show in the figure 4 (right). The proposed wall was incorporated into the bathymetry and a new simulation was executed. The results shows it will have the desired effect on the area, and do only have minor effect on the opposite side of the “river”.
Verification:
In order to verify the results from the MIKE21 HD model, a simple MIKE 11 model was set up and 2 simulation was carried out, one using the Energy Equation method and one the Dam break equation method.
Figure 3 shows a comparison of the three outflow hydrograms from, MIKE21 HD, MIKE11-Energy Eq., and MIKE11 DB Eq.. As it can be observed from the figure, (although they do not display the exact same form) there is excellent correspondence between the 3 different methodologies.
Figure 4: Comparison of three different modelling approaches – (BLACK) MIKE 21 HD, (BLUE) MIKE 11 Energy Eq., and (GREEN) MIKE11 Dam break Eq.
Conclusion, advantages and disadvantages .
For this study in particular the method has been very useful, given that the geometry of the dam itself occupies an important part of the area. For studies of a larger scale the approach might not be appropriate.
The main advantages of the method;
- the entire study is made in 2D, hence spatial results of water levels and velocities (Xv, Yv) will be available.
- all data handling are made through the MIKE 21 interface, which means that the problems that might occur when combining 1D with 2D is avoided.
- More detailed and “correct” results, compared to what´s possible to get with a 1D model.
- The outflow from the reservoir can be described with a higher degree of detail, thus creating a more realistic picture of the floodwave.
The main disadvantages are,
- the 2-D approach will normally require longer calculation times, although some of this time can be saved on the model configuration process.
- the definition of the reservoir can be tricky, due to the fact that normally (at least here in Spain) the reservoir definition is given as a Height/Area Curve, which is difficult to translate to the bathymetry.
- Its not possible to included structures such as culverts, gates etc directly in MIKE21, to do this MIKEFLOOD should be used.