In the following you can see a short summary of the MSc studies I have been supervising. Please contact me in case you are interested in a copy of one or more studies.

Sonja Huber, Armor layer in gravel bed rivers, 2013

In progress….
Sonja is conducting experiments in our new hydraulic flume. She develops an armor layer from an initially well mixed sediment mixture. In addition she is investigating the possibility to use the photogrammetric method “Basegrain“, developed by Detert&Weitbrecht at the ETH in Zurich, for laboratory purposes.

Armor layer development in progress


Laura Linzano, 2D modeling of sediment transport in river bends, 2013

In progress…
Laura is from Costa Rica and she is currently taking her Masters in Hydro power development at our department.
Laura is currently using CCHE2D to run the Yen case (Yen, 1995) and the famous physical model study of the BAWs river Oder model


Felix Hahn, TERRESTRIAL LASER SCANNING in short-term surveying of steep riverbanks, 2013


The path of a river is not stable over time, except if it is being regulated by human constructional measures. There are many determining factors, such as local or regional climate and the amount of available drainage water, an underground consisting of rock or loose sediments and the slope of the surface, that influence the path of a river. These factors all have a variable strength of influence on erosion and their influence might not occur at the same time. This circumstance is referred to as a complex response. The complexity can be caused by a time delay of reaction by the different processes and system components. Also the initial response to change may be different from the final adjustment that leads to a new equilibrium condition. It is difficult to document such a complex response in large natural settings because long periods of time might be needed until a final adjustment is accomplished by a sequence of events (RITTER, KOCHEL AND MILLER, 2011: 14). In some cases though it might be necessary to find out in which way the path of a river is responding and changing and how fast it is doing so. The change in land use, for example the replacement of forest cover by intensely cultivated land, would be an example of human influence on the system components (RITTER, KOCHEL AND MILLER, 2011: 196ff). Also very often rivers today are physically manipulated in order to improve aquatic habitat and ecosystem health. In addition, river sediments often function as a transportation medium for toxic substances since some of them represent the largest contaminated sites in the world (RITTER, KOCHEL AND MILLER, 2011: 212). The suspended sediment load that was removed from fertile agricultural lands is also often contaminated with agricultural nutrients and pesticides (NASERMOADDELI AND PASCHE, 2008). These examples show why there is a need to understand the dynamics of fluvial processes. Riverbank erosion is an important factor within the dynamics of fluvial systems and the lateral migration of a river channel. Riverbank erosion also contributes to the suspended sediment load in the rivers (NASERMOADDELI AND PASCHE, 2008). Since it contributes to the sedimentary load and through its control on channel width it also has an influence on other channel processes. Bank erosion implies three major types of processes that are fluvial entrainment, mass wasting, and the weakening and weathering of bank materials.
The laterally working bank erosion involves a unique combination of such processes depending on the individual settings in the fluvial system. Modern techniques offer a way to determine the continuous changes of an individual river. Aerial LiDAR (Light Detection And Ranging) for example made it possible to survey areas the size of entire river reaches. Still in some cases it might be necessary or useful to get a closer, more detailed look. In such a case it could be an achievement to use a Terrestrial Laser Scanner (TLS). MILAN ET AL. (2010) used a TLS for mapping hydraulic biotopes and O’NEAL AND PIZZUTO (2011) investigated the amount of sediment contribution of mercury-contaminated riverbanks using a TLS. O’NEAL AND PIZZUTO (2011) mention that other methods, like erosion pins, for measuring erosion of riverbanks do exist, but despite offering high accuracy are often limited in spatial coverage and data density. They ascribe a poor understanding of bank erosion along an upstream-downstream extent to the limitations of traditional field methods. In addition BRIDGE (2003: 400f) demurs that erosion pins disturb the bank material and interfere with the water flow past them. Photogrammetric techniques on the other hand allow a rapid production of topographic models from large volumes of data but often result in less precision and accuracy than traditional survey methods (O’NEAL AND PIZZUTO, 2011). A TLS could therefore be an advantage whenever the sediment budget of a river needs to be acquired. With the laser scanner it should be possible to identify and quantify sediment mobilization and sediment production, meaning the amount of sediment given access to the channel. The efficiency and cost effectiveness of this technology have made it useful for applications in many fields where surface investigations are needed.
The intention of this study was to empirically validate processes of riverbank erosion and to find out whether a TLS could be used to determine processes like fluvial entrainment and mass wasting. In addition the volume of material that was removed from the riverbank during the study period should be validated. In this context causes for the observed riverbank erosion were to be determined. The vegetation should be removed from the collected data in order to obtain the bare earth surface of the scanned riverbanks.



This study is about 3D Numerical Investigation of Settling basin layout by using numerical modeling program SSIIM. This study is carried out by using SSIIM windows version 1 (SSIIM 1.0). SSIIM is numerical modeling software, developed at NTNU by Professor Nils Reidar B. Olsen. This program has been used for investigation numerical modeling of hydraulic and sediment transport for different layouts geometry of settling basin.
In this study a case study has carried out on settling basin layout of Mai Khola Hydropower Project, Nepal. Hydraulics performance of proposed layout and one alternative layout with shorter approach is numerically investigated by water and sediment flow computation. For the Numerical investigation structured grid for
settling basin layout has developed with the help of drawing provided and excel spread sheet program. Hydraulics performance is investigated for design discharge with constant flow. The hydraulic performance of closing of one chamber and operation of remaining chamber with design discharge of power plant is also investigated. Based on water flow computation result, sediment computation was carried out for one settling chamber, proposed, alternative and modifications of proposed layouts. Effect of approach geometry on distribution of sediment on four chambers of settling basin and sediment trap performance were studied by sediment flow simulation. Effects of closing of chamber on distribution of sediment concentration were also investigated with the help of sediment simulation. Trapping efficiency is evaluated for one settling chamber, proposed alternative and modification layouts and closing mode models. Trap efficiency of one settling chamber model is compared with trap efficiency by analytical method.
Based on hydraulic performance, sediment distribution performance and trap efficiency performance; recommendation of modification on approach geometry has made. Also, studied result shows that SSIIM 1.0 version can be used for investigating performance of hydraulic structures and settling basin.




There are many creeki ntakes in Norway and tyrolian intakes is usually a good choise as intake construction. In creeks with large amounts of bedload transportation these intakes will fill up until the edge and then start to flow over the intake. To empty these intakes for sediments it is usual to go in manually with an excavator and dig it out, a time-consuming process and these intakes is not always easily reachable. In this report the goal has been to find a solution to clean out the intake basin using flushing. The report focuses on a tyrolian intake belonging to Statkraft, located close to the city of Narvik.
To examine different alternatives, a model has been built-in the hydrolab of NTNU. Tests were done with a straight channel going through the intake construction and also on a solution with the flushing channel passing on the side of the intake.
Results from the experiments show that it is entirely possible to construct a working flushing construction even for coarse sediments, in this case the biggest being up to 1.5 m diameter on the longest axis. Best solution for this project turned out to be a 3 meter wide channel led straight under the intake. A side channel were not as effective but could be improved by constructing leading walls inside the basin to guide the flow, it was then performing satisfactory.

Shreeja Shrestha, 3D Numerical Investigations of Mai Khola Headworks, 2012


Hydropower development is the key for the success of economy of the country like Nepal. Improper design and planning of hydropower may lead to loss in energy production and thus loss in economy. Headwork is an important part of the hydropower and it should be strong and safe enough to handle any sort of problem in future. Thus the proper design should be undertaken prior to construction phase to construct the optimal design as well as to reduce failure. With the advancement of computer technology and development in CFD modeling, various softwares are now available to study the effect and the performance of hydraulic structures.
CFD modeling is now advancing as the best tools for the modeling of complex flow pattern in natural rivers and streams. The numerical modeling is applicable in the calculation of water flow and water quality. Numerical modeling can also be applied for the computation of the sediment deposition pattern and amount of the sediment deposited in the reservoirs and rivers. Thus this quality of the CFD modeling has makes the significant use of numerical modeling in the hydropower sector. SSIIM model is a type of CFD model developed at NTNU by Professor Nils Reidar B. Olsen which is use to simulate the sediment movement in the river/channel geometries. This thesis is focused on the application of SSIIM model in Himalayan River with complex structures and high sediment load.
The report is based on study carried out on the Mai hydropower project (22MW), Nepal. The Mai hydroelectric power project is now under construction and the physical model study of the project was carried out by Hydro lab, Nepal in 2011. The study is more focused on computation of the water flow and sediment deposition pattern at upstream of the headworks by the numerical modeling. The results were reviewed and the comparison is made with that from the physical model.


Øyvind Pedersen, 3D Numerical Modelling of Hydropeaking Scenarios in Norwegian Regulated Rivers, 2012


The objective of this master’s thesis has been to investigate the capabilities of the CFD packages Star CCM+ and SSIIM to model hydro-peaking scenarios in regulated natural rivers in 3D using a RANS method. Two Norwegian regulated rivers are modeled, Surna and Lundesokna. In Surna, flow fields and bed shear-forces are
compared for the Star CCM+ and SSIIM models. In Lundesokna flow fields and water surface elevations from simulations in Star CCM+ are compared to field data at steady flow conditions. For unsteady flow a Star CCM+ simulation are compared to video footage of a hydro-peaking event.
The Surna Star CCM+ numerical model predicts a comparable flow field to SSIIM for both steady and unsteady flow. The magnitudes of velocities and shear-forces, however, deviate. Unphysical velocities and shear forces were found in parts of the cells in the Star CCM+ model. A review of the model shows that the unphysical
velocity errors are likely caused by a too coarse grid and problems with the VOF method. Efforts to run simulations on a finer grid were discontinued because of a lack of available computational resources.
Both the Surna numerical models achieved convergent solutions for unsteady flow. However, due to the quasi-unsteady treatment of the flow in SSIIM the SSIIM model is not able to capture wave effects. In consequence the SSIIM model predicts shearforce peaks about 600 seconds earlier than the Star CCM+ model at the outlet.
When accounting for this effect the models show similar flow fields but deviating velocity and shear-force magnitudes as for the steady flow.
Comparison to field data show that the Lundesokna Star CCM+ model is able to predict flow fields and water surface elevation with high accuracy for steady flow between 10 m3/s and 20 m3/s discharge. The unsteady flow simulation shows visual resemblance with the video footage, however, field data measurements are
required to quantify the accuracy of the numerical model for transient conditions.


Caroline Speth, Physical Model Studies for optimization and planning of the Dabbsjö spillway & velocity measurements in the Sarpsfossen Model, 2011

The spillway of Dabbsjö flood control system was investigated. Two modifications were done in order to optimize the discharge over the dam and through the channel. The capacity remained the same for all modifications, because the structure of the dam and the reservoir did not change. The capacity factor was determined, so it was possible to extrapolate the water level over the dam for extreme floods. Furthermore, water depths and flow velocities were measured to get a general overview of the flow situation especially in the channel. It was ob-served, that there was no significant change in all modifications. In addition, some flow velocities were measured upstream the hydro power plants in Sarpsfossen. At first, the measuring methods, ADV and hydrometric propeller, were compared; thereafter the change in the flow velocity for two different water levels was determined. The two measuring methods gave accurate results, if the model was adjusted correctly. This was also valid for the two different water levels at the ratio of the calculated and the measured change. Also the measured change in the flow velocity at different water levels complied with the calculated values.


Diwash Lal Maskey, CFD modeling of hydraulics and sediment at Nyadi Hydropower project, Nepal, 2011


Recent developments in computer speed, capacity and cost effectiveness have added advantages for numerical hydraulic modeling over the traditional scaled physical modeling to study the performance of different hydraulic structures. Different types of numerical model have been developed at different institutes for the same purpose.
Similarly, SSIIM is one of such programs developed at NTNU by Professor Nils Reidar B. Olsen which is capable of performing mathematical computations of hydraulics and sediment transport for different conditions.
In this thesis work, SSIIM 2 is used to model a physical model of diversion structures and intake of Nyadi Hydropower Project in Nepal. This version has been chosen for its advantage of simulating wetting and drying conditions in complex geometries. Hydraulic and sediment simulations (only bed load) were performed for different discharge conditions and different river cases. The simulations were performed for existing river situation and then with two types of diversion structures. The results obtained from the simulation were compared with the results from the physical model. For the hydraulic simulations, detail comparisons of hydraulic parameters could be made as they were available, but the results of the bed load simulations were done with the descriptions in the physical model study report. The simulated results on hydraulic parameters showed good agreement with the measured parameters. Although, the performance for low flow values was not good, accurate and efficient results were obtained for the high flows. Some challenges were faced during the sediment (bed load) simulation in completer convergence of the program. The results were improved for higher flows with changes in the bed cell parameter, accuracy is still lagging. Even though, the model resulted in fairly similar results with the physical model.
Based on the studies, it can be concluded that SSIIM 2 can be used in the study of steep rivers with fair accuracy. The efficiency can be improved with more similar studies.

Martin Honsberg, Advanced study to link grain size distribution parameters and shelter frequency for juvenile Atlantic salmon, 2011

Freeze core

The further development of hydropower in Norway goes on, due to national programs and external impulses initiated by changes in the European energy markets. In this context, the effects of hydroelectric production on the (eco)systems of rivers are deeply investigated to ensure a more sustainable planning, building and running of the facilities. The environmental impacts of frequent alterations of flow induced by hydropower operation (hydropeaking) are investigated within the program EnviPEAK. The wish of flexible operation patterns will consequently lead to strong variations, which will change the living conditions for fish, benthic animals and plants living in and near rivers, which are regulated for hydropower. Since the sediment household is assumed to be affected by hydropeaking, the modeling of the changes of the sediment composition in the river bed due to the flow variations is one task within EnviPEAK. The possible decrease of fine sediment would strongly affect the habitat of one of the most famous fish in the Norwegian water bodies, the Atlantic salmon. To estimate this impact, a correlation between the composition of a river bed (i.e. grain size distribution parameters) and the shelter availability for the Atlantic salmon during its juvenile live stage (most critical age) is desired to be found. The quantification of these shelters was made possible by Finstad (2007), who presented a simple method to classify shelter categories and total numbers. To gain data for this study, three river sites in Norway with different impact situations regarding hydropower were focused. Since sediment sampling in fish habitats requires techniques for submerged conditions, a freeze core method was chosen. However, it turned out that a common freeze core sampler was not suitable for the given task, so several methods and modifications were studied, developed and tested. The aim was to find a freeze core based method, which allows the relatively undisturbed sampling of all particle fractions of the upper river bed layer. With a final modified method, several sediment samples were taken and analyzed in the laboratory. Besides the grain size distribution parameters frequently used in scientific publications, an own parameter is determined, including the sediment properties which are believed to play a major role for the abundance of shelters. The total number of shelters and this parameter reach a Pearson coefficient of 0.88 and a R² of 0.78, which is promising. Good correlation values are also obtained for the geometric mean and the kurtosis (after nth square approach).

Stefan Jocham, An approach to link shelter abundance and grain size distribution for the assessment of sediment quality for juvenile Atlantic salmon, 2010


In the backdrop of resource shortage and climate change the interest of strengthening renewable energies rises. New technologies are under development and proved ones are expanding. Impacts of future developments have to be investigated in advance regarding their societal and ecological challenges.
The understanding and modeling of ecological impacts induced by hydro power plants are an important aspect in finding future operation patterns which consider both economical and ecological interests. It is assumed that frequent alterations of flow induced by hydro power operation (denoted as hydropeaking) cause elevated degrees of fine sediments in river beds which affect negatively habitat quality for juvenile Atlantic salmon.
A concept called embeddedness was introduced by Klamt (1976) to quantify the amount of fine sediments relative to the large particles in river beds. Later developments led to different approaches quantifying embeddedness but although it was initially chosen as a parameter to measure habitat space for juvenile states of stream water fish, the results of the methods are physical descriptions of river bed and do not give direct output to the requirements of fish. In this context Finstad et al. (2007) developed a method to measure interstitial space (shelter abundance for fish) in running waters. Shelter availability is a candidate mechanism for survival of juvenile Atlantic salmon due to reduction of predation risk and influenced by the degree of embeddedness. Basic parameters considered in the assessment of habitat quality for fish are the physical descriptions of running waters flow velocity, water depth and substrate composition. It is hypothesized that within the last-named parameter, embeddedness and interstitial space can be regarded as a spatial distribution of the underlying grain size distribution. The goal of this study was to find a relationship between shelter abundance (measured with the method by Finstad et al., 2007) and grain size distribution which could be implemented in predictive models for the assessment of the ecological status in running waters.
The developed approach to reach this aim was conducted by gathering data in representative places both for shelter abundances and grain size distributions. In squares of 0.5 m x 0.5 m shelter abundance was measured with the help of a rubber tube of 13 mm outside diameter. After that the areas were excavated to the depth of the largest particle visible from top and mechanically sieved and analyzed.
The results of the study indicate that particle distribution parameters generated in grain size analysis are correlated to shelter abundance to a certain degree. It was shown that percentiles describing the fine tail of the distribution are highly correlated to shelter abundance whereas percentiles describing the middle part and the coarse tail of the distribution are weakly correlated to shelter abundance. Distribution parameters show relatively high correlation quality, however under certain restrictions. The findings are used to recommend an improved field working procedure and might be useful for later developments of assessment and management tools for hydropeaked as well as regulated rivers.
The study contributes to the general understanding of sedimentation processes in regulated running waters and their potential ecological impacts.

Joern Wildhagen, Applied Computational Fluid Dynamics with Sediment Transport in a Sharply Curved Meandering Channel, 2005


Jagadishwar Man Singh, 2D modeling of sediment transport in river bends, 2004


Jerome Molliex, 2D modeling of sediment transport in river bends, 2004


Nasib Pradhan Man, 2D modeling of sediment transport in river bends, 2004

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s