Bio-Engineered Earth Retaining Structure (BEERS)

A timber sheet pile-vegetation system for stream bank protection

More Info


The Netherlands has an extensive network of rivers and canals systems serving purposes like irrigation, transportation and water removal. The banks along the canals are either protected by earth retaining structures such as sheet pile walls or left unprotected. A bulk of the engineered sheet piles used to protect the canal banks in the Netherlands are made of timber. Tropical hardwoods such as Azobé (Lophira alata) are used to make these timber sheet piles durable, owing to its high biological resistance to decay. Pine from North-west Europe is also used, but need to be treated chemically for sufficient durability. Even though roots of vegetation are known to increase the shear strength of soil, the positive effects of vegetation are not quantified in depth. Vegetation roots in a root-soil composite primarily act in tension when subjected to load, thus acting similar to steel in reinforced concrete. This thesis summarizes the efforts to study a bio-engineered earth retaining structure made of non-durable locally available wood species and vegetation to protect the canal banks as an alternative to the currently used bank protection structures. Such a retaining structure would not only reduce the need for durable hardwoods, but are also more environmentally friendly than the ‘hard’ retaining structures currently in use.
Two vegetation types, Humulus lupulus L. and Salix fragilis L. were chosen for investigation based on their potential to reinforce canal banks, nativity, root characteristics and growth conditions such as presence of high ground water table. An extensive laboratory campaign was planned and conducted to characterize the strength of roots, root-soil composite and to study the behavior of the timber sheet pile-vegetation combination as a system. The experimental results were further extended to develop approaches in the design of timber sheet pile-vegetation system.
To study the root-soil composite behavior in shear, a large scale direct shear apparatus was built. The apparatus was built in the view of conducting tests in dual loading modes, displacement controlled and load controlled shear. In order to simulate the canal bank conditions as closely as possible, the samples were tested in saturated conditions at low confining pressure. Bare soil, rooted samples of Humulus lupulus L. and Salix fragilis L. were tested in both displacement controlled and load controlled conditions. The roots were excavated after the test and analysis of root orientation, diameter and root biomass was conducted. Rooted samples showed a higher friction angle when compared to bare soil. Contractive behavior was shown by rooted samples and the peak stress ratio vs displacement trend of rooted samples were seen to diverge from the trend
for bare soil. Burger model was seen to be able to capture the time dependent behavior under loading mode, when simulated over the experimental results.
A tensile testing program was devised to test roots in tension in a displacement controlled and load controlled tests. Roots of Humulus lupulus L. and Salix fragilis L. were tested in both wet and dry conditions in displacement controlled tests. Load controlled tensile testing was conducted on samples of Humulus lupulus L. of two diameters. Power law was observed to estimate volume-effect and fit all the tensile strength-diameter variations. Comparison of tensile strength in dry and wet conditions revealed that significant difference in tensile strength was observed for Salix fragilis L. while no significant difference in tensile strength was observed for Humulus lupulus L. Further, time to failure of roots were studied using a power law model.
A physical modelling approach was attempted to study the behavior of the timber sheet pile-vegetation system. A root system similar to Humulus lupulus L. was 3D
printed using PLA material to be used in physical model. A comparison of unreinforced bank and bank reinforced with root analogues revealed that the presence of roots increase the volume of soil that needs to be mobilized for failure to occur. It was also observed that when root analogues were placed in the most efficient spatial pattern, among the conducted tests, are able to sustain twice the drawdown pressure. Subsequently, finite element modelling was conducted by including the effect of roots as an increased cohesion parameter. The results from modelling were seen to be able to capture the failure. Parametric analysis revealed that the influence of spatial distribution of the roots on forces acting on the sheet pile is higher, after a threshold value of additional cohesion is reached. That implies any additional cohesion after the threshold value might
not provide any additional benefits to the stability. The results thus indicate that vegetation with more spatially distributed roots will be more suitable to be used in timber sheet pile-vegetation retaining system.
Finally, two different perspectives in design approach of a timber sheet pile-vegetation system are investigated. The system approach is based on the concept that the mechanical reinforcement of the soil with growth of vegetation could result in a reduction of horizontal pressure against the sheet pile, bending moments and shear stresses acting on the sheet pile over time. This results in decreasing the duration of load effect in the timber and counteracting the effects of slow biological degradation of wood in air-water-soil conditions. Timber sheet pile components that are below the water level are less prone to decay. However, those components that are at the air water-soil interface are more susceptible to decay. In the discrete approach the vegetation is perceived as supporting the top parts of the stream bank (< 2meters) after timber decay has occurred. The effect
of vegetation is analyzed as both increase in internal friction angle and increase in cohesion, on stream banks of retaining height of 2m and 3m. An increase in service life of nthe timber sheet pile-vegetation system is achieved in a system approach compared to when only timber sheet pile is present. In the discrete approach, it was observed that modification to the landscape by changing the slope angle of the bank might be necessary when the influence of vegetation is incorporated as an increase in friction angle of soil.
In laboratory scale, as in this study, timber sheet pile-vegetation earth retaining systems show promises to be used for stream bank protection. Future studies need to focus on field scale system level studies and quantification of reinforcement of vegetation in presence of multiple species of vegetation.