Erosion Control in a River Meander Using Engineered Log Jams: An Experimental Study

During the last decades bank degradation has become more evident in the watercourses of the Amazon lowlands, particularly near populated areas or where major infrastructure has been built. Traditional materials used in erosion control measures such as rock or cobbles are not available in these areas. Therefore, alternative solutions for controlling riverbank erosion must be found. One alternative is the use of Engineering Log Jams, given the abundance of woody vegetation near the riverbanks which may also be found as floating debris. ELJs have been successfully used in small watercourses in North America and other places. However, this type of solution is yet to be used in large rivers.

Lateral displacement accelerated from 2 m/year to 6 m/year in the last years in the Madre de Dios River near Puerto Maldonado in SE Peru. The right bank was also the outer bank of the reach and it was approaching the Interoceanic Highway. Several engineering solutions were tested in the laboratory and finally one was chosen. However, in a second phase of the study, ELJ solution was tested as part of a research program, taking advantage of the already built hydraulic model. Four kilometers of the Madre de Dios River were represented in a 1/60 undistorted, mobile bed, scale model. The physical model was 50 m long, 6.9 m wide and the radius of curvature was 13.8 m. Therefore, the radius of curvature to width ratio, R/B, was approximately 2. Maximum flow depth was approximately 0.25 m and flow remained subcritical throughout the experimental phase.

A total of five experiments were conducted. The first two runs were conducted without including ELJ groynes. Three different ELJ groyne configurations were tested. In these runs, ELJ groyne permeability was approximately 35 %. Water levels and depths and 3D velocity fields were recorded using an ADV Profiler and a laptop computer. Bed topography was gathered using a 3D laser scanner.

Experiments showed that when groyne length was in the order of 15 % of the watercourse width, spacing varied between two and three times the length of adjacent groynes, and these structures were approximately perpendicular to the axis of the outer river banks and are unsubmerged during the tests, results are optimal in terms of protection of the riverbank, scour reduction at the groynes´ toe and sediment deposition between groynes.