In the spring of 2022, large parts of an approximately 80-year-old earth fill dam in Burvattnet Jämtland were demolished. The demolished parts consisted of a homogeneous dust body of glacial moraine that was founded on soil. As part of this thesis, soil has been collected during the demolition of the earth fill dam to map whether internal erosion has occurred along the bottom outlet. Sampling was carried out in sections perpendicular to the longitudinal direction of the outlet to investigate whether differences in the proportion of fine soil and fine sand could be identified with an increased distance from the culvert. Furthermore, six flow-through tests have been carried out in the laboratory at LTU on the moraine that was collected along the bottom outlet. The purpose of this was to investigate whether eroded material along the edge could be distinguished in the experimental setup used. Two factors were varied in the experiments, the surface roughness along the edge of the permeameters and the diameters of the permeameters. The trend in 2 out of 3 examined sections along the bottom outlet in the field suggests that material transport because of internal erosion may have occurred. No clear pattern in the proportion of fine soil could be discerned with increased distance from the culvert with the number of samples available in the study. More samples in more sections had to be analyzed to clarify whether transport of fine material and fine sand has occurred along the culvert and whether the fine soil content varies with increased distance. The experimental set-up used could not demonstrate where in the compacted soil material transport had taken place. To be able to evaluate the influence of the cylinder diameter on the flow-through tests, the material from the moraine stripes must be able to be separated, which was the approach, but did not work, in the test setup used. During the flow-through experiments in the laboratory, no material was obtained in the collection vessel of the experimental set-up. The reasons for this have been derived to a series of parameters that may have influenced that no compacted soil left the fine filter. Further studies with the same experimental set-up are thus required to be able to determine the influence of edge roughness on the erosion tendency of compacted soil.