When water travels along a channel is the bottom which the water is in direct contact with, defined by a coefficient named Manning’s coefficient of friction. This coefficient explains the roughness of the channel bed. A rough surface equals a low Manning’s coefficient. The purpose of this thesis is to study how Manning’s coefficient differs when different waterflows and depths is obtained. The Manning equation is depend-ent on using a constant Manning’s coefficient based on experience and tables. The experience in the line of business knows that the coefficient holds a different value in channels with similar bathymetry and different flow conditions, and the thesis will cover these questions. At Vattenfall Research & Development in Älvkarleby, a flume made for fluid me-chanic experiments has been built, which is perfectly suitable for the experiments for this thesis. The flume has a roughness that has been programmed mathematically with a purpose of looking like a real channel bed or a blasted rock-tunnel. This is where all the measurements were made. The measurements were made in totally 14 cross-sections with a few different flows and depths. The depth was controlled by a hatchway downstream the flume and the flow were controlled by the pumps. Depending on which setting the hatchway ob-tained measurements between 1.6 l/s to 280 l/s were made. The different settings that were used for the hatch was 100 %, 85 %, 75 %, 60 % and 30 % open hatch. The most obvious results appeared when the hatch was 100 % open, since the differ-ence in depth between each cross-section was large and the faults in the equipment weren’t that important. 30 % open hatch generated such calm flow and a deep depth, so the errors were too big to include in the report. These different flow conditions were used to gain a big variation and accuracy, with the purpose of getting as good result as possible. The measurements show that for every hatch-setting, the Manning coefficient gets lower when a lower depth and flow are held. If the results are plotted from all the hatch-settings, especially 100 % open in a graph with the Manning coefficient as a function of the flow, the results follow an obvious shape with a trend where the Man-ning coefficient reduces with lower flows. The same thing with the depth. Shallower depth equals lower Manning coefficient. This is because of the roughness elements that has a bigger part of the entire cross-sectional area, so that a bigger part of the cross-sectional area is affected by phenomenon caused by the roughness elements. Key words: Manning’s coefficient, streaming losses, energy losses, Vattenfall