Krueng Baro Irrigation is focused on increasing the productivity of food crops in Pidie District, Aceh Province, Indo-nesia. However, due to the age of the irrigation infrastructure (more than 30 years) and its large networks, it is necessary to investigate the actual water conveyance efficiency. This study aimed to evaluate the conveyance efficiency of primary and secondary channels of the irrigation system, as well as to create a water balance model based on the actual water convey-ance efficiency. The model by using Excel Solver with its objective function is to maximize the area of the irrigated land. Based on the optimization model of the water balance, the design condition can irrigate an area of 9,496 ha (paddy-I), 4,818 ha (paddy-II), and 11,950 ha (onion). The measurement results reported that the actual efficiency of Baro Kanan and Baro Kiri was 56% and 48% smaller compared to the efficiency of the designs (65%). The water loss was due to the damage to the channel lining and channel erosion resulting in the high sedimentation, leakage, and illegal water tapping. These lead to a decrease in the area of the irrigated land. Based on the optimization model of the actual water balance, the irrigated land was reduced to 7,876 ha (paddy I) and 3,997 ha (paddy-II) while it remained the same for onion. Therefore, to increase the efficiency, the regular maintenance and operations are required by fixing the damaged irrigation structure and channels, the maintenance of sedimentation, and the strict regulation of illegal water tapping.

Potentially hazardous side-channels of complex geometry need to be investigated using detailed hydraulic physical models. This study aims to analyse the cross-waves pattern and pulsating flow using a side-channel spillway physical model. This study compares the cross-waves pattern were measured using an experimental installation set to generate cross-waves on the surface (original series) with another structure that did not produce cross-waves (modified series). The results showed that the geometry of the left wall caused instability in flow patterns and secondary flows. The starting point of Q ₂ discharge was detected by minor turbulence on the water surface near the left wall at a water depth of 3.3 m at the starting point of the wall, but with no overtopping. Cross-waves formed downstream at the right wall crosswise, lower than at the left wall. The height of the cross-wave increased substantially from Q ₁₀₀ to Q ₁₀₀₀ discharges leading to overtoppings near the left wall at a water depths of 4.2 and 5.0 m at the starting point of the wall, and near the right wall at a water depths of 3.8 and 4.0 m at the upstream point of the wall. The modifications provided optimal hydraulic conditions, i.e. elimination of cross-waves and non-uniform flows. The Vedernikov and Montouri numbers showed that both original and modified series did not enter the area where the pulsating flow occurred. This indicated that both series were free from the pulsating flow.