On the 29th of December 2016 a large Rhine ship loaded with 2000 ton benzene sailed through the weir at Grave, located in the river Meuse in the Netherlands. The accident created a large gap in the weir and within 12 hours the upstream water level decreased by a maximum of 3 meters. Due to this decreased water level shipping was limited, houseboats were skewed and there was a risk of instability of the river banks and upstream weirs.

Several alternative measures were considered to close the damaged part of the weir in order to reinstate the upstream water level, and make it possible to repair the weir. Finally it was decided to build a rock-fill dam behind the damaged part of the weir which closed nearly 55% of the weir. Due to this closure, all the discharge had to pass the weir through the remaining 50 meters-wide southern opening. As a result of this upstream water levels were set-up faster than during normal operations. High upstream water levels, relatively high discharges and a low downstream water level caused extreme flow situations with even supercritical flow (hydraulic jump) over the bed protection. At weir Grave the existing bed protection was injected with colloidal concrete and an additional ballasting sill, made of 3-6t carefully placed stones and 4t rock nets was constructed, to prevent additional damage to the bed protection.

In this study the stability of the measures which were used to prevent additional damage to the bed protection are analyzed. This study focusses on the 3-6t carefully placed stones, rock net sill and the bed protection behind it. The 4t rock nets were not able to withstand discharges up to 450m3/s and flow velocities up to 7.0m/s due to the internal instability of the small rock material inside the nets. The 3-6t stones were able to withstand discharges up to 850m3/s and flow velocities up to 6.7m/s. Due to the fact that the stones were carefully placed, the stability increased. A stability correction factor (Φsc) up to 0.73 can be used instead of 1.5 as prescribed by Pilarczyk for stones on the edge of a bed protection. The 40-200kg bed protection with colloidal concrete has withstood extreme flow velocities up to 7.2m/s whereby only small damages occurred. The 40-200kg quarry stones without colloidal concrete also showed limited damages after the 850m3/s flood wave and depth averaged flow velocities up to 5.2m/s. The conclusions of the analysis are projected to the other six weirs in the Meuse and to the possible calamities which could occur here.

In order to perform the stability analysis of the bed protection measures at weir Grave, the Meuse and the seven weirs in the Meuse were analyzed. The Meuse is a controlled river in the Netherlands and is kept navigable by a seven weir system. These weirs differ in dimension, head loss and operating system. Based on the characteristics of these weirs one simplified fictive weir is chosen and tested on the consequences of various calamities. Besides, also the long-term maintenance strategy is considered in this study. Both have the similar result that part of the weir is non-operational for a longer period, discharge capacity is limited and all highly variability in occurring discharges is possible.

In this study the consequences to the water management of the fictive weir are considered. To guarantee an upstream water level to a certain target level, panels and slides are removed from a weir for increasing discharge. If part of the weir is not available due to a closure, panels and slides have to be removed earlier than in a normal situation. Therefore, the limit discharge, the discharge for which the weir is completely open, is reached earlier for the different closure situations. If the limit discharge is reached, the upstream water level starts to increase. In this thesis the discharge, according to an upstream water level equal to the flood plain level, is the maximum considered discharge. For further increasing discharge, and water levels, the flood plains start to discharge water and the flow velocities downstream of the weir will no further increase.

The flow velocities for three closure situations and three discharge situations are considered in this study. It turned out that the flow velocities which are calculated in this thesis are always lower than the flow velocities which occurred during the Grave weir calamity. Therefore it is concluded that the measures used during the Grave calamity, are all applicable to the other weirs in the Meuse. But lastly is must be concluded that due to a relative large closure of a weir (>50%), severe flow circumstances like supercritical flow, can occur. In case of a future calamity or long-term closure, it is advised to minimize the closure width as much as possible.