Effectiveness of sensors in flood defences

Introduction The popularity of monitoring dikes with sensor techniques is rising. It is claimed that sensor techniques lead to significant cost savings and can predict an upcoming dike collapse. But a technical foundation to use the sensor monitoring information in flood safety assessment is lacking. This research investigates the contribution of sensor monitoring information to flood safety and the cost-effectiveness of sensor monitoring. Sensor techniques have been tested in full-scale dike failure experiments at the IJkdijk, trying to predict an upcoming dike collapse. The sensor techniques are capable of monitoring deformation, temperature, water pressure, vibrations and moisture. The state of the sensor techniques is doubtful due to subjective analyses, controlled test conditions and a wide variety in failure prediction times: from 1,5 to 102 hours. Implementation of sensor information Water pressure is the only variable that constitutes an input for dike safety assessment models. Monitoring water pressures affects the epistemic uncertainty of the water pressure schematization which is caused by the translation from the hydraulic load to water pressures. One must be aware that sensor monitoring either leads to an increased assessment of flood safety if the prior schematization turns out to be done conservatively or a decrease in flood safety if the prior schematization turns out too optimistic. One would expect an increased assessed flood safety due the intended conservative approach. But prior schematization mistakes imply a decreased assessed flood safety. Moreover, monitoring water pressures has minimum impact on the flood safety assessment if other uncertainty aspects dominate the stability assessment. A case study of the canal of Nauerna denotes that the water pressure has resulted in a higher assessed flood safety, but the uncertainties regarding the soil conditions dominate the stability assessment. However, sensor monitoring on itself does not affect the real flood safety: only physical measures affect the real risk of flooding. Important information is obtained from monitoring high water events such that water pressure models can be calibrated to determine design loading conditions for the periodic safety assessment. Also, an additional application is to identify unforeseen risks. Cost-benefit analysis Conceptual cost-benefit models have been set up to determine the cost-effectiveness of sensor monitoring over the long-term. The monitoring costs consist of installation, maintenance and operational costs. The benefit from permanent dike reinforcements is gained from specifying the long-term optimal investment strategy based on the minimum sum of flood risk and reinforcements costs. The monitoring information from relevant high water events affects the assessed flooding probability. If the sensor monitoring reduces the assessed flooding probability, the assessed flood risk lowers and savings on permanent dike reinforcements. If a higher flooding probability is obtained by sensor monitoring, this financially leads to additional investments and negative benefits. But the value of knowing this higher flood risk is rationally beneficial. The benefit from temporary measures is gained from timely execution of emergency measures based on the early warning of the sensor system. This benefit depends on the prediction time of the sensor monitoring system and the reaction time to execute the emergency measure. Additional costs for executing the emergency measure must be incorporated. The cost-benefit models have been worked out in case studies for dike-ring 48 and 14. Conclusion The conclusion of this research is that sensor monitoring can be implemented in the flood safety, by specifying dike reinforcements in both the periodic safety assessment, as well as the operational situation. However, the investments in sensor monitoring have to be made while a long waiting time is expected before benefits turn out. Then these benefits can financially be disappointing, but do have a certain value of information.

Subject

sensor monitoring
cost-benefit analysis
macro instability
flood risk
probabilistic design
flood defence

To reference this document use:

http://resolver.tudelft.nl/uuid:ad0a5aaa-9493-4ed0-a02e-1c0a2d2c6156

Embargo date

2012-11-09

Part of collection

Student theses

Document type

master thesis

Rights

Files

PDF

Spaargaren2012.pdf

7.61 MB

Close viewer