Landfills represent a significant threat to groundwater contamination due to their nature of operation and their abundance. Monitoring well networks at these sites are of vital importance in detecting leakage plumes. This study presents a reliability assessment to estimate the performance of groundwater monitoring systems at landfill sites. A hypothetical problem is presented where the detection probability of several monitoring systems is compared. A Monte¿Carlo approach is used to incorporate uncertainties due to subsurface heterogeneity and the leak location. Hydraulic conductivity and leak location are considered as random variables with prescribed probability density functions. A finite difference groundwater model coupled with a random walk particle-tracking model simulates a contaminant plume released from the landfill for each Monte¿Carlo realization. The analysis shows that lateral dispersivity of the medium has a significant influence on the reliability of the monitoring system, since it is the primary parameter controlling the width of the contaminant plume. Furthermore the number and the location of the monitoring wells are dependent on the heterogeneity of the medium and size of the contaminant leak. It is concluded that the reliability of the common practice of three downgradient monitoring wells is inadequate from the point of view of prevention of groundwater contamination due to landfills. Keywords: Detection probability; Groundwater monitoring; Landfill; Contaminant plume; Monte¿Carlo analysis@en

Rectangular mudmat foundations for subsea structures are commonly designed according to ISO and API guidelines. However, the modern trend towards deeper waters and heavier structures is currently challenging traditional approaches to foundation design, especially in presence of particularly poor soil conditions. The intrinsic conservatism carried by design codes may not only lead to higher fabrication/installation costs, but also conflict w ith the operational limits of existing installation vessels. Based on the analysis of four design cases, it is argued that ISO/API safety factors may be from 1.5 to 2.3 times as low as the predictions of elastoplastic finite element (FE) simulations. Such a conservatism is also found to considerably increase with the embedment depth, whereas the influence of the foundation aspect ratio is nearly negligible. A simple calculation example shows that the systematic use of numerical analysis in design may result in a 47% reduction of the total foundation area, and an overall decrease in fabrication costs of approximately 16%.@en