Experimental uncertainty analysis of monopile scour protection stability tests

Hydraulic experiments using physical scale models of monopile scour protections typically have high ex-perimental uncertainties and data scattering. These uncertainties may severely affect the accuracy of the experimental results and need to be analysed quantitatively. This paper presents a study on the quantification of experimental uncertainties in monopile scour protection damage tests following the Guide to the expression of Uncertainty in Measurement (GUM) (ISO, 2008). The uncertainty analysis is performed using the three-dimensional damage number S3D and the widely applied STAB number. Through the analysis of the S3D number from individual test and repeated tests, it is found that the uncertainty analysis method for an individual test can be efficiently applied to predict the experimental uncertainty. The wave peak period and the current velocity are identified as the two major sources of uncertainties for the S3D number. The flow turbulence and correlations between input parameters can be neglected when estimating the uncertainty. The uncertainty analysis of the STAB number shows that the experimental uncertainty due to measurement can be up to 5% to 7% of the obtained STAB result. This uncertainty range is wide in comparison with the narrow margin of a dynamic scour protection design in DNV's recent recommended practice (DNV-RP-0618). The uncertainty of the STAB number is more affected by the armour stone density, wave peak period and significant wave height.