Installation of large subsea structures - Lowering of suction anchors through the splash zone

Abstract

In this thesis the dynamic air cushion pressure inside suction anchors during low- ering through the splash zone has been studied. A non-linear continuity equation for the air mass is used to establish a theoretical model for the dynamic air cush- ion pressure. Buoyancy forces due to dynamic air cushion pressure inside the suction anchors are calculated until the wave elevation inside the suction anchor reach the top. At impact Wagner s wedge method together with Von Karman s wetted length is used to calculate the impact force on the cone-shaped suction anchor top. Important parameters affecting the dynamic air cushion pressure and thus the buoyancy forces have been investigated. Force measured in lifting wire during splash zone transition of a single suction anchor is used as basis for comparison. The parameters that are investigated are introduction of a time-varying lowering velocity, changing characteristics of the ventilation hatch outlet and introducing regular and irregular waves. By varying these parameters a best-fit comparison with 2.5 [%] deviation between calculations and measurements have been ob- tained. Further, a comparison between theoretical calculations and force measurements from a lifting operation of a subsea structure with suction anchors as foundation has been performed. The buoyancy forces calculated prior to impact are due to the dynamic air cushion pressure alone. The total force in the lifting wire for this phase are corresponding well to the magnitude of the measured force. During this comparison the effect of changing perforation ratio clearly show an increase of buoyancy forces for decreasing size of ventilation hatch. Including both inflow and outflow of air from the suction anchors when oscillating in large waves have also been studied. When comparing with force measurements the im- portance of including both inflow and outflow of air to account for both suction and compressive pressure has been shown. This lead to negative and positive buoyancy forces respectively. Long crested waves accounting for wind and swell sea are calculated with the JONSWAP- and Torsethaugen-spectra. By using empirical formulas for the lim- iting case between wind and swell sea it is shown that wind sea is the dominating part for the sea state measured during splash zone transition of the suction an- chors. Using Wagner s wedge model together with Von Karman s wetted length for cal- culating impact force on a flat suction anchor top is also performed. Impact forces with impact angles in the range of 5-15 degrees between the wedge-shaped water and top overestimate the force compared with measurements. A best-fit for the impact forces with the measurements has been obtained with deadrise angle of 17.5 [deg] in addition to impact forces calculated with a time-varying slamming coefficient. Possible resonant phenomena connected with oscillations of suction anchor and pressure measurements at impact have been studied. Simplified models for cal- culating A) natural period of oscillating air cushion, B) piston mode of water column and C) coupled piston mode of air cushion and water column has been developed. These have been compared with oscillation-period of pressure mea- surements observed after impact of a single suction anchor. These calculations have also been compared with oscillations of the suction anchor observed from videos of experiment. Case B) proved to give the most similar results with the suction anchor oscillations while case A) and C) were not in the proximity of the observed oscillations. This showed that the air cushion were not causing any resonant motion. Using a common safety criterion for avoiding slack in lifting wire, the calculation model have been used to give a recommendation for perforation ratios for the suction anchors used on the installation of the subsea structure studied here. Perforation ratios in the range 1-3[%] used on suction anchor with diameters in the range 3-6[m] proved to be well within the slack-wire limit. When using the theoretical model developed in this thesis a design recommen- dation for using ventilation hatches with a certain height have been made. Also uniform hatches are advised to ensure as similar conditions as possible for outflow and inflow of air.