Innovative technologies to accurately model waves and moored ship motions

Late in 2009 CSIR Built Environment in Stellenbosch was awarded a contract to carry out extensive physical and numerical modelling to study the wave conditions and associated moored ship motions, for the design of a new iron ore export jetty for BHP Billiton, at Port Hedland in north-west Australia. International consultants WF Baird and Associates are reviewing all technical aspects of the project and also have staff attending the tests. The project had tight deadlines and required extreme accuracy of modelling due to the small nature of the long-period design waves. It is this long-wave energy that has the ability to excite the large moored bulk-ore carrier vessels, which can be in excess of 300 000 Dead Weight Ton (DWT). A numerical model of the 1:100 scale laboratory wave basin of 35 m by 60 m has also been set up to verify the numerical modelling approach. Direct use of the physical model results for vessel downtime calculations is not realistic since the area of interest is much larger than the modelled domain. Therefore the calibrated numerical model will be used, after verification, to model all the different combinations of extreme wave, wind and current conditions. Waves in the scale model are generated by a 24-m-long bank of shallow water wave generators (recently imported from the UK). Research was also carried out to ensure that the boundaries of the physical model were covered with a wave-absorbent slope to eliminate erroneous wave reflections. Model waves were also accurately measured (to 0,2 mm) by a CSIR-developed Keofloat system at each berth and in the basin, using video image processing technology. A similar technology was used to measure the moored ship motions. Strain gauges were used to measure mooring and fender forces. Another aspect of the model study was to physically and numerically model the effect of loaded ships passing close to the moored vessels. This proved to be a critical aspect of the terminal design, as the allowable passing distance would have a significant effect on the area and volume of seabed required to be dredged. The operational safety of the moored vessels during loading also had to be determined as part of the design. Any downtime or delays to these large vessels could be very costly to the terminal operators. This project highlights the many technologies developed and used by the CSIR to undertake specialist studies, both physical and numerical, to support the design of safe harbours and terminals. This paper will briefly describe those key technologies. 1 Physical modelling 1.1 Wave basins available at the CSIR – deep and shallow, plus flumes – scales The CSIR’s Hydraulics Laboratory at Stellenbosch has one of the largest model halls available worldwide, in comparison with the other top international coastal engineering laboratories. This allows the choice of a larger scale and/or larger coverage of the area to be modelled. The 35 m by 60 m model basin (see Figure 1) represented a prototype area of 3,5 km wide by 6 km long, which covered the whole proposed new dredged basin and approach channels at a 1:100 scale. This is important as it permitted the inclusion of adequate offshore and adjacent bathymetry in front of the wave generators to obtain a true representation of the local wave field, allowing the full effects of free surface gravity waves. Two other large 3-dimensional (3D) basins and a number of 2-dimensional (2D) flumes are also available within the CSIR Hydraulics Laboratory. Figure 1: CSIR model wave basin 1.2 New wave generators and their capabilities The basin was orientated so that the main incident wave direction could be generated perpendicular to the 24 m bank of wave paddles (new movable wave generators imported from the UK, see Figure 2). The wave conditions that were tested focused on the long swells that could excite low-frequency ship motions. The paddles are driven by signal-generation software capable of creating short crested waves with setdown compensation to simulate second-order boundary conditions, thereby forming the theoretical bound long waves required to test the motions of the moored ship. The measured target wave spectrum was smoothed such that the spectral shape and total wave height were retained, but the high-frequency tail of the spectrum and any sea waves were discarded, because these waves have little effect on ship motions. Figure 2: Movable wave generators 1.3 Wave absorption slopes and basin resonance The wave generators are equipped with active wave absorption, however, this feature was turned off for these particular model tests because the required absorption of very small long waves (2 to 3 mm model wave height) were better absorbed at the model boundaries. This absorption was achieved by placing a wide slope of small stones around the model boundary walls. The optimum width, slope and size of stone were tested in a 2D wave flume before being placed around the 3D basin. The achieved wave reflection off the boundary walls was less than 15%, which allowed accurate simulation of the prototype waves at the moored ship’s jetty location. Another strategy to improve the accuracy of the model was to place loose stones behind the wave generators and wave guides on the sides of the model. This had the desired effect of eliminating any spurious basin resonance. 1.4 Wave gauges – capacitance probes and Keofloats Due to the small size of the waves (long waves as small as 0,2 mm model scale) and the extreme accuracy of measurement required, two separate wave-measurement systems were employed in the physical model. Capacitance probes were used for the larger waves closest to the wave generators, while Keofloats were used for the smaller waves close to the moored vessels and at the back of the dredged basin. Figures 3A and 3B show a capacitance probe and a Keofloat on the left and right, respectively. Capacitance probes have an accuracy of about 0,5 mm and consist of twin wire gauges attached to an amplifier. Through calibration, the voltages obtained as output are coupled to the corresponding water level. Figure 3A: Capacitance probe Figure 3B: Keofloat When it became necessary to measure waves much smaller than 10 mm in the model, for which the noise levels on the signal become significant for traditional resistance and capacitance wave gauges, a new system was developed by the CSIR, consisting of small lightweight floating blocks, called Keofloats (see Figure 3B. These floats are tracked by a standard video camera and are insensitive to erratic gauge drift. Noise levels are therefore very low and they do not require separate calibration. The accuracy is estimated at 0,2 mm. The Keofloat system has been tested in a flume and compared with capacitance probes (Terblanche et al., 2009). It was concluded from the flume tests that the Keofloats are superior to traditional wave gauges for wave heights of less than 5 mm. Keofloats are equivalent to wave gauges for wave heights between 5 mm and 20 mm. The results of the flume test comparison are shown by the plots in Figure 4. 119