Page 478 - Practical Design Ships and Floating Structures

453

300wx200hx1200’(mm). A schematic diagram of the fence model and coordinate system used in this
study is shown in Figure 1. The d&(D) and width(W) of the oil fence are 40mm and 300mm,
respectively and the fence model has a flat end with a thickness(B) of 5mm made of acrylic material.
The free stream velocity was uniform at lOcm/s and its corresponding Reynolds number (Re) and
Froude number (Fr) based on the fence draft are about 4000 and 0.16, respectively. The free stream
turbulence intensity is about 0.5% at this velocity. The experiment is conducted for two free-surface
conditions i.e. open and closed free surface between the tandem fences.

4.2 Dropiet Tracking
The experiment to obtain the trajectories of paraffin balls“’ of different sizes which are released in the
upstream of a model fence is carried out at a larger circulation water channel of lmw x lmh x 4.2m’ in
order to increase the Reynolds number to the order of lo4 - los and make the balls to escape beneath
the fence, which cannot be done with the smaller channel. The fence draft was increased to 15Omm
and the Froude number to 0.292. The tracking of the motion of the balls is done by CCD camera as
schematically shown in Figure 2a and a sample photograph of the trajectory in Figure 2b.

5 RESULTS AND DISCUSSIONS
In Figure 3 the variations of the horizontal velocity profiles at several longitudinal locations obtained
by the PIV method and computations are shown.
Since our major interest lies in the flow behavior in the upper layer close to the free surface, attention
should be given to this region i.e. Y/D>-I .O, Due to the light density of the oil, the oil leaked below the
fore fence would tend to float up to the free surface. Thus, if it is assumed that the oil layer covers the
free surface between the tandem fence, it is our focal subject to find out if the computational method of
the flow field with the closed free-surface assumption would be valid or not.
Several interesting phenomena can be observed from Figure 3. In Figure 3a, a comparison of the
computed and experimental results is shown. If we focus our observation in the region of Y/D > -1 .O,
we can conclude in general that the computational method with the closed free-surface condition is
reasonable. In Figure 3b in the region of Y/D > -1.0 and 1.81 4 X/D (5.85, back flows exist for the
tandem fence and the magnitude is greater for the open free surface than the closed free surface. The
greater back flow in the upper fluid zone is favorable in view of trapping the leaked oil. The
computational results in Figure 3b show under predicts the back flow velocity, which implies that the
computational results are on the more conservative side than the experimental results in assessing the
tandem-fence effectiveness.

When a fence is deployed, the fence will block the movement of an oil slick, and the accumulated oil
in front of the fence forms a triangular shape. The initial stage of the leakage takes place by the oil
droplets sheared off the surface of the oil layer. Depending on where these droplets start their motion,
they could be reattached to the oil layer or leaks below the fence. The computational method can
predict a zone in the upstream of the fence in which oil droplets having started their movement will
escape below the fence.

In Figure 4 the demarcation lines called as “Leakage Line” are shown for two oil droplet sizes and two
Froude numbers. If the droplets begin their movement below the leakage line, they could leak below
the fence. The dotted lines are for the deformed fence. One can observe from this figure that the bigger
a droplet, the lesser the current speed and the lesser the fence deformation, the lesser the chance of oil
leakage.

(‘) The laboratory experiments using a paraffin ball and the equivalent diameter oil droplet were conducted by Lee et
al.( 1998) and found they exhibited the same trajectories