HSEHealth & Safety
Executive
A series of experiments to studythe spreading of liquid pools with
different bund arrangements
Prepared byAdvantica Technologies Limited
for the Health and Safety Executive
CONTRACT RESEARCH REPORT
405/2002
HSEHealth & Safety
Executive
A series of experiments to studythe spreading of liquid pools with
different bund arrangements
P S Cronin and J A EvansAdvantica Technologies Limited
Ashby RoadLoughboroughLeicestershire
LE11 3GRUnited Kingdom
Assessment of the hazards posed by the storage of flammable or toxic liquids in large tanks can beassisted by the use of mathematical models to calculate the consequences of leakages. Theseconsequences may include fires or explosions from dispersion of flammable vapours, or harm topersons from inhalation of toxic vapours. One component of such models is a mathematicalrepresentation of the spreading of a liquid pool, and in recent years a number of pool spread modelshave been proposed and implemented.
However, there are a number of issues in the formulation of the spreading models that have yet to beresolved. In particular, one area of uncertainty is the boundary condition applied at the front of thespreading pool. Boundary conditions which are generally accepted as applicable to the spread of oil onwater and to the dispersion of a cloud of dense gas in air may not be applicable to the spread of aliquid on land, as the balance of competing physical phenomena at the spread front is quite different.Resolution of the uncertainties this creates has been hampered by a lack of reliable experimental dataat a large scale. As a result, the Health and Safety Executive have contracted Advantica to carry out aseries of ‘liquid spread on land’ experiments at their Spadeadam Test Site to provide a sufficientlydetailed database to help to resolve this issue. This report describes the resulting programme of58 experiments.
The experiments studied the flow of liquid across a bund floor and measured the amount of liquid thatescaped the bund for a wide range of bund geometries.
This report and the work it describes was funded by HSE. Its contents, including any opinions and/orconclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.
HSE BOOKS
ii
© Crown copyright 2002Applications for reproduction should be made in writing to:Copyright Unit, Her Majesty’s Stationery Office,St Clements House, 2-16 Colegate, Norwich NR3 1BQ
First published 2002
ISBN 0 7176 2255 X
All rights reserved. No part of this publication may bereproduced, stored in a retrieval system, or transmittedin any form or by any means (electronic, mechanical,photocopying, recording or otherwise) without the priorwritten permission of the copyright owner.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page iii
Executive Summary Assessment of the hazards posed by the storage of flammable or toxic liquids in large tanks can be assisted by the use of mathematical models to calculate the consequences of leakages. These consequences may include fires or explosions from dispersion of flammable vapours, or harm to persons from inhalation of toxic vapours. One component of such models is a mathematical representation of the spreading of a liquid pool, and in recent years a number of pool spread models have been proposed and implemented. However, there are a number of issues in the formulation of the spreading models that have yet to be resolved. In particular, one area of uncertainty is the boundary condition applied at the front of the spreading pool. Boundary conditions which are generally accepted as applicable to the spread of oil on water and to the dispersion of a cloud of dense gas in air may not be applicable to the spread of a liquid on land, as the balance of competing physical phenomena at the spread front is quite different. Resolution of the uncertainties this creates has been hampered by a lack of reliable experimental data at a large scale. As a result, the Health and Safety Executive have contracted Advantica to carry out a series of ‘liquid spread on land’ experiments at their Spadeadam Test Site to provide a sufficiently detailed database to help to resolve this issue. This report describes the resulting programme of 58 experiments. The experiments studied the flow of liquid across a bund floor and measured the amount of liquid that escaped the bund for a wide range of bund geometries. From a large tank with an initial fill height of up to 1.8 m, water was released through a slot at the base and was free to spread over distances of up to 10m on a specially constructed horizontal concrete surface. The rate of release was such that the vessel emptied in about 30 seconds. The experiments investigated the flow rate of water from the vessel, the initial rate at which the pool spread and the quantity of water which overtopped the bund. The data collected are summarised in this report. The large scale of the test rig means that the experiments also have a direct bearing on issues of a practical nature. Most tanks used for storage of flammable or toxic liquids in the United Kingdom have capacities less than 50,000 cubic metres, although several tanks do exist with capacities in excess of 100,000 cubic metres. The test tank constructed at Spadeadam replicates a tank of about 150,000 cubic metres capacity at a linear scale of one-twentieth, or alternatively a tank of about 20,000 cubic metres capacity at a linear scale of one-tenth. Hence, the experiments give information that can be related to behaviour on onshore sites.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page iv
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page v
Contents 1 Introduction .......................................................................................................1 2 EXPERIMENTAL DETAILS ................................................................................2
2.1 THE TEST RIG ................................................................................................2 2.2 INSTRUMENTATION..........................................................................................2
3 Experimental Programme.................................................................................3 3.1 PHASE 1 ........................................................................................................3 3.2 PHASE 2 ........................................................................................................4 3.3 EXPERIMENTS TO VIDEO THE WATER FRONT ....................................................4
4 Experimental Results........................................................................................5 4.1 WATER FLOW.................................................................................................5 4.2 TIME OF ARRIVAL OF THE LIQUID FRONT ...........................................................6 4.3 WATER OVERTOPPING ....................................................................................7 4.4 IMAGES OF THE WATER FRONT ........................................................................8
5 Discussion .........................................................................................................8 5.1 PHASE 1 ........................................................................................................8 5.2 PHASE 2 ........................................................................................................9
6 Summary..........................................................................................................10 7 References.......................................................................................................10
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page vi
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 1
1 INTRODUCTION Assessment of the hazards posed by the storage of flammable or toxic liquids in large tanks can be assisted by the use of mathematical models to calculate the consequences of leakages. These consequences may include fires or explosions from dispersion of flammable vapours, or harm to persons from inhalation of toxic vapours. One component of such models is a mathematical representation of the spreading of a liquid pool, and in recent years a number of pool spread models have been proposed and implemented [1,2].
However, there are a number of issues in the formulation of the spreading models that have yet to be resolved. In particular, one area of uncertainty is the boundary condition applied at the front of the spreading pool; in this respect the modelling described in [1] is fundamentally different to that described in [2], for example, and in a number of other software models known to be in use. Boundary conditions which are generally accepted as applicable to the spread of oil on water and to the dispersion of a cloud of dense gas in air may not be applicable to the spread of a liquid on land, as the balance of competing physical phenomena at the spread front is quite different. Although information is available from a variety of sources ([3,4,5], for example), resolution of the uncertainties this creates has been hampered by a lack of reliable experimental data at a large scale. As a result, the Health and Safety Executive have contracted Advantica to carry out a series of “liquid spread on land” experiments at their Spadeadam Test Site to provide a sufficiently detailed database to help to resolve this issue, as well as to allow more general validation. This report describes the resulting programme of 58 experiments.
The experiments studied the flow of liquid across a bund floor and measured the amount of liquid that escaped the bund for a wide range of bund geometries. From a large tank with an initial fill height of up to 1.8 m, water was released through a slot at the base and was free to spread over distances of up to 10m on a specially constructed horizontal concrete surface. The size of the slot was such that the tank emptied completely in about thirty seconds. More rapid releases can be imagined, and indeed have occurred in practice when storage tanks have failed catastrophically. However, the rate of release in these tests was judged adequate for the intended purpose of the tests. The experiments investigated the flow rate of water from the vessel, the initial rate at which the pool spread and the quantity of water which overtopped the bund. The data collected are summarised in this report and full copies of the data are provided with the report in electronic format.
The large scale of the test rig means that the experiments also have a direct bearing on issues of a practical nature. Most tanks used for storage of flammable or toxic liquids in the United Kingdom have capacities less than 50,000 cubic metres, although several tanks exist with capacities in excess of 100,000 cubic metres. The test tank constructed at Spadeadam replicates a tank of about 150,000 cubic metres capacity at a linear scale of one-twentieth, or alternatively a tank of about 20,000 cubic metres capacity at a linear scale of one-tenth. Hence, the experiments give information that can be related to behaviour on onshore sites.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 2
2 EXPERIMENTAL DETAILS
2.1 The Test Rig The release vessel (Figures 1 and 2) was designed to represent at 1/20th scale a quadrant of a 70m diameter storage tank. The lower section of the vessel was constructed from a section of curved mild steel with a 3.5m diameter. Elsewhere the cross-sectional shape was modified slightly in order to strengthen the tank whilst maintaining the same cross-sectional area, as shown on Figures 1 and 2. For all of the tests, the vessel was filled with ambient temperature water, fed through a 3” hose, from the site fire water main. The tank was also fitted with a 2” nominal bore gate-valve, which could be used to drain it.
The liquid release mechanism is shown in Figure 3. A slot was cut around the full circumference of the lower section of the tank wall, and into it was fastened a quarter-circular strip of mild steel containing a slot 25 mm high through which the water could be released. The bottom of the slot was 20mm above the concrete surface. The water was initially held in the tank by an array of five closed flaps. On the outside of the tank was a mechanism whose activation opened the flaps and allowed the water to flow freely out of the tank. The release mechanism was controlled by a PC based SCADA system connected to a sequence timer system which co-ordinated the operation of the release mechanism and the instrumentation.
Bund walls were fabricated from mild steel and set up on a flat 15m square concrete pad with the release vessel at one corner. Thirteen different bund configurations were used, 9 circular and 4 square, see Figures 4 to 10. All configurations were sized to give nominally the same containment volume. The two sides of the 90o quadrant were enclosed with 250mm deep flat plate, and were assumed to act as walls of symmetry. Perspex sections were fitted into one of the side walls to allow video records to be made of the water flow.
For some tests, 1 circular and 2 semi circular sections were installed within the bund to represent a regular array of additional tanks, see Figures 10 and 11. The sections were fabricated from 200mm wide flat steel strip rolled to a 3.5m diameter.
2.2 Instrumentation A scale was marked on the inside of the tank to show the level of water within it above the bottom of the slit. In addition a pressure transducer was fitted on the back of the tank at a height level with the slit. During the tests the signal from the pressure transducer was recorded on a transient recorder sampling at 200Hz.
The movement of water across the bund floor was monitored using up to 60 resistance probes, fixed in position on the concrete bund floor using terminal block. The positions of the resistance probes used for these tests are shown in Figures 12 and 13 and the polar co-ordinates of the probes are listed in Table 1. The resistance probes consist of two electrodes separated by a gap (typically 5-10 mm), which provides a high electrical resistance. The arrival of the water lowers the resistance across the gap and triggers a TTL voltage step output from a purpose-built electronic circuit, which acts to terminate a computer based counting register on a counter
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 3
board. Counting was started when the release was initiated, and thus, using the known count frequency, a measure of the time of arrival of the water front at a specific probe location was obtained. The accuracy of the system has been checked by comparison with cine records and has been shown to indicate arrival times with an accuracy of better than 1 millisecond. As there may be a slight variation in the time at which the release valves opened after initiation of the release, time zero for the resistance probe records has been taken as the time when the first resistance probe was triggered.
Water that overtopped the bund was caught in a polythene sheet attached to the (outside) edge of the bund, then pumped into calibrated containers to measure its volume.
Up to four video cameras were used in the experiments; the images being recorded on MiniDV video tape. Video timers provided a timescale for the images on the video tape. One camera was used to provide general images of the release, another provided an overhead view of the floor of the bund whilst the other two cameras were used to monitor items of specific interest, for example, the time of arrival of the water front at the bund wall, or the release of water in the vicinity of the tank.
3 EXPERIMENTAL PROGRAMME The Test Programme was made up of two Test Phases. The experiments in Phase 1 were conducted using circular bunds and the experiments in Phase 2 were conducted using square bunds. The experiments conducted in each phase are detailed in Tables 2 and 3.
In both Test Phases, three nominal fill heights were used for the experiments: 1.45m, 1.60m and 1.75m, corresponding to water inventories of approximately 3.9 cubic metres plus and minus 10%. The actual fill height in each experiment is given in Tables 2 and 3. The water inventory of the release vessel at the three nominal fill heights is given in Table 4.
3.1 Phase 1 In the first of the two Test Phases, 37 experiments were carried out to examine liquid spread over flat uninterrupted terrain and interaction with a single, circular bund wall. Different combinations of bund location and height were investigated and bunds with face angles, relative to the oncoming water front, of either 30 degrees, 45 degrees or 90 degrees were used. Figures 4 to 6 show the different bund arrangements used in the experiments in Phase 1. They were chosen so that the capacity of the bund was close to 3.9 cubic metres in each case. Table 5a gives the actual volume of the bund for the different experimental arrangements.
The resistance probes were positioned as shown in Figure 12. Probes deployed along two radial lines monitored general progress of the front, and probes located along circular arcs gave information about the radial symmetry.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 4
3.2 Phase 2 In the Phase 2, 21 experiments investigated releases into square bunded areas, in some cases with obstructions present. Figures 7 to 10 illustrate the different configurations. In all of these cases, the bund walls presented a vertical face to the oncoming water. Again each configuration was chosen so that the capacity of the bund was close to 3.9 cubic metres. Table 5b gives the actual volume of the bund for the different experimental arrangements
Configurations SQ1 and SQ2 each represented one-quadrant of the case of a square bund surrounding a concentric circular tank, a situation that is often encountered in practice; the two configurations differed because they made use of the different lines of symmetry in that full-plan arrangement. If the edges of the test pad had no impact on the water flows then bunds SQ1 and SQ2 would give identical overtopping results. Significant differences between results would show that edge effects were present, suggesting that a degree of caution would be necessary in extrapolating these quadrant results to the full-plan situation.
In addition, because each of configurations SQ1 and SQ2 had the same plan area and the same height as the highest of the circular bunds, appropriate comparison of results would address the question of how, for a given release into a bund of given capacity, the retention capability is affected by the shape of the bund. Similarly, square configuration SQ3 was chosen to have the same plan area and the same height as the lowest of the circular bunds, allowing the same question to be addressed in these different circumstances.
Finally, square arrangement SQ4 was defined to have the same (free) area and the same height as square arrangement SQ3, the only difference being the introduction of obstacles. Appropriate comparison would allow conclusions on the effect of obstacles on the flow and overtopping behaviour.
The resistance probes were positioned as shown in Figure 12 for the experiments with bund arrangements without obstacles (SQ1, SQ2 and SQ3) and as shown in Figure 13 for the experiments with obstacles present (bund arrangement SQ4).
3.3 Experiments to Video the Water Front Three further experiments were carried out to film the shape of the profile of the water. The water front was filmed using a Kodak Motioncorder 500 high speed camera system operating at 125f.p.s., set-up as shown in Figure 14. For all three releases an initial water head of 1.60m was used and no bund walls were used so the water spread unrestricted across the concrete pad. Video images of the water front were taken at 5m, 7.1m and 10m radii.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 5
4 EXPERIMENTAL RESULTS
4.1 Water Flow The water pressure (and hence water depth) was measured by the pressure transducer located at the base of the release vessel in each experiment. The liquid head data for all the tests carried out is included with this report in electronic format.
Figure 15 shows the variation of water depth in the release vessel with time, for three typical tests with the three initial fill heights. Also shown on this figure are predicted values obtained using the Bernoulli equation for the outflow velocity and the measured value for the area available for outflow, assuming a discharge coefficient (Cd) with a value of 0.64:
( )2
0
221
��
�
�
��
�
� ×××××−=
cross
outd
A
tgACHtH
where
H0 is the initial height of the water, nominally 1.75m, 1.60m or 1.45m
Aout is the cross-sectional area of the hole = (0.025 * 2 * 1.75 * π)/ 4 m2
Across is the cross sectional area of tank = 2.411m2 (from tank capacity),
g is acceleration due to gravity = 9.81ms-2, and
t is the time, in seconds.
The agreement between the values inferred from the measurements and the predictions is very close. This gives some confidence that the measurements were made correctly and that the release mechanism worked as intended.
The measurement of the liquid head can be used to determine the water flow rate from the vessel. Figure 16 shows the variation in water flow rate with time inferred from the measurements of pressure for the same three experiments.
The data recorded by the pressure transducer inevitably contain some random, high frequency noise and, possibly, other more physically-based, oscillations, superposed on the signal recording the changing head of water as the tank empties. Such oscillations are particularly apparent in the first few seconds immediately following the opening of the flaps to release the water. As a result, some form of time averaging is required to remove this high frequency component from the data if they are to be differentiated numerically in order to infer a representative flow rate from the vessel. Figure 15 and Figure 16 show the liquid level and the flow rate from the vessel inferred from the numerical differentiation of 'smoothed' pressure data for three specific experiments. The curves shown in this figure were produced using 0.5 second time averaging to remove the high frequency oscillations. As a check, it was found that integrating the resulting flow rate over time gave agreement to within 1% of the original volume of water recorded as being in the tank at the start of the experiment.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 6
4.2 Time of Arrival of the Liquid Front Resistance probes data were collected for thirteen of the experiments conducted with circular bunds and sixteen of the experiments conducted with square bunds. Tables 2 and 3 show the experiments in which the resistance probes were used. The time of water arrival at each resistance probe in these experiments is given in Tables 6 to 9. As there may be a slight variation in the time at which the release valves opened after initiation of the release, time zero for the resistance probe records has been taken as the time when the first resistance probe was triggered. The complete resistance probe data are included with this report in electronic format.
Figure 17 shows the time of arrival of the liquid front against distance from the release vessel for representative tests with a circular bund with the three initial fill levels. These data were collected in experiments where the bund wall was located 10m from the tank centre.
The time of detection of water at the resistance probes was used to infer the velocity of the front. Using the time difference between arrival of water at adjacent resistance probes along a radial line was found to give a local velocity with a significant amount of variability. Consistent with this, small local variations in progress can be observed on the video records. However, a smooth polynomial curve was fitted to all of the time of arrival data collected in experiments with the same initial fill height in the vessel and this curve was differentiated to obtain an ‘averaged’ front velocity. A comparison of this inferred ‘average’ velocity with one particular set of ‘local’ velocity data is shown in Figure 18.
It should be noted that an examination of the video records and time of arrival records of the experiments suggests that the disruption to the flow caused by the presence of the resistance probes may have slowed the progress of the front slightly along the line of the probes. The size of the terminal block used to hold the resistance probe wires close to ground level was approximately 14mm and it was observed that there was a small wake-like region created in the lee of each probe. The video records suggest that these wake-like regions did not spread to affect the radial outflow at other locations. The maximum effect this produced appears to have occurred in the tests with the greatest initial fill level in the tank. In this case, the maximum difference in the time of arrival of the flow along the line of the probes and elsewhere at a distance of 10m (i.e arrival at the bund wall in the relevant tests) is about 0.5 secs. This suggests that the average speed of progress inferred from the resistance probe measurements systematically underestimates the true value by 0.05 m/s at most (c.f. observed value of about 0.37 m/s – corresponding to a discrepancy of about 13% in the values obtained from the resistance probe data along the 45 degree radial line). Whilst such differences should be born in mind if the data are used to compare with the predictions of mathematical models, they are of a smaller magnitude than differences in the predictions from the models referred to in [1] and [2].
Figure 19 shows the time of arrival of the liquid front along the radial line from the release vessel at 45 degrees for bund geometries SQ3 and SQ4 for the three initial fill heights. Examination of the video records of these experiments reveals that the flow was channelled between the adjacent obstacles, resulting in a faster, deeper
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 7
flow between these ‘model’ tanks. The radial symmetry was disrupted and the flow appeared to separate from the obstacle tank sides, producing two tongues of fluid from between the model tank obstacles. This flow hit the retaining walls and spread out along them, meeting at the sides and in the corner and then moving back towards the rear of the tank obstacles. The data show that the effect of the obstacle is similar in all of the experiments, with the water reaching the rear face of the obstacle (in its wake) after it has reached the far bund wall.
The time of arrival of the liquid front along the 22.5 degree radius against distance from the release vessel for bund geometries SQ3 and SQ4 for the three initial fill heights is shown in Figure 20.
4.3 Water Overtopping The volume of water that overtopped the bund in each experiment was found to vary with the initial fill height, the profile of the bund wall and its distance from the vessel. The volume of water that overtopped the bund in each test is given in Tables 10 and 11. In order to compare the information from different experiments in graphical form, the volume of water that overtops has been normalised by reference to the bund capacity for a 90 degrees bund wall (3.9 m3). Figure 21 shows the results for a number of the Phase 1 experiments in which the tank was filled to a height of approximately 1.45m. The volume of water released in these experiments corresponds to approximately 90% of the capacity for a 90 degrees bund wall. In all cases, even allowing for the reduced capacity as a result of the sloping bund walls, the volume was less than the capacity of the bund. The figure shows that the amount of water overtopping the bund varies with the wall angle and distance of the bund from the vessel, with most overtopping occurring for a 30 degrees angle bund wall situated at 5m.
In the experiments carried out in configuration SQ4, with the obstacles present, it was observed that water overtopped the cylindrical sheeting that formed the obstacles and flowed into the obstacles. The depth of water retained in the cylindrical obstacles could not be measured and was less than 1 mm, corresponding to a loss of less than 19 litres. Hence, the amount of overtopping recorded in the later interaction with the bund walls could have been underestimated by up to this amount. This is noted subsequently on the relevant Tables of data and Figures.
For the experiments carried out with initial water heads of approximately 1.75mm (nominally 110%), the volume of water that was released was greater than the capacity of the bund. For these experiments, two values were determined:
• Initial: This is the amount of water collected from the initial surge of water, the release being shut-off before the bund was full.
• Total: This is the total amount of water collected after the release vessel was allowed to empty without interruption.
Figure 22 shows the percentage of water overtopping the bund for the different square bund arrangements and the different nominal fill heights. The figure shows that there is more water overtopping the bund when the bund is further away from
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 8
the vessel. Also the presence of obstacles within the bunded area has little effect on the volume of water overtopping, although it should be noted that up to an additional 19 litres, or 5% of the bund capacity, could have been lost through water overtopping the obstacles in bund arrangement SQ4.
Plotting the percentage of water overtopping the bund for the circular and square bunds with the same bunded area (5.0m circular bund, SQ1 and SQ2 as one group and 10.0m circular bund and SQ3 and SQ4 as another) gives an indication of the effect of the shape of the bund. Figure 23 shows the percentage of water overtopping the bund for arrangements SQ1, SQ2 and a 5.0m circular bund, with a 90 degrees bund wall, for the different nominal fill heights. The figure shows that the percentage of water overtopping the bund is approximately the same for the different bund shapes.
Figure 24 shows the percentage of water overtopping the bund for bund arrangements SQ3, SQ4 and a 10.0m circular bund, with a 90 degrees bund wall, for the different nominal fill heights. This figure shows that for fill heights of 90% and 100% less water overtopped the circular bund than the square bunds.
4.4 Images of the Water Front Figures 25 to 27 show still images of the water front recorded at 5.0m, 7.1m and 10.0m radii. These images were downloaded from the high-speed video system that was operated at 125f.p.s with the shutter set at 1/1000th s.
5 DISCUSSION
5.1 Phase 1 Phase 1 has provided data on the spreading of water released from the base of a tank, that at a scale of 1:20, represents a 70m diameter storage tank. The release and the surrounding geometrical arrangement are of an idealized nature. This creates a flow that can be simulated in a straightforward manner by mathematical models for liquid spread over land. Hence, the results from the experiments provide an ideal dataset for the development or validation of mathematical models.
The interaction of the spreading front with a number of different bunds has been examined and the amount of water, if any, overtopping the bunds has been determined. A number of observations can be made on the data, as follows.
Firstly, as can be seen from Figure 15, the outflow from the vessel is in close agreement with expectations from theory. Also, the initial velocity of the water flowing out of the vessel is consistent as a starting value for the inferred spreading velocity data, such as that shown in Figure 17.
Within the programme, a number of experiments were carried out with the same initial head in the tank. A comparison of the spreading behaviour in these experiments gives some idea of the repeatability of the observed behaviour. It is found that there is some radial variation in the time of arrival at a given distance from the tank (of the order of 10 to 20%) in any one experiment. However, the results from
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 9
all of the experiments, taken before interaction with the bund wall, show similar behaviour. Even taking into account the observed slower rate of progress along the line of the probes, if the time of arrival is plotted against distance from the tank centre, all of the measurements lie in a band of relatively narrow width, indicating a good degree of reproducibility in flow behaviour. Figures 17 and 18 illustrate this for the results obtained with one particular initial fill level in the vessel. Results such as those shown in Figures 17 and 18 provide the data that can be used to help investigate a number of outstanding issues, such as those concerning the boundary conditions to be used at the front of the spreading pool on land.
The measurements of the amount of water overtopping the bund have also been examined. As Figure 21 demonstrates, if a large release occurs from around the base of the tank, the water is capable of overtopping the bund walls, even if the bunded area has the capacity to hold all of the water that is released. Based on these results, it appears that generally the amount that overtops the bund decreases as the distance of the bund wall from the vessel increases (note: the bund heights were chosen to ensure that the bunded area had nominally the same capacity in each case). The amount also appears to be sensitive to the bund wall angle in this case. Even allowing for the differences in capacity of the bunds for the different slopes of the bund wall, it appears that a vertical wall is more effective than a sloping bund wall at retaining the water within the bund.
5.2 Phase 2 The experiments in Phase 2 have provided data on the interaction of the spreading front on square bunds and allow a comparison of different geometry square bunds and circular bunds. Experiments were also conducted to study the effect of additional tanks within the bunded area.
As expected, for the experiments without additional obstacles, the spreading of the liquid front prior to interaction with the bund wall was the same for square and circular bunds allowing for a radial variation of 10 to 20%, see Section 5.1.
The presence of additional tanks within the bunded area interrupts the spreading of the liquid front towards the bund wall. The results show that as the initial fill height increases the liquid front reaches the region between the additional tank and the bund wall at an earlier time, see Figure 19. The presence of the additional tanks appears to channel the water between the tanks at a greater velocity as shown in Figure 20 where the time of arrival of the liquid front occurs earlier at certain parts of the bund wall than in the immediate wake of the additional tanks.
The measurements of the amount of water overtopping the bund show that with additional tanks in the bunded area the amount of water overtopping the bund is similar to when there are no additional tanks present (although there is an uncertainty of up to approximately 5% of the bund capacity introduced because some of the flow overtopped the model tanks themselves in the experiments in which the obstacles were present). It should be noted that although the bunded area is the same in the two experimental set-ups, when the obstacles are present, the bund wall is further away from the release vessel. This is to allow for the area of the additional tanks within the bund and to maintain the same area available for liquid spreading.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 10
Phase 1 experiments have shown that the amount of water overtopping the bund reduces as the distance from the release vessel increases, so this may explain part of this result.
The measurements show that for a bund area nominally equivalent to a circle of radius 5m, the amount of water overtopping a bund is similar, independent of the bund geometry (Figure 23). However, it appears (Figure 24) that for a bund area equivalent to a circle of radius 10m, less water overtopped a circular bund compared to a square bund.
6 SUMMARY A series of 58 experiments has been carried out to study the flow of water released from a slit at the base of a cylindrical tank into a bund. During these experiments the water pressure (liquid head) in the vessel, the initial flow of water across the bund floor and the quantity of water which overtopped the bund were measured. The data collected have been summarised in this report and full copies of the data are provided with the report in electronic format.
7 REFERENCES 1. Webber, D.M., A Model for Pool Spreading and Vaporisation and its
Implementation in the Computer Code G*A*S*P”, Report SRD/HSE/R507, September 1990.
2. Linden, P., Daish, N., Dalziel, S., Halford, A., Jackson, M., Hirst, I., Perroux, J., Wiersma, S. 'LSMS: A New Model For Spills Of LNG And Other Hazardous Liquids', Proceedings of 1998 International Gas Research Conference, San Diego, November 8th - 11th.
3. Greenspan, H.P. and Johansson, A.V. An experimental study of flow over an impounding dyke. Studies in Applied Mathematics, 64, 211-233, 1981.
4. Moorhouse, J and Carpenter, R.J. Factors affecting vapour evolution rates from liquefied gas spills. I Chem E North Western Branch, Hazards Symposium, 1986.
5. Sharifi, T. An experimental study of the catastrophic failure of storage tanks. Ph D Thesis. University of London, Imperial College of Science and Technology, Dept. of Chem Eng. And Chem. Technology, 1987.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 11
Table 1: Coordinates for Resistance Probes Using a Polar Coordinate System, see Figures 12 and 13
Coordinates Instrument Number θ (θ (θ (θ (οοοο)))) R (m)*
RP01 45 1.88 RP02 45 2.16 RP03 45 2.44 RP04 45 2.72 RP05 45 3.00 RP06 45 3.28 RP07 45 3.56 RP08 45 3.84 RP09 45 4.12 RP10 45 4.40 RP11 45 4.68 RP12 45 4.96 RP13 45 5.24 RP14 45 5.52 RP15 45 5.80 (13.06) RP16 45 6.08 (13.32) RP17 45 6.36 (10.20) RP18 45 6.64 (10.46) RP19 45 6.92 (10.72) RP20 45 7.20 (10.98) RP21 45 7.48 (11.24) RP22 45 7.76 (11.50) RP23 45 8.04 (11.76) RP24 45 8.32 (12.02) RP25 45 8.60 (12.28) RP26 45 8.88 (12.54) RP27 45 9.16 (12.80) RP28 45 9.44 RP29 45 9.72 RP30 45 9.93 RP31 22.5 1.88 RP32 22.5 2.50 RP33 22.5 3.12 RP34 22.5 3.75 RP35 22.5 4.38 RP36 22.5 5.00 RP37 22.5 5.53 RP38 22.5 6.05 RP39 22.5 6.57 RP40 22.5 7.10 RP41 22.5 7.83
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 12
RP42 22.5 8.55 RP43 22.5 9.27 RP44 22.5 10.00 RP45 11 5.00 RP46 34 5.00 RP47 56 5.00 RP48 67.5 5.00 RP49 79 5.00 RP50 7 7.10 RP51 15 7.10 RP52 30 7.10 RP53 37 7.10 RP54 52 7.10 RP55 60 7.10 RP56 67.5 7.10 RP57 75 7.10 RP58 83 7.10 RP59 11 10.00 RP60 34 10.00
* Radii in brackets are for tests in bund configuration SQ4
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 13
Table 2: Test Conditions for the Experiments Conducted in the Circular Bunds
Test Bund
Height (mm)
Bund Radius
(m)
Bund Angle
(o)
Nominal Fill (%)
Actual Fill Height
(m) Resistance Probe Data
H1 50 10.0 90 90 1.449 Yes H2 50 10.0 90 100 1.588 Yes G2 50 10.0 90 110 No data Yes
G2_1 50 10.0 90 110 1.808 No H3 50 10.0 45 90 1.428 No H4 50 10.0 45 100 1.579 No G3 50 10.0 45 110 1.797 Yes
G3_1 50 10.0 45 110 1.802 Yes H5 50 10.0 30 90 1.438 No H6 50 10.0 30 100 1.586 No G4 50 10.0 30 110 1.804 Yes
G4_1 50 10.0 30 110 1.806 No
H7 100 7.1 90 90 1.428 No H8 100 7.1 90 100 1.577 No G5 100 7.1 90 110 1.817 Yes
G5_1 100 7.1 90 110 1.813 No G5_2 100 7.1 90 110 1.814 No
H9 100 7.1 45 90 1.431 No H10 100 7.1 45 100 1.593 No G6 100 7.1 45 110 1.811 Yes
G6_1 100 7.1 45 110 1.806 No H11 100 7.1 30 90 1.437 No H12 100 7.1 30 100 1.592 No G7 100 7.1 30 110 1.808 Yes
G7_1 100 7.1 30 110 1.806 No
H13 200 5.0 90 90 1.442 No H14 200 5.0 90 100 1.593 No G8 200 5.0 90 110 1.758 Yes
G8_1 200 5.0 90 110 1.808 No H15 200 5.0 45 90 1.439 No H16 200 5.0 45 100 1.585 No G9 200 5.0 45 110 1.751 Yes
G9_1 200 5.0 45 110 1.788 Yes H17 200 5.0 30 90 1.441 No H18 200 5.0 30 100 1.588 No G10 200 5.0 30 110 1.819 Yes
G10_1 200 5.0 30 110 1.753 No
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 14
Table 3: Test Conditions for the Experiments Conducted in the Square Bunds
Test Bund
Height (mm)
Bund Configuration
Bund Angle
(o)
Nominal Fill (%)
Actual Fill Height
(m) Resistance Probe Data
J1 200 SQ1 90 90 1.430 Yes J2 200 SQ1 90 100 1.585 Yes J3 200 SQ1 90 110 1.733 Yes
J3_1 200 SQ1 90 110 1.732 No
J4 200 SQ2 90 90 1.437 Yes J5 200 SQ2 90 100 1.586 Yes J6 200 SQ2 90 110 1.730 No
J6_1 200 SQ2 90 110 1.734 Yes
J7 50 SQ3 90 90 1.439 Yes J8 50 SQ3 90 100 1.588 Yes J9 50 SQ3 90 110 1.731 Yes
J9_1 50 SQ3 90 110 1.730 No
J10 50 SQ4 90 90 1.436 Yes J10_1 50 SQ4 90 90 1.447 Yes J10_2 50 SQ4 90 90 1.438 Yes
J11 50 SQ4 90 100 1.593 Yes J11_1 50 SQ4 90 100 1.587 Yes J11_2 50 SQ4 90 100 1.638 Yes
J12 50 SQ4 90 110 1.716 No J12_1 50 SQ4 90 110 1.740 Yes J12_2 50 SQ4 90 110 1.738 No
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 15
Table 4: Release Vessel Water Inventory
Fill Height/Initial Head Water Inventory
1.45m 3.50m3
1.60m 3.86m3
1.75m 4.22m3
Table 5a: Volume of the Circular Bunds in Phase 1
Volume of Bund (m3) Nominal Radius (m) 30o bund angle 45o bund angle 90o bund angle
5.0 3.66 3.77 3.93
7.1 3.86 3.91 3.96
10.0 3.89 3.93 3.93
Table 5b: Volume of the Square Bunds in Phase 2
Bund Configuration Volume of Bund (m3)
SQ1 3.88
SQ2 3.88
SQ3 3.90
SQ4 3.88
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 16
Table 6: Time of Water Arrival for Experiments with a 10.0m Circular Bund Probe Test H1
(ms) Test H2
(ms) Test G2
(ms) Test G3
(ms) Test G4
(ms) RP01 0 0 10.8 3.1 - RP02 69.7 62.5 108.9 96 79.4 RP03 180.8 138.7 190.6 160.6 146.1 RP04 248.9 257.7 242.2 221.2 216.7 RP05 256.9 286.9 283.6 288.9 282.4 RP06 404.7 373.3 340.4 343.4 331.6 RP07 541.3 480.8 406.9 413.1 412.9 RP08 643.7 592.8 459.5 536.7 523.8 RP09 728.3 696.6 596.4 670.1 597.9 RP10 829.8 780.3 627.4 735.6 725.4 RP11 931.4 868.9 812.5 803.4 786.3 RP12 1033.1 997.1 887.6 867.9 898.4 RP13 1159.8 1126.6 972 979.6 971.8 RP14 1284.7 1218.7 1076.4 1056.7 1056 RP15 1422.3 1339.4 1183.3 1181.6 1150.5 RP16 1559 1475.9 1270.1 1276.8 1263 RP17 1704 1606.2 1337.3 1345.4 1339.1 RP18 1863.5 1778.7 1429.8 1430.2 1431.8 RP19 2030.5 1923.9 1548.5 1542.2 1552.9 RP20 2428.3 2271.5 1824.7 1884.9 1832.7 RP21 2652.8 2483.5 1998.6 2064.1 2032.7 RP22 2874.4 2653.6 2180.2 2223.9 2200.3 RP23 3081.7 2842.4 2360.7 2376.3 2367.4 RP24 3265.5 3091.4 2550.8 2579.4 2543.5 RP25 3474.2 3339.4 - - 2476.2 RP26 3675.5 3491.5 2960.3 3004.4 2939.7 RP27 4070.1 3816.2 3157.1 3183.1 3142 RP28 4384.3 4079 3385.7 3400.2 3350.4 RP29 5050.7 4439 3580.1 3574.2 3558.8 RP30 5029.6 4641.1 3734 3754.7 3721.9 RP31 23.4 7.7 0 0 0 RP32 144.3 145.5 174.9 164 156.9 RP33 309.5 306.4 284.5 273.9 275.9 RP34 495.7 496.1 417.1 435.6 416.6 RP35 725.2 780.7 615.4 609.6 619.2 RP36 909.2 850.2 792.2 796.2 819.8 RP37 1101.2 1040.8 1003.5 990 968.7 RP38 1351.5 1276.4 1152.1 1150 1127.7 RP39 1565.2 1509.7 1325.4 1331.9 1321.5 RP40 1934.9 1853.1 1526.1 1552.2 1532.2 RP41 2720.4 2519.6 2090.5 2114 2053.9 RP42 3325.1 3092.2 2477.7 2488.2 2425.5 RP43 4387.9 3670.2 2940.9 2938.1 2900.5 RP44 4670 4336.3 3217.2 3226.7 3167.1 RP45 908 849.1 731.9 759.1 778.6 RP46 951.6 875.8 873.8 857.9 863 RP47 977.3 917.9 870.9 843.4 845.3 RP48 951.3 862.8 875.2 889.7 840.5 RP49 866.2 875.5 686.9 784.5 730.2 RP50 2019.2 1859.8 1512.5 1499.8 1499.4 RP51 2038.7 1889.3 1546.2 1525.7 1507.3 RP52 1922.4 1883.5 1584.4 1591.5 1535 RP53 2029.1 1915.8 1580.6 1578 1568.8 RP54 1884.1 1780.3 1586.5 1571.4 1531.8 RP55 2040.6 1864.6 1632.1 1621.3 1641.1 RP56 2073.1 1909.6 1651 1667.6 1621.4 RP57 1887.2 1714.3 1590.7 1587.2 1559.6 RP58 1836.2 1664.3 1542.3 1547.4 1512 RP59 4983.7 4509.6 3378.4 3367.9 3380.5 RP60 4895.1 4222.5 3350.6 3272.9 3326.4
‘-‘ means no data were collected by that resistance probe.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 17
Table 7 Time of Water Arrival for Experiments with a 7.1m Circular Bund Probe Test G5
(ms) Test G6
(ms) Test G7
(ms) RP01 3.8 0 0 RP02 87.8 107.7 87.7 RP03 180.5 197 159.4 RP04 234.6 261.1 214.4 RP05 286.6 310.5 270.9 RP06 351.4 395.8 353.7 RP07 460.7 468.1 451.1 RP08 490.2 613.3 539.8 RP09 667.3 707.4 658.7 RP10 724 774.1 729.5 RP11 788.8 839.3 817.8 RP12 822.1 933.8 851 RP13 961.1 1023.3 988.1 RP14 - 1112.7 1069 RP15 1177.5 1213.2 1162.9 RP16 1260.4 1279.7 1236.1 RP17 1354.8 1353.8 1327.1 RP18 1451.4 1451.1 1440.9 RP19 1566.9 1566.1 1542.3 RP20 - - - RP21 - - - RP22 - - - RP23 - - - RP24 - - - RP25 - - - RP26 - - - RP27 - - - RP28 - - - RP29 - - - RP30 - - - RP31 0 42 0 RP32 163.1 203.4 160.7 RP33 267.1 312.8 268.6 RP34 418.6 462.4 430.3 RP35 630.3 649.7 637.2 RP36 844.8 874.7 808.8 RP37 994.1 1045.2 994.7 RP38 1157.5 1209.1 1143.1 RP39 1319.7 1366.1 1329 RP40 1515.1 1579 1513.1 RP41 - - - RP42 - - - RP43 - - - RP44 - - - RP45 776.9 670.2 778.8 RP46 855.3 903.2 850.9 RP47 859 895.9 860.3 RP48 834.1 890.8 869.7 RP49 692.1 806.8 652.3 RP50 1501.8 1538.3 1495.1 RP51 1530.9 1539.1 1504.3 RP52 1561.1 1613.6 1571.1 RP53 1577.7 1605.2 1591.4 RP54 1565.7 1566.2 1532.4 RP55 1599.2 - 1597.2 RP56 1637.2 - 1613.7 RP57 1521.8 1616.6 1490.7 RP58 1494 1568.9 1498.4 RP59 - - - RP60 - - -
‘-‘ means no data were collected by that resistance probe.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 18
Table 8: Time of Water Arrival for Experiments with a 5.0m Circular Bund
Probe Test G8 (ms)
Test G9 (ms)
Test G9_1 (ms)
Test G10 (ms)
RP01 6.5 0 0 0 RP02 84.6 60.1 74.3 79.7 RP03 162 149.2 154.1 193.1 RP04 223.3 233.8 211.1 265.3 RP05 269.6 256.6 283.2 309 RP06 332.6 323.7 347.4 362 RP07 415.1 419.7 462.6 481.9 RP08 486.3 527.4 494.6 600.1 RP09 518.2 669.8 561.9 699.6 RP10 584.8 731.5 738.5 773.5 RP11 799.2 790.3 801.6 824.3 RP12 920.9 954.1 935.3 929.8 RP13 - - - - RP14 - - - - RP15 - - - - RP16 - - - - RP17 - - - - RP18 - - - - RP19 - - - - RP20 - - - - RP21 - - - - RP22 - - - - RP23 - - - - RP24 - - - - RP25 - - - - RP26 - - - - RP27 - - - - RP28 - - - - RP29 - - - - RP30 - - - - RP31 0 0 0 24.6 RP32 169 150.2 158.9 191 RP33 275.7 257.7 270.8 304.4 RP34 427.5 415.8 416.5 437.5 RP35 606.8 615.2 549.4 RP36 822.2 806 807.8 839.6 RP37 - - - - RP38 - - - - RP39 - - - - RP40 - - - - RP41 - - - - RP42 - - - - RP43 - - - - RP44 - - - - RP45 783.7 772.1 746.9 795.8 RP46 - 840.4 843.3 868.2 RP47 845.4 851.5 829.6 868.8 RP48 846.6 844.3 851.4 873.8 RP49 698.4 735.9 - - RP50 - - - - RP51 - - - - RP52 - - - - RP53 - - - - RP54 - - - - RP55 - - - - RP56 - - - - RP57 - - - - RP58 - - - - RP59 - - - - RP60 - - - -
‘-‘ means no data were collected by that resistance probe.
R
epor
t Num
ber:
R 4
418
Issu
e:01
C
OM
MER
CIA
L IN
CO
NFI
DEN
CE
Pa
ge 1
9 Ta
ble
9: T
ime
of W
ater
Arr
ival
for E
xper
imen
ts w
ith S
quar
e B
unds
Test
s Pr
obe
J1
(ms)
J2
(m
s)
J3
(ms)
J4
(m
s)
J5
(ms)
J6
_1
(ms)
J7
(m
s)
J8
(ms)
J9
(m
s)
J10
(ms)
J1
0_1
(ms)
J1
0_2
(ms)
J1
1 (m
s)
J11_
1 (m
s)
J11_
2 (m
s)
J12_
1 (m
s)
RP0
1 0
- 0
0 0
0 0
0 0
0 -
0 0
0 0
0 R
P02
59
118.
9 72
.7
70
87.6
90
.9
74.7
95
97
.2
69.1
90
10
2 80
.6
88
96
61
RP0
3 15
2.5
189.
6 13
4.7
137.
9 16
9.4
235.
2 17
5.4
215.
9 20
8.3
170.
9 18
4 17
1 17
2.3
178
208
163
RP0
4 18
8.1
240.
9 20
7.5
188.
3 19
4.2
220.
4 32
0.7
237.
1 26
9.2
293.
5 28
0 24
1 28
3.1
258
262
221
RP0
5 27
3.4
327.
6 28
6.6
326.
6 31
1.3
304
364.
5 31
4.2
323.
9 33
2.5
333
397
312.
3 30
7 31
8 28
4 R
P06
397.
5 38
5.3
369.
2 39
8.3
377.
8 35
9.8
470.
6 38
6.8
384
403.
1 40
4 45
9 36
6.1
355
443
336
RP0
7 55
3.3
464.
8 46
5.6
541.
1 48
2.8
445.
2 60
2.4
558.
6 47
9.7
561
503
572
520.
9 46
6 59
8 44
8 R
P08
626.
2 55
6.5
643
646.
9 62
2.7
473.
9 68
6.5
663.
2 59
7.6
653.
2 55
7 69
7 63
1.9
605
662
555
RP0
9 72
3.1
704.
1 70
9.6
730.
1 70
5.6
577.
1 78
0.7
723.
4 70
0.1
736.
7 73
8 78
2 69
8.1
714
744
651
RP1
0 -
- -
844.
5 80
4.3
778.
3 91
5.7
814.
6 78
5.3
830.
5 83
7 88
0 78
3.7
791
865
778
RP1
1 -
- -
1091
.4
899.
1 85
0.3
1009
.2
934.
6 90
8.3
944.
8 96
0 99
7 91
3.4
924
955
845
RP1
2 -
- -
1032
.1
1020
.4
963.
9 10
67.9
99
2 99
3.1
1031
.5
1068
11
30
1067
.2
985
1022
92
9 R
P13
- -
- 11
74.7
11
15.3
11
02.1
12
31.4
11
54.5
11
32.4
11
82.1
12
05
1233
11
83.7
11
34
1174
10
92
RP1
4 -
- -
1308
.2
1230
.2
1112
.5
1364
.5
1282
.1
1288
.7
1284
.7
13`3
3 13
43
1296
.7
1269
12
76
1174
R
P15
- -
- 14
36.1
13
31.8
13
23.7
14
35.9
13
75.5
13
80.1
-
- 10
567
- -
1005
8 97
73
RP1
6 -
- -
1553
.6
1445
.7
1429
.9
1573
.6
1534
14
89.3
-
- 10
820
- -
1034
6 10
029
RP1
7 -
- -
- -
- 17
44.3
16
59.2
15
81.2
-
- 11
774
- -
1135
5 10
967
RP1
8 -
- -
- -
- 19
12.8
18
09.8
17
83.5
-
- 12
725
- -
1137
5 11
125
RP1
9 -
- -
- -
- 20
66.4
20
08.7
19
49
- -
1132
7 -
- 10
830
1075
8 R
P20
- -
- -
- -
2503
.1
2338
.3
2273
.7
- -
1104
3 -
- 10
659
1046
2 R
P21
- -
- -
- -
2746
.2
2591
.9
2500
.4
- -
1063
1 -
- 10
001
1060
7 R
P22
- -
- -
- -
2958
.1
2818
.8
2700
.2
- -
1038
2 -
- 96
80
9552
R
P23
- -
- -
- -
3175
.5
3013
.4
2895
.5
- -
1028
8 -
- 97
56
9437
R
P24
- -
- -
- -
3377
.6
3180
.8
3082
.3
- -
1046
9 -
- 99
13
- R
P25
- -
- -
- -
3603
.3
3403
.3
3274
.9
- -
1043
0 -
- 99
06
9679
R
P26
- -
- -
- -
3838
.4
3632
.9
3508
-
- 10
276
- -
9834
96
18
RP2
7 -
- -
- -
- 40
41.6
39
13.2
37
30.9
-
- 10
472
- -
1003
6 95
16
RP2
8 -
- -
- -
- 40
41.6
42
08.5
40
31
- -
1340
8 -
- 13
467
1289
1 R
P29
- -
- -
- -
- 45
65.1
43
34.2
-
- 12
287
- -
1218
9 11
784
RP3
0 -
- -
- -
- -
4848
.8
4564
.1
- -
1195
5 -
- 11
427
1144
1 R
P31
- -
- 22
.1
19.7
13
12
.4
31.4
38
.3
13.9
37
55
22
.1
39
57
- R
P32
- -
- 15
5.1
152.
1 15
4.7
178.
8 17
2.2
162.
3 14
6.3
168
178
149
168
193
- R
P33
- -
- 32
6 31
3.5
278.
6 33
5.9
335.
8 30
8.7
319.
1 32
0 34
8 30
3.9
301
317
- R
P34
- -
- 53
4 49
6.3
462.
4 52
4.2
490.
2 45
5.1
483.
7 49
5 51
6 47
9.3
474
514
- R
P35
- -
- 71
7.1
663.
3 59
8.9
- 71
6.5
651.
5 69
6.9
740
744
650.
5 65
2 75
5 -
RP3
6 -
- -
- -
- 91
2.1
894.
7 83
2.1
887.
5 90
7 95
1 88
5.3
891
939
- R
P37
- -
- -
- -
1100
.9
1055
10
29.4
10
72
1115
11
74
1081
.5
1048
11
33
-
R
epor
t Num
ber:
R 4
418
Issu
e:01
C
OM
MER
CIA
L IN
CO
NFI
DEN
CE
Pa
ge 2
0
RP3
8 -
- -
- -
- 13
39.4
18
99
1225
.2
1335
.4
1365
15
87
1312
.9
1339
13
88
- R
P39
- -
- -
- -
1640
.3
1585
.8
1479
.4
1629
.4
1643
17
33
1551
.3
1499
16
90
- R
P40
- -
- -
- -
- 19
56.8
17
62.7
19
50.5
19
67
2088
18
59.8
17
98
1923
-
RP4
1 -
- -
- -
- 28
28.7
27
89.9
26
11.3
28
01.4
28
12
2906
26
51.8
25
84
2799
23
97
RP4
2 -
- -
- -
- 33
19.9
32
14.2
30
44
2998
.8
2997
31
08
2862
.2
2836
30
47
2686
R
P43
- -
- -
- -
- 37
91
3570
.2
3329
.2
3391
34
84
3216
.6
3150
33
91
3015
R
P44
- -
- -
- -
4731
.9
4571
47
39.1
37
48.5
37
77
3883
35
99.3
35
08
3756
33
19
RP4
5 -
- -
665.
1 70
0.7
635.
1 -
884.
6 84
6.2
2282
90
03
1369
7 -
9124
14
261
1024
2 R
P46
- -
- 79
2 75
9.3
755.
6 10
42.9
95
4.5
932.
9 10
08.9
99
0 10
95
930.
9 91
1 10
26
852
RP4
7 -
- -
808.
9 63
8.8
735.
8 10
03.2
94
8.7
906.
6 97
2.7
994
982
914.
6 90
7 10
02
844
RP4
8 -
- -
743.
3 73
7.8
720.
1 98
6.4
955
870.
9 94
3.9
981
958
896.
9 89
2 96
2 83
1 R
P49
- -
- 69
2.6
614.
7 56
5.5
916
863.
2 85
9.5
2282
.1
9003
13
697
- 91
24
1426
0 10
277
RP5
0 -
- -
- -
- 20
42.5
20
05.3
18
38.5
12
4 63
66
6709
-
6916
72
42
6809
R
P51
- -
- -
- -
- 19
75.4
19
27.5
45
58.7
46
80
4780
47
87.7
45
67
4562
44
82
RP5
2 -
- -
- -
- 20
59.5
20
11.9
18
93
1996
.5
2008
21
54
1917
19
25
2022
18
07
RP5
3 -
- -
- -
- 20
57.8
20
00.6
19
09.3
22
84.3
90
05
1370
0 -
9127
14
264
1031
0 R
P54
- -
- -
- -
1945
18
73.6
17
77.9
-
9005
13
700
- 91
27
1426
4 10
310
RP5
5 -
- -
- -
- 20
67.5
19
54.6
18
78.5
19
02.5
19
14
1930
18
33.8
18
35
1922
17
56
RP5
6 -
- -
- -
- 20
30.3
19
41
1893
.6
1884
.8
2060
19
86
1777
.9
1817
18
64
1681
R
P57
- -
- -
- -
1885
.2
1781
.1
1713
.2
5150
.3
4909
51
37
2955
.6
3104
37
84
2705
R
P58
- -
- -
- -
1878
.5
1756
.6
1688
.9
322.
1 65
67
7689
-
6829
74
22
7389
R
P59
- -
- -
- -
- 35
88.8
-
- 55
16
5758
-
5138
54
69
5126
R
P60
- -
- -
- -
- 44
39.2
42
38.5
93
1.4
8681
80
35
- 65
90
7866
84
26
‘-‘ m
eans
no
data
wer
e co
llect
ed b
y th
at re
sist
ance
pro
be.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 21
Table 10: Results of the Volume of Water Overtopping the Bund in the Experiments Conducted in the Circular Bunds
* Nominal inventory quoted, as no fill height data available for this test
Test Bund
Height (mm)
Bund Radius
(m)
Bund Angle
(o)
Tank Inventory
(cu.m) (from Tables
2 and 4)
Bund Volume (cu.m) (from
Table 5a)
Water Overtopping
(litres)
Initial Wave or
Total Amount
H1 50 10.0 90 3.50 3.93 None Total H2 50 10.0 90 3.83 3.93 55 Total G2 50 10.0 90 4.22* 3.93 443 Total
G2_1 50 10.0 90 4.36 3.93 476 Initial H3 50 10.0 45 3.45 3.93 None Total H4 50 10.0 45 3.81 3.93 39 Total G3 50 10.0 45 4.33 3.93 323 Total
G3_1 50 10.0 45 4.34 3.93 323 Initial H5 50 10.0 30 3.47 3.89 7 Total H6 50 10.0 30 3.83 3.89 67 Total G4 50 10.0 30 4.35 3.89 468 Total
G4_1 50 10.0 30 4.35 3.89 419 Initial
H7 100 7.1 90 3.45 3.96 37 Total H8 100 7.1 90 3.80 3.96 160 Total G5 100 7.1 90 4.38 3.96 545 Total
G5_1 100 7.1 90 4.37 3.96 654 Total G5_2 100 7.1 90 4.37 3.96 272 Initial
H9 100 7.1 45 3.45 3.91 99 Total H10 100 7.1 45 3.84 3.91 215 Total G6 100 7.1 45 4.37 3.91 252 Total
G6_1 100 7.1 45 4.35 3.91 218 Initial H11 100 7.1 30 3.47 3.86 85 Total H12 100 7.1 30 3.84 3.86 160 Total G7 100 7.1 30 4.36 3.86 590 Total
G7_1 100 7.1 30 4.35 3.86 511 Initial
H13 200 5.0 90 3.48 3.93 22 Total H14 200 5.0 90 3.84 3.93 40 Total G8 200 5.0 90 4.24 3.93 194 Total
G8_1 200 5.0 90 4.36 3.93 76 Initial H15 200 5.0 45 3.47 3.77 177 Total H16 200 5.0 45 3.82 3.77 344 Total G9 200 5.0 45 4.22 3.77 599 Total
G9_1 200 5.0 45 4.31 3.77 520 Initial H17 200 5.0 30 3.48 3.66 325 Total H18 200 5.0 30 3.83 3.66 502 Total G10 200 5.0 30 4.39 3.66 984 Total
G10_1 200 5.0 30 4.23 3.66 827 Initial
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 22
Table 11: Results of the Volume of Water Overtopping the Bund in the Experiments Conducted in the Square Bund Configurations
Test Bund
Configuration
Bund Height (mm)
Bund Angle
(o)
Tank Inventory
(cu.m) (from Tables
2 and 4)
Bund Volume (cu.m)
(from Table 5b)
Water Overtopping
(litres)
Initial Wave or Total Amount
J1 SQ1 200 90 3.45 3.88 19 Total J2 SQ1 200 90 3.82 3.88 55 Total J3 SQ1 200 90 4.18 3.88 170 Total
J3_1 SQ1 200 90 4.18 3.88 21 Initial
J4 SQ2 200 90 3.47 3.88 25 Total J5 SQ2 200 90 3.83 3.88 59 Total J6 SQ2 200 90 4.17 3.88 38 Initial
J6_1 SQ2 200 90 4.18 3.88 249 Total
J7 SQ3 50 90 3.47 3.90 175 Total J8 SQ3 50 90 3.83 3.90 286 Total J9 SQ3 50 90 4.17 3.90 235 Initial
J9_1 SQ3 50 90 4.17 3.90 485 Total
J10 SQ4* 50 90 3.47 3.88 160 Total J10_1 SQ4* 50 90 3.49 3.88 120 Total J10_2 SQ4* 50 90 3.47 3.88 125 Total J11 SQ4* 50 90 3.84 3.88 225 Total
J11_1 SQ4* 50 90 3.83 3.88 294 Total J11_2 SQ4* 50 90 3.95 3.88 292 Total J12 SQ4* 50 90 4.14 3.88 443 Total
J12_1 SQ4* 50 90 4.20 3.88 445 Total J12_2 SQ4* 50 90 4.19 3.88 63 Initial * In experiments carried out in configuration SQ4, up to approximately 19 litres of water may have overtopped and flowed into the model obstacles, and hence have not been available to overtop the constraining bund walls.
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 23
Figure 1: Photograph of the Release Vessel
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 24
Figure 2: Schematic Diagram of the Release Vessel
Flap valve opening mechanism
Water tank
Water tank
2.6m
1.1m
1.8m
Steel sheet
Elevation
Plan
Concrete pad
Concrete pad
1.75m
Report Number: R 4418 Issue:01
COMMERCIAL IN CONFIDENCE Page 25
Figure 3: Photograph of the Release Mechanism
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
26
Fi
gure
4: B
und
Arra
ngem
ent f
or a
Circ
ular
Bun
d w
ith a
Dia
met
er o
f 5.0
m
5.0m
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
3.5
Dia
200m
m
200m
m
45 d
egre
es
200m
m
30 d
egre
es
Bund
Wal
l Geo
met
ries
Bund
wal
l
Plan
of B
und
Layo
ut
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
27
Fi
gure
5: B
und
Arra
ngem
ent f
or a
Circ
ular
Bun
d w
ith a
Dia
met
er o
f 7.1
m
7.1m
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
3.5
Dia
100m
m
100m
m
45 d
egre
es
100m
m
30 d
egre
es
Bund
Wal
l Geo
met
ries
Bund
wal
l
Plan
of B
und
Layo
ut
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
28
Fi
gure
6: B
und
Arra
ngem
ent f
or a
Circ
ular
Bun
d w
ith a
Dia
met
er o
f 10.
0m
10.0
mR
elea
se
Ves
sel
Wal
l of
sym
met
ry
3.5m
Dia
50m
m
45 d
egre
es
50m
m
30 d
egre
es
Bun
d W
all G
eom
etrie
s
Bund
wal
l
Plan
of B
und
Layo
ut
50m
m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
29
Fi
gure
7: B
und
Arra
ngem
ent f
or S
quar
e B
und
Arra
ngem
ent S
Q1
6.27
m
200m
m v
ertic
albu
nd w
all
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
30
Fi
gure
8: B
und
Arra
ngem
ent f
or S
quar
e B
und
Arra
ngem
ent S
Q2
4.43
m
200m
m v
ertic
albu
nd w
all
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
31
Fi
gure
9: B
und
Arra
ngem
ent f
or S
quar
e B
und
Arra
ngem
ent S
Q3
8.89
m
50m
m v
ertic
albu
nd w
all
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
32
Fi
gure
10:
Bun
d Ar
rang
emen
t for
Squ
are
Bun
d Ar
rang
emen
t SQ
4
3.5m
9.89
m
5.3m
50m
m v
ertic
albu
nd w
all
Rel
ease
Ve
ssel
Wal
l of
sym
met
ry
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
33
Fi
gure
11:
Pho
togr
aph
of th
e Te
st R
ig fo
r Bun
d Ar
rang
emen
t SQ
4
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
34
Fi
gure
12:
Loc
atio
n of
Res
ista
nce
Prob
es fo
r All
the
Bun
d Ar
rang
emen
ts W
ithou
t Obs
tacl
es W
ithin
the
Bun
ded
Area
wate
r tan
k
10m
7.1m
5m
orig
in
O
rCoor
dina
te S
yste
m θ=
0θ
5857
5655
54
4948
47
46
53
52
3132
3334
3536
3738
3940
4142
43
44
45
51 50
59
60
Probes
1-30 A
long G
antry
Not
e: S
ee T
able
1 fo
r pr
obe
pola
r co-
ordi
nate
s
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
35
Fi
gure
13:
Loc
atio
n of
Res
ista
nce
Prob
es fo
r the
Bun
d Ar
rang
emen
t With
Obs
tacl
es in
the
Bun
ded
Area
w
ater
tank
9.93
m7.
1m5m
orig
in
O
r
Coor
dinat
e Sy
stem θ
= 0
θ58
5756
55
54
4948
47
46
53
52
3132
3334
3536
3738
3940
4142
43
44
45
51 50
59
60
Probe
s 1-30
Alon
g Gan
try
Not
e: S
ee T
able
1 fo
r pr
obe
pola
r co-
ordi
nate
s
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
36
Fi
gure
14:
Dia
gram
Sho
win
g th
e Se
t-up
of th
e H
i Spe
ed V
ideo
Cam
era
Gra
duat
ed s
teel
rule
rH
igh
spee
d vi
deo
cam
era
Gan
try
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
37
Figu
re 1
5: V
aria
tion
of M
easu
red
and
Cal
cula
ted
Hei
ght o
f Wat
er R
emai
ning
in th
e Ve
ssel
with
Tim
e fo
r Diff
eren
t Nom
inal
Fi
ll H
eigh
ts
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
05
1015
2025
Tim
e (s
)
Head (m)
Fill
Hei
ght 1
.45m
Fill
Hei
ght 1
.60m
Fill
Hei
ght 1
.75m
Fill
Hei
ght 1
.45m
cal
cula
ted
Fill
Hei
ght 1
.60m
cal
cula
ted
Fill
Hei
ght 1
.75m
cal
cula
ted
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
38
Fi
gure
16:
Var
iatio
n of
Wat
er F
low
Rat
e In
ferr
ed fr
om H
eigh
t of W
ater
in th
e Ve
ssel
with
Tim
e fo
r Diff
eren
t Nom
inal
Fill
H
eigh
ts
050100
150
200
250
300
350
05
1015
2025
3
Tim
e (s
)
Water Flow Rate (litres/s)
Fill H
eigh
t 1.4
5m
Fill H
eigh
t 1.6
0mFi
ll Hei
ght 1
.75m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
39
Figu
re 1
7: P
lot o
f Tim
e of
Wat
er A
rriv
al a
t the
Res
ista
nce
Prob
es in
a 1
0m C
ircul
ar B
und
for
Expe
rimen
ts w
ith D
iffer
ent
Nom
inal
Fill
Hei
ghts
(Not
e: P
robe
s w
ere
loca
ted
at 5
m, 7
.1m
and
10m
to te
st fo
r rad
ial s
ymm
etry
of t
he fl
ow.
The
spre
ad
of th
e da
ta re
cord
ed a
t the
se p
oint
s gi
ves
som
e in
dica
tion
of th
e va
riabi
lity
and
effe
ct o
f the
pro
bes
them
selv
es.)
0
1000
2000
3000
4000
5000
6000
02
46
810
12
Radi
us (m
)
Time (ms)
Fill
Hei
ght 1
.45m
Fill
Hei
ght 1
.60m
Fill
Hei
ght 1
.75m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
40
Figu
re 1
8:
Com
paris
on o
f Fr
onta
l Spe
ed In
ferr
ed f
rom
Res
ista
nce
Prob
e D
ata
Alon
g a
Rad
ial L
ine
in T
est
G3
with
the
Va
lues
Infe
rred
from
a C
urve
Fit
to a
ll th
e D
ata
Col
lect
ed in
Sim
ilar E
xper
imen
ts.
Spee
d of
Pro
paga
tion
of F
ront
as a
Fun
ctio
n of
Dis
tanc
e
012345678
02
46
810
12
Dis
tanc
e (m
)
Front speed, m/s
Test
G3
Infe
rred
from
cur
vefit
tosp
read
ing
data
from
sim
ilar t
ests
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
41
Fi
gure
19:
Plo
t of
Tim
e of
Wat
er A
rriv
al a
t th
e R
esis
tanc
e Pr
obes
Alo
ng t
he 4
5 D
egre
e R
adiu
s in
a B
und
Arra
ngem
ents
SQ
3 an
d SQ
4 fo
r Exp
erim
ents
with
Diff
eren
t Nom
inal
Fill
Hei
ghts
0
2000
4000
6000
8000
1000
0
1200
0
1400
0
1600
0 0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
Rad
ius
(m)
Time (ms)
SQ 4
, Fill
Hei
ght 1
.45m
SQ 4
, Fill
Hei
ght 1
.60m
SQ 4
, Fill
Hei
ght 1
.75m
SQ 3
, Fill
Hei
ght 1
.45m
SQ 3
, Fill
Hei
ght 1
.60m
SQ 3
, Fill
Hei
ght 1
.75m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
42
Figu
re 2
0: P
lot o
f Tim
e of
Wat
er A
rriv
al a
t the
Res
ista
nce
Prob
es A
long
the
22.5
Deg
ree
Rad
ius
in a
Bun
d Ar
rang
emen
ts
SQ3
and
SQ4
for E
xper
imen
ts w
ith D
iffer
ent N
omin
al F
ill H
eigh
ts
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
02
46
810
1
Rad
ius
(m)
Time (ms)
SQ
4 Fi
ll H
eigh
t 1.4
5m S
Q4
Fill
Hei
ght 1
.60m
SQ
4 Fi
ll H
eigh
t 1.7
5m S
Q3
Fill
Hei
ght 1
.45m
SQ
3 Fi
ll H
eigh
t 1.6
0m S
Q3
Fill
Hei
ght 1
.75m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
43
Figu
re 2
1: C
ompa
rison
of V
olum
e of
Wat
er O
vert
oppi
ng th
e B
und
for C
ircul
ar B
unds
for D
iffer
ent A
ngle
s of
the
Bun
d W
all
(Cas
es s
how
n ha
ve in
itial
nom
inal
fill
heig
ht o
f app
roxi
mat
ely
90%
of t
he c
apac
ity o
f the
90
degr
ees
bund
wal
l cas
e).
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
010
2030
4050
6070
8090
100
Faci
ng a
ngle
of b
und
wal
l (de
gree
s)
Percentage overtopping (%)
bund
radi
us 1
0.0m
bund
radi
us 7
.1m
bund
radi
us 5
.0m
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
44
Fi
gure
22:
Com
paris
on o
f Vo
lum
e of
Wat
er O
vert
oppi
ng t
he B
und
for
Squa
re B
unds
for
Diff
eren
t N
omin
al In
itial
Wat
er
Dep
ths
in th
e Ve
ssel
0.0
2.0
4.0
6.0
8.0
10.0
12.0
1.4
1.45
1.5
1.55
1.6
1.65
1.7
1.75
1.8
Vess
el F
ill H
eigh
t (m
)
Percentage overtopping (%)
Arra
ngem
ent S
Q1
Arra
ngem
ent S
Q2
Arra
ngem
ent S
Q3
Arra
ngem
ent S
Q4
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
45
Fi
gure
23:
Per
cent
age
of W
ater
Ove
rtop
ping
the
Bun
d fo
r th
e Sa
me
Size
Bun
ded
Area
s fo
r D
iffer
ent S
hape
d B
unds
and
D
iffer
ent N
omin
al F
ill H
eigh
ts
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
1.4
1.45
1.5
1.55
1.6
1.65
1.7
1.75
1.8
Vess
el F
ill H
eigh
t (m
)
Percentage overtopping (%)
Arra
ngem
ent S
Q1
Arra
ngem
ent S
Q2
5.0m
circ
ular b
und
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
46
Figu
re 2
4: P
erce
ntag
e of
Wat
er O
vert
oppi
ng th
e B
und
for
the
Sam
e Si
ze B
unde
d Ar
eas
for
Diff
eren
t Sha
ped
Bun
ds a
nd
Diff
eren
t Nom
inal
Fill
Hei
ghts
0.0
2.0
4.0
6.0
8.0
10.0
12.0
1.4
1.45
1.5
1.55
1.6
1.65
1.7
1.75
1.8
Vess
el F
ill H
eigh
t (m
)
Arra
ngem
ent S
Q3
Arra
ngem
ent S
Q4
10.0
m c
ircul
ar b
und
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
47
Figu
re 2
5: V
ideo
Imag
e of
the
Wat
er F
ront
5.0
m fr
om th
e R
elea
se V
esse
l
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
48
Fi
gure
26:
Vid
eo Im
age
of th
e W
ater
Fro
nt 7
.1m
from
the
Rel
ease
Ves
sel
R
epor
t Num
ber:
R 4
418
Issu
e: 0
1
CO
MM
ERC
IAL
IN C
ON
FID
ENC
E
Page
49
Fi
gure
27:
Vid
eo Im
age
of th
e W
ater
Fro
nt 1
0.0m
from
the
Rel
ease
Ves
sel
Printed and published by the Health and Safety ExecutiveC30 1/98
Printed and published by the Health and Safety ExecutiveC1.25 01/02
CRR 405
£20.00 9 780717 622559
ISBN 0-7176-2255-X