Hazards from wave overtopping: field measurements on the Zeebrugge breakwater PDF

Preview

Summary

HAZARDS FROM WAVE OVERTOPPING: FIELD MEASUREMENTS ON THE ZEEBRUGGE BREAKWATER J. Geeraerts1, C. Boone1, J. De Rouck1, A. Kortenhaus², L. Van Damme³, L. Franco4 1 Ghent University, Department of Civil Engineering, Technologiepark 904, 9052 Zwijnaarde, Belgium; ² Leichtweiss Institut für Wasserbau der...

Description

HAZARDS FROM WAVE OVERTOPPING: FIELD MEASUREMENTS ON THE ZEEBRUGGE BREAKWATER J. Geeraerts1, C. Boone1, J. De Rouck1, A. Kortenhaus², L. Van Damme³, L. Franco4 1

Ghent University, Department of Civil Engineering, Technologiepark 904, 9052 Zwijnaarde, Belgium; ² Leichtweiss Institut für Wasserbau der Technischen Universität Braunschweig, Beethovenstrasse 51A, 38106 Braunschweig, Germany; ³ Flanders Community, Maritime Access, Tavernierkaai 3, 2000 Antwerpen, Belgium; 4 Modimar S.r.l., Via Monte Zebio 40, 00195 Rome, Italy; e-mail: [email protected]; [email protected]; [email protected]; [email protected], [email protected]; [email protected]

keywords: wave overtopping, field measurements, hazard analysis, rubble mound breakwater, wave induced loads.

1

INTRODUCTION

One of the main objectives of the research project CLASH (Crest Level Assessment of coastal Structures by full scale monitoring, neural network prediction and Hazard analysis on permissible wave overtopping, 2002-2004, funded by the European Commission EVK3-CT2001-00058) was to refine guidance on tolerable mean or individual overtopping discharges and on various levels of hazard on people and infrastructure by overtopping waves. To accomplish this objective, field measurements on wave overtopping, including visual interpretation of hazards, are carried out at three different locations in Europe: a vertical seawall with rubble mound toe protection at Samphire Hoe, United Kingdom (Pullen et al., 2003), a rock armoured rubble mound breakwater in shallow water at Ostia, Italy (Franco et al., 2003) and a rubble mound breakwater armoured with flattened Antifer cubes at Zeebrugge, Belgium (Troch et al., 2004). Moreover, at the Zeebrugge site also measurements to quantify hazards resulting from wave overtopping are carried out. The design and set-up of this measurement system is presented in Geeraerts et al. (2003). The main objective of this paper is to present the measurement results from this unique system gathered during winter 2003 – 2004. Section 2 provides a short description of the measurement site and the measurement infrastructure together with the analysis and main results of the field measurements. Section 3 provides a discussion on the measured loads.. Finally section 5 summarizes the main conclusions.

2 2.1

FIELD MEASUREMENTS Measurement site

The measurement site is located at the western breakwater of the Zeebrugge harbour, which is situated at the Belgian North Sea coast (Fig. 1). The Zeebrugge breakwater has been used for more than a decade for full scale measurements of wave interaction with a breakwater. A prototype monitoring system for acquisition of wave characteristics and induced structural

1

response, has been developed since the early nineties. Troch et al. (1998) give a general description of the monitoring system related to infrastructure, instrumentation and dataacquisition. The breakwater has been instrumented for the specific measurement and analysis of wave induced pore pressures inside the breakwater core (Troch, 2000), and wave run-up on the breakwater slope (Van de Walle, 2003). More recently the breakwater has been instrumented to measure wave overtopping at full scale (De Rouck et al., 2003 and Troch et al., 2004) and to measure hazards resulting from wave overtopping (Geeraerts et al., 2003). In this paper results of the hazard measurements gathered during winter 2003 – 2004 are presented.

Zeebrugge Blankenberge Oostende

Nieuwpoort

FR

BELGIUM

AN CE

Fig. 1 a) Location of Zeebrugge harbour at the Belgian North Sea Coast

Location of the field site at Zeebrugge harbour

The Zeebrugge breakwater is a conventional rubble mound breakwater with a low-crested superstructure. The armour layer is composed of 2 layers of flattened grooved cubes (25 ton). The breakwater core consists of quarry run (2-300 kg). The filter layer is composed of rock (1-3 ton). An access road allows access to the breakwater. Design conditions for the breakwater are: significant wave height H s = 6.20 m, maximum period T = 9.0 s and still water level SWL = Z + 6.75 (Z + 0.00 is chart datum). A tide cycle lasts for 12 hours and 24 minutes, the tidal range varies between Z + 4.62 m (M.H.W.S.) and Z + 0.32 (M.L.W.S.). Fig. 2 shows the cross-section of the Zeebrugge breakwater at the location of the wave overtopping tank. The design slope of the breakwater is 1:1.4. The design crest level of the armour layer is Z + 12.40 m, the level of the crest crown wall is Z + 10.20, the mean access road level is Z + 8.30.

Fig. 2

Cross-section of the Zeebrugge rubble mound breakwater at the location of the overtopping tank. 2

2.2

Measurements to identify overtopping hazards

The measurements include force measurements on people (dummy persons), pipelines and structures (vertical wall) on the breakwater’s crest. Since measurements of incident waves, overtopping discharges ánd hazards are carried out at Zeebrugge, it is possible to directly link between wave conditions on the one hand and overtopping discharges and quantified hazards on the other hand. The measurement devices and their design are described in Geeraerts et al. (2003) and are briefly summarized here. 2.2.1 Forces on instrumented dummies Three dummies have been installed and instrumented. The dummies consist of an aluminium ‘body plate’ on a steel framework. Two of them (dummy 2 and 3, Fig. 3; 1.70m x 0.50m) are placed on the crest wall directly behind the armour units (level Z + 10.20, Fig. 1). The smaller one (dummy 1; 1.40m x 0.50m) is placed at the landward side of the access road on top of the breakwater’s crest as shown in Fig. 3. Forces on each dummy are measured by three S-shaped Tedea Huntleigh load cells which are placed in a triangular scheme (Fig. 4). These load cells are suited for use in both tension and compression. Capacity for the sensors of the dummies is –1000 kg to +1000 kg. Total error is 0.05 % of the force applied on the cell.

LC7

LC8 Fig. 3

View at the three instrumented dummies on Fig. 4 site.

LC9 Instrumented dummy with position of load cells.

2.2.2 Forces on instrumented pipeline A section of a dredging pipe (length = 6.00m, diameter = 0.65m) has been mounted on top of the crest wall. Horizontal and vertical force components on the pipeline are measured by means of four S-type load cells as used for the dummies (Fig. 5). The capacity of the load cells for this application is from – 2000 kg to + 2000 kg. The total error is 0.03% of the applied load on the cell.

3

Fig. 5

LC16

LC15

LC14

LC13

View at the instrumented pipeline on site with indication of the position of the load cells.

2.2.3 Forces and pressures on a vertical wall These measurements are carried out by measuring the force on an aluminium plate, with the same dimensions as the body plate of the largest dummies (1.70 m x 0.50 m), mounted to the concrete column supporting the measurement jetty which is built on top of the breakwater (see Troch et al., 1998). This column serves as vertical wall. Fig. 6 shows a detail of the mounted body plate, indicating a load cell. Forces are measured by three S-shaped load cells, with the same positioning and capacity as for the dummies. Moreover, pressures are measured by five flush-mounted pressure sensors positioned along a vertical line in the centre of the aluminium plate (indicated by ‘PS’ in Fig. 7).

PS 5 LC12

PS 4

PS 3 LC11 PS 2 LC10 PS 1 Fig. 6

Mounted plate on the vertical wall, Fig. 7 indicating a load cell.

Mounted plate on the vertical wall, indicating positions of load cells and pressure transducers.

4

2.3

Measurement signals

The measurement system to identify hazards is operational since January 2003. Since then, three storms with wave overtopping have been measured at the Zeebrugge breakwater: Oct. 7th, 2003, Dec. 22th, 2003 and Febr. 8th, 2004. The wave characteristics and overtopping discharges for these storm events are given in Table 1. Table 1: Wave characteristics and overtopping discharges during storm events of Oct. 7th, 2003, Dec. 22th, 2003 and Febr. 8th, 2004 Date Oct. 7th, 2003 Dec. 22th, 2003 Febr. 8th, 2004

Time 12.00 - 14.00 00.00 – 02.00 14.45 – 16.45

MWL [Z+ …m] 4.77 5.26 5.32

Hm0 [m] 3.23 3.03 3.59

Tp [s] 7.91 8.57 8.57

q [l/s/m] 0.09 0.07 0.60

2.3.1 Force measurements on instrumented dummies During the three measured storm events no forces were recorded on the small dummy. On the two large dummies on top of the crest wall, forces were measured during all three storm events. Fig. 8 gives an overview of the forces measured by the load cells on dummy 3. Fig. 8 (a) shows signals originating from the storm dd. Oct. 7th, 2003, during a period of 1.5 s. Fig. 8 (b) and (c) show 2 different impacts on the dummy during the storm dd. Febr. 8th, 2004 each also during a period of 1.5 s. On Fig. 8(b) a clear impact can be distinguished: at 693.6 s all three load cells measure a compression (compression means that the signal becomes more negative than the default value for no load). a)

b)

Fig. 8 Measurements by the load cells on dummy 3 on (a) Oct. 7th, 2003 (1.5 s duration), on (b) Febr. 8th, 2004 (1.5 s duration; 15h38) 5

c)

Fig. 8

Measurements by the load cells on dummy 3 on (c) Febr. 8th, 2004 (1.5 s duration; 16h23).

2.3.2 Force measurements on instrumented pipeline Also the pipeline was subject to wave overtopping forces during all three measured storm events. Fig. 9 gives an overview of measured force signals. The signals in Fig. 8(a) originate from the storm dd. Febr. 8th, 2004 and show a period of 7 s. The impact only takes 1 second. The time signals in Fig. 9(a) show that the overtopping wave smashed the pipeline with a large impact at the bottom side: The two lower load cells (LC13 and LC14) indicate tension while the two upper load cells (LC15 and LC16) indicate pressure. Fig. 9(b) shows a smaller impact on the pipeline during storm 8 (Dec. 22th, 2003). a)

b)

Fig. 9

Measurements by the load cells on the pipeline and the velocity meter near the pipeline, on (a) Febr. 8th, 2004 (1.5 s duration) and (b) Dec. 22th, 2003 (7 s duration).

6

2.3.3 Force and pressure measurements on vertical wall Fig. 10 gives an overview of the forces and pressures measured by the load cells and the pressure sensors. Part (a) of the graphs show the signals measured by the load cells and part (b) of the graphs shows the measurements of the pressure sensors. The relation between the numbering used in Fig. 6 and the numbering used in the graphs is given in Table 2. The signals in Fig. 10 originate from the storm dd. Febr. 8th, 2004 and show a period of 1.5 s. This graph shows that an impact on the lower part of the wall has been measured by the load cells (LC10 and LC11) as well as by the pressure sensors (especially by PS1 and PS2 which are the lowest positions). a)

Load cells

b)

Pressure sensors

Fig. 10

2.4

Pressure sensor numbers related to numbers as used in Fig. 6. N° in Fig. 6 N° in graphs PS1 PR1243821 PS2 PR1250429 PS3 PR1237910 PS4 PR1237920 PS5 PR1250430 Measurements by (a) the load cells and (b) the pressure sensors on the vertical wall on Febr. 8th, 2004 (1.5 s duration)

Analysis and results

2.4.1 Instrumented dummies For each storm, the two highest impact loads per load cell are determined. These individual maxima or given in Tables 2 and 3 for dummy 2 and dummy 3 respectively. For these impacts 7

the total impact on the dummy is then calculated. The two maximum total impacts are also given in the tables. It should be noted that the total impacts correspond to only 1 individual maximum load since not all individual maxima occur at the same time. The maximum load (8110 N) on dummy 2 is measured on Febr. 8th 2004. The impact of the maximum load is located on the upper part of the dummy (LC1 > LC2 + LC3). The maximum load on dummy 3 is measured during the same storm (8835 N) and the upper load cell experiences the highest individual loads. Table 2: Total and individual impacts measured by load cells (LC) on dummy 2 during resp. storms. 1 LC4 LC5 LC6 Total impact Date (N) (N) (N) 2 (N) 3 Oct. 7th, 2003 Dec. 22th, 2003 Febr. 8th, 2004

2520 610 375 315 5590 2405

120 800 405 430 1335 1790

655 590 390 155 1185 1235

3295 2000 1170 900 8110 5430

Table 3: Total and individual impacts measured by load cells (LC) on dummy 3 during resp. storms. Date Oct. 7th, 2003 Dec. 22th, 2003 Febr. 8th, 2004

LC7 (N) 1245 665 970 270 4970 1950

1

LC8 (N) 1005 995 250 385 1640 1015

2

LC9 (N)

3

1375 1050 490 740 2225 1340

Total impact (N) 3625 2710 1710 1395 8835 4305

Different pressure distributions have been tried for dummy 2 and dummy 3 for the three highest total impacts on Febr. 8th, 2004. Following distributions have been considered -

rectangular distribution starting from upper border of dummy; rectangular distribution starting from lower border of dummy; triangular distribution starting from upper border of dummy; triangular distribution starting from lower border of dummy; trapezoidal distribution over whole height of dummy; point load

Fig. 11 and Fig. 12 give two possible pressure distributions corresponding to the measured loads for the respective dummies. It is clear that for each situation there is a distribution which fits the measurements better. The shown pressures are considered to be uniformly distributed over the width of the dummy (0.5 m).

8

(a)

(b)

(c)

Fig. 11 Pressure distributions for the three highest impacts on Febr. 8th, 2004 on dummy2: (a) 8114 N, (b) 5432 N and (c) 4655 N.

(a)

(b)

(c)

Fig. 12 Pressure distributions for the three highest impacts on Febr. 8th, 2004 on dummy3: (a) 8114 N, (b) 5432 N and (c) 4655 N.

9

2.4.2 Instrumented pipeline Similar as for the dummies, the two highest impact loads during three storms are determined for each load cell on the pipeline and for each storm (Table 4). For these impacts the total impact on the pipeline and the direction of the total impact is calculated. Fig. 13 explains the direction (angle α) of the impact and the numbering of the load cells (bird’s view at the pipeline). α

Sea side 1

2 4

Fig. 13

Land side

3

(a) (b) Direction of the impact load (a) and localisation of the load cells (b) on the pipeline.

Table 4: Total and individual impacts measured by load cells (LC) on the pipeline during resp. storms. LC1 LC2 LC3 LC4 Total impact α Date (N) (N) (N) (N) (N) (°) -450 -450 410 600 1350 -41.6 Oct. 7th, 2003 465 525 285 215 1110 63.3 805 815 645 565 2020 53.3 Dec. 22th, 2003 725 890 645 1310 2535 39.5 -2820 -2755 2460 2790 -46.7 7660 Febr. 8th, 2004 1095 2830 1265 2705 44.6 5585 Comparable to the measurements of the dummies and the vertical wall, the highest total impact appears at Febr. 8th 2004. Impacts up to 5585 N and 7660 N are calculated over the whole length of the pipeline. These values correspond to line loads of resp. 930 N/m and 1300 N/m. The highest impact is located at the lower part of the pipeline (ca. 45° below the horizontal), while the lower impact is located at the upper part of the pipeline (ca. 45° above the horizontal). 2.4.3 Vertical wall Also for the verticall wall, the two highest impact loads during three storms are determined for each load and for each storm. For these impacts the total impact on the vertical wall is calculated. In Table 5 the most relevant total impacts (2 per storm) on the vertical wall during the given storms are shown. For each impact, the loads per load cell are given. The maximum load (1427 N) on the vertical wall is measured on Febr. 8th, 2004. The impact of the maximum load is located near load cell 10 (LC10 > LC11, LC12). The pressures measured at the same moments are given in Table 6. The maximum impact measured at Febr. 8th, 2004 is, as registered by the pressure sensors, located at the bottom side of the vertical wall. On Febr. 8th, 2004 also an impact of 730 N is measured and is located at the upper side of the vertical wall (pressures of 0.6 and 0.9 kPa resp. measured by PS4 and PS5).

10

Table 5: Total and individual impacts measured by load cells (LC) on the vertical wall during resp. storms. LC12 LC10 LC11 Total impact Date (N) (N) (N) (N) 45 305 275 625 Oct. 7th, 2003 135 55 105 295 10 115 135 260 Dec. 22th, 2003 205 50 -10 245 411 630 386 1427 Febr. 8th, 2004 505 115 110 730 Table 6: Total and individual impacts measured by load cells (LC) and pressure sensors (PS) on the vertical wall during resp. storms. LC12 LC10 LC11 Total impact PS1 PS2 PS3 PS4 PS5 Date (N) (N) (N) (N) (kPa) (kPa) (kPa) (kPa) (kPa) 45 305 275 625 7.5 1.0 0 0 0 Oct. 7th, 2003 135 55 105 295 1.2 2.0 0 0 0 th Dec. 22 , 10 115 135 260 3.9 3.0 0.8 0 0 2003 205 50 -10 245 0 0 0 0 0 411 630 386 2.9 1.2 0.7 0.3 0 1427 Febr. 8th, 2004 505 115 110 0 0 0 0.6 0.9 730 Different pressure distributions are also investigated for the vertical wall for the three highest total impacts on Febr. 8th, 2004. Fig. 14 shows two different pressure distributions for each of the impacts. The measured loads are again considered to be uniformly distributed over the width of the wall (0.5 m). The pressures measured at the same moment of the given impacts are indicated in italic in the figures. The highest impacts (drawing (a) and (b)) on the wall are located at the lower part of the dummy. Pressure sensors are installed in the middle of the wall. When no pressures are measured (see Fig. 14 (b)), the overtopping water has hit the plate at the border of the plate. The impacts are clearly lower than the impacts on the dummies. This is probably due to the higher position of the wall as compared to the dummies.

11

(a)

(b)

(c)

Fig. 14 Pressure distributions for the three highest impacts on Febr. 8th, 2004 on the vertical wall: (a) 1427 N, (b) 1415 N and (c) 730 N.

3

DISCUSSION

From the reported measured loads, it can be concluded that for average overtopping rates smaller than 1 l/s/m, which are generally considered as “small overtopping”, rather high forces due to overtopping waves on both dummies and structures are measured during three distinct storms at the Zeebrugge site. Forces of up to 8.8 kN on a person dummy and up to 1.3 kN/m on an instrumented pipeline were recorded. Why calling the measured forces “rather high”?: Endoh & Takahashi (1994) conducted tests at prototype scale to investigate the stability of a person under several flow conditions. The current force on a person dummy was measured and the loss of balance was observed. Two types of loss of balance were considered: the “tumbling” and “slipping” type. In the first case a person is knocked over by the overtopping flow in the downstream direction, while in the latter case a person is knocked over in the upstream direction. A critical value of 140 N is found for the loss of stability (slipping type loss of balance). The reported prototype results are clearly higher than this value, which indicates a considerable hazard during storm conditions at the Zeebrugge site. It should however be noted that Endoh & Takahashi used steady flow conditi...