Evaluation of performance of scour protection and edge scour development : Offshore Windpark Egmond aan Zee

(1)

(2)

(3)

Evaluation of performance of scour

protection and edge scour development

Offshore Windpark Egmond aan Zee

(4)

(5)

Title

Evaluation of performance of scour protection and edge scour development Client MT Kennis Deltares Project 1200160-002 Reference 1200160-002-HYE-0001 Pages 58 Keywords

offshore windpark, scour protection, edge scour, Egmond Windpark

Summary

In 2006 the first Dutch offshore windpark was installed approximately 10 to 18 kilometres offshore from Egmond aan Zee. This first offshore windpark should contribute to an increase of the use of renewable energy and is therefore designated as a “demonstration project” by the Dutch Ministry of Economic Affairs. Knowledge and experience that is gained in this project will be used in the development of future offshore windparks.

The Offshore Windpark Egmond aan Zee (hereafter referred to as OWEZ) covers an area of

about 27 km2 and consists of 36 wind turbine generators on top of monopile foundations. The

seabed around the monopiles is protected against local seabed erosion by a scour protection consisting of two granular layers (a filter layer and an armour layer). In 2007 Ballast Nedam and Delft Hydraulics set up a joint research project “Evaluation of scour protection Offshore Windpark Egmond aan Zee” with the following main objectives:

I. To improve the understanding of the long-term behaviour of dynamically stable scour protection under influence of waves and currents

II. To improve the understanding of edge scour development

III. To verify knowledge regarding stone stability and edge scour development gained from laboratory experiments against field observations

Since 2009 this research program is part of the Deltares-road map “Water and soil around structures” and will continue until 2011. This report contains an evaluation of the present state of the scour protection of the monopile foundations of all 36 piles in OWEZ as well as an evaluation of edge scour development around the scour protections.

The primary focus of this report is on the analysis of bathymetric survey data, provided by Ballast Nedam. Based on the observations, dedicated analysis of the hydrodynamics and calculations of the sediment transport will be executed in the next part of this project. The evaluation is based on new bathymetric survey data (May 2009) and previous surveys, all provided by Ballast Nedam.

(6)

(7)

1 Introduction

1.1 Offshore Windpark Egmond aan Zee

In 2006 the first Dutch offshore windpark was installed approximately 10 to 18 kilometres offshore from Egmond aan Zee. This windpark should contribute to an increase of the use of renewable energy and is therefore designated as a “demonstration project” by the Dutch Ministry of Economic Affairs. Knowledge and experience that is gained in this project will be used in the development of future offshore windparks.

The Offshore Windpark Egmond aan Zee (hereafter referred to as OWEZ) covers an area of

about 27 km2 and consists of 36 wind turbine generators on top of monopile foundations

(hereafter referred to as WTGs) with a total capacity of 118 MW and an anticipated lifetime of 25 years. The water depths in the windpark area vary between 16 and 21 metres relative to MSL, see Figure 1.1.

Figure 1.1 Location of 36 WTGs (and one Meteo Mast)and bathymetry surrounding OWEZ

The owner is Noordzeewind, which is a consortium of Shell and Nuon. Construction works were carried out by Bouwcombinatie Egmond (hereafter referred to as BCE), which consists of Ballast Nedam and the Danish wind turbine manufacturer Vestas. The design of the scour protection was made by Infra Consult + Engineering. Delft Hydraulics carried out laboratory experiments in 2005 to verify and to advice on further optimisation of the scour protection design.

(10)

1.2 Research program

In 2007 Ballast Nedam and Delft Hydraulics set up a joint research project “Evaluation of scour protection Offshore Windpark Egmond aan Zee” with the following main objectives:

• To improve the understanding of the long-term behaviour of dynamically stable scour

protection under influence of waves and currents

• To improve the understanding of edge scour development

• To verify knowledge regarding stone stability and edge scour development gained from

laboratory experiments against field observations

All gained knowledge will be used in the development of future offshore windparks.

In 2008 Delft Hydraulics merged into Deltares. Since 2009 this research program is part of the Deltares-road map “Water and soil around structures” and will continue until 2011. This report contains an evaluation of the present state of the scour protection of the monopile foundations of all 36 WTGs in OWEZ as well as an evaluation of edge scour development around the scour protections. The primary focus of this report is on the analysis of bathymetric survey data, provided by Ballast Nedam, see Paragraph 1.4. Based on the observations, dedicated analysis of the hydrodynamics and calculations of the sediment transport will be executed in the next part of this project.

1.3 Scour protection layout

The scour protection that was recommended by Delft Hydraulics and mainly followed by BCE consists of two layers:

• a granular filter layer with Dn50,F 0.05m, minimum thickness of 0.4m, minimum extent of

24m (measured at the average height)

• an armour layer with Dn50,A 0.4m, minimum thickness 1.4m, minimum extent of 18m

(measured at the average height)

4.6 m

armour layer (Dn50=0.4m)

minimum extent 18m

filter layer (Dn50=0.05m)

minimum extent 24m

monopile

CROSS-SECTION

1 .4 m 1 .8 m

(11)

A schematic cross-section is presented in Figure 1.2. The installation procedure that was adopted by BCE at nearly all WTGs was:

Dump filter material

Drill pile and install the internal transition piece Mount tower and turbines

Connect electricity cables

Complete scour protection by dumping armour material 1.4 Available surveys

To be able to assess the quality of the installed scour protection, BCE executed surveys before and after installation of each protection layer. As part of the monitoring program, BCE has also been surveying the scour protection on an annual basis since 2007. In Table 1.1 an overview of all surveys that were made available to Deltares is presented.

survey ID survey description average survey date # surveyed WTGs

SU01 initial seabed May 2006 33 / 36

SU02 out survey filter 2006 June 2006 36 / 36

SU03 control survey 2006 June 2006 3 / 36

SU04 in survey armour 2006 July 2006 15 / 36

SU05 out survey armour 2006 October 2006 36 / 36

SU06 check survey 2007 June 2007 36 / 36

SU07 out survey additional armour August 2007 20 / 36

SU08 check survey 2008 May 2008 36 / 36

SU09 check survey 2009 May 2009 36 / 36

Table 1.1 Overview of available surveys provided by Ballast Nedam to Deltares

By substracting different surveys we obtain installed layer thicknesses or level changes within a certain period. The installed layer thickness of filter material, for instance, can be calculated by subtracting survey ‘SU01’ from ‘SU02’. Likewise, the change in vertical level between May 2008 and May 2009 is obtained by subtracting ‘SU08’ from ‘SU09’.

1.5 Performance criteria of scour protection

The design philosophy of BCE was based on a fixation of the initial bed material by means of a scour protection. The fixation should be obtained by a dynamically stable scour protection. This means that some deformation of the armour layer can be tolerated as long as the pile fixation level is maintained.

According to the pile design, the minimum pile fixation level may drop by 1.9m within the lifetime of 20 years. Assuming a linear relation, this means that the required bed level in 2009 is thus 3yr * 1.9m / 20yr = 0.29m below the initial bed level in 2006. The average layer thickness should be at least 1.8m (0.4m filter + 1.4m armour), see Figure 1.2. Or in other words, the average top level of the scour protection should be 1.5m above the initial bed level. Small local gaps (in the order of a few square meters) can be compensated by a surplus of scour protection material in the immediate vicinity. These evaluation criteria were first described in memo MCI26677/H4932/DR, dated 24 August 2007. Preliminary results on performance of the scour protection and edge scour development within one year after installation can be found in Raaijmakers et al. (2007).

(12)

1.6 Content of report

In Chapter 2 the performance of the scour protection until May 2009 is discussed, while Chapter 3 contains the analysis of edge scour outside the scour protection. The reader who is primarily interested in which scour protections are “fit for purpose” is referred to Chapter 4, in which the present state of the scour protections of all WTGs is evaluated against the performance criteria and summarized in Table 4.1. More detailed background information about this judgement can be found in Chapters 2 and 3. Finally, the conclusions and recommendations are described in Chapter 5.

(13)

2 Analysis of scour protection

2.1 Overview of characteristics of scour protection for all WTGs

For the assessment of the performance of the scour protection, the bed level changes close to the WTGs were investigated, both graphically and analytically. To investigate bed level changes since installation, the area within a distance of two times the pile diameter from the pile centre was studied. The pile-average and local maximum bed level changes, the average and minimum thickness of the scour protection in this area and the relative rock volume reduction since installation are presented in Table 2.1. To assess the state of the scour protection close to the WTG, where the amount of protection is most critical, these parameters were also calculated for the area within a distance of just one pile diameter from the pile centre, see Table 2.2.

For a graphical presentation of the bathymetries around all 36 WTGs, the reader is referred to the Appendix. In Appendix A the differences between May 2008 and May 2009 are presented. Appendix B illustrates the present state of the scour protection by subtracting the initial bathymetry in 2006 from the May 2009-surveys. In the main text some more plots are presented for selected WTGs.

The average bed level change around all 36 WTGs between May 2008 and May 2009 appears to be very limited according to the second column of Table 2.1. The average bed level change is 1cm with a maximum average rise of the protection of 9cm (at WTG-12) and a maximum average drop of 6cm (at WTG-8). The values of this second column appear to be of the same order as typical errors in the correction for the tidal water level at the moment of the survey. This second column tells us that all major changes, if any, occurred within a zone of 2D from the pile centre; the settling of the armour layer, which occurred in the first year after installation, appears to have stopped.

The local maximum changes in scour protection level within 2D from the pile centre are in the order of 0.2-0.5m with the largest value at WTG-10. The local maximum changes are probably partly caused by redistribution of armour material within the protected area (local heaps of armour material are spread out) and maybe also caused by small inaccuracies in the survey data.

The average layer thickness is still more than 2m for most of the WTGs. Only WTG-13 (1.75m) and WTG-15 (1.83m) have a significantly smaller layer thickness. The local minimum thickness of the scour protection is on average smaller than 1m (range 0.52-1.26m). Because these minimum values are mostly located at the edge of the armour layer, which not always extends up to 2D from the pile centre, the consequences of these local minima are expected to be negligible. Therefore, all parameters of Table 2.1 are also calculated for an area of 1D from the pile centre, see Table 2.2.

Within 1D, the average bed level changes remain small, which is an indication that the past storm season was relatively mild, since during severe storms armour stones would be transported from the dynamic zone between 0.5-1.0D towards the 1.0-2.0D zone. The local maximum bed level change still yields the same value of 0.52m at WTG-10. The average value of the maximum bed level change is comparable to the 2D-value.

(14)

1200160-002-HYE-0001, Version 1.0, 3 September 2009, final WTG nr. average bed level change 2008-2009, positive means lowering [m] maximum (local) bed level change 2008-2009, positive means lowering [m] average thickness of scour protection (filter+armour) in May 2009 [m] minimum thickness of scour protection (filter+armour) in May 2009 relative volume reduction (filter + armour) in period 2006-2009 [%] 1 0.05 0.36 2.13 1.12 21 2 0.01 0.22 2.17 1.13 9 3 -0.01 0.31 1.94 0.55 17 4 -0.01 0.36 1.97 0.74 13 5 -0.06 0.37 2.09 0.70 14 6 -0.04 0.35 2.24 0.65 12 7 -0.06 0.23 2.13 0.83 10 8 0.06 0.38 2.17 1.19 3 9 0.05 0.45 2.13 0.90 13 10 0.03 0.52 2.05 1.01 15 11 0.04 0.33 2.08 0.73 16 12 -0.09 0.36 2.05 0.68 11 13 0.03 0.39 _1.92 _0.64 ₁₀ 14 0.02 0.44 2.03 1.03 9 15 0.02 0.33 1.83 0.86 14 16 0.04 0.41 2.01 1.18 12 17 0.03 0.41 2.00 0.91 9 18 0.01 0.27 1.97 0.66 7 19 0.05 0.28 1.97 1.05 11 20 0.00 0.32 2.06 1.21 5 21 0.05 0.38 2.11 0.99 4 22 0.00 0.43 2.52 1.26 10 23 0.04 0.35 2.07 1.20 6 24 0.00 0.35 2.04 0.97 17 25 0.00 0.37 1.97 0.86 18 26 0.01 0.33 2.07 0.81 9 27 0.00 0.35 2.18 1.05 15 28 0.01 0.26 2.06 1.12 5 29 -0.08 0.32 2.30 1.21 7 30 0.05 0.38 2.02 0.95 18 31 0.02 0.39 2.19 0.95 -2 32 0.04 0.50 2.00 0.73 12 33 0.02 0.40 1.99 0.52 10 34 0.05 0.33 2.17 1.07 12 35 0.04 0.44 2.12 1.03 12 36 0.04 0.32 2.26 1.13 13 mean 0.01 0.36 2.08 0.93 11 min -0.09 0.22 1.83 0.52 -2 max 0.06 0.52 2.52 1.26 21

Table 2.1 Overview of selected parameters, describing the state of scour protection, calculated for an area within a distance of two pile diameters (9.2m) from the pile centre

(15)

1200160-002-HYE-0001, Version 1.0, 3 September 2009, final WTG nr. average bed level change 2008-2009, positive means lowering [m] maximum (local) bed level change 2008-2009, positive means lowering [m] average thickness of scour protection (filter+armour) in May 2009 [m] minimum thickness of scour protection (filter+armour) in May 2009 relative volume reduction (filter + armour) in period 2006- 2009 [%] 1 0.04 0.36 2.25 1.64 26 2 0.01 0.22 2.20 1.72 17 3 -0.05 0.23 2.11 1.52 20 4 -0.04 0.36 2.08 1.45 17 5 -0.07 0.14 2.33 1.69 17 6 -0.06 0.31 2.60 2.11 14 7 -0.08 0.23 2.42 1.67 10 8 0.11 0.38 2.10 1.50 9 9 0.08 0.45 2.21 1.74 14 10 0.08 0.52 2.09 1.68 19 11 0.05 0.33 2.30 1.76 13 12 -0.10 0.09 2.21 1.73 9 13 0.08 0.39 2.02 1.20 13 14 0.04 0.44 2.06 1.56 14 15 0.02 0.28 1.96 1.37 14 16 0.07 0.39 1.96 1.64 16 17 0.02 0.27 1.93 1.44 13 18 0.01 0.26 2.11 1.81 5 19 0.03 0.28 1.97 1.65 14 20 0.00 0.32 2.11 1.57 7 21 0.04 0.3 2.04 1.39 10 22 0.00 0.43 2.29 1.60 17 23 0.05 0.35 2.13 1.67 10 24 -0.03 0.15 1.92 1.44 20 25 -0.01 0.23 1.93 1.55 19 26 0.01 0.27 2.07 1.72 16 27 -0.02 0.23 2.20 1.92 19 28 -0.03 0.26 2.10 1.82 9 29 -0.10 0.32 2.29 1.74 12 30 0.04 0.38 2.03 1.67 23 31 0.04 0.39 2.23 1.66 4 32 0.00 0.50 2.13 1.60 12 33 0.00 0.30 1.93 1.44 10 34 0.03 0.33 2.19 1.94 15 35 0.03 0.23 2.10 1.79 14 36 0.03 0.2 2.31 2.05 14 mean 0.01 0.31 2.14 1.65 14 min -0.10 0.09 1.92 1.20 4 max 0.11 0.52 2.60 2.11 26

Table 2.2 Overview of selected parameters, describing the state of scour protection, calculated for an area within a distance of one pile diameter (4.6m) from the pile centre

(16)

The average layer thickness, calculated over 1D from the pile centre is only slightly larger compared to the thickness, calculated over 2D, although the plots in Appendix B show that the largest layer thickness generally occurs at the area between 1.0 and 1.5D from the pile centre. Apparently, the smaller layer thickness close to the pile (0.5-1.0D) is balanced out against the smaller layer thickness at the edge of the armour layer (1.5-2.0D). The minimum layer thickness of total scour protection (filter + armour) is larger within 1D than within 2D, which is favourable for the stability of the protection close to the pile, except for WTG-13. The relative reduction of the volume of scour protection (i.e. total dumped scour protection minus present volume of scour protection divided by total dumped protection) is somewhat larger for the 1D-area than for the 2D-area. If we assume that settlement of the stone layers occurs evenly distributed over the entire protection, this is an indication that scour protection material is indeed transported from the dynamic zone (0.5-1.0D).

However, the relatively small difference between the values in the last columns of Table 2.1 and Table 2.2 also shows that the majority of ‘loss of scour protection material’ is likely to be caused by settlement of the stone layers and not by deformation during severe storms, see also Paragraph 2.4. However, at a later stage we still have to study the severity of the already occurred storm events.

After this general evaluation of all 36 WTGs, we will further investigate some of the mentioned WTGs, which showed conspicuous values. The selected examples that will be discussed in Paragraph 2.2 are:

1 WTG-08: largest average bed level lowering within 1D and 2D between May 2008 and

2009, see Paragraph 2.2.1

2 WTG-10: largest local bed level lowering within 1D and 2D between 2008 and 2009, see

Paragraph 2.2.2

3 WTG-24: smallest average layer thickness within 1D, see Paragraph 2.2.3

4 WTG-15: smallest average layer thickness within 2D, see Paragraph 2.2.4

5 WTG-13: smallest local layer thickness within 1D, see Paragraph 2.2.5

6 WTG-33: smallest local layer thickness within 2D, see Paragraph 2.2.6

7 WTG-07: worst distribution of protection material around WTG, see Paragraph 2.2.7

WTG-01, which shows the largest relative reduction in scour protection volume, is not further discussed, because it was already demonstrated in 2008, when the relative volume reduction was also largest at WTG-01, that this pile is sufficiently protected with values for the average layer thickness and minimum layer thickness all above the all-pile average values.

2.2 Selected examples

2.2.1 WTG-08: largest average bed level change between 2008 and 2009

According to Table 2.1, the largest average bed level change within two pile diameters from the pile centre is only 0.06m and occurred at WTG-8. In order to investigate whether this bed level change is mainly caused by a small error in the correction for the vertical tidal level at the moment of surveying, we compare this value with the average bed level change within one pile diameter from the pile centre. Within one pile diameter, the average bed level change appears to be slightly larger (0.11m). See also Appendix A.08 and B.08 for a comparison

(17)

between the surveys of 2008 and 2009 (App. A) and a comparison between the initial seabed survey of 2006 and the survey of 2009 (App. B).

To investigate the deformation of the scour protection around WTG-08, we compared the survey of 12 October 2006 just after installation of the armour layer (SU05) with the survey of May 2009, see Figure 2.1. It can be observed that since armour installation, the scour protection has been smoothened by wave action and that the area just southeast to west of the pile (in clockwise direction) experienced some deformation. This deformation in this area is approximately 0.5m on average close to the pile and occurred mainly during the storm of 1 November 2006.

Figure 2.1 Bed level around WTG-08 just after installation of armour material (left) and in May 2009 (right)

The average layer thickness at WTG-08 however is still 2.10m within one diameter from the pile centre (ranking 22 out of 36 WTGs), with a minimum protection thickness of 1.50m (ranking 29/36). And within two diameters from the pile centre a surplus of protection material is present with an average layer thickness of 2.17m (ranking 7/36), which is an indication that the extent of the scour protection is good compared to the other WTGs.

In Figure 2.2, the present armour layer thickness is plotted, by subtracting the survey just after installation of filter material (10 May 2006) from the 2009-survey. Note that no additional armour was installed at WTG-08 in 2007. The left plot shows the bed level difference between the surveys in top view. The blue colours represent rising of the seabed (negative values), which in this case represents the placement of armour. The red colours represent a lowering of the seabed (positive values), which represents the development of edge scour since the installation of the filter layer (and later also the pile and armour layer).

The right plot shows the bed level change, averaged over 8 directional sectors, so-called “pie pieces”. For example, “Ray 23°” represents the area located between 0°N and 45°N, so the area located NNE of the pile centre. Negative values mean an upward movement of the bed level, in this case caused by installation of the armour layer. By averaging over pie parts, it can be seen whether the average protection is sufficient in all directional sectors around the pile perimeter and if significant changes in deformation patterns occur around the pile.

(18)

According to Figure 2.2, most protection material is present in the ENE-section (“Ray 68°”), the maximum thickness of the armour layer, averaged in the ENE-pie piece, is about 2.2m and is located at a distance of approximately 5.5m from the pile centre.

This figure also shows that the armour layer close to the pile is 1.4m on average (bold black line “complete”) and still over 1m in the WSW-, SSW- and SSE- pie piece. Furthermore, it was calculated that the smallest armour layer thickness is approximately 0.92m (at 3.0m from the pile centre at an angle of 222°N).

Figure 2.2 Layer thickness of armour material, in top view (left) and averaged along rays (right)

For WTG-08 we can conclude that the thickness of scour protection is still sufficient, that the shape of the scour protection is most likely caused by deformation during storms and that the majority of this deformation probably already occurred during the storm of 1 November 2006.

2.2.2 WTG-10: largest local bed level change between 2008 and 2009

Although the average bed level change in the period May 2008 - May 2009 was largest at WTG-08, the largest local change occurred at WTG-10 (see App. A.10). At a distance of 2.9m from the pile centre and an angle of 239°N, the top level of the protection dropped 0.52m. Also the maximum bed level drop, averaged over a “45°-pie piece”, occurred at WTG-10: the WSW-pie piece (ray 248°) within a distance of one diameter from the pile centre dropped 0.24m on average, see the grey line in the right plot of Figure 2.3.

(19)

To investigate whether these observations are not caused by local erroneous values in the survey data, we also compared the survey of 2009 with the survey of 14 August 2007 (just after additional armour was installed), because the distribution of newly dumped armour material may be uneven, which causes redistribution of armour material towards a more stable shape. In the left plot of Figure 2.4 the bed level difference between 2007 and 2009 is plotted.

Figure 2.4 Bed level change in the period 14 August 2007 and 15 May 2009

This figure shows that the WSW-pie piece experiences no significant deformation. The apparent large local bed level change is most likely caused by an error in the 2008-survey of WTG-10. This is confirmed by comparison of the 2007- and 2008-surveys, which showed a rise of the bed level at this location.

It is, however, peculiar that the protection seems to have increased since the additional armour dump in 2007 (especially in ray 338°) and shows relatively bumpy cross-sections in the right plot. It is not known to Deltares whether additional repair works, possible related to cable protection, were executed. Just outside the scour protection in the SSpart and W-part some increase in bed level can be deduced from the light blue colours, which could have the shape of block mattresses, installed to protect the electricity cables.

Finally, we investigated the present state of the scour protection by subtracting the initial bathymetry from the present bathymetry, see Figure 2.5. The shape of the scour protection is according to expectations: in the dynamic zone between 0.5-1.0D from the pile centre the protection level is somewhat smaller compared to the zone between 1.0-1.5D.

Furthermore, there is a surplus of armour material in the WNW- and NNW-pie pieces. The rather strange pattern of filter material in the NE-section is discussed in more detail in Paragraph 3.3.

(20)

Figure 2.5 Present state of the scour protection around WTG-10 in 2009

2.2.3 WTG-24: smallest average layer thickness within 1D in 2009

The smallest average layer thickness within 1D can be found at WTG-24 and is 1.92m, see Appendix B.24. The 68°-pie piece pays a large contribution to this small layer thickness with an average layer thickness of 1.68m. This weaker spot was already noticed in 2007, when additional armour was dumped. Unfortunately, the majority of the armour was dumped a little too far from the pile, see the upper right graph of Figure 2.6. As a consequence, the weaker spot was not completely repaired. Since 2007 negligible deformation of the scour protection has occurred at WTG-24.

(21)

In Figure 2.7, it can be observed that the initially installed scour protection in 2006 (filter+armour) already showed less surplus material in the eastern part of the scour protection. So after settlement of the protection material and deformation towards the “dynamic equilibrium shape”, the “63°-pie piece” developed a small layer thickness close to the pile. Although the repair of the weaker spot did not fully succeed in 2007, a surplus of material is now present between 1D and 2D from the pile centre. Therefore, it is more likely that during a severe storm redistribution of this heap of armour material will cause infilling of the weaker spot than that a further lowering of the weak spot will occur. Occurrence of a severe storm will be a good opportunity to verify this hypothesis. So, the present scour protection is fit for purpose, because the weakest spot is surrounded by a surplus of armour material.

Figure 2.7 Comparison between total dumped protection until July 2007(left) and state of scour protection in 2006 (right) around WTG-24

2.2.4 WTG-15: smallest average layer thickness within 2D in 2009

The smallest layer thickness within 2D in 2009 is 1.83m and is located around WTG-15, see Appendix B.15. The two right plots show that the scour protection only has a somewhat weaker spot in the 68°-pie piece with an average thickness of 1.69m (within 2D).

If we look at the total dumped scour protection around WTG-15 (114m3 within 1D and 527m3

within 2D) in Figure 2.8, it appears that both the average thickness of the protection and the

extent are smaller than the average values: the average protection volumes are 123m3 (1D)

and 584m3(2D) and only the western side of WTG-15 is nicely protected to an extent of 2D.

The right plot in Figure 2.8 shows that the deformation of the scour protection is not dramatic.

Because of deformation at the SSW-side of the pile, some additional armour (7.5m3 within 1D

and 15.6m3 within 2D) was dumped in 2007 at this location. This explains the small surplus in

the left plot and the large bed level difference in the right plot at this location.

WTG-15 thus is sufficiently protected and the unfavourable average layer thickness is merely a combination of an overall small dumped volume and a relatively small extent at the N-, E- and S-side of the pile and not caused by excessive deformation of the protection.

(22)

Figure 2.8 Total dumped scour protection around WTG-15 and the changes until 2009

2.2.5 WTG-13: smallest average layer thickness within 1D

The smallest average layer thickness within 1D is 1.20m and can be found around WTG-13. This point is located at 2.5m from the pile centre at 106°N. The average protection thickness of “113°-pie piece” within 1D is 1.68m.

In 2008, some questions were raised on the possible worse quality of the filter layer around WTG-13, which could have caused washing out of seabed sand through the pores of the armour material. These observations were based on comparison with the initial seabed before pile installation. However, the situation at WTG-13 is somewhat different from the other WTGs. While at all other WTGs first the filter layer was installed and then the pile was drilled, this occurred in opposite order at WTG-13: the initial seabed was surveyed at 23 March 2006, the pile was drilled at 17 April 2006, seabed scour up to 1-1.5m just SSE of the pile occurred until 22 April 2006, when another survey was executed. The filter material was dumped 8 days later at 30 April 2006 and the “filter-out”-survey was executed at 6 May 2006. Figure 2.9 shows that the filter material was dumped inside already developed scour holes, see upper left graph (also in Appendix B.13). The upper right graph, representing the layer thickness of dumped filter material, implies that everywhere a layer thickness of 0.5m was installed. However, the upper left graph shows that such a significant scour pattern can occur within a time frame of only 5 days (i.e. the time between pile installation and this survey). This means that between 22 April (in-survey filter) and 30 April 2006 (filter installation) also significant bed level changes could have occurred. As long as scour depths were increasing in this period, the filter layer thickness would actually be larger compared to the upper right graph.

However, during wave-dominated conditions or less severe current conditions backfilling of the scour hole can occur. Especially hydrodynamic conditions with high sediment mobility, such as storm conditions, cause relatively fast backfilling. If between 22 and 30 April 2006 the local deep spot at the SE-side of the pile was backfilled with only a few dm, the actual thickness of the filter layer could be minimal (or even “negative”, which means that more backfilling has occurred between 22 and 30 April than filter material was supposedly dumped).

(23)

Figure 2.9 Installed filter layer in already developed scour hole around WTG-13; (upper left) seabed bathymetry before filter installation; (lower left) bed level after filter installation; (upper right) difference between two left plots, indicating installed filter layer; (lower right) cross-sections along 45°-rays and complete protection

To investigate this hypothesis, we compared the total dumped protection around WTG-13 with the absolute bed level change until 2009, see Figure 2.10. At WTG-13 no additional armour was dumped in 2007, so the total dumped protection only consists of filter and armour that was dumped in 2006. Although the drop of the protection in the 113°-pie piece is approximately 0.5m close to the pile, this value is not excessive at all and, moreover, comparable to the other WTGs. This is an indication that the filter layer has no weak spots. So, although the protection just east from WTG-13 appeared to have a weak spot, this can be explained by the different construction method and the fact that scour protection is dumped in the already developed scour holes. Nevertheless, it is remarked that the “in-survey” should be executed only hours (or a few days at most) before the scour protection is installed, since hydrodynamic conditions with high sediment mobility will cause a relatively fast (partial) backfilling of the scour holes, resulting in a too positive estimate of the installed volume of protection material.

If the time between in-survey and installation of filter material is too long or it is known that a storm occurred in between, it is recommended to completely fill developed scour holes with filter material, at least to the level of the initial seabed, and place the armour layer on top of it. In this case, it is certain that the thickness of the filter layer is sufficient all around the pile.

(24)

Figure 2.10 Comparison between total dumped scour protection and present state of scour protection at WTG-13

2.2.6 WTG-33: smallest local layer thickness within 2D in 2009

The smallest layer thickness within 2D from the pile centre in 2009 is located around WTG-33, just as in 2008. While the smallest layer thickness was 0.45m (at a distance of 9.1m at an angle of 10°N) in 2008, the minimum layer thickness is approximately the same in 2009 (0.52m). It can be seen in Figure 2.11 that the scour protection in the two most northern 45°-pie 45°-pieces (ray 23° and ray 338°) has a significantly smaller extent.

(25)

In the 2008-memo, the reason for the local change of the bathymetry north-east of the pile was not completely clear. Possible explanations comprised passing bed forms, variations in seabed sediment characteristics (cohesiveness of material) or maybe nearby construction activities. Although it was assessed that WTG-33 was still well protected, it was recommended to pay special attention to seabed changes in the future, especially north-northeast of the WTG.

To further investigate the situation at WTG-33, we calculated all installed scour protection at this WTG, which at this pile only consisted of filter and armour material dumped in 2006, since no additional armour was dumped in 2007, see the upper left graph in Figure 2.12. The filter layer extended on general up to 2.5D from the pile centre, which is comparable to the other WTGs. The present state of the scour protection (lower left graph) at WTG-33 clearly shows the deformation of the scour protection: 1) the overall layer thickness is reduced due to settlement of the stone layers and underlying seabed and (possibly) spreading out of protection material over the surrounding seabed; 2) the protection transformed towards a dynamic profile (armour stones are transported from the dynamic zone (0.5-1.0D) towards the area between 1.0 and 2.0D); and 3) the extent of the scour protection in the north is significantly reduced.

(26)

Note that especially the third deformation pattern is deviating from the other WTGs, see Figure 2.13, which illustrated cross-sections of the NNE-pie piece (ray 23°). This figure shows that the installed filter layer (red lines) and the total installed protection (blue lines) are quite similar for WTG-33 (solid lines) and the average WTG (dashed lines). There is, however, a small difference. The total protection at WTG-33 already shows some deformation between 11 and 22m from the pile centre. The filter material that was present within 11 and 15m from the pile centre is transported to the area 15-22m from the pile centre. This deformation probably occurred during the 1 November 2006-storm, since the “armour out survey” was executed just after this storm. This statement is confirmed by the fact that the deformation close to the pile is also somewhat more distinct at WTG-33 (compare the blue lines between 3 and 6m from the pile centre).

The present state, represented by the black lines, shows a rather large difference between WTG-33 and the average pile. Both the armour layer between 6 and 11m from the pile centre and the part of the filter layer that was not protected by armour stones (roughly between 11 and 14m) show a considerable lowering. If we assume that the steepness of the profile is an indication of the composition, it seems that the edge of the armour layer moved from 11m to 8m from the pile centre.

-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 0 5 10 15 20 25 30

Distance from pile centre [m]

V

ertical level [m]

filter at WTG-33 average filter

total protection at WTG-33 average total protection present state at WTG-33 average present state WTG

Figure 2.13 Cross-sectional profiles for “23°-pie piece” (NNE) for both WTG-33 (solid lines) and the average WTG (dashed lines)

In the present state, filter material seems to be present between 8 and 12m from the pile centre, since this part of the profile has a more gentle slope compared to the armour layer. The hump in the profile of the “present state at WTG-33” suggests that the heap of filter material, initially located at 15-22m from the pile centre, is still present, though it is spread out somewhat. In order to further study the retreat of the scour protection in this “pie piece”, the development in time is plotted in Figure 2.14.

(27)

1200160-002-HYE-0001, Version 1.0, 3 September 2009, final -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 0 5 10 15 20 25 30

V er ti cal l evel [m] November 2006 June 2007 May 2008 May 2009 WTG

Figure 2.14 Development in time of cross-sectional profiles for “23°-pie piece” (NNE) for WTG-33 between 2006 and 2009

This figure illustrates that the deformation at the edge of the scour protection mainly occurred between 2006 and 2008. The deformation between November 2006 and June 2007 (blue and green line) seems to be primarily caused by settlement and spreading out of protection material, which is confirmed by the almost constant drop of the scour protection between the pile and 12m from the pile centre. The deformation between June 2007 and May 2008 (green and red line) is more located at the edge of the protection and is possibly caused by redistribution of protection material at the edge of the protection towards the edge scour hole, since there is hardly any deformation close to the pile, what you would expect during severe storm events.

Furthermore, it can be concluded that the edge scour development mainly took place between 2006 and 2008. It is very likely that the redistribution of protection material caused an overall better protection against edge scour development, since the edge scour development within about 22m from the pile centre appears to have stopped after May 2008. Only farther away from the pile centre an increase of the edge scour depth can be observed. The question remains why deformation at the edge of the scour protection at WTG-33 is so much more distinct compared to the other WTGs. If we consider the bathymetry (see Figure 1.1), we observe that WTG-33 is one of the more shallowly located WTGs at the offshore side of the most shallow sandbar. At a more shallow location, the scour protection is more easily redistributed by the wave action. It is also obvious that the edge scour development at this WTG is relatively severe, see Figure 2.15. The total eroded volume in pie piece 23°N (NNE)

within 3-6D from the pile centre is 153m3 at WTG-33 against 116m3 at the average WTG.

Further research is needed to investigate the edge scour patterns throughout OWEZ against the local hydrodynamics.

(28)

Figure 2.15 Cross-sectional profiles for WTG-33 and the average WTG between 2006 and 2007

Concluding, it appears, based on this year’s survey data, that the scour protection and edge scour hole around WTG-33 reached a more or less stable shape and WTG-33 thus remains sufficiently protected. In the coming years we will closely follow if the retreat of the scour protection indeed has stopped around WTG-33. It is also possible that this retreat is mainly driven by severe storm events which are able to redistribute protection material. Therefore, the hydrodynamics between May 2008 and May 2009 still need to be checked.

2.2.7 WTG-07: worst distribution of scour protection over “45°-pie pieces”

The final WTG that will be considered is WTG-07, see Appendix B.07. This WTG is characterized by the worst distribution of scour protection material over the “45°-pie pieces”, both within 1D and within 2D, see Figure 2.16.

The upper left graph shows that a significant larger volume is dumped in the south-western

quadrant (186 m3 within 2D and 42m3 within 1D) compared to the north-eastern quadrant

(118 m3 within 2D and 25m3 within 1D). Fortunately, we can consider the volume in the

south-western quadrant to be a surplus: the volume in the north-eastern quadrant is more than sufficient.

The upper right graph shows that the area with a large volume of scour protection also showed a larger absolute “loss of volume” between 2006 and 2009. However, if we consider relative values, we can conclude that the relative loss is 10-12% for all “pie pieces” around the WTG. This means that 10-12% of the total dumped volume within 2D is lost by the combination of settlement and transport out of the scour protection.

(29)

Figure 2.16 Difference between total installed protection and present state of the scour protection at WTG-07

2.3 Electricity cables

At numerous piles, elevation of the seabed can be observed at the southern and western edge of the scour protection (see for instance App. A.02, A.05, A.10, A.14. Since the diameter of the cables is only 0.15m, the seabed elevation is most likely caused by block mattresses, possibly in combination with deposited sediment. Figure 2.17 shows the bathymetries of August 2007, May 2008 and May 2009 around WTG-05. It can be observed that between August 2007 and May 2008 rectangular-shaped cable protection is installed at the SSW- and WNW-side of the WTG.

(30)

The difference plot between the 2007- and 2008-survey (Figure 2.18) illustrates that 1) there were only minor changes in the scour protection around WTG-05, 2) in general, the seabed lowered around the protection due to edge scour and 3) the ‘seabed’ rose with up to 0.3m at the locations of the cables.

Figure 2.18 Difference between bathymetries of August 2007 and May 2009 at WTG-05

The characteristics of the additional cable protection (at which WTGs, location around WTG, extent from the pile, width of protection) are not known to Deltares. However, this additional protection might be important for the estimation of edge scour depths and patterns (see Chapter 3).

Whether the protection of the electricity cables is sufficient or whether local scour around the mattresses occurs can not be assessed on the basis of the available surveys. ROV surveys can provide insight in the interaction between mattress, cable, seabed sediment and adjacent scour protection.

2.4 Average deformation of scour protection

Because the deformation pattern around an individual WTG is very much dependent on the shape and volume of installed protection, we also translated all 36 surveys to one coordinate system and calculated z-levels relative to the initial seabed. After averaging these translated surveys, Figure 2.19 was processed.

The upper left graph shows that the scour protection around an average pile was quite evenly distributed around the pile. Please note that the “armour out survey” (SU05) was executed before the severe storm of 1 November 2006 for 14 WTGs and after 1 November 2006 for 22 WTGs. This means that some storm-induced deformation of the scour protection is already incorporated in the upper left graph, which can be noticed by the small local lowering of the protection around the pile. The upper right graph shows some generally lower protection levels in 2007.

(31)

Then at 20 WTGs additional armour was dumped and the lower left graph shows the average situation in May 2008. A dynamic zone within 0.5-1.0D is clearly visible as well as the ridge of armour stones between 1.0-1.5D. The dynamic profile appears to be somewhat more pronounced in the western and northern part with slightly lower layer thicknesses close to the pile and a larger extent of the lowered area close to the pile. The southern and eastern side of the pile appear to be more protected against storm attack. Negligible changes occurred between May 2008 and May 2009 (lower right graph). The average effect of additional cable protection at some of the piles is also visible in Figure 2.19 at the southern edge of the scour protection in the two lower graphs (2008 and 2009).

Figure 2.19 Bed level change relative to initial seabed, averaged over all 36 WTGs, after (upper left) 122 days on average since pile installation, (upper right) 371 days since pile installation, (lower left) 699 days since pile installation and (lower right) 1057 days since pile installation

(32)

The majority of the loss of protection volume so far occurred between November 2006 and June 2007, so within 1 year since pile installation. Possible explanations for this loss are: A. the soil underneath the scour protection settles due to the weight of the protection. B. the soil close to the pile is lowered due to pile driving.

C. seabed sediment is washed out through the pores of the filter layer and armour layer. D. the filter layer is compacted, among others, by cyclic wave loading.

E. the armour is compacted, among others, by cyclic wave loading.

F. the filter and armour layer are mixed (armour stones sink in the armour layer and filter material fills the pores of the armour rocks.

G. filter material is washed out through the pores of the armour layer.

H. filter material that is not covered with armour stones is spread out over the seabed during moderate to severe wave conditions.

I. armour material is spread out over the surrounding seabed during severe wave conditions.

J. the armour layer is transformed to its ‘dynamic shape’ with a lower area close to the pile and a ring of stones at a somewhat larger distance.

K. due to edge scour development filter material rolls into the edge scour hole and becomes trapped.

L. due to edge scour development and retreat of the filter layer armour material rolls into the edge scour hole and becomes trapped.

The above mentioned mechanisms that can explain the loss of protection volume assume that the protection material itself is of good quality. No degradation of the material is taken into account. Furthermore, it is assumed that the pile itself is hardly moving during severe conditions and has no effect on the scour protection.

It is important to understand that many of the 12 possible mechanisms occur simultaneously, especially just after installation of the scour protection, which makes distinction between different processes difficult. Therefore, an attempt is made to classify the mechanisms both in time and in hydrodynamic condition, see Table 2.3. The darkness of the grey colour is a measure for the estimated correlation.

id filter installation pile drilling armour installation edge scour holes anytime normal conditions causing sediment transport moderate wave conditions causing filter movement severe wave conditions causing armour movement A B C D E F G H I J K L

time: possible from the moment of hydrodynamics: possible during

(33)

Then it was assessed at which location these mechanisms can probably best be observed, see Table 2.4. It is remarked that in some cases, bathymetrical surveys alone are not sufficient to distinguish between the different processes and visual observations (e.g. by ROV) are required. Again the darkness of the grey colour is a measure for the estimated correlation between mechanism and location.

id close to the pile entire armour layer transition between armour

and filter layer

exposed filter layer outside armour layer

transition between filter layer and surrounding seabed surrounding seabed only in edge scour holes A B C D E F G H I J K L

Table 2.4 Probable locations where mechanisms for volume loss can best be observed

Occurrence of different mechanisms can have different negative consequences for the pile fixation level, the quality of the filter layer (the part that is covered by armour as well as the part that is exposed) and the quality of the armour layer, see Table 2.5. It can be observed that mechanisms A, B and C have a direct effect on the pile fixation level, while they probably have no negative effect on the scour protection. This effect may even be positive, because the obstruction height of the scour protection will become smaller, which reduces the hydrodynamic load on the protection.

Most mechanisms have an indirect effect on the pile fixation level, since a reduction of the quality of the scour protection increases the risk on failure of this protection. Only then, the pile fixation level might be reduced. Mechanisms K and L can have a negative effect on the pile fixation level, but this will probably take quite some time: first an edge scour hole has to develop, then some storm events are required to transport protection material into the scour holes. Nevertheless, with an anticipated lifetime of 25 year these mechanisms may turn out to be important.

id pile fixation level filter layer thickness underneath armour

exposed filter

layer thickness filter layer extent

armour layer thickness armour layer extent A B C D E F G H I J K L

Table 2.5 Possible negative consequences for different protection dimensions of all mechanisms for volume loss

(34)

In particular, mechanisms A, B, D, E and F will primarily occur in the first year after installation of pile and scour protection. Of these 6 mechanisms, the latter three will be enhanced by moderate to severe wave events. All mechanisms except A and B can occur during severe wave conditions. Of these 10 mechanisms, only K and L are not likely to occur within the first months after armour installations. Unfortunately, a rather severe storm occurred just after armour installation (i.e. 1 November 2006 storm), which complicates an accurate estimation of the magnitude of the mentioned mechanisms.

The only mechanism that can be studied without a detailed analysis of the hydrodynamics is mechanism B, since this process occurs at a distinct moment in time at a distinct location, i.e. at the moment when the pile is drilled and close to the pile. An estimate of this effect can be obtained from the comparison between surveys SU04 (after filter installation and before pile drilling) and SU02 (after filter installation and pile drilling). In Figure 2.20 the difference between the surveys just before and just after drilling of WTG-19 is plotted. This figure shows that within 2m from the pile, the bed level is dropped with 0.2-0.4m with an average value of 0.33m at the pile perimeter. Similar values are observed at the other WTGs where survey SU04 was executed: the range of the bed level drop close to the pile is about 0.2-0.5m. The lowering extends typically to 2m from the pile perimeter and the side slope of this local deepening is about 1:4.

Figure 2.20 Bed level change between 9 June 2006 and 10 July 2006 at WTG-19; the pile was drilled at 3 July 2006

Although we have to be careful with a quantification of the other mechanisms, we can make some estimations, based on the survey data, averaged over all WTGs. The correlation with occurred hydrodynamics has to be studied at a later stage.

Therefore, we translated the lower right graph of Figure 2.19 into rays averaged over the complete scour protection and over 45°-pie pieces. 4 of the in total 8 pie pieces and the “complete 360°” are plotted in Figure 2.21 (upper lines). In order to study the average deformation we also calculated the total installed scour protection (filter+armour+additional armour), averaged over all WTGs. These lines are also plotted in Figure 2.21 (lower lines). Please note that already some influence of storm-induced deformation is visible, since part of the “armour-out” surveys was executed after occurrence of the 1 November 2006 storm.

(35)

1200160-002-HYE-0001, Version 1.0, 3 September 2009, final -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 2 4 6 8 10 12 14 16 18 20

E le vatio n wrt in itia l s e abe d [ m ] complete 360°

pie piece-1 (NNE, 23°N) pie piece-3 (ESE, 113°N) pie piece-5 (SSW, 213°N) pie piece-7 (WNW, 293°N) avrg WTG drop of protection installed protection state of protection in 2009

Figure 2.21 Elevation of protection, averaged over all WTGs, for the complete 360°-area and 4 selected 45°-pie 45°-pieces (NNE, ESE, SSW, WNW); upper lines are based on surveys of May 2009, lower lines are based on the sum of the three installed protections (filter in 2006 + armour in 2006 + additional armour in 2007), middle lines represent the difference between the initially installed protection and the present situation in 2009

The difference between the present state of the scour protection and the total installed protection is the average change of the scour protection. This figure clearly shows that:

1 The region where armour material was installed (roughly between 2.3 and 11m)

experienced a relatively constant loss of around 25-30cm.

2 The dynamic zone between 2.3-4.6m (or 0.5-1.0D) experienced only a slightly larger

loss with an average drop of the protection of 39cm and values up to 50cm close to the pile.

3 Based on visual inspection there appears to be no correlation between the installed

height of scour protection and the absolute loss of protection.

4 The region where only filter material was installed (roughly between 11 and 15m)

showed hardly any loss of material for almost all pie pieces.

5 The 23°-pie piece (NNE-side of WTG) experienced an average drop of 15-20cm in the

region where only filter material was installed;

6 The loss of filter material in this pie piece seems strongly correlated to the measured

edge scour (see region between 15-20m from the pile centre)

7 The initial rather sharp boundary between filter and armour material (see the point of

(36)

Note that these findings are all based on averaged data over all WTGs. The benefit is that some trends are more easily visible. However, it should be realized that averaged values tell nothing about minimum and maximum values and also do not account for differences between the individual piles. In order to quantify possible relations between calculated deformation parameters (such as average level of protection around an individual WTG, minimum remaining layer thickness, maximum drop of protection level) and parameters that are related to the load on and strength of the scour protection (such as local water depth, location inside OWEZ, initially installed total volume of scour protection, initially installed volume of filter material etc.), we compared multiple parameters. These investigations were executed for different effective areas (e.g. within 1D or 2D from the pile centre and for 45°-pie pieces or for the complete 360°-area). However, no clear relations could be established (so far):

1 There is no clear relation between deformation and local water depth;

2 There is no clear relation between the average thickness of installed filter material within

a certain pie piece and the absolute or relative drop of the scour protection. This might be an indication that mechanisms C (seabed loss through filter) is not dominant. The compaction of the filter layer (mechanism D), which is likely to be correlated to the filter layer thickness, is probably also not prominent.

3 The reduction of the height of the protection is fairly similar for all 45°-pie pieces (both

within 1D and 2D), which is an indication that scour protection, located at the side of direct storm attack will not experience significantly more deformation than scour protection located at the lee of the pile. This behaviour can best be observed in the averaged bathymetries in Figure 2.19: the dynamic zone extends all around the pile.

There is no clear relation between the relative obstruction height of the protection (hobs/hw)

and the deformation.

A possible explanation for the lack of clear relations could be the fact that the mentioned mechanisms all occurred primarily within the first year after installation of the scour protection, since this year had a relatively severe wave climate. Nature, of course, can not be controlled, but for the distinction between the different processes, it would be better if the rather severe 1 November 2006-storm occurred one year later instead of in the middle of the “armour-out surveys”.

The following parameters should be investigated in the future in order to try to improve the relations between deformation and installed protection:

1 vertical difference between lowest point of ray, averaged over a pie piece, within 1D and

highest point, averaged over a pie piece, within 2D. This difference is a measure for the profile deformation and might be related to the wave action and thus the water depth;

2 distribution of wave heights and wave periods inside OWEZ;

3 large-scale morphological patterns (mega ripples, sand waves) which in the long term

can affect the stability of the scour protection.

4 current velocity distribution in relation to the large-scale bathymetry inside OWEZ, taking

into account the above mentioned morphological patterns;

5 translation of as-built scour protection of each WTG (with irregular pattern of armour and

filter stones) into one representative obstruction height and extent of both armour and filter layer. This ‘simplified’ representation of the real situation is necessary for the comparison with laboratory results and the derivation of a generic formula for the prediction of local scour depth inside a dynamically stable scour protection.

(37)

6 division of all 36 WTGs into subgroups that are more comparable, such as:

a WTGs that had their “armour-out” survey before or after 1 November 2006

b WTGs where additional armour was dumped in 2007 and WTGs without additional

armour

c classification of WTGs in a few subgroups based on water depth

d classification of WTGs in a few subgroups based on edge scour depth

e Removal of WTGs with a ‘different history’, e.g. WTG-13, WTG-10 and WTG-22

2.5 Comparison with laboratory tests

Although at this stage when no clear relations between deformation on the one hand and installed volume, local water depth and location of the pile inside OWEZ on the other hand could be established, it is useful to compare the field observations with the laboratory observations.

The fact that the initial state of the scour protection in the laboratory is much more evenly distributed makes a direct comparison with the field case, where initially small heaps of stones are present, difficult. However, if we look at the schematization in Figure 2.22 (presented in Delft Hydraulics, 2005), we see that the shape of the “profile after test”, represented by the red dashed line, is very similar to the observed profiles in the field. Both the dynamic zone and the overall lowering of the scour protection are clearly visible.

Figure 2.22 Schematization of local scour inside scour protection and displaced armour as observed in laboratory tests [source: Figure 10 in Delft Hydraulics, 2005]

In the laboratory tests, two sizes of armour material were investigated (Dn50=0.25m and

Dn50=0.40m) and various extents of the armour layer (two layers with an extent of 3, 4 and 5

times the pile diameter as well as one layer with an extent of 3D on top of another layer with an extent of 5D; all extents measured between two opposite points located halfway the outer slope). The average protection that was installed in the field can be described as follows: - average extent of filter layer = 6D;

- average extent of armour layer = 4D;

(38)

If we consider all measured (non-dimensional) test results in Figure 2.23, the average scour protection in OWEZ is most similar to the black “ -symbols”. This figure also shows quite some scatter, because some variables are not normalized in this graph, such as layer thickness and extent of scour protection. Nevertheless, some scatter is also present in these results obtained under controlled conditions with almost perfectly installed scour protections.

For the design condition for Dn50=0.40m, the best estimate of the predicted local scour depth

will be around 0.1*Dpile, which is about 0.5m, but for smaller extents this value may rise to

1.0m (upper boundary). According to Figure 2.21 the average drop of the protection close to the pile is also approximately 0.5m, which would imply that already a storm event comparable to the design condition occurred.

This statement is confirmed by Delft Hydraulics (2008b) in which it was concluded that with respect to the scour protection “the relative mobility at the disturbed seabed (close to the pile) was close to the design value during the 1 November 2006-storm, but due to limited available knowledge on the relation between KC-number and relative mobility in the vicinity of a structure, it can not be firmly stated that regarding the scour protection the design condition was exceeded.”

Figure 2.23 Relation between the mobility parameter and the maximum non-dimensional scour depth as deduced from laboratory tests [source: Figure 3 in Delft Hydraulics, 2005]

Another parameter that was investigated is the average armour layer thickness reduction, see Figure 2.24. An increasing trend can be observed between the displaced volume and the hydraulic conditions (presented by the dimensionless mobility parameter), although the scatter is relatively high. Also the differences in the displaced volume between the two armour

sizes (Dn50=0.25m and Dn50=0.40m) appeared to be marginal.

It is remarkable that also in the laboratory tests only a very weak correlation could be found between the “slenderness of the protection” (i.e. the obstruction height) and the average volume reduction. In the field, the average reduction of the armour layer varied between 5 and 33% with an average value of 18%. These values correspond nicely to Figure 2.24.

(39)

On the other hand, this also implies that even if we correct the field measurements for 1) the local water depth and 2) the individual initial shapes of the scour protection in terms of extent and layer thickness and then distinguish between I) WTGs with the “armour-out-survey” before and after the 1 November 2006-storm and between II) WTGs with and without installation of additional armour in 2007, we will still have to deal with significant scatter in the scour depth predictions / hindcasts.

The final conclusion to be drawn from this figure is that even under the 1 and 10 year extreme condition (solid markers left of the green dashed line) considerable displacement of armour material can be expected. For the design condition with a return period of 100 year (see the

green dashed line that is valid for Dn50=0.40), the average armour layer reduction is

somewhere between 10 and 30%.

Please note, that the laboratory test results are all based on a single storm event with a certain return period. Cumulative deformation can not be deduced from this figure. The field measurements in OWEZ hopefully will provide information on this cumulative effect when multiple storms occur during the lifetime of 25 year.

Figure 2.24 Relation between mobility and the average armour layer reduction as deduced from laboratory tests [source: Figure 11 in Delft Hydraulics, 2005]; note that the design condition is valid only for Dn50 = 0.40m

(40)

Evaluation of performance of scour protection and edge scour development : Offshore Windpark Egmond aan Zee