Subsea noise modelling of impact piling operations to ... · PDF filehave led to changes in the calculation of propagation with respect to some ... In reality a piling operation will

COMMERCIAL IN CONFIDENCE

COMMERCIAL IN CONFIDENCE

Submitted to: Submitted by:

Trevor Baker Tim Mason Galloper Offshore Wind Farm Subacoustech Environmental Ltd Auckland House Chase Mill Lydiard Fields Winchester Road Great Western Way Bishop’s Waltham Swindon, Wiltshire Hampshire SN5 8ZT SO32 1AH Tel: +44 (0)1793 474109 Tel: +44 (0)1489 892 881 e-mail: [email protected] e-mail: [email protected] website: www.rwe.com website: www.subacoustech.com

Document No. Date Written Approved Distribution

E505 Appx1 13/04/2015 R Barham T Mason Trevor Baker (Galloper)

This report is a controlled document. The Report Documentation Page lists the version number, record of changes, referencing information, abstract and other documentation details.

Subsea noise modelling of impact piling operations to install monopiles at

the Galloper offshore wind farm

R. Barham, T. Mason

13th April 2015

Subacoustech Report No. E505 Appx1

COMMERCIAL IN CONFIDENCE Subsea noise modelling of impact piling operations to install monopiles

at the Galloper offshore wind farm

Subacoustech Environmental Ltd. 2 COMMERCIAL IN CONFIDENCE

1 Introduction

The Galloper Offshore Wind Farm is located off the Suffolk coast and the east coast of England. It lies adjacent to the Greater Gabbard Offshore Wind Farm. Since being consented, the Galloper project has been the subject of a detailed design optimisation, and as a consequence, the maximum pile diameter requires a minor increase, up to 7.5 m in diameter, as well as an increase in the maximum blow energy used to drive it.

Subacoustech Environmental has undertaken an analysis of the proposed changes to pile diameter and maximum blow energy, reported here, and how they potentially impact on the M-Weighted SEL injury criteria from Southall et al (2007).

In addition to using the latest version of the INSPIRE model, a more detailed, realistic ramp up and strike rate has since been applied to the modelling presented in this report. These new assumptions are detailed in the following section.

2 Modelling Approach

2.1 The INSPIRE Model

The INSPIRE model (currently version 3.4.3) is a semi-empirical underwater noise propagation model based around a combination of numerical modelling and actual measured data. It is designed to calculate the propagation of noise in shallow, mixed, coastal water, typical of the coastal conditions around the UK. The model provides estimates of the unweighted peak, peak-to-peak and RMS level of noise as well as various other noise metrics along 180 equally spaced radial transects. The model meets the general requirements set out by the guidance for good practice in underwater sound modelling by Robinson et al, 2014 (although the guidance does not go into detail regarding the requirements of semi-empirical models):

• Depth dependence;

• Sound absorption;

• As the model is semi-empirical, sea conditions (e.g. sediment type, sea temperatures, salinity) are intrinsically taken into account.

It has been validated by comparison with another model (RAMSGeo) and by comparison with further experimental data (e.g. Thompson et al, 2013), which Robinson et al (2014) state provides considerable extra confidence in the model.

The INSPIRE model has been developed using over 50 underwater piling noise measurement datasets; comparing this to the modelling undertaken for the 2011 report, where approximately 20 datasets were incorporated into the model. It is also worth noting that since 2011, more measurements of larger monopiles have been made; for example, in 2011 the predicted noise levels for a 7 m diameter pile were extrapolated using measured data from piles measuring up to 4.9 m in diameter, whereas the current data includes piles in excess of 6.1 m in diameter, which are some of the largest to have been installed for wind turbine foundations around the UK. As measurements of larger foundations have been used, fewer assumptions have had to be made in order to extrapolate the noise impacts of piles measuring over 7 m in diameter.

For each scenario, a criterion level can be specified allowing a contour to be drawn, within which a given effect may occur. These results are then plotted over the bathymetry data so that impact ranges can be clearly visualised and assessed as necessary.

To give a most realistic worst case set of results, all INSPIRE modelling in this report has assumed a tide of +3.9 m above LAT (Lowest Astronomical Tide); this is the same tidal level that was assumed for the previous modelling at Galloper undertaken in 2011.




Since 2011, refinements have been made to the model, particularly for the installation of large monopiles, in light of new data acquired. This has led primarily to changes in the level of noise output at the pile, the ‘source level’. The differences in source level between the past and present modelling, in terms of unweighted peak-to-peak levels are noted in Section 3. Other refinements have led to changes in the calculation of propagation with respect to some metrics. All of these factors bring the predictions made by INSPIRE in line with the growing database of measured data.

2.2 Modelling confidence

In order to provide confidence in the accuracy of the INSPIRE model, comparisons have been made between the outputs from the model and measured data from operations similar to those proposed in this study.

To compare the INSPIRE model against measured data, INSPIRE has been run retrospectively against similar projects for which measurements have been taken. By plotting the estimated propagation from INSPIRE with the measured data points from the survey, the accuracy of the model can be attained.

The graphs in Figure 2-1 and Figure 2-2 give the estimated noise levels with range from the INSPIRE model for impact piling operations to install a 5.33 m diameter monopile and a 6.0 m diameter pile plotted over unweighted peak-to-peak data points from actual measurements. The comparison shows a good, conservative, agreement with the measured data and indicates a high degree of confidence in the INSPIRE modelling output.

Figure 2-1 Comparison between measured data and an estimate using INSPIRE for piling along transect (5.33 m diameter pile)




Figure 2-2 Comparison between measured data and an estimate using INSPIRE for piling along transect (6.0 m diameter pile)

The new INSPIRE noise level fits are somewhat lower when compared with the predictions using INSPIRE from 2011 (see also Figure 2-3 and Figure 2-4 below regarding source levels). However, when comparing the fit to the measured data, it still clearly remains conservative.

Figure 2-3 below shows the complete 50 site source noise level dataset with the new INSPIRE fit. Figure 2-4 shows the fit used for the original Galloper underwater noise assessment modelling.

Figure 2-3 Revised INSPIRE fit including full 50-set dataset of source noise levels with pile size




Figure 2-4 INSPIRE fit of source noise levels with pile size as used for original 2011 underwater noise modelling

It should be noted that the source levels, when used for modelling, are still highly dependent on blow energy used as well as the pile size. The line of best fit will increase or decrease as a consequence of a change in blow energy: an increase in blow energy will lead to an overall increase in noise output (see also Table 2-2 for a description of the effect of changes in blow energy specifically at Galloper).

2.3 INSPIRE Input Parameters

Three modelling positions have been specified from the 2011 modelling for comparison. These are: Locations “A”, “C”, and “G”. The coordinates of these three locations are presented in Table 2-1 and shown graphically in Figure 2-5.

Table 2-1 Coordinates of the three modelling locations at Galloper

Longitude Latitude Location A 51.9927°N 001.9767°E Location C 51.9283°N 002.0465°E Location G 51.7505°N 002.0526°E

Two maximum blow energies have been modelled: blow energies of 2000 kJ or 3000 kJ. For comparative purposes, Table 2-2 shows the predicted source levels for piles using the 2011 modelling for 1100 kJ, and the new modelling using 1100 kJ, 2000 kJ and 3000 kJ single strike.

Table 2-2 Single strike source levels using 2011 and new modelling

7.0 m, 1100 kJ 7.0 m, 1100 kJ 7.5 m, 2000 kJ 7.5 m, 3000 kJ

0

50

100

150

200

250

300

0 1000 2000 3000 4000 5000 6000 7000 8000

Pe

ak

to

pe

ak

So

urc

e L

ev

el

(dB

re

. 1

µP

a @

1m

)

Pile diameter (mm)

Cubiclaw Fit

Measured Data

Predicted Source Level for a 2.5 m diameter pile

Predicted Source Level for a 7.0 m diameter pile




(2011) (new) (new) (new) Source level (re

1 µPa peak to peak) 254 dB 246.8 dB 249.4 dB 251.1 dB

For noise metrics that require Sound Exposure Levels (SELs), values are calculated for multiple pulses over time.

Figure 2-5 Map showing the Galloper site boundary along with the three modelling locations used in this study




The 2011 modelling assumed a short soft start period of 20 minutes followed by the remaining duration at the maximum blow energy. In reality a piling operation will gradually build up over time before reaching maximum blow energy. The ramp up profile that has been used for this modelling is detailed in Table 2-2. All the assumptions made are in accordance with the performance and capacity of the piling hammers proposed for the installation.

Table 2-2 Summary of piling ramp up scenarios modelled for multiple pulse based on a more realistic blow energy profile

Energy (%) 20 % 40 % 60 % 80 % 100 % Energy for a 2000 kJ hammer 400 kJ 800 kJ 1200 kJ 1600 kJ 2000 kJ Energy for a 3000 kJ hammer 600 kJ 1200 kJ 1800 kJ 2400 kJ 3000 kJ

Duration 20 mins (1200s)

28.3 mins (1700s)

28.3 mins (1700s)

28.3 mins (1700s)

105 mins (6300s)

Strike rate 20 strikes per min

30 strikes per min

45 strikes per min

45 strikes per min

32 strikes per min

Also modified from the previous modelling is the animal fleeing speed; the 2011 modelling assumed a flee speed of 1 m/s, however data from Otani et al (2000)1 showed an average swimming speed of a harbour porpoise to be around 1.5 m/s, which is what has been used as a fleeing speed for this modelling. This change in speed alone can make a substantial difference to exposures as an animal will be much further from the noise as it increases in level. It is also likely to be slower than a real-life fleeing speed.

INSPIRE results

Table 2-3 and Table 2-4, below, show the modelled mean impact ranges for the new modelling parameters using the M-Weighted SEL injury criteria from Southall et al (2007)2 for the high frequency cetaceans and pinnipeds (in water) hearing groups.

In each case the predicted impact ranges for the new 7.5 m modelling are reduced from those predicted in 2011 for a 7 m pile using the old ramp up assumptions and revised levels.

Table 2-3 Summary of the mean M-Weighted SEL impact ranges using the injury criteria from Southall, et al (2007), for high frequency cetaceans (including harbour porpoise) at the three modelling locations using the new ramp up parameters

High Frequency Cetaceans

(198 re 1 µPa2s (Mhf))

7.5 m, 2000 kJ (new ramp-up)


7.0 m, 1100 kJ (2011)

Difference between 7 m, 1100kJ and

7.5 m, 3000kJ Location A < 100 m < 100 m 340 m - 0.24 km Location C < 100 m < 100 m 320 m - 0.22 km Location G < 100 m < 100 m 450 m - 0.35 km

Table 2-4 Summary of the mean M-Weighted SEL impact ranges using the injury criteria from Southall, et al (2007), for pinnipeds (in water) (including common seal) at the three modelling locations using the new ramp up parameters

Pinnipeds (in water)



7.0 m, 1100 kJ (2011)

Difference between 7 m,

1 Otani S, Naito T, Kato A, and Kawamura A. (2000). Diving behaviour and swimming speed of a free-

ranging harbour porpoise (Phocoena phocoena). Marine Mammal Science, Volume 16, Issue 4, pp 811-814, October 2000. 2 Southall B L, Bowles A E, Ellison W T, Finneran J J, Gentry R L, Green C R, Kastak D, Ketten D R, Miller

J H, Nachtigall P E, Richardson W J, Thomas J A, and Tyack P L. (2007). Marine mammal noise exposure criteria: initial scientific recommendations. Aquatic Mammals Vol. 33, No. 4, 411-521.




(186 re 1 µPa2s

(Mpw)) 1100kJ and

7.5 m, 3000kJ Location A 11 km 14 km 15 km - 1 km Location C 12 km 15 km 16 km - 1 km Location G 14 km 17 km 18 km - 1 km

Figure 2-6 to Figure 2-11 show contour plots of the impact ranges summarised in the preceding tables for the criteria proposed by Southall et al (2007). The solid lines represent the new modelling using a pile diameter of 7.5 m and the new ramp up parameters, and the dotted lines represent the 2011 modelling using the smaller, 7.0 m, diameter pile.

It should be noted that due to the small predicted impact ranges for the High Frequency Cetacean hearing group, the relevant contours are too small to be seen clearly at this scale.




Figure 2-6 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position A, comparing a 7.5 m diameter pile installed with a maximum blow energy of

2000 kJ and the new ramp-up assumptions, with the 2011 modelling




Figure 2-7 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position A, comparing a 7.5 m diameter pile installed with a maximum blow energy of





Figure 2-8 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position C, comparing a 7.5 m diameter pile installed with a maximum blow energy of





Figure 2-9 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position C, comparing a 7.5 m diameter pile installed with a maximum blow energy of





Figure 2-10 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position G, comparing a 7.5 m diameter pile installed with a maximum blow

energy of 2000 kJ and the new ramp-up assumptions, with the 2011 modelling




Figure 2-11 Contour plot showing the impact ranges, using the Southall et al (2007) criteria at Position G, comparing a 7.5 m diameter pile installed with a maximum blow

energy of 3000 kJ and the new ramp-up assumptions, with the 2011 modelling

Subsea noise modelling of impact piling operations to ... · PDF filehave led to changes in the calculation of propagation with respect to some ... In reality a piling operation will

Documents