ABSTRACT
Environmental risk assessment for impact pile driving requires characterization of the radiated sound field. Damped cylindrical spreading (DCS) describes propagation of the acoustic Mach cone generated by striking a pile and predicts sound exposure level (LE) versus range. For known water depth and sediment properties, DCS permits extrapolation from a measurement at a known range. Impact assessment criteria typically involve zero-to-peak sound pressure level (Lp,pk), root-mean-square sound pressure level (Lp,rms), and cumulative sound exposure level (LE,cum). To facilitate predictions using DCS, Lp,pk and Lp,rms were estimated from LE using empirical regressions. Using a wind farm construction scenario in the North Sea, DCS was applied to estimate ranges to recommended thresholds in fishes. For 3500 hammer strikes, the estimated LE,cum impact ranges for mortal and recoverable injury were up to 1.8 and 3.1 km, respectively. Applying a 10 dB noise abatement measure, these distances reduced to 0.29 km for mortal injury and 0.65 km for recoverable injury. An underlying detail that produces unstable results is the averaging time for calculating Lp,rms, which by convention is equal to the 90%-energy signal duration. A stable alternative is proposed for this quantity based on the effective signal duration.
Subject(s)
Noise , Sound , Acoustics , Animals , Fishes , Noise/adverse effectsABSTRACT
Sound produced by marine pile driving activities poses a possible risk to marine life. The assessment and mitigation of this risk requires a precise prediction of the expected levels. An analytical approach to estimate the radiated sound exposure levels is presented, based on the axial symmetry of the problem, resulting in damped cylindrical spreading. The approach is verified against numerical results from the recently held COMPILE benchmark workshop and validated with data from three different wind farm construction sites in the North Sea. In addition, found to yield more accurate estimates of the sound exposure level than an empirical decay formula sometimes used to evaluate the impact of marine pile driving.
ABSTRACT
Numerical models of underwater sound propagation predict the energy of impulsive signals and its decay with range with a better accuracy than the peak pressure. A semi-empirical formula is suggested to predict the peak pressure of man-made impulsive signals based on numerical predictions of their energy. The approach discussed by Galindo-Romero, Lippert, and Gavrilov [J. Acoust. Soc. Am. 138, in press (2015)] for airgun signals is modified to predict the peak pressure from offshore pile driving, which accounts for impact and pile parameters. It is shown that using the modified empirical formula provides more accurate predictions of the peak pressure than direct numerical simulations of the signal waveform.
ABSTRACT
This paper presents an empirical linear equation to predict peak pressure level of anthropogenic impulsive signals based on its correlation with the sound exposure level. The regression coefficients are shown to be weakly dependent on the environmental characteristics but governed by the source type and parameters. The equation can be applied to values of the sound exposure level predicted with a numerical model, which provides a significant improvement in the prediction of the peak pressure level. Part I presents the analysis for airgun arrays signals, and Part II considers the application of the empirical equation to offshore impact piling noise.
ABSTRACT
Due to the construction of offshore wind farms and its potential effect on marine wildlife, the numerical prediction of pile driving noise over long ranges has recently gained importance. In this contribution, a coupled finite element/wavenumber integration model for noise prediction is presented and validated by measurements. The ocean environment, especially the sea bottom, can only be characterized with limited accuracy in terms of input parameters for the numerical model at hand. Therefore the effect of these parameter uncertainties on the prediction of sound pressure levels (SPLs) in the water column is investigated by a probabilistic approach. In fact, a variation of the bottom material parameters by means of Monte-Carlo simulations shows significant effects on the predicted SPLs. A sensitivity analysis of the model with respect to the single quantities is performed, as well as a global variation. Based on the latter, the probability distribution of the SPLs at an exemplary receiver position is evaluated and compared to measurements. The aim of this procedure is to develop a model to reliably predict an interval for the SPLs, by quantifying the degree of uncertainty of the SPLs with the MC simulations.
