Marine piling - energy conversion factors in underwater radiated sound: review

5. Summary and Recommendations

This section provides a summary of the review contents and the recommendations for the prediction of the underwater sound fields resulting from impact piling.

• The point-source equivalent ECF method comprises:
  • generating a single value broadband ESL for which the inputs are the hammer energy and the point-source equivalent ECF.
  • Calculating a source spectrum using recorded results scaled to the generated ESL.
  • Using a point source propagation model to generate the per-pulse SEL sound fields from which impacts are assessed.
• Benefits of the method include:
  • its simplicity in that it requires only the hammer energy and a value for β to generate a source function.
  • the speed at which one can generate source inputs and modelling outputs.
  • its exploitation of a powerful physical principle, i.e., conservation of energy
• Shortcomings of the method exist in the source generation:
  • The ratio of hammer input energy to radiated acoustic energy in the water column is not a fixed universal value. Recorded values range from 0.17 % to 1.56 %, which equates to a range of 9.6 dB.
  • The dependence of this ratio on input parameters based on the pile, the hammer, the environment, and the geometry is not well understood.
• Shortcomings also exist in the modelled propagation:
  • The nature of propagation from point source models is substantially different from one suitable for piling noise. It is also noted that a source level does not exist for a pile, and that it is unhelpful to attempt to characterise it as such.
  • Predictions of distances to sound level thresholds can often be out by orders of magnitude, with examples showing errors up to 10 dB within 5 km of the pile.
• Effects are compounded when both are used:
  • In one reported numerical modelling example, combining the point-source equivalent ECF generated point source using β = 0.5 % with the point source propagation model yielded underestimates of the per-pulse SEL of between 9.5 and 12.1 dB between 100 and 1000 m from the pile when compared against the dedicated pile model.

The following recommendations are made:

• Modelling using the point-source equivalent ECF method described above should not be used, having been entirely superseded by more modern approaches and shown to be inaccurate for sound field predictions.
• Numerical modelling provides the greatest flexibility in terms of selection of hammer, pile, and environment and is considered the leading method, provided:
  • The Mach wave nature of piling is taken into account in the source modelling, and
  • This is coupled to a propagation model that will support the Mach wave.
• Genuine values of the ECF, however, could be used for sound level predictions provided they are used with a model that supports them, e.g., DCS model.
• If a measurement or measurements exist for the scenario in question, the DCS model can be used to predict sound levels at other distances up to approximately 5 km using the reference sound level.
• Where measurements exist of similar scenarios, these may be used with adjustments to apply to alternative scenarios with caution.
  • Possibilities include change in hammer energy, change in ram mass, change in pile diameter, and change in water depth.
  • The difficulty of this approach is knowing how to determine whether changes are too numerous or great. If in doubt, it is recommended to remodel using numerical techniques.