As new sources of anthropogenic noise, arrays of current- and wave-energy converters (CECs and WECs, respectively) will add to other man-made and natural sources in the marine environment. Regulatory constraints and environmental concerns over the sound pressure levels produced by these devices may be a hurdle to their deployment and operation. The Marine Mammal Protection Act in the USA, for example, limits disturbance and injury, where the present classification of total noise levels permissible are below 120 dB and 180 dB re 1 μPa, respectively . While initial studies from individual devices are promisingly below 120 d B [2-6] , alternative device designs, equipment failure, or array deployment could potentially surpass these levels. Each device type will create unique noise profiles and levels, and how those spectra change between individual deployment sites will need to be predicted (either through experiments or modeling efforts) in order to meet regulatory requirements for permitting and deployment.
Shallow water environments are likely to be louder due to increased anthropogenic and natural noise sources in a constrained space, such as from shipping traffic, breaking waves, sediment/debris transport, and environmental variability [7, 8]. Few measurements from controlled experiments  and in-situ studies [3-5] or predictions from models  exist to build a conclusive understanding of the types of noise produced from these devices. The addition of new marine hydrokinetic (MHK) sources will couple to the total sound level of the system and has the potential to push that total over regulatory and environmental limits. However, dependent upon the sound pressure level (SPL) produced by individual MHK devices, their arrangement in an array, and the local site characteristics, these additions may have little contribution relative to all the sources even considering their additive nature. Pref can be calculated in octave or 1/3-octave bands (commonly used when the frequency content of the noise is needed, such as for biological studies ) or as a broadband of all frequencies. The latter is chosen for this work and follows the definition of the NOAA Interim Sound Threshold Guidance ; however, a draft of updated guidelines is currently under review .
The modeled, or measured, SPL can then be used to understand how particular species respond [10-13]. Insonification studies of marine species detail what sounds, levels , and frequencies are impactful [10, 14]. Eulerian Lagrangian-agent methods (ELAM) coupled with hydrodynamic models can provide a mechanism to statistically predict changes to behavior of marine animals or fish populations in response to particular stimuli [12, 13]. Based upon the model, temporary and permanent shifts in hearing levels can then be determined [11, 14]. However, these conclusions will be highly dependent up on the calculated SPL from an array of devices. Farcas et al. highlights the overall infancy of underwater noise modeling and integration into environmental impact assessments .
This work investigates the broadband noise generated from an array of CECs and how the array shape and size impact the sound environment. Extension to particular frequency bands and updated NOAA guidelines for marine mammals will be performed in a more detailed analysis to follow.