Page 14 - 2016Winter

Seismic Surveys
Figure 3. Cutaway view of a compressed air sound source (airgun). See text for an explanation of source operation. From Schlumberger Ltd., with permission.
sources are used (see How Seismic Arrays Are Used on page 15). Standard industry practice is to express airgun volumes, pressure, and other measures in American units like cubic inches, pounds per square inch (psi), or bars, so this review follows that convention but also gives the SI units in paren- theses.
Amplitude also varies with air pressure. An air pressure of 2,000 psi (13,789.5 kPa) is most commonly used but can range from 1,500 to 3,000 psi. For reference, 3,000 psi is the typical fill pressure of a scuba tank, and 1,600-2,000 psi is the output pressure of household pressure washers.
The size and shape of the ports through which the air is re- leased also influences the characteristics of the sound (Coste et al., 2014). In addition to the sound frequencies of interest for seismic surveys (under 100 Hz), higher frequencies are also created (Landrø et al., 2011). Minimizing acoustic en- ergy at higher frequencies is therefore desirable from a geo- logical imaging perspective and to reduce concerns about marine species such as dolphins, which use high-frequency sound.
Alternative Seismic Survey Sound Sources
Due to concern about the effects on marine life and to re- duce source energy not used in geophysical imaging, a vari- ety of novel sources are being explored as potential replace- ments for airguns (Rassenfoss, 2016). Vibroseis, a formerly trademarked name for a technology no longer in use, is often used today as a shorthand rubric for all innovative acoustic source technologies.
Generally speaking, these new sources are only viable due to advances in computer signal processing, enabling a tone series several seconds long to be “reconvolved” during data processing as if all frequencies had been produced at the same time. Because the acoustic energy is spread in time, the peak amplitude is lower than that of an impulse source, but the total energy is typically comparable to that of the compressed air source. Demonstration of the anticipated environmental benefits and of the cost, reliability, and safety will likely take some time, but there is clearly widespread motivation to try to find such a source (Rassenfoss, 2016).
Arrays
Use of a single airgun for geophysical surveys is rare; more often 18-48 airguns will be arranged in a rectangular con- figuration: a planar array oriented parallel to the sea surface (Figure 4).
An array serves several purposes. First, it is the simplest way to increase the nominal level of the source, although it should be noted that the nominal source level of an array is an imaginary number, calculated by extrapolating mea- surements at a distance back to a hypothetical point. Actual measurable levels around the array are typically 10-20 dB sound pressure level (SPL) lower than the nominal source level in the downward direction and an additional 10-20 dB lower at increasing angles away from the vertical (Caldwell and Dragoset, 2000).
Second, the arrangement of the elements in a planar array enables the added energy of the individual elements to be directed primarily downward (Figure 5). At all angles out- ward other than straight down, there are varying degrees of frequency-dependent interference between the elements (Dragoset, 2000). This is an important point because the nominal “source level” of seismic arrays is an idealized value projected to a hypothetical point within the array. Thus a “nominal” source level of 260 dB peak SPL (SPLpeak; re 1 μPa at 1 m) would not produce a measurable sound pressure at that level anywhere (a fact that nonacousticians rightly find difficult and frustrating). Sophisticated modeling is, how- ever, able to characterize the actual sound field well and is described in more detail in Sound Propagation.
The third and perhaps most important reason for using seis- mic sources in an array is the cancellation of sound from the oscillating bubbles after their initial formation. Any sound after the initial pulse clutters the return signal as well as adding high-frequency energy that is both useless for imag-
12 | Acoustics Today | Winter 2016