Elsevier

Chemical Engineering Science

Volume 131, 28 July 2015, Pages 187-196
Chemical Engineering Science

Can acoustic emissions be used to size bubbles seeping from a sediment bed?

https://doi.org/10.1016/j.ces.2015.03.058Get rights and content

Highlights

  • We generated air bubbles in a sediment bed in the laboratory.

  • Gas interactions are recorded simultaneously by hydrophone and high speed camera.

  • The acoustical signals are analyzed by power spectra and sonogram methodologies.

  • Sediment interaction, elastic resistance and asymmetric gas detachment are showed.

  • Underwater acoustic sensors could warn of critical changes in bubble-seep behavior.

Abstract

Experiments on bubble formation from a granular medium are presented. The granular medium consisted of immersed stones giving a sediment bed thickness of 10 cm. High speed photographic and acoustic passive wireless techniques were employed to obtain the bubble size. The power spectra of the acoustic data for bubble production of 0.5 bubbles/s showed the existence of two principal peaks which are correlated with two discrete events. Firstly, a signal peak at 0.497 kHz, representing asymmetric bubble detachment, and secondly, a peak at 2.070 kHz, which corresponded to interactions of the granular media as inter-particle gaps advance, representing elastic resistance to orifice creation. Provided the signals were windowed to the bubble-detachment event only, the classical bubble-acoustical Minnaert relation for the bubble size agreed with optical data within about 10%. Since seepage generates discrete bubble pulses, appropriate acoustic analyses could both count and size the bubbles formed. The results of this study lead to the proposition that underwater acoustic sensors could warn of critical changes in bubble-seep behavior over time. This could lead to the possibility of remote early warnings, because shifts in production-rate regimes from seeps may herald alterations in the progression of global warming, or impending earthquakes and tsunamis.

Introduction

Recently it has been found that methane trapped in permafrost is escaping the sea bed from wide areas of the East Siberian shelf (Shakhova et al., 2005) and is an important accelerator of global warming (MacGuire et al., 2006). Methane may migrate through the sediment as either a dissolved or a free gas phase, creating bubble plumes, and gases emitted from the decay of organic matter in marine sediments also may contribute to global greenhouse gas budgets (Arndt et al., 2013). The methane bubbles are generated in geological structures called marine seeps or pockmarks which contribute 0.4–48 tera grammes per year of methane to the hydrosphere and atmosphere (Judd and Hovland, 2007). The marine seeps are located on the seabed and generally percolate through sediments (siliceous, calcareous, pelagic clay, land-formed, etc.). The ability to monitor such seeps over long periods of time would give information about local gas fluxes and their temporal and spatial variation (Judd, 2004).

Furthermore, it has been recognized for some time that monitoring of the emission of gases near faults may assist earthquake prediction (Sugisaki, 1981, Hartmann and Levy, 2005). Changes in undersea gas seepage may be earthquake precursors (Khilyuk and Chilingar, 2000). Emissions of radon vapor and its decay products from the Earth’s crust, in wells and spring water along active fault zones, have been recognized as a potential tool in earthquake prediction (Klusman and Webster, 1981, Papstefanou, 2007). Thus, underwater bubble formation in terrestrial springs and lakes should also benefit from monitoring.

Zones of the ocean bottom which generally show microseismic activity can be monitored by an array of geophones (Webb et al., 2001). This technology, commonly used to studying the structure of the upper mantle, is now applied to the oceanic crust using natural earthquakes as sources. The Ocean-Bottom Seismometers (OBS) record high-fidelity seismic data for long periods of time. Gas seeps offer an additional source of data enhancing the predictive power of long-term monitoring. In the course of a dedicated research cruise, it is possible to locate gas seeps using active acoustical methods such as vessel-mounted sonar (Matveeva et al., 2003, Schneider von Deimling et al., 2007, Artemov et al., 2007, Nikolovska et al., 2008). However, active devices are expensive and inherently require significant power to operate and hence cannot be left for long periods of time on the seabed.

Some studies (Nikolovska and Waldmann, 2006, Leighton and White, 2012) have used the Passive Acoustical (PA) technique to obtain the bubble-size populations from undersea bubble plumes, with the objective of gas flux quantification. Generally, sparged bubbles are generated using rigid orifices (capillary tubes, perforated plates, ceramic stones, etc.) but in the granular case, the orifice behavior is affected by the air flow rate, particle diameter, particle density and the frictional forces (mobility) between particles (Gostiaux et al., 2002). When air is injected into a sediment bed, the gas penetration is resisted by capillary pressure as the gas invades the interstitial spaces. In addition, it is also known that if the bubbles are generated in microbial clay sized particles of≈3.9–62.5 μm (a by-product of metabolism by methanogenic bacteria methane; see Floodgate and Judd, 1992), the particle size can affect the bubble dynamics, that is, if the bubbles are small relative to particle size, they remain within the pore fluid and behave as bubbles in water, on the other hand, if bubbles are large relative to particle size the structure of the sediment frame interacts with the bubbles and changes the bubble compressibility (Wilkens and Richardson, 1998). PA techniques are also applied in the laboratory (Vazquez et al., 2008) and in industries (Boyd and Varley, 2001, Manasseh and Ooi, 2009) in systems in which bubbles are generated by diverse devices: capillary tubes, ceramic stones, perforated plates, plastic membranes, metallic mesh, granular materials, sediments, etc. The bubble generation can have a great influence on the formation size and hence subsequent behavior when bubbles rise through the water (Leighton, 1994, Leighton et al., 1991).

The simplest class of data likely to be of practical relevance to long-term monitoring is the volumetric gas flow from a seabed source, Q, although the gas composition may also be relevant (Fu et al., 2005). The flow rate is equal to the volume of each bubble produced multiplied by the rate at which bubbles are produced; in situations where variable bubble sizes are generated, this can be expressed asQ=iN43πR0i3fbi,where R0i is the equivalent-spherical radius of the ith bubble, N is the number of bubbles and fbi is the rate at which the bubbles are formed. In order to reliably and accurately monitor gas seepage, accurate measures of both bubble size and the rate of bubble production are needed. To achieve this, the acoustic signals corresponding to the formation of individual bubbles generated at low gas flow rates must be separated from other confounding signals due to sediment motion or background noise (or other marine noises such as that due to the breaking wave, or bubbles bursting the sea surface, etc.). Low gas flow rates are studied to preclude errors due to overlapping bubble signals, i.e. to preclude bubble coalescence or jetting regimes (Manasseh et al., 2008, Leighton and White, 2012). Since bubble-acoustic signals from bubble formation points fall off rapidly with distance from the source (Manasseh et al., 2008) appropriate placement of hydrophones close to individual, low gas-flow seeps would provide the most accurate data and hence the greatest dynamic range. The bubbles could then be counted and sized.

The core of the present PA methodology is based on bubble detachment from an orifice or capillary tube, i.e., when the bubbles are growing attached to an orifice. The subsequent inflation leads to the appearance of a neck, which connects the bubble’s body to the orifice (Czerski and Deane, 2011, Deane and Czerski, 2008, Deane and Stokes, 2008). When the buoyancy force exceeds the other applicable forces (surface tension, drag, and pressure) (Vazquez et al., 2010), the neck finally breaks and a pulse of sound propagates through the fluid (the origin of the sound has been explained in the literature by three separate mechanisms, summarized in Manasseh et al. (2008)); one mechanism possibly relevant to the present experiments is compression of gas in the bubble by a radial inrush of liquid as the pinch-off occurs (Deane and Czerski, 2008). The Minnaert relation (Minnaert, 1933) is used to relate the bubble size with the sound wave frequency,f0=1R0(3γPA4π2ρ)where f0 is the frequency in Hz, PA is the absolute liquid pressure, ρ is the liquid density and γ is the ratio of specific heats for the gas assuming adiabatic compression and expansion, which is valid for millimetric bubbles. It is interesting to note that γ takes the ideal-gas values of 1.67 (53) for monatomic gases (e.g. Rn), 1.4 (75) for diatomic gases (e.g. N2), 1.3 (97) for triatomic gases (e.g. CO2) and 1.2 (13/11) for CH4 (Langes Handbook (Lange, 1999)). Thus, changes in gas composition may also cause changes in f0, e.g. obtain low values of the frequency if CH4 concentration were to increase. The Minnaert relation assumes several aspects such as a spherical bubble shape, that the mass-inertial contribution is only due to the liquid (the liquid density is much greater than the gas density), that the surface tension forces are negligible, the liquid is incompressible and that the bubble is surrounded by pure liquid and not particles with their own dynamics. Nonetheless, the Minnaert frequency has been widely and successfully used by many researchers (Boyd and Varley, 2001, Al-Masry et al., 2005, Chicharro and Vazquez, 2014) to obtain bubble sizes in diverse situations. What happens if the bubble is surrounded by a complex micro-environment of mobile solid particles as well as liquid? The answer to this question remains unclear so far, and leads to the goal of the present paper: a preliminary assessment of the feasibility of precise, long-term quantitative monitoring of a sediment bed.

We realized a preliminary and simple experiment (detailed below) using a granular-sediment bed in a Plexiglass tank. Air was injected at a rate (0.1 cm3/s) at which there was only a single bubble in the tank at a time (the solitary flow rate), and at low (0.5 cm3/s), medium (2.5 cm3/s) and high (4.5 cm3/s) rates. Bubble-detachment images and sound were recorded simultaneously by a high speed camera and wireless hydrophone. Finally the passive acoustic signal obtained from bubble generated from sediment bed was compared to the equivalent signal for a rigid orifice.

Section snippets

The equipment

The experimental setup (Fig. 1) consisted of a 5.3 W air pump (Atlantis Force, Mod. 12858, Taiwan) which sent an air flow to a gas chamber (2×10−3 m3) controlled by a bronze coarse valve. Thus, the bubbles were air (i.e. γ=1.4). The acoustic signals generated during the bubble growth and departure from the sediment bed were measured with a custom built wireless piezoelectric hydrophone (Vazquez et al., 2005) and recorded by a data-logger (USB-6501, National Instruments, USA). The tank dimensions

Acoustical comparison response for rigid orifice and gates

The acoustic response given by the hydrophone signal for both rigid and gate orifices for a solitary bubble are shown in Fig. 2. The left-hand signal shows a typical freely-oscillating lightly-damped bubble acoustic emission as a bubble escapes from a rigid glass capillary tube; here the acoustic amplitude decays as exp[−(f0δt/2)] as in any classical damped oscillator, in where f0 is the resonant frequency and δ the damping factor. The maximum amplitude is reached in the first few milliseconds

Discussion

The synchronous images and audio signal recording shown in Fig. 5 suggest that the sediment bed affects the bubble growth and that the various signatures in the signal can be understood as the granular media interacting with the gas. Considering first the results on the solitary bubble behavior (Fig. 2), we divided the behavior into two stages using a lateral video recording (Fig. 7). In the first stage (Fig. 7-I), an oscillatory sound signal (Fig. 2a) is registered even though no bubble is

Conclusions

The present study has identified the acoustical signature when an air bubble is formed in a granular-sediment bed. We found that the air bubble growth through a sediment bed under quiescent water is affected by the mechanical proprieties of granular media. Moreover, the acoustic signal revealed two phenomenological stages: fracture advancing-elastic resistance to orifice creation and asymmetric bubble detachment.

Careful observations using both spectral and spectrogram approaches as well as

Nomenclature

    A

    image area, mm2

    fbi

    bubble rate, bubbles s−1

    f0

    resonant frequency, Hz

    N

    number of bubbles, bubbles

    PA

    liquid pressure, Pa

    Q

    volumetric gas flow, cm3 s−1

    R0

    equivalent-spherical bubble radius, mm

    t

    time, s

    γ

    specific heats ratio, dimensionless

    δ

    damping factor

    ρ

    liquid density, kg/m3

Acknowledgments

The authors gratefully acknowledge Prof. Ira Leifer for the technical assistance in the MatLab routines by the sound signal analysis.

References (45)

  • Y.G. Artemov et al.

    Methane emission to the hydro - and atmosphere by gas bubble streams in the Dnieper paleo-delta, the Black Sea

    Rep. Natl. Acad. Sci. Ukr.

    (2007)
  • B.P. Boudreau et al.

    Bubble growth and rise in soft sediments

    Geological Society of America

    (2005)
  • W.R.J. Boyd et al.

    The uses of passive measurement of acoustic emissions from chemical engineering processes

    Chem. Eng. Sci.

    (2001)
  • H. Czerski et al.

    The effect of coupling on bubble fragmentation acoustics

    J. Acoust. Soc. Am.

    (2011)
  • G.B. Deane et al.

    A mechanism stimulating sound production from air bubbles released from a nozzle

    J. Acoust. Soc. Am.

    (2008)
  • Deane, G. B., & Stokes, M. D. (2008). The acoustic excitation of air bubbles fragmenting in sheared flow. J. Acoust....
  • C.C. Fu et al.

    Reconnaissance of soil gas composition over the buried fault and fracture zone in southern Taiwan

    Geochem. J.

    (2005)
  • L. Gostiaux et al.

    Dynamics of a gas rising through a thin immersed layer of granular material: an experimental study

    Granul. Matter

    (2002)
  • J. Hartmann et al.

    Hydrogeological and gasgeochemical earthquake precursors – a review for application

    Nat. Hazards

    (2005)
  • A.G. Judd

    Natural seabed gas seeps as sources of atmospheric methane

    Environ. Geol.

    (2004)
  • A.G. Judd et al.

    Seabed Fluid Flow

    (2007)
  • L.F. Khilyuk et al.

    Gas Migration: Events Preceding Earthquakes

    (2000)
  • Cited by (32)

    • A non-invasive, low frequency resonant method to detect bubbles in liquid media

      2021, Applied Acoustics
      Citation Excerpt :

      Contrary to optical methods, acoustic methods have proven to be a more versatile approach capable of performing well for all types of fluid media and appropriate for portable designs [9,16–21]. Although some acoustic bubble detection techniques are based on non-linear bubble oscillations, most of them rely on the ability of a bubble to act as a small amplitude linear oscillator with a well-defined resonant frequency [19,22–24]. Typically, as in the ultrasonic Doppler shift technique, acoustic measurements require an ultrasonic emitter for excitation of bubbles within a certain radius range and a hydrophone or transducer signal collector.

    • The heart signal: An acoustic signature observed during a second-bubble entrainment

      2020, Chemical Engineering Science
      Citation Excerpt :

      In Fig. 10 it is observed that the values are also grouped in two zones; in the left-hand zone, the Froude number is in the range of 5.25 to 6.90, corresponding to eccentricity values between 0.29 and 0.39, where the crater shape tends to an oblate ellipsoid, while for the right-hand zone the Froude number is 4.82 with an eccentricity of 0.61, corresponding to a tendency for the cavity shape to form a prolate-ellipsoid. The signals observed in Section 3 show a similar behavior to those reported in many other studies (Pumphrey and Elmore, 1990; Ma and Nystuen, 2005; Czerski and Deane, 2011; Howe and Hagen, 2011; Vazquez et al., 2015). In particular, there is the common observation of the sound produced when the mother drop impacts the water surface, which appears as a sudden pulse or peak of short duration and large amplitude, commonly called the hammer blow or shock pulse.

    • Numerical modeling for characterization of CO<inf>2</inf> bubble formation through submerged orifice in ionic liquids

      2019, Chemical Engineering Research and Design
      Citation Excerpt :

      This important information requires further improvement in VOF model. During bubble formation, vortices generate that produces pulses of acoustic pressure in flow field (Pandit et al., 1992; Vazquez et al., 2015; Yasui et al., 2015). So, discretization of acoustic pressure equations will be useful to understand bubble formation stages.

    • Experimental study of sound emission in subcooled pool boiling on a small heating surface

      2018, Chemical Engineering Science
      Citation Excerpt :

      Acoustic measurement techniques for investigations of boiling sound are summarized in Table 1. These acoustic detection techniques are also widely used in marine, metallurgy, and chemical industries (Vazquez et al., 2015; Moghadam et al., 2017; Zhou et al., 2018). Hydrophone is a microphone designed to be used underwater for the detection of underwater sound, with a sampling frequency up to tens of MHz.

    View all citing articles on Scopus
    View full text