A real-time method for sensing suspended dust concentration from the light extinction coefficient

https://doi.org/10.1016/j.jlp.2020.104242Get rights and content

Highlights

  • A real-time, low-cost suspended dust concentration measurement method using light extinction coefficient was developed.

  • The effect of changing ambient light conditions and camera noise was removed through a calibration process.

  • The relationship between light extinction coefficient and suspended dust concentration was established.

  • Suspended dust concentrations ranging between 10 and 90 kg/m3 were tested using the proposed method.

Abstract

In powder handling and processing industry, location of dust emission can vary, with the suspended dust concentration assessment requiring installation of an immovable or wired equipment. For increased dust sensing, not limited by location within the facility, a portable suspended dust concentration measuring system is needed. In this study, a new method of sensing suspended dust concentration under daylight environment using the change in light extinction coefficient was developed. The method involves capturing images of the suspended dust cloud and then analyzing the light extinction coefficient. This method mitigated the environmental light scattering and absorption and eliminated the noise from the images obtained through a camera by calibration between two targets. Cornstarch, corn dust, and sawdust were used as test materials in this study. The light extinction coefficient (ε) was found to correlate with the suspended dust concentration, and the ε values depended on the dust properties. Mass extinction coefficient (K) was obtained for cornstarch, saw dust and corn dust, from known suspended dust concentrations using image analysis. The mass extinction coefficient of the three sample materials tested in this study were in the range of 0.03–0.04. This method of using light extinction coefficient can be used for real-time sensing of suspended dust concentration in both open and confined spaces.

Introduction

Increasing production demands have exacerbated the health and safety risks to workers in particulate material processing and handling facilities. Over 70% of the dust produced in industry are explosive, and the suspended dust concentration is one of the most important factor that leads to a major explosion (Abbasi and Abbasi, 2007). Suspended dust concentration, an indicator of explosion risk in industrial environments (Hinds, 2012; Zhao and Ambrose, 2019), is normally expressed as particle counts per unit volume or mass per unit of volume. However, the suspended dust concentration measurement to evaluate explosion risk, with the minimum explosive concentration (MEC) that could be as low as 15 g/m³ , is less developed.

To monitor the MEC within the industry, high mobility method with large measurement area is required. In environmental science and occupational safety related dust concerns, the well-developed gravimetric and light scattering methods are widely used for measuring airborne dust concentrations, and form the basis of European and US standard methods for monitoring outdoor concentration of PM10 and PM2.5 (Tasić et al., 2012). These methods are mostly used for respirable dust with concentration of micro-grams per unit volume.

In order to prevent dust explosion, a method to measure the suspended grain dust in silos using the Lambert–Beer law is proposed by Hauert et al. (1996). But the probe that includes a laser and a photodiode must be calibrated before every use. Light scattering with a portable dust track aerosol monitor (Dacunto et al., 2015) or using an optical fiber method (Zhong and Li, 1988) are the other approaches used to measure MEC. But the optical beam may be a potential ignition source for combustible dust, for example Nd-YAG laser can ignite the cornstarch cloud with 1.9 W incident power on the suspended starch cloud (Proust, 2002). Other techniques use electrostatic interactions, including scanning mobility particle sizers like the differential mobility analyzer. 49 Even though electrostatic equipment can measure the number of aerosol particles per unit volume, these devices are generally very expensive and used more for testing particle-size distribution (Kousaka et al., 1985).

In the market today, nearly all methods for measuring dust concentration require the purchase and installation of new equipment in an industrial facility. However, suspended dust clouds are dynamic and move with air currents in the facility, and dust can be emitted from a variety of locations in a processing facility. So, there is a need for a portable and inexpensive dust concentration measurement method/probe.

Smoke and suspended dust particles can reduce visibility as the particles scatter and absorb light. The reduction in intensity of light passing through a dust cloud is referred to as extinction. The effects of dust concentration on visibility through the atmosphere have been studied widely. Several empirical relationships between dust concentration and visibility have been proposed (Chepil and Woodruff, 1957; Patterson and Gillette, 1977; Chung et al., 2003; Wang et al., 2008; Baddock et al., 2014; Camino et al., 2015). These relationships are used widely in environmental science but have not been studied for use in industrial environments. Each empirical relationship between visibility and dust concentration was developed for a different specific environment and type of dust. Therefore, to estimate the dust concentration from visibility in an indoor environment, new empirical relationships need to be developed. Atmospheric science generally classifies dust clouds by their cause, like dust storms or fuel burning, so the materials making up the dust vary and are usually underspecified. However, industrial dust emissions are primarily from known products, so the components and size of dust will be consistent and known. Light will be affected predictably by such clouds of dust, which should make empirical relationships between dust concentration and light extinction relatively easy to establish.

The development and widespread acceptance of smartphones means that many people have ready access to digital cameras. Changes in light intensity can be detected using these cameras’ CCD/CMOS sensors to infer dust concentrations. This paper presents a novel method for measuring dust concentrations using smart-phone cameras. The specific objectives are to explore the relationship between dust concentration and extinction coefficients, and develop a empirical linear regression model for several types of aerosolized materials.

The extinction coefficient represents the rate of diminution of transmitted light via scattering and absorption for a medium. The particle concentration affects the extinction coefficient (ε) of the atmosphere (Ogle, 2016):ε=πdp2Nq4where dp is the particle diameter, N is the number of particles per unit volume and q is the dimensionless extinction efficiency of a single particle. For the same dust sample, the aerosol particle-size distribution and q are considered constant at all mass concentrations.

Thus, the mass concentration C can be calculated as follows:C=2dpρ3qεwhere ρ is the particle density. To obtain the dust concentration, the extinction coefficient ε value is required, and to calculate this value a dimensionless mass extinction coefficient K (m2/g) is introduced, where1K=2dpρ3q

As the particle diameter and chemical composition are constant for the same material in a dust cloud, so the dimensionless extinction efficiency, and density are also considered as constant.

The extinction coefficient can be calculated based on atmospheric light scattering models which describe the observed light intensity of a target and a background, as that intensity is affected by the extinction coefficient at distance r (Graves and Newsam, 2011):J0r=J0eεr+JA(1eεr)where, J0r is the observed target light intensity with R the distance from target to observed location, J0 is the real target light intensity and JA is the ambient light intensity.

The ambient light intensity highly depends on the environmental conditions, therefore the background reference Jg is introduced in order to calibrate the extinction coefficient:ε=ln(J0rJgrJ0Jg)rwhere, Jgr is the observed background light intensity, Jg is the real background light intensity, target and background reference are at distance r from the observed location.

The light intensity (J) can be obtained through a camera. J0 and Jg are the intensity without particles between target and camera. This intensity will differ with the ambient light conditions. So, during sensing the dust cloud concentration, obtaining the real-time J0 and Jg will be a challenging task and inaccurate values will affect the prediction accuracy. Therefore, another target was used to negate the effect of ambient light conditions and to calculate the real-time ε. Using two targets at different distances from the observed location enables calculation of ε without knowing J0 and Jg value.

On the other hand, most cameras use a charge coupled device (CCD) sensor, and there will be noise signals when sensing the light signal. Thus, a calibration to remove the noise signal is important before using the light intensity obtained from a CCD. J is linearly related with the intensity value obtained from a CCD sensor (G) (Healey and Kondepudy, 1994):G=A(J+NDC+NS+NR)where A is the amplifier to increase the signal power from CCD sensor, NDC is the dark current noise, NS is the zero mean Poisson shot noise, and NR is the readout noise. Noise is unstable in most cases and makes the result inconsistent, especially when a response intensity is lesser than the noise, then the intensity will not be accurate or even undetectable. So, a two-target method can overcome the noise effect since the region of interest lies between the specifically patterned targets.

The extinction coefficient calculated using the intensity value measured from two targets can be used to eliminate the effects of noise:ε=ln(G0r1Ggr1G0r2Ggr2)Rwhere G0r1 and Ggr1 are the first target and its background intensity value calculated from the image by averaging the grey value of all pixels, respectively, and G0r2 and Ggr2 are the second target and its background intensity value calculated from the image, respectively. R (= r1 – r2) is the distance between the two targets.

Section snippets

Experimental dust dispersion

A transparent 0.3 × 0.3 × 0.45 m³ chamber, with two targets placed inside, was used for suspended dust concentration measurement tests. Cornstarch (Clabber Girl Corporation, IN, USA), sawdust (System Three Resins, Inc, WA, USA) and corn dust obtained from a local grain elevator were used in the all experimental measurements. Dust samples of 0.5, 1.0, 1.5, 2.0, or 2.5 g were placed inside the chamber, in front of the nozzle, where the theoretical concentration corresponds to 17.5, 25, 42.5, 50,

Dust concentration and extinction coefficient during dispersion

Dust was dispersed fully within 3 s of introducing the compressed air. For 2 g dust dispersion, suspended dust concentrations measured using the laser and the extinction coefficient values were plotted in Fig. 3, Fig. 4, Fig. 5. The peak concentration of suspended dust was observed around 0.5–1.5 s, and then decreased as the particles settled. During the dust dispersion process, particles are suspended by air movement, and only settle once they hit the chamber wall due to gravitational force

Conclusion

A two-target method for measuring the concentration of suspended dust was developed and tested with cornstarch, corn dust, and sawdust powder. This method used the light extinction coefficient of a dust cloud between two targets using a digital camera. This extinction coefficient is linearly related to the suspended dust concentration, and the mass extinction coefficient is the key value for this measurement method. The mass extinction coefficient (K) is highly dependent on the physical and

Credit author statement

Yumeng Zhao: Methodology, Formal analysis, Writing-Original Draft, Visualization. Kingsly Ambrose: Conceptualization, Resources, Writing – Review & Editing, Project administration.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

This research was supported by Purdue Process Safety and Assurance Center (P2SAC), Purdue University, West Lafayette, Indiana, USA. We thank undergraduate research assistant Pranav Vashisht for assistance with experiments.

References (28)

  • W.S. Chepil et al.

    Sedimentary characteristics of dust storms; Part II, Visibility and dust concentration

    Am. J. Sci.

    (1957)
  • Y.S. Chung et al.

    Atmospheric loadings, concentrations and visibility associated with sandstorms: satellite and meteorological analysis

    Water Air Soil Pollut. Focus

    (2003)
  • L. Colangeli et al.

    Mass extinction coefficients of various submicron amorphous carbon grains: Tabulated values from 40 NM to 2 mm

    Astron. Astrophys. Suppl. Ser.

    (1995)
  • P.J. Dacunto et al.

    Determining PM 2.5 calibration curves for a low-cost particle monitor: common indoor residential aerosols

    Environ. Sci.: Processes Impacts

    (2015)
  • Cited by (0)

    View full text