Michael L Anderson
Black Oak Engineering
28 Dec 24
Why not just measure optical opacity?
For over a century there have been techniques for measuring the optical opacity of flue gas. The first formal method to be devised, the Ringelmann scale (Fig. 1), is based upon a subjective, eyeball comparison of the observed smoke plume versus a set of cards that have been printed with progressively thicker black grids. This method is still widely used and is legitimately useful for spot checks. The Ringelmann scale itself is still often the metric by which emissions are measured, even if better technology is available.
An improvement is obtained by measuring the optical transmission density of the flue gas. This is loosely referred to in the industry as opacity, or sometimes Continuous Opacity Monitoring Systems (COMS). A constant source, preferably monochromatic, of visible light is collimated into the flue gas. This light beam is either reflected back to a collocated receiver (double pass) or is detected at the opposite end (single pass). The logarithm of the ratio of a fiducial intensity (‘zero gas’ or ‘clear stack’ level) over the sample intensity is known technically as optical density. Opacity is a percentage derived from this logarithmic attenuation. Fig. 2 shows an industry standard stack-based instrument.
If we assume that the particles act as pure absorbers and there is no secondary re-radiation or scattering of any type, then we have a simple Beer-Lambert Law situation. Transmissivity
where σ is absorption cross section, l is total path length, and A is the density of absorbers per unit volume. As in any Beer-Lambert analysis, we assume that the absorber density is low enough that there is no overlap of the absorber cross sections. Is this reasonable? Consider a typical case. An exhaust stack is emitting a modest dust load of δd = 10mg/m3 of monodisperse 10μ diameter spherical particles of SiO2. At that diameter the particle cross section σp is 7.85 × 10-11 m2 . A single pass opacity instrument is set up across 2 m of stack width (l). Assume our instrument’s beam width is σbw = 1 cm2 . As with any such sampling measurement, we assume that the sample cross section is uniform and representative of the whole. The mass density of SiO2 is 2634 kg/m3 . The mass of each 10 μ particle is mp = 5.23 × 10-16 kg. The number of particles in the sample beam is na = σbw l δd / mp = 3.82 × 109
For there to be no significant overlap in absorber profile, na σp / σbw must be « 1. In our case we have ≈ 3000, and this is far from a difficult case. Unfortunately, opacity measurement is based on some bad assumptions. There is also a problem with opacities less than ten percent. Here the difference between fiducial and sample signals is so slight that it is difficult to resolve accurately, given typical electronic and optical drift effects. For one thing, the fiducial level is only set during a known ‘clear stack’ condition (or its optical simulation). It may be several days at least between such conditions and the resetting of the reference level. Typical drift phenomena take place on the time scale of seconds. The simple accumulation of dust on optical surfaces (despite purge air systems) is another error source.
While the air pollution monitoring industry is presently rife with relatively inexpensive opacity instruments, their inherent limitations are becoming problematic. It is telling that many instruments do not even allow for the user to enter the optical path length. Without this information we may only be doing a purely comparative assessment, and certainly not a complete Beer-Lambert Law analysis. And we are certainly not addressing the concerns raised above concerning particle size discrimination. Opacity measurements are a poor substitute for particulate measurement (PM) and they are expected to wane in popularity.
Note – references to any specific company’s products are not to be construed as an endorsement. All content and claims herein are believed to be correct; any misstatements are due to the author’s ignorance and are not to be taken as criticisms of any other party.
Sources
ASTM, D6216-07 details protocol for measuring opacity < 10%.
Pui, David YH & David L Swift, Direct-Reading Instruments For Airborne Particles, in Air Sampling Instruments for evaluation of atmospheric contaminants, 8th ed, eds.
Beverly S Cohen & Susanne V Hering, 1995. USEPA, Office of Research & Development, Continuous Emission. Monitoring Systems for Non-criteria Pollutants, EPA/625/R-97/001, Aug 1997.
Michael L Anderson, manderson @ blackoakeng.com, holds an MSEE and an MS Physics. He served in the USMC. He now works as a consulting design engineer and business executive in New York.
© 2024 Black Oak Engineering. All rights reserved.
Comments