Field evaluation of Digital Optical Method to quantify the visual opacity of plumes.

ABSTRACT

Visual Determination of the Opacity of Emissions from Stationary Sources (Method 9) is a reference method established by U.S. Environmental Protection Agency (EPA) to quantify plume opacity. However, Method 9 relies on observations from humans, which introduces subjectivity. In addition, it is expensive to teach and certify personnel to evaluate plume opacity on a semiannual basis. In this study, field tests were completed during a "smoke school" and a 4-month monitoring program of plumes emitted from stationary sources with a Method 9 qualified observer to evaluate the use of digital photography and two computer algorithms as an alternative to Method 9. This Digital Optical Method (DOM) improves objectivity, costs less to implement than Method 9, and provides archival photographic records of the plumes. Results from "smoke school" tests indicate that DOM passed six of eight tests when the sun was located in the 140[degrees] sector behind one of the three cameras, with the individual opacity errors of 15% or less and average opacity errors of 7.5% or less. DOM also passed seven of the eight tests when the sun was located in the 216[degrees] sector behind another camera. However, DOM passed only one of the eight tests when the sun was located in the 116[degrees] sector in front of the third camera. Certification to read plume opacity by a "smoke reader" for 6 months requires that the "smoke reader" pass one of the smoke school tests during smoke school. The average opacity errors and percentage of observations with individual opacity errors above 15% for the results obtained with DOM were lower than those obtained by the smoke school trainees with the sun was located behind the camera, whereas they were higher than the smoke school trainee results with the sun located in front of the camera. In addition, the difference between plume opacity values obtained by DOM and a Method 9 qualified observer, as measured in the field for two industrial sources, were 2.2%. These encouraging results demonstrate that DOM is able to meet Method 9 requirements under a wide variety of field conditions and, therefore, has potential to be used as an alternative to Method 9.

INTRODUCTION

Visible plumes caused by emission of particulate matter (PM) to the atmosphere from stationary sources raise public concern about the effects of PM on human health (1) and atmospheric visibility. (2) The Regional Haze Rule was adopted during 1999 to protect visual air quality by reducing anthropogenic emissions to the extent that visibility is not noticeably degraded more than it would be under natural conditions. (3) Standards have been developed and enforced by U.S. Environmental Protection Agency (EPA) and local authorities to regulate particulate emissions from anthropogenic sources. These emissions can be regulated with an opacity-based standard and/or a mass-based standard. Opacity of a plume has been described as "the fraction of a light beam, which during its passage through a plume, is removed from that beam by absorption and/or scattering." (4) Opacity standards are established primarily because plumes can be more readily monitored by their optical rather than their mass-based properties. Both mass- and opacity-based standards should be met for stationary sources of PM. However, opacity standards established by EPA are reportedly more lenient than the corresponding mass emission standards, and, thus, a violation of opacity standard is an indicator but not proof of a violation of a mass emission standard. (5)

Methods for measuring the opacity of a plume emitted from a stationary source include the following: in-stack transmissometers, light detection and ranging (LIDAR) systems, visual observations by humans (Method 9 (6)), and digital photography-based techniques. An in-stack transmissometer quantifies plume opacity within the exhaust stack of the source and requires installation and maintenance of each transmissometer at each source. An in-stack transmissometer can cost more than U.S. $10,000. LIDAR is used as a research instrument to quantify opacity. LIDAR is even more expensive than in-stack transmissometers, is complicated to use, and is reported to underestimate plume opacity. (7,8) Method 9 is the most common way to quantify the opacity of plumes that are emitted from stationary sources because of EPA regulatory requirements.

EPA requires "qualified observers" to pass a "smoke school" test by achieving an individual opacity error (IOE, [d.sub.i]) of 15% or less and an average opacity error (AOE, [bar.d]) of 7.5% or less for all 50 black and white plumes evaluated during a particular test. IOE with unit of percent is defined as the absolute difference between an opacity value, [O.sub.1,i], that is obtained by a human or a digital camera observation and a corresponding opacity value, [O.sub.2,i], that was measured by a reference in-stack transmissometer, as described by:

IOE [equivalent to] [.sub.i] = |[O.sub.1,i] - [O.sub.2,i]| x 100 (1)

where subscript i represents each corresponding observation and measurement. AOE is defined as follows:

AOE [equivalent to] [bar.d] = [1/N] x [N.summation over (i = 1)] [d.sub.i] (2)

where N is the total number of corresponding observations and measurements for a particular test. The accuracy of a Method 9 qualified observer to determine a plume's opacity, as indicated by IOE and AOE, is influenced by the plume's background, as well as ambient lighting conditions (e.g., orientation of the sun with respect to the camera, the time of day, and the geographic location of the source (9)). Passing the test qualifies the observer to make plume opacity measurements for 6 months.

Digital photography-based methods to determine plume opacity have been under development for a few years. (10) The cost can be reduced when compared with other methods, because digital photography-based methods do not require expensive instruments, such as a transmissometer or LIDAR. Plume opacity can be determined with digital photography with a low-cost digital camera (e.g., $300) and a personal computer to interpret the photographs (e.g., $500). Furthermore, travel to conventional smoke schools to teach and certify "smoke readers" every 6 months is not needed. Protocols to use digital photography to determine plume opacity and to certify "digital smoke readers" can be provided through the Internet. The objectivity of the measurement to determine plume opacity is also improved, because a predesigned algorithm is used to compute the plume's opacity. Finally, permanent photographic records of the plume and its surroundings at ambient conditions are also available. These records can be useful for demonstrating compliance to regulators and providing evidence during possible legal actions.

One research group has used digital photography to quantify the opacity of plumes from stationary sources with the Digital Opacity Compliance System (DOCS). (10) DOCS was developed to use a specific digital camera that self-calibrates for clear-sky backgrounds. The user can quantify a plume's opacity with DOCS software by downloading digital images that describe the plume and its background to a computer and then select an area in the photograph that includes a part of the plume where opacity is to be determined, as well as its background. The plume opacity is then calculated using the computer algorithm provided by DOCS. DOCS was tested with clear skies at a high mountain desert, (10) cloudy skies with mild temperature and moderate wind, (11) and overcast skies with freezing temperature and light rain that was mixed with snow. (11) DOCS was also tested at commercial and industrial sites during rainy and misty conditions. (12) Other than being tested at smoke schools, DOCS was also tested under regulatory enforcement conditions and was compared with results from Method 9. (13) Most recently, DOCS was tested using a range of commercially available cameras in lieu of a specific digital camera that was required to be used for the previous field campaigns. (14)

A new digital photography-based method, Digital Optical Method (DOM), was developed recently to quantify plume opacity. (15-17) DOM can be used with readily-available, low-cost commercial digital cameras, with a wide range of existing backgrounds or by installing a more optimal background and with a wide range of meteorological conditions by using either a contrast model or a transmission model. DOM was initially tested under carefully controlled conditions by completing a specifically designed field campaign independent of a smoke school and under more realistic conditions by participating in a smoke school. In addition, differences in the performance of the cameras were characterized by yielding consistent results as long as they are calibrated for response curves. (16) For example, the average absolute difference (AAD) is 3% between the results obtained during the previous study (16) using two cameras (Canon G3 and Sony DSC S30). More thorough studies were completed later to evaluate the feasibility and flexibility of applying DOM during those more realistic conditions by comparing the opacity values obtained from DOM to opacity values obtained from a reference in-stack transmissometer and smoke school trainees during a smoke school. Opacity measurements from DOM and from a qualified observer were also compared when monitoring industrial sources in the field. Results from these two later field campaigns are presented in this paper.

EXPERIMENTAL WORK

DOM