|Use of Open-Path FTIR Spectroscopy During Site Remediations|
Assessment of the air migration pathway represents a significant aspect of many hazardous waste site remediations. Compliance with pre-established health-based action levels must be demonstrated in order to protect on-site workers and nearby residents. This can be an especially difficult task based on use of traditional point monitoring.
The nature of atmospheric plume dispersion, in conjunction with the need to consider acute health impacts arising from short-term contaminant exposure, has often resulted in either the implementation of ineffective remediation air monitoring programs which, unfortunately, are not protective of human health or, conversely, the performance of site remediations at a painstakingly slow pace due to an excessive level of conservatism in the air monitoring results. This over-conservatism arises directly from an inability to adequately address the need for real-time data or the need for spatially representative data, or both. Analytical methods which require sample collection and subsequent off-site laboratory analysis cannot meet the requirements for real-time data. Similarly, point monitors (or samplers) which can characterize the air only at a single point in space cannot meet the requirements for spatially representative data, unless many such monitors are employed at a considerable cost.
Fourier-transform infrared (FTIR) spectroscopy can be used together
with on-site meteorological data to provide on-going assessment of
action-level compliance, in real time, for a virtually unlimited downwind
receptor field, thereby overcoming the limitations associated with
use of point monitors. As discussed in detail, this method involves
first, the continual back-calculation of site-specific emission rates,
and second, the prediction of downwind concentrations (and, thus,
assessment of action-level compliance) along the site perimeter and
at all identified "sensitive
Open-path FTIR spectroscopy is able to provide real-time, simultaneous analysis of up to several dozen gaseous contaminants. The technology is identical in principle to classical laboratory FTIR spectroscopy, except the cell into which a sample would be injected is extended to the open atmosphere. A beam of light spanning a range of wavelengths in the near-IR portion of the electromagnetic spectrum (approximately 2 to 14 microns) is propagated from the transmitter portion of the instrument. In the most common configuration, a "retroreflector," comprised of an array of corner-cubed mirrors, is positioned to intercept this radiation and redirect it back upon itself to the receiver portion of the instrument.
As described by Grant,(1) an interferometer splits the returning beam of radiation into two paths, and then recombines them in a way to generate an interference from the phase differences. The phase difference, and thus the interference, is dependent on the wavelengths present in the beam. In one of the paths, the radiation is reflected off of a moving mirror, resulting in an intensity variation which is measured at the detector as a function of the path difference between the two mirrors. The result is an interferogram.
The interferogram obtained from a monochromatic beam is simply a cosine wave. The broadband interferogram is a sum of cosine waves (the Fourier series) for each spectral component as a function of mirror pathlength separation. A spectrum in the optical frequency units, cm-1, is obtained by performing a Fourier transform on the interferogram.
Contaminants of concern are identified and quantified via a computer-based spectral search involving sequential, compound-specific analysis and comparison to the system's internal reference spectra library. The most widely employed technique for analyzing FTIR spectral data is the multicomponent classical least squares (CLS) technique developed by Haaland and Easterling.(2) Any gaseous compound which absorbs in the IR region is a potential candidate for monitoring using this technology.
One-way pathlengths can range from less than 10 meters (as in the case for combustion source stack monitoring) to several hundred meters or more (as may be required for many ambient air applications).
Gaseous contaminant concentrations are generally reported in units of mass of contaminant per volume of gas, such as micrograms per cubic meter (ug/m3), or volume of contaminant per volume of gas, such as parts per billion (ppbv) or parts per million (ppmv). Path-integrated concentrations, however, are usually reported in units of parts-per-million-meters (ppm-m). For reasons which will become apparent, it is often desirable to convert path-integrated concentrations (ppm-m) to units of milligrams per cubic meter times meter (mg/m3 x m), or mg/m2.
For an open-path FTIR spectrometer, the total contaminant burden is measured within the approximate cylinder defined by the finite cross-sections of the light beam at each end and the length of the beam itself. This contaminant burden is then normalized to a pathlength of 1 meter. If, for example, a path-integrated concentration of 30 ppm-m is reported, no information concerning the contaminant distribution within the beam can be directly inferred, and the instrument response would be identical whether there was a uniform concentration of 30 ppmv over a distance of 1 meter, 3 ppmv over a distance of 10 meters, 300 ppbv over a distance of 100 meters, or 30 ppbv over a distance of 1 kilometer.
It is immediately evident that the integrated concentration reported is directly proportional to the total pathlength for a given uniform contaminant concentration. It also follows that for a site from which contaminants are emanating in a plume of narrow width (e.g., 10 meters), the same path-integrated concentration will be reported regardless of pathlength, as long as the narrow plume remains contained within the observing pathlength and there is no upwind (or background) contaminant contribution.(3)
The generation of a path-integrated concentration yields contaminant information along the entire pathlength, and not just at a single point (or collection of points) in space as with traditional point-monitoring methods. This solves the issue of spatial data representativeness, as a non-buoyant ground-level plume cannot pass through the beam path undetected.
One may divide the path-integrated concentration by the pathlength to obtain an average concentration along the pathlength, but this concentration representation is of limited value when dealing with action-level averaging times typical of acute exposure assessment.
The following benefits are identified for use of this technology in site clean-up:
A general perception exists that open-path FTIR spectroscopy is an expensive alternative to traditional air monitoring methods for site clean-up applications. This is a misconception arising from what turns out to be an "apples and oranges" comparison. When compared to a traditional air monitoring program which is able to meet the necessary site clean-up data quality objectives, an open-path FTIR-based program is far less expensive. A typical cost for a 1-month program involving a single open-path FTIR unit with full upwind/downwind coverage would be on the order of about $45,000. This includes all mobilization and demobilization activities, labor and equipment, and QC activities to ensure the technical validity and legal admissibility of the data.
The same program based on an automated gas chromatography network consisting of one upwind and eight downwind monitors would cost on the order of $85,000. However, even with this number of downwind monitors, data representativeness is only marginally achievable, even for a small site. It would be difficult, if not impossible, to ensure the plume does not migrate off-site undetected, especially under stable atmospheric conditions. By way of illustration, even at a downwind distance of 100 meters, one need only move 12.8 meters away from plume centerline (i.e., normal to the wind direction) to see a full 90% reduction in concentration (point-source release) when the atmosphere is stable.
Speed and Versatility
Library spectra exist for several hundred compounds, and new ones can be created within a few days for virtually any gaseous compound which exhibits IR absorption. Today, more than 40 compounds can be monitored simultaneously, with quantitation available within 30 seconds of data collection. Off-site contaminant exposure, via back-calculation of emission rates and subsequent modeling of downwind concentrations, can be assessed within about 1 minute.
Path-averaged minimum detection limits (MDLs) are generally in the single-digit-ppb range based on a pathlength of 100 meters. This is usually more than sufficient for assessment of action-level compliance for acute exposure.
An infinite Asample holding" time exists, as analysis information is stored as an electronic document. This means that the data can be reexamined at a later date for evidentiary reasons, or even reanalyzed should an additional target contaminant be later identified. Any sample collection error is eliminated, as there is no "sample" per se; the media is unaffected by measurement method.
Finally, no calibration is required as the instrument is intrinsically calibrated. Only daily precision and accuracy assessments need to be made in accordance with procedures set forth in Toxic Organic Compendium Method 16 (Compendium Method TO-16).
Documentation of Contaminant Exposure
The ability to generate a continual assessment of action-level compliance for an unlimited downwind receptor array can be important in reducing responsible-party or government liability associated with unsubstantiated future claims involving exposure (worker or public) to unknown contaminants during site clean-up. Another benefit of exposure-documentation capabilites concerns personal protection. For example, field decisions to downgrade personal protection levels (e.g., from Level B to Level C) can be supported by generation of real-time action-level compliance data.
It has been our experience that the "high-tech" nature of the open-path FTIR technology invariably leads to community appeal and positive public perception. Total fenceline coverage (the "eye which never sleeps") allays public fear. Such community appeal, in turn, benefits regulatory agencies, as there is less opposition to the selected cleanup remedy.
EMISSION-RATE ESTIMATION TECHNIQUES
The inability to assess acute exposure based on the direct use of path-integrated data would, on first thought, seem to be a drawback. However, when coupled with onsite meteorology, this type of data is actually unparalleled, as all of the limitations associated with traditional point-monitoring approaches are eliminated. Action-level compliance can be assessed, in real time, for a virtually unlimited downwind receptor field.
The cornerstone of this methodology is the back-calculation of contaminant emission rates. Rather than relying on receptor monitoring for a direct assessment of action-level compliance, an accurate emission-rate estimation facilitates application of traditional dispersion modeling in predicting action-level compliance for any locations of concern (e.g., site perimeters and sensitive off-site receptors such as residences and schools). Because 5 minutes of coadded spectra are more than sufficient from a precision and accuracy perspective, it is a straightforward task to generate a new, site-specific emission rate -- and a corresponding assessment of action-level compliance -- up to 12 times each hour.
To estimate the health impacts to downwind receptors, reliance upon some type of conservative dispersion model offers the only practical alternative. Actual concentrations could be continuously measured at each receptor of concern, but this activity is generally both cost- and labor-prohibitive. All dispersion models rely upon accurate estimates of emission rates. The ability to provide accurate emission-rate estimates continually and in real time is the key to the power of the path-integrated concentration.(4)
specific back-calculation techniques appropriate for action-level
compliance are discussed below.
Within classical Gaussian dispersion theory, the general equation for concentration calculated at ground-level (z = 0) for a continuously emitting point source is given as follows:(5)
χ (x,y,0;H) = Q (πσyσzu)-2 exp [-2 (y/σy)2] exp [-2(H/σz)2] (Eq. 1)
= downwind distance to a receptor, m
= vertical distance to a receptor, m
= uniform emission rate of contaminant, g/s
σz = standard deviation of
plume concentration distribution in the vertical direction at the
distance of measurement, m
This relationship forms the basis for many of the USEPA atmospheric dispersion models currently employed for estimating downwind air quality impact.
Examination of this relationship shows that the downwind concentration at a given location increases with increasing source strength, but decreases with increasing wind speed and horizontal and vertical dispersion (as determined via σy and σz). The standard deviations of the plume concentrations in the horizontal and the vertical are, in turn, functions of atmospheric stability and the distance downwind of the source. Nomographs which define σy and σz as a function of downwind distance for each of six stability classes are frequently used to estimate these parameters. Larger σy and σz values are associated with unstable atmospheric conditions (greater dispersion) and greater downwind distances.(6)
If one integrates Equation 1 in the y (cross-plume) direction, the resultant representation is a crosswind-integrated concentration instead of a point concentration. Performing this integration with respect to y, from y = - 4 to +4, yields:
C(x,0;H) = 2Q [(2π)2 σz u]-1 exp [-2(H/σz)2] (Eq. 2)
C = ground-level crosswind-integrated contaminant concentration at distance x, g/m2
Equation 2 has historically been employed in diffusion experiments to determine vertical dispersion coefficients (standard deviations of the plume concentration in the vertical direction), σz, from ground-level data where the source strength, Q, was known and the ground-level crosswind-integrated concentration was determined from a crosswind line or arc of point-sampling measurements made at some predetermined downwind distance.(7)
The effective height of emissions, H, is defined as the sum of the actual height of emissions and the buoyancy-induced height increment arising from an elevated effluent temperature. Because most site remediation activities occur at ground level and without elevated effluent temperatures, H generally equals zero and Equation 2 reduces to:
C(x) = 2Q [(2π)2 σz u]-1 (Eq. 3)
Rearranging, Equation 3 may be written as:
Q = 2(2π)2 C(x) σz u (Eq. 4)
Equation 4 is the general emission-rate equation for a point source involving path-integrated measurement data. For a measured crosswind-integrated concentration at some specified downwind distance, the emission rate, Q, depends only upon σz at that distance and on wind speed, u. The point-source emission-rate technique is applicable for those site disturbance activities which may be approximated as point sources (e.g., excavations).
The tracer-ratio technique is appropriate for estimating emission rates from any type of site disturbance activity (i.e., point source or area source) and, in contrast to the point-source technique, does not rely on the contaminant distribution in the plume being Gaussian.(8) The tracer-ratio technique involves the release of an appropriate tracer gas (such as sulfur hexafluoride) at a known, controlled flow rate from locations which adequately simulate the source geometry. Assuming that the tracer and source plumes are fully contained by the downwind FTIR beam, the following ratio applies:
/ Q = CT / QT (Eq.
= ground-level crosswind-integrated concentration of tracer
at distance x, g/m2
Equation 5 simply states that the ratio of the path-integrated concentration of the contaminant to its emission rate is equal to the ratio of the path-integrated concentration of the tracer to its emission rate. (It is important to note that all concentrations must be expressed in units of g/m2 or mg/m2, as use of ppm-m units will yield erroneous results owing to the fact that molecular weights are unaccounted for.) Rearranging Equation 5 and solving for Q yields:
Q = (QT C) / CT (Eq. 6)
the emitting source is not too large, a tracer will typically be released
from a single point positioned at the source edge furthest upwind.
The simplicity of such a source simulation generally outweighs the
resultant increased conservatism (i.e., higher emission rates).
area-source technique is simple to implement and can be used to estimate
emission rates from area sources which are too large for simple treatment
via the tracer-ratio technique. The technique is applicable for both
homogeneous and nonhomogeneous sources (i.e., sources which emit uniformly
and sources which have "hot
spots"). However, for nonhomogeneous sources,
some information on the extent and magnitude of the hot spots is required.
If no hot-spot information exists, it is possible to generate reasonable
bounds upon the site emission rate. Like
the point-source technique, the area-source technique does not involve
use of a tracer gas and the plume is generally assumed to obey Gaussian
dispersion theory. The following four-step methodology is employed.(9)
step involves making ground-level FTIR measurements upwind and downwind
of the source to identify source attribution. The instrument background
will typically serve as the upwind measurement, and site attribution
is obtained by subtraction. It is essential that the downwind pathlength
be of a magnitude sufficient to encompass the entire width of the
This step involves use of an appropriate dispersion model, preferably the ISCST (Industrial Source Complex Short-Term) Model, to predict point concentrations along the downwind FTIR measurement path at a nominal receptor spacing of 1 or 2 meters. Relative emission rates are modeled (i.e., unity emissions, with hot-spot subareas represented as multiples of unity) based on actual meteorology and source configuration.
Site-specific σz values based on tracer releases are generally preferable to model (textbook) σz values, and should be substituted to back-calculate emission rates whenever possible. Equation 4 can be rearranged, as follows, to facilitate site-specific σz calculation:
σz = [(2π)2 QT] / πCT u (Eq. 7)
By knowing QT, CT, and u, a site-specific σz value is calculated directly. However, because σz is a function of stability and downwind distance, a curve comprised of σz values at several downwind distances should be generated for the range of stability classes expected to be encountered. Similarly, the downwind distances at which σz is measured should span the range of downwind distances to be encountered during site-disturbance activities. All tracer work should be carried out in advance.
3. Integrate the Function Defined by the Point Concentrations Along the Measurement Path
Some type of rudimentary numerical technique will generally be required to integrate this function (e.g., Simpson's Three-Point Rule, in which the line representing the value of the function is replaced by a second-order equation, y = ax2+ bx + c). The resultant path-integrated concentration is what the FTIR is predicted to "see" based on the relative emission rates used in the dispersion modeling.
4. Scale Modeling Results to Estimate Area Emission Rate
The actual contaminant emission rate, Q, is estimated in a manner which is conceptually similar to the tracer-ratio technique:
CM / Q = CP / QR (Eq. 8)
CM = measured ground-level crosswind-integrated contaminant concentration at distance x, g/m2
CP = predicted ground-level crosswind-integrated contaminant concentration at distance x, g/m2
QR = relative emission rate of contaminant, g/s
Equation 8 simply states that the ratio of the measured path-integrated concentration to its emission rate is equal to the ratio of the predicted path-integrated concentration to its emission rate. Rearranging Equation 8 and solving for Q yields:
Q = (QR CM) / CP (Eq. 9)
The Michigan Avenue Dump Site, a 1.8-acre hazardous waste site located in Canton, Michigan, was used by 3M Corporation during the 1960s to dispose of industrial wastes. In 1993, an imminent threat to public health was identified by the USEPA, Region 5, due to large volumes of waste materials entering the Rouge River, which cut through the center of the site. In July of that year, an emergency removal action was initiated, and wastes were excavated and hauled away for offsite disposal while contractors shored up the riverbank with sheet piling.
The USEPA identified a potential for significant offsite exposure to airborne gaseous contaminants generated during excavation and stockpiling of contaminated waste materials. After extensive Agency review of available monitoring methods and based upon ongoing consultation with USEPA-ERT, open-path FTIR spectroscopy was selected as the technology to Adrive" the action.
While working for Blasland, Bouck and Lee, we were retained by 3M's consultant, Roy F. Weston, to design and implement the air monitoring program. The objective was to ensure that emissions generated during the excavation and off-site transport of waste materials did not exceed the health-based property-line exposure levels established by the USEPA for this site, and to support the application of vapor suppressants whenever action levels were approached.
Exhibit 1 identifies a total of 15 target contaminants and associated 30-minute action levels developed specifically for this emergency removal action.
Exhibit 1. Target Contaminants and Associated 30-Minute Action Levels
* As approximated by the sum of n-octane and iso-octane.
Open-path monitoring was performed, using a single FTIR unit, in such manner as to provide full coverage of the site perimeter, regardless of the wind direction. The instrument was positioned at the NW corner of the nearly rectangular site and could pivot to monitor along either the W or the N leg of the site. Flat mirrors were placed in the NE and SW corners to "bend" the beam along the E and S legs, respectively, and retroreflectors were positioned in the SE corner to send the beams back upon themselves to the FTIR for analysis. Up to six 5-minute-averaged (70 coadded spectra) path-integrated downwind measurements were made each hour.
The tracer-ratio technique was used to back-calculate emission rates for the 15 target contaminants. The source (area of site disturbance) was represented as a virtual point conservatively positioned at the upwind site perimeter.
A proprietary plume dispersion model software package (SPECTRAMET) was used to assess action-level compliance based on the back-calculated emission rates and on meteorology supplied by a portable 3-meter meteorological tower equipped to generate 5-minute averages of wind speed and wind direction.
SPECTRAMET was configured to generate maximum predicted fenceline concentrations (mg/m3) in near real-time (within 15 or 20 minutes of actual occurrence) approximately twice each hour for the duration of waste disturbance or vapor suppressant activities, or on demand by the Weston field manager. Whenever an action level was exceeded, waste disturbance activities were immediately stopped and a vapor suppressant applied. Activities could not recommence until maximum fenceline concentrations fell to background levels.
The local press was successful in gaining community support for the technology. ALike something right out of the Star Wars defense initiative, the Fourier Transform Infra-Red system has been doing some surreptitious defense work of its own in Canton.
ABy quietly and inconspicuously testing the air at a Michigan Avenue dumpsite, the system -- introduced by 3M to monitor its cleanup of the site -- has been defending residents against the possible inadvertent release of any harmful chemicals."(10)
Because of the proximity of the emission sources to the site perimeter, action-level exceedances occurred frequently and 11 months was required for completion of the entire emergency removal action. However, during the course of the project, the USEPA stated that if open-path FTIR spectroscopy had not been utilized in the manner it was, the whole operation would had to have been performed under an enclosure, at a greatly increased cost to 3M.
1. Grant, W.B. (1992). Optical Remote Measurement of Toxic Gases. Journal of Air & Waste Management Association. Vol. 42, No. 1, p. 18.
2. Haaland, D.M. and Easterling, R.G. (1982). Application of New Least-Squares Methods for the Quantitative Infrared Analysis of Multicomponent Samples. Applied Spectroscopy. Vol. 36, p. 665.
3. Minnich, T.R., Scotto, R.L., Leo, M. R., and Solinski, P.J. (1993) "Remote Sensing of Volatile Organic Compounds (VOCs): A Methodology for Evaluating Air Quality Impacts During Remediation of Hazardous Waste Sites." Sampling and Analysis of Airborne Pollutants. Chapter 15, p. 237. Lewis Publishers.
5. Turner, D.B. (1994) "Work Book of Atmospheric Dispersion Estimates - An Introduction to Dispersion Modeling." Lewis Publishers.
6. Minnich et al., 1993.
8. Minnich, T.R., Scotto, R.L., and Kricks, R.J. (1992) "Field Standard Operating Procedure for the Use of Open-Path FTIR Spectroscopy at Hazardous Waste Sites." USEPA preliminary draft.
10. The Community Crier. (1993) "Clean Up Date - 3M Gets EPA Approval; Begins Cleanup of Canton Dumpsite." Alex Lundberg. July 14, 1993.
© 2002 Minnich and Scotto, Inc.