Toxicologic Assessment of the Army's Zinc Cadmium Sulfide Dispersion Tests (1997)

PREPARED BY

EDMUND CROUCH

CAMBRIDGE ENVIRONMENTAL, INC.

CAMBRIDGE, MASSACHUSETTS

This document has been prepared for the Subcommittee on Zinc Cadmium Sulfide to summarize some estimates of concentrations and potential exposures (time-integrated concentrations) of zinc cadmium sulfide (ZnCdS) that were achieved during the use of this compound in certain air-dispersion tracer tests. The basic source materials were a subset of the references included in an appendix entitled "Zinc Cadmium Sulfide Testing Documents," which listed all known Army-sponsored ZnCdS tests. All reference numbers herein refer to the list for this appendix. The description of the aim of this report is best described by the scope of work:

Dr. Bakshi has provided Cambridge Environmental Inc. with approximately 50 documents (see the reference list for this appendix), most of which contain details of various U.S. Army-sponsored tests using ZnCdS tracer (some provide details of multiple tests in one or more series).

Cambridge Environmental, Inc., reviewed these documents and abstracted summaries of them, providing the following details for each test or series of tests whenever such details were available in the documentation:

Not all such details were available in all cases, although certain details always appear to have been provided. In all cases examined, measurements and estimates were made of time-integrated concentrations at various locations.

The two measurements of exposure-time—integrated concentration and concentration—are those required for estimating risks using the standard paradigms. In all cases, the release times were relatively short (<24 h), so the maximum concentration was compared with a short-term safety standard. Estimates of lifetime cancer risk are generally made using

where R is the lifetime risk estimate, U is the unit risk (m³/µg), T is a standard lifetime (70 years), C is the time-varying concentration, and t is time. This estimate involves the time-integrated concentration directly.

Throughout this document, various abbreviations and standard nomenclatures are used. The most important of these are the terms used to describe the two measurements of exposure or potential exposure evaluated. These two terms are

All the documents examined use the nomenclature "dose" to represent the time-integral of the concentration measured in particles of ZnCdS per unit volume (usually measured as particle-minutes per liter).

An attempt has been made to distinguish between concentrations and exposures measured in (humanly) populated areas and in unpopulated areas, although this attempt necessarily is limited by the available information.

Individuals, particularly those involved in the experiments, might have been subjected to the concentrations or exposures listed as being in unpopulated areas; in some cases, no one might have been subjected to the concentrations and exposures listed as being in populated areas.

The material of interest is fluorescent ZnCdS. This nomenclature is used for convenience only. The material used should probably be considered an alloy of zinc and cadmium sulfides, with the fluorescence color determined by minor additions of other elements. The stoichiometry is approximately Zn_0.8Cd_0.2S, but all masses used in this document refer to the total mass of the ZnCdS—no correction to obtain a mass of cadmium (the principal component of concern to the subcommittee) has been applied.

Some of the abbreviations used in the following summaries are

The entries in Table B-1 are the following:

See the summary reports that follow for important caveats on all the reported values. In particular, all exposures are likely to be accurate to a factor of 2 at best, all concentrations will be substantially less accurate than that, and the magnitude of the values given often depends strongly on the placement of the sampling devices. All exposures and concentrations represent the total mass of ZnCdS present. (There is no correction to the mass of only cadmium.)

Exposure and concentration values are not directly comparable between experiments because of the different areas affected and the different placement of sampling points. To give an idea of the variation, the table includes (for the non-Dugway releases) some order of magnitude estimates for the areas affected within approximately an order of magnitude of the maximum concentrations or exposures reported. The areas given are approximate upper bounds on the areas affected, rounded up to the nearest power of 10.

A missing entry generally indicates a lack of, or irrelevance of, a measurement or estimate for that entry (e.g., exposures to unpopulated areas are irrelevant for releases made over populated areas).

In Chapter 5, Table 5-1 is derived from Table B-1. To obtain the cadmium dose, the maximal exposure of ZnCdS (expressed as µg-min/m³ as shown in Table 5-1) is multiplied by 0.0166 m³/min (the volume of air inhaled by an active person in one minute). The product is then multiplied by 0.156 (the mass fraction of cadmium in ZnCdS). The corresponding cadmium doses are 6.8 µg in Minnesota, 14.5 µg in Winnipeg, 24.4 µg in St. Louis, and 390 µg in Biltmore Beach.

An initial scan of the approximately 50 documents available showed that there were too many and too varied a set of experiments to be able to perform complex or even simple analyses beyond those performed by the original experimenters and reported in the references. As far as possible, this report therefore contains simple abstracts of the most relevant data from the references, with as little interpolation as possible, and with only the simplest extrapolations.

The original researchers usually provided substantial documentation. A much more thorough job of reconstruction of exposures would in most cases be possible with substantially more resources, but, in view of the orders of magnitude of the exposures, such a reconstruction would probably not add much value for the committee.

In the references examined, there were two principal purposes for the experiments performed:

• Experimental confirmation of the practicality of dispersion of biologic warfare agents, empirical observations of the ranges and exposures that might be achieved, and obtaining correlations between the dispersal of biologic agents and inert aerosols. The fluorescent tracer was used as safe indicator of the potential extent of spread of biologic organisms under similar meteorologic conditions.

• Tracer tests in the atmosphere, to determine how the atmosphere behaves and derive correlations relating the dispersion of gases and aerosols to observable meteorologic conditions.

Of course, if correlations could be obtained between dispersal of biologic organisms and inert aerosols, the second purpose would also serve the first, although the second also has much wider application.

The fluorescent-particle (FP) tracer technique is extensively described in Ref. 48. The idea is to release into the atmosphere a very large number of small particles as an aerosol that then disperse. The concentration of these particles after dispersion downwind is then measured by counting the individual particles collected by a collection device that samples a known volume of air. ZnCdS, originally developed as a fluorescent paint pigment, was found to be suitable for this application. The useful particle size range is about 0.5 to 5 µm, with the lower end limited by the detectability of the particles, and the upper end limited by the rate of fall-out from the air. The typical experiment used particles with a mass mean diameter of 1 to 2 µm, and an effective dispersal rate of around 1 x 10¹⁰ particles per gram (ppg).

It was found that the best available fluorescent ZnCdS particles could be reliably detected by eye under an optical microscope (with ultraviolet illumination to excite the fluorescence) at sizes down to around 0.3 µm diameter, allowing measurement of concentrations by counting individual particles collected on filters or other collection devices. Background counts in most areas were very low. No automatic particle-counting

methods were described in the references examined. All particle counts were apparently obtained by human microscopic observations, although mechanical methods to improve the efficiency of sample preparation and presentation to the human observer are mentioned.

In a typical experiment, grams to kilograms of ZnCdS powder would be released over periods ranging from seconds to hours using an air-driven dispenser designed to efficiently produce an aerosol cloud. The dispenser would be mounted at a particular location (point source), or on the back of a moving vehicle (line source)—trucks, aircraft, and boats were used. The method of dispersion affected the dispersion efficiency, a measure of how much clumping of the particles occurred during release. Typical particles-per-gram values for particular manufacturers' lots of ZnCdS were apparently¹ obtained by a standardized test involving the dispersal of a small quantity of the material in an enclosed chamber. The particle count would then be obtained by sampling the aerosol with a filter-type sampler. The dispersion efficiency in an experiment is then the ratio of the particles per gram achieved in the experiment to the standardized value. A blending agent was typically used to improve the flow characteristics of the ZnCdS powder and reduce clumping, but environmental conditions could obviously affect the dispersion efficiency attained.

Downwind of the release point or line, sampling stations were set up at locations of interest. These sampling stations sampled the air by drawing a known volume flow rate (typically 6 liters (L)/min) through a filter (an approach that collects practically 100% of the particles on the filter) or by using a Rotorod collector. The latter consists of an H shaped wire (0.016 x 1/16 in.) with vertical members approximately 2.5 in. long, and crossbar 4.75 in. long, rotated upright about the center of the cross-bar. Particle collection is by impact on the 0.016-in.-wide leading edge of the vertical members of the H, which were coated with silicone (vacuum) grease to improve collection efficiency. The effective volume swept by the

Rotorod was about 40 L/min, and collection efficiencies (which were calibrated against filter collectors) ranged from 30% to 80%, depending on the ZnCdS particle size distribution. Two independent samples could be obtained on each Rotorod by reversing the direction of rotation and using the other edge of the wire to obtain the second.

Sampling was typically started just before the release, and continued for a period lasting longer than the expected time during which any cloud of particles would be expected to be present. In some cases the time course of the cloud of aerosol was followed by using a sequential drum sampler, a type of filter sampler with a set of filters that could be automatically sequenced mechanically.

Filters and Rotorods would then be taken to a laboratory, and the collected particles counted. Reported results were always given in terms of exposure—typically as particle-minutes per liter.

The advantages of the FP tracer technique include its very high sensitivity, and its applicability over a wide range of scales. Statistically reliable particle counts require only 40 detected particles (for 15-20% accuracy), and with a collection volume flow rate of 6 L/min, this corresponds to an exposure of around 6 particle-min/L. At a count of 1 x 10¹⁰ ppg, this corresponds to 0.6 µg-min/m³, or an average of 0.04 µg/m³ for a 15-min sampling period. The technique was applied over scales that ranged from almost backyard to multistate.

All the references examined reported exposures in particle-minutes per liter or gave tabular data of raw particle counts (e.g., particles per filter). These values were, in general, taken at their face values (in one case (Ref. 19) a clearly incorrect description of tabular data was corrected). Reported values of particle-minutes per liter had generally been corrected

for the collection efficiency of the sampling device, and these corrections were not questioned. Where only raw particle counts were given, the collection efficiency was also available, so that a correction could be applied for collection efficiency. The Dugway Proving Ground tests often gave only raw count data, but in the references examined these always used filter-type collectors with effectively 100% collection efficiency. In some of these cases, the sampling rate for the filter sampler was not given, and was assumed similar to the sampling rate in previous or subsequent experiments. Details are given in the individual summaries that follow.

The conversion from particle-minutes per liter to microgram-minutes per cubic meter (the unit of exposure used throughout this document) requires a particles-per-gram measurement for the dispersed aerosol in each experiment. Where this value was provided in the references, it was used, taking account also of any cited or cross-referenced dispersion efficiency for the device used in the particular experiment.² Where no dispersion efficiency was cited, but only a particles-per-gram count was given, that particles-per-gram count was generally taken to be the effective particles-per-gram count for the experiment (i.e., incorporating the dispersion efficiency), provided that this interpretation was consistent with the language of the reference and the result was within a reasonable range. Where no particles-per-gram count or dispersion efficiency was given, generic estimates based on the ZnCdS lot number or supplier and other experimental results were used to estimate the effective particles per gram of the dispersed aerosols. More details are given where necessary in the individual summaries that follow.

The estimates of particles per gram for the dispersed aerosols is necessarily approximate, and applying this estimate to the aerosols that were collected is doubly so. For these reasons, the exposure estimates are not considered accurate to better than a factor of two even in the best circum-

stances. All exposure estimates in this document have thus been scaled (to microgram-minutes per cubic meter) to take account of both dispersion and collection efficiencies.

Exposures were often measured at many points, and the exposures varied substantially with location—particularly with distance from the sources. All exposure estimates given in this document correspond as closely as possible to the maximum exposures measured at any given point in any particular experiment. Where there were multiple releases, the exposures due to each release were summed at each measurement point, and the value abstracted here is the maximum of those sums. In a few cases, documented in the summaries that follow, it was impossible³ to obtain the sum of the exposures at every measured point. In such cases, the best estimate available, or a conservative overestimate, for the maximum sum at any given point was used instead (e.g., in some cases, the maximum exposures—at different points—from each release were summed to obtain an overestimate of the maximum that could occur at any point).

In several cases, no tabular data were available, and it was necessary to estimate maximum exposures from contour data plotted on maps. Although individual measurement points were always also plotted, such individual points were often unreadable with the copies of the maps available. Because the contours plotted were widely spaced (often only in decades of particle-min per liter) and not completely consistent from map to map (some maps interpolated extra contours between some decades of particle-min per liter) and because the legends were often unreadable, this technique necessarily introduced further errors. The combined errors of reading from maps could amount to a factor of 5 in the exposure estimate in some cases.

The convention of using the highest measured value results in substantial

variability due to the selection of sampling points. Exposures might vary dramatically with distance from a source, particularly from ground-level point or line sources, and particularly at small distances from the source. However, this approach was adopted because it requires the minimum extrapolation from the reported data. As mentioned in the Introduction, it would probably be possible in many cases to perform modeling to obtain a more standardized value for exposure, but such an approach would require resources beyond those available for this document.

Where appropriate and possible, separate estimates have been made for exposures in populated and unpopulated areas. Unpopulated areas were defined as areas not containing any buildings or street layouts indicating a built-up area on contemporary maps (if they were provided in the original reference) or on a modern computerized street map. ''Unpopulated'' areas included Dugway Proving Ground, the middle of a national forest, and the middle of a desert area. Caveats about such interpretations are provided in the individual summaries that follow.

The methods used in these references did not allow direct measurements of concentrations, so that some inferences are required to provide estimates of concentration. The approach generally was to use as little extrapolation as possible, so that the concentration estimates are based as closely as possible on the measurements. However, the uncertainties in maximum concentrations are necessarily substantially larger than those in exposures.

For the purposes of this committee, the instantaneous concentration is of little consequence, since all concentrations are sufficiently low that no acute effects are expected under any circumstances (see Ref. 48 in Section 4.2, for example, where a short discussion of potential toxicity is abstracted from Ref. 48). Thus, whenever time-resolved data from an experiment were presented in the references, those data were used to estimate average concentrations over the resolution period of the data, even though the resolution varied from 10 s to 6 min.

The varied data collection and presentation methods given in the references necessitated varied approaches to estimating concentrations. The following techniques, or combinations of them, were used in different cases to estimate the maximum concentration that occurred during any single release at any measurement point. The methods used in each case are mentioned in slightly greater detail in the summaries that follow.

• If time-resolved data were available, those data were used. Time resolution varied from 10-s to 6-min averages, and those data were obtained with drum samplers. Where the time-resolved data did not coincide with the maximum exposure, the time-resolved data were adjusted upwards in proportion to relative exposures. This technique assumes a similar concentration-versus-time variation at both such points.

• For extended period point sources, the release periods were used to estimate the time the aerosol cloud could be present at any particular sampling point, and the exposures measured at those sampling points divided by the time periods to obtain an average concentration.

• For some instantaneous point sources, measured plume sizes allowed an estimate of the cross-wind standard deviation (σ_y) for an assumed gaussian shape. A similar standard deviation (σ_y = σ_x) was assumed for the downwind direction. Combining an assumption of downwind gaussian shape with the measured wind speed then allows an estimate for the peak concentration.

• For short releases from point or line sources, a combination of the two previous methods was applied. The release period was increased by twice the estimated downwind standard deviation in time (standard deviation in space divided by wind speed) to estimate the period the aerosol cloud might be present, and the average concentration obtained by dividing exposure by this period.

• Where no time-resolved data or aerosol-cloud-size data could be estimated from the measurements, but meteorologic data were available,

the cloud was assumed to be gaussian in time, and the standard deviation in time of the concentration estimated by dividing σ_x by the wind speed, obtaining σ_x by assuming σ_y = σ_x, and estimating the former σ_y using the meteorologic conditions (using the standard estimates used in all EPA air-dispersion models).

• If no other methods could be used and some measurements of plume spread and wind speed were available, a standard line-source dispersion model was used for long-distance plume travel under inversion conditions from a line source. The measured value of σ_y was matched to the model to select the stability class, or the total exposure was matched to the measured exposure, and the model then provided an estimate of maximum concentration.

These approaches do not exhaust all possibilities used, because almost all references required a slightly different approach. For example, some methods used can be summarized as follows:

An estimate of the area affected was not included within the original scope-of-work, and much less effort has been devoted to these estimates, which are provided only in the summary table in Section 2 for the non-Dugway releases. The estimates of areas affected have been obtained by roughly estimating the length by width of the measured or inferred plumes in each experiment (or the angle and radius affected for point sources). The area so obtained (in square miles) was rounded up to the nearest power of 10, and entered in the table with a "less than" (<) indication. Where plumes went in substantially different directions in different releases, the area affected listed in the "Exposure" column in the table might be larger than that in the "Concentration" column, the latter indicating the size of individual plumes. The "size" of an aerosol plume is fairly arbitrary, since it depends on the definition given to the edge of the plume—much of the material in these plumes could travel over thousand mile distances if not rained out. As a rough approximation, wherever an explicit bound was needed, the distance to a concentration or exposure approximately 10 times lower than the maximum was used.

There are some cases in which a rough estimate of the plume size can be made (by using details of the experimental design), even though no estimate can be made of the concentration or exposure.

These references are probably available.

There is evidence in the paper that experiments up to 10 years old have been re-examined, but no data are presented.

Rotorod efficiencies observed to range from 28% to 73% for FP 2266 with different particle-size distributions. First for MMD of 1.8 µm, 7.9 x 10¹⁰ ppg, second for MMD 3.1 µm and 1.6 x 10¹⁰ ppg.

Several tables contain particle counts per gram.

The last lot is described as being used extensively by the GE Hanford group. The report mentions a series of small-scale trials (SAL FE 132, 29 Aug. 1960) (FE, field experiment) in which mixtures of FP 2267 and FP 2210 were aerosolized and sampled downwind. These were to be described in a subsequent report. Six releases were completed in an afternoon, involving release rates of 2-3 g/min (period not specified) with samplers 700 ft downwind.

Stanford University (Grinnell, Perkins, Webster, later Hutchison)

Jan. 1946

Contract W-18-035-CWS-1256

Original 2-year contract, running 1/1/46 through 12/31/47

June 1947

Contract W-18-035-CM-147

Next 2-year contract

Dec. 1947

Contract W-18-035-CWS-1256

Original contract extended to 2/28/48 (Ref. 36 covers the results of Contract W18-035-CWS-1256 through the end of 1947.)

Feb. 1948

Contract W-18-035-CM-147 Both contracts combined to run under this number until 12/31/49

The results obtained in W-18-035-CM-147 are not documented in any of the available reports. From subsequent experiment numbering and discussions, it appears that up to 14 experiments were performed over populated areas (numbered FE 1 through FE 14), probably in Palo Alto.

In Ref. 37, the following bimonthly reports are referred to: BMR 4, Oct.-

Nov. 1948; BMR 9; BMR 11, Apr.-May 1949; BMR 13; BMR 14, Oct.-Nov.-Dec. 1949.

Ref. 37 also mentions that experiments FE 13 and FE 14 showed the feasibility of using FPs in 1-lb quantities for a dispersion over a distance of 4 miles or more.

Feb. 1950 Contract DA-18-108-CML-450 Continuation of studies under W-18-035-CM-147.

These are documented in Ref. 37 quarterly reports. However, QR 4 is missing, and any quarterly reports beyond QR 7 are also missing (including any final report).

This chronology incorporates many of the loose ends that were noticed but is almost certainly not complete. It was compiled by noting in each reference any cross-references to other experiments. All dates are said to be those of FP releases described in the cross-reference, or of documents that are said to describe FP releases in the cross-reference. In some cases, there are guesses based on nomenclature, indicated by a ? under the reference.

The notation ?(5) indicates a cross-reference to some release, and it is possible that the unavailable reference given in the parentheses corresponds to that cross-reference. It is also possible that these two are distinct.

The list below shows the following: (1) a range of dates during which releases were said to take place, (2) the reference-list number (Ref.) in which that set of releases is described in detail, or (3) the reference-list number (XRef.) in which that release is mentioned but is not described in detail. A reference number in parentheses indicates that the reference is in the list but was not in the set of references available for this report.

The unnumbered reference is indicated with an asterisk (*).

Note that the people originally at Stanford subsequently or concurrently founded or worked at or for Metronics.

FP use appears to be routine and hardly commented on at Dugway by this time.

The series of field experiments named FE 1 upwards was started by Stanford, and apparently continued by Metronics. Note that there are gaps in the numbering in the documents available.

Ref. 32 used a Mark IX disseminator and refers to the "T3-665" trials testing the efficiency of this disseminator. However, there is little, if any, reference to earlier "Mark" disseminators. Other references are to the D- (L-20) and the L-23 disseminators or to a Stanford generator or a Skil blower.

1. T.B. Smith and K.M. Beesmer. 1967. Dissemination and Evaluation of

a Tracer Material Release. West Site (U) Vol. 2—Part B. Report to Deseret Test Center, Fort Douglas, Utah, from Meteorology Research, 464 West Woodbury Road, Altadena, Calif. Project DESERET. Contract DA 42-007-AMC-162(Y) Phase II. MRI Document MR165 FR288. Mar. 1967.

2. Special Report 273. Comparison of Simulant Decay Rates in Field Tests (U). Program Research, Field Operations and Meteorology Branches, Assessment Division, Fort Detrick, Frederick, Md., Nov. 1956 (supersedes Assessment Division Test Report A-294).

3. Comparative Diffusion of Dissimilar Agents. PGR 149 W5, 6-52. Downwind travel of simulant agents released simultaneously, BW 5 and 6-52. Dugway Proving Ground Report 149. Projects 4-98-05-005 and 4-98-01-001. 5 Feb. 1954.

4. Trial Reports from Dugway Proving Ground, Relationship of Dosages to Source Strength for BG and an Inert Tracer. 24 Feb. 1954.

   Report DPG, BW 8A-1-54, 24 Feb. 1954

   Report DPG, BW 8A-2-54, 9 Mar. 1954

   Report DPG, BW 8A-4-54, 22 Mar. 1954

   Report DPG, BW 8A-5-54, 12 Apr. 1954

5. Dugway Proving Ground, Utah. Oct. 1954 (not available for this report).

6. Trial Report from Dugway Proving Ground. BWAL, BW 8B-7 and 854, Operation "Trouble Maker." Relationship of dosages to source strength for BG and an inert tracer. BWALTR 27. 18 July 1955.

7. Dugway Proving Ground, Utah. 1956 (not available for this report).

8. An Experimental Study of Long Range Aerosol Cloud Travel. Special Report 162. U.S. Army Chemical Corps Biological Laboratories, Camp Detrick, Frederick, Md., 1 Aug. 1952.

9. Behavior of Aerosol Clouds Within Cities (Classified) (82 pp.). Minneapolis, Minn., 15 Jan.-24 Mar., St. Louis, Mo., 1953 (not available for this report).

10. Preliminary Field Trials of the Aerosol X1A. (Classified) (78 pp.) Abn dry agent dissemination unit. Rosemont, Minn., Sept.-Oct. 1953 (not available for this report).

11. San Francisco Bay, 21 and 26 Mar. 1956, Redwood City, Calif. (not available for this report).

2. Continental U.S., 30 Nov. 1957, East of Rocky Mountains, 6 Feb. 1958, 25 Apr. 1958, 20 Mar. 1958 (not available for this report).

3. T.B. Smith and M.A. Wolf. Intermediate Scale Particulate Cloud Travel—Project WINDSOC—Part 1—Technical Report. Meteorology Research, 2420 N. Lake Ave., Altadena, Calif. MRI Report 6023. Prepared for U.S. Army Chemical Corps Proving Ground under Contract DA-42-007-403-CMI-432. July 1960.

4. D.A. Haugen and J.J. Puquay, eds. The OCEAN BREEZE and DRY GULCH Diffusion Programs. Vol. 1. Research Report. Meteorology Laboratory, Air Force Cambridge Research Laboratories, Office of Aerospace Research, U.S. Air Force. AFCRL-63-719(I). Hanford Doc. HW-78435. Nov. 1963.

5. D.A. Haugen and J.H. Taylor, eds. The OCEAN BREEZE and DRY GULCH Diffusion Programs. Vol. 2. Research Report. Meteorology Laboratory, Air Force Cambridge Research Laboratories, Office of Aerospace Research, U.S. Air Force. AFCRL-63-719(II). Dec. 1963.

6. T.B. Smith and M.A. Wolf. Vertical Diffusion from an Elevated Line Source over a Variety of Terrains. Part A. Final Report to Dugway Proving Ground. Meteorology Research, 2420 North Lake Ave., Altadena, Calif. Contract DA-42-007-CML-545. Mar. 31, 1963.

7. F. Pooler. A Tracer Study of Dispersion over a City. Paper 66-28 at the 59th Annual Meeting, Air Pollution Control Association, San Francisco, Calif., June 20-24, 1966 (from the Air Resources Field Research Office, ESSA, at the Robert A. Taft Sanitary Engineering Center, Division of Air Pollution, U.S. Department of Health, Education, and Welfare, Cincinnati, Ohio).

8. Calibration of the Model D-1 Dry Particulate Disseminator with Green Fluorescing FP, DPGTP 508C. USATECOM Project 5-39030-10. DPGTM 1060. Case File 3164. Biological Branch, Test Design and Analysis Division, Technical Plans and Evaluation Directorate, U.S. Army Test and Evaluation Command, Dugway Proving Ground, Utah. Nov. 1963.

9. Deciduous Forest Diffusion Study, Report 3004, Final Report. Applied Science Division, Litton Systems, Minneapolis, Minn. Contract DA-42-007-AMC-48(R). Project 1T062111A128, for the

Meteorology Division, Deseret Test Center, Building 103 Soldiers Circle, Fort Dugway, Utah. June 1969.

Vol. 1: M.H. Tourin and W.C. Shen. Diffusion studies in a deciduous forest.

Vol. 2: M.H. Tourin, R.J. Rickett, and W.C. Shen. Detailed program description.

Vol. 3: M.H. Tourin and W.C. Shen. Detailed technical data supplement.

0. J.A. Murray, T.S. Brown, and F.X. Webster. FP Tracer Co-Dispersal and Sampling Studies. Technical Report 135. Metronics Associates, 3201 Porter Drive, Stanford Industrial Park, Palo Alto, Calif. Contract DA-42-007-AMC-240(R). Task IV025001 A128, Meteorological Aspects of CB Program. U.S. Army Dugway Proving Ground, Utah. Dec. 1968.

1. Diffusion under a Jungle Canopy. Final Report. Vol. 3: Mathematical Model. Meteorological Research Laboratory, Melpar, 7700 Arlington Blvd., Falls Church, Va. Contract DA 42-007-AMC33(R) for U.S. Army Dugway Proving Ground, Dugway, Utah. Oct. 1967.

2. G.R. Hilst and N.E. Bowne. A Study of the Diffusion of Aerosols Released from Aerial Line Sources Upwind of an Urban Complex. Vol. 1, Final Report; Vol. 2, Data Supplement. Travellers Research Center, 250 Constitution Plaza, Hartford, Conn. Contract DA-42007-AMC-37(R) under RDT&E Project 1V025001A128 Meteorological Aspects of CB Program, for U.S. Army Dugway Proving Ground, Salt Lake City, Utah. July 1966.

3. R.L. Miller, ed. Victoria Diffusion Trials. MR166 FR-374. Meteorology Research, 464 West Woodbury Road, Altadena, Calif. Contract DA 18-064-AMC-422(A) for Fort Detrick, Frederick, Md. May 1966.

   Vol. 1: Final Report (not available for this report)

   Vol. 2: Parts A and B (not available for this report)

4. T.B. Smith, and B.L. Niemann. Shoreline Diffusion Program, Oceanside, Calif. Report MRI 169 FR-860. Meteorology Research, 464 W. Woodbury Road, Altadena, Calif. Contract DA 42-007-

AMC-180(R) for Deseret Test Center, Fort Douglas, Utah. Nov. 1969.

5. G.R. Hilst and G.T. Csanady. Aerosol Diffusion over Woodlot Complexes. Theoretical and Empirical Assessments of Atmospheric Dispersion and Deposition Processes over Inhomogeneous Terrain (Woodlots). Phase I Report. DAAD09-67-C-0100(R). Travellers Research Center, 250 Constitution Plaza, Hartford, Conn. RDT&E Project IV025001A128 Meteorological Aspects of CB Program for U.S. Army Dugway Proving Ground, Dugway, Utah. Aug. 1968.

6. G.P Ettenheim, Jr. and C.L. Crum. MATAGORDA Deposition Trials. Field Report. MR167 FR-468. Meteorology Research, 464 West Woodbury Road, Altadena, Calif. Contract DA 18-064-AMC422(A) for Fort Detrick, Frederick, Md. 31 Jan. 1967.

7. Dispersion of Air Tracers into and Within a Forested Area. College of Forest Resources, University of Washington, Seattle, for U.S. Army Electronics Command, Atmospheric Sciences Laboratory, Fort Huachuca, Ariz. Grant DA-AMC-28-043-68-G8. Technical Report ECOM-68-G8-1. Sept. 1969.

   Vol. 1: L.J. Fritschen, C.H. Driver, C. Avery, J. Buffo, R. Edmonds, and R. Kinerson (objectives, methods, site description, preliminary data).

   Vol. 2: L.J. Fritschen, C.H. Driver, C. Avery, J. Buffo, R. Edmonds, R. Kinerson, and P Schless (tabulated data and dispersion patterns).

   Vol. 3: L.J. Fritschen, C.H. Driver, C. Avery, J. Buffo, R. Edmonds, R. Kinerson, and P Schless (analysis and interpretation).

8. J.K. Allison, A.V. Duffield, and J.M. Morton. Meteorological Analog Test and Evaluation: First Field Test Operation. Final Report. Meteorological Research Laboratory, Melpar Division of American Standard, 7700 Arlington Blvd., Falls Church, Va. Contract DA 42007-AMC-339(Y), Task VIII B, for U.S. Army Deseret Test Center, Fort Douglas, Utah. June 1970.

9. Comparison of Decay Rates for Bacillus globigii and Zinc Cadmium Sulfide. BW 1A-56, Operation "GOOF." DPGR 175. Dugway Prov-

ing Ground Report, BW Assessment Directorate, Project Order 0016. 27 Apr. 1956; Trial Report, BWAL, BW 1A-1, 2, and 3-56. Comparison of Decay Rates for BG and FP. Operation "GOOF." BWALTR 38. 7 Nov. 1955; Trial Report, BWAL, BW 1A-4 and 556. Operation "GOOF." CMLRE-DU-MBW. BWALTR 40. 21 Dec. 1955.

0. An Experimental Investigation of Viable Aerosol Travel from Sea to Land. Special Report 193. U.S. Army Chemical Corps Biological Laboratories, Camp Detrick, Frederick, Md. 24 Sept. 1953.

1. Trial Record 230, BW 398-A, Trials 8, 9, 10, and 11. Test Design and Analysis Office, Technical Operations Directorate, U.S. Army Chemical Corps Research and Development Command, U.S. Army Chemical Corps Proving Ground, Dugway, Utah. May, 1958.

2. A.T. Hereim and J.E. Frese. Check Test of West Vertical Grid, Dugway Proving Ground, Utah. Summary Report. RDTE Project 1-X-665704-D-634-06. USATECOM Project 5-CO-413-000-013. DTC Project DTC B-008. Deseret Test Center, Fort Douglas, Utah. Oct. 1970.

3. P.B. MacCready, T.B. Smith, and M.A. Wolf. Vertical diffusion from a low altitude line source--Dallas Tower Studies, Vol. 1. Final Report MR161 FR-33 from Meteorology Research, 2420 North Lake Ave., Altadena, Calif., for the U.S. Army Chemical Corps, Dugway Proving Ground. Contract DA-42-007-CML-504. Dec. 1961. (Note: only two title pages and pp. I-iii and 1-3 were available for this report.)

4. Operation Moby Dick. Sept. 1950. San Francisco Bay, Calif. (not available for this report).

5. Behavior of Aerosol Clouds within Cities. Joint quarterly reports submitted by Stanford University and the Ralph M. Parsons Company to the U.S. Army Chemical Corps. Contract Nos. DA-18-064CML-1856 (Stanford) and DA-18-064-CML-2282 (Parsons).

   A. JQR No. 1, July-Sept. 1952

   B. JQR No. 2, Oct.-Dec. 1952

   C. JQR No. 3, Jan.-Mar. 1953

   D. JQR No. 4, Apr.-June 1953 (missing some pages)

   E. JQR No. 5, July-Sept. 1953 (missing some pages)

F. JQR No. 6, Oct.-Dec. 1953, Vols. 1 and 2 (Vol. 2 is missing some pages)

6. S.W. Grinnell, W.A. Perkins, F.X. Webster. Bimonthly Report 11 submitted by Stanford University to the U.S. Army Chemical Warfare Service Research and Development Program. Contract W-18035-CWS-1256, Sept.-Oct. 1947. (BMR 1 through 13 and the final report are all available. Only BMR 11 describes any field release. The others imply small (milligram quantity) releases into an air chamber, and ''field work of a minor nature.'')

7. Quarterly reports submitted by Stanford University to the U.S. Army Chemical Corps Research and Development Program. Contract DA18-108-CML-450.

   A. QR 1, Feb.-Mar. 1950 (tests at Palo Alto)

   B. QR 2, May-June 1950 (tests at Dugway Proving Ground)

   C. QR 3 Aug.-Sept.-Oct. 1950 (San Francisco and adjacent area)

   D. QR 4, Missing

   E. QR 5, Feb.-Mar.-Apr. 1951 (measurement of urban temperature gradients)

   F. QR 6, May-June-July 1951 (meteorologic studies)

   G. QR 7, Aug.-Sept.-Oct. 1951 (meteorologic studies; FP size measurement)

8. F.X. Webster. The FP Atmospheric Tracer Technique. Metronics Associates, Stanford Industrial Park, Palo Alto, Calif. (Abstract of paper presented at Conference on Air Pollution in California, held Oct. 30, 1967 at San Jose State College, Calif.) (see also Ref. 42).

9. H.E. Cramer, F.A. Record, and H.C. Vaughan. The Study of the Diffusion of Gases or Aerosols in the Lower Atmosphere. AFCRC-TR58-239. ASTIA Doc. 152582. Massachusetts Institute of Technology, Department of Meteorology. Final Report. Contract AF 19(604)-1058. Sponsored by the Geophysics Research Directorate of the Air Force Cambridge Center, Air Research and Development Command. 15 May 1958.

10. L.M. Vaughan and W.A. Perkins. The Washout of Aerosol Particles and Gases by Rain. Technical Report 88, Aerosol Laboratory, Stan-

ford University, Stanford, Calif. Contract DA-42-007-403-CML448. U.S. Army Chemical Corps Research and Development Program. Jan. 1961.

1. F.X. Webster. Collection Efficiency of the Rotorod FP Sampler. Technical Report TR 98. Aerosol Laboratory, Metronics Associates, 3201 Porter Drive, Palo Alto, Calif. Contract DA 42-007-CML-543 for the U.S. Army Chemical Corps Research and Development Program. 31 Jan. 1963.

2. P.A. Leighton, W.A. Perkins, S.W. Grinnell, and F.X. Webster. 1965. The fluorescent particle atmospheric tracer. J. Appl. Meteorol. 3:334-348.

3. F. Pooler, Jr. 1966. A tracer study of dispersion over a city. J. Air Pollut. Control Assoc. 16:627-631.

4. M.H. Tourin and W.C. Shen. Diffusion Studies in a Deciduous Forest. Vol. 1, Final Report. Report 3004, AD 856703L. June 1969. 325 pp. (part of Ref. 19).

5. Y.C. Athanassiadis. Preliminary Air Pollution Survey of Cadmium and Its Compounds. Contract PH 22-68-25. U.S. Department of Health, Education, and Welfare. Oct. 1969. 325 pp. (not available for this report).

6. FP Tracer Counting Manual. Technical Manual 163-2. Aerosol Laboratory, Metronics Associates, Stanford Industrial Park, Palo Alto, Calif. Prepared for U.S. Weather Bureau, Dept. of Commerce, Contract CWB-10635. 17 May 1963 (rev., 11 Aug. 1964).

7. QR No. 448-4. July-Sept. 1960. Aerosol Laboratory, Stanford University, Stanford, Calif. Contract DA-48-007-403-CML-448. U.S. Army Chemical Corps Research and Development.

8. P.A. Leighton, director. The Stanford Fluorescent-particle Tracer Technique—An Operational Manual. Dept. of Chemistry, Stanford University. June 1955 (second printing, June 1958).

9. H.E. Cramer, R.N. Swanson, and A.G. Tingle. Review of the Meteorological Aspects of a High Altitude Release. GCA Corp., GCA Technology Division, Bedford, Mass., for U.S. Army Deseret Test Center, Fort Douglas, Utah. 9 Sept. 1970.

10. Duplicate of Ref. 48.

    Unnumbered. W.D. Crozier and B.K. Seely. 1955. Concentration distri-

butions in aerosol plumes three to twenty-two miles from a point source. Trans. Am. Geophys. Union 36:42-53.

DESCRIPTION OF TEST

Name of test: Comparison of Simulant Decay Rates in Field Tests

Reference-list number: 2

Reference: Special Report 273. Comparison of Simulant Decay Rates in Field Tests (U). Program Research, Field Operations and Meteorology Branches, Assessment Division, Fort Detrick, Frederick, Md. Nov. 1956 (supersedes Assessment Division Test Report A-294).

Principal object:

To obtain a definitive estimate of the downwind physical decay of aerosols of BG and FP and their relative decay rates; to compare estimates of these relationships obtained from three kinds of dissemination; to relate any differences between estimates to weather factors; to compare these estimates with those obtained elsewhere.

Site selection:

"Of the many sites considered for this test, Camp Cooke, California, offered the best combination of characteristics for efficient field testing." The principal factor was climate, but other factors included terrain, military control of the land, and operational factors. The prevailing wind from the northwest during summer ensured that trials could be run on consecutive days shortly after sunset.

Number and nomenclature of releases:

Three phases (A, B, and C) with different release mechanisms (for FP or BG) in each phase. Thirty-nine trials were conducted, numbered 1 through 39, with 1-20, 26-27 in Phase A, 21-25 in Phase B, and 28-39 in Phase C. Difficulties with the sampling of BG led to a complete dis-

carding of data from trials 1-4 and 9-12, with trials 5-8 considered somewhat unreliable for BG.

Test conditions:

Phase A: Two E61R4 (bomblets) filled with 40 g each of FP, and located 40 ft apart, were fired simultaneously.

Phase B: A single Stanford generator (80 g hopper) disseminated at approximately 8 g/min for 4 min at ground level. Phase C: As Phase B, but at 13 ft above ground.

Samplers were placed on a 90-degree circular sector grid at 100, 200, 400, 600, 1,200, 1,760, and 2,640 yd, with 20, 40, 60, 80, 80, 80, 120 stations at each radius, respectively. The samplers were Millipore type with a flow rate of 12.5 L/min.

Test material:

ZnCdS from New Jersey Zinc Company, CEP 8000, Lot 0002 with 2% magnesium silicate. No estimates of particles per gram or dissemination efficiencies are given in the report, but this lot was tested as having 4.51 x 10¹⁰ ppg in Ref. 47 (QR SAL 448-4; see Section 4, "Miscellaneous Notes."

Place of release:

Camp Cooke, California, in scrub-brush desert, 2.5 miles downwind from the ocean.

Dates, times, and quantities of release:

No information is given on the dates or times of releases. All that can be inferred is that they were some time in 1955. No more accurate information is given on release quantities than presented in "Test conditions:" above. With these values, each trial in Phase A (22 trials) released 80 g of FPs, and each trial in Phases B and C (17 trials) released 32 g—a total release of 2304 g.

Communities affected:

Unknown. The location of Camp Cooke was not provided.

Distances from releases to affected communities:

Greater than the maximum radius of the sampling grid, 1.5 miles.

Weather conditions:

A neutral temperature gradient prevailed consistently during the hours after sunset.

Other materials released:

There was simultaneous release of BG.

Maximum time-integrated concentrations and concentrations:

The report was concerned with estimating decay curves for FP and BG. No measured counts, concentrations or exposures were presented. Cross-wind-integrated-exposures were tabulated at each sampling radius, and cloud widths (distances between 1/10 maximum exposures) were given. These two measurements have been used to estimate the following maximum exposures and concentrations, by assuming gaussian shaped plumes and (for the instantaneous releases) equal dispersion along and across the plume. Dissemination efficiency has been assumed to be 50% for the bombs, and 100% for the Stanford generators.

The highest measured exposures and estimated concentrations occur at the 100 yd sampling line, and these are the values given for "Unpopulated areas." The upper bound for the "Populated area" was obtained from the estimates for the sampling line at greatest distance from which estimates were available, generally 1.5 miles from the dissemination point for Phase A, and 1 mile from the sampling point for Phases B and C.

The cumulative estimates have been estimated by summing the individual exposures—the wind direction was always similar (from the north-

west)—and adjusting the total for Phase A by assuming that the average of the missing trial data was similar to the average of the available trial data. As usual, the "cumulative" maximum concentrations are simply the largest estimated concentrations—the trials were assumed to occur at distinct times. The large difference in concentrations between Phase A and Phases B and C is due to the different dissemination mechanisms—an instantaneous release versus a continuous release.

Other comments:

This report supersedes Assessment Division Test Report A-294. That earlier report might contain the missing raw data.

There are general references to FP use:

• "Dispersion of Aerosols and Travel of Aerosol Clouds." Contract DA-18-108-CML-450, Chemistry Department, Stanford University, 1953.

• "Research Concerning the Propagation of Airborne Agents for Work over Cities." Contract DA-18-064-CML-1856, Chemistry Department, Stanford University, 1953.

These describe experiments over Palo Alto and San Francisco, and so probably duplicate Ref. 37.

There are references to other operations involving FP and BG:

• Summer of 1950, FP, BG, and AP1 generated simultaneously at Dugway.

• 1951, at Dugway, FP with AB 1 and bacterium Tularense (more than four trials).

• 1954, at Dugway, "A Number of Trials Producing Aerosols of BG and FP."

• Operation MOBY DICK, "conducted in California in 1954." Special Report 237, "Moby Dick: Sea-to-Land Travel of Simulant Aerosols Generated by the XB-14B Mine (C)." Assessment Division, Camp Detrick, Frederick, Md. Aug. 1955. This is listed as Ref. 34, except that it is said to have occurred in Sept. 1950.

• The operation off Georgia, North Carolina, and South Carolina is

referred to as Operation DEW I. The reference given is the document in Ref. 8 (Dugway Special Report 162).

• An operation named Operation DEW II, involving release of FP and Lycopodium spores from an aircraft. This is Dugway Special Report 179 ("An Experimental Study of Long Range Aerosol Cloud Travel Involving Ground Deposition of Biological Spore Material," F&MR Division, Camp Detrick, Frederick, Md., 1 June 1953) and appears to be the second (classified) document in Ref. 8 (not available).

• The Minneapolis, St. Louis, and Winnipeg trials are referred to as "Research Studies of Aerosol Cloud Travel over Cities." Contract DA-18-064-CML-2282, the Ralph M. Parsons Company. This appears to be the study examined in Ref. 35.

• Other referenced trials involving FP are Special Report 142. BW trials at San Francisco, Calif., Sept. 1950, PD Division, Camp Detrick, Frederick, Md., 22 Jan. 1951. "Studies in Aerosol Cloud Behavior." Contract DA-42-007-403CML-111, Chemistry Department, Stanford University, 1953.

There is a reference (p. 3) to the "St. Jo Munitions Expenditure Panel" that might indicate a named operation.

DESCRIPTION OF TEST

Name of test: Comparative Diffusion of Dissimilar Agents

Reference-list number: 3

Reference: Comparative Diffusion of Dissimilar Agents. PGR 149 W5, 6-52. Downwind travel of simulant agents released simultaneously, BW 5 and 6-52. Dugway Proving Ground Report 149. Projects 4-98-05-005 and 4-98-01-001. 5 Feb. 1954.

Principal object:

To compare the distribution and downwind-travel characteristics of SM

(Serratia marcescens), C2 (Aspergillus fumigatus), and FP (2266) released simultaneously; to compare the viability and decay rates of the SM and C2 aerosols; to compare the distribution and downwind-travel characteristics of aerosols of SM, BG, and FP released simultaneously; and to compare the viability and decay rates of the SM and BG aerosols.

Site selection:

Dugway Proving Ground.

Number and nomenclature of releases:

Two releases: BW 5-52 and BW 6-52.

Test conditions:

A portion of the Crops Grid No. 2 was used. The grid sector used lies along an axis of 135° and is 13 miles long by 3.5 miles wide, with grid laterals placed at 0.5-mile intervals. It is bounded to the southwest by Simpson Buttes and Camel Back Mountain, but is flat in the sampling area. Four generators were used to simultaneously disseminate the FP in 90 to 105 s. Filter-type samplers were used.

Test material:

ZnCdS FP type 2266, with 1 x 10¹¹ ppg.

Place of release:

Both tests: approximately 6 miles east-northeast of Simpson Buttes.

Dates, times, and quantities of release:

Communities affected:

No attempt has been made to estimate concentrations in populated areas.

Distances from releases to affected communities:

No attempt has been made to estimate concentrations in populated areas.

Weather conditions:

During BW 6-52, the wind was approximately south-southwest at 4 to 5 mph, carrying the FP over the sampling grid. During BW 5-52, the wind shifted and apparently carried the FP cloud west and then southwest out of the sampling grid.

Other materials released:

In BW 6-52, BG and SM were simultaneously released with the FP. In BW 5-52, AF and SM were simultaneously released with the FP.

Maximum time-integrated concentrations:

BW 6-52:61 µg-min/m³

BW 5-52:16 µg-min/m³

These were measured at 1 mile from the source. The wind shift during BW 5-52 probably resulted in a low measurement. In BW 6-52, the exposure was as high as 20 µg-min/m³ at 13 miles from the source.

Maximum time-integrated concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Maximum concentrations:

No time-resolved data were available to allow estimation of concentration.

Maximum concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Other comments:

The north direction arrows on figures 6 and 7 are reversed with respect to the other figures, and with respect to reality. For various reasons, not discussed here, both these tests were complete failures from the point of view of their objectives.

DESCRIPTION OF TEST

Name of test: Relationship of Dosages to Source Strength for BG and an Inert Tracer

Reference-list number: 4

Reference: Trial Reports. Relationship of Dosages to Source Strength for BG and an Inert Tracer:

Principal object:

Not stated, except by implication in the title. The test plan (BW 8-54) was not available.

Site selection:

Dugway Horizontal grid.

Number and nomenclature of releases:

One release per report (four releases). The reports were numbered in the same way as the trials (BW 8A-1-54 through 8A-5-54, omitting 8A-354).

Test conditions:

Tests were according to a test plan that is not available (see ''Principal object'' above). A Stanford Blower was used, presumably for FP dissemination. Filter-type samplers were used.

Test material:

FP (nature unspecified). Particle counts were given as 1.96 x 10¹⁰ ppg for the last three experiments, and not specified for the first.

Place of release:

Dugway horizontal grid.

Dates, times, and quantities of release:

The dissemination time is not specified, but was probably short, since the experiments were apparently attempting to compare dispersion of FP with BG from bombs.

Communities affected:

No attempt has been made to estimate concentrations in populated areas.

Distances from releases to affected communities:

No attempt has been made to estimate concentrations in populated areas.

Weather conditions:

Maximum time-integrated concentrations:

No calibration data were provided, only the raw counts on Millipore filters. No information on the samplers was available. The following estimates of total exposure assume a sampling rate of 6.7 L/min, identical to that in the releases in May and June of 1953. The particles-per-gram count for the first experiment has been assumed to be identical to that for the other three experiments.

Maximum concentrations:

No time-resolved data, or even release-timing data, were available to allow estimation of concentration.

Maximum time-integrated concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Maximum concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Other comments:

All data are noted as preliminary and subject to revision. Each is a 9-10-page report of a release, including raw counts for organisms and FP.

The nomenclature suggests that a report for trial 8A-3-54 is missing.

DESCRIPTION OF TEST

Name of test: Operation "Trouble Maker"

Reference-list number: 6

Reference: Trial Report. BWAL, BW 8B-7 and 8-54, Operation "Trouble Maker." Relationship of dosages to source strength for BG and an inert tracer. BWALTR 27, 18 July 1955.

Principal object:

Not specified. The test plan (test plan BW 8B-54, 29 Nov. 1954, 6 pp., amended 12 Jan. 1955, 4 pp.) was not available.

Site selection:

Dugway west vertical grid.

Number and nomenclature of releases:

Two releases: BW 8B-7-54 and BW 8B-8-54.

Test conditions:

Tests were conducted according to a test plan that is not available (see "Principal object," above). A total of 318 Millipore filters were used for FP collection. A Skil blower was used, presumably for FP dissemination.

Test material:

FP (not otherwise specified). Particle count estimated as 2 x 10¹⁰ ppg.

Place of release:

Dugway west vertical grid.

Dates, times, and quantities of release:

No information on dissemination times is provided. BG was dispersed (presumably simultaneously) from bomblets; therefore, the dissemination time is likely to have been short.

Communities affected:

No attempt has been made to estimate concentrations in populated areas.

Distances from releases to affected communities:

No attempt has been made to estimate concentrations in populated areas.

Weather conditions:

Wind was approximately from the north, with a speed of about 16 mph

(at 2 m) during BW 8B-7-54. It was northerly with a speed of about 11 mph (at 2 m) during BW 8B-8-54.

Other materials released:

BG was released from bomblets.

Maximum time-integrated concentrations:

No calibration data were provided, only the raw counts on Millipore filters. No information on the samplers was available. The following estimates of total exposure assume a sampling rate of 6.7 L/min, identical to that in the releases in May and June of 1953.

BW 8B-7-54: Maximum exposure, 180 µg-min/m³

BW 8B-8-54: Maximum exposure, 420 µg-min/m³

Maximum concentrations:

No time-resolved data, or even release-timing data, were available to allow estimation of concentration.

Maximum time-integrated concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Maximum concentrations in any populated area:

No attempt has been made to estimate concentrations in populated areas.

Other comments:

This report was a 20-page summary, subject to revision, giving raw counts and summary meteorologic data only.

DESCRIPTION OF TEST

Name of test: Operation DEW I

Reference-list number: 8

Reference: An Experimental Study of Long Range Aerosol Cloud Travel. Special Report 162. U.S. Army Chemical Corps Biological Laboratories, Camp Detrick, Frederick, Md. 1 Aug. 1952.

Principal object:

To test the possibility of achieving long-range aerosol-cloud travel and consequent coverage of large areas (several thousand square miles) at ground level beneath a slow-moving frontal system.

Site selection:

The southeastern states of North Carolina, South Carolina, and Georgia were selected as a test area after "consideration of meteorologic situations, terrain and modes of dissemination."

Number and nomenclature of releases:

Five releases, named Trial 1 through Trial 5.

Test conditions:

Releases were from a naval surface craft (USS TERCEL, a converted mine sweeper) operating at 15 knots parallel to the coast and about 5 to 10 miles offshore on a line segment approximately 100-150 nautical miles long off the North Carolina, South Carolina, or Georgia coasts on the total 390 nautical miles track between Jacksonville, Fla. and Hatteras, N.C. Stanford-type aerosol generators were used. A total of 46 sampling stations were established in the coastal plain area of Georgia, South Carolina, and North Carolina; sampling was set up for 15 consecutive 2-h periods during each trial.

Trial 1 occurred under conditions that were not correctly anticipated, so that the plume traveled principally parallel to the coast and out to sea. Trials 2 through 4 achieved the objectives, with the plumes traveling over large areas of Georgia, South Carolina, and North Carolina.

Test material:

ZnCdS (New Jersey Zinc Company, No. 2266), particle size 2.25 µm mass median diameter. Mean particle count approximately 4.2 x 10¹⁰ ppg (see "Other comments" below). Examination of this material also showed about a 1% contamination with soluble particles that fluoresced very intensely green, but were not otherwise identified.

Place of release:

For all trials, along a track about 5-10 miles offshore, between Jacksonville, Fla. and Hatteras, N.C.

Trial 1: Along 138 statute miles, starting approximately 35 statute miles from Jacksonville, northward direction of travel.

Trial 2: Along 121 statute miles, starting approximately 121 statute miles from Jacksonville, northward direction of travel.

Trial 3: Along 121 statute miles, starting approximately 173 statute miles from Jacksonville, southward direction of travel.

Trial 4: Along 121 statute miles, starting approximately 331 statute miles from Jacksonville, southward direction of travel.

Trial 5: Along 156 statute miles, starting approximately 173 statute miles from Jacksonville, southward direction of travel.

Dates, start times, and quantities of release:

Trial 1:3/26/52 0600-1400 EST 240 lb (109 kg, 4.6 x 10¹⁵ particles)

Trial 2:3/30-31/52 1900-0200 EST 250 lb (113 kg, 4.8 x 10¹⁵ particles)

Trial 3:4/4/52 1100-1800 EST 250 lb (113 kg, 4.8 x 10¹⁵ particles)

Trial 4:4/9/52 1100-1800 EST 200 lb (91 kg, 3.8 x 10¹⁵ particles)

Trial 5:4/21/52 1400-2300 EST 450 lb (204 kg, 8.6 x 10¹⁵ particles)

Total release to the environment: 1,390 lb (630 kg).

Communities affected:

The following describes areas affected at a cumulative dosage greater than 1 particle-min/L (approximately 2.7 x 10^-2 µg-min/m³). The areas are so large that only a general description can be given.

Trial 1: Approximately 1,000 square miles. Almost everywhere within approximately 5-10 miles of the coast of Georgia, and within a few miles of the southernmost 30 miles of the coast of South Carolina.

Trial 2: Approximately 21,800 square miles, covering approximately the southern half of South Carolina and an abutting swathe of east Georgia approximately 25 miles wide at the coast, widening to 75 miles at 150 miles inland.

Trial 3: Approximately 13,100 square miles, roughly encompassed by a north-south stripe across the central one-third of South Carolina.

Trial 4: Approximately 28,900 square miles covering the southern two-thirds of South Carolina, together with an abutting 60-mile-radius semicircle extending into Georgia, with its base on the state line and extending inland 120 miles from the coast.

Trial 5: Approximately 34,800 square miles, a coastal strip 75 miles wide from the southern most tip of the Georgia coast to Cape Hatteras, N.C.

Distances from releases to affected communities:

The releases were 5-10 miles offshore and 5-10 miles from the nearest affected communities. Measurable plume concentrations extended up to 175 miles inland.

Weather conditions:

The releases were designed to be released with onshore winds under frontal inversions and, taking advantage of nocturnal inversions, to ensure long-distance travel with minimum vertical dispersion. During Trial 1, the winds remained parallel to the shoreline instead of turning onshore as predicted, and so the plume missed most land. Trial 2 had

onshore winds and nocturnal inversions throughout the area, but no frontal inversion. Low clouds formed during the early morning, followed by showers, light rain, and drizzle. The plume (disseminated from 1900 through 0200 h) was observed to increase in area until morning when the nocturnal inversions dissipated. Trial 3 achieved substantial dissemination in the cold air under a warm front that was moving northward, accompanied by unstable air with small tornadoes, followed by rain. Measurements indicated substantial trapping under the frontal system. In Trial 4, dissemination (1100-1800 h) was into a neutral atmosphere with a weak warm front that dissipated rapidly. Coastal clouds dissipated later in the day, with clear inland weather and onshore winds. A wind shift across the weak front apparently resulted in two effective plumes from different parts of the dissemination line. Trial 5 was conducted when no frontal conditions were present, but a radiation inversion formed and was present at night (dissemination 1400-2300 h). The early part of the release moved rapidly, but the later (nocturnal) part of the release was into relatively light winds in neutral conditions. The nocturnal plume spread slowly until morning, when it dissipated rapidly as the inversion was destroyed by daytime heating.

Other materials released:

There is no mention of other releases in the unclassified document.

Maximum time-integrated concentrations:

These were measured over 30 h from the beginning of dissemination, ensuring complete plume capture (except for the very minor re-entrainment from deposited material). Measurement accuracy was of the order of ±40%, so the 2-digit precision given here is for calculation purposes only. The values shown might be underestimated by a factor of 7 (see "Other comments" below).

Combined: 98 µg-min/m³ (The maximum time-integrated concentrations for all trials at each individual measurement point.)

Maximum time-integrated concentrations in any populated area:

Same as the maximum time-integrated concentrations anywhere, because the dispersion area was so large. The highest cumulative concentration was measured at Hunter Air Force Base, Ga.

Maximum concentrations:

These concentrations are 2-h averages, as measured in the tests. Once again, accuracy is approximately ±40%, so the two digit precision given here is for calculation purposes only. The values shown might be underestimated by a factor of 7 (see "Other comments" below).

Maximum concentrations in any populated area:

Same as the maximum concentrations anywhere, since the dispersion area was so large. Maximum concentration was measured at Hunter Air Force Base, Ga.

Other comments:

Estimates of deposition indicate that only 5 % to 6% of the material was deposited within the measurable plume area (out to 150 miles).

The calculation of number of particles released was not discussed. It appears that the cited particle "mass median" diameter of 2.25 µm was used, together with a particle density of 4 g/cm³, and an assumption of spherical particles. This gives the approximate counts cited above, but

it ignores the efficiency of dissemination and the difference between mean and median diameters. (Note that ''mass median'' is a misnomer—this is also the number, surface area, and every other median diameter as well.) This could have led to an underestimate of a factor of approximately 7 in the concentration and exposure estimates given above (assuming number counts and dissemination efficiencies close to Ref. 19).

DESCRIPTION OF TEST

Name of test: Project WINDSOC

Reference-list number: 13

Reference: T.B. Smith and M.A. Wolf. Intermediate Scale Particulate Cloud Travel—Project WINDSOC—Part 1—Technical Report. Meteorology Research, 2420 N. Lake Ave., Altadena, Ca. MRI Report 60-23. Prepared for U.S. Army Chemical Corps Proving Ground under Contract DA-42-007-403-CMI-432. July 1960. (Only the first 18 pages of this over 73-page report were available.)

Principal object:

Informally:

"To study particulate cloud travel under night release conditions on a distance scale intermediate between the LAC operation and the small scale diffusion studies usually performed."

Formally:

Principal objective: "To describe and determine empirically the relationships between meteorologic forces acting on elevated line sources of particulate aerosols and the resulting quantitative distribution of those particulates at ground level."

Secondary objective: "To provide data for the extension of existing dosage predictive techniques, or the development of new techniques, appropriate for the aerial line-source dissemination of particulates."

Number and nomenclature of releases:

A total of 13 tests using FP tracer. They are referred to as Test 1 through Test 13.

Test conditions:

Release was into stable air over relatively flat terrain in central Texas. The test area (125 x 125 miles south and west of Dallas, including Dallas, Fort Worth, Waco, and Fort Hood) was selected for such weather and terrain characteristics. A total of 84 ground samplers were used in the test area, together with balloon samples at Fort Hood for some tests, five samplers (at 300, 600, 900, 1,200, and 1,400 ft above ground) on a TV tower 20 miles southwest of Dallas for Tests 9 through 13, groups of Rotorod samplers at various points on the ground for various tests, and samplers aboard four L-20 aircraft. FP tracer was released from a C-119 aircraft flying crosswind along the upwind side of the test area, at a height of 600 to 1,300 ft. above ground level, along 200 miles (i.e., extending well beyond the edges of the test area. Tests 5, 9 and 11 were not fully analyzed, since meteorologic conditions did not meet requirements.

Test material:

ZnCdS from U.S. Radium Corp. (Contract DA-18-064-404-CML-427). Average particle count per gram of five lots was 2.09 x 10¹⁰; differences between lots were not considered meaningful. The mode of the particle diameter was near 2 µm, with 10-25% outside the range 1-4 µm.

Place of release:

Along a 200-mile line over-running the 125-mile upwind edge of the test area.

Dates, times, and quantities of release:

The release rate varied during each test, with a maximum of 4.5 lb/mile (Test 1). From Test 3 onward, it was stabilized at approximately 3.2 lb/ mile. The implied total releases were thus:

Other information is unavailable in the section of report obtained.

Communities affected:

All those within the 125 x 125 mile test area, including Dallas, Fort Worth, Fort Hood, and Waco. In addition, communities outside the test area but within the plume from the 200-mile release.

Other information is available in the section of the report obtained.

Distances from releases to affected communities:

Unavailable in the section of the report obtained.

Weather conditions:

Unavailable in the section of the report obtained.

Other materials released:

Unavailable in the section of the report obtained.

Maximum time-integrated concentrations:

Unavailable in the section of the report obtained.

Maximum time-integrated concentrations in any populated area:

Unavailable in the section of the report obtained.

Maximum concentrations:

Unavailable in the section of the report obtained.

Maximum concentrations in any populated area:

Unavailable in the section of the report obtained.

Other comments:

The full report contains all the information unavailable in the 18 pages that are available. Mention is made in this report of a series of 11 trials at Dugway, 21-23 Sept. 1959, to test the dissemination efficiency of the plane. Measured particle number efficiency was 13% to 72%.

DESCRIPTION OF TEST

Name of test: Vertical Diffusion from an Elevated Line Source (no code name identified)

Reference-list number: 16

Reference: T.B. Smith and M.A. Wolf. Vertical Diffusion from an Elevated Line Source over a Variety of Terrains. Part A. Final Report to Dugway Proving Ground. Meteorology Research, 2420 North Lake Ave., Altadena, Calif. Contract DA-42-007-CML-545. 31 Mar. 1963

Principal object:

Determination of the applicability of the dosage prediction technique of the Dallas Tower study under differing conditions of other areas.

Site selection:

All four selected sites were required to be in areas where adverse conditions (rain, fog, low clouds, extreme wind speeds, variable wind direction) were minimized, and where there were existing roads parallel to the persistent wind direction. Two sites, in Oklahoma and Washington, were selected on the basis of their homogeneous rolling terrain (providing greater surface roughness than the Dallas Tower study). The two other sites were selected for single features—the Texas site offers a flat coast

line normal to the wind flow; the Nevada site presents a cross-wind ridge on an otherwise smooth surface.

Number and nomenclature of releases:

A total of 36 releases, identified as Test 1 through Test 36, with three considered unsuccessful (Test 10, disseminator malfunction; Tests 20 and 34, wind shifts). There were nine releases at the first three sites (Oklahoma, Tests 1 through 9; Texas, Tests 11 through 19; Washington, Tests 20 through 28), and eight at the last site (Nevada, Tests 29 through 36).

Test conditions:

At each site, the releases were made from an Aero Commander airplane flying at 500 to 1300 ft. FP were metered for release at 1.5 lb/mile over a flight path of 30 miles, yielding a total mass released of 45 lb per release. Sampling was performed at 1-mile intervals along approximately 25 miles in the downwind direction, generally starting a short distance upwind of anticipated initial touchdown point of the plume.

Test material:

ZnCdS containing 1% by weight of micronized Valron Estersil for increased fluidity (Lot 14, produced Nov. 1959 by U.S. Radium Corp.). Median diameter approximately 2.4 µm. Mean particle count 2.16 x 10¹⁰ per gram, with a release efficiency assumed to be the same (39%) as in the Dallas tests, yielding a particle emission rate of 1.11 x 10⁹ per foot.

Place of release:

Dates, times, and quantities of release:

Communities affected:

The nearest communities (on current maps) that were potentially most affected were the following:

Oklahoma: Cushing, Ripley, Drumright, Oilton, Yale, Jennings, Hallett, Glencoe, Pawnee, and nearby towns, principally those to the north.

Texas: Port Aransas, Aransas Pass, Ingleside, Gregory, Portland, Corpus Christi, Rockport, Bayside, Taft and nearby towns, principally those to the northwest.

Washington: Colfax, Palouse, Garfield, Steptoe, Thornton, Oakesdale, Malden, Rosalia, Tekoa, Farmington, Latah, and nearby towns, principally those to the north and north-northeast.

Nevada: Goldfield, and other towns to the south.

Distances from releases to affected communities:

Oklahoma: Releases were approximately 4 or 5 miles upwind of Ripley, Cushing, and Drumright. Other towns were at least 5 miles further downwind.

Texas: Releases were approximately 2 to 5 miles from Port Aransas and at least 8 miles from other communities.

Washington: Releases were mostly slightly north (downwind) of Colfax, with apparently one release slightly (a mile or so) upwind of Colfax. Most releases were probably about 5 miles upwind of Palouse and more than 8 miles upwind of other communities.

Nevada: Releases were approximately 4 to 11 miles north (upwind) of Goldfield. Other towns were at least 30 miles downwind of the release.

Weather conditions:

In all cases, the tests were carried out in the desired clear weather, with specific meteorologic regimes that established relatively high turbulence in the lower layers of the atmosphere.

Oklahoma: Tests were conducted in typical summer weather, with southerly flow conditions (wind direction recorded as south two times, south-southeast four times, and southeast three times), when turbulence is gen-

erated in the lower 1,000-1,500 ft, with little turbulence above 1,000-1,500 ft. In Test 4, however, there was insufficient vertical wind shear to generate turbulence in the layers near the release height.

Texas: Tests were conducted with south to southeasterly flow conditions (wind direction recorded as southeast five times, south-southeast two times, and east-southeast two times). Solar heating and ocean conditions were such that the turbulence in the lower 1,000-1,500 ft was relatively low compared with the other test areas.

Washington: Tests were conducted in south to southwesterly wind flow conditions (wind direction recorded as south-southeast three times, south one time, south-southwest three times and southwest two times). The desired relatively high turbulence in the lower levels of the atmosphere is not typical of the area, and a 3-week period was required to complete the tests.

Nevada: The tests were conducted in the steady north to northwest flow regime that often develops in the late afternoon and night (recorded wind directions were northeast one time, north five times, and north-northwest two times). This wind flow developed despite easterly pressure-gradient winds (above 7,000 ft), and has a depth of about 2,000 ft, passing easily over the Goldfield ridge.

Other materials released:

No other materials are mentioned in the report.

Maximum time-integrated concentrations and maximum concentrations:

The report gives the measured cumulative exposure. The concentration estimates here were made by assuming a gaussian plume shape with along-wind dispersion of 1,000 m (corresponding approximately to C stability at 10 to 20 km downwind), and using the average measured wind speed at 10 m height. These estimates should thus be considered very rough estimates. Accuracy of measurements was not discussed. By com-

parison with other reports, the measurement accuracy cannot be better than about a factor of 2 (so the precision of the entries in the table should be retained only for checking calculations).

Maximum time-integrated concentrations and maximum concentrations in any populated area:

In Oklahoma, Texas, and Washington, the density of communities is high enough in the affected area that it is likely that populated areas experienced concentrations and time-integrated concentrations as high as were

measured anywhere. In Nevada, the only nearby community was the town of Goldfield, which experienced time-integrated concentrations (and probably maximum concentrations) between one-third and one-fourth of the maxima reported above for Nevada.

DESCRIPTION OF TEST

Name of test: Tracer Study of Dispersion over a City (no code name identified).

Reference-list numbers: 17 and 43

Reference: F. Pooler. A Tracer Study of Dispersion over a City. Paper 6628 at the 59th Annual Meeting, Air Pollution Control Association, San Francisco, Calif.; June 20-24, 1966 (from the Air Resources Field Research Office, ESSA, at the Robert A. Taft Sanitary Engineering Center, Division of Air Pollution, U.S. Department of Health, Education, and Welfare, Cincinnati, Ohio). Also published in 1966 in the Journal of the Air Pollution Control Association (Vol. 16, pp. 627-631). (The journal article appears to have been slightly edited.)

Principal object:

To obtain ''direct experimental evidence to indicate at least order-of-magnitude urban area dispersion parameters.''

Site selection:

"St. Louis was chosen as the experimental area for study primarily because a more general study of air pollution in the St. Louis metropolitan area was then being initiated. In addition," the city is reasonably flat and not affected by significant topographic features, the Weather Bureau Office operated a weather radar, and St. Louis was easily accessible from Cincinnati (experimenter's home office).

The areas principally affected by releases discussed in this reference are distinct from the areas affected by releases discussed in Ref. 35.

Number and nomenclature of releases:

Seven series of experiments, with a total of 42 releases. Each release is labeled by a series number (1 through 7) and a test number from 2 to 43. (Test 1 was a dry run with no tracer release.)

Test conditions:

Samplers used included 60 Rotorods, 30 membrane filters, and 10 pulsed drum samplers. A total of 316 sampling sites were used during some or all tests, placed in arcs around the sources at nominal distances of 0.5, 2, and 4.5 miles from Forest Park, and 1.25, 2.5, and 5 miles from the Knights of Columbus Building (see below). Two generators were available that were capable of release rates from 0.5 to 200 g/min. Forty of the 42 tests had a 1-h release time, so that the maximum release possible was 24 kg per test (2 x 0.2 kg/min x 60 min); the other two tests had a release period of 0.5 h, giving a maximum possible release of 12 kg.

Test material:

Not discussed, except by reference to the literature.

Place of release:

Releases were from a site in Forest Park (Tests 2-4, 9, 16-20, 22-25, 29, 31-41, 43) or from the roof of the Knights of Columbus Building. The available map is of low resolution, but places the release sites in the vicinity of the junction of Clayton Road and Faulkner Road in Forest Park and near the junction of S. Grand Blvd. and Gravois Ave. for the Knights of Columbus Building. Those two sites are approximately 2.8 miles apart.

Dates, times, and quantities of release:

Dates, day of week, times, and source location were the following: