Prediction of vulnerable zones for reactive substances.

The determination of the plume associated with the atmospheric dispersion of a substance from a point source has been well researched (2,3). Modeling plume generation to estimate the distribution of pollutant substances in the atmosphere has been used in a number of different areas, most notably the emissions from power plants and automobile exhausts. Pollutant dispersion models are now being used by local community emergency planning committees to predict vulnerable zones for sites in their communities as a result of an accidental release of an extremely hazardous substance. This application does not take into account long term atmospheric reactions like oxidation and photochemical reactions. Rather, it is concerned with one or two hours after a hazardous substance has been released.

Included in the Superfund Amendment and Reauthorization Act of 1986 (SARA) is Title 3, also known as the Emergency Right-to-Know Act. The Act requires the state-designated community planning committees to analyze sites in their communities for the potential effect of airborne release of extremely hazardous substances. When a release of an airborne extremely hazardous substance (EHS) occurs the concentration of the substance can approach a level which is potentially dangerous to the health of individuals in the area of the release. For emergency planning purposes, the concentration of the EHS which can cause serious irreversible health effects or death as a result of a single exposure is defined as the level of concern (LOC) by the Technical Guidance for Hazards Analysis (1). The LOC is usually determined from the immediately-dangerous-to-life-and-health (IDLH) levels developed by the National Institute for Occupational Safety and Health (NIOSH). In most cases the LOC is arrived at by taking one-tenth of the IDLH level.

A dispersion model is used to estimate the vulnerable zone for the airborne release of the toxic substance. The Gaussian dispersion model, or variations thereof, are the usual model employed to analyze the release of an EHS. The model calculates the distance from the source that the concentration of the substance is above the LOC, i.e., the vulnerable zone. If the EHS reacts with the moisture in the air, oxygen or other material present to form a more hazardous substance, the dispersion model will predict a vulnerable zone that is significantly less than is required. The vulnerable zone of the new reactant EHS release can be greater than the dispersion model will predict for the non-reactant EHS. The local planning committees' hazard analysis of a site must take into account this reactive effect.

Determination of the vulnerable zone

The vulnerable zone is determined by application of the Gaussian model of the plume spread in both the horizontal and vertical directions. The derivation of the Gaussian model for dispersion estimates originated with Hay and Pasquill, Cramer and Gifford (2). The model typically used for emergency planning applications assumes a ground level release over a rural flat terrain. The atmospheric stability classification is F for moderately stable conditions. Turner's equation (1), for the ground level concentration of a substance is given by:

C = QR/p ||Sigma~.sub.y~||Sigma~.sub.z~ u 1)

where C = airborne concentration, gm/|m.sup.3~

QR = rate of release, gm/sec

||Sigma~.sub.y~ = standard deviation of the plume concentration distribution in the y direction, m

||Sigma~.sub.z~ = standard deviation of the plume concentration distribution in the z direction, m

u = wind speed, m/sec

This equation assumes the substance is a stable gas or aerosol. The radius of the vulnerable zone is found by inserting the LOC for the substance under consideration as C in Turner's equation. The rate of release QR is estimated based on the amount of the EHS stored at the site. This could become the volume of one or more connected vessels or other container that could rupture and release the EHS. The wind speed is known or estimated; the lower the value the higher the airborne concentration. The values of ||Sigma~.sub.y~ and ||Sigma~.sub.z~ have been determined experimentally and refined empirically by McElroy and Pooler (2). The TABULAR DATA OMITTED graphs depicting these parameters are on pages eight and nine of the reference. The values of ||Sigma~.sub.y~ and ||Sigma~.sub.z~ depend on x, the downwind distance from the source. Entering the estimates of the other variables allows the calculation of this radius which defines the envelope where the concentration of the substance is above the LOC.

Determination of the vulnerable zone for fluorine 

Fluorine is a pale yellow gas at standard pressure and has a LOC of 0.039 gm/|m.sup.3~ (4). When gaseous fluorine is released into the atmosphere it will react with the moisture in the air and form hydrogen fluoride following either of these two equations (5).

2|F.sub.2~+2|H.sub.2~O|right arrow~4HF+|O.sub.2~ 2)

3|F.sub.2~+3|H.sub.2~O|right arrow~6HF+|O.sub.3~ 3)

This hydrolysis process produces a more toxic EHS, hydrogen fluoride, from an EHS, fluorine. Hydrogen fluoride has a LOC of 0.0016 gm/|m.sup.3~ (1). Using equation 1) the vulnerable zone can be determined by substitution of the LOC for C, the unknown concentration. A number of tables have been developed to show this relationship between the variables of Turner's equation.

Table 1 (1) depicts the vulnerable zone distances for different rates of release and levels of concern for a rural setting, atmospheric stability F, and a wind speed of 3.4 miles per hour.

Entering the table at a rate of release of 10 lbs. per minute and interpolating for the LOC of 0.039 gm/|m.sup.3~ for fluorine, the vulnerable zone distance or radius is 0.6 of a mile. This severely underestimates the vulnerable zone necessary to model the release. A graphic example of the results of modeling the fluorine LOC and not the hydrogen fluoride LOC is shown in Figure 1, superimposed on a map of South Dade County, Florida.

As can be seen from equation 1), the windspeed greatly influences the vulnerable zone for the EHS. A higher wind speed will reduce the vulnerable zone radius.

In addition the level of ozone goes up. Ozone is a reactant that is greatly involved in photochemical processes that can intensify atmospheric pollution. Ozone corrodes metals, cracks certain rubber compounds, and reacts with nitric acid to form nitrogen dioxide.

Determination of the vulnerable zone for boron trisulfide 

Boron trisulfide, |B.sub.2~|S.sub.3~, is a white crystalline compound at standard temperature and pressure. It is not listed on the EPA list of extremely hazardous substances. |B.sub.2~|S.sub.3~ reacts readily with moisture in the atmosphere to decompose per the following equation (6):

|B.sub.2~|S.sub.3~+6|H.sub.2~O|right arrow~2B|(OH).sub.3~+3|H.sub.2~S 4)

The reaction of boron trisulfide with moisture produces boric acid and hydrogen sulfide. |H.sub.2~S is listed as an extremely hazardous substance. Its LOC is 0.042 gm/|m.sup.3~. This is an example of a non-listed substance reacting and producing a listed substance. A 10 pound per minute reaction of |B.sub.2~|S.sub.3~ will produce the equivalent of approximately a 10 pound per minute leak of |H.sub.2~S. Table 1 indicates that a vulnerable zone of at least 0.5 mile is required to accommodate the release of |H.sub.2~S.

Discussion and conclusion

Fluorine reacts with the moisture in the air to produce hydrogen fluoride. This more toxic substance requires a larger vulnerable zone to adequately defend against its effects. About half a mile radius is estimated for fluorine, but the product of the reaction, hydrogen fluoride, requires 9.0 miles based on the EPA listed LOC. This is a 1,700 percent underestimation of the zone required.

Boron trisulfide is not included on the EHS list in the Technical Guidance for Hazards Analysis (1), but its reaction with moisture produces hydrogen sulfide, a listed substance. A vulnerable zone of 0.5 mile radius is calculated for hydrogen sulfide. There may be other such substances that should be included in the list.

Planning committees using models for the release of extremely hazardous substances should take into consideration the reactivity of the substance, because the Technical Guidance for Hazards Analysis (1) does not take this into account. The substance can become more of a threat to human life than the Technical Guidance for Hazards Analysis may indicate.

References

1. ... U.S. Environmental Protection Agency (1987), Technical Guidance for Hazards Analysis, Federal Emergency Management Agency, U.S. Dept. of Transportation, Washington, DC.

2. ... U.S. Dept. of Health, Education and Welfare (1970), Workbook of Atmospheric Dispersion Estimates, Public Health Service Pub. No. 999-AP-26, 3rd printing, Washington, DC.

3. Hanna, S.R., G.A. Briggs and R.P. Hosker, Jr. (1982), Handbook on Atmospheric Diffusion, Technical Information Center, U.S. Dept. of Energy, Washington, DC.

4. Stacey, M., J.C. Tatlow and A.G. Sharpe (eds.) (1961), Advances in Fluorine Chemistry, Vol. 2, Butterworths, London.

5. Jacobson, C.A. (ed.) (1959), Encyclopedia of Chemical Reactions, Reinhold Publishing Corp., New York, NY.

6. Trotman-Dickenson, A.F. (ed.) (1973), Comprehensive Inorganic Chemistry, Pergamom Press Ltd., Elmsford, NY.

Michael K. Miller, Greiner,Inc., 5805 NW 11th St., Miami, FL 33126.