Editor’s Note: Recently the US Demilitarization Program recommended that in the event of an accident at any CW destruction facility, local citizens should close their windows and use duct tape and plastic to seal their rooms and homes. Please read on. The ASA Newsletter will accept response to this paper.

Ventilation Kinetics of Sealed Shelters

by Evan E. Koslow Ph.D.

Introduction
The kinetics of vapor penetration into sealed enclosures has been extensively studied. In most cases, the sealed enclosure is supplied with purified air from a combined HEPA and activated carbon filter and a net positive pressure is maintained into the shelter to substantially eliminate the diffusion of toxic vapors from the external environment.

However, the special case of shelters that have no air purification equipment has also been studied. In these cases, the shelter simply delays the diffusion of toxic vapors, gases, or particulates. It is assumed that some residual leakage is always present. Therefore, this type of protective shelter can provide only short-term protection from an external threat.

The general theory of ventilation kinetics into shelters is provided by R. L. Helmbold (1982). One of the fundamental results of this model, and all other ventilation models, is that given sufficient time, the dosage received inside a shelter lacking positive pressure and filtration will equal the dosage one would receive outside the shelter, i.e., there is no benefit if the occupants stay too long within the shelter. Although this result is derived from essentially all models of limited-diffusion shelters, this concept is often misunderstood and the value of such shelters is often exaggerated compared to its actual benefits. This type of shelter is usually of value only for protection against large chemical agent particulates and droplets and would be of limited value against a vapor threat.

Detailed Analysis
The dosage received inside a shelter having limited, but finite, diffusion of toxic vapors from outside the shelter approaches the dosage received outside the shelter as time (t) ---> infinity. For example, let us assume that a high-concentration pulse of chemical agent passes through the environment outside the shelter. Although the concentration of toxic vapor that penetrates into the shelter is lower than experienced outside the shelter, the limited rate of diffusion also dictates that this low concentration of chemical agent inside the shelter can not readily be dispersed. Unless the occupants of the shelter emerge within a reasonably short period of time, their total exposure will approach that experienced outside the shelter.

In 1980, the U.S. Army completed a study of Soviet armored fighting vehicles (Ferriter, J.M. and K. Bellone 1980). Because of the very low air leakage rates for these vehicles, the U.S. Army concluded that these vehicles would be able to traverse NBC-contaminated terrain, even if their cockpits were not equipped with NBC protective filtration systems. However, a reexamination of the data by Koslow (1984), demonstrated that the dosage actually measured inside the vehicles was essentially equal to that outside the vehicle within less than 100 minutes. Hence, even well sealed armored vehicles provided no substantial protection against a chemical threat.

Another important result of ventilation kinetic models is that the dosage within a shelter exposed to evaporating liquid chemical agent is equal to the quantity of chemical agent evaporated (milligrams, mg) divided by the volume ventilation rate in cubic meters per minute (m3/minute). This result remains true regardless of the size of the shelter. As such, if contamination is introduced into a shelter in liquid or vapor form, the lack of adequate pure-air ventilation will result in a maximum received dose. This is confirmed by the work performed by McAndless et al. (1983) where the impact of small volumes of liquid agent introduced into a model shelter was measured.

Zeller (1970) performed an extensive study of ventilation kinetics of military enclosures, including both a review of existing models and test data. Essentially all of the models examined, the Uniform Mixing Model, the Nonuniform Mixing with No Diffusion Model, and the Nonuniform Mixing with Diffusion Model, all converge on the same result. As time (t) approaches V/Q Volume of the shelter divided by the ventilation rate (leakage rate), the dosage inside the shelter approaches the dosage outside the shelter. Therefore, if a shelter is 1000 cubic feet in volume and has a leakage rate of 20 cubic feet per minute, the dosage inside the shelter will approach the dosage outside the shelter in approximately 50 minutes.

Actual shelter leakage given as a ventilation period is equal to the shelter volume divided by the volume flow rate. The volume flow rate is dependent upon both the number and size of individual leakage paths and certain environmental factors, especially wind speed, and secondarily on wind direction and turbulence. Low wind speeds reduce agent dispersion, but may also limit the diffusion of chemical agent into shelters.

These models and empirical data do not take into account certain complexities, such as the adsorption of chemical agent onto surfaces within the shelter environment, subsequent desorption of this chemical agent, and the ability to potentially emerge from the shelters after the concentration outside the shelter is lower than the concentration inside the shelter.

Karlsson et al. (1992) have studied the possible impact of adsorption within the shelter environment and both modeled the impact of adsorption and measured adsorption with simulants in actual household environments. Their conclusion is that the protective effect of adsorption increases for small volumes of air exchange (i.e., for well-sealed shelters). In their work, trialkylphosphoacetate (TA PA), a VX simulant, was evaporated into a room together with an inert tracer gas. The TAPA showed a strong initial deposition on shelter surfaces. However, after the initial period, TAPA was reentrained into the shelter air by desorption. Textile surfaces appeared to participate in this weak adsorption behavior. As such, the benefits of the transient adsorption occurring during the first hour will be lost during the subsequent ten hours.

Shelters with NBC Purified Air
If we assume that a shelter is ineffective unless provided with a source of purified air, and that this pure air source is used to both pressurize the shelter to prevent the diffusion of chemical agent into the shelter and to maintain an essentially zero concentration of chemical agent within the shelter, we must obtain an estimate of the air flow required to both maintain shelter pressurization and to maintain a healthy condition for the occupants within the shelter.

The Defense Civil Preparedness Agency (DCPA) Attack Environment Manual, Chapter 7 (DCPA, 1973 describes the minimum standards for nuclear shelters). The first requirement is to provide a minimum of 10 square feet (65 cubic feet) of shelter volume for each person. Hence, a shelter for 10 persons must be approximately 10' x l0' in size and have a height of approximately 6.5'. A similar shelter for 25 persons must be 10' wide and 25' long and standard 6.5' high. However, this amount of space is only suitable for short term confinement.

The minimum air supply is dictated by either the amount of air required to dissipate heat, remove carbon dioxide, or to obtain adequate pressurization of the shelter. If the shelter consists of a well-sealed plastic enclosure, the leakage rate can be assumed to be very small compared to the flow required to obtain adequate ventilation.

The accumulation of carbon dioxide is usually the most serious shelter problem. Normal air contains approximately 0.04% carbon dioxide. The latest standards for shelters allow the accumulation of up to 1% carbon dioxide and have desirable goal of 0.5% carbon dioxide. This minimum requirement can be met with an air flow of 3 cubic feet per minute per person within the shelter.

Another requirement is to maintain shelter ventilation sufficient to prevent the buildup of heat and humidity. Each shelter occupant releases approximately 500 Btu/hour. To prevent excessive heat buildup within the shelter environment, it is assumed that the shelter operates within a building often equipped with air conditioning. Regardless, a warm weather region often has a requirement of 20 SCFM per person to obtain a 90% probability of maintaining an effective temperature (ET) less than 82•F, when the system operates within a building without air conditioning.

The shelter requires an over-pressurization valve to simultaneously obtain a controlled over-pressure and ventilation rate. This valve is normally sealed prior to actuation of shelter deployment, to allow maximum blower pressure to be applied to the inflation of the shelter or to maintain the shelter in a positive pressure condition. Thereafter, the occupants of the shelter must remove the protective cover from the over-pressure valve to allow the shelter to operate at the optimum ventilation and over-pressurization.

Conclusion
Shelters without filtration and positive pressure are very limited in their use. They are limited not only by the diffusion of contaminated outside air into the interior, but also by the unsafe CO2 and heat buildup. Advocating their use in all but extreme, short time circumstances is promoting unsafe practices.

Bibliography:

1. Ferriter, J.M. and K. Bellone 1980. Leakage assessment and challenge testing of the BTR-5OPU, BTR-6OPB, and the BRDM-2 SAGGER Antitank Guided Weapon (ATGW). Report AD-E41 0253
2. Helmbold, R.L. 1982. A general mathematical treatment of hazards to NBC collective protection systems with applications to particular cases. Naval Air Development Center, Warminster, PA Report No. NADC-82255-20
3. Karlsson, E., T Bergland and L. Rittfeldt. 1992. On the protective capacity of sealed rooms against vapour clouds. Symposium on Protection Against Chemical Warfare Agents, 8-11 June 1992, Stockholm, Sweden. National Defence Research Establishment, Department of NBC Defence, S-901 82 Umeå, Sweden.
4. Koslow, E.E. 1984. Reexamination of aerosol and vapor infiltration data on the Soviet BTR-5OPU, BTR-6OPB, and BRDM-2 SAGGER Antitank Guided Weapon (ATGW) armored fighting vehicles. Private Report.
5. McAndless, J.M. M.E. Galloway, and C.L. Chenier 1983. Removal of irritant vapours from enclosures by ventilation. Defense Research Establishment Suffield, Canada, Suffield Report No. 360, PCN No. 13E30.
6. Zeller, H.W. 1970. Investigation of the ventilation kinetics of military enclosures and vehicles. Applied Science Division, Litton Systems, Inc., Minneapolis, MN. Contract No. DA18-035-AMC-370(A), Report No. 3426.

Editor’s Note: Dr. Evan Koslow is President and Chief Executive Officer, KX Industries, the world’s largest producer of extruded carbon. Although Dr. Koslow had written this paper some time ago, the dynamics of the current situation dictate that this short paper be provided as a consideration in developing strategies for protection of and in surrounding communities. Dr. Koslow can be reached at tel: 1-203-799-9000 and fax: 1-203-799-7000 and e-mail .

00-5, issue no. 80