pollution. Because England imports limited quantities of high-grade fossil fuels, the typical London "pea souper" consists mainly of sulfur compounds, particularly SO, which are produced by the combustion of bituminous coal, low-grade heating oil, acid manufacture, ore smelting, and other industrial manufacturing processes.

In the United States, New York and Chicago have record quantities of SO, in their atmospheres that are second only to London's. In all, about 60 percent of the American population is exposed to continuous peril from atmospheric contaminants (see Fig. 7). And it does not require a medical opinion to suggest that pollutants capable of corroding metal, darkening white paint (Fig. 8), disintegrating stone, dissolving nylon hose, and cracking rubber are somewhat less than beneficial to human lung tissue. There is ample circumstantial evidence to link air pollution with asthma, pneumonia, tuberculosis, pulmonary emphysema, lung cancer, and even the common cold. In 1962, the chairman of a panel of medical experts at the National Conference on Air Pollution Control stated: "The evidence that air pollution contributes to the pathogenesis of chronic respiratory disease is overwhelming."

During the symposium on the clean air problem at the recent American Power Conference, Dr. Dinman, in his opening statement, gave a concise description of the pathological effects of sulfur dioxide:

"To understand the effects of SO, on health, it is neces

sary to delineate those mechanisms whereby sulfur oxides alter human function. On a mechanistic basis, we may conceive of air conduction tubes (the tracheobronchial tree) as a series of interconnecting ducts. These ducts have the unusual capacity of changing their crosssectional area. This is accomplished by contraction of circumferentially aligned muscles. Thus, given a proper stimulus at certain receptions in the wall of this air conduction system, input from these receivers arrives at the brain. A flow of impulses, in turn, is transmitted to these surrounding muscles, which leads to their contraction with a decrease in cross-sectional area.

"The consequences of such decrease in cross-sectional area are apparent. Consider that a fixed volume of air per unit time must be available for oxygen extraction by the blood. Therefore, an increase in velocity is the only method whereby this fixed volume may be moved through this attenuated system. Obviously, the energy required per unit time to obtain this work function is increased. In individuals who have cardiac disease, these increased demands are met with difficulty and subsequent deterioration. Another complication stems from one other consequence of SO, or SO, (sulfur trioxide) or H,SO, (sulfuric acid) impingement on the lining of the gas-blood exchange surfaces. Impingement of these irritant polar compounds stimulates the release of a diluent at such affected surfaces. While this dilution increases pH toward normal levels, at the same time it imposes a thickened barrier to gas transfer across the membrane. Since this barrier is but a few hundred microns thick, this imposes no significant load on gas transfer in the normal person.

However, in those persons with already thickened gas exchange membranes due to chronic lung disease, this increased thickness essentially causes a barrier to oxygen diffusion with resultant asphyxia, and further deterioration as a result of decreased oxygen availability to the vital organs.

"... While there is much data from animal experimentation, there is relatively little human data even in normal persons. In such persons, 4–6 ppm of SO, produces consistently reproducible changes in airway resistance within 10 seconds to 4 minutes..."

Nitrous oxide is no laughing matter. Some medical experts regard nitrous oxide (NO), the common "laughing gas" administered as an anesthetic by dentists for tooth extraction, in the same insidious category as SO2. Recent evidence indicates that NO has teratogenic-and possibly carcinogenic-effects on animal and human receptors. In air pollution, this gas is emitted as a byproduct of hydrocarbon combustion.

In therapeutic medicine, bone marrow depressant effects have been observed as a result of the protracted administration of NO to control convulsions in tetanus victims. And, although any conceivable concentration of the gas as an air pollutant would be very low, the cumulative effects of continued exposure to this gas may be deleterious to human receptors.

Catalysts and buffers. Dinman further observed that with the addition of any suspended particulates, of less than 5 microns in size, the response of the human receptor to SO, may be accentuated. The reasons for this physiological reaction are twofold, and may be ascribed to

1. The interposition of particulate surfaces of relatively large areas for irritant gas adsorption, which would increase the gas concentration per unit volume.

2. The increased probability of impingement, which is a result of the different kinetic behavior of particulate vs. gas phase.

Although many particulates tend to aggravate the biological damage potential of SO2, it is apparent from actual case histories that the biopotency of these particulates is a function of their chemical properties. It is known that manganese catalyzes the conversion of SO, to H2SO1, but it is also possible for some particulates to buffer the physiological reaction of SO.

In the phenomenon of buffering it is known that inhaled hydrogen sulfide (HS) and other sulfhydryl compounds can protect mice from otherwise lethal exposures to ozone. Inhaled formaldehyde and SO, produced much more resistance to air flow in guinea pigs when applied with a physiologically inert aerosol than did the same concentrations of the gases alone.

Knowing the concentration of a pollutant does not necessarily indicate its physiological effect upon a receptor. The presence of other contaminants may either inhibit or increase the expected effect. Yet, almost no research has been conducted to determine the effects on re

ceptors of long-term exposure to known admixtures of pollutants. Nevertheless, many proposed air standards for urban areas are predicated on consideration of the exposure effects of a single pollutant at a time.

The final medical factor is individual sensitivity or allergic reaction. Heretofore this has been a rather vague and nebulous concept, but recent evidence is accumu lating to verify this phenomenon.

Methods of reducing SO, and SO, emissions from coal Based upon statistics available for the year 1962, and by combining a knowledge of the sulfur content of coal seams being mined in various states, Table I indicates the range of sulfur contents and the corresponding percentages consumed in the United States.

In view of the very large estimated demand for electric power generation in geographic areas where coal is the preferred fuel, the quantities of SO, released to the atmosphere will rise alarmingly unless means are developed either to remove the sulfur before combustion or to remove the SO, from the stack gases. One of the prime difficulties in achieving the former objective is that a portion of the sulfur content (20-60 percent) is chemically bound as organic sulfur, and this can only be removed by very complex and expensive chemical processes.

At present, three high-temperature processes' are being operationally tested for electric utility applications. These are: the alkalized alumina, the Reinluft, and the Pennsylvania Electric.

Bureau of Mines' alkalized alumina process. In this process, flue gas containing SO, and SO, are absorbed by alkalized alumina-Al(OH), in a vessel at a temperature of 625°F (see Fig. 9). The alkalized alumina is regenerated in a second vessel at 1200°F, by using producer gas or re-formed natural gas. The product gas from the regenerator is then introduced to a sulfur recovery plant in which elemental sulfur is produced.

The flue gas used in the Bureau's pilot plant is made from the combustion of powdered coal and it contains all the impurities that might affect the absorption and regeneration cycles. Tests have been conducted to establish the optimum conditions of temperature and time for both the absorption and regeneration, and various procedures for preparing the alkalized alumina have been tried because, in the repeated cycles, physical and chemical changes occur that may degrade or poison the absorbent.

The experimentation with this process has led to the construction of a larger pilot plant in which variables can be studied more efficiently. The advantages of the process include a low pressure drop of the flue gas during absorption, operation over a wide temperature range of 250°-650°F, and the ability to obtain elemental sulfur as the end by-product.

Reinluft process. In this method, flue gas at 300°F is forced upward through a filter bed of activated charcoal (see Fig. 10) that is slowly descending through the adsorber. The SO, is adsorbed directly and the flue gas is then cooled to 220°F. At this temperature, SO, is oxidized to SO, which is then adsorbed on the activated charcoal. The SO, combines with the adsorbed water from the flue gas to form dilute H,SO,.

The activated charcoal, with the adsorbed dilute H-SO,, is next regenerated in a separate vessel by the recirculation of product gas heated to 700°F. The dissociated H,SO,

products react chemically with a portion of the carton to form a gas that contains a high concentration of CO and SO2. The latter gas is converted to H,SO, in a contact acid plant. After cleansing, the regenerated char is recycled to the adsorber.

At present, two commercial plants are under construction in Germany to use this method of SO, removal. One of these plants will service flue gases produced from lowgrade fuel oil, and the other will be used in connection with a coal-fired installation.

The Reinluft process is particularly feasible if there is a nearby industrial requirement for sulfuric acid.

The Pennsylvania Electric process. The Pennsylvania Electric Company has constructed a pilot plant at its Sewart generating station to remove SO, by the catalytic conversion of SO, to SO1. Sulfuric acid is formed and collected on the cooling water stream that contains the SO1. The objectives of the pilot plant are fourfold:

1. To determine the effect of actual flue gas, with time, on catalyst activity.

2. To establish the degree to which the flue gases must be precleaned to prevent catalyst fouling.

3. To calibrate the rate of catalytic oxidation of SO, and flue gas pressure drop so that large-scale plants can be sized.

4. To determine removal methods for the acid and the quality of the acid produced.

Reports indicate that the pilot plant has been operated successfully and that adequate data have been gathered for the design and construction of a full-scale plant. Many variables, however, still must be investigated, such as the life expectancy of the equipment, required construction materials, and the character of the maintenance problems that will be experienced.

At the present time there is insufficient information

available to make accurate cost estimates for the comparison of the three methods. Each of the described processes has its unique advantages, but each requires additional development to verify critical process variables that have an important bearing on the eventual economics. Also, market studies are needed to evaluate the industrial requirements of the manufactured by-products. Magnetic separation of sulfur. An interesting new process, called magnetic separation, has been reported by the Russians. Essentially, this method involves a flash coking step in which the iron pyrites containing the sulfur fraction in the coal are modified into a magnetic form to simplify the separation procedure.

In this technique, pulverized coal is heated to about 640°F for a period of 2-5 minutes, during which time a jet of steam and air is blown through the heated coal. This treatment produces a surface layer of magnetized material on each pyrite crystal, and separation is made in a magnetic field of 10 000 gauss. By discarding this magnetic fraction, the sulfur content of the coal can be reduced by about 0.5 percent and the ash content by 50 percent. The U.S. Bureau of Mines is investigating this process to determine the feasibility of its application to the treatment of coal in the United States.

So, monitoring and control-TVA experience

During the past 15 years, the Tennessee Valley Authority has added 52 coal-fired, steam-electric generating units, located in eight plants and ranging in size from 125 to 650 MW, to its power production facilities. And a single-unit plant of 900-MW capacity (Bull Run) is scheduled to go on the line in 1966.

Beginning with the first large steam-electric station, TVA conducted extensive air pollution studies at each plant site. The experience and data obtained by these studies have been applied in planning air pollution control at subsequent plants and for additional units in existing plants.

Long-term records of meteorological data and SO, concentrations from permanent monitoring stations have

Total time that SO2 equals or exceeds indicated concentration, percent (0.1% frequency is approximately two times per year)

Fig. 14. Vertical-view diagram of a two-stage electrostatic precipitator. "1" indicates grounded cylinders; "2," corona wires; "3," grounded collector plates; "4," charged plates are similar to "2" in polarity.

3

3

3

4

Polluted gas

Cleansed gas

3

3

3

3

been augmented by extensive mobile sampling by car and helicopter equipment and by full-scale atmospheric dispersion studies. Since most of the steam-electric plants in the TVA system are geographically located in areas remote from other significant sources of SO, it is felt that the test results are most closely representative of the flue gas distribution patterns of modern coal-fired power plants. The SO concentrations in these tests refer to 30-minute average concentrations.

Frequency distribution of SO concentrations. A logarithmic plot of frequency of SO, concentrations at fixed monitoring stations has consistently indicated a fairly straight line. Figure 11 shows this distribution as measured by an autometer situated where maximum concentrations occurred at ground level in the vicinity of a 4-unit plant, with two 500-foot-high stacks, and with a total generating capacity of 1050 MW. For the sampling period approximately 19 months-the highest recorded concentration was about 0.6 ppm for three 30-minute periods. And SO; concentrations were 0.2 ppm or above for only eighty-four 30-minute periods, or approximately 0.40 percent of the time.

Similar data obtained from Public Health Service studies in Nashville, Tenn., are also plotted in Fig. 11. Although the maximum SO, concentration recorded was only about 0.3 ppm, the concentrations of this gas were 0.2 ppm or more 14.1 percent of the time. And the estimated SO, emissions in the urban area were less than half of that recorded at the power plant.

Although higher concentrations of pollution in urban areas tend to occur during periods of low wind velocity and temperature inversion, the higher levels of pollution in the vicinity of large power plants tend to occur during moderate to high wind speeds and neutral atmospheric stability conditions. Since none of the TVA plants are located in large urban areas, the data do not provide a direct quantitative measurement of the contribution of a large power plant to an urban pollution problem. But data analysis from an autometer located in a small town near one of the large plants indicated that SO, in detectable amounts was present 14 percent of the time.

Effect and influence of stack height. Monitoring data and data obtained from full-scale dispersion studies have been used in estimating stack height requirements for TVA plants. An example of estimates made by empirically derived formulas, based upon monitoring data, is shown in Fig. 12. The subject of this graphic plot was a two-unit, two-stack plant, with a generating capacity of 1800 MW, and an estimated SO, emission rate of 810 tons per day. This emission rate is calculated empirically from SO, monitoring data at plants with 250-, 300-, and 500-foot-high stacks. The curve for a 400-foothigh stack is interpolated, while the curves for the 600and 800-foot-high stacks were extrapolated.

The electrostatic precipitator-TVA experience. The practical application of the electrostatic precipitator was first demonstrated by F. G. Cottrell in 1906. Essentially, it is a device that is used to remove liquid chemical mists or solid particulates from a gas in which they are suspended. Electrostatic precipitation is a two-stage process. In the first step, the gas containing the suspended particulates is passed through an electric, or corona, discharge area in which ionization of the gas occurs. The ions produced collide with the suspended particles and impart an electric charge to them. These charged particles

then drift toward an electrode of opposite polarity, and they are deposited upon this electrode, where their electric charge is neutralized.

In its most elementary form, the precipitator configuration may consist merely of a vertical tube that contains an insulated concentric wire (see Fig. 13). When a dc potential of 10-100 kV is applied to the central wire, a corona discharge occurs in a small area surrounding the wire. The suspended particulates are ionized in the corona discharge and migrate to the tube wall. If the suspension is liquid, it will accumulate on the wall and coalesce into droplets that can be drained from the base of the tube. Suspended solid particulates can be removed from the tube wall by mechanical vibrators or scrapers, and then discharged into a cyclone or dust collector at the bottom of the apparatus.

In more complex configurations, the ionization may occur in one vessel and the deposition and precipitation in another. Figure 14 shows the plan view of a simplified two-chamber apparatus. In the first chamber, the particles become charged but are prevented from depositing