Click for next page ( 204


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 203
CHAPTER VI The Risk/Benefit Equation in Chemistry Gasoline is a mixture of hydrocarbons, mainly alkanes and alkenes, but with aromatics or tetraethy! lead added to improve combustibility. It is toxic to drink, the aromatics are carcinogenic, the tetraethy! lead can cause lead poisoning, the mixture is extremely flammable, and, when burned, it produces the noxious and toxic substances nitric oxide and carbon monoxide. Gasoline is probably the most dangerous compound that the average person will ever encounter on a daily basis. Yet this same average person stores 5 to 10 gallons at home (in the car's gas tank), and he or she might purchase some 500 gallons of this perilous liquid every year, and then combust it to release into the atmosphere 250,000 cu.ft. of oxides of nitrogen and carbon. Every city has dozens of repositories for this fluid, gasoline stations, each storing perhaps 10,000 gallons, always in a crowded neighborhood. These supplies must be regularly replenished by trucks that carry 20,000 gallons each through the city amongst the normal traffic. This is risky business! But evidently the public (that's you and me) believes that the benefits are important enough to justify the risks. Every time you start your car, you are implicitly stating that the r~skJbenefit equation in our reckless use of gasoline comes down on the side of benefit. Dichiorodipheny} trichioroethane is an insecticide that has saved over 50 million human lives and prevented untold suffering. It did so by almost eliminating malaria. In 1959, at the height of its use, over 156 million pounds of this chemical, commonly called DDT, were produced in the United States alone. Many thousands of people were dusted with DDT, literally from head to toe, without apparent harm. Such programs reduced the number of malaria victims in India alone from 75 million in 1952 to less than 100 thousand by 1964. In Sri Lanka, the number of malaria cases dropped from about 3 million, with 12,000 deaths per year, to less than 100, with no deaths at all. But production of this lifesaving chemical has been sharply cut back, and in the United States it has almost gone out of use. The reason is that DDT has been judged to be potentially dangerous because of its persistence in the environment and because of its accumulation in the tissues of living organisms. This is a dilemma. There is no doubt that human lives could be saved, but that ecological disturbances will accompany continued use of DDT. The ban against its use indicates that the public has decided that the risks exceed the benefits. 203

OCR for page 203
204 THE RISKIBENEFIT EQUATION IN CHEMISTRY Gasoline and DDT are only 2 of the 70,000 or so chemicals that have come into widespread use. These chemicals range from aspirin to Vitamin C, flea powder to household detergents, Dacron shirts to Teflon-coated frying pans. Our quadity of ~ . ~ . ~ , . ~ . . ~ . . ~ . , ~ . ... lute Is determined, sustained, and constantly Improved by our ability to control chemical reactions and to make chemical compounds that are useful in everyday life. But gasoline and DDT are excellent examples because they point out vividly the fact that handling all of these chemicals must involve some risks. We have only recently realized that minimizing risks must be considered to be as important a dimension of progress as maximizing benefits from technological change. We must learn to deal wisely with the risk/benefit equation. FEARS OF THE MODERN AGE In a recent, highly publicized incident, the compound trichioroethylene, TCE, was found at several parts per million in the Unnking water from 35 private wells near Palo Alto, California. A lawsuit caused these wells to be closed. Yet the available evidence implies that these wells are probably quite safe if that is so, then why are they closed? Part of the answer is that there is a growing alarm in this country over chemicals and their possible effects on individuals. The populace is concerned about chemical pollutants, additives, waste, by-products, residues in short, any chemical that is the direct result of technological change. Some of this current fear stems from a general lack of information a fear of the unknown; some is triggered by sensa- tional or overzealous reporting in newspapers and on television. Many people fee! a sense of helplessness, a feeling they have no say in the control of these new substances entering the environment. There is also a vague feeling of mistrust of the priorities and interests of those with a vested financial stake in producing, distributing, and utilizing these substances. Yet aD of these chemicads came into being to establish and maintain our high quality of life. For decades, we have been blithely enjoying the fruits of our technological success without thought of the possible incumbent hazards and undesired ejects. Now, suddenly, our society has become "chemically aware"; we have become hypersensitive about all chemicals, no matter what the source, amount, degree of hazard, or intended purpose. ~ . . . . , , , untortunate~v such fear can lead us to overreact. so that situations presenting minimal danger can divert attention and resources from real dangers that must be corrected and eliminated. We must alleviate this chemical insecurity so that we can find the optimum balance in the risk/benefit equation. Then we can continue to enjoy the growing benefits of the chemical age while guaranteeing that we protect the health and well-being of our planet and its occupants. WHAT IS TOXIC? "Everything is poisonous. The dose alone determines the poison." Paracelsus, the sixteenth century chemist, physician, and healer, made this assertion, one that comes to mind when we read that a healthy diet should not contain too much salt, sugar, or butterfat. Nitrogen is 80 percent of the air we breathe, but too much

OCR for page 203
THE RlSKlBENEFIT EQUATION IN CHEMISTRY nitrogen acts as an anesthetic, and it can give deep sea divers a dangerous feeling of euphoria called "rapture of the depths." Selenium is essential to human and animal health, but in excess can cause a variety of ailments. And if too much salt, sugar, butterfat, and nitrogen is unhealthy, perhaps we should believe Paracelsus- everything is poisonous if taken in too large amounts. This news can be a bit unnerving since everything around us and in us is chemical, including everything we eat and drink. But it is reassuring, too, because we, and all other life forms, evolved and thrived in the presence of these chemicals. Perhaps the persistence of life is evidence that the corollary to Paracelsus's remark is true as well things that are poisonous in too large doses are not necessarily poisonous in small enough doses. With that premise, we are faced with two fundamental questions. First, we must find out what hazard levels we face risk assessment- and then we must decide what to do about themrisk management. The Environmental Protection Agency wisely recognizes these as separate dimensions of their responsibility. Risk assessment is mainly connected with the known scientific facts about a given possible hazard. Risk management requires choices among options, as well as consideration of costs and social consequences. We will discuss them in turn. RISK ASSESSMENT . . . To begin with, there are two kinds of toxicity to worry about. A toxic chemical can cause illness soon after exposure, which is called acute toxicity. Another chemical can have no immediate effect, but it may be injurious much later, after continuous, long-term exposure; this is called chronic toxicity. For example, phosgene, (2, iS an acutely toxic gas that is accidentally produced by use of a carbon tetrachIoride fire extinguisher to put out an electncal fire. A concentration of 5 parts per million (ppm) will cause eye irritation in a few minutes, whereas greater than 50 ppm would probably be fatal. On the other hand, benzene, C6H6, is chronically toxic its vapor inhaled at the same level, 50 ppm, would cause no immediate effect, but if inhaled every day for many months or years, benzene can cause decreases in red blood cell count, hemoglobin level, and the number of leukocytes in the blood. Unfortunately, it isn't easy to get such detailed information. The most definitive way, the hard way, is to have enough people exposed to a given chemical to show that it is safe or to show the exposure at which toxicity begins to appear. Plainly, it is hardest to learn about chronic toxicity. Very large populations must be exposed for long periods to give a statistical chance of establishing something useful. That is what epidemiology is all about. What Is Epidemiology? Historically, epidemiology is the study of epidemics, contagious diseases that spread rapidly. But today, epidemiology is also used in a statistical way to try to detect acute or chronic toxicity even when the ejects on health are quite small. For example, vinyl chIonde, CH2CHCI, is known to be carcinogenic (a cause of cancer). The reason is that a very rare form of liver cancer, angiosarcoma, is found 205

OCR for page 203
206 THE RISK/BENEFIT EQUATION IN CHEMISTRY statistically to be concentrated in a small number of workers who have been continuously exposed over long periods to high levels of viny! chloride, in the hundreds of ppm. Here we can reach the confident epidemiological conclusions that this chemical has a toxic effect and that the degree of hazard to the general public is extremely small. What Causes What? Unfortunately, epidemiological data can be misinterpreted, even when the statistics are firm. It is one thing to show (1) that colon cancer is much more prevalent in the United States than it is in India, and (2) that Americans eat more dairy products than people in India do. Before jumping to the conclusion that dairy products cause colon cancer, we must remember that colon cancer appears in older people and that U.S. citizens live a lot longer (on the average) than citizens of India. Thus, the opposite conclusion might be reachedthat eating dairy products allows one to grow old enough for colon cancer to show up (from other causes). Epidemiology can show "association" but not necessarily "causality." The epidemiologists' joke is that the twentieth century growth of population in Western Europe has decreased at about the same rate as the decrease in the number of storks. Few of us would conclude that the human birth rate is going down because there aren't enough storks for the deliveries. Animal Tests These difficulties have driven us to the use of laboratory test animals as substitutes for humans. Without debating the ethics of such practice, we observe that among the animals so used are mice, rats, guinea pigs, monkeys, hamsters, dogs, cats, pigs, and even fish. Mice and rats are used most, probably because they are inexpensive, they breed rapidly, and their use is generally accepted. In a typical study, groups of a few thousand mice might be exposed to two or three different doses of a particular chemical every day for 2 years (including a zero dose for a control group). Then these mice are killed ("sacnficed" is the term used) and every mouse is autopsied to search for tumors. The statistical differences between the control group (zero dose) and the exposed groups are taken as a measure of hazard. Such an experiment might show that one milligram of chemical X eaten every day by a half-pound mouse causes a 14 percent increase in stomach cancer. To decide what this means for us, we usually assume that a human weighing about 150 pounds, 300 times more than the average mouse, will need to eat about 300 milligrams per day of chemical X to have about the same 14 percent cancer probability. Thus, toxic doses are expressed in terms of milligrams per kilogram of body weight (mg/kg). For acute toxicity, a substance legally earns the title "poison" if 50 percent of a test group of animals dies from a dose of 50 mg/kg or less. This dose the one that causes a 50 percent death rate is called the LD50 dose (lethal dose, 50 percent death rate). Thus, strychnine is a poison it takes only 1.2 mg/kg body weight to kill 50 percent of a rat population. On the other hand, tnchioroethylene (TCE) (recently found in water wells at ppm concentrations) is not called a poison since a rat has to eat 7,200 mg/kg of body weight to reach the LDso level. Transferred to

OCR for page 203
THE RlSKlBENEFIT EQUATION IN CHEMISTRY humans, this means that a 150-pound adult would have to eat about 3 pounds of TCE per day to receive this same dose, based on body weight. A 50-pound child would have to drink 4,000 gallons per day of well water containing 25 ppm of TCE to receive this dose. Since we cannot deliberately use human populations to test poisons and possible cancer-causing chemicals (carcinogens), this use of animal subjects is at least a rational approach. Even then, it raises the difficult question of whether animal responses provide a reliable estimate of human responses. After all, we aren't really trying to protect mice with this testing we have in mind the health of human beings, who come from an entirely different rung on the evolutionary ladder. The uncertainty comes when we have to rely only on this method to make decisions. As an example, the compound 2,3,7,8-tetrachiorobenzo-p-dioxin (popularly called "dioxin") is extremely poisonous to guinea pigs. For these little fellows, the LD50 is only 0.6 milligrams per kg body weight. In astonishing contrast, it takes 10,000 times larger doses to reach the LD50 level for hamsters! From species to species, we find enormous variation in toxic responses to dioxin. For this substance, we have had many documented human exposures. Among 400 severe- exposure cases that occurred 20 to 35 years ago, painful skin lesions were the only definite injury that appeared, and no deaths attributable to the exposures have yet occurred. In this case, we can't even predict usefully the lethal dose of dioxin to a 50-gram hamster from measurements on a 200-gram guinea pig, let alone a 150-pound human. This type of uncer- tainty is part of every regulatory decision in which tolerance limits must be estab- lished for humans with only animal tests as a guide. Is There a Dose-Time Relationship? TABLE VI-1 Dioxin's Lethal Dose Varies from Species to Species Animal Guinea pig (male) Rat (male) Rat (female) Mouse Rabbit Dog Bullfrog Hamster LD50 (mg/kg) 0.6 22 45 114 115 >300 >500 5,000 There remains one further question to complicate this perplexing but crucial issue of risk assessment. It is natural to wonder, if a large dose of something is poisonous in a short time, whether a small dose of the stuff is also poisonous, but over a longer time? For example, suppose we want to eliminate disease-spreading rats with the fumigant ethylene dibromide, C2H4Br2. Lab experiments show that rats are killed by breathing 3,000 ppm of this gas after 6 minutes of exposure. If that is so, how long would it take to have a lethal dose at 300 ppm? The simplest guess we can make is that at one-tenth the concentration, it takes just 10 times as Tong, 60 minutes, one hour. Actually, this linear assumption with its one-hour estimate is pretty close to what is found in lab tests. Does this, then, let us predict the lethal concentration for a 6-month exposure? Six months is 4,320 hours, so our linear model predicts lethality at the very low concentration of 0.07 ppm (300/4,3201. This time, however, experiment shows that the 6-month LD50 exposure for rats is 50 ppm. For prolonged exposure, the rats can tolerate much more ethylene dibro- 207

OCR for page 203
208 THE RlSKlBENEFIT EQUATION IN CHEMISTRY mice 700 times more than we estimated. In this case, the linear assumption has failed. This is not an isolated case. We have already mentioned selenium, which, at low concentrations, is essential for both human and animal health. At higher concen- trations, selenium produces serious health effects. Evidently, this contradicts the linear model, which gives no clue to the beneficial effects of selenium but, instead, would lead us to expect selenium to be toxic at any level if exposure is long enough. Carbon monoxide, a treacherous poison, provides another clear example. In the blood, CO bonds to hemoglobin and renders it useless as an oxygen carrier. If about one third of the hemoglobin is so tied up, the victim dies. This would happen to the average person after one hour of exposure to 4,000 ppm of CO (a partial pressure of 3 torr). From this evidence, the linear mode] would predict that 1 ppm would be lethal in about 4,000 hours, i.e., in about 6 months. However, the natural atmosphere we breathe all of our lives always contains about 1 ppm of CO and is clearly not lethal. There are examples in the opposite direction as well. Liquid mercury has a rather low vapor pressure, about one millito~T, and breathing it continuously has no immediate effect on health. However, the body cannot efficiently eliminate mercury, once ingested, so it accumulates over time. After many years of continuous exposure, venous undesirable symptoms begin to appear, including unsteadiness, inflammation of the gums, general fatigue, and headaches. In this case, the absence of short-term effects gives no warning about the chronic, fong-term danger of continuous exposure to the vapor of liquid mercury. All of these examples suggest that we must be very careful when trying to extrapolate data concerning potentially harmful substances. It is not possible to predict with confidence the long-term, low-exposure toxicity of a given chemical just from evidence about toxicity at high exposures for short times. Summary We see that risk assessment is a difficult business. Nevertheless, it is an essential part of maintaining a healthy environment. We want to enjoy the benefits of technological advances, so we must learn to evaluate possible undesired side effects. We cannot afford to ignore the possibility that something might be hazardous, but at the same time, we cannot be paralyzed by indecision or fear. RISK MANAGEMENT Our everyday lives are full of risks, but this is not a new situation. We evolved in a threatening environment. Nature provides many risks free of charge: tornados, hurricanes, avalanches, earthquakes, floods, fires, volcanic eruptions. But many of the risks were "invented" by the human race from catching one of the plagues which were spawned in the Middle Ages by the growth of cities, to falling off your horse while riding off to war. Our modern list is getting longer and longer it includes automobile accidents, airplane crashes, ferryboats sinking, muggings in Central Park, smoggy air, and catching colds in the subway. Some risks we choose to take, like risking skin cancer while getting a suntan; others we may choose to

OCR for page 203
THE RISKIBENEFIT EQUATION IN CHEMISTRY avoid, like smoking cigarettes. Some risks we prefer not to take, but cannot find a way to avoid, like living under the ominous threat of nuclear war. Thus, the sheer volume of risks we face seems to be steadily increasing with time. Yet a look at one particular measure, life expectancy, reassures us that we are not only holding our own in this battle, but perhaps actually winning. In the United States, life expectancy has steadily risen throughout this century, and it continues to rise at a rate of 3 years of additional age per decade. This impressive statistic can ease some of the anxiety about modern, technological nsks, and steady our resolve to deal with them in an attentive, pru- dent, and rational way. Dealing with them begins with thinking about what is an "acceptable risk." Acceptable Risk Notice that we propose to "think about," not to "de- cide," what is an acceptable risk. That is because risk ac- ceDtance is a highly personal O() 1: Llle Expectancy at Age 4 5 _ 50t 209 - _o~ ~ Late Expectancy at Blrth ~0 I , ~ I , 1 900 1 920 94(3 YEAR 960 1980 LIFE EXPECTANCY IN THE UNITED STATES TIlREE YEA RS 1 NCREASE EV ERY DECA DE ~ _ . and subjective thing. It can be arbitrary and even contradictory. An individual can decide to avoid the dangers of hang gliding but regularly Unve an automobile at 60 mph with seat belt unfastened. Another individual may smoke cigarettes but vigorously shun marijuanaor vice versa. Some risks we take unthinkingly because they are familiar. Californians are used to earthquakes, Nebraskans to tornados, and Floridians to hurricanes. Everyone who can adore it goes from here to there in airplanes equipped with life rafts but not parachutes. What factors are at work as each of us decides what to worry about? While the answer to this question varies enormously from person to person, there seems to be one factor that is generally important. Most people are highly sensitive about taking a risk involuntarily; they like to have a choice in the matter. The same individual who may choose to try skydiving, smoke cigarettes, consume caffeine, take birth control pills, or just cross a busy street in the middle of the block is likely to object strongly to the news that a pesticide might be found on his or her fresh fruit. People just fee! better making their own choices. The current nervous- ness about chemicals probably stems largely from the feeling that chemical risks are being taken involuntanly and that they are increasing. On the other hand, it is difficult for most people to deal with the magnitude of a risk. How does one evaluate the news that eating peanut butter might involve a risk of one in 500,000 that its 2 parts per billion (ppb) aflatoxin content might result in

OCR for page 203
210 THE RISKIBENEPIT EQUATION IN CHEMISTRY cancer? The numbers tend to blur in the face of the news that a dreaded disease may be involved. But there is one way to evaluate risks that most people can appreciate and use as a guide in decision making. It is to compare the risk for an unfamiliar hazard with another risk that is usual and similar in kind. That type of comparison calibrates for each of us what is an acceptable risk. . Comparable Risks Let us begin by examining quantitatively the magnitude of a risk almost everyone takes voluntarily and daily: riding in an automobile. Ample statistics indicate that the chance of being killed while driving 3 miles is about one in a million. That means that a city dweller who travels that much every day for a year takes an annual risk of about 4 in 10,000. Over such an individual's 60-year lifetime, the chances of a fatal accident in an automobile are about one percent. This single example says that we regularly expose ourselves to an "acute risk" of one in a million and probably without realizing it, a "chronic risk" of one in a hundred. Now let's look at some comparisons between similar risks. For example, 10 years ago, careful estimates were made of the possible effect of spray can propellants and refrigerator fluids on the stratospheric ozone. The outcome at that time was that the ozone would be lowered in the next few decades by at most 20 percent if chIorofluorocarbon use were continued. Since stratospheric ozone shields us from ultraviolet radiation, this depletion, if it were to occur, would cause a predictable increase in the number of nonfatal skin cancers in the U.S. population. To what could this new risk be compared? Is it large or small? Here we have a firm basis for an answer. The increased probability that an individual might end up with such a skin cancer is about the same as the increased probability of skin cancer associated with moving from San Francisco to the more sunny Los Angeles or from Baltimore to Miami. That comparison does not decicle the question of whether the chiorofluorocar- bons should or should not be used. That depends upon the benefits to be lost as well. But it gives a person with a nonscientific background a tangible measure of what is at stake. The Palo Alto water wells contaminated with trichIoroethylene provide a second example. These wells, used for drinking water, were found to contain up to 3 ppm of tnchloroethylene (TCE). At this concentration level, TCE is known to be a weak carcinogen through laboratory animal tests. Is there a comparison risk we can use to guide us? Tests show that the TCE contaminant is 1,000 times less hazardous than drinking an equal volume of cola, beer, or wine, each of which also contain weak carcinogens. Cola and beer, for example, both contain the carcinogen formaldehyde, cola at ~ ppm and beer at 0.7 ppm. These can be compared, in turn, to human blood, which also contains fo~-~naIdehyde at about 3 ppm from normal metabolism. These contrasts allow a lay person to compare, in everyday terms, just how big (or small) the TCE risk is in those water wells. What about pesticides? Man-made pesticide residues are found in our food at about 0.1 ppm, and most of them are classified as noncarcinogenic. Here, a useful comparison can be made to the presence of natural pesticides that are also present in our foods, but at 10,000 times higher concentrations than the man-made

OCR for page 203
THE RISKIBENEFIT EQUATION IN CHEMISTRY pesticides. These natural pesticides, present in every plant, are toxic chemicals that Nature evolved to protect the plant from fungus, insect, and animal predators. Some of these have been tested on rats or mice and found to be carcinogenic at sufficiently high dosages: estragole in basil, safrole in various herbs, psoralins in parsley and celery, hydrazines in mushrooms, and ally! isocyanate in mustard. That is useful and relevant evidence to be considered when we are considering the current hazard posed by agncultural pesticides. We see that comparable risks provide us with an easily understood measuring stick that can help us decide which potential hazards are serious enough to warrant corrective or restraining action. This kind of evidence can help us sort out those hazards small enough to be ignored so that we can concentrate our efforts (and resources) on the hazards that deserve and require attention. Who Is at Risk? Risk management is connected with getting general agreement on the balance between risks and benefits. But sometimes the people who see themselves at risk are different people than the ones who benefit. Thus, the society at large wishes to dispose of radioactive waste by storing it safely in some remote area. Unfortu- nately, it is necessary to transport it there on trucks that pass through many small towns on the way. The people who live in these towns are the ones at risk that a truck might overturn right downtown at the intersection of 1st and Main. They are likely to agree with Los Angeles that the wastes should be stored in some desert, but not be ready to endorse carrying that dangerous material through the town. As another example, lots of people in Logan, Utah, depend for their livelihood on their jobs in the smelters, and people all over benefit from the useful materials that come from these smelters. However, the smelters make their contribution to the air pollutants that generate acid rain a thousand miles away in the Northeast and in Canada. The neonIe in the regions suffenne from acid rain see the issue entirely ~ - r -a _ _ _ ~ . _ .. . .. .. ~ ~ ~ .' . _ 1 ._ _ ~ dluerently than those in the region whose economy Depends upon Ine vl~allly OI those industries. There is no easy answer to such dilemmas except to say that the interests of both groups must be considered in shaping public policy. Recognizing the nature of the problem helps, though, in see- ing why diametncally opposed positions might be persua- sively argued by reasonable and sincere people on each side. TABLE VI-2 Media Treatment of Three Chemical Spills Incident 10,000 gallons of toluene leaked into bay Newspaper Headline Chemical Scare Blocks Estuary Toluenethe ``T'' in TNT __. White solid substance Giant Traffic Jam spilled on roadway Scare Closes Bay Bndge MEDIA TREATMENTS OF ''Mystenous white Chemical Scare on Gate Bndge CHEMICAL SPILLS substance" found in roadway Table VI-2 lists three inci- dents that occurred during 1983 in the San Francisco Bay Area. In the first of these, both the television and the local newspapers opened their reports with the misleading remark: "Toluenethat's the T in TNT.'' What the press did not report 211

OCR for page 203
212 THE RISKIBENEFIT EQUATION IN CHEMISTRY was what all this toluene was doing in the area. Toluene is widely used as a solvent for many useful products such as paints and lacquers. Because of its wide use, large volumes of toluene are regularly transported over large distances. It has come into use as a substitute for benzene because toluene is safer: it is less flammable than benzene, it has a lower vapor pressure, and benzene is considered to be carcino- genic. These facts and the outcome- that no one was injured- add perspective to this undesirable accident, but they were not effectively communicated to the public. The second incident caused the closure of the Bay Bridge during the peak commute hours, trapping 20,000 autos and disrupting the plans of their 40,000 occupants heading to work, to the airport, to the hospital, or to visit the San Francisco Art Museum. The spired white substance proved to be lime, used in making concrete, handed daily by construction workers. The third incident in Table VI-2 was reason for closing the Golden Gate Bridge for 3 hours. This "chemical scare" was associated with a big bag of cornstarch that must have dropped off the back of a truck. These are two of the five closures of these bridges due to "chemical spins" since 1980. As in the three examples in Table VI-2, the other three closures were reported in the press and on television with emphasis on the worst-case scenano. These three chemicals were iron oxide (used as a pigment; it has the composition of rust), calcium phosphate (a fertilizer and component of detergents), and talcum powder. The talcum powder case closed the bndge for 10 hours! In none of the five bridge closures was there a newspaper foDow-up article reassuring the public that not one individual had been harmed and, indeed, that no one was ever In danger. What do we learn from these examples? The first lesson to be learned is that reporters rushed to the scene of an accident cannot be expected to be chemists. They will be reporting information received from other individuals, such as police, who are also unlikely to be chemists but whose responsibility is to act in the public interest in the face of the limited information they have. These latter officials can do nothing other than assume the worst possible situation. Consequently, media reports of chemical spills will usually overestimate the danger. We might hope, though, that the reporters would feel enough responsibility to avoid unjustifiable, fear-laden expressions (the " 'T' in TINTS. We might expect them to tell us what the chemicals are used for (once identified) and to let the public know, in retrospect, that this incident, at least, did not pose any real public hazard. As for the officials, they did what they had to do. In our social climate, they must act as though every white powder is as lethal as, say, sodium cyanide. What we can do to help them is to look ahead to avoidance of repeats. How might these spills have been handled after the first example? First, a plastic sheet held down over the spill would reduce wind-blowI1 distribution. Then, bridge personnel wearing fine-pore dust masks could sweep the solid material into a pile for later removal by a cleanup truck equipped with a vacuum cleaner. Turning from bndges to the larger context, it win be, more often than not, fire personnel and police who will deal with the immediate consequences of a chemical spin. Such personnel must (and many already do) have specific training in how to deal with chemical spills. There should be particular emphasis on the industnal feedstocks and chemical products in local use. A college chemistry course with a laboratory

OCR for page 203
THE RISKIBENEFIT EQUATION IN CHEMISTRY would be of immeasurable value and should be a normal criterion for advancement in a fire department. LARGE-SCALE USES OF CHEMICALS Any large-scale human activity carries with it a special consideration. While an unexpected and undesired outcome may have a quite low probability, the fact that very large numbers of people may be affected must influence our thinking. This special consideration is obviously applicable to nuclear war, reactor meltdowns, and genetic engineenng; it also is awakened by large-scale uses of particular chemical substances. We have already discussed in Chapter I! the possible global impact of the widespread use of chiorofluoromethanes as spray-can propellants and air-conditioner refngerants. The worldwide use of DDT provides a second, informative case history and will be discussed here. Large-scale industnal acci- dents fall in this category as well. Most chemical spills are well-handled and contained, but there have been, and will continue to be, rare but serious industnal accidents in which there is the possibility of catastrophe. While the potential dangers do not approach the centur~es-Iong and worldwide impact of a Chernobyl reactor meltdown, we are reminded by Bhopal and the recent chemical spills into the Rhine River that large-scale industrial operations pose real public risks. Two large-scale accidents occurred within the last 10 years in which large populations were put at risk. One of these took place at Seveso, Italy, in 1976, and the other at Bhopal, India, in 1984. These two catastrophic events deserve review. Seveso and Dioxin A Swiss-Italian chemical firm, Industrie Chemiche Meda Societa, Anonima (ICME-SA), manufactured the effective herbicide 2,3,5-trichIorophenoxyacetic acid (2,4,5-T) in large quantities at its plant in northern Italy, near the town of Seveso. The herbicide, used worldwide to increase food supply, is made from the simpler compound 2,4,5-trichIorophenol (TCP). In manufacturing TCP, an unde- sired impurity is formed in small quantity. The impurity, 2,3,7,8-tetrachIorodi- benzo-p-dioxin, popularly called "dioxin," has been discussed earlier in this section because of its extreme toxicity for certain small animals and the strong species dependence of its toxicity (see Table VIM. This large-scale accident at the ICME-SA plant began in July 1976, when cooling water was turned off to a chemical reactor making TOP. The temperature and pressure rose until a safety valve opened, releasing the reactor contents into the atmosphere over a densely populated area. The reactor was estimated to contain several pounds of the dioxin impurity. Among the toxic chemicals that have received notoriety during the last 20 years, dioxin may well be the one for which there exists the largest amount of and most systematic epidemiological data. Since 1949, there have been eight large industrial accidents, two of them involving U.S. companies. Table VI-3 shows that 823 workers were exposed, two thirds of whom suffered chIoracne, a very unpleasant skin lesion, as a result. Much smaller numbers suffered liver dysfunction, elevated 213

OCR for page 203
214 THE RISKIBENEFIT EQUATION lN CHEMISTRY TABLE VI-3 Industrial Accidents Causing Dioxin Exposurea Number of Number of Deaths Workers Chloracne Number Expected Number Date Exposed Location Cases without Exposure Observed 1949 250 Nitro, West Virginia 132 46.4 32 1953 75 Ludwigshaften, West 55 ~18 l7b.c Germany 1963 106 Amsterdam, The 44 13 8c Netherlands 1964 61 Midland, Michigan 49 7.8 4 1965-1969 78 Prague, Czechoslovakia 78 ? 5a,b,~ 1966 7 Grenoble, Prance 21 ? ? 1968 90 Derbyshire, United 79 ? 1 Kingdom 1976 156 Seveso, Italy 134 ? Normalab'~ Total 823 592 aAll at plants manufacturing TCP; dioxin was an undesired impunty. bImpaired mobility; fatigue, neurological symptoms. CLiver damage. Elevated blood cholesterol levels. lipid and cholesterol levels in the blood, and neurological damage. All of these conditions gradually recover. Astonishingly, for the 492 workers exposed before 1964 (22 years ago), the number of individuals who have died (61) is 30 percent lower than expected from normal causes. No one concludes, of course, that dioxin increases longevity, but it is difficult to conclude that it has a lethal effect on humans. To summarize the outcome of the widely reported Seveso accident, hundreds of townspeople and ICME-SA workers were evacuated, many of these having received severe exposure to the chemicals released. An estimated 37,000 people received minor exposure. The health of 500 people who received the largest exposures is being carefully monitored. To date, there have been no known deaths, involuntary abortions, or birth defects attributable to the exposure. Many small animals and rodents were killed in the town of Seveso. The disposal of 41 drums of toxic waste from the cleanup was in the European news for the next few weeks. A French waste disposal company contracted to move the drums from Italy to an authonzed waste disposal site in West Germany. Then began a strange odyssey for the 41 drums occasioned by the fact that everyone wanted to get rid of the wastes but no one wanted to have them pass through their town, let alone be stored nearby. Dioxin has received much attention, both in the courts and in the press. Certainly, a factor has been the extreme and well-documented toxicity for guinea pigs and mice, coupled with the fact that there are some well-established, though most often temporary, human ailments caused by severe exposure. It does help to show that the '~chronic risk" to any individual, even those living near TCP chemical plants, is negligible compared with the chronic risks involved in driving a car, smoking cigarettes, eating peanut butter sandwiches, or drinking beer or wine. However, these are familiar and voluntary risks. As we have said before, the public is extraordinarily sensitive to any risks to which it is exposed involuntanly.

OCR for page 203
THE RISKIBENEPI T EQUA T10N IN CHEMISTR Y Bhopal and Methyl Isocyanate Near midnight on Sunday, December 2, 1984, the impoverished residents of the squatter community called J.P. Nagar slept, unaware of the tragedy about to strike them. They occupied huts and hovels in a crowded shantytown built in the safety zone surrounding the Union Carbide plant just outside of Bhopal. Bhopal is a city of 800,000 and the capital of the agricultural State of Madhya Pradesh, the largest State in India. The plant was sited near Bhopal to manufacture pesticides, an essential element of the "green revolution" in a country trying to come to grips with its most cntical national problems, starvation and maInour- ishment. This plant had three large, underground storage tanks containing the volatile and toxic liquid methyl isocyanate, which is an immediate precursor to several effective herbicides. Late in the evening of December 2, the weekend plant personnel found that the pressure in one of these tanks, number 610, was abnormally low. Then, the temperature and pressure in 610 began to nse, this dangerous development being accentuated by the fact that the protective refrigeration unit may have been switched off. The plant personnel panicked as the temperature began to rise precipitously. The vapor pressure of the volatile liquid soared upward until it ruptured first a safety disk and then a relief valve designed to alleviate such an emergency. However, the vent line to the flare tower, where such releases are burned to harmless products, was closed off for repairs. The torrent of gas passed into, and overwhelmed, chemical scrubbers intended to neutralize any methyl isocyanate release not handled by the (inoperative) flare tower. Pressurized sprinklers designed to form a "water curtain" over such a release did not function because the water pressure was too low. In an accelerating calamity, tank 610 discharged 41 tons of lung-searing methyl isocyanate gas on the people of I.P. Nagar. The wind carried the lethal cloud south toward the Railway Station, which had its own shanty community. Before that terrible night ended, about 14,000 of the 800,000 inhabitants of Bhopal had been seriously exposed. Perhaps 1,500 men, women, and children died within the first few hours. Without question, the world had seen the worst mass exposure to toxic chemicals since the deliberate chemical warfare of World War I. The repercussions of this tragic event are still with us, as both societies and chemical industries around the world work to ensure that it will not happen again. The Chemistry Behind Bhopal Methyl isocyanate (MIC) is a volatile, reactive, toxic, and flammable liquid. It boils at 39C, and its vapor pressure is almost half an atmosphere at 20C. It is shipped only in stainless steel or glass-lined containers under a slight ove~pressure of dry nitrogen to prevent the entry of atmospheric moisture. In bulk storage, it should be cooled, preferably to 0C. It is toxic to rats with an LD50 of 21 ppm for 2 hours of exposure and 5 ppm for 4 hours of exposure. In 1965 (when it was still permitted), four human volunteers in West Germany were exposed to low levels of MIC. At 0.4 ppm, none of the subjects detected it, but at 2 ppm, there was nasal irritation and their eyes watered. 215

OCR for page 203
216 METHYL ISOCYANATE tWHAT IT S FOR | N= C~ ~ + ROH - CH3 WHAT IT S NOT FOR| THE RISKIBENEFIT EQUATION IN CHEMISTRY At 21 ppm, the irritation became extreme and the test was terminated. There were no lasting aftereffects. When water comes into contact with MIC, it reacts rapidly to form methyl amine and carbon dioxide. The reaction releases heat, so if there is no cooling, the temperature rises and the reaction speeds up. As the temperature rises (or in the presence of catalysts such as Fe, Cu. Sn, or Zn), MIC reacts with itself to form a tnmer, again liberating heat so that the temperature rises still more and reactions accelerate. These un- desired reactions have the po- tential for a thermal "run- awav.'' so safe handling of MIC N_C O ~ H20 CH3 ~2 N= CO CH3 o 11 ROCNHCH3 CARBONATE PESTICIDES + CO2 aH =-231ccal o 11 H3C_ SCOW ACHE I I An=-54lccal in ~ requires careful temperature control, avoidance of moisture, and meticulously clean contain- ers (to avoid catalysts). With all of these hazardous ~B=-54]ccal properties, why does anyone O^' ANN to use MIC, let alone store 40 H TRIMETHTt tons in a single storage tank? 3 I SOCIANURATE The value of MIC is that it readily reacts with alcohols to form carbamates, which are extremely effective pesticides. It is used by Union Carbide to make the pesticide Levine (1-naphthyI-N-methy! carbamate), by Shell to make Nun (methonyI), by DuPont to make Lennate@, and by FMC to make Furudon~ (carbofuran). Union Carbide has its biggest MIC production plant at Institute, West Virginia (10 times bigger than the Bhopal plant). Because of the importance of pesticides in raising the food supply for the 700 million population of India, Union Carbide established near Bhopal the most advanced Research and Development Center for pesticides in all of Asia. In addition, Union Carbide built the plant at Bhopal so that pesticides for use in India could be manufactured in India by Indian personnel. ~ ~ , w. ~3 T ROCYANURATE The Victims One year after the Bhopal accident, the official government cleath toll was placed at about 1,800, although it is possible that this estimate is too low by as much as 500 to 1,000. Among the most exposed victims, lung damage was persisting, eye damage was generally recovering, and liver dysfunctions were prevalent (due, in part, to complications caused by the drugs administered). In the exposed population, with a normal death rate of 250 per month, deaths were being registered at 265 per month. No birth defects had been associated with the exposure.

OCR for page 203
THE RISKIBENEFIT EQUATION IN CHEMISTRY lessons To Be [earned Governments in developing countries are generally enthusiastic, sometimes insistent, that manufacturing of essential products be conducted within the country's borders. Some countries require majority local ownership, local engineering and construction, and local operating and maintenance staff. These requirements can affect safety adversely because of cultural differences in work attitudes, understanding of industrial concepts, and response to training. Some of these factors may have contributed to the magnitude of the Bhopal disaster. Whether that is so or not, these are real problems that must be recognized and handled effectively while giving proper attention to national sensitivities. More generally, this catastrophe draws acute attention to the importance of safety in chemical operations. There should be an enforced safety zone around a chemical plant, and care should be exercised in the placement of chemical plants. There is now an increased awareness in communities near chemical plants, and in this country, many chemical industries have responded to this awareness with active communication programs that directly involve the local citizenry. These efforts are leading to improved readiness plans for various emergency situations that might develop. Perhaps the most important single lesson of Bhopal is, however, that particularly dangerous intermediate chemicals should not be stored in unnecessarily large quantities. Processes must be designed to manufacture these intermediates at the time of use and only in the quantity needed. This is a time-honored principle in any chemistry research laboratory, and it is even more important when the lives and health of many people are involved. The DDT Story It all began in 1939 when a Swiss chemist, Paul Muller, synthesized dichIoro- dipheny! trichIoroethane (DDT) during a systematic exploration for new insec- ticides. At the outset DDT appeared to be a miracle compound; it was extremely potent against a wide range of insect pests, and it did not have the acute human toxicity problems associated with the lead and arsenic compounds widely used at the time. The Benefits The United States first used DDT extensively in 1944 dunng World War I] to counter a growing typhus epidemic among troops and the civilian population in Italy. Typhus is cawed by body lice, and thousands of people were liberally dusted from head to toe with DDT to eliminate these pests. The epidemic was stopped, preventing a potentially devastating loss of human life. In the light of this massive success, DDT was put into service against the Anopheles mosquito, which spreads malaria across many parts of the globe. Before the use of DDT, malaria was responsible worldwide for 2 m~lion-3 million deaths per year and recurrent, periodic suffering for a much larger number. After a decade of use, malana has been removed as the primary scourge of human existence in several countnes. In India, the number of malana cases was lowered from 75 217

OCR for page 203
ago ~ -I E0~ ~ ~~y mUbonin1932lolOO,OOO in 1964. In 1be Soviet Union, 1be number of cases dropped Mom 33 mUbon in 1936 to 13~000 in 1966. In Sh Lanka over 1bb same time period, 1be meads toU went Mom 12~000 deaths per year 10 zero! The Wodd BeaRb O~ganizabon of the Onited Nabons teas credited this wonder cbemica1 Bob saving possibly 30 miDion Eves ham mama alone. For big accompliAbments, Dr. Paul MOHer was amazed the 1948 global Faze in Medicine.

OCR for page 203
THE RISKIBENEFlT EQUATION IN CHEMISTRY addition, some insects and pests became resistant to DDT after its prolonged use, and some beneficial insects were locally exterminated unintentionally. The Risk/Benefit Equation Here we see the evolution of a classic risk/benefit case. At the outset, it is clear that short-term benefits are great (in this case, the saving of human lives), and there are no known costs to be weighed against these benefits. But, despite the realization of the anticipated benefits, vigilant monitoring revealed environmental disturbances that were too pervasive to be ignored. Even though no human ailment has ever been connected with exposure to DDT, it is plain that some of its properties are incompatible with our desire to protect the world around us DDT's incredible stability, its mobility, and its amenity for living systems. However, at the same time that these special problems worked against its continued use, they defined the properties needed for a substitute. These now exist insecticides that are much more species specific, that are also nontoxic for human exposure, and that degrade in the environment after a few days or a few weeks. While DDT was saving millions of human lives, it was also guiding us to better solutions to the risk/benefit equation. CONCLUSION The most resounding message that emerges is that risk assessment is a difficult business. Paracelsus told us: "Everything is poisonous. The dose alone deter- mines the poison." Yet it is extremely difficult to determine the dose. Tests with humans are not permitted, and animal tests have questionable applicability to humans. Epidemiology shows association but not necessarily causality. There are strong subjective elements, too. One person's negligible risk is another's unacceptable hazard. Worse yet, often the group at risk is different from the group that benefits. Finally, everyone is sensitive about any risk to which the exposure is involuntary. Despite these difficult and sometimes perplexing aspects, nskJbenefit trade-offs have become a common element of countless decisions that affect us all. Some of these are decided for us by our elected officials in our state capitols and in Washington, D.C. Some of them we decide for ourselves in voting booths. Wherever these decisions are made, they should reflect both the common good and the common will. To make this possible, we need to improve scientific literacy throughout our population. Plainly, this must begin early in our schools; science education must receive more attention. In conclusion, we should be reminded that our quality of life and steadily increasing longevity are directly attributable to our technological advances in chemistry. An approach to chemical hazards based on unreasonable fear can deprive us of health-restoring drugs, essential sources of energy, increased food supplies, useful commodities, and industrial productivity. To avoid paralysis and the loss of these benefits, we need calm, wise, and rational decisions in deciding when and how much regulation is needed. We can achieve this by dealing wisely with the risk/benefit equation. 219

OCR for page 203
220 THE RISKIBENEFIT EQUATION IN CHEMISTRY SUPPLEMENTARY READING Chemical & Engineering News "BhopaI" by W. Lepkowski (C.&E.N. staff), vol. 63, pp. 18-32, Dec. 2, 1985. "Stringfellow Cleanup Mishaps Show Need to Alter Superfund Law" by L.R. Ember (C.& E.N. staff), vol. 63, pp. 11-21, May 27, 1985. "Bhopal, A C. and E.N. Special Issue" (C.& E.N. staff), vol. 63, pp. 1~63, Feb. 11, 1985. "Dioxin, A C. and E.N. Special Issue" (C.& E.N. staff), vol. 61, pp. 20-64, June 6, 1983. "Acid Pollutants: Hitchhikers Ride the Wind" by L.R. Ember (C.& E.N. staid, vol. 59, pp. 20-31, Sept. 14, 1981. "William Lowrance: Probing Societal Risks," Interview, W. Lowrance, vol. 59, pp. 13-20, July 6, 1981. Science "Risk Assessment and Comparisons: An In- troduction" by R. Wilson and E.A.C. Crouch, vol. 236, pp. 267-270, April 17, 1987. '`Ranking Possible Carcinogenic Hazards' by B.N. Ames, R. Magow, and L.S. Gold vol. 236, pp. 271-289, April 17, 1987. "Perception of Risk" by P. Slovic, vol. 236, pp. 280-285, April 17, 1987. "Risk Assessment in Environmental Policy- Making" by M. Russell and M. Gruber, vol. 236, pp. 28~290, April 17, 1987. "Health and Safety Risk Analysis: Informa- tion for Better Decisions" by L.B. Lave, vol. 236, pp. 291-295, April 17, 1987.

OCR for page 203
CHAPTER VII Career Opportunities and Education in Chemistry

OCR for page 203
Library into Space Unbelievable though it sounds, we may have to place whole libraries in a space-like environment over the next decade! This strange proposal is not made because our orbiting astronauts need more reading maters, but because if we don't do that or something' similar, most of our books won't be around very long for the rest of us to read. An Warming and little know problem faces mankind today the vast majority of books, those printed since the 1850s, are relentlessly yellowing and crumbling to dust. The library at the University of California at Berkeley alone stands to lose 60,000 books and periodicals per year to decomposition. This is not because of air pollution; the source of the destruction lies in the very paper on which the books are pnnted. Now, some clever chemists have discovered that, surpnsingly, a trip into an environment similar to space provides at least one solution to this vexing problem. Papennaking processes used since the 1850s universally employ an alumrosin sizing to keep irk from "feathenng" or spreading on the paper. Slowly, this papermaker's alu~aluminum suIfate~ombines with moisture in the pages and in the air to form sulfuric acid. This aggressive substance, in turn, facilitates attack on the cellulose fibers in the paper, breaking them into smaller arid smaller fragments and, ultimately, to dust. Between 75 and 95 percent of the deterioration in ``modern'' paper is caused by such acid attack. In recent years, chemists have developed a number of acid-neutralizing processes for books. One of these, developed in the Library of Congress research laboratory, suggests that the chemical diethyl zinc may be ideal for the job. Diethy} zinc is a gas, so its molecules can easily permeate even a closed book. Once inside, the substance deacidifies each book and then looks ahead to the future by leaving an alkaline residue of zinc oxy-carbonate. This residue, uniformly distributed throughout the paper fibers, protects the book Tom any future acid attack. Iron~caBy,- this life-sav~ng agent, diethy} zinc, bursts into flame on contact with air and explodes when it touches water. How does a chemist work with a compound that cannot be exposed to am or water? In a deep space environment, of course. A suitable location was fourth at NASA's Goddard Flight Center, where 5,000 books from the Library of Congress took a simulated flight, not on a rocket into space, but in a laboratory space-simulating vacuum chamber. First, the books were thoroughly dried by wing under vacuum for about 3 days. Then, with all oxygen removed from the chamber, gaseous diethyl zinc was introduced and allowed to defuse into the books. As the neutralizing ~ '7 Jo reaction proceeds, harmless ethane gas is produced and pumped away. ,~ (I >~~ Then the protective zinc oxy-carbonate is fanned. The results have _ \~ been extremely promising, and, as the technology Is perfected, Em, ~ libraries across the nation will be looking to instaI1 huge deacidifi- ` ~ ~ cation facilities. These countermeasures, coupled with the new ,,_ "alkaline reserve" papers now used in modern printing, promise that the precious heritage of the worId's libraries, including the A, ~ enormous Library of Congress, will be preserved for future genera- ~ tions to enjoy, and profit from, just as we do today. 222