6 Better Vulnerability through Chemistry

On july 18, 2001, baltimore was at risk of devastation when a sixty-car train owned by CSX caught fire in a tunnel in downtown Baltimore. Several of the cars had pulp material, which caught fire, burning other cars with dangerous chemicals, one of which was hydrochloric acid. The train derailed at three in the afternoon, and by nine that night all major highways into the city were closed to incoming traffic. Local telephone service was disrupted, a baseball game canceled, and parts of the backbone of the Internet were slowed. A water main broke, flooding and collapsing streets and knocking out electric power to 1,200 people. Even weeks later it was not over; a combustible chemical appeared in city sewers but fortunately was not ignited.

But as we always say, following a disaster, we are lucky it wasn’t worse. Had a railcar carrying ninety tons of chlorine (a routine event on the railroads) been one of the cars on that train, an estimated one and a half million people could have been at risk. As a FEMA report put it, “The derailment was unusual; the shipment was not. CSX reports that 40 freight trains run through Baltimore on an average day. Some days, all of them carry hazardous materials.” (Clarke 2005, p 117) Among the hazardous materials carried by rail are caskets of high-level nuclear waste from our power plants, and if the Yucca Mountain or some other facility to store the waste for centuries goes online, there will be daily shipments going through cities and tunnels. We have been reassured that the caskets could withstand extreme heats, up to 1,475 degrees for thirty minutes. But the fire in the Baltimore tunnel (which would handle shipments from Maryland’s Calvert Cliffs nuclear power station) exceeded that temperature for a full twenty-four hours. Experts estimated that twenty-four square miles around the accident (Baltimore, neighboring towns, and much more) would be highly contaminated or worse. “Amazingly,” say Clarke, whose account in Worst Cases we are drawing on, “officials in the Nuclear Regulatory Commission think the Baltimore experience was actually a success story.” (118) They said the robust nature of the casket was evident and the exposure in the tunnel fire would not release any radioactivity. The data Clarke reports flatly contradicts our Nuclear Regulatory Commission.

INTRODUCTION

Really devastating chemical-plant disasters have been rare in the last few decades. There is, of course, Bhopal, which we will come to, but the yearly deaths from individual chemical accidents in the United States is relatively small. The industry kills from 250 to 260 people a year, but this is hardly noticed since it is usually only two or three a time—and they are almost always workers—and most of these are transportation accidents rather than those that occur at industrial facilities. (Johnson 1999) (Think of the outrage if we had comparable fatalities from two 727 jetliner crashes each year, especially since those are victims with higher status than workers and can afford air travel.) Accidents are very frequent, averaging five a day, but so far the devastation has been remarkably small, given the potential. (Purvis and Bauler 2004) Hurricanes have rarely made direct strikes on chemical facilities. None of the chemical facilities accidents to date have resulted in catastrophic emissions, though evacuations are frequent. Even Bhopal, supposedly a wake-up call for the industry, did not prompt any industry-wide efforts of significance. But the rise of terrorism in the late 1990s prompted a great deal of concern about the security and safety of chemical plants and other concentrations of hazardous chemical materials. The chemical industry took notice and established safety standards, a government board increased its activity, public interest groups sounded alarms, and after 9/11, journalists wrote about walking up to unguarded tanks that could endanger seven million people with one terrorist suitcase bomb. The catastrophic potential of this vital industry is now on the map. (We will not consider the long-term health effects of other aspects of the chemical industry. We are only interested here in large catastrophes.)

The themes of this chapter are familiar:

(1) Organizations are always imperfect tools, and the best of them cannot provide complete security from the determined terrorist, avoid the devastations of weather, or avoid the inevitable mistakes that lead to accidents.

(2) The chemical industry, as with the nuclear power industry and others we will consider, will have a preponderance of firms that do not try very hard to prevent the occasional disaster. Prevention costs money in the short run, and in competitive capitalism, short-run advantages pay off, even if they do not in the long run.

(3) The institutional framework (Congress, regulatory agencies, trade associations, unions, etc.) is not conducive to security, safe operations, or protection from extreme weather. The self-regulation scheme of the chemical industry is ineffective, and our local and national government either is insufficiently aware of the catastrophic potentials, too beholden to the political campaign blandishments of the industry, or, as many organizations will always be, indifferent to the public welfare. All this would change if, instead of a dribble of the equivalent of two 727 crashes each year, we had two Bhopals or two horrendous terrorist attacks a year for, say, five years.

(4) There is an alternative that would not depend on having perfect organizations, and is not inconceivable even with our flawed institutional framework: reduce the size of the concentrations of hazardous materials, and reduce their toxicity and potential for fires and explosions. A few changes in this direction are being made; they could be extended to many of our 1,500 risky sites, at some economic costs, of course. With smaller targets and less-potent substances, terrorist acts would have less effect; the five-aday accidents would be even less consequential; and the weather would take a much smaller toll.

(A note on terminology. I will generally cite the danger as “toxicity,” but in strict terms there are four categories: toxicity, or the lethality of a chemical to human health; volatility, or the tendency to disperse the gas phase; flammability; and reactivity, or the tendency of a chemical to release a large quantity of energy or explode spontaneously. (Ashford et al. 1993, III-18) The term “hazmats” embraces the lethality of these four, and toxicity will be the general term.)

BHOPAL, THE MOTHER OF THEM ALL

The largest chemical-plant disaster in history, the release of poisonous gas on December 3, 1984, at the Union Carbide plant in Bhopal, India, had “prompt deaths” of at least 2,000, and perhaps as many as 6,000 to 8,000 prompt deaths, and left another 100,000 injured, many of them permanently. Following the explosion, a toxic cloud of methyl isocyanate drifted over the sleeping citizens in an adjacent shantytown, penetrating their porous homes. No alarms were sounded and when some phoned the plant, officials denied there had been a release, even as most workers were fleeing upwind. The cause was a commonplace, prosaic industrial accident of the kind that will continue to occur. Union Carbide still insists that it was sabotage, thus denying responsibility. Others say the immediate cause was an “operator error” in connecting some hoses, though the source of this explanation was the Arthur D. Little Company, which is a consultant to Union Carbide. A U.S. Chemical Safety and Hazard Investigation Board (CSHIB) report targets the company, rather than a hapless operator, saying that there was an explosion due to a nonfunctional critical part of the system and a high-temperature alarm that had been disconnected. (Kosal 2005; Kosal 2006) Whether an operator error or management error we should expect such errors; they are unavoidable. What was avoidable was the wretched condition of the plant that allowed the error to propagate. Supposedly, this was a wake-up call to the industry in the United States. But one doubts that the alarm rang loudly enough.

If there ever was a plant waiting for an accident to happen it was Bhopal in December 1984. Several of the major safety devices were inoperative and had been so for some months, waiting for repairs. There was inadequate instrumentation to detect risky conditions, as had been pointed out in the past by Union Carbide’s own inspection teams, and numerous gauges were not calibrated or were inoperative. To save money, the refrigerant had been removed from the key tank that exploded from overheating. The vent gas line, carrying methyl isocyanate (MIC) to an emergency scrubber, leaked most of the MIC to directly to the atmosphere instead. The gas that did get to the scrubber was not neutralized because of a lack of alkali in the scrubber. The vent gas scrubber could only handle five to eight tons of product, while the MIC tank’s capacity was seventy tons. Neither the temperature indicator on the MIC tank nor the flare tower for burning off the released MIC was functioning, and the water curtain (high-pressure water sprayers) for neutralizing MIC could reach a height of only ten meters whereas MIC leaked from the vent gas line to about thirty-three meters. The plant staff had been cut back because of business conditions and they were inadequately trained. Management was short of staff and itself lacked experience and training. The alarm system for warning the population was almost nonexistent, and there were no devices to detect the presence or direction of toxic clouds. (Everest 1985; Shrivastava 1987; Ashford et al. 1993, II-9)

Shortly after the Bhopal accident, Union Carbide performed an inspection of a somewhat similar Union Carbide plant in Institute, West Virginia. It found the conditions there to be exemplary and said that such an accident would be impossible. The Occupational Safety and Health Administration then conducted a thorough investigation and “saw” what they expected—a safe plant. However, Union Carbide’s own inspection had disclosed a few things that needed remedy, so it spent about $5 million, principally on a gas leak detection system that would check the prevailing winds and predict where the leak might travel and an increase in the height of a venting tower.

Just eight months after Bhopal, on August 11, 1985, a fairly similar accident occurred at the Institute plant. Aldicarb oxime, dangerous but not as dangerous as the methyl isocyanate at Bhopal, was transferred to a standby tank that was being pressed into service because of some other problems. Unfortunately the operators did not know that this tank had a heating blanket and that it was set to come on as soon as it received some product. Furthermore, a high-temperature alarm on the tank was out of service and a level indicator in the tank was broken, giving no warning that there was excessive temperature and excessive product. With warning systems failing to activate, and the product being mistakenly heated, the tank blew. The new venting tower was not high enough to perform adequately, and the gas warning system was not programmed to react to this particular gas, aldicarb oxime. A few other failures took place, not the least of which was that the number of gas masks in the only protected room, the control room, was set for the normal number of control personnel, but not for the many that flocked there after the explosion to seek protection. The operators survived by lying on the floor and passing the few gas masks back and forth among them in order to breathe, like hippies sharing a “joint.” Fortunately the weather and wind was favorable for a gas leak and not many people in the four neighboring communities that were affected were harmed, though about 135 were hospitalized. (Everest 1985; Ashford et al. 1993, II-10)

The Occupational Safety and Health Administration (OSHA) conducted a careful examination of the plant and, as we now can predict, found that this was an “accident waiting to happen.” It cited hundreds of “constant, willful violations” in the plant and levied a small fine of $1.4 million, which was later reduced. As mentioned above, the plant had been inspected by both Union Carbide and OSHA a few months before the accident as a result of the Bhopal “wake-up call” and had been given a clean bill of health. Inspections tell us little. If the plant has been running safely, the managers and agencies such as OSHA will find all is well. If an accident has occurred, the inspectors will find numerous causes for “an accident waiting to happen.”

But the Bhopal tragedy lives on. Twenty years later, the tiny $470 million civil suit fine that Union Carbide negotiated with a compliant Indian government still has not been disbursed; indeed, the fund has nearly doubled from accrued interest payments. (Baldauf 2004) Worse still, the Indian government’s inaction on ground water contamination is affecting thousands more by leaking poisons, such as heavy metals, nickel, chromium, mercury, and lead, along with other toxic materials such as dichlorobenzene, all of which were used at the Union Carbide plant. The contamination has spread two to three miles from the plant, into the nearby city, where wells are the source of drinking water.

The problem in Bhopal was not a third-world plant failing to measure up to first-world standards; the Institute, West Virginia, case contradicts that. The problem is the institutional structure that regulates the plant and oversees the disbursement of the fines. As flawed as the structure is in the United States, it continues to fail massively in India, which cannot seem to disburse funds to the victims or clean up the mess in Bhopal.

The Bhopal accident is extreme in all respects, but it could happen here. All the elements of that tragedy are replayed yearly on a less grand scale in the United States—the failure of a company to manage its plant safely, its evasion of responsibility, the slight wrist slap it receives from the government, and walking away from the contamination problem. A 2002 Rand report said “[M]any of the chemicals used or produced in plants throughout the U.S. have the potential to match or exceed the 1984 disaster in Bhopal, India. This risk is compounded by the frequent movement of these chemicals, typically by rail, through densely populated areas such as Baltimore and Washington.” (Karasik 2002) The threat of a chemical plant failure from a terrorist attack is second only to the threat of a biological weapons attack by terrorists, according to experts. Evidence of al Qaeda’s interest in chemical attacks is well-known— copies of U.S. chemical trade publications were found in an Osama bin Laden hideout in November 2001. (Grimaldi and Gugliotta 2001) As always, the source of the tragedy could be nature, industry, or terrorists.

HURRICANES AND CHEMICAL PLANTS

Before turning to the industrial accident potential and the terrorist potential, we should briefly examine nature’s threat—primarily hurricanes in the southeastern United States, where so many plants reside. The issue is explored in a paper by Marc Levitan. (Levitan 2000) Chemical plants abound in areas vulnerable to hurricanes, but direct hits are not common. Hurricane Andrew came close, but it turned aside from the concentration of plants on the lower Mississippi River and was a rapidly weakening category three storm at landfall, with a small eye and small radius of hurricane-force winds. It curved toward Baton Rouge, which has many chemical plants, but its winds of only seventy miles per hour caused only minor damage. However, the category four Hurricane Hugo in 1989 hit the island of St. Croix and produced significant oil spills from damaged petroleum storage tanks and fuel oil tanks; Houston-area plants incurred damage from Hurricane Alicia in 1983; and Katrina, as noted earlier, produced 575 hazardous petroleum and chemical spills. Chemical plants are not safe from hurricane-force winds.

Surprisingly, there are no uniform standards they must conform to. Most of the structures in chemical and petrochemical plants are not addressed in the standards set by the American Society of Chemical Engineers (ASCE). Worse yet, following decades of low hurricane activity, the number of hurricanes increased in the 1990s; the wind speeds were increasing; and the natural barrier of land that weakened the storms was inexorably receding due to development, so that the chemical plants were in effect marching closer to the Gulf of Mexico each year. Hurricane activity is fairly cyclic, with twenty- to thirty-year stretches of increased activity followed by similar lengths of decreased activity. The recent period of low activity has “led to what appears to be an unfounded expectation that significant damage due to high winds is very unlikely.” That, the author of the article says, is a great mistake. As we saw in chapter 2, two large oil companies failed to protect their storage tanks by filling them.

THE INDUSTRY: BIG AND DANGEROUS

It is not the gross size of the industry that is of concern; we need a lot of chemistry for a lot of better living. For our concerns, it is the inevitably capital-intensive nature of the industry that is important; its capital investment is more than twice the size of all manufacturing. This means huge facilities. Economies of scale abound in this business. A doubling of plant capacity increases the capital costs by only about 60 percent, so bigger is cheaper. (Ashford et al. 1993, III-9) Bigger facilities also mean bigger targets and bigger catastrophes. Accidents have been increasing. Ashford and others wrote in 1993: “A survey of the largest property losses over the past 30 years indicates an increase of 500 percent in the average size of the loss, holding prices constant. . . . A study by the Organization for Economic Cooperation and Development (OECD) notes that, in the Post World War II period, the incidence of major industrial accidents was only one every five years or so, until 1980. Since 1980, the incidence has risen to two major accidents per year.” (III-9) The industry claims that it is safer than manufacturing industries, but the workers most at risk are contract workers, hired for short-term jobs from companies that the government does not classify as being in the chemical industry but in various construction industry categories. Since 30 percent of the chemical industry’s workforce are contract workers, and they are the most likely to have accidents, the industry’s claim is without merit. (Perrow 1999, 361–62)

GOVERNMENT ACTION

The Chemical Safety and Hazard Investigation Board (CSB) was authorized by the Clean Air Act Amendments of 1990 to investigate chemical accidents and recommend steps to prevent them. However, it did not become operational until 1998. It has no authority to do anything but investigate, but at least that mandate is clear. The legislation that established it no doubt reassured Congress that the investigations would be carried out: “In no event shall the Board forego an investigation where an accidental release causes a fatality or serious injury among the general public, or had the potential to cause substantial property damage or a number of deaths or injuries among the general public.” But Congress and the Office of Management and Budget gave it only a $8 million budget in 2004—so small that it could only conduct six to eight full investigations, in a nation where 250 or more persons are killed yearly and hundreds more receive serious injury. (Rosenthal 2004) Nevertheless, its accident investigations provide evidence of the routine failures of plants. I will summarize just one small one here, quoting freely without attribution from their report, to give the flavor of a chain of failures at the plant, the firm that owns the plant, the local authorities, and the chemical industry. (Board 2002)

The DPC Enterprises facility, in Festus, Missouri, had twelve full-time employees. The facility repackages chlorine, transferring it from railroad tank cars into smaller containers for commercial, light-industrial, and municipal customers. On the morning of August, 14, 2002, six workers put the operation on standby and took a break. Twenty minutes later, three men who were smoking outside the building heard a loud popping sound and saw chlorine escaping from a tank car. They rushed to evacuate the area. Three men inside a break room heard a chlorine leak detection alarm and saw chlorine entering the building through an open door and quickly rushed outside. The operations manager pushed the emergency chlorine transfer shutoff button on his way out, but it failed to close the valves on the railcar. An automatic emergency shutdown system, triggered by chlorine sensors, also failed to close the valves. (We must expect such failures of emergency devices.)

The release continued for about three hours. Emergency responders, with protective clothing, were finally able to stop the leak by crossing through a four-foot-deep yellowish-green fog of chlorine, climbing on top of the tank car, and closing several manual shutoff valves. By that time, some 48,000 pounds of the gas (a small amount, given the tank size) had been released to the environment. While the slow leak continued, fire department personnel notified residents of the Blue Fountain mobile home park and another neighboring area to evacuate. Authorities ordered hundreds of residents, office workers, occupants of an assisted-living facility and a learning center, and students in a local school to remain “sheltered in place” for four hours, staying inside for protection from the toxic gas. Police halted traffic on nearby Interstate 55 for nearly one and one-half hours to keep vehicles from driving into the dangerous cloud. Although light westerly winds kept most of the chlorine away from residential areas, some of the gas likely drifted into the Blue Fountain mobile home park, the CSB said. The accident caused sixty-three people from the surrounding community to seek medical evaluations for respiratory distress; three were admitted to the hospital for overnight observation. In addition, three workers received minor skin exposure to chlorine during cleanup operations after the event. The chlorine also caused trees and other vegetation around the facility and in the plume path to turn brown until the following spring.

How could such a thing happen? The factory manufacturing the hoses “mislabeled” them, so DPC got ones with only a stainless steel braiding, rather one with a more expensive alloy that would resist chlorine corrosion. Both types of hoses look identical, though a simple non-destructive test would have told DPC that it was not getting what it ordered. Neither the fabricator nor DPC had the equipment to do the testing, known as “positive materials identification.” It is necessary where different materials look alike and where a mix-up can lead to a highly hazardous event. The mislabeled hose lasted fifty-nine days before bursting.

But there are safety devices that should prevent a runaway accident. The DPC Festus facility had chlorine monitors, safety alarms, and automatic shutoff valves in place that were supposed to stop the flow of chlorine in case of accidental discharge. But the CSB found that four of the five safety shutoff valves failed to close completely when the emergency occurred, due to corrosion and lack of maintenance. Although workers believed they were effectively testing the emergency shutoff system by activating it once each day, they were not required to verify that the valves were fully closed when the system was triggered, a rather obvious requirement.

As DPC employees rushed to evacuate, they notified emergency authorities about the release; but neither the company nor local agencies had an effective system for notifying neighbors. There were no community sirens and no system for making telephone alert calls to area homes, called “reverse 911” calls. Instead firefighters, who arrived about ten minutes after the leak began, had to go door-to-door through neighborhoods with bullhorns ordering residents to evacuate. The company’s own emergency training and drills were also inadequate, the CSB said. In addition, chlorine protective gear was stored in the chlorine-packaging building, too close to the tank car unloading station. Once the chlorine leak began, the gear became engulfed and inaccessible, leaving workers no alternative but to flee the plant. The CSB found that emergency responders from Jefferson County, Missouri, were also unprepared for the accident. It took about an hour and a half to assemble all of the county’s volunteer hazardous materials team members. The hazmat team determined that chlorine concentrations at the scene were greater than 1,000 parts per million—life-threatening to personnel without proper respiratory equipment. It then took another forty-five minutes to plan entry to the site and don protective suits. By the time team members reached the railcar and were able to shut off the leak, a massive amount of chlorine had been released. (This was a slow leak; imagine if the whole tank burst from a small bomb or a derailment. Or imagine this leak occurring in a large facility, which had gotten the faulty shipment, perhaps creating collateral damage when workers were forced to abandon their stations for an hour or more. We can better “tolerate” accidents in small plants because there is less collateral damage.)

This was a tawdry series of completely ordinary, prosaic organizational failures. It could happen in a crowded metropolitan area under weather conditions that would spread the toxic gas much more widely, perhaps delaying the response of the hazmat team well beyond the two and a quarter hours it took to notify them and get them there and suited up. This was a small facility, so the damage was limited. But similar errors occur in large facilities with more destructive capability, as documented in the chapter on chemical plants in Normal Accidents. (Perrow 1999) What can be done? The CBS issued pious safety recommendations for the facility (it is all they can do), for its owners (the DX Distribution Group), the hose fabricator, the Jefferson County Emergency Management Agency, and the Chlorine Institute, a trade association (“develop an industry-wide system to allow positive identification of chlorine transfer hoses.”). (Board 2002) We may be sure it will happen again, and indeed, the DPC had another accident at its repackaging facility, this time in Glendale, Arizona, on November 17, 2003. The absorbent chemicals in a scrubber were exhausted, and fourteen people required treatment for chlorine exposure. (Board 2004)

We are told nothing about any fines in these cases, but we may be assured that they were trivial. Occupational Safety and Health Administration (OSHA) violations in the chemical industry averaged less than $750 in fines during the period from 1972 to 1979, and the average fell during the Reagan administration. The median OSHA fine following a death or serious injury accident in 1990 was reduced, when adjusted for inflation, to less than half of the 1970s median. Furthermore, although relatively large fines have been imposed recently in a few cases—such as the $5.6 million penalties OSHA has proposed for Phillips Petroleum in connection with its 1989 explosion in Pasadena, Texas—even then, the fines are only a minute fraction (typically well less than 1 percent) of the total damage caused by the chemical accident. (Ashford and Stone 1991) Perhaps fines should be scaled to the size of the facility such that the greater the risk the facility presents, the greater the fine for each injury or violation. This would be an incentive for reducing the size of the targets.

On the night of July 5, 1990, an explosion at the Atlantic Richfield Chemical Company’s (ARCO) Channelview production facility generated a fireball and a blast that was heard ten miles away. There were seventeen deaths and five injuries to plant workers. ARCO agreed to pay a $3.48 million fine relating to “willful” violations of federal safety law. A year later, Phillips Petroleum Company was fined $4 million for similar violations that contributed to a 1989 accident at another Houston-area chemical production facility in which twenty-three workers were killed and more than one hundred injured. (Swoboda 1991; Staff 1991b)

WORST CASES

As ineffectual as it may seem to be for the CSB to do no more than investigate and tell everyone to play safe, merely making its six to eight careful investigations a year will do some good. Much more good was done when, in 1999, the CSB conducted a census of chemical industry accidents, the first ever. It covered the period from 1987 to 1996. The CSB was uncertain that all the regulations imposed on the industry were having any positive impact on health and safety. Chemical accidents happen more frequently than most of us would ever imagine, they observed, and occur in every state—and not just in chemical plants but in other kinds of plants, including those that provide our water supply—and on railways, highways, and waterways. In ten years, there were 605,000 reported chemical incidents (many additional accidents are never reported), and almost one-third caused fatalities, injuries, evacuations, or property damage. There were more than 60,000 incidents per year and almost 2,600 deaths in the ten-year period. Unfortunately, the report was never published. It exists as “The 600k Report: Commercial Chemical Incidents in the United States, 1987– 1996, Special Congressional Summary,” and I have drawn on Lee Clarke’s summary for the data given above. (Clarke 2005, chap. 4) But another report did make an appearance for a time on the World Wide Web that is more relevant to our concerns than just the number of accidents. It asked: What are the hazard potentials of chemicals? As a result of the 1990 Clean Air Act, facilities with significant amounts of chemicals had to prepare worst-case scenarios, including projections of potential consequences, and report them to the Environmental Protection Agency (EPA), which consolidated them and issued the report. In 1999, Congress agreed with chemical companies to restrict the Internet availability of the EPA’s worst-case scenarios for individual plants on security grounds, but the scenarios remained on the Web for some time, and a remarkable Washington Post story summarizes some of the findings. (Grimaldi and Gugliotta 2001)

A single railcar of chlorine, if vaporized near Los Angeles, could poison four million people. Four million people could be harmed by the release of 400,000 pounds of hydrogen fluoride that a refinery near Philadelphia keeps on hand. A chlorine railcar release near Detroit would put three million at risk. Union Carbide’s Institute, West Virginia, plant near Charleston once again has 200,000 pounds of methyl isocyanate—the chemical that did so much damage in Bhopal—which would threaten 60,000 people. (Union Carbide had removed the methyl isocyanate storage from the Institute plant after the Bhopal accident, but the Institute plant was reported to be storing it again in 2001.) And close by New York City, a plant in New Jersey has 180,000 pounds of chlorine and sulfur dioxide, which could create a toxic cloud that would put twelve million people at risk. (Grimaldi and Gugliotta 2001) A terrorist attack on such facilities would not require the sophistication needed for an attack using a biochemical weapon.

The chemical industry spokespersons dispute the findings, noting that for the damage figures to be accurate, the plume of, say, chlorine gas, would have to spread evenly over the area, whereas it would possibly be a cone that covers only a part of the area. The EPA and the DHS have agreed with this reservation, but the Government Accountability Office, in addressing the issue, points out that a “worst case” was assumed to involve the release from only one source: the most deadly one, such as the biggest tank or the most heavily inventoried process. But the GAO points out that other parts of the site might be disabled by the failure of the one source, bringing much more inventory into play. An attack that breached multiple chemical vessels simultaneously could result in a larger release with potentially more severe consequences than those outlined in worst-case scenarios. (Stephenson 2005; GAO 2004)

It is worth visiting the Web page of D.C. Councilmember Kathy Patterson on hazmat transport regulation http://www.dccouncil.washington.dc.us/patterson/pages/prinfo/HazmatPhotos.htm. It shows the calculated plume from a chlorine release in the Washington DC area, which the U.S. Naval Research Lab estimates could kill 100,000 people in half an hour, and it is enormous. It also shows photos of chlorine tank railroad cars parked near the Capitol, traveling next to highways in the Capital area, with graffiti on them (indicating their accessibility); a Chlorine Institute estimate of 180,000 pounds of release in ten minutes traveling 14.8 miles downwind; liquefied propane tank cars on railroad bridge at Seventh Street, SW, in Washington DC; and a parked chemical tank car at Ninth Street, SW. It also quotes a railroad engineer saying a handheld grenade launcher, rifle, or bazooka could easily puncture a railcar of deadly gasses, killing thousands within minutes. And, of course, a commonplace railroad accident could do the same.

One can easily see how an accident, storm, or terrorist attack would release many hazmats from their containers through destructive explosions or fires. Or, take the vinyl chloride accident on April 23, 2004, where five workers were killed, four towns evacuated, several highways closed, a no-fly zone declared, and three hundred firefighters from twenty-seven surrounding communities battled the flames for three days. It took that long because the company, Formosa Plastics, ran the wells that provided the town’s water and the shock wave of the blast (a hundred-foot fireball) disabled the town’s water supply. Thus, the figures released by the EPA are likely to be underestimates, as frightening as they already are. (Steingraber 2005)

One last disaster, this one in Toulouse, France, occurred two weeks after 9/11. A massive explosion at a fertilizer plant killed thirty people and injured another 3,500. At least ten of the dead were people who lived near the plant; and 11,000 homes were destroyed and an additional 16,000 structures were damaged. The explosion registered 3.4 on the Richter scale. It is unwise to live close to a fertilizer plant. But the most interesting aspect of the accident is that the French judge who presided over the inquiry charged the company with deliberately encouraging investigators to find terrorists responsible in an effort to divert attention from negligent safety practices and insufficient storage facilities at the plant. The editor of the leading industry publication in the United States sagely noted “it will enormously help the chemical industry’s reputation if it becomes possible to prove that the blast was not an accident.” (Hunter 2001)

THE INDUSTRY RESPONSE

Of course the chemical industry is aware of these dangers. Accidents are very expensive and might provoke costly governmental regulations. The industry has taken some steps over the years to mitigate the risk. The principal organization representing chemical manufacturers was founded in 1972 and initially called the Chemical Manufacturers Association. It changed its name to the more friendly one of American Chemistry Council (ACC) in 2000, and has played an active role in promoting safety among the members, claiming that its members have spent $2 billion on safety in recent years. It represents approximately 140 companies that manufacture basic chemicals in some 2,000 facilities in the United States, accounting for more than 85 percent of basic chemical production in the country. In 1988 it borrowed a safety program that Canadian firms had developed, expanded it, and trademarked it as the Responsible Care Management program. The program has been praised by the U.S. Environmental Protection Agency. All its members have voluntarily agreed to abide by the plan, and citing the competitive advantage afforded to those that do not, the ACC would like to see legislation to require the remaining 15 percent of chemical manufacturing companies to abide by the Responsible Care Management program. (Jenkins 2005, 2, n. 3)

But critics of the program note some important flaws. According to the ACC document regarding verification of inspections performed on plants, the companies can select the verifiers, for example, firefighters and police in the local community, security consultants they hire, and local insurance auditors. This could mean that the company picks the “low-hanging fruit,” verifiers that may not be qualified and are most easily influenced by the local plant, which is generally the most powerful economic entity in the community. The GAO, in its report “Protection of Chemical and Water Infrastructure” (Jenkins 2005) raises this issue, as do critics from public interest organizations. The latter groups want a bill that has the federal government either do the verification or provide a list of approved verifiers according to national standards for verification. The president of the ACC disagrees, and was quoted shortly after the 9/11 attack as saying, “Additional regulations, stronger enforcement—that isn’t going to do the trick,” he said. “What you need is the industry stepping up on its own, preventing the worst from happening.” (Grimaldi and Gugliotta 2001)

“Stepping up on its own” did not impress the GAO. Not only was it concerned about the independence of the verifiers, worse still, the Responsible Care Management system certification “does not require third parties to verify that a vulnerability assessment was conducted appropriately or that actions taken by a facility adequately address security risks.” (Jenkins 2005, 18, 27 table 2, n. b) It is like taking a test and then grading it yourself. Companies volunteer to join the ACC and abide by its program, but can do an inappropriate assessment and make no changes. This would make the ACC claim that all their members follow the guidelines rather meaningless. Some evidence of this appears in a recent survey by the Paper, Allied-Industrial, Chemical and Energy Workers International Union (PACE). It found that, according to a journalist with Chemical Engineering News, fewer than 17 percent of the industrial chemical facilities have enacted “fundamental changes that would lower the impact of an accident or attack by making chemical processes inherently safer or by storing smaller amounts of hazardous materials on-site.” The article also reported: “What actions were taken were mainly in terms of guards and security (76%), and only 17% dealt with making processes inherently safer or reducing hazardous material storage.” (Johnson 2004)

The GAO asked ACC members for volunteers to be interviewed by the GAO staff for a study. Ten plants agreed. The results were mixed, even for this self-selected sample (presumably the most progressive): in only seven of the ten were they making process or inventory changes that would reduce hazardous chemicals on site, and only five of the ten had established system redundancies such as backup pumps, backup power systems, and storage capacity. (Jenkins 2005, 21) That seven of the ten were reducing storage is good news, but presumably these are the showcase plants; the PACE survey found nothing of the sort.

Some of the company’s problems deserve sympathy, and tell us something about interorganizational complexities. Three of the ten had trouble trying to install fences because the Army Corps of Engineers controlled the adjacent property. Part of the corps’ legal mandate is to protect the wetlands, so building a fence required a permit and the corps was not cooperating. (29) In two cases the chemical company complained that it had no control over the security practices of the railroads that served the facility, whose security levels were below those the plants; two other chemical plant officials complained that the contractors they hired for periodic major maintenance jobs did not have an appropriate level of background screening. (22) (They could hire and train their own workers, of course, but contract workers are much cheaper and nonunion.)

Nevertheless, the efforts of the ACC do not match those of the U.S. Coast Guard, which is responsible for chemical facilities in ports and harbors. The Coast Guard inspected 2,900 regulated facilities in the last six months of 2004, took 312 enforcement actions against owners or operators, and imposed “operational controls” over twenty-nine facilities, including suspending their operations. (Jenkins 2005, 8) Their actions included unscheduled spot checks as well as scheduled inspections, developing detailed checklists for the inspectors to use, reviewing the enforcement actions taken by inspectors, and training them. (30) Nothing comparable occurs in the voluntary program of the rest of the chemical industry. If the Coast Guard can do it in six months for 2,900 facilities, it should not be hard to fund another government agency such as the EPA or OSHA to do an equally careful job on 4,000 dangerous chemical plants and storage sites.

The effectiveness of the chemical industry’s voluntary Responsible Care Management program has been questioned by those examining the data on emissions by members of the program versus those organizations that have not joined it. The results are discouraging, and the analyses suggest a “gaming” of the system. Participation in the program means that the EPA is not likely to check on the emissions from the plants. In a survey of 3,606 facilities involving 1,500 firms over the period from 1987 to 1996, the researchers found that while emissions overall declined significantly, the firms that were not members of the program reduced their emissions considerably more than firms that were members. The report says: “Our data provide no evidence that Responsible Care has positively influenced the rate of improvement among its members. Indeed, we found evidence that members of Responsible Care are improving their relative environmental performance more slowly than nonmembers.” (King and Lenox 2000, 709) Belonging to the much acclaimed and EPA-praised voluntary program seems to have encouraged the members to make fewer improvements then they would have if they knew the EPA was watching. Similar findings appear with studies of other industries who have been allowed to engage in voluntary self-regulation, a practice that started in the 1990s and has steadily increased. (Delmas and Keller forthcoming; Delmas 2000; Harrison 1999; Welch, Mazur, and Bretschneider 2000) What goes for emissions is very likely to be the case with self-regulation to improve security.

In response to the 9/11 attack, Senator Jon Corzine (D-NJ) introduced a bill in 2002 requiring the federal government to set up security standards for chemical plants and to have plants not just enhance their security but reduce the amount of hazardous materials in storage and make efforts to substitute safer materials. The ACC strongly opposed it, favoring voluntary actions by the chemical manufacturing community, and the bill died in committee. Corzine attributes the defeat of the bill to a strong lobbying effort. (This may account for the unfortunate fact that since taking office as governor of New Jersey in 2006, he has made no move impose such standards on the facilities in his own state, even though the governor has to power to do so.) So did the EPA administrator of the time, Christine Todd Whitman. In her book, It’s My Party Too (2005), she charged that industry lobbyists worked with key Republican lawmakers to sabotage new security regulations for chemical plants after the 9/11 attacks. One newspaper account of her battles with the White House indicates that she and Tom Ridge, when he was heading up domestic security from the White House, before DHS was born, worked on a modest effort to require high-risk plants—especially the 123 factories where a toxic release could endanger at least one million people—to enhance security. But industry groups intervened, warning President Bush’s political adviser, Karl Rove, that giving new regulatory power to the Environmental Protection Agency would be a disaster. As noted earlier, Rove wrote in a reassuring letter to the president of BP Amoco Chemical Co.: “We have a similar set of concerns.” In her book, Whitman confirms the White House veto of the proposal but singles out two people from her own party for blame, Senator James Inhofe (R-OK) and Representative Billy Tauzin (R-LA) “Although both Tom and I agreed such legislation was necessary, strong congressional opposition, led by some Republicans on the Senate Environment and Public Works Committee and the House Energy and Commerce Committee, to giving EPA even this modest additional statutory authority made it difficult to secure administration support.” (Lane 2005) But the pressure has been increasing with strong editorials in the New York Times and other papers and now the ACC calls for federal legislation. (Staff 2005h)

A weak bill was still under consideration in March 2006, advocating uniform national standards. But compliance would not be checked by a government agency, or necessarily by a nonprofit professional association, but could be validated by a for-profit organizations. Still, DHS secretary Michael Chertoff seemed opposed to allowing some states such as New Jersey to set higher standards than the national ones, and he was opposed to forcing companies to switch to less dangerous chemicals even where this was feasible. This annoyed a spokesperson for the Greenpeace Toxics Campaign who pointed out that there were forty-five sewage treatment plants in urban areas, as well as power plants and refineries in a number of states, that used extremely dangerous chemicals like chlorine gas and hydrochloric acid while there were widely used alternatives that did not impose excessive costs. (Lipton 2006c)

THE MEDIA COMES ABOARD

The next major actor in the institutional framework of firms, public interest groups, Congress, and federal agencies such as the Chemical Safety Board, EPA, and GAO is the media. An enterprising reporter, Carl Prine, of the Pittsburgh Tribune-Review began probing security at chemical plants six months after 9/11. Chemical companies had been warned by the government that they were potential targets. Prine visited sixty plants all over the country and found that despite the government warning, security was extremely lax, even nonexistent. He walked onto plant grounds and up to tanks of dangerous chemicals and took pictures of them. Rarely was he stopped or questioned; more often, if anyone noticed him, it was to give him a cheerful wave. (Prine 2002) Of course, he was not dark-skinned and wearing a turban. People like that were being arrested in Washington DC and other parts of the country for taking photographs of river views and monuments, as a result of the Patriot Act. His investigation was fairly widely reported, but Prine doubted that much had changed after his stories ran, and about two years later he again went into the field. He teamed up with CBS’s 60 Minutes crew and investigated fifteen plants around Pittsburgh and Baltimore (CBS went on to four more states). (Prine 2003)

Prine’s findings are very depressing, but given this book’s view that for-profit organizations are weak vessels when the public good is concerned and that even when they try hard the failures will be many, we should not be surprised. Some plants had installed the usual fences, cameras, lighting, and additional security guards, but to no avail. Prine and his photographer went through unlocked gates, holes in the fencing, or down railroad tracks, past inattentive guards, and up to huge tanks and railroad cars filled with extremely dangerous hazmats, occasionally joking with workers.

The facilities included the warehouse of the grocery giant Giant Eagle, where the journalists went through a fence hole and up to a tank holding 20,000 pounds of anhydrous ammonia, a coolant for refrigeration, which could put nearly 43,000 people—including children in twenty-four schools—at risk of death, burns, or blindness, according to company filings with local emergency planners. At the mammoth Sony Technology Center in Westmoreland County, Pennsylvania, an unsecured gate, distracted guards, and unconcerned employees let them reach 200,000 pounds of chlorine gas. (Recall the damage done by the leak—not a full breach of the tank— of just 48,000 pounds that the Chemical Safety Board investigated.) No one stopped them as they touched train derailing levers, waved to security cameras, and photographed chlorine tankers and a nitric acid vat. If ruptured, one Sony railcar could spew gas thirteen miles, endangering 190,000 people. A water treatment plant spent more than $100,000 on electric gates, cameras, and identification badges, but the reporter went through a fence hole and an unlocked door that led to twenty tons of chlorine gas. The director of the plant was very concerned with the security breach; he noted that terrorists can hit natural gas or electricity supplies and people will survive, but that they can’t live without water. (True, but it is hard to live with twenty tons of released chlorine gas.) An industrial chemical distributor erected fences, instituted round-the-clock guards, installed cameras and even fortified its river dock. But federal safety laws would not allow the railroad to fence off the track where a chlorine tanker parks daily. The reporter and the CBS crew were able to make four undetected trips up the rails to ninety tons of chlorine gas. Railroad cars filled with toxic substances are so easy to get to that they are frequently covered with graffiti. (Kocieniewski 2006)

The journalists then went to the companies which, according to workers, put more effort into making sure toilet paper was not stolen than they did in protecting the public, and most company officials declined to comment. They then went to Tom Ridge, head of the DHS, who was concerned but also said that he was optimistic that long-term federal reforms will provide protection.

Senator Corzine, whose state of New Jersey has more than its share of chemical plants, joined the CBS camera outside a chemical facility that was a couple of hundred feet below a well-traveled highway overpass, near the New York border and Manhattan. The footage shows them walking up to an unguarded tank, its gate ajar. According to government records, nearly one thousand tons of deadly chlorine gas are stored here—the first agent ever used in chemical warfare during World War I. Chlorine gas does not burn or explode itself, but a full release following, say, an explosive charge on just one of the tanks might create conditions that could release the gas in others (e.g., abandonment of monitoring controls, enforced operating errors); stranger things have happened in this industry.

“This is one of the main thoroughfares for commuters who come in and out of New York City every day,” says the senator on the tape. “You know, it looks to me like you could drive a truck through some of these fences if you wanted to pretty quickly.” According to the plant’s worst-case estimate, the number of people at risk—those living within a seven-mile radius of the plant—was fourteen million. CBS then interviewed the head of the American Chemistry Council, telling him of their visit; curiously the official expressed reassurance. He said what they found was “totally unacceptable,” but that “it underscores the point that you and I have been discussing—that we work every day on being better at security. One security breach in one facility or several facilities is unacceptable to us.” It was an exception, he insisted, but the reporters had found dozens of such exceptions. (CBS News 2004)

The initial Pittsburgh Tribune-Review story appeared in 2002, the next Prine story and the CBS 60 Minutes report in 2004. In the spring of 2005, the tale continued. A New York Times reporter and a cameraman looked over the two miles of New Jersey that are said to be the most dangerous two miles in America, home to three major oil and natural gas pipelines, heavily traveled rail lines and more than a dozen chemical plants, a prime example of the concentration we are concerned with in this book. A congressional study in 2000 by a former Coast Guard commander deemed it the nation’s most enticing environment for terrorists, providing a convenient way to cripple the economy by disrupting major portions of the country’s rail lines, oil storage tanks and refineries, pipelines, air traffic, communications networks, and highway system. The reporters focused on a particularly deadly plant, a chemical plant that processes chlorine gas. It remained loosely guarded and accessible. Dozens of trucks and cars drove by on the highway within one hundred feet of the tanks. The reporter and photographer drove back and forth for five minutes, stopping to snap photos with a camera the size of a large sidearm, then left without being approached by plant employees. (Kocieniewski 2005)

In 2005, the federal Department of Homeland Security cut New Jersey’s financing to about $60 million from $99 million in the previous year. (Kocieniewski 2005) (Because of congressional politics, Montana, with almost no terrorist targets, gets as much federal aid for security as New Jersey, one of the top three most vulnerable states.) One would think that Michael Chertoff, the new head of the DHS would be sympathetic to the state’s situation because he is a native of Elizabeth. But when he visited New Jersey during a terror drill at this time (April 2005), he was noncommittal about restoring cuts. He told the reporter, “Frankly, it’s not a matter of spending a great lot of money, it’s a matter of taking resources we have and having a plan in place so we use them effectively.”

New Jersey officials see it differently. They say that the cuts will force them to reduce surveillance of possible targets, cancel training sessions for first responders and counterterrorism experts, and forestall the purchase of equipment to detect chemical, nuclear, or biological agents. They would have to scale back plans to fortify storage facilities and rail lines near the Pulaski Skyway, an area known as Chemical Alley. A DHS spokesperson was not worried; the DHS had visited more than half of the nation’s three hundred most dangerous plants and had “urged” the companies to enhance perimeter security and to switch to less hazardous chemicals and processes. In a remarkable non sequitor, the newspaper quotes her as saying that as a result of this effort, she believes North Jersey is “one of the safer areas because it has received the most attention in terms of protective measures.” (Kocieniewski 2005) Perhaps that is why federal funding was cut. Observers have noted that New Jersey is a “blue” state.

The legislative battle is illuminating. Senator Corzine introduced his Senate bill in 2002, and Senator Inhofe introduced a competing bill in May 2003, which had wide support from the chemical sector and the security industry. The Corzine bill emphasized using “inherently safer technologies.” The Inhofe bill did not; it focused strictly on physical security. As noted, the Corzine bill required companies to submit response plans to the government, which would review them; the other bill only said that the government could request them at times and for places that the government officials deemed appropriate. The first bill designated the EPA as the lead agency because of its experience with existing accident-prevention requirements and would do the implementing in coordination with the DHS. But the Inhofe bill limits any other agencies than the DHS from involvement; they can only providing “technical and analytic support” on request by DHS and specifically bars them from any “field work.” The Corzine bill would set standards “to eliminate or significantly lessen the potential consequences of an unauthorized release”; the other would set no standards, such that extending a guard’s hours would suffice to “reduce vulnerability.”(NRDC 2003) In July 2002 the Corzine bill was adopted unanimously by the Senate Environment and Public Works Committee (EPW). But more than thirty trade associations, led by the ACC, signed a letter opposing it, and within a month seven members of the EPW reversed their position, and the bill died. (Jarocki and Calvert 2004) The Inhofe bill also died in committee the next year. The ACC supported the Inhofe bill, and President Bush later excluded the EPA from the issue of plant security.

But all may not be lost. Encouraging news appeared in June 2005. Back in 2003, the administration favored legislation restricted to voluntary standards self-assessed by plant owners. But in 2005, in testimony to congress, Robert Stephan, a top deputy to DHS head Chertoff, said, “The existing patchwork of authorities does not permit us to regulate the industry effectively.... It has become clear that the entirely voluntary efforts of those companies alone will not sufficiently address security for the entire sector.” Federal standards may finally be forthcoming, but opposition was voiced at the June hearings. A representative of the American Petroleum Institute said, “Industry does not need to be prodded by government mandates. . . . Chemical security legislation would be counterproductive.” (Lipton 2005) It also seems likely that any legislation will focus primarily on security and hardly at all on reductions in volume and toxicity.

WILL INSTITUTIONAL PRESSURES CHANGE THE INDUSTRY?

It will take tough legislation to capture the attention of the chemical industry, which is so well situated with their massive campaign financing that it will be difficult to get Congress to act. However, it is not impossible, since Congress has acted in the past as a result of public pressure and especially that of environmental groups. Andrew Hoffman explores the development of environmental concerns by the petroleum and chemical industries from 1970 to 1993, arguing that they moved from considering environmental concerns as heresy to accepting them as dogma. Over just three decades, the chemical and petroleum industries “moved from a posture of vehement resistance to environmentalism to one of proactive environmental management.” (Hoffman 1997, 6) His enthusiasm for the changes in the industry is premature, I believe, since major instances of falsifying or hiding data continue to appear, and the industry even goes to great lengths to discredit critical scholarly works that expose their record, as we shall see. But his account of the pressures on the industry is revealing and deserves a summary. Public pressure can make a difference, so all is not lost.

In the 1960s, there was hardly any serious legislation concerning emissions, pesticides, or carcinogenetic substances. From 1960 to 1970, government regulation was minimal and environmentalists had little influence, so the chemical and petroleum industries were free to establish their own conception as to what safety and environmentalism meant. The oil industry flatly denied that emissions were harmful and that oil spills damaged the ecosystem, and said that no regulation was needed. Even the publication of Rachel Carson’s Silent Spring in 1962, probably the most influential environmental tract ever published, had no effect in the 1960s on the practices of chemical firms, which denied that pesticides did any harm.

But oil spills, air pollution, and the dangers of Agent Orange— which was used in the Vietnam War and affected our troops—kept emerging. The “sixties generation” was stirring things up, and protest movements appeared, bringing the first Earth Day celebration in 1970. Local and state regulations were being enacted with confusing and contradictory requirements, so the chemical industry called for a national policy on chemical pollution. The result was two federal acts: the Clean Air Act of 1970, targeting the oil industry, and more significant, an act establishing the EPA, signed by President Nixon in 1970. The EPA was at first welcomed by the chemical industry, to bring “order out of confusion.” But the only appreciable impact of the 1960s’ growing environmental concerns on the industries was that some firms formed their first environmental departments. These had little power in the organizations and focused strictly on legal requirements that the new federal laws were establishing. (Hoffman 1997, 51–55)

The formation of the EPA was a major restructuring for the federal government at the time, bringing in people from many other organizations that were concerned with air, water, pesticides, radiation, and solid waste. Nearly six thousand employees were moved to the new department, small in comparison with the Department of Homeland Security but large at the time. That it worked much better than similar reorganizations may have a lot to do with its strong-minded first administrator, William Ruckelshaus. During its first sixty days, the agency brought five times as many enforcement actions as the agencies it inherited had brought during any similar period. Hoffman looks askance at this “penchant for strong enforcement,” and says it was justified to establish agency credibility, rather than saying it was needed in its own right. Unfortunately, he says, it established an “adversarial, ‘command-and-control’ type of relationship between government and industry.” (65) Given industry’s denial that its pollution was at all harmful and its reluctance to change, there might have been no other alternative.

Industry responded to the outpouring of a string of tough laws by bringing its own lawsuits against the new agency, filing forty to fifty suits a year between 1976 and 1982 and charging that the government was the “biggest predator of them all.” (68, 75) President Reagan came to power on a antigovernment platform and in 1981 appointed a new director, Ann Burford Gorsuch, who proceeded to cut the agency’s budget and reverse its rulings. Industry was delighted, but the reversal was so extreme that Reagan was forced to fire her in 1983, her assistant resigned, and another key official in the EPA went to prison for lying to Congress. Industry was outraged but found it hard to sustain the outrage when, in 1984, the horrendous Bhopal disaster occurred.

More legislation, and more pollution, such as the Exxon Valdez oil spill (1989) followed. Environmental expenditures by corporations increased at a steady rate of $250 million per year from 1973 to 1993. Hoffman does note that this was not voluntary on industry’s part; it was because industry was “forced to react.” (6) By 1992, firms in both industries were forced to spend 10 percent of their capital budgets on environmental projects, but even so, the chemical industry remained far and away the number one polluter and the petroleum industry, number seven, according to the EPA. (10)

I have not dealt with the chemical companies’ problems with toxic consumer products, or the exposure of workers to dangerous substances. There are disasters here, of course, but not the catastrophic ones that concern us in this book. However, a number of recent exposes and lawsuits suggests that the rather benign neoinstitutional viewpoint expressed by Hoffman—that the public and the government will eventually compel sound environmental practices that large companies will not only accept but the companies will become “proactive” environmentalists—leaves something to be desired. One such case is worth reviewing, since twenty chemical companies, led by the ACC, broke new ground in defending themselves from the charge that they knew they were poisoning their employees and perhaps poisoning a vast number of the U.S. population. The aggrieved widow of a chemical worker brought a lawsuit that could have devastating consequences for the companies since, if successful, it would be trigger massive class-action suits. An important part of the evidence came from a book by two academics. The companies charged that the research used by the plaintiff was flawed and the two eminent university professors that conducted it had engaged in unethical behavior, so the evidence should be thrown out.

The evidence came from a warehouse full of company documents that a lone lawyer had collected in the process of filing an earlier suit. Apparently, the companies intended to drown this lawyer and his tiny firm in documents and discourage the suit. They succeeded and the lawyer gave up. But he offered the documents to the two historians, Gerald Markowitz and David Rosner, who published a book, Deceit and Denial: The Deadly Politics of Industrial Pollution (2002) based on the files. The book got rave reviews in the scientific press. Another suit was brought by the wife of a worker who died of an extremely rare disease caused by exposure to vinyl chloride monomer on the job. The chemical companies named in the suit charge that the research of Markowitz and Rosner is “not valid” despite being based on company documents that, for example, acknowledged they risked being charged with “an illegal conspiracy” by concealing the vinyl chloride–cancer link. The companies charged that the publisher’s review process, even though using an unprecedented eight reviews, was “subverted,” and that the two authors “frequently and flagrantly violated” the code of ethics of the American Historical Association. The companies subpoenaed all the records of the publishers (the University of California Press and the Milbank Memorial Fund), and even went so far as to subpoena five of the eight reviewers of the manuscript. They subjected Markowitz to five and a half days of grueling deposition conducted by fifteen different chemical companies. A key question: Had the reviewers checked all 1,200 footnotes? They had not checked all of them; it is not a requirement for reviewers. But attorneys for PBS and HBO, both of whom ran specials on cancer caused by chemicals in consumer products such as hair spray and vinyl food wrap did check them. (Wiener 2005) The case was settled in 2006, but Markowitz and Rosner were not involved in the settlement. Markowitz is set to testify in another vinyl chloride plant worker suit, but the chemical companies have moved to have his testimony excluded, again on the grounds that his book violated the ethics of the American Historical Association. (Personal communication)

The actions of the ACC and its members might not be quite as bad as denying that pollution and pesticides harm people, as they did forty years ago, but they seem not to have come very far in “embracing environmentalism,” as Hoffman claimed.

The advances that have been made, and there are many, have been in limited areas: obvious pollution, and exposure of workers and communities to persistent or occasional hazmats. Hopefully the next phase in Hoffman’s sequence of periods will be a concern with large-scale disasters that kill many people or contaminate large areas. The origins of the disasters could be, as we have relentlessly repeated, natural, industrial, or deliberate in source. The first two still remain under the radar, even for most activist groups. But 9/11 has brought the third, terrorism, into view. And as yet, it seems that the institutional field has not mobilized enough to move the chemical industry from “heresy to dogma” regarding the terrorist risks to which they are subjecting us.

CAN REDUCTIONS AND SUBSTITUTIONS BE MADE?

Are we being realistic in calling for reductions and substitutions? Is the size of plants and of chemical companies the basic problem? There is no doubt that chemical companies are among the biggest in the world, and many of those operating in the United States are foreign owned and certainly global in scope. The size of the companies operating in this country is already so enormous that increases in recent decades are beside the point. I could find no data indicating any increase in the size of storage facilities, per se, in the last fifty years or so, but again, the amount of hazmats stored at each site is already so large that no such data are required. However, such an increase seems quite likely to have occurred. A rather old study of the size of chemical plants found a startling fivefold increase in size in only twenty-five years, from 1957 to 1982. (Lieberman 1987) It seems likely that the larger the plant, the larger the number of storage vessels, and probably the larger the vessels themselves.

A study by Nicholas Ashford and associates in 1993 provides evidence of the storage/size problem. Analyzing accidents, Ashford found that storage releases were far more severe than releases from processing, valves and pipes, disposal, and other activities. For primary processors, storage releases were six times greater than the next category (valves and pipes); for secondary producers, twenty times larger. (Ashford 1993, III-15) Nevertheless, Ashford does not think that risks will be reduced by dispersing storage tanks because of the risk of transportation accidents. While favoring primary prevention by using less-hazardous materials and safer processes he does not see much future for inventory reductions. “Some reductions in inventory of hazardous substances, while heralded as primary prevention, may simply shift the locus of the risk and increase the probability of transport accidents.” (Ashford 1993, i, VII-16)

This is a widespread belief that should be challenged, since it is true only in some cases but not in most. First, reductions of inventory do not necessarily mean more transport of inventory. Inventory reduction can occur through closed-loop production, where the only hazardous materials are in the production loop rather than in storage, and more efficient processing. Second, the location of the risk is quite important. If sixty tons of hazardous materials are stored in six places rather than one, the probabilities of an accident in each of the six is much lower than the probabilities of an accident in the one concentrated storage. If there is an accident in one of the six, the damage will be only one-sixth of that of an accident in the concentrated one, and probably a great deal less because the collateral damage of one big tank accident on adjacent facilities will be a bigger event and probably in a bigger plant. Finally, the issue of transport accidents is not clear-cut. If each of six plants requires ten tons of the hazardous material to be in storage, and it is made on each site, there is no transportation hazard but there is a substantial deconcentration. If they buy it from a producer, the shipping hazards would probably be same whether the producer shipped it to one big plant or six small ones. (It could be even less in the latter case, since smaller containers would be more likely to be used.) Production economies may well be affected by dispersed storage, but these should be compared with the substantial reductions in the risk of natural, industrial, and terrorist dangers.

One might argue that the bigger the firm, or the plant, the more notice it will attract, and thus it will spend more resources on safety than will smaller plants. The limited evidence we have suggests the opposite. Paul Kleindorfer, of the University of Pennsylvania, found that the larger the plant, the greater the number of unwanted releases, even controlling for size of inventory. (Kleindorfer et al. 2003) Professor Don Grant, of the University of Arizona, found the same thing for chemical company plants, and for the size of chemical companies themselves. The bigger the plant, the more the pollution; the bigger the organization that owns the plants, the more the pollution. (Grant and Jones 2003; Grant, Jones, and Bergesen 2002) These studies dealt with accidental releases or just sloppy procedures, but they are related to catastrophic potentials. Whether it is that the size of the facility makes it more unmanageable and more prone to accidents, or that the power that comes with size makes it less concerned with damage to its reputation or community inhabitants, and thus it neglects safety, is not known. But big is not necessarily safer. A fire or explosion in one vessel is likely to spread to others in the plant, and a release of a toxic substance is likely to disable employees who are needed to keep other processes stable or, as we saw above, disable the water supply needed to fight the fires. Large sections of chemical plants are typically destroyed by collateral damage. Their very size is a hazard, and the size of the storage the greatest hazard.

We are not likely to either break up big firms or require them to have three or four dispersed plants instead of one concentrated one. But a reduction in storage volume and toxicity is certainly feasible, and have little or no economic impact. Here are a few examples, copied verbatim from the U.S. Public Interest Research Group Education Fund publication, “Irresponsible Care.” (Purvis and Bauler 2004, 10)

 

• A Massachusetts state law requires companies to disclose the chemicals used by their facilities, including the amounts on site, transported in products, released to the environment, and generated as waste. Companies also are required to produce toxics-use reduction plans. As a result, between 1990 and 1999, facilities reduced their use of toxic chemicals by 41%, while at the same time production increased by 52% and companies saved $15 million.

• Reductions in storage volume and toxicity may have played a role in the reduction of releases of carcinogenic chemicals by 41% between 1995 and 2000, as a result of the federal Toxic Release Inventory program. The program requires several industry sectors to report the toxic chemicals they release into our air, water, and onto our land.

• In Washington, DC, the Blue Plains Sewage Treatment Plant switched from volatile chlorine gas, which could have blanketed the nation’s capital in a toxic cloud, to sodium hypochlorite bleach, which has almost no potential for an off-site impact. In the wake of September 11th, 2001, the facility completed the switch in a matter of weeks. The expected cost to consumers will be 25 to 50 cents per customer per year.

• In Cheshire, Ohio, American Electric Power selected a urea-based pollution control system rather than one involving large-scale storage of ammonia that would have endangered the surrounding community.

• In Wichita, Kansas, the Wichita Water and Sewer Authority’s sewage treatment plant switched from using chlorine gas to ultra violet light in its disinfection processes. The plant expects to save money in the long run as a result of the change, as there is about a 20% anticipated cost savings in energy costs versus chemical costs.

• In New Jersey, more than 500 water treatment plants have switched away from or are below threshold volumes of chlorine gas as a result of the state’s Toxic Catastrophe Prevention Act.

• In 2003 in Wilmington, California, the Valero Refinery switched from hydrofluoric acid, which when released forms a toxic cloud that hovers over surrounding communities, to modified hydrofluoric acid, which is less hazardous. This change was largely due to decades of community pressure after a devastating accident at a near-by refinery in the area.

 

Hydrofluoric acid was involved in a 1987 release in Texas City that sent more than one thousand people to the hospital and forced three thousand residents to evacuate their homes for three days. (TexPIRG 2005) (That unfortunate city also had a horrendous series of explosions in 1947 that leveled large parts of the city and killed thousands.) Refineries could switch to sulfuric acid, which has a lesser off-site threat, or modified hydrofluoric acid, which is less toxic and which the Wilmington, California, refinery switched to after an accident and community pressure.

Nicholas Ashford and his research team, writing in 1993, report on other instances of substitutions that are encouraging. Monsanto cut its total volume of highly toxic gases in storage facilities throughout the world in half by shifting to just-in-time deliveries of raw materials. Rohm and Haas replaced its batch processing system to continuous processing, cutting the inventory of toxic materials from 3,000 gallons to 50 gallons. Hoffman-LaRoche went from 15,000 gallons of liquid ammonia to 2,000 gallons. In a decoupling example, Hoffman-LaRoche hired an outside contractor to make large-volume phosgene-based materials, thus eliminating the use of the extremely hazardous substance in a large plant where the collateral damage would be extensive. DuPont found a way to avoid keeping 40,000 to 50,000 pounds of deadly MIC, purchased from Union Carbide, in storage. It will produce MIC and use it immediately in a closed-loop operation, with a maximum of two pounds on premise at any one time. (Ashford et al. 1993, II-18)

As noted, chemical plants are expanding their presence in the United States. One thing we should do is to pay close attention to the location of new facilities. Adding them onto existing plants increases the risk of “collateral damage,” putting both the new and old facilities at risk. New facilities should also be located far from rivers and harbors. The recent explosion in China makes the point. This occurred on November 2005 in the large city of Jilin, China. The plant, owned by CNPC, China’s largest oil producer, is located right next to the Songhua River and just upstream from the city of Harbin, a city of 3.8 million people (roughly the size of Los Angeles). Chemical plants are frequently located next to such vectors, causing their accidents to have much farther-reaching consequences than is necessary, even to the point of being international in scope, as the Jilin accident demonstrates. The November 13 explosion released one hundred tons of benzene, nitrobenzene, and other toxins into the river, which is a primary source of drinking water for Harbin’s citizens. Because benzene can cause leukemia and ulcers of the mouth, Chinese officials were forced to shutdown Harbin’s drinking water supply for five days until the toxins passed downstream. However, they did not do this until eleven days after the spill, and even then said it was for “pipe maintenance.” Unfortunately, the toxins did not dissipate rapidly and in December reached the Amur River in Russia, of which the Songhua is a tributary. Once the toxins reached the Russian town of Khabarovsk, officials there were forced to shut off drinking water to the town’s 580,000 residents. Additionally, officials imposed a ban on fishing on the Amur that could last as long as two years. Thus, an accident that caused only one immediate death ultimately affected millions of people due to the plant’s location next to a river. (Spaeth 2005)

CONCLUSIONS

Facilities that contain large amounts of hazardous materials, such as chemical plants, water treatment plants, and railroad tank cars constitute the best targets for terrorists. Nuclear power plants could create greater destruction, but it is not as easy to disable a spent storage pool as it would be to apply explosives to one of the hundreds of large tanks along Chemical Alley in New Jersey, setting off other explosions and fires just a few miles from Manhattan. It could be a drive-by shooting, and all the terrorists would have to do is to then escape upwind. Of course, radiological weapons and biological weapons can kill more people than the one-toseven million that might be killed in a particularly successful attack upon a chemical plant or railroad train, but those attacks are much more difficult to stage. (Kosal 2006)

The chemical industry has made some progress in reducing concentrations of hazardous materials and in using less-toxic substances in its operations. But the instances of these are in the dozens rather than the thousands that we need. The industry has resisted mandatory inspections and standards and members of Congress, dependent on the industry for campaign financing funds, have allowed this resistance to be successful. Local communities all too often are unprepared for even small accidents and overwhelmed by substantial ones. There is no indication that DHS funds have improved our ability to respond. If anything, they have diverted funds from first responders to those concerned with terrorism. The industry should do more with fences and guards and surveillance cameras, but it is an almost hopeless task in the face of dedicated terrorists. As is evident from the oil disasters in the Katrina hurricane, some in the industry do not even follow obvious safe practices. This is a wealthy, highly profitable industry. We should expect more from it, but can’t reasonably do so until we hear from Congress. Despite congressional ability to oversee the DHS, it has failed to give it “authority to require chemical facilities to assess their vulnerabilities and implement security measures,” something the GAO says it requires. A January 2006 GAO report notes: “DHS has stated that its existing authorities do not permit it to effectively regulate the chemical industry, and that the Congress should enact federal requirements for chemical facilities.” (GAO 2006)