The chemical industry consists of more than 12,000 manufacturing plants
in the U.S., producing a total of more than 66,000 different commonly-
During the past 30 years, individual chemical plants have grown larger-
-often increasing 10-fold in size--to take advantage of economies of
scale. A chemical plant today will typically produce 300,000 to 600,000
metric tonnes (660 million to 1.3 billion pounds) of product per year.
Storage tanks at a plant may hold as much as 1.5 million barrels (75
million gallons) of product or raw material.
Chemical plants are complex, tightly-coupled operations. Tightly-
coupled means one event or process affects another event or process
directly and quickly, thus making human intervention difficult when
something goes wrong. Complex means that events cannot be predicted
reliably because many different things can go wrong in many different
ways. This is why every chemical plant in the world is an accident
waiting to happen.
During the last 30 years, while plants have grown larger, many
corporate managers have adopted policies that increase the chance of an
** The spacing of process equipment within plants has been reduced, to
save money on energy, piping, and instrumentation.
** During the 1980s, faced with the worst recession in 50 years, the
chemical industry laid off thousands of workers, thus reducing the
numbers of knowledgeable people paying attention to various processes.
** In order to diminish the power of labor unions, petrochemical
corporations in the 1980s began replacing regular employees with
contract laborers, temporary workers who move from site to site, rarely
gaining in-depth experience. According to Nicholas Ashford at MIT
[Massachusetts Institute of Technology], "well over 30%" of the hours
worked in the petrochemical industry are now logged by contract
employees. Contract employees lack the process-specific experience of
many of the workers they have replaced.
** Meanwhile in recent years, chemical plants have been run, in some
cases, in excess of 100% of their "official" capacities.[1,pg.III-8]
** Finally, in some instances, it is obvious even to an outsider that
corporate owners are letting their plants decay. You need only to drive
along the Niagara River in Niagara Falls, N.Y., to see fully-operating
plants that are not being maintained, or repaired properly, or even
painted, as they rust, decay and deteriorate with time.
As you might expect, there is evidence that the frequency of chemical
accidents has increased in recent years. According to Nicholas Ashford,
a survey of the largest property losses over the past 30 years
indicates a five-fold increase in the average loss (calculated in
constant dollars, taking inflation into account). 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 5 years or so, until 1980.
Since 1980, the incidence has risen to two major accidents per year.
Thus, according to the OECD, the frequency of major chemical accidents
has increased 10-fold in recent years.[1,pg.III-9]
However, it is not major accidents that affect the largest number of
people. It is the smaller, more routine accidents that expose workers,
emergency response personnel, the general public, wildlife, soil and
water to toxic chemicals.
When a reporter sets out to examine routine chemical accidents, he or
she quickly learns that there are no reliable data. Three databases are
nationwide in scope and contain information on toxic chemical
accidents, but each database is flawed in some major way. The federal
Bureau of Labor Statistics (BLS), for example, maintains records of
worker accidents--but they omit injuries to contract workers who, as we
have seen, represent at least 30% of the total workforce in the
chemical industry. (Injuries to petrochemical contract workers are
reported under various categories in the construction industries.)
Since contract workers are more accident-prone than regular employees
who have learned the ropes during years spent at a plant, the BLS
database substantially under-reports workplace injuries in
The other databases suffer from flaws that are similar or worse.
The National Environmental Law Center (NELC) examined one of the
national accident databases for the period 1988-1992. They studied the
Emergency Response and Notification System [ERNS] database, maintained
by EPA [U.S. Environmental Protection Agency].
The NELC staff did not summarize the entire ERNS database; for example,
they only gathered data on about 800 out of 66,000 chemicals commonly
in use. Furthermore, they omitted data on accidents involving only
petroleum products (fuel oil, gasoline and so forth)--which means they
omitted 52% of the ERNS data.
Still, what they found is impressively large. NELC reports that during
the 5-year period, 34,500 accidents involving toxic chemicals were
reported to ERNS; this means that, on average, a toxic chemical
accident was reported 19 times per day in the U.S., or nearly once
To get an idea of how badly the ERNS database may underestimate the
chemical accident picture, its data can be compared to a more thorough
study conducted by the Attorney General's office in New York state.
The New York study examined records from EPA, the New York Department
of Conservation, the U.S. Coast Guard, the New York Office of Fire
Prevention and Control, and a private newsletter.
The New York study found that a total of 3496 toxic chemical accidents
occurred in New York state during the 3-year period, 1988-1990. During
this same 3-year period, only 496 New York accidents were reported to
the ERNS database. If this kind of under-reporting is typical of the
ERNS database, then the true picture of chemical accidents in the U.S.
may approach 240,000 chemical accidents per year, or 130 per day, or 5
per hour around the clock.
Despite astonishingly poor data, it seems safe to say that chemical
accidents are a large, and probably growing, problem. Nicholas Ashford
at MIT has described a fundamental solution, which he calls primary
prevention. Primary prevention means designing chemical processes to
eliminate the possibility of an accident; secondary prevention reduces
the probability of an accident.[1,pg.III-1]
The bulk of Ashford's report (which was funded by EPA) describes how
industry might change from its present emphasis on secondary prevention
(add-on safety systems, such as sprinklers, leak detectors and double-
walled tanks) to an emphasis on primary prevention (designing plants so
they rely less on toxic, volatile, flammable, or reactive chemicals,
and so they operate at reduced temperatures and pressures).
Ashford finds that technology-forcing regulations can help, i.e.,
regulations that require industry to do things that they say they
cannot do today. Ashford finds that the chemical industry has a decent
record for innovating when the law says it must.
Toxic tort (injury) lawsuits can be another beneficial force, prompting
industry to innovate, Ashford says.
Government enforcement of worker safety laws is presently NOT an
effective inducement for industry to change its ways, Ashford finds.
Regulatory fines for OSHA [Occupational Safety and Health Act]
violations averaged only $75 per violation during the period 1970-1980
and then dropped even lower during the Reagan years. The median OSHA
fine following a death or serious injury accident in 1990 was only $890
(less than half what it had been in 1972, after the figures are
adjusted for inflation).[1,pg.VI-1] Thus, at present, it appears to be
cheaper to kill and injure people than to make chemical plants
Some argue that corporations, as presently structured, simply cannot do
the right thing. For example, when Union Carbide killed somewhere
between 3000 and 8000 citizens of Bhopal, India in 1984 and injured
somewhere between 200,000 and 600,000 others, they aggressively
resisted demands for compensation. They ultimately paid $520 million
total ($470 million to the victims, plus $50 million to Carbide's
lawyers). Just the economic damage from the event (leaving aside the
question of fair compensation to the victims) was estimated to be $4.1
billion, so Carbide got off incredibly lightly. Carbide said it
accepted "moral responsibility" for the catastrophe at Bhopal, yet it
refused to pay the real dollar costs.
Given the legal structure of corporations, Carbide had no choice, says
Ward Morehouse: "Had they been genuinely forthcoming and made truly
disinterested offers of help on a scale appropriate to the magnitude of
the disaster, they would almost certainly have been confronted with
suits by shareholders seeking to hold the management accountable for
mishandling company funds...."[4,pg.490]
Furthermore, the stock market sends perverse signals to corporations
that create disasters. The day the $470 million penalty against Carbide
was announced, its stock rose $2.00 per share.[4,pg.487] Last week when
a jury awarded $5 billion to victims of the Exxon Valdez oil spill,
Exxon's stock rose $1.50 per share.
Whatever the solution to these problems turns out to be, the nation
will need more reliable data on chemical accidents. The Clean Air Act
of 1990 created a Chemical Safety and Hazard Identification Board,
charged with "investigating the causes of chemical accidents, issuing
reports to Congress and other federal and state agencies, and proposing
ways to reduce the risk of injuries arising from the production and use
Unfortunately, at present the Board has no effective budget, so the
very substantial benefits that it could provide may be lost unless
Congress takes action soon to give the Board the resources it requires.
 Nicholas A. Ashford and others, THE ENCOURAGEMENT OF TECHNOLOGICAL
CHANGE FOR PREVENTING CHEMICAL ACCIDENTS: MOVING FIRMS FROM SECONDARY
PREVENTION AND MITIGATION TO PRIMARY PREVENTION (Cambridge, Mass.:
Massachusetts Institute of Technology, 1993). This insightful and even-
handed report is available free from the federal government's EPCRA
hotline (1-800-535-0202); request "The Ashford Report"; expect delivery
in 2-3 weeks.
 Joel A. Tickner and Hillel Gray, ACCIDENTS DO HAPPEN (Boston,
Mass.: National Environmental Law Center [29 Temple Place, Boston, MA
02111; phone: (617) 422-0880], August 1994.
 Peter N. Skinner, TOXIC CHEMICAL ACCIDENTS IN NEW YORK STATE 1990
DATA SUMMARY (Albany, N.Y.: Environmental Protection Bureau, New York
State Department of Law, 1991).
 Ward Morehouse, "The Ethics of Industrial Disasters in a
Transnational World: The Elusive Quest for Justice and Accountability
in Bhopal," ALTERNATIVES Vol. 18 (1993), pg. 487.
 Keith Schneider, "Exxon is Ordered to Pay $5 Billion for Alaska
Spill," NEW YORK TIMES September 17, 1994, pgs. A1, A10.
Descriptor terms: chemical industry; accidents; spills; leaks;
occupational safety and health; job safety; nicholas ashford; mit;
oecd; bureau of labor statistics; national environmental law center;
emergency response and notification system; epa; studies; databases;
primary prevention; regulation; toxic torts litigation; osha; chemical
safety and hazard identification board; union carbide; bhopal; fines;
exxon valdez; oil spills; corporations;