Deadly Surprise.

Will Y2K cause chemical disasters?

HERE'S A SOBERING THOUGHT: AT least 85 million Americans live within five miles of a facility that handles toxic chemicals. And there is no information on what steps a huge number of these companies have taken to ensure they won't have catastrophic accidents because of Year 2000 computer bugs lurking in their operations.

The 1984 Bhopal disaster demonstrates how bad an accident at a chemical plant can be. Forty metric tons of methyl isocynate, a highly toxic organic chemical used to produce pesticides, were released into the air at the Union Carbide plant in the densely populated city of Bhopal, India, after water got into the tank and reacted with the chemical. Two thousand people died and 100,000 were seriously injured. Ten years after the event, 50,000 remained partially or totally disabled, according to the International Medical Commission on Bhopal.

Could the Y2K bug cause similar disasters here? The answer appears to be yes. The larger multinational companies have made a serious effort to discover, test, and fix their Y2K bugs, according to Gerald V. Poje, a member of the U.S. Chemical Safety and Hazard Investigation Board. But there is little information about how--or whether--vast numbers of small and medium-sized companies that manufacture, process, handle, and/or dispose of toxic chemicals are dealing with the task. More disturbing yet, Poje, who sits on the government body charged with both investigating chemical disasters and working with industry and communities to prevent them, said nobody even knows exactly how many of these companies exist or where they all are.

As almost everyone knows by now, the Y2K bug itself is a simple mistake. Computer programmers abbreviated the date in their computer codes so that, say, 69 was understood to mean 1969. When the date rolls over to 00 at the end of this year, computers won't know it is 2000, not 1900, which could cause them to malfunction or crash. The glitch resides in billions of lines of mainframe computer code, numerous personal computers, and an unknown but substantial percentage of the world's 20 to 70 billion microchips, fingernail-sized stand-alone computers "embedded" into everything from nuclear missiles, cars, and traffic lights to pacemakers, coffee makers, chemical plants, and reservoirs.

How could Y2K cause a chemical disaster? Computers run the "distributive control systems" that keep vats of volatile chemicals swirling at the appropriate levels of flow, pressure, and temperature at most plants. At midnight on January 1, 2000, a chip that failed to recognize the date could freeze or "zero out," causing the whole computer to stop working. The chemicals would no longer be circulating safely, and if there were no backup systems (or if they too failed) an "uncontrolled reaction" might be the result. "That could generate its own heat and pressure," as one engineer puts it, which might soon be too much for the vessel containing the chemicals. They could spill out, and turn into vapor. If the chemicals were highly flammable, like natural gas, styrene, or toluene, they could ignite, causing deadly fires or explosions.

Engineers insist that this scenario is extremely unlikely, because someone would be on hand to help, and most spills would be funneled off by safety valves before they became a real danger. Still, they are possible, especially because the Y2K bug exists on so many chips and could cause multiple simultaneous failures. The only way to be sure they won't occur is to remove the bug--and that is far from easy.

Consider the task facing a medium-sized chemical company, which typically has 8,000 computer programs with 12 million lines of code, according to Dennis Grabow, who has written on the Y2K problem for Chemical Engineering magazine. Since one out of every 50 lines contains a date, that works out to about 240,000 lines of code that must be changed. For smaller companies, that translates into three years of labor and $30 million just to fix their software. This estimate does not include dealing with embedded chips, which are expected to give industry its biggest Y2K headaches because they can be hamer to find and test. To make matters worse, most companies' systems contain layers of different computerized product from several manufacturers. This complicates the task, because companies must check with their suppliers on whether the equipment has a Y2K fault, and answers aren't always quick to come. All these obstacles mean that most, if not all, companies are fixing only their "mission critical" systems, those they deem essential to their operations, says Grabow.

Despite the threat, the government appears to be asleep at the wheel. President Clinton's Y2K czar, John Koskinen, claims that Y2K is the biggest management problem the world has faced in the past 50 years. Yet neither the U.S. Environmental Protection Agency nor the Occupational Safety and Health Administration, the two federal agencies responsible for protecting the public and workers from environmental hazards, have surveyed industry to see what they are doing to find and fix their Y2K bugs. Most federal public servants either downplay the threat or say it is up to the states and local governments to make sure chemical catastrophes don't occur within their borders. But it appears that most states and local governments either don't recognize the threat or are not prepared.

The following computer-related incidents--some of which occurred during Y2K testing--portend what Y2K might bring:

--The Olympic Pipeline exploded on June 10 in Bellingham, Wash., after the company's control-room computer at one of its plants crashed, killing three people. The Seattle Times reported that after the backup computer came on, a valve shut, building up pressure until the pipeline burst.

--Four million gallons of raw sewage spilled into a suburban park in Los Angeles on June 17 after a computer closed an underground gate that sent the muck spewing through a manhole during a Y2K test. The Los Angeles Times reported clean-up costs for the spill were estimated at $100,000.

--Computers crashed at Unit 2 of the Peach Bottom Atomic Power Station in Pennsylvania on Feb. 8 during a Y2K test. Control-room operators had to rely on manual gauges for the seven hours it took to get the monitoring computers back online.

There have been concerted efforts within the chemical industry during the past three to four years to try to prevent Y2K problems. Rob Bradford, a special consultant to the Chemical Manufacturers Association on Y2K, says chemical companies are being careful because of their liability in case there is an accident, and he's confident there won't be any major disasters. "We don't think the chemical industry's going to be on the front pages of the newspaper," he says.

But others don't share Bradford's optimism. Gerald Poje points out that the 6,000 questionnaires sent out to other segments of the chemical sector by trade associations representing them only garnered 300 responses. Although he said those 300 "seemed diligent in their response and effort to meet Y2K-compliance deadlines," they hardly provide a scientific extrapolation of what the smaller and mid-sized companies are doing. To get an idea of the size of the chemical universe out there, consider that the EPA estimates that more than 275,000 establishments nationwide handle hazardous chemicals. And that estimate may be low, since California alone lists 130,000. In trying to assess chemical companies' Y2K status for his clients, Dennis Grabow says he often finds companies very reluctant to talk about it. He recounts one exchange with the Y2K program manager of a large corporation: "I told her that I'd studied [the Y2K issue] for several years, had probably given 150 speeches in technical areas and know all about it, so I'd like to just cut to the chase. Her response was `well, you probably do, and the questions I know you're going to want to ask, I probably can't answer.' She said she'd direct my questions to the company's lawyers, which inspires absolutely no confidence"

Power Failure

Part of the problem is that chemical companies are extremely dependent on electricity, not only to keep their business operations going but also to prevent accidents. Even if a company's own computers are prepared, a glitch in the electrical network caused by the Y2K bug could disable the complicated systems these companies rely on to keep their chemicals in a stable condition.

Many questions have been asked already about the reliability of the electricity supply in the face of Y2K. The general consensus--led by the North American Electric Reliability Council (NERC), an industry group charged with coordinating electricity sector efforts on Y2K, and President Clinton's Y2K czar, John Koskinen--is that Y2K will not cause widespread power outages.

But not everyone accepts this rosy scenario. Utility industry Y2K consultant Rick Cowles--echoing Gerald Poje in the chemical sector--says not enough information is known to rule out significant power blackouts. First, he says, the 251 entities NERC has collected information on only comprise 3 percent of those that are listed as suppliers by the Edison Electric Institute, the trade association of for-profit electric utilities. Second, based on his own professional experience, he believes a "significant minority" of the companies NERC lists as ready are being "a little fast and loose with their Y2K readiness statements" Third, despite assertions to the contrary, he says NERC has almost no information on independent power producers (IPPs). Although IPPs only account for about 12 percent of the power-generating capacity nationwide, they are the fastest-growing segment of the electric industry and in some areas--particularly the Northeast--account for a much greater proportion of the power provided. Cowles estimates that IPPs provide about 50 percent of the power in his home state of New Jersey, a densely populated major chemical producer with the country's highest concentration of toxic air and water releases.

Like most other businesses, chemical and power plants depend on things outside their direct control. In fact, according to the Institute of Electrical and Electronics Engineers, Inc. (IEEE), the oldest and largest international nonprofit association of engineers and computer scientists in the world, this web of interdependency makes Y2K unsolvable. In an open letter to several congressional committees on June 9, 1999, the institute's Year 2000 Technical Information Focus Group said the way industry and government have been attacking the problem is technologically all wrong. "If interconnected systems are made compliant in different ways, they will be incompatible with each other. In fact, efforts to become `Y2K compliant' in one place could be the direct cause of such failures in others," the IEEE wrote. The letter also points out that many Y2K failures are inevitable because some may not be able to be detected before they fail. Furthermore, the sheer amount and complexity of all the software that runs most systems makes it impossible, considering the time constraints, for complete testing, which the IEEE says is "a recipe, almost a commandment, for widespread failures"

What, Me Worry?

Whatever the dangers, it's clear that chemical companies can't work all this out by themselves. Yet the government's response thus far has been little more than passing the buck. The EPA and John Koskinen, who has chaired the President's Council on Year 2000 Conversion since its inception in February 1998, say the federal government has no authority to mandate Y2K fixes in chemical plants. Instead, both have developed "cooperative partnerships" with industry, which have consisted of providing information. Koskinen says the council has not pushed any government agency to regulate on Y2K, because that would require a specific statute. "Our biggest obstacle in dealing with some of the industry groups was getting over their concern that somehow we were going to be telling them what to do or pretend we had authority to regulate that we didn't," he said.

In the meantime, Koskinen acknowledges that not such is known about what many small and medium-sized chemical companies are doing about Y2K. At the urging of the Chemical Safety Board and Sens. Robert Bennett (R-Utah) and Christopher Dodd (D-Conn.), the chair and co-chair of the Senate Special Committee on the Year 2000 Technology Problem, he hosted an industry roundtable at the White House during the last week in August to find out. Koskinen doesn't seem to find the information gap very disturbing, and he says that "small and medium-sized companies, by definition, have small and medium-sized problems with challenges that can be met if they deal with them" (The Chemical Safety Board's Gerald Poje disagrees, saying that small and medium-sized chemical companies could have Bhopal-sized accidents if they were working with highly toxic material and were located in heavily populated areas.) Getting the companies to address Y2K is, of course, the challenge. Koskinen also acknowledges that many smaller companies are waiting to see what breaks during the rollover before trying to fix their systems. Although he calls that "irresponsible behavior," he says ultimately it is up to the companies to decide to address Y2K.

Like other government departments, the EPA has focused its efforts on trying to get information out to facilities at risk. According to Don Flattery, who is responsible for liaisons with industry on Y2K issues, the EPA cannot require Y2K status reports from companies that deal with hazardous chemicals, let alone demand that they find and fix their Y2K bugs and verify that it's been done. The Occupational Safety and Health Administration (OSHA) produced a single-page fact sheet on the problem last year and has been distributing it to workplaces. When asked what workers are supposed to do to ensure their safety, OSHA Deputy Director of Policy Francis Frodyma says it isn't up to his agency. "It's the employers who have the responsibility to protect their employees" He says OSHA presumes employers are doing enough Y2K testing to meet their requirements to provide safe workplaces.

What about the workers? After all, OSHA does have a regulation, the Process Safety Management Standard, which requires that workers get a chance to participate when procedures involving hazardous materials are altered. But this is not happening in any systematic way with regard to Y2K, says Deborah Solomon, a lawyer with the Paper, Allied-Industrial, Chemical and Energy Workers' International union (PACE) who is also working on a program to educate workers on Y2K issues in their workplaces for the Clearinghouse on Hazardous Waste Worker Trainings. "Many employers are not training their workers in Y2K awareness," she says. "They're not involving them in the contingency planning process, and they're not letting them know how they should be prepared personally for changes in their job functions" Solomons observation is not based on hard statistical data but on her daily conversations with workers.

As for Congress, Rep. Stephen Horn of California and Sens. Bennett and Dodd have by many accounts done a yeoman's job of trying to move the government to action via their committee on Y2K. But their efforts have been hampered by small resources and lack of authority. "The committee, by design, cannot challenge anybody else's turf," one source said. "It can only refer to another committee" The source added the committee would never have been created if it had been given the power to write legislation, appropriate money, or direct government departments to do anything.

Don't Think Local

So who is really in charge when it comes to protecting us from possible chemical accidents on and after Jan. 1, 2000? Those in the federal government--Koskinen, the Senate Y2K committee staff, and the EPA--say it is up to states and local governments. And what are they doing about it? Fred Millar, the environmental program director for the Center for Y2K and Society, spends much of his time contacting and trying to prepare local communities throughout the country on how to deal with Y2K, and he's not hopeful about what he sees. "State and local governments aren't doing a damn thing about what's going on outside their own agencies--they're not doing independent auditing of electric, gas, food, or water," he says.

The National Association of State Information Resource Executives (NASIRE), an organization representing state chief information officers, has been working to coordinate state government efforts to deal with Y2K since 1996. In testimony before the Senate special committee on Y2K in July of this year, NASIRE president Mike Benzen said 20 states were 90 percent done with fixing their "mission critical systems"--those the states consider necessary for maintaining their business--and 38 were at least 75 percent complete. NASIRE'S information was self-reported by the states.

Others claim that many states are simply unaware of the extent of the problem. Testifying before the Senate special committee on Y2K this past May, Jane Nogaki of the New Jersey Work Environment Council said the New Jersey Department of Environmental Protection seemed ill-informed about the state's chemical plants and had not bothered to send anyone to the conference the Chemical Safety Board held last December on Y2K. "Thus it appears the DEP, the agency charged with preventing toxic disasters, has put its head in the sand when faced with challenges posed by the `millennium bug,"' she concluded. She added that no other state agency seemed to be independently verifying "even the most basic assertions" from chemical facilities.

As for counties, they appear to be even less well-prepared. The National Association of Counties (NACo) has conducted two surveys of Y2K readiness among 500 randomly selected counties. Fifty percent of the counties polled in the first survey last November said they had no plans to address Y2K issues. In the follow-up survey in May, that number dropped to 26 percent. Yet NACo spokesperson Shawn Bullard says counties are well-prepared because they have civil defense plans and deal with emergencies all the time. In other words, they are prepared to deal with the disaster after it happens, not to prevent it.

By then it may be too late for an effective response. If there are multiple disasters, says Eric Lamar, the hazardous materials director for the International Association of Fire Fighters, which represents the nation's emergency responders, we're pretty much on our own: "Most communities are only set up to deal with one or two significant emergencies. So if you operate a big [chemical] plant, when the date rolls over to 2000, you should probably assume that emergency responders will not be available"