In 1989, I was working as an at editor at IEEE Spectrum when I was assigned to write a feature on Bhopal. The thirtieth anniversary of that industrial disaster that killed thousands is tonight. My article back then began:

On arriving at work on Dec. 3, 1984, Rick Horner, a chemical safety engineer with the U.S. Environmental Protection Agency, was ordered to rush over to the State Department. The embassy in New Delhi, India, had been calling frantically for information on an industrial chemical called methyl isocyanate (MIC). The night before, an MIC vapor from a Union Carbide pesticide plant had spread like a low-lying fog over a heavily populated portion of the central Indian city of Bhopal. Already dozens—maybe hundreds—of casualties were being reported.

Horner knew MIC well from his days as a safety officer at Eli Lilly Corp. in the mid-1970s. The substance is superbly reactive, readily adding methyl to other organic molecules to produce pharmaceuticals or Sevin, a pesticide, in Bhopal. But what Horner recalled as vividly was the skull and crossbones on the data sheets accompanying the small amounts used in the Lilly laboratories. "The chemical can cause instantaneous death," said Horner, which meant to him that it was "on a par with phosgene, chlorine, or any of those other bad actors."

The article I wrote looked at the risk planning, such as it was, that Union Carbide had undertaken before Bhopal. It also documented the endless legal wrangling as to whether a disgruntled employee sparked the incident or whether negligent plant design was at fault. What I had written was one of a countless number of articles examining the lessons to be taken away from the catastrophe.

The continuing study of the incident—and the consequences of operating complex, hazardous technologies in developing countries—still merit consideration thirty years after. One example: My friend, the brilliant writer and former Sciam editor Madhusree Mukerjee, has written in the current issue of Dissent about the role of geopolitics in undercutting the legal case against Union Carbide, now absorbed into Dow Chemical.

Even in 1989, it was clear that the memory of Bhopal would linger. My article in Spectrum referenced Union Carbide's continuing problems after Bhopal at a plant in Institute, West Virginia:

For six months after December 1984, EPA inspectors were in and out constantly at the Institute plant, trying to ensure its MIC [methyl isocyanate] process would not turn into another Bhopal. "We did a complete cradle-to-grave analysis on the MIC facility," said Rick Horner, the EPA safety engineer. Union Carbide spent $5 million or more to improve the MIC safety systems. Both the inspection team and the corporation were convinced that everything possible to make it safe to process and store MIC had already been done when the operation started again at mid-year.

In August of 1985, two months after the agency completed its review, a cloud of aldicarb oxime, another chemical used for making pesticides, wafted over the city. The gas sent 134 people to the hospital. The aldicarb and MIC facilities sit side by side. All attention during the review had been focused on the MIC plant.

When the accident occurred, a computer operator in the Institute control room, lacking a gas dispersion profile for aldicarb, entered a data file for another chemical. Crunching the faulty data, the computer projected that the leak would stay within the confines of the plant. It did not. The data file entered was the one for methyl isocyanate.