It’s the end of the working week. You finish your tasks, shut down your equipment and head for home. Your thoughts are with friends and family, and you don’t have to think about work until Monday rolls around.
But what if something wasn’t properly shut down before you left? And what if you happened to work somewhere that a mistake like that could devastate your neighbourhood?
The chemical plant near Seveso in Italy was owned by the Industrie Chimiche Meda Societa Anonima, or ICMESA. They were a subsidiary company of a Swiss company called Givaudan, who were in turn owned by Hoffman-La Roche, an international manufacturer.
The chemicals they produced at Seveso were mostly intermediate compounds – things that would be taken somewhere else for further processing into a final product.
The people of Seveso didn’t think too much about the chemical plant. Most of them would have had no idea what was made there.
That changed in July 1976.
On Saturday, the 10th, just after 12:30pm, a white cloud burst from the reactor with a loud hissing, screeching sound. It spread out across the neighbourhood and descended like a fog. The cloud spread, pushed by a northerly wind, and settled not only over Seveso, but also the nearby towns of Meda, Desio and Cesano Maderno.
Stuff came down from the cloud in white flakes. It looked like snow. Children played in it.
But it wasn’t snow. Where people encountered it, their eyes burned. They started coughing. Later, they got headaches, felt dizzy, had diarrhoea.
Despite that, they still didn’t think too much about it. One resident said, “At that time, we didn’t pay too much attention to it, because there was always some cloud, some leak. And although there was a terrible smell, we didn’t worry about it; we ate, we collected vegetables and flowers from the garden, as if nothing had happened.”
According to some accounts, people who were concerned enough to contact the police or the plant for information were told that it was nothing to worry about. According to other reports, officials from the plant told nearby residents not to eat any garden produce, as it may be polluted, and asked police to repeat that warning. The police refused to do so, as they could only take such instructions from the municipal health officer, who could not be located that day. It was, after all, the weekend.
Sunday came, and throughout the area there was an acrid smell. Shiny, oily crystals were everywhere.
Paolo Pauletti, the director of production at ICMESA, was reached that day and told of the incident. He met with the mayor of Seveso, Francesco Rocca. As the health officer still couldn’t be reached, he asked the mayor to instruct the police to pass on the warning against eating fruit and veg from the local area. Fabricio Malgrati, mayor of nearby Meda, under whose jurisdiction the plant actually stood, was also informed.
But, it was a warm summer Sunday, so children still played outside, people enjoyed the sunshine and tried to ignore the smell. Having either not received any warnings, or failing to take them seriously if they had, people made their Sunday dinners using vegetables, chickens and rabbits from their gardens.
On the Monday, two days after the incident, the manager of the plant, Herwig von Zwehl, met with Mayor Rocca and the local health officer, Francesco Uberti, who was now back on duty. A letter was sent later confirming what they had discussed.
The incident had taken place in a reactor used to make a chemical called trichlorophenol.
Although the letter implied that the vapour cloud was largely comprised of chlorinated phenol, von Zwehl added a caveat:
“We are not in a position to evaluate the substances present in the vapour or to predict their exact effects, but knowing that the final product is used in manufacturing herbicides, we have advised householders in the vicinity not to eat garden produce.”
While true, this wasn’t actually what ICMESA was producing trichlorophenol for; they were sending it on to a Givaudan plant in America where it would be converted into a bacteriacide called hexachlorophene. This had been widely used from the 1950s in consumer products like soap, talc, deodorants and more. However, in the late 60s and early 70s studies started to demonstrate that it wasn’t as safe as originally thought. Large doses of hexachlorophene could result in brain damage and paralysis; in 1972 it was linked to the deaths of more than twenty babies in France who had been treated with talc containing high levels of hexachlorophene. By the time of the Seveso incident, the US had banned its use in consumer products – it was only used for surgical soaps.
And when Von Zwehl had mentioned herbicide in his letter? The true scale of the risk might have been clearer if he’d mentioned the name of one particular herbicide to use trichlorophenol as an ingredient. That name was Agent Orange, as used by the American military during the Vietnam War.
Following their meeting, and that rather vague letter, health officer Uberti came to the conclusion that there was no risk to the area. He passed that conclusion on to the mayors of Seveso and Meda, and to the provincial health officer in Milan.
Not everybody shared that conclusion, though. Workers at the chemical plant suspected that the company hadn’t revealed the full extent of the risks, and threatened to go on strike.
Whilst the workers obviously felt that their direct management was not reacting appropriately, some in the company hierarchy had taken the incident very seriously. When he learned what happened on the day after the cloud was released, Dr Jorg Sambeth, technical director and chief chemist of parent company Givaudan, had the reactor sealed off and a number of soil and dirt samples from the area sent for analysis. Givaudan’s laboratories in Switzerland were routinely used to check the trichlorophenol produced at Seveso, so that was where these samples were sent. However, the tests couldn’t be done quickly. Even though the lab’s director, Dr Bruno Vaterlaus, arranged for them to be processed straight away, it would take until at least the following Thursday to reach any conclusions. Whilst waiting for the results, Vaterlaus and Sambeth went to Seveso to see the damaged reactor for themselves. They felt it was serious enough to escalate it to their parent company, Hoffman-La Roche. Clinical research director Dr Giuseppe Reggiani now became involved, at a meeting with Sambeth in Switzerland.
Back in Seveso, a number of children were developing skin rashes. They were hospitalised as a precautionary measure, and Reggiani visited them there. He would later recall seeing twelve children there, four of whom had very swollen faces. Those four, he suggested, should be moved to a hospital which was better equipped to provide intensive care, should their condition deteriorate. They were accordingly moved to the Niguarda Hospital in Milan. Treatment with corticosteroids reduced the water retention which had caused their faces to swell, but they weren’t yet out of the woods. According to Reggiani, some of the doctors there brought up the idea of Agent Orange.
By the 14th of July – the Wednesday, just four days after the cloud was released – there were reports that animals in the area had died. According to an article in TIME magazine, written a month afterwards:
“One farmer saw his cat keel over, and when he went to pick up the body, the tail fell off. When authorities dug the cat up for examination two days later, said the farmer, all that was left was its skull.”
On the 15th of July, the mayors of Seveso and Meda declared that an area south of the chemical plant was polluted, and placed a ban on consuming local fruit and vegetables.
On Friday, the 16th of July, the workers at the chemical plant put their threatened strike into action. It wasn’t the first time they’d taken such action – there had been a strike just two months earlier, to force management into allowing the regional factory health inspectorate in to assess health hazards in the plant. Workers had been asking for this for two years.
Signs were placed around the vicinity of the plant that day.
“DANGER AREA. CONTAMINATED. DO NOT EAT VEGETABLES, FRUIT, OR ANIMALS THAT EAT GRASS FROM THE GROUND.”
For some residents, this was the first warning they’d received. But they still didn’t know what they were dealing with.
Saturday, 17th of July saw the pronunciation that the ICMESA plant was polluted, and orders to have vegetable and fruit products and animals in the contaminated area destroyed.
It also saw the story on the front page of the Milanese newspaper, Il Giorno.
That day, a health official from Milan by the name of Dr Aldo Cavallaro visited the plant. ICMESA representatives told him what had caused the release – a temperature increase in the trichlorophenol reactor. On returning to his lab, Dr Cavallaro read up on trichlorophenol production and found out what that could mean.
There was a certain byproduct of trichlorophenol production. If temperatures went over 200 degrees C (392 F) it could form tetrachlorodibenzeno-p-dioxin, more simply known as just dioxin. This was known to be a highly toxic substance. It was linked to reports of cancers, systemic poisoning and death. And yes, it was present in Agent Orange. However, the literature available talked about its toxicity in animals. Little was known about its effects on humans.
Although the worker’s strike had already brought work at the plant to a halt, it was officially closed on Sunday the 18th of July by order of Mayor Malgrati of Meda. A local magistrate ordered that the doors to the reactor be sealed to prevent tampering.
On the following Monday, the 19th of July, Dr Cavallaro asked ICMESA officials a straight question: was dioxin present in the cloud? The response was, it might have been.
He then flew to Switzerland to speak directly to management at Givaudan. They had the results from the samples that had been taken; they did contain dioxin, but they said the levels were unclear.
Although the company had confirmed nothing publicly, the media made the same connections. They checked with technical experts, and reported that as little as five pounds of dioxin was enough to kill more than 100,000 people. That was the amount believed to have been released at Seveso. People reading these reports, naturally, began to panic.
For his part, Reggiani was trying to establish what exactly should be done. He called other companies who had experienced accidents with trichlorophenol to get their advice. They all said the same thing – evacuation of the local population was necessary. He and his colleagues began to try and work out where exactly the contamination had reached, which parts of the area were at risk and which could be considered safe.
On the 20th of July, a map of the contaminated area was handed to Dr Cavallaro and Dr Giuseppe Ghetti, the municipal health officer for Seveso and Meda. They had been sent to Switzerland to get the results of the laboratory analysis, and reported back to Dr Vittorio Carreri, the provincial health officer, and his superior, Dr Vittorio Rivolta, the Minister of Health for Lombardy.
More people were being hospitalised, throwing up and with rashes on their skin after coming into contact with produce from the Seveso area. Carreri sent a circular out to tell doctors to watch out for particular complaints that might be connected to the incident, but it was increasingly obvious that more needed to be done.
On the 21st of July, Rivolta announced a number of measures to contain the contamination. All traffic on the two major roads of Seveso was banned. Desio and Cesano Maderno restricted the consumption of fruit and vegetables. A commission of experts was to be set up to assess the health implications.
Because it seemed to be the children of Seveso who were suffering the most, Mayor Rocca asked to have 90 of the youngest sent to a convalescent camp.
At a meeting on the morning of Friday the 23rd of July, Rivolta explained what had happened at Seveso, and what measures he had put in place, to an assembly of leading scientists and doctors, as well as local officials. His actions were deemed to be appropriate; everything was under control. Rivolta said as much in a television interview that day.
Reggiani tried to intervene in that meeting, but was informed that it was not open to the public. Before being forced to leave, he left a message that he had important information for Rivolta and Ghetti.
He reached Rivolta that evening, at a meeting of local administrators where they were finalising the details of their public health measures, and quite bluntly told him that, despite what their experts had agreed that morning, the area did indeed need to be evacuated. Extreme measures would need to be taken, perhaps even the destruction of homes.
Rivolta was dismissive; Reggiani had travelled in a hurry and hadn’t brought any credentials with him.
“This man has been parachuted in; nobody was expecting him… I have the impression he is bluffing.”
Rivolta did, however, agree to reconvene the experts at a second meeting the next day. In order to ensure everyone’s attendance, he asked the police to locate them all, and guard Reggiani to ensure his attendance. Reggiani was tipped off by a colleague and instead left Milan to return to Switzerland. In his place, Bruno Vaterlaus from the Givaudan laboratories in Switzerland attended, armed with a letter from the company’s director and a comprehensive map of the contaminated areas. Reggiani’s exit may seem a little odd, but it makes more sense in light of the fact that Von Zwehl and Pauletti, the managing and production directors at the plant, had already been arrested on the 21st and charged with causing the disaster.
The meeting that Saturday – a full two weeks after the incident – also considered evidence from Dr Cavallaro, who now reported that he, too, had identified dioxin in samples from the Seveso area. The decision the experts had made only the previous day was now reversed, and evacuation was at last ordered.
On the 26th of July, 170 people from Seveso and 55 from Meda were evacuated. Another 511 people were evacuated on the 2nd of August.
For those people, it was an extraordinarily hard time. They said their lives had become brutissima. It wasn’t merely the hardship of being forced to leave with only the clothes they were wearing and whatever they could fit into a suitcase; it wasn’t just the heartbreak of being forced to abandon homes that, in many cases, they had quite literally built with their own hands, or the fact that the accommodation found for them was cramped and often required families to be split up. On top of all that, they were treated as if they were, themselves, contaminated. As if they were bringing the poison with them.
They were compensated for the property and produce that they had lost, and promised homes of equivalent value to those they’d left behind, but it was little comfort to many. Housewife Caterina Rivolta said a couple of years later,
“I’d give anything to move back. My husband and I saved for 16 years to buy our home. Nothing will ever replace it.”
An area of some 269 acres was designated Zone A; this was the most heavily contaminated area, from which the residents were evacuated. Barbed wire surrounded the area, and on the highway passing through large signs warned:
“CONTAMINATED AREA. ROLL UP WINDOWS. CLOSE VENTS. DO NOT STOP. DRIVE SLOW.”
This was not, however, the full extent of the contamination or the disruption to residents. Another 669 acres, stretching out to the south of Seveso, was designated Zone B. Although the 4,800 residents of this area were not evacuated, they were still at risk from the dioxin.
Based on the available evidence, it was thought that the most at-risk groups were children and pregnant women. For that reason, all pregnant women and children under the age of fifteen within Zone B would be evacuated for the daylight hours, and only return at night, so that their exposure would be limited. They were also told not to eat any fruit, vegetables or animals and not to touch the ground.
A further area of several thousand acres and housing 20,000 people was called Zone R or the Zone of Respect. In this area, people were to be monitored, but not restricted.
Local doctors warned women not to get pregnant; autopsies of poultry from the Seveso area had shown unfamiliar pathological conditions, and it was simply not known how dioxin might affect people. Although abortion was illegal in Italy at the time – and the powerful Catholic Church opposed all abortions – a recent court decision had made it possible for an abortion to be performed if the mother’s health was in danger, provided a panel of two doctors and a psychiatrist gave permission.
The New Scientist in 1983 reported that 90 Seveso women had decided to have abortions. German pathologists examined 30 of the foetuses; 29 were found to be normal and one too badly damaged to tell. It’s impossible to tell how many women may have had illegal abortions.
Two years later, local residents were still wary. A building contractor and father of one named Ugo Basilico, told TIME magazine:
“I thought it was about time we had another child, but the doctor says better wait a while. If you have a baby with some defect, the baby is there for life.”
With the residents moved out, officials moved in to try and work out what to do with the polluted area. For the next year, the air in schools throughout the area would be sampled and tested; for the next ten years the soil would be monitored. This needed expensive equipment which was in short supply – mass spectrometers, for example, of which there were only five in the entire country. Information about the health effects of the dioxin had to be collected, and at first this was not done in any kind of organised way.
The chemical plant was completely shut down one month after the disaster, but it would take six years to clean it up. People objected to the first plan, which was to encase it in a concrete structure, so the decision was made to dismantle the whole thing using the tactics that would be used for a nuclear facility. The plant was sealed, and workers in air-tight suits took everything apart and packed the contaminated stuff into lead drums. When this process was completed, there were forty-one drums full.
Because people wanted to return to their homes, the authorities did what they could. Some of Zone A was irretrievable, but it was possible for about 60% of the evacuees to return. The homes they came back to looked very different, though. Everything that was replaceable – curtains, carpets, clothes, furniture, appliances, even wooden floors – had been thrown away. The exteriors were washed and the topsoil was removed.
It was the early 80s before the worst areas of Zone A were tackled. There, the buildings were demolished, vegetation and topsoil was removed, and any area that couldn’t be decontaminated to a certain level was fenced off. Over 270,000 cubic yards of contaminated soil had to be disposed of. The area is now a park, the contaminated soil buried in concrete beneath the ground.
In the end, the health effects of the Seveso disaster weren’t as bad as it had been feared they could be. There were no deaths that were linked directly to the disaster. A 1998 study concluded that the only issue that could be linked to the dioxin release with certainty was chloracne. This is a condition like acne which causes the skin to develop blackheads, papules and pustules. It was found in nearly 200 individuals, mostly children. Even in Zone R, where the contamination was thought to be low, increased incidence of chloracne was found. In most, if not all, of those affected, the condition cleared up within a couple of years.
Other studies have suggested an increased mortality from cardiovascular and respiratory diseases, increased occurrence of gastrointestinal, lymphatic and breast cancer, and lower than average sperm counts in males whose mothers had been exposed to dioxin during pregnancy. However, because the data was limited – there was a relatively small population exposed, and there is no individual exposure data – it’s hard to quantify these effects and link them directly to the disaster.
One death can be, tangentially, linked to the Seveso disaster. Paolo Pauletti, the plant’s director of production, was assassinated in February 1980 by the radical left-wing terrorist organisation Primea Linea.
Obviously, in the aftermath there were questions as to how exactly the dioxin had been released.
When the plant shut down that Friday night, in accordance with a law that required production to stop for the weekend, the trichlorophenol reactor hadn’t been stopped correctly.
The process involved mixing tetrachlorobenzene with sodium hydroxide in an ethylene-glycol solvent. Then, the mixture would be distilled to take out the solvent.
This time, the reactor had been shut down before the distillation was complete. The steam used to heat the reactor was switched off, and the agitation of its contents was stopped. It was at that point at a temperature of 158 degrees C; although usually several thousand litrs of water would be used to cool it down, technicians reasoned that it would cool down by itself if left to its own devices.
It did not.
The temperature increased, and the resulting rise in pressure blew out a safety valve in a vent pipe that projected above the roof of the chemical plant. The wind did the rest, taking the dioxin south and depositing it across the area.
It was a tragic error, made even more tragic in hindsight when the company attempted to simulate the same conditions. They found that, had the stirrer been left running for just ten minutes longer, the accident could have been avoided.
Ignorance, confusion and miscommunication were also at the heart of this incident.
An article in the New Scientist in September 1983 – seven years after the disaster – said:
“It can be argued that, in the end, lack of information had a worse effect on the people of Seveso than the dioxin itself. Whenever someone is sick in the area now, poison is suspected. Mothers are afraid to let their children play outside.”
It took ten months for the first report on the Seveso accident to be released – and that was written in highly technical language, and published in a professional journal. Local residents were in the dark.
Gianni Tognoni, a researcher at Milan’s Mario Negri pharmacology institute, worked with two colleagues to put together a booklet which would explain the situation to those who were actually affected by it, in clear language that they could understand without a chemistry degree. All it needed was official approval from the public health authorities so it could be published.
They delayed that approval for so long that it wasn’t worth publishing at all.
It seems astonishing that it took so long for the company to acknowledge the risk of the chemical release. Scientists at Givaudan had discussed the extreme toxicity of dioxin with managers at ICMESA as early as 1970. Despite that, they had no contingency plans in place, and hadn’t made the local authorities aware of any potential risk either.
As a direct result of this disaster, the EU introduced new, standardised industrial safety regulations, known as the Seveso Directive. Now on its third iteration, it means that national legislation and national chemical safety authorities exercise much greater control over dangerous substances, their use and their production.
There were also criminal charges brought; five former employees of ICMESA or its parent company Givaudan were initially sentenced in September 1983. They all appealed, and in May 1985 three were found not guilty. The remaining two had their sentences confirmed in May, 1986. One of those was Jorg Sambeth, who told a German reporter that he “was absolutely stupid”.
“Of course we should not have made the mistakes… the Seveso accident was a prime example of the stupidity of management. Including me.”
Thanks this episode go to:
I’d like to say a special thank you to Patreon supporters Skeleheron and Mish Liddle, and to all of you for listening and reading.
Supporting the Great Disasters podcast on Patreon can give you access to exclusive content, including at least one mini-episode per month, and helps the show keep going.