This number eventually rose to about 100,000 and for decades, scientists wondered what caused what later became known as Gulf War syndrome: fatigue, persistent headaches, muscle pain, confusion and even difficulty speaking.
Some studies even said the soldiers suffered higher rates of motor neuron disease and a type of brain cancer known as glial brain cancer.
On May 14, a milestone study by scientists at the University of Texas concluded that Iraqi weapons storage sites filled with nerve gas and other explosives in central Iraq — and one site not far from the southern town of Nasiriyah, had exposed the soldiers to the deadly chemicals when they were blown up by Coalition jets.
This raises the question: if 100,000 Coalition soldiers could be sickened by these toxic explosions despite being 600 kilometres away, what happened to Iraqis near the ammunition sites?
The mystery of Al Muthanna
The evening of January 19, 1991, was the third night of fear for millions of Iraqis in a month-long Coalition bombing campaign.
The Al Muthanna State Establishment, an industrial complex for chemical weapons production, was near the top of the target list.
For two nights, nearly 30 cruise missiles smashed into bunkers and buildings at the remote site, 80 kilometres north-west of Baghdad. But the biggest bunkers, built with thick reinforced concrete, were impervious.
Coalition jets returned on January 19 with “bunker buster" bombs, each filled with nearly a tonne of powerful explosives, piercing one of the enormous structures.
After the war, about four Iraqi chemical sites — including the three largest ― were identified as having been hit in the air raids.
Iraq had widely dispersed its arsenal of about 700 tonnes of declared chemical weapons, burying it or storing it alongside regular ammunition — later saying the US bombed about five per cent of the arsenal in 1991, before Saddam Hussein ordered it to be destroyed.
“We now know that the Iraqis really were trying to account for all their chemical weapons when working with UN weapons inspectors in the 90s," says Charles Duelfer, the former deputy head of UN weapons inspection teams through much of the 1990s.
"The problem was their accounting was not very good. For example, they often did not mark chemical rounds differently from normal high explosive or white phosphorus. This was true of artillery and artillery rockets."
In bunker 2 at Al Muthanna as many as 2,500 sarin-filled rockets were destroyed — nearly 17 tonnes of nerve agent, later UN inspections confirmed.
It remains the largest air strike on a nerve agent stockpile in history.
About 50 milligrams of sarin in a cubic metre of air can be lethal, but even a milligram per cubic metre can cause illness, inhaled over the course of one minute.
Scientists have now confirmed even smaller doses can cause lifelong health complications.
Multiple inquiries struggled to find the cause of Gulf War Syndrome; some believed new vaccines for troops against biological weapons, or pills distributed to protect against chemical attacks could be the culprit.
Others suggested oil fires set ablaze by Saddam’s forces in Kuwait — that caused respiratory illnesses in thousands of Kuwaitis — was another possibility.
Depleted uranium — a metal used in armour-piercing ammunition — could also have sickened the troops, or dangerous pesticides used to control insects on Coalition bases.
Some said it was a toxic cocktail of all of the above.
By 1995, research zoomed in on the bombing of chemical sites.
Models using meteorological data and simulated explosions of chemical weapons at five bases across Iraq showed that nerve agents could have travelled hundreds of kilometres.
Initial studies concluded that Coalition troops were exposed to chemical vapour near a destroyed bunker in southern Iraq known as bunker 73 at Khamisiyah, close to where the forces had advanced.
There were no reports of civilian casualties near any of the sites, so civilians apparently escaped the toxic clouds.
Meanwhile, because sites like Al Muthanna in central Iraq were 600 kilometres north of the Coalition, the soldiers were thought to be out of harm’s way.
“The intelligence community's assessment was that potential plumes from these sites could not have reached US troops,” a 2004 report by the American government said.
Subsequent inquiries focused on the Khamisiyah explosions that were closer to US forces.
For Robert Haley, a scientist at the University of Texas, these explanations weren’t satisfactory.
Mr Haley and his team noted an intriguing coincidence: when air strikes hit Al Muthanna on 19 January, within 24 hours thousands of chemical alarms used by Coalition forces went off 600 kilometres to the south.
The alarms had picked up low levels of sarin, but it could not have come from Khamisiyah, which was destroyed by US engineers in early March.
Mr Haley’s suspicions were well-founded — inquiries in the early 2000s found the US Department of Defence models of the plumes were based on flawed data.
On May 14 this year, after decades of research, his team released a study on the illnesses, saying that very low doses of the nerve gas, previously thought to be too small to cause harm, were responsible for the Gulf War illness.
Until now, there has been little research on very low doses of sarin, raising the question whether tens of thousands of Iraqis escaped chemical exposure.
The chemical alarms
Mr Haley’s first breakthrough, collaborating with US intelligence analyst James Tuite, looked at newly available meteorological data.
Their meteorological study highlighted how the explosion inside bunker 2 would create extreme heat and upward pressure, sending a hot plume of chemicals about one kilometre into the sky, boosted by exploding weapons of mass destruction rocket fuel, piercing a cold layer of air known as the nocturnal boundary layer (NBL) and taking it into higher, southerly winds.
“Mr Tuite’s finding was that when the blast penetrated the nocturnal boundary layer of cold air near the ground, the high-level winds carried it SSE downwind 600km to our troop positions without coming to the ground due to the NBL barrier until the sun came up the next morning and mixed the layers right over our troop positions. Thus, no chemical weapons exposure of civilian populations in between,” Mr Haley tells The National.
“This is exactly what happened when Chernobyl exploded and the radioactive fallout was first detected the next morning in Sweden. The IAEA actually had airborne and ground sensors along the path that documented only high-level transit with the prevailing high-level winds.”
If this was the case at Al Muthanna, what about other sites?
US models in the early 2000s described the bombing of another sprawling base.
Reports on Al Muhammadiyat are murky. UN inspectors visiting the site eight months after the war found that chemical bombs had been buried between buildings among 40 warehouses, in a complex saturated with a bombardment of 1,200 unguided and 42 laser-guided bombs.
Lines of craters were still visible in satellite imagery 12 years later.
“Almost all the structures in the chemical weapons storage area were damaged or destroyed. UNSCOM found damaged and destroyed mustard-filled bombs as well as damaged sarin/cyclosarin-filled bombs. Almost all of these bombs were found outside of the buildings,” a US government report said.
Inspectors found chemical contamination in the ruins.
“The bombs were not subjected to a fire as earlier assessed, causing us to increase our estimate of the release percentage (of chemicals) from 10 to 50 per cent,” a 2001 US report said.
In other words, there was no vertical white-hot plume, as happened at Al Muthanna.
“There’s a difference between a bomb landing on a chemical munition and landing next to it,” says Dan Kaszeta, a former US government chemical weapons expert and RUSI fellow.
“If you don’t have all that right, you get model output that’s not correct.”
Satellite images of the site taken in 2003 show lines of bomb craters across the entire area.
US intelligence later said any of 17 attacks over 15 days were likely to have released sarin but it would be almost impossible to know the exact strikes and the wind direction at the time.
Any one of several Iraqi towns — or empty desert — could have been hit by Al Muhammadiyat’s chemicals.
The amount released, at first feared to be three tonnes, was also in question, because sarin’s purity degrades over time.
The UN believed the bombs were filled with the chemical agent in 1988 and might have degraded to 15 per cent purity by 1991.
“Iraqi sarin was notorious for decomposing quickly,” says Mr Kaszeta.
Factoring in this decrease in purity, US models of Al Muhammadiyat show potential “hazard areas” streaking over the provinces of Karbala and Babil, home to nearly three million Iraqis.
Even if the models were unreliable — as Mr Kaszeta suggests, it would be difficult to discount civilian harm.
“It would be important to note that there were populations of Iraqis much closer to where nerve agent sites were attacked and that they, too, may have had some exposure to nerve agents,” says Alistair Hay, a professor of environmental toxicology at Leeds University and an adviser to the Organisation for the Prohibition of Chemical Weapons (OPCW).
There are several difficulties in determining if and how many Iraqi civilians may have been exposed to sarin. One reason is the lack of tracking data, making it hard to trace patterns of symptoms. Another complication is the sheer scale of other health problems linked to the conflict.
One Iraqi surgeon who worked in Baghdad after the 1991 Gulf War contacted by The National — who didn’t wish to speak on the record — said it would be difficult to determine the cause of illnesses linked to pollution in Iraq in the 1990s.
He said there were no indications of significant rises in neurological illnesses but rising breast cancer rates in young women alarmed health workers. Yet the causes were not studied as pollution in Iraq rose and the health system collapsed under sanctions.
One of these pollution sources that could have caused nerve damage similar to sarin exposure, was the presence of lead in refined fuel that was phased out in 2010.
In 1991, three of Iraq’s biggest refineries at Baiji in northern Iraq, Basra in the south and Dora in Baghdad, were bombed by Coalition jets, along with nearby storage tanks.
The resulting clouds of toxic chemicals could have included lead and benzene, both known to cause neurological illness.
“Tracking healthcare is really tricky in Iraq. Health statistics were often absent, incomplete, or had no good baseline data, lack of diagnosis, no data on exposure,” says Wim Zwijnenberg, a researcher who has studied conflict pollution in Iraq.
But even if a rise in neurological illnesses or brain cancer had occurred, it might still be difficult to work out why. A major challenge with investigating Gulf War syndrome has been pinpointing the exact cause.
Mr Haley’s study has been called groundbreaking because it focused on a gene called PON1, which produces an enzyme that combats poisons in the blood.
His team found that Gulf War veterans with a stronger version of the gene who had likely been exposed to sarin suffered a much lower incidence of illness than those with a weaker gene.
Studies in Iraq would therefore need to track down Iraqi sufferers of Gulf War syndrome who may have left Iraq or were internally displaced by the conflicts since 1991.
“There are many overlapping exposures that may have been experienced by local civilians, including pesticides, possibly sarin and oil well fire smoke. It also seems likely that they may have had greater and longer-term exposures to these kinds of environmental mixtures, which could well have produced symptoms and systemic disorders similar to those experienced by veterans, especially in genetically vulnerable individuals,” says Dr Roberta White, who advised a US Congressional inquiry into Gulf War illness.
“This report will stimulate a lot of thinking in this particular area. We've always been concerned, more on toxic industrial chemicals and the impact of that in war,” says Hamish De Bretton Gordon, a Gulf War veteran who remembers hearing the chemical alarms in 1991, an experience which spurred his later career as an expert on chemical weapons.
In some ways, Mr De-Bretton Gordon typifies those who may have been exposed — his main problems after the war, which he describes candidly in his book Chemical Warrior, related to trauma.
For Mr Haley, the research could point to further work on treating exposure to nerve gas.
He warns that several states still possess these banned weapons, so this is not merely a historical issue for survivors.
“The strong evidence for nerve agent in Gulf War illness from our paper emphasises the need for more evidence on issues that would lead to a treatment to reverse the Gulf War syndrome symptoms, he tells The National.