New Analysis Shows Risk in Intel Silica Dust

From the Corrales Comment of 9/23/06: Scientific analysis of Intel's claim that it does not release a potentially lethal form of silica (suspected in the deaths of two Corrales residents) concludes that Intel is probably wrong.

Contradicting an oft-repeated assertion by Intel officials, Los Alamos Air quality specialist serving on Intel's Community Environmental Working Group has reported that Intel's thermal oxidizers (incinerators) do create airborne crystalline silica when the industrial chemical hexamethyldisilazane (HMDS) is burned.

Mike Williams, serving on Intel's committee as a representative of New Mexicans for Clean Air and Water, disagrees with Intel's most recent risk assessment which states flatly that the microchip maker releases no crystalline silica.

No one disputes that Intel releases silica particulate into the air when it routinely burns HMDS as a pollution abatement measure. But Intel has maintained it only emits amorphous-shaped silica.

There's a major public health distinction between the two forms. When people breath in the round-cornered amorphous silica, the tiny particles generally get exhaled right back out.

But with the sharp-edged crystalline form, the particles tend to lodge in the lungs creating much increased risk for lung cancer and lung fibrosis.

Silica emissions from Intel were suspected as a cause of death of former Village Councillor Larry Vigil in 2001. He died of pulmonary fibrosis; doctors listed its cause as unknown. Vigil himself, a long-time resident below Intel, was convinced his lung disease was caused by Intel's emissions.

A second Corrales resident who lived even closer to Intel also died from lung fibrosis a few months later.

Intel's reports to the N.M. Environment Department for that period show the microchip maker was releasing an average of 11 tons a year of silica particulate into the air.

The question arose, then, as to how much of a health threat the silica emissions are, and that depended largely on whether the particles going into the air over Corrales are in amorphous or crystalline form.

As corporate officials had done in the past when villagers reported health problems from Intel's air pollution, they hired a consultant to reassure the public that Intel's operations did not constitute a significant health threat.

Silica emissions were part of the new risk assessment produced for Intel by the Colorado-based Environmental Resources Management (ERM) in 2005.

Assurances that the silica released is amorphous and not crystalline are based on Intel's claim that the heat in the oxidizers that burn HMDS is not sufficient to produce the crystalline form.

ERM's toxicity tables for the risk assessment show that crystalline silica is produced at heat "above 1000 degrees to over 1400 degrees."

An Intel-Rio Rancho webpage displays a report, "Durr Thermal Oxidizer Research" dated April 27, 2004, stating that the combustion chamber temperature in the incinerators here is 1,400 degrees F.

Even so, ERM states categorically that "Intel does not emit crystalline silicon dioxide," without citing evidence.

ERM's toxicological profiles section, Appendix D of the 2005 risk assessment, notes that "Certain forms of silica are known to be more toxic than others," particularly the crystalline form. It points out that crystalline silica "may cause coughing, dyspnea (trouble breathing) and fibrosis in humans, and is considered to have twice the toxicity as that of other forms of silica in causing silicosis."

Intel is allowed by state permit to release to the air up to 14.2 tons of silica particulate each year. Reported emissions have been as high as 12.2 tons per year.

In an interview this summer, Williams said he had examined the apparent inconsistency in Intel's data about the effect of the incinerator's heat on HMDS.

He phrased the questions this way: "ERM states that no crystalline silicon will be formed in the thermal oxidizer although oxidizer temperatures are 1400 degrees and crystalline silica forms at 1000 to 1400 degrees."

Furthermore, "The ERM risk assessment uses a ratio of two between toxicities of amorphous and crystalline silicon, while the actual values should be about 500," Williams said.

The Los Alamos scientist said his research indicated that crystalline silica is 100 to 500 times more dangerous.

As to whether Intel's incinerators produce crystalline silica, Williams concluded they probably do: "Available information suggests that the risk assessment's treatment of crystalline silica is inadequate. The contention that no crystalline silica is emitted is unsupported and quite possibly wrong."

He based that finding on the incinerator manufacturer's data and the properties of silicon dioxide: "Combustion at low temperatures (less than 700 degrees Centigrade (1292 Fahrenheit) gives no crystalline silica, but 800 degrees C (1472 F) can produce crystalline silica. Sustained temperatures above 750 C (1382 degrees F) lead to a significant quantity of crystalline silica in ash in other contexts," he pointed out.

He noted that Intel has claimed actual tests of silica from the incinerators show it to be amorphous, rather than crystalline.

However, that test methodology has been called into question. The silica samples tested were said to have been scraped from parts of the incinerators rather than taken from the air leaving the incinerators.

A new initiative by the Community Envirornmental Working Group, on which Williams serves, is expected to use silica sampling and testing as a pilot study for a "Citizen Protocol" to independently verify testing carried out on Intel's behalf.