IV. Mineralogical Considerations

The following is a discussion of the mineralogical evidence submitted to this record concerning defining and differentiating the types of minerals commonly designated as "asbestos", "asbestiform" and "non-asbestiform". OSHA's position, expressed in the proposal and in the 1986 standards, was that precise mineralogic definitions are helpful in describing the scope of the standard, but absent strong evidence that mineralogic distinctions are biologically relevant, such distinctions by themselves, should not dictate regulatory health based decisions. In the 1986 standards, OSHA defined "asbestos" and "non-asbestiform ATA" separately, but covered both varieties based on health effects evidence.

Much evidence and testimony in this proceeding related to the extent to which different mineral varieties can be distinguished . OSHA's overall regulatory approach to this issue is shaped by its mandate to protect employee health, and to err on the side of protection when presented with a close scientific question. The Agency believes that mere difficulties in differentiating between these mineral varieties should not dictate uniform regulatory treatment, unless such difficulties reflect the fact that the varieties, in biologically relevant respects, behave the same. Of course, misidentification of mineral type affects the confidence in and usefulness of studies reporting the biological potential of different mineral types. Also, the extent of analytical difficulty in distinguishing even well characterized mineral types, would be relevant to OSHA in making feasibility determinations concerning analytic methods.

In general there was agreement concerning the broad definitions of these mineral classifications. Thus, asbestos is not a precisely defined chemical compound, but rather, a collective term given to a group of similar silicate minerals having commercial significance. Historically six silicate minerals have made up the group of minerals which has been collectively referred to as Asbestos. These six minerals are chrysotile, crocidolite, amosite (which is mineralogically known as cummingtonite - grunerite asbestos), tremolite asbestos, anthophyllite asbestos, and actinolite asbestos. Chrysotile belongs to the family of minerals called serpentine minerals. The remaining five minerals belong to the family of minerals called amphiboles.

Dr. Arthur Langer pointed out in his testimony and comments to OSHA, that the definition of asbestos is comprised of a mineralogical definition and an economic geology definition. Langer states:

Asbestos is described in the mineralogical literature as several silicate minerals with the following characteristics: Minerals occurring in nature as fibers; Fibers are bundles composed of "hair-like" (filiform) fibrils, each with a high length-to-width ratio; Fiber bundles are polyfilamentous and the fibril strands may be easily separated by hand. Unit fibrils cannot be resolved by [the] unaided eye; ...In addition to the mineralogical criteria, the economic geology literature contains additional descriptive terms, mostly pertaining to properties exhibited by asbestos which render it useful in commerce. Among these are: fibers exhibit stability in acids and alkalies; act as electrical insulators; act as thermal insulators; fibers are highly flexible and can be woven into asbestos cloth or rope; fibers possess diameter dependent high tensile strength. Together, both geological disciplines have defined what asbestos is mineralogically. (Ex. 517, Tab 5)

Dr. Ann Wylie, testified that "Asbestos is a commercial term applied to a group of highly fibrous silicate minerals that readily separate into long, thin, strong fibers of sufficient flexibility to be woven, are heat resistant and chemically inert, and possess a high electrical insulation and therefore are suitable for uses where incombustible, nonconducting, or chemically resistant material is required." (Ex. 479-23).

Similarly, the Bureau of Mines stated in comments to the NPRM that a correct mineralogical definition of asbestos was:

A term applied to six naturally occurring serpentine- and amphibole-

group minerals that are exploited commercially because they crystallize into long, thin, flexible fibers that are easily separable when crushed or processed, can be woven, are resistant to heat and chemical attack, and are good electrical insulators. The six serpentine- and amphibole-group minerals commonly referred to as asbestos are chrysotile, cummingtonite-grunerite asbestos (amosite), riebeckite asbestos (crocidolite), anthophyllite asbestos, tremolite asbestos, and actinolite asbestos (Ex. 478-6).

The above minerals which are collectively termed asbestos, are also described as being asbestiform. Asbestiform is a mineralogical term describing a particular mineral habit. The habit of a mineral is the shape or form a crystal or aggregate of crystals take on during crystallization and is dependent on the existing environmental/geological conditions at the time of formation. The National Stone Association (NSA) and the American Mining Congress (AMC)state that, "The asbestiform habit can be defined as a habit where mineral crystals grow in a single dimension, in a straight line until they form long, thread - like fibers with aspect ratios of 20:1 to 100:1 and higher. When pressure is applied, the fibers do not shatter but simply bend much like a wire. Fibrils of a smaller diameter are produced as bundles of fibers are pulled apart. This bundling effect is referred to as polyfilamentous." (Ex. 467) Dr. Wylie testified that the asbestiform habit can be recognized by certain characteristics using light microscopy. For example she testified that:

Populations of asbestiform fibers, and this would include all, not just commercial asbestos, but all asbestiform fibers that I have looked at, they have mean aspect ratios greater than twenty to one for particles longer than five microns -- and again, it's very important that we qualify, when speaking of aspect ratio, length, because aspect ratio by itself as a population characteristic has no meaning -- very thin fibrils that are usually less than half a micrometer in width. And you will see in any population of asbestiform fiber[s] at least two of the following characteristics. Normally they are all present, but two, I think is enough to convince me. Parallel fibers occurring in bundles, fibers displaying splayed ends, the matted masses of individual fibers, and fibers showing curvature. (Tr. 5/9, p. 92)

However Dr. Wylie emphasized that these are characteristics which apply to populations of asbestiform fibers and not a particular particle. She states that "The characteristics that were listed were population characteristics, not characteristics on a fiber by fiber discriminator. They weren't meant to say a particular particle must meet all these criteria in order to say that this is an asbestos particle or population present. And that's the way that definition is approached that if we have a bulk sample, and we are looking in that sample for the presence of -- asbestos," (Tr. 5/8, p. 144) In further clarification of the asbestiform habit Dr. Tibor Zoltai, a professor of mineralogy at the University of Minnesota, states that:

The development of the asbestiform properties is a gradual process, [and] depends on the extent of the appropriate conditions of crystallization. Consequently, there are variable qualities of asbestiform fibers. The poor quality asbestiform fibers of amphiboles are called byssolite, or brittle asbestos. The high quality asbestiform fibers because of their highly developed flexibility, strength and physical - chemical durability, constitute desirable industrial materials and are exploited under the generic term of asbestos. Although practically all amphiboles and most other minerals are known to occur in asbestiform habit, only a few amphiboles are known in sufficient concentration and quantity to produce commercial asbestos:... (Ex. 546).

Thus, asbestos is a collective term composed of both mineralogical and economic elements which has been used to refer to a specific set of asbestiform minerals which are, or were in the past, regarded as being commercially significant. The term asbestiform is a mineralogical term used to refer to those minerals which are found in a particular mineral habit. That is, while all asbestos is asbestiform, not all asbestiform minerals are asbestos.

As the above discussion shows, the term "asbestos" is based on more than mineralogic criteria, and its meaning also reflects to a certain extent the interests of the affected commercial communities. Non-asbestiform mineral varieties have a different commercial history. For the most part, they have had little commercial significance. This is related to their different crystallization habit. Because, unlike asbestos, they do not grow unidirectionally, into long thin fibers, therefore they often do not possess properties such as weavibility or high tensile strength which make them valuable for asbestos - like uses. For the most part non-asbestiform minerals are not mined for any special property, but rather, they are mined generally with other minerals as a basic stone product. However, non-asbestiform tremolite when mined with talc, results in enhanced usefulness to industries such as ceramic manufacturing, because of the other properties specific to non-asbestiform minerals.

The record makes clear, that from a mineralogic perspective the crystallization growth pattern of these minerals determines whether they develop as asbestos, or as non-asbestiform varieties. In joint comments to the record, the NSA and the AMC stated that "in the non-asbestiform variety crystal growth is random, forming multi-dimensional prismatic patterns. When pressure is applied, the crystal fractures easily, fragmenting into prismatic particles. Some of the particles or cleavage fragments are acicular or needle - shaped as a result of the tendency of amphibole minerals to cleave along two dimensions but not the third" (Ex. 467).

In his comments to the record, Dr. Zoltai notes that: Both asbestiform and non-asbestiform amphibole minerals have the same chemical composition and crystal structure. They are not distinguishable by instrumental analysis and x-ray diffraction. The difference between them is in their respective crystallization habit, that is, in their respective condition of crystallization. Non-asbestiform prismatic crystals are the common crystal habits of amphiboles. The asbestiform crystallization habit is the unusual one, it requires unique temperature and pressure conditions inducing unidirectional and rapid crystal growth. (Ex. 446)

In the NPRM, OSHA stated that unlike asbestiform minerals, non-asbestiform minerals do not separate into fibrils but, during processes such as mining, milling and/or processing can be broken down into fragments resulting from cleavage along the minerals two or three dimensional plane of growth. OSHA also stated that particles thus formed, are generally referred to as cleavage fragments and these fragments may occur in dimensions equal to asbestiform fibers.

Various commentors agreed with OSHA's definition of a cleavage fragment but objected to OSHA's characterization that non-asbestiform cleavage fragments and asbestiform fibers occur in similar dimensions. In testimony to OSHA, Kelly Bailey, an Industrial Hygienist with Vulcan Chemical Company speaking for the NSA stated:

The NSA believes that this statement is deliberately misleading in that it fails to take into account the population characteristics of both cleavage fragments and asbestiform fibers. It is true that there are some cleavage fragments that may have dimensions of 10:1, 20:1 or higher in aspect ratio when examined with PCM and that there may be a few asbestos fibers that have low aspect ratio dimensions similar to cleavage fragments: however, to imply that cleavage fragments do not differ from asbestiform fibers in an observable, dimensional way is poppycock! (Ex. 479-23).

Similarly, in earlier testimony to OSHA during the rulemaking for the 1986 revised standards, Dr. Wylie stated :

A particle of any mineral which is formed by regular breakage is called a cleavage fragment. Mineralogically, a fiber or fibril is a crystal which has attained its shape through growth, in contrast to a cleavage fragment which has attained its shape through regular breakage. The shape of amphibole cleavage fragments is somewhat variable depending upon the history of the mineral sample. Some amphiboles when crushed will produce a population of particles which may have the average aspect ratio of 5 to 1 or 6 to 1, whereas other amphibole samples when crushed may produce a population of particles whose aspect ratios average closer to 8 to 1 or 10 to 1. And in almost any population of amphibole cleavage fragments, it is possible to find a few particles whose aspect ratios may extend up to 20 to 1 or perhaps even higher. Amphibole asbestos populations, on the other hand, are characterized by aspect ratios which are considerably greater than this." (Ex. 230, Docket # H-033c).

Dr. Ann Wylie reiterated her earlier opinions in the current rulemaking stating:

Throughout OSHA's Notice of Proposed Rulemaking, they imply that cleavage fragments are similar in size to asbestos fibers, and the distinctions between them are fuzzy. In most cases, this is simply not so. Asbestos crystallizes from a fluid medium; growth takes place rapidly in one direction; the chemical makeup of the fluid may inhibit growth laterally. ...These fibrils are single or twin crystals and they have very, very narrow widths and long lengths. It is the narrow width and long lengths that give asbestos flexibility and high tensile strength. Fibrils share a common axis of growth, but they are randomly [ar]ranged in the direction perpendicular to the fiber axis, and when disturbed, they are easily desegregated. Because their origin is different, population of cleavage fragments and fibers of the same minerals are simply different. Dr. Wylie adds that: While, there may be some cleavage fragments that cannot be distinguished from asbestos solely on dimensions, and there are some particles in asbestos samples that can't be distinguished from cleavage fragments, the populations are as wholes easily distinguishable. (Tr. 5/9, pp. 102-103)

As evidence of these differences Dr. Wylie cited to her paper entitled "An Analysis of the Aspect Ratio Criterion for Fiber Counting". Dr. Wylie testified:

As a part of the record, I have prepared a paper entitled "An Analysis of the Aspect Ratio Criterion for Fiber Counting: and that is part of OSHA's record. The paper reviews the distribution of aspect ratio for fiber and fiber bundles of amosite, crocidolite, chrysotile, and they clearly show that for those fibers and fiber bundles, again, that are longer than five micrometers, 100 percent or close to it, have aspect ratios greater than ten to one, and in every population that I have ever looked at that has the asbestiform habit, more than 50 percent have aspect ratios in excess of twenty to one...but most of them are 90 percent.

Also included in that paper are data from bulk and airborne samples of cleavage fragments, and there are cleavage fragments [with] aspect ratios greater than ten to one, and there are some that have aspect ratio[s] greater than twenty to one, but they are in much lower abundance, as a population. (Tr. 5/9, pp. 94-95)

While Dr. Wylie notes that there are differences in the distribution of aspect ratios when one looks at populations of asbestos fibers and non-asbestiform cleavage fragments, she also states that "aspect ratio is a dimensionless parameter" and "..it lacks information about the size particles; it only describes shape." (Tr.5/9, p. 95). Rather than aspect ratio, Dr. Wylie stressed that "width is a much more fundamental parameter of asbestos fibers, and perhaps will shed some light on how we tell particles that are elongated, whether they are cleavage fragments, or whether they are asbestos." (Tr. 5/9, p. 95).

To illustrate this point Dr. Wylie presented data in her testimony on the widths of various populations of asbestos fibers and non-asbestiform cleavage fragments from both bulk and airborne data (Transcripts, May 9, pp. 2-95 to 2-98). This data showed that in the populations of asbestos fibers she studied, the majority of fibers had widths less than one micrometer. For example, 85-90 percent the crocidolite fibers she studied had widths less than one micrometer and 60 percent had widths less than 0.5 micrometers. In amosite samples, greater than 90 percent had widths less than one micrometer and 75 percent had widths less than 0.5 micrometers. In tremolite asbestos samples, 85-95 percent of the fibers had widths less than one micrometer and 75 percent had widths less than 0.5 micrometers. Wylie stated that when looking at these fiber populations "... it really doesn't make any difference, much, whether you look at particles longer than five micrometers, or all particles in a population, when you look at width. Because of the nature of asbestos, width changes very little as length increases,..." (Transcripts May 9, p. 2-96). Dr. Wylie acknowledged, however, that asbestos fiber bundles may have widths greater than one micrometer, but she added that even in these cases the majority of particles are less than one micrometer.

Dr. Wylie was criticized for inconsistencies in her comparative population: i.e. sometimes using all fibers, other times citing only those exceeding certain dimensions, e.g.longer than 5 micrometers. Dr. Wylie agreed that "depending upon which of those qualifiers you put forth, you get vastly different datasets. Now, I took all my cleavage fragment data and I first looked at the particles that are longer than five micrometers, and of these -- I'm just going to use a ten to one as aspect ratio -- 11 percent have aspect ratios greater than ten to one. If we look at that dataset... and only at the particles that have aspect ratios greater than three to one... and are longer than five micrometers, then we would say it's six percent are longer than five micrometers and have aspect ratios greater than ten to one. And finally if we look at particles that are both longer than five micrometers, and have an aspect ratio greater than three to one, we have 19 percent with aspect ratios greater than ten to one."(Tr. 5/9 at 106-107).

The record contains some additional, but less comprehensive evidence on comparative dimensions of non-asbestiform cleavage fragments and their asbestiform analogues. For example, in 1979, the Bureau of Mines compared 8 samples of ground tremolite of varying habit. It concluded that "based on this limited study, there is a relationship between the number of particles of "critical" dimensions, >10 um in length and < 0.5 um in width, and the habit of the tremolite - actinolite prior to grinding. ... Only the asbestos variety gave long, thin particles of the dimensions established by some medical scientists as necessary to produce adverse biological effects in laboratory animals. " (See RI 8367, p.17 as part of the NIOSH pre hearing submission Ex. 478-15).

A critical dimensional distinction between asbestiform fibers and ATA appears to be their widths. Thus, Dr. Wylie stated that her analyses of width show that "About 80 percent of the amphibole cleavage fragments longer than five micrometers, have widths greater than one micron, and none have widths less than 0.25." (Tr. 5/9, p. 98) Dr. Wylie also pointed out how the width of asbestos fibers will influence their aspect ratio. She states that "the mean width of asbestos fibers is less than half a micron, and if you have five micrometer particles, you have to have an aspect ratio of at least 10 to 1. (Tr. 5/9, p. 101-102). Moreover in her comments to the NPRM she states that "while low aspect ratio fiber (or fiber bundles) are present in asbestos populations, they are characteristic of short asbestos fibers...Since the mean width of asbestos fibers is less than 0.5 micrometers, the mean aspect ratio of a 5 micrometer fiber is about 10:1." (Ex. 479-23).

Dr. R.J. Lee, a microscopist and mineralogist with R.J. Associates, also noted the importance of width in distinguishing asbestos fibers from non-asbestiform cleavage fragments. Dr. Lee testified the following:

First asbestos -- airborne asbestos is less than one micrometer in diameter, unless it's present as bundles or cluster, which exhibit the characteristic fibrillar structure of asbestos, or as Dr. Wylie indicated, the hallmark of asbestos. Asbestos larger than a half a micron is a bundle --.

Second, non-asbestos particles longer than five micrometers in length are generally [more] than one micrometer in diameter, and only rarely less than half a micrometer in diameter. When larger than one micrometer in diameter, they do not exhibit the fibrillar structure of asbestos. (Tr.5/9 , pp. 114-115).

Similarly in their joint comments to the record the NSA and the AMC stated the following observations about particle width:

Due to the straight line fibrillar crystal growth of asbestos, the width of an asbestos fiber is essentially independent of its length and is not easily altered by processing. In contrast, cleavage fragment populations show increasing width as particle length increases due to the characteristics imparted from normal three dimensional crystal growth. The result of this difference is cleavage fragments with widths rarely less than 0.5 micrometers and almost never less than 0.25 micrometers. Asbestos tends to show a high proportion of fibers less than 0.25 micrometers in width. (Ex. 467)

Dr. Charles Spooner, a microscopist and mineralogist with Charles Spooner and Associates Inc., concurred in his testimony that asbestos fibrils have widths less than 1 micrometer and that most cleavage fragments have low aspect ratio (Tr. 5/8, pp. 120-121). However he also noted that cleavage fragments may also have high aspect ratios. Dr. Spooner stated that "In the universe of amphibole cleavage fragments it seems likely that a greater proportion will exist as more or less equant bodies, however there will be those instances where high aspect ratio respirable cleavage fragments will be generated upon crushing of the amphibole bearing rock." (Ex. 512).

As noted earlier in this discussion, Dr. Wylie acknowledged that one may find a few cleavage fragments with high aspect ratios, but she added that populations of asbestos fibers and cleavage fragments, as a whole, are distinguishable from one another. However Dr. Spooner points out that "...from the industrial hygiene perspective, very often we are dealing with air samples. We are looking at an airborne fiber and trying to assess its respirability. And again, we are often in the industrial hygiene setting, we don't have the opportunity to know where the material is coming from, nor do we have the opportunity to look at a very large population of fibers..." (Tr. 5/8, pp. 117-118). Thus OSHA believes that while one can differentiate between mineral types when populations of particles are examined, when single, isolated particles are examined (e.g. particles from air samples) the ability to differentiate may become more difficult.

In the NPRM OSHA stated that at the microscopic level, on a particle by particle basis, differences in gross growth characteristics may not be readily observable. Similarly, Dr. Art Langer acknowledged that "...in some instances single, isolated particles may be impossible to distinguish, i.e. acicular cleavage fragment from asbestiform fibril" (Ex. 517, Tab 5). Dr. Langer also noted however that while there are some particles which defy mineralogical identification, the percentage of particles that comprise this group is a small percentage (Tr. 5/11, p. 230).

Identification of fibers is confounded by the existence of particles which do not fit a precise mineralogic definition. For example, some samples of industrial talc have been shown to contain "intermediate fibers." Dan Crane, a microscopist at OSHA's Salt Lake City Technical Center, describes these intermediate fibers which are found in industrial talc samples and notes that "It is only by a combined optical/electron optical approach can the nature of the intermediate fibers can be determined. Even at that, they defy definite description." (Ex. 410-23). Mr. Crane goes on to explain that:

When one looks at the industrial talcs in the microscope, he sees large numbers of particles that are much longer than 20 to 1 even to nearly 100 to 1 in aspect ratio. The first reaction is to say these are the asbestos fibers of tremolite and anthophyllite indicated by the known presence of those minerals in the products. Unfortunately, this is a false assumption. They are for the most part fibers of industrial talc. They have been dubbed intermediates by us, as talcboles by Malcom Ross and fibrous biopyriboles by David Veblan. What they are not is anthophyllite or tremolite. (Ex. 410-23)

In his description of these intermediate fibers Crane notes that examining these particles by light microscopy (e.g. using indices of refraction and dispersion oils) one would not call these particles anthophyllite. However when one use electron microscopy one would conclude that these particles are indeed anthophyllite. Mr. Crane explains why this difference occurs:

The fault can be corrected when the analyst realizes that in this particular mineral, the deposit was anthophyllite at one time. The particular mechanics of this are beyond the scope of this letter. Suffice it to say that it is being done in such a way as to leave the more major structure of the anthophyllite fibers intact while transforming them to talc. This residual structure has given rise to electron diffraction patterns that mimic amphibole patterns. Very careful measurement and calibration of these patterns reveal subtle strains in the structure leading to a mineral with similar features to talc and to anthophyllite and yet the numbers fall in between. ...I have described these other fibers because they are the fibers with the closest morphological similarity to asbestos. They do have splintering and bundle of sticks and frayed ends as characteristics. These are characteristics which we often ascribe to truly asbestiform minerals. All the samples we have examined have been crushed prior to our receiving them. Therefore, we cannot say whether they grew in nature as asbestos fibers. They do look like asbestos and if morphology is the major role in toxicity or carcinogenicity these should be considered more important tha[n] the non-fibrous cleavage fragments of tremolite and anthophyllite. (Ex. 410-23)

Dr. Arthur Langer, in his testimony, also discussed the difficulties in identifying these intermediate fibers. He stated that:

...some of us might call this a pyrobole, pyroxene and amphibole. This has also been described in various deposits, and you're going to ask me about the Vanderbilt talc deposit. That's fine because they're intergro[wthes] like this in the Vanderbilt talc deposit. These are the complex fibers that we have talked about that defy mineralogical classification. (Ex. Tr. 5/11, pp. 170)

The significance of "intermediate" or "transitional" fibers was also addressed by Dr. Langer, who stated that OSHA's major question should be "how common are they in the work place?" and answered "I don't think they're terribly common in the work place. They are only described in certain specific locales." (Tr. 5/11, p. 219).

OSHA notes that even those mineralogists who contend that asbestos is a separate mineral entity from non-asbestiform ATA, agree that intermediate forms exist. Dr. Tibor Zoltai, Professor of Geology at the University of Minnesota, explained that "...(T)he development of asbestiform properties is a gradual process, (and) depends on the extent of the appropriate conditions of crystallization. Consequently, there are variable qualities of asbestiform fibers. The poor quality asbestiform fibers of amphiboles are called byssolite, or brittle asbestos. The high quality asbestiform fibers, because of their highly developed flexibility, strength and physical - chemical durability, constitute desirable industrial materials and are exploited under the generic term of asbestos."(Ex. 546). Dr. Langer testified that based on Dr. Wylie's work, it is known that byssolite is not composed of unit fibrils. "So we would not classify byssolite as an asbestos mineral. Now some people consider this as a transition kind of mineral in characteristics."(Tr. 5/11 at 518) Other mineral forms exist which are intermediate between anthophyllite and talc, as discussed above.

In summary, the discussion indicates that populations of fibers and populations of cleavage fragments can be distinguished from one another when viewed as a whole. For example one can look at the distribution of aspect ratios or even widths for a population of particles and can then generally identify that population of particles as being asbestiform or non-asbestiform. However when one looks at individual particles, (e.g. particles from air sampling filters) sometimes these mineralogical distinctions are not clear. Unfortunately the data in the record is insufficient at this time to precisely determine how often these situations occur.

The record also describes the presence of various kinds of "intermediate" fibers, which "defy mineralogic classification". Various participants have requested OSHA to base its regulatory decisions on precise mineralogic definitions. Clearly, any significant presence of mineral types which "defy classification", would defeat such an approach. Although these transitional fibers exist OSHA does not believe that independent evidence of their health effects exists which would support regulation. Dr. Langer testified that there are some fibers which "defy mineralogical identification" but they are a "small percentage" (Tr. 5/11, p. 230) Thus, although their presence lends credence to the explanation that asbestos minerals and non-asbestiform varieties developed on a continuum it does not change the fact that for most mineral deposits, asbestos and non-asbestiform habits are distinguishable.

OSHA finds, based on this record that while these intermediate fibers do exist, the record indicates that they are minor constituents of most mineral deposits. In general, when observed in their natural habit of growth, the two habits of asbestiform and non-asbestiform minerals are distinctly different. The record also indicates that populations of particles derived from mining, crushing or processing these minerals, are also distinctly different (e.g. in the distribution of widths and aspect ratios). However on an individual particle basis, which is often the case for particles from air monitoring samples, these distinctions may become less clear. The record indicates that there are situations where individual particles of asbestiform and non-asbestiform minerals may be indistinguishable. These situations are likely to be rare in occupational contexts but OSHA has little information upon which to make such a determination.

The regulatory implication of these findings are as follows: Several participants suggested that all forms of asbestos and their non-asbestiform analogues should be treated as a single mineral entity for purposes of regulation because the forms of ATA cannot be distinguished, and there is no clear mineralogic dividing line between various varieties of ATA. Dr. Charles Spooner, a witness for OSHA, a geochemist, a mineralogist and an industrial hygienist, in response to a question concerning how his laboratory distinguishes asbestos from fibers that are not asbestos, stated that "at this point if we identify the mineral tremolite, we make no distinction on the basis of fiber." (Tr. 5/8, p. 119). Dr. Spooner's post - hearing submission again noted that distinguishing asbestiform and non-asbestiform cannot be made reliably either on the basis of a hand sample or microscopic examination: hand - specimen characterization of mineral habit does not necessarily carry over to mineral habit on the micro scale; and, on the micro scale, high - aspect ratio cleavage fragments and asbestiform fragments can co-exist. Dr. Spooner recommended that "the issue must be resolved on the basis of biological activity and aspect ratio of the respirable fibrous bodies." (Ex. 512).

Dr. Bruce Case, in a letter to the British Journal of Industrial Medicine, November, 1990, provides a clear summary of the mineralogic argument for considering asbestiform ATA and non-equant non-asbestiform ATA to be a single substance for purposes of regulation:

The major flaw in the substitution of mineralogical definitions for microscopical characteristics is a reliance of the former on gross morphology. For regulatory and health assessment purposes, it is microscopical morphology that counts: there is no evidence that potentially affected cells can distinguish between "asbestiform" and "non-asbestiform" fibers having equivalent dimensions. The lack of agreement as to what is and what is not"asbestiform" tremolite would be less critical if those who advocate such a definition could show that there is a clear line between the two forms when they present `fibrous' morphology. Unfortunately, this is not the case. Pooley has noted that the differences in structure between massive, acicular and fibrous morphology are not "sharply defined", but rather represent points on a continuum. So-called cleavage fragments may, in a strict morphological sense, be fibrous in their appearance in microscopic fields, and there is no convincing evidence that these `fibers' are of no public health concern. (Ex.529.4)

The ATS's report also concluded that mineralogic distinctions between different forms of anthophyllite, actinolite and tremolite were not clear: "It became apparent both from our review of the literature and from submissions made to this committee by experienced mineralogists, that the distinction between cleavage fragment and asbestiform fibers, although theoretically clear, is in practice extremely murky." (Ex. 525 at 3) As noted above, other participants took issue with these statements. In particular, in a post - hearing submission, the R.T. Vanderbilt Company directly took issue with the ATS statement quoted above as follows: "(a)t the OSHA hearing, Dr. Wylie, Dr. Langer and Mr. Addison explained that the distinctions at issue were in no way "murky" (theoretically, practically or otherwise). While we do not disagree that some gray areas exist (i.e., at the single crystal level), the important day - to day distinctions at issue in this rulemaking simply do not fit this "murky" characterization".(Ex. 529-6 at 3). Other presenters made similar statements.(See e.g. testimony of Dr. Wylie at Tr.5/9, at 103 and Dr. Lee at Tr. 5/9, at 1).

OSHA has determined that non-asbestiform ATA and asbestos anthophyllite, actinolite, and tremolite should be defined separately for regulatory purposes to conform to common mineralogic usage. As discussed above, the testimony of Dr. Wylie, Dr. Langer, Dr. Nolan, Dr. Campbell, the Bureau of Mines and others agreed that populations of asbestos and non-asbestiform ATA are separate mineral entities, which for the most part have widely diverging population characteristics which are the result of its habit of crystallization in nature. In addition, these characteristics, such as high fibrosity, fiber shape and size, and easy separability appear to be biologically relevant in producing disease. The agency notes the position it adopted in the 1986 standards, where it stated: "(t)he Agency recognizes that the minerals tremolite, actinolite and anthophyllite exist in different forms", and therefore required that warning signs and labels for ATA need not include the term "asbestos" (See 51 FR at 22679, 29 CFR 1910.1001 (j)(2)(iii), 1926.58(k)(1)(iii), recognized the mineralogic distinctions, but did not distinguish the minerals based on biologic effects. Thus, the difference between the Agency's 1986 and its current positions is not mineralogical and as explained above, is related to its view of the health effects evidence. Thus although the Agency now reaches a different conclusion than it did in 1986 concerning the evidence of health risks of non-asbestiform ATA, it continues to believe that the mineralogic forms are sufficiently distinctive to be treated differently for regulatory purposes. Also, unlike its determination in 1986, which was based on a far less extensive review of health effects evidence, the Agency now finds that differences in biologic effect between asbestos and its non-asbestiform analogues are likely related to the distinctions which define the two groups as separate mineral entities.

[57 FR 24310, June 8, 1992]