The GHS has no testing requirements; therefore, in converting MSDSs to SDSs or producing hazard labels, companies are not required to perform lab tests to determine what hazards a chemical poses. Instead, they can rely on past tests and use old data to quantify the hazard potential of each substance.

This system works well for pure substances for which test data is available, but it does not apply to mixtures. For mixtures, companies can use what are known as bridging principles, which are rules for determining new hazard classifications. However, in a presentation at the STLE Annual Meeting in May, Howell warned this approach may result in a more severe classification than necessary. In contrast, actual testing can help avoid overly severe classifications for skin corrosion and irritation, eye irritation and skin sensitization.

In addition, he noted, This work is complicated by the fact that upstream suppliers have been slow to develop safety data sheets for their products, creating a data gap.

Overcoming the Data Gap

Absent test data for mixtures, compounder/blenders must classify their products for hazards based on component data. Hazard classification data can be found on the European Chemicals Agency Classification and Labelling Inventory, said Howell, but little data is available on certain components critical to the lubricants industry like emulsifiers and alkanolamine organic acid reaction products.

So, what do we do? Howell offered several strategies to get the ball rolling. The first strategy is product rationalization, he said. If you havent rationalized your product lines recently, do it now. This will greatly reduce the number of products that must be tested and classified.

Second, if you havent already, immediately classify the top 20 percent of your products according to the mixture rules in Appendix A of 29 CFR 1910.1200, advised Howell. These products likely account for 80 percent of sales. Refer to the ECHA C&L inventory for classification information.

Third, chemists should begin thinking of their formulations in terms of already neutralized reaction products, Howell advised. That way, regulatory affairs personnel wont have to search for Chemical Abstract Service (CAS) numbers for reaction products that may not be listed on critical national chemical inventories.

Fourth, seriously consider product or component testing as a strategy. But, understand that it is going to cost a lot of money! Despite, that, he said, Testing is surprisingly affordable.

Tiered Approach

Because mixtures are combinations of several different components, hazard test data for a particular mixture typically is not available. If data is available, great, Howell said. If not, use the bridging principles outlined in Appendix A where they apply. If bridging principles do not apply, hazards must be estimated.

To illustrate this approach, Howell went through the process of determining a classification according to the GHS C&L guidelines for the skin and eye corrosion/irritation for a product based on the criteria in Tables A.2.3, A.2.4 and A.3.3 in Appendix A. He then described how the label for the example product would be constructed.

After going through the process, Howell asked, Are we better off if we test the product for eye and skin corrosion/irritation and for skin sensitization? The answer is, Probably yes because using the principles outlined in Appendix A often results in a more severe rating than actual testing would produce.

He explained that there are two testing strategies. The first is to test the components in your products to assess the effects of likely irritants for which you do not have data. The second strategy is to test finished products. In this approach, Howell said, You assess key products; namely, those that have the highest concentrations of components that are or that are likely to be irritants. Then, apply bridging rules.

Bridging rules are explained in detail in 29 CFR 1910.1200, Appendix A. They cover a variety of situations, including classification of mixtures where test data is not available for the complete mixture, dilutions and concentrations of mixtures. They also cover how to interpolate data within one toxicity category and how to handle substantially similar mixtures.

Bridging rules are very specific, Howell said. You cant just test one product in your product line and assume all other formulations with different components are OK. However, by working within the rules, he added, you can extrapolate data from tested formulations to other similar, but untested, formulations.

Testing Protocols

Formulators often ask where to find information on what test protocols to use, Howell said. He noted that ASTM E1302-13, Standard Guide for Acute Animal Testing of Water-Miscible Metalworking Fluids is a useful reference for test methods referenced in U.S. regulations. Also, the Organization for Economic Cooperation and Development (OECD) has issued a number of protocols, covering acute dermal irritation/corrosion, acute eye irritation/corrosion and skin sensitization.

In irritation testing, there is an assumption of additivity for the action of two or more irritants in a mixture; that is, the irritant potential of each component is added together to determine the rating for the entire mixture. However, Howell noted, Appendix A addresses some simple cases where additivity does not apply.

He cited a simple example in which two irritants are used alone or in combination up to a maximum concentration of 5 percent in a formulation. Test data can be generated on three mixtures: 5 percent of either irritant in the matrix and 5 percent of the two irritants together in the matrix.

This testing will indicate whether either substance triggers the irritation hazard at a 5 percent concentration and whether results for the combination are more severe than the most severe result with either irritant alone. This would justify whether or not you apply the additivity rule for combinations of the two irritants, Howell concluded.

This approach applies to acute hazards, especially skin corrosion and irritation; eye irritation Classes 1, 2A, 2B; and skin sensitization, he said. However, it is impractical and not scientifically valid for chronic hazards.

As stated in Appendix A, It is important to note that the OSHA rule specifies an exception … for carcinogenicity, germ cell mutagenicity and reproductive toxicity. For these three hazard classes, mixtures shall be classified based upon information on the ingredient substances, unless on a case-by-case basis, very rigorous justification can be provided for classifying based upon the mixture as a whole.

For complex mixtures like [finished lubricants], it may be more straightforward to test the worst case product; that is, the one with the highest concentration of irritants, Howell stated. This approach may help confirm that a large part of a product line is irritating or not. If the results for the worst case mixture warrant an irritant classification, you can step down and test lower concentrations until you find a level that is nonirritating.

This approach is especially useful with combinations of irritants that are acids and bases. For example, Howell said, it may be possible to demonstrate with a modest amount of test data a pH range where the irritation potential of the acids and bases is neutralized.

Skin Sensitivity

The GHS regulation contains new definitions for Class 1A and 1B skin sensitizers. The rule specifies a 0.1 percent concentration threshold for Class 1A and 1.0 percent for Class 1B, Howell said. Under the earlier version of the Hazard Communication Standard, sensitizers were not specifically identified as a health hazard. Fortunately, a lot of information on new sensitizers is now available.

For example, the European Union has defaulted to Class 1A for all sensitizers that have a mandatory classification. The list includes about 1,000 chemicals, and many commonly used biocides are listed as Class 1A skin sensitizers, Howell noted.

In addition, the EU Classification and Labeling Inventory lists hazard classifications for a variety of nonpolymeric chemicals. And more than 8,000 chemicals are listed in the new Local Lymph Node Assay in REACH registration dossiers.

Skin sensitization follows the same rules as common allergens such as poison ivy, Howell said. The first exposure may not affect your skin but sensitizes your immune system, which leads to skin rashes of various types when you are exposed a second time.

The recommended testing guidelines for skin sensitizers or allergens used guinea pig tests for decades, said Howell. More recently, the Local Lymph Node Assay in mice was recommended because it is faster and uses fewer animals.

This move was first viewed as a positive step among toxicologists because guinea pig tests did not yield consistent results between labs or, sometimes, from week to week. However, after about a decade of experience with the lymph node assay, the prime discussion topic at toxicology meetings is the alarming false positive rate for a growing number of chemical classes, as well as a good deal of variable and inconsistent results similar to those for guinea pig testing, Howell cautioned. To illustrate the point, he compared animal test criteria for skin sensitization, showing that the rating criteria vary significantly depending on the test used.

To obtain reasonable results in skin sensitization testing, Howell emphasized, Ask your suppliers to provide data that correctly classifies particular chemicals as Class 1A or Class 1B. Also, get expert guidance from an independent lab experienced in sensitization testing.

Another important point is to assess the best diluent system for water-based and oil-based lubricants. For example, an emulsifiable oil, also known as a soluble oil, may yield more consistent test results when diluted in an oil matrix rather than a water matrix. Also, use the maximum concentration possible in a formulation for these tests.