Using data from non-CLP Safety Data Sheets
Mixture data is not as readily available as substance data, and the main source is from non-CLP Safety Data Sheets. (If the mixture has a CLP Safety Data Sheet is available, it has already been classified for CLP!).
Data from non-CLP Safety Data Sheets (SDSs) can be used to classify either substances or mixtures for CLP, as long as the data is valid for CLP, or can be converted into an equivalent format.
GHS SDSs follow the same format as the current EU (REACH) SDS, which means that relevant information is available as listed here: https://ttenvironmenta.wpengine.com/clp-knowledgebase/safety-data-sheet-layout/ .
CHIP-only SDSs use the original 16-section CHIP layout:
- Composition/ information on ingredients – Section 2
- CHIP classification – Section 15
- Physical and chemical properties – Section 9
- Stability and reactivity – Section 10
- Toxicological information – Section 11
- Ecological information – Section 12
- Transport classification – Section 14
Old format USA SDSs (and Canada, which tended to copy the USA), and other non-GHS SDSs are likely to have a different format, and you will need to read them carefully to extract relevant information.
Once you have found data in the non-CLP safety data sheet, it needs to be checked, and if necessary:
- converted into the units used within CLP (SI units rather than Imperial; and the relevant units for a test method, eg mg/m3 rather than ppm)
- reviewed to see if it is relevant for CLP (the test method may be out of date or inapplicable)
- checked to see whether it needs to be converted into a different format, e.g. you have toxicology results from a 1 hour exposure when CLP requires a 4 hour exposure. (Details on how to convert test results made over different timescales are often given within the CLP regulation as part of the classification information for each hazard, and you should refer back to the regulation itself if you are in this situation).
For example, you may have a flash point generated using the open-cup method, whereas CLP requires a closed-cup method. There can be differences between the two methods of several degrees, as open-cup flash points mimic a spill in the open, whereas closed-cup flash points mimic ignition in a confined space such as the head-space of a vessel. You might decide to use the open-cup flash point in the absence of closed-cup flash point data, and subtract e.g. 5 degrees C to account for the difference in test methods.
When you are happy that the data is in the correct format, it can be used to classify the hazards from first principles, see: https://ttenvironmenta.wpengine.com/knowledgebase_category/first-principles-clp-classification/ .
When you are making these types of assessments, you should keep any original SDSs (and labels, where relevant), and all of your reasoning as part of the classification process.