Machinery Lubrication magazine published by Noria Corporation
Issue link: http://www.e-digitaleditions.com/i/290460
analyzers are highly automated and can offer amazing precision, which enables test flask volumes to be reduced. Another advantage of GC analysis is that it allows the extraction of residual oil at the end of the test through the use of ordinary, non-toxic hydrocarbon solvents. However, such solvents are not helpful for introducing the oil sample into the aqueous test medium. An effective biodegradation test must provide equally good dispersion conditions for all types of materials and products. The iden- tification of such a "co-solvent" was an important step in the development of the new test method. This was followed by the exchange of mercury chloride in the abiotic (non-biological) "poisoned flasks" in favor of a mild but efficient organic biocide. As prescribed for every CEC test-develop- ment process, the new procedure had to prove sufficient repeatability and discrimination with two reference oils. In the first phase of development, five repeat tests with two refer- ence oils were conducted in each of the participating laboratories. Very good repeat- ability was seen in one of the cooperating labs, and clear discrimination of the two tested products was evident in all three laboratories. The average repeatability of all three labs was better than in the CEC-L-33 test. Round-robin Exercise The second phase of the CEC test-development process involved the demonstration of the method's reproducibility in a minimum of five test labora- tories. Three additional labs with suitable equipment and skills were located. The active labs in this phase included three independent commercial test labs, one additive supplier, one oil supplier and one governmental lab. A round-robin exercise was then conducted according to the guidelines set by the CEC with guidance from the Statis- tical Development Group (SDG). Each lab tested four oils in parallel. One month later, the same oils were tested again using fresh inoculum. This would provide a final judgment on the new test method's discrimination, repeatability with the same inoculum (batch of micro-or- ganisms), repeatability with another inoculum in the same lab and reproducibility in six different labs using six different inocula. The average biodegradability results of the six labs were in the range of 35 to 40 percent with mineral base oils and formulations, 70 to 90 percent with synthetic base stocks and oils, and 90 to 100 percent with vegetable oil prod- ucts. These results showed clear discrimination of high and low biodegradability, with ranges similar to many results associated with the CEC-L-33 test. The repeatability of the four samples was calculated at 4 to 7 percent using identical inoculum and 7 to 14 percent using different inocula, with 12 to 19 percent reproducibility. The best figures were seen at high degrada- tion, and the worst at lower degradability. This was not surprising considering the degradation/time curve in which the daily GC RL110 RL245 36% 71% 36% 75% 34% 74% 36% 72% 36% 72% 38% 72% 35% 69% 34% 71% 34% 74% 41% 72% 38% 70% 38% 74% 39% 74% 37% 74% 39% 75% Data from one lab: 5 repeat tests with both reference oils Results 35% 73% 37% 72% 34% 71% 39% 72% 37% 74% Standard Deviation 1.7% 1.3% Overall TDG-L-103 Results Repeatability Low Ref. Oil RL 110 High Ref. Oil RL 245 5.6% 3.2% Average: 4.6% CEC-L-33: 6.7% 4.5% Average: 5.6% Results of the repeatability exercise in phase one of the test-development process CEC-L-103: Repeatability Exercise www.machinerylubrication.com | March - April 2014 25