Celite and filter paper, to the 200-ml blender jar 

 plus about 15 g anhydrous MgSO, and 75 ml I PA. 

 The desiccating step with MgS04 is not effective if 

 volumes of IPA are excessive. In addition, all 

 volumes should be maintained as specified to allow 

 rinsing without exceeding the capacity of the 

 250-ml evaporating flask. The IPA should be 

 measured in the liter graduate and shaken or 

 rotated to wash cylinder. Blend at high speed for 5 

 min, then pour contents of blender jar into 150-ml 

 dry filter funnel (M-porosity), apply vacuum until 

 dripping ceases, rinse briefly with PE (wash 

 bottle), then repeat extraction with 75 ml PE. The 

 second extraction with PE removes about 2.5% of 

 the total oil. 



Quantitatively transfer filtrate to a pre-dried 

 and weighed 250-ml 24/40 standard taper round- 

 bottom flask, and flash evaporate using a rotating 

 vacuum evaporator and warm water bath. This 

 method takes from 5 to 10 min, but other tech- 

 niques of evaporating would be suitable. When 

 solvents are removed, add about 10 ml PE to 

 determine if water or solid materials are present. 

 If clean, evaporate to dryness, wipe outside of 

 flask, and place in drying oven for exactly 30 min 

 to remove traces of solvent or water. Cool in air for 

 1 h and weigh. Subtract tare weight and record 

 weight of oil directly as milligrams per liter. The 

 common practice of storing the dry flasks in a 

 desiccator was not necessary because there was 

 little change in weight with subsequent exposure 

 to air. The oil apparently reached nearly constant 

 weight (oxidation) during the 0.5-h drying step. 

 Exposure of the dry oil and the flasks to air for 15 

 and 40 min resulted in 2.2 and 2.6 mg gain in 

 weight for 1,684 mg oil and only 3.2 mg gain with 

 overnight exposure. Consequently, because the 

 250-ml round-bottom flasks were difficult to weigh 

 in a rapid manner, weights were obtained after 

 oven drying for 0.5 h and air cooling for 1 h. 



If the above PE solution is not free of water or 

 solid particles, add 10-15 g anhydrous sodium 

 sulfate and sufficient PE to mix well. Let sit a few 

 minutes, and filter through sodium sulfate on a 

 60-ml medium- or fine-porosity fritted-glass fun- 

 nel, rinse with PE, and transfer back to evaporat- 

 ing flask. The pre-weighed 250-ml flask should be 

 washed out with water and solvents before reuse. 

 This step is time-consuming and is never neces- 

 sary if the previous extraction and desiccating 

 steps are done properly. 



Accuracy and Precision 



The results of replicate analyses on eight 

 effluent samples indicate that the proposed meth- 

 od gives acceptable precision (Table 1). 



The mean standard deviation of these data on 

 three difl:'erent species is 5.3, and the mean is 552 

 mg/liter. The published mean standard deviation 

 for the three methods given in the Environmental 

 Protection Agency (EPA) manual is 1.1, with a 

 mean of 15.0 mg/liter. To compare standard 

 deviations with different means, the coeflficient of 

 variation (CF) is used, and for the data in this 

 paper the CV is 1 as compared with 7 for the data 

 given in the EPA manual. This means that a 

 sample of waste effluent having 100 mg oil and 

 grease/liter will have a comparative standard 

 deviation of 1 or 7 mg/liter, depending on the 

 method used. 



The accuracy of the proposed method was 

 evaluated by comparing the EPA Soxhlet method 

 with the method given in this paper, using seven 

 grab samples of king crab, snow crab, and shrimp 

 waste efl^uents. The data in Table 2 show that the 

 oflicial EPA Freon 113 Soxhlet method gave oil 

 and grease values that were consistently low, 

 varying from 6 to 48% and averaging about 30%. 



The filtrates from the EPA method of filtration 

 from samples 3, 4, 5, 6, 7 were precipitated and 



Table l.-Oil and grease values expressed as milligrams per liter 

 for eight effluent samples. 



Replicate oil and grease values 



Table 2.-Comparison of oil and grease values expressed as 

 milligrams per liter determined by the EPA Soxhlet method and 

 the proposed method. 



'Samples 3A, 3B, and 4A could not be filtered except by 

 changing filters. 



682 



