Cement, Energy and Environment

process i.e. no shift, then it amounts to 2 defects per billion. Control is exercised on the variability of process conditions, while the specification range remains fixed. It is shown in Exhibits 2 and 3. Why 6-sigma?: Because +1- 6-sigma span covers almost the entire distribution under the normal curve and a negligible portion of the area falling outside the control-span is equivalent to a rejection rate of 2 per billion opportunities or 3.4 ppm with the process shift of 1.5 sigma away from the mean. Question arises whether we should aim at 6- sigma control or the rejection rate of 3.4 ppm. Obviously, the answer is to achieve a state of zero defect i.e. 3.4 ppm rejection rate. But it is achievable even with less than 6-sigma control depending on the customer requirement and condition or capability of the machine or process producing it. Fix Sigma Values: The above table acts as a guide for fixing the sigma value for a given situation. Depending on the acceptable rate of defects in ppm and the anticipated shift from the mean ranging from 0.5 to 2.0 sigma distances, the desired sigma value for control can be fixed. For example, when the plant machinery is new, the accuracy and precision are easily maintained of the manufactured product within the specified limits; the raw materials are assumed of acceptable quality and the shift of the process from the mean is negligible. In this case, a lower sigma value is recommended instead of six- sigma. Variations in the process conditions are produced by chance and by an assignable cause . Chance variations fall 6()){()7 I .7 pplll Pill II ~ -6 , j~l11\l I'R hhihil - 2 E'hihit-J 1. 7 Tabl, for fixing sigma value for the rate of defects in ppm and shift from the process mean in multiples of sigma distances. (Refer to table on last page for Z-value) 3.0 2700 158655 66807 22750 6210 3.5 465 66807 22750 621 0 1350 4.0 63 22750 6210 1350 233 4.5 8 6210 1350 233 32 5.0 0.58 1350 233 32 3.4 5.5 0.02 233 32 3.4 0.29 6.0 0.002 32 3.4 0.29 0.02 17