Page 43 - Glass Machinery Plants & Accessories no. 3-2019
P. 43
Figure 11 - unexplained noise
percentage versus real specific
consumption deviation
ADDITIONAL OBSERVATIONS firms that, on average, 76 per We can say that, below 500
Other observations can be cent of the deviance can be MJ/ton, the non-explained val-
made. For example, it is useful explained and removed basing ues are above 30 per cent, while
to look at the deviance not just on a model that considers the beyond this value they decrease
in percentage terms, but also in three main production parame- to 20-25 per cent.
absolute ones. If we graph the ters (pull, cullet, boosting), and An EEI-based evaluation
unexplained deviance as a func- the furnace age. method enables not only to mon-
tion of the initial one we see We also notice that, for lim- itor the performance of the melt-
how well it responds to a linear ited values of the base deviance, ing furnace when the operating
model (with obvious intercepts corresponding to very-stable conditions change, but also to
0.0) with a slope of 0.24 and a operating conditions, the per- better appreciate and compare
correlation index of 0.89. This centage of non-explained devi- the aging trend on the different
means that this approach con- ance appears to be limited. furnaces.
glass machinery plants & accessories 3/2019
41
