• :

    Chemical and other process industries are always watchful for improvements to increase productivity and decrease waste. A number of methodologies have evolved over the years that focus on the overall organization but few assign priorities to quantifying process input and output as a measure of efficiency by using measurement errors.

     

    Even for process industries, the problem remains to isolate specific areas within the extensive network of chemical manufacturing that are the actual culprits for loss of material and the resulting cost of that material. There may very well be a reluctance to invest in various equipment and methodologies to measure significant differences between input and output throughout the process.

     

    This can be resolved by using two basic statistical approaches. The first is a simple measurement and calculation of material quantities and its variation. The second is to use error propagation techniques to combine the different process components into easily manageable relative terms. Upon completion of all calculations, a simple “student’s t-test” is all that is necessary to determine if there is a significant difference between input and output that requires management attention.

     

    Significant advantages to this approach are iteration and segmentation, that is, the significance level can be initially determined at the original input and the final output, and significance levels can be determined for input and output at any segment of the process. With segmentation, the efficiency at different points in the process can be evaluated and combined into the total throughput to establish the risk potential and planning future capital investments.

  • : Joseph Mikolajczak
  • : Skraeling
  • : Joseph Mikolajczak
  • : statistics
  • : intermediate
  • : jem11sr@aol.com
  • : 4107077040
Process Efficiency Using Error Propagation