Fall 2003 Capstone Project: Tracking of Honeywell's Downtime

This optimization project comes from the Honeywell Chemical Plant in Geismer, LA, a major producer of HF. The plant is seeking to optimize the effectiveness of plant operations by examining their hourly operation data to find what factors are limiting production and increasing unnecessary downtime.

This group seeks to combine the specialties of students from the computer science, mathematics, and mechanical engineering departments and use the different viewpoints provided by their academic experience to solve a real world problem Initially the group was given six months of data (in hourly increments) from four separate interrelated plant functions. From this mess of data we have then:

  1. Separated good data from data which is suspect due to equipment malfunctions.
  2. Created charts which reflect the trends in data and how these trends show the effects of production demands, seasonal changes, and equipment downtime.
  3. Attempted various mathematical models which might serve to predict the fluctuations in data.

Currently we hope to determine the constraints within the plant process and introduce a detailed mathematical analysis of the production process in order to optimize overall production efficiency.

Drip Stripper Temp vs Flow
This illustration correlates the relationship between the Drip Stripper Temp and the Flow Into Drip Stripper.
HF vs H2SO4
This illustration correlates the relationship between HF in the Feed vs. H2SO4 in the Feed.