Asset Management and Other Data-Driven Improvements

ChemCo also has used their PI System data to develop a fleet-wide CBM program designed to do the following:

  • 1. Continuously monitor their plant equipment for sensor malfunctions, such as freezing of data values.
  • 2. Detect offset or drift from optimum operating ranges, caused by conditions such as heat exchanger fouling, excessive vibration, or increased times for normal open/close operations.
  • 3. Totalize the number of times a device turns on or off, which may indicate improper operation or that it has reached the maximum usage time and should be serviced.
  • 4. Synchronize maintenance operations with other software systems, such as their SAP maintenance module.
  • 5. Benchmark similar assets to compare performance of equipment from different suppliers.

Recently, ChemCo has developed new analytic solutions using PI System data that is ingested by a software analytics product that recognizes patterns in large amounts of data. ChemCo used PI System data and analysis tools, together with a third-party analytics software product, to identify and solve several issues:

  • Emissions monitoring. By identifying when a certain valve opened, there was a pressure drop in off-gas treatment, leading to increased emissions. Once resolved by modifying their control system logic, it reduced off-gas emissions by about 60%.
  • Energy consumption. By layering five years of energy consumption data (2013-2017) and comparing several specific items, ChemCo was able to reduce energy consumption.
  • Efficient production. By layering data and comparing good-quality period data with bad-quality period data, ChemCo observed that during the bad-quality periods, one of the flows to the reactor was significantly higher than during good-quality periods. By improving their control system and making sure there would be notifications when this occurred, they significantly reduced bad-quality periods.
  • Batch-to-batch variability. When producing a specific product, the process went to a hold state when in the reactor. By creating an ideal batch profile or fingerprint from many past batches of a specific product, they were able to identify when batches were deviating from ideal operation. ChemCo made another control system correction that resolved this problem and averted a yearly product loss of over 100 tons.

Lessons Learned and Next Steps

In using real-time and archived plant data for several decades to continuously improve operations, ChemCo recommends the following steps:

  • 1. Leverage existing archived data and look for new ways to analyze information to uncover what is difficult to see or explain. It will turn into business value.
  • 2. Use the information in a company's value chain both horizontally and vertically, integrated with partners and customers.
  • 3. Having a common data infrastructure is an advantage, but a good consistent asset model is needed for fast deployment of initiatives. Big data analysis needs more than just a data lake.
  • 4. Focus on environmental, safety, and sustainability issues. ChemCo has reduced C02 emissions by over 35% through PI System data analysis in energy management.
  • 5. Use existing data with emerging analytics software to solve problems by viewing and analyzing the data in new and innovative ways.

Reference

Harclerode, C. 2017. "Data operations transforms fuels value." PTQ Magazine, Ql.

9__

 
Source
< Prev   CONTENTS   Source   Next >