As a process and product engineer for a fortune 100 company that manufactured high value chemical products, my job involved analysis of a tremendous amount of process and product information.
I remember coming to work on mornings, feeling anxious about what might have happened overnight. Five to twenty orders of product ($5,000 to $50,000 an order) were on my team’s production schedule since I left work the day before. Were they successful? Were there any product parameters out of spec.? Were the orders on hold, or were they released for downstream production?
Typically, multiple orders would be on hold for abnormal parameter(s). My work would then begin, attempting to evaluate what was different about those questionable orders and determining proper disposition. There were reams of production information, available to be analyzed. Alarms were identified, and each order had over a hundred parameters that could be trended. Notes written by operators related to the orders could be read.
I would typically spend several hours each morning evaluating all available information before dispositioning the product. Monday mornings were tough since production didn’t stop over the weekend. During my morning ritual every day, I would either: release orders for downstream production, hold an order to see if the same abnormality would repeat on future orders, or send orders to waste for disposal.
The product disposition process was very important to our bottom line. If I released an order for downstream production, but it was later found that the final product had to be thrown out because of “my bad order,” our company would waste 5 to 10 times the value of my order!
My quality of life was significantly improved when we started use of a new tool, called Process Monitor (PM), that efficiently displayed our process and product information. PM utilized a control chart for every critical process and product parameter for each production order. The control limits were automatically created, based on recent production information and specs. limits could also be displayed. And, information was summarized for all recent production orders on a single screen. Any “abnormal” orders would be displayed with a red alarm bell next to the order on the display screen. If no red alarm bell existed for an order, it was “good to go.”
If a red alarm bell did exist for an order, it was simple to display what was different about that order. In the screenshot below, order 80060 was out of control for the HEATUP_TEMP_TMX tag, as displayed in the Order Number control chart. The TEMP Overlay trend chart on the lower right section of the display shows that the order in alarm had a delay on the heat up timing compared to other batches that were in control.
The TEMP time-trace chart on the lower right portion of the display above was created by clicking my mouse 5 times: one click on the order in alarm, and a click on each of the 4 batches highlighted in the control chart. The time-trace chart on the lower right shows that the order in alarm had a delay before the product temperature increased compared to “normal batches.”
For this batch of a specialty chemical, I was able to release the product for final production, because no product parameters were out of specification. But the alarm let us know that something was abnormal, and after further analysis we learned that the heat transfer surface of a jacked vessel was beginning to foul. Cleaning the heat transfer surface fixed the problem before any costly product was lost.
The time required by me to disposition orders was typically reduced to less than a half hour a day. And, I was able to make the decisions with good information that was easily displayed!
Our company was able to implement PM on diverse types of operations to display and ultimately eliminate process variability, resulting in huge productivity improvements. The operations included: manufacture of bulk and precision chemicals, extrusion and coating of liquids onto wide rolls of optical grade film, manufacture and operation of high-speed printers, and even production and distribution of supplied utility services including steam, electricity and chilled water.
The basic ideas behind Process Monitor are not new. Statistically based process and product information organized in a simple manner to display abnormalities can lead to great improvements for you and your company’s bottom line.
{{cta(’81fd131b-9eec-46da-980a-04bbbce8cd72′)}}