Personnel can easily obtain all of the information associated with this sample from the Process Scheduler. Although all of the data are within the limits, which are represented by the red lines, the sharp change may indicate a problem.
Abandoning manual trends
The Pasadena plant has a history of continual improvement. Statistical process control data are an important part of day-to-day decision-making in the plant, and the ability to efficiently gather and organize these data has become increasingly important.
The plant originally shared the cost of a host-based LIMS with several corporate siblings. Laboratory test results would be printed on paper and sent to unit operations for recording and manual plotting. Laboratory statistical quality-control checks also were maintained manually, which requires additional time. The system provided basic functionalities, but nothing close to the real-time plotting capabilities the company now receives.
When the plant decided to upgrade its LIMS, it chose the PC version of Labworks from PerkinElmer Inc., Wellesley, Mass. Labworks was chosen because it is a configurable product that is designed to meet the needs of chemical plant laboratories without customization. The LIMS provides built-in tools that can be used to adapt the software to most any lab process. This made it possible to upgrade the system in just three days with minimal disruption to plant operations.
Move to automated plots
"The main reason we decided to do the upgrade was to take advantage of the online plotting program," Cook says. The Labworks Trend Plot program provides users with the ability to graphically display a series of result values for groups of selected parameters in real time. These trend plots can be used by personnel outside the laboratory for tracking manufacturing or environmental processes in a plant. They also can be used by workers in the lab for tracking instrument performance and calibration.
Cook says the software was configured to show alarm limits and specification limits in different colors on the plots, which are automatically updated every time a sample is analyzed. As a result, lab and operations personnel can see when the process units have a problem. They also can see when the units are heading toward a problem when there is still time to make corrections.
Ethyl Corp. has set up five or six plots for supervisors and process operators in each of the plant's four production units. Software automatically scans from plot to plot, making it possible to instantly review trends. The intervals between tests vary within the production units. The continuous units might receive data every six hours, whereas the blending units, which operate in batch mode, only receive data at the end of a batch.
"Our objective in implementing trend plotting was to improve our process capabilities by getting a better understanding of the variation in our process," Cook says. "Right now, we are looking closely at the variation of different parameters. If a parameter varies more than usual, even if it meets the specification, we will investigate.
Streamlined workflow
The laboratory workflow proceeds as follows: unit operations deliver the samples on a defined schedule. Each sample has a unique unit code. This unit code is defined in the LIMS with all the required test parameters and specification limits. When a sample is received, an analyst enters the unit code, date and time, and all the test parameters for the sample are quickly displayed. As a test is completed, an analyst enters the result. When all the tests are completed, the analyst validates the results. As results are saved, the value also is displayed on the trend plots.