Process Engineering Discussion Community

Expand all | Collapse all

How do you use your data?

  • 1.  How do you use your data?

    Posted 11-04-2019 18:57
    Hello fellow process engineers! I am fairly new to the field, I've been in my plant for about a year and I worked in O&G for the previous 5 years in very different roles. I'm hoping to spark some discussion about data and get some ideas for improvement and just learn what other people out there are doing!

    What type of plant do you work in, what metrics do you track, and what software has been most helpful in monitoring/tracking the efficiency of your process?

    Also if your process is manual, how do you track that process? Do your operators use timers for each batch? Do you have programming to track the time the tank levels change/agitators are on until the tank is empty again?

    Last question, once you've analyzed your data, how do you utilize it? Do you have a daily meeting with operators to say what went well/what needs to be improved?

    My facility is a batch process for bulk chemicals. We use DeltaV for process control, Aspen for viewing/storing continuous and batch data (IP21, process explorer, and production record manager), and we also use JMP to perform statistical analysis on specific products by importing the continuous data from Aspen.

    I think we're on to something with JMP but I would like to see some ideas from others, I know Aspen has OEE monitoring in IP21 but we haven't used it yet and I'm not sure if its worth exploring further or if there are better solutions? JMP also requires a lot of programming up front to analyze a specific product, I think we could send better data to JMP to make it a bit easier/to add more products to track. I have also seen some standalone OEE products but I'm not sure how they would work with our process.


    Amber Manzi
    Process engineer II
    INEOS Composites
    Pittsburgh PA

  • 2.  RE: How do you use your data?

    Posted 11-05-2019 10:23
    Amber and fellow Process Engineers:?? If your company has JMP from SAS
    (Statistical Analysis Systems), they are most likely plotting the data
    by using Statistical Control Charts.?? It will not only produce the
    charts, but it will calculate the standard deviation of the data over
    time and determine the upper and lower control limits which is typically
    2.66 times the standard deviation for Shewart Individuals Control
    Charts.?? Changes to set points should only be made when a data point
    lies outside the control limits.?? Otherwise, chasing the "needle" will
    cause more process variation.

    The institute used to have an eLearning course, ELA128, Statistical
    Process Quality Control.?? The course consists of two lectures on
    statistics, one lecture on "run" charts, two lectures on variables
    control charts with Gaussian, one lecture on attribute control charts
    for binomial distributed variables, one lecture on sampling and ANOVA
    (ANalysis Of VAriance), and a final lecture on the "Real World" which
    includes both JMP from SAS and Statistica from SoftStat.

    My recommendation would be to contact the AIChE Academy which can be
    found at:

    Good luck and please let me know if you were able to find the course.??
    It has not been listed for a couple of years.

    David Griffith

  • 3.  RE: How do you use your data?

    Posted 11-06-2019 08:34
    Apologies for the question marks when a double space is used after
    period in my previous reply.  There is a bug in Mozilla Thunderbird with
    their default of "False" for the "mail.strictly_mime" setting when
    aol/verizon is the SMTP server.  Unfortunately, posting this reply is
    the only way I can test to verify that switching the setting from
    "False" to "True" has fixed the problem.

    T. David Griffith

  • 4.  RE: How do you use your data?

    Posted 11-14-2019 06:35
    Thanks David! I am familiar with the concept your describing and we certainly don't make knee jerk changes due to the variability in our process. What it has helped us with is identifying some maintenance issues like the heating oil pump started to go bad and we noticed it took longer to heat up several batches in a row. This spurred an investigation and quick repair of the pump! That's what I love about data but I just feel like there are so many more ways to use Aspen and other tools to look at and interpret the data.

    Amber Manzi
    Process engineer
    INEOS Composites
    Pittsburgh PA