Rephrasing engineering benefits for the bean counters

This back-to-basics tutorial quantifying the benefits of automation puts engineering in language even a bean counter can understand: What have you done for me lately? And how much?

At its recent users’ group meeting, Emerson’s John Dolenc, principal consulting engineer, offered a back-to-basics tutorial quantifying the benefits of automation in language even a bean counter can understand: What have you done for me lately? And how much?

View related content on PlantServices.com

The principles he described readily transfer to investments in maintenance, reliability and other efforts that with the potential to improve production output, quality, energy consumption, etc. Everyone knows these investments pay off, but it seems that few users can calculate these gains and express them in dollars and cents.

Maybe you’re so focused on keeping equipment running that you forget to keep score, or perhaps you see the benefits as so obvious that it’s not necessary to calculate them. Whatever the reason, specific benefits analyses often don’t reach management levels. This is unfortunate because administrators need to be reminded frequently about why it’s crucial for them to fund their automation and reliability projects.

“The justification for process automation modernization is one of the most difficult steps to complete during the planning process,” says Dolenc. “You have to create the business case, obtain operational data, develop the benefits, estimate degree of improvement and establish the financial gain.”

Dolenc demonstrated three main financial tools for estimating process automation benefits: the best-operator method, the data-reduction method and the percent-limit-violation method. Just employing these accounting practices can help make a project’s case because they’re published methods that are widely used in business.

To begin the search for benefits, Dolenc says engineers should try to think of their plant as a financial asset or entity. This means looking at the facility much like a balance sheet with raw materials, net utilities, operating expenses, maintenance costs and new capital going in, and primary and secondary products and waste streams coming out.

Basically, to calculate return on invested capital (ROIC), users must divide profit by invested capital: ROIC = profit/invested capital.

For the best-operator method, Dolenc continued, users need to identify and retrieve historical data, including key performance indicators, important manipulated variables, constraint variables, and material and energy balance parameters. They must obtain data while the plant is running at normal operating conditions and make sure that the data-sampling frequency matches the type of data collected. This historical data can then be used to compare, for example, operating data between the historical best versus average results or two similar lines or plants with different levels of reliability.

Best-operator is a conservative method because it doesn’t account for improvements. It’s expressed mathematically as:

∆ value = (|average variable – best operator variable|) x average rate

Similarly, the data-reduction method begins by collecting raw data over an appropriate time interval, such as hourly averages/week, shift averages/month or daily averages/year, and then removing startup/shutdown and upset data. Next, the data is normalized to a production rate, plotted in a histogram, and the resulting values are subtracted at the median and the 15% points. This difference indicates potential economic benefit.

The limit-violation method involves calculating old and new performance averages for a given process, identifying how much of each falls beyond the process limit or product specification, and comparing the two. The same percent-limit-violation method involves calculating the average and then calculating the standard deviation. These are expressed as:

  • Z = (L -AvgO)/SDO
  • Z = (L -AvgN)/SDN
  • (L - AvgN)/SDN = (L -AvgO)/SDO
  • AvgN = L – SDN/SDO x (L – AvgO)

“Financial justification is difficult, but it’s necessary for automation to receive the approval it needs,” concludes Dolenc. “To justify a project, it’s vital to collect historical data, build a base case and provide some prediction of results.”

Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments