Ways Of Employing Calculations

Why create calculations

  • Combine manual / automatic variables

  • Drive disposition, or alarm conditions

  • Summarize variables into key performance measures

  • Implement site specific logic for triggering events

There are three key attributes of a calculation, the "Event Type," the "Sampling Type," and the "Sampling Interval." The Event Type defines which type of Plant Applications Event drives the calculation, when is the calculation triggered and what is the time scope of the calculation. For example, a calculation with Event Type = Production Event will be triggered any time a Production Event is created or updated on a Production Unit. The time scope of the calculation will be between the start time and the end time of the Production Event.

The Sampling Interval applies only to Time Based variables. Event Based variables inherit their time scope from the event they are derived from, and Time Based variables must have it explicitly defined. The Sampling Interval is specified in minutes.

The Sampling Type defines how data is transformed or aggregated from its raw form. Simple examples of Sampling Type include "Average," "Standard Deviation," and "Total." More complex examples of Sampling Types include "%Limit" or the percentage of time a variables was controlled within its Reject limits, or "Raw Count" which is the count of values found over a time period.

There are two types of calculation definitions, ones that use the predefined sampling types (which typically aggregate or summarize data), and User Defined Calculations. User defined calculations allow almost unlimited capability to create new sampling logic.

Predefined Calculations (Sampling types)

  • Sampling - Last Good Value, Next Good Value, Interpolated

  • Simple Aggregation - Avg, Std, Min, Max, Total

  • Complex Aggregation - Raw Count, Unique Count, Raw Average, Total, Min, Max

  • Specification Based - Cpk, Cp, Ppk, Pp, %Warning, %Limit

  • Other - Last Event Time, Increase

For detailed descriptions of sampling types, see Variable Property Definitions.

See Also