When was the last time you heard someone say they were going to 'calibrate' a control valve?

To most people the term 'field instrumentation calibration' applies to analyzers, switches, and transmitters. The reality is, field instrumentation also includes final control devices, such as variable-frequency drives and control valves.

It's unfortunate when final control devices are ignored because the cause of process variability frequently has little or nothing to do with controller tuning, operator actions, or raw material influences, and everything to do with how well these devices perform.

Recently revised standards are designed to establish control valve evaluation and acceptance testing criteria, available from the Instrumentation, Systems, and Automation Society (ISA, Research Triangle Park, N.C.).

The standards do not define acceptable control valve performance nor do they restrict selecting a control valve for any application. The standards do provide detailed instructions that establish a uniform means of testing and analyzing control valve cradle-to-grave performance.

For example, ANSI/ISA-75.25.01-2000, 'Test Procedure for Control Valve Response Measurement from Step Inputs,' defines how to test, measure, and report control valve response characteristics for three test types-bench, laboratory, and in-process.

• Bench tests are conducted with the control valve out of the line or out of service. Because there is no flow through the valve body, results can be significantly different than results produced by laboratory and/or in-process tests for the same valve. Despite the simplicity of bench testing, useful information can be obtained.
• Laboratory tests are performed with media flowing through the valve but under highly controlled conditions (i.e., flow is fully turbulent and not choked). Because of the controlled testing environment, control valve laboratory performance results maynot be achievable under actual in-process conditions.
• In-process tests provide valve response results in actual, or close-to-actual conditions. The limiting factor to obtaining meaningful in-process test results is usually the availability of suitable measurement devices and/or process factors that may limit the range of tests that can be conducted.

Regardless of the test type used, ANSI/ISA 75.25.01 defines three procedures designed to obtain the maximum quality and quantity of valve performance information-baseline, small step, and response time.

Baseline testing is considered optional, but when used it evaluates measurement noise, presence of valve limit cycling, and establishes baseline response time. Baseline testing is very useful in establishing a control valve's initial installation performance 'signature.'

Small-step testing is performed to determine a valve's deadband and resolution. Testing begins after setting the input signal to a nominal value and waiting a specified time. The test continues with a series of small up and down step changes with defined wait periods between each step change.

Response time step testing provides a comparison of response time to step size as well as approximate values for dead band and resolution.

Despite their rugged appearance, control valves are every bit as much a precision instrument as an analyzer or transmitter. Show them a little respect and they will return big dividends.

Comments? E-mail dharrold@cahners.com