AMERICAN NATIONAL STANDARD

ANSI/ISA–75.25.01–2000

Test Procedure for Control

Valve Response Measurement
from Step Inputs

Approved 10 January 2001

TM

ISA–The Instrumentation,
Systems, and
Automation Society
ANSI/ISA-75.25.01-2000
Test Procedure for Control Valve Response Measurement from Step Inputs

ISBN: 1-55617-742-9

Copyright Ó 2000 by ISA–The Instrumentation, Systems, and Automation Society. All rights reserved. Not
for resale. Printed in the United States of America. No part of this publication may be reproduced, stored
in a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photocopying,
recording, or otherwise), without the prior written permission of the Publisher.

ISA
67 Alexander Drive
P. O. Box 12277
Research Triangle Park, North Carolina 27709
USA
 —3— ANSI/ISA–75.25.01–2000

Preface

This preface, as well as all footnotes and annexes, is included for information purposes and is not part of
ANSI/ISA-75.25.01-2000.

The standards referenced within this document may contain provisions which, through reference in this
text, constitute requirements of this document. At the time of publication, the editions indicated were valid.
All standards are subject to revision, and parties to agreements based on this document are encouraged to
investigate the possibility of applying the most recent editions of the standards indicated within this
document. Members of IEC and ISO maintain registers of currently valid International Standards. ANSI
maintain registers of currently valid U.S. National Standards.

This document has been prepared as part of the service of ISA–The Instrumentation, Systems, and
Automation Society, toward a goal of uniformity in the field of instrumentation. To be of real value, this
document should not be static but should be subject to periodic review. Toward this end, the Society
welcomes all comments and criticisms and asks that they be addressed to the Secretary, Standards and
Practices Board; ISA; 67 Alexander Drive; P. O. Box 12277; Research Triangle Park, NC 27709;
Telephone (919) 549-8411; Fax (919) 549-8288; E-mail: [email protected].

The ISA Standards and Practices Department is aware of the growing need for attention to the metric
system of units in general, and the International System of Units (SI) in particular, in the preparation of
instrumentation standards. The Department is further aware of the benefits to USA users of ISA standards
of incorporating suitable references to the SI (and the metric system) in their business and professional
dealings with other countries. Toward this end, this Department will endeavor to introduce SI-acceptable
metric units in all new and revised standards, recommended practices, and technical reports to the
greatest extent possible. Standard for Use of the International System of Units (SI): The Modern Metric
System, published by the American Society for Testing & Materials as IEEE/ASTM SI 10-97, and future
revisions, will be the reference guide for definitions, symbols, abbreviations, and conversion factors.

It is the policy of ISA to encourage and welcome the participation of all concerned individuals and interests
in the development of ISA standards, recommended practices, and technical reports. Participation in the
ISA standards-making process by an individual in no way constitutes endorsement by the employer of that
individual, of ISA, or of any of the standards, recommended practices, and technical reports that ISA
develops.

CAUTION — ISA ADHERES TO THE POLICY OF THE AMERICAN NATIONAL STANDARDS

INSTITUTE WITH REGARD TO PATENTS. IF ISA IS INFORMED OF AN EXISTING PATENT THAT IS
REQUIRED FOR USE OF THE STANDARD, IT WILL REQUIRE THE OWNER OF THE PATENT TO
EITHER GRANT A ROYALTY-FREE LICENSE FOR USE OF THE PATENT BY USERS COMPLYING
WITH THE STANDARD OR A LICENSE ON REASONABLE TERMS AND CONDITIONS THAT ARE
FREE FROM UNFAIR DISCRIMINATION.

EVEN IF ISA IS UNAWARE OF ANY PATENT COVERING THIS STANDARD, THE USER IS
CAUTIONED THAT IMPLEMENTATION OF THE STANDARD MAY REQUIRE USE OF TECHNIQUES,
PROCESSES, OR MATERIALS COVERED BY PATENT RIGHTS. ISA TAKES NO POSITION ON THE
EXISTENCE OR VALIDITY OF ANY PATENT RIGHTS THAT MAY BE INVOLVED IN IMPLEMENTING
THE STANDARD. ISA IS NOT RESPONSIBLE FOR IDENTIFYING ALL PATENTS THAT MAY
REQUIRE A LICENSE BEFORE IMPLEMENTATION OF THE STANDARD OR FOR INVESTIGATING
THE VALIDITY OR SCOPE OF ANY PATENTS BROUGHT TO ITS ATTENTION. THE USER SHOULD
CAREFULLY INVESTIGATE RELEVANT PATENTS BEFORE USING THE STANDARD FOR THE
USER’S INTENDED APPLICATION.
ANSI/ISA–75.25.01–2000 —4—

HOWEVER, ISA ASKS THAT ANYONE REVIEWING THIS STANDARD WHO IS AWARE OF ANY
PATENTS THAT MAY IMPACT IMPLEMENTATION OF THE STANDARD NOTIFY THE ISA
STANDARDS AND PRACTICES DEPARTMENT OF THE PATENT AND ITS OWNER.

ADDITIONALLY, THE USE OF THIS STANDARD MAY INVOLVE HAZARDOUS MATERIALS,

OPERATIONS OR EQUIPMENT. THE STANDARD CANNOT ANTICIPATE ALL POSSIBLE
APPLICATIONS OR ADDRESS ALL POSSIBLE SAFETY ISSUES ASSOCIATED WITH USE IN
HAZARDOUS CONDITIONS. THE USER OF THIS STANDARD MUST EXERCISE SOUND
PROFESSIONAL JUDGMENT CONCERNING ITS USE AND APPLICABILITY UNDER THE USER’S
PARTICULAR CIRCUMSTANCES. THE USER MUST ALSO CONSIDER THE APPLICABILITY OF ANY
GOVERNMENTAL REGULATORY LIMITATIONS AND ESTABLISHED SAFETY AND HEALTH
PRACTICES BEFORE IMPLEMENTING THIS STANDARD.

THE USER OF THIS DOCUMENT SHOULD BE AWARE THAT THIS DOCUMENT MAY BE IMPACTED
BY ELECTRONIC SECURITY ISSUES. THE COMMITTEE HAS NOT YET ADDRESSED THE
POTENTIAL ISSUES IN THIS VERSION.

The following people served as members of ISA Subcommittee SP75.25:

NAME COMPANY

C. Langford, Chairman Cullen G. Langford, Inc.

W. Weidman, Managing Director Parsons Energy & Chemicals Group
J. Beall Eastman Chemical Co.
D. Bennett Sampson Controls, Inc.
W. Bialkowski Entech Control Engineering, Inc.
W. Black Cashco, Inc.
M. Boudreaux Exxon Mobil Chemical
S. Boyle Neles Automation
D. Buchanan Union Carbide Corporation
F. Cain Flowserve Corporation
N. Cammy UOP LLC
R. Champagne Consultant
J. Jamison Bantrel Inc.
S. Kempf Harold Beck & Sons, Inc.
P. Maurath Procter & Gamble Company
R. McEver Bettis Corporation
N. McLeod Elf Atochem
G. McMillan Solutia, Inc.
J. Reed Norriseal
K. Senior Dupont Dow Elastomers

The following people served as members of ISA Committee SP75:

NAME COMPANY

D. Buchanan, Chairman Union Carbide Corporation

W. Weidman, Managing Director Parsons Energy & Chemicals Group
A. Abromaitis Red Valve Company, Inc.
H. Backinger J. F. Kraus & Company
G. Barb Retired
H. Baumann H B Services Partners LLC
H. Boger Masoneilan Dresser
G. Borden Consultant
 —5— ANSI/ISA–75.25.01–2000

S. Boyle Neles Automation

R. Brodin Fisher Controls International, Inc.
F. Cain Flowserve Corporation
C. Corson Fluor Daniel Inc.
M. Coughran Fisher Controls
A. Engels Praxair, Inc.
H. Fuller Valvcon Corporation
J. George Richards Industries
A. Glenn Flowserve Corp.
L. Griffith Consultant/Retired
B. Guinon Shell Chemical
F. Harthun Retired
B. Hatton DeZurik Division Unit
J. Jamison Bantrel, Inc.
R. Jeanes TXU Electric
J. Kersh M. W. Kellogg Company
C. Langford Cullen G. Langford, Inc.
A. Libke DeZurik Valve Company
R. Louviere Creole Engineering Sales Company
O. Lovett Consultant/Retired
A. McCauley Chagrin Valley Controls, Inc.
R. McEver Bettis Corporation
H. Miller Control Components, Inc.
T. Molloy CMES Inc.
L. Ormanoski Frick Company
J. Ozol Commonwealth Edison
W. Rahmeyer Utah State University
J. Reed Norriseal
K. Schoonover Con-Tek Valves, Inc.
A. Shea Copes-Vulcan, Inc.
E. Skovgaard Leslie Controls, Inc.
H. Sonderegger Tyco Flow Control
R. Terhune Retired

This standard was approved for publication by the ISA Standards and Practices Board on
30 December 2000:

NAME COMPANY

M. Zielinski, Vice President Fisher-Rosemount Systems, Inc.

D. Bishop Consultant
P. Brett Honeywell, Inc.
M. Cohen Senior Flexonics, Inc.
M. Coppler Ametek, Inc.
B. Dumortier Schneider Electric SA
W. Holland Southern Company
A. Iverson Ivy Optiks
R. Jones Dow Chemical Co.
V. Maggioli Feltronics Corp.
T. McAvinew Bateman Engineering, Inc.
A. McCauley, Jr. Chagrin Valley Controls, Inc.
G. McFarland Westinghouse Process Control Inc.
D. Rapley Rapley Consulting Inc.
R. Reimer Rockwell Automation
ANSI/ISA–75.25.01–2000 —6—

J. Rennie Factory Mutual Research Corp.

H. Sasajima Yamatake Corp.
R. Webb Altran Corp.
W. Weidman Parsons Energy & Chemicals Group
J. Weiss EPRI
J. Whetstone National Institute of Standards & Technology
M. Widmeyer EG&G Defense Materials
R. Wiegle CANUS Corp.
C. Williams Eastman Kodak Co.
G. Wood Graeme Wood Consulting
 —7— ANSI/ISA–75.25.01–2000

Contents

1 Scope and limitations ....................................................................................................................... 9

2 Purpose and limitations .................................................................................................................... 9
3 Definitions......................................................................................................................................... 9
4 General test procedures ................................................................................................................. 14
4.1 Test valve conditions............................................................................................................. 14
4.2 Test system........................................................................................................................... 14
4.3 Measuring instrument requirements...................................................................................... 15
4.4 Process variable.................................................................................................................... 15
4.5 Nominal test positions ........................................................................................................... 16
4.6 Examples of step response................................................................................................... 16
5 Tests specified for each of three test environments ....................................................................... 18
5.1 Bench tests ........................................................................................................................... 18
5.2 Laboratory tests .................................................................................................................... 19
5.3 In-process tests..................................................................................................................... 19
6 Detailed test procedures................................................................................................................. 20
6.1 Baseline test.......................................................................................................................... 20
6.2 Small step test....................................................................................................................... 21
6.3 Response time tests.............................................................................................................. 23
7 Presentation of test results ............................................................................................................. 24
7.1 General information............................................................................................................... 24
7.2 Test results............................................................................................................................ 25
8 References ..................................................................................................................................... 27
Annex A — Sliding Friction Measurement........................................................................................... 29
This page intentionally left blank.
 —9— ANSI/ISA–75.25.01–2000

1 Scope and limitations

This standard defines the testing and reporting of step response of control valves that are used in throttling
closed loop control applications. A control valve consists of the complete, ready-to-use assembly of the
control valve body, actuator, and any required accessories. The most probable accessory is a valve
positioner. For background, refer to technical report ISA-TR75.25.02-2000, Control Valve Response
Measurement from Step Inputs.

If this standard is used for evaluation or acceptance testing, the parties may agree to documented
variations from these requirements.

This standard does not define the acceptable control valve performance for process control nor does it
restrict the selection of control valves for any application.

This standard defines methods and criteria for performing response tests and evaluating test results for
three alternative environments – “bench testing”, “laboratory testing”, and “in-process testing”. “Bench
testing” is testing without flow such as in a plant instrument shop, laboratory, or control valve
manufacturing site. “Laboratory testing” is testing with flow in a laboratory. “In-process” testing is
performed in a plant during normal plant operation with process flow.

2 Purpose and limitations

The purpose of this standard is to define how to test, measure, and report control valve response
characteristics. This information can be used for process control applications to determine how well and
how fast the control valve responds to the control valve input signal. It does not define acceptable process
control, or restrict the selection of control valves to any application, since process requirements and user
objectives vary widely.

The information is specifically applicable to closed loop feedback control but may have some application to
open loop control applications. It does not address valves used in on-off control service.

Tests specified in this standard may not be sufficient to measure the performance required for all
applications. Not all control valve applications will require this testing.

3 Definitions
The technical report ISA-TR75.25.02-2000 explains applicable terms and explores control valve static and
dynamic response characteristics important for process control. That information will aid correct
interpretation and application of the test results obtained from the tests defined in this standard. Other
relevant standards are included in clause 8.

This standard and ISA-TR75.25.02-2000 make use of terms as defined in ANSI/ISA-51.1-1979 (R1993),
Process Instrumentation Terminology, and some of the essential terms are repeated here for convenience.
In the specific area of nonlinear dynamics, it was determined that some terms defined in ANSI/ISA-51.1-
1979 (R1993) lacked the precision desired for these documents. Others were inconsistent with the
terminology used in the nonlinear control literature. A common set of definitions is used in this standard
and ISA-TR75.25.02-2000. Those used only in ISA-TR75.25.02-2000 are marked with (*).

3.1 backlash:*
in process instrumentation, a relative movement between connected mechanical parts, resulting from
looseness when motion is reversed [ANSI/ISA-51.1-1979 (R1993)]. Sometimes also referred to as slop,
lost motion, or free play.
ANSI/ISA–75.25.01–2000 — 10 —

Output

b
a c
d
a < resolution £ b

Input c £ dead band < d

Amplitude

Time

Dynamics are not shown

Figure 1 — Dead band and resolution

3.2 closed loop time constant:*

the time constant of the closed loop response of a control loop, used in tuning methods such as Internal
Model Control (IMC) and Lambda Tuning. The closed loop time constant is a measure of the performance
of a control loop.

3.3 dead band:

the range through which an input signal may be varied, with reversal of direction, without initiating
an observable change in output signal [ANSI/ISA-S51.1-1979 (R1993)]. In this standard and
ISA-TR75.25.02-2000, it is defined in percent of input span. Note that in some other literature this
definition is used for dead zone.

3.4 dead time (Td):

the time after the initiation of an input change and before the start of the resulting observable response.

3.5 dead zone:*

a zone of input for which no value of the output exists [ANSI/ISA-S51.1-1979 (R1993)].

3.6 dynamic response:

the time-dependent output signal change resulting from a defined time-dependent input signal change.
Commonly used input signal changes include impulse, pulse, step, ramp, and sinusoid [McGraw-Hill
"Dictionary of Scientific and Technical Terms", fifth edition, 1994]. Dynamic means that the control valve is
 — 11 — ANSI/ISA–75.25.01–2000

moving. Dynamic response can be measured without process loading in bench top tests with simulated or
active loading in a flow laboratory or under normal process operating conditions.

3.7 gain ratio GZ / GZ02:

the response gain GZ divided by the response gain GZ02 determined from the multi-step test performed with
a step size of 2 percent. The ideal gain ratio equals 1.0 for tests about any nominal position.

GR = GZ / GZ02

3.8 hunting:*
an undesirable oscillation of appreciable magnitude, prolonged after external stimuli disappears
[ANSI/ISA-S51.1-1979 (R1993)]. Hunting can have two forms: oscillations occurring near the stability
limit of a linear system or the limit cycling tendency of a nonlinear system.

3.9 limit cycle:

an oscillation caused by the nonlinear behavior of a feedback system. These oscillations are of fixed
amplitude and frequency, and can be sustained in a feedback loop even if the system input change is zero.
In linear systems, an unstable oscillation grows theoretically to infinite amplitude, but nonlinear effects limit
this growth [Van De Vegte, J., "Feedback Control Systems", 2nd ed, Prentice Hall, 1990, p14]. See also
hunting [ANSI/ISA-S51.1-1979 (R1993)].

3.10 memory:*
in the context of small signal nonlinear dynamics, is that property of a nonlinearity which makes it sensitive
to the current direction, and the history of the input signal. Memory requires the inclusion of direction
arrows on those line segments of an X-Y plot that are directionally sensitive.

3.11 nonlinear system:*

a nonlinear system is one whose response depends on the amplitude and the nature of the input signal, as
well as the initial conditions of the system. As an example, a nonlinear system can change from being
stable to unstable by changing the size of the input signal.

When a nonlinear system is driven towards a setpoint by feed back control action, it is likely to develop a
limit cycle. The amplitude and frequency of such limit cycles are a function of the nature of the
nonlinearities which are present, and the effective gain of the feed back control action. As the gain of the
feed back is increased, the frequency of the limit cycle is likely to increase. More aggressive gain
increases may produce behavior such as bifurcation, frequency doubling and eventually chaotic behavior.

3.12 nonlinearity:*
there are many types of nonlinearities, although they can be generally grouped into two main groups:
simple nonlinearities without memory and more complex nonlinearities with memory [Van De Vegte,
above, Gibson, J. E. "Nonlinear Automatic Control" McGraw-Hill 1963]. Not the same as in ANSI/ISA-
S51.1-1979 (R1993), linearity: the closeness to which a curve approximates a straight line.

3.13 overshoot:
the amount by which a step response exceeds its final steady state value. Refer to figure 24 of
ANSI/ISA-S51.1-1979 (R1993). Usually expressed as a percentage of the full change in steady state
value.

3.14 position Z:
the position of the closure member relative to the seated position. In this standard and ISA-TR75.25.02-
2000, expressed as a percent of span.
ANSI/ISA–75.25.01–2000 — 12 —

3.15 resolution:
smallest step increment of input signal in one direction for which movement of the output is observed.
Resolution is expressed as percentage of input span. The term in this document means: the tendency of a
control valve to move in finite steps in responding to step changes in input signal applied in the same
direction. This happens when the control valve sticks in place, having stopped moving after the previous
step change.

3.16 response:
the time history of a variable after a step change in the input. In this standard and ISA-TR75.25.02-2000,
the step response can be stem position, flow, or another process variable.

3.17 response flow coefficient CvR:*

apparent flow coefficient as determined by testing in an operating type environment. The data available in
the operating environment may differ from the laboratory data required by valve sizing standards.

3.18 response gain GZ:

the ratio of the steady state magnitude of the process change DZ divided by the signal step Ds that caused
the change. One special reference response gain is defined as that calculated from the 2 percent step
size response time test. This is designated as GZ02.

GZ = DZ / Ds

GZ02 = DZ02 / Ds02

3.19 sampling interval Dts:

the time increment between sampled data points. It is the inverse of the sampling rate, f0. DtS = 1/f0. As
used in this standard and ISA-TR75.25.02-2000, since more than one variable is being sampled, it is the
time between the sets of sampled data. Ideally, all variables in one set are sampled at the same time. If
data is recorded using analog equipment, the time constant for the recording equipment shall be less than
or equal to the maximum allowed Dt.

3.20 sampling rate f0:

the rate at which data samples are taken or the number of samples per unit time. See sampling interval.

3.21 shaft windup:*

in rotary valve systems, the tendency of the drive shaft to twist under load while the closure member is
stuck at a given position.

3.22 sliding friction Fr or Tr:

the force or torque required to maintain motion in either direction at a prescribed input signal ramp rate.

3.23 static:
means without motion or change [McGraw-Hill "Dictionary of Scientific and Technical Terms", fifth edition,
1994]; readings are recorded after the device has come to rest. Static performance can be measured
either without process loading (bench top tests), with simulated or active loading, or under process
operating conditions. This kind of test is sometimes called a dynamic test [McGraw-Hill above], which may
cause confusion. The static behavior characteristics identified as important to the control valve
performance are the dead band, the resolution, and the valve travel gain.
 — 13 — ANSI/ISA–75.25.01–2000

3.24 steady state:

a condition of a dynamic system when it is at rest at a given value. In testing the responses of a dynamic
system, step test methods are often used. The resulting system transitions from an initial steady state
value to a new final steady state value.

3.25 step change:

a nearly instantaneous step change made to an input signal of a dynamic system with the intention of
stimulating a step response of the dynamic system. Such a test is used to characterize the step response
of the dynamic system.

3.26 step change time Dtsc:

the time between the start of a signal input step and when it reaches its maximum value.

3.27 step test:

the application of a step change to an input signal in order to test the step response dynamics.

3.28 step response time (T86):

the interval of time between initiation of an input signal step change and the moment that the response of a
dynamic reaches 86.5% of its full steady state value. The step response time includes the dead time
before the dynamic response.

3.29 step size Ds:

the difference between the beginning and ending signal in a step change expressed as a percent of the
signal span.

3.30 stiction (static friction):*

resistance to the start of motion, usually measured as the difference between the driving values required to
overcome static friction upscale and downscale [ANSI/ISA-S51.1-1979 (R1993)].

3.31 stick/slip:*
a term that attempts to explain jerky or “sticky” motion by postulating that static friction differs substantially
from sliding friction. However, friction is rarely directly measured, and “sticky” behavior can be caused by
other physical effects (e.g., positioner behavior, at small amplitudes).

3.32 stick/slip cycle:*

a term that attempts to describe a limit cycle caused when the control valve “sticks” and suddenly “slips”
during a change in input signal. It is the result of static friction combined with a positioner and actuator
system that does not provide enough force to overcome friction at low positioner error values.

3.33 time constant t:

for first order dynamic systems, the interval of time between initiation of an input signal step change and
the moment that a first order dynamic system reaches 63.2% of the full steady state change. The term is
used in this standard and ISA-TR75.25.02-2000 to describe the dynamic characteristics of the analog
measuring instruments.

3.34 valve travel gain:*

the change in closure member position divided by the change in input signal, both expressed in
percentage of full span.

GX = DX / Ds
ANSI/ISA–75.25.01–2000 — 14 —

3.35 valve system approximate time constant (t'):*

the time constant of a first order response without dead time, which may fit the actual control valve step
response reasonably well. The approximate time constant is defined to provide a basis for comparison of
the valve with other time constants, such as the closed loop time constant for the control loop. A first order
system reaches 86.5% of its final step response value in two time constants; the approximate time
constant is considered to be one half of the step response time, T86. The use of the approximate time
constant in no way implies that the response of the control valve is first order. The step response of the
control valve is typically complex, having dead time initially, followed by potentially complex dynamics
before the steady state is achieved. T86 includes the dead time in the initial part of the response, as well
as the possibility of slower settling in the last portion of the response. Some valve positioner designs
attempt to achieve a slow-down in the final part of the response in order to limit overshoot. t' attempts to
produce a simple linear time constant approximation of the control valve dynamic response, which can be
compared to the closed loop time constant of the control loop on the same basis in time constant units.
Note that as the portion of T86 that is dead time increases, this approximation becomes less ideal.

3.36 velocity limiting:*

the maximum rate of change that a system can achieve due to its inherent physical limitations.

3.37 wait time Dtw:*

the time spent after a step input change waiting for the response to come to the new steady state value.

3.38 X-Y plot:*

a plot of the output excursions plotted against input excursions. Input-output plots are useful for defining
the steady state characteristics of nonlinearities.

4 General test procedures

4.1 Test valve conditions

The test valve shall be set to its desired test configuration. For in-process testing, the valve may already
be configured properly or further adjustment may not be possible. This includes configuring the valve
assembly with the desired packing type and condition, positioner if applicable, and actuator configuration.
The positioner configuration shall include any applicable adjustments such as gain setting. In some cases,
preliminary tests may be performed such as testing to assure there is no excessive overshoot. (Excessive
overshoot is not defined here and the amount allowed may vary according to the application but is to be
reported.) All applicable characteristics of the valve configuration that would affect test results shall be
reported.

4.2 Test system

Testing to determine the response of a control valve requires a signal generator or source and instruments
to measure the input signal, the position of the closure member and, for laboratory testing or in-process
testing, the desired response variable. The response variable could be derived from other variables that
may need to be measured as well.

The tests can be performed manually with appropriate instrumentation but computers are recommended
for all or at least part of the testing and analyses.

When measuring response time, data shall be collected fast enough to give good time resolution using the
requirements below for Dts. Measurement of static behavior (dead band, gain, and resolution) generally
does not depend on sample interval and can be performed using existing field instrumentation, with the
sample interval reported.
 — 15 — ANSI/ISA–75.25.01–2000

For a control valve with a pneumatic input signal, the input signal shall be measured as close as possible
to the device input port to avoid input distortion caused by the piping. The total time for the complete input
signal step change, Dtsc, shall meet the requirements given below.

Valve position should be measured as close as possible to the closure member or at least at a location that
closely approximates the closure member position. Care should be taken to avoid measurement errors
due to excessive elastic deformation, clearances, linkages, etc. In all cases, the location of measurement
points shall be reported.

4.3 Measuring instrument requirements

The measurement of each output variable, which includes the combined effects of transducers, any signal
conditioning equipment, and recording equipment shall meet the following minimum requirements.

T
Time constant £ --------
86
- *
20

Resolution -----------------------------------------ö , preferably £ æ -----------------------------------------ö

£ æè valve resolution
ø è
valve resolution
ø
3 10

Uncertainty £ 5 percent of full scale value (where the full scale value is the range of the measured variable
known or estimated as the control valve goes from 0 to 100 percent open, preferably £ 2 percent of full
scale value)

T 86ö
Sampling interval, Dt s £ æ lesser of 0.5 sec or --------
- *
è 20 ø

T 86
Input step change time, Dt sc £ --------- *
20

NOTE — *Instrumentation used to measure the static parameters dead band, gain, and resolution need not meet these requirements
T 86
but time constants, Dt s , and Dt sc , must be reported. For in-process tests, the flowmeter time constant would not have to be £ --------
-,
20
unless it is used to measure T 86 . But if installed in-process instrumentation used to measure T 86 does not meet these
requirements, an external position transducer and recording equipment which meet the above requirements are recommended.

4.4 Process variable

For laboratory and in-process dead band and resolution testing, a process variable shall be measured, if
possible, in addition to the input signal and the position. ISA-TR75.25.02-2000 provides guidance for
choosing the best process variable out of those that may be available at a specific plant or laboratory. The
response flow coefficient, CvR, shown below, is a simplified flow coefficient recommended for use as the
process variable, if measurement of the variables necessary to calculate it is possible. It is used here
because an accurate determination of Cv is outside the scope of this standard and may not be feasible in
many plant and in some laboratory environments. Measurements of dead band and resolution using CvR
would equal those using Cv since changes would be equal. This assumes the flow through the control
valve is fully turbulent and not choked. If that is not the case, such as may occur during “in-process
testing”, the actual Cv could be calculated using the normal ISA sizing equations [see ISA-75.01-1985
(R1995), Flow Equations for Sizing Control Valves]. This response flow coefficient is calculated according
to equations 1 or 2 below.
ANSI/ISA–75.25.01–2000 — 16 —

For liquid flow

q
Eq. 1 C vR = -----------
Dp

or for compressible fluid flow

w 1
Eq. 2 C vR = ----------- ---------
63.3 p 1 x

where, for equation 1, q is the liquid flow rate and Dp is the pressure drop across the valve. For
equation 2, w is the mass flow rate, p1 is the upstream absolute pressure, and x is the pressure drop
Dp
ratio or ------- where Dp is the pressure drop.
p1

To calculate the percent change of the process variable when using the response flow coefficient, defined
by equations 1 and 2 above, the maximum value of CvR must be determined (or estimated) and reported.
The CvR of the valve at 100 percent open can be measured, estimated, or determined from manufacturer-
supplied data. The value used must be stated in the results.

The measured process variable will often fluctuate significantly during the course of the testing because of
normal fluctuations due to disturbances etc. in the process itself or because of electrical noise in a plant
environment and instruments. Curve fitting or averaging routines can therefore be applied to the data
around key points such as the point where T86 occurs and where the total magnitude of the step change is
measured. In all cases, the raw data shall be plotted and if curve-fitting procedures are applied, the curve
fit data should be plotted along with the raw data.

4.5 Nominal test positions

The tests shall typically be performed at 50 percent open and at other positions that may be specified in
lieu of or in addition to this position. Testing at additional, or other, positions may be desirable for valve
types known to have anomalies at openings other than 50 percent. In-process testing may require testing
only at the current operating position plus and minus allowed step sizes. All nominal positions at which
tests are performed shall be recorded.

4.6 Examples of step response

Figure 2 and figure 3 show examples of responses due to input step changes. The response shown in
figure 2 has no overshoot while the one in figure 3 does. In these figures, there is some measurement
noise superimposed on the signal. The input signal is shown along with the response which could be the
valve position or a process variable.
 — 17 — ANSI/ISA–75.25.01–2000

51.2
Dtw
Dtsc
51.0

50.8 Input signal

Response
Ds
DY
50.6

0.865DY
50.4

50.2

50.0

Td
T86
49.8
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Time

Figure 2 — Typical step change and response without overshoot

NOTE — The response gain, Gz, equals DY/Ds.

When the valve input signal suddenly changes, the valve begins to respond (if the input signal change is
large enough) after some delay or dead time,Td. The response then begins moving toward its final value
like that shown, often exponentially. The signal is held constant after the step for a specified amount of
time, Dt w, to allow the response to reach its final new steady state value. The response time, T86, is
defined as the time it takes for the response to reach 86.5 percent of its final value from the initiation of the
step.

Since measurement noise is often superimposed on the response, averaging and curve fitting may be
helpful. If performing the tests manually, this may have to be done visually from a plot. Plots of the raw
data shall be made which could be used later or by others to verify calculations as required.
ANSI/ISA–75.25.01–2000 — 18 —

51.2
settling time

51.0

50.8 overshoot
magnitude

Input Signal
50.6 Response

DY
50.4

50.2

50.0

49.8
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

Time

Figure 3 — Step response with some overshoot

5 Tests specified for each of three test environments

Detailed test procedures required for each of the three test environments are listed below along with notes
and precautions.

5.1 Bench tests

Bench tests are usually the simplest to perform and often provide much useful information. The results,
however, can be significantly different than results from laboratory or in-process tests because there is no
flow (see ISA-TR75.25.02-2000). Requirements are listed below.

Valve configuration — Complete valve with packing configuration that would be specified for intended
service. The valve may or may not be pressurized but packing should be tightened as it normally would be
for typical, or specially defined, conditions. The procedure used for tightening the packing shall follow the
manufacturer’s instructions and be documented. The nominal valve position shall be set at 50 percent
unless otherwise specified.

Actuator assemblies can also be tested separately (not attached to the valve body assembly) when
permitted by the user and preconditioned to all applicable points. Actuator assemblies shall also be
installed in a test fixture that includes a normal control valve packing box unless manufacturer and user
agree to alternate procedures. The packing shall be tightened as per manufacturer specifications. The
test report shall clearly identify the actuator tested, the stem friction measured or estimated as available,
the procedure used to tighten the packing, and the operating temperatures and pressures.
 — 19 — ANSI/ISA–75.25.01–2000

If a valve is tested in a condition other than that described above, that condition shall be described.

Special considerations — Tapping or vibrating the valve under test is not allowed unless required and
specified in the test report.

Measured variables — Input signal and valve position (see 4.2)

Applicable test procedures — Valve cycling (10 times if possible), friction measurement (see annex A),
baseline test (see 6.1), small step tests (see 6.2), and response time tests (see 6.3)

5.2 Laboratory tests

Laboratory tests are performed in a laboratory with flow. Flow shall be fully turbulent and not choked unless
otherwise specified and noted. These tests represent in-process tests more closely than bench tests.
Applicable requirements are listed below.

Valve configuration — Complete valve mounted in flow line set at a nominal position of 50 percent unless
otherwise specified. Packing tightened as it normally would be for typical conditions unless specified
otherwise. With valve pressurized with test fluid, cycle valve open then closed 10 times while measuring
pressure drops and flows. Then measure total friction (see annex A).

Special considerations — No tapping or extra vibration is permitted. However, there will be some vibration
with the flow, which may be measured especially if it appears to influence the test results.

Measured variables — Input signal, position (see 3.14), and process variable for baseline, small step, and
small step tests (see 6.2).

Applicable procedures — Baseline test (see 6.1), small step test (see 6.2), and response time test
(see 6.3).

5.3 In-process tests

In-process tests give valve response in actual, or close-to-actual, process conditions. The range of test
conditions may be more limited than that possible in laboratory testing however. It may also be more
difficult to get good measurements. Valve input and measurements of some process variables can
sometimes be taken directly from existing plant instrumentation if it has the required time constant,
sampling rate, resolution, and accuracy. The following requirements apply.

Valve configuration — Complete valve running at allowed process conditions. Measure or estimate, giving
method of estimation, total friction. Perform tests at positions and conditions required. Sometimes, only
operation close to the existing operating conditions may be permitted.

Special considerations — Limitations in plant operation procedures or safety requirements may not allow
the complete test as defined here.

Measured variables — Input signal, position (see 4.2), and process variable for baseline, small step, and
response time tests (see 4.4).

Applicable procedures — Baseline test (see 6.1), small step test (see 6.2), and response time test
(see 6.3).
ANSI/ISA–75.25.01–2000 — 20 —

6 Detailed test procedures

6.1 Baseline test

The baseline test is normally conducted first but is an optional test. It is used to evaluate measurement
noise, the presence of limit cycling of the valve or other similar behavior, and to determine the baseline
response time,T86b. Figure 4 shows the input signal and an example of the position and the response
during the test. The following steps are included in this test.

Set control signal to desired base value and allow the valve to come to its steady state condition. When
performing in-process tests, the control signal will normally already be at the desired setting and the
controller will be put on manual.

Monitor variables for three minutes using a sample interval, Dts, no larger than 0.5 seconds or
T 86
- , whichever is shorter.
--------
4

Step input up two percent and continue monitoring variables for another one minute or longer.

Repeat stepping up until there is movement, then step one more time to get the full response.

Step input down by two percent and continue monitoring variables for another minute.

Repeat stepping down in two percent increments until the valve position returns to approximately its
starting position.

Evaluate the data for evidence of limit cycling. If there is any, estimate the peak-to-peak magnitude and
period of the limit cycling.

For the last one-minute segment in the up direction, determine the response time,T861. If there is
overshoot, measure its magnitude and the elapsed time from the step initiation until reaching the final
position.

For the last one-minute segment, after stepping down, determine the response time,T862. If there is
overshoot, measure its magnitude and the elapsed time from the step initiation until reaching the final
position.

Determine the base response time T86b as the greater of T861or T862 (see figure 4).
 — 21 — ANSI/ISA–75.25.01–2000

56

54

DYZ
Input Signal, 0.865DYZ
52 Position Response, YZ
T861

50
T862

48

Other Response, Y
46

44
0 60 120 180 240 300 360 420
Time (sec)

Figure 4 — Example step and response during baseline test

If there was any overshoot, determine the overshoot and the overshoot time from the largest overshoot for
the increasing or decreasing input signal steps.

6.2 Small step test

The small step test is performed to determine dead band and resolution. This test may be omitted by
agreement if the response time test will provide the required information to the required accuracy.
Figure 5 shows the signal versus time for a typical small step test. It begins after setting the input signal to
the nominal value and waiting at least three minutes if the optional baseline test was not run or 30 seconds
if the baseline test was run (it can begin right after the optional baseline test) at which point, step up Ds and
then wait a specified time, Dt w. Monitor input signal, position, and (for laboratory and in-process testing)
process variable (or variables necessary to determine response coefficient, CvR) with a sampling interval
of Dt w. Continue this process for n steps. At this point, wait another two time periods, 2Dt w, then decrease
the signal n number of steps, and wait Dt w after each step. Wait another 2Dt w again, and repeat the same
series of steps up and down. The parameters Ds, Dt w, Dt s, and Dt sc must meet the following criteria:

Ds £ ½ (smaller of resolution or dead band)

Since approximate resolution and dead band may not be known beforehand, one can use Ds = 0.1% full
scale and then verify afterwards that specified conditions have been met. It is possible that dead band and
resolution are smaller than the limit cycle peak-to-peak magnitude. If that is the case, the true dead band
and resolution can not be measured but their values can be stated to be no greater than the limit cycle
peak-to-peak magnitude.
ANSI/ISA–75.25.01–2000 — 22 —

n ³ greater of 4 or the quantity 1.2 (dead band + resolution)

---------------------------------------------------------------------------
Ds
This requirement should assure that there be at least one step in addition to the step causing initial
movement.
T 86b
Dt s £ lesser of æ ------------- or 0.5 secö
è 20 ø

If T86b is not available because the optional baseline test was not run, an approximate T86 may have to be
determined using Dts = 0.5, or a lower value, during this test and then adjusting Dts accordingly using 2T86
in place of T86.

Dt w ³ 4 (process measurement time constant) or ³ 30 sec if the process measurement time constant is not
known

T 86b
Dt sc £ ------------
20

See note above if baseline test was not performed.

50.7

2Dtw

50.6

50.5

D tw
50.4

50.3

Ds
n steps
50.2

50.1

50.0
³ 30 sec

49.9
-60 0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900

Time

Figure 5 — Signal sequence for small step test

 — 23 — ANSI/ISA–75.25.01–2000

6.3 Response time tests

The response time tests consists of a series of steps designed to determine the response time, T86, versus
step size, Ds, at each of an increasing sequence of step sizes. Response time tests can also provide
approximate values for dead band and resolution. If less precise values for dead band and resolution are
acceptable, response time tests can be used in place of small step tests. The wait time at the nominal
input signal at the beginning can also be increased to two minutes to determine limit cycling rather than
performing the optional baseline test.

54

53
Dtw
(next cycle)
52

Dtw Ds
51 for next cycle
(2X previous cycle)
Ds
50

49

48

one cycle next cycle

47

46
0 1 2 3 4 5 6 7 8 9 10

Time

Figure 6 — Sample signal step sequence for response time tests.

NOTE — The normal nominal position of 50 percent is shown but may have other values as required.

For each step size, a step (or series of steps) up, two (or more) steps down, and then a step (or series of
steps) up, each followed by a wait, shall be taken. The number of steps down, ndown, is two times the
number of steps up so the signal returns the nominal value after the second set of steps up. Figure 6
shows a case where the number of steps up, nup, equals one, the number of steps down, ndown, equals
two, and the number of steps back up, nup, equals one. The input signal and position are recorded during
each step. The step size for the first response test shall be set equal to the smallest step size equal to or
just greater than the resolution (or dead band if it is smaller) from the step size sequence listed below. If
the small step test was not run and the step response test is used to determine the approximate dead band
and resolution, the complete standard step size sequence below (unless limited by operating conditions),
starting with a step size of 0.1 percent and ending with 10 percent, shall be performed unless dead band
and resolution magnitudes or requirements are known to be higher. The minimum number of steps up,
nup, required for each step size are set below so the total input signal change (the number of steps times
the step size) in one direction exceeds the quantity 1.2 (dead band + resolution). This should assure that
ANSI/ISA–75.25.01–2000 — 24 —

the signal changes enough to overcome deadband and to move at least one more time in the same
direction. Very small step sizes may therefore require more steps. Subsequent step sizes continue with
the sequence listed below starting with the next larger step size and increasing until 10 percent steps (or
biggest step allowed by laboratory or in-process constraints). Values of T86 are then determined for each
step size from the position data and can be tabulated or presented in a plot. The requirements are listed
below.

Data from each response time test is used to determine the gain Gz by taking the response, DZ from a step
near the end of a series of steps in the same direction where there appears to be a full response, and
dividing by Ds. The gain found from the 2 percent response step test, GZ02, is used as a reference gain to
GZ
determine the gain ratio ------------
-.
G Z02

Ds — for the first response time test, set equal to the smallest step size equal to or just greater than the
dead band from the step size sequence.

0.1, 0.2, 0.5, 1.0, 2.0, 5.0, and 10.0 percent of input signal span. (If the small step size test was not
performed, use this whole sequence if possible unless dead band and resolution magnitudes or
requirements known to be larger.) Larger step sizes such as 20 and 50 percent may be used if desired.

Then continue until a response time test has been performed with the step size set to each succeeding
step size in the above sequence. For example, if the resolution is found to be 0.3 percent and is smaller
than the dead band, perform response time tests with Ds set to 0.5, 1, 2, 5, and 10 percent of input signal
span. Extra tests can be performed at other step sizes as required. (The maximum step size may have to
be limited in laboratory and in-process testing to avoid water hammer, etc.)

Dt w ³ 5(T86)

Since T86 is not known for the particular step size in advance, the minimum allowed Dt w for the previous,
smaller step can normally be used to meet this requirement.

T 86
Dt s £ ---------
20

T 86
Dt sc £ ---------
20

1.2 (dead band + resolution)

n up ³ ---------------------------------------------------------------------------
Ds

n down = 2n up

7 Presentation of test results

7.1 General information

The following general information about the test article is required in the presentation of test results. Any
other conditions affecting the test results (such as deviations from recommended conditions) shall be
reported.
 — 25 — ANSI/ISA–75.25.01–2000

Description of the tested valve, actuator, and positioner, including name of manufacturer, model, serial
number, single- or double-acting, and air action.

Description of the test equipment used, including time constants of transducers and signal conditioning
instruments used to measure each variable, names of persons testing and reducing data, and date of test.

Valve calibration results: signal range, range of valve position, and packing system (e.g., tightened to the
normal operating state, as defined by the manufacturer), positioner gain, and cam characteristic as
applicable. Where possible, the friction load shall be measured (annex A) and reported. Identify the point
on the valve assembly where position measurement is taken relative to closure member. Record the
settings of any other adjustable parameters that could affect the rest results. For laboratory and in-process
tests, identify the process variable including the location of measurement devices and the range used to
calculate percent of span. The description of the process variable should also include an estimate of the
response time between valve movement and the measured change in response variable.

7.2 Test results

Test results shall be presented in tabular and graphical formats that identify the minimum requirements
indicated below as applicable. Figure 7 and figure 8 are example plots showing results from small-step
tests and step response tests.

7.2.1 Baseline test

7.2.1.1 Nominal position

7.2.1.2 Test parameters Dt s and Dt sc

7.2.1.3 Test results T86, limit cycling peak-to-peak magnitude and period (if any), and overshoot
magnitude, and settling time

7.2.2 Small step test

7.2.2.1 Nominal position

7.2.2.2 Test parameters Dt s, Dt sc, Dt w, Ds, and n

7.2.2.3 Test results dead band and resolution

7.2.2.4 Limit cycling peak-to-peak magnitude and period (if any) it baseline test not performed

7.2.3 Response time tests

7.2.3.1 Nominal position

7.2.3.2 Test parameters Dt s, Dt sc, Dt w, nup, ndown, and Ds for each step size used

GZ
7.2.3.3 Test results — for each step: T86, overshoot magnitude, and time, and gain ratio ------------
-
G Z02

7.2.3.4 Limit cycling peak-to-peak magnitude and period (if any) if baseline test not performed

7.2.3.5 Range of dead band and resolution if small step tests not performed
ANSI/ISA–75.25.01–2000 — 26 —

7.2.4 Results of additional tests or additional data or characteristics may be included at the
manufacturer’s option or user’s request.

Z(%) I(%) Q(%) CR(%)

85.5 84.5 80.0 55.5
Example Valve S / Actuator T / Positioner U, packing per instructions
85.0 84.0 Tested at Nominal 748 gpm, 3 psid 79.5 55.0

84.5 83.5 0.3% < Resolution £ 0.4% 79.0 54.5

1.3% £ Dead Band < 1.5%
84.0 83.0 78.5 54.0

83.5 82.5 78.0 53.5

Actuator Position, Z
83.0 82.0 (% of transmitter span) 77.5 53.0
no response
82.5 81.5 upon reversal 77.0 52.5

82.0 81.0 Input Signal, I (% of 4-20 mA) 76.5 52.0

CR (% of max. Cv), Q/Ö(DP/G)
81.5 80.5 76.0 51.5
no response while continuing
81.0 80.0 75.5 51.0
in same direction FlowRate, Q (%)
80.5 79.5 75.0 50.5

80.0 79.0 74.5 50.0

79.5 78.5 74.0 49.5

1100 1300 1500 1700 1900 2100 2300 2500
Time (seconds)

Figure 7 — Sample data from small-step test (Ds=0.13%)

performed in a process loop

NOTE — Calculated values of dead band and resolution shown based on the response coefficient CR calculated from Q (which was
not a true representation of the valve response) and the valve pressure drop (DP). Note that the measured position, Z, would not be
a true representation of valve response in this case, possibly because of lost motion between the position measurement point and the
closure member.
 — 27 — ANSI/ISA–75.25.01–2000

Step Response, T86, Maximum

Four Valves

100

Valve A
Valve B
Valve C
Valve D

10

0.1
0.1 1 10
Step Size, Percent

Figure 8 — Sample plot showing step response, T86, versus

step size for four different valves

8 References
ANSI/ISA-TR75.25.02-2000, Control Valve Response Measurement from Step Inputs

Control Valve Dynamic Specification, Version 2.1, 3/94, published by EnTech Control Engineering Inc.,
Toronto

ISA-75.01-1985 (R1995), Flow Equations for Sizing Control Valves

ANSI/ISA-75.05.01-2000, Control Valve Terminology

ANSI/ISA-51.1-1979 (R1993), Process Instrumentation Terminology

ANSI/ISA-75.13-1996, Method of Evaluating Performance of Positioners with Analog Input Signals and
Pneumatic Output
This page intentionally left blank.
 — 29 — ANSI/ISA–75.25.01–2000

Annex A — Sliding Friction Measurement

The procedures below can be used to measure sliding friction in a control valve system with a pneumatic
actuator. Before performing the test, packing should be tightened using manufacturer-recommended
procedures and the valve should be cycled the specified number of cycles. This method assumes that the
friction and spring forces (or torques) at any particular position are the same whether stroking open or
stroking closed.

a) Increase (or decrease) input signal at a rate of 2 percent per second until reaching the desired stroke
while measuring the position and the pressures on the top and bottom of the piston or diaphragm as
the actuator moves. Ordinarily the control valve will be stroked from fully-open to fully-closed (or fully-
closed to fully-open) in this step.

b) Stroke valve in opposite direction by decreasing (or increasing) input signal at a rate of 2 percent per
second until reaching the desired stroke while, again, measuring the same variables measured in
step 1 above.

c) Calculate the friction force at any specified position for a linear valve using the formula

[ ( p b A b – p t A t ) incr – ( p b A b – p t A t ) decr ]
Eq. A1 F f = ----------------------------------------------------------------------------------------------------
2

where

Ff = friction force, lbf

Ab = effective area on bottom of piston or diaphragm

At = effective area on top of piston or diaphragm

Pb = measured pressure on bottom of piston or diaphragm at specified position

Pt = measured pressure on top of piston or diaphragm at specified position

incr, decr = subscripts meaning increasing, and decreasing signal respectively

or, for a rotary valve, use the formula

L [ ( p b A b – p t A t ) incr – ( p b A b – p t A t ) decr ]
Eq. A2 T f = -------------------------------------------------------------------------------------------------------
-
2

where

L = effective moment arm at the particular position

Tf = torque due to friction

and other variables are the same as defined above. (Note that pressures can be absolute or gage but
must be consistent. Also note that for single-acting actuators or if A b equals A t, equations A1 and A2 can
be simplified significantly). No more than two significant figures on friction should be reported.
This page intentionally left blank.
This page intentionally left blank.
Developing and promulgating technically sound consensus standards, recommended practices, and
technical reports is one of ISA’s primary goals. To achieve this goal, the Standards and Practices
Department relies on the technical expertise and efforts of volunteer committee members, chairmen,
and reviewers.

ISA is an American National Standards Institute (ANSI) accredited organization. ISA administers
United States Technical Advisory Groups (USTAGs) and provides secretariat support for
International Electrotechnical Commission (IEC) and International Organization for Standardization
(ISO) committees that develop process measurement and control standards. To obtain additional
information on the Society’s standards program, please write:

ISA
Attn: Standards Department
67 Alexander Drive
P.O. Box 12277
Research Triangle Park, NC 27709

ISBN 1-55617-742-9