Professional Documents
Culture Documents
ANSI/ISA-75.25.01-2000 (R2010)
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISBN: 978-1-936007-49-3
Copyright © 2010 by the International Society of Automation. 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
--`,,```,,,,````-`-`
Preface
This preface, as well as all footnotes and annexes, is included for information purposes and is not part of
ISA-75.25.01-2000 (R2010).
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 the International Society of Automation (ISA),
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: standards@isa.org.
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 DOES NOT TAKE ANY POSITION WITH RESPECT TO THE EXISTENCE OR
VALIDITY OF ANY PATENT RIGHTS ASSERTED IN CONNECTION WITH THIS DOCUMENT, AND
ISA DISCLAIMS LIABILITY FOR THE INFRINGEMENT OF ANY PATENT RESULTING FROM THE
USE OF THIS DOCUMENT. USERS ARE ADVISED THAT DETERMINATION OF THE VALIDITY OF
ANY PATENT RIGHTS, AND THE RISK OF INFRINGEMENT OF SUCH RIGHTS, IS ENTIRELY THEIR
OWN RESPONSIBILITY.
OTHER PATENTS OR PATENT CLAIMS MAY EXIST FOR WHICH A DISCLOSURE OR LETTER OF
ASSURANCE HAS NOT BEEN RECEIVED. ISA IS NOT RESPONSIBLE FOR IDENTIFYING
PATENTS OR PATENT APPLICATIONS FOR WHICH A LICENSE MAY BE REQUIRED, FOR
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISA REQUESTS THAT ANYONE REVIEWING THIS DOCUMENT WHO IS AWARE OF ANY PATENTS
THAT MAY IMPACT IMPLEMENTATION OF THE DOCUMENT NOTIFY THE ISA STANDARDS AND
PRACTICES DEPARTMENT OF THE PATENT AND ITS OWNER.
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 ISA75.25 and approved
ANSI/ISA-75.25.01-2000:
NAME COMPANY
The following people served as members of ISA Committee ISA75 and approved
ANSI/ISA-75.25.01-2000:
NAME COMPANY
--`,,```,,,,````-`-`,,`,,`,`,,`---
ANSI/ISA-75.25.01-2000 was approved for publication by the ISA Standards and Practices Board on
30 December 2000:
NAME COMPANY
--`,,```,,,,````-`-`,,`,,`,`,,`---
C. Williams Eastman Kodak Co.
G. Wood Graeme Wood Consulting
The following people served as members of ISA Subcommittee ISA75.25 and reaffirmed
ISA-75.25.01-2000 (R2010):
NAME COMPANY
The following people served as members of ISA Committee ISA75 and reaffirmed
ISA-75.25.01-2000 (R2010):
NAME COMPANY
This standard was approved for reaffirmation by the ISA Standards and Practices Board on
--`,,```,,,,````-`-`,,`,,`,`,,`---
3 June 2010:
NAME COMPANY
--`,,```,,,,````-`-`,,`,,`,`,,`---
Contents
3 Definitions ............................................................................................................................................ 11
8 References........................................................................................................................................... 29
--`,,```,,,,````-`-`,,`,,`,`,,`---
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.
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 (R2010) 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 (R2010) make use of terms as defined in 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
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 (R2010). Those used only in ISA-TR75.25.02-2000 (R2010) are
marked with (*).
3.1 backlash:*
in process instrumentation, a relative movement between connected mechanical parts, resulting from
looseness when motion is reversed [ISA-51.1-1979 (R1993)]. Sometimes also referred to as slop, lost
motion, or free play.
Output
b
a c
d
a < resolution ≤ b
Time
--`,,```,,,,````-`-`,,`,,`,`,,`---
G R = G Z / GZ02
3.8 hunting:*
an undesirable oscillation of appreciable magnitude, prolonged after external stimuli disappears [ISA-
51.1-1979 (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.
--`,,```,,,,````-`-`,,`,,`,`,,`---
p. 14]. See also hunting [ISA-51.1-1979 (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.
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
ISA-51.1-1979 (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
ISA-51.1-1979 (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 (R2010), expressed as a percent of span.
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
(R2010), the step response can be stem position, flow, or another process variable.
G Z = ΔZ / Δs
3.23 static:
means without motion or change [McGraw-Hill, "Dictionary of Scientific and Technical Terms," sixth
edition, 2002]; 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.
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).
G X = ΔX / Δs
--`,,```,,,,````-`-`,,`,,`,`,,`---
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.
--`,,```,,,,````-`-`,,`,,`,`,,`---
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.
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 Δt s . 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.
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, Δt sc , 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.
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.
T86
Time constant ≤ *
20
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 ⎞
Sampling interval, Δt s ≤ ⎜⎜ lesser of 0.5 sec or 86 ⎟⎟ *
⎝ 20 ⎠
T86
Input step change time, Δt sc ≤ *
20
NOTE — *Instrumentation used to measure the static parameters dead band, gain, and resolution need not meet these
requirements but time constants, Δt s , and Δt sc , must be reported. For in-process tests, the flowmeter time constant would not
T86
have to be ≤ , unless it is used to measure T86 . But if installed in-process instrumentation used to measure T86 does not
20
meet these requirements, an external position transducer and recording equipment which meet the above requirements are
recommended.
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 (R2010) 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 C v 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 C v 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 C v could be calculated using the normal ISA sizing equations [see ANSI/ISA-
75.01.01 (60534-2-1 Mod)-2007, Flow Equations for Sizing Control Valves]. This response flow
coefficient is calculated according to equations 1 or 2 below.
--`,,```,,,,````-`-`,,`,,`,`,,`---
q
Eq. 1 C vR =
Δp
w 1
Eq. 2 C vR =
63.3 p 1x
where, for equation 1, q is the liquid flow rate and Δp is the pressure drop across the valve. For
equation 2, w is the mass flow rate, p 1 is the upstream absolute pressure, and x is the pressure drop
Δp
ratio or where Δp 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.
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.
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.
51.2
Δtw
Δtsc
51.0
0.865ΔY
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
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, Δt 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.
51.2
settling time
51.0
50.8 overshoot
magnitude
Input Signal
50.6 Response
ΔY
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
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 (R2010)). 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.
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.
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)
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).
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).
--`,,```,,,,````-`-`,,`,,`,`,,`---
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, Δt s , no larger than 0.5 seconds or
T86
, 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 T861 or T862 (see Figure 4).
--`,,```,,,,````-`-`,,`,,`,`,,`---
56
54
ΔYZ
Input Signal, 0.865ΔYZ
52 Position Response, YZ
T861
--`,,```,,,,````-`-`,,`,,`,`,,`---
50
T862
48
Other Response, Y
46
44
0 60 120 180 240 300 360 420
Time (sec)
If there was any overshoot, determine the overshoot and the overshoot time from the largest overshoot
for the increasing or decreasing input signal steps.
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 Δs and then wait a specified time, Δt w . Monitor input signal, position, and (for laboratory and in-
process testing) process variable (or variables necessary to determine response coefficient, C vR ) with a
sampling interval of Δt w . Continue this process for n steps. At this point, wait another two time periods,
2Δt w , then decrease the signal n number of steps, and wait Δt w after each step. Wait another 2Δt w
again, and repeat the same series of steps up and down. The parameters Δs , Δt w , Δt s , and Δt sc must
meet the following criteria:
Since approximate resolution and dead band may not be known beforehand, one can use Δs = 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 cannot be measured but their values can be stated to be no greater than the limit
cycle peak-to-peak magnitude.
If T86b is not available because the optional baseline test was not run, an approximate T86 may have to
be determined using Δts = 0.5, or a lower value, during this test and then adjusting Δt s accordingly using
2T86 in place of T86 .
Δt w ≥4 (process measurement time constant) or ≥ 30 sec if the process measurement time constant is
not known
T86b
Δt sc ≤
20
See note above if baseline test was not performed.
50.7
2Δtw
50.6
50.5
Δtw
50.4
50.3
Δs
n steps
50.2
50.1
50.0
≥ 30 sec
--`,,```,,,,````-`-`,,`,,`,`,,`---
49.9
-60 0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900
Time
The response time tests consists of a series of steps designed to determine the response time, T86 ,
versus step size, Δs , 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
Δtw
(next cycle)
52
Δtw Δs
51 for next cycle
(2X previous cycle)
Δs
50
49
48
46
0 1 2 3 4 5 6 7 8 9 10
Time
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 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 G z by taking the response, ΔZ 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 Δs . The gain found from the 2 percent response step test, G Z 02 , is used as a reference
GZ
gain to determine the gain ratio .
G Z02
Δs — 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 Δs 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.)
Δt w ≥ 5( T86 )
Since T86 is not known for the particular step size in advance, the minimum allowed Δt w for the previous,
smaller, step can normally be used to meet this requirement.
T86
Δt s ≤
20
T86
Δtsc ≤
20
ndown = 2 nup
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.
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.
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.
7.2.1.3 Test results T86 , limit cycling peak-to-peak magnitude and period (if any), and overshoot
magnitude, and settling time
7.2.2.4 Limit cycling peak-to-peak magnitude and period (if any) it baseline test not performed
7.2.3.2 Test parameters Δt s , Δt sc , Δt w , nup , ndown , and Δs for each step size used
GZ
7.2.3.3 Test results — for each step: T86 , overshoot magnitude, and time, and gain ratio
G Z 02
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
7.2.4 Results of additional tests or additional data or characteristics may be included at the
manufacturer’s option or user’s request.
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 ( ΔP ). 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.
--`,,```,,,,````-`-`,,`,,`,`,,`---
Valve A
Valve B
Valve C
Valve D
10
0.1
0.1 1 10
Step Size, Percent
8 References
ISA-TR75.25.02-2000 (R2010), Control Valve Response Measurement from Step Inputs
Control Valve Dynamic Specification, Version 2.1, 3/94, published by EnTech Control Engineering Inc.,
Toronto
ANSI/ISA-75.01.01 (IEC 60534-2-1 Mod)-2007, Flow Equations for Sizing Control Valves
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
where
where
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.
--`,,```,,,,````-`-`,,`,,`,`,,`---
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: 978-1-936007-49-3
--`,,```,,,,````-`-`,,`,,`,`,,`---