Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

3.2 Design Basis Valve Stem Thrust Fstem rej = stem rejection load = Pup Astem

For rising-stem valves (gate and globe Fdisc = f ( Pup − Pdown ) A DP

valves), the thrust needed to operate the
valve is a summation of these types of loads: Note the ± sign for the stem rejection
load. This appears because the stem
a. Loads not dependent on the fluid rejection load is always out of the valve
conditions. These include loads such body, thus it resists closure and assists
as disc and stem weight and packing opening.
drag. The disc and stem weight is
often small compared to the packing 3.2.1 Gate Valve Stem Thrust
drag, therefore, this load is often
referred to as the packing load. Figure 3-3 is a data plot showing
b. Loads caused by the valve internal stem thrust measurements taken during
pressure trying to expel the valve testing of the subject 6-inch gate valve. The
stem from the valve body in a piston- thrust measurement is in the negative
like effect. This is called the stem convention, a feature of the instrumentation
rejection load. The stem rejection used to obtain the data (negative thrust
load includes all loads dependent on indicates compression in the stem, and
system pressure. positive thrust indicates tension in the stem).

Loads caused by the differential

pressure acting on the valve disc surfaces.
This is called the disc load or valve DP load.
It includes all loads dependent on valve
differential pressure.

For rising-stem valves:

Equation (3-1)
Where

Fstem = total stem load At the beginning of the closing

stroke, the measured stem thrust is low, but
Fpack = stem packing load as the disc approaches the seat and restricts
the flow, the thrust increases. Flow isolation

USNRC Technical Training Center 3-13 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

is indicated by the point in the trace where load for a valve operating in the closing
the thrust stabilizes momentarily. At this direction:
point, the disc is riding fully on the
downstream valve body seat. The thrust Fstem = Fpack + Fstem rej B Felps + Fnet stem
increases abruptly a moment later, when the
disc wedges tightly between the upstream Equation (3-2)
and downstream valve body seats, and the Where
thrust continues to increase until the torque
switch trips. Motor inertia and other effects Fstem = total stem load
cause a small increase in the thrust after
torque switch trip. In this test program, the Fpack = stem packing load
valve actuator's torque switch was set higher
than normal, to ensure that the valve would Fstem rej = stem rejection load = Pup Astem
close all the way before the torque switch
tripped, so that we could obtain all the data Pup = upstream pressure
we needed from the test. With the exception
of the high torque switch setting (in this data
plot, torque switch trip occurs at a higher Dstem = stem diameter
thrust than normal), this figure is typical of
valve thrust data, and serves to illustrate Felps = elliptical pressure load
how the magnitude of the stem thrust = ∆P Adisc tan α
changes during the closing stroke. The
crucial load occurs at or just after flow ∆P = differential pressure across disc
isolation, when the maximum differential = Pup – Pdown
pressure is applied to the maximum surface
area of the disc. Pdown = downstream pressure

The various loads that contribute to

the total stem load are illustrated in Figure
3-4, a sketch showing the disc engaged Dmean = mean diameter of the disc seat
against the downstream seats during a
closing stroke, after flow isolation but α = seat angle from vertical
before the disc is fully wedged in the seats.
The angle of the disc in this sketch is tan α = 0.08749 for a seat angle of
exaggerated for clarity. The following 5 degrees
formula mathematically represents the stem Fnet stem = net stem load (discussed in detail
later)

USNRC Technical Training Center 3-14 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

Using this formula, we can begin As stated earlier, the stem rejection
calculating an estimate of the stem load for load is the load caused by the pressure inside
our example valve, using valve dimensions, of the valve trying to expel the stem. It
design basis conditions, and other resists motion during valve closure but
parameters as input to the formula. assists during opening. The stem rejection
load resists motion during valve closure (but
Minor loads assists during opening). Stem cross-sectional
area times upstream pressure equals the stem
The most significant load rejection load (Fstem rej in Figure 3-4).
represented in the previous equation is the
net stem load, namely, that portion of the
stem load caused by resistance to motion at
the disc/seat interface. Our discussion
proceeds with an evaluation of the other
three, less significant loads.

In precise terms, the stem packing

load applicable to the closing direction is the
friction at the stem packing seal, minus the
weight of the disc and stem. The packing
load can be directly measured in an
instrumented test with no flow or pressure,
or it can be calculated from the results of a Additional stem rejection forces can
test with pressure but no flow (by occur in wedge-type gate valves where the
subtracting the stem rejection load from the upstream and downstream pressures acting
total load). Assuming no fluctuation in the on the angled wedge surface cause
actual packing friction during the no- additional vertical forces. Often, full
pressure, no-flow test sequence, the differential pressure is used because the
measured packing load will be a little higher point of interest during the closing stroke is
during opening than during closing, because at or immediately after flow isolation, when
of the weight of the disc and stem. Actual the maximum upstream pressure is pushing
packing loads can vary considerably, against the maximum applicable disc area,
depending on the type of packing and on the and the disc is sliding on the downstream
packing adjustment. Often, a default value valve body seat, just before wedging. The
of 1,000 lbf per inch of stem diameter is elliptical pressure load on the top of the disc
used. (Felps in Figure 3-4) is the differential
pressure times the area of the ellipse. The

USNRC Technical Training Center 3-15 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

length of the ellipse is defined by the mean Net stem thrust

diameter of the seat, The elliptical pressure

load assists during valve closure (but resists
during opening).

(A note about differential pressures:

Surface sliding load Normal component
occasionally results of tests using cold water
Total normal load Sliding component
or water with high subcooling show a
downstream pressure less than 0 psi gauge,
Sliding component
approaching 0 psi absolute. The occurrence
of this negative gauge pressure immediately
Net valve
downstream of the disc at flow isolation is a horizontal load
Normal component
phenomenon due to the momentum effects
Valve disc/seat α
of the downstream fluid. In these uncommon interface surface

instances, the differential pressure is greater 1Figure 3-5 Net Stem Load During Valve
than the upstream pressure.) Closure

Net Stem Load Figure 3-5 shows the complex

interaction of vertical forces, horizontal
The discussion so far has addressed forces, and disc/seat angles involved in the
the three minor loads imposed on the valve net stem load during valve closure. Figure
stem during closure: the packing friction 3-5 shows only the net stem load, ignoring
load, the stem rejection load, and the the minor loads discussed in the previous
elliptical pressure load. The major stem load paragraphs. The term net stem thrust
as the valve approaches and achieves flow (indicated at the top of Figure 3-5) is
isolation is the net stem load, that is, the equivalent to the net stem load. (In precise
stem load (a vertical load) created by the terms, thrust refers to the force imposed by
disc as it slides on the downstream valve the actuator on the top of the stem, pushing
body seat. If a valve is instrumented to the stem downward; load refers to the
measure stem thrust, one can calculate the resistance to motion at the bottom of the
net stem load as follows: total stem thrust stem. Sometimes, however, the terms thrust
minus packing load minus stem rejection and load are used interchangeably.)
load plus elliptical pressure load equals the
net stem load. If we were to look at the net stem
load in very simple terms, ignoring the disc
angle, we could treat the net stem load as the
differential pressure times the disc area

USNRC Technical Training Center 3-16 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

times the disc friction (coefficient of friction We now proceed with an evaluation
at the disc/seat interface). However, this of the net stem load. The mean seat diameter
simple view has a history of distorting the is used for this calculation. Disc area times
analysis of test results and producing differential pressure equals the horizontal
inaccurate estimates of valve operating force on the disc (see Figure 3-4). Note that
requirements. The following discussion is because of the angle of the seat, this
based on the more complex, more accurate horizontal force is not exactly the same as
view indicated in Figure 3-5. In this view, the normal force (a force that is
the friction at the interface between the disc perpendicular to the seat), which holds the
and the valve body seats is represented as a disc against the seat. The horizontal force is
true coefficient of friction, rather than a disc based on the area of the circular profile of
factor that includes other unknown or hidden the seat when viewed from the horizontal
variables along with the friction. direction.

The following discussion describes Figure 3-5 shows how we resolve the
mathematical formulas that represent the horizontal force (defined above) and the net
view shown in Figure 3-5. These formulas stem thrust (described earlier) into normal
have two important applications. components and sliding components. The
normal component (a) of the horizontal
• In an evaluation of valve operability. force (b) combines with the relatively small
Where the net stem load is the normal component (c) of the net downward
unknown variable, standard or thrust of the stem (d) to create the total
known variables serve as input to the normal load (e). The total normal load times
formulas to produce an estimate of the disc friction produces a sliding load that
the net stem load. represents the most significant load that the
• In an evaluation of test data. Where valve actuator must overcome. That load
the net stem load can be determined plus the sliding component (f) of the
from test measurements, the data can horizontal force equals the total sliding load
be analyzed in terms of the single (g), which is the total load that the sliding
unknown variable, namely, the disc component (h) of the net stem thrust must
friction. overcome during valve closure.

The following discussion

incorporates both of these applications (but
not in a simple two-step sequence).

USNRC Technical Training Center 3-17 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

The mathematical representation of following representation of the net stem

the normal and sliding forces is: load:

Fn = ∆P Adisc cos α + Fnet stem sin α

Fnet stem = net stem load =
Equation (3.3)
Equation (3-5)
Fs = ∆P Adisc sin α - Fnet stem cos α
f = Fs / Fn
Equation (3.4)
= disc friction (coefficient of
friction)
Where

This formula serves as the basis for

Fn = normal force
making predictions of the net stem load and
for evaluating the results of instrumented
Fs = sliding force
tests. Remember, if you know the
differential pressure (and valve dimensions
∆P = differential pressure across disc
and geometry), and if you assume a
conservative value for the disc friction, you
Adisc = disc area( based on mean seat
can estimate the net stem load. If you
diameter)
measure the stem thrust and the differential
pressure, you can calculate the net stem load
α = seat angle from vertical
using the approach described earlier in this
chapter (net stem load equals total stem
cos α = 0.99619 for a seat angle of 5º
thrust minus packing load minus stem
rejection load plus elliptical pressure load),
sin α = 0.08716 for a seat angle of 5º
and you can use the result to determine the
actual disc friction.
Fnet stem = net stem load (the resistance to
motion imposed on stem by disc as it
Estimating the total stem load. The
slides on the seat)
following equation shows the process for
performing the calculation. As presented
Since the sliding force Fs is equal to
here, the equation includes the minor loads
the normal force Fn times the coefficient of
(packing load, stem rejection load, and
friction f, the coefficient of friction at the
elliptical pressure load) discussed earlier in
interface between the disc and the seat can
this chapter, then evaluates the net stem load
be represented as Fs/Fn. Thus, substituting,
as described in the previous paragraphs.
rearranging, and simplifying yields the
Thus, with inclusion of the additive

USNRC Technical Training Center 3-18 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

bounding term, the formula for estimating wedging. If the results of the test thus
the total stem load is: recorded, and expressed in terms of
normalized sliding load versus normalized
Fstem = Fpack + Fstem rej B Felps + normal load, fall within the bounds of the
INL correlation as shown in, Figure 3-6, the
valve can be considered typical. If the
Equation (3-6)
results fall outside those bounds, the valve
Where
cannot be considered typical, and it might be
necessary to use some other method to
f = 0.40 for fluid subcooling less than
estimate the disc friction for the valve.
70ºF

f = 0.50 for fluid subcooling 70ºF or

greater

This is an estimate of the maximum

stem thrust needed to close the valve at
design basis conditions. The estimate is
reasonably conservative without being
excessively so.

Determining applicability of the INL

correlation.

Use of the INL formula to make

estimates like this one requires some reliable Note that use of the INL correlation
strategy for determining that the INL in this manner does not constitute an
correlation, and the corresponding disc extrapolation. With an extrapolation, the
friction value of either 0.4 or 0.5, is disc friction determined from a low-load test
applicable to the valve in question. We is used to estimate the stem thrust at a higher
recommend that the valve be tested at a load. With the INL correlation, the low-load
differential pressure of about 300 to 400 test is used to determine typicality. If
psid or greater. The valve should be typicality is successfully determined, then it
equipped with instrumentation to monitor is acceptable to use the INL formula
upstream pressure, differential pressure, and (Equation 3-7) and the applicable disc
stem thrust, with these parameters recorded friction term (either 0.40 or 0.50) to estimate
during the seating portion of the stroke, that the stem thrust at a higher load.
is after flow isolation but before full

USNRC Technical Training Center 3-19 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

The INL correlation, as shown in For DP < 415 psi (330 psi if the fluid
Figure 3-6, includes a secondary equation subcooling is greater than 70°F)
that narrows the acceptable bounds for the
data scatter at lower loads. At the lower Fstem = Fpack + Fstem rej B Felps +
loads, the upper and lower bounds are not
defined as constant offsets relative to the
Equation (3-8)
normalized normal disc load, but instead are
represented as a ± 30% variation in the
The development of the portion of
friction factor. The threshold that separates
the INL correlation that addresses the lower
the lower loads from the moderate to high
loads is described in NUREG/CR-6100,
loads for the 0.40 disc friction is a
Gate Valve and Motor-Operator Research
normalized normal force Fnn of 415 lbf per
Findings.
in.2 disc area, a value that corresponds with
a differential pressure of about 400 psid in a
If low-load testing shows that the
gate valve with a seat angle of 5 degrees.
valve is not typical of the valves tested by
For the 0.50 disc friction, the threshold is a
INL and used to develop the correlation, the
normalized normal force Fnn of 330 lbf per
reason might be one of the following:
in.2 disc area, corresponding with a
differential pressure of about 320 psid. The
• New valves and newly refurbished
mathematical representation of the
valves tend to operate with unusually
correlation, showing both equations and
low disc friction values. Such valves
including the upper and lower bounding
generally operate at more typical
terms to account for data scatter, is:
disc friction values after they have
been subjected to many strokes at
For DP > 415 psi (330 psi if the fluid
cold water conditions or exposed to
subcooling is greater than 70°F)
hot water or steam conditions. This
process of achieving a stable disc
Fstem = Fpack + Fstem rej B Felps +
friction via multiple strokes and/or
exposure to hot water or steam is
Equation (3-7) sometimes referred to in the
literature as preconditioning.
• Some valves, particularly those
manufactured by Borg-Warner, tend
to operate with higher disc friction.
This issue discussed in NRC
Information Notice 89-61.

USNRC Technical Training Center 3-20 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

• The disc guide clearances in some

valves are such that during closure
against high differential pressure, the
disc tips on the guides enough that
an unusual pressure distribution
around the disc occurs, and the disc
seat does not engage smoothly
against the valve body seat.

The performance of this last group of

valves is called nonlinear, because the
differential pressure times the area of the
disc exposed to pressure does not produce a
linear relationship when plotted against the Figure 3-8 illustrates some of the
stem thrust measurement. In extreme cases, flow and pressure dynamics involved during
damage to the guides and seats can occur, a valve closure in which the disc tips. These
especially if the leading edge of the disc flow dynamics might contribute to the
gouges the valve body seat as the valve nonlinearity in the stem thrust measurements
approaches flow isolation. in a valve closure in the absence of valve
damage.
Gate valves with nonlinear performance
An instrumented, best-effort test at a
Figure 3-7 shows a stem thrust trace differential pressure of at least 200 psid or
from a test in which valve damage did not half the design basis differential pressure
occur, but the performance of the valve was (whichever is smaller) is probably sufficient
nonlinear, as indicated by the hook shape to determine whether the valve’s
just before flow isolation. Compare this performance is linear or nonlinear. Of
figure with Figure 3-3, a stem thrust trace course, the 300 to 400+ psid test used to
displaying the classic, linear response. determine typicality for applicability of the
Valves with nonlinear performance cannot INL correlation (as described above) would
be evaluated using the INL correlation serve this purpose. For valves with non-
described earlier in this chapter. linear performance, it might be possible,
under some circumstances, to use the results
of the best-effort differential pressure test to
make a prediction of the valve's response at
full design-basis differential pressure. A
brief description of that method is presented

USNRC Technical Training Center 3-21 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

here. Additional discussion of this issue is cold water, and hot water near boiling tends
presented in NUREG/CR-6100 Gate Valve to produce a lower disc friction than hot
and Motor-Operator Research Findings. water with a large degree of subcooling. The
test must be conducted with adequate
measurements of stem thrust, upstream
pressure, and differential pressure to provide
input to the following formula:

Fstem = Chook ∆P + Pup Astem + Fpack

Equation (3-9)
Where

Fstem = stem thrust

Chook = hooking factor

∆P = differential pressure
Use of this method for valves with
non-linear performance assumes that valve
Pup = upstream pressure
damage is not a concern. (In some instances,
non-linear performance is accompanied by
Astem = stem area
damage to the guides and/or seats.) In
instances where damage does not occur, the
Fpack = packing drag
best effort differential pressure test
described above (200 psid or half the design
The hooking factor is simply a
basis differential pressure) is probably
numerical value that broadly represents the
sufficient to produce a baseline non-linear
combined effects of the disc area, disc
response that can be used in an
friction, and idiosyncrasies of the pressure
extrapolation. The flow must be sufficient to
distribution, disc/guide contact, and disc/seat
produce the minimum differential pressure
contact for that particular valve. To use this
before flow isolation. In cases where the
formula, solve for the hooking factor, using
design basis conditions include a low degree
the actual packing drag (measured in a no-
of subcooling, it might be a good idea to
load test), the maximum stem thrust
approach or match that degree of subcooling
measured at the bottom of the hook during
in the best effort test, so that any effects due
the best effort test, and the differential
to the flashing of steam in the orifice are
pressure and upstream pressure measured at
duplicated. Remember, too, that hot water
the point in time when the maximum stem
tends to produce a lower disc friction than

USNRC Technical Training Center 3-22 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

thrust was measured. Then use the hooking

factor thus derived, along with the design
basis packing drag (not the same as the
measured packing drag) and the design basis
differential and upstream pressures, to
calculate an estimate of the stem thrust
required at design basis conditions. Note
that this method has not been validated.
Note also that the results of this kind of
extrapolation are applicable only to the
single valve that produced the baseline data.

Some valves are subject to damage

to the guides or seats when exposed to very
high design-basis differential pressures.

Figure 3-9 shows stem thrust traces

from tests in which valve damage occurred
(compare with Figure 3-7 and with Figure 3-
3). Of the six valves we tested in our 1989
NRC/INL valve tests, two suffered
significant guide and/or seat damage during
closure against very high design basis loads. Figure 3-10 is a photograph showing
typical damage.

Figure 3-11 illustrates how tipping of

the disc during closure at high flow loads
can cause damage to either the seats or
guides, depending on which components
bear the excessive load at the point of
contact. We know of no reliable method for
determining in advance which valves are
likely to experience such damage. Similarly,
we know of no reliable method for
predicting the requirements of such valves.

USNRC Technical Training Center 3-23 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

ability of a valve to open can be as

important in the mitigation of an accident as
the ability to close. This was true, for
example, in the loss of feedwater incident
(near-accident) at the Davis Besse plant in
1985.

Figure 3-12 shows a typical stem

force trace from a valve opening test. The
highest load occurs while the disc is sliding
Opening Gate Valves on the seat, while the full disc area is
exposed to the full differential pressure,
The design basis operating before flow initiation. This opening response
requirements for some valves include a corresponds with the typical, linear closing
requirement to be able to open against response described earlier in this section. A
specified loads and at specified conditions. spike occurs at unwedging, a plateau during
In general, however, any valve in a safety- unseating, and a decline in the load after
related system needs to have the capability flow initiation.
to open against any conceivable differential
pressure that might occur in the system in Figure 3-13 shows a stem force trace
which the valve is installed; such capability illustrating the occasional nonlinear opening
might prove essential in a scenario in which response. In this case, the highest load
a valve is intentionally or inadvertently occurs after flow initiation, creating the
closed and then needs to be reopened. The appearance of a "hump" in the trace. We

USNRC Technical Training Center 3-24 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

performed an analysis that showed that this Pup = upstream pressure

hump is not a manifestation of an unusually
high load after flow initiation, but is instead Astem = stem area
a manifestation of an unusually low load
before flow initiation. This analysis is
documented in NUREG/CR-6100, Gate Dstem = stem diameter
Valve and Motor-Operator Research
Findings. Felps = elliptical pressure load
= ∆P Aseat tan α

∆P = differential pressure across disc

= Pup – Pdown

Pdown = downstream pressure

Adisc = disc area

Dmean = mean diameter of the disc seat

α = seat angle from vertical

The formula we use to estimate the tan α = 0.08749 for a seat angle of 5º
opening load is:
f = disc/seat coefficient of friction =
Fstem = Fpack B Fstem rej + Felps 0.50

cos α = 0.99619 for a seat angle of 5º

Equation (3-10)
sin α = 0.08716 for a seat angle of 5º
Where

This formula is basically the same as

Fstem = total stem thrust load
the formula for estimating the closing load,
except for sign changes. Further, the term 80
Fpack = stem packing load
Adisc, an additive term that provides
conservatism to the estimate, is a higher
Fstem rej = stem rejection load = Pup Astem
numerical value than the one used in the

USNRC Technical Training Center 3-25 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

formula for closing. The term for disc

friction in the formula for opening is 0.50
regardless of temperature or degree of
subcooling.

A valve evaluation can use this

formula, with a disc friction value of 0.50, to
estimate a valve's opening thrust, if there is
evidence that the valve's behavior is typical
of the INL test valves that produced the
correlation described below. As with the
formula for closing, we recommend that a
best effort test be performed with
instrumentation sufficient to monitor the The INL correlation, as shown in
stem thrust, upstream pressure, and Figure 3-14, includes a secondary equation
differential pressure, with a focus on that narrows the acceptable bounds for the
collecting data near the beginning of the data scatter at lower loads. The threshold
opening stroke, while the disc is riding on that separates the lower loads from the
the downstream seat before flow initiation. moderate to high loads is a normalized
If the results of the test, in terms of normal force Fnn of 450 psi, a value that
normalized sliding load versus normalized corresponds with a differential pressure of
normal load, fall within the bounds of the about 425 psid in a valve with a seat angle
INL correlation as shown in Figure 3-14, the of 5º. The mathematical representation of
valve can be considered typical, and the INL the correlation, showing both equations and
correlation can be considered applicable. If including the upper and lower bounding
the results fall outside those bounds, the terms to account for data scatter, is:
valve cannot be considered typical, and
some other method must be used to estimate For Fnn > 450 psi
the disc friction for the valve. Figure 3-14
shows the slope and the bounds of the INL Fstem = Fpack B Fstem rej + Felps +
opening correlation, along with the test data
that we used to develop the correlation. Equation (3-11)

USNRC Technical Training Center 3-26 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

For Fnn < 450 psi wedge-shaped disc seals by seating against a
pair of matching valve body seats. The
Fstem = Fpack B Fstem rej + Felps + parallel disc design employs a different
mechanism for sealing. For example, in the
Anchor Darling parallel disc gate valve,
Equation (3-12)
sealing occurs when the disc assembly
reaches the bottom of the valve, after the
The terms ± 80 Adisc and 1.0 ± 0.35
two parallel discs have lowered far enough
provide the correlation with the means to
to cover the two parallel faces of the valve
bound the scatter in the test results upon
body seats. At that point, further downward
which the correlation is based.
motion of the stem actuates a pair of
interfacing wedges that pry the two discs
In general, experience has shown
apart (they are separate pieces), creating a
that the thrust needed to unwedge a valve is
tight fit by locking them in place against the
lower in magnitude than the thrust that
valve body seats. Some parallel disc gate
wedged the valve during the previous
valves of other manufacture use other
closing stroke. The point of interest for an
mechanisms for prying the discs apart toseal
opening stroke is after unwedging but before
the seats, and some use no such mechanism
flow initiation, when the full differential
at all, relying on the upstream pressure to
pressure and the full disc area are effective.
hold the disc assembly against the
For most gate valves, the available thrust at
downstream seat.
this point is determined by the design basis
stem friction and by the actuator capability
at design basis motor conditions, not by the
torque switch setting; most valve actuators
are equipped with a bypass switch,
controlled by stem position, that bypasses
the torque switch during the early part of the
opening stroke, until after flow initiation.

Parallel Disc Gate Valves

Of course, the main difference

between flexible wedge gate valves and
parallel disc gate valves is the shape of the
disc. Figure 3-15 shows a cross section of a
The mathematical representation of
parallel disc gate valve. In the flex-wedge
the closing and opening loads experienced
design discussed earlier in this textbook, the

USNRC Technical Training Center 3-27 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

by parallel disc gate valves is similar to that

of flex-wedge gate valves, but simpler. In ∆P = differential pressure across disc
the absence of a wedge shape, there is no = Pup – Pdown
elliptical pressure load on the top of the disc. Pdown = downstream pressure
More significant, with the parallel design the
normal load on the disc is simply the Adisc = disc area
differential pressure times the disc area, and
the sliding load is simply the net stem thrust.
Thus, the net stem load is simply the Dmean = mean diameter of the disc seat
differential pressure times the disc area
times the coefficient of friction at the f = disc friction (coefficient of
disc/seat interface (disc friction). friction at the interface between the
disc and downstream valve body
Thus, the formula for estimating the seat)
required stem thrust for closing a parallel
disc gate valve (or for analyzing the results For the opening direction, the
of an instrumented test) is: formula is:

Fstem = Fpack + Fstem rej + Fnet stem Fstem = Fpack B Fstem rej + Fnet stem
Equation (3-13) Equation (3-14)
Where
The difference being that the stem
Fstem = total stem thrust load rejection load assists during opening but
resists during closing.
Fpack = stem packing load
The main concern related to the use
Fstem rej = stem rejection load = Pup Astem of these formulas in valve evaluations is the
choice of an appropriate value for the disc
Pup = upstream pressure friction. Before the late 1980s, the U.S.
nuclear industry accepted the use of values
Astem = stem area as low as 0.20, though little or no testing had
been performed to verify the adequacy of
this value. It is likely that the use of this low
Dstem = stem diameter a friction value in valve evaluations could
produce inaccurate, unconservative
Fnet stem = net stem load estimates of the net stem load. The use of
= ∆P Adisc f higher disc friction values, typically 0.40 but

USNRC Technical Training Center 3-28 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

ranging from 0.3.to 0.50, is more typical

today. The industry's shift to the higher Pup Astem = stem rejection load
friction values followed publication of the = upstream pressure times
results of the NRC/INL flexible wedge gate stem cross-sectional area (added to
valve testing and the results of parallel disc the load during closing and
gate valve testing conducted by researchers subtracted from the load during
in the Federal Republic of Germany (results opening)
described in U.S. NRC Information Notice
90-72). The higher friction values are more ∆P = differential pressure
consistent with the results of the NRC/INL
and German tests. Adisc = disc area

Typical Industry Methods and µd = disc factor

Alternative Methods
As used in this formula and applied
Since the mid-1980s, valve to wedge-type gate valves, the disc factor µd
researchers, utilities, and utility is not a true friction coefficient, but serves
organizations have developed several new instead as a catch-all variable that covers the
analytical tools for evaluating the disc friction along with any other parameters
performance and the requirements of motor- not specifically covered by other
operated valves in general and wedge-type components of the formula, including any
gate valves in particular. Before that time, idiosyncrasies associated with this formula's
several tools were in use for evaluating or failure to account for the angle (from
predicting the stem thrust load for a flex- vertical) of the disc/seat interface.
wedge gate valve. Though these tools varied
somewhat in certain details (the definition of Before the mid-1980s, use of a disc
the disc area, the default disc friction factor), factor of 0.30 with the standard industry
they represented basically the same formula. equation was common in U.S. nuclear
We refer to this formula as the standard industry evaluations of valve requirements.
industry equation: By the early 1990s, some evaluations used
disc factors as high as 0.50. The definition
Fstem = Fpack  Pup Astem +∆P Adisc µd of the disc area was not consistent. Use of a
Equation (3-15) larger value for the disc area (one based, for
example, on the outside disc area) produced
Fstem = total stem thrust load a higher estimate of the stem thrust than use
of a smaller value (based, for example, on
Fpack = stem packing load the valve orifice diameter). Use of a smaller

USNRC Technical Training Center 3-29 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

disc factor along with a larger disc area methodology, very similar to the INL
might produce a stem force estimate about correlation.
the same in magnitude as use of a larger disc
factor with a smaller disc area. Pressure Locking and Thermal Binding
of Gate Valves
The standard industry equation is
still used by the U.S. nuclear industry in Pressure locking occurs when the
many valve evaluations. We describe it here high pressure of fluid trapped in the valve
for completeness, so the reader can easily bonnet causes the valve to be difficult to
compare the standard industry equation with open. Thermal binding occurs when thermal
the INL formulas we recommend. Part of expansion/contraction effects squeeze the
our research effort in the early 1990s, as we valve disc between the two seats, likewise
analyzed the results of our gate valve tests, causing the valve to be difficult to open.
consisted of an appraisal of the standard
industry equation. We identified the In general, when a gate valve opens
following deficiencies: against an ordinary differential pressure
load, the single major load the actuator must
• The disc factor of 0.30 was far too overcome is the resistance created by
low; in some instances a disc factor friction at the downstream disc/seat
of 0.50 was too low. interface. Under differential pressure
• It failed to consistently specify the conditions, the upstream pressure tends to
mean seat diameter as the basis for decrease or eliminate the load at the
determining the disc area upstream disc/seat interface and apply the
• It failed to account for the elliptical entire load to the downstream disc/seat
pressure load on the top of the disc interface. Typical formulas for estimating
• It failed to isolate the disc friction, valve operating requirements are based on
instead including it in the disc factor differential pressure across only one
along with other unknown or disc/seat interface.
unspecified variables.
Pressure locking occurs when the
At about the same time as we valve bonnet pressure is higher than both the
published the INL correlation, the EPRI’s upstream and downstream pressures. In most
Nuclear Maintenance Assistance Center gate valves, including most flexible wedge
(NMAC) was developing an improved gate valves, split wedge gate valves, and
equation for gate valve evaluations. The double-disc gate valves, the bonnet cavity
NMAC equation is a proprietary communicates with the area between the
disc faces. The effect is that the pressure of

USNRC Technical Training Center 3-30 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

the fluid between the discs acts on both the both discs respond equally and
upstream and the downstream discs, independently to the pressure of the fluid
introducing resistance to motion at both between the discs. However, the disc
disc/seat interfaces rather than just one. This assembly in a flex-wedge valve is made
adds to the total force necessary to from a single piece of metal, with the
unwedge/unseat the valve disc, compared upstream and downstream halves of the disc
with the ordinary differential pressure case. connected in the center by a hub. As with
The various forces involved in a pressure the double-disc design, the two discs
lock (in a flexible-wedge gate valve) are respond to the pressure of the fluid between
indicated in Figure 3-16; notice in particular them, but the area exposed to the pressure is
the pressure forces between the two disc smaller, because of the presence of the hub.
halves. At its worst, pressure locking causes
the valve to be locked in the closed position, In addition, not all of the area
such that the actuator does not have exposed to the bonnet pressure (in the flex-
sufficient output capacity to open it (the wedge design) responds in a way that results
motor stalls). in additional force at the disc/seat interface.
When the disc assembly is exposed to
pressure locking loads, part of the pressure
load deforms the disc and presses it against
the valve body seat, and part of the pressure
load is reacted in the hub. The effect,
theoretically, is that for valves of a given
size, the more flexible disc design is more
likely to be affected by pressure locking
loads, with a greater stem force necessary to
unwedge a pressure-locked disc assembly.

One other feature of the flex-wedge

gate design contributes to the effects of
Pressure locking loads are much pressure locking. The angle of the disc,
more difficult to predict than ordinary usually about 5 degrees from vertical (in an
differential pressure loads (mentioned above upright valve), creates a horizontal area on
and discussed earlier in this textbook), the top of the disc that is acted on vertically
especially with flex-wedge gate designs. by the difference between the bonnet
(The flex-wedge design is the most widely pressure and the downstream pressure, and
used of all gate valve disc designs.) With another acted on by difference between the
double-disc and split wedge gate valves, bonnet pressure and the upstream pressure,

USNRC Technical Training Center 3-31 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

as shown in Figure 3-17. These areas are cause the thrust needed to open a pressure
typically modeled as elliptical areas defined locked valve to be higher than the value
by the downstream and upstream seat typically calculated by industry formulas for
orifices in the valve when viewed from the design basis differential pressure
above (the axis parallel to the stem). The conditions. Since most valve actuators are
corresponding forces, indicated as Ftop - sized and set according to design basis
Fbottom in Figure 3-16 and as Felps in Figure conditions, the higher thrust demands due to
3-17, resist opening in much the same way pressure locking can exceed the capability of
that the stem rejection load assists opening. the actuator, causing the valve to fail to open
(that is, the torque switch trips or the motor
stalls before the actuator successfully
unseats and opens the valve).

The bonnet pressure that causes

pressure locking can be either hydraulically
or thermally induced. Hydraulically induced
pressure locking can result from various
operational sequences involving low-
pressure system interface with high-pressure
systems, or from system depressurization
during an accident. In a typical scenario, a
valve closed at high pressure might
When a valve opens against normal experience pressure locking if an attempt is
differential pressure, only the downstream made to reopen the valve after both the
Felps force is active; the upstream Felps force upstream and downstream piping has been
is zero, because the bonnet pressure is equal depressurized, and with the high pressure
to the upstream pressure. However, in the remaining in the bonnet. Such a scenario
pressure locked case, both the upstream and occurred, for example, in 1991 at the
downstream Felps loads act to resist opening, Fitzpatrick Nuclear Power Plant (discussed
resulting in yet another increase in the in the next paragraph). Thermally induced
opening load, as compared to the ordinary pressure locking can occur by thermal
differential pressure opening situation. The expansion of water trapped in the bonnet. A
magnitude of the load increase depends on valve closed under cold conditions might
the bonnet, upstream, and downstream experience pressure locking if the valve
pressures. were later heated by a slug of hot fluid
coming into contact with the closed disc, by
Taken together, the load increases convection or conduction from adjacent hot
described in the preceding paragraphs can

USNRC Technical Training Center 3-32 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

piping, or by steam from a line break inside damage had been caused by high pressure
the containment. An instance of thermally produced by heatup of fluid trapped in the
induced pressure locking occurred, for valve bonnet. The heat source was the hot
example, in 1995 at the Susquehanna Steam water in the feedwater system, which has a
Electric Station (discussed in a subsequent connection about three pipe diameters from
paragraph). the HPCI valve. No known attempt to open
the valve was made while the valve was in a
The instance of pressure locking that pressure locked condition, but the utility’s
occurred at the Fitzpatrick station was analysis revealed that the actuator for the
hydraulically induced (NRC Information HPCI valve did not have sufficient thrust
Notice 92-26). The utility hydro-tested the capability to open the valve against the
piping between the inboard and outboard pressure locking load that caused the valve
low-pressure coolant injection (LPCI) damage.
valves. The inboard LPCI valve is a 24-in.
flexible-wedge motor-operated gate valve. Other examples of pressure locking
After the hydro-test, the utility depressurized are discussed in NUREG-1275.
the piping between the valves and filled and
vented the system to return it to service, The magnitudes of possible loads
unaware that high-pressure fluid remained due to pressure locking depend on the valve
trapped in the valve bonnet. About 10 hours design and on the pressures prevalent in the
later the utility commanded the inboard specific system where the valve is installed.
valve to open. The valve actuator was Leaking valves tend to be less susceptible to
energized for about 30 seconds, and then the pressure locking, because the leakage may
circuit breaker tripped. (The normal stroke prevent the bonnet from becoming or
time for this valve is 120 seconds.) The remaining pressurized. However, variations
valve had failed to open. The root cause of due to valve seating conditions, pressure
the failure was pressure locking. conditions, and pressure changes cause
valve leakage to be inconsistent, so that a
An instance of thermally induced valve that leaks under some conditions
pressure locking was reported at the might not leak under other conditions.
Susquehanna Steam Electric Station in 1995 Typical modifications to gate valves to
(NRC Information Notice 96-08). When prevent pressure locking include venting the
performing a valve modification to eliminate bonnet to the high-pressure side by drilling a
the potential for pressure locking, the utility hole through the disc, or by installing a
discovered internal damage to a high- small vent line between the bonnet and the
pressure coolant injection (HPCI) valve in upstream piping. The vent line might or
Unit 1. The utility determined that the

USNRC Technical Training Center 3-33 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

might not be equipped with a check valve or NRC/INL laboratory tests showed pressure
a block valve. locking loads 2.05 to 2.40 times the ordinary
opening pressure load. We do not know why
Thermal binding is a term describing the loads were higher than expected, but we
the effects of heatup and cooldown on wonder if the wedge-to-wedge mechanism
differential expansion and contraction of the in the Anchor Darling valve we tested failed
valve internals. Valves closed in the hot to function normally, such that the disc
condition might experience thermal binding assembly failed to collapse.
loads when the seat rings contract against
the disc after cooldown, as shown in Theoretically, the pressure locking
Figure 3-18. These binding loads may be load in a flexible wedge gate valve should
minor, or they may be so large that the valve be twice the ordinary opening load against
must be reheated in order to free the disc. In the same differential pressure, minus the
theory, a more flexible disc design is less differential pressure times twice the area of
likely to be affected by thermal binding the cross-section of the hub, minus an
loads. (As mentioned earlier, the opposite is unknown variable to account for the
true of pressure locking loads.) stiffness of the disc/hub assembly. Results
from our NRC/INL laboratory tests of a
flexible wedge gate valve showed pressure
locking loads 1.1 to 1.5 times the ordinary
opening pressure load. (These values are
applicable only to the specific valve that
produced the data, and are not applicable to
any other valve.) The U.S. nuclear industry
is preparing a formula for estimating
pressure locking loads; this formula includes
a calculation that accounts for disc/hub
stiffness. The calculation is very
complicated and is still undergoing
validation (in 1999).

Precisely estimating a pressure

We know of no method for
locking or thermal binding load is not a
estimating thermal binding loads. An
simple task. Theoretically, the pressure
analysis of some NRC/INL test data showed
locking load in a parallel disc gate valve
that in instances where valves were
might be as high as twice the ordinary
subjected to thermal and depressurization
opening load against the same differential
cycles that might induce thermal binding
pressure. However, results from our

USNRC Technical Training Center 3-34 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

conditions, the unseating loads ranged from

about 40 to 120% of the previous seating
load. In contrast, the ordinary unseating
load, in the absence of any such thermal
cycles, is typically about 20% of the
previous seating load. The obvious risk is
that a high unwedging load might cause the
motor to stall, leaving the valve closed. The
results of the NRC/INL pressure locking and
thermal binding tests are documented in
NUREG/CR-6611, Results of Pressure
Locking and Thermal Binding Tests of Gate
Valves.

USNRC Technical Training Center 3-35 05/10

Motor-Operated Valves Course Manual Theory of MOV Design Basis Operation

USNRC Technical Training Center 3-36 05/10