Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

PV Limit

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

TECHNOLOGY BULLETIN

COMPANY CONFIDENTIAL

ISSUE : 27

PV Limit Recommendations for Material Combinations


Introduction:
The PV limits noted in this bulletin are general recommendations for the combinations
shown. There is no one PV level that can be provided for all applications as
combinations of slow or high speeds at extremes will cause the limits to change.
Environmental conditions (fluid, presence of abrasives, process pH level, temperature,
presence of a buffer fluid, batch versus continuous operation) also play an important role
in establishing PV limits. The flush arrangement, type of mating ring, mating ring support
surface and surface exposure of the mating ring to the product or buffer/barrier fluid will
have an effect that can extend or lower the PV limit. Finally, the overall distortion of the
mating face pair will have a major impact on the PV limit. Excessive divergent faces will
lead to face damage at considerably lower levels than noted, while convergent faces can
extend the limits (at the expense of leakage). The limits established are based upon a
mating pair face profile in the divergent condition not exceeding 4 minutes of rotation.
The PV limits are general recommendations that under normal conditions should
provide for a minimum seal life of 3 years.
Table 1 shows the nominal allowable PV values for common mating pairs. This is based
upon a rotating primary ring against a square section mating ring. It is normalized to a
base of a single point flush at minimum flow. The Seal Duty Parameter is 1 X 10-7. For
carbon materials, Table 1 is based upon wear for 3 years and for hard materials, the
basis is thermal shock or face damage.
PV levels can be increased (or decreased) by design features of the application. For
example hydropadded seals have higher limits than plain face seals. Improvements in
heat transfer (reducing the face temperature) will increase PV limits. Proven applications
that exceed these limits can be used as guidelines that over-ride the limits established.
DATE: 2/25/03 PREPARED BY: Gordon Buck & Ralph Gabriel APPROVED BY: Joe Haas
_____________________________________________________________________________

Page 1 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade.

The PV level for a specific application is developed by applying factors to the nominal PV
limit noted in Table 1.

PVNew = (PVLimit)m 1m2m 3m4m5m6m 7m8


These multiplication factors are:
m1 = factor for lubricity (Figure 1)
m2 = factor fo r temperature (Table 2)
m3 = factor for face design (Table 3)
m4 = factor for heat transfer (Table 4)
m5 = factor for flush rate
m6 = factor for flush design (Table 5)
m7 = factor for radial clearance (Table 6)
m8 = factor for arrangement
Table 1 - Nominal PV Limits (Non-Lubricating Fluid)
Primary Ring

Mating Ring

PV Limit
(psi x fpm)

Face
Condition

Comments

General Purpose
Resin Grade
Carbons
(9031, 9051)
Premium Performance
Resin Grade Carbons
(9003, N905, 9007,
9036, 9055, 9091)
Antimony Grade
Carbons (F48)

Ni-Resist
Ceramic
Tungsten Carbides
Silicon Carbides
Ni-Resist
Ceramic
Tungsten Carbides
Silicon Carbides
Ni-Resist
Ceramic
Tungsten Carbides
Silicon Carbide (O95)
Silicon Carbide (O58)
Tungsten Carbide
Silicon Carbide
Tungsten Carbide
Silicon Carbide
Tungsten Carbide
Tungsten Carbide
Silicon Carbides
Silicon Carbides

100,000
100,000
400,000
400,000
100,000
100,000
500,000
500,000
100,000
100,000
550,000
600,000
650,000
400,000

Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish
Polish/Polish

Limit is based
on Mating Ring

500,000

Polish/Polish

120,000
185,000
350,000
500,000

Matte/Polish
Matte/Polish
Matte/Polish
Polish/Polish

Ceramic
Silicon Carbide

25,000

Lapped/Polish

Bronze (O33, O41)


Bronze (O35) &
Ni-Resist (7)
Tungsten Carbide
Tungsten Carbide
Silicon Carbide
SiC Graphite Loaded
Glass Filled PTFE

Limit is based
on Mating Ring

Limit is based
on Mating Ring

For Lubricating
Fluids only
For Lubricating
Fluids only
Similar Grades
Dissimilar Grades
Potential for High
Leakage
Temp Dependent
250F Maximum

___________________________________________________________________________

Page 2 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade.

Lubricity (m1):
The lubricity factor is based upon the Seal Duty Parameter. In the science of tribology,
a dimensionless number generally described as a duty parameter, is often used to
correlate performance with regards to lubricity. Figure 1 defines the factor m 1.
Remember that Table 1 is defined as having a Seal Duty Parameter of 1 X 10-7 (nonlubricating fluid). Lubricating fluids, defined as having a Seal Duty Parameter of 1 X 10-6,
has a multiplier of 1.6. This is in line with historical practice. Additional information on
Seal Duty Parameter is located in the appendix.

2
1.6

1.5

0.5

0.5
0.1

0
1E-10

1E-09

1E-08 1E-07 1E-06


Seal Duty Parameter

1E-05

Figure 1 Variation of the lubricity factor, m 1, with the Seal Duty Parameter
Temperature (m2):
The temperature in the seal chamber will affect the wear rates of carbon versus any
mating ring material that it runs against. For hard face materials, such as tungsten
carbide, silicon carbide, etc., the temperature is assumed to have no effect on wear rate;
therefore, the factor m 2 is 1. Table 2 lists the derating factors for temperature.
Table 2 Temperature Factor
Sealing Temperature
Fluid temperature below 175F (80C)
Fluid temperature 175F to 250F (80C to 120C)
Fluid temperature 250F to 350F (120C to 180C)
Fluid temperature above 350F (180C)

m2
1.00
0.90
0.80
0.65

_____________________________________________________________________________

Page 3 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

Face Design (m3):


The factor m 3 approximates the effects of the face design. Hydropadded seal faces
generate less heat and have less wear than plain face seals. In viscous services a
hydropadded face may actually produce enough hydrodynamic load support that the
counteracting pressure is nearly zero; however this can also result in excessive leakage.
Use of hydropadded faces in viscous fluids must be evaluated using more
comprehensive analysis tools. Face design factors are listed in Table 3.
Table 3 Face Design Factor
Design/treatment
m3
Plain Face
1.0
Standard hydropadded face 1.3
Heat Transfer (m4):
Mechanical seal performance is often limited by heat transfer. The heat transfer factor,
m 4, approximates the effects of variations in heat transfer configuration. The base case
is a standard balanced seal using a carbon primary ring versus a square section hard
face mating ring such as tungsten carbide or silicon carbide. In this configuration, the
majority of the heat transfer occurs through the stationary mating ring even though only a
small amount of the surface is directly exposed to the sealed fluid. When two hard faces
are used (such as tungsten carbide versus silicon carbide) the factor m 4 = 1.
The Extended mating ring design was originally developed for the Type 48 seals. In
this design much of the surrounding gland material is removed so that the wetted
(exposed) area on the outer diameter of the mating ring is increased. As a rule of thumb,
the wetted area of the primary and mating rings should be at least twice the seal face
area. Further increases in the wetted area are not as effective.
Both the standard square section and Extended mating ring designs may be used in
either a stationary or rotating configuration. In general, heat transfer for a rotating mating
ring is better than a stationary ring. Some of the improvement is due to the improved
convection from the effects of rotation. In addition, the configuration of a rotating mating
ring is often better from a heat transfer point of view. When dealing with a vaporizing
fluid it is advantageous to use a chamfered mating ring so that there is not a close
vertical space above the outer diameter of the primary ring that can trap vapor bubbles.
While this is standard on a Distributed flush design, it is good practice to utilize this
feature with single point flush stationary mating rings as well as rotating mating rings.

_____________________________________________________________________________

Page 4 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

Table 4 lists the heat transfer factors.


Table 4 Heat transfer factor
Configuration
Standard square section mating ring
(wetted area face area)
stationary
rotating
Standard O-ring mating ring
(wetted area face area)
stationary
rotating
Extended mating ring
(wetted area 2x face area)
stationary
rotating
Distributed flush mating ring
Cup mounted mating ring

m4

1.00
1.15

0.90
1.05

1.30
1.70
1.15
0.50

Flush Rate (m5):


The nominal PV limit assumes that the seal is flushed at the minimum flush rate. The
minimum flush rate is based upon an energy balance computation that assumes the heat
generated at the seal faces and the heat soak coming into the seal chamber is absorbed
by the flush into the seal chamber through ideal mixing. This raises the temperature of
the flush. The allowable temperature increase depends upon the type of fluid sealed.
The following is a rule of thumb for allowable temperature increases:
Allowable temperature increase (F)
30
15
5

Fluid type
Lube oils
Water, chemicals and low VP hydrocarbons
High VP hydrocarbons

The temperature rise can be calculated using the following equation:


T = BTU/hr x Q x 500 x SG x cp
The flush rate can be calculated by changing the equation around:
Q = BTU/hr / (500 x SG x cp x T)
Where,
T = Temperature rise (F)
BTU/hr = heat load (1 hp = 2545 BTU/hr)
Q = flush rate (gpm)
SG = fluid specific gravity (dimensionless)
Cp = fluid specific heat (Btu/lb F)
_____________________________________________________________________________

Page 5 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

The flush rate factor (m 5) can then be calculated based upon the following equation:
m 5 = Qactual / Q minimum
If the factor m 5 is less than 0.5 or if the application is operating without a flush a more
detailed analysis is required. The flush rate factor, m 5, is limited to a maximum value of
1.2.
Flush Design/Method (m6):
In addition to the effects of flush rate, it is well known and accepted that certain flush
designs are better than others. In general, most product bulletins are based on a single
point injection design. Designs using better flow distribution have improved performance.
The multiplier for flush design, m 6, is given in Table 5.
Table 5 Flush design/method factor
Configuration
M6
Radial injection
single port
1.00
multiport (note 4) 1.10
distributed
1.15
Radial Withdrawal
single port
0.70
multiport (note 4) 1.00
distributed
1.00
Axial flow path
0.30
Note:
1. Radial injection refers to a piping plan, like Plan 11, or any other plan where the fluid enters over the seal faces.
2. Radial withdrawal refers to a piping plan, like Plan 13, where the fluid exits over or near the seal faces.
3. Glands with inlet or outlet ports drilled at a 90 degree angle with the final path parallel to the shaft would have an
axial flow path.
4. More ports than mean face diameter (in inches)

Effects of radial clearance (m7 ):


Tests have proven that seal performance improves with increasing radial clearance.
ANSI pump manufacturers now offer large bore and large tapered bore seal chambers
versus the old stuffing box. API 682 requires that the minimum radial clearance over the
rotating parts of the seal be 1/8 minimum. The improved performance results from
increased liquid volume as well as improved flow patterns with the seal chamber. The
multiplier factor, m 7 in Table 6, is an attempt to quantify the complex relationships within
the seal chamber.

_____________________________________________________________________________

Page 6 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

Table 6 Effects of radial clearance


Radial Clearance over rotating seal parts (inches)
0.125
0.093
0.078
0.062
0.046

m7
1.00
0.90
0.80
0.70
0.65

Note: For clearances in-between values listed extrapolate a new factor

Arrangements (m8):
All of the previous discussions of PV limits and factors are for a single seal with liquid on
the outer diameter and atmosphere on the inside diameter. For multiple seals the inner
seal has liquid on both the i nside and outside diameters. In addition to improving heat
transfer, the buffer/barrier fluid may be at a lower temperature than the process fluid. For
these reasons, the estimated seal face temperature used to estimate m 2 should be
computed from the average of the process and buffer/barrier temperature. In the case of
hot applications, this may make a significant difference in the seal rating.
On the other hand, when the process is a low VP hydrocarbon, the buffer/barrier fluid
temperature can increase the temperature of the process decreasing the VP margin and
limiting the life of the seal. In these applications a more detailed analysis is required.
When the buffer/barrier temperature is less than the process temperature:
m 8 = 1.1

_____________________________________________________________________________

Page 7 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

Appendix
Within this document the Seal Duty Parameter is defined as:

G = (7.6e-9) (D/w) (N/Pf)


Where:
G = Seal Duty Parameter (dimensionless)
D = Mean face diameter (inches)
w = Face width (inches)
= Viscosity (centipoises)
N = shaft speed (rpm)
Pf = Face pressure (psi)
The face pressure uses the classic equation; P f = P(b-k) + P sp. The pressure gradient
factor, k, is 0.5 for liquids and 0.67 for gases so long as the face profiles are parallel. If
the fluid is changing from a liquid to a gas as it moves across the seal face, the pressure
gradient factor could be greater than 0.67.

1.00
Mixed

Solid Friction

Hydrodynamic Lubrication

0.50
Friction Coefficient

0.30
0.20
0.10
0.05
0.03
0.02
0.01
1E-10

1E-9

1E-8

1E-7

1E-6

1E-5

1E-4

1E-3

Seal Duty Parameter

Figure 2 Friction coefficient and lubrication regimes for mechanical seals


_______________________________________________________________________

Page 8 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

Figure 2 shows how the friction coefficient varies with the Seal Duty Parameter. The
data points are from mechanical seal tests. Closer examination shows that
hydrodynamic lubrication is considered to begin near a Seal Duty Parameter of 1x10-6.
Mixed film lubrication begins near a Seal Duty Parameter of 1x10-8. Note that the
coefficient of friction has its lowest values in the mixed lubrication regime, but that there
is a lot of scatter in the data.
An interesting aspect of the mixed film regime is that the friction coefficient is unstable
with respect to operating conditions. For example, suppose that the friction coefficient for
a given seal is represented by the heavy curve. Further suppose that the operating point
is at Seal Duty Parameter of 1x10-7. If some temporary upset, say in heat transfer,
causes a temperature increase, then the viscosity and hence the Seal Duty Parameter
would decrease. According to this figure the friction coefficient would increase. This
would cause a further increase in temperature, a further decrease in viscosity, etc. On
the other hand, if the temporary upset causes a temperature decrease, then the viscosity
and hence the Seal Duty Parameter would increase. In this case, according to the figure,
the friction coefficient would decrease. This would cause a further decrease in
temperature, a further increase in viscosity, etc. The limiting factor fo r these excursions
must be the change in viscosity with respect to temperature. Fluids like water would
exhibit much more stability than fluids like lubricating oils.
Because of problems when stick-slip at the sea faces excites undamped vibration of the
bellows or forces a lateral offset, the preferred application of metal bellows seals is in, or
near, the hydrodynamic lubrication regime. Stick-slip refers to a lubrication condition
where there is a rapid change in the coefficient of friction at the sealing faces. The
hydrodynamic lubrication regime is considered to begin at a Seal Duty Parameter of
1x10-6 for plain face seals or 0.5x10-6 for hydropadded seals.
For face combinations using carbon, metal bellows seals may be used at Seal Duty
Parameters as low as 0.5x10-6 for a plain face or down to 0.1x10-6 for hydropadded
faces. This is because carbon is considered to have some self lubricating qualities.
When one of the faces is not carbon, the lubrication regime must be hydrodynamic; that
is, the Seal Duty Parameter must be at least 1x10-6 for plain face seals or 0.5x10-6 for
hydropadded seals.

_____________________________________________________________________________

Page 9 of 9
This John Crane technology bulletin is intended solely for the use of John Crane personnel and is not to be distributed to the trade

You might also like