ADELAIDE

BRISBANE

PERTH

LEAKAGE ACCEPTANCE RATES COMPARISON METAL & SOFT

SEATED VALVES ISO 5208/API 598/API 6D/MSS SP-61/FCI 70-2
TESTING & LEAKAGE RATES OVERVIEW
From an engineering point of view, almost all valves leak by some measure. Valves may be said to be bubble tight or
zero leakage, but in actuality that is just a term that species the allowable leakage of that classication.
In the past many products were rated with Cold Working Pressure (CWP) (i.e. 800 psi WOG which meant 800 psi
working pressure for water, oil or gas service), instead of the ASME/ANSI pressure classes used today for most steel
valves. Of course many commodity, low pressure screwed end valves up to 2000 psi are still produced in CWP WOG
ratings.
The base test specication for most steel valves is API 598 Valve Inspection & Test. Most metallic seated valves
larger than ANSI 50 NB (2) size have an allowable leakage rate allowable in API 598. Soft seated valves such as ball
valves must be zero leakage. Bronze gate, globe & check valves are usually only tested per MSS SP-61 Pressure
Testing of Steel Valves.
Pipeline Valves such as trunnion mounted ball valves, pressure balanced lubricated plug valves and pipeline slab
gate valves are tested to API6D - Pipeline Valves. The testing requirements differ slightly from API 598. The primary
difference being 6Ds zero allowable leakage on closure (seating) tests which is achieved as most of the valves built
to API 6D are resilient seated. However, pipeline full port metal seated API 6D check valves do have a seat leakage
allowance. API 598 does incorporate a zero leakage requirement for soft seated valves used in reneries and
downstream but API 6D is generally specied for pipeline valves.
There are no API standards for globe valves or check valves (except full port API 6D style swing check valves).
Hence valve manufacturers make globe valves to British standard BS 1873 Steel Globe and Check Valves for The
Petroleum, Petrochemical and Allied Industries. BS 1868 is more commonly used for check valves. However, most
manufacturers supplying cast steel ANSI B16.5 anged or ANSI B16.25 buttweld globe and check valves test them
to API 598 and often specify wall thickness to API 600 or API 603 or dual conforming to either/or: - ANSI B16.34, BS
1868, BS 1873 or API 600.
EXPLANATION OF ZERO AND LOW LEAKAGE TEST STANDARDS
In general, specication such as API598 or MSS SP-61 that govern leakage for soft seated valves call for 0 bubbles
of air or 0 drops of water under the specied test conditions over the minimum test time period. These valves are
therefore sometimes referred to as zero leakage valves. In reality, there really is no such thing as zero leakage,
since microscopic amounts of material may indeed cross the seat or packing boundaries, especially if helium or
hydrogen or other small molecule gases are used. Another common term for soft seated valves is bubble tight.
Less frequently, the leakage performance for soft seated valves is referred to as Class VI, which is the tightest leakage
under FCI 70-2, and generally applies to resilient seated control valves (as opposed to metal seated control valves or
soft seated shut off valves). In fact, FCI 70-2 Class VI (formerly ANSI B16.104) allows a small number of bubbles per
minute, increasing with valve size, during the test, whereas API598 and MSS SP-61 do not (for soft seat, but does for
metal seat). FCI 70-2 Class VI is for soft seated control valves but is frequently used as a leakage acceptance test
criteria for metal seated isolation valves such as ball and buttery valves. FCI 70-2 only requires a low pressure test,
consequently closure and seat tests should also be done per API 598 or MSS SP-61.
In actual fact even for metal seat valves the API 598 leak acceptance criteria for metal seated valves (excluding check
valves) allows less leakage than FCI 70-2 Class VI above 150NB (6) and 50NB (2) and under. For zero leakage
metal seated valves API 6D or API 598 soft seated zero leakage criteria can be specied (such as triple offset metal
seated buttery valves and some metal seated ball valves). Special provisions for zero leakage gate valves can also
be specied to BS 6755 (now EN 12266-1) and ISO 5208 under special zero leakage classes. However, metal seated
Global Supply Line Pty Ltd ABN 86 008 134 512
Head Office - 1-15 Barndioota Rd Salisbury Plain South Australia 5109 Ph +61 (0)8 8285 0000 Fax +61 (0)8 8285 0088 www.globalsupplyline.com.au
Page 1

V A LV E S A N D P I P E L I N E S U P P L I E S

valves usually have some level of acceptable leakage when tested, dened rst as some acceptable amount of liquid,
under test conditions and over the time period of the test. Knife gate valves refer to MSS-SP81 leakage rates for metal
seated (for resilient seat its up to manufacturer and client to nominate leakage).
ISO 5208 (EN 12266-1) supersedes BS6755-1 and species acceptable leakage rates such as Rate A and Rate B.
Rate A allows no visible leakage, similar to API 598 for resilient seated valves. API 6D now references leakage rates
in IS 5208 which in turn corresponds to leakage rates in EN 12266-1 (but only for class A, B, C, D, E, F & G).
COMMON VALVE TYPES & RELATED TEST STANDARDS
Valve type

Common test standard

Steel ball, gate, globe and check valves

API 598

Steel ball, gate, globe and check valves

BS 6755*, ISO 5208 (EN 12266-1)

Cast Iron gate valves

API 598, MSS SP-70

Bronze gate, globe and check valves

MSS SP-80

Steel gate, globe and check valves larger than NPS 24

ASME B16.34

Pressure seal gate, globe and check valves

ASME B16.34

Pipeline valves

API 6D, ISO 5208

Cast iron checks

API 598, MSS SP-71

Cast iron globes

API 598, MSS SP-85

Cast iron plugs

API 598, MSS SP-78

Steel ball valves

API 598

Steel butterfly valves

API 598

Cryogenic valves

API 598, BS 6364

Control valves

FCI 70-2, ISA-S75,

Pressure relief valves

API 527, ASME PTC 25

*ISO 5208 (EN 12266-1) supersedes BS 6755

ANSI/FCI 70-2-1976 VALVE LEAKAGE CLASSIFICATIONS

(supersedes ANSI B16.104)
There are six seat leakage classications dened by ANSI/FCI 70-2-1976 (supersedes ANSI B16.104). The six valve
leakage classications are as follows:
Class I.

Identical to Class II, III, and IV in construction and design intent, but no actual shop test is made.

Class II.

Intended for double-port or balanced single-port valves with a metal piston ring seal and metal-to-metal
seats. Air or water at 45 to 60 psig is the test uid. Allowable leakage is 0.5% of the rated full open capacity.

Class III. Intended for the same type of valves as in Class II. Allowable leakage is limited to 0.1% of rated valve
capacity.
Class IV. Intended for single-port and balanced single-port valves with extra-tight piston seals and metal-to-metal
seats. Leakage rate is limited to 0.01% of rated valve capacity (known as metal to metal). Test uid is air or
water, pressure 45 and 50 psig or the operating pressure if lower.
Class V.

Intended for the same types of valves as Class IV. The test uid is water at 100 psig or operating pressure.
Leakage allowed is limited to 5 x 10-4 ml (= .0005ml) per minute per inch of orice diameter per psi
differential.

Class VI. Intended for resilient-seating valves. The test uid is air or nitrogen. Pressure is the lesser of 50 psig or
operating pressure. The leakage limit depends on valve size and range from 0.15 to 11.5 ml per minute for
valve sizes 1 through to 12 inches (known as soft seat classication). Class VI is also frequently used as a
leakage bench mark for metal seated control and isolation valves where a tight degree of shut off is
required.

Page 2

G L O B A L S U P P LY L I N E

Table 1 - FCI 70-2 Leakage Rates Class VI

Nominal Port Diameter

Allowable Leakage

Allowable Leakage

(inches)

(ml Per Minute)

(Bubbles Per Minute)

0.15

1.5

0.30

0.45

2.5

0.60

0.90

1.70

11

4.00

27

6.75

45

10

9.00

63

12

11.5

81

14

21.6

16

28.4

Bubbles per minute is a suggested alternative to ml per minute; in which a 0.25 inch OD x 0.032 inch wall tube is submerged in water to a depth of
1/8 to 1/4. The tube is to be cut square and smooth with no imperfections and should be perpendicular to the surface of the water.

API 598 VALVE LEAKAGE RATES

API 598 (9th edition 2009) from the American Petroleum Institute also covers testing/inspection requirements for gate,
globe, check, ball, plug and buttery valves. API 598 indicates that for shell and backseat tests (backseat only applies
in the case of gate & globe valves), no visible leakage is permitted through body, body liner or joints; i.e. no visible
evidence of drops or wetting of the external surfaces. In the case of metal to metal buttery valves, the client can
specify the acceptable leakage rate. API 598 nominates a low pressure seat test of 80 psi and a high pressure seat*
hydrostatic test of 1.1x the maximum valve cold working pressure rating and a high pressure hydrostatic test of 1.5x
the maximum valve cold working pressure rating.
*The requirements on the body for seat testing under API 598 do vary according to valve type, size & class. For instance trunnion valves 150 NB
(6) and over seat testing requirements are more stringent. Having said that API 6D/ISO 5208 is the required standard for trunnion ball valves.

API 598s allowable seat leakage rates are as follows:

Table 2 - API 598 (9th edition 2009) Valve Seat Leakage Rates
All Metal-Seated Valves (except Check Valves)

Metal-Seated Check Valves

DN
(mm)

NPS
(in.)

All Resilient
Seated Valves

Liquid Test (drops/min)

Gas Test (bubbles/min)

Liquid Test (cc/min)

Gas Test (m3/hr)

50

0b

0b

0.08

65

2 1/2

10

7.5

0.11

80

12

0.13

100

16

12

0.17

125

10

20

15

0.21

150

12

24

18

0.25

200

16

32

24

0.34

250

10

20

40

30

0.42

300

12

24

48

36

0.50

350

14

28

56

42

0.59

400

16

32

64

48

0.67

450

18

36

72

54

0.76

500

20

40

80

60

0.84

600

24

48

96

72

1.01

650

26

52

104

78

1.09

700

28

56

112

84

1.18

750

30

60

120

90

1.26

800

32

64

128

96

1.34

900

36

72

144

108

1.51

1000

40

80

160

120

1.68

1050

42

84

168

126

1.76

1200

48

96

192

144

2.02

a) 1 millilitre is considered equivalent to 16 drops for the liquid test. (b) There shall be no leakage for the minimum specied test duration. For liquid
test, 0 drop means no visible leakage per minimum specied test duration. For gas test, 0 bubbles means less than 1 bubble per minimum specied
test duration.

Page 3

G L O B A L S U P P LY L I N E

MSS SP-61 VALVE LEAKAGE RATES

MSS SP-61-2009 - The Manufacturers Standardisation Society also has a section on valve leakage in Section 5 of
their Pressure Testing of Steel Valves - MSS SP-61-1999.
The following leakage rates are deemed acceptable by this testing standard:
METAL SEATED - GATE, GLOBE, BALL OR BUTTERFLY VALVES: 10 cc/hr per inch of nominal pipe diameter per
hour. (e.g. A 6 globe valve is allowed to leak 60 cc/hr in a test). Metal Seated knife gate valves are not included, refer
to MSS-SP81 which allows 40ml/min/inch at 40 psi. All shut off or isolation valves specied to MSS-SP-61 must pass
the above standards. The seat closure test must be performed at a uid (liquid or gas) pressure no less than 1.1 times
the 1000F (380C) rating rounded to the next 5 psi (0.5 bar).
METAL SEATED CHECK VALVES: Do not require a seat test (except for stop check valves), however some
manufacturers commonly specify a guaranteed maximum leakage.
API 6D/ISO 5208 VALVE LEAKAGE RATES
API 6D follows (see table 3) leakage rates specied in ISO 5208 covers pipeline valves. API 6D pipeline standards are
a mirror/joint venture with ISO 14313 pipeline valve standards. API 6D and ISO 5208 nominate several leakage rate
levels (API 6D now refers to ISO 5208 leakage rates). The most common being zero leakage Rate A for soft seated
ball valves, pipeline (6D style) gate valves and lubricated plug valves. Refer to table 3 for leakage rates. API 6D and
ISO 5208 require a high pressure hydraulic body test (body 1.5x, seat 1.1x maximum valve cold working pressure)
and a low pressure pneumatic test on the seat. No body leakage is allowed during body test. Low pressure and high
pressure gas tests are regarded as special tests. ISO 5208 (EN 12266-1) supersedes BS 6755.

Unit
Test fluid leakage
rates

Rate A

Rate AA

Rate B

Rate C

Rate CC

Rate D

Rate E

Rate EE

Rate F

Rate G

0,01
x
DN

0,03
x
DN

0,08
x
DN

0,1
x
DN

0,3
x
DN

0,39
x
DN

1
x
DN

2
x
DN

0.000 16
x
DN

0,000 5
x
DN

0,001 3
x
DN

0,001 6
x
DN

0,004 8
x
DN

0,006 2
x
DN

0,016
x
DN

0,032
x
DN

0,18
x
DN

0,3
x
DN

3
x
DN

22,3
x
DN

30
x
DN

300
x
DN

470
x
DN

3 000
x
DN

6 000
x
DN

0,003
x
DN

0,0046
x
DN

0,045 8
x
DN

0.340 7
x
DN

0,458 4
x
DN

4,583 7
x
DN

7,129 3
x
DN

45,837
x
DN

91,673
x
DN

0,006
No visually
x
detectable
DN
leakage for
the
0,000 1
duration of
drops/s
x
the test
DN
3

mm /s

Liquid

No visually
detectable
leakage for
the
duration of
bubbles
the test
3

mm /s

Gas

Table 3 - Maximum Allowable Closure Test Seat Leakage Rate ISO 5208:2008(E)/API 6D* (2008)
*API 6D - 2008 refers to ISO 5208 (2008-E) for leakage rates. EN 12266-1 also corresponds to ISO 5208 leakage rates but only for class A, B, C, D, E, F & G.
NOTE 1 The leakage rates only apply when discharging test uid to the atmosphere.
NOTE 2 The closure leakage rate that applies is either that identied in a valve product standard or a leakage rate identied in a purchasers valve
procurement purchase order that is more stringent than that specied in the product standard.
NOTE 3 The meaning of No visually detectable leakage is that there is no visible weeping or leakage in the form of drops or bubbles.
NOTE 4 There is a loosely dened correspondence between the leakage rate acceptance values of API 598 and the leakage values Rate A as
applied to DN < 50, Rate AA-Gas CC-Liquid for other than metal seated check valves and for check valves Rate EE-Gas and Rate
G-Liquid. Rates A, B, C, D, E, F and G correspond to values in EN 12266-1.
NOTE 5 Double block & bleed leakage test is optional only and the client can specify the allowable leakage past the rst seat for larger sizes.
NOTE 6 Valves tested in both direction.

Table 4 - Variations in Basic Test Parameters

Standard
ASME B16.34

Gas
80 psi

API 598

60 - 80 psi

MSS SP-61

60 - 100 psi

ISO 5208

6 bar +/-1 bar

(73 - 102 psi)

API 6D

NA

Test water
Temperature
chlorides
NA
100*
NA
100*
30**
(by mass)

< 125 deg F

41 - 122 deg F
125 deg F
5 deg C 40 deg C
(41 - 106 deg F)
5 deg C 40 deg C
(41 - 106 deg F)

* austenitic stainless steel valves, ** wetted components austenitic & duplex stainless steels
Page 4

G L O B A L S U P P LY L I N E

Table 5 - Leakage Comparison Metal vs Soft Seat

Diameter

API 598
(9th ed. 2009)
Metal Seated

mm.

inch.

Liquid

Air (*)

50

2"

80

3"

12

100

4"

16

API 6D
& ISO 5208***
Metal Seated

API 598/API 6D
& ISO 5208****
Soft Seated

Liquid
Air**
ml/min bubbles/min

Liquid/Air

Air
bubbles/min

Liquid
ml/min

0.45

0.30

Smaller sizes
refer to
Table 3
0.61
0.46

FCI 70-2
(2006)
Class VI

0.90

11

1.70

150

6"

12

24

0.91

27

4.00

200

8"

16

32

1.21

45

6.75

250

10"

20

40

1.52

63

11.10

300

12"

24

48

1.83

81

16.00

350

14"

28

56

2.14

21.60

400

16"

32

64

2.44

450

18"

36

72

2.74

500

20"

40

80

600

24"

48

750

30"

900
1000

Leakage rate
class client
to select
for larger
sizes

28.40

3.05

96

3.66

60

120

4.57

36"

72

144

5.48

40"

80

160

6.40

FCI 70-2
(2006)
Class V

MSS SP-61
(2009)

Liquid

Liquid or Air

5 x 10 -4ml
(=.0005ml)
per minute
per inch of
valve size
per PSI
differential
pressure

10 ml liquid
or 10 cc gas
per inch
of valve size
per hour

API 598 leakage is in drops/min (liquid) and bubbles/min (air)

API 6D corresponds to ISO 5208 - 2008 leakage rates, ISO 5208 in turn corresponds to EN 12266-1 leakage rates but only for class A, B, C, D, E, F & G
FCI 70-2 Class VI does not dene leakage above 400 NB (16)
FCI 70-2 Class VI test is performed always at 3 bar independently from the valve rating
(*) API 598 air test is required for metal seated ball and buttery valves
(**) API 6D air test optional for metal seated
(***) API 6D refers to ISO 5208 Rate D for metal seated, leakage rates which also aligns with EN 12266-1
(****) API 6D refers to ISO 5208 (2008 - E) Rate A for soft seated, zero leakage which aligns with EN 12266-1
Rough guide: - 1 ml = approximately 16 drops, 1 cc = approximately 25 bubbles, 1cc = 1ml

Copyright Global Supply Line 2013

Page 5
Valve Leakage Rates & Test Standards R2 - AS

G L O B A L S U P P LY L I N E