Tank Storage Istanbul 2011

29-30 November 2011

Grand Cehavir Hotel, Istanbul, Turkey

News regarding venting of atmosheric

and low-pressure storage tanks

News regarding use of flame arresters

(why do p/v-vents not function as flame arrester?)

Dipl.-Ing. Axel Sommer

PROTEGO Braunschweiger Flammenfilter GmbH

1
 New International Standard:
Venting of atmospheric and low-pressure
storage tanks ISO 28300

ISO 28300
API 2000
Petroleum, EN 14015
5th edition petrochemical and
Annex L
natural gas industries
Venting of atmospheric
and low-pressure
storage tanks TRbF 20

API 2000
6th edition

2
Background and development of ISO 28300 Standard

ISO 28300 was mainly developed based on the API 2000

standard 1998 6th Edition and the EN 14015 Standard
Annex L and the German TRbF 20
Contradiction towards the venting requirements for
normal venting
Contradiction towards the use of vents as flame
arresters
Committee goal:
This standard shall consider all state of the art
knowledge concerning tank venting and safety and
provide best practice to the user

3
Why new calculation methods for determining normal
venting requirements?

Prof. Salatino from the University of Napoli predicted

that the calculation method of API 2000 may
underpredict thermal breathing
The German TRbF 20 standard developed by Dr. Hans
Foerster from the Federal Institute if Physiks (PTB) also
results in higher values for thermal breathing
The EN 14015 Standard developed by Dr. Wheyl from
BASF also results in higher breathing values
All the above methods depend on a detailed
thermodynamic model and provide higher inbreathing
rates than the API 2000 standard

4
 Validation of results for inbreathing
Prof. Salatino Model calculation at University of
Napoli, 1999
Tank: V = 63000 m3; D = 70 m; H = 15 m
T = 40 C
Water (rain) flow density
Refined model calculation
- Dynamic simulation (pressure
differential at vent)
- Different start temperatures of roof, shell
and product

5
 Validation of results for inbreathing

API 2000
TRbF 20
ISO 28300

Prof. Salatino Model calculation at University of Napoli, 1999

6
 Validation of results for inbreathing

Life field tests and model calculation at

Hoechst in 1980 and 1981
Tank: V = 617 m3; D = 8,5 m; H = 10,6 m
17 C T 26 C
Water (rain) flow density: about 60 kg/m2h
TRbF-model calculation

7
Validation of results for inbreathing

8
Summary
The new section is based on the European EN 14015 Standard.
The approach used is more general than API (the API approach
is centered around hexane or similar services).
Calculated vent rates with the new approach can be substantially
higher for certain tank sizes than what is shown in API-2000.
A research paper from Prof. Salatino and research results from
Hoechst in Frankfurt, which had been provided by Dr. Hans
Foerster from the PTB justified this change.
An advantage of the new calculation method is that it does allow
to consider full and partial insulation of the tank for normal in- and
out-breathing.

9
 ISO 28300 venting requirements
Normal out-breathing and normal inbreathing flows
are defined as the combination of tank vent flows
due to:
Liquid flows into and out of the tank

Ambient weather (thermal) effects

=V
V +
V
out thermal out pump in

=V
V +
V
in thermalin pump out

10
 Liquid filling capacities - out-breathing
out-breathing rate = filling rate

special calculation have to be done for spike products, and at storage

temperature > 40C and vapour pressure > 50 mbar:

11
Liquid filling capacities - inbreathing
in-breathing rate = discharging rate

12
Basis: Model calculations for a fixed roof above
ground storage tank of steel
General assumptions and approximations:
Uniform (time dependent) temperatures of wall, tank
atmosphere, ambient air and rainwater-film
Primary result is the temperature of the tank atmosphere in
dependence on time ; volume flow rates are then deduced
by an isobaric approach (constant ratio of volume to
temperature)
Tank atmosphere is air at ambient pressure
Wall thickness is according to common tank standards (
4 mm)
No heat flux via tank bottom
13
Determining of normal & emergency venting requirements

General Basic Equation for ISO 28300 Model:

V dTg
V=
Tg dt
Energy balance to describe temperature
distribution with respect to time

Q = k A (Tg Ts ) = V g c g
dTg
dt

14
Heat flows during heating by solar radiation
(outbreathing)

solar irradiation
far IR radiation loss

convection convection

15
 Solution if solving differential equation:
80

Maximum volume flow W

30

G
60
Volume Flow V in m3/h

Temperature in oC
25

40
VG,B

20
20
Maximum volume flow occurs
at maximum delta T
0 15
0 900 1800 2700 3600 4500 5400 6300 7200
Time t in s 16
Thermal out-breathing simplified as in ISO 28300

V = C V 0,9
R in
thermal out out T

Cout = 0,2 latitude : > 58

Cout = 0,25 latitude : 58-42
Cout = 0,32 latitude : < 42

Rin = reduction factor insulation

Vt = tank volume

17
 Heat flows during cooling by rain
(inbreathing)
Rain water flow to wall

conduction

convection convection
and
evaporation

Rain water drain from wall

18
 Solution if solving differential equation:
300 Maximum volume flow 55

VG,B

240
45
Volume Flow V in m3/h

Temperature in oC
180 G

35

120

W
25
60
Maximum volume flow occurs
at maximum delta T

0 15
0 180 360 540 720 900
Time t in s 19
Thermal - inbreathing

Vthermalin = Cin VT R in
0,7

vapour pressure
Cin haxane or similar higher than hexane,
or unkown
storage temperature
latitude < 25 C 25C < 25 C 25C
> 58 2,5 4 4 4
42 - 58 3 5 5 5
< 42 4 6,5 6,5 6,5

Rin = reduction factor insulation

Vt = tank volume

20
Calculation Examples
Tank 1
Tank:
Height: 5m
Diameter: 7m
Tank volume: 192.4 m3
Pump in rate: 96 m3/h
Pump out rate: 96 m3/h
Vertical tank
No insulation
MAWP: + 7.5 mbar
MAWV: - 2.5 mbar

21
 Inbreathing Requirements (Total) for Tank 1

Inbreathing requirements Tank 1

400

350
Venting requirements [m3/h]

300

250

200

150

100

50

0
API 2000 EN 14015, EN 14015, EN 14015, 42- EN 14015, 42- EN 14015, EN 14015, TRbF 20
North, VP North, VP> 58, VP Hexane 58, VP> South, VP South, VP>
Hexane Hexane Hexane Hexane Hexane

Pump out Thermal

22
 Outbreathing Requirements (Total) for Tank 1

Outbreathing requirements Tank 1

250
226
200
Venting requirements [m3/h]

150 H/D = 0.5

H/D = 0.71
123 130 H/D = 2

116 117
100 118 122
109

50

0
API 2000, FP API 2000, FP EN 14015, EN 14015, 42- EN 14015, TRbF 20 TRbF 20-2 TRbF 20-3
<37.8C >=37.8C North 58 South

Pump in Thermal

23
Calculation Examples
Tank 2
Very Large Size Tank (outside of scope of API 2000):
Height: 15 m
Diameter: 75 m
Tank volume: 66,268 m3
Pump in rate: 1,400 m3/h
Pump out rate: 1,400 m3/h
Vertical tank
No insulation
MAWP: + 7.5 mbar
MAWV: - 2.5 mbar

24
 Inbreathing Requirements (Total) for Tank 2

Inbreathing requirements Tank 5

18000

16000
Venting requirements [m3/h]

14000

12000

10000

8000

6000

4000

2000

0
API 2000 EN 14015, EN 14015, EN 14015, 42- EN 14015, 42- EN 14015, EN 14015, TRbF 20
North, VP North, VP> 58, VP 58, VP> South, VP South, VP>
Hexane Hexane Hexane Hexane Hexane Hexane

Pump out Thermal

25
 Outbreathing Requirements (Total) for Tank 2

Outbreathing requirements Tank 5

10000

9000

8000
Venting requirements [m3/h]

7000

6000

5000

4000

3000

2000

1000

0
API 2000, FP API 2000, FP EN 14015, North EN 14015, 42-58 EN 14015, South TRbF 20
<37.8C >=37.8C

Pump in Thermal

26
Calculation example considering insulation:

Tank volume 592,000 barrel (94.120 m)

Stored liquid Bitume
Pump in 4542 barrel/h (722 m/h)
Pump out 5458 barrel/h (867 m/h)
Insulation Calciumsilicate
Insulation thickness 2

27
 Overview Venting Requirements
(API 2000, ISO 28300)

API 2000 (without consideration of insulation):

Inbreathing: 6.600 Nm3/h
Outbreathing: 4.200 Nm3/h

ISO 28300 (without consideration of insulation):

Inbreathing: 16.020 Nm3/h
Outbreathing: 7.920 Nm3/h

28
How to consider insulation during thermal in-
and out-breathing
Reduction factor for insulation according to ISO 28300
1
RIN =
h LIN
1+
IN
= heat conduction coefficient

h = heat transfer coefficient

LIN = thickness of insulation

Here: RIN = 0.2145
29
 Overview Venting Requirements
(API 2000, ISO 28300)

API 2000 (without consideration of insulation):

Inbreathing: 6.600 Nm3/h
Outbreathing: 4.200 Nm3/h

ISO 28300 (with insulation):

Inbreathing: 4.140 Nm3/h (vs. 16.020 Nm/h)
Outbreathing: 2.280 Nm/h (vs. 7.920 Nm/h)

30
How to determine inert gas blanketing rates
ISO 28300 Annex F provides guidance for inert
gas blanketing of tanks for flashback protection
This guidance is based on the German TRbF 20
standard
The concept has provided proven safety to the
industry for decades
It is simple way to assure sufficient inert gas
blanketing levels
The amounts result from the inbreathing rates of
the ISO 28300 equations
31
 3 different levels of inert gas blanketing

Level 1
minimum inert gas blanketing requirements in
combination with a specific flame arrester classification
Level 2
more stringent inert gas blanketing requirements with
different flame arrester classification
Level 3
the highest inert gas blanketing requirements with no
flame arrester

32
Tank inbreathing needs to be considered

inbreathing due to changes in weather

inbreathing due to emptying of tank

Inert gas supply needs to be determined

minimum amount of inert gas volume flow VI

amount of reserve inert gas VI

33
 Inert gas level 1:

VI = 0,1 C VT + Vpe
0,7
expressed in m3/h

=
VI 0, 04 VT expressed in m3

Additional conditions:
monitor inert gas supply
alarm shall be triggered when set pressure of the vacuum vent is
reached.
inside of the tank can be classified as Zone 1 (according to the
IEC)
An end-of-line flame arrester shall be installed which has been
tested for atmospheric deflagration and endurance burning for IEC
explosion group IIA (NEC Group D) vapours.
34
 Inert gas level 2:

VI = 0, 2 C VT + Vpe
0,7
expressed in m3/h

=
VI 0, 08 VT expressed in m3
Additional conditions:
The alarm specified under inert gas stage 1 shall activate the
shutdown of the liquid outflow.
At this level of inert gas blanketing the inside of the tank can be
classified as Zone 2 in accordance with IEC 60079-10.
An end-of-line flame arrester shall be installed which has been
tested for atmospheric deflagration for IEC explosion group IIA
(NEC Group D) vapours.

35
Inert gas level 3:
VI = 0,5 C VT + Vpe
0,7
expressed in m3/h
=
VI 0,12 VT expressed in m3
Additional conditions:
The tank pressure shall be kept above atmospheric pressure and the monitoring
system shall have redundancy in the design.
The inert gas supply shall be kept above the tank pressure and in particular the
required flow rate of shall be achieved with a tank pressure at least equal to the
atmospheric pressure.
The trip pressure at which the liquid outflow will be shut down shall be set above
atmospheric pressure. (Pump Shut Off)
Alarms shall be triggered at the trip pressure.
At this level of nitrogen blanketing the inside of the tank can be classified as
Zone 2 in accordance with IEC 60079-10. At this level of inert gas blanketing no
additional protection against flame propagation from the outside to the inside of
the tank is required.

36
Why conservation vents do not function as flame arresters:

37
Why conservation vents do not function as flame arresters:

API 2000 5th Edition 1998:

A flame arrester is not considered necessary for use in
conjunction with a pressure vacuum valve venting to
atmosphere because flame speeds are less than vapor
velocities across the seat of the pressure vacuum valve
TRbF 20 (German standard):
clearly calls for flame arresters for tanks that contain liquids
that can create an explosive atmosphere
Factory Mutual (Insurance and approval company)
requires installation of flame arresters on tanks which store
liquids with a flash point at or below 43 C or on tanks which
heat the stored liquid to its flash point
38
Conclusion for ISO 28300 committee regarding
atmospheric explosion protection of storage
tanks:

Research work is needed due to contradicting

standards and opinions on the ISO 28300
committee
ISO 16852 shall apply as test standard
Two types of test are needed:
A) atmospheric deflagration test
B) continuous burn test

39
5 major vent manufacturers where tested

pressure
pallet

vacuum
pallet

40
 Typical settings for API 650 tanks

set pressure:
10 mbar
( 4.0 in WC)

set vacuum: -2 mbar (-0.8 in WC)

41
Atmospheric Deflagration - Test set-up

1 ignition source
2 plastic bag 1,2 m, length 2,5m
foil thickness >0,05 mm
3 conservation vent
4 explosion proof container
5 mixture inlet with shut-off valve
6 mixture outlet
7 bursting diaphragm

atmospheric deflagration test of end-of-

line flame arrester as described in ISO
16852 part 7.3.2.1 and EN 12874 part
6.3.2.1
42
 Atmospheric Deflagration - Test set-up

1 ignition source
6
2 plastic bag 1,2 m,
length 2,5m foil
2 thickness >0,05 mm
3 conservation vent
4 explosion proof container
1
5 mixture inlet with shut-off
valve
3
6 mixture outlet
7 bursting diaphragm

5 4
7
43
44
 Atmospheric
Deflagration
Test No 1

P/V VALVE

4,2 vol% propane in air

28.03.2007

45
 Atmospheric
Deflagration
Test No 2

P/V VALVE

5,5 vol% propane in air

28.03.2007

46
 Atmospheric
Deflagration
Test No 3

P/V VALVE

6,0 vol% propane in air

28.03.2007

47
 High Velocity Burning - Test set-up
1 continuous flame
2 pressure vacuum valve
3 explosion proof container
4 mixture inlet
5 bursting diaphragm
7 pilot flame
10 shut-off valve

Flame transmission test for high velocity vent valves

as described in ISO 16852 part 9.2. and EN 12874
part 9.2.
48
 High Velocity Burning
Test No 4

P/V VALVE

stoichiometric propane
air mixture

V= 85 m/h

28.03.2007

49
 High Velocity Burning
Test No 5

P/V VALVE

stoichiometric propane
air mixture

V= 100 m/h

28.03.2007

50
Recommendation of ISO 28300 regarding
explosion prevention:
Different tank selection
Inert gas blanketing
Flame arresters

Flame propagation through pressure/vacuum valves

(4.5.4)
Testing has demonstrated that a flame can propagate through a pressure
vacuum valve and into the vapour space of the tank. Tests have shown
that ignition of a PV's relief stream (possibly due to a lighting strike) can
result in a flash back to the PV with enough overpressure to lift the
vacuum pallet causing the flame to enter the tank's vapour space. Other
tests have shown that, under low flow conditions, a flame can propagate
though the pressure side of the PV.

51
Summary
- p/v valves cannot stop an atmospheric deflagration
- p/v valves are not able to stop a flame by dynamic
effects
hence:
- p/v valves cannot substitute flame arresters
- p/v valves are not high velocity vent valves
- only devices approved according flame arrester
standards* are flame arresters
* ISO 16852, EN 12874, USCG 33 CFR part 154, CSA Z343-98

52
Thank you very much
for your attention !

Any questions?

53