SPEMEFACULTY OF ENGINEERINGSPEMEFACULTY OF ENGINEERING

Vented Gas Explosions in SmallVessels with an L/D of 2

Fakandu, B., Kasmani, R.M, Kumar, P., Andrews, G.E.and Phylaktou, H.N.

Energy and Resources Research Institute

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

Explosion Safety is a legal requirement in industry in mostcountries.

The venting technique is a popular and most cost effectiveexplosion protective measure for reducing explosionhazards in industries containing of flammable gases,liquids, mist and dust.

The design guidance on the design of vents in Europeand the US as specified in NFPA 68 is based on theoriginal work of Bartknecht

Introduction

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

Explosion Venting

Explosion venting is process of channelling excessivepressure into the atmosphere as a result of explosion toprotect the integrity of the process vessel.

Most installations have openings in the vessel walls of area Av

that fails at a static burst pressure Pstat that is well below thefailure pressure of the vessel to be protected. Experimentallybased design equations are used to predict the reducedvessel peak pressure, Pred.

Explosion venting (cont)

VentPanel

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

1/Kv = a Pred-b --------------- (1)

where Kv = V2/3/Av

Pred = The maximum overpressure during the vented explosion

V = The vessel volume

1/Kv = (0.1265 log KG – 0.0567) Pred-0.5817 ---------- (2)

Av/As = C1Pred0.5 --------------------------- (3)

Av/V2/3 = 1/Kv = C1C2 Pred

0.5 = a Pred0.5 ----------- (4)

Design Equations

1/Kv = C3 C2 Su (Ep-1) Pred-0.5 ------------ (5)

C3 = [ρu0.5 / (Cd 20.5)] -------------- (6)

A/S = [12.3/Pred]0.5 = (Cd/KvC2)[av/Su(Ep-1)] --------- (7)

where A = Cd Av/As= Cd /Kv C2

S = Su(Ep-1)/av where av is the velocity of sound at the initial conditions (343 m/s).

1/Kv = 0.0169 C2 Su(Ep-1)Pred-0.5 -------------- (8)

Design Equations (cont)

Bartknecht’s vent constants from Eq. 1 with comparison with Swift andlaminar flame theory.

Gas - air KG

bar/msa Eq.1(bar)0.5

b Eq.1 Su

m/s

a(bar)0.5

Swift

a Eq. 1LaminarFlame Eq.5bar0.5

aEq.8bar0.5

Methane 55 0.164 0.5729 0.4 0.22 0.063 0.26

Propane 100 0.200 0.5797 0.45 0.27 0.077 0.32

Coal Gas 130 0.212 0.5900

Hydrogen 550 0.290 0.5850 3.1 0.515 2.14

Ethylene 177 0.228

Eq. 2

0.5817 0.80 0.154 0.645

Design Equations (cont)

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

• Most theories of explosion venting have had to invoke a turbulenceparameter to force agreement between predictions and experimentalmeasurements of overpressures in vented explosions.

• The phenomenon of self acceleration of flames with the developmentof cellular flame structure can occur. This is a function of the vesselsize or diameter, as self acceleration increases with vessel diameter ordistance from the spark to the vent.

• The presence of a bursting disc on a vent can give rise to thegeneration of pressure waves that disturb the laminar flame front, butthe bursting pressure has to be large for this to be significant. The maineffect of vent burst pressure is that the initial spherical flame is largerwhen the vent bursts, so that the initial mass burning rate is higher.

Possible Problems

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

To separate the turbulence generation by venting from thatof self acceleration, by undertaking venting experiments ona very small scale with free venting, where self accelerationof flames and vent flow turbulence and pressure wavegeneration could be considered not to occur.

Aim

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

Conclusion

A vented explosion vessel of 0.0067 m3 (L= 0.324 and D= 0.162m) wasused which was a cylinder with length to diameter ratio, L/D, of 2.

Experimental Methods

324

T3

Flame Speed Measurements

Pressure time record for ethylene/air vented explosions with the time ofarrival at the outer wall marked at 35ms.

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

0 .0

0 .1

0 .2

0 .3

0 .4O

ve

rpre

ssu

re(b

ar)

T im e (m s)

P ressu re -T im e fo r 6 .5% E thylene

T3

T2

T1

Scope

Introduction

Venting Theory

Vent Design equations

Possible Problems

Aim

Experimental Method

Results

Conclusions

the 0.2 m3 vessel that for a stratified explosion in a vesselwith L/D of 2, a Kv of 16.4 and a vent pipe attached that themaximum overpressure occurs for end ignition opposite thevent and not for central ignition , the same was found forpremixed combustion in the same geometry.

The present work investigated central and end ignition forfree venting at Kv = 16.4 with no vent pipe attached.

Influence of Ignition Position

Ignition Positions

Influence of Ignition Positions

CH4/air

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.6 0.8 1.0 1.2 1.4 1.6

Equivalence ratio

Pm

ax

(ba

rg)

End ignition

Centre ignition

C2H4/air

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.5 0.7 0.9 1.1 1.3 1.5 1.7

Equivalence ratioP

ma

x(b

arg

)

End ignition

Centre ignition

Comparison of end and central ignition on Pred for a cylinderwith V=0.2 m3 and L/D=2 (0.5m diameter) for a) methane andb) ethylene for Kv = 16.4

Maximum Flame Speed- oncentreline upstream of Vent

Maximum flame speeds on the vent centreline just upstream ofthe vent for V=0.2 m3, L/D = 2 and Kv=16.4 with end and centralignition for a) methane=air and b) ethylene-air.

CH4/air

0.0

5.0

10.0

15.0

20.0

25.0

0.8 0.9 1 1.1 1.2 1.3 1.4 1.5

Equivalence ratio

Fla

me

sp

ee

ds

(m/s

)m

End ignition

Centre ignition

C2H4/air

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

EquivalenceratioF

lam

es

pe

ed

(m/s

)m

End ignition

Centreignition

Flame Movement

X

Ignitionposition

Diagram on the flame movement at end ignition andthe unburnt gases left inside the vessel

Influence of Kv on Pred

Experimental result compared with other correlations showing mixture reactivity.

Note that the measured overpressures are below the laminar flame predictions with the flamearea = As, thus the actual maximum flame area must be <As and no turbulence factor or selfacceleration factor is required. All other predictions are an order of magnitude too high.

1 1 0

0 .0 1

0 .1

1

1 0P

red

(ba

r)

K v

B a rtk n e c h tS w if tB ra d le y a n d M itc h e n s o nE x p e rim e n ta lL a m in a r F la m e

P re d -K v fo r 4 % P ro p a n e

1 10

0.1

1

10

Pre

d(b

ar)

Kv

BartknechtExperimentalLaminar Flame

Pred-Kv for Hydrogen

Experimental result of hydrogen compared with

Bartknecht and laminar flame theory.

Influence of Kv on Pred

Overpressures in small explosion vesssels are much lowerthan for design methods based on the Bartknecht ventingequation for methane, propane and ethylene.

These small volume explosions are laminar explosions andthere is no need for turbulence factors to predict theoverpressure.

The results for hydrogen in a small explosion vessel withL/D of 2 and end ignition were substantially higher than theoverpressures measured by Bartknecht in a 1m3 cubic vesselwith central ignition.

Self acceleration of flames depends on vessel volume.

Conclusions

There is need for more work to be carried out on themixture reactivity for hydrogen explosion venting.

Recommendation

THANK YOU FOR YOU ATTENTION

Questions