- Email: [email protected]
Practical and computer assessments of ignition hazards during tank washing and during wave action in part-ballasted obo cargo tanks
Journal of Electrostatics, 1 (1975) 61--70 © Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands
PRACTICAL AND COMPUTER ASSESSMENTS OF IGNITION HAZARDS DURING TANK WASHING AND DURING WAVE ACTION IN PARTBALLASTED OBO CARGO TANKS*
J.N. CHUBB
UKAEA Research Group, Culham Laboratory, Abingdon, Oxon. (Great Britain) (Received July 12, 1974; in revised form September 16, 1974)
Summary The paper describes the theoretical, experimental and practical studies which were carried out at Culham Labor~tory up to March 1973 on the mechanisms by which electrostatic hazards arise in the operation of large oil tankers. Theoretical calculations show that hazardous sparks (with energy above 0.2 mJ) can be expected when fairly modest sized slugs of water, perhaps as small as 100 mm long, 20 mm diameter, arrive at a projection such as a washing machine in a tank filled with a typical charged mist (30 nC m-3). Hazards can also be expected during sloshing in part-ballasted tanks. Aerodynamic sweep-up of charge on isolated bodies is not able to produce hazardous conditions. Experimental studies are reported showing that the incendivity of electrostatic sparks is affected by the velocity of approach of the discharge surfaces. Radio techniques have been developed for monitoring the occurrence of electrostatic sparks. These techniques have been used in shipboard studies to show the occurrence of sparks during tank washing operations. By using the radio signals to trigger flash photography, information has been obtained on the physical circumstances in the cargo tank at the instant of individual spark events.
1. Introduction
Electrostatic sparks are a possible source of ignition for the explosions which have damaged a number of very large crude oil tankers and OBO's** over the last few years. A number of people have shown that electrostatically charged mists can be generated in cargo holds by tank cleaning operations with high pressure water jets [1,2,3,4] and by the sloshing of water in partballasted tanks due to ship motion [ 5,6 ]. It seems plausible that electrostatic sparks occur when conducting bodies such as slugs of water approach projections in the tank space where electric fields are concentrated, or when bodies leave such projections and move to the tank wall. Under the right * Paper presented in the Special Symposium on Tanker Explosions at the 2nd International Conference on Static Electricity, Frankfurt, 6 April, 1973. ** OBO -- oil or bulk ore carrier
62 conditions, and with an inflammable atmosphere present, these sparks might form a source of ignition. In our work at Culham on the problem of tanker explosions, which is described in this paper, we have calculated the quantities of electrostatic energy which may be expected to be released from conducting bodies inside cargo tanks and have examined, experimentally, constraints on the incendivity of discharges from such bodies. We have also developed radio techniques for monitoring the occurrence of low energy electrostatic sparks and have used these in shipboard studies in conjunction with flash photography to find the practical circumstances associated with the occurrence of sparks during tank washing and during sloshing in part-ballasted tanks. 2..Computer studies
2.1 Background A charged mist within a cargo tank gives rise to an electric field at the tank wall which will concentrate at any projections into the tank space. Electrostatic charges may appear on conducting bodies in the tank space by the following main processes: (a) induction charging as a b o d y leaves a projection where the electric field is concentrated; (b) by polarisation effects in the electric field near a projection; (c) by corona or water spray neutralisation of polarisation charges; or (d) by direct aerodynamic sweep-up of charged mist particles. At best, corona or water spray charging only achieves the same charging as would be achieved inductively at a projection, b u t without direct physical contact -- so the hazard will be no greater than with induction or polarisation charging. The a m o u n t of electrostatic energy released when a b o d y , inductively charged at a particular projection, is discharged at the tank wail is the same as that which would be released by the same b o d y uncharged coming from this point on the tank wail up to the projection. Either of these routes can, thus, be used for assessing the hazards involved. Calculations of the quantities of electrostatic charge which conducting bodies may be expected to acquire by induction in a variety of circumstances within cargo tank spaces have been made using a c o m p u t e r program, POTENT, which was developed at Culham [ 7 ]. This program enables potential and electric field distributions to be calculated by relaxation solution of Poisson's equation in a wide variety of two-dimensional and axi-symmetric structures. Different potentials may b e ascribed to individual parts of the structure and regions of space charge m a y be introduced which do n o t need to be constrained by c o n d u c t o r boundaries. The results of calculations may be stored and used as input for further investigation of regions of special interest. As a charged b o d y approaches an earthed surface the capacitance increases, and the potential and the electrostatic energy fall. For spherical bodies, these variations have been calculated in each case studied, and the potential of the
63 b o d y compared with the b r e a k d o w n strength of the remaining gap to find the gap and energy at which a spark discharge may occur. If the breakdown gap is t o o small (below a b o u t 1 mm -- which is equivalent to a breakdown voltage of a b o u t 4.5 kV) the spark would not be able to ignite inflammable vapour mixtures of the hydrocarbon gases arising from crude oil, which have ignition energies above about 0.2 mJ.
2.2 Tank washing Calculations have been made on the charge and energy acquired by slugs of water leaving a washing machine in a cylindrical simulation of an OBO cargo tank and leaving the ends of stiffeners in a two-dimensional model of a VLCC* cargo tank. Figure 1 shows the potential distribution in a cylindrical model o f an OBO cargo tank with a charge density of 30 nC m-3 -- typical of that for charged water mists during tank washing. A cylindrical slug of water 1-m-long, 20-mmdiameter leaving the 4.25-m-long, 100-mm-diameter cylindrical projection simulating a washing machine would take away an induced charge of 450 nC
heic. (n"
0
20.77 kV/m
6
ro.diu5 (m)
Fig.1. P o t e n t i a l d i s t r i b u t i o n in O B O cargo t a n k w i t h slug o f water at end o f w a s h i n g machine. Charge density 30 nC m -3. E q u i p o t e n t i a l s a t 4 . 3 9 k V intervals. * VLCC - - very large c r u d e carrier
64 and have a free space energy o f 5.9 mJ. If this w a t e r slug r e a c h e d the base o f the t a n k w h e r e the a m b i e n t field is a b o u t 21 kV m -1 w i t h o u t breaking up, it w o u l d release a b o u t 3.2 m J o f energy at a p o t e n t i a l o f 19.2 kV. T h e radial electric field along the slug at the washing m a c h i n e is fairly u n i f o r m at 0 . 8 - 1 MV m -1, so as the charge and c a p a c i t a n c e are r o u g h l y p r o p o r t i o n a l t o length, it appears t h a t a discharge above the m i n i m u m energy for ignition (0.2 m J ) can be e x p e c t e d f o r slugs as small as 100-mm-long, 2 0 - m m - d i a m e t e r . Figure 2 shows t h e p o t e n t i a l d i s t r i b u t i o n in a cross-section o f a VLCC cargo t a n k with a u n i f o r m charge density o f 30 nC m -3. Calculations were m a d e o f the charge i n d u c e d o n 100-mm- and 200-mm-radius b o d i e s leaving the webs at the ends o f a s t i f f e n e r in t h e m i d d l e o f o n e side o f t h e t a n k . Analytical studies and separate calculations in a x i - s y m m e t r i c geometries i n d i c a t e d t h a t the charge o n spherical bodies w o u l d be a b o u t twice t h a t o n o n e d i a m e t e r length o f c y l i n d e r in t h e t w o - d i m e n s i o n a l calculation. O n this basis, the charges o n t h e 1 0 0 - m m - and 200-mm-radius spheres were 65 and 240 nC. These bodies have free space energies o f 0.19 and 1.3 mJ. If t h e y r e t a i n e d their spherical shape o n arrival at the t a n k wall the increase in t h e i r c a p a c i t a n c e w o u l d r e d u c e these energies t o 0.047 and 0.33 m J and t h e i r b r e a k d o w n gaps w o u l d be 0.2 and 0.5 m m . A l t h o u g h these discharges w o u l d be n o n - i n c e n d i v e because the discharge gap w o u l d be less t h a n the q u e n c h i n g distance, this result c o u l d have been m o d i f i e d if the b o d y h a d c h a n g e d shape b e f o r e arrival at the wall.
TABLE I Electrostatic energies released by bodies leaving wave tips during sloshing Body
Wave
Charge (nC)
Free space energy (mJ)
Breakdown gap (ram)
Discharge energy (mJ)
1-m-long cylinder, 40-ram-diameter
3-m-high wave at tank centre
449
1.1
2
1.1
100-ram-radius sphere
3-m-high wave at tank centre
362
5.9
3.5
2.2
150-mm-radius sphere
3-m-high wave at tank centre
464
6.7
2.5
2.1
100-rnm-radius sphere
1-m-high wave at tank centre
166
1.25
1
0.37
200-ram-radius sphere
1-m-high wave at tank centre
418
3.9
1.25
1.1
150-ram-radius sphere
(Fig.4) 192
1.1
0.58
0.28
300-rnm-radius sphere
(Fig.4) 588
5.2
1.0
1.2
65
, : z : Y - - : ..... :;;
~ - ~ ? |/.'$~-~,x ',114
lO ~8 E E 6
8
12 (metres)
16 ~,
20
24
Fig.2. Potential distribution in a VLCC cargo tank with a charge density of 30 nC m -3. Equipotentials at 2.34 kV intervals. Maximum potential 46.9 kV.
2. 3 Sloshing in part-ballasted tanks Calculations have been made on the charge and energy acquired by slugs of water leaving the tips of waves at the centre and at the side of a half-ballasted OBO cargo tank. Figure 3 shows the potential distribution within a cylindrical model of a part-ballasted OBO cargo tank with a wave at the centre of the tank and a charge density of 30 nC m -3. Table 1 lists the quantities of charge induced on spherical and cylindrical water slugs in this model tank, and the breakdown gaps and energies released if these bodies, w i t h o u t breaking up, fell back to the general surface of the water, or reached the tank wall, where the ambient electric field was about 15 kV m -1. Figure 4 shows the potential distribution in a cross-section of a two-dimensional model OBO cargo tank with a wave near one side of the tank and a charge density of 30 nC m -3. Table 1 lists the quantities of charge induced and the breakdown gaps and energies which can be expected in this model tank when the bodies either fall back to the water surface or reach the tank wall where the ambient field is around 15 kV m -1. The charge on spherical bodies in these two-dimensional calculations was taken as before (section 2.2) as being twice that on a diameter length of a cylindrical body.
2. 4 Aerodynamic sweep-up o f charged mist particles The mobility of charged mist particles in an electric field is very low (3 X 10 -~ m 2 V-1 sec -1 [1] ) so the collection of particles by a body moving in the tank space will be determined by aerodynamic processes with little influence by the ambient electric field or the charge already on the body. If the collection efficiency of a body were unity over its projected cross-section the minimum size of sphere to collect the minimum ignition energy of 0.2 m J over a 50-m-path-length in a charge density of 30 nC m -3 would be 260-ram-diameter. Using equations derived by Langmuir in his studies of aircraft icing [9], and taking a mean droplet radius of 10 microns, we have calculated the maxi-
66
T h~t
12
10
8 5 ~---radial distance (m)
4
"z
~J
Fig.3. Potential distribution in an OBO cargo tank with a 1-m-high wave at centre. Charge density 30 nC m -3.
iO
2
4
6
B 10 12 1/* 16 horizontol distance (m)
18
20
22
Fig.4. Potential distribution in rectangular tank with 5-m-high wave near tank centre. Charge density 3 X lO-SC m -3. Potential contours at 3.0 kV intervals. Maximum potential 42 kV. m u m energies which w o u l d b e a c q u i r e d b y spherical a n d cylindrical b o d i e s t h r o w n o n a m a x i m u m t r a j e c t o r y across an O B O t a n k 4 0 m l o n g a n d 20 m high. A 3 0 0 - r a m - d i a m e t e r s p h e r e w o u l d a c q u i r e a b o u t 3.5 × 10 -6 J a n d a 1-mlong, 2 0 - m m - d i a m e t e r c y l i n d e r a b o u t 15 × 10 -6 J. S m a l l e r energies w o u l d b e a c q u i r e d w i t h smaller m i s t particles. I t is clear t h a t a e r o d y n a m i c s w e e p - u p o f c h a r g e d m i s t particles d o e s n o t p r e s e n t a n y ignition hazard.
67 3. Constraints on incendivity of electrostatic discharges The discharge of charged conducting bodies in a cargo tank space is likely to involve breakdown to a water surface or layer, and to involve m o v e m e n t of the discharge surfaces. Some studies we have done on the time delay to breakdown between a sphere and a horizontal water surface show that when a charged body approaches slower than about 0.5 m sec -1 the breakdown involves mechanical instability [9] of the water surface. Above this velocity, breakdown seems to be as between smooth rigid electrodes. Ignition tests suggest that the incendivity of a discharge from an isolated charged body to a water surface is lower at low speeds than for a similarly charged body discharging to a metal surface. If an incendive spark occurs from a conducting body approaching a surface the flame front may fail to reach the surrounding inflammable gas mixture if the approach speed is too high in comparison to the flame propagation speed. Some calculations of this quenching effect were carried o u t for the case of a sphere approaching a plane surface with a simple cylindrical flame front propagating at a constant velocity of 0.4 m sec -1 through a gas volume being displaced outward by the advancing surface. The results are shown in Fig.5. Some ignition experiments were carried out in which 12.5-, 25- and 62.5-mmradius surfaces were dropped into water in a 5 % propane air mixture. The results are included in Fig.5. Although the agreement between theory and experiment is not very good, it is clear that quenching is a real effect and involves approach velocities of only a few metres per second. The practical importance of this is that bodies thrown upwards towards the top of a tank are more likely to approach surfaces sufficiently slowly for flame propagation to occur from an incendive energy spark than bodies falling to the b o t t o m of a tank. 4. Practical studies Laboratory and o u t d o o r studies [10,11] have shown that radio observations using a tuned loop aerial provide a sensitive way to monitor the occurrence of low energy electrostatic sparks, and to separate sparks from corona discharges. Since corona is, in general, non-incendive -- whereas sparks may be -- radio observations form a useful technique to m o n i t o r the type of events which could present an ignition hazard. In initial studies we used simple superhet receiver units at 150 MHz and 10 MHz. To avoid the shot-to-shot variation in o u t p u t signal associated with frequency changing in a superhet, and to take better advantage of the bandwidth of the aerial system, we have now changed to simple straight-through receivers in which integrated circuit elements are used to provide wide band amplification of the signal induced in the aerial circuit for transistor detection. The overall gain of these detector systems at 40 MHz varies from
68
IOO~
'
'
'
'
'
'
' ' 1
r
,
,
,
,
,
,,
& 60 4O
13°
m 52.5
~2G E E m
45.0 31.2
~
25.6
8
"Q 6
/
2 ms " 1 ~ ' ~ / /
~
1 ms -1 21-ns"1 19,9
~4
14.0
k9
3
i
1 ms"
Approachvelocity
;
I
' 6
Radius of c u r v a t u r e
I
I
I
I
8 io
2b
3'o J o ' 60 ' " 8O 100
Of dlscharge surface ( r a m )
ExperimentQI o b s e r v a t i o n s on ionition probQbil~ty High ignition p r o b a b i l l t y Low ;gnitionprobabitity
m-
Approach v e l o c i t y 1 msec -1 2 msec -1 e°
m
Fig.5. Gaps and breakdown voltages at which quenching occurs for various approach velocities.
around 46 to 72 db over their operating range. We have used this radio technique in shore-tank studies at the Shell Laboratories in Amsterdam (KSLA) and during shipboard studies on m.v. "Furness Bridge" and on m.v. "Jedforest". During studies in March 1973 on m.v. "Jedforesf", radio equipment was used to monitor the occurrence of sparks during tank washing and during sloshing in part-ballasted cargo tanks due to ship motion. Two aerials, spaced well apart, were mounted inside a cargo tank, and each inside aerial had a separate outside aerial mounted nearby. Sparks were considered to occur in the tank only if coincident signals were observed on the two inside aerials without signals being observed at that time on either outside aerial. This arrangement eliminated interference from external sources and the effect of small-scale events very close to either aerial. At the instant of occurrence of a definite spark within the tank space the circumstances inside the tank were photographed by flash photography using two cameras with wide angle lenses (28 mm f/2.8 with HP4 film) and Braun F700 flash units for illumination. In the tank-washing studies, a number of low energy sparks were recorded and the photographs taken showed that
69
Fig.6. Flash photograph u........ were the tank space. when sparks occurred the opposed pair of jets of the washing machine directed straight down to the tank bottom and straight up to the underside of the hatch cover, with water cascading down from the hatch cover into the tank space. An example of such a spark.triggered photograph is shown in Fig.6. Although individual photographs do not pinpoint the location of individual sparks, and the radio observations cannot yet give a quantitative measure of their energY, it is possible to learn about the practical circumstances responsible for the occurrence of sparks by looking for common features m a series of photographs. These flash photographs also provide information on the size of water slugs and the position of cleaning jets at the time of occurrence of individual sparks. This information can be fed back into computer calculations of electrostatic field distributions to provide estimates of the energies of spark discharges which could have occurred.
5. Conclusions Computer calculations show that it is not difficult for fairly modest sizes of conducting bodies, such as slugs of water, to acquire incendive quantities of electrostatic energY during tank washing operations and during sloshing m part-ballasted tanks when charged mists are present in the cargo tank space. The hazard presented by these charged bodies depends upon the speed of their approach to the discharge surface. Since the combined probability of a body with enough energy approaching a surface slowly enough to avoid
70
quenching in an atmosphere which is in the inflammable range is likely to be very low, it is not surprising that explosions have been fairly infrequent occurrences. Radio detection can be used to monitor the occurrence of low energy electrostatic sparks in practical operations, and flash photography triggered from the radio equipment can show the physical circumstances associated with individual sparks.
Acknowledgements I would like to express m y gratitude to John Houlder, of Furness Withy, and to the International Chamber of Shipping for their support and encouragement and for permission to present the information and results contained in this paper. I would also like to thank Dr. E.S. Hotston, Dr. S.K. Erents and I.E. Pollard for their contributions to the work described. References 1 J.M. van der Weerd, Electrostatic charge generation during washing of tanks with water sprays, II. Measurements and interpretation, Proc. 3rd Conf. Static Electrification, London, May 1971, Inst. Phys. Conf. Set. No. 11, p.158. 2 R.F. Lange, Static electrification studies during tank washing on s.s. "Mobil Daylight", Mobil Research and Development Corp., 71.22 -- AD, August 1971. 3 W.M. Bustin, Electrically charged mists produced by water washing, Esso Engineering Rep. EE. 8TMR. 72, March 1972. 4 R.F. Klaver and V.A. Dayot, Investigation of electrostatic charge generation during cargo tank washing on the "Ralph B. Johnson", Chevron Research Co. Rep., 27 July, 1970. 5 R. Tangen, Some investigations of electrostatic charge production on board an OBO carrier, Personal communication, March 1972. 6 J.M. van der Weerd, R.L., Lindbauer and W.A. van Laar, Static electricity from sloshing water bottoms in tanks of oil-bulk ore carriers: Tests on board "Hoegh Rover", Shell Rep. AMSR 0044.72, December 1972. 7 C.L. Thomas, POTENT -- A package for the numerical solution of potential problems in general 2D regions, Proc. Conf. on Software for Numerical Mathematics and its Applications, Loughborough University, April 1973, Academic Press. 8 I. Langmuir, Mathematical investigation of water droplet trajectories. In G. Suits (Ed.), The Collected Works of Irving Langrauir, Pergamon, Vol.X, Rep. No. RL--224, 1961. 9 G.I. Taylor and A.D. McEwan, The stability of a horizontal fluid interface in a vertical electric field, J. Fluid Mech°, 22 (1965) 1. 10 J.N. Chubb, S.K. Erents and I.E. Pollard, Radio detection of low energy electrostatic sparks, Nature, 245 (5422) (1973) 206. 11 J.N. Chubb, Radio signals from low energy electrostatic sparks, Paper presented at IEE Conf. on Gas Discharges, London, 9--12 Sept. 1974, Culham Preprint CLM-P390.
PRACTICAL AND COMPUTER ASSESSMENTS OF IGNITION HAZARDS DURING TANK WASHING AND DURING WAVE ACTION IN PARTBALLASTED OBO CARGO TANKS*
J.N. CHUBB
UKAEA Research Group, Culham Laboratory, Abingdon, Oxon. (Great Britain) (Received July 12, 1974; in revised form September 16, 1974)
Summary The paper describes the theoretical, experimental and practical studies which were carried out at Culham Labor~tory up to March 1973 on the mechanisms by which electrostatic hazards arise in the operation of large oil tankers. Theoretical calculations show that hazardous sparks (with energy above 0.2 mJ) can be expected when fairly modest sized slugs of water, perhaps as small as 100 mm long, 20 mm diameter, arrive at a projection such as a washing machine in a tank filled with a typical charged mist (30 nC m-3). Hazards can also be expected during sloshing in part-ballasted tanks. Aerodynamic sweep-up of charge on isolated bodies is not able to produce hazardous conditions. Experimental studies are reported showing that the incendivity of electrostatic sparks is affected by the velocity of approach of the discharge surfaces. Radio techniques have been developed for monitoring the occurrence of electrostatic sparks. These techniques have been used in shipboard studies to show the occurrence of sparks during tank washing operations. By using the radio signals to trigger flash photography, information has been obtained on the physical circumstances in the cargo tank at the instant of individual spark events.
1. Introduction
Electrostatic sparks are a possible source of ignition for the explosions which have damaged a number of very large crude oil tankers and OBO's** over the last few years. A number of people have shown that electrostatically charged mists can be generated in cargo holds by tank cleaning operations with high pressure water jets [1,2,3,4] and by the sloshing of water in partballasted tanks due to ship motion [ 5,6 ]. It seems plausible that electrostatic sparks occur when conducting bodies such as slugs of water approach projections in the tank space where electric fields are concentrated, or when bodies leave such projections and move to the tank wall. Under the right * Paper presented in the Special Symposium on Tanker Explosions at the 2nd International Conference on Static Electricity, Frankfurt, 6 April, 1973. ** OBO -- oil or bulk ore carrier
62 conditions, and with an inflammable atmosphere present, these sparks might form a source of ignition. In our work at Culham on the problem of tanker explosions, which is described in this paper, we have calculated the quantities of electrostatic energy which may be expected to be released from conducting bodies inside cargo tanks and have examined, experimentally, constraints on the incendivity of discharges from such bodies. We have also developed radio techniques for monitoring the occurrence of low energy electrostatic sparks and have used these in shipboard studies in conjunction with flash photography to find the practical circumstances associated with the occurrence of sparks during tank washing and during sloshing in part-ballasted tanks. 2..Computer studies
2.1 Background A charged mist within a cargo tank gives rise to an electric field at the tank wall which will concentrate at any projections into the tank space. Electrostatic charges may appear on conducting bodies in the tank space by the following main processes: (a) induction charging as a b o d y leaves a projection where the electric field is concentrated; (b) by polarisation effects in the electric field near a projection; (c) by corona or water spray neutralisation of polarisation charges; or (d) by direct aerodynamic sweep-up of charged mist particles. At best, corona or water spray charging only achieves the same charging as would be achieved inductively at a projection, b u t without direct physical contact -- so the hazard will be no greater than with induction or polarisation charging. The a m o u n t of electrostatic energy released when a b o d y , inductively charged at a particular projection, is discharged at the tank wail is the same as that which would be released by the same b o d y uncharged coming from this point on the tank wail up to the projection. Either of these routes can, thus, be used for assessing the hazards involved. Calculations of the quantities of electrostatic charge which conducting bodies may be expected to acquire by induction in a variety of circumstances within cargo tank spaces have been made using a c o m p u t e r program, POTENT, which was developed at Culham [ 7 ]. This program enables potential and electric field distributions to be calculated by relaxation solution of Poisson's equation in a wide variety of two-dimensional and axi-symmetric structures. Different potentials may b e ascribed to individual parts of the structure and regions of space charge m a y be introduced which do n o t need to be constrained by c o n d u c t o r boundaries. The results of calculations may be stored and used as input for further investigation of regions of special interest. As a charged b o d y approaches an earthed surface the capacitance increases, and the potential and the electrostatic energy fall. For spherical bodies, these variations have been calculated in each case studied, and the potential of the
63 b o d y compared with the b r e a k d o w n strength of the remaining gap to find the gap and energy at which a spark discharge may occur. If the breakdown gap is t o o small (below a b o u t 1 mm -- which is equivalent to a breakdown voltage of a b o u t 4.5 kV) the spark would not be able to ignite inflammable vapour mixtures of the hydrocarbon gases arising from crude oil, which have ignition energies above about 0.2 mJ.
2.2 Tank washing Calculations have been made on the charge and energy acquired by slugs of water leaving a washing machine in a cylindrical simulation of an OBO cargo tank and leaving the ends of stiffeners in a two-dimensional model of a VLCC* cargo tank. Figure 1 shows the potential distribution in a cylindrical model o f an OBO cargo tank with a charge density of 30 nC m-3 -- typical of that for charged water mists during tank washing. A cylindrical slug of water 1-m-long, 20-mmdiameter leaving the 4.25-m-long, 100-mm-diameter cylindrical projection simulating a washing machine would take away an induced charge of 450 nC
heic. (n"
0
20.77 kV/m
6
ro.diu5 (m)
Fig.1. P o t e n t i a l d i s t r i b u t i o n in O B O cargo t a n k w i t h slug o f water at end o f w a s h i n g machine. Charge density 30 nC m -3. E q u i p o t e n t i a l s a t 4 . 3 9 k V intervals. * VLCC - - very large c r u d e carrier
64 and have a free space energy o f 5.9 mJ. If this w a t e r slug r e a c h e d the base o f the t a n k w h e r e the a m b i e n t field is a b o u t 21 kV m -1 w i t h o u t breaking up, it w o u l d release a b o u t 3.2 m J o f energy at a p o t e n t i a l o f 19.2 kV. T h e radial electric field along the slug at the washing m a c h i n e is fairly u n i f o r m at 0 . 8 - 1 MV m -1, so as the charge and c a p a c i t a n c e are r o u g h l y p r o p o r t i o n a l t o length, it appears t h a t a discharge above the m i n i m u m energy for ignition (0.2 m J ) can be e x p e c t e d f o r slugs as small as 100-mm-long, 2 0 - m m - d i a m e t e r . Figure 2 shows t h e p o t e n t i a l d i s t r i b u t i o n in a cross-section o f a VLCC cargo t a n k with a u n i f o r m charge density o f 30 nC m -3. Calculations were m a d e o f the charge i n d u c e d o n 100-mm- and 200-mm-radius b o d i e s leaving the webs at the ends o f a s t i f f e n e r in t h e m i d d l e o f o n e side o f t h e t a n k . Analytical studies and separate calculations in a x i - s y m m e t r i c geometries i n d i c a t e d t h a t the charge o n spherical bodies w o u l d be a b o u t twice t h a t o n o n e d i a m e t e r length o f c y l i n d e r in t h e t w o - d i m e n s i o n a l calculation. O n this basis, the charges o n t h e 1 0 0 - m m - and 200-mm-radius spheres were 65 and 240 nC. These bodies have free space energies o f 0.19 and 1.3 mJ. If t h e y r e t a i n e d their spherical shape o n arrival at the t a n k wall the increase in t h e i r c a p a c i t a n c e w o u l d r e d u c e these energies t o 0.047 and 0.33 m J and t h e i r b r e a k d o w n gaps w o u l d be 0.2 and 0.5 m m . A l t h o u g h these discharges w o u l d be n o n - i n c e n d i v e because the discharge gap w o u l d be less t h a n the q u e n c h i n g distance, this result c o u l d have been m o d i f i e d if the b o d y h a d c h a n g e d shape b e f o r e arrival at the wall.
TABLE I Electrostatic energies released by bodies leaving wave tips during sloshing Body
Wave
Charge (nC)
Free space energy (mJ)
Breakdown gap (ram)
Discharge energy (mJ)
1-m-long cylinder, 40-ram-diameter
3-m-high wave at tank centre
449
1.1
2
1.1
100-ram-radius sphere
3-m-high wave at tank centre
362
5.9
3.5
2.2
150-mm-radius sphere
3-m-high wave at tank centre
464
6.7
2.5
2.1
100-rnm-radius sphere
1-m-high wave at tank centre
166
1.25
1
0.37
200-ram-radius sphere
1-m-high wave at tank centre
418
3.9
1.25
1.1
150-ram-radius sphere
(Fig.4) 192
1.1
0.58
0.28
300-rnm-radius sphere
(Fig.4) 588
5.2
1.0
1.2
65
, : z : Y - - : ..... :;;
~ - ~ ? |/.'$~-~,x ',114
lO ~8 E E 6
8
12 (metres)
16 ~,
20
24
Fig.2. Potential distribution in a VLCC cargo tank with a charge density of 30 nC m -3. Equipotentials at 2.34 kV intervals. Maximum potential 46.9 kV.
2. 3 Sloshing in part-ballasted tanks Calculations have been made on the charge and energy acquired by slugs of water leaving the tips of waves at the centre and at the side of a half-ballasted OBO cargo tank. Figure 3 shows the potential distribution within a cylindrical model of a part-ballasted OBO cargo tank with a wave at the centre of the tank and a charge density of 30 nC m -3. Table 1 lists the quantities of charge induced on spherical and cylindrical water slugs in this model tank, and the breakdown gaps and energies released if these bodies, w i t h o u t breaking up, fell back to the general surface of the water, or reached the tank wall, where the ambient electric field was about 15 kV m -1. Figure 4 shows the potential distribution in a cross-section of a two-dimensional model OBO cargo tank with a wave near one side of the tank and a charge density of 30 nC m -3. Table 1 lists the quantities of charge induced and the breakdown gaps and energies which can be expected in this model tank when the bodies either fall back to the water surface or reach the tank wall where the ambient field is around 15 kV m -1. The charge on spherical bodies in these two-dimensional calculations was taken as before (section 2.2) as being twice that on a diameter length of a cylindrical body.
2. 4 Aerodynamic sweep-up o f charged mist particles The mobility of charged mist particles in an electric field is very low (3 X 10 -~ m 2 V-1 sec -1 [1] ) so the collection of particles by a body moving in the tank space will be determined by aerodynamic processes with little influence by the ambient electric field or the charge already on the body. If the collection efficiency of a body were unity over its projected cross-section the minimum size of sphere to collect the minimum ignition energy of 0.2 m J over a 50-m-path-length in a charge density of 30 nC m -3 would be 260-ram-diameter. Using equations derived by Langmuir in his studies of aircraft icing [9], and taking a mean droplet radius of 10 microns, we have calculated the maxi-
66
T h~t
12
10
8 5 ~---radial distance (m)
4
"z
~J
Fig.3. Potential distribution in an OBO cargo tank with a 1-m-high wave at centre. Charge density 30 nC m -3.
iO
2
4
6
B 10 12 1/* 16 horizontol distance (m)
18
20
22
Fig.4. Potential distribution in rectangular tank with 5-m-high wave near tank centre. Charge density 3 X lO-SC m -3. Potential contours at 3.0 kV intervals. Maximum potential 42 kV. m u m energies which w o u l d b e a c q u i r e d b y spherical a n d cylindrical b o d i e s t h r o w n o n a m a x i m u m t r a j e c t o r y across an O B O t a n k 4 0 m l o n g a n d 20 m high. A 3 0 0 - r a m - d i a m e t e r s p h e r e w o u l d a c q u i r e a b o u t 3.5 × 10 -6 J a n d a 1-mlong, 2 0 - m m - d i a m e t e r c y l i n d e r a b o u t 15 × 10 -6 J. S m a l l e r energies w o u l d b e a c q u i r e d w i t h smaller m i s t particles. I t is clear t h a t a e r o d y n a m i c s w e e p - u p o f c h a r g e d m i s t particles d o e s n o t p r e s e n t a n y ignition hazard.
67 3. Constraints on incendivity of electrostatic discharges The discharge of charged conducting bodies in a cargo tank space is likely to involve breakdown to a water surface or layer, and to involve m o v e m e n t of the discharge surfaces. Some studies we have done on the time delay to breakdown between a sphere and a horizontal water surface show that when a charged body approaches slower than about 0.5 m sec -1 the breakdown involves mechanical instability [9] of the water surface. Above this velocity, breakdown seems to be as between smooth rigid electrodes. Ignition tests suggest that the incendivity of a discharge from an isolated charged body to a water surface is lower at low speeds than for a similarly charged body discharging to a metal surface. If an incendive spark occurs from a conducting body approaching a surface the flame front may fail to reach the surrounding inflammable gas mixture if the approach speed is too high in comparison to the flame propagation speed. Some calculations of this quenching effect were carried o u t for the case of a sphere approaching a plane surface with a simple cylindrical flame front propagating at a constant velocity of 0.4 m sec -1 through a gas volume being displaced outward by the advancing surface. The results are shown in Fig.5. Some ignition experiments were carried out in which 12.5-, 25- and 62.5-mmradius surfaces were dropped into water in a 5 % propane air mixture. The results are included in Fig.5. Although the agreement between theory and experiment is not very good, it is clear that quenching is a real effect and involves approach velocities of only a few metres per second. The practical importance of this is that bodies thrown upwards towards the top of a tank are more likely to approach surfaces sufficiently slowly for flame propagation to occur from an incendive energy spark than bodies falling to the b o t t o m of a tank. 4. Practical studies Laboratory and o u t d o o r studies [10,11] have shown that radio observations using a tuned loop aerial provide a sensitive way to monitor the occurrence of low energy electrostatic sparks, and to separate sparks from corona discharges. Since corona is, in general, non-incendive -- whereas sparks may be -- radio observations form a useful technique to m o n i t o r the type of events which could present an ignition hazard. In initial studies we used simple superhet receiver units at 150 MHz and 10 MHz. To avoid the shot-to-shot variation in o u t p u t signal associated with frequency changing in a superhet, and to take better advantage of the bandwidth of the aerial system, we have now changed to simple straight-through receivers in which integrated circuit elements are used to provide wide band amplification of the signal induced in the aerial circuit for transistor detection. The overall gain of these detector systems at 40 MHz varies from
68
IOO~
'
'
'
'
'
'
' ' 1
r
,
,
,
,
,
,,
& 60 4O
13°
m 52.5
~2G E E m
45.0 31.2
~
25.6
8
"Q 6
/
2 ms " 1 ~ ' ~ / /
~
1 ms -1 21-ns"1 19,9
~4
14.0
k9
3
i
1 ms"
Approachvelocity
;
I
' 6
Radius of c u r v a t u r e
I
I
I
I
8 io
2b
3'o J o ' 60 ' " 8O 100
Of dlscharge surface ( r a m )
ExperimentQI o b s e r v a t i o n s on ionition probQbil~ty High ignition p r o b a b i l l t y Low ;gnitionprobabitity
m-
Approach v e l o c i t y 1 msec -1 2 msec -1 e°
m
Fig.5. Gaps and breakdown voltages at which quenching occurs for various approach velocities.
around 46 to 72 db over their operating range. We have used this radio technique in shore-tank studies at the Shell Laboratories in Amsterdam (KSLA) and during shipboard studies on m.v. "Furness Bridge" and on m.v. "Jedforest". During studies in March 1973 on m.v. "Jedforesf", radio equipment was used to monitor the occurrence of sparks during tank washing and during sloshing in part-ballasted cargo tanks due to ship motion. Two aerials, spaced well apart, were mounted inside a cargo tank, and each inside aerial had a separate outside aerial mounted nearby. Sparks were considered to occur in the tank only if coincident signals were observed on the two inside aerials without signals being observed at that time on either outside aerial. This arrangement eliminated interference from external sources and the effect of small-scale events very close to either aerial. At the instant of occurrence of a definite spark within the tank space the circumstances inside the tank were photographed by flash photography using two cameras with wide angle lenses (28 mm f/2.8 with HP4 film) and Braun F700 flash units for illumination. In the tank-washing studies, a number of low energy sparks were recorded and the photographs taken showed that
69
Fig.6. Flash photograph u........ were the tank space. when sparks occurred the opposed pair of jets of the washing machine directed straight down to the tank bottom and straight up to the underside of the hatch cover, with water cascading down from the hatch cover into the tank space. An example of such a spark.triggered photograph is shown in Fig.6. Although individual photographs do not pinpoint the location of individual sparks, and the radio observations cannot yet give a quantitative measure of their energY, it is possible to learn about the practical circumstances responsible for the occurrence of sparks by looking for common features m a series of photographs. These flash photographs also provide information on the size of water slugs and the position of cleaning jets at the time of occurrence of individual sparks. This information can be fed back into computer calculations of electrostatic field distributions to provide estimates of the energies of spark discharges which could have occurred.
5. Conclusions Computer calculations show that it is not difficult for fairly modest sizes of conducting bodies, such as slugs of water, to acquire incendive quantities of electrostatic energY during tank washing operations and during sloshing m part-ballasted tanks when charged mists are present in the cargo tank space. The hazard presented by these charged bodies depends upon the speed of their approach to the discharge surface. Since the combined probability of a body with enough energy approaching a surface slowly enough to avoid
70
quenching in an atmosphere which is in the inflammable range is likely to be very low, it is not surprising that explosions have been fairly infrequent occurrences. Radio detection can be used to monitor the occurrence of low energy electrostatic sparks in practical operations, and flash photography triggered from the radio equipment can show the physical circumstances associated with individual sparks.
Acknowledgements I would like to express m y gratitude to John Houlder, of Furness Withy, and to the International Chamber of Shipping for their support and encouragement and for permission to present the information and results contained in this paper. I would also like to thank Dr. E.S. Hotston, Dr. S.K. Erents and I.E. Pollard for their contributions to the work described. References 1 J.M. van der Weerd, Electrostatic charge generation during washing of tanks with water sprays, II. Measurements and interpretation, Proc. 3rd Conf. Static Electrification, London, May 1971, Inst. Phys. Conf. Set. No. 11, p.158. 2 R.F. Lange, Static electrification studies during tank washing on s.s. "Mobil Daylight", Mobil Research and Development Corp., 71.22 -- AD, August 1971. 3 W.M. Bustin, Electrically charged mists produced by water washing, Esso Engineering Rep. EE. 8TMR. 72, March 1972. 4 R.F. Klaver and V.A. Dayot, Investigation of electrostatic charge generation during cargo tank washing on the "Ralph B. Johnson", Chevron Research Co. Rep., 27 July, 1970. 5 R. Tangen, Some investigations of electrostatic charge production on board an OBO carrier, Personal communication, March 1972. 6 J.M. van der Weerd, R.L., Lindbauer and W.A. van Laar, Static electricity from sloshing water bottoms in tanks of oil-bulk ore carriers: Tests on board "Hoegh Rover", Shell Rep. AMSR 0044.72, December 1972. 7 C.L. Thomas, POTENT -- A package for the numerical solution of potential problems in general 2D regions, Proc. Conf. on Software for Numerical Mathematics and its Applications, Loughborough University, April 1973, Academic Press. 8 I. Langmuir, Mathematical investigation of water droplet trajectories. In G. Suits (Ed.), The Collected Works of Irving Langrauir, Pergamon, Vol.X, Rep. No. RL--224, 1961. 9 G.I. Taylor and A.D. McEwan, The stability of a horizontal fluid interface in a vertical electric field, J. Fluid Mech°, 22 (1965) 1. 10 J.N. Chubb, S.K. Erents and I.E. Pollard, Radio detection of low energy electrostatic sparks, Nature, 245 (5422) (1973) 206. 11 J.N. Chubb, Radio signals from low energy electrostatic sparks, Paper presented at IEE Conf. on Gas Discharges, London, 9--12 Sept. 1974, Culham Preprint CLM-P390.