EXPLOSION HAZARDS AND PROTECTION IN THE USE OF INTERMEDIATE BULK CONTAINERS (PART III)
Where the vessel is relatively weak, explosion relief venting is an option but does not appear to be used very often in this type of vessel. Where explosion venting is used, the preferred method is to install the explosion vent to the ducting immediately upstream of the IBC. Since personnel are likely to be working in the vicinity of an IBC filling point, the vent is either ducted to a safe area or a flameless venting device is fitted.
Explosion suppression is important in cases where explosions could cause the emission of gases, vapours or dusts which are toxic or otherwise harmful to the surroundings. These systems are designed to prevent the creation of unacceptably high pressure by gas or dust explosions within enclosures that are not designed to withstand the maximum explosion pressure. The explosibility data of the dust must lie within the range of applicability of suppression systems. The design of the suppression equipment will depend on a range of factors including the volume of the plant and the severity of the dust explosion hazard. Suppression is likely to be most suited to a permanent installation, it is not applicable to plant that is moved from site to site. Typically, detection is by threshold membrane detectors that provide an electrical contact when the pressure load exceeds a preset value. A range of suppression hardware is available including hemispherical explosion suppressors, mounted inside the plant, which are suited to small volume plant components. Other types include single exit and dual exit suppressors which use a high speed valve to release suppressant. For the more violent dust explosion hazards, high rate discharge suppressors (HRDs) are more appropriate.
Inerting is often done as part of the product quality control procedures where oxidation of the product is to be avoided. This fortunately has the advantage of acting as an explosion prevention measure. Where there is an explosion risk it is common practice to ground IBC, the filling stations and discharge valves, and use nitrogen blanketing to inert the atmosphere within the IBC. Inerting a flammable dust/air mixture can be achieved by replacing the oxygen in the air by an inert gas such as nitrogen. Dust/air mixtures can explode only if the oxygen required for combustion is available in the immediate vicinity of each dust particle. The maximum oxygen concentration at which dust explosions are just not possible cannot be predicted; it depends on the nature of the combustible material and has to be determined
experimentally in each case and will vary with the type of inert gas. Where inerting is used, effective monitoring of the atmosphere should be provided, with automatic action such as shutdown, if the oxygen concentration exceeds a predetermined level.
The metal frame and body of the IBC is, in many cases, grounded as the IBC is loaded onto the metal support frame. However, to ensure that earthing is achieved, a system is used where earthing is provided by automatically engaging a spring plate as the IBC is mounted on the support frame. Apart from the benefit of automation, this also eliminates any doubt associated with a manual system of earth clamps.
There are no known examples of explosion venting or explosion suppression applied to FIBC installations.
Inerting is used in FIBC filling operations and many of the basic principles applicable to rigid IBC, is also applicable to FIBC. Typically during FIBC filling operations the FIBC is filled with the product via a spouted filling opening. Inside the FIBC is a separate carbon impregnated conducting liner. Before filling, the liner is grounded to earth and the bag and liner are sealed at the filling neck using an inflatable seal and finally clamped with an outer
steel clamp ring. The seal is dust tight. The inner conducting liner is inflated with nitrogen via an annular inlet at the neck for a timed period before the purge lines are closed using inlet and outlet valves. The purged nitrogen can be recycled to the plant. It is not necessary to monitor the oxygen level as the degree of inerting is determined by timing the inert purge flow rate through the known volume of the system. Following the introduction of
nitrogen, the bag retains a slight positive pressure which removes creasing in the bag and promotes the complete filling of the bag with no empty voids. Inerting can also be carried out during discharge.
Type C FIBC are electrically conductive and require permanent grounding of the bag during the whole period while the bag is filled or discharged. Typically, earthing clamps are attached to earthing tags on the FIBC, enabling charge to dissipate before it builds up to an excessive level. Earthing relies on a) the operator correctly fitting the earthing clamp, b) in the case of a multi-trip bag, that the bag has not degraded and lost some of its anti-static properties, and c) that the correct type of bag has been used. It is considered that there is a need to earth conductive objects in the vicinity of Type D FIBC in order to avoid corona charging.
Plant operators are generally in close proximity to the FIBC when opening the outlet spout at the base of the FIBC and later the operator may shake the bag to release residual powder. In the event of ignition of the powder, the operator is likely to be caught in the flames. If the flames propagate into the FIBC, a deflagration may occur inside the FIBC causing the bag to rupture due to the pressure development from the ensuing explosion. The following
examples illustrate how incidents involving different types of intermediate bulk containers can occur. Whilst it is well known that the electrostatic hazard posed by FIBC can ignite flammable vapours, these examples illustrate the hazard posed by flammable dust clouds.
- Dust explosion events almost always start inside the process where powder is being moved or processed. However, a recent incident in the UK illustrated that ignitions can be initiated outside the dust containment system. In this incident, at a factory producing plastic packaging, an ignition started with a fire in an air filter associated with a pneumatic conveyor system. An adjacent small aluminium hopper was fed from above by the FIBC. It was located on the top of the hopper with its outlet spout untied. The powdered product was a fine free flowing polymer additive. The fire at the filter spread to the aluminium hopper, which failed, probably when some of themetal melted. This allowed powder to escape into the fire already present, and more powder to flow into the top of the hopper from the FIBC above. The fire then spread to the FIBC, melted the fabric and caused the contents to spill into the fire. A dust cloud formed in the hopper which then resulted in an explosion.
- In an incident in 1989, described by Britton(3), a fine organic herbicide comprising 6–8 micron particles was being discharged from a 100% plastic FIBC. There were no flammable liquids involved. The material was being discharged from the FIBC into a steel chute, approximately 0.45m diameter 5m long, and then into a weigh bin with an attached dust collector. The FIBC needed to be beaten with a rod to loosen the contents. An operator was emptying the FIBC and when he observed that the FIBC was emptying very fast. As it completed emptying, he saw a mushroom cloud of smoke around the FIBC and then a wall of flame travelling towards him. A second employee, approximately 6 m away, heard a rumble and saw a fireball engulfing the FIBC emptying area. The two employees were injured, one with second degree burns. Flash fire damaged equipment, there was structural damage to the building walls and ventilation ductwork was damaged due to overpressure. According to Britton a possible scenario was that the FIBC may have been wet due to rain entering the suppliers truck. The water may have created a conductive patch on the FIBC capable of yielding sparks.
- A company was transferring the contents of a batch of 100% plastic FIBC into a row of drums(3) on a concrete floor. While one operator worked the hoist, a second held two vacuum hoses near the top of each drum to minimize dust leakage into the room and a third operator regulated flow from the FIBC. At the time of the incident, the drum was being filled and the FIBC was being “puffed” to shake out residual powder. The three operators observed the material on fire, in the drum. Flame propagated into the FIBC and all three operators received first degree burns. It was clear that “puffing” the FIBC, when almost empty, created a dust cloud. The operators and drums were not grounded and a spark may have occurred between the operator holding the vacuum hoses and the ungrounded top of the drum.
- An incident(4) occurred in which product was being transferred from a hopper to an IBC via a butterfly valve and a flexible hose. The IBC was a rigid metal container which was sitting on a metal floor and was adequately grounded. The dedusting system attached to the IBC had a 75 mm flexible low conductivity rubber hose between the metal cover and metal pneumatic duct leading to a bag filter. The hose was not earthed. The 200 mm diameter transfer hose was made of a flexible low conductivity rubber with a spiral metal wire but this was not earthed. Having observed that the rate of flow was decreasing, the operator fully opened the butterfly valve.
The material then began to flow rapidly into the IBC, such that the metal cover resting on top of the IBC lifted and “danced about”. An explosion then occurred and a fireball expanded into the workplace inflicting some first and secondary degree burns on the operator. The subsequent incident investigation found that the explosion was not as violent as predicted by the dust explosibility data. The dust concentration was thought to have been below the optimum concentration which resulted in a weak explosion. There was little damage to the equipment; the fireball did not propagate from the IBC to the bag filter, and there was no secondary fireball in the
workplace. The investigation revealed that even though the butterfly valve had been fully opened, bridging caused material flow to cease and there existed simultaneous electrostatic charging and dissipation. The charges were caused by movementof material into the loading hopper, above the IBC. At the same time, charges were being dissipated to the walls of the loading hopper. New material was added to the top of the pile in the hopper. When the bridge collapsed, a conglomerate of electrostatically charged material fell into the IBC. The displacement of air in the IBC caused the cover to “dance about”. At this point, the cover was ungrounded and
was in the presence of an electric field from the falling charged conglomerate. This caused an electrostatic charge to be induced into the metal cover which then discharged an incendive spark to the grounded IBC.
- A Type B FIBC, with a breakdown voltage ,4 kV to prevent propagating brush discharges, and was considered suitable for this duty, was being filled with a plastic product(5). A short time before the explosion, there was a greater throughput of product and possibly a greater proportion of fines in the bulk material. The product was described as a low-conductivity product, the resistivity of the bulk material was .1012 ohm.m and the MIE was 3–10 mJ. The material was conveyed pneumatically from a mill through a dust separator into the FIBC. The system was made from a conductive material and was grounded. The potential sources of ignition such as the mill overheating or ingress of foreign objects could be excluded. Past experience had shown that the product could produce strong electric fields emanating from the FIBC. This fact did not, however, cause concern since propagating brush discharges were not expected. Analysis of the explosion incident suggested that while the highly charged bulk material was being filled into the FIBC, a cone discharge originated in the product. As the plastic product had a broad particle size distribution, with particles about 1 mm diameter down to a fine powder, cone discharges with an energy up to multiples of 10 mJ could occur. Following the incident the FIBC was replaced with a conductive and grounded Type C FIBC. The report stated that cone discharges may not be completely prevented in this type of FIBC. It is believed that the changed direction of the electric field (the wall of the FIBC now at zero potential) means that the energy of the cone discharges are considerably reduced. There was no detail relating to the extent of the fireball or the damage sustained although the report did indicate that there was “little damage to the plant”.
The trend towards IBC in preference to smaller containers can markedly change the characteristics of the dust cloud. Filling times in small drums are relatively short, the powder concentration within the filling stream is normally above the Upper Explosive Limit, and little, if any, separation of “fines” occurs in the ullage space. All these factors mitigate against the initiation of a propagating explosion. However, in IBC, the filling stream forms a smaller part of the volume and a dispersed cloud can be formed around it. The dust cloud characteristics will depend on the physical form of the powder and the filling procedure but in many cases there will be an increased tendency for a dust cloud within the flammable limits to be formed. Furthermore, if the filling procedure is such that the “fines” can separate from the coarser material, then a dust cloud requiring a lower energy for ignition may be formed. The major hazard in any dust explosion arises when the dust enclosure ruptures, the dust is dispersed in the general plant area and a secondary explosion is initiated. The use of IBC will increase the possibility and extent of secondary dust explosions in that (a) they are mechanically weaker than drums and will rupture more readily with possibly a large split and (b) a greater quantity of powder is available for dispersion.
Many of the explosion incidents involving FIBC are attributed to spark or electrostatic ignition sources and so called “anti-static” fabrics have been developed. However, these are solely intended to prevent electrostatic discharges from the FIBC fabric itself, that is, make the FIBC electrostatically equivalent to metal IBC. This does not preclude the possibility of incendive discharges from the bulk powder. Nor does it provide protection against other sources of ignition e.g. burning powder from upstream units. Many process plant designs integrate closely the powder processing stages. The packaging of materials into an IBC can result in a more direct link between units (e.g. mill, blenders, dryers), in which combustion of bulk powder may be initiated. This may increase the possibility of burning material initiating a dust explosion in the IBC. Explosions may be prevented by inerting the atmosphere to control the oxygen content to a safe level. However, it appears that monitoring of the oxygen content is rarely done in preference to blanket inerting or relying on known volumes and gas flow rates. Explosion suppression is not normally used to protect FIBC but it is often used for the protection of rigid IBC. However, inerting is often preferred where it is an economic option.
Explosion venting of rigid IBC is used but is not common, mainly due to the inconvenience of fitting explosion vents to the body of an IBC. A practical alternative is to install the explosion vent to the ducting immediately upstream of the IBC; this will only be effective during the filling operation. During product discharge, venting is more difficult, but an explosion vent fitted to a rigid IBC can be achieved by using removable top section. A telescopic duct over the top section of the IBC would relieve the explosion.
1. Hazardous Cargo Bulletin. Cs and Ds fight it out, P59–60. July 1998.
2. HSE, Health and Safety series booklet HS(G)103, Safe handling of combustible dusts,
precautions against explosions, ISBN 0 7176 2726 8, 2003.
3. Britton L. G. Static hazards using flexible intermediate bulk containers for powder handling.
Process Safety Progress (Vol. 12, No. 4), October 1993.
4. Pratt T. H. Static electricity in pneumatic transport systems: three case histories. Process
Safety Progress (Vol.13, No. 3) July 1994.
5. Glor M. Overview of the occurrence and incendivity of cone discharges with case studies
from industrial practice. Journal of Loss Prevention in the Process Industries Vol. 14
No. 2, March 2001 123–128.
SYMPOSIUM SERIES No. 150 # 2004 Crown Copyright
Source: P Holbrow Health and Safety Laboratory, Harpur Hill, Buxton, SK17 9JN SYMPOSIUM SERIES No. 150