Abstract of Special Research Report (SRR-No.34)

National Institute of Occupational Safety and Health, Japan

A Study on the Prevention of Explosions and Fires for Recycle Industries (Final Report)


Takayuki ANDO

: With the recent promotion of the conservation of energy and resources and reduction of waste, many materials are being recycled. Some aspects of recycling, however, are dangerous, and frequent incidents of explosion, fire, and poisoning have been reported. These hazards are attributed largely to the increase in the amount and types of recycled substances, the replacement of chlorofluorocarbons with liquefied petroleum gas, which is done to protect the ozone layer and to ward off global warming, and the use of flammable cleaning liquids instead of nonflammable ones in an effort to reduce environmental pollution.
    The objective of this research was to present accident-prevention measures that are applicable to the industry by evaluating the fire and explosion hazards related to the recycling process, developing techniques to evaluate the risk of such hazards, and designing technology that is capable of controlling or preventing the development of explosions.
    The research subjects of this project were as follows:
  (1) Study of the fire and explosion hazards resulting from mixing chemicals during the recycling process.
  (2) Study of explosibility and combustibility of discarded flammable substances.
  (3) Development of an explosion mitigation system applied to crushing and milling equipment.
    This "Final Report" shows the results of this research obtained after the "Midterm Report" published in 2002.
    Chapter 2, 3 and 4 are related to the research subject (1). The purpose of Chapter 2 was to propose a simplified equipment to test the incompatible hazard of materials. Chapter 3 and 4 dealt with the evaluation of the hazard associated with the catalytic effect of metals on decomposition of chemicals.
    Chapter 5 is related to the research subject (2). Chapter 5 has shown the importance of burning velocity to estimate the energy of deflagration in open space.
    Chapter 6 and 7 are related to subject (3). The purpose of Chapter 6 was to explore the fundamental aspect of the spreading fire in the pile of RDF(Refuse-Derived Fuel). Chapter 7 is about the results of full-scale experiments planned to verify the revised edition of the technical guide for explosion venting.

Methods for Evaluating Reaction Hazards


: The disposal of industrial waste materials is potentially hazardous, especially as the blending of recycled chemicals (liquids etc.) is gradually increasing. It is thus possible that unexpected chemical reactions could occur and fires and/or explosions result. One reason leading to this situation would be scarce and/or incorrect information about waste materials; an indication that sufficient safety information can no longer be obtained from literature searches. Methods of screening the blending need to be performed, but these tests must come at a reasonable cost, have good performance, and be easy to carry out because most chemical waste disposal companies are small, are not well capitalized and have few experts available to evaluate safety. Easily-operated and low-cost screening methods for reaction hazards that have reasonable performance are useful for such companies. This research details methods that are easier and lower-cost than a typical high performance reaction calorimeter, the Mettler-Toledo RC1. We also show that these methods are more effective than those carried out in typical waste disposal workplaces where waste chemicals are deposited. Two tests are applied as useful screening methods. The first one is the OmniCal Super-CRC, a commercially available differential reaction calorimeter, while the other is a custom-made differential calorimetry system using a principle similar to that of Super-CRC (labeled as CCS).
    The liquid-liquid heterogeneous reaction _ such as the blending of waste oil and water _ was investigated in this paper for examining calorimetry performances.
    As the accidents caused by the unexpected heterogeneous reaction, the nitration of aromatics is very famous in the chemical industry. And the same types of accidents have happened in waste disposal fields. An example of a heterogeneous reaction is the explosion that occurred on a lorry that was laden with waste of concentrated acid taken from drums; unbelievably, there was waste water in one of the drums. The water in the drum was pumped out into the lorry and the lorry then exploded.
    In the study of heterogeneous reactions, the availability of screening tests are examined in respect to the prediction formula for the heat release rate; on the basis that the heat release rate can be estimated using the conditions of the agitation (agitating speed, etc.) if it is diffusion controlled.

The Investigation of Induction Periods in the Reaction of Hydrogen Peroxide


: Chemical industries and laboratories have utilized various kinds of chemicals. In disposal process, they are gathered into a waste disposal tank for following disposal treatment such as incineration. The gathering involves risk of unexpected reactions; the waste processes generally proceed without enough care while development and production process require for careful quantitative hazard estimation.
    Hazard increases in waste disposal tank when induction period exists prior to exothermic reactions. In a disposal process, a tank is left unobserved after disposal since chemicals appear not to start a reaction until induction period completes. Then unexpected runaway reaction or explosion cause severe damage. Even if personnel know necessity of screening test with small amount, the hazard is not obvious during a short period.
The test result allows for mixing of large amount of chemicals and causes runaway reaction consequently.
    This research focuses on induction period in the reaction between hydrogen peroxide and copper chloride by which a tank car exploded during transporting waste disposal chemicals. This report describes the experimental investigation of the effect of CuCl2 on hydrogen peroxide and comparing it with two other copper(II) compounds (nitrate: Cu(NO3)2; and copper sulfate: CuSO4 ) and three iron(III) compounds (chloride: FeCl3; nitrate: Fe(NO3)3; and sulfate: Fe2(SO4)3).
    The experiments were performed using a reaction calorimeter. During the experiments at 35 oC, 2 x 10-5 mol of copper compounds slowly reacted with H2O2 and generated a precipitate. The iron compounds allowed the hydrogen peroxide to violently decompose. A 1 x 10-4 mol solution of CuCl2, however, produced a violent decomposition at 35 oC. At 15 oC, a moderate heat release occurred.
    Based on these results, the concentration and temperature dependence of the catalytic ability of CuCl2 were postulated to contribute to the induction period observed in the accident.

The Effect of Chemical Structure on the Interaction with Metal Ion in DIsposal Process


: Some hazardous materials show characteristic features in the reaction with transition metals; they are unstabilized, and sometimes decompose in contact with metal ions. This character may cause severe damages especially in waste disposal process of chemicals. Waste disposal process is less attractive rather than research and development. They sometimes pour chemicals into a waste tank and mix with ones others disposed of beforehand. That may lead unexpected reaction and explosion. The knowledge of the reaction between metal ions - unstable chemicals enable to prevent such disasters as well as thoroughly implementing confirmation of safety.
    This research relates some investigation to deepen understanding of the nature of structural effect on reactivity between unstable chemicals and metal ions by using of NH2OH and Fe(III) ion. The investigation can be divided into two parts: steric hindrance on Fe3+ by using of ligands, and steric hindrance on NH2OH by using of substitutent methyl group. In the reaction with hydroxylamine, three types of Fe(III) species exhibited different behavior and suggested that steric effect can prevent precipitate. Besides, ligands completely capping Fe(III) enable to catalytic decomposition triggered by direct interaction between metal ions and unstable chemical molecules.
    In the research on hydroxylamine and its derivatives, all the substituted chemicals exhibited lower reactivity than hydroxylamine. As a whole, acidic condition and steric hindrance on oxygen decelerate the interaction with metal. Based on ab initio molecular orbital calculation, interaction with iron(III) through the oxygen atom caused extend the N-O bond in NH2OH ; the length of N-O is 1.4692 A, and 1.4899 A in the complex. The coordination bond between Fe3+ and the oxygen atom is 2.0733 A. In the optimized structure simulating interaction through the nitrogen, however, 2.2185 A is the most stable distance between Fe3+ and the nitrogen atom. The interaction through oxygen has a more influence on bond length in hydroxylamine than through nitrogen.

A Computational Verification for Deflagration Blast in Open Space

Teruhito OTSUKA

: When the speed of energy release, such as the flame velocity and expansion speed, is much greater than the ambient sound speed, such as in a high explosive's detonation or a detonation of unconfined vapor, all released energy is concentrated on the shock surface. Therefore, the Sachs' scaling law could well explain the effects in these cases.
    However, it requires an enormous amount of energy to initiate the detonation of an unconfined vapor cloud and air mixture, and this is unlikely to happen. The greatest difference between a high explosive explosion and gaseous deflagration, is the burning speed which contributes to the generation of a blast. In the case of a vapor cloud deflagration, the burning velocity is not fast enough to apply blast theory. The TNT equivalence ratio is typically used for fixing this difference. There is arbitrariness about the TNT equivalence ratio because of an uncertainty of the burning velocity. In addition, some reports include all vapor and gas leakages in the blast estimation. Therefore, they also included the vapor and gas that leaked during the early stage, and were thinned to be lower than the minimum explosion concentration by diffusion. This makes the TNT equivalence more unclear.
    In this paper, it was simulated that the hydrogen-air mixture deflagration with k-ε and Eddy-Break-Up as turbulent and combustion model. The simulation showed that the small differences of the ignition delay give the large influence to the energy estimation, almost 10 times without considering burning velocity variance which leads to the low reproducibility of the explosion experiments.

Fire Spred in RDF(Refuse-Deriverd Fuel) Pile


: The technology of waste handing and resource recycling advances, and RDF or RPF are used as fuel for the incinerator in order to obtain heat energy and electricity energy in Japan. The RDF used in Japan is RDF-5 in ASTM. Fire or explosion is being occurred with the increase in use of the RDF at manufactories or storage facilities. The spreading of combustion zone in the RDF pile is the primary phenomenon in an enlarging fire. The phenomenon is called smoldering. In this study, temperature change in the pile and flame established over the pile were examined in order to explore the fundamental aspect of the spreading fire. In addition, a point to notice in extinguishing the RDF was examined. The RDF sample was piled up to 300 mm using a cylindrical container (inner diameter; 160 mm, height; 350 mm) which consisted of heat insulating material. The cylindrical container was put on a hot plate, and the RDF ignited in the hot surface (bottom). Upward spreading fire is a subject in this study. Considering the effect of the permeability or the natural draft on the spread of fire, the experiment was carried out under next 3 conditions.
  a)Experiment No.1: without permeability of the bottom face / with permeability of the side
  b)Experiment No.2: without permeability of the bottom face / without permeability of the side
  c)Experiment No.3: with permeability of the bottom face / without permeability of the side
    The results are summarized as follows:
  (1) In the condition without permeability of the bottom face, mean spread rate of the combustion zone is from 0.24 to 1.8 mm/min. The rate of this range is equivalent to those of general smoldering.
  (2) The temperature distribution in the RDF pile is examined, and it is clarified that the combustion region are wide.
  (3) When the pile would collapse with smoldering, then the clearance between the RDF pile increases, the inflow of the air from the top to the pile became easy and that the flame became easy to establish.
  (4) When the flame establishes in the clearance in the RDF pile, the spread rate increases, since the preheating of RDF of the upper layer is promoted.
  (5) In the final stage that the combustion zone reaches the upper surface in the pile, it is found that large flame can establish suddenly over the pile.
  (6) The RDF consists of the multicomponent, and it has the characteristics of absorbency and large heat capacity. Therefore, if spraying water is not appropriately carried out in the fire extinguishing, otherwise subsidiary reaction of the incomplete combustion generated combustible gas and CO, and it may become a cause of re-ignition or gas explosion.
  (7) It should care about differing on the component of the gas which arises from the RDF in fire and thermal analysis in laboratory, because the rate of temperature rise in smoldering combustion is small.

Full-scale Experiment on Dust Explosion Venting


: The technical recommendation for explosion venting of revised edition, NIIS-TR-NO.38 (2005), was published referring to NFPA 68 (2002) mainly in June, 2005 in NIIS. A full-scale experiment was planed with the purpose of verification of vent area and flame length for dust explosions based on the technical recommendation. In addition, it was planed to obtain a useful knowledge on the protective measure to the circumference. In the experiment, 4 vessels ( 0.2, 1, 6, and 20 m3 ) and wood powder (particle size; 8.8 μm ) were used. In the measurement, the following were examined: internal pressure of the vessel, propagation velocity of the flame in the vessel, velocity of dust cloud ejecting from the vent opening, flame length, sound level. Results are as follows.
  (1) Although the experiment using the vessels of 6 m3 and 20 m3 were carried out, the experiments using the vessels which exceeded 1 m3 was for the first time in Japan. In the experiment on 6 m3 vessel, the experiment was perfectly controlled.
  (2) It was found that the length of the flame which ejecting from the vent opening was shorter than that in the technical recommendation.
  (3) The calculation formula of vented area on dust explosion in the technical recommendation is seemed to be appropriate. However, in the workplace, maximum explosion pressure ( P max ) and explosion index ( K St ) on handling powder are carefully evaluated, and the technical recommendation should be used.

Other Publications

Kiyose District Map

Umezono 1-4-6,Kiyose,
Tokyo 204-0024 Japan
TEL: +81-42-491-4512
FAX: +81-42-491-7846

Noborito District Map

Nagao 6-21-1, Tama-Ku,
Kawasaki 214-8585 Japan
TEL: +81-44-865-6111
FAX: +81-44-865-6124