Showing papers in "Journal of Loss Prevention in The Process Industries in 1993"

Determination of self-accelerating decomposition temperatures for self-reactive substances

Abstract: A self-accelerating decomposition temperature ( T SADT ) is the lowest ambient air temperature at which a self-reactive substance undergoes an exothermic reaction in a specified commercial package in a period of seven days or less. The same substance and package must be able to survive for seven days at a temperature within 6°C of the temperature at which the reaction occurred. A T SADT is determined for the purpose of deciding whether a self-reactive substance should be subject to temperature control during transport. We have re-examined the established T SADT test methods to clarify their application, affirm their validity and extend them to a wider range of materials. Two of the four test methods recommended by the United Nations Orange Book 1 require calculations to determine a T SADT . Our proposed method, which uses Accelerating Rate Calorimeter (ARC) data, requires similar calculations to determine a T SADT . To ensure determination of conservative values when using any of the methods for self-reactive substances, the investigator should 1. consider pressure increases and potential package failure due to non-condensable gas formation and 2. account for the effects of a non-uniform temperature distribution within a viscous liquid, paste or solid by using the Frank—Kamenetskii thermal explosion model. Methodologies are discussed which address both of these areas. Also discussed is a method to account for the effect of reactant depletion on predicted T SADT values for both the Semenov and Frank—Kamenetskii thermal explosion models. This correction can be significant and is required to ensure that analytical values agree with experimental values.

Use of the accelerating rate calorimeter and the thermal activity monitor to estimate stability temperatures

Abstract: Self-accelerating decomposition temperatures derived from accelerating rate calorimetry measurements and heat accumulation or Dewar tests are compared with those from the definitive United States self-accelerating decomposition temperature test. Neither simulation is biased on average relative to the United States test. However, the use of the accelerating rate calorimeter, as recently proposed, or the Dewar test, as currently recommended, to estimate self-accelerating decomposition temperatures requires wide safety margins to be applied. Closer and more consistent agreement with US test results is achieved by estimating self-accelerating decomposition temperatures from accelerating rate calorimeter and thermal activity monitor data together.

Gas explosions in linked vessels

Abstract: In process plants, vessels handling flammable mixtures are often interlinked through pipeline systems. Explosions communicating between vessels connected in this way have occurred, often with devastating results in the secondary vessel. There is very little data on the explosion development in these geometries and no guidelines for their protection against this hazard. Some data are presented here for methane/air explosions in a 0.5 m diameter, 0.5 m long vessel connected to an identical vessel through a 76 mm diameter, 1.7 m long tube. Detailed pressure measurements were taken in both chambers and the flame propagation in the system was recorded. The initial laminar combustion in the first vessel induced a gas flow through the linking pipe into the second vessel. For 10% methane/air mixtures the pipe gas velocities ranged from 70 to 130 m/s. This high velocity flow generated high turbulence levels in the pipe and in the second compartment. When the flame encountered this turbulence it was accelerated to 370 m/s and resulted in a violent explosion in the second vessel. This in turn generated an even faster flow back into the ignition vessel resulting in an equally violent deflagration. With ignition at the centre of one of the vessels a 4 to 6-fold increase in the violence (compared to what it would have been in a single vessel) was measured in both vessels. With ignition at the end of the vessel the enhancement factor in the second vessel was about 17. This was twice the maximum reported in the literature. The venting implications were discussed and it was concluded that prevention of the explosion transmission to the second vessel should be a priority, in order to avoid an explosion which could be impossible to vent.

Evaluating the probability of major hazardous incidents as a result of escalation events

Abstract: A method has been developed in which the frequency contributor to major hazardous releases from escalation events has been quantified The originating events which may escalate into a larger event with greater consequences include jet flame impingement, pool fire engulfment, vapour cloud explosion effects, and missile effects For each originating event, its frequency of occurrence is calculated This, in combination with the likelihood of its consequences causing a failure of the equipment being studied, is the frequency of the escalation contributor to the overall failure frequency

Experimental study of thermal hazards during the hydrogenation of aromatic nitro compounds

Abstract: Catalytic hydrogenations of aromatic nitro compounds are known to be potentially hazardous reactions. Their safety depends both on the properties of compounds and on the operating conditions. The focus of this paper is on thermal hazards. Firstly, the thermal stability of the reaction mixtures determines the temperature range in which the reaction can be run safely. Secondly, accumulation of hydroxylamine is a potential hazard as it can lead to undesired side-reactions, which are highly exothermic and, in contrast to the hydrogenation, cannot be controlled by the hydrogen supply. Using a reaction calorimeter, the thermal data of the reactions, i.e. overall heat and heat flow during the reaction, can be measured. Conclusions on the accumulation of intermediates in the reaction mixture can also be drawn from these data. Finally, the thermal behaviour of the reactor in the case of a cooling failure can be predicted with a high accuracy.

Explosion enhancement through a 90° curved bend

Abstract: Junctions and bends are commonly incorporated in pipeline systems conveying potentially explosive mixtures. Very little data exists on how these affect the development and transmission of an accidental explosion. There is therefore great uncertainty when designing explosion protection measures for such systems. A 3 m long, 162 mm diameter tube, closed at both ends and incorporating a 90° curved bend was used to investigate the influence of the bend on the development of gaseous explosions. Methane/air mixtures of 10% and 5.7% by volume were used. The mixture was ignited at one end of the tube at 5 mm or 53 mm from the flange. It was found that for all explosions, the flame moved faster around the inner wall of the bend than the outer and hence it was elongated. This gave rise to an overall acceleration of the flame and a significant increase in the rate of pressure rise. The enhancement factor due to the bend ranged from 4 to 6 for the 10% mixture and it was equivalent to the effect of an orifice plate with a 20% blockage in the path of the flame. These findings highlight the need for more work on bends of different shapes and in different layouts.

Pipeline leak detection based on mass balance: Importance of the packing term

Abstract: After a brief survey of pipeline leak detection techniques, some mass balance systems are presented. Such systems use a pipeline flow model in order to compute the change in pipeline inventory during a transient flow. The packing term, whose importance is highlighted, is a function of the speed of sound in the pipeline. However, generally, the parameters on which the speed of sound depends are poorly defined. The consequences on leak detection thresholds are described.

A simplified method for predicting the effect of ducts connected to explosion vents

Abstract: A phenomenological model of explosion venting through ducts has been developed. The simplifed one-dimensional model includes the effects of inertia of the duct contents, flame propagation in the duct, and skin friction, while ignoring the details of the wave motion in the duct. The model is only slightly more complicated than the widely used algebraic quasi-steady discharge equations, since it is in the form of a first order ordinary differential equation. The model has been validated against the method of characteristics solution of the full partial differential equation for the case of unsteady discharge from an infinite reservoir. The model has also been found to provide good agreement with data obtained for the unsteady development of ejector flow. The ducted vent model coupled with a simplified explosion model is used to perform detailed comparisons of the experimental and calculated pressure—time traces. The model is shown to predict the effect of the presence of the ducts remarkably well for the low reduced pressure explosions for which it was developed. The model correctly predicts all the qualitative features of the data, reproducing details that have not been predicted by previous phenomenological models. The Helmholtz oscillations commonly observed in ducted vent experiments arise naturally from the formulation of the model. The ducted vent model, if used with a more accurate explosion model, is capable of predicting the peak negative pressure in the vented volume, which can cause the implosion damage sometimes observed after an explosion.

Marine Transport of Dangerous Goods. Risk Assessment Based on Historical Accident Data

Abstract: In this paper marine transport of dangerous goods is surveyed on the basis of 151 accident case histories. Accident frequencies have been estimated for the different accident types (collisions, groundings, fire/explosions and structural damage) and are in the range of 1 × 10−3 to 2 × 10−2 per ship per year. Further estimates of the probabilities of spillage of at least 100 tons of cargo and/or fatalities in connection with the accidents have been made. The consequences measured in number of fatalities were compared for each cargo type (oils and chemicals) and it was shown that accidents involving oils were twice as frequent as accidents involving chemicals. However, the distribution of number of fatalities seems to be similar for the two types of goods. Concerning the local surroundings (port, coastal waters and open sea) it was shown that most accidents, both with small and large consequences, happen in coastal waters. The sizes of the spills have been modelled by linear regression based on accident type and the size of the ship. Fairly good correlation between the spill size and the size of the tanker was shown for groundings, structural damage and fire or explosions.

Preliminary safety analysis

Abstract: Various major safety studies are carried out at appropriate stages during a project. Many companies do some form of preliminary analysis at points between initial project concept and when the process design is completed. These studies aim to ensure that the decisions on process design and site selection take full account of process safety requirements and related risk and environmental constraints. Methods have been incorporated and developed during this work to take account of best industrial practice for such safety studies. These are listed under the general heading of preliminary safety analysis (PSA) and are carried out from the time of the concept safety review until such time as reasonably firm process flow diagrams or early P & I diagrams are available. The methods included are as follows: • concept safety review (CSR) • critical examination of system safety (CE) • concept hazard analysis (CHA) • preliminary consequence analysis (PCA) • preliminary hazard analysis (PHA) These have been developed from a model of the plant and its interpretation as part of an incident scenario. The emphasis throughout is on utilizing the best points to start the search to identify undesired events contributing to the development of accidents. For the main method described, preliminary hazard analysis, this search has as its starting point and fulcrum the ‘dangerous disturbances of plant’ which arise at a point in the incident scenario just after emergency control measures have failed to control the situation. The study should be conducted using risk evaluation sheets which model each stage of the incident scenario and allow for a short-cut assessment of risk when this is desired. The above methods are demonstrated by part of a simplified case study. The methods function well and provide not only a good model of incident scenarios but are readily developed into fault and event trees and operating procedures. They are invaluable for the development of safety reports for regulatory authorities. Furthermore, by not imitating HAZOP methods they strengthen the effectiveness of the search process.

The FIRES project: Experimental study of thermal runaway due to agitation problems during toluene nitration

Abstract: The mononitration of toluene by mixed acid has been taken as a specific case of heterogeneous liquid-liquid reaction, and the influence of stirrer malfunction, as a potential cause for hazardous situations, has been experimentally studied in a 100 litre pilot reactor. The results and their implications for safe processing of discontinuous aromatic nitration are discussed and analysed using a mathematical model that simultaneously takes into account mass transfer phenomena and chemical kinetics.

Prediction of pressure and flame effects in the direct surroundings of installations protected by dust explosion venting

Abstract: When applying dust explosion venting as a measure to protect installations one should always bear in mind the secondary effects of application of venting, viz blast waves and flames emerging from the vent. These two effects pose hazards to the direct surroundings of explosion vents and these hazards should be considered during the design phase of the equipment. This paper will give an overview of the data and methods that are available to predict flame and blast effects associated with vented dust explosions.

Thermal stability screening and reaction calorimetry. Application to runaway reaction hazard assessment and process safety management

Abstract: This paper describes a methodology for runaway reaction hazard assessment and process safety management in this field. This methodology includes: o ⊙ a preliminary hazard assessment method based on information from the literature ⊙ a testing methodology to determine the product thermal stability as well as the relevant kinetic and pressure information ⊙ a method of process safety analysis devoted to runaway reaction hazard, including homogeneous and heterogeneous situations as well as Arrhenius and non-Arrhenius reactions ⊙ recommendations for prevention and mitigation measures, based on process situations and on the experimental information obtained This methodology should be applied to any complex situation for the control of runaway reaction hazard

Fire spread along structural metal pieces in oxygen

Abstract: Recent results of studies on fire spread along structural metal pieces in high pressure oxygen are described. The results of fire spread along cylinders of iron and aluminium, which are representative of heavy and light structural metals, are presented. The parameters, on which fire spread depends, are identified. The process by which structural metal combustion occurs is described. A suitable model of fire spread along metal pieces in high pressure oxygen is proposed.

Evidence of the oxidation of deposits in heated bitumen storage tanks

Abstract: Thermal analytical techniques have been used to examine the exothermic nature of the deposits found in heated bitumen storage tanks and to determine if they could act as an ignition source to flammable vapours. The deposits were found to undergo an oxidation/smouldering reaction at temperatures found on the underside of the roof of the tanks. Onset temperatures for exothermic behaviour ranged from 337 to 406 K. Deposits taken from refinery tanks were found to have an activation energy in the range of 80–90 kJ mol −1 . The self-heating found in deposits constitutes a smouldering reaction which could deplete the oxygen concentration and so contribute to the conditions suitable for the formation of pyrophoric material. In a situation where cross-ventilation can occur and air is allowed to pass over the surface of the deposits, the rate of reaction would increase leading to higher temperatures and increasing the risk of fire.

Mitigation of hydrofluoric acid releases: simulation of the performance of water spraying systems

Abstract: This paper describes an assessment of the effectiveness of two water spray configurations in mitigating an accidental release of hydrofluoric acid (HF) in an actual facility. This assessment is based on mathematical simulations using the model HFSPRAY 1 and wind tunnel modelling of releases in a scaled down prototype of the actual facility 2 . The spray configurations considered are: (i) two headers at about the same height, one equipped with spray nozzles pointing upwards and the other with the same nozzles pointing downwards, and (ii) two headers at different elevations with nozzles pointing horizontally toward the release. In these simulations, HF releases at two elevations and flow rates were considered to bracket the range of potential releases. At grade (1 m above the ground) the release rate was 43 kg s −1 and at a 15 m elevation the release flow rate was 37 kg s −1 . The water flow rate in the entire system was 33 000 gpm. According to the HFSPRAY simulations, the two-tier horizontal configuration removed HF more effectively than the up-and-down configuration for these specific release heights. Effectiveness ranged from 70% for high wind speeds (e.g. 17 m s −1 ) to 96% for average wind speeds (e.g. 5 m s −1 ).

Environmental impact of a warehouse fire containing ammonium nitrate

Abstract: Burning of materials is one of the most investigated reactions. However, there are some areas, such as warehouse fires, which are still not well understood because of the complexity of the problem. The problem in warehouse fires is the combinations and reactions of different chemicals and substances the nature of which is not always known either by the experts or by the people extinguishing the fires. Hence, warehouse fires can have very serious effects on the environment in air, water and soil, as experienced in the Sandoz fire, for example. The objectives of this study were to gain experimental data on combustion products at the laboratory scale and to demonstrate the use of these results by introducing a case study where 50 000 kg of ammonium nitrate is stored in a warehouse.

Velocity of a single small missile ejected from a vessel containing high pressure gas

Abstract: A single small missile ejected from a vessel containing high pressure gas is driven by the under-expanded gas jet issuing from the breach. A model is developed to predict the velocity of such a missile and the results are compared with data obtained in a parallel experimental investigation. Using the predicted dependence of missile velocity on the jet source pressure as a basis for correlating the experimental data, upper limit velocities for jet driven missiles are defined.

Study of iron sulfide as a possible ignition source in the storage of heated bitumen

Abstract: It is thought that pyrophoric iron sulfide may be able to form in heated bitumen storage tanks and act as a source of ignition to the flammable vapours in the ullage space. Electron microscopy has been used to indicate the possible presence of pyrophoric material within a sample of bitumen deposit taken from a refinery tank. High levels of sulfur were found in areas rich in iron and part of this sulfur was in the form of a sulfate, a product of the weathering of pyrite. The sulfur found in the iron-free areas of the deposit was in the form of elemental sulfur.

The phenol + formaldehyde runaway reaction. Vent sizing for reactor protection

Abstract: The purpose of this paper is to demonstrate the use of the experimental techniques of the process safety laboratory to investigate the runaway reaction hazard of the phenol + formaldehyde reaction. A preliminary study of this reaction, using DTA and isothermal calorimetry helps in understanding the various aspects of the runaway reaction hazard and allows us to recognize the worst credible deviation leading to the most severe consequences. A vent is sized to protect the reactor from exploding due to this runaway reaction. The DIERS methodology is applied to determine the vent size: • characterization of the reaction behaviour • acquisition of the data necessary for vent sizing from adiabatic calorimetric techniques. The choice of the best calorimetric technique, to obtain experimental data for vent sizing, is discussed following the results of the various calorimetric methods. • taking into account the system behaviour and the experimental data obtained, a vent size is determined using various vent sizing equations and a relevant two phase flow model. This work demonstrates that a careful study is necessary to obtain reliable expeimental data for vent sizing. A single experiment could be misleading. This supports the conclusion that a good safety level is better achieved by the careful study of a limited number of dangerous reactions than by limited work on a great number of reactions.

Characterization of fire products from organophosphorus pesticides using the DIN 53436 method

Abstract: Fires in warehouses containing chemical substances such as pesticides occur frequently causing harm to humans and the environment due to the toxic fire effluents generated and dispersed with the fire plume. It is very difficult to assess the potential risks, since only limited documentation exists concerning the quantities and nature of these toxic products. Consequently, there is an urgent need to develop methods to characterize the fire products from fires involving chemical substances. This paper describes a series of experiments carried out in order to evaluate the combustion products from three organophosphorus pesticides (dimethoate, azinphos-methyl, and parathion-methyl) using the DIN 53436 method. The concentrations of the combustion gases such as carbon dioxide, carbon monoxide, sulfur dioxide, nitrogen oxides and hydrogen cyanide have been determined and the organic combustion products have been identified.

Determination of the thermokinetic parameters of an exothermic reaction using isothermal, adiabatic and temperature-programmed calorimetry in conjunction with spectrophotometry

Abstract: Spectrophotometry was used to determine the isothermal kinetics of the esterification reaction between sec. butyl alcohol and propionic anhydride. The effect of adding sulphuric acid as a catalyst was investigated and the results compared with rates of heat generation measured using isothermal reaction calorimetry (RC1). Adiabatic calorimetry (PHI-TEC) and differential scanning calorimetry (Perkin Elmer DSC7) were used to measure rates of heat generation over a wide range of temperature and composition. Values for the heat of reaction were determined using the calorimetric data, and kinetic parameters for the reaction were derived from the isothermal concentration-time profiles in conjunction with the results of isothermal and adiabatic calorimetry. The suitability of differential scanning calorimetry for kinetic studies is assessed.

HGSPRAY: A complete model of spraying unconfined gaseous releases

Abstract: The model presented here, HGSPRAY, is an expanded and refined version of the GAS-SPRAY model presented by this author in an earlier study. HGSPRAY simulates the momentum, mass and energy interactions between several water sprays and a turbulent plume of a heavy gas (HG) in air; the model can predict the flow velocities, temperature, water vapour and gas concentration fields in two-dimensional large geometries for spraying in any direction. The model has been verified with the data obtained from the Hawk field experiments of hydrofluoric acid (HF) mitigation, at the DOE Nevada Test Site.

DIERS churn—turbulent disengagement correlation extended to horizontal cylinders and spheres

Abstract: For a vertical cylinder and non-unity distribution parameter (C0), an analytical solution is given for the DIERS (Design Institute for Emergency Relief Systems) churn—turbulent disengagement model (i.e. axial void fraction profile under volumetric heating model). This solution validates the earlier approximation of Fauske et al.1. Horizontal cylinders and spheres were investigated using numerical integration and the results plotted against a generalized dimensionless superficial vapour velocity. Thus, the analytical solution is shown to hold for varying cross-sectional vessels. Therefore, this correlation predicts liquid disengagement regardless of vessel shape. An alternative correlation form is introduced, and examples for design and for rating are included.

Report of an explosion during the manufacture of an azodicarbonamide formulation

Abstract: A damaging explosion occurred during the manufacture of an azodicarbonamide formulation which was not explicable in terms of existing knowledge and experience. Three scenarios were advanced which could account for the facts of the incident. Two of these were eliminated by subsequent experimentation but the third was demonstrated to be a viable mechanism. It is concluded that the event was a bottom-initiated spreading decomposition or deflagration which achieved very high mass rates due to a critical degree of confinement.

Mathematical model for the self-heating of deposits found in heated bitumen storage tanks

Abstract: A mathematical model has been developed to simulate the conditions occurring within the deposit layer found on the underside of the roof of a heated bitumen storage tank. The deposit was found to undergo low temperature smouldering, reaching maximum temperatures of around 650–750 K, due to the fact that the reaction took place at the surface of the deposit and any heat generated was easily lost to the surroundings. The rate of reaction was found to be dependent on the rate of diffusion of oxygen into the surface of the deposit. The model can be used to identify potentially hazardous deposits and to determine a maximum safe thickness from which the necessary frequency of removal of deposits can be estimated to ensure a safe system.

Design of long safety valve vent line systems for gas relief

Abstract: The proposed fluid dynamic design methodology for long safety valve vent line systems with gases at high velocities comprises two steps: the procedure for calculating the pressure loss in the inlet line and that for the computation of the pressure drop in the outlet line to obtain the back pressure in the outlet of the safety valve. The hole design procedure is based on a one-dimensional approach for a gas flow assumed to be at steady-state at the instant where the pressure in the equipment to be protected is at its highest. It considers multiple chokes of flow, which often occur in long safety valve vent lines with a high driving pressure difference. Recommendations are given to simplify the calculation routine and the results of a design calculation for a vent line system of a real production plant are shown.

Industrial fuel ignition conditions by flame radiation

Abstract: With the aim of developing and testing the model of liquid fuel ignition by flame heat flux a simplified approach is presented to evaluate the ignition conditions on the basis only of well-known fuel characteristics. The case of a sufficient size of heat source which enables boundary effects to be ignored and the case of local ignition are considered. In accordance with the presented model, critical parameters of ignition (heat flux power, exposure time, heat source size) of typical heavy fuels are determined. Direct experimental validation for the model is presented. Experimentally-determined critical parameters for diesel fuel are in a good agreement with calculated results.

Vent line void fractions and mass flow rates during top venting of high viscosity fluids

Abstract: Experimental data for the initial mass flow rate and vent line void fraction during top venting are presented. The fluids used were water, foamy water and fluids with viscosities up to 750 Pa s (4190 times that of water at 150°C). The results show that the vent line void fraction is much higher than the maximum void fraction in the vessel.

Elliptic integral solutions for fluid discharge rates from punctured horizontal cylindrical vessels

Abstract: The problem of determining fluid discharge rates from the side (as opposed to the bottom) of punctured cylindrical and spherical vessels was recently addressed by Crowl1. Both vertical and horizontal cylindrical configurations were examined. Analytical solutions to the spherical and vertical cylindrical problems were obtained in rather straightforward fashion. However, the geometry is considerably more complex in the case of the horizontal cylindrical configuration. Crowl does not present an analytical solution to this more complicated problem, but employs numerical methods in order to integrate the governing differential material balance equation. The purpose of this communication is to present a direct analytical solution to the problem of side drainage from a punctured horizontal cylindrical vessel with the use of elliptic integrals, values of which are tabulated in numerous mathematical handbooks2–4. Because of their inherently greater accuracy, analytical solutions can serve as standards or benchmarks against which to check the reliability of numerical solutions and, in addition, are generally readily programmable on computers and/or spreadsheets.