CHAPTER 4 THERMAL HAZARD ASSESSMENT OF FIREWORKS MIXTURE USING ACCELERATING RATE CALORIMETER (ARC)

Transcription

1 68 CHAPTER 4 THERMAL HAZARD ASSESSMENT OF FIREWORKS MIXTURE USING ACCELERATING RATE CALORIMETER (ARC) 4.1 INTRODUCTION Accelerating Rate Calorimeter (ARC) is one of the versatile experimental tools available to study the self-propagating and thermally sensitive reactions of fireworks mixtures. It operates in an adiabatic condition. By using thermal techniques such as DSC and ARC in complement, it is possible to address thermal hazard potentials of fireworks mixtures (Lightfoot et al 2001, Brown and Gallagher 2003). Although there are many reported studies on impact and friction sensitivity of fireworks mixtures, studies of the thermal stability of fireworks mixtures and their thermo kinetics are few. In the present study, thermal hazards of fireworks mixtures were studied using ARC to understand the thermal behavior of fireworks mixture. ARC studies on fireworks mixtures were used to determine ceiling temperatures and pressure limits for safe operation, storage and transportation.

2 69 In this chapter, the thermal hazards of three cracker mixtures viz. atom bomb cracker, Chinese cracker and palm leaf cracker and two tip mixtures viz. flower pot tip mixture and ground spinner tip mixture by adiabatic tests in ARC are studied. Further the water induced thermal hazards of tip mixtures are also investigated by iso-ageing tests in ARC to deduce the field conditions for triggering accidents. 4.2 MATERIALS AND METHOD Accelerating Rate Calorimeter (ARC) Accelerating Rate Calorimeter (ARC) has gained importance since the 1980 s for studying the self heating reactions that cause thermal runaway. It was employed (Aurbach et al 2003, Bunyan et al 1999, Carr 1983, Fenlon 1984, Fisher 1991, Iizuka and Surianarayanan 2003, Lee et al 2002, Liao et al 2011, Ottaway 1986, Roduit et al 2008, Smitha et al 2013, Sivapirakasam et al 2006c, Surianarayanan et al 2001) for studying the runaway characteristics of chemical reactions. ARC used in this study was an esarc supplied by Thermal Hazard Technology, UK ARC Construction Figure 4.1 illustrates the calorimeter part of ARC. It is a container with its contents maintained at adiabatic conditions with respect to its environment. This is accomplished by constant monitoring of its temperature and suitably adjusting the surrounding temperature to minimize the heat gains (or) losses from the container. In order to achieve an adiabatic environment over a temperature range of ambient to 425 C, the ARC is equipped with a sophisticated digital control for the heater system.

3 70 The calorimeter can be divided into three temperature control zones viz. Top, middle and bottom with each of them equipped with its own control instrumentation. The sample container or bomb is attached to a pressure transducer on the top of the chamber for close monitoring of pressure responses. The radiant heater located at the bottom of the adiabatic chamber, is meant for heating the sample container at the start of the experiment. Shield Insulation Pressure Transducer Control Thermocouple Control Heater Bomb Radiant Heater Sample Thermocouple Figure 4.1 Accelerating Rate Calorimeter chamber setup ARC Principle The working principle, design description, and operational details of ARC were well cited in literature (Townsend and Tou 1980). ARC measurements were made using a sample bomb, i.e. a metal sphere in a 2.5cm diameter, typically made of Titanium. The sample mass usually (1-2 g) would depend upon the expected energy release and type of sample container

4 71 (known as a bomb) used. The sample bomb was attached to the lid section on the calorimeter assembly by a Swagelok pressure fitting and a pressure line that led to the pressure transducer. A thermocouple was attached to the outer surface of the bomb and the lid of the calorimeter positioned on the base section. The calorimeter has three separate thermal zones. The top (lid section) contains two heaters and a thermocouple, the side zone of the base section has four heaters and a thermocouple and the bottom zone at the base section has two heaters and a thermocouple. After set up and connection, the calorimeter was sealed with an explosion proof containment vessel. After defining experimental conditions on the PC, the test commenced. Two types of tests can be performed in ARC viz., heat-wait-search and iso-ageing method. ARC experiments were performed both under heat-wait-search and iso-ageing methods. The heat-wait-search method (Figure 4.2) is briefly discussed below. The test conditions were a start and end temperature and choosing the size of heat steps', 'wait time' and detection sensitivity'. The system will heat to the start temperature. A small heater in the calorimeter, the radiant heater was used to heat the sample, bomb and its thermocouple. The calorimeter was cooled and the temperature difference observed by the three calorimeter thermocouples. The system then applied power to the calorimeter heaters to minimize the temperature difference. This was to continue as the temperature rose to the start temperature. When this start temperature was reached the system would go into a wait period and during this time no heat was provided by the radiant heater. This allowed the temperature differences within the calorimeter to be reduced to zero. The calorimeter operates adiabatically allowing the calorimeter temperature to track sample

5 72 temperature. These wait periods (typically minutes) was followed by a search or seek period. Again during this period (typically 20 minutes) no heat was provided by the radiant heater, and any temperature drift, upwards or downwards, was observed. If there was upward temperature drift it was caused by a self-heating reaction. The heat-wait-seek procedure, the normal mode of operation of the ARC was to continue until an upward temperature drift observed an exothermic reaction, greater than the selected sensitivity (normally C min -1 ). The system automatically switched to the exothermic mode; it would apply heat to the calorimeter jacket to keep its temperature the same as the bomb/sample. The adiabatic control is the key feature of the ARC. The system continues in the exotherm mode until the rate of self heating is less than the chosen sensitivity and at this stage the heat-wait-seek procedure resumes. When the end temperature is reached (or an end pressure is reached) the test automatically stops and cooling, by compressed air, begins. The aim of the ARC is to complete the test to get a full time, temperature, and pressure profile of the exothermic reaction in a safe and controlled manner. In iso-ageing method the sample is screened for exothermic activity at a specified temperature, on initiation of exothermic activity, the self heating was followed under adiabatic control mode. The difference between iso-ageing method and heat-wait-search method is that, in the iso-ageing method the sample is subjected to remain at a pre determined temperature. On exothermic onset ARC follows the exothermic reaction adiabatically as in heat-wait-search method.

6 73 Figure 4.2 Heat-wait-search operation of ARC Following data plots can be obtained from an ARC experiment. Self heat rate vs. temperature This plot provides information on the onset temperature of the exothermic activity and qualitative indication of the rate of energy liberation. Temperature vs. time It provides information on the vigour of the exothermic reaction and also the available time span from the onset of exothermic activity to the end of the reaction. Pressure vs. temperature Information on the rate of pressure and temperature rise will be most useful for estimating the vent area required for the safe operation of a reaction mixture.

7 Adiabatic Thermo Kinetics The first assumption in the interpretation of ARC experimental data is the representation of concentration in terms of temperature differences (Townsend (1977 and 1991), Townsend and Tou (1980)). The equivalence of temperature and concentration for a simple well-defined chemical reaction is established using the ratio (Bunyan et al 1999, Davis et al 1991, Fisher 1991, Kossoy et al 1994, Lee and Back 1986, Nalla et al 2012, Ottaway 1986, Smitha et al 2013, Wedlich and Davis 1990,): = = (4.1) Here C is the concentration of the reacting substance, and T the temperature. The subscript 0 indicates some initial condition, and F a final state in which the substance has been consumed. Then T = T F T 0 is the temperature rise for the reaction. It is also equal to the ratio of enthalpy to average specific heat. As the reaction proceeds, the disappearance of the reacting species produces a proportional increase in the heat energy. The heat of reaction, H r, can be calculated from H = mc T (4.2) where C P, is the average heat capacity, m the mass of the sample. The heat generated in an exothermic reaction is used to heat the material, the container or bomb, and the surroundings. The heat being used up in heating the sample mass depends on specific heat. The proportion of heat used in heating the container is called thermal inertia ( ), expressed as Heat capacity of sample(s)and container or bomb(b) = (or) Heat capacity of sample

8 75 = m C + m C m C (or) (4.3) = 1 + m C m C (4.4) Incorporating the effects of thermal inertia ( ), the absolute temperature rise is given by T ab = T (4.5) The corrected heat of reaction, Hr, is calculated using Hr = mc T (4.6) The question that is basic to the study of relationship of time to explosion is the measurement, and extrapolation of data. Extrapolation must involve a concept of concentration since no material can continue to self-heat forever. The time dependence of concentration for an nth order reaction rate is expressed as follows dc = kc (4.7) dt Where C is the concentration, k the rate coefficient, and t the time. When Equations (4.1) and (4.7) are used, additional temperature dependence appears. dc dt = C d (T T) = C dt T T dt dt (4.8) m = dt dt = k(t T T ). C T (4.9)

9 76 Here m T, is defined as the rate of temperature increase (or slope of the graph of T vs. t), i.e. the self-heat rate. To remove this extra temperature dependence, a modified rate is defined as the pseudo rate constant, k*. k = k. C = m T. ( T T T ) (4.10) It is defined in such a way that its dimensions for any order reaction are reciprocal of time. In practice, k* is evaluated from experimental data using the right hand side expression. With the proper choice of N, k* has the same temperature dependence as k, and yields a straight-line graph. The Arrhenius relationship for determining the rate coefficients is k = Ae (4.11) where T is the absolute temperature in Kelvin, E the activation energy, R the universal gas constant, and A is the pre-exponential factor First order model (i.e. N=1) kinetics was assumed for the decomposition of fireworks tip mixture. For N =1 Equation (4.10) becomes k = k = m (T T) (4.12) Pseudo rate constants (k*) were calculated using Equation (4.12). Then E/R, and pre-exponential factor were found out by plotting ln k* versus inverse of temperature. The slope of the plot was equal to the intercept gave the pre-exponential factor, A. E/R, and

10 RESULTS AND DISCUSSION ARC Studies for Cracker Mixtures Under adiabatic conditions the atom bomb cracker mixture decomposed slowly (Figure 4.3) until 1750 min (285 C) and beyond this, the temperature rise is sudden and sharp until the end of the exothermic activity. The entire activity is recorded within a time span of 300 minutes. The sharp and sudden rise in temperature shows the vulnerability of this mixture to undergo violent decomposition. The discontinuity in both the self heat rate plot and time versus temperature plot is indicative of multiple exothermic activities Time, (min) Figure 4.3 Temperature rise profile for different types of crackers (Atom bomb cracker ( ), Chinese cracker ( ), Palm leaf cracker ( ))

11 78 The self heat rate plot for thermal explosive decomposition of atom bomb cracker consisting of KNO 3, S and Al in the ratio of 60:20:20 is shown in Figure 4.4 and the data summarized in Table 4.1. The onset for thermal explosive decomposition was observed at 275 C and extended up to 340 C. The self heat rate plot shows a maximum heat release rate of C min -1 at 320 C. The observed large heat release rate confirms the vigour of exothermic explosive process of atom bomb cracker mixture Temperature, ( C) Figure 4.4 Self heat rate profile for different types of crackers (Atom bomb cracker ( ), Chinese cracker ( ), Palm leaf cracker ( )) The exothermic activity is accompanied by a considerable quantity release of gaseous components as shown in Figure 4.5. It is interesting to note that one gram of sample could contribute to a peak pressure rise of 25.9 bar at C. The adiabatic temperature rise for this process was 66.7 C.The heat of reaction for the exothermic activity was calculated as Jg -1. The ARC data showed that the fireworks mixture decomposition process under adiabatic condition was vigourous and therefore dangerous.

12 79 Although explosive potential was required for it to be a cracker, the same property could be dangerous if the event occurred during handling, manufacturing, storage or transport. Evidence is available (Chapter 6) that the onset temperature attainment could be possible due to mechanical stimulus Temperature, ( C) Figure 4.5 Pressure rise profile for different types of crackers (Atom bomb cracker ( ), Chinese cracker ( ), Palm leaf cracker ( )) The ARC results for Chinese cracker and palm-leaf crackers are shown in Figures Both Chinese and palm-leaf crackers show delayed onset temperatures of 295 C and 290 C respectively as compared to the atom bomb cracker mixture. The delay in initiation of the exothermic activity can be related to their mixture, especially the quantity of the oxidiser KNO 3. While the percentage of KNO 3 in palm-leaf is more or less equal to that of Chinese cracker, early onset is perhaps due to the presence of another oxidiser Ba(NO 3 ) 2 and a large quantity of different grades of aluminium. Palm-leaf

13 80 cracker also shows multiple exothermic activities extended up to 420 C. Both these mixtures liberate peak heat rates more than 100 C min around 320 C. The time versus temperature plot (Figure 4.3) shows that the exothermic activity is sudden and sharp as observed with atom bomb cracker mixture. Although the heat rates of the second exothermic activity of palm-leaf mixture were within 1 C min -1, the decomposition process raises the system temperature suddenly to 410 C from 345 C. The decomposition process contributes to a system pressure rise up to 15 bar. The ARC data showed that the Chinese cracker and palm-leaf cracker decompositions under adiabatic conditions were vigourous and therefore dangerous Thermo Kinetics of Cracker Mixtures Pseudo rate constants (k*) were calculated using Equation (4.12). Then ln k* versus inverse of temperature was plotted. The straight line obtained confirms the assumption that the pyrotechnic mixture follows first order kinetics. The slope of the plot is equal to E/R. The activation energy was calculated as kj mol -1, kj mol -1 and kj mol -1 for atom bomb cracker, Chinese cracker and palm leaf cracker respectively. The pre-exponential factor was evaluated as 1.38X10 30, 8.35X10 33, and 1.05X10 44 min -1. Thus the Arrhenius rate law for thermal decomposition of atom bomb cracker, Chinese cracker and palm leaf cracker mixture can be given as Equations (4.13), (4.14), and (4.15) respectively. k = 1.38X10 30 exp ( /RT) (4.13) k = 8.35X10 33 exp ( /RT) (4.14) k = 1.05X10 44 exp ( /RT) (4.15)

14 81 The heat rates determined were compared with the experimentally obtained heat rates in Figure 4.6 and they did show a good agreement between the two Temperature, ( C) Figure 4.6 Self heat rate profile experimental and predicted for different types crackers (Atom bomb cracker ( ), Chinese cracker ( ), Palm leaf cracker ( ), Predicted ( )

15 82

16 ARC Studies of Ground Spinner Tip and Flowerpot Tip Mixtures The ARC results for ground spinner tip mixture are presented in Figures Compared to DSC results, adiabatic calorimeter showed interesting results for ground spinner tip mixture and ground spinner tip mixtures with varying water content. The ground spinner tip mixture without added water is marked as 0% in Figures The time temperature curve (Figure 4.7) shows two discontinuous exothermic activity at C and C. The first self heating rate process proceeds slowly compared to the second event Time, (min) Figure 4.7 Time vs. Temperature plot for ground spinner tip mixture and ground spinner tip mixture with different % of water content (0% ( ), 10% ( ), 20% ( ), 30%( ), 40%( ))

17 Temperature, ( C) Figure 4.8 Temperature vs. Self heat rate plot for ground spinner tip mixture and ground spinner tip mixture with different % of water content (0% ( ), 10% ( ), 20% ( ), 30%( ), 40%( )) The self heating process for ground spinner tip mixture started at 220 C and peaked at 320 C with a heating rate of 100 C min -1. The self heating process ended at this point perhaps due to complete depletion of the sample. The sudden jump in self heat rate from C and finally 100 C min -1 (in a discontinuous manner) indicated the explosive nature of the process. As the water content increased the exothermic onset point decreased. The mixture with 40% water content showed the lowest onset point of 80 C. The time required for exothermic initiations under dynamic heat-wait-seek pattern was 300 minutes. Once the exothermic initiation occurred, in less than 100 minutes, the temperature rose to 300 C. Although the onset temperature (120 C) for the mixture with 30% water content was higher, the behavior of the exothermic event was similar to that of the mixture with 40% water content, but with a higher end point i.e. 340 C. The mixtures with no water

18 85 and 10-20% water (w/w) content underwent multiple exothermic events in the temperature range of 140 C-380 C. The self heat rate pattern observed for fireworks tip mixture presented in Figure 4.8 showed the vigour of exothermic behavior of these samples. The neat sample showed a maximum self heat rate of 100 C min -1 at 320 C. The sample with 40% water (w/w) content released a maximum heat of 40 C min -1 at 300 C. The maximum heat releases for mixture containing 30, 20, 10% water (w/w) content were within 15 C min Temperature, ( C) Figure 4.9 Temperature vs. Pressure plot for ground spinner tip mixture and ground spinner tip mixture with different % of water content (0% ( ), 10% ( ), 20% ( ), 30%( ), 40%( )) The pressure release profiles for mixtures are presented in Figure 4.9. The self heating exothermic process for 40 and 30% water (w/w) containing samples accompanied a maximum pressure release of 65, and 75

19 86 bar respectively. The samples with less than 30% water (w/w) content showed pressure release of less than 20 bar, though that was comparatively less than these of samples with 30 and 40% water (w/w) content. The ARC results of flowerpot tip are presented in Figures Compared to DSC results, adiabatic calorimeter showed interesting results for flowerpot tip mixture and mixture with 40% water content. The neat sample showed an onset of C but the sample with 40% water content showed the lowest onset point of C. The time required for exothermic initiations under dynamic heat-wait-seek pattern was 475 minutes. Once the exothermic initiation occurred, in less than 100 minutes, the temperature rose to 210 C. Both the samples underwent multiple exotherm at two different temperatures as shown in Figure Time, (min) Figure 4.10 Time vs. Temperature plot for flowerpot tip mixture and flowerpot tip mixture with 40% of water content (Flowerpot tip ( ), Flowerpot tip with 40% water ( ))

20 87 The self heat rate pattern observed for flowerpot tip mixture presented in Figure 4.11 showed the vigour of exothermic behavior of these samples. The neat sample showed a maximum self heat rate of C min -1 at 339 C. The sample with 40% water (w/w) content released a maximum heat of 4.19 C min -1 at C Temperature, ( C) Figure 4.11 Temperature vs. Self heat rate plot for flowerpot tip mixture and flowerpot tip mixture with 40% of water content (Flowerpot tip ( ), Flowerpot tip with 40% water ( )) The pressure release profiles for mixtures are presented in Figure The self heating exothermic process for neat sample and sample with 40% water (w/w) containing samples accompanied a maximum pressure release of 32, and 20 bar respectively.

21 Temperature, ( C) Figure 4.12 Temperature vs. Pressure plot for flowerpot tip mixture and flowerpot tip mixture with 40% of water content (Flowerpot tip ( ), Flowerpot tip with 40% water ( )) Thermo Kinetics of Ground Spinner Tip Mixtures Thermo kinetics for the thermal behavior of ground spinner tip for dynamic heat-wait-search, and iso-ageing ARC tests were determined. The thermo kinetic data are presented in Tables 4.2 and 4.4 respectively. The reliability of Arrhenius kinetics has been checked by predicting self heat rate by employing Arrhenius equation. The activation energy was calculated for fireworks tip with 0%, 10%, 20%, 30%, and 40% water (w/w) are 81.04, , , , and kj mol -1 respectively. The pre-exponential factor was evaluated as 6.78 x 10 20, 3.57 x 10 23, 1.47 x 10 13, 6.68 x 10 21, and 9.17 x min -1. Thus the Arrhenius rate law for thermal decomposition of ground spinner tip for different quantities of water (w/w) 0%, 10%, 20%, 30%, and 40% can be given as Equations (4.16) to (4.20) respectively.

22 89 k = 6.78 x exp (-81.04/RT) (4.16) k = 3.57 x exp ( /RT) (4.17) k = 1.47 x exp ( /RT) (4.18) k = 6.68 x exp ( /RT) (4.19) k = 9.17 x exp ( /RT) (4.20) The heat rates determined were compared with the experimentally obtained heat rates in Figure 4.13 and they did show a good agreement between the two Temperature, ( C) Figure 4.13 Experimental and predicted self heat rates of thermal decomposition of ground spinner tip mixture with different % of water (0% ( ), 10% ( ), 20% ( ), 30%( ), 40%( ), Predicted( ))

23 90

24 Thermo Kinetics of Flowerpot Tip Mixtures Thermo kinetics for the thermal behavior of flowerpot tip for dynamic heat-wait-search, and iso-ageing ARC tests were determined. The thermo kinetic data are presented in Table 4.3 and 4.5. The reliability of Arrhenius kinetics has been checked by predicting self heat rate. The activation energy was calculated for flowerpot tip mixture and flowerpot tip mixture with 40% water (w/w) are and kj mol -1 respectively. The pre-exponential factor was evaluated as 3.11 x and 2.35 x min -1 respectively. Thus the Arrhenius rate law for thermal decomposition of flowerpot tip mixture and mixture with 40% water (w/w) can be given as Equations (4.21) and (4.22) respectively. k = 3.11 x exp ( / RT) (4.21) k = 2.35 x exp ( / RT) (4.22) The heat rates determined were compared with the experimentally obtained heat rates in Figure 4.14 and they did a good agreement between the two. For kinetic treatment, the initial portion of exotherm alone is sufficient; since the interest lies in knowing the Arrhenius constants for the onset of exothermic activity and not for the entire exothermic region. Furthermore, as the reaction progresses, complexity may arise due to onset of secondary decompositions.

25 Temperature, ( C) Figure 4.14 Experimental and predicted self heat rate plot for flowerpot tip mixture (Flowerpot tip ( ), Flowerpot tip with 40% water ( ), Predicted ( ))

26 93

27 ARC Iso-Ageing Studies for Ground Spinner Tip Mixture The ARC outputs for ground spinner tip mixture containing 40% water (w/w) were subjected to iso-ageing test in ARC in the temperature range of C (Figures ). The temperature range chosen here was based on the dynamic ARC results presented in the previous section and involvement of these mixtures in frequent accidents in the industry Time, (min) Figure 4.15 Time vs. Temperature plot for ground spinner tip mixture at various iso-ageing temperatures (40 C ( ), 47.5 C ( ), 50 C ( ), 55 C ( ), 60 C ( ), 65 C ( ), 75 C ( )) A simple view of the time versus temperature profile shown in Figure 4.15 at various iso-ageing temperatures confirmed the vigourous thermal behavior of the mixtures. Irrespective of the iso-ageing (thermal history) conditions all the samples show a single sharp rise to a final temperature within a short duration from the onset point. The onset point for

28 95 the thermal event was dependent on the initial iso-ageing temperature. Lowest onset of 40 C was recorded for the mixtures. The time needed to initiate the event depended on the initial iso-ageing temperature. At the lower iso-ageing temperature (40 C), longer time was taken to initiate the event. Thus, this study, perhaps for the first time, could successfully correlate the accident occurring conditions. More or less close to ambient conditions considered here exist in the southern districts of Tamilnadu, India, where a major firework manufacturing is carried on Temperature, ( C) Figure 4.16 Self heat rate plot for ground spinner tip mixture at various iso-ageing temperatures (40 C ( ), 47.5 C ( ), 50 C ( ), 55 C ( ), 60 C ( ), 65 C ( ), 75 C ( )) The self heat rate plots presented in Figure 4.16 depict the vigour of the exothermic event. Interestingly the experimental runs at lower iso-ageing temperatures (although too long to initiate the event) showed maximum self heat release rates i.e. 40 C is 37 C min -1, 47.5 C is 30 C min -1 and 50 C is

29 96 25 C min -1. The self heat rate release curves also showed multiple steps in their pattern; the first trend from onset to maximum heat release rate followed by a drop at the temperature range C, the second slower exothermic activity picking up from C to reach a final temperature range of C. This behavior was a clear indication of operation of two different reaction mechanisms for the exothermic events. The heat rates determined were compared with the experimentally obtained heat rates in Figure 4.17 and they did show a good agreement between the two Temperature, ( C) Figure 4.17 Experimental and predicted self heat rates plot for ground spinner tip mixture at various iso-ageing temperatures (40 C ( ), 47.5 C ( ), 50 C ( ), 55 C ( ), 60 C ( ), 65 C ( ), 75 C ( ), Predicted ( ))

30 97 Figure 4.18 confirmed the pressure release during the exothermic events. In fireworks industry, the unused water mixed with ground spinner tip mixture was left overnight at the end of the day for next day use. Explosive accidents were reported after shutting down the factory. The iso-ageing studies could be correlated to blast accidents in fireworks industry involving ground spinner tip mixture Temperature, ( C) Figure 4.18 Temperature vs. Pressure plot for ground spinner tip mixture at various iso-ageing temperatures (40 C ( ), 47.5 C ( ), 50 C ( ), 55 C ( ), 60 C ( ), 65 C ( ), 75 C ( ))

31 98

32 ARC Iso-Ageing Studies for Flowerpot Tip Mixture The ARC outputs for flowerpot tip mixture containing 40% water (w/w) were subjected to iso-ageing in the temperature range of 40 C (Figures ). A simple view of the time versus temperature profile shown in Figure 4.19 at iso-ageing temperature confirmed the vigourous thermal behavior of the mixtures Time, (min) Figure 4.19 Time vs. Temperature plot for flowerpot tip mixture with 40% water at 40 C iso-ageing temperature The self heat rate plots presented in Figure 4.20 depict the vigour of the exothermic event. Interestingly the experimental runs at 40 C iso-ageing temperatures (although too long to initiate the event) showed maximum self heat release rates of C min -1. The self heat rate release curves also showed multiple steps in their pattern; the first trend from onset to maximum

33 100 heat release rate followed by a drop at the temperature range C, the second slower exothermic activity picking up from C to reach a final temperature range of C. This behaviour was a clear indication of operation of two different reaction mechanisms for the exothermic events. Pressure profile in Figure 4.20 confirmed the pressure release during the exothermic events Temperature, ( C) Figure 4.20 Self heat rate plot and pressure plot for flowerpot tip mixture with 40% water at 40 C iso-ageing temperature (Self heat rate ( ), Pressure ( ))

34 Temperature, ( C) Figure 4.21 Experimental and predicted self heat rate plot for flowerpot tip mixture with 40% water at 40 C iso-ageing temperature (Self heat rate ( ), Predicted ( ) The heat rates determined were compared with the experimentally obtained heat rates in Figure 4.21 and they did show a good agreement between the two. Only the initial portion of the exothermic onset is considered for prediction of the exothermic behaviour of the mixture. The ARC studies confirmed the vigour of water induced exothermic behavior of flowerpot tip mixtures.

35 102

36 PROCESS SAFETY ARC Results and their Impact on Process Safety of Cracker Mixtures Fireworks mixtures are vulnerable to thermal hazards. ARC data are used for determining the ceiling temperature for processing, handling and transportation of hazardous materials. The practice adopted is that the process/handling temperature should be 100 C below the onset temperature observed in ARC. This rule has been in practice in process chemical industry for safe and successful operation of process plants, storage systems and transportation. On these considerations, in case of atom bomb cracker, Chinese fire cracker and Palm leaf cracker the ceiling temperature should never exceed 175 C, 195 C, and 191 C respectively. Although there is no possibility of reaching this temperature during normal mixing and packing process of these mixtures, the chances of thermal explosions cannot be ruled out. This is because the onset temperature can be achieved under situations like heat radiation from neighbouring area or ignition from unknown sources, subjecting these mixtures to impact (dropping the mixture containing boxes leading to impact stimulus) and friction (dragging the mixture compositions leading to the generation of sparks as a result of friction stimulus). During such abnormal situations the cracker mixture is vulnerable to hazard. Further, impact and friction sensitivity can also lead to triggering of explosive decompositions. As of now, there is no direct correlation available between thermal, impact and frictional sensitivity, either to predict one or the other or to predict which of these forces can come together to trigger a thermal explosion. We hypothesize that impact or frictional stimulus onsets the thermal stimuli for the fireworks mixture to undergo thermal explosion. Under severe impact or friction stimuli, thermal stimuli can occur

37 104 immediately, and this can lead to a catastrophic thermal explosion. Irrespective of the nature of the stimuli, explosion occurs through thermal mechanism only. This means that, for these mixtures, the impact or any other stimulus can only initiate the thermal mechanism by providing the minimum threshold energy needed/necessary to raise the onset temperature observed experimentally for thermal explosion in ARC. Therefore it is possible to relate the mechanical form of energy with the threshold energy ( E) observed in the ARC. ARC study provides a means of suggesting a predictive correlation in such explosive systems. Nevertheless the degree of explosivity also depends on other factors such as compactness, particle size and shape and other environmental conditions ARC Results and their Impact on Process Safety of Tip Mixtures The ARC studies for fireworks tip mixture confirmed that the trigger of exothermic activity could be possible at ambient conditions. The fireworks tip mixture is involved in triggering accidents often in fireworks industry, when the remaining quantities of these mixtures are left over night. The water in the mixture slowly dries; the exothermic activity induced during drying of water appears to be responsible for accidents. Most of the fireworks industry accidents (reported in media) have occurred either late at night or early in the morning. Therefore, from this study the following precautions are given to the fireworks industry. The unutilized fireworks tip mixture should be appropriately disposed off at the end of each day. During the manufacture of fireworks, it should be ensured that the ambient temperature should be less than 30 C.

38 105 Effect of impact and friction stimulus should be analysed, as these could trigger accidents. As ARC studies presented here have shown such acts have the potential to initiate exothermic activity at ambient conditions. For the first time the reasons for accident triggering at closer to ambient conditions and the culprit mixture have been identified. 4.5 SUMMARY ARC studies of the atom bomb cracker, Chinese cracker and palm leaf cracker have revealed that all the above mixtures are susceptible to thermal decomposition. The thermal decomposition contributes to substantial rise in system pressure. The onset temperature for explosive decomposition observed in accelerating rate calorimeter is however the minimum temperature for triggering an accident. In practice this temperature can be attained by thermal and mechanical stimulus. Therefore these mixtures are to be handled carefully. The thermo kinetic studies revealed that the mixtures follow first order Arrhenius Kinetics and the kinetic data are validated by comparing the predicted self heat rates with the experimental data. The sharp and sudden rise in temperature shows the vulnerability of these mixtures to undergo violent decomposition. The discontinuity in both the self heat rate plot and time versus temperature plot is indicative of multiple exothermic activities. The self heat rate plot shows a maximum heat release rate of C min -1 at 320 C. The observed large heat release rate

39 106 confirms the vigour of exothermic explosive process of atom bomb cracker mixture. One gram of atom bomb cracker mixture can contribute to a peak pressure rise of 25.9 bar at C. The ARC data showed that the fireworks mixture decomposition process under adiabatic condition was vigourous and therefore dangerous. Although explosive potential is required for it to be a cracker, the same property can be dangerous if the event occurs during handling, manufacturing, storage or transport. Both Chinese and palm-leaf crackers show delayed onset temperatures of 295 C and 290 C respectively as compared to the atom bomb cracker mixture. The delay in initiation of the exothermic activity can be related to their composition, especially the quantity of the oxidiser KNO 3. While the percentage of KNO 3 in palm-leaf is less than that of Chinese cracker, early onset is perhaps due to the presence of another oxidiser Ba(NO 3 ) 2 and a large quantity of different grades of aluminium The time versus temperature plot shows that the exothermic activity is sudden and sharp as observed with atom bomb cracker mixture. ARC studies of the fireworks tip mixture have revealed that the above mixture is susceptible to thermal decompositions at ambient conditions. The thermal decomposition contributes to substantial rise in system pressure.

40 107 The onset temperature for explosive decomposition observed in accelerating rate calorimeter is however the minimum temperature for triggering an accident. In practice this temperature can be attained by thermal and mechanical stimulus. Therefore, this mixture is to be handled carefully. The thermo kinetic studies revealed that the mixture follows first order Arrhenius kinetics and the kinetic data are validated by comparing the predicted self heat rates with the experimental data. A second self heating process for the ground spinner tip mixture started at 220 C and peaked at 320 C with a heating rate of 100 C min -1 ; the self heating process ended at this point perhaps due to complete depletion of the samples. The sudden jump in self heat rate from 10 to 50 C min -1 and finally 100 C min -1 (in a discontinuous manner) indicated the explosive nature of the process. The time vs. temperature plot of ground spinner tip shows two discontinuous exothermic activities C and C. The first self heating rate process proceeds slowly compared to the second event. Thus the self heating nature of the ground spinner tip mixture and its reactive thermal hazards with water necessitated the study of the behaviour of these samples under iso-ageing conditions. The neat mixture showed a maximum self heat rate of C min -1 at 339 C. The sample with 40% water (w/w) content released a maximum heat of 4.19 C min -1 at C.

41 108 The self heating natures of the ground spinner tip mixture and its reactive thermal hazards with water under iso-ageing conditions showed a lowest onset temperature of 40 C. The self heating exothermic process for neat sample and sample with 40% water (w/w) containing samples accompanied a maximum pressure release of 32, and 20 bar respectively.