Calorimetric Approach on the Thermal Hazard Assessment of Cumene Hydroperoxide


Chemical processing industries require the use of extensively reactive chemicals for preparation, processing, sampling, storage, transportation, and even during disposal. The various perturbed conditions, such as mischarging, overdosing, or exposure to external fire at any stage, could lead to the generation of heat caused by the reactive material decomposition or polymerization (NFPA 43B, 1999). The peroxy group (-0-0) present in the organic peroxides is the reason for thermal runaways due to its intrinsic unstable and reactive nature, which is similar to cumene hydroperoxide (CHP) referred to in this chapter (Hou et al., 2000). When the rate of heat generated is more than the rate of heat removed in a reactive system, thermal runaway occurs. The peroxy group in CHP is most elusive to heat and the exothermic threshold is as low as ambient temperatures and sometimes below 120 °C. Since the 1970s, due to the highly unstable nature of CHP it has caused several fires and explosions over industries in Taiwan. In common with all other alkyl-peroxides, hydroperoxides contains a hydroxy group/acidic hydrogen, which makes it more reactive than others. Moreover, hydroperoxides are extremely sensitive towards impurities, metal ions, acids, and bases and thus hydrogen peroxide (H-,0,), CHP, and tert-butyl hydroperoxide (TBHP), are handled carefully (Shu et al., 2000 & Boundy and Boyer, 1952). For peroxide systems during a runaway reaction, solvents with a higher boiling point will have no effect on the reaction kinetics and pressure relief systems. The different concentration of 30, 60, and 81 mass% of commercial CHP is produced industrially. In the production of dicumyl peroxide, CHP is used as a catalyst and is sensitive to pH. The oxygen balance, functional groups of explosives, and incompatibility charts are used to evaluate the hazardous storage of different materials that estimate the amount of oxygen released during oxidation. In past cases, the instability property of CHP has led to a number of fire hazards explosions in various industries.

Thermal Runaway of Cumene Hydroperoxide

CHP is known to be widely used in polymerization reaction, where it acts as an initiator. It has been used extensively for synthesizing acrylonitrile-butadiene-styrene (ABS) co-polymer in various industries across Taiwan. Its usage is also known in synthesizing phenol and acetone through the catalytic cleavage reaction mechanism (Boundy and Boyer, 1952). The National Fire Protection Association (NFPA) has classified CHP to be a class III type flammable (fire hazard). The emphasis in research studies has been on understanding the pressure relief for organic peroxides by the members of the Design Institute for Emergency Relief System (DIERS) under the American Institute of Chemical Engineers (AIChE) (Wang et al., 2001). In order to design the measures for preventing such hazardous reaction by peroxides and also to lower the generated pressure through adequate vent sizing, various aspects need to be addressed. The most important perspectives such as thermal decomposition, reactive incompatibility, the external fire in process as well as the storage area should be taken into account for handling such hazards, along with other credible scenarios. Many previous works have been reported of various process conditions regarding the runaway hazards, reaction mechanisms and decomposition kinetics of CHP (Wang et al., 2001; Ho. Duh and Chen, 1998 & Duh et al., 1997).

In the production of dicumyl hydroperoxide (DCPO) and phenol, CHP has been widely used in Taiwan. During the polymerization synthesis of the ABS, it is applied as an initiator. The after-reaction process includes the release of toxic elements with runaway reactions uncontrolled in nature. Such reaction leads toward fires and explosion which further results in severe casualties and damages. In many cases, thermal runaways are unavoidable since the high calorific capacity of gaseous substances and their decomposed materials incur different hazards due to their specific physical and chemical properties. In order to mitigate the related hazard and damages incurred due to the runaway reactions it is important to analyze the hazardous chemicals with adequate instruments, such as differential scanning calorimetry (DSC) and a thermal activity monitor (TAM). Such procedures could proactively help in preventing possible disaster, or, if it happened, could reduce it to an acceptable level.

In previous researches, investigation was carried out for auto-catalytic nature utilizing DSC and TAM test at low temperature conditions, but with /;th order reaction behaviors under high temperature ranges (Duh et al., 1997). From the preventive measure point of view, during thermal runaway reactions, it was suggested that apart from the related thermokinetic parameters of the peroxides and isothermal conditions, safety alertness during manufacture and transportation should be viewed as a top priority before the design process stage. DSC and TAM testing primarily evaluates the data from thermokinetic parameters utilizing experimental thermal curves. The thermal hazards of reactive chemicals could be determined with the help of these two calorimeters, under normal or perturbed operating conditions. Utilizing such experimental results, along with data from several other sources, CHP’s thermal decomposition phenomenon, along with other contaminants, could be easily calculated using a micro amount of the sample. Although DSC and TAM utilizes a sample quantity of milligrams or grams, such obtained experimental curves could be extrapolated through the adoption of a technological approach for commercial applications.

This information could be utilized for safety information during process design and could provide guidance for any specific reactive chemical during storage and transportation.

Experimental Studies

The 80 mass% solution of CHP in cumene was purchased directly from Merck Co. The density and concentration was measured and then stored in a refrigerator at 4°C for subsequent use. The contaminants that could be easily encountered in process or storage such as H2S04, NaOH, and FeCl3,were chosen to be from 0.1 to 0.5 mass% for DSC experiments. The CHP samples of 35 mass% was diluted in cumene and 80 mass% were used in DSC and TAM experiments, respectively.

DSC (Differential Scanning Calorimeter)

DSC is the heat-flux type, w'hich could measure as well as acquire relationships between micro-heat transform and temperature difference of materials of interest. Practically speaking, it is placed with sample and reference in two small crucibles in a heat flow furnace. By the configuration, a heater around the furnace is employed to control the temperature of the crucibles (Duh et al., 1997). Typically, there is a heat flow sensor between sample and reference, which has a pair of thermocouple wires measuring the temperature difference between these two. As the sample temperature rises to the conversion point, such as the glass transition point, the fusion point, the boiling point or the heat changes of decomposition reaction, the sample absorbs the energy supplied by the heater (an endothermic reaction) or releases energy (an exothermic reaction) to indicate the temperature difference between the sample and the reference. Without this step, the balance resulting from generating or liberating heat flow could not be suitably maintained.

TAM (Thermal Activity Monitor)

Structurally, TAM possesses two parts: a 25-liter thermostat water bath, and an external water circulator. Normally, although not always, water is continuously circulated by pump upwards into the thermostated water bath and connected to the external water circulator, w'hich keeps the sample under an excellent constant temperature (or isothermal) environment (The Isothermal Calorimetric Manual for Thermometric, 1999). The ampoules containing sample and reference were inserted into a measuring cup between a pair of Peltier thermopile sensors. These sensors are in contact with a pair of metal heat sinks, where it is indicated that a pair of metal heat sinks of all cylinders are immersed into the thermostated water bath. Meanwhile, a pair of metal heat sinks carry messages to the Digital Voltmeter (DVM) of the heat difference between the sample and reference. DVM can continuously provide all messages to users by showing the related information, along with the situation of the experiments. As for data acquisition, the data, as shown in DVM, are simple heat powers from the sample, which illustrates the typical thermal curves of simple heat power (pW/g) versus time (day) from samples.


In essence, DSC and TAM are two independent calorimeters with different purposes. Although both have different approaches in terms of design principles, they can continuously detect small heat flow (pW) transform from a sample. If working suitably, both DSC and TAM could be smoothly coordinated to complement each other. Because, sample by TAM could not release all of its own calorific value under an isothermal experiment, there is typically 20% calorific value which could be detected by the follow'-up DSC scanning experiment to determine the remaining calorific value. In terms of the principles of instrumental structure for both DSC and TAM, either one has both its advantages and its disadvantages. Through a comprehensive understanding of these experimental principles and operating procedures, optimal usages or control by both DSC and TAM can be accomplished to fit a specific application, such as pressure relief designs by DIERS technology for the safer venting of runaway reactions incurred by organic peroxides in the incipient stages of a runaway reaction (Wang et al., 1999 & Leung & Fisher, 1998).

Results and Discussion

The DSC experiments were performed to determine the acquired exothermal phenomenon of CHP, associated with its catalytic decomposition reaction w'ith contaminants, such as H2S04, NaOH, and FeCl, was selected to be as contaminants. Table 14.1 and 14.2 indicates the DSC experimental data for the decomposition of 35 mass% CHP with incompatible chemicals and 93 mass% of CHP with NaOH. respectively. The heat flow vs temperature curves of 35 mass% CHP with and without contaminants were presented in Figure 14.1.

The addition of incompatible chemicals to CHP decreased the exothermic onset temperature when compared to 35 mass% CHP without contaminants increasing the exothermal heat of reaction. The results demonstrate the substantial impacts of CHP with potentially incompatible contaminants encountered during preparation.

TABLE 14.1

Experimental Data of Decomposition for CHP 35 Mass% and Contaminants Conducted by DSC Tests



Mass of CHP 35 mass% (mg)


Dosage (mg)








H,S04(1 N)






NaOH (1 N)






FeCI, (1 M)





(Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health)

TABLE 14.2

Experimental Data of Decomposition for CHP 93 Mass% and CHP 93 Mass% with NaOH 99 Mass% by DSC Tests


93 mass% CHP




93 mass% CHP + 99 mass'/! NaOH (0.5 mg) (first peak)




93 mass% CHP + 99 mass'/! NaOH (0.5 mg) (second peak)




(Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health)

Comparisons with heat power vs. temperature for CHP 35 mass% and contaminants by DSC tests (Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health)

FIGURE 14.1 Comparisons with heat power vs. temperature for CHP 35 mass% and contaminants by DSC tests (Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health).

processing, storage, transportation, or even disposal, in which runaway reactions and accompanying hazards cannot be overlooked. The isothermal TAM experimental data of CHP demonstrated the autocatalytic runaway reaction features at low temperatures, as shown in Table 14.3.

During the induction period, the runaway reaction was accelerated after reaching the autocatalytic runaway onset temperature, reducing the possibility of controlling the catastrophic fire and explosion. The thermograms of 80 mass% CHP at five different isothermal temperatures of 75, 80, 83, 88, and 90 °C are shown in Figure 14.2. The results indicated that the runaway reaction temperature of CHP was less than 75 °C, which is due to the initial endothermicity and is much lower than the exothermic onset temperature of 110 °C by VSP2 (Vent Sizing Package2) (Hou, Shu & Duh, 2001 & VSP2 Manual and Methodology, 1997). The results demonstrated that the various stages in operation of CHP could accumulate a considerable amount of heat increasing the hazards that may result in unexpected thermal runaway reaction.

TABLE 14.3

Experimental Data of Autocatalytic Reaction for CHP 80 Mass% Conducted by TAM Tests

Isothermal Temperature (°C)

Sample Mass (g)




Time (Hours)

Time (Days)




































(Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health)

Comparisons with heat power vs. time of CHP 80 mass% under different isothermal temperatures by TAM tests (Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health)

FIGURE 14.2 Comparisons with heat power vs. time of CHP 80 mass% under different isothermal temperatures by TAM tests (Obtained from Lin et al. (2007) © National Institute of Occupational Safety and Health).

Significance and Applications of CHP Derived by DSC and TAM

The use of various calorimetric techniques to determine the reactive hazardous characteristics of CHP during the various process stages could help in understanding the potential unexpected hazards from thermal runaway reactions. The significant observation drawn from both the DSC and TAM experiments of CHP in estimating the thermal characteristics due to its unstable nature and reactive structure were: l.

l. Addition of NaOH or FeCl, with CHP resulted in a higher heat of reaction with a considerable decrease in exothermic onset temperatures. The formation of OH- or Fe-,+ suggests the possible different reaction pathways for the increase in the reaction hazard.

  • 2. The contamination of CHP with incompatible substances such as acids/ bases, metal, ions and dusts during the sampling, preparation, and storage/ disposal could result in runaway reactions. Critical care should be taken to avoid contamination with reactive chemicals during each stage
  • 3. The prolonged storage of CHP at ambient temperatures results in thermal runaway owing to an autocatalytic reaction. The TAM experiments at 75,
  • 80, 83, 88 and 90 °C proved that over a period of time, CHP could initiate a thermal runaway reaction at high temperatures. Therefore, it is crucial to avoid direct heat accumulation and external fire exposure to all the CHP storage vessels and warehouses. However, CHP hazards could not be detected by the conventional adiabatic VSP2 calorimeter.
  • 4. The DSC could adequately simulate the thermal runaway reaction of CHP as the heat liberated during the thermal decomposition is less than the exothermic heat liberated by the self-heating rate of CHP. It would be very difficult for the emergency rescue team to evacuate the personnel without any knowledge of that temperature of no return (TNR).

Comparison of Thermokinetic Parameters for CHP Derived From DSC and TAM

The kinetic models could simulate, and predict, the degree of hazards, physical property changes, and reactions, including oxidation, crystallization, polymerization, and decomposition, involved in a chemical process. It could also evaluate the potential reaction hazards of materials during various stages of operation. The fire and explosion from CHP due to quick pressure rise during thermal decomposition (by itself/ contaminants) has resulted in fatalities, injuries and property loss. Table 14.4 represents the thermokinetic parameters in both isothermal and non-isothermal conditions of TAM and DSC. The experimental results were adequate to depict the real-time conditions during processing, transportation or storage. In addition, the application of DSC and TAM is expected to predict the CHP runaway reaction proactively during the early stages of process development.


To evaluate the degree of hazards of reactive materials, one normally cannot be only dependent on a single instrumental analysis. Therefore, assessment of the related thermal hazard using different technologies will play a crucial role in chemical

TABLE 14.4

Comparisons on Thermokinetic Parameters for CHP 80 Mass% Derived from DSC and TAM Tests






9.26 x 10”

4.5 x 1(H





9.064 x Iff'

1.618 x 10-*

(Obtained from Lin et at. (2007) © National Institute of Occupational Safety and Health) processing industries. In theory, physicochemical properties of hazardous materials such as thermal stability, sensitivity, amount of test material, range of operating temperature and materials of test cell could be analyzed by calorimeters. The thermal hazard nature of CHP, along with its in incompatibilities, could be evidently performed by DSC and TAM. The results indicated that DSC and TAM can be used appropriately based on the advantages, disadvantages and applications corresponding to specific conditions with this new approach. The thermokinetics and potential hazards of reactive chemicals could be readily understood based on the accurate data provided by these calorimeters for the process industries that demonstrate the hazardous behavior during handling and usage that could effectively lessen the probabilities and consequences of hazards generated during perturbed conditions among all the process stages. Through adequate testing, the results could be employed to alleviate any chemical disasters by apprehending the inherent thermal hazard.


Boundy, R. H.. Boyer, R. F. 1952. Styrene, Its Polymers, Copolymers, and Derivative. Rinehold. New York.

Duh. Y.S., Kao. C.S.. Hwang, H.H., Lee. W.L., 1998. Thermal Decomposition Kinetics of Cumene Hydroperoxide. Trans IChemE. 76 (Part B), 272-276.

Duh. Y.S.. Kao. C.S., Lee, C., Yu, S.W.. 1997. Runaway Hazard Assessment of Cumene Hydroperoxide from the Cumene Oxidation Process. Proc. Safety Environ. Protec. 75(2), 73-80

Но, T.C., Duh, Y.S.. Chen, J.R.. 1998. Case Studies of Incidents in Runaway Reactions and Emergency Relief. Proc. Safety Prog. 17. 259.

Hou. H.Y., Shu, C.M.. Duh.Y.S., 2001. Exothermic Decomposition of Cumene Hydroperoxide at Low Temperature Conditions. AlChE J 47, 1893-1896.

Hou, H.Y., Shu, C.M., Duh. Y.S.. Yeh, P.Y., Peng, D.J., 2000. North American Thermal Analysis Society (NATAS). Ottawa. Canada.

Leung, J.C., Fisher. H.G., 1998. Runaway Reaction Characterization: A Round-Robin Study on Three Additional Systems. Process Integration on Runaway Reactions. Pressure Relief Design, and Effluent Handling, New Orleans, LA, USA 1998; 109.

Lin, S.-H., Li. J.-C.. Chang. C.-М.. Jr Peng, D.. Shu, C.M.. 2007. Basic Thermal Hazard Assessment on Cumene Hydroperoxide by Calorimetric Approaches. J. Occup. Safety Heal. 15,297-307.

Lin, W.H., Shu, C.M., 2005. Reactive Hazards of Cumene Hydroperoxide Incompatible with Sodium Hydroxide. National Yunlin University of Science and Technology, 5-22.

NFPA 43B, 1999. Code for the Storage of Organic Peroxide Formulations, National Fire Protection Association, Quincy, MA. USA.

Shu, C.M., Hou, H.Y., Peng. D.J., Duh, Y.S., 2000. The 2nd International Conference of EDUC, Ludwigs, Germany.

The Isothermal Calorimetric Manual for Thermometric AB. Jarfalla, Sweden; 1999. Merck & CO., NJ, 2006.

VSP2 Manual and Methodology, 1997 Fauske and Associates, Inc., Burr Ridge, IL, USA. Wang. Y.W.. Shu, C.M.. Duh, Y.S.. Kao. C.S., 1999. Incompatibilities on Thermal Runaway Hazards of Cumene Hydroperoxide CHP. Methodology of Reaction Hazards Investigation and vent sizing, St Petersburg, Russia; 1999: 1-15.

Wang. Y.W., Shu. C.M.. Duh.Y.S., Kao, C. S.. 2001. Thermal Runaway Hazards of Cumene Hydroperoxide with Contaminants. Ind Eng Client Res.. 40(4), 1125-1132.

"| IT Evaluation of the

< Prev   CONTENTS   Source   Next >