Consider Chemical Reactivity in Process Hazard Analysis

Managing chemical reactivity hazards is more complex than other major hazards. Learn how to use process hazard analysis (PHA) to identify such hazard scenarios and determine whether additional risk-reduction measures are needed for your process.

Many chemicals used in the chemical process industries (CPI) present chemical reactivity hazards. Such materials can undergo uncontrollable or unintended chemical reactions that could cause serious harm to people, property, and the environment. Addressing chemical reactivity hazards is more challenging than dealing with toxic, flammable, or explosive chemicals, because reactivity hazards are not as widely recognized or understood; and because hazard scenarios involving them are more involved and harder to identify.

This article discusses the different types of chemical reactivity hazards and how process hazard analysis (PHA) can be used to identify these hazards. The article explains how PHA for chemical reactivity hazards differs from the use of PHA to address other types of hazards. Finally, it identifies common pitfalls in performing chemical reactivity PHAs and offers suggestions for avoiding them.

Chemical reactivity: The basics

Chemical reactivity is the tendency of a chemical to react with other chemicals and materials (Figure 1). Unlike other hazards posed by chemicals, chemical reactivity is not solely an intrinsic property of the chemicals involved - it also depends on the conditions under which the chemicals are used. In the realm of process safety, the term chemical reactivity often includes chemical instabil- ity - i.e., the self-reactivity of a chemical or its tendency to change chemical form. Reactivity incidents occur when control of intended chemical reactions is lost or when unintended chemical reactions take place.

Chemical reactivity hazards can be categorized as follows:

Self-reacting or unstable chemicals. A single unstable compound can undergo uncontrolled decomposition, rearrangement, or polymerization. For example, organic peroxides can pose fire and explosion hazards.

Runaway reactions. Intended reactions of chemicals to produce desired products can become uncontrollable. For example, the reaction of phenol with formaldehyde to produce phenolic resins is subject to runaway.

Incompatibilities. Inadvertent mixing of two or more process chemicals may lead to an unintended chemical reaction. For example, mixing acids with cyanide salts generates highly toxic hydrogen cyanide gas, and mixing silver salts with ammonia in the presence of a strong base may generate an explosively unstable solid. Also, process chemicals may react with materials present in the process such as water, air, materials of construction, lubricating oils, utility fluids, etc. For example, acetylene reacts with copper to form explosive compounds, and phosgene reacts corrosively with stainless steel, which may result in loss of containment.

Most chemical reactions liberate energy, and many reactivity incidents involve the release of energy in quantities or at rates too high to be absorbed by the immediate environment. Potential consequences of reactivity incidents include:

* overpressurization and rupture of closed containers and vessels, which can result in an explosion and projectiles that can cause injuries and property damage and/or the dispersal of hazardous materials (possibly violently) and toxic exposures, fires, and explosions

* generation of gases (possibly toxic) or other hazardous materials (e.g., the generation of hydrogen sulfide by the inadvertent mixing of an acid with a sulfide solution)

* generation of heat that causes thermal bums and ignition of combustible materials

* fire in the absence of an additional ignition source

* initiation of other chemical reactions.

The possible consequences of reactivity incidents are more diverse and complex than those incidents involving other types of major hazards. Consequences vary according to the type of chemical reactivity hazards present and the circumstances involved.

Process hazard analysis

Process hazard analyses (PHAs) are conducted to identify potential hazard scenarios and determine whether additional risk-reduction measures are needed to meet tolerable risk criteria. A hazard scenario begins with an initiating event and proceeds through intermediate events to a consequence that impacts receptors such as people, property, and the environment. Scenarios may involve the failure of safeguards (e.g., inhibitor addition, quench system) and-the effects of enablers (e.g., disabled alarms, high ambient temperature) that make the scenario possible. In process safety, the focus is typically on scenarios that involve the major hazards of toxicity, flammability, explosivity, and reactivity.

In addressing toxicity, fire, and explosion hazards, PHA studies concentrate on identifying hazard scenarios resulting from loss of containment and the release of toxic, flammable, or explosive materials. In contrast, uncontrolled chemical reactivity hazards often cause loss of containment, which then produces various adverse effects that include toxic exposures, fires, and explosions. Thus, the nature of chemical reactivity scenarios is different than those of the other types of major hazards (Figure 2).

In scenarios dealing with toxicity, fire, and explosivity, the hazards are realized after containment has failed, and receptors are impacted directly by the realized hazards (toxic exposure, fires, explosions). In chemical reactivity scenarios, the hazard is realized as the proximate result of the initiating event, which causes effects - including possible loss of containment - that produce impacts on receptors.

PHA should identify reactivity hazard scenarios for each operation in a facility where chemical reactivity hazards may be present. PHA teams must be able to recognize the potential for incidents involving the three types of chemical reactivity hazards (self-reaction, runaway reactions, and incompatibilities). They should understand how such incidents can be initiated and their consequences, know how they can be recognized and detected, be aware of safeguards for protecting against them, and know what response actions are appropriate (1-5).

For self-reactive chemicals, the PHA team must explore possible triggers for self-reaction. For intended chemical reactions, the team must consider various causes of runaway, as well as possible side reactions. For incompatible materials, the team must consider the possibility of inadvertent mixing of process chemicals with other chemicals that may be present in a facility, and the possibility that process chemicals may react with other process materials, such as materials of construction. PHA teams will likely need to explore the effects of variables such as charging rates and concentrations. Some quantitative analysis may be needed, for example, to quantify the rate and quantity of heat or gas generated during a chemical reaction. Such analysis is usually performed outside of the PHA sessions.

The identification of the consequences is more complex for reactivity hazard scenarios than for other types of major hazards. While a principal concern with reactivity incidents is loss of containment of hazardous materials due to reactivity excursions that produce overpressures (e.g., equipment leaks or rupture or the operation of relief devices), other effects also may be important, including the generation of toxic gases and other hazardous materials, production of energetic projectiles from damaged process equipment, heat generation, fires, and explosions. PHA teams should consider the consequences that significantly impact receptors of concern.

The presence of chemical reactivity hazards within a process must be determined before the PHA is started. The PHA team then examines whether circumstances could arise that allow the chemical reactivity hazards to be realized and identifies the reactivity hazard scenarios.

Hazard identification

Chemical reactivity hazards should be addressed for all modes and phases of process operation, such as startup, normal operation, shutdown, charging, transfer, and discharging, and all steps in processes where chemical reactivity hazards are present. Storing, handling, repackaging, dispensing, physically processing, producing, and using or transporting materials that are chemically reactive should be considered. Reactants, reaction mixtures, byproducts, waste streams, and products also should be addressed, as appropriate.

For each operation involving chemically reactive or potentially chemically reactive materials, all three types of reactivity hazards must be considered - unstable chemicals, chemical reactions subject to runaway, and incompatible materials (i.e., chemicals that are reactive with process materials and chemicals that cause a reactivity incident if inadvertently mixed).

Considerable data are available on chemical instabilities; information on runaway reactions can be obtained through laboratory experiments; and incompatibilities should be determined via a chemical interaction matrix (5). Often, a significant amount of work is needed to develop this information for a process and put it into a form suitable for use in a PHA. The National Oceanographic and Atmospheric Administration (NOAA) has developed a free software tool, the Chemical Reactivity Worksheet (available at http://response.restoration.noaa.gov/reactivityworksheet), that can be used to create a chemical interaction matrix.

The reactivity hazard identification step produces a list of the reactivity hazards present in a process throughout its lifecycle and the locations where and times when they may occur. The list should include information on the circumstances under which reactivity hazards may be realized, for example, the onset temperature for an exotherm. Reference 6 describes the critical role of chemical reactivity data for PHA, and References 5 and 7 provide procedures and tools for identifying and screening chemical reactivity hazards.

Some iteration may be needed between the PHA and the hazard identification activity (8). PHA studies clarify the circumstances that chemically reactive materials will experience; as the chemical reactivity hazards of a process are better understood, the need for ftirther evaluation and testing may arise. Thus, it is important that the PHA team includes personnel who understand chemical reactivity hazards. In addition, they should be familiar with the chemicals and materials present in the process - their chemical and physical behavior under both normal and upset process conditions.

Identification of potential chemical reactivity scenarios

Chemical reactivity hazards are sometimes not addressed in PHA owing to the lack of awareness of such hazards by the PHA team and/or because the PHA technique used does not effectively address chemical reactivity hazards. Few PHA approaches designed specifically to identify hazard scenarios involving chemical reactivity hazards have been described in the literature. The Center for Chemical Process Safety (CCPS) guidelines on hazard evaluation (9) contain only a brief discussion of the evaluation of chemical reactivity hazards* *. One paper discusses the issue, but does not expand upon the identification of scenarios (8).

PHA methods differ primarily in the way they approach the identification of initiating events for hazard scenarios. The identification of other scenario elements is performed in essentially the same way for all methods. Consequently, reactivity hazard scenarios require a PHA approach that best addresses their causes. The challenge is greater than for other major hazards owing to the multiple types of chemical reactivity hazards and their diversity of causes.

One PHA approach called chemistry hazard analysis (CHA) applies specifically to chemical reactivity hazards (10). Derived from the hazard and operability (HAZOP) study method, CHA applies HAZOP guide words to the reaction chemistry. The focus is on consequences rather than causes of scenarios, but the method could be extended to identify complete scenarios. However, the HAZOP study method is not as well suited to the identification of chemical reactivity hazard scenarios as other major hazard scenarios (because of the differences discussed earlier). Also, HAZOP is susceptible to the incomplete consideration of design intent for a process (11), which may result in reactivity scenarios being missed. What-if analysis directly addresses the causes of accidents and could include reactivity scenarios, but it it depends entirely on brainstorming by the team or, alternatively, the use of lengthy checklists.

Another PHA approach, the major hazards analysis (MHA) method (12-13) uses structured brainstorming to identify initiating events directly and is well suited for identifying reactivity hazard scenarios. The overall approach in MHA is similar to that of conventional PHA. The PHA team works through each process mode, operation, and step, evaluating the potential for chemical reactivity and other major hazards in each part of the process. All major hazards are addressed in an equivalent way. Processes are divided into sections of suitable size, usually called nodes, for the purpose of the analysis. For each process section, the team uses information from the hazard identification and considers ways that chemical reactivity and other major hazards may be realized in the section. Each of the different types of chemical reactivity hazards that are present must be addressed in turn.

The MHA technique uses a concise checklist of causes of scenarios organized into categories to identify hazard scenarios. This approach provides more guidance than the simple brainstorming in a HAZOP study, and it does not require the completion of lengthy lists of questions as what-if checklist studies do. Also, MHA checklists, which can be tailored for the specific process being studied, include only items that can result in a major hazard scenario. Table 1 is an example of an MHA checklist for immediate causes of chemical reactivity hazards. The checklist can be incorporated into PHA software to allow the team to select items for entry into the MHA worksheet for each process section studied.

As with other PHA methods, immediate causes such as those shown in Table 1 must be evaluated to identify their underlying causes (14). For example, in the case of a selfreactive chemical, possible underlying reasons for heating that cause a reactivity excursion may include higher-thannormal ambient temperature, excessive heating-medium temperature, and loss of agitation. In the case of a runaway reaction, possible underlying reasons for the presence of contaminants that cause a runaway reaction may include residual chemicals from a previous batch, use of a reactant from a new supplier that has a different specification, and the presence of cleaning chemicals after a maintenance operation. Of course, these causes will also have further underlying causes, which must be evaluated until a level of causality is reached that is suitable for deciding on the development of additional risk-reduction measures.

Once causes of scenarios have been identified at a suitable level of causality, other scenario elements can be addressed in the usual way, including safeguards for present and possible scenario consequences. Risks are then ranked to help determine the need for further risk reduction. Figure 3 demonstrates the application of MHA to chemical reactivity hazards.

It is possible to address chemical reactivity hazards and other types of major hazards in separate PHA studies. For example, the HAZOP technique could be used to address toxicity, flammability, and explosivity, and MHA or another suitable technique could be used for chemical reactivity. Alternatively, all major hazards could be addressed in a single study using a method that treats all major hazards equally well, such as the MHA method. The advantages of a single study are that hazard scenarios are recorded in one place and the analysis is more efficient. The advantages of separate studies are that different PHA methods can be used, and specialty team members, such as chemists, may need to be present only for the chemical reactivity hazards study.

Pitfalls of chemical reactivity analysis

Several pitfalls await PHA teams that are inexperienced at addressing chemical reactivity hazards. Here are some ways to avoid them.

Lack of understanding of chemical kinetics and thermodynamics. The behavior of chemically reactive materials can be complex and highly dependent on the circumstances under which the chemicals are used. Inadequate understanding of process chemistry and lack of awareness of causes of reactivity problems are sources of reactivity incidents (15). Consequently, a knowledgeable chemist should be part of the PHA team if complex or unusual chemical reactivity hazards may be present.

Inadequate process design. Inadequate designs, safety systems, and control systems have been cited as the cause of incidents (15). A useful precursor to a PHA study is a review of the process by its designers with assistance from process chemists to ensure that the process is properly designed to address chemical reactivity hazards. The PHA team can then focus on how deviations from the design intent can produce reactivity hazard scenarios.

Problems with procedures. Both inadequate operating procedures and failure to follow procedure have been identified as sources of reactivity incidents (15). Procedures may not specify safe operating limits for avoiding chemical reactivity hazards. A review of procedures to ensure they address chemical reactivity hazards should be conducted prior to embarking on the PHA study. PHA teams should not be burdened by identifying problems that should have been addressed by others beforehand.

Misconceptions about chemical reactivity ratings. The National Fire Protection Association's standard NFPA 704 rates chemical reactivity (16). However, NFPA ratings address inherent instability only, not reactivity with other chemical substances (except water), nor chemical behavior under nonambient conditions. The PHA team members should be aware of this and the broader meaning of chemical reactivity.

Insufficient consideration of chemicals with low reactivity ratings. Some chemicals may be characterized as lowreactivity materials, as they are not particularly reactive under normal operating conditions. However, such chemicals can be very reactive under some conditions, for example, in combination with other chemicals, or at temperatures that are reached only as a result of deviations from normal operating conditions. Where chemical reactivity hazards may be present in a process, the PHA team should be briefed on how they may arise, and a chemist should be part of the team.

Incomplete safety data sheets (SDSs). SDSs for raw materials may not identify all hazards that may be encountered upon mixing with other chemicals or contact with other materials. This issue should be addressed during the screening of chemical reactivity hazards for the process, and the chemical interaction matrix for the process should be provided to the PHA team.

Incomplete list of sources of chemical reactivity hazards. Chemical reactivity hazards may derive not only from raw materials, but also from intermediates, products, and byproducts of chemical interactions. The personnel performing the chemical reactivity hazards screening must search for such hazards and communicate them to the PHA team.

Neglecting chemical reactivity hazards that develop over time. Some chemical reactivity hazards build slowly over time, for example, the formation of peroxides in stored chemicals. The personnel performing the chemical reactivity hazards screening must search for such hazards and communicate them to the PHA team.

Failing to consider a combination of causes. Reactivity incidents may result from a combination of factors. For example, the explosion and fire that occurred at a Morton International facility in Paterson, NJ, on April 8, 1998, is believed to have been triggered by a combination of events: a chemical reaction was started at a higher-thannormal temperature; steam used to initiate the reaction was left on for too long; and cooling water used to control the reaction rate was not turned on soon enough (17). Identification of hazard scenarios involving multiple failures is challenging for PHA teams, but studies should address them to the extent feasible (18).

Misunderstanding of the importance of runaway reactions. A common, but incorrect, assumption is that chemical reactivity hazards in processes are dominated by runaway reactions in reactor vessels. The U.S. Chemical Safety and Hazard Investigation Board (CSB) found that only 35% of the reactive incidents it studied were attributable to runaway reactions, and reaction vessels accounted for only 25% of the equipment involved (15). Such reactivity incident data should be included in briefings of PHA teams to ensure they do not focus unduly on the types of reactivity hazard scenarios that they perceive to be the most important when data are available that contradict their perceptions.

To sum up

Managing chemical reactivity hazards is more complex than other major hazards encountered in the CPI. PHA teams are less familiar with these hazards and reactivity hazard scenarios are more difficult to identify than other major hazards. Traditional PHA methods, such as HAZOP, are not as well suited to addressing chemical reactivity hazards as other methods such as MHA. Various pitfalls may be encountered in addressing chemical reactivity hazards during PHA. Awareness of pitfalls can help PHA teams avoid them.

The management of chemical reactivity hazards should be integrated into a company's overall process safety management program. Such hazards must be addressed in all relevant elements of process safety, including not only PHA but also process safety information, hazard communication and training, management of change, incident investigation, emergency response, and audits.

* Editor's Note: Other CCPS publications, e g.. Refs. 2 and 3, deal with chemical reactivity in detail.

Literature Cited

1. Yoshida, T., "Safety of Reactive Chemicals," Elsevier Science Ltd., Amsterdam, Netherlands (1987).

2. Center for Chemical Process Safety, "Guidelines for Chemical Reactivity Evaluation and Application to Process Design," AIChE, New York, NY (1995).

3. Center for Chemical Process Safety, "Guidelines for Safe Storage and Handling of Reactive Materials," AIChE, New York, NY (1995).

4. Barton, J., and R. Rogers, "Chemical Reaction Hazards - A Guide to Safety," 2nd ed., IChemE, London, U.K. (1997).

5. Johnson, R. W., et at,, "Essential Practices for Managing Chemical Reactivity Hazards," John Wiley and Sons, Hoboken, NJ (Mar. 2003).

6. Berkey, B., and W. Workman, "Relating Chemical Reactivity to Process Hazards," International Symposium on Preventing Major Chemical Accidents, CCPS, New York, NY (1987).

7. Davis, E. M., et aL, "The CCPS Chemical Reactivity Evaluation Tool," Process Safety Progress, 31 (3), pp. 203-218 (Sept. 2012).

8. Johnson, R. W., and S. D. Unwin, "Addressing Chemical Reactivity Hazards in Process Hazard Analysis," presented at the 18th Annual International CCPS Conference, Scottsdale, AZ (Sept. 2003).

9. Center for Chemical Process Safety, "Guidelines for Hazard Evaluation Procedures," 3rd ed., John Wiley and Sons, Hoboken, NJ (2008).

10. Mosley, D. W., et al., "Screen Reactive Chemical Hazards Early in Process Development," Chem. Eng. Progress, 96 ( 11 ), pp. 51 -65 (Nov. 2000).

11. Baybutt, P., "Requirements for Improved PHA Methods: Addressing Weaknesses in HAZOP and Other Traditional PHA Methods," presented at the 2014 Spring Meeting and 10th Global Congress on Process Safety, New Orleans, LA (Apr. 2014).

12. Baybutt, P., "Major Hazards Analysis -An Improved Process Hazard Analysis Method," Process Safety Progress, 22 ( 1 ), pp. 21-26 (Mar. 2003).

13. Baybutt, P., and R. Agraz-Boeneker, "A Comparison of the Hazard and Operability (HAZOP) Study with Major Hazard Analysis (MHA): A More Efficient and Effective Process Hazard Analysis (PHA) Method," presented at the 1st Latin American Process Safety Conference and Exposition, CCPS, Buenos Aires, Argentina (May 2008).

14. Baybutt, P., "Initiating Events, Level of Causality and Process Hazard Analysis," Process Safety Progress, 33 (3), pp. 217-220 (Sept. 2014).

15. U.S. Chemical Safety and Hazard Investigation Board, "Hazard Investigation: Improving Reactive Hazard Management," wwvv.csb.gov/assets/1 /19/ReactiveHazardInvestigationReport.pdf, CSB, Washington, DC (Oct. 2002).

16. National Fire Protection Association, "Standard System for the Identification of the Hazards of Materials for Emergency Response," NFPA 704, NFPA, Quincy, MA (2012).

17. U.S. Chemical Safety and Hazard Investigation Board, "Investigation Report: Chemical Manufacturing Incident, Morton International, Inc., Paterson, NJ, April 8, 1998," Report No. 199806-1-NJ, CSB, Washington, DC (Aug. 16, 2000).

18. Baybutt, P., "Treatment of Multiple Failures in Process Hazard Analysis," Process Safety Progress, 32 (4), pp. 361-364 (Dec. 2013).

ADDITIONAL RESOURCES

Baybutt, P., "Analytical Methods in Process Safety Management and System Safety Engineering - Process Hazards Analysis," in Haight, J. M., ed., "Handbook of Loss Prevention Engineering," DOI: 10.1002/9783527650644.ch21, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany (2013).

Johnson, R. W., and P. N. Lodal, "Screen Your Facilities for Chemical Reactivity Hazards," Chem. Eng. Progress, 99 (8), pp. 50-58 (Aug. 2003).

Liu, Y.-S., "Screening Reactive Chemical Hazards," Chem. Eng. Progress, 102 (5), pp. 41-47 (May 2006).

Murphy, J. R, "Understanding Reactive Chemical Incidents," Chem. Eng. Progress, 100 (3), pp. 31-33 (Mar. 2004).

Saraf, S. R.,etaL, "Classifying Reactive Chemicals," Chem. Eng. Progress, 100 (3), pp. 34-37 (Mar. 2004).

UJS. Environmental Protection Agency, "General Guidance on Risk Management Programs for Chemical Accident Prevention," 40 CFR Part 68.

U.S. Occupational Health and Safety Administration, "Process Safety Management of Highly Hazardous Chemicals," 29 CFR 1910.119.

U.S. Occupational Health and Safety Administration, "Process Safety Management and Prevention of Major Chemical Accidents," Federal Register, 78 (236), pp. 73756-73768, www.osha. gov/pls/oshaweb/owadisp.show_document?p_table=FEDERAL_ REGISTER&p_id=24053 (Dec. 9,2013).

Paul Baybutt

Primatech, Inc.

PAUL BAYBUTT is the president, CEO, and founder of Primatech, Inc. (Email: [email protected]). He has worked in the fields of risk analysis and process safety since 1975. He received BSc, MSc, and PhD degrees in chemistry from the Univ. of Manchester in England. He also conducted postdoctoral research on chemical reaction systems at Battelle Memorial Institute. He is a member of AlChE; the Society for Risk Analysis; the Safety and Reliability Society; the Instrumentation, Systems and Automation Society; and the Royal Society of Chemistry.