lable at ScienceDirect

Journal of Loss Prevention in the Process Industries 22 (2009) 757–763

Contents lists avai

Journal of Loss Prevention in the Process Industries

journal homepage: www.elsevier .com/locate/ j lp

A little knowledge is a dangerous thing – Unexpected reaction case studiesmake the case for technical discipline

John Brennan a,1, Gregg M. Kiihne b,*

a BASF Corporation, 8404 River Road, Geismar, LA 70734, USAb BASF Corporation, 602 Copper Rd, Freeport, TX 77541, USA

a r t i c l e i n f o

Article history:Received 26 January 2009Accepted 22 July 2009

Keywords:Runaway reactionsOperational disciplineProcess safetyDisciplineCase studies

* Corresponding author. Tel.: þ1 979 415 6323.E-mail addresses: john.brennan@basf.com (J. Bren

(G.M. Kiihne).1 Tel.: þ1 225 339 7087.

0950-4230/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.jlp.2009.07.012

a b s t r a c t

The old saying, ‘‘what you don’t know can’t hurt you,’’ implies that ignorance is bliss. ‘‘A little knowledgeis a dangerous thing,’’ may be closer to the truth; however, it is not the little that we know that isdangerous, but that which is not known. By design, the processes used in the chemical industry arereactive, and the intended reaction receives much scrutiny. However, other reactions occur, oftenunexpectedly, and possibly with severe consequences. The lessons we learn from these reactions mustdrive the improvement of our process development and technology management processes and theculture that shapes those processes, a culture of Technical Discipline.

Technical Discipline, analogous to Operating Discipline in the manufacturing organization, is a culturecommitted to fully identifying and characterizing chemical and reaction hazards, and properly doc-umenting and communicating those hazards to create a permanent knowledge and understandingwithin the organization operating that process.

A culture of Technical Discipline will reveal reaction hazards that might otherwise remain unknownuntil being unveiled in a dramatic and unexpected fashion. Until you fully identify and characterize thehazards of the materials you handle in your processes.what you don’t know can hurt you.

� 2009 Elsevier Ltd. All rights reserved.

1. Background

While reviewing process safety-related incidents, members ofthe process safety expertise team noted a trend of incidentsinvolving unexpected reactions. The incidents encompassed reac-tions in all stages of processing, including reactors, productcontainers, purification and drying steps. In looking for common-alities among the incidents, team members found a patternrelating to the key technical knowledge or information behind thereactions. They were able to classify the incidents in one of fourcategories based on the type of knowledge ‘error’ that led to theunexpected reactions.

1. We did not know the basic facts – one or more basic facts aboutthe unexpected reaction was not previously known. The technicaldevelopment itself did not adequately identify the potential forthe unexpected reaction.

nan), gregg.kiihne@basf.com

All rights reserved.

2. We knew the facts but did not fully understand them – basicknowledge was not applied adequately. Technical developmentwas not sufficiently understood so that adequate controlmeasures were implemented.

3. We knew the facts, but did not communicate them wellenough –the basic knowledge was known somewhere within theorganization/business group, but was not adequately communi-cated to the personnel operating the unit. A breakdown occurred inthe communication/training process so that the personneloperating the process were not adequately aware of the hazards.

4. We knew the facts, but forgot them – the basic knowledge wasknown within the organization in the past, but was no longerfound in current training or procedures at the location thatexperienced the incident. A breakdown occurred in the docu-mentation process that allowed the knowledge of this potentialreaction to be lost or forgotten over some period of time.

No matter which of these four basic shortcomings allowed theincident to occur, ultimately there was a breakdown in the tech-nical/know-how development process. To prevent such break-downs and future process safety incidents due to unexpectedreactions, a culture of Technical Discipline must be developed.

Technical Discipline, analogous to the concept of OperatingDiscipline for manufacturing units, is defined as a culture committed

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763758

to fully identifying and characterizing chemical and reaction hazards,and properly documenting and communicating those hazards tocreate a permanent knowledge and understanding within the orga-nization operating that process.

2. Case studies

2.1. Case study 1: unexpected reaction in a flash dryer

2.1.1. What happenedIn a production operation, the final product (a solid powder) is

obtained by filtration from an aqueous suspension followed byremoving the remaining water from this wet product (presscake) ina flash drying operation (see Fig. 1).

The presscake is fed to the flash drying chamber where it iscontacted with hot air in a highly turbulent environment, rapidlyevaporating the water and forming fairly small solid particles.These dry, solid particles are conveyed out of the flash dryingchamber with the air stream to a filter where the product iscollected, then sent for further downstream processing. Theproduct is a combustible organic solid, so there is a known potentialdust explosion hazard.

Several days leading up to the incident, there had been opera-tional difficulties that had required multiple start-ups, shut-downsand cleanouts of the system. The unit was finally stabilized andrunning well for several hours when the presscake feed wasinterrupted due to loss of supply.

The sudden loss of feed resulted in a significant upset in theflash dryer. High temperature and high pressure interlocks shut theunit down immediately, without the cool down normally associ-ated with a controlled shutdown.

Operators saw what they thought to be smoke in the flashdrying chamber and believed there was a fire inside the unit. Theyevacuated the area, initiating the emergency alarm that evacuatedthe entire production building. The emergency response team wasassembling on the ground floor of the building when a boom washeard and felt. They evacuated the building immediately. Simul-taneous with the boom, a cloud of material was ejected from theexhaust fan outlet.

Eight minutes later, a similar but more powerful event occurredsending a cloud of material out of the inlet fan, causing significantdamage to the fan, the inlet ductwork and the inlet filter assembly.During the next hour three more similar events occurred, each oneless powerful than the previous.

Fig. 1. Flash dryer pr

2.1.2. How it happenedThe operations team believed these events to be dust explo-

sions, as this was the primary hazard associated with this opera-tion. However, the incident investigation told a different story.What had occurred was the thermal decomposition of organicsolid. Due to inadequate cleaning and inspection over a period ofseveral months, a substantial amount of solids had accumulatedbetween the air heater and the drying chamber, a portion of thesystem that was never intended/expected to contain solids, andcoincidentally, where the highest process temperatures existed.

Normal flash dryer operation exposes presscake to the heatedair stream. Because of the significant presscake water content, thetemperature that the product sees is limited to approximately100 �C. Additionally, the residence time in the flash dryer is veryshort. The solids that had accumulated in the flash dryer, however,were exposed to much high temperatures for an extended period oftime (weeks to months). Post-incident experimental testingshowed that the solids would decompose violently with substantialheat and non-condensible gas evolution.

2.1.3. Why it happenedAs with most significant incidents, there were numerous causal

factors and root causes. They included inadequate design andplanning for emergency response, as well as not followingapproved procedures for cleaning and inspection of the flash dryerunit. However most importantly, the product thermal behavior wasnot well understood and the decomposition reaction violence was notknown. The major focus of the safety concept that had been appliedto this operation was preventing dust explosions, and a sophisti-cated system was installed for protection. This strategy providedessentially no protection against a significant decomposition eventlike the one that occurred.

This incident demonstrates a case where we did not know thebasic facts.

2.1.4. What could have been done to avoid this incident?

- Develop comprehensive thermal stability data for the product,which would have:

B revealed the decomposition hazardsB allowed for a design and operating strategy to manage the

hazard effectivelyB emphasized the safety-related importance of cleaning and

inspecting effectively

ocess schematic.

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763 759

- Enforce the established cleaning and inspection protocols,which would have eliminated the solids accumulation in thesystem

2.2. Case study 2: unexpected reaction in a product recovery unit

2.2.1. What happenedIn a continuous process plant, the chemistry results in side-

products that must be removed to meet product quality require-ments. These side-products are removed in a distillation train andare collected from the sump of the last tower. The bottoms productfrom this last distillation tower also contains a considerableamount of the desired product, so this stream is sent to a productrecovery unit (see Fig. 2) to recover as much desired product aspossible.

The concentrated side-products stream is referred to as residue.Residue can solidify in the piping and equipment if the temperatureis too low. However, if the temperature is too high, it can dimerize,polymerize and decompose, forming a solid.

On the day of the incident, the unit downstream of the productrecovery unit tripped off-line, interrupting the flow of residue. Theoperations team attempted to operate valves that would open analternate flow path for the residue, but were not successful. Theproduct recovery unit continued to operate, maintaining circula-tion through the heater and evaporating product. The feed to theunit was being reduced as the evaporator level increased due to theloss of the residue discharge flow. Eventually feed to the unit wasstopped and steam to the heater was interrupted.

Sections of line were then flushed with product and solvent inan attempt to free the inoperable valves. After approximately 2 h ofattempting to open a residue discharge flow path, a loud boom washeard, followed by the release of material from within the productrecovery unit.

The unit was immediately evacuated and emergency responsecommenced. Ten minutes after the initial event a second release

Fig. 2. Product recover

occurred accompanied by a sound like a PSV relieving, followedminutes later by another loud boom and a further release ofmaterial from within the product recovery unit structure.

There were no injuries, but there was substantial damage to theproduct recovery unit equipment and piping, resulting in signifi-cant downtime for that unit and a major financial impact to theoperation.

2.2.2. How it happenedThe interruption of feed to and residue discharge from the

product recovery unit resulted in over-concentrating the residue inthe recirculation loop, accompanied by increased viscosity andreduced flow. The residue concentration was monitored by the loadon the recirculation pump motor. However, in this abnormaloperating condition (no net flow of residue out of the system) thismonitoring strategy failed because the expected increased motorload due to increased viscosity was offset by the reduced flow. Asa result, the pump continued to run for a significant time witha drastically reduce flow, resulting in substantial heat input to thefluid in the pump. This eventually caused the solidification of thematerial in the recirculation loop, which continued to polymerizeand decompose.

The decomposition reaction evolves CO2 gas. This over-pres-sured the piping and heat exchanger in the recirculation loop, sincethe solidified mass blocked the available vent paths. The two boomsheard were the over-pressure failure of two sections of the recir-culation piping. The PSV-like release was the upper head of theheater lifting and venting as the flange bolts stretched.

2.2.3. Why it happenedThe resulting violent decomposition/gas evolution that over-

pressured and failed the piping and heat exchanger was completelyunexpected, because the hazards associated with hot, highlyconcentrated residue were not well understood by the design teamand the operations team.

y unit schematic.

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763760

There were numerous causal factors and root causes identifiedfor this incident including an advanced process control scheme thatdid not consider upset conditions and the lack of a reliable venting/pressure relief system. However most importantly, the operationsteam did not know that the residue could undergo an exothermic,gas-evolving reaction under conditions of high concentration andtemperature. This was knownwithin the technical community, but itwas not identified as a hazard in the safety reviews for that section ofthe plant. If operating personnel had understood this hazard, theirresponse to the upset conditions would have been different.

This incident demonstrates a case where we knew the facts, butdid not fully understand them.

2.2.4. What could have been done to avoid this incident?

- A more thorough and comprehensive safety review for thissection of the process.

This was accomplished as a part of the corrective actions. Thisunit was down for five months to implement major design andoperational changes including:

B Measuring additional process parameters to bettermonitor the unit condition,

B Implementing a Safety Instrumented System to preventthe occurrence of unacceptable conditions,

B Implementing a significantly revised and more robustcontrol strategy that responds correctly even duringabnormal conditions,

B Revising procedures for responding to abnormal condi-tions, and

B Retraining the operations team

2.3. Case study 3: unexpected reaction in a waterpre-treatment process

2.3.1. What happenedIn a facility that manufactures pigments, the water removed

from the product must be pre-treated to de-color the water beforebeing sent to the waste treatment facility. This de-coloring process

Fig. 3. Multi-purpose diaphragm pump u

includes the addition of 50% hydrogen peroxide (H2O2). Theperoxide was charged with a wand from 5 gallon jugs using a teflonair-driven diaphragm pump.

In the course of a normal charge of peroxide, the dedicatedperoxide pump failed. Failure to treat the wastewater would resultin a shutdown of production; therefore, a search began for a suit-able replacement. There was no dedicated spare for the pump andno replacement was available from the warehouse.

Finally after much searching, a general-purpose teflon dia-phragm pump was located in another part of the facility, (see Fig. 3).The pump was rinsed out and connected to the piping for thetreatment system. The operator resumed charging the peroxide bystarting airflow to the diaphragm pump. After only a few seconds ofoperation, the replacement pump stopped pumping. Seconds later,the pump ruptured violently. The operator charging peroxide viathe wand sustained chemical burns from the released peroxide anda broken leg from pump shrapnel.

2.3.2. How it happenedThe operating technicians knew that hydrogen peroxide was

reactive and needed to be pumped using a plastic pump with Teflondiaphragms, but they did not fully understand the reactive hazardsof the peroxide. The general-purpose pump that was secured toreplace the failed peroxide pump had been used in dozens ofdifferent services for more than 15 years. The pump casing hadmany internal fissures that contained traces of these materialspreviously pumped.

However, the most likely cause of the rapid decomposition ofthe peroxide was the water used to rinse out the pump. The pumpwas rinsed with water from a supply located on a building adjacentto the peroxide shed. This building had been out of service forthe last two years and this supply had not been used for manymonths. The water in the line contained significant levels of rust(iron oxide), which is an excellent catalyst for the decomposition of50% hydrogen peroxide.

2.3.3. Why it happenedThere were several causal factors and root causes that allowed

this incident to occur. The operating technicians were aware thathydrogen peroxide was reactive, but did not understand the extent

sed to pump 50% hydrogen peroxide.

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763 761

of the reactivity, nor did they know all of the materials that couldcause the peroxide to decompose. A management of change wasnot completed since the technicians considered this to bea replacement-in-kind. Furthermore, the operations supervisor didnot consult with the engineers or operations manager regardingthe replacement and the measures needed to prepare the pump forservice. There is a wealth of information available regarding thereactivity of hydrogen peroxide, but ultimately weaknesses in thetraining and management of change programs prevented thishazard from being recognized and properly managed in this case.

This incident demonstrates a case where we knew the facts, butdid not adequately communicate them.

2.3.4. What could have been done to avoid this incident?

- Develop a comprehensive training program for techniciansthat includes the following aspects:

B Discussion of material and process hazards, with a strongfocus on highly reactive materials and processes,

B Exercises or activities to demonstrate understanding ofthe training material, and

B Refresher training to ensure the pertinent points of thetraining are not forgotten over time.

- Ensure that a technically qualified person is involved in theevaluation of every change within the operating facility, andthat all personnel are aware of changes that are made ina facility.

2.4. Case study 4: unexpected reaction with storage and transferof an intermediate

2.4.1. What happenedDue to complications with the start-up of a modified unit,

several tons of an intermediate material were cleared from processvessels and temporarily stored in a railcar. Once the unit wasrunning properly, the intermediate was to be transferred from therailcar to smaller containers for processing. The intermediatematerial solidifies at temperatures in the range of 80–90 �C, so theheating/cooling coils in railcar were connected to the plant’s 4 barg

steam system to melt the intermediate for transfer.

Fig. 4. Remains of the railcar contain

Plant personnel began heating the material and had drawn offall but about 10,000 lb of material from the railcar when therelief valve on the railcar lifted. The relief valve vented for severalminutes then reseated. Personnel in the area claimed that theysmelled ammonia, but there was no ammonia stored in therailcar.

Several minutes later, the relief valve lifted once again, but thistime was relieving at a much higher velocity. One of the engineersoverseeing the operation called for an evacuation of all personnelfrom the area. Seconds after the last employee reached cover, therailcar failed catastrophically in an explosion that was felt up to10 miles away (see Fig. 4).

2.4.2. How it happenedThe operations team had previously handled only small ‘drum’

quantities of this material (less than 55 gallons/200 L), which wastypically melted with hot water when solidified. The location of therailcar only had utility connections for 4 barg steam instead of hotwater. The 4 barg steam has a temperature of approximately 160 �Cinstead of the 95–100 �C found in the hot water system. As thesteam was applied to the coils in the railcar, the material nearest thewall began melting and was slowly drained from the railcar. Forthe first few hours of this operation, there was still a large mass ofsolid material in the railcar. As long as this solid material waspresent, the bulk temperature of the liquefied material remainedclose to the melting temperature of approx. 80 �C.

Once all of the solid material in the railcar melted, the bulktemperature began approaching the temperature of the steam.When the bulk temperature reached a temperature of approxi-mately 120 �C, the intermediate material began to decompose. Thefirst stage of the decomposition was slightly endothermic andproduced ammonia as one of the reaction products – thus anammonia odor was noted when the relief device on the railcar firstvented.

The second stage of the decomposition is highly exothermicand produces non-condensible gases among the reaction products.This caused the second, more energetic venting from the reliefdevice. The energy of the decomposition reaction soon overcamethe relief device and the integrity of the railcar, which rupturedexplosively.

ing the intermediate compound.

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763762

2.4.3. Why it happenedAs with most significant incidents, there were numerous causal

factors and root causes. However, during the investigation, it wasnoted that all other ‘sister’ facilities to this plant had a hot watersystem installed for melting this intermediate if it solidifies,whereas this plant had only a 4 barg steam system. Upon furtherinvestigation, older analytical reports and MSDSs indicated thatthis material could decompose violently at temperatures above120 �C. The hazard that resulted in this explosion had been iden-tified decades earlier, but this information was forgotten; theorganizational memory was lost.

This incident demonstrates a case where we knew the facts, butforgot them.

2.4.4. What could have been done to avoid this incident?

- Develop permanent documentation/technology transfersystems to accomplish the following:

B Ensure significant hazards of a process, such as this one,are effectively communicated to each ‘generation’ ofemployee,

B Ensure that special design measures, such as a hot watersystem, are maintained as part of the process safetyconcept for the technology, and

B Ensure employee training materials include training onvital safety aspects of the process.

3. Technical discipline

The incidents described in the preceding case studies all hadmultiple contributing causes, but all could have been avoided byproperly identifying the potential for the particular unexpectedreaction, improving the technology or ‘‘know-how’’ transfer tooperations to ensure an adequate understanding of the reactions, orbyenhancing the documentation and communication systems withinthe units to ensure the knowledge of the reaction is understood by allassociated with the operation of the process, not only now, but for thelife of the facility. In short, the incidents could have been prevented ifa culture of strict Technical Discipline had been in place.

The concept of Operational Discipline is now fairly well-knownwithin industry, and has been defined by DuPont ‘‘as a deeply rooteddedication and commitment by each member of the organization tocarry out each task the right way, each time.’’(Klein, 2005) TechnicalDiscipline follows the same concept as Operational Discipline, butapplies it specifically to development and design processes.

Technical Discipline, therefore, may be defined as a culturecommitted to fully identifying and characterizing chemical andreaction hazards, and properly documenting and communicatingthose hazards to create a permanent knowledge and understandingwithin the organization operating that process.

4. A culture of technical discipline

What does a culture of Technical Discipline look like? Whatare some of the key characteristics of such a culture? Thefollowing is not a comprehensive listing, but rather a few keypractices, processes, and systems that support a culture of Tech-nical Discipline.

1. Standardized Process Development Process/Practices/Proce-dure – Reduces the likelihood that pertinent information willbe overlooked.

A well-structured, consistent approach for developing thereaction hazards knowledge associated with raw materials,

products, side-products and residue streams is the founda-tion of Technical Discipline.Reaction hazards can be organized in two categories: (Basf,1997; Basf, Berthold, Giesbrecht, Loffler, & Maurer, 1992; Basf,stark, Riede, Maurer, & Schoenherr, 1998).

Self-reactivityDeveloping comprehensive data describing the materialbehavior as a function of temperature is key to avoidingunexpected reactions.Time-temperature dependent behavior and the effects ofimpurities should be investigated to determine if thematerial exhibits non-Arrhenius behavior.Storage is not trivial. Storage stability studies are essen-tial and should cover residence times and temperaturesthat exceed the expected limits, to account for problemsor delays in manufacturing or storage.Recommended testing for reactivity hazards includes:For liquids, differential scanning calorimetry (DSC),thermogravimetric analysis (TGA) and adiabatic calo-rimetry can be used to accurately identify the onsettemperature for exothermic behavior and the associatedenergy release.For solids, the same tests as for liquids plus minimumignition temperature of a thin layer, Grewer Oven orequivalent, and Drahtkorb (wire basket) testing, to char-acterize behavior of thin layers, small quantities, andlarger quantities, respectively. It is important to match thetesting method to the actual conditions that will beexperienced in the process under normal and upsetconditions.For reactive materials that will be stored or handled incontainers, the self-accelerating decomposition temper-ature (SADT) should be determined.

A consistent strategy for applying adequate temperaturebuffers from onset temperatures is essential, and should bebased on the reaction characteristics and the typical upsetconditions that the process will experience (not only rulesof thumb, i.e. the 50 �C rule).

Reaction with other materials

Comprehensively identifying all materials that can come incontact with one another in a process is essential. For multi-purpose plants, materials from other processes should alsobe evaluated. A matrix illustrating all situations that canresult in an unwanted reaction can then be developed. Thismatrix is used to establish storage and operating strategiesthat eliminate or minimize the potential for unintendedmixing of non-compatible materials.

2. Robust Technology/Know-How Transfer Systems – Ensures thehazard and safeguard information from the first step is prop-erly documented and communicated to the design andmanufacturing teams.

Standard report formats for reactivity test results are highlyrecommended, as this simplifies interpretation and mini-mizes the chance of misinterpretation of the data.

3. Process Safety and Exothermic Reactions Competency/Training –Ensures the basic knowledge and competency needed to inter-pret the hazard data and to properly eliminate or manage thosehazards.

Roles and responsibilities at every level within a company, asthey relate to reaction and reactivity issues must be clearly

J. Brennan, G.M. Kiihne / Journal of Loss Prevention in the Process Industries 22 (2009) 757–763 763

defined. Those whose function involves the identificationand/or management of reactivity hazards must have theappropriate technical competency.A training program identifying job categories and therequired reactivity hazards training is vital to ensuringcompetency throughout the organization.Training content should include basics of exothermic reac-tion behavior, testing for characterization of reactionhazards, Arrhenius and auto-catalytic behavior, heat transfer,scale up issues, and typical mitigation strategies.

4. Standardized Process Hazard Documentation and Communi-cation Tool – Reduces the Likelihood that Pertinent Informationwill not be properly documented and communicated.

A ‘‘Process Safety Concept’’ is a tool used to document all basichazards of a process and communicate the minimumrequirements for safe operation. A Process Safety Conceptmust include a section on reaction/reactivity hazards. Thissection should contain the analysis and interpretation of testresults, and also the raw test data. Use of a standardized formatmakes information in this document readily accessible. AProcess Safety Concept should be developed for every process.As companies become more globalized, often technology isdeveloped in one country, but implemented in another.Technology transfer documentation must be provided in thenative language of those that will need to understand,interpret, and apply the reaction/reactivity hazard informa-tion, and those charged with implementing and maintainingappropriate safeguards. Any translation that is necessaryshould be made in cooperation with the receiving group toensure suitability of the language and terminology used.Regular updates of reaction information are essential.Responsibility for this must be clearly assigned within theR&D, process development, technical, engineering andmanufacturing communities to ensure that changes that arisefrom incidents, optimization efforts, waste minimizationefforts, etc. are effectively communicated in a timely way.

5. Robust Communication/Documentation Systems (Training,Process Safety Information) – Ensures the hazard and safeguardinformation is properly maintained (organizational memory)for the life of the process.A robust system is one that is reviewed periodically to ensure:

B all content is up-to-date and reflects the latest ‘know-how’for the process,

B all changes and revisions are documented, andB it is well-integrated into other aspects of a safety

management system.

6. Stage Gate Review/Check SystemReaction and reactivity hazards need assessment at everystep from development to commercialization. The focus of

assessment will shift as the project moves from conceptionthrough design and into implementation.A staged review process will allow/drive evaluation ofreactivity hazards and their impact on the safety of theprocess being implemented. The following reviews shouldbe carried out during the development and implementationprocess:

a. Process Development – application of inherently saferprocess principles

b. Project/Design Concept – identification of reaction andreactivity hazards

c. Initial Facility Design – countermeasures for control orelimination of hazards

d. Design Check – detailed review of the intended designto ensure that all reaction/reactivity hazards have beenappropriately addressed

e. Implementation Confirmation – prior to start-up,confirm that all hazards are adequately managed.

7. Experienced Multi-Disciplinary Team InvolvementProcess safety involves many aspects of chemical processingtechnology. All of these aspects are important, and as such,representatives from each much have a seat at the assess-ment team table. Team members need to be highly experi-enced in their field in order to accomplish a thoroughanalysis and implement a robust hazard control strategy.

5. Summary

A culture of operational discipline is important to the safeoperation of a facility; we must be able to count on everyone tocarry out their functions properly and consistently. However, inorder to ensure that the disciplined operating team has the infor-mation necessary to operate the facility safely, the process designfunction must demonstrate a culture of Technical Discipline toensure the process hazards are correctly identified, interpreted,controlled, documented, and communicated.

References

Basf, A. G. (1997 edition). Safety-compliant process development. Issued by the SafetyEngineering and Process Safety Departments.

Berthold, W., Giesbrecht, H., Loffler, U., Maurer, B., & Schiephake, V. (1992 edition).Safety through testing. Process Engineering Department.

Basf, A. G., Stark, Riede, Maurer, & Schoenherr. (December 1998). Safety in dryingprocesses. Process Safety and Safety Engineering Departments.

J. A. Klein, E.I. du Pont de Nemours and Company, Wilmington, DE 19898, Opera-tional discipline in the workplace, Process Safety progress, American Institute ofChemical Engineers (AIChE), Vol. 24 Issue 4, Pages 228 – 235, Published Online:26 Sep 2005, Copyright� 2005.