VERSION 1/3/99

ENVIRONMENT, HEALTH AND SAFETY COMMITTEE

NOTE ON

SAFETY ISSUES IN THE SCALE UP OFCHEMICAL REACTIONS

Enquiries received by the Royal Society of Chemistry indicate that there is a need toprovide some basic guidance on safety issues raised by the scale up of chemicalreactions from laboratory scale to full sized commercial plant. The Society believesthat chemists should be fully aware of the safety problems associated with the scaleup of chemical processes and should contribute to the scale up procedure in order toensure the health and safety of all persons involved.

This Note is designed to provide guidance to members who either managelaboratories or are employed in laboratories where the scale up of chemical reactionsis undertaken. Therefore it assumes familiarity with basic terms such as ‘risk’ and‘hazard’. The Note is not intended to be a full or definitive guide and readers areurged to obtain more detailed information and/or expert advice if this is required.

The Note was prepared by a Working Party of the Environment, Health and SafetyCommittee (EHSC) of the Royal Society of Chemistry. The Society is a registeredCharity. Its Royal Charter obliges it to serve the public interest by acting in anindependent advisory capacity. In order to meet this obligation the members of theEHSC are drawn from a wide range of backgrounds and serve on the committee asindividual experts and not as representatives of their employers.

The EHSC is eager to receive feedback on this Note. Comments can be sent to theCommittee Secretary whose details are given on the last page.

1. INTRODUCTION

Of the chemical incidents reported to HSE each year as dangerous occurrencesunder RIDDOR some 20 to 30 are exothermic runaways. There may be otherchemical incidents which are either not reported or which do not fit the legaldefinition of a ‘dangerous occurrence’. It is likely that many of these areattributable to inadequate scale up procedures.

2

Chemical processes usually originate on a small scale in the laboratory andtheir development is normally carried out by conducting the reactions onsuccessively larger scale before transferring to full size production plant. Inthe early stages of process development limited quantities of chemicals areused and full information about the reaction or side reactions is not usuallyavailable. Risk assessment studies and the scale up procedure enable furtherinformation to be obtained which is relevant to the final process plant design.

Failure to scale up properly and to take appropriate precautions may lead tothe loss of process control which in turn may result in a runaway exothermicreaction and /or the generation or release of toxic materials. There have been anumber of such failures in the past which have led to multiple fatalities, severedamage to property, environmental damage and business loss.

Other problems which have arisen from inadequate scale up proceduresinclude for example :

• dust and vapour explosions inside vessels due to the mishandling of reactantsand solvents

• fires due to overfilling of vessels• failure to correctly assess electrical equipment for use in hazardous areas

identifying all possible sources of ignition• autoignition of flammable vapours, usually resulting in a flash fire.

The first and most critical step in the scale up procedure is to undertake a riskassessment of the proposed chemical process at the concept stage. Whereappropriate this should include a study of the thermochemistry of the proposedreaction.

2. ASSESSMENT OF REACTION HAZARDS

The risk assessment undertaken on all new or modified reactions should berecorded and should involve the consideration, where appropriate, of thefollowing:

• the physical and chemical properties of the reactants, intermediates,products and wastes

• the possibility of hazardous reactions if there are deviations from setparameters

• the potential for the substances in use to corrode plant

• the fire and explosion hazards

• the health hazards and associated COSHH and Personal ProtectiveEquipment (PPE) assessments

• the thermochemistry of both of the desired and potential undesiredreactions, in particular the rate of heat output

3

• the rate of gas generation and the potential for vessel over-pressurisation

• scale-up problems (see below)

• plant design characteristics, including control systems, service needs (egcooling water, inert gas) and the adequacy and reliability of safety devicessuch as pressure relief valves, bursting discs, vents, etc

• the application of hazard identification methods, such as ‘What If’studies,Hazard and Operability Studies (HAZOPS) and quantified risk assessmenttechniques;

• environmental effects;

• new or modified process instructions; and

• the training needs of operators.

The first stage in the risk assessment is data collection. This will enabledesktop screening of the hazards of the proposed reaction to be undertaken andif possible thermochemical calculations to be carried out. In this exerciseconsideration should be given to the reactive group(s) of the reactants andcomparisons should be made with analogous substances or reactions. Anumber of publications are cited in the Bibliography which may be useful inthis desktop screening exercise.

Where there is potential for an exothermic reaction, particularly a runawayreaction, it is essential that a study of the thermochemistry of the proposedreaction is undertaken at an early stage. It is current best practice to subject allnew processes to thermochemical investigations. Various calorimetrictechniques are available which can be used in such a study. For exampledifferential scanning calorimetry or Carius tube studies can be used for basicthermal stability screening tests. Isothermal calorimetry can be used mainly tomeasure reaction kinetics and heats of reaction. Adiabatic calorimetry can beused to examine the potential for runaways. Relief vent sizing tests can also becarried out.

As a result of the risk assessment the principles of elimination, reduction andcontrol should be applied. If the risks associated with the proposed reaction areunacceptable then an alternative route to the product should be considered, ifpossible using inherently safer methods. If however the reaction is to proceedthe risk assessment should specify the control measures and safety procedureswhich should be adopted to ensure the safe operation at all stages of the scale-up process.

3. SCALE-UP PROCEDURE

The design of a commercial plant can be accomplished by scaling up fromlaboratory equipment using pilot plant. In some cases several pilot plants ofincreasing size may be used to effect the best design for the larger plant. Riskassessment is an evolving process as scale-up progresses. At each stage of the

4

scale up procedure the information generated should be used to review the riskassessment to enable a decision to be made as to whether or not to proceed tothe next stage. If the process is to proceed the risk assessment should specifythe controls and operating conditions required.

The rate of a chemical reaction is fixed at any given temperature buttemperature may be influenced by mass transfer and heat transfer, which are inturn affected by the size and design of the reactor. It is not always possible totheoretically assess these effects on a quantitative basis and in such cases it isusual to carry out trials with water or inert substances in the intended plant. Ifthis is not feasible purpose built reactor calorimeters will be required.

Laboratory experiments, carried out in test tubes, small flasks etc., produce arequired chemical or product but do not necessarily indicate side effects of thereaction, i.e. by-products, release of gases or vapours which may be toxic orflammable. Heat releases may be absorbed by the equipment or surroundingsand not noticed. The chemicals used may be pure materials rather than bulkcommercial chemicals, which may contain traces of impurities. Therefore inorder to overcome this problem it is essential that the apparatus, materials andchemicals used at all stages of the scale up accurately reflect those that will beused in the final plant. In the laboratory reactions are usually carried out inglass vessels but the scaled up process may well be carried out in containersmade of other materials. With some reactions such changes may be importantand could result in unexpected reactions or problems including catalytic orinhibition effects.

As scale increases the time required to carry out each operation is also likelyto increase and appropriate allowance should be made for this.

Pilot plant is used to assist in the scale up of the chemical process designrather than the mechanical design. It provides information for economicdesign, operating parameters, and safety considerations. In addition pilotplants can be used in small scale production for evaluation and trial marketing.Information should be obtained from pilot plant studies to confirm thedecisions made as a result of the risk assessment stage in relation to :

• operating conditions;

• design parameters;

• reactor problems, design, materials of construction;

• unit operations problems;

• materials handing and sampling problems;

• thermal instability and other decomposition;

• Phase problems;

• impurities;

• corrosion;

5

• fouling;

• analytical problems;

• operating procedures;

• working and environment problems; and

• effluent and waste disposal problems.

Impurities in the chemicals used can cause many problems which can beidentified in pilot plant studies. Impurities can occur in the feedstock or arisefrom side reactions, decompositions, polymerisations, etc. which can causeunexpected effects. Leaks into the system may bring in unwanted materialssuch as pump lubricant, seal fluids or heat transfer media, including water.These may lead to blockages and other problems. Some impurities cancatalyse undesirable explosive reactions or may be thermally unstablethemselves.

Pilot plant studies can reveal corrosion problems. These can be associated withminor components such as gaskets and diaphragms or with impurities in thereactants.

Scaling up in various size plants can produce variations in reactions, andapparently identical reactors can give slightly different products withapparently the same feed materials and operating conditions. This is mostmarked in fermentation processes.

Pilot plants should be operated by technically qualified and appropriatelytrained competent personnel. The extra unknown or unforeseen hazardsassociated with pilot plant should be compensated for by betterinstrumentation and technical control by the operators. It must beremembered, however, that accidents on pilot plant despite their size can stillhave serious consequences.

It is important, therefore, that scale-up of chemical reactions is done correctlyso that the eventual reactions can be carried out safely in full size productionplant.

4. SAFE OPERATION

Information obtained from the risk assessment of the chemical process and thescale-up studies will enable decisions to be made on the most appropriatecontrols to ensure a safe operation. The main options which could beconsidered for a safe operation are:

• inherently safer methods which use less hazardous materials or alternativereaction routes to eliminate or reduce hazards;

• preventative measures such as process controls using sensors, trips, alarms(temperature, pressure, stirrer, cooling water failure) and other controlsystems which initiate automatic remedial action; and

6

• protective measures such as pressure relief valves, vents, reactioninhibition, crash cooling, drown out/quenching or secondary containmentwhich will limit the consequences of a runaway reaction.

There is no single best option for controls and protective measures that can beapplied in all cases. Whichever option is chosen it must cater for the worstcredible scenario and reduce the risk as far as is reasonably practicable.

The option chosen must also be supported by appropriate occupational healthand safety management systems which cover operating procedures, thetraining and supervision of operators and maintenance personnel, the

maintenance of plant and equipment, the control of plant modifications,software and emergency procedures. It is essential that there is goodcommunication and co-operation between the risk assessment team, theprocess plant team and maintenance personnel so that everyone concerned isfully aware of the operating parameters and limitations of the process plant.

Once the process is operational it is important to adhere to the agreedoperating parameters and procedures. If there is to be any deviation or changein raw materials, further risk assessments should be undertaken.

5. CONCLUSION

It is essential that suitable and sufficient risk assessments are undertaken of allnew and modified reactions during the scale-up of laboratory processes to fullsized commercial plant. Appropriate process controls and protective measuresare needed to reduce the risk of a runaway exothermic reaction and/or thegeneration or release of toxic materials to a level which is as low as reasonablypracticable. This not only ensures that legal requirements are complied withbut also avoids the disruption, cost, potential damage and injuries that can becaused by the loss of control of chemical reactions.

6. BIBLIOGRAPHY

• Royal Society of Chemistry, “EHSC Note on Individual LegalResponsibility for Health and Safety”, 1995.

• Royal Society of Chemistry, “EHSC Note on Risk Assessment at Work”,1997.

• Health and Safety Executive [HSE] leaflet IND(G)254 “ChemicalReaction and the Risk of Thermal Runaway”.

• Association of the British Pharmaceutical Industry [ABPI] “Guidelines forChemical Reaction Hazard Evaluation”, 2nd Edition.

• J G Lowenstein “The Pilot Plant”, Chemical Engineering, Dec 9/23, 1985pp 62-76.

• HSE/Institution of Chemical Engineers [IChemE] video “Control ofExothermic Chemical Reactions”.

• IChemE “Chemical Reaction Hazards – a guide to safety”, Ed J Bartonand R Rogers, 2nd Edition 1998.

7

• “Bretherick’s Handbook of Reactive Chemical Hazards”, 5th Edition, Ed.Urben, Butterworth Heinemann, 1995

• National Fire Protection Association “Manual of Hazardous ChemicalReactions”, 4th Edition, 1971.

• A Starkie and S Rowe ‘Taking the heat’, ‘Chemistry in Britain’, February1996.

• 'Pilot Plants and Scale-Up of Chemical Processes II', Ed. W. Hoyle, RSC,March 1999.

This Note was prepared by a Working Party of the RSC Environment, Health and Safety Committee [EHSC].The members of the Working Party were :

Mr G J Dickes

Mr T G R Farthing

Mr R W Hazell [Secretary]

Dr G V MacHattie [Co-chairman]

Dr R Owen

Mr D M Sanderson

Dr I Wrightson [Co-chairman]

The Working Party gratefully acknowledges the assistance of the following individuals and organisations inthe preparation of this Note, though of course the views expressed remain the responsibility of the RSC :

• The Health and Safety Executive (HSE)

• Dr J Bickerton

• Mr K D Jackson

• Dr J D Jones

• Mr P G Lambert

• Dr I M McRobbie

• Dr A J Starkie

• Dr W Hoyle

This Note is also available on the RSC website http://www.rsc.org

The Environment, Health and Safety Committee welcomes comments on this Note. Comments shouldbe sent to the Committee Secretary:

Mr R W HazellHealth, Safety and Environment OfficerThe Royal Society of ChemistryBurlington HousePiccadillyLondon W1V 0BN• Direct Tel : 0171 440 3337• Switchboard : 0171 437 8656• Fax : 0171 437 8883• E-mail : hazellr@rsc.org

Web Address : http://www.rsc.org or http://www.chemsoc.org

1

1 l:\ehscnote\Scale up of Chemical Reactions 1 March 99