Download Health and Safety Services

Safety Services
Guidance on risk assessments for chemical reactions –
interpreting journal protocols,
changing parameters and scaling up
Introduction
1. Chemists and other researchers routinely refer to journal accounts of
experimental protocols carried out by others to inform their current work. In
designing their own protocols, they interpret these often very brief accounts;
amend and vary key parameters, and scale up quantities and volumes. The
original accounts may have been based on acceptable practices and scientific
knowledge many years old, and by institutions with various risk appetites,
complying with very different national legislative requirements. Competent and
informed risk assessments are the key to converting this information into a safe
system of working today; Principal Investigators are responsible for ensuring
assessments are carried out, and kept up-to-date.
2. There have been a few serious incidents at HEEs involving run-away exothermic
reactions and/or the uncontrolled release of toxic substances. Some of these are
known to have arisen from errors in interpreting data from others and scaling up.
Interpretation of journal protocols
3. Journal accounts seldom provide a step-by-step guide to all the practical
considerations that need to be determined. They often do not include safetyrelevant information such as size of vessel, or size of stirring device. They may
not explain the reason for the selection of a particular solvent, or know that choice
of a different solvent, perhaps offering better solubility, may completely alter the
characteristics of the reaction.
Changing reaction parameters
4. The adjustment of journal procedures to suit a different research direction can
have a devastating impact of the overall character of the reaction. In one
example, a research student changed the solvent of his reaction from
dichloromethane to toluene. Toluene is less dense than water; DCM is not.
Hence the heat transfer characteristics of this particular 3 phase system altered
fundamentally giving rise to local hot spots and detonation of hydrogen azide.
Scaling up
5. Increasing the volume /weight of reagents without considering the effects on a
wide range of reaction parameters is known to be a factor in many catastrophic
reaction failures, even at relatively small scale.
6. The main concerns include :
Page 1 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
•
runaway reactions
•
over-pressurisation
•
dust and vapour explosions inside vessels
•
fires due to over-filling of vessels
•
failure to identify all possible sources of ignition (including static)
•
failure to assess the ATEX 1 requirements for electrical equipment (refers to the
degree of intrinsic protection of electrical equipment used in potentially
explosive atmospheres
•
auto-ignition of flammable vapours
•
inadequate mixing
•
inadequate removal of heat
7. The Health & Safety Executive’s analysis 2 of the main causes of chemical reactions
that run out of control lists the following:
•
inadequate understanding of the process chemistry and thermochemistry;
•
inadequate design for heat removal;
•
inadequate control systems and safety systems; and
•
inadequate operational procedures, including training.
Initial assessments
8. For any chemical reaction, there are numerous risks to consider and many of
these will need to be included in your COSHH assessments and general
laboratory/ safe systems of working. Remember that risk assessments must
include hazards that arise from the chemical reaction itself, most importantly:
•
thermal instability of any or all of the reactants, reactant mixtures and
products, including intermediates, contaminants and by-products
•
exothermic reactions leading to increased temperatures and production of
decomposition reactions or violent boiling, and
•
evolution of gas in sealed systems, leading to pressurisation.
Sources of useful information for initial assessment include literature survey,
predicting reactivity and stability from molecular structure and knowledge of the
thermochemistry involved. HSG143 3 provides more guidance on each of these
(pp16-18 in particular). Colleagues from other academic areas often possess
expert knowledge about aspects of chemical reactions. For example, chemical
engineers will be more familiar with the issues of scaling up, mixing, heat transfer
than those used to working at a smaller scale.
1
2
3
http://www.hse.gov.uk/electricity/atex/general.htm
See http://www.hse.gov.uk/pubns/indg254.htm
Health & Safety Executive, 2000, (pp16-18 in particular)
Page 2 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
9. Certain molecular groupings should prompt more careful consideration. These
include:
•
double and triple bonded hydrocarbons
•
epoxides
•
hydrides and hydrogen
•
metal acetylides
•
many nitrogen-containing compounds, eg amides, imides, azides, diazo,
diazeno compounds, hydrazine-derived N compounds, nitrates, nitrites,
nitroso- and nitro- compounds, N-metal derivatives (and more)
•
oxygenated compounds of halogens
•
peroxides.
Check list for scale up from laboratory to pilot plant (> 2 litres, > 100 g)
10. The following is based largely on information from the American Institute of
Chemical Engineers
http://www.aiche.org/uploadedFiles/CCPS/Publications/SafetyAlerts/CCPSAlertChe
cklist.pdf This lists the following key points to establish as part of the risk
assessment process (more information about each is provided on the webpage):
•
Know the heat of reaction for your intended reaction, and any other potential
reactions.
•
Calculate the maximum adiabatic temperature for the reaction mixture.
Assume a worst case scenario of no heat removal and 100% reactants actually
react. If this temperature exceeds the boiling point of the mixture, then in
theory at least, the vessel could become pressurised.
•
Determine the stability of all individual components of the mixture at the
maximum adiabatic temperature (worst case), to identify possible
decompositions. Could the decompositions produce gaseous products, again
leading to pressurisation of the reaction vessel?
•
Understand the means of heat transfer within the mixture (often 2 or 3
phases) and heat removal. Consider agitation, introduction of energy by the
mixing technique, means of monitoring (is temperature representative of the
mixture, could hot spots occur, can a thermometer be seen accurately through
coolant?)
•
Avoid using a single temperature as the only means of monitoring reaction
rate, especially for exothermic reactions. The monitoring system is vulnerable
to mechanical failure or operator error. Introduce multiple monitoring
positions to be representative of the whole system.
•
Identify possible reaction contaminants and their effect on the reaction.
Include ubiquitous ones such as water, metal ions such as Na, Ca and others
usually present in process water. Eg azides react explosively with heavy metal
ions.
•
Ask “what if” questions – what if the water/electricity/heating/ etc fails? What
if one reactant quantity is doubled by mistake?
•
Page 3 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
•
What if you increase the scale of reaction? As scale, and the ratio of volume
to surface area, increases, cooling may become inadequate. Heat generation
increases with the volume of the system – by the cube of the flask diameter.
Heat removal capability increases with the surface area of the system, because
heat is generally only removed through an external surface of the reactor. Heat
removal capability increases with the square of the linear dimension (not the
volume). Heat removal is less efficient in larger vessels.
•
What if the solvent is changed? What effect will the change in density, vapour
pressure have? What is the effect of temperature on the different solvent?
•
Identify all heat sources connected to the reaction vessel and asses their
maximum temperature (what if the thermostat on the heating plate is
defective?)
•
Determine the minimum temperature which could occur, and what hazards
might arise? If increased viscosity or freezing could occur, what happens to
the process of heat transfer?
•
For larger scale reactions, consider the potential for temperature gradients, eg
higher near the point of mixing?
•
Understand the rate of all chemical reactions that might occur – how quickly
reactants are consumed and how the rate of reaction increases with
temperature.
•
Consider possible vapour phase reactions eg vapour phase decomposition of
materials such as organic peroxides.
•
Consider the properties of reactants and products, both intended and
unintended, and whether they could degrade or attack materials such as the
reaction vessel, tubing and pipework, gaskets, etc
Check list for scale up in laboratory setting (bench top, < 2 litres, < 100g)
11. It may not be practicable to implement the above check list fully for smaller bench
top scale-ups, but the same principles should be borne in mind. The following
check list however should be reasonably practicable in most circumstances. If
any measure cannot be implemented, the risk assessment should document the
reason, and steps taken to address the risks from uncertain or unknown factors in
other ways.
•
Calculate changes in bond strengths between reactants and products which can
predict whether strong exotherms are to be expected.
•
Ensure that when a new procedure is being trialled, work is carried out on
small scale initially
•
Consult journals such as Organic Syntheses (available on line and free) or
Inorganic Syntheses (not yet in electronic format but available in the JRULM)
to ascertain if there is a tested procedure similar to the one to be attempted.
In these journals, the apparatus is detailed, there are always safety notes and
it is rare to be unable to find a model system. When a system is found it is
better not to mix methods – keep to solvents with similar properties to those
used in the published methods and appropriate to that scale.
•
Do not scale up ‘non-hazardous’ reactions more than five times that of a
previously safe and successful reaction.
Page 4 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
•
Scale up of hazardous reactions should be carried out in smaller steps.
(Hazardous reactions are defined here as exothermic and /or heterogeneous
(involving more than one phase) reactions.
•
When carrying out the trial reaction, note all physical aspects of the reaction,
including colour and phase changes, and in particular, temperature changes
within the reaction medium using an internal thermometer, preferably digital.
It is also a good idea to monitor the temperature of any cooling or heating
baths which can indicate significant thermal events during the reaction (thus it
is preferable to work with the initial trial reaction vessel in a heating/cooling
bath rather than a heating mantle to enable these changes to be noted).
•
When the chemical risk assessment is being carried out, consider the impact of
scale on the potential exposure to toxic, mutagenic and carcinogenic chemicals
(those using the chemical risk assessment pro forma for the Chemistry building
and the MIB will be taking this into consideration if the form is completed
correctly). Sources of data include Broderick’s Handbook of Reactive Chemical
hazards and Material Safety Data Sheets (MSDS).
•
Consider the practicalities of large scale disposal of chemical residues as well
as methods for containing and clearing up any potential spillages.
•
In an exothermic reaction, be aware that the heat produced in reaction mass
increases with volume, i.e. is proportional to the cube of the reaction vessel
diameter; whilst the heat lost to the surroundings depends on surface area of
the vessel available for heat transfer, and i.e. is proportional to the square of
the reaction vessel diameter. For example 4, the time taken for a 1 degree
drop in temperature due to natural cooling from an initial temperature of 80oC
when the surroundings are 20oC:
•

10 mL test tube – 10 seconds

100 mL glass beaker – 20 seconds

1000 mL glass dewar – 60 seconds
Be aware of both the mechanistic and physicochemical nature of the reaction,
including unavoidable or potential side-reactions, which may vary depending
on the purity of the starting materials (note bulk chemicals often contain more
impurities than laboratory grade chemicals, mild steel contains more impurities
than stainless steel etc.).
Practical Tips
12. Mixing
4
•
Without having baffles in a flask or beaker, fast stirring doesn’t actually mix
the contents.
•
Magnetic stirrers should not be used on large scale as they will not stir
efficiently and may damage the flask.
•
An overhead (preferably compressed air) stirrer should be used for reaction
volumes of 2 L or more.
•
For larger reactions slower agitation will provide adequate mixing compared to
that used on a small scale.
Adapted from HSG 143, para 50
Page 5 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
•
Overhead stirrers are essential for heterogeneous reactions which are heated
to reflux.
13. Apparatus
•
With larger volume quick fit flasks, the minimum ground glass joint size should
be B29 size or greater (this makes it easier to remove solid products and to
prevent condensers from flooding.
•
Do not overfill flasks. Calculate the total volume of all reagents plus solvents,
including quench solutions before deciding flask size – the total volume of
reactants should not exceed 50% of the flask volume.
•
A large bore condenser (B.24 or more) must be used for reactions heated to
reflux to avoid flooding of the condenser.
•
When using a solvent bath cooled with cardice, industrial methylated spirit
(IMS) is preferable to acetone (subject to the temperature required) as
addition of cardice to IMS causes less foaming.
•
Some exothermic reactions with long initiation times, e.g. nitrations can be
better scaled up if the reaction is run at elevated temperature so that when
reactants are added in aliquots they are consumed at a faster rate and the
initiation period is reduced. However, during aliquot addition of a reagent
always allow the reaction temperature to return to a predetermined level
before introducing more material.
•
For large flasks, separating funnels and chromatography columns, standard lab
bosses and clamps are inadequate and more robust support is required.
•
Catheters (flexi-needles) should be used to transfer, under inert gas pressure,
air/moisture sensitive liquids from tared bottles to the reaction vessel. The use
of a suitable pressure regulator to control the inert gas pressure is required to
allow a steady and controllable flow of liquid from one vessel to another.
•
When transferring viscous solutions or suspensions do not use a normal
cannula, instead use a wide bore glass/plastic tubing device. Use as wide a
bore needle for the argon inlet as possible in order to reduce the pressure of
inert gas to transfer the fluid.
•
Do not leave a fluid which is being transferred by cannula unattended. Always
ensure the vessel from which you are transferring the liquid is positioned below
the reaction flask as otherwise as soon as a flow is established the transfer will
continue by a siphoning effect regardless of whether there is a positive
pressure of inert gas applied or not.
Thermal runaway
14. An exothermic reaction can lead to thermal runaway, which begins when the heat
produced by the reaction exceeds the heat removed. The surplus heat raises the
temperature of the reaction mass, which causes the rate of reaction to increase.
This in turn accelerates the rate of heat production. An approximate rule of thumb
suggests that reaction rate - and hence the rate of heat generation - doubles with
every 10°C rise in temperature.
15. Thermal runaway can occur because, as the temperature increases, the rate at
which heat is removed increases linearly but the rate at which heat is produced
increases exponentially. Once control of the reaction is lost, temperature can rise
rapidly leaving little time for correction. The reaction vessel may be at risk from
Page 6 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
over-pressurisation due to violent boiling or rapid gas generation. The elevated
temperatures may initiate secondary, more hazardous runaways or
decompositions.
16. If thermal runaway is a possibility, the risk assessment must address all
reasonably practicable means of eliminating the risk though process controls, but
will probably also have to specify precautionary measures to minimise the damage
caused by such an event. These might include:
•
designing the equipment to contain the maximum pressure – eg fit emergency
relief vents and ensure vented material goes to a safe place;
•
having a facility to crash cool the reaction mixture if it moves outside set
limits;
•
being able to add a reaction inhibitor to kill the reaction and prevent runaway;
or
•
having a dump of quenching fluid to hand (and the means to move the reaction
to it!)
Training & information
17. All those at risk from the reaction should be given sufficient information about the
risk to be able to respond appropriately. The researchers directly involved will
obviously need to know most about setting up the procedure, understanding the
reaction thermodynamics (quantifying where necessary) and should receive
information and training in responding to emergencies or non-planned outcomes.
Those working nearby should also be given information to enable them to take the
safest course of action in the event of an emergency.
18. Training & information given should be recorded, either as part of the risk
assessment, or in laboratory notebooks.
Acknowledgements
Safety Services acknowledge with thanks the assistance of Dr Elaine Armstrong, safety
advisor in the School of Chemistry, in the preparation of this guidance.
References and Sources of Further Information:
The Royal Society of Chemistry Environment, Health and Safety Committee, has
produced a very informative guidance note on “Safety Issues in the Scale-up of Chemical
Reactions”, at
http://www.rsc.org/images/Scale%20Up%20of%20Chemical%20Reactions190707_tcm1
8-97905.pdf All researchers and their supervisors should refer to this document whilst
preparing their risk assessments.
Designing and operating safety chemical reaction processes, HSG143, Health & Safety
Executive 2000 (available from JRUL via the A-Z of Electronic Databases, select O for
Occupational Health & Safety Information Systems).
http://www.hse.gov.uk/foi/internalops/fod/oc/400-499/431_15.pdf Internal HSE advice
and guidance to its inspectors on exothermic reactions.
Page 7 of 8
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor
Bretherick’s Handbook of reactive chemical hazards, 7th ed, Urben & Pitt, 2007 ISBN 0
123725831
Document control box
Guidance title:
Guidance on risk assessments for chemical reactions - interpreting
journal protocols, changing parameters and scaling up
Date approved:
Version:
Supersedes:
Previous review
dates:
Next review date:
Related Policies:
Related
Procedures and
Guidance:
Related
information:
Issued by Safety Services Feb 2012
1.1
v1.0, Mar 2009
n/a
Policy owner:
Lead contact:
Head of Safety Services (Dr M J Taylor)
Head of Safety Services (Dr M J Taylor)
Page 8 of 8
Upon significant change
Health & Safety Policy
A-Z of documents on specific health & safety topics, at
http://www.campus.manchester.ac.uk/healthandsafety/CoPs&Guidance.htm
Safety Guidance
Version 1.1
Lead contact: Melanie Taylor