Safety first

C&I Issue 11, 2016

During the 1960s to 1980s, ICI personnel were responsible for major innovations in chemical process safety, which ultimately affected the safety culture in all sections of the chemical industry. This change improved the quality and robustness of operating processes, not only resulting in improved profits, but also saving lives.1

Industrial safety has  been of importance for longer than most people would imagine. One of the first reports is in an extract from the Canterbury Tales by Geoffrey Chaucer in 13862, included in the Canon Yeoman’s Tale:

It happens, like as not,
There’s an explosion and good-bye the pot!
These metals are so violent when they split
Our very walls can scarce stand up to it.
Unless well-built and made of stone and lime,
Bang go the metals through them every time and some are driven into the ground –
That way we used to lose them by the pound-
And some are scattered on the floor;
Some even jump into the roof, what’s more........

I’ve no idea why things went wrong;
Recriminations though were hot and strong.
“Well” said my lord “there’s nothing more to do.
I’ll note these dangers for another brew:
I’m pretty certain that the pot was cracked,
Be that as it may, don’t gape! We’ve got to act.
Don’t be alarmed, help to sweep up the floor
Just as we always do, and try once more!”

But six centuries later, the attitude to safety and accident investigation had changed very little, the emphasis still being on blame!

In the introduction to his well-known book3, published in 2003, Still Going Wrong! Case Studies of Process Plant Disasters and How They Could be Avoided, Trevor Kletz, formerly of ICI,  provocatively suggests that ‘a high price was paid for the information in this book; people were killed or injured and billions of dollars worth of equipment was damaged. Someone has paid tuition fees. There is no need for you to pay them’.

Sadly, although his messages were widely known and distributed in the period 1970-2010, plants are still going wrong, particularly in Asia, and loss of life and equipment damage is still all too common. The legacy he left from developing innovative ideas whilst a technical safety officer at ICI in the 1960s-1980s, and sharing them with the rest of industry, is still strong in many European and American companies, but unfortunately has never been learned or has been forgotten by many organisations in the rush for profits.

One of Kletz’s mantras was to ensure that lessons from accidents should never get forgotten, since nearly every incident that occurs has happened before. He used to say to people following an accident, ‘don’t write a report, I have  it on file already’.

Kletz began life as a chemist but after joining ICI in 1944, he spent time in R&D and as a plant manager, where he was transformed into a chemical engineer4. After experience in process investigation and plant commissioning, he developed an interest in process safety, a term not widely used at that time, and in 1968, he was appointed the first technical safety advisor in ICI, but with responsibility only for  ICI’s Heavy Organic Chemicals (later renamed Petrochemicals) Division manufacturing facilities at Billingham and Wilton in North East England.

The reason for his appointment was ICI’s poor process safety record in the 1960s, with serious fires occurring in 1965, 1966 and 1967, all resulting in loss of life. In the past, safety was left to elderly foremen, retired army officers and graduates in non-technical subjects but their focus was on trips, spills and falls, rather than process safety.

In 1968, ICI decided that it was time to appoint a technical safety advisor with a strong background in science or engineering and wide experience of plant matters. Within a few months of Kletz’s appointment to this new role, there were two major incidents occurring on other ICI sites, and Kletz was involved in the incident investigation. It became clear to him, and to some senior managers, that there was a need for change in the whole company, in particular a change in culture towards process safety. Out of this change in attitude, several innovations developed within ICI and these have been magnanimously shared with the rest of industry, to the benefit of all.

Hazard Analysis (Hazan)

Hazan, now often known as Quantitative Risk Assessment, is the application of numerical methods to obtain an understanding of hazards in terms of a) how often a hazard will manifest itself and b) with what consequences for people process and plant5.

Although the methodology started in the nuclear industry in the 1960s, ICI was the first company to apply the technique in the chemical process industry. It is now widely used in all industries under the guise of Quantitative Risk Analysis, which has led to techniques such as Failure Mode and Effects Analysis (FMEA) and Fault Tree Analysis (FTA).

Hazard and Operability Studies (Hazop)

HAZOP is a structured and systematic examination of a complex planned or existing process in order to identify and evaluate problems that may represent risks to equipment or personnel5. The idea of HAZOP is to review the design/process to identify issues that would otherwise not be found; it is carried out by an experienced multidisciplinary team, which breaks the complex design/process into simpler parts (nodes) that are individually reviewed. It is a qualitative technique requiring intuition and good judgement from the team.

The methods were developed at ICI in the early 1960s, when a team of three people met for three days/week for four months to study the design of a new phenol plant. Starting with a critical examination of the whole process, the team looked for possible deviations and the possible consequences.

Gradually this methodology, which ultimately became HAZOP, was refined within the company and applied to all processes. It was not until 1974, however, that courses on HAZOP were offered outside ICI, by the Institute of Chemical Engineers with ICI tutors, and, coming shortly after the Flixborough disaster at the Nypro plant in the UK, the first course was fully booked.

Trevor Kletz published the first book on the subject in 1983 entitled HAZOP and HAZAN; Notes on the Identification and Assessment of Hazards and this helped to spread the word. A fourth edition5 came in 2006, when he was in his 80s and still active as a guru and consultant.

The technique is now widely used throughout all industries and recommended by most authorities involved in health and safety.

Inherent Safety and Safe Design

Kletz has suggested that the official enquiry on the Flixborough explosion in 1974, from which many lessons were learned, omitted the most important one, which was also missed by many commentators. He suggests that the explosion was so big because the inventory of flammable materials in the plant was large (several hundred tonnes), partly because only 6% of the raw material fed to the reactor was used and the rest was repeatedly recovered and recycled.

Kletz argues that the best way of preventing such large explosions is to improve the process so that the inventory of the process is minimised. If this is not possible or reasonably practicable, maybe one could substitute less hazardous materials or use the hazardous materials in the least hazardous way. These are the principles of inherent safety, in which the terms minimisation, substitution and attenuation/moderation are used to design inherently safer processes. These concepts also lead to process intensification, which became a popular concept in the 1990s onwards.

ICI in the 1970s made inherently safer design into a procedure that was applied to every project and allowed Kletz to publish an article in C&I on the subject, entitled What You Don’t Have, Can’t Leak.6 Subsequently, after he left ICI, he published books on the subject; the most recent of those in 2010.

The principles of inherent safety have also been taken up by chemists in the design of green syntheses and processes by avoiding the use of hazardous reagents and reactions, and avoiding processes that are hazardous to scale.

Batch and Semi-batch

Although the work of Kletz and his colleagues at the Billingham and Wilton sites of ICI in the North-East UK made great contributions to industrial process safety, the emphasis was on the continuous processes operated on those sites. In the early 1970s, after a series of explosions and near misses, both in ICI and other similar companies in Europe and the US, there was a recognition that the existing approach to safety of batch processes was deficient. 

Many of the incidents were concerned with nitration processes or with the subsequent conversion of the resultant nitro compounds to amines by reduction, often involving hydrogenation with heterogeneous catalysts. Many of the amines were converted to unstable diazonium compounds, which were coupled with complex aromatics to give a wide range of useful and stable coloured compounds. These products were particularly prevalent in the dyestuffs and colour chemicals applications used in textiles, pigments for plastics,  in the photographics industry or occasionally as food dyes.

In batch processes, one of the dangers is when the vessel is full and the contents become overheated, which on occasion, particularly with nitro compounds, can lead to decomposition of the product with release of heat. It can then be imposssible through cooling measures to control the exotherm and the reaction runs way with build up of pressure at increasing temperature. If the pressure relief system cannot cope, the vessel may rupture releasing the contents, sometimes with explosive force. It is then that loss of life can occur, as well as considerable damage to equipment.

Another dangerous situation is when reagents are added to a vessel, for example, nitric acid in the case of nitrations, but the reaction does not start for some reason; the temperature may be too low or the agitator is not working. When the operator tries to perform a correction, the heat generated by the reaction is too great for the vessel coolant to cope and the reaction again runs out of control.

These runaway reactions caused by the accumulation of unreacted reagents and starting materials were recognised in the 1970s to be a major cause of incidents in the batch process industries.

Building on the principles of Kletz at ICI Petrochemicals Division, a group under the direction of Norbert Gibson at ICI Organics Division in Blackley, Manchester was formed in the early 1970s to elucidate chemical reaction hazards with a focus on batch processes.8 The group, the Hazard and Process Studies Group, recognised that a key part of safety is in understanding the chemical reactions and processes, particularly the kinetics and rates of the desired reaction as well as the reactions that produce side products. The stability of the desired product and side products to heat was a major emphasis of their studies as well as decompostion reactions if the reaction became stressed. The group often had to build its own equipment to study these processes and devise novel ways to measure parameters whilst reactions were running away or decomposing. The safety testing itself could therefore be hazardous if not carried out correctly.

Shortly after the group was set up in the early 1970s, I joined ICI at the Corporate Laboratory in Runcorn, Cheshire, working in R&D on chemicals for electronics. Shortly afterwards, I became a part-time safety advisor for the 400-strong laboratory, and then became aware of the work being done in Gibson’s group at Blackley.

After a few years at Runcorn, I transferred to Blackley to work on development and scale up of agrochemicals and it was during discussions regarding the safety of scale up of some novel diazonium chemistry I had developed that I first hand encountered the Hazard and Process Studies Group. Here I learned so much that is not taught to chemists in university about the hazards of  scale up. Sadly it is still not taught in many chemistry schools, but chemical engineers usually get a grounding.

So when I left ICI in 1979 to join Smith Kline and French to study the development and the scale up of pharmaceutical batch processes, I was able to introduce some of the ICI principles of Hazard Testing into the new organisation. In fact, when I left SK&F in 1989 to set up my own scientific training company Scientific Update, the first course on Chemical development and scale up that I taught included a module on process safety and hazard evaluation.

By 1989, of course, many of the principles developed by ICI, and by other companies such as Ciba-Geigy, Sandoz, Hoechst, Dow etc, were becoming better known through the presentation of case studies at events mostly organised by the Institute of Chemical Engineers.

In the 1970s, the UK’s Health and Safety Executive (HSE) had also become interested in the topic and set up a unit devoted to Fire and Explosion Hazards at Buxton, Derbyshire. John Barton of HSE Buxton, along with Richard Rogers of ICI’s Process Hazard Studies group began to teach a course in chemical reaction hazards for the IChemE and these principles later became enshrined in an important book on the subject, published in 1993 by IChemE.2

So although the principles of Chemical Reaction Hazard Analysis were known in the 1970s, they were not so widely disseminated, particularly amongst chemists.

It was not until the availability of automated reaction calorimeters, such as Mettler’s RC1 in the late 1980s, that chemists began to realise that they could carry out easily measurements, such as heats of reaction,  whereas previously cumbersome Dewar flasks had been the order of the day. Scientists working on new processes could now begin to evaluate the thermochemistry and hazards of scaling up prior to venturing on to kilogram and tonne scale, learning a lot more about the process and its kinetics that were useful in process development.

Since the publication of Barton and Rogers in 1993, many more books have been published in this area and the subject has advanced into an important new discipline. Nevertheless, in some companies, the pressure to get new products on the market quickly means that corners get cut and it is sad that I still hear of many chemical runaways and explosions, leading to loss of life as well as expensive equipent loss, which occur, not only in developing countries, but also in the US and Europe.

Thus a widely publicised explosion in the USA in the 1990s9 during a reaction with a nitro compound designed to produce a fuel additive could have been avoided if the known principles of reaction hazrd analysis had been followed.

In this explosion, it was clear the company had a lack of knowledge of the stability of its own product and it was also apparent that  the process for many years had been operating on the edge.

Even some earlier near misses had not given this company a warning to carry out further safety evauation of the process. Sadly, a process operator lost his life because of the lack of understanding of the company’s process by senior technical management.

For this reason, the continuing education in safety of chemists and engineers, which began in the 1970s in ICI under Kletz’s influence, is still relevant today and is being carried out, not only by professional bodies such as IChemE, but also by private companies too. The latter group often think more internationally and educate all over the world, including tailored in–house course for specific companies.

In conclusion, the legacy of ICI’s innovations in process safety is immense and the chemical industry should be grateful for ICI’s willingness to share these innovations by their publication in books and  scientific journals and magazines.


1 ICI’s contribution to process safety, T.A. Kletz, Hazard XXI Symposium Series, (I Chem E) No 155, 39-42, 2009

2 This extract first came to my notice when reading the excellent text Chemical Reaction Hazards: A Guide to Safety, J.A Barton and R.L Rogers, 2nd Ed, 2004, I Chem E(Rugby).

3 Still Going Wrong: Case Histories of Process Plant Disasters and How They Could Have Been Avoided, T.A. Kletz, 2003, Oxford, GPP/ Elsevier

4 Chemical Engineers Who Have Changed the World; A Lifetime Spent Saving Lives, The Chemical Engineer, Oct 2012, 53-56

5 HAZOP and HAZAN: Identifying and Assessing Process Industry Hazards, T.A. Kletz, 4th Ed, 2006, I Chem E, Rugby.

6 T A Kletz, Chemistry and Industry, 1978, 6th May, 278

7 A Chemical Approach to Inherent Safety, R. L Rogers and S Hallam, I Chem E Symposium Series No 124, 235-241, 1991.

8 Chemical Reaction Hazards, An Integrated Approach, N Gibson, R.L Rogers and T.K Wright, I Chem E Symposium Series, NO 102, 61-83

9 document/Morton_Report. pdf

Become an SCI Member to receive events discounts

Join SCI