Hazop & Hazan

Identifying and Assessing Process Industry Hazards, Fouth Edition

Trevor A. Kletz

• Éditeur: CRC Press
• Année: 2018
• ISBN:
  9781351441346
• Édition: 4
• Sujet: Tecnologia e ingegneria
• Sous-sujet: Ingegneria chimica e biochimica

This revised edition provides the basics of applying hazard and operability study (Hazop) and hazard analysis (Hazan). Hazop is a creative but systematic method of identifying hazards in process plants. Hazard analysis is then used to quantify the risks from these hazards, and to assess how far to go in reducing them. This book is presented in easy-to-read style and explains: what a Hazop is, who carries it out, when, and how long it should take; points to watch during a Hazop; an example of a Hazop; Hazops on flowsheets; the stages of Hazard analysis; the Fatal Accident Rate; risks to the public; estimating how often an accident will occur, with examples; and pitfalls in Hazan.

Hazard identification and assessment ‘The great end of life is not knowledge but action.’ T.H. Huxley (1825–1895)

1

1.1 Introduction

The techniques for identifying hazards — for finding out what hazards are present in a plant or process — and the techniques for assessing those hazards — for deciding how far we ought to go in removing the hazards or protecting people from them — are often confused. Figure 1.1 may help to make the differences clear.

The left-hand side shows some of the methods used for identifying hazards — and problems that make operation difficult.

Some hazards and problems are obvious. For example, if we manufacture ethylene oxide by mixing oxygen and ethylene close to the explosive limit we do not need a special technique to tell us that if we get the proportions wrong there may be a big bang.

Figure 1.1 Methods of identifying and assessing hazards

The traditional method of identifying hazards — in use from the dawn of technology until the present day — was to build the plant and see what happens — ‘every dog is allowed one bite’. Until it bites someone, we can say that we did not know it would. This is not a bad method when the size of an incident is limited but is no longer satisfactory now that we keep dogs which may be as big as Bhopal (over 2000 killed in one bite) or even Flixborough (28 killed). We need to identify hazards before the accidents occur.

Check-lists are often used to identify hazards but their disadvantage is that items not on the list are not brought forward for consideration and our minds are closed to them. Check-lists may be satisfactory if there is little or no innovation and all the hazards have been met before, but are least satisfactory when the design is new.

For this reason the process industries have come to prefer the more creative or open-ended technique known as a hazard and operability study or Hazop. It is described in Chapter 2. It is now widely used on designs for new plants and plant extensions but, because of the effort involved, has been less widely used on existing plants.

Samuel Coleridge described history as a ‘lantern on the stern’, illuminating the hazards the ship has passed through rather than those that lie ahead. It is better to illuminate the hazards we have passed through than not illuminate them at all, as we may pass the same way again, but we should try to see them before we meet them. Hazop can be a lantern on the bow.

Unfortunately we do not always learn from the hazards we have passed through, but that is outside the scope of this book^1,2.

Other methods of identifying hazards are described in Lees, Chapter 8, and are summarized in Section 2.11, page 54.

After we have identified the hazards we have to decide how far to go in removing them or in protecting people and property. Some of the methods used are listed on the right-hand side of Figure 1.1. Sometimes there is a cheap and obvious way of removing the hazard; sometimes our experience or a code of practice tell us what to do. Sometimes it is less easy to decide. We can then try to work out the probability of an accident and the extent of the consequences and compare them with a target or criterion. This method is called hazard analysis or Hazan in this book. Sometimes a five-minute estimation is sufficient. On other occasions detailed studies can take many weeks.

Hazop can and should be applied to all new designs, unless we are making an exact copy of an existing plant which has been proved satisfactory, as we need to know all the hazards and all the problems that can prevent efficient operation. Hazan on the other hand should be used selectively — there are neither the need, the data nor the resources to attempt to quantify every problem on every plant. Carling³ has described a Hazop which produced 326 recommendations of which only seven justified a detailed hazard analysis.

In the development of a design the Hazop comes first. We identify the hazards and the problems that prevent efficient operation and then decide what to do about them. However, if there is an obvious major hazard we may start on the Hazan before the Hazop is carried out. In a Hazop the operability part is as important as the hazard part. In most studies more operating problems are identified than hazards.

Table 1.1 The differences between Hazop and Hazan

Hazop	Hazan
Identifies hazards	Assesses hazards
Preferred technique: use on every project	Selective technique: use when others fail
Qualitative	Quantitative
Done by a team	Done by one or two people
Also called:	Also called:
‘What if?’	Risk analysis
	Risk assessment
	Probabilistic risk assessment (PRA)
	Quantitative risk assessment (QRA)

Hazop and Hazan are often confused, and Hazop is sometimes used to describe any technique for identifying hazards. Figure 1.1 and Table 1.1 should make the difference clear. However, if someone asks you to carry out a Hazop or Hazan on a design, first make sure that the questioner is clear on the difference between them and is using the terms correctly.

The techniques described in later chapters are sophisticated techniques which enable companies to use their resources more effectively. They assume that the general level of management is competent, that the plant will be operated and maintained in the manner assumed by the design team and in accordance with good management and engineering practice. In particular they assume that protective systems will be tested regularly and repaired promptly when necessary.

If these assumptions are not true then Hazop and Hazan are a waste of time. It is no use identifying hazards or estimating their probability if no-one wants to do anything about them; it is no use installing trips and alarms if no-one is going to use or maintain them. The time spent on Hazop and Hazan would be better spent on bringing the safety consciousness of employees and management up to standard. The following is a summary of a paper by Atallah and Guzman on doing this in developing countries⁴ (and perhaps elsewhere):

• Be patient when you are waiting for data, prompt when asked for advice.

• Include in your team someone who speaks the local language.

• Submit your report in draft for comment; justify your criticisms and recommendations.

• Photograph problem areas.

• Visit the plant at night.

• Wear all the required protective clothing and follow all the safety rules.

• Expect to be asked about subjects not covered in the remit.

• Provide the client with copies of references, codes, and so on, not just a list of them.

• Involve the client in your audit.

• Learn as much as you can beforehand about the customs and culture of the country, expect a cultural shock and do not discuss politics or religion.

If you wish to introduce Hazop and/or Hazan into an organization in which they have not been used before, you should start small. Do not try to set up a large team capable of studying all new and existing designs. Instead apply the methods to one or two problems. If your colleagues find that the methods are useful they will ask for more and the use of the techniques will grow. If, on the other hand, the methods do not suit your organization, little has been lost.

Despite all our efforts we shall fail to foresee every hazard and some will result in accidents. We should learn from these accidents, not only from those that result in serious injury or damage but also from those that do not — for example, leaks that do not ignite. If these ‘near-misses’ are not investigated and the lessons made known to those concerned, next time injury or damage may result.

In my former company, ICI, Hazop and Hazan form part of a series of six hazard studies carried out on new projects as they progress⁵. They are:

(1) Exploratory phase — identification of basic hazards and assessment of suitability of possible sites.

(2) Flowsheet phase — identification and assessment of significant hazards, using Hazan.

(3) Detailed design — Hazop.

(4) Construction — a check that decisions made in earlier studies have been implemented.

(5) Commissioning — final inspection.

(6) Post-commissioning — safety audit and review of modificati...