5.2.1 The Team Leader/Facilitator

The selection of a team leader should be made at an early stage of the planning for a HAZOP study. An essential role of the team leader is to ensure the HAZOP methodology is used effectively and productively and so the leader needs to have a deep understanding and considerable experience of HAZOP studies. HAZOP study teams work best when there is clear leadership from an experienced leader. Desirable attributes for the leader include:

Independence from the project allows the leader to stand back and be able to take an objective, unbiased view. The leader acts, and is recognized to act, as an impartial person within the team. Freedom from other responsibilities is important because of the multiplicity of tasks already resting upon the leader. It is also advisable to check the suitability of the leader for the magnitude of the task, with only experienced and proven leaders given studies involving major hazards or other difficulties.

A good leader will have interpersonal skills that can be deployed to help the study group to function as a true team. These include the ability to listen, guide, and encourage individual contributions as and when required, to sense the unspoken feelings of individuals—perhaps indicated by their body language—to move the group toward a conclusion when a consensus is emerging and to work efficiently with the scribe. Within the HAZOP process, the leader directs the route taken through the stages, the parameters and guidewords. It may be necessary to vary from the preset plan by introducing new parameters and guidewords, to question doubtful issues and to defer examination—perhaps to gather more information—and to periodically review and remind the team of the key issues. The leader has responsibility for ensuring that the HAZOP study is functioning to a high standard and, if it is not, to call for changes which will restore these standards. In an extreme case, this could mean terminating a study until acceptable drawings, project information, or other team members are available.

A good leader working with a badly-selected team may be able to produce an adequate HAZOP study. The quality of the leader is critical and a good team cannot always negate the impact of a poor leader. If the team loses confidence in the leader, the study is doomed.

5.2.2 Scribe (Scribe/Recorder)

The scribe is another key individual in the team. The person chosen must be familiar with the HAZOP study method and usually has a technical background so that special explanations are not required. The scribe must be a good listener, always paying attention to details. In a major study, the scribe has no other role, as recording is a full-time task. In a small study—for example, a modification—or on a self-sufficient small site, the recording may be done by the team leader or a team member. Note that experience shows that the scribe cannot be contributing to a discussion and writing at the same time so this inevitably slows the study.

The scribe should establish a working relationship with the leader to become a helper, not just a recorder. Help can be given by noting suggestions and deferred problems and bringing them up at a later time, and by assisting in the selection of guidewords and parameters. More directly, the scribe starts to record as a consensus appears, without waiting for instruction or dictation from the leader. To do this the scribe must learn when to write if recording manually (or to type if recording electronically)—starting too soon is a waste of effort while waiting for a final version to be dictated can slow the progress of the study and interferes with the creative flow of the discussion. The team should know exactly what has been recorded. This may be done by the leader reviewing the entry, the scribe reading it back or, if a computer recording program is in use, having it projected so all of the team can see the record.

The choice of computer recording over a simple written record is a matter of preference as both methods can produce good records. If a projector is used—as is common practice—it must be understood that during the meeting the priority is to capture the essential points of the analysis; spelling is not a high priority as it can be improved later but there must be no ambiguity or incompleteness in the first record.

It is important the study records are produced consistently and speedily after each session, particularly in a lengthy study, and that final and agreed versions are produced.

5.2.3 Team Size and Membership

The size and composition of the team varies according to the complexity of the study. The minimum size for even the smallest study is probably four with some of the roles combined. Usually there is a core team of five or six. Occasionally, the number may be as large as eight or nine; team membership should be reviewed if this is exceeded. A typical team might involve the following personnel:

The essentials for the team members are that they have sufficient experience in the area of the operation and, as a group, have a comprehensive knowledge of the intended operation. In addition, the basic disciplines represented should ensure that a fundamental analysis of identified problems can be made. The key is the assembly of individuals who, together, provide the correct combination of basic disciplines and expertise, and general experience of the type of operation to cover all stages of the project. The disciplines needed may be a selection from chemical, mechanical, and electrical engineering, control systems engineering, chemistry, biochemistry, and other disciplines which offer specialist knowledge, when necessary, of reactors, piping, instrumentation, software, and metallurgy. General experience within the industry should average several years per member, including direct relevance to the intended operation with regard to the site, the method of control, and type of process. These latter aspects may be covered by the inclusion of a senior supervisor. Direct project involvement ranges from the initial research and development, earlier hazard studies, process design and the intended construction, commissioning, operations, control and maintenance groups.

It is possible that in complex processes which have a number of operating interests that the membership may be “dynamic.” If this is the case the new team members must be given sufficient time to appreciate the direction and dynamics of the HAZOP study and be able to fit into the team at the appropriate time. Likewise a departing member should be satisfied that there is no further use for his/her team membership before departing from the team.

Compromises often have to be made when selecting the team. The ideal persons may not be available and there can be difficulties if a big team results from an attempt to meet all of the criteria suggested above. In a large team—say nine or more—separate conversations may start, making it difficult for the leader to capture everything and disrupting the unity of the team. The leader must make a balance to achieve the best possible results. The potential for conflict is greater in a large team; it is a responsibility of the leader to be alert to developing problems and to prevent them from becoming major issues.

Whenever a new design or a modification to an existing plant is studied, the team should always include a representative from the design group, from the operations team, and it is likely to include an instrumentation engineer. Other team members are then chosen according to the nature of the project. Possible further members could be a mechanical engineer, development chemist, electrical engineer, software specialist, and safety specialist. It may be possible for the safety specialist to also act as the team leader. A good team is likely to have several members who are closely involved with the project and one or more who have little or no previous connection. It should not consist solely of those who have worked full time on the design nor should it be totally independent of the project. The interaction between these two groups can improve the quality of the study. Some companies like to have an independent process engineer present to act as a critical questioner and challenger of assumptions, another role that can be filled by a truly independent team leader who has the necessary background.

In a normal HAZOP study where several meetings are required, it is important to have the core team members present at all the study meetings. Apart from the development of a team style and spirit, this maximizes the benefits from the growing understanding of the operation that comes as the study progresses. If one member of the team is from an operating team working on shifts, however, it may be impractical to have the same individual present at all the meetings. Specialists are added to the team when their expertise is needed for a particular section or stage.

It is important that some of the team have previous experience of HAZOP study; if some have not participated before, then training should be considered. Provided a few members are experienced, the others can be introduced by a short example or, if a single individual is involved, by a briefing and then by participation within the otherwise practiced team.

5.3.1 Continuous Processes

For a good HAZOP study to be possible, it is necessary for the team to have a complete and accurate process representation. For continuous operations this is based upon the P&ID, supplemented by design specifications and other details used in the design work, including flow diagrams, and material and energy balances. Further items are the intended operating conditions—usually expressed as a range—the operating and control philosophy and methods, and equipment and instrument specifications. Relief settings are essential to allow verification of pipe and vessel specifications, and alarm and trip settings should be normally available. In addition, there should be material properties, hazards and known operating problems together with the basis for safety, site details, and demands on the site services, and the results of earlier hazard studies including any material and reaction investigations. Information on vendor packages needs to be available.

Once the required level of detail has been obtained, the next planning task is the division of the P&IDs into sections and items to which the guidewords can be applied. Alternative terms may be used to describe the blocks selected for examination. The term “section” is used in this guide for part of a continuous process to which the guidewords are applied; another frequently used term is “node.” The term “item” is sometimes used for vessels, exchangers, and so on, to which restricted or modified guidewords are applied. In a batch process, the terms step or stage are commonly used. A HAZOP study is a rigorous but time-consuming, and hence expensive, exercise. The higher the number of sections or steps, the more time-consuming is the study. Ideally, no sections should feature a process line junction and no steps should cover more than one element of the batch sequence. Experience suggests that some branching within a section and some step combination can be tolerated on systems other than those of very high hazard potential. Defining the intention helps demonstrate that the team have selected a worthwhile manageable node. Designation of the sections and steps is best undertaken by the leader at the outset of the study. Since this task is critical to the integrity of any HAZOP study, it is essential the leader has a proven track record in HAZOP management (see Section 5.2.1, page 31).

The selection of the sections is usually based upon the line diagram together with a description of the operating conditions. Often there are several possible ways of doing this division based upon the variation of parameters such as flow, pressure, temperature, and composition, at control points or junctions of lines with vessels. It is essential each section can be given a clear design intention and that a conceptual model can be constructed for the system in its given mode of operation. The size of the section contributes to effective analysis as there are problems in choosing either too small or too large a section. In the former case, there will probably be causes of deviations outside the section and consequences occurring upstream or downstream. There is a danger of these being overlooked. If too large a section is taken, then the design intention may be imprecise or very complicated so that it becomes difficult for the team to cover all possible meanings of each deviation. No simple, universal method can be given for the division into sections; experience is essential. The example in Appendix 3 provides an illustration.

It can be useful to mark each line and item of plant equipment as it is studied during the operating sequence. A final check can then be made to ensure all lines and items have been considered.

A further feature of continuous processes is that a high proportion of incidents and near misses occur when the plant is operated at conditions other than the steady state—for example, during start-up after maintenance, during commissioning, or shutdown. Therefore, it is important for the plan to give sufficient attention to these conditions and to the deviations that might occur in these situations—for example, too rapid heating or cooldown to the operating temperature, inadequate purging, or protective systems overridden. The team should also be mindful of the need for diagnostic instruments that will facilitate process performance monitoring and allow the assessment of local and wider heat and mass balances. These may be needed to allow assessment of heat exchanger performance, efficiency of pumps and compressors, and fouling of filters and piping, and can be of great importance during incidents.¹⁹

During start-up, it is sometimes necessary to override protective systems such as low temperature or pressure trips during what is an unsteady operation with a high risk of upset. The HAZOP study must pay particular attention to the process dynamics and risks with the protective system overridden and how the protective system will be reset (see also Section 4.5, page 19).

5.3.2 Batch Processes and Sequential Operations

The preparation work for a batch or sequential operation is usually more demanding than for a continuous process. In addition to the same range of background material, descriptions are also required of the detailed method of operation, the draft operating procedure, and the outline of the control sequence for a computer-controlled process. This time-dependent information is used in the division of the operation into stages for study, as illustrated in the example in Appendix 4. If a reaction is studied, the team needs information on the reaction process including heat and gas flows, exotherm onset temperatures, and the physical and chemical properties of the mixtures as well as the individual components. Again, the ability to define a design intention for each stage is an essential requirement.

One major difference between continuous and batch operations is that a physical item may need to be considered several times in a batch HAZOP study as the status, conditions, and design intention change. Thus, a reaction vessel may be considered during the addition of each component of the reaction mixture, during the reaction stage itself, and then during cooling and discharge. Another example is the common pipeline operation of pigging, where a typical sequence involving the loading of a scraper pig into a pig launcher at the start of the operation might be:

The pig launcher is at the center of each of these steps but each one involves a different configuration and usually different contents and pressures. The main issues will differ from step to step as will the design intention. Hence, the parameters will vary between the steps. Each of these steps should be considered separately, using the full round of guidewords each time to ensure that the hazards are fully examined.

Further requirements for batch process HAZOP studies are intended trip settings for both the control system and the safety system, including those that take place between successive stages of the process. The HAZOP team must take into account the manner and timing of the resetting of the control and trip systems, covering both safety and operability aspects.

Another difference from continuous processes is the need for the team to picture the whole operation and the changing conditions throughout the plant at all stages of the batch. Without this understanding, interactions and consequences may be overlooked. Indeed, as the progress of one batch is followed through its various stages, a previous batch may still be in a connected section of the plant and preparations for a following batch may be underway. Thus, a deviation in the batch being followed through may cause an interaction with the previous or the following batch and so have consequences for these batches. The HAZOP study team needs to have a clear overall picture to enable it to identify such cross-connections; time-related flow charts, for example, a bar chart, are needed to give this picture. Such charts are unlikely to be available from the design work and, if needed, must be prepared for the HAZOP study. Good preparation and organization of the data can greatly enhance the prospects of success in batch operations.

5.3.3 HAZOP Study of a Procedure

Many operations, both simple and complex, are dynamic and are done by following a strict, predetermined procedure. Any failure to strictly adhere to the procedure or to ensure all the steps are done in the correct order may have the potential for a serious consequence. Start-up and shut down operations normally follow a sequence set out in a procedure, and these operations are known to be common occasions for problems and incidents. Hence, all procedures should be considered for evaluation by HAZOP study. The general principles of HAZOP study, as discussed in this monograph, hold good but the approach has to be altered to get the best results. This is described below and illustrated in Appendix 5.

The major difference from the examination of a steady-state operation is that procedures are carried out in a chronological order and so each step must be considered in turn and be challenged by appropriate guidewords. Thus, instead of traditional parameters such as flow, and guidewords such as more, potential problems are more likely to be uncovered by deviations such as operation mistimed or guidewords such as too early/too late or out of sequence. However, this does not preclude the use of the usual parameters and guidewords used in the HAZOP study of steady-state processes.

As ever, the team must have a clearly defined node and design intention for the step being examined. The node may be a piece of equipment such as a pump or a pig launcher. It may be a section of line, as might be selected for steady-state HAZOP study, but where the flow is not fixed but is dynamic due to moment-by-moment changes during the procedure. The node may include a number of simple operations that comprise a clear section of the procedure, for example, a series of valve changes in starting a pump or changes of flow, quantity, and temperature for the start-up, shut down, or conditioning of a reactor. Arriving at the best selection of nodes is often more difficult than when planning for a continuous process HAZOP study.

The choice of guidewords is also more challenging. The guidewords too early/too late, out of sequence or incomplete, could identify missed venting in the pump start-up or residues left in the reactor. Further possible guidewords are tabulated in Section 10.2 on Human Factors. Of course the standard guidewords given in Table 4.1 should also be used if they generate meaningful deviations with the parameters of the node.

Good practice, as with HAZOP study of steady-state operations, is to identify appropriate guidewords and parameters before the study is commenced, and it is the responsibility of the team to adapt the method of study to suit the problem being examined.