HAZARD & OPERABILITY STUDIES (2 of 2)
Mike Lihou - Lihou Technical & Software Services
The team who will conduct the Hazop study should consist of personnel with a good understanding of the process and plant to be reviewed. The group should ideally contain about six members, with perhaps an absolute upper limit being set at nine. In a study in which both contractor and client are participating, it is desirable to maintain a balance between the two in terms of team membership so that neither side feels outnumbered.
The participants should consist of people from a range of disciplines, and this aspect is one of the strengths of the Hazop methodology:
- With a team of people, each with differing backgrounds and experience, potential problems are likely to be identified that might be missed by one or two people working on their own.
- It is often the case that one person's solution can become a problem to another department within the project. For example, a Process Engineer conducting his own review in isolation may identify a potential problem for which he considers that another instrument and alarm would be desirable. When this requirement is passed to the Control & Instrumentation Engineer, it transpires that no suitable channels are available within the appropriate section of the electronic control system, which has already been ordered and is currently being manufactured by the vendor. A protracted inter-departmental discussion and correspondence then ensues as to possible alternative remedies, and the potential cost penalty of re-specifying the control system. All of this could have been settled within a few minutes had both departments participated in the study.
- A spirit of co-operation and common purpose is engendered that crosses departmental boundaries, and this will persist even after the Hazop Study has been completed. Personnel will understand better the views, concerns and constraints within which other disciplines have to work, and will take these into account when making decisions affecting the project.
The actual composition of the Hazop team will vary according to the type of plant being reviewed. One person who should always be included is a representative from Operations. He or she should have first hand experience of day-to-day operations on either the plant being reviewed, or one that is very similar in nature. The contribution of this team member to the discussion can be invaluable, as it introduces an operational perspective to other participants who may have never, for example, had to climb down into a vessel wearing breathing apparatus to carry out repairs or an inspection.
To summarise, a team should be selected so that a balanced approach to the study is ensured. In addition, the intention should be that questions raised during the meeting can be answered immediately, rather than having to resort to the time consuming process of referring to outside expertise. It is not of course necessary for the same people to participate in the study from beginning to end. If the "core" of the group consisted of five people, for example, additional members could be called in from session to session as and when their particular expertise was needed.
As with all group activities, there needs to be a person appointed who will be in overall charge; with Hazop Studies this person is usually called the Chairman or Study Leader. Ideally, he should not have been too closely associated with the project under review as there might be a risk of him not being sufficiently objective in his direction of the team. As the Chairman's role is of vital importance in the smooth and efficient progress of the study, he should be carefully chosen and be fully conversant with the Hazop methodology.
Another important member of the team will be the Secretary. His contribution to the discussion may well be minimal, as his main function during the sessions will be to record the study as it proceeds. He will therefore need to have sufficient technical knowledge to be able to understand what is being discussed.
(Note to the reader: The above few paragraphs, as well as some that follow, are rather male-specific, referring as they do to 'he', ' him' and 'his'. This is purely to avoid the grammatical contortions involved with repeatedly referring to 'he or she', 'his or her', etc., up to three times in a single sentence. Note that there are many excellent female Study Leaders and Secretaries, who have organised, guided and participated in very successful Hazop Reviews).
It is most important that, before a study commences, work that can be conveniently done beforehand is carried out. This is not only essential in some respects for the proper structuring of the study and the team, but will also greatly increase the efficiency of the Hazop and thus retain the interest and enthusiasm of the participants.
This preparatory work will be the responsibility of the Chairman, and the requirements can be summarised as follows:
- Assemble the data
- Understand the subject
- Subdivide the plant and plan the sequence
- Mark-up the drawings
- Devise a list of appropriate Keywords
- Prepare Node Headings and an Agenda
- Prepare a timetable
- Select the team
All relevant documentation should be collected beforehand. Typically this might consist of:
- A Process Flow Diagram.
- A comprehensive Process Description containing operating parameters, flow rates, volumes, etc., as well as a brief summary of how each plant item functions.
- Cause & Effect Charts setting out how control and trip systems operate.
- Details of vendor packages if available.
- Plant layout diagrams.
The Chairman should take as much time as is necessary to gain a good understanding of how the plant is meant to operate, by studying the assembled data and if necessary talking to the design personnel involved. As he performs this task, it is very likely that he will notice potential problem areas. Private notes should be made of these, as they might possibly be missed during the course of the study. In such an event, it can only serve to enhance the Chairman's status within the group if he demonstrates his grasp of the subject by pointing out potential problems that the team have overlooked.
This stage of preparation is perhaps the most important, because it is the foundation upon which the other steps in the preparation process will be built. Without a reasonable understanding of how the plant functions, it will be impossible to plan a sensible study strategy, decide how long the review is likely to take, or who needs to be included in the study team.
Some proponents of the Hazop methodology state that there is no need for the Chairman to have any knowledge of the plant being reviewed, his function being only to ensure that the meeting progresses smoothly. An analogy to this approach would be a leader attempting to guide an expedition without a map, no plan of action other than to get to the destination, and with no knowledge of the terrain to be traversed. Such a person would command very little respect from other members of the team, and at the first sign of trouble he would likely be sidelined and marginalised by those with a better understanding of the situation. Once that has happened it will be almost impossible for him to regain control of the group.
In all but the simplest of plants, it is too much to expect any study team to deal with all aspects and operations in the process simultaneously. Therefore, it must be split into manageable sections (commonly referred to as Nodes, but sometimes called Tables because of the tabular means of recording the study). Also, the sequence in which these sections are studied is important.
With continuous plant, one usually progresses from upstream to downstream, with services such as drains, vent headers, instrument air, cooling water, etc. being considered separately and last. With regard to splitting the plant into sections, there is no need to consider each line and every single minor item of equipment under a separate Node. This will be wasteful of time, and boring and tedious for the team.
Instead, endeavour to group smaller items into logical units. Therefore, a minor pump with its suction, discharge and kick-back lines might be grouped together in a Node. However, with a major compressor, the recycle line and its in-line cooler should perhaps be studied separately. Also, when studying a vessel the Node should encompass those inlet/outlet lines up to and including any control/isolation valve/s, all level bridles, as well as vent lines up to the PSV.
If a number of streams converge on a vessel, the study sequence should if at all possible deal with all of those streams before the vessel is considered. The rule is "never study a vessel until the incoming deviations are known".
With batch operations, an entirely different approach is needed. In such a case the plant drawings are an accessory rather than the prime focus of the study. Of greater importance instead will be a detailed flowchart or operational sequence of steps to be accomplished. It is these batch sequences that will need to be split into manageable sections, and keywords may well target sequential operations such as Prepare, Charge, React, Transfer, Centrifuge, Dry etc. This methodology is required because an individual plant item is very likely to be put into differing states and serve different purposes at various stages of the sequence.
When the study strategy has been decided, the plant items encompassed by each Node should be marked in distinctive and separate colours, with the Node Numbers alongside in the same colour. Lines should be paralleled, and equipment and vessels outlined in the chosen colour. Where a Node spans two or more drawings, the colour used should remain constant.
This prior marking is a departure from the more usual practise of doing such work whilst the study progresses. However, it serves two purposes. Firstly, it will save time during the meeting, both in the actual marking and the discussion as to where a Node should begin and end. Secondly, the Chairman will be assured that in planning the study strategy nothing has been inadvertently missed.
Having completed the work above, it will be a simple matter to formulate a comprehensive list of the Keywords required to cover all aspects of the process to be studied.
Some companies, because most of the plant that they operate is of a similar nature, will have a standard set of Keywords. Such a list should be checked to ensure that it is covers all aspects of the system to be studied. Any redundant Keywords should be removed. For example, if the subject of the review is to be a pumping station, the inclusion of a keyword such as 'Absorb' is unnecessary.
The finalised list should be duplicated and a copy given to every team member. Also included should be a schedule of appropriate keyword combinations (i.e. which Secondary keywords will be applied with each Primary keyword). Where there are likely to be semantic problems as to what meaning/s a particular combination is intended to convey, then a full explanation should be given.
When devising the list, bear in mind that the smaller the number of words utilised, the more speedy the study. That is not to say that aspects of the process should be discounted. Instead, to illustrate what is meant, imagine a plant containing a separation vessel, some pump suction filters, and an environmental scrubber. Rather than have three keywords 'Separate', 'Filter', 'Absorb', have instead one keyword 'Separate'... that, after all, is the basic function of all those equipment items. Similarly, 'Temperature' can cover heat transfer aspects of Heaters, Coolers, and Heat Exchangers.
Node Headings reference the relevant drawings, and contain a brief description of the design intent of the relevant plant section, with process parameters, flow rates, and any other potentially informative details.
The agenda is a list of those headings. A copy should be handed to each team member. In addition to being informative and an aid to full participation, it will serve to put into perspective the amount of work to be accomplished in the time allotted. Hopefully this will induce an appropriate sense of urgency.
For all but a one day study, the Chairman should devise a timetable showing what needs to be accomplished at each study meeting if the schedule is to be maintained. In devising this schedule he will need to call upon his experience when assessing how much time the review will take. A great deal will depend upon the complexity of the plant as well as the experience of the team.
As a rough guide, with straightforward plant and with P&IDs that are not too 'cluttered', on average three drawings can be studied in a day. If the system to be reviewed is complex, or if each P&ID seems to have been drawn with the intention of not wasting any space (i.e. as many plant items as would fit are included on the drawing), then almost certainly only two or perhaps even one drawing will be completed in a day.
Be prepared for time slippage at the start of the study. Progress is always slow to begin with, whilst the team are acclimatising themselves to this novel role of casting critical eyes over their own or their colleague's design efforts. After the first day everything will speed up, and the schedule should be on target by the end of the week. Do not, however, allow the timetable to reflect this expectation of a slow start... better for the team to realise that they must increase their efforts, rather than go home thinking that this first slow day is the norm.
Having gained a good appreciation of what will be involved in the study, both in terms of content and timetable, the Chairman can ensure that the core team members have suitable expertise and will be available for the duration of the review. In addition, he can also ascertain which personnel with additional expertise are likely to be needed during the course of the meetings, and when their assistance will be required. With regard to the latter aspect, in certain circumstances the study sequence may need to be tailored around the availability of such personnel.
After all the above preparation, the Chairman should be in a position to easily guide an efficient and comprehensive study through to a successful conclusion. However, there are a few guidelines to remember:
- It is always a temptation for team members to illustrate their ideas by quickly drawing on the master P&ID that has been so carefully marked up. Establish the rule right at the beginning that this is forbidden, even in pencil.
- Similarly, with tie-ins and vendor packages, a team member may endeavour to help by roughly illustrating the upstream/downstream plant or the internal workings of the package. Be firm in the rejection of such help... it is dangerous to pretend to have studied something when all that is available is a few scribblings on a sheet of paper.
- If the schedule is slipping, resist the temptation to hasten the process by listing potential causes and consequences yourself. All that results is that the team sits back and listens to you dictating to the Secretary, and they will continue to do so until you force them to participate again.
- Do not allow a separate meeting to develop, with two team members conversing in low voices at the corner of the table. If this happens, stop the general discussion and ask them to share with the rest of the team the benefit of their deliberations (always assume that they are discussing something directly relevant to the study, although the likelihood is otherwise). This will usually elicit an apology and bring them back to full participation.
If they persist, request that the rest of the team members be completely silent whilst the private discussion continues. If even this does not produce the required result, call a coffee break. Then speaking privately to the persons concerned, politely but firmly insist that they leave the meeting. Such members usually have nothing to contribute to the study, and they will only irritate and demotivate the remainder of the team.
- Ensure that all team members participate, even those who might well feel unsure of themselves. Do this by asking questions such as "Do you agree with that solution, Bob?", or "What severity would you attach to this consequence, Fred?". Alternatively, and less potentially contentious, you could request "John, could you help the Secretary by summarising in a few words the agreed action". Once such team members realise that they are not going to be contradicted as soon as they open their mouths, they participate to the best of their ability.
- Recognise and reward with praise the team member/s who contribute to the discussion wholeheartedly and sensibly. However, do not allow them to overshadow the rest of the team.
- If discussion wanders away from the matter under consideration, re-focus the attention of the team either by requesting that the Secretary read out what he has recorded, or by asking for an action to be formulated. The latter usually concentrates the mind and encourages the team members to get to the heart of the problem.
- Where a particularly intractable problem arises, or consequences of a serious nature are uncovered, too often an inordinate amount of time is devoted to formulating potential remedies. Solutions and counter solutions are proposed and discussed, there is much speculation as to costs and other related aspects, and generally no satisfactory conclusion is reached. Before too much time is wasted, such situations should be dealt with by placing an action upon a specific person to investigate and report upon what alternatives are available, together with the advantages/disadvantages of each. Any discussion, gathering of additional data, reliability calculations, etc. can thereby be accomplished outside of the Hazop meeting, allowing the team to progress steadily with the review.
- The Chairman should be independent and unbiased, and should not be perceived as constantly favouring one section of the team as opposed to another. This is of particular importance when personnel from both client and contractor are participating. If a difficult situation arises, where, for example, there is a heated dispute over whether an action should be undertaken, in some cases one of the parties to the dispute will request that the Chairman makes the final decision. If, in the Chairman's estimation, the reasons on one side of the argument are so strong as to be indisputable, then he should say so. On the other hand, should the situation be finely balanced, then the dispute can be defused by careful wording of an action.
Take as an example the situation where the client wishes to have an additional High Level Alarm, but the contractor strongly disputes its necessity. Consider the following actions:
- "Fit a High Level Alarm". In the view of the contractor, the Chairman has sided with the client. He may, wrongly or otherwise, perceive this to be a biased decision.
- The action "Justify the requirement for a High Level Alarm" is addressed to the client. The Chairman favours the contractor's argument, but is not dismissing altogether the views of the client. Both parties are likely to be content with this formula.
- The action "Justify the absence of a High Level Alarm" is addressed to the contractor. The Chairman favours the client's argument, but is not dismissing altogether the views of the contractor. As before, neither party will have cause to feel aggrieved.
By effectively postponing a final decision until a later review of Action Responses, it is often the case that the two sides will get together after passions have cooled to discuss the matter rationally. Almost invariably the situation will then be amicably resolved.
The Hazop Report is a key document pertaining to the safety of the plant. The number of man-hours spent on the study is usually considerable. It is crucial that the benefit of this expert study is easily accessible and comprehensible for future reference in case the need arises to alter the plant or its operating conditions.
The major part of such a report is of course the printed Minutes, in which is listed the team members, meeting dates, Keywords applied, and of course every detail of the study teams findings. However, it is usual to include with this a general summary. The contents of such a summary might typically be:
- An outline of the terms of reference and scope of the study.
- A very brief description of the process that was studied.
- The procedures and protocol employed. The Keyword combinations applied should be listed, together with the explanatory meanings given to the team at the start of the study. Also the fact that Action Sheets have been produced and responses will be recorded should be explained. A brief description of the Action File (described in the following section) should be included.
- General comments. If, for example, the team were assured that high point vents and low point drains would be universally provided, mention that statement and its source. If certain details of vendor packages were not available, explain and list the items that were not reviewed.
- Results. This usually states the number of recommended actions.
Also included in the Hazop Report would be an Appendix containing:
- Master copies of the drawings studied.
- Copies of technical data used.
- Cause and Effect charts (i.e. matrices showing the executive action of safety related instruments and trips).
- Any calculations produced.
- Relevant correspondence between departments, from contractor to vendor, or client to contractor.
Each of the above should be signed and dated by the Chairman.
The Hazop Report is compiled as soon as possible after the end of the study, and once completed does not change. On the other hand the Action File is only started at the end of the study, and its contents will continue to change perhaps for many months, until the very last action has been reviewed and accepted as having been satisfactorily discharged.
Essentially, this Action File is a binder, or perhaps some electronic form of storage. Initially, at the end of the study, it will be empty. As completed and signed Action Response Sheets are returned, they are housed in the binder folder. Periodically, the returned responses will be input into the data file (either manually or electronically, according to the system being used).
By the time the first review meeting is convened there should be no outstanding (i.e. overdue) responses. The Secretary would prepare a listing of all responses received, making a copy for each review team member. During the review meeting responses will either be accepted and marked as having been discharged, or in a small number of cases further action would need to be taken.
At the end of the first review, where further action had been required, Action Sheets for these would be produced for distribution. In due course these would be completed, signed and returned, and these further responses would be input into the data file and housed in the Action File as before.
The procedure for the second review meeting is the same as for the first, except that the number of responses would of course be much smaller. If some of those responses were still not found to be satisfactory, then the process as outlined above would be carried out again.
It can be seen that the Action File represents a record of the state of completion of Hazop recommendations at any point in time. When all action responses have been reviewed and accepted, it finally becomes a static record containing the complete history of the implementation of the Hazop Study's findings.