Safety check
Transcription
Safety check
Assessing industrial hazards with software by Ellen Gardner We wouldn’t want to board a plane that hadn’t undergone an in-depth hazard and operability analysis. But the process is both time-consuming and costly. An Ontario risk management company has developed software that reduces the headaches. A s necessary as the process is, a hazard and operability analysis or HAZOP has traditionally been arduous and painstaking. Those who perform HAZOPs have gone so far as to describe them as the “industrial equivalent of a trip to the dentist.” “There is no denying these reviews make our world a safer place, but they also place considerable demands on the time and energy of engineers, facilitators and risk managers,” says Nigel Hyatt, P.Eng., director of technical services for Dyadem International Limited in Markham, Ontario. Dyadem is in the business of risk management, and its engineering services division has performed and documented HAZOPs for clients for several years. “We’ve used every possible tool you can think of,” says Nick Colella, engineering division manager. “Everything from manual transcripts to spreadsheets.” “We wanted to find a way of doing HAZOPs that was easier, faster and more versatile than the standard methods,” says Hyatt. They also hoped the project would create a meaningful bridge between the programmers developing the software 32 E N G I N E E R I N G and the teams executing HAZOPs. The project began in the Hyatt family’s back bedroom in 1992, and took two years to develop. With marketing and regular updates of the software, Dyadem’s software business has grown from five employees to 25, now occupying an office space of 450 square metres. PHA-Pro 5® —the fifth generation of HAZOP software—is Dyadem’s made-in-Canada answer to all that was ailing safety review teams. Safety reviews are a fact of life in just about any environment where hazards exist. They typically take the form of a checklist. “HAZOPs look at the com- D I M E N S I O N S – J U L Y / A U G U S T 2 0 0 0 plete process,” says Colella, “and ask: Are people following the correct procedures? Are the proper codes and standards being adhered to? Are steps in place to reduce or safeguard against potential hazards?” HAZOPs examine every node within a system to identify possible deviations (hazards). “The HAZOP doesn’t replace the safety review, it just goes much deeper,” says Colella. “You take the complete process, break it down into nodes and each one of those nodes is broken down even further in terms of potential deviations.” The review team for the HAZOP—engineering staff, production staff, maybe a chemist, health and safety reps, risk managers— work together to find the causes and the consequences of the deviations, the risks of those deviations taking place, and then finally, what safeguards are in place to ensure they don’t happen. Some potential deviations might be: What happens if you have back flow versus forward flow? What happens if it goes up, if it’s supposed to go down? What happens if you have too little ammonia and too much hydrogen? Documenting all this information is not an easy task at the best of times, but it has been made a lot easier with computerization. “When you’re looking for causes, you can use your imagination, but at the end of the day, people get tired and lose their concentration,” says Hyatt. Dyadem’s HAZOP software has various knowledge-based libraries, some with over 2000 causes and safeguards for the main methodologies, to help establish what can cause those deviations and what can be done about them. “For experienced teams,” he says, “it’s like a memory aide, since it prevents important topics from being overlooked. And for less experienced teams, the libraries provide a learning exercise.” In many cases, the company conducting the HAZOP will use a number of people from E N G I N E E R I N G SOFTWARE BUYERS’ CHECKLIST BEFORE YOU BUY HAZOP SOFTWARE, CONSIDER THESE QUESTIONS: Will you want your software to run on single PCs or on a local area network? How much disk space will the software occupy? What technical support does the software vendor offer? How often does the software vendor update the software? Can the software vendor provide support at your company’s different sites? Does the software vendor offer multiple licensing options? What process hazards analysis methodologies does my company wish to use, and will the software support them? Is the software compatible with other relevant software on the operating system? Does the software have features that will make my company’s HAZOPs (or other process hazards analyses) more efficient? To identify what’s available on the market, type the words “HAZOP” or “HAZOP software” into search engines on the Internet. Before purchasing software, try to obtain a working demo or a copy made available on loan and put it through its paces. You should also ensure that the software is written as a 32-bit application for the latest generation of computing systems. different facilities–a move that keeps the HAZOP more interactive and prevents workers from having tunnel vision. “For that reason, you need software that is going to be more versatile, and preferably has hyperlink capabilities,” says Colella. “Hyperlink is the most effective means of data management, whether it’s MSDS, CAD drawings or action items updates.” Computerization has other benefits, notably: ◆ consistency of analysis; ◆ access to stored data and information, on an as-needed basis; ◆ documentation of results. This ensures that accurate and representative records are created. “This last point is extremely important,” says Hyatt. “There is perhaps nothing more frustrating than conducting a meaningful PHA (process hazards analysis) session, only to find that the documentation fails to adequately record or encapsulate what has transpired, or that it cannot be understood.” To promote ease of application, the software is Windows®-based. It provides several options for performing the analysis and, D I M E N S I O N S – J U L Y / A U G U S T 2 0 0 0 like Windows, caters to a variety of individual preferences. Users can enter data by using either keyboard combinations, the mouse, a tool bar or pull-down/pop-up menus. The HAZOP software also features what’s known as the “naming convention.” In a standard HAZOP, you create columns with such headings as “causes,” “consequences,” “recommendations” and “safeguards.” Users in different jurisdictions may prefer to use different terminology, substituting the word “controls” for “safeguards,” for example. The naming convention permits the renaming of these key items, and is thus adaptable to many other applications. HAZOP methodology is comprehensive and intended to list concrete recommen33 dations, with the goal of mitigating identified hazards and reducing operating risks. “It’s possible to get caught up with listing causes or hazards, which may never lead to credible consequences,” says Colella. For that reason, the team at Dyadem strongly recommends the use of a facilitator, or HAZOP team leader, to help keep discussions in perspective. “The facilitator can be someone within the company or a third party from another facility, but they should have a general understanding of the industry, and be able to ask the right types of questions,” he points out. “It’s the facilitator’s job to take each and every suggested cause and ask: ‘Does that really apply to your operation?’ This person ensures there is credibility to the causes and consequences being suggested.” Based on its past experience, Dyadem applied a simple semi-quantitative methodology to compare the various methods of transcribing HAZOPs or safety reviews, both computer-based and manual. Table 1 ranks various features of three methods from 10 to 1, with 10 being the best and 1 the worst. “Because risk is defined as the product of consequence (severity of an incident) times the frequency (likelihood), once the user has identified the hazard, they can estimate how severe it would be,” says Hyatt. For example, a risk could be very severe but highly unlikely, or of minor severity and very likely. Both scenarios would be assigned a low risk matrix. A much higher value would be placed on a hazard that was moderately severe and very frequent. Most process hazard analyses are, by nature, qualitative. “We built a risk matrix into the HAZOP software to add some partial quantification, but it can still be customized to your own standards,” notes Hyatt. Since it came to market six years ago, the HAZOP software has undergone numerous updates and revisions. The people who rely on PHA methodology now include manufacturers of medical devices, the aerospace industry, water treatment plants and automotive manufacturers, who have all provided suggested improvements to the software. Dyadem has also relied on constructive criticism from within its own ranks. The teams doing the HAZOP analysis say the process has become easier in all respects, and they report improved documentation and downstream-tracking. “The program saves time, which results in considerable cost-savings,” adds Hyatt. There are three major misconceptions about HAZOP software: M I S C O N C E P T I O N 1 : “The lowestpriced software is also the best value.” Since the cost of HAZOP software is usually minuscule compared to the cost of performing the HAZOP, you will, invariably, get the best financial return by using the best product on the market. M I S C O N C E P T I O N 2 : “There isn’t much difference between one piece of HAZOP software and another.” Some types of software are less versatile than others and may be less user-friendly. A program needs to be comprehensive in its capabilities, while also being simple to operate. One that directs the user into an inescapable maze or pays little attention to navigation is unlikely to be a winner with HAZOP or safety review teams. M I S C O N C E P T I O N 3 : “Those who perform HAZOPs are best suited to writing HAZOP software.” Not so. Organizations mainly oriented toward executing HAZOPs may be less successful at producing good software. “We’ve seen many programs created by people who want to solve their problems through programming and are ill-suited to the task,” says Nigel Hyatt. “In the ideal scenario, the programmers ask what the users (typically engineers) want, and then the two parties working together. To make the process of performing HAZOPs less arduous, we need all the help that programmers, engineers, facilitators and other team members can pro◆ vide,” he concludes. Ellen Gardner is a Toronto freelance writer. TABLE 1. SCORECARD FOR HAZOP METHODS SCORES FEATURE Learning Curve Required User Friendliness Productivity Customization Features Copying Capabilities Team Participation Quality Assurance Assists Auditing of HAZOP Good Documentation "Power" Features Overall Assessment 34 MANUAL TRANSCRIPTION SCORES (OUT OF 10) SPREADSHEETS None (10) Maximum (10) Poor (1) Very Good (10) None (1) No (1) None (1) No (1) No (1) None (1) Poor (37) E N G I N E E R I N G Little (7) Good (7) Fair (4) Good (7) Some (4) Some (4) Little (4) Little (4) Fair (4) Some (4) Fair (49) WINDOWS®-BASED SOFTWARE Some (4) Fair (4) Very Good (10) Very Good (10) Very Good (10) Very Good (10) Very Good (10) Very Good (10) Very Good (10) Very Good (10) Very Good (88) D I M E N S I O N S – J U L Y / A U G U S T 2 0 0 0