OSHA SAFETY REGULATION CALLS FOR STEP-BY-STEP APPROACH

P.J. Bellomo
Arthur D. Little Inc.
Houston

The U.S. Occupational Safety & Health Administration's long-awaited process safety management (PSM) regulation mandates the implementation of a PSM program at facilities handling highly hazardous materials, including oil refineries and petrochemical plants.1

This article presents a step-by-step PSM program compliance strategy, delineated and explored through practical examples.

OSHA's PSM regulation-effective May 26, 1992-arrives in the wake of many similar codes, standards, and practices. And it will likely establish guidelines for the safe management of process facilities for years to come. 2

It has become customary to speak of PSM in terms of discrete "elements." Along these lines, OSHA's final rule has 14 discrete elements, listed in Table 1, along with key compliance requirements and timing aspects.

The notion of PSM compliance raises many questions, not the least of which is, "How do you get there from here?" The guidelines presented here will help answer this question. However, before describing the strategy, it is wise to first acknowledge the complex nature of the issue.

While companies share the same PSM problems, in many instances individual company solutions will be unique. One can learn from successes and failures elsewhere, but existing organizational structures and operating practices at individual facilities will likely demand custom-designed PSM programs.

The intent of this article is to provide order to what otherwise might be a patchwork approach, by offering a PSM compliance strategy.

PSM COMPLIANCE

The road to compliance can be divided into four phases:

Phase 1 -Identification of needs
Phase 2-Design of policies and procedures
Phase 3-Implementation of policies and procedures
Phase 4-Maintenance of program.

Each phase will be discussed, along with the final products one can expect to generate during each phase.

In addition, certain pitfalls will be identified. These will be placed in the context of practical examples of successes and failures in industry.

PROGRAM DEVELOPMENT

Conceptually, the strategy is simple, though perhaps not immediately obvious. In fundamental terms, it is a call for careful planning before action. If PSM compliance is currently a goal instead of a reality, then a methodical approach to the problem is needed, with these four phases in mind.

Identification of needs. Survey your facility; establish a PSM baseline; discover the fine details of where you are today. (It will take a long time to get there from here if you do not know where "here" is.)
Design of policies and procedures. Address PSM deficiencies, e.g., design a management of change (MOC) procedure; create a process hazard analysis (PHA) policy and obtain PHA expertise; finish your emergency response plan (ERP); etc. The key objective of this phase is to establish the necessary PSM management systems. (This is the easiest phase.)
Implementation of policies and procedures. Institute the MOC procedure; conduct PHAs; carry out ERP drills; etc. Expect the need for fine tuning the management systems developed in Phase 2. (This is called "execution" and it is undoubtedly the most difficult phase.)
Maintenance of program. Conduct periodic PSM audits and follow-up on the findings; monitor the industry and stay ahead of new codes, standards, and regulations. (Your overall program must stay "evergreen.")

Again, bear in mind that each facility will carry out these phases differently. The similarity is that all programs must, in one way or another, move through each of these phases. 3 4

PHASE 1

The final product of this phase is a PSM compliance work plan. Typically, the way to produce a work plan is to conduct a baseline survey of a facility; this might also be called a needs assessment. This is a snapshot in time that will reveal the status of the current safety management situation.

There are numerous ways to accomplish a facility baseline survey. The different approaches are linked to factors such as facility size and resources. Irrespective of the specific approach, it is important to strive for a thorough and objective assessment.

For companies with numerous sites, it may be worthwhile to develop a single, detailed survey protocol for use at all locations. This should ensure both thorough and consistent baseline surveys across all sites.

If such a self-assessment approach is used, it will be important to remain sensitive to issues of objectivity. It may be useful to have representatives from Site A participate in the baseline survey at Site B, and representatives from Site B participate in the baseline survey at Site C.

Another possibility is to have a composite team survey all sites. The obvious benefits include knowledge/technology transfer from site to site as well as the Generation of cost-effective baseline surveys.

But some words of caution about potential pitfalls are necessary. Experience has revealed that there are mistakes to be made in this phase. The optimistic surveyor may be too easily convinced by formal policies. Existence of a written procedure does not necessarily mean it is followed. It is important to spot check and verify procedures.

On the other hand, the pessimistic surveyor may miss the informal procedure. Although written procedures may not exist, interviews with operators may reveal that there are accepted and consistently followed ways of doing things. In such an instance, the solution might be to formalize the procedure by documentation.

EXAMPLES

A classic example of the formal policy not being followed concerns pressure safety valves (PSVs). It is common for a facility to produce a written PSV installation and testing policy. There will be explicit statements that require any block valves upstream and downstream of PSVs to be either car-sealed or locked open.

There will also be a comprehensive list of all PSVS, along with an explicit testing schedule. Finally, there will be guidance for variances from the PSV policy. Despite the thorough documentation, the surveyor may want to spot-check the PSV policy for thorough implementation: frequent implementation breakdowns in PSV policies are hardly uncommon.

An example of an informal policy was revealed during a site-wide assessment at a facility handling strong oxidizers. An operator was questioned about the reliability of some braided stainless steel flex hoses at the inlet of a vessel.

Because the hoses were not on any inspection list or preventive maintenance schedule, it was anticipated that there might have been past loss of containment incidents involving the oxidizer. At the very least, some operator concern was expected.

This, however, was not the case. Once a year, the operators had each of the flexes replaced, despite the absence of a formal preventive maintenance and testing policy. In such cases where an informal policy exists, it is generally worthwhile to formalize the policy by modifying and/or building on the existing one.

BASELINE SURVEY

The essential point here is that the idea of a baseline survey can easily be oversimplified. Such thinking will lead to a poor survey and a poor start to a facility's PSM initiative.

Before undertaking a baseline survey, evaluate the magnitude of the effort required and proceed accordingly. If a sizable task confronts you, it may be advisable to obtain training or outside assistance.'

Incidentally, the word "audit" has been purposely avoided. Although good auditing techniques are essential, an audit usually only identifies deficiencies with respect to strict codes, standards, and regulations.

Because the final product of this phase is a PSM work plan, such an audit will not suffice. In the area of process safety, there is no standard approach that is universally practiced. Consequently, a thorough assessment is required to establish a baseline.

Following the baseline survey, a PSM work plan should be generated. Again, there are many forms this can take, but there are indispensable items with respect to content.

Sequencing PSM tasks will be an important step. Likewise, it will be important to estimate task durations and resource requirements. Internal and external labor requirements should be estimated, along with associated costs. Finally, the PSM work plan should include a schedule for completing PSM tasks, with PSM milestones specified to help measure progress and facilitate the overall management of the effort.

PHASE 2

The final products of this phase are written PSM policies and procedures (i.e., management systems). In one way or another, all PSM elements should be addressed, ideally by modifying existing policies and procedures instead of creating new ones. 1 6 7 PSM guidelines are performance-based, and they require management systems to consistently achieve the desired level of process safety performance. Organizational accountability and paper trails will be necessary.

A case in point is the ubiquitous PSM element involving PHAS. It simply is not enough to send a handful of good technical people for PHA training and expect compliance to follow. A company needs a PHA procedure. This procedure should pay explicit attention to the task of follow-up on all PHA action items.

A related example involves PHAs on capital projects. As a minimum, the process for appropriating and spending capital monies should require a PHA on any process with hazardous materials at or above threshold quantities. In addition, it is advisable to set guidelines for the PHA review technique(s) and the timing of PHA implementation during the project cycle.

What should become clear is that PSM systems cannot be designed in isolation. Existing management systems should be modified when practical. And whether the policies and procedures are existing or new, it is important to receive input from key functional groups.

For instance, both operations and maintenance have an interest in MOC, and consequently, an MOC policy/procedure should not be finalized without their input. An example comes from experience at a small facility in the fertilizer industry.

EXAMPLE

At the conclusion of several days of PHA training, conversation turned to other elements of PSM. In a discussion of MOC, the plant manager hastily concluded that the solution was simple: form a committee of his best people and establish a policy that the committee had sole responsibility for authorizing any plant changes.

Although this was a small plant, it ran 24 hr a day, 365 days a year. The outburst from both operations and maintenance supervisors soon made it apparent to the plant manager that more thought was needed.

This shows why it is important to plan, for each element, a program implementation phase that is separate from a program design phase. However, it may not be apparent that these phases will inevitably overlap. All PSM elements do not Necessarily have to be designed at the same time.

It may make the most sense to design policies for maintaining comprehensive and up-to-date process safety information (PSI) and standard operating procedures (SOPS) concurrent with MOC. Design of the remaining elements, in the meantime, may not begin until PSI, SOPS, and MOC have undergone initial implementation, adjustment, and finalization.

In other words, highest priority elements may move through Phase 5 into Phase 3 before lower priority elements enter Phase 2.

PHASE 3

The final product of this phase is a fully functional PSM compliance program. Engineering, operations, maintenance, and safety functions should be routinely practicing PSM as part of their everyday responsibilities and assignments. The greatest pitfall in this phase is to simply omit explicit implementation efforts. It is one thing to generate a policy or procedure; it is entirely another to ensure that it is functioning as designed.

Another example concerns a large polymer facility where releases of toxic gases were a chronic problem. A brief history is in order.

EXAMPLE

The facility's ERP was essentially restricted to fire brigade activities. Although the brigade was well-organized and trained, functioning superbly in emergency situations, breakdowns at the interface between these emergency responders and the remainder of the plant population were not uncommon.

While the brigade was identifying and isolating the source of a release, others in the plant would sometimes unwittingly expose themselves to the toxic gas cloud.

Following an episode that resulted in first aid treatment for more than 20 people remote from a gas release, management decided to modify the ERP. The updated ERP included a communications interface between the brigade and area personnel.

Management even generated ERP training for the entire plant. However, periodic drills were never fully instituted. The plan that was understandable on paper and during brief training seminars became less clear during emergency situations.

Unlike the well-trained and drilled brigade, the plant population found emergency situations confusing. Implementation was never carried to its logical end.

Implementation will mean different things for different PSM elements. For the ERP, it may mean training and drills; for PHAS, it may require a pilot process where a specific hazard identification methodology is used and all action items are driven to completion during follow-up.

Often, the meaning of implementation will be mundane: for mechanical integrity testing programs, it will likely mean that testing will begin as prescribed by some policy/procedure and all results will be documented.

The crucial point regarding implementation is that adjustments will be needed. Attention to fine details will successfully, close-out this phase of the program.

Ideally, PSM will result in a cultural change inside an organization. Although a step-by-step approach will bring an organization from Phase 1 to Phase 4, such a "task orientation" must disappear. That is to say, ultimately PSM should not require the execution of deliberate compliance tasks; instead, it should be a normal part of everyday business.

Depending on the advancement of PSM ideas in the existing organizational culture, different levels of effort will be required. As indicated earlier, this phase is essentially nothing more than the onerous task of execution.

PHASE 4

As mentioned, the final product of this phase is an evergreen PSM program. If the previous three phases were successful, Phase 4 should be relatively routine. Clearly, the level of effort to maintain the program should be less than that for Phases 1-3, though greater than the effort required prior to the formal PSM program.

The obvious task here involves periodic PSM auditing. It will take time to determine the correct frequency for this auditing. The minimum standards set by regulators-currently hovering at about 3 years-may prove too infrequent.

In that time, PSM system breakdowns may grow beyond the point where they are easily corrected. For such cases, the audit schedule could be increased in frequency, or different levels of audits could be established.

One pitfall in this maintenance mode is the failure to close out audit findings. Because auditing uncovers system breakdowns, management must be prepared to support measures to fix those breakdowns.

Another pitfall involves sheer complacency: the fact that a PSM program is working today does not mean that industry standards will not surpass it.

Like total quality management, PSM is a process that should strive for continuous improvement. Consequently, a function should be in place to monitor the regulatory environment. It is more cost-effective to stay ahead than it is to play catch-up.

WORKING ANALOGY

Convincing engineers and technical managers to buy into something as "soft" as a strategy can be difficult. Such professionals relate more to hardware than to a strategic concept.

A relatively crude comparison between the compliance strategy and a well-accepted technical management approach can be made. This comparison is illustrated in Table 2. Although the comparison is crude, the similarities are nevertheless provocative.

Everyone in industry has a horror story about a project that was undertaken with a loose, unclear, and/or ill-conceived scope. And no schedule is complete if time has not been allotted for start-up. Finally, the best design can be undermined by a poor maintenance program. And so it goes with management systems in general, and PSM in particular.

In the form of codes, standards, and regulations, the PSM compliance "map" has been laid out before industry. With excellent technical people and committed management, any company can get there from here.

REFERENCES

"Process Safety Management of Highly Hazardous Chemicals; Explosives and Blasting Agents; Final Rule," 29 CFR 1910.119, Feb. 24, 1992.
Ozog, Henry and Stickles, R. Peter, "Process hazard management documents, practices compared," OGJ, Jan. 28, 1991, P. 86.
Wade, D.E., "Manage for Process Safety," Hydrocarbon Processing, September 1990.
Hawks, J.L., and Merian, J.L., "Create a Good PSM System," Hydrocarbon Processing, August 1991.
Greeno, J.L., Hedstrom, G.S., and DiBerto, M.A., Environmental Auditing: Fundamentals and Techniques, 2nd edition, Arthur D. Little Inc., Wiley & Sons, New York.
"Management of Process Hazards," API Recommended Practice 750, 1st edition, American Petroleum Institute, Washington, D.C., January 1990.
Guidelines for Technical Management of Chemical Process Safety, Center for Chemical Process Safety, American Institute of Chemical Engineers, New York, 1989.