“I don’t have a particular recommendation other than that we base decisions on as much hard data as possible.”  — Dorothy E. Denning

The best waiters are the ones who are willing to make recommendations when you ask.  Anything but, “Well, everything is good.” Because, admit it, everything is not. And then, when you try their recommendation, it actually is good.

Coming up with good recommendations is an art.

Especially for PHAs and LOPAs. But without good recommendations, there is not much chance of a process hazard analysis or a layer of protection analysis doing what they are intended to do, which is to find hazards and then address them to reduce the risk of a process.

If we are not going to reduce the risk of a process—to make the process safer—then what’s the point? Without implementing recommendations, the process is no safer than it was before the PHA or LOPA, in which case we needn’t have bothered.

But what is a good recommendation for a PHA or LOPA?

S.M.A.R.T. Goals and Objectives

In November 1981, George T. Doran published a brief piece in Management Review called, “There’s a S.M.A.R.T. way to write management’s goals and objectives.” Doran defined the SMART acronym for the first time with these words:






Since then, a cottage industry has developed where management consultants tweak the acronym by substituting other words into the acronym, e.g. achievable for assignable, or relevant for realistic, or time-anything for time-related. They all achieve much the same thing, however, and the lasting value of Doran’s little piece is that most of us still get performance reviews where our managers are admonished to develop goals and objectives with us that are SMART.

S.M.A.R.T. Process Safety Recommendation

More recently, process safety consultants have applied the SMART acronym to recommendations from PHAs and LOPAs, adjusting it make it more applicable. I recently encountered one that so closely aligned with my own thoughts on process safety recommendations that I thought it worth sharing.






Let me offer a few words about each attribute.


It has become increasingly clear since the original promulgation of the Process Safety Management (PSM) Standard (29 CFR 1910.119) that while employees should all have access to PHA documentation, the individuals charged with responsibility for implementing a recommendation from PHA or LOPA often have little more to work from than the recommendation itself.  So, it is important that the recommendation contain enough information so that the person assigned to carry it out know what it is they are supposed to do. That is not to say that the recommendation has to be overly prescriptive. It is great to give some latitude and flexibility in the recommendation, as long as it is clear what is to be achieved. That’s why the final compilation of the recommendations should say something about why the recommendation is necessary.


In the context of process safety recommendations, this attribute is not much different than when used performance reviews: clearly define what “done” looks like, in terms of an objective, quantifiable standard.


Recommendations don’t implement themselves; someone has to see to it that they get done. Often, the team members of the PHA or LOPA review are not the right people to assign responsibility for implementing a recommendation or to allocate the resources necessary to implement a recommendation, that still must happen. If there is no one to whom the recommendation can be assigned or no resources with which to implement it, then it is not a good recommendation.

A good recommendation is one to which someone in the organization can assign personnel and resources. And then those personnel and resources must be assigned.


PHAs sometimes get a reputation as boondoggle magnets.  Various pet-peeves which haven’t been approved in the past suddenly get re-branded as “safety recommendations” with the expectation that now they must be approved.

Not so.

The PSM standard requires that recommendations must be resolved; it doesn’t require that the resolution be implementation as proposed. However, every resolution must be documented, which means that every recommendation makes work for someone. Therefore, it is important that recommendations actually contribute to reducing the risk of a particular hazard. Otherwise, the recommendation doesn’t belong in the PHA or LOPA.


In the PSM Standard, in 29 CFR 1910.119(e)(5), OSHA requires of PHA that “the employer shall establish a system to promptly address the team’s findings and recommendations; assure that the recommendations are resolved in a timely manner and that the resolution is documented; document what actions are to be taken; complete actions as soon as possible; develop a written schedule of when these actions are to be completed; communicate the actions to operating, maintenance and other employees whose work assignments are in the process and who may be affected by the recommendations or actions.”

In other words, recommendations must be tracked.  So, any good recommendation will lend itself to the tracking system that the employer has established, and actually be tracked.

What Do You Recommend?

The return on the time and effort we invest in a PHA and LOPA is highest when they yield good recommendations.  Not necessarily a lot of recommendations, but good recommendations.  Learning how to make SMART recommendations will go a long way to make good recommendations. It may not come easily, but it will come with practice, so practice. That way, you can be the person others turn to and ask, “What do you recommend?”


  • Mike Schmidt

    With a career in the CPI that began in 1977 with Union Carbide, Mike was profoundly impacted by the 1984 tragedy in Bhopal and has been working on process safety ever since.