Management of Change (MOC) has been one of the hardest of the PSM requirements for industry to master. When we analyzed MOC failures we discovered it isn’t technical incompetence. So, what is the root cause for such failures? It’s the time involved by key personnel in processing multiple changes and their effects combined with juggling other duties. What is missing is an appreciation for how change affects the workflow of operators and managers. We have seen examples of these problems in several familiar industrial accidents including Flixborough, Union Carbide’s Bhopal, BP’s Texas City, ExxonMobil’s Longford and Texaco’s Pembroke.
This is an important piece of information if we consider that very few companies apply Management of Organizational Change (MOOC) procedures. A brief review of industrial accidents is all that is needed to show how important it is manage the human element during process changes. The companies that are doing MOOC are doing it in the same spirit as MOC; we often see a superficial discussion of a badly conceived plan. The best we’ve seen is a human factors checklist (Figure 1) from the Chemical Manufacturer’s Association (CMA), “CMA Management of Safety and Health During Organizational Change - A Resource and Tool Kit for Organizations Facing Change.”
Figure 1. Splitting an operator between field and control room could have consequences.
The checklist is intended as a motivational tool. There is more to an MOOC review than the checklist. As a credit to the CMA, they defined how a multi-disciplined team should review and understand a change and it’s implications. One fly in the ointment is that like a PHA review, the value of a MOOC study relies on the quality of the team searching out potential problems and its preparation prior to the discussion. Often such a review would benefit from a “gap analysis” to ensure that the new organization fulfills the needs of the people living within its structure and the goals of the company. This goes well beyond the “Yes” and “No” of a generic checklist.
I recently reviewed one of these completed checklists at a major U.S. refinery. The team could say “Yes” to most of the questions but they never debated the subject. If they had done so they would have discovered that the existing training was inadequate. The only reason that the operators are competent is that, after 20 years on the job, they had found their own ways to get the job done. Contrary to company policy, procedures weren’t documented, validated or tested. On one of those rare occasions when the holes in the Swiss cheese line up and an accident occurred operators were condemned by managers who didn’t understand the process the way the operators did.
The failure of this method shouldn’t surprise anyone in the chemicals industry. During the evolution of the PHA some companies tried to automate the process by using checklists. It was a rude awakening when they discovered that the HAZOP was a better approach — compelling engineers to consider integration of their design into a plant system. A formal HAZOP has its own checklist, a list of guidewords asking, for example, “Is the temperature too high, or too low?” The power of the guidewords is that they provoke a debate: What could cause the temperature to be too low? Is the cause significant — could someone be hurt or would equipment be damaged?
Now, let’s consider how a human factor checklist would be used in practice (Table 1).
Table 1. In this example training gets a 'D' because no MOOC plan exists and there is no simulator training. (Click to enlarge.)
Suppose an organization that has roaming operators that work inside and outside of a control room. Half of their day involves working in the field and the other half with a distributed control system (DCS). The organization plans to change this. In the new structure operators are to be dedicated exclusively to inside work.
After a cursory review managers recognize that the current DCS workload represents only about 40% of a shift’s workload. The easy solution is to consolidate this job position with similar jobs until under normal operation the operator is doing 100% of a job. The manager must then consider the outside portion of those consolidated jobs and ensure that all the field positions are filled. This isn’t always as easy as consolidating console jobs which are all in the same room. The field can be geographically challenging and units can be miles apart. So added to the field operators workload is travel time.
The checklist may expose some of the assumptions about common management systems but it won’t identify all of the problems. It’s clear from this example that the training is going to be very different and that console operators will need extensive cross-training. Even if the written procedures are accurate, operators working as a close team have their own ways of implementing a procedure, and every shift tends to do it differently. We aren’t saying that this is a good practice only that this is a fact. Teamwork, in this case, is an impediment to cross-training. The danger of this impediment can be countered by preparatory meetings before procedures are carried out. This new organization would benefit from a more formal procedure with assigned jobs against job posts (descriptions).
After shuffling the deck a new organization may find that some tasks aren’t being done or that they are being done incorrectly. So many things in the past may have been communicated by face-to-face communication. Now, in the new structure, radio communication is relied on. Communication standards should be reviewed and the potential loss of the radio system should be thoughtfully considered. A generic checklist won’t work. This should serve as a caution against reliance on avoiding discussion while implementing an MOOC. Fortunately, a better guideline exists.