Using TapRooT to Improve Incident Investigation and Identify Fixable Causes

Challenge
Prior to implementing TapRooT in 1993, we performed incident investigations but we often stopped at a level above the root cause; we missed root causes that were important; we placed blame rather than finding fixable system problems; and we didn’t have a consistent, convincing means to present our information to management. Therefore, our challenge was to improve our investigation system so that we could better prevent repeat incidents (that can be dangerous and expensive) by implementing effective corrective actions that would be approved by management and implemented in the field.

Action
In 1993 we trained six people in a 5-Day TapRooT Course. The training helped us to decided to go forward with implementing TapRooT, rewriting our investigation procedure, and training a large number of people to use TapRooT to investigate problems.
We decided to license our site to use TapRooT so that we could conduct our own courses based on the System Improvements copyrighted training material. One of the initial trainees became a certified instructor and over an 18 month period he trained about 200 people to investigate problems by using TapRooT.
After our initial implementation, we also developed the capability to record our results in a database and trend the incidents’ causes over time.
Specific Example of How TapRooT Helped
How did TapRooT help us? The best way I can explain the impact of TapRooT is to describe some specific incidents that the TapRooT System helped us investigate and prevent (by identifying root causes that we could fix by implementing practical fixes).
One example was the investigation of a large fire in a process unit. The fire resulted in a complete unit shutdown. The systematic TapRooT investigation helped us identify the misapplication of steel in the process as one of the causes.
But TapRooT didn’t let us stop there. We continued to look for the system causes that led to why the wrong metal was used. This allowed us to prove that we needed to perform a complete material verification (PMI or Positive Material Identification) to make sure that there were no other misapplications of steel in other similar parts of the process.
This verification was expensive. Without the detailed proof and logical presentation tools that are built into the TapRooT System, management might not have perceived the need to spend the money needed (and commit to the plant down time needed) to complete this verification.
What did we find? The inspections identified three additional areas that needed repair. Any one of these could have resulted in an additional fire and unexpected shutdown. Worse yet, a large fire could cause injuries or fatalities. By using TapRooT we avoided these future problems that would have eventually occurred.
As part of the TapRooT process we institutionalized our corrective actions by updating the PMI Policy. We also looked beyond the specific problem of this application of metallurgy to the generic problem of pipe wall and pipe joint thinning. (Having a database helps you develop a convincing argument that this isn’t just a one-time problem but rather is a repetitive problem that needs a refinery-wide system fix.)

Longer Term Results
In January 1997, when performing the inspections implemented as a result of our previous TapRooT investigation, we found thinning of a piping joint on a process unit furnace. We shutdown the furnace to make repairs that cost about $270,000. If the inspection (implemented after the incident investigation described above) had not been performed, the pipe would have failed. The cost of repairing a catastrophic failure would have been much higher (possibly ten times as much) and could have led to personnel injuries.
The total impact has been so broad and pervasive that it’s hard to measure in dollars. Why? Because some improvements are made by people in the field and aren’t “documented” as being attributed to implementing TapRooT. Also, it would be unfair to say that the only improvement initiative is TapRooT. Therefore in any large facility the total progress being made is not the result of a single initiative but rather the cumulative impact of all improvement initiatives. Therefore we can’t just look at our improved performance and attribute it all to TapRooT.
However, I can point to specific incidents (like the one I described here) that TapRooT helped us develop effective corrective actions that would not have been developed before we started using TapRooT.

Lessons Learned
Here at the Torrance Refinery, we are firm believers that the TapRooT System provides us with the necessary tools to determine the true root cause of an incident. This saves us investigation time and saves us the pain of having repeat incidents that we know would cost millions of dollars and cause personnel injuries.
We have also learned that we needed to apply TapRooT “across the board.” If we tried to apply TapRooT only occasionally, we lost proficiency and missed opportunities to catch problems when they were small (before big accidents happened).
We also learned that we needed to trend root cause data from the incidents so that we could identify problem areas (generic causes), allowing us to eliminate whole classes of incidents by implementing generic system fixes. The database also helps us make a convincing argument that helps management see the need to implement fixes that we previously would have failed to convince them of their desirability.
We are expanding the use of TapRooT beyond safety and production issues to problems with project planning and implementation. This is leading us to new ways to improve our project initiation and approval processes and save even more money by stopping problems before they are “set in concrete.”