The horrors of Hurricane Harvey as bad as it’s been the last week are, I’m afraid, about to eclipse what we’ve witnessed so far. In addition to the loss of life and flood damage, East Texas is home to the chemical and oil and gas industries. The signature oil sheen on the floodwaters is visible now. And if you think about the garbage, construction materials, oils and other fluids used for the plants that is now floating out into the area, the hazards are just beginning to be known.
We discovered this morning that the best laid plans for shutdown of these plants can go awry. Loss of power followed by flooding of the back up generators has resulted in rising temperatures in trailers filled with a reactive chemical and subsequent explosion.
While social media fills with naysayers and finger pointers at lax regulations for these industries, I go back to my own experience in chemical plants as we tried to identify the ‘what ifs’ for hazards.
I mentioned this in a previous post the process the chemical industry included in their designs as a result of a 1984 Union Carbide accident in India. I worked for a consulting firm in Alabama for a few months. Their contract with Amoco involved performing HAZOP, HAZardous OPerations, analysis of their existing facility to identify deviations and their causes. Then we looked at the likelihood of any of these situations occurring and the consequences to property and life should a design fail. The results of that analysis then may drive a design or instrumentation change. It may require an expenditure of capital to ensure the consequences are controlled.
For the scenario at the Arkema plant in Crosby, Texas, I’m thinking, regulations aside, the HAZOP analysis may have identified a scenario in which the power failed AND the generator failed due to running out of fuel, or flooding. While the consequences are high for this particular chemical, the likelihood that they would experience a hurricane causing the flooding at levels never experienced before may have been rated very low.
Alternatively, the engineers and safety personnel performing the analysis may have recommended a third back up that was rejected by the company as too expensive to install given a low likelihood of occurrence.
It will be interesting to watch how the root cause analysis results turn out. I’m assuming there will be one and that it will go public. Lack of regulations and oversight by the local and state government could be a possibility, as could a company decision to save the chemical stores ahead of the storm.
After all, how bad could it be?