Chemical Safety Board Releases Factual Update On Storage Tank Fire at Intercontinental Terminals, Houston, TX

Publication date, 10/30/2019

Here is the latest from the US Chemical Safety and Hazard Investigation Board on the Intercontinental Terminals explosion last March.  Every time I read these reports I get a creepy feeling on my skin.  The question pops into head, “Are these people engineers, or just bureaucratic, dead weight, English majors.  Take a close read.

ITC Tank.png

The first of two scheduled butane deliveries began unloading at about 7:23 pm and was completed by about 8:15 pm, with approximately 170 barrels of butane added to Tank 80-8. The second butane delivery of roughly 193 barrels was unloaded between 9:29 pm and 10:29 pm. Following completion of these two butane deliveries, the pump remained on to circulate the product. ITC expected a ship to arrive the following day, and the company planned to transfer all the contents from Tank 80-8 to the ship.

On the morning of March 17, 2019, ITC distributed control system (DCS) data indicates a series of unanticipated changes to the monitored pump operating pressures and tank volume. These fluctuations suggest a mechanical problem developed in the pump circulation system. DCS data indicates that at approximately 7:25 am, the pump discharge pressure readings began to rise slowly; by around 8:45 am the pump discharge pressure reading had increased from 80 to 84 pounds per square inch (psi). Consistent with a naphtha product release, beginning at roughly 9:34 am, the recorded tank volume began to decrease steadily. At about the same time, the pump discharge pressure reading suddenly dropped back down to 80 psi. At about 9:45 am DCS data indicates a second sudden decrease in discharge pressure from 80 to 75 psi, and the recorded tank volume continued to decrease during this time. Between 9:34 and 10:01 am, the DCS data shows that the tank volume decreased by approximately 221 barrels. The tank farm was not equipped with a fixed gas detection system, so no alarms were activated to warn ITC personnel of a release. The reduction in tank level and volume that occurred as naphtha product released from Tank 80-8 did not trigger any alarms in the ITC control room. As a result, ITC personnel were unaware of the naphtha product release before the fire erupted.

  • The first red section states that the pump was left on to recirculate the product.  Hmmm, although I am a chemical engineer, I am not an operations guy, I am a designer; yet, seems to me that if there is a leak, the first source for ignition would be an unmanned pump.  I know these areas require explosion proof electrical components, but they have explosions there all the time.  Engineers always say, “it somehow found an ignition source”.  Gee, I wonder if an unmanned pump could have done it?  Wow, you think?  Also, butane and naphtha are some of the most ideal fluids that I know of:  non-polar, clear, water like when a liquid.  Why would they have to circulate the tank for about 10 hours?   Only about 400 barrels were added to a 80,000 barrel tank, was so much circulation required for substances that mix naturally?
  • In the second red section, the states that the pump discharge pressure began to rise.  Yes, and so what?  Without the suction pressure, I do not think that information is very helpful.  A pump is a differential pressure device:  the system determines the suction pressure, the pump only guarantees a pressure differential at a specified flow rate.  If the pressure is rising, that would signify to me that the tank is being filled, that would cause the suction pressure to rise, and thus, the discharge pressure.
  • In the third red section, the report also notes that there was no way to detect the leak because no gas monitoring systems were in place and no alarms went off in the control room.  I still think, “so what?”.  People have been doing transfers at tank farms for over 100 years – it is a routine, a no brainer.  As far as I know, storage tanks do not have “low liquid alarms”.  And, why put in an expensive gas detection system for an activity that is so remote from the process operations and is something so routine?  To put it simply, the pumping activity was for a closed loop system:  it is static.  The drop in tank level alone should have told the operators that something is wrong (level was being measured).  Looks like the operators were simply “asleep at the wheel”.

Oh and one last thing, let us just assume that the liquid static pressure of the tank increased from 80 to 84 psi.  According to my rough calculations, that is a 16 ft rise in the tank level.  For a static system, that sounds serious.

Like I said, I am not an operations guy; but the phraseology of these “English majors” is confusing me.  Their writing just does not look right.  Any counter-point views are welcome.  Please feel free to comment.

Full report can be found at this URL:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s