Pre-Planning for Refinery Vapor Releases

 

Vapor releases at industrial facilities handling flammable liquids or gases can result in some of the most challenging incidents an industrial responder can face. Process vapor releases can result in three primary types of incidents:

1. A vapor release without ignition,
2. A vapor release with delayed ignition resulting in a vapor cloud explosion (VCE) (discussed in part 1),
3. A vapor release with immediate ignition resulting in a jet fire.

This is intended to be a starting point for discussion and to illustrate how existing chemical safety analyses can be used in pre-planning. 

Most facilities have had some kind of vapor release analysis conducted. The output is frequently expressed as downwind vapor concentration, often in parts per million (ppm). It is important not to mix units when looking at ignition potential and personnel exposure. Ignition potential is related to the explosive limits, stated in percent vapor concentration.

An example value for the lower explosive limit (LEL) is 2 percent. Two percent corresponds to 20,000 ppm. The LEL could be well over tolerable exposure limits for many chemicals. Therefore, the toxicity threat can extend far beyond the ignition threat area.

There are many vapor release models available with varying degrees of sophistication. The more sophisticated they are, the more variables they can account for. Besides the model’s variables, there will be a wide range of input assumptions. It is important for responders to understand the assumptions.

Responders should ask what would happen if the release scenario was more severe. Additional models may be appropriate. One of the most common models used in the United States is the Computer-Aided Management of Emergency Operations (CAMEO) software suite and its program Areal Locations of Hazardous Atmospheres (ALOHA). CAMEO and ALOHA are easy to use and excellent for planning. Details, along with examples of their outputs, are readily available from an internet search.

An actual incident can unfold in a number of ways. A model is intended to be the basis of discussion about what might be encountered. For example, if control rooms are exposed to toxic vapors, some alternative means of process shutdown may be needed.  

The third type of release, a jet fire, results from immediate ignition of a pressurized flammable vapor release. This differs from a pool fire (or dike or tank fire) in that the force of the pressurized release can project the flames towards hazards and that foam is not effective on a jet fire. The pressurized flame can result in direct flame impingement or radiant heat exposure.

Jet fire radiation threats can be modeled with the same types of software used for vapor release and vapor cloud explosions. They can then be used to plan for BLEVE (boiling liquid expanding vapor explosion) threats from vessels exposed to the jet, cooling water needs, and for personnel exposure. Manual calculation methods can be found in the SFPE Handbook of Fire Protection Engineering.   

This article is a simplification of a very complex topic. Even the most sophisticated models cannot determine exactly how an incident could unfold. Combinations of VCEs and then subsequent jet fires can occur. These jet fires can then in turn expose pressure vessels with a resulting BLEVE threat.

The issues discussed here, along with this basic model, can help guide discussions and serve as a starting point for more in-depth evaluation. Flexibility during an operation is essential, but the skills and thought processes developed in preplanning sessions can guide decisions during an operation, no matter how it unfolds.