Consequence Analysis for Chemical Releases

Consequence analysis is the systematic process of predicting the potential impacts of an accidental chemical release. It forms the critical technical backbone of quantitative risk assessment (QRA), informing decisions on plant design, safety systems, and emergency response planning. Without accurate consequence modeling, we cannot reliably understand the hazards we design against or effectively protect people, property, and the environment.

Source Term Modeling: Defining the Release

Every consequence analysis begins with defining the source term, which describes the rate, duration, and physical state of the chemical release. This step is paramount, as errors here propagate through all subsequent calculations. The release scenario is dictated by the failure mode (e.g., hole size, full rupture) and the chemical's physical state.

For a liquid spill, pool evaporation is a key model. It calculates the rate at which a spreading liquid pool evaporates into the atmosphere. The rate depends on pool area (driven by ground properties and containment), liquid temperature, and the chemical's vapor pressure. A common simplifying approach uses the pool's boiling point; if the chemical is above its boiling point, evaporation is very rapid (boiling), whereas for a cryogenic or ambient-temperature liquid, evaporation is governed by heat transfer from the ground and air.

A jet release (or two-phase jet) occurs when a pressurized liquid or gas escapes through an orifice. For a pure gas, the flow is sonic (choked) initially and is calculated using standard thermodynamic equations. A two-phase release is more complex and common, occurring when a pressurized liquid (e.g., LPG) flashes upon release due to sudden pressure drop. A portion vaporizes instantly, cooling the remaining liquid, which may then form an aerosol. The total mass flow rate and the fraction that is vapor versus aerosol are critical outputs for the next stage of analysis.

Atmospheric Dispersion: Tracking the Cloud

Once the source term is defined, the next step is predicting how the released material disperses downwind. The choice of model hinges on whether the released vapor is buoyant or dense relative to air.

The Gaussian dispersion model is the most common for neutrally buoyant or positively buoyant gases. It assumes the dispersing cloud spreads in a statistically normal (bell-shaped) distribution both horizontally and vertically from a central axis. Key inputs include the release rate, meteorological conditions (wind speed, atmospheric stability), and height of release. While it involves simplifying assumptions, its relative simplicity makes it a standard for initial screening assessments. It outputs downwind concentration contours, which are used to define hazard zones for toxicity or flammability.

For chemicals heavier than air (e.g., chlorine, cold propane), a dense gas dispersion model must be used. A dense cloud slumps and spreads radially under gravity, resisting turbulent mixing. Its behavior is fundamentally different from a Gaussian plume. Specialized models (like SLAB or DEGADIS) account for this gravity-driven spreading, ground-level heating, and eventual transition to passive dispersion. Ignoring dense gas effects can lead to severe underestimation of ground-level hazard ranges.

Fire Consequence Models

If a flammable release ignites immediately, the result is a fire. The type of fire depends heavily on the release mode and ignition timing.

A pool fire results from the ignition of a burning vapor cloud above an evaporating liquid pool. Its primary hazard is thermal radiation. Modeling involves calculating the pool's steady burning rate, flame geometry (height, tilt), and the radiant heat flux at various distances. View factors (the geometric relationship between the flame and a target) are crucial for accurate radiation estimation.

A jet fire occurs when a pressurized flammable release ignites. It is a turbulent diffusion flame with significant momentum, often causing intense localized heating and flame impingement. Models calculate flame length and shape (which may be tilted by wind) and the radiation field, which is typically more focused and hazardous near the source than a pool fire.

A flash fire is the combustion of a flammable vapor cloud that has dispersed some distance before finding an ignition source. It does not produce significant overpressure but results in a rapidly moving flame front that burns within the flammable limits of the cloud. The consequence is primarily fatal burn injury to anyone within the cloud's footprint at the moment of ignition. The hazard range is therefore defined by the earlier dispersion modeling.

A BLEVE (Boiling Liquid Expanding Vapor Explosion) is a complex and devastating event. It occurs when a vessel containing a pressurized liquefied gas (like LPG) fails catastrophically, often due to fire impingement weakening the shell. The sudden pressure loss causes violent boiling and explosive vaporization of the entire contents. This results in two primary hazards: a powerful pressure vessel burst mechanical explosion generating projectiles and a blast wave, and, if the substance is flammable, a massive fireball. Fireball models predict diameter, duration, and intense thermal radiation.

Explosion Models

Explosions generate damaging overpressure (blast waves). Two main types are modeled in consequence analysis.

A VCE (Vapor Cloud Explosion) occurs when a large, partially confined flammable vapor cloud ignites. The flame front accelerates through turbulence created by obstacles (pipe racks, buildings), transitioning from deflagration to a high-speed combustion that generates significant overpressure. Models like the TNT-equivalent method or more advanced multi-energy methods estimate peak overpressure and impulse at distances from the blast. Key influences are cloud size, congestion/confinement level, and fuel reactivity.

A pressure vessel burst (or physical explosion) involves the sudden failure of a gas-filled or superheated liquid vessel. The stored mechanical energy is released almost instantaneously, generating a blast wave. The overpressure decay with distance is typically modeled using methods derived from the Brode equation, which relates released energy (based on vessel pressure and volume) to blast parameters.

Common Pitfalls

1. Misapplying Dispersion Models: Using a standard Gaussian plume model for a heavy gas like cold propane. This will predict a narrow, elevated plume, drastically under-predicting the ground-level hazard area. Correction: Always check the density of the released vapor relative to air. If it's denser, use a dense gas dispersion model.
2. Ignoring Release Duration: Treating all releases as instantaneous or continuous without justification. A small hole from a large inventory leads to a continuous release, while a catastrophic vessel failure is instantaneous. Correction: Define the release duration based on the specific failure scenario and isolation system response time. The duration directly affects the total quantity released and the dispersion outcome.
3. Confusing Fire Types: Assuming a jet release will always cause a flash fire. If a pressurized flammable jet ignites immediately at the breach, it is a jet fire. A flash fire requires delayed ignition of a dispersed cloud. Correction: Clearly link the ignition timing and location to the physical release scenario defined in the source term.
4. Overlooking Blast Load Interactions: Calculating explosion overpressures in isolation without considering subsequent effects like structural response or fragment hazards from a BLEVE. Correction: In a full risk assessment, combine blast, projectile, and thermal effects from a single event where applicable, as they can act synergistically to increase damage.

Summary

• Consequence analysis is a sequential modeling process: first define the source term (release rate and phase), then model atmospheric dispersion, and finally apply appropriate fire and explosion models based on material properties and ignition conditions.
• The choice of dispersion model (Gaussian vs. dense gas) is critical and depends on whether the released vapor is heavier or lighter than air; using the wrong model invalidates the hazard range prediction.
• Fire consequences are scenario-specific: pool fires and jet fires present severe thermal radiation hazards, a flash fire danger zone is defined by the pre-ignition vapor cloud dispersion, and a BLEVE combines a destructive blast wave with a massive fireball.
• Explosions from a VCE depend heavily on vapor cloud size and the degree of congestion/obstruction to cause flame acceleration, while a pressure vessel burst releases stored mechanical energy directly as a blast wave.
• Accurate modeling requires precise definition of the failure scenario—hole size, release duration, chemical state, and ignition parameters—as small changes in these inputs can lead to large differences in predicted consequences.