Ignition Mechanism

What is Ignition? [2] 

"ignition" is defined as "the sudden transition to a self-sustained, high-temperature oxidation reaction."

 

How does ignition occur? [2] 

Ignition can occur as the result of "autoignition" or "forced" ignition, and in some cases, materials can self-ignite. 

• Autoignition - by temperature, pressure, or volume
• Other self-induced ignition- by static discharge (Upon release, this charge may be imparted to the
  released material)
• Forced ignition- by flame 

HC이 대기으로 누출되어 ignition이 되고 화재/폭발이 발생하는 mechanism은 아래와 같이 구분할 수 있다. [1]

(1)  Operating Temperature> AIT, a spontaneous combustion

(2)  Operating Temperature < AIT, ignited by ignition source such as flame, spark or hot surfaces.

Ignition does not take place at the flashpoint unless there is an ignition source. If the liquid is heated further, at some point (AIT) it will ignite without the application of an external ignition source. This temperature is higher than the flashpoint.  Ignition source가 제공하는 energy가 누출된 유체의 온도를 AIT까지 올려서 ignition된다.  

 

Ignition 시기의 구분 

사전적의미는 하기를 참조하면 된다.

Risk analysis를 위해 각 화재/폭발 시나리오의 발생 빈도/확률을 estimation하기 위해 구분하는 Immediate/delayed의 구분 기준은 충분히 누출되어 gas cloud를 형성할 정도로 지연된 후 점화(Vapor cloud explosion)가 되는 것을  delayed ignition이라고 정의하고  그렇지 않다면 immediate ignition이다. 

Immediate ignition의 가능성에 영향을 주는 인자는 AIT, MIE, Discharge Velocity, Pressure 

Dealyed ignition의 가능성에 영향을 주는 인자는 Strength, duration of exposure, release rate/amount, material being released. distance from the point of release to ignition source, meteology, Events originating indoors

 

Immediate ignition 

CCPS [2] 

Early Ignition. Ignition that occurs closely enough to the source of a flammable release in space and time such that the possibility of a flammable cloud sufficient to create a vapor cloud explosion is precluded. 

 

QRA에서 ignition probability를 정하기 위해 어떤 회사에서는 누출 후 1min 내에 점화되는 경우를 immediate ignition이라고 정의하기도 한다. (ADNOC)

 

Moosemiller [3] 

The term “immediate ignition” is widely used in event tree work and so is utilized here, although “prompt ignition” is a more accurate description of this branch of the event tree. This point is defined as ignition at or near the point of release, and occurring quickly enough to preclude formation of an appreciable vapor cloud. In this category, we would expect to include ignitions due to the released material being near or above its autoignition temperature, and events that are ignited due to static or other energy sources at or near the point of discharge.

 

Delayed ignition 

Moosemiller [3] 

For the purposes of this discussion, a “delayed ignition” is defined as any ignition that does not meet the definition of an
“immediate ignition” as defined earlier. Delayed ignition allows time for a flammable vapor cloud to form. In this category, we would expect to include ignitions from the released material contacting an external ignition source such as rotating equipment, fired heaters, passing vehicles, etc.

 

Conservatism to considering each ignition. 

Conservatism vs Accuracy [2]

As a general principle, risk analyses of whatever type should strive for accuracy,  but where accuracy is not absolute, the analyses should err on the side of conservatism  (overstating the risk). This principle could be incorporated in two ways in this book: 
1. Overall Conservatism—The final result should err on the side of conservatism.
2. Conservatism with Respect to Expected Level of Accuracy—Since more effort is put into a Level 3 analysis than a Level 1 analysis, it would be preferred that a benefit in terms of reduction in conservatism would be associated with the extra effort needed for the Level 3 analysis.

 

Case 1- Overall ignition probability in perspective from Rigorous vs Simple 

Generic value - overall ignition probability  0.3 (Simple approach -historical data) 

만약 HC source 근처에 fired heater가 있다면 ignition은 훨씬 확실하다. (certain) 그러므로 rigorous estimation이 훨씬 보수적인 결과일 수 있다. 

 

Case 2-  Probabilities of Immediate Ignition vs. Delayed Ignition/Explosion

Ignition probability 측면에서는 immediate ignition이 100% 발생하는 것이 보수적인 접근이지만, delayed ignition을 무시하면 최악의 consequence가 될 수 있는 VCE의 가능성을 배제하는 것이라 이를 보수적인 접근이라고 보기 어렵다. 

 

Events being "conservatively" predicted to have a higher probability of immediate ignition than might be expected based on historical data. Particularly if there are buildings/populations in the immediate vicinity that are vulnerable to fire.
However, the conservative assumption of 100% immediate ignition precludes the possibility of a delayed ignition. The consequences of a delayed ignition could be much more severe than an immediate ignition, particularly if the outcome is an explosion. Thus, eliminating the possibility of an explosion by developing a conservative algorithm for the probability of immediate ignition defeats the intent of erring on the side of conservatism.

 

개별적으로 계산하는 가장 정확할 것이다. 특히 Immediate는 Delayed에 비해 훨씬 그렇다. 아무리 그렇다고 해도 Historical data 혹은 assumption 들이 적용되기에 어느 정도 이상의 conservatism이 적용된다. Accuracy를 헤치는 정도는 아니어야 함을 잊지 않아야 한다. 이는 Moosemiller [3]의 계산법에서도 알 수 있다. 

 

References

1. Are hot pipes and equipment surfaces ignition sources in the HPI? Hydrocarbon Processing
2. CCPS, GUIDELINES FOR DETERMINING THE PROBABILITY OF IGNITION OF A RELEASED
   FLAMMABLE MASS
3. Development of algorithms for predicting ignition probabilities and explosion frequencies, Moosemiller