441487 Dynamic Simulation Modeling to Protect Flare Headers from Air Ingress

Wednesday, April 13, 2016: 9:00 AM
362 A, B, D & E (George R. Brown )
Jesse Mumm1, Satyajit Verma1 and Freeman Self2, (1)Process Engineering, Bechtel OG&C, Houston, TX, (2)Process, Bechtel OG&C, Houston, TX

The pressure inside a flare header fluctuates depending upon the nature of relief and the ambient temperature. Because the flare tip is open to the atmosphere, the flare header system is prone to air ingress from outside even at small vacuums. This can result in potentially explosive air-hydrocarbon mixtures inside the header system, and because sources of ignitions are ubiquitous—the flare pilot being a guaranteed source—explosions are likely. The consequences of an explosion can range from damage to the flare header system to significant injury of plant personnel, lost production, and harm to company reputation

Air ingress into a flare header system is a distinct possibility when the cooling or condensation of hot vapors results in vacuum formation in the flare stack and header. For example, after a hot release event is over and the header is stagnant, vapors in the header will contract in volume as cooling or condensation occurs due to heat loss to ambient. If this rate of volume reduction is greater than the continuous seal purge flow to the flare stack, air from outside will start to flow into the flare stack and the flare header. It is only a matter of time until an explosive mixture of air and hydrocarbons will be formed inside the stack and the header system.

In order to avoid the reverse flow of atmospheric air into the flare header, make-up gas must be injected into the flare header to compensate for the reduction in volume of vapors in the header.

The paper will present the findings from dynamic simulations performed using a common commercial software package (details will be provided in the full paper).  The paper will address the five major considerations for selecting, sizing and specifying make-up gas requirements. 

  • The reduction in volume is a result of either cooling or condensation or both.  Cooling is slow and produces smaller reductions, while condensation results in rapid and significant volume reductions and is strongly affected by composition of the release.
  • The heat sink that causes cooling or condensation is primarily the ambient air, although cooling from a cold pipe wall, such as in cryogenic headers, may also occur.  
  • The method of measuring the rate of cooling depends on whether cooling or condensation is occurring.  Since heat loss is a function of the difference between ambient and gas temperatures, DT is a logical choice.  However, one may also use the condensing temperature for selected cases.  
  • Instrumented controls are used to start and stop the make-up gas injection. Sensors employed in control schemes can comprise of temperature difference, absolute temperature, header pressure, gas flow, oxygen concentration, and time.  Various combinations of these sensors may be used, and the implications for each will be discussed.
  • The locations of sensors and make-up gas injection points affect time lags and the rate of make-up gas addition required.  Practical rules for selecting locations of sensors will be explained.

These concepts will be illustrated using graphs and charts showing pressure and temperature profiles along the header, make-up gas rates, vacuum formation as a function of time, as well as other parameters that will help clarify the concepts.

Extended Abstract: File Uploaded
See more of this Session: Modeling of Relief and Flare Systems
See more of this Group/Topical: Global Congress on Process Safety