Chemical & Process Technology

Recent article "Designing for pressure release during fires - Part 1" by Saeid & Sanaz released in Hydrocarbon Processing, November 2007, has reveal two (2) major points in determining if an overpressure is possible for vessel contains liquid expose to fire and should the relieving flow calculate base on wetted basis or unwetted basis.

In this article, it proposes if total liquid heating and boiling time (deta t) after fire is initiated is higher than effective fire team response time (i.e. 20 min) and operating pressure of vessel exposed to fire at fire team taking response (P,atm) still lower than pressure set pressure (Ps), then the vessel will not subject to overpressure. Details may refer to following image.


As proposed in this article, the effective team response team (deta,t) shall be determined by design engineer but should not more than 15-20 minutes with reference to an article "Fires, vessels and pressure relief valve", Chemical Engineering May 2000 by W.Y. WONG. In additional API Std 521 section 5.12.1.2.2, has given an example where an unwetted ASTM A 515 Grade 70 Steel plate, 1 inch think, would take about 19.5 minutes to rupture.

This is a very good effort to address some liquid which is having very high initial boiling point like dead crude, hot oil, etc.

My two cent comment...

Effective fire response time
The definition and support for the effective fire response time is rather weak. Does the effective fire response time mean the arrival of fire rescue team ? Or it include the fire rescue team arrive plus control or put-off the fire ? How to evaluate and quantify if a fire is considered "control" or "put-off" ? Does it includes the automatic fire protection system ? There is no clear definition of this term. Personal interpretation all above are included.

Effective fire response time subject to many factors such as reliability and availability of fire protection system, location of fire, type of fire, environmental conditions like wind velocity and wind direction, level of difficulties assess to the location of fire, fire rescue team ability and skill, fire rescue team psychology condition, etc. There are many unknown and uncontrolled parameters to work out an effective fire response time. It need a lot of testing, investigation, modeling, history, real world example, etc in order to properly define correctly this parameter. Relying justification by design engineer would be a great risk.

Code Compliance
Fire scenario has been an issue discussed many times in many papers, forums and meetings. No concrete decision and solution from any party. As far as we are aware, any vessel designed to ASME and in compliant to API, Pressure relieving device shall be provided on the vessel to protect the vessel from fire attack. This clearly stated in ASME code. Adopting above procedure would potentially leads to conflict with vessel design code. This is be extended to other code such BS, GB, JIS, etc

Local Regulation Compliance
Most of the local code and standard would request vessel to be protected by pressure relieving device in case of fire attack. Similarly adopting above procedure would potentially leads to conflict with local code and standard.

Self Decomposition and Reaction
There are many products and reactants possibly decomposed and continue reaction even after a fire is controlled. Continuous decomposition and reaction (especially exothermic reaction) would further increase the system internal pressure and eventually reach system or vessel maximum allowable working pressure. This scenario will demand a pressure relieving device for overpressure protection.

Inert Gas Expansion prior to Liquid Boiling
There are many vessel containing high initial boiling point liquid are blanketed with inert gas i.e. nitrogen. In the event of fire attack, entire system will be total isolated by plant emergency shutdown system. Fire may engulf entire vessel and part of the heat may transfer into inert gas. This leads to gas expansion even though the liquid temperature still far below it initial boiling temperature. Inert gas expansion may leads to increase of internal pressure and potentially reach vessel maximum allowable working pressure. This scenario will demand a pressure relieving device for overpressure protection.

Summary
Above procedure is considered a good effort to address fire attack for high initial boiling liquid. However, there are still many factors and conditions shall bed assessed prior to application of this procedure. The proposed method shall be used with EXTRA CAUTIOUS. In addition, there are still a assessment procedure remain open for discussion and a lot of work and effort required to back-up this procedure.

Related Topic

• Should we consider JET FIRE for Pressure Relief Valve (PSV) load determination ?
• Should we install Butterfly valve for Pressure Relief Valve (PSV) isolation ?
• Another descrepancy found in API Std 521Jan 2007
• Requirement of overpressure protection devices on system design to PIPING code
• ERRATA - API Std 521, Pressure Relieving and Depressuring Systems
• Workbook for Chemical Reactor Relief System Sizing
• A must have book...Emergency Relief System Design Using DIERS Technology