H. RODRIGUES, Siemens, Lisbon, Portugal; A. SAFAR, Siemens, London, UK; and A. TIJANI, Saudi Aramco, Dhahran, Saudi Arabia
Process operating facilities typically process large inventories of gas and volatile hydrocarbon liquids at high pressures, which have the potential of posing major hazards during emergency incidents—such as fire, loss of containment or operating excursion. In such situations, pressure relief devices will not adequately protect or safely depressurize the high-pressure system, as it can only prevent overpressure. For example, the mechanical integrity of an unwetted vessel exposed to fire may fail at a pressure lower than the relief valve set pressure, causing a hazardous vapor cloud. Generally, depressurizing systems are designed to:
Given the importance of depressurizing systems, it is imperative to adequately determine the required depressurizing for a given period in conjunction with any constraints in the flare system, such as capacity or metrology.
Depressurizing valve sizing and optimization. The specific design of emergency depressurizing systems depends on the risk involved (such as fire or leakages) and the depressurizing design philosophy for unit operation. Based on plant experiences in oil, natural gas and refining facilities, the depressurizing rate is mainly dictated by mitigating the consequences of equipment failure due to fire exposure. In such cases, the depressurizing system is designed to reduce the equipment pressure rapidly to a safe level—typically to 100 psig or 50% of the equipment design pressure, whichever is lower—in 15 min, as specified in API STD 521 4.6.6. This criterion is applicable for pool fire scenarios and for wall thicknesses greater than 1 in., but it may not be suitable for lower wall thicknesses or for jet fires that require further analysis to ensure equipment survivability.
Jet fire exposure is a localized flame impingement exerted on vessels or piping systems that may cause rapid equipment failure, often without increasing the system pressure to the maximum design pressure. As such, it is important to evaluate the equipment survivability when designing the emergency depressurizing system. This is a dynamic analysis involving multiple variables such as heat flux, material type, thickness and fluid thermodynamic behavior. Updated industry standards—including the 7th edition of API STD 521—now recognize the need to consider vessel stress during blowdown by considering the analytical method (API 521 Annex A) and calculating stress, as per API or other methodologya. The following general acceptance criteria—outlined in methodologya—are that exceedance of any of the following makes rupture unacceptable:
Furthermore, the flare system’s capacity may influence the depressurizing system design philosophy. For example, excess flare capacity should be utilized by increasing the depressurizing rate (depressurizing faster than 15 min). Conversely, capacity constraints might be mitigated by using staggering logic or by limiting the initial opening of the valve, inclining it to the fully open position over specified periods, such that the average blowdown rate is maintained and the system depressurized. This article will explore the staggering option and showcase how vessel fire survivability may be used to justify a (simpler) design with a single restriction orifice and a depressurization longer than 15 min to 100 psig.
Problem statement. FIG. 1 illustrates a process segment consisting of a propane condenser, propane accumulator vessel and piping system that are all part of the propane refrigeration system. The segment is bounded by two isolation valves—used in emergency cases—to isolate the system. The design of an emergency depressurizing valve for the process segment should consider the flare system capacity constraint of 25 MMft3d. The following are the depressurization conditions:
Analysis approach. A detailed geometric representation that reflected the distributed nature of the blowdown segment was created using proprietary modeling softwareb that enabled the full assessment of the transient behavior of the system during depressurization. The required relief orifice size was determined based on depressurizing the system to 100 psig within 15 min. The sizing analysis was conducted under the condition of a pool fire, with a time averaged heat flux as prescribed in API 521 Section A.3.2. Further dynamic optimization analysis was conducted to reduce the depressurizing peak flow to meet the flare system constraint of 25 MMft3d.
Two alternative solutions were explored. The first was to use staggering logic restriction orifices operating stepwise, which does not require any additional emergency shutdown activation buttons. The second alternative solution was to extend the depressurizing time beyond 15 min, supported by rigorous equipment survivability analysis to determine the required wall thickness and materials to avoid equipment failure.
Results. The dynamic depressurizing analysis revealed that the required restriction orifice (RO) size needed to depressurize the system to 100 psig within 15 min was 1.82 in., with a peak flow of 33.5 MMft3d, which exceeded the flare capacity constraint of 25 MMft3d (FIG. 2).
Using multiple ROs. To reduce the depressurization peak flow, an alternative depressurization solution was investigated using staggering logic ROs operating stepwise. The first valve opens at t = 0 min, while the second and third valves open at approximately t = 5 min and t = 10 min, respectively. The resulting total peak flow was 24.5 MMft3d. The calculated RO sizes are tabulated in TABLE 1. The profiles of the flowrates and pressures during depressurization are shown in FIGS. 3 and 4.
Fire survivability analysis. Although the multiple ROs option can achieve both goals of meeting the conventional 15 min to 100 psig guideline without exceeding the flare capacity constraint (25 MMft3d), using a single RO would be preferable. To this end, vessel stress was evaluated in case of a jet fire impinging on the cylindrical wall. The RO size was set to 1.56 in. The propane accumulator thickness was 1.97 in., and the material was carbon steel 235LT. The dynamic analysis revealed that rupture occurred at 12.5 min. At the time of rupture, the pressure in the vessel exceeded 65.3 psig, which violated the acceptance criteria outlined in methodologya. Therefore, this design was unacceptable.
Two other sensitivity cases were evaluated to demonstrate the impact on the material selection and vessel wall thickness during the fire survivability study—the results are summarized in TABLE 2. In addition, a graphical demonstration of the wall stress, tensile strength and temperature profiles during jet fire blowdown is shown in FIG. 5. Both sensitivities demonstrated how, by changing the material or wall thickness, the rupture becomes tolerable. Another alternative to mitigate the depressurization requirements—not demonstrated here—was the use of passive fire protection, since this option can introduce other challenges such as increased maintenance, cost and weight.
Takeaways. The intent of this article was to summarize the design considerations for adequate sizing of emergency depressurizing systems during fire exposure while considering flare capacity constraints. This article also provided a practical sizing example in which peak flow optimization and a fire survivability analysis were discussed. It was demonstrated that the flare capacity constraint can be met by considering stepwise staggering logic ROs that will reduce the peak flow during depressurization.
Another alternative was to extend the depressurization time beyond 15 min, supported by rigorous equipment survivability analysis to determine the wall thickness and materials to avoid equipment failure. Based on experience, care should be taken when analyzing stainless-steel vessels, as these vessels can withstand jet fire exposure for long periods before rupturing. In such cases, project guidelines should set a maximum allowable survivability criterion—e.g., a maximum depressurization time to 100 psig in 25 min—to avoid design issues later in the design phase when fire risk analysis studies highlight increased risk due to prolonged inventories that can sustain and escalate in the event of a fire. HP
NOTE
a Scandpower
b Siemens gFLARE software
REFERENCES
Hugo Rodrigues is a Principal Consultant Engineer at Siemens. He has more than 10 yr experience in flare, relief and blowdown studies. He also is responsible for the development of gFLARE software and integrated digital solutions. Rodrigues earned a MS degree in chemical engineering from Instituto Superior Técnico in Portugal. He is a chartered member of the IChemE.
Anas Safar is a Senior Consultant Engineer at Siemens. He has more than 9 yr of experience with Aramco in multiple technical areas such as refining operations, NGL fractionation, gas plants and utilities. Safar worked as a Flare and Relief Systems Engineering Consultant, providing technical consultations to all Aramco operating facilities, capital projects and affiliates. He earned a BEng degree from Heriot Watt University and a MSc from the University of Leeds. Safar is also a licensed professional engineer in Texas.
Abdulaziz H. Al-Tijani is an Engineering Specialist at Saudi Aramco’s Process and Control Systems Department in Dhahran, Saudi Arabia. He earned a BS degree in chemical engineering from King Fahd University of Petroleum and Minerals (KFUPM) and a MS degree in oil and gas surface facilities from KFUPM (in partnership with IFP). Al-Tijani supports company operations and project design primarily in flare and relief systems and flare gas recovery applications. He also supports Saudi Aramco and JV oil and gas operational facilities, pipelines, process simulations and various phases of projects.