Petrol rafinelerindeki yangınların CFD ile modellenmesi = CFD modeling of fires in oil refines

Abstract

Petrol tesislerinde yangınları modellemek için çeşitli yöntemler bulunmaktadır. Katı alev modeli, tek nokta kaynak modeli ve HAD modeli. Katı alev modeli ve tek nokta kaynak modeli, petrol tanklarındaki yangınları modellemek için yaygın olarak kullanılan iki yöntemdir. Katı alev modeli, tank içindeki yangının, iyi tanımlanmış şekli ve boyutu olan katı bir alev olduğu varsayımına dayanmaktadır. Bu model, yangını katı bir nesne olarak ele alır ve alev yüksekliği, şekli ve sıcaklık dağılımı gibi faktörleri dikkate alır. Tek nokta kaynak modeli ise, yangının ısı ve yanma ürünleri salan tek nokta kaynakla temsil edilebileceğini varsayar. Bu model, yalnızca ısı yayma oranını ve ışınımsal ısı akısı dağılımını dikkate alarak yangın davranışını basitleştirir. Hem katı alev modeli hem de tek nokta kaynak modelinin güçlü yanları ve sınırlamaları vardır. Modelleme yönteminin seçimi, çalışmanın özel amaçlarına, mevcut verilere ve istenen doğruluk düzeyine bağlıdır. Güvenilirliğini ve uygulanabilirliğini sağlamak için seçilen modeli deneysel verilere veya gerçek hayattaki yangın olaylarına karşı doğrulamak önemlidir. Petrol depolama tankları için HAD'de (Hesaplamalı Akışkanlar Dinamiği) bir yangını modellemek, yangın güvenliğini değerlendirmede ve etkili yangından korunma önlemleri tasarlamada önemli bir araçtır. Bu çalışma, petrol depolama tanklarındaki yangınların davranışını anlamak ve çevredeki ortam üzerindeki termal etkileri tahmin etmek için HAD simülasyonlarını kullanmaya odaklanmaktadır. Araştırma, depolama tanklarının geometrisini ve fiziksel özelliklerini doğru bir şekilde temsil eden kapsamlı bir HAD modeli oluşturarak başlar. Model, tank boyutları, malzeme özellikleri, yakıt özellikleri ve atmosferik koşullar gibi faktörleri içerir. Yangın kaynağı, yakıt-hava karışımı, tutuşma ve alev yayılımı dikkate alınarak uygun yanma modelleri kullanılarak simüle edilir. HAD simülasyonları aracılığıyla, alev şekli, sıcaklık dağılımı, ısı akısı ve duman üretimi dahil olmak üzere yangın davranışının çeşitli yönleri araştırılabilir. Simülasyon sonuçları, kütle yanma oranı için iki farklı değerle değerlendirildi. SFPA referans değeri olan 0.055 kg.m-2s-1 ile karşılaştırıldığında, 1.5, 3 ve 4 metre çapına sahip tanklar için sırasıyla 1.3, 1.28 ve 0.6 kW/m2 radiant ısı akısı elde edildi. Bu koşullar altında yayma gücü, sırasıyla 50.03, 67.95 ve 81.67 kW/m2 olarak hesaplandı. Aynı simülasyonlar, daha yüksek bir kütle yanma oranı olan 0.083 kg.m-2s-1 değeri ile de değerlendirildi. Sonuçlar, 1.5, 3 ve 4 metre çapına sahip tanklar için sırasıyla 1.84, 1.72 ve 0.7 kW/m2 radiant ısı akısı elde edildiğini gösterdi. Bu koşullar altında yayma gücü, sırasıyla 56.37, 73.8 ve 84.12 kW/m2 olarak hesaplandı. Simülasyonlardan elde edilen sonuçlar, petrol depolama tanklarındaki yangınlarla ilgili potansiyel tehlikeler ve riskler hakkında değerli bilgiler sağlar. Bu bilgi, yeterli xx havalandırma, yangın söndürme sistemleri ve acil durum müdahale planları gibi yangın önleme ve azaltma stratejilerinin tasarımına rehberlik edebilir. Ek olarak, HAD simülasyonları, değişen yangın boyutları, tank konfigürasyonları ve yangın söndürme stratejileri gibi farklı senaryoların değerlendirilmesine olanak tanır. Sonuçları karşılaştırarak ve analiz ederek, petrol depolama tesislerinin güvenliğini artırmak için en uygun tasarım seçimleri yapılabilir. HAD modelinin doğruluğu ve güvenilirliği, simülasyon sonuçlarının mevcut deneysel veriler ve gerçek hayattaki yangın olayları ile karşılaştırılmasıyla doğrulanır. Herhangi bir tutarsızlık dikkatli bir şekilde analiz edilir ve yangın davranışını tahmin etmedeki doğruluğunu sağlamak için modelde iyileştirmeler yapılır. Sonuç olarak, benzin depolama tankları için HAD'de bir yangının modellenmesi, yangın güvenliği analizi ve tasarım optimizasyonu için güçlü bir araç sağlar. Yangın davranışını doğru bir şekilde simüle ederek, potansiyel riskler belirlenebilir ve petrol depolama tesislerinde yangınların etkisini en aza indirmek için etkili yangından korunma önlemleri uygulanabilir.

A simulation was carried out to predict the possibility of a liquid hydrocarbon storage tank fire, its subsequent heat transfer to adjacent tanks through thermal radiation, and the response of the contents of the adjacent tank to this radiation. To predict whether an adjacent tank will ignite and when, it is crucial to quantify the radiant heat received by the adjacent tank from the fire. In tank farm scenarios, fires occurring in large-scale storage tanks exhibit two prominent characteristics. Firstly, extinguishing these large-scale tank fires poses significant challenges, often resulting in substantial material losses. Secondly, these fires can trigger domino effects due to the propagation of thermal radiation, leading to a chain reaction of incidents. Given that pool fires are common accidents in the process industry and consistently trigger domino effects, it becomes essential to model and simulate the risks associated with pool fires in order to accurately predict their behavior. Extensive research has been conducted on domino effects, exploring various aspects of their occurrence. Analyses of these effects in different industrial installations have revealed that the majority of domino accidents within process industries occur in atmospheric and cryogenic storage tanks. Furthermore, historical accident scenarios have highlighted fire as the predominant initiator for more than half of industrial domino incidents, with pool fires accounting for 44% of these fire-triggered accidents. A domino effect is characterized by an initial event originating in a specific item or unit, which then sets off a sequence of subsequent events in nearby items or units. These subsequent events are triggered by escalation vectors such as blast waves and heat loads. In the context of a pool fire, the primary escalation vector is the heat load, resulting from the combination of heat transferred through convection and radiation. For tank fires, the contribution of radiation to the total heat load ranges from 92% to 100%, while convection's contribution is between 0% and 8%. Consequently, in open pool fires and tank fires, radiation heat flux emerges as the principal driver of escalation. A range of mathematical models are available in the literature to calculate the radiant heat flux to a specified target, with each model being based on assumptions about the fire. Extensive literature discusses both experimental and theoretical investigations concerning thermal radiation emitted from flames. However, the majority of these studies have primarily centered around small-scale pool fires, which exhibit notable differences when compared to larger turbulent fires. Nonetheless, accurately predicting the radiative properties of these substantial flames remains challenging due to uncertainties associated with certain parameters unique to large turbulent diffusion flames. In the case of substantial hydrocarbon pool fires, substantial volumes of smoke are generated, enveloping the fire and leading to a reduction in thermal radiation. This phenomenon is commonly referred to as the "smoke blockage effect." Currently, xxii calculations and guidelines for determining safe distances between fires and structures adopt highly cautious approaches due to the lack of a dependable methodology for factoring in the smoke blockage effect. Furthermore, the significant impact of soot on the radiation emitted by hydrocarbon fires is widely acknowledged. Regarding oil fires, several methods are available for modeling fires in oil reservoirs, including the solid flame model, the single point source model, and the CFD model. The solid flame model and the single point source model are two commonly used methods for modeling fires in oil tanks. The solid flame model operates under the premise that the fire inside the tank assumes the form of a solid flame, characterized by well-defined dimensions and shape. This approach treats the fire as a cohesive entity, taking into consideration variables such as flame height, shape, and temperature distribution. Despite its consideration of the wind's impact, the solid flame model is subject to limitations. Primarily, these models fail to capture the intricate dynamics of pool fires at smaller eddy scales, owing to the absence of advanced turbulence models. Furthermore, their applicability falls short when it comes to accommodating flame tilting in scenarios involving wind influences. Additionally, these models do not provide accurate predictions for the behavior of pool fires in situations involving complex geometric configurations. On the other hand, the single point source model assumes that the fire can be represented by a single point source that releases heat and combustion products. This model simplifies fire behavior by considering only the heat release rate and the distribution of radiative heat flux. However, this model has several limitations, such as its applicability being limited for fires with a diameter larger than 5 m. Additionally, when modeling large pool fires with smoke effects, it does not account for wind velocity and direction, leading to underestimation of thermal radiation at shorter distances. Both the solid flame model and the single point source model have their strengths and limitations. The choice of modeling method depends on the specific objectives of the study, available data, and desired level of accuracy. It is important to validate the chosen model against experimental data or real-life fire incidents to ensure its reliability and applicability. While experimental investigations into pool fires have extensively covered small and medium scales for various hydrocarbon fuels, including crude oil, gasoline, kerosene, jet propellant (JP-4), heptane, diesel, liquefied natural gas (LNG), and ethanol, The large-scale pool fire experiments remain scarce. This scarcity can be attributed to the considerably higher costs associated with conducting large-scale pool fire studies. As a result, there is a growing emphasis on utilizing Computational Fluid Dynamics (CFD) for fire modeling, primarily due to the inherent challenges posed by experiments, especially for larger-scale fires. But It's important to note that CFD simulations can demand significant computational time. One notable advantage of CFD-based simulations is their potential to offer more accurate outcomes compared to traditional empirical models like solid flame and single-point source models. This improved accuracy arises from the consideration of factors such as geometric obstructions and the ability to handle complex accident scenarios. The ongoing progress in computational technology has further bolstered the development of fire-specific computer codes. CFD codes, such as the Fire DynamicsSimulator (FDS), possess the capability to offer temporal resolutions that yield valuable insights when evaluating potential scenarios. Nonetheless, successful CFD fire simulations require valid assumptions and precise boundary conditions. Despite their potential, these simulations are not devoid of challenges, and their accuracy is contingent on various factors, including model setup, grid resolution, and turbulence modeling. As computational technology continues to evolve, the application of CFD for fire modeling holds promise for delivering more detailed and contextually nuanced results. Modeling a fire using CFD (Computational Fluid Dynamics) for petrol storage tanks is an essential tool in assessing fire safety and designing effective fire protection measures. In this study, the Fire Dynamics Simulator (FDS) is adopted to simulate tank and dike pool fires in a tank farm. This study focuses on utilizing CFD simulations to understand the behavior of fires in petrol storage tanks and predict the thermal effects on the surrounding environment. The research begins by constructing a comprehensive CFD model that accurately represents the geometry and physical properties of the storage tanks. The model incorporates factors such as tank dimensions, material properties, fuel characteristics, and atmospheric conditions. The fire source is simulated using appropriate combustion models, considering the fuel-air mixture, ignition, and flame propagation. Experimental data was taken from the 2004 Munoz study in which a series of large outdoor pool fire experiments were conducted using gasoline and diesel fuel placed on top of a layer of water. Five concentric circular pools made of reinforced concrete (1.5, 3, 4, 5, and 6 m in diameter) were used. The experiments were filmed with at least two video cameras registering visible light (VHS) and a thermographic camera (IR). In Munoz's study, thermographic images were used to determine the flames' distribution of emissive power, the mean emissive power, and the radiant heat flux. The contribution of each part of the flame to the total radiated energy was analyzed. A method is presented combining the IR images and the visible images; it offers further insight into the relationship between the heat emitted by the luminous part and the obscured, non-luminous part of the flame. This study analyzed the impact of two different mass burn rate values: 0.055 kg.m-2s-1 (the SFPA reference) and 0.083 kg.m-2s-1 (the value used in Munoz's study). The results demonstrate the effect of these two different mass burn rates on the radiant heat flux for storage tanks of diameters 1.5m, 3m, and 4m. When the value of 0.055 kg.m-2s-1 was applied, radiant heat fluxes of 1.3 kW/m2, 1.28 kW/m2, and 0.6 kW/m2 were obtained for the 1.5m, 3m, and 4m diameter tanks, respectively. These results were then calculated as emissivity values of 50.03 kW/m2, 67.95 kW/m2, and 81.67 kW/m2, respectively. On the other hand, when a mass burn rate of 0.083 kg.m-2s-1 was applied, radiant heat fluxes of 1.84 kW/m2, 1.72 kW/m2, and 0.7 kW/m2 were obtained for the 1.5m, 3m, and 4m diameter tanks, respectively. The emissivity values were calculated as 56.37 kW/m2, 73.8 kW/m2, and 84.12 kW/m2, respectively, under these conditions. Through the CFD simulations, various aspects of fire behavior can be investigated, including flame shape, temperature distribution, heat flux, and smoke generation. The results obtained from the simulations provide valuable insights into the potential hazards and risks associated with fires in petrol storage tanks. This information can guide the design of fire prevention and mitigation strategies, such as adequate ventilation, fire suppression systems, and emergency response plans. xxiv Additionally, CFD simulations enable the evaluation of different scenarios, such as varying fire sizes, tank configurations, and fire suppression strategies. By comparing and analyzing the results, optimal design choices can be made to enhance the safety of petrol storage facilities. The accuracy and reliability of the CFD model are validated by comparing the simulation results with available experimental data and real-life fire incidents. Any discrepancies are carefully analyzed, and improvements to the model are implemented to ensure its accuracy in predicting fire behavior. In conclusion, modeling a fire in CFD for petrol storage tanks provides a powerful tool for fire safety analysis and design optimization. By accurately simulating fire behavior, potential risks can be identified, and effective fire protection measures can be implemented to minimize the impact of fires in petrol storage facilities. Therefore, the quantitative results obtained through FDS modeling can be used to quantitatively assess the risks of a tank farm and determine safe separation distances between tanks.