Pulverize Kömür Kazanında Yakıcı Açılarının Alev Yapısı Üzerine Etkisinin İncelenmesi

Abstract

Kütahya ili sınırları içersinde yer alan Tunçbilek termik santralinin 150 MW kapasiteye sahip olan 5 numaralı kazanı sayısal olarak incelenmiştir. 5 numaralı kazan teğetsel yanma tipi olup, 3 farklı yükseklikte ve her yükseklikte 6 adet olmak üzere toplam 18 yakıcıya sahiptir. Her sıradaki yakıcıların 4 adeti köşelerde ve 2 adeti de karşılıklı yüzeylerde yer almaktadır. Kazanın katı modellemesi ANSYS Design Modeller R16.2 versiyonu kullanılarak yapılmıştır. Hazırlanan katı modelde Ansys Meshing R16.2 programı kullanılarak ve bölgesel iyileştirme yapılarak ayrık uzay oluşturulmuştur. Oluşturulan ayrık uzayda toplam 4.4 milyon eleman kullanılmaktadır. Yapılan analizlerde burgaç hareketini daha iyi yakalamak adına realizable k-ε çalkantı modeli ve Menther-Lencher duvar fonksiyonu kullanılmıştır. Kömür yanma odasına tanecikler halinde yollanmış ve kömür taneciklerinin takip edilmesinde Lagrangian yöntemi kullanılmıştır. Yanma reaksiyonu için ön karışımsız yanma modeli kullanılmış ve PDF tabloları oluşturulmuştur. Işınım modeli olarak P1 modeli seçilmiş ve sayısal analizde yanma odasındaki sıcaklık, katı ve gaz emisyon dağılımı hesaplanmıştır. PDF tablosu hazırlanırken santralden alınmış olan kömür elemental analizleri kullanılmış ve bu analizin sonuçları Fluent programı altında yer alan kömür hesaplama aracı kullanılarak PDF tablosu oluşturmada kullanılmıştır. Yapılan sayısal analizelerde gerçek boyutlarda kazan kullanılmıştır. Kazan hava ve kömür debileri sınır şartları belirtilirken sahadan alınan gerçek koşul verileri kullanılmıştır. Duvar sıcaklıkları ölçümler sonucunda suyun doyma sıcaklığından 50K yüksek olarak sabit sıcaklıkta tanımlanmıştır. Kazanın içindeki boruları modellemenin zorluğu ve işlem gücünün kısıtlı olması nedeniyle gözenekli ortam yaklaşımı yapılmış ve analizler bu yaklaşım ile tamamlanmıştır. Bu çalışmada 4 farklı durum üzerinde durulmuştur. Mevcut durum, yakıcı açılarının yukarıya kaldırıldığı ve aşağıya indirildiği durumlarla karşılaştırılmış ve yanma gazları çıkış alanı 2 katına çıkarılmıştır. Yapılan bu incelemelerde öncelikle mevcut durum analizleri test sonuçları ile karşılaştırılmış ve sonuçlar kontrol edilmiştir. Sonuçların doğrulanması sonrasında ise sırasıyla sıcaklık ve hız dağılımları incelenmiş, yanma gazları olan karbonmonoksit ve karbondioksit gazları karşılaştırılmış ve son olarak kirletici gazları olan hidrojen sülfür, azot oksit ve diazot monoksit gazları salınımları karşılaştırılmıştır.

Today’s World wealth level measured with the usage and generation of the energy, mostly the electric energy. In Turkey, we have coal beds so the coal-fired power plants became very important. When making a literature survey, you can see that there is lots of new publication about this subject. It also show that coal-fired power plants still very important for electricity producing. In this thesis, the Tunçbilek coal-fired power plant’s number five boiler was investigated. This boiler is tangential boiler and it could provide 150 MW power. It has three level of coal injectors and each level has six injectors, four of them stays on the corners and two of the stay on the opposite sites. Solid model of the boiler was modelled with Ansys Design Modeller R16.2. The original geometry was modelled for the analysis. The boiler has 69 m length, 12 m width and 12m depth. It is modelled from slag zone to exhaust outlet zone. The tangential boiler has 3 level of burner. First level stays on the 10.4 meter height, second level stays on 13.03 meter and the last one stays on 15.66 meter. First level’s burners stay parallel to the ground, but the second and third ones have upper 7° angle regarding to the ground. Meshing of the geometry was prepared with Ansys Meshing R16.2 program. Tetrahedral mesh method was chosen because of the automatic meshing procedure. Model have 4.4 million elements. On the model, both coarse and fine elements were used. On the fire region, and inlet and outlet boundaries, fine mesh was used. Other sides, of course with the reasonable growth rate, coarse mesh used for the calculation time and required memory. For computation fuid dynamics analysis, realizable k-ε turbulence model was used. Realizable k-ε turbulence model can be used wide range of application and tested lots of time. Realizable k-ε turbulence model predict better results for the rotational flow compared to the standard k-ε turbulence model so for this problem, rotational flow is dominant. Also two equations models required less computational power and computational time. For near wall treatment model, Menther-Lencher method was used. According to the Fluent, this model is independent y+ model. Because of the size of the geometry, boundary layer could not be used results of the computational power. Boudary layer was modelled with Menter Lencher method. Non-Premixed PDF model was used to mimic combustion reaction. Emprical coal fuel was calculated with coal calculater tool under the Fluent R16.2. The required data for coal calculater was getting from elemental analysis of the coal. For the radiation, P1 radiation model was used for the analysis. Also radiation from coal particles included to analysis. P1 model can provide reasonable results with less computation power regarding to the other models on the Fluent. Calculation of the pollutants, NOx calculation model was used. NO, HCN and N2O gases investigated for the analysis. Lagrangian method of particle track was used for coal particles. Injection of the coal was defined for the injectors. %43 of coal injected from bottom injectors, %35 of coal injected from the middle injectors and %22 of coal injected from the upper injectors. Rosin-Rammler method was used to define coal distribution between measured maximum and minimum diameter. Also mass flow, velocity magnitude and temperature was defined for the coil injector boundary condition. Pipes and economizer of the boiler was not modelled since they have very complex geometries and they need very fine mesh to resolve. Porous zone was used to simulate pressure drop. Porosity ratio was taken as 0.8 to model it. For the walls, no slip boundary condition was defined. The temperature of the walls are 659 K. %43 of air was sent to the lower injectors, %43 of the air was sent to the middle injectors and %14 of total air was sent to the upper ones. Mass flow inlet condition used to define. For the analysis, total 4 different case was prepared. The first one is the standart case. The second case, case 2, injectors have a bigger angle between ground. This case’s injecteros look up. The third case, case 3, injectors have smaller angle (negative angle for bottom injectors) regarding to the standard case. Its injectors look down. And the forth case, case 4, have a bigger outlet are compared to the standart case. The outlet area was doubled for the analysis. First of all, case 1 compared to the experimental results. While the results examined, there were temperature value for 8 points and average temperature for outlet. Just one point have %34 uncertanity. This point stays at the top side of the furnace. It could be a result of the porous region approximation. After the verification of the model was complited, parametric studies were done. While we investigated different cases, It can be seen that the maximum temperature was 1500K. On the case 3 temperature is lower than the other cases below injectors, but its temperature also increase to 1500K slower than others. It is the result of the higher fuel accumulation below to the injectors. Also when the different cases analysis, It can be easily seen that the bottom injectors create the center of the air-fuel vortex. The middle lines can penetrate a little to the center of the vortex however the top line cannot penetrate the vortex. Also when the residence time of the coil particules were investigated, case 4, case which has the bigger outlet, had the smallest residence time. Carbon dioxide emmision is the similar for all of the four cases. However corbon monoxide and unburned carbon emmissions is the biggest for case 4. It can be the result of small residence time. Also the poisonus gases, hydrogen cyanide (HCN), nitrogen monoxide (NO), and nitrogen dioxide (N2O) emissions were analysed. All of the four cases have similar emission for these gases. On the analysis, It can be seen that dross area was the biggest cause of this emissions. On the real life, dross are regularly cleaned. However for the analysis, the boundary condition of the bottom furnace, which is the dross are, used as wall. All of the cases are examined and it can be seen that the best case is the standart case, case 1. Temperature distrubiton on the contours are the similar for the cases. However, the emmissions are smallest on the case 1. On the future works, it can be changed air fuel ratio on the injectors. For this study, one group of injektors send only air. Also each line of injecktors have different fual – air ratio. And as it seen on the analysis, middle and top injectors could not penetrate the vortex. Their speeds and their tangental angle can be changed to investigate the flame characteristic of the furnace.