Heat transfer in oxy-fuel fluidized bed boilers.pdf

Heat transfer in oxy-fuel fluidized bed boilers

Bolea Agüero, Irene
Romeo Giménez, Luis Miguel (dir.)

Universidad de Zaragoza, 2013

Resumen: In spite of the stabilization of coal demand in developed countries, the role of coal in the next decades energy mix is still essential. Particularly relevant will be in the great developing economies, such as India or China, where this fuel is abundant and avoid external energy dependences. In parallel, the international community needs to drive its efforts towards politics that commit fossil fuels energetic companies to drop their CO2 emissions drastically for 2015. In this regard, great advances have been made towards gaining plant efficiency and therefore, reducing the tones of CO2 per produced kWh. Still, emissions need a more drastic reduction if we want to avoid an increment of atmosphere temperature higher than 2ºC. Here, the CO2 capture and storage (CCS) technologies will have the potential of reducing up to 25% of CO2 from stationary sources as soon as they will be commercially available. Among the CO2 capture technologies, grouped in pre-combustion, post-combustion and oxy-fuel combustion, this last one is receiving outstanding support by the national and European authorities. The possibility of implementing oxy-fuel combustion into circulating fluidized bed technology, contributes to approaching the concept of clean-coal technology. Fluidized bed combustors have the outstanding feature of offering the possibility of burning a wide variety of fuels They have the possibility to capture SO2 emissions, adding in-bed limestone. Their working temperature is lower than in pulverized fuel boilers, which avoids thermal NOx formation. Additionally to these characteristics, already exploited under air-firing, applying oxy-fuel combustion technology and being able to capture the CO2 emissions from the coal combustion, or even from blends of coal and other fuels, makes oxy-fuel combustion in fluidized bed a great opportunity to turn the coal sustainable in the future power plant designs. About the implications derived of applying oxy-fuel technology to a commercial scale CFB boiler, scarce literature exits, especially when considering high O2 concentrations at inlet. A one dimensional model has been developed. The overall modeling strategy, in which the model has been based on, is explained in the first part of Chapter 2. It is based on the already known and validated air-firing semi-empirical expressions. The model has been divided into three sub-models interacting with each other: fluid-dynamics, combustion and energy balance of plant. For attributing reliability to the developed model, the scarce public experimental measurements of real air-firing boilers have been compared with the model results. Additionally, three studies regarding the modeling of large oxy-fuel CFB boilers have also been used for comparing the model predictions. In spite of having insufficient information about the published models details, the model developed in this work fairly fits the predictions in the literature. This has allowed making the sensibility analysis, trying to draw the main consequences of oxy-fuel deployment in CFB boilers. For retrofitting purposes, i.e. with no changes on an air-firing boiler configuration, the adequate O2 proportion of oxygen at entrance should be around 30%. Higher O2 concentrations lead to smaller cross sectional areas of the boiler. For a given fuel power required in a boiler, feeding 45% O2 in the comburent, would reduce the cross sectional area down to 54% of the original one. This involves a reduction of heat transfer surface along the boiler walls of 23% approximately. The immediate consequence is the need of resorting to external heat transfer surfaces, i.e., external heat exchangers (EHE). This device would need to remove almost 50% of the total heat of combustion in the case of feeding comburent with 60% O2 content. The importance of the EHE resides not only in compensating the reduction of heat transfer surface in the riser, but in managing higher amount of elutriated solids. The simulations have shown that higher solids densities in the boiler will enhance heat transfer coefficients to the riser walls. For certain boiler geometry, if increasing boiler load, higher recycled solids rate will be required. Feeding 60% of O2 at inlet, fuel input can be increased from 600 to 800 MW if elutriated solids increase from 25 to 40 kg/m2s. This refers us again to the higher solids crossing the EHE. An increase of 10% of heat removal will be required in this device for said changing load. Applying EHEs to conventional boilers was not essential during air-firing operation. But for oxy-fuel combustion it was here demonstrated to be crucial for accomplishing the boiler energy balance. However, several operational and design uncertainties will need to be solved, before deploying first demonstration oxy-CFB boiler. The design of the future EHE will imply two relevant distinguishing features of oxy-firing operation: the influence of gas composition on the determination of the heat transfer coefficients and the greater amount of elutriated solids, cooled down in the EHE. The CIRCE bubbling fluidized bed pilot plant presents the adequate bubbling working regime to obtain results of heat transfer coefficient for a wide range of oxy-fuel conditions and extracting further conclusions on possible effects of gas composition on heat transfer coefficients. The range of O2 concentration at inlet reached values as high as 60%. Such a high concentration was scarcely achieved in pilot plants due, in most cases, to the limiting bed cooling capacity. Measurements of heat transfer coefficients were taken when cooling was needed to control the combustion temperature. Water could circulate through one or more of the four cooling jackets, depending on the cooling requirements. Heat transfer coefficients were indirectly measured by energy balance with the water mass flow and temperatures. There are no previous results on heat transfer measurements under oxy-fuel combustion, up to date. The pilot plant is characterized by two important performance parameters: the fluidizing velocity and the bed temperature. These two parameters are common for all the fluidized bed plants working on combustion. Particularly for characterizing oxy-fuel combustion, the composition of the oxidant gas is the other key parameter in the plant operation. These three factors have been analyzed and their influence on heat transfer was examined. The three of them are, however, interrelated. O2 concentration and bed temperature varied the gas density and thus, the fluidizing velocity. At the same time, the fluidizing velocity will affect the heat transfer coefficients and consequently, bed temperature would be influenced. For accounting for this kind of dependences, non-dimensional numbers have been used for comparison. It was detected no dominant effect of non-dimensional numbers on the heat transfer. This is mainly offset by the different fluidization velocities in AF and OF operation. In the former, uf was kept over 1 m/s, whereas OF required lower velocities, around 0.9 m/s. It was then determined the adequate semi-empirical correlations for the effective thermal conductivity and the residence time of particles at the heat transfer surface. Hence, a semi-empirical mechanistic approach is recommended for a good agreement with the experimental heat transfer coefficients obtained during oxy-fuel operation. It was demonstrated the relevance of the gaseous film resistance in the oxy-fuel tests, and a new empirical coefficient was deduced for both modes. As examined in Chapter 3, section 3.5, the recommended expressions to predict heat transfer coefficients during oxy-fuel combustion modified the thermal film resistance, fitting the empirical parameter M with experimental data. Where: M=6.51 for oxy-firing and M=11.33 for air-firing The larger amount of solids arriving at the EHE will influence the values and distribution of the average and local heat transfer coefficients, respectively. A review of the difficulties associated with the estimation of heat transfer to the tubes of a heat exchanger has been examined. By the use of a scaled-down EHE, it was possible to experimentally confirm the influence of heat transfer coefficients when horizontal movement of solids took place. The increase of solids rate stressed the inequalities of the local heat transfer coefficient, whereas the longer residence time taken by particles to travel through the EHE allows higher average heat transfer coefficient. The contribution of this parameter to the average heat transfer coefficient was correlated by means of a new expression, as developed in Chapter 4, section 4.4. This expression allows modifying the heat transfer coefficient previously deduced for stationary conditions, and therefore, accounting for the enhancement of heat transfer when recirculation of solids takes place. A real design of an EHE was then simulated and integrated in the existing CFB model previously developed. This is the first time that such a model is developed to predict the heat transfer area required in oxy-fuel operation. The EHE sub-model must fulfill the energy balance requirements previously set for the CFB model. The temperature, at which solids must be recycled back into the boiler, in order to keep the desired boiler temperature, is accomplished with this sub-model. The expressions for the heat transfer coefficient and the enhancement due to recycled mass flow of solids were included in the EHE sub-model. Hence, it was possible to determine the increase on the heat transfer surface, for different O2 concentration in the oxidant stream, and two ranges of boiler temperature required. It was then recognized that, in spite of doubling the heat transfer surface requirements, when O2 concentration increased 10%, the heat transfer surface increases less than expected if solids flow influence were not included in the heat transfer evaluation. This thesis demonstrates that heat transfer surface design, arrangement and allocation, will differ in future oxy-fuel CFB boilers. Particularly, the heat transfer in the EHE will need address the influence of fluidizing gas composition and recycled solids, for an adequate and efficient heat exchanger configuration.