The enrichment of alkali in the blast furnace has been proven to be a catalyst of coke gasification and is thus a key parameter in the degradation of coke. Alkali also directly destroys the carbon structure, increases the risk of scaffold formation, increases the load and attacks the refractory. It is thus important to decrease the recirculation of alkali in the blast furnace and the gasification of coke to ensure sufficient strength of the coke. The aim of the present study was to examine possible ways of alkali control in the blast furnace. This was done by investigating if coke with a modified ash composition contributed to a higher capacity of binding alkali in stable phases, which can be drained via the slag. This would decrease the recirculation of alkali in the blast furnace and prohibit coke degradation. Two campaigns were studied to determine the distribution of alkali in the shaft when the charging differed, this to improve the understanding of alkali control in the blast furnace with respect to the charging practice. Three different test cokes were produced in pilot scale with a mineral addition of kaolin, silica or bauxite. The test cokes were together with base coke used as a reference, charged in baskets to LKAB’s Experimental Blast Furnace (EBF) at the end of a campaign. When the campaign was finished the EBF was quenched with nitrogen and the charged baskets were excavated. The influence of alkali on coke with a modified ash composition was examined with XRF, XRD, SEM-EDS and TGA. This was done in order to confirm any difference between the test cokes and the base coke in terms of chemical composition, phases in the coke ash, degree of graphitisation and reactivity. The results showed that the base coke in most cases had collected more alkali compared to the test coke with a mineral addition of kaolin and silica. For the test coke with addition of bauxite the alkali content was higher in three out of four samples compared with the corresponding base coke. Unreacted grains with bauxite were detected, which indicates that bauxite was completely or partly inactive in the capturing of alkali. All aluminosilicates detected in the coke samples contained alkali, which indicates that aluminosilicates contributes in the capturing of alkali in the EBF. The main mineral phases containing potassium in the coke were kalsilite, leucite and other aluminosilicates with varying alkali content. The carbon conversion and thus the reactivity increased with the alkali content in both the test coke and the base coke. The reactivity of the test coke was thus not decreased due to the mineral addition. No indications of an increased capacity of capturing alkali in stable phases could be seen in the test cokes, this could be due to the low amount of minerals added. The uptake of alkali in the different coke types was dependent of the horizontal and vertical position in the EBF, and thus the conditions the baskets had been exposed to and the distribution of alkali within the EBF. It was concluded that the charging had an impact on the alkali distribution in the EBF. During campaign 31 the alkali content was more evenly distributed over the horizontal section in the upper part of the furnace, in the lower shaft the alkali content increased towards the centre. During campaign 32 the alkali content was increasing towards the walls in shaft of the EBF. The content of alkali in the lower shaft was higher during campaign 32 compared with campaign 31.

The maximum yield taken in an oil refinery can not exceed 70% without including Delayed Coker Unit (DCU) as part of unit operations in the refinery. This implies naturally a big attraction on investing of such a unit for refiners. However, during the past decades, there were few refiners included Coker unit in the refinery, due to the fact of its large capital investment with a high marginal profitability. On the other hand the technologies developed to operate a coker unit, involve a series of process steps that require highly trained and well experienced operators with a state of art of design to overcome all the challenges with this unit operation. Safety, as a prime factor of design and operation requires much attention in the design of this unit. Among different safety consideration in the design and operation of Coker Unit, this project thesis focuses on the dynamic behaviour of Coke Drums PSV (Process Safety Valve) relief and its interaction with Blowdown section of the unit that leads also to the PSV relief of Blowdown section with change of temperature versus time during the first 15 minutes that is considered as the time required for operators intervention. The main findings in this thesis are about the complications in the design aspects of delayed coker unit as well as the importance and role of safety of operation of this unit. It also gave me an insight of cascade relief during the upset condition in an online coke drum and the importance of a reliable piping system to handle the hydraulics as well high temperature conditions.

A unidade de coqueamento retardado é um processo térmico de conversão, utilizado pelas refinarias, para converter cargas residuais em produtos de baixo peso molecular e com alto valor agregado (gases, nafta e gasóleo) e coque verde de petróleo. Um pequeno aumento no rendimento líquido da unidade de coqueamento retardado proporciona benefícios económicos consideráveis, especialmente no destilado líquido. A concorrência no mercado, as restrições sobre as especificações do produto e gargalos operacionais exigem um melhor planejamento da produção. Portanto, o desenvolvimento de novas estratégias e modelos matemáticos, focados em melhores condições de operação do processo industrial e formulações de produtos, é essencial para alcançar melhores rendimentos e um acompanhamento mais preciso da qualidade do produto. Este trabalho tem como objetivo o desenvolvimento de modelo matemático do conjunto forno-reator do processo de coqueamento, a partir de informações obtidas em uma planta industrial. O modelo proposto é baseado na caracterização da carga e dos produtos em pseudocomponentes, modelos cinéticos de grupos e condições de equilíbrio liquido-vapor. Além disso, são discutidos os principais desafios para o desenvolver o modelo matemático do forno e do reator, bem como a caracterização rigorosa do resíduo de vácuo e dos produtos para determinar os parâmetros que afetam a morfologia do coque e a zona de reação no interior do reator de coque. / The delayed coke unit is a thermal conversion process, used by the crude oil refineries, to convert residual feedstocks into products of low molecular weight and high aggregated value (gases, naphtha and gasoil) and green coke. A small increase in the net yield in the delayed coke unit results in considerable economic benefits, particularly in the liquid distillates. The market competition, the restrictions on the product specifications and the operational bottlenecks require a better production planning. Therefore, the development of new strategies and mathematical models, focused in better industrial process operating conditions and product formulations, is essential to achieve better yields and a more precise product quality monitoring. The objective of this work is the development of a furnace-reactor mathematical model of the delayed coke process based on industrial plant information. The proposed model is based on the feed and product characterization as pseudo components, group kinetical models and liquid-vapor equilibrium. Furthermore, the main challenges to develop the furnace and reactor mathematical model are discussed, as well as the vacuum residual and the coke unit products rigorous characterization to determine the parameters that impact the coke morphology and the reaction zone inside the coke reactor.

A unidade de coqueamento retardado é um processo térmico de conversão, utilizado pelas refinarias, para converter cargas residuais em produtos de baixo peso molecular e com alto valor agregado (gases, nafta e gasóleo) e coque verde de petróleo. Um pequeno aumento no rendimento líquido da unidade de coqueamento retardado proporciona benefícios económicos consideráveis, especialmente no destilado líquido. A concorrência no mercado, as restrições sobre as especificações do produto e gargalos operacionais exigem um melhor planejamento da produção. Portanto, o desenvolvimento de novas estratégias e modelos matemáticos, focados em melhores condições de operação do processo industrial e formulações de produtos, é essencial para alcançar melhores rendimentos e um acompanhamento mais preciso da qualidade do produto. Este trabalho tem como objetivo o desenvolvimento de modelo matemático do conjunto forno-reator do processo de coqueamento, a partir de informações obtidas em uma planta industrial. O modelo proposto é baseado na caracterização da carga e dos produtos em pseudocomponentes, modelos cinéticos de grupos e condições de equilíbrio liquido-vapor. Além disso, são discutidos os principais desafios para o desenvolver o modelo matemático do forno e do reator, bem como a caracterização rigorosa do resíduo de vácuo e dos produtos para determinar os parâmetros que afetam a morfologia do coque e a zona de reação no interior do reator de coque. / The delayed coke unit is a thermal conversion process, used by the crude oil refineries, to convert residual feedstocks into products of low molecular weight and high aggregated value (gases, naphtha and gasoil) and green coke. A small increase in the net yield in the delayed coke unit results in considerable economic benefits, particularly in the liquid distillates. The market competition, the restrictions on the product specifications and the operational bottlenecks require a better production planning. Therefore, the development of new strategies and mathematical models, focused in better industrial process operating conditions and product formulations, is essential to achieve better yields and a more precise product quality monitoring. The objective of this work is the development of a furnace-reactor mathematical model of the delayed coke process based on industrial plant information. The proposed model is based on the feed and product characterization as pseudo components, group kinetical models and liquid-vapor equilibrium. Furthermore, the main challenges to develop the furnace and reactor mathematical model are discussed, as well as the vacuum residual and the coke unit products rigorous characterization to determine the parameters that impact the coke morphology and the reaction zone inside the coke reactor.

This serves to introduce the present research and is chiefly historical. It deals briefly with methods of assessing the coking ability of coals and also with early modern, and contemporary investigations upon coke formation. The author's earlier work on the distillation of oil from coals is described fully since it forms both the starting point and the basis of the present research. This work showed the power of retention of oil which is capable of 'wetting' the surface of the coal, up to temperatures of 420 degrees C., was an essential characteristic of coking coals. Part II deals with attempts to add more oil mechanically to improve the coking performance of various coals.

A laboratory scale delayed coking process was used to produce petrol precursors, diesel precursors, methane rich gas, green and calcined coke from five previously untested South African heavy petroleum residues. The ash content of the heavy petroleum residues was found to be detrimental to the microstructure of the green coke and Coefficient of Thermal Expansion (CTE) of the calcined coke. The sulphur content of the calcined cokes produced was found to be in-line with typical global anode grade cokes. De-ashing of the feedstock would be necessary to produce anode grade fillers for the aluminium industry. The local production of anode grade coke would serve to reduce imports and supply the aluminium smelters on the east coast of South Africa. The heavy petroleum residues (also known as heavy fuel oil) are currently used as bunker fuel in the shipping industry and are responsible for substantial air pollution. Delayed coking of these residues could extend the production of petrol and diesel per barrel of imported crude oil and reduce the effect on South Africa’s balance of payments deficit and impact the environment in a beneficial manner with respect to carbon dioxide and sulphur emissions. The research also evaluated the replacement of heavy fuel oil with marine diesel produced by delayed coking of the former. Marine diesel was found to emit less sulphur oxides and have a higher energy density per unit of carbon dioxide emitted. While seawater scrubbing of the heavy fuel oil would be more cost effective in reducing the sulphur oxide emissions, it would not contribute to carbon dioxide reductions. The research created a hypothetical scenario to determine the required value of Clean Development Mechanism credits for a marine diesel replacement, were shipping to be incorporated under the Kyoto Protocol in future

The need for more efficient blast furnaces is even greater now that there are stricter environmental regulations on greenhouse gas (GHG) emissions. Coke within the blast furnace not only supports the furnace bed and allows gas flow, it also carburises liquid iron. The carburisation of iron is one of the most important reactions and must be better understood if the ironmaking process in the blast furnace is to be made more sustainable. By understanding what coke properties influence the rate at which coke dissolves in iron we can predict a coke?s performance and use it to determine its quality. As carbon dissolution rates have only been determined for a few cokes, a systematic and comprehensive study was conducted on the dissolution of carbon from nine Australian cokes into liquid iron. The kinetics of carbon dissolution from Cokes A to I was measured and a range of experimental techniques were used to elucidate the dominant rate influencing factors. The role of coke structure, coke inorganic matter composition and yield and temperature were investigated. Furthermore, the influence of interfacial products and dynamic wettability studies were also conducted. The carburiser cover method was used to measure carbon pick-up as a function of time over the temperature range of 1450-1550 ??C. Fundamental data on the apparent carbon dissolution rate constant (K) in molten iron at 1550 ??C for Cokes A to I were obtained and ranged from K (x 103 s-1) = 0.47 to K (x 103 s-1) = 14.7. The wide variation in K showed that not all cokes dissolve at similar rates. In fact one of the nine cokes in this investigation dissolved at a rate comparable to graphite dissolution rates. The apparent carbon dissolution activation energy, Ea, for two of the nine cokes plus synthetic graphite (SG) was also determined. The Ea obtained for SG (Ea = 54 kJ / mol) was in agreement with literature values and was consistent with a diffusion controlled mechanism. The observed Ea values for Cokes D and F (313 kJ / mol and 479 kJ / mol respectively) are an order of magnitude larger than the Ea obtained for SG. The difference in Ea between cokes and SG does not appear to be solely due to differences in the structure of the carbon source. The difference in Ea between the cokes was attributed to differences in their inorganic matter composition. The interfacial contact area is a function of inorganic matter yield and composition, which in turn is a function of temperature. Therefore, as temperature decreases the slag / ash layer produced at the carbon / iron interface can increase in area and viscosity and thus hinder carbon dissolution and transfer, and increasing the apparent activation energy for carbon dissolution. Thus, the differences in viscosity and melting temperature of the interfacial product play a key role. Wettability experiments were carried out using the sessile drop technique. The wettability of Cokes D, F and G with liquid iron at 1550 ??C was measured as a function of time. All three coke samples showed non-wetting behaviour with contact angles ranging between 123-129?? in the initial stages and between 109-114?? after two hours of contact. The differences in the wettability of the three coke samples could not explain the large differences in dissolution rates observed between these cokes. Thus, the wettability of these coke samples was not considered a dominant factor in influencing the rate of carbon dissolution. The sessile drop technique was also used to study the interfacial products formed at the coke / iron interface. The interfacial products formed on the underside of the iron droplet after contact with Cokes F and G were initially different in regards to the morphology and chemical composition. The interfacial product formed with Coke G had a network or mesh like structure that seemed to wet the iron droplet much better than the interfacial product formed with Coke F. In contrast, Fe globules and discrete interfacial products were observed in Coke F. It was suggested that this was due to differences in inorganic matter content, especially in calcium (Ca) and sulfur (S) content in the coke. Formation of interfacial products containing sulfides, such as calcium sulfide (CaS) and manganese sulfide (MnS), were observed on the iron side of the interface of both Cokes F and G. As a result, the interfacial products can act as a physical barrier blocking iron and coke contact, thus reducing the contact area for carbon dissolution and decreasing the rate of carbon dissolution. The presence of MnS may act to lower the liquidus temperature of the interfacial product, which in turn can affect the overall viscosity of the interfacial layer. Thus, the deposition of reaction products at the interfacial region can have a significant effect on carbon dissolution rates. The mineral pyrrhotite was also identified as a significant factor in influencing the rate of carbon dissolution. Electron dispersive X-ray analyses of Coke F identified iron to be in close association with sulfur. These Fe / S species have atomic ratio similar to pyrrhotite (Fe1-xS) or troilite (FeS). Pyrrhotite in coke can decompose to release gaseous sulfur and metallic iron, which can be carburised by carbon in the surrounding area to form Fe-C particles. Thus, carburisation of liquid iron can occur via Fe-C particles. There was little difference in structure between the nine coke samples and therefore the high dissolution rates of Coke F, cannot be explained on the basis of crystallite size or anisotropic carbon content. Inorganic matter yield and composition were identified as the dominant rate influencing factors on carbon dissolution. More specifically: - High content of iron phases, such as iron oxides and pyrrhotite, can lead to an increase in carbon dissolution rates. This maybe due to increased amounts of Fe-C particles that are formed upon the reduction of magnetite and decomposition of pyrrhotite, and carried through the slag layer to carburise the bulk liquid iron. - High aluminium oxide content can lead to a decrease in carbon dissolution rates. This maybe due to higher ash fusion / melting temperature or decrease in wettability, both of which lead to a decrease in carbon / iron contact area. - Formation of interfacial products, such as CaS and MnS, can lead to a decrease in carbon dissolution rates. Such products can act as a physical barrier blocking iron and coke contact, thus reducing the contact area for carbon dissolution. However, the presence of MnS may act to lower the liquidus temperature of the interfacial product. - An increase in temperature increases the rate of carbon dissolution. This dependence is predominantly due to the composition of the inorganic matter present in cokes, which influences the viscosity and melting point of the interfacial product formed and hence contact area between coke and iron.

Coke drums are vertical pressure vessels used in the delayed coking process in petroleum refineries. Significant temperature variation during the delayed coking process causes the useful life of coke drums to be shortened. In order to better understand the failure mechanisms, a experimental study of elastic/plastic mechanical properties and deformation behaviors of typical coke drum materials was performed. A new biaxial thermal-mechanical material testing system has been successfully developed. Basic characterization of mechanical properties of coke drum materials is achieved through uniaxial monotonic and cyclic loading tests. In addition, strain-rate dependence and creep of coke drum materials were further experimentally investigated. Complex thermal-mechanical cyclic tests were conducted. The experimental findings help us to understand the damage mechanisms of coke drums such as bulging. In addition, experimental data serve as benchmark data to verify the predictions of the temperature dependent elastoplastic constitutive model.

Chars were obtained from individual fuels and blends with different blend ratios of coal, coke and biomass in Drop Tube Furnace at different temperatures. Based on TGA experimental data, it was shown that the effect of the blending ratio of biomass to other fuels on the reactivity of the co-pyrolyzed chars is more pronounced on the chars prepared at lower temperature, due to the presence of synergetic effects originating from the interaction of the two fuels. SEM images showed differences in shapes and particle size of char particles from biomass and coal/coke. These also show the agglomeration of coal and coke chars with biomass char particles at high temperatures. The agglomeration may be the reason for the non-additive behavior of the blends. BET analysis showed increase in the surface area with an increasing temperature for biomass and coal, but the trend for coke was inversely related to the temperature. / Chemical Engineering

Microwave heating has unique advantages compared to convection-radiation heating methods including fast heating rate and selective heating of objects. This thesis studied two applications of microwave heating in the environmental field: adsorbent regeneration and oil sands coke activation. The thermal behavior during microwave heating of select adsorbents when dry or saturated with selected adsorbates was studied to assess the potential for using microwave heating to regenerate adsorbents. Strong microwave-absorbing adsorbents depicted faster heating rate when dry. Weakly microwave-absorbing adsorbents depicted faster heating rate when saturated with polar adsorbates. Fast activation of oil sands coke using microwave heating and KOH was successfully completed. The iodine number of the activated delayed coke obtained after 10 minutes of microwave activation was 1130 mg/g. The short activation time and simplicity of the process demonstrate that microwave-activation is a promising approach to convert oil sands coke into activated carbon adsorbent with high adsorption capacity. / Environmental Engineering