A investigation of the literature was made to determine the scope of previous investigations into the possibilities of obtaining drying oil characteristics in tall oil. Refined tall oil was esterified in an inert atmosphere with pentaerythritol at temperatures of 250, 300, and 320°C. Carbon dioxide was used to maintain an inert atmosphere. The esterification was carried out in a glass, Wolfe, three-necked flash heated in a constant temperature oil bath using S.A.E. 250 gear oil as a hosting medium. Agitation was supplied by means of an air driven agitator. The runs were six hours in length with the exception of one four hour run made at 32C°C. The procedure was repeated using activated alumina as a catalyst. the same procedure was repeated using dipentaerythritol in place of pentaerythritol. Crude tell oil was esterified with both pentaerythritol and dipentaerythritol at 250°G for six hours in an inert atmosphere of carbon dioxide. / Master of Science

A dissertation submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science in Engineering, 2016 / The sulfur removal methods from petroleum products have become an important research topic. Sulfur poisons the catalysts found in vehicles engines and it is also a major air pollutant (Nehlsen, 2005). Recent sulfur specifications require refineries to produce ultra-clean products (Ma et al., 2002). This work aims at exploring a batch adsorptive desulfurization technique using a polymer-supported imidation agent (PI) as an adsorbent. The test was carried out at atmospheric pressure and on two commercial diesel fuels with sulfur contents of 5200 (Case 1) and 670 (Case 2) mg/kg which resembles the feed and outlet streams from the hydrodesulfurization (HDS) reactor respectively. The adsorbent was synthesized according to the procedure described by Shiraishi et al. (2003), BET, FTIR, SEM equipped with EDS and TGA were used for charaterization of the adsorbent. The PI was successfully synthesized and its surface area was 0.5333 m2/g which was incredibly lower than that of the PI synthesized by Fadhel (2010). Hence carbon nanotubes (CNTs) were added to the solution with the aim of improving the sulfur removal efficiency of PI. The obtained results indicated that PI with CNTs yield better results than PI without CNTs. In overall, the lowest sulfur content of 3462 mg/kg (33% removal efficiency) and 26 mg/kg (96% removal efficiency) for Case 1 and Case 2 respectively were obtained. Furthermore, the adsorbents were most effective at lower mixing rates (150 – 400 rpm), longer contact time (30 – 40 hours), practically high adsorbent amount (1 g) and moderate lower temperatures (25 – 50 ºC). The Freundlich adsorption isotherm model was the best fit to the experimental data in both Case 1 and Case 2. The kinetic model that best fitted well the experimental data is the pseudo-second-order model for both Case 1 and Case 2. The kinetic rate constant for Case 2 (4.079 x 10-3g/mg.min) was greater than that for Case 1 (6.75 x 10-5g/mg.min) thus indicating that fuel with low sulfur content has a higher sorption uptake than fuel with high sulfur content. Based on the results obtained in this study, it is suggested that the adsorption of sulfur at high sulfur content fuel is not capable to be used as a complimentary method to the HDS process. On the other hand, at low sulfur content fuel, there is an opportunity for combining this method with the traditional HDS method to achieve ultra-clean fuel. / GR2016

[pt] O conhecimento da composição química das águas conatas ou de formação é de grande relevância para a produção de petróleo. Atualmente, considera-se a salinidade da água amostrada no aqüífero como representativa de todo o reservatório. Há, entretanto, comprovadas discordâncias e poucas informações sobre a composição química dessas águas. Isso pode levar à estimativa incorreta da reserva de um determinado reservatório, nas previsões de incrustação dos poços e, portanto, nas ações preventivas de elevado custo nos projetos de engenharia. Neste trabalho desenvolveu-se uma metodologia para extração e análise de água conata, a partir de amostras de petróleo. A metodologia se baseia na adição de quantidade conhecida de água sobre um determinado volume de óleo e recuperação dessa água, após separação, para posterior análise. Foram otimizados os parâmetros relevantes para a separação óleo- água, permitindo a recuperação quantitativa de águas associadas, mesmo em concentrações muito baixas (< 0,5%). Estabelecidas as condições, a água extraída foi submetida à análise multielementar por ICP OES e ICP-MS. Os elementos menores e traços determinados apresentaram a seguinte faixa de concentração (em mg L(-1)): Na: 1.880 - 63.794; K- 41,9 - 3.710; Li: 0,48 - 35; Mg: 15,2 - 991; Ca: 143 - 5.867; Sr: 1,12 - 676; Ba: < 0,005 - 243; B: < 0,09 - 149; Zn: < 0,042 - 76,2; Mn: < 0,006 - 4,02; Co: < 0,085 - 465. Os ânions foram determinados por cromatografia iônica, sendo cloreto (4632 - 102.757 mg L(-1)), brometo (2,7- 261 mg L(-1)) e sulfato (15,0 - 3.292 mg (L-1)) os mais abundantes. Os ânions provenientes de ácidos orgânicos, formiato (116 - 7.324 mg L(-1)), acetato (144 - 24.630 mg L(-1)) e propionato (34,3 - 4.891 mg L(-1)), foram investigados também pela técnica de cromatografia iônica em algumas amostras. Na grande maioria das águas extraídas, as concentrações desses cátions e ânions são ordens de grandeza mais altas do que as reportados para água do mar. O excelente balanço iônico calculado para a maioria das águas extraídas (maior ou igual 5%) reforçou a confiabilidade na qualidade analítica dos dados aqui apresentados. Foi otimizada também uma metodologia para pré-concentração de elementos do grupo das terras raras (ETR) através da coprecipitação com hidróxido férrico, a qual permitiu a determinação quantitativa de La, Ce, Nd, Sm, Eu, Gd e Dy nessas águas. O perfil normalizado de concentração desses elementos pelo padrão NASC (North American Shale Composite) mostrou anomalias muito diferentes de água do mar, indicando a potencialidade dos ETR para caracterização geoquímica de águas conatas e dos ambientes geológicos associados. Aperfeiçoou-se também uma metodologia para determinação de elementos menores e traços em petróleo por ICP OES e/ou ICP-MS com introdução de amostra em forma de microemulsão. Ela foi aplicada na análise de óleos e não permitiu apenas a sua caracterização elementar, mas também, um balanço de massa entre a amostra original de óleo e do seu extrato aquoso, o qual se mostrou satisfatório para a maioria das amostras analisadas. Uma vez introduzidas e validadas, as metodologias foram usadas para caracterização de amostras de óleo (e águas associadas), sendo a maior parte proveniente da Bacia de Campos (RJ). Com auxílio de gráficos de correlação, razões elementares e da análise de componentes principais (PCA) foi possível identificar grupos de amostras (águas e óleos) com características comuns (mesmo campo; maior ou menor intrusão de água do mar), mostrando a utilidade destas ferramentas neste e para futuros estudos. / [en] Knowledge on the chemical composition of connate or formation waters is of great relevance in oil production. Presently, the salinity of water sampled from the aquifer is considered generally as being representative of the whole reservoir. This simplification and the few and contradictory results available on the composition of connate water may result in incorrect estimation of the oil production potential of a reservoir, and in the evaluation and use of the cost- intensive actions to prevent scale formation. In this work, a methodology for the extraction and analysis of connate waters from petroleum samples was developed. The extraction method is based on the addition of known quantities of water to the oil sample and its recovery, after phase separation, for subsequent analysis. Relevant experimental parameters were optimized allowing the quantitative separation of water, even when present in very low concentrations (< 0.5%) Once established, the methodology was applied to extracted water samples, which were analyzed by ICP OES and ICP-MS. Minor and trace element content of these waters showed the following range of concentrations (in mg L-1): Na: 1,880 - 63,794; K: 41.9 - 3,710; Li: 0.48 - 35; Mg: 15.2 - 991; Ca: 143 - 5,867; Sr: 1.12 - 676; Ba: < 0.005 - 243; B: < 0.09 - 149; Zn: < 0.042 - 76.2; Mn: < 0.006 - 4.02; Co: < 0.085 - 465. Concentrations of anions (same unit as before) were determined by ion chromatography, being chloride (4,632 - 102,757), bromide (2.7 - 261) and sulfate (15.0 - 3,292) the most abundant ones. Anions of organic acids (also in mg L-1), such as formiate (116 - 7,324), acetate (144 - 24,630) and propionate (34.3 - 4,891) were also determined by ion chromatography in some samples. In most of the extracted water samples, the concentrations of cations and anions were orders of magnitude higher than reported for ocean water. The excellent ionic balance (equal greater or 5%) calculated for most of the extracted waters reinforced our confidence on the reliability of the here produced data. Additionally, a procedure for the preconcentration of rare earth elements (REE) by coprecipitation with iron (III) hydroxide was developed, and which permitted the quantitative determination of La, Ce, Nd, Sm, Eu, Gd and Dy in these waters. Their normalized concentration profile against the NASC (North American Shale Composite) standard showed patterns very different from ocean water, indicating the potential of REE for the characterization connate waters and associated geological environments. A method for the determination of minor and trace elements in oil samples by ICP OES and ICP-MS, after sample preparation by emulsification, was also introduced and applied to the analyses oil samples, allowing the calculation of elemental mass balances between the original oil sample and its aqueous extract, which were satisfactory for most of the samples analyzed. Once established and validated, the methods and procedures were applied to oil samples, mainly for the Campos Basin (RJ). By means of correlation plots, elemental ratios and the Principal Component Analysis (PCA), different groups of waters and oils with common characteristics (same oil field; minor or major intrusion of water) could be identified showing the utility of these tools for this and further studies.

[pt] A determinação de diferentes níveis de silício e alumínio em petróleo é importante porque pequenas quantidades destes elementos podem produzir efeitos adversos nas refinarias devido à corrosão de equipamento ou afetar a qualidade dos produtos refinados. A espectrometria de fluorescência de raios-X por dispersão de energia (EDXRF) foi utilizada para desenvolver um método para a determinação de Si e Al em petróleo, onde estes elementos se encontram principalmente sob a forma de partículas sólidas de aluminossilicatos dispersas na amostra. Uma vez que os analitos não puderam ser determinados sem interferência diretamente no óleo, a fusão das amostras de petróleo foi realizada utilizando tetraborato de lítio. Em seguida, as amostras fundidas foram colocadas no centro de papéis de filtro de 10 mm de diâmetro, que foi colocado entre duas folhas de filme de polipropileno, para então ser fixado à cubeta para as medições no instrumento. A quantificação foi feita por meio de curvas analíticas no intervalo de concentração de 0 a 40 mg kg(-1) (para ambos os elementos) no material fundido final. O método desenvolvido, não sofreu interferência matriz uma vez que as amostras foram completamente decompostas e, posteriormente secas no substrato. Os resultados para os elementos foram estatisticamente comparáveis com os obtidos por espectrometria de absorção atômica com chama (FAAS). Além disso, os resultados concordaram com os obtidos nas amostras de programas de proficiência da ASTM. Amostras reais de petróleo, fornecidas pela Petrobras, também foram analisadas. As recuperações alcançadas se encontram entre 71 e 100 por cento para as diferentes amostras, o qual pode ser considerado satisfatório devido à dificuldade em se determinar esses elementos. Os limites de quantificação (10 vezes o desvio-padrão, n=10) encontrados para o Si e Al no petróleo foram de 0,7 e 1,1 mg kg(-1), respectivamente, mostrando o potencial do método proposto para a determinação de amostras com valores mais elevados destes elementos. / [en] It is important to determine silicon and aluminum at different levels in crude oils because of trace amounts of these elements may produce adverse effects in oil refining either by causing corrosion or by contaminating and affecting the quality of the refined products. Energy dispersive X-ray fluorescence spectrometry (EDXRF) was used to enable a reliable method for determination of Si and Al in crude oil, where these elements are found mainly in the form of solid aluminosilicate particles dispersed in the sample. Since the analytes could not be determined directly in the oil without interference, the fusing the crude oil samples was made using lithium tetraborate. Then, the fused samples were placed in the center of 10 mm diameter filter paper that were sandwiched between two polypropylene film foils and attached to the instrument cell for measurements. Quantification was made by using analytical curves in the concentration range from 0 to 40 mg kg (-1) (for both elements) in the final fused material. The method developed did not suffer from matrix effect once the sample matrix was completely decomposed and the sample solution dried in the substrate before measurements. The results for the elements were statistically comparable to the ones obtained by flame atomic absorption spectrometry (FAAS). In addition, for proficiency test samples, the results were in accordance to the ones reported by ASTM proficiency programs. Real samples of oil provided by Petrobras Company also were analyzed. Recoveries were achieved between 71 and 100 percent for different samples, which can be considerable satisfactory because of the difficulty in determine those elements. The limits of quantification (10 times the standard deviation. N = 10) found for silicon and aluminum in the oil were of 0.7 and 1.1 mg kg (-1), respectively, showing the potential of the proposed method to screen for samples with higher amounts of these elements.