Computer Aided Evaluation of Solvent Extraction for Light Hydrocarbon Using Carbon Dioxide

 

Evaluación Asistida por Computadora de la Extracción de Solventes para Hidrocarburos Livianos usando Dióxido de Carbono

 

Avaliação Auxiliada por Computador da Extração por Solvente para Hidrocarboneto Leve usando Dióxido de Carbono

 

Viktor Oswaldo Cárdenas Concha1*; Luz Sheyla Cárdenas Concha2; Leandro Lodi3; Juliana Otavia Bahú4; Diana Paola Figueredo-hernandez5; Willian Amaro-Marchioli1; Nelson Antonio Moreno Monsalve5; Jeffrey Leon-Pulido5*

 

1Chemical Engineering Department, Federal University of São Paulo (UNIFESP),

Diadema, São Paulo, Brazil, 09913-030

2Universidad Nacional de Trujillo, Avenida Juan Pablo II, Trujillo, Perú, 13011

3Institute of Science and Technology, Federal University of Alfenas (UNIFAL), Poços de Caldas 37715‐400

4School of Chemical Engineering, University of Campinas (UNICAMP),

Campinas, São Paulo, Brazil, 13083‐852

5Faculty of Engineering, Chemical Engineering Program, EAN University, St. 79 #11-45, El Nogal, Bogotá, Cundinamarca, Colombia, 110221

 

*viktor.cardenas@unifesp.br

 

Recibido: 11 de enero, 2023. Aprobado: 26 de enero, 2023. Versión final: 14 de abril, 2023.

 

 

Abstract

 

Different process of separation was used in the chemical industry, in particular, extraction is a process used to increase the quality of resins in oil removing impurities like organics solids and heavy metals. Supercritical carbon dioxide offers high selectivity at the end of the extraction process of light hydrocarbons from heavy oils mixture. A simulation technique in Aspen Plus ® software was used to develop the process and sensitivity analysis of the extraction configuration. The simulation of extraction process includes two output streams: the first one, a top stream (unpaved oil), and the second one a bottom stream (asphalt residue). A steady state methodology was implemented for process simulation. The sensitivity analysis was used to assess the influence of variables such as solvent flow rate, temperature and pressure. It was found a significant increase in the flow rate of unpaved oil when the solvent flow rate is increased. Optimal extraction values were selected depending on temperature and pressure effects over the process. An increase in temperature directly enhances the quality of API gravity. In certain occasions, an increase in pressure affects the light oils extraction because of product drag.

 

Keywords: Computer Aided-Design, extraction, hydrocarbon, carbon dioxide, simulation, process.

 

Resumen

 

En la industria química se utilizaron diferentes procesos de separación, en particular, la extracción es un proceso utilizado para aumentar la calidad de las resinas en el aceite eliminando impurezas como sólidos orgánicos y metales pesados. El dióxido de carbono supercrítico ofrece una alta selectividad al final del proceso de extracción de hidrocarburos ligeros a partir de una mezcla de aceites pesados. Se utilizó una técnica de simulación en el software Aspen Plus ® para desarrollar el proceso y el análisis de sensibilidad de la configuración de extracción. La simulación del proceso de extracción incluye dos flujos de salida: el primero, un flujo superior (petróleo sin pavimentar), y el segundo, un flujo inferior (residuo de asfalto). Se implementó una metodología de estado estacionario para la simulación de procesos. El análisis de sensibilidad se utilizó para evaluar la influencia de variables como el caudal de disolvente, la temperatura y la presión. Se encontró un aumento significativo en la tasa de flujo de petróleo cuando se incrementa la tasa de flujo de solvente. Los valores de extracción óptimos fueron seleccionados dependiendo de los efectos de temperatura y presión sobre el proceso. Un aumento en la temperatura mejora directamente la calidad de la gravedad API. En determinadas ocasiones, un aumento de presión afecta a la extracción de crudos ligeros por efecto del arrastre del producto.

 

Palabras clave: Diseño asistido por computadora, extracción, hidrocarburo, dióxido de carbono, simulación, proceso.

 

Resumo

 

Diferentes processos de separação são usados na indústria química, em particular, a extração é um processo usado para aumentar a qualidade das resinas no óleo removendo impurezas como sólidos orgânicos e metais pesados. O dióxido de carbono supercrítico oferece alta seletividade ao final do processo de extração de hidrocarbonetos leves da mistura de óleos pesados. Uma técnica de simulação no software Aspen Plus ® foi utilizada para desenvolver o processo e a análise de sensibilidade da configuração da extração. A simulação do processo de extração inclui dois fluxos de saída: o primeiro, um fluxo superior (óleo não pavimentado) e o segundo um fluxo inferior (resíduos asfálticos). Uma metodologia de estado estacionário foi implementada para simulação de processos. A análise de sensibilidade foi utilizada para avaliar a influência de variáveis como: vazão do solvente, temperatura e pressão. Foi constatado um aumento significativo na vazão de óleo quando a vazão de solvente é aumentada. Os valores ótimos de extração foram selecionados dependendo dos efeitos da temperatura e pressão sobre o processo. Um aumento na temperatura aumenta diretamente a qualidade da gravidade API. Em certas ocasiões, um aumento de pressão afeta a extração de óleos leves devido ao arraste do produto.

 

Palavras-chave: Desenho assistido por computador, extração, hidrocarboneto, dióxido de carbono, simulação, processo.

 

Cómo citar: Cardenas-Concha, V., Cardenas-Concha, L., Lodi, L., Otávio Bahú, J., Figueredo-hernandez, D. P., Amaro-Marchioli, W., Moreno Monsalve, N. A., & Leon-Pulido, J. (2023). Computer Aided Evaluation of Solvent Extraction for Light Hydrocarbon Using Carbon Dioxide. Fuentes, el reventón energético, 21(1), 39-48. https://doi.org/10.18273/revfue.v21n1-2023003

 

---

1. Introduction

 

The production of lubricants has been of great importance for the petroleum industry through solvent route, which implies the identification of new loads to feed the existing units, as well as, to study several solvents to be used in the extraction process. During the last two decades, the average quality of oil derived from refineries has decreased by approximately 2 points in the American Petroleum Institute gravity scale - API scale, along with an increase in inorganic solids, heavy metals and heteroatoms [1,2]. Petroleum consists of colloidal systems containing hydrocarbons which can be classified into alkanes, naphthalene, and aromatics, the energy consumption and exergy of process describes the thermodynamics relation and equilibrium for process simulation [18, 19, 20]. There are two types of particles: dispersed in the colloidal solution (asphaltenes) and petroleum resins; asphaltenes are hydrocarbons with high molar weight and contain mainly oxygen, sulfur, organic particles, inorganic salts, among other components. One of the most used processes to “purify” resins is the extraction [3].

 

Extraction is a process for reducing the asphaltenes and metals of desired component through the solubility of the appropriate solvent (e.g. propane, pentane, heptane or carbon dioxide) resulting in its final flocculation [4]. Wilson, Keith, & Haylett [5] developed a separation process based on phase equilibrium, which became the basis for the deasphalting process to propane still in use during refining of lubricating oils. Such process makes use of the change in the solvency power of a liquid near its critical point. Several systematic studies of phase equilibrium of carbon dioxide determined a characteristic point for each substance along the liquid-vapor equilibrium zone and with the gradual increase of pressure and temperature, that is, the critical point, located in Figure 1 [6, 7]. CO2 is a solvent capable of extracting light hydrocarbons from a mixture of heavy oils. It is used as solvent for supercritical extraction due to its high selectivity at the end of the process [8].

 

 

Figure 1. Pressure x Temperature diagram of a substance in its supercritical state. Source: Adapted from [6].

 

1.1. Thermodynamic modeling

To perform extraction simulation, it is necessary to define the best thermodynamic model that can explain appropriately solvent extraction. Based on current contributions, it was selected the thermodynamic model Predictive Soave-Redlich Kwong (PSRK), showed in Equation 1 [1, 7, 9]. The PSRK template is based on the Soave-Redlich-Kwong- Equation of State (SRK-EOS), with modifications suggested by Mathias & Coperman [10], and it makes use of the Universal Functional-group Activity Coefficients (UNIFAC) Original method to calculate the parameters of the mixture, with the first order modified by Huron & Vidal mixing rule (MHV1). However, PSRK uses a value of q1 different from that suggested by Michelsen in Equation 1, which leads to better results at high pressure. Where GE is the Gibbs free energy in excess and is related intrinsically with the pressure of the half together with the α property defining the mixed pattern according to the UNIFAC properties.

 

(1)

 

At high pressures, PSRK predicts liquid-vapor equilibrium with greater accuracy than Modified Huron-Vidal seconder-order model (MHV2); however, the quality of their predictions deteriorates when the mixing components differ in molecular sizes. PSRK also presents all the advantages of a state equation since it can be applied to supercritical component systems and allows the calculation of densities, enthalpies, and other properties, even in systems with polar components [11, 12]. Asymmetric systems are commonly found in processes with supercritical fluids, where the solutes are larger in size than the solvent, usually CO2 or paraffin of low molar mass, which is why the state equations are modified, often empirically, to achieve a better representation of the equilibrium of these complex systems.

 

2. Methods and materials

 

2.1. Simulation of the extraction process

 

In the assembly of a process to be simulated, it is necessary the characterization of raw material. For this study, petroleum with characteristics of heavy oil (low-grade API = 7.1) was used. Tables 1 and 2 show the physical-chemical characterization, the PEV curve or true boiling point of the oil, as well as the API gravity for each distillate percentage. These data were fed to the simulation software in order to generate the pseudo-components, which will be used in the simulation.

 

Table 1. Pseudo-Components of raw material

 

Pseudo-components	Temperature (°C)	°API	Specific gravity	MW	Critical Temperature (°C)	Critical Pressure (bar)
PC312C	311,96	14,48	0,97	223,21	528,11	20,52
PC323C	322,60	13,84	0,97	232,22	537,93	19,88
PC336C	336,48	13,03	0,98	244,32	550,65	19,10
PC350C	350,35	12,25	0,98	256,80	563,26	18,36
PC364C	364,21	11,50	0,99	269,66	575,74	17,66
PC378C	378,06	10,78	0,99	282,90	588,12	17,00
PC392C	391,91	10,08	1,00	296,51	600,40	16,38
PC406C	405,88	9,39	1,00	310,62	612,69	15,79
PC420C	419,71	8,75	1,01	324,99	624,75	15,23
PC441C	440,85	7,75	1,02	347,53	643,11	14,44
PC467C	467,42	6,65	1,02	377,26	665,70	13,52
PC497C	497,03	5,93	1,03	413,49	689,43	12,50
PC519C	518,99	5,19	1,04	440,66	707,39	11,86
PC552C	551,59	2,95	1,05	476,69	736,91	11,23
PC580C	580,36	1,16	1,07	508,98	762,49	10,70
PC608C	608,04	-0,50	1,08	539,75	786,90	10,23
PC637C	636,69	-2,24	1,09	570,36	812,25	9,80
PC683C	682,82	-5,47	1,12	612,21	854,55	9,27
PC720C	720,17	-7,40	1,14	647,37	886,66	8,79

 

Table 2. Main simulation equipment and their process variables

 

TAG	Equipments	Temperature (°C)	Pressure (bar)
MIXER	Mixer	-	-
TC1	Exchange	150	300
EXT	Extractor	150	300
TC2	Exchange	150	150
TC3	Exchange	150	150
SEP1	Separator 1	150	150
SEP2	Separator 2	150	150

 

The most assertive method for determining the fractions of charge is through experimental measurements. However, this method is not valid for all petroleum mixtures; for these cases statistical and simulated methods are employed [13, 21]. A pseudo-component represents each temperature range. It is understood that the normal boiling temperature of it corresponds to the average temperature of the interval [14]. Through the data of Table 1 generated, it is possible to cross the information and characterize the pseudo-components generated as a function of the molar mass and its API gravity. Figure 2 shows the correlation between both and again we note the inverse proportion of each characteristic.

 

 

Figure 2. Characteristics of Pseudo-Components

 

The virtual CO2 extraction plant was built in the Aspen Plus ® commercial simulator, where the main equipment and process streams were added and their operating parameters were identified in Figure 3 and Tables 3, 4 and 5.

 

 

Figure 3. Flowchart of the deasphalting process.

 

Table 3. Characterization of the main streams of the process

 

	Flow IN	Flow OUT
	PET	ODES	RASF
Temperature (°C)	135,00	150,00	150,00
Pressure (bar)	300,00	150,00	150,00
Flow rate (kg/hr)	200,00	9,10	182,20
Vapor Fraction	0,00	0,00	0,00
Average MM (g/mol)	318,10	179,50	350,50
°API	7,10	43,90	11,90

 

Next, Tables 4 and 5 present the main operating parameters used in the process equipment (DAO separation vessels).

 

Table 4. Characterization of streams: Input and output of the separator 1

 

	Flow IN	Flow OUT
SEP1	OIL+SLV2	LIGHTEND	OIL
Temperature (°C)	150,00	150,00	150,00
Pressure (bar)	150,00	150,00	150,00
Flow rate (kg/hr)	1062,10	1053,00	9,10
Enthalpy (Gcal/hr)	-2,20	-2,20	0,00
Vapor Fracion	1,00	1,00	0,00
Average MM (g/mol)	44,60	44,30	179,50

 

Table 5. Characterization of the streams: Input and output of the separator 2

 

	Flow IN	Flow OUT
SEP2	RAF2	RAF+SLV	ASPHALTE
Temperature (°C)	150,00	150,00	150,00
Pressure (bar)	150,00	150,00	150,00
Flow rate (kg/hr)	187,90	5,80	182,20
Enthalpy (Gcal/hr)	-0,10	0,00	-0,10
Vapor Fracion	0,20	1,00	0,00
Average MM (g/mol)	288,80	44,00	350,50

 

Table 6. Final yields of the process.

 

Deasphalted Oil Yield	4,38%
Asphalt Yield	93,60%
Solvent Recovery	99,27%

 

As can be seen, the yield is low, however, Liu et al.15 describes that CO2 extractions present low yields and high selectivity for light components (hydrocarbons). This phenomenon of low CO2 extraction yields is due to the solvent’s own characteristics. In normal conditions the CO2 is apolar, but as it happens changes in the temperature and pressure of the system the solvent happens to present/display polarity quadripolar, that is, that defines four moments dipole, bringing it to an opposite plateau of the compounds to be extracted. Another point to be evaluated is that the carbon dioxide, when worked at temperatures above 66 ° C, shows a decrease in density, leading to a sharp fall in the extraction, around 20% to 30%, thus suggesting the follow- solvents. These two factors, together, justify the low extraction of deasphalted oils using CO2 a critical solvent7. In this way, once the characterization of the residue was carried out and the process flowchart was assembled, the tests were carried out in the simulator.

 

3. Results and discussion

 

The following are results obtained in the simulation of the extraction process, as well as the sensitivity analysis performed for the process. Figure 6 clearly synthesizes the incoming stream distillation curves compared to the ODES (deasphalted oil) and RASF (asphaltic residue) generated at the end of the process.

 

 

Figure 4. PEV curves of the products and the sample fed

 

In Figure 4, it can be observed that the profiles of the PEV curves generated for the process output streams (extract and raffinate) are different between them. It can be noticed that the PET of the main operating parameters used in the process equipment (DAO) is in the inferior part as expected because it presents components of low molar mass, consequently, the lower boiling point. In contrast, it is observed that the stream corresponding to the asphalt (raffinate) residue is in the upper part, showing that the composition of the same is mostly composed of heavy compounds, thus raising the boiling temperature thereof, and as already expected, the oil fed is in the middle of the two curves mentioned above.

 

3.1. Sensitivity analysis

 

Figure 5 (A), (B) and (C) show the behavior of deasphalted oil (ODES) extraction flow in relation to CO2 flow, extraction temperature, and extraction pressure.

 

 

Figure 5. Effect of solvent flow rate, temperature and pressure on ODES flow

 

As expected, as the solvent flow increases the amount of extracted oils also increases ODES, which indicates the possibility of finding the ideal amount of CO2 in the extraction process. Figure 5 (B) and (C) show that the temperature and pressure parameters directly influence the extraction flow, and to obtain the best yield the ideal is to work at temperatures below (100 ° C) and at higher pressures (350 to 400bar), supercritical CO2 conditions [16, 17, 22, 23]. However, when evaluating the API gravity that is extracted under these conditions of temperature and pressure (100 ° C and 400bar), it is seen that there is some contradiction. Figures 6 (A) and (B) show that under these conditions API gravity are slightly lower than if working under reverse conditions. However, they would still be high-grade API oils, so there would be a high index extraction of light oils.

 

 

Figure 6. Effect of solvent flow rate, temperature and pressure on ODES API gravity

 

Figure 6 (A) shows the behavior of the API gravity as a function of the extraction temperature. As it can be seen there is a tendency for the API gravity to increase with increasing temperature, which means a selectivity of the top product as a function of temperature. However, Figure 6 (B) shows that the pressure exerts a decrease of the API gravity, which implies in the drag of components of high molar mass to the top stream (extract), not being favorable for the extraction process of light oils. Figure 6 (C) corroborates the results obtained in Figure 6 (A), showing that the molar mass of the top stream decreases with increasing temperature, which means extraction of light (high-grade API) components.

 

4. Conclusions

 

The process of extraction of light hydrocarbons using CO2 under supercritical conditions presented data consistent with those found in the literature. It can be observed that it had a low yield extraction as observed in Table 6 and high selectivity, which was contacted by the high API gravity of the extract stream. Two output streams (extract = DAO and raffinate = RASF) were obtained, which presented different PEV curves, showing the CO2 efficiency as a solvent. In addition, the sensitivity analysis confirmed that temperature is a parameter that influences the yield and quality of the DAO stream and that the pressure has no significant influence on the process. However, it should be emphasized that the characterization of the raw material is an important factor in the extraction process.

 

5. Acknowledgments

 

Authors would like to thank Federal University of São Paulo, Universidad Nacional de Trujillo, EAN University, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) for providing financial support to successfully conclude this work.

 

---

References

 

1. Abdulredha, M.M., Hussain, S.A., & Abdullah, L.C. (2020). Overview on petroleum emulsions, formation, influence and demulsification treatment techniques. Arabian Journal of Chemistry, 13 (1), 3403-3428. https://www.sciencedirect.com/science/article/pii/S1878535218302442

    

2. Ahn, S., Shin, S., Im, S.I., Lee, K.B., & Nho, N.S. (2016). Solvent recovery in solvent deasphalting process for economical vacuum residue upgrading. Korean Journal of Chemical Engineering, 33(1), 265–270. https://link.springer.com/article/10.1007/s11814-015-0146-3

    

3. Budisa N., & Schulze-Makuch, D. (2014). Supercritical Carbon Dioxide and Its Potential as a Life-Sustaining Solvent in a Planetary Environment. Life, 4(3), 331–340. https://www.mdpi.com/2075-1729/4/3/331

    

4. Cárdenas, V.O.C. (2010). Estudo experimental e modelagem matemática da separação de asfaltenos e óleos lubrificantes provenientes do resíduo de petróleo por desasfaltação supercrítica. Universidade Estadual de Campinas. https://repositorio.unicamp.br/acervo/detalhe/782233

    

5. De Souza R. A., De Oliveira A. A., Vieira Y. M., Zaparoli A. T., Wolf Maciel, M. R., Filho R. M., & Cárdenas-Concha, V. O. (2017). Sensibility analysis of propane deasphalting process variables. Petroleum Science and Technology, 35(1), 22–29. https://www.tandfonline.com/doi/abs/10.1080/10916466.2016.1225086?journalCode=lpet20

    

6. Eckermann B., & Vogelpohl, A. (1990). Deasphaltization and demetalling of heavy crude oils and distillation residues with CO2. Chemical Engineering & Technology, 13(1), 258–264. https://onlinelibrary.wiley.com/doi/abs/10.1002/ceat.270130135

    

7. Espinoza, S.N. (2001). Procesamiento supercrítico de productos naturales. Modelado, análisis y optimización. Universidad Nacional del Sur. http://repositoriodigital.uns.edu.ar/handle/123456789/2181

    

8. Freitas A., Santana C., Silva R., & Silva, G. (2007). Investigação das facilidades e métodos utilizados atualmente no processamento primário de petróleo em campos onshore e offshore. 4o PDPETRO, 1–8. http://www.portalabpg.org.br/PDPetro/4/resumos/4PDPETRO_2_3_0322-1.pdf

    

9. Grziwa. (1982). The chemistry and technology of petroleum: by James G. Speight Marcel Dekker, New York, Fuel Processing Technology, 5(3–4), 325–326. https://www.sciencedirect.com/science/article/pii/0378382082900261

    

10. Holderbaum T., & Gmehling J. (1991). PSRK: Eine Zustandsgleichung zur Vorhersage von Dampf/Flüssig‐Gleichgewichten bei mittleren und hohen Drücken. Chemie Ingenieur Technik, 63(1), 57–59.

     

11. Hosseinifar, P., Shahverdi, H. (2021). A predictive method for constructing the distillation curve of petroleum fluids using their physical bulk properties. Journal of Petroleum Science and Engineering, 200, 108403. https://doi.org/10.1016/j.petrol.2021.108403

     

12. Leal-Navarro, J., Mestre-Escudero, R., Puerta-Arana, A., León-Pulido, J., & González-Delgado, Á. D. (2019). Evaluating the Exergetic Performance of the Amine Treatment Unit in a Latin-American Refinery. ACS Omega, 4(26), 21993 –21997. https://pubs.acs.org/doi/10.1021/acsomega.9b03051

     

13. Li, J., Fischer, K., & Gmehling, J. (1998). Prediction of vapor–liquid equilibria for asymmetric systems at low and high pressures with the PSRK model. Fluid Phase Equilibria, 143(1–2), 71–82. https://www.sciencedirect.com/science/article/abs/pii/S0378381298002064

     

14. Liu, Z.M., Yang, G.Y., Lu, Y., Han, B.X., & Yan, H.K. (1999). Phase equilibria of the CO2–Jiangsu crude oil system and precipitation of heavy components induced by supercritical CO2. The Journal of Supercritical Fluids, 16 (1), 27-31. https://www.sciencedirect.com/science/article/abs/pii/S0896844699000182

     

15. Martins S., Azevedo M., Silva M., & Silva, V. (2015). Produção de petróleo e impactos ambientais: algumas considerações. HOLOS, 6, 54-76. https://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/2201/1212

     

16. Mathias, P.M., & Copeman, T.W. (1983). Extension of the Peng-Robinson equation of state to complex mixtures: Evaluation of the various forms of the local composition concept. Fluid Phase Equilibria, 13, 91-108. https://www.sciencedirect.com/science/article/abs/pii/0378381283800843

     

17. Peng, H., Yang, J., Peng, J., Han, H., Gou, X., Jia, Y., Li, Y., & Kadet, V. (2022). Influence of supercritical CO2 on the physical property of tight sandstone. Petroleum. https://www.sciencedirect.com/science/article/pii/S2405656122000645

     

18. Pulido, J.L., Martinez Arias, E.L., Wolf Maciel, M.R., & Maciel Filho, R. (2011). Heat Integrated Reactive Distillation Column (r-HIDiC): Implementing a New Technology Distillation. Chemical Engineering Transactions, 24, 1303-1308. https://www.cetjournal.it/index.php/cet/article/view/CET1124218

     

19. Riazi, M.R. (2005). Characterization and properties of petroleum fractions. ASTM International. https://www.astm.org/mnl50-eb.html

     

20. Rincón, F.D. (2009). Representação do petróleo através de componentes reais para a simulação de processos de refino. Universidade de São Paulo. http://www.teses.usp.br/teses/disponiveis/3/3137/tde-18082009-142836/es.php

     

21. Rudyk, S., Spirov, P., Al-Hajri, R., & Vakili-Nezhaad, G. (2017). Supercritical carbon dioxide extraction of oil sand enhanced by water and alcohols as Co-solvents. Journal of CO2 Utilization, 17, 90–98. https://www.sciencedirect.com/science/article/abs/pii/S2212982016304449

     

22. Wilson, R.E., Keith Jr, P.C., & Haylett R.E. (1936). Liquid propane: Use in Dewaxing, Deasphalting, and Refining Heavy Oils. Industrial and Engineering Chemistry, 28(9), 1065–1078. https://pubs.acs.org/doi/10.1021/ie50321a022

     

23. Yuan, Q., Zhang, Y., Wang, T., & Wang, J. (2023). Characterization of heavy metals in fly ash stabilized by carbonation with supercritical CO2 coupling mechanical force. Journal of CO2 Utilization, 67, 102308. https://doi.org/10.1016/j.jcou.2022.102308