Analysis of Influence Factors for FCC Slurry Separation by Mixed Solvent Extraction

Sun Yudong; Zhao Xiaoning; Liu Ziyuan; Feng Yi; Han Zhongxiang

(College of Chemical Engineering, China Uniνersity of Petroleum (East China), Qingdao 266580)

Abstract: Daqing FCC slurry was treated using mixed solvent containing N,N-dimethylformamide (extraction solvent) and n-paraffin (stripping-agent). The influence of solvent formulation (ratio of DMF/stripping-agent), temperature, time and solvent-oil ratio on yield and aromatics content of the extract oil were analyzed by a four-factor and three-level orthogonal experimental design. The polynomial regression models between the extract oil yield and aromatics content with the four factors were established. Response surface analysis showed that solvent formulation, temperature and solvent-oil ratio have significant effects on extraction result and there is an interaction between temperature and solvent-oil ratio on aromatics content, when the FCC slurry is extracted by mixed solvent. However, the extraction time has no significant effect. The optimal conditions for FCC slurry extraction covered: a solvent mixed ratio of 2.3, a temperature of 62.8 °C, a solvent-oil ratio of 3.2,and a time of 35 min. The result of verification experiment was in good agreement with the prediction of the model.

Key words: solvent formulation; FCC slurry extraction; response surface method; extract oil yield; aromatics content

1 Introduction

FCC slurry oil is one of the by-products of FCC unit and its yield has risen to approximately 5% to 10% of the feed, as a consequence of the increasingly heavy and inferior FCC feedstock[1]. FCC slurry oil is mainly composed of aromatic compounds with 3-6 rings and a few saturated hydrocarbons, which can be separated by extraction to obtain high aromatic components and saturated hydrocarbons, and can be further processed into high value-added products to improve the economic benefit of FCC unit[2-3]. Conventional FCC slurry extraction is mainly single solvent extraction, which inevitably dissolves some saturated components while dissolving aromatic components, resulting in a low aromatic content of the extracted oil[4-6]. In this study, the mixed-solvent extraction was used to separate FCC slurry, and the effect of solvent formulation and other conditions on the extraction result was discussed, and the optimal conditions were obtained by Response Surface Method (RSM)[7].

2 Experimental

2.1 Samples

Daqing FCC slurry oil was selected as the raw material.The slurry oil is firstly subjected to tailing treatment through vacuum distillation in order to remove the catalyst powder, which is harmful to the subsequent extraction and sample analysis. The fraction of ＜340 °C(approximately 600 °C under atmospheric pressure),which was cut under an operating pressure of 1 mmHg,was used as the extraction feedstock, and its yield was approximately 80% of the FCC slurry oil. The properties of the FCC slurry oil, distilled FCC slurry oil (FCC slurry instead hereafter) and extraction products obtained under the optimal conditions are shown in Table 1.

2.2 Experiment method

The mixture ofN,N-dimethylformamide (DMF) andn-paraffins (stripping-agent) was used as the extraction solvent based on previous study[8]. In the extraction process, a certain amount of FCC slurry and mixed solvent was placed in a 250-mL three-necked flask, the temperature was controlled by an electric heating jacket and mass transfer was enhanced by an agitator operating under the given conditions. The mixed solution wastransferred into a separating funnel after a specified period of time and was then put into an incubator (set at the extraction temperature) for 30 min. After the mixed solution was separated into two layers, the extraction phase in the bottom layer and the raffinate phase in the upper layer of the separating funnel were transferred into two 250-mL flasks, respectively. Finally the extract oil and raffinate oil were obtained, respectively, by removal of solvent through distillation. The composition and properties of the extraction product were analyzed, and the extraction result was evaluated by measuring the extract oil yield and the aromatic content.

Table 1 Properties of FCC slurry oil and extraction products

The influence of solvent formulation (DMF/strippingagent), extraction temperature, solvent-oil ratio and extraction time was investigated in this study.

2.3 Selection of factors and their levels

Solvent formulation, extraction temperature, solventoil ratio and extraction time are the main influencing factors for FCC slurry extraction using the mixed solvent. Orthogonal experiment of four factors and three levels was designed by using the Box-Behnken Design (BBD) method according to the result of single factor experiment[5], the influence of temperature (X1),solvent-oil ratio (X2), time (X3) and solvent formulation(X4) was investigated, respectively. There were 29 sets of experiments, 5 of which were central (zero level) experiments. The factors and its levels for the experimental design are listed in Table 2.

Table 2 Factors and its levels for experimental design

2.4 Calculation of extract oil yield

Extract oil yield is calculated by Equation (1).

wheremis the mass of extract oil (g),m0is the mass of FCC slurry feedstock (g).

3 Results and Discussion

The FCC slurry was extracted by mixed solvent containing DMF and a stripping agent having better solubility to saturates. The addition of stripping agent not only could significantly improve the aromatics content in extract oil, but also increased the saturates fraction in raffinate oil.

3.1 Results of orthogonal experiment

The results of BBD experimental design scheme are listed in Table 3, whereX1-X4are factors as those listed in Table 2,Yis the extract oil yield (%) andCis aromatics content (%) which are selected as response values.

3.2 Model of extract oil yield

The Design-Expert software can be used to conduct regression analysis on the experimental data of extract oil yield in Table 3, and a quadratic polynomial model of extract oil yield on temperature, solvent-oil ratio, time and solvent formulation was obtained, as shown in equation(2).

The variance analysis results of the extract oil yields model is shown in Table 4.

It can be seen from Table 4 that thePvalue of the extraction rate model is less than 0.0001, indicating that the polynomial model is very significant. ThePvaluesofX1,X2, andX4are less than 0.0001, and thePvalue ofX3is greater than 0.05, which means that the extraction temperature, solvent-oil ratio and solvent formulation have very significant effects on the extract oil yield model,but the extraction time has no significant effect. According to thePvalue, it can be concluded that the interactions between any two of the factors are not significant, and the square terms of factorX1,X2, andX4have a very significant impact exceptX3. ThePvalue of the lack of fit tests is 0.0585, which is greater than 0.05, indicating that the model fits well. The coefficient of determination(R2=0.9982) and the adjustedRsquare (R2(adj)=0.9965)obtained by the Design-Expert software also showed that the extract oil yield model fits the experimental data well.In a word, the polynomial model can be well adopted to predict the extract oil yield for FCC slurry separation.

Table 3 Results of orthogonal experiment

Table 4 Variance analysis of the extract oil yields model

The optimal operating conditions for the extraction on the basis of extract oil yield obtained by the Design-Expert software cover: an extraction temperature of 70 °C, a solvent-oil ratio of 3.5, an extraction time of 28.9 min,and a solvent formulation of 3.0. The extract oil yield is 63.8% and the aromatics content is 74.1% obtained under the optimal conditions.

3.3 Model of aromatics content in extract oil

Through regression analysis of the experimental data of aromatics content in Table 3, the quadratic polynomial model of aromatics content on temperature, solvent-oil ratio, time and solvent formulation was also obtained by the Design-Expert software. See Equation (3).

Figure 1 shows the experimental and predicted values of aromatics content. The predicted values showed a better consistency with the experimental data, since the points were distributed quite close to the diagonal line.

Figure 1 Predicted and experimental values of aromatics content model

The variance analysis results of the aromatics content model are shown in Table 5.

Table 5 Variance analysis of aromatics contents model

It can be seen that the polynomial model is very significant because thePvalue of the aromatics content model is less than 0.0001 according to the data depicted in Table 5. ThePvalues of factorX1,X2,X4and the interaction between factorX1andX2, as well as the square terms of factorX1,X2andX4, are less than 0.05, which indicate that temperature, solvent-oil ratio, and solvent formulation have significant effects on the aromatics content model, and there is an obvious interaction between temperature and solvent-oil ratio. ThePvalue of factorX3that is greater than 0.05 showed that the effect of time is not significant. ThePvalue of the Lack of fit tests is 0.0515 (＞0.05), verifying that the model fits well.The coefficientsR2=0.9791 andR2(adj)=0.9581 also showed that the aromatics content model fits the experiment data well. Hence the established model can be used to analyze the aromatics content in the extract oil for FCC slurry separation because of its practical significance, minimum errors, and better fitting.

3.4 BSM analysis of aromatics content

The 3D response surface and contour plots of aromatics content vs. different influencing factors were drawn according to the polynomial model by the Design-Expert software. The effects of temperature, solvent-oil ratio and time on aromatics content of extract oil are similar to the single-solvent extraction[9], so this part focuses on the influence of interactions among solvent formulation,temperature, solvent-oil ratio and time on the aromatics content.

When the solvent-oil ratio is 3.0 and time is 30 min,the 3D response surface and contour plots of aromatics content vs. solvent formulation and temperature are shown in Figure 2. It can be seen from Figure 2 that the aromatics content increased with the increase of solvent formulation and then decreased when the solvent formulation reached a certain extent. The stripping agent could extract some aromatics in the extract oil into raffinate oil when the saturates were subjected to counter-extraction, because there was more stripping agent having high solubility to saturates in mixed solvent when the solvent formulation is small. With the increase of solvent formulation, aromatics extracted by the stripping agent decreased and aromatics content of extract oil increased. However, the aromatics content of extract oil also decreased when the solvent formulation increased to a certain level, because the stripping agent content in mixed extractant is so small that can result in a reduction in efficiency of anti-extraction of saturates. The interaction of solvent formulation and temperature was not significant, according to Figure 2.

Figure 2 3D response surface and contour plots of aromatics content to compound raio and temperature

When the extraction temperature is 60 °C and the time is 30 min, the 3D response surface and contour plots of aromatics content vs. compound raio and solvent-oil ratio are shown in Figure 3. It can be seen from Figure 3 that the influence of compound raio and solvent-oil ratio on aromatics content was similar. With the increase of compound raio and solvent-oil ratio, the aromatics content first increased and then decreased, and there is no obvious interaction between solvent formulation and solvent-oil ratio.

When the extraction temperature is 60 °C and the solventoil ratio is 3.0, the 3D response surface and contour plots of aromatics content vs. compound raio and time are shown in Figure 4. The aromatics content increased at the beginning and then decreased with the increase of solvent formulation, but the time had no significant effect on aromatics content. There is no obvious interaction between the solvent formulation and the time as well.

3.5 Optimal extraction conditions and their verification

Figure 3 3D response surface and contour plots of aromatics content to compound raio and solvent-oil radio

Figure 4 3D response surface and contour plots of aromatics content to compound raio and time

In actual production, it is necessary to comprehensively consider the extract oil yield and aromatics content in order to determine the operating conditions that should suit with both indexes better. The optimal operation conditions obtained by software simulation using the fitted model cover a compound raio of 2.3, a temperature of 62.8 °C, a solvent-oil ratio of 3.2, and a time of 35 min. The expected extract oil yield was 58.5% and the aromatics content was 80.9%, respectively.

The reliability of RSM results was verified by three repeated experiments performed under the optimal operation conditions, with the results listed in Table 6. The average of extract oil yield is 58.3% and the aromatics contents is 80.7%, which is very close to the values that are predicted by the models. The model established in this paper was accurate and reliable and could predict the results of FCC slurry extraction by mixed-solvent well.

Table 6 Repetitive verification experiment conductedunder optimal operation conditions

4 Conclusions

(1)The influence of temperature, solvent-oil ratio, time and compound raio on extraction result of FCC slurry was investigated by a BBD method orthogonal experiment of four factors and three levels. And the quadratic polynomial models for extract oil yield and aromatics content were established.

(2)The results of RSM showed that temperature, solventoil ratio and compound raio have significant effects on extraction results, but the effect of time is not significant.There is obvious interaction between temperature and solvent-oil ratio during the extraction process.

(3)The optimal operating conditions of FCC slurry separation by mixed-solvent extraction cover: a compound raio of 2.3, a temperature of 62.8 °C, a solventoil ratio of 3.2 and a time of 35 min. The extract oil yield is 58.5% and the aromatics content is 80.9%, respectively,under the optimal conditions.