Prediction method of methane content change in cyclic hydrogen after desulfurization

2.1 Hydrogen-facing system flow

The flow chart of the hydrogen facing system is shown in Fig. 1.

Close

Fig. 1

Flow diagram of hydrogen facing system.

Export to PPT

The reflective effluent from R5101 (hydrogenation reactor) is heat exchanged by E5101 (feed oil/reaction effluent heat exchanger), and then the temperature drops to 250 °C and enters V5102 (hot and high pressure separator). After heat transfer by E5102 (high separation gas/mixed hydrogen heat exchanger), the hot high separation gas is cooled by A5101 (hot high separation gas air cooler) to 50 °C, and then entered into V5104 (cold high pressure separator) for oil, water and gas three-phase separation (Van et al., 2018). The high and cold gas coming from the top of V5104 is separated by V5106 (circulating hydrogen coalescer) and then enters the bottom of T5104 (circulating hydrogen desulfurization tower). After desulfurization, the circulating hydrogen comes out from the top of the tower and is separated into C5102 (circulating hydrogen compressor) by V5107 (replacing and clearing the inlet separating tank of compressor) to boost pressure. Then it is divided into two ways, one as quenching hydrogen to control the bed temperature of reactor, the other is mixed with new hydrogen from the outlet of C5101 (new hydrogen compressor) to form mixed hydrogen.

2.2 Prediction of methane content change

According to the process of hydrogen facing system, the methane content in circulating hydrogen after desulfurization is mainly affected by the solubility of methane, the density of feed oil, the composition of new hydrogen (hereinafter referred to as new hydrogen), reaction temperature, the consumption of hydrogen per ton of oil and the composition of catalyst, etc. (Hristov et al., 2018). Therefore, in the process of predicting the change of methane content in circulating hydrogen after desulfurization, it can be rooted. According to the above factors, the methane content in circulating hydrogen after desulfurization was measured by spectral absorption method and methane content device under different working conditions (different influencing factors). The change of methane content in circulating hydrogen after desulfurization was predicted according to the influence of various influencing factors on methane content.

2.2.1

2.2.1 Prediction principle

Spectral absorption method is often used to predict the change of methane content in circulating hydrogen samples after desulfurization. Spectral absorption method is based on Lambert Bill's law. The content of gas can be retrieved by detecting the change of projected light intensity. As shown in Fig. 2, when the incident light intensity is $I_0$ passes through the methane in the circulating hydrogen after desulfurization, the methane in the circulating hydrogen after desulfurization will scatter and absorb the incident light (Pardakhti et al., 2017). Because the methane in the circulating hydrogen after desulfurization absorbs part of the light energy, the intensity of the emitted light $I$ will be weakened.

Close

Fig. 2

Transmission of light.

Export to PPT

From Lambert-Beer law, we can see that:

(1)

$$\frac{I}{I_0}=e^{\alpha l}$$

That is:

(2)

$$\frac{I_0}{I}=e^{\alpha l}$$

Among them, $I_0$ is the light intensity of incident laser, $I$ is the laser intensity of methane absorbed by gas when passing through circulating hydrogen after desulfurization, $\alpha$ is the turbidity of methane in circulating hydrogen after desulfurization, $l$ is the light path of methane in circulating hydrogen after parallel light passing through desulfurization.

After taking logarithms on both sides of Eq. (2), we can get that:

(3)

$$ln\left(\frac{I_0}{I}\right)=\alpha l$$

When the diameter of suspended particles is the same and the distribution is uniform, the turbidity $\alpha$ of methane in circulating hydrogen after desulfurization is proportional to the number of particles $N$ and the diameter $D$ of suspended particles per unit volume (Jung and Park, 2017).

(4)

$$\alpha =N\frac{\pi }{4}{D}^{2}E$$

$E$is the extinction coefficient, which is the intrinsic property of matter. It is only related to the wavelength $\lambda$ of incident light, the diameter $D$ of particles and the relative refractive index $m$ of particles (Todic et al., 2017).

(5)

$$\beta =\frac{\pi D}{\lambda }$$

Then $E$ is only related to $\beta$ and $m$, and formula (4) is introduced into formula (3) to obtain:

(6)

$$ln\left(\frac{I_0}{I}\right)=N\frac{\pi }{4}{D}^{2}E\left(\beta ,m\right)l$$

For different wavelengths, ${\lambda }_1$ and ${\lambda }_2$ are introduced into the upper formula respectively.

(7)

$${\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_1}=N\frac{\pi }{4}{D}^{2}E\left({\beta }_1,m\right)l$$

(8)

$${\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_2}=N\frac{\pi }{4}{D}^{2}E\left({\beta }_2,m\right)l$$

Under the same experimental conditions, the number of particles$N$, the diameter $N$ of suspended particles and the optical path $l$of methane in circulating hydrogen after desulfurization by parallel light are all certain (Ai et al., 2017). So dividing Formula (7) by Formula (8) can be obtained:

(9)

$$\frac{{\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_1}}{{\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_2}}=\frac{E\left({\beta }_1,m\right)}{E\left({\beta }_2,m\right)}$$

Among them, ${\beta }_1=\frac{\pi D}{{\lambda }_1}$, ${\beta }_2=\frac{\pi D}{{\lambda }_2}$, ${\lambda }_1$ and ${\lambda }_2$are known quantities, ${\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_1}$ and${\left[ln\left(\frac{I_0}{I}\right)\right]}_{{\lambda }_2}$ can be obtained from formula (9), and $D$ of particle diameter can be calculated. The methane content in circulating hydrogen after desulfurization can be obtained by bringing $D$ of particle diameter into formula (4):

(10)

$$N=\frac{ln\left(\frac{I_0}{I}\right)}{\frac{\pi }{4}{D}^{2}E\left(\beta ,m\right)}$$

2.2.2

2.2.2 Selection of monochromatic light source and detection device

There are four kinds of fixed vibration modes for methane molecule (Rong et al., 2018). At the intrinsic absorption spectra of ${\lambda }_1=3.43\mu m$, ${\lambda }_2=6.78\mu m$, ${\lambda }_3=3.31\mu m$, ${\lambda }_4=7.66\mu m$, there are strong vibration absorption peaks in the wavelength range 3 μm –44 μm. Although the lead salt diode laser can produce light in this band, the light source and detector need low-temperature refrigeration, which is expensive and unsuitable for the field. Considering that the binding band $v_2+2v_3$ of methane absorption line is located near (Cui et al., 2018) 1.3 μm, which belongs to the low loss region of quartz optical fiber (Feng et al., 2018), the light source technology is relatively mature, and it is the best choice under the current technical conditions. The prediction of methane content change in circulating hydrogen after desulfurization is based on the S3FC 1310 laser produced by Thorlabs Company of the United States. The laser can produce 1310 μm laser, and the output laser precision error will not exceed 1 μm, that is, the output power will be between 1309 μm and 1311 μm.

From the Hitran database of Harvard University, it can be found that there are 251,440 absorption lines of methane gas (Yang, 2018), of which 667 absorption lines are absorbed from 1312 μm (7627 cm⁻¹) to 1307 μm (7627 cm⁻¹). It can be concluded that the absorption intensity of methane gas at 1309 μm (7639 cm⁻¹) and 1311 μm (7627 cm⁻¹) is the weakest up to 0.6*10⁻²³.

The device for predicting the change of methane content in circulating hydrogen after desulfurization by monochrome spectral absorption method based on optical fiber coupling technology is shown in Fig. 3.

Close

Fig. 3

Prediction device for methane content change.

Export to PPT

The device consists of light source, temperature regulator, current regulator, optical fiber coupler, isolator, optical fiber detector, data acquisition card, computer and signal processing system. The application of optical coupler in spectral absorption method can ensure that the light intensity received at any site and the light intensity of the light source remain unchanged after circulating hydrogen after desulfurization under different working conditions (Wang et al., 2017), thus realizing real-time on-line remote detection.

According to the real-time test results, the influence of different working conditions on the methane content in circulating hydrogen after desulfurization can be obtained. Based on this, the change of methane content in circulating hydrogen after desulfurization can be predicted under the fluctuating conditions of various influencing factors.

3.1 Detection result

This method is used to predict the change of methane content in circulating hydrogen after desulfurization from March to November 2017. The specific data are shown in Fig. 4.

Close

Fig. 4

Prediction results of methane content change.

Export to PPT

From Fig. 4, it can be seen that the methane content in circulating hydrogen after desulfurization has been increasing gradually since March, reaching the maximum content of 10.76% in May. This is because with the prolongation of operation time, the methane content in circulating hydrogen after desulfurization will gradually enrich and cause the content to increase. The methane content in recycled hydrogen decreased to 7.99% in June due to the exhaust of hydrogen in May. The hydrogenation unit was shut down for overhaul in August for the end of operation in July. After maintenance, the methane content in recycled hydrogen increased from 2.16% in September to 7.38% in November. The experimental results show that this method can effectively predict the change of methane content in circulating hydrogen after desulfurization.

3.2 Effect of methane solubility on prediction of methane content change

The effect of Methane Solubility (methane molar fraction) on the change of circulating hydrogen methane content after desulfurization is shown in Fig. 5.

Close

Fig. 5

Effect of Methane Solubility.

Export to PPT

Fig. 5 shows that the solubility of methane increases with the increase of high cooling temperature, and the relationship between the solubility of methane and the circulating hydrogen methane content after desulfurization is inverse. The higher the solubility of methane, the higher the hydrogen content in the equilibrium gas component, that is, the lower the methane content in circulating hydrogen after desulfurization. The experimental results show that the higher the solubility of methane, the lower the methane content in circulating hydrogen after desulfurization.

3.3 Effect of raw oil density on the change of methane content

In order to analyze the influence of feed oil density on the change of circulating hydrogen methane content after desulfurization, the conditions of close volume fraction of new hydrogen methane and reaction temperature and large difference of feed oil density were selected for analysis. The results are shown in Table 1.

Comparing and analyzing the data of No.3 and No.4, No.3 and No.6 in Table 1, it can be seen that methane is more produced when the density of feed oil component is small, but the data of No.2 and No.7, No.1 and No.5 show the opposite trend. The reason is that when the density of feed oil is 892–909 kg/m³, the change of composition is not obvious, and it has little effect on the circulating hydrogen methane content after desulfurization.

3.4 Effect of new hydrogen methane content on the change of methane content

In order to analyze the effect of new hydrogen methane content on the change of methane content in circulating hydrogen after desulfurization, sample data with stable reaction temperature and close density of feed oil were selected for analysis. The results are shown in Table 2.

Table 2 shows that the recycled hydrogen methane content ranges from 3.40% to 10.06% when the temperature is 330 °C, the density of filtered feed oil is 891.1–905 kg/m³ and the content of new hydrogen methane is 2.08%–4.74%.

By analyzing the data of the first four working conditions and the last three working conditions in Table 3, it is found that the increase of methane content in new hydrogen will lead to the increase of methane content in circulating hydrogen. The comparative analysis of seven groups of data in Table 3 shows that when the reaction temperature is the same and the density of feed oil changes little, the content of circulating hydrogen methane is mainly affected by the content of new hydrogen methane, and increases with the content of new hydrogen methane.

3.5 Effect of reaction temperature on the change of methane content

In order to analyze the influence of reaction temperature on the prediction of methane content in circulating hydrogen after desulfurization, samples with similar density of feed oil and new hydrogen methane content and larger difference in reaction temperature were selected for analysis. The results are shown in Table 3.

Table 3 shows that the methane content in recycled hydrogen is 3.46%–4.52% when the reactor temperature is 310 °C, the density of feed oil is 889.3 kg/m³ and the content of new hydrogen methane is 1.99%–3.36%. At 355 °C, the density of feed oil is 891.1 kg/m³ and the content of new hydrogen methane is 2.04%–3.32%, the methane content in recycled hydrogen is 8.16%–9.46%.

By comparing and analyzing the data of working conditions (1) and (6), 2 and 4, 3 and 4 in Table 3, it can be concluded that the increase of reaction temperature will lead to the increase of methane content in circulating hydrogen. The reason may be that when the hydrogenation reaction temperature is adjusted, the temperature rises, which causes deep cracking of feed oil and more methane.

3.6 Effect of hydrogen consumption per ton of oil on the change of methane content

In view of the gradual decrease of catalyst activity, the research object began to carry out daily catalyst replacement according to the catalyst replacement scheme and the instructions for the use of catalyst. The replacement amount was 0.6 t/d of new catalyst. Five periods of similar working conditions before and after the overall change were selected for comparative analysis. The effect of hydrogen consumption per ton of oil on the change of methane content in circulating hydrogen was shown in Fig. 6.

Close

Fig. 6

Effect of oil and hydrogen consumption on methane content in recycled hydrogen.

Export to PPT

As can be seen from Fig. 6, the methane content in recycled hydrogen decreased from 1.86% to 1.14% and the reduction rate was 38.7% after the catalyst was replaced as a whole. The methane content in recycled hydrogen decreases with the constant feed property and load, which indicates that the selectivity of catalyst after overall replacement increases and the selectivity of oil from tail fracture decreases.

3.7 Effect of catalyst components on the change of methane content

The effect of catalyst composition on methane content in circulating hydrogen after desulfurization is shown in Fig. 7.

Close

Fig. 7

Effect of catalyst composition on methane content in circulating hydrogen after desulfurization.

Export to PPT

Analysis of Fig. 7 shows that FF-24 catalyst has high hydrodesulfurization activity, and its hydrodesulfurization activity is significantly higher than that of FF-18 and two other catalysts FFT-18 and HRK-658 before maintenance. FF-24 catalyst is supported by special alumina and Mo-Ni-Co is used as active metal component of catalyst. It has the characteristics of good dispersion of active metal, simple preparation process and low cost. The hydrodesulfurization activity of FF-24 catalyst results in the hydrocracking reaction of feed oil, which produces more methane.

3.8 Accuracy of calculation of methane content in circulating hydrogen after desulfurization

The methane content in circulating hydrogen after desulfurization is calculated by this method. The results are shown in Table 4.

Table 4 shows that this method can not only predict the change of methane content in circulating hydrogen after desulfurization, but also the error between the predicted results and the actual content is less than 1.0%. It shows that this method has high accuracy in predicting the change of methane content in circulating hydrogen after desulfurization.

3.9 Prediction efficiency analysis

In order to test the prediction efficiency of this method, this method and three other models were used to predict the change of methane content in circulating hydrogen after desulfurization in different months of 2018, and the time required for different methods was compared. The results are shown in Table 5.

Analysis Table 5 shows that the time used to predict the change of methane content in circulating hydrogen after desulfurization in different months of 2018 is 1.53–2.22 s, and the average time spent is 1.95 s. The average time spent by the other three models is 5.47 s, 4.37 s and 7.33 s, respectively. The experimental results show that this method takes less time to predict the change of methane content in circulating hydrogen after desulfurization and has the highest efficiency.