Article Cite This: Ind. Eng. Chem. Res. 2019, 58, 12301−12311
pubs.acs.org/IECR
Electrostatic Distribution in the Riser of the Multizone Circulating Fluidized Bed for Polypropylene Zhedong Lou,† Shiyi Ge,† Yao Yang,*,† Zhengliang Huang,† Jingyuan Sun,† Jingdai Wang,‡ Yongrong Yang,‡ Lei Xie,§ Hongye Su,§ and Yunzhong Gao∥ †
Downloaded via NOTTINGHAM TRENT UNIV on August 7, 2019 at 08:04:19 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Zhejiang Provincial Key Laboratory of Advanced Chemical Engineering Manufacture Technology, College of Chemical and Biological Engineering, ‡State Key Laboratory of Chemical Engineering, College of Chemical and Biological Engineering, and § College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, People’s Republic of China ∥ SINOPEC SABIC TianJin Petrochemical Company Limited, Tianjing 300000, People’s Republic of China S Supporting Information *
ABSTRACT: This work investigates the electrostatic distribution in the riser of a multizone circulation fluidized bed for polypropylene by various electrostatic measurements. Results show all measured electrostatic parameters have a strong dependence on flow patterns with increasing bed height. They remain constant in dense-phase pneumatic transportation, while they decrease in fast fluidization. However, their polarities are different. The charge-to-mass ratio is negative under all conditions, but the negative electrostatic potential only appears in the lower part of the riser at lower gas velocities. Further signal source analysis shows that only contact electrostatic potential generated by charge transfer has a one-to-one correspondence with real charge on particles. In addition, compared to the induced potential, the contact potential in regions with higher solids concentration tends to be more negative under stronger effects of charge transfer, while in regions with lower solids concentration it tends to be more positive under strong effects of contact charging.
1. INTRODUCTION The Spherizone process developed by Lyondell Basell is the most innovative and advanced process for the production of polypropylene. In this process, a multizone circulating fluidized bed reactor is used instead of the loop reactor in the original process, which realizes the cyclic polymerization of polymer particles between two reaction zones controlled by different conditions, thereby producing a polymer product with uniform mixing but wide adjustable molecular weight distribution.1 However, in the multizone circulating fluidized bed reactor, a large amount of charges are easily generated and accumulated on the particle surfaces because of intensive friction and collisions between particles and the wall in an environment with relative low humidity and high insulation.2 Extra accumulation of electrostatic charges can cause various problems, including particle sticking, agglomeration, and even abnormal sparking;3 thus static electricity becomes a severe problem that cannot be ignored definitely in the circulating fluidized bed reactor. In the existing industrial process, due to the lack of understanding of the electrostatic phenomenon in the circulating fluidized beds, the electrostatics up to several thousand of volts during the production of polypropylene (the industrial data can be found in the Supporting Information) can only be solved by adding an excessive amount of antistatic agent, although a riser made of metal is used in the industry. © 2019 American Chemical Society
Considering that the antistatic agent will negatively impact the catalyst productivity or change the product molecular weight and molecular weight distribution,3 the addition of an excessive amount of antistatic electrostatic agent will inevitably affect the activity of the catalyst. Therefore, it is necessary to study the electrostatic distribution in the circulating fluidized bed reactor and reveal the generation mechanism of electrostatic charges, and then supply a fundamental rule for the addition strategy of antistatic agent in the real industrial process. Electrostatic phenomena are ubiquitous in gas−solid systems involving dry particles.4,5 A large number of researchers have conducted detailed studies on the electrostatic phenomena in gas−solid fluidized beds. In the early stage, researchers focused on understanding the electrostatic distribution in the fluidized bed and analyzing the generation mechanism of electrostatics in the fluidized bed. The results show that bipolar charging is the main cause for the generation of electrostatic charges in fluidized beds with materialsidentical particles. Wang et al.6 used a contact-type electrostatic probe to detect the axial and radial electrostatic potential Received: Revised: Accepted: Published: 12301
January 31, 2019 May 10, 2019 June 10, 2019 June 10, 2019 DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research distributions in a bubbling fluidized bed with polyethylene particles, and explained the uneven electrostatic potential distribution along the axial direction by the “bipolar charging” theory. That is, the difference in surface work function between small particles in the upper bed layer and large particles in the lower bed layer leads to the opposite electrostatic polarities in the fluidization process, which also causes the electrostatic potential distribution in the bed. Chen et al.7,8 studied the charge distribution around a single bubble during the ascending process in a fluidized bed. By continuously optimizing the detection method and the algorithm model, the predicted results are in good agreement with the experimental results. This research makes the study of electrostatics in gas−solid fluidized beds from the macroscale into the mesoscale phenomena such as bubbles. After fully understanding the electrostatic distribution in the fluidized bed, researchers mainly focused on the influence of electrostatics on the hydrodynamic behaviors in the fluidized beds in recent years. Dong et al.9,10 studied effects of electrostatic forces on bubbles and particle motions in the bubbling fluidized bed. Results showed that the existence of electrostatic force obviously affected the bubble size, the average collision angle, and the particle velocity. The more the particles were charged, the smaller the bubble size, the larger the average collision angle, and the higher the particle velocity. Partial experimental results of Dong et al. were in accordance with simulated results of Hassani et al.11 Besides the hydrodynamics in the dense-phase bed layer, electrostatics also plays an important role in the process of particle elutriation, and even affects heat transfer ability.12−16 It is not difficult to find that aforementioned studies mainly focus on dense-phase fluidized beds with higher particle concentrations (commonly above 0.3); however, studies involved in the electrostatics in circulating fluidized beds with lower particle concentrations (commonly lower than 0.1) are scanty. Cheng17 studied the electrostatics in a large circulating fluidized bed by detecting the electrostatic current. It was found that, under the pneumatic transport flow pattern, the electrostatic current in the fully developed section of the riser was inversely proportional to the solid flow rate. In their work, the study of electrostatics in gas−solid two-phase system is extended from a dense-phase fluidized bed to a circulating fluidized bed with a relatively low particle concentration, which is of great significance. However, this research only covers the electrostatics at limited positions and the electrostatic distribution along the whole riser is not included; meanwhile, the operating conditions involved are relatively simple and cannot fully reveal the electrostatic distribution in the circulating fluidized bed and clarify the relationship between complex flow patterns and electrostatic distributions. In addition, the circulating fluidized bed used in this study is a three-stage circulating fluidized bed consisting of three parts riser, downer, and bubbling sectionwhich is totally different from the circulation fluidized bed for production of polypropylene. Therefore, this study has only a weak reference for understanding and solving the electrostatic problems in polypropylene circulating fluidized bed reactors. In addition to the above research, the dilute-phase pneumatic transport system is also a low-concentration gas− solid two-phase flow system close to the riser of the circulating fluidized bed, and electrostatic research in pneumatic transport systems may also show some guidance for the understanding of the electrostatic problems in the circulating fluidized bed.
Joseph and Klinzing18 found that, in vertical pneumatic transport, the presence of electrostatic forces increased the gas velocity for the minimum transportation, thereby increasing the energy consumption during the transporting process. Zhang et al.19 studied effects of electrostatics on particle motions in the dilute phase transport system. It was found that the electrostatics significantly affected the particle velocity distribution in the radial direction. This conclusion coincides with results of Dong et al.9 in the fluidized beds. Yao et al.20 observed that, in the presence of electrostatics, halfring-shape and ring-shape agglomerates could be formed when large particles were transported, and which further affected the flow patterns. It can be seen that these studies related to pneumatic transport proved again that the influence of electrostatics in the gas−solid flow system with low particle concentration and high particle velocities cannot be ignored, and electrostatics strongly couples with the flow pattern of particles. Investigation of electrostatics in the gas−solid twophase flow must be combined with the flow pattern. The basis of the electrostatic investigation is a reliable and stable detection method. Currently, based on different detection methods, parameters used to characterize the electrostatic level include the charge-to-mass ratio, contact/ induced electrostatic potential, and contact/induced electrostatic current. Comparing them, the influence of the contact between particles and the probe on the electrostatic potential signals can be revealed. The average charge-to-mass ratio of particles based on the Faraday cup is the most intrinsic parameter describing the electrostatic level, while other parameters are indirect. Therefore, how to establish the qualitative and quantitative relationships between the average charge-to-mass ratio of particles and other detection parameters is extremely important. In the dense-phase fluidized bed, the researchers generally directly use the electrostatic potential signals detected by the contact probe to characterize the electrostatic level.21 This is because in the dense-phase system the relative velocity between particles and the probe is small, and signals detected by the probe mainly come from the charge transfer of charged particles around the probe;22 thus the detected electrostatic potential signal is basically related to the surface charge density of particles. However, for a fast fluidized system, contacts between particles and the probe are completely different and the influence of contact charging on the detection becomes non-negligible. Therefore, the applicability of various indirect online detection methods used for fluidized beds should be checked, which is also the prerequisite for the study of electrostatics in circulating fluidized beds. Above all, the purpose of this work is to study the electrostatic distribution in a polypropylene circulating fluidized bed reactor and investigate the applicability of various electrostatic detection methods. The main research contents are as follows. First, different flow patterns were constructed by changing the superficial gas velocity in the riser, and the relationship between the electrostatic level and the flow pattern in the riser was analyzed by the charge-to-mass ratio detection. Second, the contact/induced electrostatic potential was simultaneously measured and compared with the charge-tomass ratio measurement to check the applicability of the contact/induced electrostatic potential detection. Finally, the detection mechanism of contact/induced electrostatic potential in the circulating fluidized bed reactor was analyzed to explain the relationship between the contact/induced electrostatic potential and the charge-to-mass ratio measurements. 12302
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
2. EXPERIMENTAL SECTION 2.1. Apparatus and Materials. The experimental device mainly consists of a multizone circulating fluidized bed and a signal acquisition system. As shown in Figure 1, the circulating
mounted in the riser with the same interval of 0.15 m. The length of the probe into the bed is approximately 5 mm. The pressure fluctuation detection system uses a Plexiglas pressure tube and CYG1219 series pressure sensors to collect pressure pulsation signals at different axial heights of the riser. The Plexiglas pressure tube has a length of 25 mm and an inner diameter of 4 mm. The front end is flush with the inner wall of the fluidized bed and is covered with a fine mesh to prevent fluidized particles from entering. The positions of the pressure tube are the same as those of electrostatic probes in the riser. The pressure sensor adopted in this work is a two-channel differential pressure sensor with a measuring range of ±5 kPa and a resolution of 0.25%. During the experiment, the two channels are respectively connected to the two adjacent pressure tubes in the riser. The Faraday cup and electrometers (NanoCoulomb Meter 284, Monroe Electronics) were used to measure the average charge-to-mass ratio of charged particles at different axial positions of the riser. Taking into account the possible impact of the sampling process on the measurement results, each measurement is based on the average value of three repeated measurements. Each sampling port is just a pore on the wall, and its diameter is 5 mm and its length is the same as the thickness of the wall (2 mm). Its position is 20 mm parallel to each electrostatic probe. The particles used in the experiment are polypropylene particles (PP), and the air is used as the fluidization medium. During the experiment, the air is at room temperature and its humidity is controlled to less than 10%. The properties of PP particles and the particle size distribution are shown in Tables 1 and 2, respectively. The particle characteristic velocities
Figure 1. Schematic diagram of the cold-model experimental system. 1, Roots blower; 2, buffer tank; 3, dryer; 4, rotameter; 5, computer; 6, data acquisition (DAQ); 7, pressure fluctuation sensor; 8, electrostatic sensor; 9, circulating fluidized bed; 10, butterfly valve; 11, electrostatic probe; 12, cyclone; 13, external cyclone.
fluidized bed comprises two Plexiglas straight pipes, wherein the left one is the riser with a height of 2.0 m and an inner diameter of 0.09 m, and the right one is the cyclone separation section connected with the downer. The Plexiglas straight pipes were chosen since Plexiglas and steel are both easier to carry positive charges than polypropylene (PP) in triboelectric series,23 and their triboelectric series was confirmed by preexperiment. The downer is 1.3 m in height and 0.06 m in inner diameter. In the experiment, the air from the Roots blower passes through the dryer and the rotameter, and then enters the bottom of the riser to fluidize the particles. The gas and particles leaving the top of the riser pass through a smooth right-angle elbow into the cyclone separation section. In the cyclone separation section, particles fall into the downer which the gas is discharged through the top outlet. Particles in the downer pass through the butterfly valve which controls the solids circulation speed and the smooth bend, and then enter the riser again. The unique characteristic of the circulating fluidized bed is that particles in the downer move downward as the form of a dense-phase moving bed. The signal acquisition system mainly includes three parts: electrostatic potential detection, pressure fluctuation detection, and charge-to-mass ratio detection. The electrostatic potential online detection system consists of an electrostatic collision probe, an electrostatic induction probe, a voltage−current conversion module, a data acquisition card (NI, USB-6351), and a computer. The main body of the probe is a stainless steel rod with a length of 125 mm and a diameter of 5 mm.6 The size of the collision probe is the same as that of the induction probe. The difference is that the front end of the collision probe is exposed to the bed, while the front end of the induction probe into the bed is covered with a Teflon insulation layer. As shown in Figure 1, starting from 0.15 m above the entrance of the riser, a total of 13 probes are
Table 1. Properties of PP particle density, kg/m3 mean particle size, mm minimum fluidizing gas velocity, m/s fast bed transition gas velocity, m/s terminal velocity, m/s
900 2.38 0.64 4.5 7.0
Table 2. Particle Size Distribution of PP particle size, μm mass fraction, wt %
2800 26.8
2000 43.2
1000 27.9
500 1.7
75 0.3
shown in Table 1 were all calculated from empirical formulas,24 and the particle size distribution data shown in Table 2 were measured by sieving. 2.2. Experimental Method. The experiments were divided into three parts depending on the purpose of the experiment. Dry polypropylene pellets were prefilled into the downer from the top of the cyclone separation section prior to each run of the experiment. At the beginning of the experiment, the fluidization gas velocity was first adjusted to the set gas velocity, then the butterfly valve at the bottom of the downer was opened to make the particles enter the riser, and the downward velocity of particles in the downer, namely the solids circulating speed, was controlled by the opening degree of the butterfly valve. Throughout the experiment, the downward velocity of particles was continuously recorded and always controlled as 2.0 cm/s. The recorded downward velocities at different gas velocities are shown in Figure 2. When the downward velocity of particles and the gas velocity were both stable, the bed would be run for more than 3 h to 12303
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
Figure 2. Variations of downward velocities of particles with time at different gas velocities.
Figure 3. Voidage distributions along the axial direction of the riser at different gas velocities.
ensure that the static electricity distribution in the bed was stable. After the static electricity distribution in the bed was stabilized, the detailed steps of each part of the experiment were as follows. Part I. In order to detect the voidage distribution along the axial direction in the riser to determine the specific flow pattern, the pressure drop between the two adjacent measuring positions was detected one by one. The sampling frequency for each test was 400 Hz, and the sampling time was 60 s. Part II. In order to detect the electrostatic potential distribution in the circulating fluidized bed, after the electrostatics in the bed is stabilized, the electrostatic potential detection system was connected to the electrostatic probes one by one to detect the contact and induced electrostatic potential. Here, the contact/induced electrostatic potential
signals were detected by the same system, and the only difference is if the metal probe is wrapped in PTFE insulation and in direct contact with charged particles. When using this system to measure the electrostatic potentials, it was observed that since the electrostatic potential signals are related to the particle motions and particle motions are fluctuating, the electrostatic signals fluctuate around a stable value. Therefore, the sampling frequency was set as high as 400 Hz, and the sampling time was set as long as 3 min to ensure the stability of the data. The average electrostatic potential was used next to decrease the uncertainties. Moreover, the experiment under the same conditions was repeated as least three times to further decrease the uncertainties, and the discussion focused on the trend of the electrostatic potential rather than the absolute value. 12304
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
Figure 4. Axial distributions of charge-to-mass ratio in the riser under different gas velocities.
divided into two areas based on the observed particle distribution in the experiments. The upper part of the riser was similar to the flow pattern at higher gas velocity, which was a dilute phase region with a larger voidage, while the lower part of the riser was a dense-phase region with a smaller voidage. The interface between the two regions was distinct. Above all, two different flow patterns were constructed in this work by selecting proper gas velocities. Below, the static electricity distribution in the riser under these two different flow patterns will be investigated. 3.2. Axial Distribution of Charge-to-Mass Ratio. Although the sampling process in the charge-to-mass ratio detection increases the complexity and difficulty of this method and the detection result is susceptible to the sampling process, the average charge-to-mass ratio of the particles is still the most direct parameter describing the electrostatic level in gas−solid two-phase systems. Therefore, in order to demonstrate the real electrostatic distribution in the riser, Figure 4 shows the axial distribution of the average charge-to-mass ratio of particles in the riser of the circulating fluidized bed at different operating gas velocities. As can be seen from Figure 4, particles at all locations within the bed carry substantially a certain amount of negative charges. According to previous research, there are two generation mechanisms for the static electricity in the gas− solid systems: one is the friction between the particles and the wall surface, and the other is the bipolar charging caused by the contacts between particles with different sizes. Lim et al.27 found that in the process of pneumatic transport, due to the low probability of contacts between particles and particles, static electricity was mainly generated by the interaction between particles and the wall. The chargeability and polarity of charged particles were determined by materials of particles and the wall. The study by Wang et al.6 found that, in densephase fluidized bed, bipolar charging caused by frequent contacts between particles and particles was the main cause of static electricity; thus it showed different electrostatic levels as well as polarities at different heights of the fluidized bed. Combined with the results of previous studies and the results shown in Figure 4, it can be concluded that the main cause of
Part III. In order to investigate the axial distribution of the charge-to-mass ratio in the riser, after the electrostatic potential distribution of the circulating fluidized bed was stabilized, particles were sampled through the sampling port and directly into the Faraday cup for charge-to-mass ratio measurements. After the charge-to-mass ratio measurement was completed, particles were added to the fluidized bed to ensure that the total amount in the bed did not change.
3. RESULTS AND DISCUSSION 3.1. Axial Voidage Distribution in the Riser. For fluidized beds, the bed pressure drop is directly proportional to the weight of the particles in the bed.25 Thus, the voidage between two measured positions of pressure signals can be calculated by eq 1. ε=1−
ΔP ρp gh
(1)
where ΔP is the bed pressure drop, ρp is the particle density, ε is the voidage, g is the acceleration due to gravity, and h is the height difference between two measured positions. In order to determine the flow patterns at different gas velocities, the axial voidage distribution in the riser was calculated based on the pressure drop measurements26 and is shown in Figure 3. It can be seen from Figure 3 that, under the five gas velocities used in this work, the flow patterns in the riser can be divided into two types. At three larger gas velocities, the axial voidage distribution conformed to the characteristics of dense-phase pneumatic transport24 which showed that the voidage in the riser was basically constant along the axial direction. In addition, the voidage in the riser increased as the gas velocity increased, and the voidage was mostly greater than 0.95. At the two lower gas velocities, the voidage distribution agreed well with the characteristics of fast fluidization; that is, the voidage increased gradually along the axial direction of the riser, and there was almost no area where the voidage was relatively stable with the increasing bed height. Under the flow pattern of fast fluidization, the riser could be 12305
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
Figure 5. Axial distribution of the contact electrostatic potential in the riser under different air velocities.
static electricity in the riser should be contacts between particles and the wall due to the low solid concentration in the riser, closer to the pneumatic transport system. Since the polypropylene is located in front of the Plexiglas in the rubbing sequence,2 the PP particles are negatively charged after contacting with the Plexiglas wall and the wall is positively charged. Therefore, the measured charge-to-mass ratio of particles at any position of the riser exhibited a negative charge. At the same time, it can be clearly seen that the particle charge-to-mass ratio shows different axial distributions under different flow patterns. At two lower gas velocities with the flow pattern of fast fluidization, the charge-to-mass ratio in the riser increased with the axial height; furthermore, when it exceeded a certain position, the charge-to-mass ratio remained basically the same showing a weak negative charge. But at three higher gas velocities, the charge-to-mass ratios of the particles in the entire riser were basically the same; only the particle charge-to-mass ratios close to the outlet and inlet were slightly lower. Moreover, the higher the gas velocity, the smaller the average charge-to-mass ratio of particles. Thus, there is more electrostatic risk under the flow pattern of fast fluidization from above. The particle charge-to-mass ratio distribution in Figure 4 is consistent with the voidage distribution shown in Figure 3. Since the static electricity is mainly generated by contacts between particles and the wall, it is severely affected by the particle−wall contact probability. When the riser is in fast fluidization (low gas velocities), the particle concentration in the riser is high, especially in the lower part of the riser; thus the particle−wall contact is extremely frequent resulting in a large negative charge in the lower part of the bed. In addition, at the same gas velocity, the particle concentration gradually decreases with the bed height; thus the particle−wall contact probability decreases gradually as well. Therefore, the negative charges carried by the particles gradually decrease with the increase of the bed height. When the flow pattern in the riser is in the pneumatic transport pattern (high gas velocities), since the particle concentration is extremely low and does not change much with the axial height, the negative charges on particles in the entire riser are low, and the measured charge-
to-mass ratio remains almost the same at any height. At the same time, as the particle concentration decreases with the increase of the operating gas velocity, the particle−wall contact probability also decreases with the increase of the gas velocity, resulting in the electrostatic level in the riser decreasing with the operating gas velocity. Based on above measurements, for the riser of the circulating fluidized bed for polypropylene, if the antistatic agent is needed, it should be added in the lower part of the riser in fast fluidization due to the following reasons. On one hand, the electrostatic level is highest in the lower part of the riser in the fast fluidization; thus injecting the antistatic agent at this position may directly influence the electrostatics. On the other hand, since the particle concentration is highest in the lower part of the riser in the fast fluidization, adding the antistatic agent at this place may ensure the complete contacts between particles and the antistatic agent. But in the densephase pneumatic transport flow pattern, it can be added at any position because the axial distribution of electrostatics and voidage are both relatively uniform. 3.3. Axial Distribution of Electrostatic Potential. The detection of the charge-to-mass ratio in section 3.2 indicates that the main cause of the static electricity in the riser of the circulating fluidized bed is the contact charging between the particles and the wall, and the electrostatic distribution has a dependence on the flow pattern. However, due to the complexity of the detection process of the particle charge-tomass ratio, it is not possible to use it to realize the online detection of the electrostatic level in the industrial process. Therefore, the applicability of different online detection methods, including the measurements of contact and induced electrostatic potential, will be checked in the following part by comparison with the measured charge-to-mass ratio. Figure 5 compares the axial distribution of contact electrostatic potential in the riser of the circulating fluidized bed at different operating gas velocities. From Figure 5, as the gas velocity decreased, the electrostatic potential at the lower measuring point decreased accordingly. The reason may be that the decrease of the gas velocity increases the particle concentration in the bottom of the riser, as well as the contact 12306
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
Figure 6. Axial distribution of induced electrostatic potential in the riser at different gas velocities.
Figure 7. Comparison of normalized particle charge-to-mass ratio (q/m), contact electrostatic potential (UC), and induced electrostatic potential (UI) at different gas velocities: (a) 5.68, (b) 6.11, (c) 6.99, and (d) 7.42 m/s. 12307
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
compared, as shown in Figure 7. Taking the contact electrostatic potential as an example, the normalized methods can be expressed as eq 2.
frequency between the particles and the wall, and then more negatively charges on the surface of particles leading to the value of the electrostatic potential. At the same time, it can be found from Figure 5 that the distribution law of the contact electrostatic potential also shows a dependence on the flow pattern, which is consistent with the detection result of the charge-to-mass ratio. At low gas velocities with fast fluidization, in the bottom of the riser, namely, the dense-phase section, the contact electrostatic potential is negative, and gradually becomes positive with the increase of the detection height. At high gas velocities, the contact electrostatic potential detected at any position is positive, and the contact electrostatic potential does not change much except for the detection positions near the inlet and outlet of the riser. Further comparing the contact electrostatic potential distribution shown in Figure 5 with the charge-to-mass ratio distribution shown in Figure 4, it can be found that although the distribution law of the contact electrostatic potential also has a dependence on flow pattern; there are large deviations between the contact electrostatic potential and the measured charge-to-mass ratio, which are mainly reflected in the electrostatic polarity. The charge-to-mass ratio shown in Figure 4 indicates that the particles in the polypropylene circulating fluidized bed are mainly negatively charged, but the detection results of the contact electrostatic potential only show negative values in the lower part of the riser at low operating gas velocity, and others all show positive contact electrostatic potential. This may be caused by the fact that, in a circulating fluidized bed where the particle concentration is low and the particle velocities are very high, the static electricity detected by the contact electrostatic probe is not completely derived from the charge transfer of charged particles, which results in the large deviation in the measured electrostatic polarity by different methods. Therefore, the above results indicate that, taking the measured charge-to-mass ratio as a basis, the contact electrostatic potential cannot be directly used to characterize the electrostatics in the circulating fluidized bed. Figure 6 gives the axial distribution of induced electrostatic potential in the riser of the circulating fluidized bed at different operating gas velocities. It can be seen from Figure 6 that the distribution law of the induced electrostatic potential is also basically consistent with the contact electrostatic potential, and shows different distributions in different flow patterns. That is, at low gas velocities, the dense phase in the lower part of the riser exhibits negative induced electrostatic potentials, and as the height increases, the induced electrostatic potential gradually becomes positive. At high gas velocities, the induced electrostatic potential is always positive, and the induced electrostatic potential is extremely low except for these positions near the inlet and outlet. The highest measuring position is near the riser exit; the voidage is almost the same at different gas velocities from our experiments and others’ study.19 Thus the electrostatic potentials at this position tend to be the same. Comparing the induced electrostatic potential distribution shown in Figure 6 with the particle charge-to-mass ratio distribution shown in Figure 4, it can be found that the polarity of the induced electrostatic potential also has no direct correspondence with the particle charge-to-mass ratio. In order to further compare these three electrostatic parameters, the average charge-to-mass ratio (q/m), the contact electrostatic potential (UC), and the induced electrostatic potential (UI) at each gas velocity were normalized and
UC, i ,normalized =
UC, i ,original − UC,min UC,max − UC,min
(2)
where UC,i,normalized is the normalized contact electrostatic potential at position i, UC,i,original is the original contact electrostatic potential at position i, and UC,min and UC,max are the maximum and minimum contact electrostatic potentials along the axial direction of the riser at a certain gas velocity. Based on methods shown in eq 1, all electrostatic parameters were distributed between 0 and 1, and they could be compared in the same level. It can be clearly seen from Figure 7 that, since all these measurements have good dependence on flow patterns, the normalized contact and induced electrostatic potentials have almost the same distribution as that of the charge-to-mass ratio except for the results at the gas velocity of 6.55 m/s. In summary, results in Figure 7 show that although the polarities of electrostatic potentials measured by online measurements differ from the charge-to-mass ratio measured by the offline method, it is possible to unify the electrostatic potentials and the charge-to-mass ratio to online characterize the electrostatic level in the riser of the circulating fluidized bed precisely. 3.4. Analysis of the Generation Mechanism of Electrostatic Potential Signals. In order to find out the difference between the measured contact/induced electrostatic potential and the charge-to-mass ratio, the source of the electrostatic potential signal is analyzed. In essence, the contact/induced electrostatic potential signals are a reflection of the charge level of the particles and the wall near the electrostatic probe. As shown in Figure 8, in the circulating fluidized bed used in this work, the electrostatic potential signals mainly have four sources:
Figure 8. Main sources of electrostatic potential signals.
(a) The Plexiglas wall surface contacts frequently with the particles, which causes the Plexiglas wall to be positively charged and generates a positive charge on the steel probe through electrostatic induction. During the experiments, we found that after the bed was run for hours, each time when our hands approached the bed wall, hairs on the arms would stand up. This feeling proved the strong electrostatics on the wall. (b) Negatively charged polypropylene particles will generate the induced negative potential signals when passing through the electrostatic probe. (c) When the negatively charged polypropylene particles in the bed directly contact the electrostatic probe, if it is the collision probe with the front end exposed to the bed, charge 12308
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
Figure 9. Difference between contact and induced electrostatic potentials at different axial positions under different gas velocities.
probe, it covers all the above four sources.22 Therefore, the difference between the induced electrostatic potential and the contact electrostatic potential mainly comes from the difference between the latter two signal sources. As to charge transfer of source c, it is found that, for charge transfer between polymer particles and metals, the amount of transferred charge is related to the collision frequency, and the transferred charge properties are consistent with the charge on the surface of the particles.29 Since the measured charge-to-mass ratio indicates that particles in the bed mainly carry negative charges, it will definitely generate negative electrostatic potential and be influenced by the contact probability between particles and the probe. This means that the increase of solid concentration will increase the effect of source c and lead to a more negative contact electrostatic potential than the induced one. For source d, the surface work function difference between the steel probe and the polypropylene determines that the contact charging will lead to positive contact electrostatic potential.4 Besides, since the contact charging will be enhanced with the increase of particle speed in a certain range, when the gas velocity is increased and the particle concentration in the riser is decreased, the total contact electrostatic potential will shift to be more positive than the induced one. In summary, based on the above analysis, when the solid concentration increases, the contact electrostatic potential will tend to be more negative than the induced one; on the contrary, the contact one will tend to be more positive than the induced one. Figure 9 further gives the difference between the contact electrostatic potential and the induced electrostatic potential at different axial positions under different gas velocities, which is in accordance with the above analysis. In Figure 9b with high gas velocities, since the effect of source d is stronger than that of source c, the contact electrostatic potential is more positive than the induced electrostatic potential and the difference between them at almost all positions is positive. In Figure 9a with low gas velocities, it can be divided into two regions. In the upper part of the riser, the effect of contact charging is larger since the solids concentration is low and then the contact electrostatic potential is larger than induced one; the difference between them tends to be positive just the same as those with high gas velocities. But in the lower part of the riser, the contact electrostatic potential is more negative than the induced electrostatic potential since the charge transfer is
transfer will happen between the electrostatic probe and charged particles and lead to a negative electrostatic potential. (d) When the polypropylene particles contact with the steel probe at a high speed, contact charging will happen between particles and the probe,4 and the probe always loses electrons generating a positive electrostatic potential since the work function of the steel is lower compared with polypropylene. Based on these four sources, only when source c dominates the whole process, the measured contact electrostatic potential is directly related to the charge density on particles, namely the real electrostatic level in the bed. In dense-phase fluidized bed, the particle concentration is very high but the particle speed is low; thus sources a, b, and d are negligible and the measured contact electrostatic potential agrees well with the measured charge-to-mass ratio. Wang et al.6 used the contact electrostatic probe to detect the electrostatic potential of the particles in the dense-phase fluidized bed, and found that the contact electrostatic potential appeared in a negative “Z” type distribution, which corresponded to the particle charge-tomass ratio detected by other researchers using Faraday cages. At the same time, Dong et al.10 also used the contact electrostatic probe to detect the electrostatic potential distribution and the charge-to-mass ratio distribution of the particles in the dense-phase polypropylene fluidized bed. It was found that the polarities also have a good correspondence. However, in the riser of the circulating fluidized bed, since particles are fluidized with high speeds which enhances the effects of sources a, b and d, the measured electrostatic potential even has the opposite polarity compared with the charge-to-mass ratio. In addition, the source analysis of electrostatic potential signals can also be used to explain the difference between contact and induced electrostatic potentials shown in Figures 5 and 6. For the induced electrostatic probe, since the entire probe is wrapped by Teflon, the detected signals are mainly composed of source a and source b. Chen et al.28 used the induction probe to characterize the movement of bubbles in the fluidized bed; it was found that the signals were mainly from the induction of charged particles at that time. However, in the following investigation, Fotovat et al.22 noticed there was a strong charge accumulated on the column wall when they used the same method; the charge on the wall significantly influenced electrostatic signals. For the contact electrostatic 12309
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research
concentration tends to be more negative under the stronger effect of electrostatic transfer, while in the region with lower solids concentration it tends to be more positive under the stronger effect of contact charging. In addition, since only the contact electrostatic potential generated by charge transfer has a one-to-one correspondence with the real charge of particles among these four sources, the contact electrostatic potential measurement is only available in the dense-phase gas−solid system.
stronger when the solid concentration is larger; therefore, the difference at any position tends to be more negative. It is worth saying that in Figure 9a there are few points near the outlet and inlet of the riser which disagree with the above analysis. This may be caused by the fact that the flow pattern near the outlet and inlet of the riser is not fully developed. It is worth saying that the riser is metallic in the industry; thus the electrostatic signals induced by the wall should be ignored in the above analysis. But from the above analysis, it can be found that the electrostatic signals induced by the wall have the same effects on the contact signals and induced signals; thus it does not affect the above conclusion about the difference between the contact signals and induced signals.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.9b00592.
4. CONCLUSION In this paper, two different flow patterns were created in the circulating fluidized bed made of Plexiglas for polypropylene by adjusting the gas velocity, and the axial distributions of the charge-to-mass ratio, the contact electrostatic potential, and the induced electrostatic potential in the riser were measured and compared to select a useful method for the electrostatic measurement in the real industrial process. Furthermore, aiming at explaining the differences between various methods, the measuring mechanism of electrostatic potential signals in the dilute gas−solid fluidization system was analyzed. The detailed results are as follows: 1. In the circulating fluidized bed reactor for polypropylene, the main cause of static electricity is the particle−wall friction, so the average charge-to-mass ratio of particles in the riser is negative at any position and any gas velocity examined herein. Furthermore, the distribution of the particle charge-to-mass ratio in the riser has a strong dependence on the flow pattern. In the dense-phase pneumatic transport flow pattern with high gas velocities, the average charge-to-mass ratio of particles is almost constant with a weaker negative charge along the axial direction. In fast fluidization with lower gas velocities, the particle charge-to-mass ratio in the upper part of the riser almost remains the same, while in the bottom dense-phase region particles are strongly negative and increase with the bed height. 2. In the riser of the circulating fluidized bed reactor, the axial distributions of the contact and induced electrostatic potential show strong dependence on the flow patterns. In the fast fluidization region with low gas velocities, the electrostatic potential in the bottom dense phase is negative, and gradually becomes positive with the increase of axial height. In the pneumatic transport region with high gas velocities, the electrostatic potentials are all positive, and they do not change much along the axial height. Comparing the axial contact/ induced electrostatic potential distribution with the particle charge-to-mass ratio distribution, it can be found that they have corresponding dependence on the flow pattern, but the electrostatic polarity is totally different. 3. In a circulating fluidized bed, the source of the electrostatic potential signals consists of four parts: the wall induction, particle induction, particle−probe charge transfer, and particle−probe contact charging. Among them, the induced electrostatic potential is only affected by the former two sources, but the contact electrostatic potential is affected by all of them. Therefore, the difference between the contact electrostatic potential and the induced electrostatic potential mainly comes from the latter two sources. The contact electrostatic potential in the region with higher solids
■
Electrostatic change trend of the opening process of SINOPEC polypropylene production line in October 2012 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Yao Yang: 0000-0003-3611-2859 Jingdai Wang: 0000-0001-8594-4286 Yongrong Yang: 0000-0002-5598-6925 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors acknowledge the support and encouragement of the National Natural Science Foundation of China (21808197), The National Science Fund for Distinguished Young (21525627), and the Science Fund for Creative Research Groups of National Natural Science Foundation of China (61621002).
■
REFERENCES
(1) Covezzi, M.; Mei, G. The multizone circulating reactor technology. Chem. Eng. Sci. 2001, 56 (13), 4059. (2) Cross, J. A. Electrostatics: Principles, Problems and Applications; IOP Publishing: Bristol, U.K., 1987. (3) Hendrickson, G. Electrostatics and gas phase fluidized bed polymerization reactor wall sheeting. Chem. Eng. Sci. 2006, 61 (4), 1041. (4) Matsusaka, S.; Maruyama, H.; Matsuyama, T.; Ghadiri, M. Triboelectric charging of powders: A review. Chem. Eng. Sci. 2010, 65 (22), 5781. (5) Yang, Y.; Ge, S. Y.; Zhou, Y. F.; Sun, J. Y.; Huang, Z. L.; Wang, J. D.; Lungu, M.; Liao, Z. W.; Jiang, B. B.; Yang, Y. R. Effects of DC electric fields on meso-scale structures in electrostatic gas-solid fluidized beds. Chem. Eng. J. 2018, 332, 293. (6) Wang, F.; Xu, Y.; Yu, H.; Wang, J.; Yang, Y. Electrostatic potential distribution in gas-solid fluidized beds and measurement of bed level. J. Chem. Ind. Eng. (China) 2008, 59 (3), 574. (7) Chen, A.; Bi, H. T.; Grace, J. R. Charge distribution around a rising bubble in a two-dimensional fluidized bed by signal reconstruction. Powder Technol. 2007, 177 (3), 113. (8) Park, A. H. A.; Bi, H. T. T.; Grace, J. R.; Chen, A. H. Modeling charge transfer and induction in gas-solid fluidized beds. J. Electrost. 2002, 55 (2), 135. (9) Dong, K.; Zhang, Q.; Huang, Z.; Liao, Z.; Wang, J.; Yang, Y. Experimental investigation of electrostatic effect on bubble behaviors in gas-solid fluidized bed. AIChE J. 2015, 61 (4), 1160. 12310
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311
Article
Industrial & Engineering Chemistry Research (10) Dong, K.; Zhang, Q.; Huang, Z.; Liao, Z.; Wang, J.; Yang, Y.; Wang, F. Experimental investigation of electrostatic effect on particle motions in gas-solid fluidized beds. AIChE J. 2015, 61 (11), 3628. (11) Hassani, M. A.; Zarghami, R.; Norouzi, H. R.; Mostoufi, N. Numerical investigation of effect of electrostatic forces on the hydrodynamics of gas-solid fluidized beds. Powder Technol. 2013, 246, 16. (12) Wang, H. T.; Hernandez-Jimenez, F.; Lungu, M.; Huang, Z. L.; Yang, Y.; Wang, J. D.; Yang, Y. R. Critical comparison of electrostatic effects on hydrodynamics and heat transfer in a bubbling fluidized bed with a central jet. Chem. Eng. Sci. 2018, 191, 156. (13) Baron, T.; Briens, C. L.; Bergougnou, M. A.; Hazlett, J. D. Electrostatic effects on entrainment from a fluidized bed. Powder Technol. 1987, 53 (1), 55. (14) Li, J.; Kato, K. Effect of electrostatic and capillary forces on the elutriation of fine particles from a fluidized bed. Adv. Powder Technol. 2001, 12 (2), 187. (15) Rokkam, R. G.; Fox, R. O.; Muhle, M. E. Computational fluid dynamics and electrostatic modeling of polymerization fluidized-bed reactors. Powder Technol. 2010, 203 (2), 109. (16) Yang, Y.; Zi, C.; Huang, Z. L.; Wang, J. D.; Lungu, M.; Liao, Z. W.; Yang, Y. R.; Su, H. Y. CFD-DEM investigation of particle elutriation with electrostatic effects in gas-solid fluidized beds. Powder Technol. 2017, 308, 422. (17) Cheng, Y.; Lim, E. W. C.; Wang, C.-H.; Guan, G.; Fushimi, C.; Ishizuka, M.; Tsutsumi, A. Electrostatic characteristics in a large-scale triple-bed circulating fluidized bed system for coal gasification. Chem. Eng. Sci. 2012, 75, 435. (18) Joseph, S.; Klinzing, G. E. Vertical gassolid transition flow with electrostatics. Powder Technol. 1983, 36 (1), 79. (19) Zhang, Y. F.; Yang, Y.; Arastoopour, H. Electrostatic effect on the flow behavior of a dilute gas/cohesive particle flow system. AIChE J. 1996, 42 (6), 1590. (20) Yao, J.; Zhang, Y.; Wang, C.; Matsusaka, S.; Masuda, H. Electrostatics of the Granular Flow in a Pneumatic Conveying System. Ind. Eng. Chem. Res. 2004, 43 (22), 7181. (21) Ciborowski, J.; Wlodarski, A. On electrostatic effects in fluidized beds. Chem. Eng. Sci. 1962, 17 (1), 23. (22) Fotovat, F.; Bi, X. T.; Grace, J. R. A perspective on electrostatics in gas-solid fluidized beds: Challenges and future research needs. Powder Technol. 2018, 329, 65. (23) McCarty, L. S.; Whitesides, G. M. Electrostatic charging due to separation of ions at interfaces: Contact electrification of ionic electrets. Angew. Chem., Int. Ed. 2008, 47 (12), 2188. (24) Jin, Y. Fluidization Engineering Principles; Tsinghua University Press: 2002. (25) Yang, S.; Yang, H. R.; Zhang, H.; Li, J. J.; Yue, G. X. Impact of operating conditions on the performance of the external loop in a CFB reactor. Chem. Eng. Process. 2009, 48 (4), 921. (26) Seleghim, P., Jr.; Milioli, F. E. Improving the determination of bubble size histograms by wavelet de-noising techniques. Powder Technol. 2001, 115 (2), 114. (27) Lim, E. W. C. Mixing Behaviors of Granular Materials in Gas Fluidized Beds with Electrostatic Effects. Ind. Eng. Chem. Res. 2013, 52 (45), 15863. (28) Chen, A. H.; Bi, H. T.; Grace, J. R. Effects of probe numbers and arrangement on the measurement of charge distributions around a rising bubble in a two-dimensional fluidized bed. Chem. Eng. Sci. 2006, 61 (19), 6499. (29) Ahuja, S. K. COLLISION MODEL OF CHARGEEXCHANGE BETWEEN METAL AND POLYMER SPHERES. J. Phys. D: Appl. Phys. 1976, 9 (9), 1305.
12311
DOI: 10.1021/acs.iecr.9b00592 Ind. Eng. Chem. Res. 2019, 58, 12301−12311