Moisture Distribution of Heterogeneous Agglomerates in Fluidized Bed

Zhou Yunlong; Yang Ning; Miao Yanan

(College of Energy Resource and Power Engineering, Northeast Dianli University, Jilin 132012)

Moisture Distribution of Heterogeneous Agglomerates in Fluidized Bed

Zhou Yunlong; Yang Ning; Miao Yanan

(College of Energy Resource and Power Engineering, Northeast Dianli University, Jilin 132012)

Agglomerates formed in the fluidized bed were studied in this paper using the TEB atomization nozzle. The multi-sieving method was adopted to distinguish the size of original particles, nucleation agglomerates, coherence agglomerates, and paste agglomerates in order to successfully identify the different growth stages and select the region for coexistence of most stable heterogeneous agglomerates as the research object. A multi-channel conductance electrical circuit experimental device was developed in this study to measure the conductance signal, which was found to have a liner relationship with the moisture content inside the fluidized bed. By adjusting the sieve mesh openings to achieve the layered isolation of heterogeneous agglomerates, the conductance signal recovered slowly as a result of the agglomerates’ fracture during the continuous fluidization process, so that particles and agglomerates moisture distribution measurements could be implemented. The device was used to measure the particles and agglomerates moisture distribution state in the heterogeneous coexistence region, when they were injected with liquid mass within the range of wi=2.8 kg to 4.4 kg. The results indicated that with the increase of liquid mass flow, the moisture content of coherence agglomerates also increased, but the moisture content of nucleation agglomerates was decreased, and that of the original particles was maintained at a relatively low level. When the experimental injection amount reached 4.4 kg, the moisture contained in coherence agglomerates could amount to 87.3%, accounting for a big percentage of moisture in the fluidized bed.

agglomerates; conductance signal; fracture; moisture content

1 Introduction

In industrial applications of fluidized bed catalytic coking and fluid catalytic cracking, which operate at about 550 ℃, bitumen is spray-atomized with steam and injected into a fluidized bed full of hot coke particles providing the heat required to crack heavy hydrocarbons into lighter products[1-4]. An imperfect liquid distribution will result in the formation of agglomerates that hinder heat transfer to the reacting liquid, thereby slowing down the cracking reactions[5-9]. Therefore research on the agglomerates’ moisture distribution in the fluidized bed has great significance for understanding the mechanism of agglomeration and controlling agglomerate formation.

In recent years, the reaction mechanism of agglomerates and liquid in fluidized beds has been extensively studied. Bruhns[10]suggested a mechanism of liquid injection into a fluidized bed by spraying ethanol that manufactures the agglomerates. The drying characteristics of the porous and non-porous agglomerates are compared, and they have found out that for porous agglomerates, most of the liquid is trapped inside the agglomerates, which can explain why drying needs a longer period, while for non-porous agglomerates the evaporation occurs from the outer surface of the agglomerates, making the liquid evaporate faster. Korina Terrazas-Velarde[11]analyzed the spherical droplet penetration process on porous agglomerates, and the results showed that the growth rate for porous agglomerates was significantly reduced due to the losses of deposited droplets into the pores. Mills[12]found out that the viscosity of the liquid affected both the rate of size enlargement and the mechanism of size enlargement. The size growth rate increased with the increase in liquid viscosity and occurred through the layering mechanism. M. Mohamad Saad[13]added different water levels to assess the changes in the size distribution and morphological parameters of the agglomerates, and the relationbetween the diameter and the solid volume fraction of the agglomerates could be modeled by a power law equation. Thomas Abberger[14]applied three spray droplet sizes, viz. 30 μm, 60 μm, and 90 μm, respectively, to evaluate the effects of the droplet size on the formation and growth of agglomerates. The results showed that the 90 μm droplets offered a potential for larger agglomerates than the 30 μm and 60 μm droplets did. Although the structure and formation mechanism of the agglomeration had been thoroughly understood, a scientific and valid method that can detect the moisture distribution does not exist. In order to develop a valid and reliable method for the determination of the moisture distribution in a fluidized bed, fluidized bed agglomerates were manufactured by TEB atomization nozzle, in order to use the multi-sieving method to identify different growth stages. The most stable heterogeneous agglomerates coexistence region was selected as the research object, while a multi-channel conductance electrical circuit experimental device was installed in the fluidized bed, and the sieve mesh openings were adjusted to achieve the layered isolation of heterogeneous agglomerates, with the particles and agglomerates moisture distribution detected based on the proportional relationship between the moisture contained in the agglomerates and the conductance signal studied in this paper.

2 Experimental

2.1 Agglomerates manufacturing process flow

The flow diagram of the experiment unit is presented in Figure 1. The experiments were performed in a Plexiglas fluidized bed column with a rectangular cross section of 0.2 m in length and 0.05 m in width. In order to optimize the fluidization quality, a rectangular porous plate distributor was installed at the bottom of the fluidized bed. A total of 24 hemispherical wind caps, 1 cm in diameter with 8 holes (1 mm in diameter of each hole) in each cap, were placed in 3 rows and were uniformly arranged on the plate distributor. When the air supplied by the air compressor after passing through a valve and a vortex flow meter entered the fluidized bed, the vortex flow meter located at the entrance of fluidized bed could monitor the air flowrate. The air flowrate in this experiment was set at 1 m/s.

Figure 1 Flow diagram of the experimental setup1—Air compressor; 2—Ball valve; 3—Vortex flowmeter; 4—Water tank; 5—Fluidized bed; 6—TEB atomization nozzle; 7— Conductance electrical circuit; 8—Data acquisition; 9—Computer

The liquid was injected into the fluidized bed through the TEB atomization nozzle, which was developed by T. E. Base[15], with its structure presented in Figure 2. The unit entrance was equipped with liquid and air pipelines, and the air-liquid ratio (ALR) was adjusted by changing the air and liquid mass flow rate. The air as the power for liquid transmission was maintained at a rate of 1.2 g/s during the experiment. The measurement indicated that the liquid was injected into the fluidized bed at a mass flow rate of 40 g/s.

The TEB atomization nozzles were 600 mm apart from the porous plate distributor. In a cold fluidized bed, sugar solution and sand were used to simulate the spraying of hot bitumen on hot coke particles in industrial fluid cokers. McDougall[16]has shown that the sugar solution-sand system is a good room temperature simulator for the hot bitumen-coke system, as there is nearly perfect wettability of the solids by the liquid in both cases. The system’s physical properties are presented in Table 1. A solution containing 8% sugar was used to match the viscosity of bitumen at the injection temperature used in the industry[17]. Meanwhile, the sugar dissolved in the solution functioned as a liquid bridge to stabilize the agglomerates during the contact between solution and sand.

Figure 2 The structure of the TEB atomization nozzle

Table 1 The medium's physical properties

2.2 Regional identification of agglomerates

The sand was mixed for 2 min without liquid addition in order to homogenize the particles. This operation was followed by a wetting/mixing stage. During this stage, an injection time t = 40 s, 50 s, 60 s, 70 s, 80 s, 90 s, 100 s, 110 s, 120 s, 130 s, 140 s, and 150 s, respectively, was adjusted to simulate the different growth cases, and the injected liquid mass was equal to wi=1.6 kg, 2 kg, 2.4 kg, 2.8 kg, 3.2 kg, 3.6 kg, 4 kg, 4.4 kg, 4.8 kg, 5.2 kg, 5.6 kg, and 6 kg, respectively. The subscript i refers to these 12 experimental injections.

Different kinds of agglomerates were produced after all the liquid was injected into the fluidized bed by the nozzle[18]. Samples of 500 g of sand were sieved by 20 different sieves with mesh openings expressed in Dj=1 mm, 1.05 mm, 1.1 mm, 1.15mm, 1.2 mm, 1.5 mm, 2 mm, 2.5 mm, 3 mm, 3.5 mm, 4 mm, 4.5 mm, 5 mm, 5.5 mm, 6 mm, 6.5 mm, 7 mm, 7.5 mm, 8 mm, and 8.5mm, respectively. The subscript j refers to these 20 mesh openings, suggesting that the agglomerates were similar to the material with spherical structure. The diameter was adopted as the feature size, and the size of remainder particles on each sieve j was equivalent to the mesh openings Dj.

With respect to the injecting liquid, its mass is wi, and the remainder particles on the sieve mesh openings Dijare weighed. The weight distribution, according to agglomerates size, is then expressed as a percentage of the total weight. The curves of cumulative agglomerates size Dij, obtained for these i experimental injections, is shown in Figure 3. The cumulative agglomerates size is mainly clustered in 4 regions, as shown below. (I) In the range of wi=1.6 kg to 2.0 kg, the size of cluster is between 1.0 mm ad 1.5 mm, which almost denotes no increase compared with the original particles, while the growth does not occur. (II) In the range of wi=2.4 kg to 3.6 kg, the size of cluster ranges from 1.5 mm to 3 mm, which is slightly increased compared with the original particles, and the structure of adhesive growth is not significant and is called nucleation agglomerates. Nucleation agglomerates form the basic element for an induced large-scale agglomeration. (III) In the range of wi= 4.0 kg to 5.2 kg, the size of cluster is between 3 mm and 5.5 mm, which exhibits an apparent growth. The agglomerates in this stage belong to the coherence agglomerates. (IV) When the liquid mass runs up to wi= 5.6 kg to 6.0 kg, the size of agglomerates shows a rapid growth. The agglomerates in this stage, which are called the paste agglomerates, display the sticky paste characteristics and do not have the ability to resist the external stress.

Figure 3 Cumulative agglomerates size curves obtained for each experimental injections■—wi=1.6;●—wi=2.0;▲—wi=2.4;▼—wi=2.8; —wi=3.2;—wi=3.6;◆—wi=4.0; —wi=4.4; —wi=4.8;★—wi=5.2;—wi=5.6; ※—wi=6.0

Figure 4 Evolution of agglomerates mass fraction according to each experimental injection—original particles; ●—nucleation agglomerates;◆—coherence agglomerates; —paste agglomerates

The evolution of 4 specific types of heterogeneous agglomerates mass fraction according to these experimentalinjections wiis shown in Figure 4. In the range of wi=1.6 kg to 2.8 kg, the amount of original particles continuously decreased, while the corresponding nucleation agglomerates increased gradually. This instance suggests that the nucleation agglomerates are caused by adhesion of original particles under the effect of sugar solution. When the injection liquid mass is greater than wi=2.8 kg, the coherence agglomerates start to emerge in the fluidized bed as a result of nucleation agglomerates association. With the increase in liquid mass, the fraction of nucleation agglomerates is decreased to 7.6%. As the liquid mass increases to wi=4.4 kg, the paste agglomerates begin to emerge, which is accompanied by a sharp decrease of nucleation and coherence agglomerates.

The results indicate that the liquid mass can be considered as a critical value to distinguish the different agglomeration region. The structure in heterogeneous agglomerates coexistence region, which is comprised of original particles, nucleation agglomerates, and coherence agglomerates, has the best stability[19]. Selecting the region as the research object by maintaining liquid mass in the region of wi=2.8 kg to 4.4 kg is favorable to the research on moisture distribution.

2.3 Principle for measurement of experimental device

A multi-channel conductance electrical circuit experimental device was developed to measure the moisture contained in heterogeneous types of agglomerates. The device is shown in Figure 5, where the conductance probes, composed of hollow and stainless steel tubes with an outer diameter of 10 mm, are located at a height of 200 mm, 400 mm, 600 mm, 800 mm, and 1 000 mm, respectively, above the porous plate distributor. These probes penetrate into the fluidized bed with a length of 200 mm. The probes, with one end of which connected to a signal generator that provides a sinusoidal, 100 Hz current at a constant voltage of 6.7 V, are connected in series with a small 100 kΩ resistor (Rm) and a common electrode, which is grounded, to form an electrical circuit. The circuit diagram of the experimental device is shown in Figure 6.

Figure 5 Multi-channel conductance electrical circuit experimental device

Figure 6 The circuit diagram of experimental device

The following equation shows how the conductance can be calculated by applying the Kirchhoff’s current law.

Then Equation (1) can be rearranged in the following order.

The inverse of this expression is defined as the local bed conductance.

The local bed conductance Rbedcould then be measured from the two voltages V1and Vm, where V1is the voltage provided by the signal generator which is known (V1in this experiment is 6.7 V). The local conductance can be measured from the voltage Vmacross the resistor Rm, which can be acquired using a data acquisition system.

Before utilizing the multi-channel conductance electrical circuit experimental device to measure the moisture contained in the agglomerates, the relationship between the conductance signal and the injection liquid mass should be obtained. However, the collision of fluidized sand particles will contribute to the porous agglomerates that trap inside a portion of sugar solution, then the unfeigned content of sugar solution will decrease. In order to preventthe sand from forming agglomerates in the fluidized bed, the effect of injection liquid mass on the conductance signal is described under the packed bed condition. Since the original particles mainly exist in the fluidized bed under that condition, the moisture is spread on the surface of the original particles that do not have the ability to trap inside the solution, so that the original particles in the fluidized bed are completely suffused with moisture. As a case of measurement by the conductance probe located at the height of 600 mm as illustrated in Figure 7, the conductance signal value is in the vicinity of 1×10-7S prior to injection of the liquid. As the injected liquid spreads through the bed, it provides additional pathways for electrical current to increase the bed conductance signal. The signal tends to stabilize when injection comes to an end, as shown in Figure 7. Conductance signal shows a liner relationship with moisture content inside the fluidized bed. Therefore, adopting the conductance signal to reflect the moisture distribution in fluidized bed is feasible, and the conductance signal demonstrates a perfect liner distribution with the increase in experimental injections wi.

To ensure that no adhesion of quartz sand was present before the experiment, the liquid mass, which was kept in the range of wi=2.8 kg to 4.4 kg, was injected into a fluidized bed through the nozzle. Meanwhile, upon activating the air compressor, the injection liquid could promote the particles to form agglomerates in a fluidized state. The original particles, nucleation agglomerates, and coherence agglomerates coexisted in the fluidized bed, when the injection came to an end. The air flow was shut down until heterogeneous agglomerates were stably formed in the fluidized bed, and, when the height of packed bed reached 270 mm, the sieve with mesh openings of 3 mm in diameter was installed at a height of 300 mm above the porous plate distributor, as presented in Figure 5.

Figure 7 The effect of injection liquid mass on the conductance signal■—wi=2.8;●—wi=3.2;▲—wi=3.6;▼—wi=4.0; —wi=4.4

3 Results and Discussion

3.1 Experiment results of agglomerates fracture

A case relating to the curves of conductance signal changing with the time at an experimental injection wi=3.6 kg is presented in Figure 8 (a) and (b). During the liquid injection into the fluidized bed, a rapid increase in the conductance signal due to the increased concentration of electrically conductive liquid was detected. In the subsequent injection period covering 60 s to 90 s, the agglomerates gradually emerged along with adsorption of the conductive liquid, which resulted in a decrease of the conductance signal increase rate. When the injection came to an end, the heterogeneous agglomerates were completelyformed, which caused the signal to drop quickly. Then, the signal recovered slowly as the moisture gradually diffused from fractured agglomerates caused by the collision of particles and agglomerates. The conclusion that can be drawn from these results would confirm that the conductance signal did not stabilize until the agglomerates were completely fractured.

Figure 8 Curves of bed conductance signal changing with time

Upon comparing the measurement results illustrated in Figure 8 (a) & (b), the recovery of the conductance signal values, which were measured by the conductance probe 1 and probes 2 to 5, show a palpable difference. This phenomenon is attributed to the sieve installed at the top of the region housing probe 1. The agglomerates were stably formed when injection came to an end, with the height of packed bed equating to 270 mm, and then the sieve was installed at a height of 300 mm. Therefore all the agglomerates were below the sieve, and the mesh openings of the sieve, 3 mm in diameter, could ensures that the coherence of the agglomerates with the cluster size ranging from 3 mm to 5.5 mm was completely insulated in the region with probe 1. Yet, the original particles and nucleation agglomerates were uniformly distributed in the region containing probes 1 to 5 to achieving the layered isolation of heterogeneous agglomerates, that is, the recovery of the conductance signal in the region with probe 1 was promoted by fracture of nucleation agglomerates and coherence agglomerates, while the recovery of the conductance signal in the region including probes 2 to 5 was merely promoted by the fracture of nucleation agglomerates. As the moisture content is proportional to the conductance signal, the proportion A of the coherence agglomerates among the original particles, nucleation agglomerates, and coherence agglomerates mixture is then expressed by Equation (4).

where C2is the average value of recovery of the conductance signal, which is proportional to the moisture content of original particles and nucleation agglomerates in the region containing probes 2 to 5; C3is the average value of recovery of the conductance signal, which is proportional to the moisture content of original particles, nucleation agglomerates, and coherence agglomerates in the region housing the probe 1.

The proportion B of the nucleation agglomerates among original particles and nucleation agglomerates is expressed by Equation (5).

where C1is the average value of the previous recovery of the conductance signal, which is proportional to the moisture content of original particles in the region housing probes 2 to 5, as fractured agglomerates would not occur. At an injected liquid mass of wi, the moisture content of coherence agglomerates can be calculated by the following equation.

The moisture content of nucleation agglomerates can be calculated according to Equation (7):

The moisture content of original particles is shown in Equation (8):

3.2 Experimental results on moisture content of agglomerates

According to the above results, the moisture distribution of heterogeneous agglomerates in fluidized bed was measured at experimental injections wi=2.8 kg, 3.2 kg, 3.6 kg, 4 kg, and 4.4 kg, respectively. The measurement results are presented in Table 2. With the increase in the liquid mass flow, the coherence agglomerates’ moisture content (wC) increased, while the nucleation agglomerates’ moisture content (wN) decreased, and the original particles’ moisture content (wY) was maintained at a relatively low level. This phenomenon is illustrated in Figure 4. The nucleation agglomerates were associated with each other to form the coherence agglomerates, resulting in an increased proportion of coherence agglomerates in the fluidized bed. The results suggested that the coherence agglomerates absorbed the ambient conductive liquid along with obtaining the moisture contained in nucleation agglomerates. On the other hand, the original particles were not involved in this process, so the moisture content tended towards stabilization at a relatively low level. Table 2 shows that the moisture content of coherence agglomerates reached 87.3% when the experimental injec-tion was wi=4.4 kg. The case corresponding to the limit of the heterogeneous agglomerates coexistence region indicated that the coherence agglomerates occupied a majority of agglomerates.

Table 2 Results on measurement of agglomerates moisture content

Table 3 The results on measurement of moisture content in agglomerates upon drying at a constant temperature of 105℃

The 105 ℃ constant temperature drying method is used to verify the above-mentioned measurement results, when heterogeneous agglomerates were stably formed in the fluidized bed as the injection time came to an end with the air compressor being shut down, and the original agglomerates, nucleation agglomerates, and coherence agglomerates were separated and placed in the oven at 105 ℃ for 24 h until the weight did not change further, and the weight difference before and after drying was the moisture content of the agglomerates. As shown in Tables 2 and 3, the results are almost the same when the 105 ℃constant temperature drying method was compared with the method proposed in this paper, which confirmed the validity of the multi-channel recovery of conductance signal measurement method.

4 Conclusions

(1) Different types of agglomerates could be obtained by maintaining the experimental injections wiin different regions. The range of wi=2.8 kg to 4.4 kg of the injected liquid mass corresponded to the region for existence of heterogeneous agglomerates, comprising the original particles, the nucleation agglomerates, and the coherence agglomerates. As the liquid mass flow increased to wi=4.4 kg, the paste agglomerates, which displayed sticky paste characteristics, would emerge.

(2) By adjusting the mesh openings of the sieves installed at the top of the region housing the probe 1 to achieve the layered isolation of heterogeneous agglomerates, the coherence agglomerates were completely insulated in the region containing probe 1, yet the original particles and nucleation agglomerates were uniformly distributed in the region including probes 1 to 5. The conductance signal had a perfect linear relationship with the moisture content inside the fluidized bed. The recovery of the conductance signal was promoted by the moisture that slowly diffused from the fractured agglomerates.

(3) Nucleation agglomerates were associated with each other to form coherence agglomerates with an increasing liquid mass flow. In this region, the moisture content of the coherence agglomerates was increased, and that of the nucleation agglomerates was decreased. The moisture content of original particles was maintained at a relatively low level. The moisture content of the coherence agglomerates reached 87.3% when the experimental injection was wi=4.4 kg. In this case, the coherence agglomerates occupied a majority of the agglomerates.

Acknowledgements:This work was financially supported by the National Natural Science Foundation of China (No.51276033).

[1] Boduszynski M M, Grudoski D A, Rechesteiner C E, et al. Deep-cut assay reveals additional yield of high-value VGO[J]. Oil and Gas Journal, 1995, 93(37): 39-45

[2] Chung H K, Xu C M, Hu Y X. Supercritical fluid extraction reveals resid properties[J]. Oil and Gas Journal, 1997, 95(3): 66-69

[3] Zhang X B. Study on Deep Catalytic Cracking of FCC Gasoline[D]. Dongying: China University of Petroleum, 2007: 1-23

[4] Li Q, Wang L, Zhang Q H, et al. Heavy oil deep catalytic cracking process in downer reactor to produce light olefins[J]. CIESC Journal, 2004, 55(7): 1103-1108

[5] Dougall S M, Saberian M, Briens C. Using dynamic pressure signals to assess the effects of injected liquid on fluidized bed properties[J]. Chemical Engineering and Processing: Process Intensification, 2005, 44(7): 701-708

[6] Jennifer M, Cedric B, Franco B. Flow stability in a downer mixing chamber[J]. Chemical Engineering and Processing: Process Intensification, 2006, 45(9): 798-805

[7] Maryam M, Majid H, Cedric B. The effects of liquid properties and bed hydrodynamics on the distribution of liquid on solid fluidized particles in a cold-model fluidized bed[J]. Powder Technology, 2014, 256: 5-12

[8] Stefan H, Jan B, Markus H, et al. Study of dynamic multidimensional temperature and concentration distribution in liquid sprayed fluidized beds[J]. Chemical Engineering Science, 2003, 23(58): 5135-5160

[9] Flora P, Cedric B, Franco B, et al. Effect of particle size, liquid content and location on the stability of agglomerates in a fluidized bed[J]. Powder Technology, 2013, 237: 376-385

[10] Stefan B, Joachim W. An investigation of the mechanism of liquid injection into fluidized beds[J]. AIChE Journal, 2005, 51(3): 766-775

[11] Korina T, Mirko P, Evangelos T. Kinetics of fluidized bed spray agglomeration for compact and porous particles[J]. Chemical Engineering Science, 2011, 66(9): 1866-1878

[12] Mills P J T, Seville J P K, Knight P C K, et al. The effect of binder viscosity on particle agglomeration in a low shear mixer/agglomerator[J]. Powder Technology, 2000, 113(1): 140-147

[13] Saad M M, Barkouti A, Rondet E, et al. Study of agglomeration mechanisms of food powders: Application to durum wheat semolina[J]. Powder Technology, 2011, 208(2): 399-408

[14] Thomas A, Anette S, Torben S. The effect of droplet size and powder particle size on the mechanisms of nucleation and growth in the fluid bed melt agglomeration[J]. International Journal of Pharmaceutics, 2002, 249(1): 185-197

[15] Terence E B, Edward W C, Douglas A E, et al. Nozzle for atomizing liquid in two phase flow: The United States, US6003789 [P]. 1999-12-21

[16] Steven L M, Mohammad S, Cedric B, et al. Characterization of fluidization quality in fluidized beds of wet particles[J]. International Journal of Chemical Reactor Engineering, 2004, 2(1): 1542-6580

[17] House P, Briens C, Berruti F, et al. Effect of spray nozzle design on liquid-solid contact in fluidized bed[J]. Powder Technology, 2008, 186(1): 89-98

[18] Barlouti A, Rondet E, Delalonde M, et al. Influence of physicochemical binder properties on agglomeration of wheat powder in couscous grain[J]. Journal of Food Engineering, 2012, 111(2): 234-240

Mechanical Completion of PO/MTBE Unit at Wanhua Chemical Group at Yantai

The construction of the PO/MTBE combination unit, which comprises a part of the PO/AE integrated project, has been completed mechanically at the Wanhua Chemical Group Company in Yantai and currently the process route of all facilities has been tested. The PO/AE integrated project mainly covers a 750 kt/a unit for manufacture of propylene from propane dehydrogenation, a 240 kt/a PO unit, a 750 kt/a MTBE unit, a 225 kt/a butanol unit, a 300 kt/a unit for producing acrylic acid and acrylic esters, and a 300 kt/a polyether polyol unit. The PO/MTBE unit will adopt the technology for cooxidation of isobutane with propylene licensed by the Huntsman Corporation.

It is told that the Huntsman technology has been licensed to the Wanhua Chemical Group in Yantai and the Jinling Hunstman New Materials Company in Nanjing. The construction of PO/MTBE unit has been mechanically completed and handed over to the Wanhua Group, while the pile foundation works will be kicked off in June 2015 at the Jinling Hunstman New Materials Company. Furthermore, the Lyondell Company intends to set up a joint venture in Ningbo engaging in the construction of a large PO/TBA production unit. According to statistical data, there are globally four existing units for producing PO via co-oxidation of isobutane with propene, and other three units are under construction or are envisaged in the plan. Lyondell has constructed each one PO/TBA unit in the US, the Netherlands and France, respectively, whereas Hunstman owns only one unit in Texas rated at 240 kt/a of PO and 750 kt/a of MTBE.

date: 2014-08-08; Accepted date: 2014-09-01.

Prof. Yang Ning, Telephone: +86-432-64807495; E-mail: 1020219438@qq.com.