Chinese Petroleum Corp.
Chiayi, Taiwan
Fu Sou Jeng
Chinese Petroleum Corp.
Kaohsiung, Taiwan
A new method of calculating catalyst circulation rate on line improved the operating efficiency of the moving bed transalkylation unit at Chinese Petroleum Corp's refinery at Kaohsiung, Taiwan.
The improved measurements increased toluene conversion by 3 wt %/pass and reduced energy usage by about $220,000/year in the xylene production unit.
AROMATICS MARKETS
The p xylene market has grown rapidly in many parts of world, especially East Asia. Demand in Taiwan has almost doubled in the past 5 years.
Among the processes that produce mixed xylenes, toluene disproportionation and toluene/nine carbon aromatics (A9) transalkylation are the major production routes.
Optimization of transalkylation unit operation to boost p xylene production and reduce production cost was a main research strategy for Chinese Petroleum Corp. (CPC).
CPC's transalkylation unit at the Kaohsiung refinery applied moving bed reactor design. The operational know how CPC gained from this unit, including an in-house method for on line catalyst circulation rate measurement, improved transalkylation unit operation efficiency and increased p xylene production capacities.
ON-LINE MEASUREMENT
Fig. 1 shows a schematic flow diagram of the moving-bed transalkylation process. Catalyst circulates between the reactor and regenerator.
Catalyst circulation rate is an important operating variable. It affects reactor and regenerator operations and unit performance. Reliable measurement of circulation rate thus is necessary to the efficient operation of a transalkylation unit.
Conventional measurement methods apply two types of correlations. One correlation between air lift pressure (PRC 304 in Fig. 1) and circulation rate was derived using a momentum balance.1 The other correlation was linked to a metering box reading (07M in Fig. 1).
CPC's operating experience showed that the PRC correlation drifted with unit age, and the metering box reading method can be used only on an off line basis.
CPC has developed a reliable and accurate on line CCR measurement method for moving bed transalkylation units. The method is based on a heat balance calculation in the reactor.
At steady state operation, coke formation rate in the reactor generated as a side reaction of the transalkylation reaction should be equal to the combustion rate in the regenerator. The heat balance equation is given in Equation 1 (Equations).
The heat balance calculation requires conversion data, which is troublesome to measure on line. CPC concluded from sensitivity analysis that neglecting the heat of reaction caused an error in the calculation of catalyst circulation rate of only about 3%, which is believed to be acceptable from a practical viewpoint.
By eliminating the heat of-reaction term, Equation 1 can be simplified to produce Equation 2, which only involves calculation constants, such as the specific heats of the hydrocarbons and catalyst, and operating variables. (The representative constant for the specific heat of a hydrocarbon stream can be obtained by summing algebraically those of its individual pure components, which are available in thermodynamic data bases.)
SPECIFIC HEAT
The specific heat of the catalyst can be obtained by a calibration method. Catalyst particles applied in the moving bed transalkylation process usually are 1/8 in. diameter.
Equation 2 assumes an even temperature profile inside the catalyst particle, which is not true in unit operation. The alterative procedure described here is to lump all the deviation into an apparent specific heat capacity. Fig. 2 shows the algorithm used to determine the specific heat capacity of the catalyst.
Using this scheme, the specific heat of the catalyst can be back calculated from a standard catalyst circulation rate value. This value was determined by measuring the catalyst accumulation rate in the catalyst hopper when the catalyst circulation was temporarily interrupted by closing the stop valve between the regenerator and the reactor in the commercial unit at the Kaohsiung refinery.
The specific heat coefficient was back calculated by fitting the circulation rate measured from an energy balance calculation to that obtained from the metering box reading. In this case, the most appropriate specific heat coefficient was 0.17 cal/g K.
A simple working equation was then obtained (Equation 3). This calculation can be adopted into an on line method.
Specific heat coefficients are subject to operating variables, catalyst compositions, and coke content. Recalibration of the coefficients is recommended at the time of unit shut down, especially when the catalyst type is changed or repair work is performed in the reactor or regenerator system.
COMMERCIAL TEST
CPC's in house circulation rate measurement method was successfully tested in its commercial transalkylation unit at Kaohsiung.
The circulation rates measured using the method and the PRC correlation were compared against that measured by the catalyst hopper level method, which is considered standard (Fig. 3). This test showed that the previous PRC correlation had poor accuracy, while the catalyst metering box reading and our in house method were very reliable.
The metering box reading correlation, however, can be monitored only in the field and must be measured off-line. On the other hand, the in house method can be applied on line as a real time indication.
The on line capability of the new method offers unit operators a way to respond quickly to circulation rate offsets. Unit performance has improved in terms of toluene conversion and unit safety.
Toluene conversion increased after applying the on line measurement method (Fig. 4). The increase in conversion was attributed to higher catalyst to oil ratio, which is an operating variable that requires reliable catalyst circulation rate measurement.
In addition, regenerator temperature is of great concern, in terms of regenerator metallurgy limitations and unit safety. The increased accuracy of the new method thus has improved plant safety by better indicating actual unit operation.
IMPURITY EFFECTS
The transalkylation process is operated within the constraints of thermodynamic equilibrium conversion. Unconverted toluene is separated from the reaction products by distillation, and recycled to extinction.
In its transalkylation unit, CPC experienced that conversion of toluene per pass decreased as the xylene impurities in the toluene increased (Fig. 5). The presence of xylene impurities can be monitored by an increased A9 product yield.
The purity of the recycled toluene depends on the distillation efficiency. By improving distillation tower efficiency to reduce xylene impurities to 0.2% from 4%, toluene conversion per pass increased 3 wt %.
ECONOMICS
CPC adopted the on line catalyst circulation rate measurement method and operation on feed with reduced xylene impurities into its commercial unit operation. This system led to increased toluene conversion.
By increasing toluene conversion, the toluene recycle rate can be reduced. Utility cost for toluene separation, therefore, also decreases. Furthermore, when unit operation is limited by downstream distillation capacity, the new technology can be applied to debottleneck a unit.
CPC's unit experienced a reduction in utility costs as toluene conversion levels increased. An energy savings of 106/kl feed (almost 17/bbl) was achieved, totaling about $220,000/year at 620 kl/day (3,900 b/d) throughput (Table 1).
DISPROPORTIONATION
Theoretically, transalkylation produces greater xylene yield than does disproportionation.
In disproportionation, 2 moles of toluene can be converted to 1 mole of xylene. Taking trimethylbenzene, however, to represent the A9 stream, transalkylation produces 1 mole of toluene per mole of xylene.
The toluene to xylene conversion factors for disproportionation and transalkylation are, respectively, 57.6 wt % and 112.8 wt %.
Comparing the economics of the two processes depends on the current market situation, separation costs, A9 availability, and environmental concerns.
Transalkylation is a maximum xylene mode and disproportionation is a maximum benzene mode. The flexibility of reaction mode enables a transalkylation unit to be operated in accordance with market dynamics.
Fig. 6 clearly shows that blending of the A9 stream can boost xylene yield. For toluene disproportionation, xylene yield increases to 62 vol % from 58 vol % for 4 wt % A9 blending, while benzene yield decreases to 30 vol % from 35 vol %.
The A9 stream typically contains some impurities - such as n propylbenzene and indane which deactivate zeolite activity. Transalkylation of the A9 stream, on the other hand, usually forms some saturates and deteriorates benzene purity.
Zeolite catalyst deactivation rates and saturate formation therefore limit the allowable concentration of A9 in transalkylation operations. The limitation of A9 stream blending is a characteristic of the catalyst type.
A9 availability favors those refinery complexes that have an associated naphtha cracker. Current applications of the A9 stream from naphtha crackers are as either gasoline blending stocks or transalkylation feedstocks. The choice depends on the differential price between gasoline and xylene.
Environmental regulation is another key factor for process economics. Benzene could be in surplus when the U.S.'s new reformulated gasoline regulations limit benzene content. Furthermore, reducing gasoline aromatics content from the current standard to 20% in the future is under heated debate.
After enforcement of the proposed regulation, blending of A9 stream into gasoline no longer would be feasible. These environmental concerns favor transalkylation over disproportionation.
As for separation costs, transalkylation usually produces greater ethylbenzene yield than disproportionation (Fig. 6). The transalkylation mode therefore has greater separation costs than does disproportionation mode.
REFERENCES
- Chang, J. Y., Hu, H. C., Chiu, C.T., and Tsai, T.C., AIChE spring meeting, Mar. 28 Apr. 3, 1993, Houston.
- Tsai, T.C., Proceedings of ROC-Japan Joint Workshop on Catalysis, (P.Y. Chen and Y.W. Chen, Eds.), Chungli, Taiwan, Dec. 10, 1992, p. 75.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
