JPH0782203A - Controlling method for hydroformylation - Google Patents

Abstract

PURPOSE:To facilitate the control of aldehyde production ratio in a method of controlling hydroformylation reaction. CONSTITUTION:In the aldehyde production process by subjecting an olefin, the oxo gas and the recycle gas to hydroformylation reaction in the reactor, the oxo gas to be fed into the reactor is controlled in its flow rate to adjust the partial pressure of carbon monoxide in the reactor. In the operation, the so-called model prediction control method is employed where the process model is installed in the computer for process control and the data are sequentially input into the system to predict the output data and the input data having the minimum difference from the target operation are used for the actual operation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a normal form of a product obtained by controlling the partial pressure of CO in a recycle gas during a hydroformylation reaction in which an oxo gas containing hydrogen and carbon monoxide (CO) is reacted with an olefin. The present invention relates to a method for controlling the production ratio of hydrid and isoaldehyde.

[0002]

2. Description of the Related Art Conventionally, a process has been adopted in which an aldehyde containing normal aldehyde and isaldehyde is produced by a hydroformylation reaction using oxo gas and olefin as raw materials. In this process, the CO partial pressure (concentration) in the oxo gas is adjusted to control the CO partial pressure in the recycle gas, so that the CO concentration in the reactor is adjusted to produce the normal aldehyde-isoaldehyde production ratio ( Hereinafter, a method for controlling the aldehyde production ratio) to a desired value is known.

FIG. 1 shows a schematic system diagram of a general process in which a hydroformylation reaction is carried out. In the figure, the system is equipped with a hydroformylation reactor 1 and a gas-liquid separator 2. In this reactor 1, raw material and catalyst are fed from an olefin supply line 11, an oxo gas supply line 12 and a catalyst supply line 14, respectively. Is being supplied. The gas-liquid separator 2 is extracted from the reactor 1.
The gaseous recycle gas separated in 1 is returned to the reactor 1 again via the recycle gas supply line 13 and reacts with the raw material olefin to produce an aldehyde. The generated aldehyde is taken out from the aldehyde product withdrawing line 15. The purge control valve 9
Is provided to suppress the accumulation of inert gas components such as methane and operates intermittently to purge a part of the recycled gas.

The target value of the CO partial pressure in the recycled gas is
It is determined by the correlation with the desired aldehyde production ratio. In order to control the CO partial pressure in the recycle gas based on this target value, the CO partial pressure in the raw oxo gas is adjusted. Since this adjustment is possible by adjusting the amount of feed hydrogen introduced into the oxo gas through the hydrogen supply line 16, the control of the amount of feed hydrogen supplied is as follows.
For example, it is performed by a PID (proportional / integral / derivative) control device.

The production rate of aldehydes is often changed according to the demand of the product. In such a case, the target value of the CO partial pressure in the recycled gas is newly determined based on the new production ratio of aldehyde. The feed hydrogen amount is newly determined based on this target value, and the determined feed hydrogen amount is given as a set value for PID control. The target value of the CO partial pressure is usually set by an operator, and then PID control is performed based on the target value so that the feed hydrogen amount matches the set value.

[0006]

The conventional methods for controlling the hydroformylation reaction have the following problems. In the hydroformylation reaction, the process time constant is large, and the CO partial pressure in the recycled gas fluctuates greatly due to changes in the feed hydrogen amount. Therefore, when determining the set value of the feed hydrogen amount, the set value is changed in multiple stages. This is a burden on the operator. For the above reason, when changing and setting the set value of the feed hydrogen amount, there are individual differences in the operation results depending on the skill of the operator, and the process tends to become unstable when the set value is changed. Amount of aldehyde produced per unit time (production rate)
Is changed, the characteristics of the process are changed, so that the control of the CO partial pressure becomes particularly unstable. When the raw material gas contains a large amount of components other than hydrogen and carbon monoxide, for example, when the partial pressure of olefin (propylene), paraffin (propane) or other inert gas is high, it is necessary to control the amount of feed hydrogen only in the recycled gas. It is difficult to control the CO partial pressure within the desired range.

In view of the problems of the conventional method for controlling the hydroformylation reaction, the present invention makes it easy to control the CO partial pressure in the reactor, and therefore the aldehyde production ratio can be easily adjusted to a desired value. It is an object to provide a method for controlling a hydroformylation reaction.

[0008]

In order to achieve the above object, a method for controlling a hydroformylation reaction according to the present invention is a feed containing olefin, hydrogen and carbon monoxide in the presence of a catalyst in a reactor. A method for controlling a hydroformylation reaction in a method of controlling a hydroformylation reaction for producing an aldehyde by causing a hydroformylation reaction between oxo gas and a recycled gas that is withdrawn from the reactor and then returned to the reactor. A target value of the carbon monoxide partial pressure in the inside is detected, the carbon monoxide partial pressure corresponding to the target value is detected, the detected carbon monoxide partial pressure is compared with the target value, and the feed oxo It is characterized by adjusting the flow rate of gas. With this configuration,
It makes it possible to produce aldehydes in the desired production ratio.

By adopting the above configuration, it is possible to prevent the CO concentration in the reactor from fluctuating greatly when the target value of the CO partial pressure is changed. In a preferred aspect of the present invention, the detection of carbon monoxide is performed by detecting the carbon monoxide partial pressure of the recycled gas returned to the reactor. Alternatively, it is also preferable that the recycled gas is merged with the feed oxo gas before being returned to the reactor, and the carbon monoxide partial pressure of the merged gas is detected to be the detected carbon monoxide partial pressure. Is.

In a particularly preferred embodiment of the present invention, a predetermined signal transfer function is provided, which has an operation amount for adjusting the flow rate of the feed oxo gas as an input, and outputs the calculated value of the detected carbon monoxide partial pressure, A process model for simulating the hydroformylation reaction is stored in a computer, and an input series consisting of a plurality of inputs sequentially generated at a predetermined cycle is sequentially selected as an input of the process model and an output of the process model. Sequentially calculate the future output series based on, among the calculated future output series,
One output series having a small difference from the expected future target value is selected, and the manipulated variable is set based on the input series corresponding to the selected one output series.

In the above particularly preferred embodiment, the selection of the output sequence is performed according to, for example, the least square method.

According to the above particularly preferred embodiment, stable process control can be performed irrespective of the skill level of the operator making the change.

Further, in the particularly preferred embodiment, the signal related to the flow rate and composition of the purge gas is detected, and the coefficient of the signal transfer function is obtained based on at least the signal related to the flow rate and composition of the purge gas. If such a configuration is adopted, a large fluctuation of the CO partial pressure in the reactor can be suppressed, and stable control can be performed.

Further, in the present invention, a signal relating to the flow rate and the composition in the feed oxo gas is further detected, and the manipulated variable is determined based on the detected signal relating to the flow rate and the composition. By adopting this configuration, the CO concentration in the reactor can be controlled within a desired range despite the large composition fluctuation of the fed oxo gas.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION In the present invention, the olefin used as a raw material for the hydroformylation reaction is not particularly limited as long as it is an organic compound having at least one olefinic double bond in the molecule. For example, ethylene, propylene, butene, butadiene, pentene, hexene, hexadiene, octene, octadiene, decene, hexadecene, octadecene and the like can be mentioned.

In the present invention, as a catalyst, a Group VIII metal complex such as iron, cobalt, nickel, ruthenium, rhodium, palladium, osmium, iridium,
A complex soluble in an organic compound-containing solution containing at least one metal selected from the group consisting of platinum and further containing an organic phosphorus compound, for example, a phosphine such as triarylphosphine or a ligand such as phosphite is used. To be

In carrying out the hydroformylation reaction,
The use of a reaction solvent is not essential, but if necessary, an inert solvent can be present in the hydroformylation reaction. Specific examples of preferable solvents include toluene, xylene, aromatic hydrocarbon compounds such as dodecylbenzene, ketones such as acetone, diethyl ketone and methyl ethyl ketone, ethers such as tetrahydrofuran and dioxane,
Esters such as ethyl acetate and di-n-octyl phthalate may be mentioned.

As the mixed gas of hydrogen and carbon monoxide used in the reaction, oxo gas or water gas is used. In this specification, these are collectively referred to as oxogas.

As the reaction conditions for carrying out the hydroformylation reaction process, the conditions conventionally used can be adopted. For example, the reaction temperature is selected from the range of room temperature to 200 ° C, preferably 50 ° C to 150 ° C, and the reaction pressure is usually atmospheric pressure to 200 atm, preferably 5 to 100 atm, particularly preferably 5 to 50 atm. Selected from the range. The molar ratio of hydrogen to carbon monoxide (H ₂ / CO) is usually 10/1 to
It is selected from the range of 1/10, preferably 1/1 to 6/1. As a reaction system of the hydroformylation reaction, a continuous system performed in a stirring reaction tank or a bubble column reaction tank can be adopted, but a stirring reaction tank is preferably used.

In the present invention, the CO partial pressure in the reactor, which is obtained in correlation with the desired production ratio of the reaction product aldehyde, is set as a target value in the PID control system, for example, and based on this setting. By adjusting the flow rate of the feed oxo gas, the CO partial pressure in the reactor is controlled so as to reach the target value.

Hereinafter, a particularly preferred embodiment of the present invention will be described in more detail. In this embodiment, in controlling the hydroformylation reaction: 1. CO model predictive control,
2. Process gain scheduling, 3. Each method of feedforward control is adopted.

1. CO Model Predictive Control First, prior to control, a process model having a predetermined transfer function is stored in the control computer. As the process model, for example, a first-order lag system model including dead time is adopted. The transfer function G (s) of this model is expressed by the following equation.

G (s) = {K _P / (1 + T _P · s)} · exp (−T _L · s) (1) where K _P is the process gain and T _P is the time constant of the past process response. _TL is a Laplace variable, where _{TL is a} dead time and a value obtained from a past process response.

FIG. 2 is a graph for explaining the control state in the CO model predictive control. In the figure,
According to the change in the production ratio of aldehyde, CO at the current time t
The partial pressure target value R is changed stepwise. In such a case, the manipulated variable for adjusting the flow rate of the feed gas required for control is determined by the process model predictive control based on the deviation between the current CO partial pressure and the target CO partial pressure. The procedure is as follows.

Step 1 A target value reference trajectory y _R having a smoother waveform is calculated from the given CO partial pressure target value R. This is to prevent the system from becoming unstable due to a sudden change in the target value. An example of the target value reference trajectory y _R is shown in FIG.

Step 2 Based on the process model, τ is used as a control period, and an input sequence from the current time t to the time t + Mτ of the input u (the manipulated variable for adjusting the feed amount of oxo gas) illustrated by the broken line is used as a parameter. , Future time t + Lτ to time t +
Output prediction series up to (L + P) τ (CO of recycled gas
Express partial pressure). This is done as follows.
Note that M, P, and L are respectively M: manipulated variable determination section (Co
ntrol Horizon), P: Output / target value match desired section (Coin
cidence Horizon), L: output / target value matching desired section start time point, which are determined when designing the control system.

Now, y _M (t + L) is a future time t + which is directly obtained by a process model from one input sequence.
Assuming that the model value of CO partial pressure at Lτ is a model prediction series Y _M including this prediction value, Y _M = [y _M (t + L), ..., Y _M (t + L + P-1 )] It can be expressed as ^T.

Further, y _P (t + L) is obtained by correcting the model predictive value y _M (t + L) of the CO partial pressure at time t + L.
If the output predicted value at τ is set, the time t + (L + P−1) τ from the future time t + Lτ including this output predicted value.
Similarly, one output prediction sequence (CO partial pressure) Y _P between Y _P = [y _P (t + L), ..., Y _P (t + L + P-1)] ^T Can be represented.

A process model correction term is calculated in order to correct the model prediction series Y _M to obtain the output prediction series Y _P. In this case, the difference between the actual measured value y (t) of the CO partial pressure at the current time t and the estimated output value y _M (t) at the current time obtained from the process model by the past data is used as the correction term.

The series Y and Y _M0 for the above correction are defined as follows. Y = [y (t), ..., y (t)] Y _M0 = [y _M (t), ..., y _M (t)] If defined as above, the output prediction sequence Y _P is It is expressed as follows. Y _P = Y _M + Y−Y _M0 (2) Here, the following relationship Y _M = Y between the model prediction sequence Y _M and the sequence Y _M0 including the output estimated value y _M (t) at the current time. _{There is M0} + A _F · ΔU _M + A ₀ · ΔU ₀ (3). The model prediction series Y _M is obtained based on the above equation (3). When the model prediction series Y _M is obtained, the output prediction series Y _P is obtained based on the equation (2).

The second term on the right side of the equation (3) is a future prediction term, and the third term is a prediction term based on past data. A _F in the second term is the response sequence of the prediction model, It is represented by. In the equation, a _k is a step response sequence determined by a transfer function, and a _K = 0 (0 ≦ kτ ≦ T _L ) a _k = K _P {1-exp (-(kτ-T _L ) / T _P ). } (T _L ≦ kτ) (4) a _k = a _s (s <k).

Further, ΔU in the second term of the equation (3)
_M is one input sequence consisting of input parameters from time t to time t + M-1, that is, a sequence corresponding to the manipulated variable for adjusting the feed amount of oxo gas at each time during that period. ΔU _M is a parameter Δu (t) representing the operation amount at each time
From Δu (t + M−1) to ΔU _M = [Δu (t), Δu (t + 1), ..., Δu (t + M-1)] ^T. By selecting each parameter, the future prediction term of equation (3) is calculated.

A ₀ and Δ in the third term on the right side of the equation (3)
U ₀ is A ₀ : process characteristic matrix, ΔU ₀ : manipulated variable input sequence, respectively. Is expressed as From this A ₀ and ΔU ₀ , the third
Terms are required.

As described above, the process computer selects each parameter of the input series ΔU _M in a predetermined order, and sequentially selects the input series one by one and inputs them to the sequential process model. . For the model prediction series Y _M sequentially obtained from the output of the process model,
The correction is performed based on the equation (2), and the output prediction series Y _P is sequentially obtained.

Step 3 Future output prediction series Y _P obtained based on the process model in Step 2 and the value of the section from future time t + Lτ to time t + (L + P) τ in the reference trajectory y _R obtained in Step 1. One input sequence that minimizes the difference between
It is calculated using the method of least squares. That is, one of the output prediction sequences Y _P calculated in sequence, which gives the output closest to the reference trajectory sequence defined by the reference trajectory y _R , is calculated using the following equation (5). To be selected. Minimize ┃Y _P −Y _R ┃ ² (5) where Y _R is a reference trajectory sequence.

The reference trajectory sequence Y _R can be obtained from the output at the current time, for example, by the following equation. Y _R = A · F · y (t) + (IA) · Γ where A, Γ, and F are A: weight vector,
Γ: target value vector in desired output / target value matching section, F: identity matrix, represented by the following equation. In addition,
In the equation, α is a parameter, 0 <α <1, γ (t) is a target value at time t, and I is a unit matrix. A = [α, α ² , ... α ^P ] ^T F = diag [1,1, ..., 1] ^T Γ = [γ (t + L), ..., γ (t + L + P-1)] ^T

In this way, when one output prediction series Y _P capable of giving the output closest to the future target value is obtained, the input series giving this output prediction series adjusts the feed amount of oxo gas. Determined as a series of quantities. The feed flow rate of the oxo gas is sequentially selected based on this series of manipulated variables.

2. Process Gain Scheduling When the production rate is changed in a hydroformylation reactor, the characteristics of the process change accordingly, especially the change in process gain, which is the coefficient K _P of the transfer function G (s) of equation (1). . Therefore, in order to maintain stable control even when the production rate is changed, the process gain in the process model of Expression (1) is automatically changed. This technique is called gain scheduling. In this embodiment, gain scheduling is performed as follows.

F _t is the total purge flow rate of the gas purged from the purge control valve, F _H2 , F _CO , and F _i , respectively.
F _t = F _H2 + F _CO + F _i (6) where H ₂ flow rate, CO flow rate, and inert flow rate in the total purge flow rate are given. Here, _assuming that A _CO , A _H2, and A _i are the CO concentration, H ₂ concentration, and inert concentration in the purge gas, respectively, A _CO = F _CO / F _t = F _CO / (F _H2 + F _CO + F _i ) (7) If A _CO is fully differentiated,

[Equation 1] Is obtained. Substituting equation (7) into equation (8), dA _CO = 1 / F _t {-(F _CO / F _t ) ・ dF _H2 + (F _H2 / F _t + F _i / F _t ) ・ dF _CO } (9) Further, from the definition, A _CO = F _CO / F _t (10) A _H2 = F _H2 / F _t (11)

In the steady state, the composition of the source gas and the reaction temperature are almost constant and the reaction is stable. Therefore, if the reaction amount y is constant and the feed gas amount is P _t , then P _t = 2y + F _t (12) source gas Since the composition of H 2 / CO = 1, if the flow rate of H ₂ in the feed gas is P _H2 and the flow rate of _CO is P _CO , then dF _t = (1/2) · dP _H2 (13) dF _t = (1 / 2) · dP _CO (14) is obtained. Substituting the equations (13) and (14) into the equation (9), dA _CO = ((1 / (2F _t )) · (1-2A _CO ) · dF _t (15), and from the equation (12), Since dP _t = dF _t (16), dA _CO / dP _t = (1 / (2F _t )) · (1-2A _co ) (17)

The left side of the equation (17) is the change in the CO partial pressure with respect to the change in the unit amount of the feed flow rate, that is, the process gain K _P defined by the equation (1). Therefore, it is expressed as K _P = (1 / (2F _t )) · (1-2A _CO ) (18). By changing the process gain according to this equation, a transfer function that can accurately represent the process response is obtained each time.

Therefore, the process data A _CO and F _t are fetched from the detector at a constant period, the value of K _P is obtained according to the above equation (18), and the value of the gain K _P of the process model is changed. Gain scheduling in the process model is possible.

3. The CO partial pressure in the feedforward controlled reactor is affected by compositional variations in the feed oxo gas. Therefore, C in the feed oxo gas
To compensate for O density variation, controls the feed rate F _in performs feedforward control. Here, when C _CO is the CO partial pressure (concentration) in the recycled gas and B _CO is the CO concentration in the feed oxo gas, dF _in / dt = − (F _in / B _{CO is obtained} from the mass balance around the reactor. ) ・ DB _CO / dt (19) holds. Therefore, when the CO partial pressure B _CO in the feed oxo gas fluctuates, the CO fluctuation is compensated by changing the feed flow rate F _in based on the equation (19).

The present invention will be described in more detail with reference to the drawings. The method for controlling the hydroformylation reaction according to the first embodiment of the present invention is performed, for example, in the system shown in the system diagram of FIG. CO in recycled gas
A desired value of the partial pressure is set as a target value in the PID control device, and the target value and the recycled gas supply line 13 are set.
The amount of oxo gas fed to the reactor 1 is adjusted based on the detected amount of CO concentration in the reactor. Adjusting the feed flow rate
This is performed by adjusting the opening degree of the oxo gas feed control valve 10, and the CO concentration in the recycle gas is controlled so as to match the target value.

FIG. 3 shows a systematic diagram of a system for carrying out the method for controlling the hydroformylation reaction according to the second embodiment of the present invention. In the figure, this system has a reactor 1, a gas-liquid separator 2, a purge control valve 9, an oxo gas feed control valve 10, each system piping, and a control unit 8, and a signal is provided at a predetermined position. A detection unit and an operation unit are provided, and these are connected to the control unit 8 via a control wiring.

The reactor 1 includes an olefin supply line 1
1. An oxo gas supply line 12 is connected for feeding a raw material, and a hydrogen gas supply line 16 is connected to the oxo gas supply line 12. The reactor 1 is further provided with a catalyst supply line 14, an aldehyde product withdrawal line 15 for removing product aldehyde, communication lines 17 and 18 with the gas-liquid separator 2, and a recycled gas from the gas-liquid separator 2. Is connected to a recycled gas supply line 13 for extracting and returning to the reactor 1.

A recycle gas purge control valve 9 is provided on the gas-liquid separator 2 side of the recycle gas supply line 13, and a purge gas composition analyzer 3 and a recycle gas purge flow meter 4 are provided on the outlet side of the purge control valve 9. Is installed. In addition, in the recycled gas supply line 13,
A CO analyzer 5 is provided for analyzing the CO concentration in the recycled gas supplied to the reactor 1, and an oxo gas feed control valve for adjusting the flow rate of the oxo gas fed to the reactor 1 is provided in the oxo gas supply line 12. 10 are arranged respectively. In the oxo gas supply line 12, a flow meter 6 for detecting the flow rate of the oxo gas and a CO analyzer 7 for measuring its components are arranged.

The control unit 8 is composed of a process computer for controlling the entire reaction process, and functionally has a data storage unit 81 for temporarily storing the data detected by each detection unit, and a composition fluctuation of the feed oxo gas. A feedforward (FF) control unit 82 for compensating for the above, a feedback (FB) control unit 83 for controlling the CO concentration in the recycled gas according to its target value, a model calculation unit 85 for calculating a process model, a calculation A model 84 in which the stored model is stored, and a feed flow rate controller 86. The target value m of the CO concentration in the recycle gas is input to the feedback control unit 83 by the operator as necessary. The control unit 8 can be configured as a DCS (distributed computer system), and the signal of each control unit can be output as a DDC (direct digital control) control signal.

In the system configured as described above, the reactor 1 is provided with each supply line 11, 12, 13, 1
Olefin, oxo gas, recycle gas, catalyst, and hydrogen gas are continuously supplied via 4 and 16 at a supply rate according to the reaction conditions. Inside the reactor 1,
Olefin and oxo gas undergo a hydroformylation reaction in the presence of a catalyst at a predetermined temperature and pressure,
Normal aldehyde and isoaldehyde are produced in the desired ratio. The obtained aldehyde product is continuously withdrawn as a liquid component from the lower portion of the reactor 1 via the aldehyde product withdrawing line 15.

On the other hand, the gaseous component in the reactor 1 is withdrawn from the top of the reactor 1 via the line 17 and the gas-liquid separator 2
And is separated into gas and liquid in the gas-liquid separator 2. The liquid phase in the gas-liquid separator 2 returns to the reactor 1 again via the liquid return line 18, and the gaseous component in the gas phase passes via the recycle gas supply line 13 to the reactor 1 Returned from its bottom.

The composition fluctuation of the feed oxo gas can be compensated by the operation signal j from the feed forward control unit 82. This feedforward control is performed based on the equation (19) derived from the mass balance around the reactor 1. CO concentration signal a in oxo gas detected by CO analyzer in oxo gas 7, CO concentration signal c in recycled gas detected by CO analyzer in recycled gas 5, and oxo gas flow rate signal detected by oxo gas flow meter 6 Each signal of b is temporarily taken into the data storage unit 81, and further, the data storage unit 8
From 1, the data is fetched into the feedforward control unit 82 as feedforward control data e. Based on this signal, the feedforward control unit 82 calculates the feedforward operation signal j.

Further, in order to set the ratio of the normal aldehyde and the isoaldehyde of the aldehyde product to the desired ratio, the target value of the CO concentration of the recycled gas returned to the reactor 1 is set. The control according to this target value is performed based on the feedback operation signal k, and is possible by controlling the opening amount of the oxo gas feed control valve 10 to control the supply amount of the CO component supplied into the reaction system. Become.

The feedback operation signal k is calculated as follows. CO analyzer in recycled gas 5
CO concentration signal c in recycled gas detected by
However, it is temporarily stored in the data storage unit 81. The feedback control unit 83 takes in these signals as feedback control data f at regular intervals. The CO partial pressure (concentration) target value m is stored in the feedback control unit 83.

The process gain is calculated as follows. The purge gas composition signal n indicating the component composition (CO and inert components) in the purge gas obtained by the purge gas composition analyzer 3 and the purge gas flow rate signal d detected by the recycle gas purge flow meter 4 are temporarily taken into the data storage unit 81. . The model calculation unit 85 takes in these signals as the process model calculation data h at regular intervals, and based on this data, the above equation (18)
Is calculated to calculate the process gain. The calculated process gain is given to the model 84 as the process model change information i, and is used in the model 84 to change the transfer function (equation (1)) of the process model.
The feedback control unit 83 receives the process model information g from the model 84 and performs arithmetic processing based on the equations (2) to (5). By this arithmetic processing, CO
The operation amount for adjusting the feed amount of the oxo gas that minimizes the difference between the predicted concentration and the target value m is obtained as the feedback operation amount k.

The feedforward manipulated variable j and the feedback manipulated variable k respectively calculated as described above are added by the feed flow rate controller 86 to obtain the set value. This set value is given to the oxo gas feed control valve 10 as a feed flow rate operation signal 1 for adjusting the opening thereof. Note that, instead of this configuration, either the feedforward manipulated variable j or the feedback manipulated variable k may be given to the feed control valve 10 as the manipulated variable.

[0056]

EXAMPLES Examples using the control method of the present invention and comparative examples for comparison therewith will be described below. However, the present invention is not limited to the configurations in these examples as long as the gist thereof is not exceeded.

Example 1 Regarding the controllability of the reaction when the amount of propylene supplied was increased by 10% using the method for controlling the hydroformylation reaction of the first embodiment with the configuration of the system diagram shown in FIG. It was investigated by simulation. In this case, the supply amount of propylene is set to 1 while maintaining the pressure (total pressure) in the reactor at a predetermined pressure by adjusting the purge flow rate of the recycled gas.
The CO concentration in the reactor was adjusted by increasing the CO concentration in the recycle gas by 10% over time and adjusting the supply amount of oxo gas so that the CO concentration in the recycle gas became the target value. As a result, as shown in FIG. 4, the reaction rate of propylene, the reactor pressure (total pressure), and the behavior of CO partial pressure in the reactor were all within the predetermined ranges, and the reaction was well controlled. It was

Comparative Example 1 The total amount of purged recycle gas was the same as in Example 1, oxo gas was supplied in proportion to the supply amount of propylene without adjusting the supply amount of oxo gas, and the amount of oxo gas in the reactor was adjusted. The simulation was performed under the same conditions as in Example 1 except that the pressure (total pressure) was adjusted to the predetermined pressure while being adjusted by the purge flow rate. The result is shown in FIG. As shown in the figure, the CO partial pressure in the reactor could not be kept within the predetermined range.

Example 2 With the configuration of the system diagram shown in FIG. 1, the propane concentration in propylene was increased from 2.5% to 5% by using the method for controlling the hydroformylation reaction of the first embodiment. The controllability of the reaction in each case was investigated by simulation. In this case, the purge flow rate is adjusted to maintain the pressure (total pressure) in the reactor at a predetermined pressure, and the supply amount of oxo gas is adjusted so that the concentration in the recycled gas reaches its target value. The propane concentration in propylene was instantaneously increased from 2.5% to 5% while adjusting the CO concentration in the reactor. Results are shown in FIG. As shown in the figure, the behavior of the reaction rate of propylene, the reactor pressure (total pressure), and the CO partial pressure in the reactor were all within the predetermined ranges, and the reaction was well controlled.

Comparative Example 2 With the total amount of recycled gas purged being the same as in Example 2, oxo gas was supplied in proportion to the amount of propylene supplied without adjusting the amount of oxo gas supplied, and the amount of oxo gas in the reactor was adjusted. The simulation was performed under the same conditions as in Example 2 except that the pressure (total pressure) was adjusted to the predetermined pressure while being adjusted by the purge flow rate. The result is shown in FIG. 7. As shown in the figure, the CO partial pressure in the reactor gradually increased and could not be kept within the predetermined pressure range.

[0061]

As described above, according to the present invention,
The desired C concentration in the reaction system is set as the target value, and the detected C
By comparing the O partial pressure with the target value and adjusting the flow rate of the oxo gas fed to the reactor, the C in the reactor can be adjusted.
By adopting a configuration for controlling the O concentration, the CO
When the target value of the concentration is changed, the CO concentration in the reactor does not largely fluctuate, and good control can be performed only by setting the target value to a desired value, thus reducing the burden on the operator.

Further, by adopting the configuration in which the process model having a predetermined transfer function is stored in the computer and the model predictive control is adopted, even if the change of the target value is particularly abrupt, a large control fluctuation does not occur. Since stable control can be performed in response to the change, stable process control is particularly easy regardless of the skill of the operator who makes the change.

By adopting a configuration in which the coefficient of the transfer function is sequentially changed according to the data obtained from the process, even if the production rate is largely changed, the process is stable without requiring any special operation by the operator. Control can be continued.

Further, by adopting a configuration in which the feed flow rate of the oxo gas is changed based on a predetermined function to compensate for the variation of the composition of the feed oxo gas, the reaction can be performed regardless of a large variation or difference of the composition of the feed oxo gas. Since the CO concentration in the vessel can be easily controlled within a desired range, a desired ratio of aldehyde can be produced.

[Brief description of drawings]

FIG. 1 is a system diagram showing a configuration of a system for carrying out a method for controlling a hydroformylation reaction according to a conventional method and a first embodiment of the present invention.

FIG. 2 is a graph for explaining the control of the present invention.

FIG. 3 is a system diagram showing a method for controlling the hydroformylation reaction according to the second embodiment of the present invention.

FIG. 4 is a graph showing the control results obtained in Example 1 in which the hydroformylation reaction was controlled according to the present invention.

FIG. 5 is a graph showing the control results obtained in Comparative Example 1.

FIG. 6 is a graph showing the control results obtained in Example 2 in which the hydroformylation reaction was controlled according to the present invention.

7 is a graph showing the control results obtained in Comparative Example 2. FIG.

[Explanation of symbols]

 1: Hydroformylation reactor (reactor) 2: Gas-liquid separator 3: Purge gas composition analyzer 4: Recycle gas purge flow meter 5: Recycle gas CO analyzer 6: Oxo gas flow meter 7: Oxo gas CO analyzer 8: Control unit 9: Recycle gas purge control valve 10: Oxo gas feed control valve 81: Data storage unit 82: Feed forward control unit 83: Feedback control unit 84: Model 85: Model calculation unit 86: Feed flow rate controller

Front page continuation (72) Inventor Noroshi Inoue 3-10 Shiodori, Kurashiki City, Okayama Prefecture Mitsubishi Kasei Co., Ltd. Mizushima Plant

Claims (7)

[Claims] 1. Hydroformylation of an olefin, a feed oxo gas containing hydrogen and carbon monoxide, and a recycle gas withdrawn from the reactor and returned to the reactor in the presence of a catalyst in the reactor. In a method of controlling a hydroformylation reaction for producing an aldehyde by reacting, an objective value of carbon monoxide partial pressure in a reaction system for controlling a hydroformylation reaction is set, and a carbon monoxide content corresponding to the objective value is set. A method for controlling a hydroformylation reaction, comprising detecting a pressure, comparing the detected partial pressure of carbon monoxide with the target value, and adjusting a flow rate of the feed oxo gas. 2. The method for controlling a hydroformylation reaction according to claim 1, wherein the carbon monoxide partial pressure of the recycled gas returned to the reactor is used as the detected carbon monoxide partial pressure. 3. The recycle gas is combined with the feed oxo gas before being returned to the reactor, and the carbon monoxide partial pressure in the combined gas is set as the detected carbon monoxide partial pressure. The method for controlling the hydroformylation reaction according to claim 1, which is characterized in that. 4. A method for simulating the hydroformylation reaction, which has a predetermined signal transfer function which receives an operation amount for adjusting the flow rate of the feed oxo gas as an input and outputs the calculated value of the detected partial pressure of carbon monoxide. A process model for performing the process is stored in a computer, and an input sequence consisting of a plurality of inputs that occur sequentially at a predetermined cycle is sequentially selected as an input of the process model, and a future output based on the output of the process model. A sequence is sequentially calculated, and one output sequence having a small difference from the planned future target value is selected from among the calculated future output sequences, and the output sequence corresponding to the selected one output sequence is selected. The method for controlling a hydroformylation reaction according to claim 1, wherein the manipulated variable is set based on an input series. 5. The method for controlling a hydroformylation reaction according to claim 4, wherein the one output series is selected according to a least square method. 6. The signal related to the flow rate and composition of the purge gas is detected, and the coefficient of the signal transfer function is calculated based on at least the signal related to the flow rate and composition of the purge gas. Or the method for controlling the hydroformylation reaction according to item 5. 7. The manipulated variable is further determined by further detecting a signal related to the flow rate and composition of the feed oxo gas, and further based on the signal related to the flow rate and composition of the feed oxo gas. A method for controlling a hydroformylation reaction according to claim 1.