Ethylene oxide chemical production plant and method

Abstract

The invention is directed to a chemical production plant (100) for producing ethylene oxide (EO), which comprises a reaction unit (102) for providing a reaction stream (110) comprising ethylene oxide (EO) and a recovery unit (112) for recovering ethylene oxide (EO). The chemical production plant (100) further comprising a linear model based predictive controller unit (150) that includes a recovery sub-controller unit (152) and a reaction sub-controller unit (154) that based on respective current variable data (VD1, VD2) and respective dynamic control sub-matrices (DCM1, DMC2), are configured to determine and provide respective setpoint values (SP1, SP2) of one or more manipulated variables (MV1, MV2) associated to the recovery unit or the reaction unit; and to generate and provide respective operation instructions (OP1, OP2) for controlling operation of the recovery unit or the reaction unit in accordance with the determined setpoint values.

Description

[0001] Ethylene oxide chemical production plant and method

[0002] FIELD OF THE INVENTION

[0003] The present invention is directed to a chemical production plant for the production of ethylene oxide, also referred to herein as EO. The invention is further directed to a method for controlling operation of a chemical production plant and to a computer program.

[0004] BACKGROUND OF THE INVENTION

[0005] Ethylene Oxide (C2H4O) is one of the most important raw materials used in large scale chemical production. It is mainly used for synthesis of ethylene glycols, including diethylene glycol and triethylene glycol. Other important products include ethylene glycol ethers, ethanolamines and ethoxylates.

[0006] Model predictive control is a method of process control that is used to control a process while satisfying a set of constraints. MPC models predict the change in the dependent variables of the modeled system that will be caused by changes in the independent variables. Dependent variables are commonly referred to as controlled variables. Independent variables are commonly referred to as manipulated variables.

[0007] R. A. R. Cano, and D. Odloak, Robust model predictive control of integrating processes, Journal of Process Control 13 (2003) 101 -1 14 discloses an example of an industrial reactor system or chemical production plant that produces ethylene oxide and has controlled outputs or variables and manipulated inputs or variables. The document discloses a model predictive control (MPC) that can be used to control operation of the industrial reactor system. | BASF SE | 231731

[0008] SUMMARY OF THE INVENTION

[0009] It would be beneficial to enable the use of a linear model predictive control for a chemical production plant for producing ethylene oxide using a reaction-recovery process.

[0010] According to a first aspect of the present invention, a chemical production plant for producing ethylene oxide in a production process is disclosed. The chemical production plant is configured to produce EO using a reaction-recovery process. Thus, the chemical production plant comprises

[0011] - a reaction unit comprising an ethylene input unit, or ethylene feed (e.g., a pipe), for receiving ethylene, an oxygen input unit, or oxygen feed for receiving oxygen and a reactor assembly for generating and providing a reaction stream comprising ethylene oxide; and

[0012] - a recovery unit comprising an absorber unit that is configured to receive the reaction stream and a lean absorbent stream, and to provide a rich absorbent stream comprising a higher ethylene oxide content than that of the lean absorbent stream. The recovery unit further comprises a stripper unit that is configured to receive the rich absorbent stream, to extract or recover ethylene oxide from the rich absorbent stream thereby generating the lean absorbent stream, to provide the recovered ethylene oxide via an ethylene oxide output unit and to provide the generated lean absorbent stream to the absorber unit.

[0013] The chemical production plant further comprises a linear model based predictive controller unit that includes:

[0014] - a recovery sub-controller unit that is configured to receive first current variable data indicative of a current operational state of the recovery unit, and based on a predetermined first dynamic control sub-matrix, to determine and provide respective setpoint values of one or more first manipulated variables associated to the recovery unit and to generate and provide first operation instructions for controlling operation of the recovery unit in accordance with the determined setpoint values;

[0015] - a reaction sub-controller unit that is configured to receive second current variable data indicative of a current operational state of the reaction unit, and based on a predetermined second dynamic control sub-matrix, to determine and provide a respective setpoint value of one or more second manipulated variables associated to the reaction unit and to gener- | BASF SE | 231731 ate and provide second operation instructions for controlling operation of the reaction unit in accordance with the determined setpoint values.

[0016] The operation of the recovery unit and of the reaction unit in accordance with the determined setpoint values advantageously corresponds to a predetermined target operation state of the chemical production plant that is determined by the linear model based predictive controller unit, in particular by the first and second dynamic control sub-matrices of the recovery sub-controller unit and the reaction sub-controller unit respectively.

[0017] The production process of chemical production plant that includes the reaction unit for generating EO and the recovery unit for extracting the EO is controlled by the linear model based predictive controller unit that is advantageously adapted to this particular chemical production part. The recovery unit is associated to the recovery sub-controller unit and the reaction unit is associated to the reaction sub-controller unit. The recovery sub-controller unit is configured to receive the first current variable data. The first current variable data pertains to the current operational state of the recovery unit. The recovery sub-controller unit includes a first dynamic control sub-matrix. Applying the first current variable data to the first dynamic control sub-matrix results in setpoint values for one or more manipulated variables associated to the recovery unit. A manipulated variable is a variable that is manipulated by the linear model based predictive controller unit to get the production process close to a predetermined optimal working point. Using the first dynamic control sub-matrix, the linear model based predictive controller unit determines or calculates setpoints values to achieve the optimal working point of the recovery unit. Further, first operation instruction for controlling the recovery unit are generated and provided. The first operation instructions are indicative of, or otherwise include or make reference to, the determined setpoint values, such that the determined setpoint values for the recovery sub-controller unit can be extracted from the first operation instructions, which may include further data, such as headers, overheads, timestamps or other payload, for example to identify to which variable the respective setpoint value refers to.

[0018] Analogously, the reaction sub-controller unit is configured to receive the second current variable data. The second current variable data pertains to the current operational state of the reaction unit, such as to the reception of ethylene, the reception of oxygen, the generation and / or provision of the reaction stream. The reaction sub-controller unit includes a second dynamic control sub-matrix. Applying the second current variable data to the second dynamic control sub-matrix results in setpoint values for one or more manipulated variables associated to the reaction unit. Using the second dynamic control sub-matrix, the linear model based predictive controller unit determines or calculates setpoints values to BASF SE 231731 achieve the optimal working point of the reaction unit. Further, second operation instruction for controlling the reaction unit are generated and provided. The second operation instructions are indicative of, or otherwise include or make reference to, the determined setpoint values, such that the determined setpoint values for the recovery sub-controller unit can be extracted from the first operation instructions, which may include further data, such as headers, overheads, timestamps or other payload, for example to identify to which variable the respective setpoint value refers to.

[0019] The first and the second dynamic control sub-matrix are extracted from, or part of, a common dynamic control matrix that reflects the interaction between the reaction process in the reaction unit and the recovery process in the recovery unit. Dynamic control matrix refers to a matrix of dynamic models that describes the interactions between the different variables of the respective process, i.e., the reaction process carried out in the reaction unit and the recovery process carried out in the recovery unit. The dynamic control matrix (and thus, the respective dynamic control sub-matrices) enables a predictive control strategy of the chemical production plant, based on dynamic matrix control (DMC) principles.

[0020] Providing distinct sub-controller units, one for the recovery unit and another one for the reaction unit, each including a respective dynamic control sub-matrix for determining optimal setpoint values in dependence on the current variable data simplifies the control of the chemical production plant based on the reaction recovery process.

[0021] In the following, embodiments of the chemical production plant of the first aspect of the invention will be disclosed.

[0022] In a particular embodiment, the linear model based predictive control unit also comprises a process controller unit configured to receive, from the recovery sub-controller unit and from the reaction sub-controller unit (also referred to as the sub-controller units), the respective setpoint values of the first manipulated variables and of the second manipulated variables. In this embodiment, the process controller unit is configured to generate and provide the first operation instructions for controlling operation of the recovery unit in accordance with the determined setpoint values and the second operation instructions for controlling operation of the reaction unit in accordance with the determined setpoint values.

[0023] The setpoint values determined by the recovery sub-controller unit and by the reaction subcontroller unit are thus provided to the process controller unit. Based on the received setpoint values, the process controller unit is configured to control operation of both the recov- BASF SE 231731 ery unit and the reaction unit. For this, the controller unit is configured to generate and provide, to the recovery unit, first operation instructions for controlling the recovery unit in accordance with the received setpoint values and to generate and provide, to the reaction unit, second operation instructions for controlling the reaction unit in accordance with the received setpoint values.

[0024] Thus, in this embodiment, the respective sub-controller unit merely provide the setpoint values determined to the process controller unit, which based on the setpoint values received, generates and provides the required first and second operation instructions for controlling operation of the reaction unit and of the recovery unit.

[0025] The operation of the recovery unit and the reaction unit in accordance with the determined setpoint values corresponds to a predetermined target operation state of the chemical production plant.

[0026] In an embodiment, the first current variable data comprises first current manipulated variable data indicative of a current value of the respective first manipulated variables associated to the recovery unit and first current controlled variable data indicative of a current value of first controlled variables that are associated to the recovery unit, wherein the current values of the first controlled variables depend on the current values of one or more first manipulated variables in accordance with the first dynamic control sub-matrix.

[0027] Optionally, the first current variable data also comprises, first current feedforward variable data indicative of a current value of respective first feedforward variables associated to the recovery unit that influence the values of the first controlled variables but are not adjustable by the sub-controller units and / or by the process controller unit.

[0028] The objective of the sub-controller units and / or of the process controller unit is to drive the controlled variables such that the chemical production plant, in particular the production process, to an optimal working point. Controlled variables are also referred to as dependent variables since their values depend on the manipulated variables as defined by the first dynamic control sub-matrix.

[0029] Manipulated variables, as stated above, are variables that are manipulated by the linear model based predictive controller unit so that the controlled variables get closer to the optimal working point. The linear model based predictive controller unit determines, based on the first dynamic control sub-matrix, setpoint values for the manipulated variables. BASF SE 231731

[0030] A feedforward variable is a variable that cannot be manipulated by the linear model based predictive controller unit but that nevertheless has an influence on the controlled variables. The feedforward variable acts as a measured disturbance on the controlled variables. By taking into account this measured disturbance (feedforward variable), the linear model based predictive controller unit can provide a more accurate control of the controlled variables.

[0031] The recovery sub-controller units determines the setpoint values for the first manipulated variables and provides the first operation instructions based thereon. When a process controller unit is provided, the process controller unit receives the setpoint values for the first manipulated variables and then generates and provides the first operation instructions based thereon.

[0032] Additionally, or alternatively, in another embodiment, the second current variable data comprises second current manipulated variable data indicative of a current value of the respective second manipulated variables associated to the reaction unit, and second current controlled variable data indicative of a current value of second controlled variables that are associated to the reaction unit, wherein the current values of the second controlled variables depend on the current values of one or more second manipulated variables in accordance with the second dynamic control sub-matrix.

[0033] Preferably, the second current variable data also comprises second current feedforward variable data indicative of a current value of respective second feedforward variables associated to the reaction unit that influence the values of the second controlled variables but are not adjustable by the reaction sub-controller unit.

[0034] In another embodiment, the recovery sub-controller unit and / or the reaction sub-controller unit are further configured to receive constraint data indicative of allowed operational ranges for one or more controlled variables and / or for one or more manipulated variables.

[0035] The optimal working point for the production process in the chemical production plant is further determined in this embodiment by plant constraints. Information pertaining to these constraints is provided as constraint data to one or more of the sub-controller units. The constraints can be for instance defined by equipment limitations, safety limits or specification limits. The linear model based predictive controller unit is configured to provide the setpoint values for the first and / or the second manipulated variables further in dependence BASF SE 231731 on the constraint data, such that the provided setpoint values stay within an adequate range.

[0036] In yet another embodiment, the reception of the first and second current variable data, and / or the determination of the setpoint values of the first manipulated variables and the second manipulates variables is performed at regular time intervals. Preferably, in an embodiment, the regular time interval is one minute. Preferably the necessary data is retrieved at the same time; subsequently the respective sub-controller unit calculates the corresponding setpoint values and provides the control instruction indicative thereof back to the reaction unit and / or the recovery unit, within a fixed cycle or time interval of 1 minute In alternative embodiments, the time interval is 10 seconds, 20 seconds, 30 seconds, 40 seconds, 50 seconds, 2 minutes, 3 minutes, 4 minutes or five minutes. In an embodiment, the current variable data is received at a different time interval from the time interval at which the setpoint values are determined and provided. Also, current variable data indicative of a first variable can be received at a different time interval from the time interval at which a current variable data of a second variable is received. In an embodiment, current variable data indicative of one or more variable is monitored and continuously or quasi-continuously provided.

[0037] In a preferred embodiment, the first current variable data includes data pertaining to one or more of ethylene oxide loss, differential pressure in the absorber unit, differential pressure over a lean absorbent valve, temperature in a top section of the stripper unit, temperature in a middle section of the stripper unit, temperature in a bottom section of the stripper unit, ethylene oxide content in the bottom section of the stripper, inert limit, opening of stripper steam valve, differential pressure in the stripper unit, density of glycol and level of demineralized water tank, preferably as values of first controlled variables.

[0038] Additionally, or alternatively, the first current variable data includes data pertaining to one or more of lean absorbent flow, stripper temperature, demineralized water flow, residual absorber pressure and glycol bleed flow, preferably as values of first manipulated variables.

[0039] Preferably, the first current variable data includes data pertaining to one or more of a lean absorbent temperature, opening of a process water valve, lean absorber to residual absorbent flow and an ethylene oxide flow to the stripper unit, preferably as values of first feedforward variables. BASF SE 231731

[0040] Preferably, according to the first dynamic control sub-matrix, the ethylene oxide loss (e.g., in ppm) is preferably associated, but not limited to, the lean absorbent flow as available handle or manipulated variable, the differential pressure in the absorber unit (e.g., in kPa) is preferably associated, but not limited to, the lean absorbent flow available handle or manipulated variable, the differential pressure over a lean absorbent valve (e.g., in kPa) is preferably associated, but not limited to, the lean absorbent flow as available handle or manipulated variable, the temperature in a top section of the stripper unit (e.g., in °C) is preferably associated, but not limited to, the lean absorbent flow as available handle or manipulated variable, the temperature in a middle section of the stripper unit (e.g., in °C) is preferably associated, but not limited to, the stripper temperature as available handle or manipulated variable, the temperature in a bottom section of the stripper unit (e.g., in °C) is preferably associated, but not limited to, the stripper temperature as available handle or manipulated variable, the ethylene oxide content in the bottom section of the stripper (e.g., in ppm) is preferably associated, but not limited to, the stripper temperature as available handle or manipulated variable, the inert limit (e.g., in kPa) is preferably associated, but not limited to, the residual absorber pressure as available handle or manipulated variable, the opening of stripper steam valve (e.g., in %) is preferably associated, but not limited to, the residual absorber pressure, the stripper temperature and the lean absorbent flow as available handles or manipulated variables, the differential pressure in the stripper unit (e.g., in kPa) is preferably associated, but not limited to, the lean absorbent flow and to the stripper temperature as available handles or manipulated variables, the density of glycol (e.g., in kg / m3) is preferably associated, but not limited to. the glycol bleed flow as available handle or manipulated variable and the level of demineralized water tank is preferably associated, but not limited to, the lean absorbent flow as available handle or manipulated variable.

[0041] Preferably, the sub-controller, or the process controller unit when provided, is configured to minimize the lean absorbent flow (e.g., in tons / hour) as much as possible, for example until a constraint is reached, to minimize the stripper temperature (e.g., until a constraint is reached), to keep the level of the lean absorbent tank in balance, to minimize the residual EO absorber pressure (e.g., until a constraint is reached) and to minimally vary the glycol bleed flow, or in other words, to keep this manipulated variable constant unless a variation is required to keep a controlled variable in a desired range.

[0042] In another embodiment, the second current variable data includes data pertaining to one or more of reactor inlet oxygen concentration (O2inlet in vol%), distance to flammable limit at reactor inlet (margin inlet), reactor outlet oxygen concentration at reactor outlet (O2outlet in vol%), distance to flammable limit at reactor outlet (margin outlet), oxygen conversion, inlet temperature, loopgas (recycled gas) pressure, pressure of a methane system, loopgas BASF SE 231731 flow, gasflow to CO2absorber, differential reactor coolant steam drum pressure, oxygen difference between reactor inlet and outlet, opening of steam drum pressure valve, and differential pressure of CO2absorber, preferably as values of second controlled variables.

[0043] Additionally, or alternatively, the second current variable data includes data pertaining to one or more of steam drum pressure, methane flow, purge gas flow, pressure at a reaction gas cooler steam drum located between the reactor unit and the EO absorber unit, opening of a compressor guide vane for recycle gas flow, and opening of CO2bypass valve, preferably as values of second manipulated variables.

[0044] Preferably, the second current variable data includes data pertaining to one or more of oxygen flow, ethylene concentration, and loopgas temperature, preferably as values of second feedforward variables.

[0045] Preferably, according to the second dynamic control sub-matrix, the reactor inlet oxygen concentration inlet (e.g., in vol%) is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the margin inlet is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the oxygen outlet (e.g., in vol%) is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the margin outlet is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the oxygen conversion (e.g., in vol%) is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the inlet temperature (e.g., in °C) is preferably associated, but not limited to, the drum pressure at B1 170 as available handle or manipulated variable, the loopgas pressure (e.g., in MPa) is preferably associated, but not limited to, the methane flow and the purge flow as available handle or manipulated variable, the pressure of a methane system (e.g., in MPa) is preferably associated, but not limited to, the methane flow as available handle or manipulated variable, the loopgas flow (e.g., in kNm3 / h) is preferably associated, but not limited to, the recycle gas compressor guide vane and bypass valve over CO2absorber as available handle or manipulated variable, the gasflow to CO2(e.g., in kNm3 / h) is preferably associated but not limited to, the guide vane and bypass valve over CO2absorber as available handle or manipulated variable, the differential drum pressure (e.g., in MPa) is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the differential oxygen reactor outlet (e.g., in vol%) is preferably associated, but not limited to, the drum pressures as available handle or manipulated variable, the opening of drum pressure valve at the drum located between the reactor unit and the EO absorber unit (e.g., in %) is preferably associated, but not limited to, the drum pressure at the drum located between the reactor unit and the EO BASF SE 231731 absorber unit as available handle or manipulated variable, and the differential pressure of CO2absorber (e.g., in kPa) is preferably associated, but not limited to, the loop gas flow as available handle or manipulated variable.

[0046] Preferably, the sub-controller, or the process controller unit when provided, is configured to minimize the drum pressures, the methane flow, the purge flow the pressure at the drum located between the reactor unit and the residual EO absorber unit and the opening of CO2bypass valve as much as possible (e.g., until a constraint is reached), and to maximize the opening of guidevane for recycle gas flow as much as possible (e.g., until a constraint is reached).

[0047] In yet another embodiment, a subset of the first and / or of the second controlled variables, and, preferably, a subset of the first and / or the second manipulated variables are critical variables, and wherein the recovery sub-controller unit and / or the reaction sub-controller unit are configured to provide operation instructions for stopping operation of the linear model based predictive controller unit, upon determining that at least one critical variable is not available for the recovery sub-controller unit and / or for the reaction sub-controller unit, or, in other words, when the value of at least one critical variable has not been received, and / or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range. For instance, when the value of at least one critical variable has not been received or has a value outside the corresponding validity range, the generated operation instructions can control the operation of the linear model based predictive controller unit such that the linear model based predictive controller unit is temporarily switched off, or cannot be switched on, for instance during a ramp up process. It is noted that the operation of the linear model based predictive controller unit can be stopped if a value of at least one of the critical value is not received or has a value outside an expected range, but the chemical production plant can continue its operation. Optionally, an alert signal can be provided, when the value of at least one critical variable has not been received, and / or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range, so that an operator is informed and the cause of the problem can be analyzed and solved before restarting operation of the linear model based predictive controller unit.

[0048] In a particular embodiment, the critical variables include one or more of the ethylene oxide loss, the differential pressure in the absorber unit, the differential pressure over a lean absorbent valve, the temperature in a top section of the stripper unit, the temperature in a bottom section of the stripper unit, opening of stripper steam valve, the differential pressure in the stripper unit, the level of demineralized water tank, the lean absorbent flow, the resid- BASF SE 231731 ual EO absorber pressure, the reactor inlet oxygen concentration (oxygen inlet, e.g. in vol%), the distance to flammable limit at reactor inlet (margin inlet), the reactor outlet oxygen concentration (oxygen outlet, e.g. in vol%), the distance to flammable limit at reactor outlet (margin outlet(, the oxygen conversion, the loopgas pressure, the oxygen concentration difference between reactor inlet and outlet, the reactor steam drum pressure, the methane flow, and the purge gas flow.

[0049] According to a second aspect of the invention, a computer implemented method for controlling operation of a chemical production plant for producing ethylene oxide in a production process is disclosed. The method comprises:

[0050] - receiving, at a recovery sub-controller unit of a linear model based predictive controller unit, first current variable data indicative of a current operational state of the recovery unit, and, at a reaction sub-controller unit of the linear model based predictive controller unit, second current variable data indicative of a current operational state of a reaction unit,

[0051] - determining and providing, based on a predetermined first dynamic control sub-matrix, respective setpoint values of one or more first manipulated variables associated to a recovery unit, and, based on a predetermined second dynamic control sub-matrix, respective setpoint values of one or more second manipulated variables associated to the reaction unit; and

[0052] - generating and providing first operation instructions for controlling operation of the recovery unit in accordance with the determined setpoint values, and second operation instructions for controlling operation of the reaction unit in accordance with the determined setpoint values.

[0053] The method of the second aspect of the invention thus shares the advantages of the chemical production plant of the first aspect or of any of its embodiments.

[0054] In the following, embodiments of the method of the second aspect will be disclosed.

[0055] In an embodiment, the first current variable data comprises

[0056] - first current manipulated variable data indicative of a current value of the respective first manipulated variables associated to the recovery unit; BASF SE 231731

[0057] - first current controlled variable data indicative of a current value of first controlled variables that are associated to the recovery unit, wherein the current values of the first controlled variables depend on the current values of one or more first manipulated variables in accordance with the first dynamic control sub-matrix; and,

[0058] - preferably, first current feedforward variable data indicative of a current value of respective first feedforward variables associated to the recovery unit that influence the values of the first controlled variables but are not adjustable by the recovery sub-controller unit.

[0059] Additionally, or alternatively, the second current variable data comprises:

[0060] - second current manipulated variable data indicative of a current value of the respective second manipulated variables associated to the reaction unit;

[0061] - second current controlled variable data indicative of a current value of second controlled variables that associated to the reaction unit; and,

[0062] - preferably, second current feedforward variable data indicative of a current value of respective second feedforward variables associated to the reaction unit that influence the values of the second controlled variables but are not adjustable by the reaction sub-controller unit.

[0063] Optionally, the method further comprises receiving, at the recovery sub-controller unit and / or at the reaction sub-controller unit, constraint data indicative of allowed operational ranges for one or more controlled variables and / or for one or more manipulated variables, and determining the setpoint values further in dependence on the constraint data.

[0064] Preferably, in an embodiment, the step of receiving the first and second current variable data, and determining the setpoint values of the first manipulated variables and the second manipulated variables is performed at regular time intervals, for instance once per minute.

[0065] In yet another embodiment, the method comprises: defining a subset of the first and / or of the second controlled variables, and, preferably, a subset of the first and / or the second manipulated variables as critical variables, and BASF SE 231731 providing operation instructions for stopping operation of the linear model based predictive controller unit, upon determining that at least one critical variable is not available for the recovery sub-controller unit and / or the reaction sub-controller unit and / or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range.

[0066] A third aspect of the present invention is formed by a computer program comprising instructions that, when executed by a linear model based predictive controller unit of a chemical production plant causes the linear model based predictive controller unit to carry out the method of the second aspect of the invention.

[0067] Another aspect of the invention refers to a linear model based predictive controller unit for use in a chemical production plant for producing EO according to the first aspect of the invention. The linear model based predictive controller includes a recovery sub-controller unit that is configured to receive first current variable data indicative of a current operational state of a recovery unit, and based on a predetermined first dynamic control sub-matrix, to determine and provide respective setpoint values of one or more first manipulated variables associated to the recovery unit, and to generate and provide first operation instructions for controlling operation of the recovery unit in accordance with the determined setpoint values. The linear model based predictive controller unit also includes a reaction sub-controller unit that is configured to receive second current variable data indicative of a current operational state of the reaction unit of the chemical production plant for producing EO, and based on a predetermined second dynamic control sub-matrix, to determine and provide respective setpoint values of one or more second manipulated variables associated to the reaction unit, and to generate and provide second operation instructions for controlling operation of the reaction unit in accordance with the determined setpoint values.

[0068] It shall be understood that the method described above, the device described above, the arrangement described above and the computer program product described above have similar and / or identical preferred embodiments, in particular, as defined in the dependent claims.

[0069] It shall be understood that a preferred embodiment of the present invention can also be any combination of the dependent claim or above embodiments with a respective independent claim. BASF SE 231731

[0070] These and other aspects of the present invention will be apparent from and elucidated with reference to the embodiments described hereafter.

[0071] BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In the following drawings:

[0073] Fig. 1 a schematic block diagram of an exemplary chemical production plant in accordance with an embodiment of the invention;

[0074] Fig. 2 an exemplary dynamic control matrix describing the effects of lean absorbent flow, guidevane position and lean absorbent temperature on the EO losses in the top of the absorber, dP over the lean absorbent valve and dP of the absorber column.

[0075] Fig. 3 a schematic block diagram depicting a control of the chemical production plant according to an embodiment of the invention;

[0076] Fig. 4 a diagram depicting an exemplary chemical production plant in accordance with an embodiment of the invention, showing the position of sensor for determining the values of manipulated, controlled and feedforwards variables;

[0077] Fig. 5 a flow diagram of an exemplary method for controlling operation of a chemical production plant in accordance with the invention.

[0078] DETAILED DESCRIPTION OF THE EMBODIMENTS

[0079] Fig. 1 shows a schematic block diagram of main parts of an exemplary chemical production plant in accordance with an embodiment of the invention. The chemical production plant 100 is suitable for producing ethylene oxide (EO) in a production process. The chemical production plant 100 comprises a reaction unit 102 and a recovery unit 1 12 connected to the reaction unit 102. The reaction unit 102 includes an ethylene input unit 104 for receiving ethylene C2H4, an oxygen input unit 106 for receiving oxygen O2, for instance in the form of compressed dried air or high purity oxygen, and a reactor assembly 108 for generating and providing a reaction stream 1 10 comprising, among other compounds, ethylene oxide EO. The ethylene stream and the oxygen stream are mixed with a gaseous mixture 105 coming from the recovery unit 1 12 and that contains ethylene and EO, among other BASF SE 231731 gaseous compounds. The reactor assembly 108 may comprise a catalytic fixed-bed tubular reactor that uses silver catalyst supported on alumina, where three main reactions occur, namely ethylene epoxidation (partial oxidation of ethylene to obtain ethylene oxide), ethylene combustion (oxidation of ethylene to obtain carbon dioxide and water), and EO oxidation (oxidation of EO to obtain carbon dioxide and water).

[0080] The recovery unit 1 12 comprises an absorber unit 1 14 that is configured to receive the reaction stream 1 10 containing EO and a lean absorbent stream 1 16, and to provide a rich absorbent stream 1 18 comprising a higher ethylene oxide EO content than that of the lean absorbent stream 1 16. The absorber unit is connected to a stripper unit 120 that is configured to receive the rich absorbent stream 1 18, to strip off or recover ethylene oxide (EO) from the rich absorbent stream 1 18, thereby generating the lean absorbent stream 1 16, to provide the recovered ethylene oxide EO via an ethylene oxide output unit 122 and to provide the generated lean absorbent stream 1 16 to the absorber unit 1 14.

[0081] In order to monitor and control the production process, the chemical production plant 100 further comprises a linear model based predictive controller (MPC) unit 150 that includes a recovery sub-controller unit 152 and a reaction sub-controller unit 154. In general, a model of the chemical production plant is the basis of the MPC unit 150. Based on this model, the MPC unit is able to predict the behavior of the plant in the near future. Then, based on these predictions, the MPC unit is configured to determine how to take action to bring the chemical production plant 100 to a desired, preferably optimal, working point.

[0082] The recovery sub-controller unit 152 is configured to receive first current variable data VD1 indicative of a current operational state of the recovery unit 1 12. The first current variable data can be for instance provided by dedicated sensors, for instance as part of one or more proportional-integral-derivative (PID) controllers. Based on a predetermined first dynamic control sub-matrix DCM1 , the recovery sub-controller unit is configured to determine and provide respective setpoint values SP1 of one or more first manipulated variables MV1 associated to the recovery unit 1 12. The dynamic control sub-matrix DCM1 includes models that describe the interaction between different related variables in the recovery unit.

[0083] Analogously, the reaction sub-controller unit 154 is configured to receive second current variable data VD2 indicative of a current operational state of the reaction unit 102. Here, again, the second variable data VD2 can be for instance provided by dedicated sensors, for instance as part of one or more proportional-integral-derivative (PID) controllers. Based on a predetermined second dynamic control sub-matrix DCM2, the reaction sub-controller unit 154 is configured to determine and provide respective setpoint values SP2 of one or BASF SE 231731 more second manipulated variables MV2 associated to the reaction unit 102. The dynamic control sub-matrix DCM2 includes models that describe the interaction between different related variables in the reaction unit.

[0084] The term manipulated variable refers to a variable whose value can be manipulated by the MPC unit, to get the chemical production plant closer to the optimal or desired working point. The MPC unit calculates, based on the respective dynamic control sub-matrix, new setpoints SP1 , SP2 for the manipulated variables. Manipulated variables are sometimes referred to as independents.

[0085] The linear model based predictive controller (MPC) unit 150 also comprises a process controller unit 155 that is configured to receive the respective setpoint values SP1 SP2 of the first manipulated variables MV1 and of the second manipulated variables MV2, and to generate and provide first operation instructions OP1 for controlling operation of the recovery unit 1 12 in accordance with the determined setpoint values SP1 and second operation instructions OP2 for controlling operation of the reaction unit 102 in accordance with the determined setpoint values SP2.

[0086] The process controller unit 155 is shown as an independent unit in Fig. 1 . However, the controller unit can also be integrated as part of the recovery sub-controller unit 152 and the reaction sub-controller unit 154, such that the respective sub-controller units 152, 154 can be configured to provide the corresponding operation instructions OP1 , OP2 for controlling operation of the corresponding one of the recovery unit 1 12 and the reaction unit 108 in accordance with the determined setpoints. For example, the operation instructions can be indicative of, or otherwise refer to, the determined setpoint value of a given flow, and the operation instruction is provided to the respective unit 1 12, 102, for operating the hardware element, e.g. a suitable valve, that controls said flow. In this exemplary case, based on the provided setpoint value, a valve is opened or closed to meet the desired flow as given by the setpoint value determined by the linear model based predictive controller unit 150.

[0087] In the exemplary chemical production plant 100, the first current variable data VD1 comprises first current manipulated variable data MVD1 that is indicative of a current value of the respective first manipulated variables MV1 associated to the recovery unit 1 12. The first current variable data VD1 also includes first current controlled variable data CVD1 indicative of a current value of first controlled variables CV1 that are associated to the recovery unit 1 12. Controlled variables are variables whose values depend on the values of the manipulated variables. Thus, the current values of the first controlled variables CV1 BASF SE 231731 depend on the current values of one or more first manipulated variables MV1 , in accordance with the first dynamic control sub-matrix DCM1 .

[0088] Preferably, the first current variable data also includes first current feedforward variable data FVD1 that is indicative of a current value of respective first feedforward variables FV1 associated to the recovery unit 1 12 that influence the values of the first controlled variables CV1 but are not adjustable by the sub-controller units 152, 154 or by the process controller unit 155, when provided. Feedforward variables are variables that cannot be manipulated by the MPC unit 150 but that nevertheless have an influence on the controlled variables. A feedforward variable acts as a measured disturbance on the controlled variables. By taking into account this measured disturbance or feedforward variable, the MPC unit 150 can control more accurately the controlled variables.

[0089] The values of the corresponding first variables (manipulated, controlled and feedforward variables) can be provided to the recovery sub-controller unit 152 by a dedicated sensor, or can be otherwise ascertained by the recovery sub-controller unit 152.

[0090] Analogously, in the exemplary chemical production plant 100, the second current variable data VD2 comprises second current manipulated variable data MVD2 indicative of a current value of the respective second manipulated variables MV2 associated to the reaction unit 102, second current controlled variable data CVD2 indicative of a current value of second controlled variables CV2 that are associated to the reaction unit 102, wherein the current values of the second controlled variables CV2 depend on the current values of one or more second manipulated variables MV2 in accordance with the second dynamic control submatrix DCM2; and, preferably, second current feedforward variable data FVD2 indicative of a current value of respective second feedforward variables FV2 associated to the reaction unit 102 that influence the values of the second controlled variables CV2 but are not adjustable by the sub-controller units 152, 154, or by the process controller unit 150, when provided.

[0091] The values of the corresponding second variables (manipulated, controlled and feedforward variables) can be provided to the reaction sub-controller unit 154 by a dedicated sensor, or can be otherwise ascertained by the reaction sub-controller unit 154.

[0092] Optionally, the recovery sub-controller unit 152 and / or the reaction sub-controller unit 154 are further configured to receive constraint data CD1 , CD2 indicative of allowed operational ranges for one or more controlled variables CV1 , CV2 and / or for one or more manipulated BASF SE 231731 variables MV1 , MV2. Typically, the optimal working point of the plant is determined by plant constraints. These constraints can be defined by equipment limitations, safety limits or specification limits. For example, the constraints can be defined by setting the operation limits of the manipulated variables and the controlled variables associates to the reaction unit 108 and the recovery unit 1 12, in the form of corresponding constraint data CD1 , CD2.

[0093] Fig. 2 shows an exemplary dynamic control sub-matrix describing the effects of lean absorbent flow, guidevane position and lean absorbent temperature on the EO losses in the top of the absorber, dP over the lean absorbent valve and dP of the absorber column. This is an example of a simplified dynamic control sub-matrix used in a recovery sub-controller unit 152, in particular for the absorber unit 1 14. For instance, in the top left corner, the effect of the lean absorbent flow, which is a manipulated variable, on the EO losses in the top of the absorber unit 1 14, which is a controlled variable, is shown. The model curve indicates that if the lean absorbent flow is increased, the EO losses decrease. In the top mid-section, the effect of the lean absorbent flow (manipulated variable) on the differential pressure in the absorber unit 1 14 (controlled variable), is shown. The model curve indicates that if the lean absorbent flow is increased, the differential pressure at the EO absorber will increase too.

[0094] In the following a simplified example of the strategy applied by the MPC unit 150 to solve a constraint problem related to the part of the model of the EO absorber unit described above with regard to the dynamic control matrix of Fig. 2 is discussed.

[0095] Assuming a current situation where:

[0096] - the EO losses are at the maximum value;

[0097] - the differential pressure of the lean absorbent valve is at the minimum value (to avoid that the lean absorbent pump trips the differential pressure of the lean absorbent, it should stay high enough)

[0098] - the ambient temperature is increasing and thus the lean absorbent temperature (feedforward variable) as well.

[0099] If the MPC unit does not take any action, the EO losses would increase further is not acceptable because it takes the production plant away from the optimal working point. BASF SE 231731

[0100] In case the MPC unit provides operation instructions for further increasing the lean absorbent flow to avoid that the EO losses become too high, the differential pressure over the valve would become too low. Thus, this is also not a suitable action according to the model.

[0101] According to the dynamic control matrix shown in Fig. 2, the MPC unit 150 is advantageously configured to provide operation instructions for decreasing the gas flow to the EO absorber unit. This will allow the MPC 150 to keep both the EO losses at the maximum limit and the differential pressure of the lean absorbent valve at the low limit.

[0102] In principle, the general objectives for the MPC unit 150 are minimizing raw material consumption and minimizing specific energy consumption. The raw material consumption is minimized by controlling the process to operate as close as possible against the reactor inlet or outlet flame limit and by increasing the gas hourly space velocity GHSV and maximizing the gas flow towards the EO absorber. The specific energy consumption will be minimized by decreasing the steam consumption in the EO stripper by decreasing the lean absorbent flow while maintaining the EO slip in the top of the EO absorber.

[0103] As explained with reference to Fig. 1 , the MPC unit 150 includes two sub-controller units, each associated to a group of manipulated variables and controlled variables.

[0104] The recovery sub-controller unit is associated to variables of the EO absorber unit and the stripper unit. The main objectives of the recovery sub-controller unit are to optimize the EO concentration in the top absorber by minimizing the lean absorbent flow and maximizing the loopgas (recycled gas) flow. In winter conditions, e.g., when sufficient cooling capacity is available to cool down the lean absorbent flow, the EO concentration is pushed towards the lower limit (e.g., 50ppm). Here, the loopgas flow will be pushed towards its maximum limit and the lean absorbent flow will be maximized until the lower limit of the EO slip in the EO concentration is reached. In summer conditions, e.g., when sufficient cooling capacity is not available to cool down the lean absorbent flow, the EO concentrations increases up to the high limit (e.g., 100 ppm). Here, the lean absorbent flow will be increased until the maximum limit is reached and subsequently, the MPC unit will slowly reduce the loopgas flow in order to maintain the EO concentration in the top of the absorber at the high limit.

[0105] Other objectives include:

[0106] - minimizing steam consumption in the EO stripper by reducing the residual absorber pressure as far as possible until the inert margin in the surge drum is hit, mainly in summer BASF SE 231731 conditions, and / or by decreasing the temperature profile in the stripper column as much as possible;

[0107] - keeping the glycol concentration (density) in a zone by only adjusting the glycol bleed flow if the density is out of its defined zone or range; and

[0108] - balancing the pressure level of demineralized water by manipulating the demineralized water flow.

[0109] In turn, the reaction sub-controller unit 154 is associated to the variables of the reaction unit 108. Also purge flow and CH4flow can be included, since these variables also influence the loopgas pressure. The main objectives of the reaction sub-controller unit are:

[0110] - stabilizing the reaction unit in order to be able to optimize oxygen concentrations at the inlet / outlet of the reactor. Optimizing the oxygen concentration in order to increase selectivity can be achieved by minimizing the steam drum pressure until the O2concentration at the inlet or the outlet reaches the upper limit, bringing the distance to the flammable limit inlet / outlet to the lower limit and / or when the oxygen conversions hit the upper / lower limit. It has to be taken into account that the oxygen concentrations at the outlet of the reactors should not deviate too much to avoid that conversions in the reactors differ too much, and that the drum pressures should not deviate too much to avoid that reactors run at activity with big difference. Also, the oxygen concentration can be optimized by minimizing the steam drum pressure until the inlet temperatures hit the lower limit (lower inlet temperature will lead to higher potential oxygen concentration at the inlet of the reactor - which is beneficial if the inlet margin is the constraint), and / or the pressure valve hits the high limit (indicating the differential pressure over the pressure valve becomes too small).

[0111] - Maintaining the loopgas pressure within well-defined low limit and high limit by increasing the methane flow if the pressure becomes lower than the low limit, increasing the purge flow if the pressure becomes higher than the high limit

[0112] - Maximizing loopgas flow. In winter (no cooling constraints) the loopgas flow will be at the high limit and in summer (cooling constraints) the loopgas flow will be slowly reduced due to the EO slip in the top of absorber unit. At the same time the gasflow to the CO2absorber will be kept constant at the maximum limit (unless the differential pressure of the CO2increases above the high limit, then the gasflow to the CO2absorber will be decreased). In principle, this optional part of the reaction sub-controller interacts with the variables of the BASF SE 231731 recovery sub-controller via the EO slip in the top of the EO absorber. This should be taken into account if one of the sub-controllers is switched off.

[0113] Fig. 3 shows a schematic block diagram depicting a control of a chemical production plant, such as plant 100 of Fig. 1 . The chemical production plant includes a plurality of sensors 160, for instance as part of one or more proportional-integral-derivative (PID) controllers, configured to determine and provide variable data VD1 , VD2 indicative of values of variables of the production process for producing EO. The exemplary MPC unit 150 can run on a computing device such as, for example, a personal computer, and it is preferably configured to perform the following steps at regular time intervals, for example, every minute.

[0114] At the regular interval, the MPC unit is configures to receive or otherwise ascertain the values of the variable data, e.g., the measured values of the controlled variables (first and second), the manipulated variables (first and second) and the feedforward variables (first and second) from the sensors. These values are fed to the corresponding dynamic control sub-matrices DCM1 and DMC2, which are configured to provide a prediction of where the plant is going to in the near future (e.g. 4 hours) given its current state (as indicated by the received variable data VD1 , VD2). In an optimization step, indicated by 01 and 02, the prediction and the optimal working point that can be achieved are combined, in particular in dependence on possible constraints provided as constraint data CD1 , CD2, for instance by an operator. Based on the prediction of the behavior of the plant in the near future and the knowledge of the optimal working point, setpoints SP1 SP2 for one or more manipulated variables are determined and operation instructions in order to push the plant towards the optimal point, which consist of the determined setpoints, are provided to the PID controllers. The operation instructions OP1 , OP2 affect the manipulated variables so that the values of the related controlled variables are pushed such that the optimal working point is reached. Thus, the MPC unit provides instructions indicative of the new setpoints for the PID controllers, that is to control the manipulated variables, included valve outputs.

[0115] By changing the operator limits of the manipulated and / or controlled variables, the operator is able to change the constraints defined for the controller and thus the operating range of the controller.

[0116] Fig. 4 shows a more detailed diagram depicting an exemplary chemical production plant 100 in accordance with an embodiment of the invention. The diagram shows the position of sensors for determining the values of manipulated, controlled and feedforwards variables. The variables will be briefly explained in the following: BASF SE 231731

[0117] For the recovery sub-controller unit, the significant variables include, as controlled variables:

[0118] - ethylene oxide loss CV01 . This is considered as a critical variable and the available manipulated variables are the lean absorbent flow MV01 and the loopgas flow CV26. In winter (sufficient cooling capacity) the MPC unit is configured to keep the EO losses at the lower limit by adjusting the lean absorbent flow (gasflow is at maximum); in summer (not sufficient cooling capacity) MPC unit is configured to keep the EO losses below the upper limit by reducing the loopgasflow to the absorber.

[0119] - Differential pressure in the absorber unit CV02. This can optionally be considered as a critical variable and the available manipulated variables are the lean absorbent flow MV01 and the loopgas flow CV26. It should be avoided that the differential pressure over the absorber exceeds an upper limit.

[0120] - Differential pressure over a lean absorbent valve CV03. This can optionally be considered as a critical variable and the available manipulated variable is the lean absorbent flow MV01 . It should be avoided that the differential pressure over the lean absorbent valve becomes too low (and might cause trip of the lean absorbent pump).

[0121] - Temperature in a top section of the stripper unit CV04. This can optionally be considered as a critical variable, and the available manipulated variable is the lean absorbent flow MV01 . It should be avoided that the EO concentration in the top of the EO stripper becomes too high, the top temperature should not decrease too much; this temperature is pressure compensated and thus the temperature used by the MPC unit deviates slightly from the one provided by the sensor. From the top of the EO stripper a stream -labeled as “Raw EO” in Fig. 4- can be extracted. The raw EO stream can be for instance transferred to an EO purification stage and / or to reaction section where the raw EO is converted to form ethylene glycols. This section can be referred to as glycol section and therein mono ethylene glycol (MEG) and / or diethylene glycol (DEG) and / or triethylene glycol (TEG) can be formed.

[0122] - Temperature in a middle section of the stripper unit CV05. The available manipulated variable is the stripper temperature MV02. If this temperature would become too low EO would breakthrough in the bottom of the EO stripper. It should be avoided as well that excessive steam is added to the EO stripper when the temperature at the bottom of the stripper (see CV06) is slightly below its lower limit. BASF SE 231731

[0123] - Temperature in a bottom section of the stripper unit CV06. This can optionally be considered as a critical variable, and the available manipulated variable is the stripper temperature MV02. If this temperature becomes too low EO would breakthrough to the lean absorbent. This temperature is pressure compensated and thus that the temperature used by the MPC unit deviates slightly from the temperature provided by the sensor.

[0124] - Ethylene oxide content in the bottom section of the stripper CV07. The available manipulated variable is the stripper temperature MV02. It should be avoided that the EO slip in the bottom of the EO stripper becomes too high; this is regarded as the main constraint to reduce the EO stripper temperature profile.

[0125] - Inert limit CV08. The available manipulated variable is the residual absorber pressure (see MV04). It should be avoided that the inert margin becomes too low.

[0126] - Opening of stripper steam valve CV09. This can optionally be considered as a critical variable, and the available manipulated variables are the residual absorber pressure MV04 and the lean absorbent flow MV01 . If the pressure in the EO stripper is increased too much steam is not condensed sufficiently in the reboiler. To avoid this the valve position of the steam flow valve has to be checked.

[0127] - Differential pressure in the stripper unit CV10. This can optionally be considered as a critical variable, and the available manipulated variables are the stripper temperature MV02 and the lean absorbent flow MV01 . It should be avoided that the differential pressure of the EO stripper becomes too high.

[0128] - Density of glycol CV1 1 . The available manipulated variable is the glycol bleed flow (see MV05). The density of the glycol should be kept within a given range.

[0129] - Level of demineralized water tank CV12. This can optionally be considered as a critical variable, and the available manipulated variable is the demineralized water flow (see MV03). The level of the demineralized water tank should be kept in balance around the setpoint.

[0130] Also, for the recovery sub-controller unit, the significant variables include, as manipulated variables: BASF SE 231731

[0131] - lean absorbent flow MV01 . This can optionally be considered as a critical variable. In winter MPC unit is configured to maximize the lean absorbent flow until the EO losses hit the lower limit: this is the minimum lean absorbent flow required to keep the EO losses at the minimum limit (this will save steam in the EO stripper). In summer the lean absorbent flow will increase until the max limit (as the EO losses will increase) and the controller will keep the EO losses around the upper limit by decreasing loopgas flow.

[0132] - Stripper temperature MV02. The temperature profile in the EO stripper should be reduced as much as possible in order to save steam (but avoid too high EO concentration in the top (top temperature) or in the bottom (bottom temperature)).

[0133] - Demineralized water flow MV03. Used to keep the level of the demineralized water tank in balance.

[0134] - Residual absorber pressure MV04. This can optionally be considered as a critical variable. Used to reduce the residual absorber pressure as much as possible (to reduce the pressure in the stripper and reduce steam consumption). The main limit for reducing pressure is the inter margin in the surge vessel.

[0135] - Glycol bleed flow MV05. The glycol bleed flow is only changed if the glycol concentration is outside the zone defined by the operator.

[0136] Additionally, data pertaining to one or more of a lean absorbent temperature, opening of a process water valve, lean absorber to residual absorbent flow and an ethylene oxide flow to the stripper unit, can be provided as values of first feedforward variables.

[0137] For the reaction sub-controller unit, the significant variables include, as controlled variables:

[0138] - oxygen inlet CV13. This can optionally be considered as a critical variable and the available manipulated variables are the drum pressures (see MV06 and MCV07). The oxygen concentration should be kept as high as possible but below the operator limit to increase selectivity.

[0139] - Margin inlets CV14, CV15. These can optionally be considered as a critical variables and the available manipulated variables are the drum pressures (see MV06 and MCV07). The inlet margins should be kept as low as possible but below the operator limit to increase selectivity. BASF SE 231731

[0140] - Oxygen outlets CV16, CV17. These can optionally be considered as a critical variables and the available manipulated variables are the drum pressures (see MV06 and MCV07). The oxygen concentration should be kept as high as possible but below the operator limit to increase selectivity.

[0141] - Margin outlets CV18, CV19. These can optionally be considered as a critical variables and the available manipulated variables are the drum pressures (see MV06 and MCV07). The outlet margins should be kept as low as possible but below the operator limit to increase selectivity.

[0142] - Oxygen conversions CV20, CV21 . These can optionally be considered as a critical variables and the available manipulated variables are the drum pressures (see MV06 and MCV07). The oxygen conversions should be kept within a zone or range delimited by an upper and a lower limit.

[0143] - Inlet temperatures CV22, CV23. The available manipulated variable is the drum pressure at the steamdrum located between the reactor and the EO absorber (see MV10). The inlet temperatures should be kept as low as possible.

[0144] - Loopgas pressure CV24. This can optionally be considered as a critical variable and the available manipulated variables are the CH4flow (see MV08) and the purge flow (see MV09). The loopgas pressure should be kept within a zone / range.

[0145] - Pressure of a methane system CV25. The available manipulated variable is the methane flow (see MV08). It should be avoided that the pressure in the CH4system becomes too low (and trips the CH4supply).

[0146] - Loopgas flow CV26. The available manipulated variables are the opening of guidevane for gas flow (see MV1 1 ) and the opening of the CO2bypass valve (see MV12). The loopgas flow should be maximized (until the max limit is hit)

[0147] - Gas flow to CO2CV27. The available manipulated variables are the opening of guidevane for gas flow (see MV1 1 ) and the opening of the CO2bypass valve (see MV12). The gas flow to the CO2absorber should be maximized until the max limit is hit; as a remark, increasing gas flow above this high limit would only increase H2O carry over resulting in decreased selectivity. BASF SE 231731

[0148] - Differential drum pressure CV28. The available manipulated variables are the drum pressures (see MV06 and MCV07). It should be avoided that the drum pressures of both steam drums deviate too much from each other in order to keep reaction severity in both reactors in pace with each other.

[0149] - Differential oxygen reactor outlet CV29. This can optionally be considered as a critical variable and the available manipulated variables are the drum pressures (see MV06 and MCV07). It should be avoided that the O2concentration at the outlet of the reactors would deviate too much from each other in order to keep reaction severity in both reactors in pace with each other.

[0150] - Opening of drum pressure valve of the steamdrum located between the reactor and the EO absorber CV30. The available manipulated variable is the drum pressure at the steamdrum located between the reactor and the EO absorber (see MV10). It should be avoided that the drum pressure at said steamdrum decreases too much in order to keep sufficient differential pressure over the steamdrum valve to keep it in control.

[0151] - Differential pressure of CO2absorber CV31 . The available manipulated variable is the opening of CO2bypass valve (see MV12). It should be avoided that the differential pressure of the CO2absorber becomes too high. The CO2absorber is part of a CO2removal section of the exemplary chemical production plant 100, as indicated in Fig. 4. The CO2removal section outputs CO2off gas as an outlet stream.

[0152] Also, for the reaction sub-controller unit, the significant variables include, as manipulated variables:

[0153] - drum pressures MV06, MV07, the drum pressures (run reactions at lowest temperatures to increase selectivity) should be minimized, The constraints for minimization are O2and margin limits at inlet and outlet of the reactors. These can optionally be considered as critical variables.

[0154] - Methane flow MV08. The CH4flow is used to keep the loopgas pressure above its lower limit; but try to keep the methane flow as low as possible. The lower limit for the methane flow is determined by the trip point of the CO2absorber (lower methane flow might trip the CO2absorber). If the control has opened the CH4-flow (above the lower limit) then the purge will be forced to be equal to its lower limit (normally 0Nm3 / h) to avoid that both purge flow BASF SE 231731 and CH4flow are opened at the same time. This can optionally be considered as a critical variable.

[0155] - Purge flow MV09. The purge flow is used to keep the loopgas pressure below its upper limit; but try to keep the purge flow as low as possible. Remark that if the MPC unit determines to open the purge flow, the CH4flow will be forced to its lower operator limit to avoid that the CH4flow and the purge flow are opened at the same time. This can optionally be considered as a critical variable.

[0156] - Pressure at a drum located between the reactor unit and the EO absorber unit MV10. The MPC unit is configured to minimize this pressure in order to decrease the inlet temperatures of the reactors slightly (higher oxygen concentration feasible). Main constraint is the valve position of the controller that should not become too high (indicating too low differential pressure over the valve.

[0157] - Opening of guidevane for gas flow MV1 1 . The MPC unit is configured to try to increase this guidevane position in order to increase the loopgas flow. Constraints are the upper limit of the gas flow and / or the EO slip in the top of the absorber. Remark that due to stiction in the valve this manipulated variable is configured to be an accumulated moves. The manipulated variable will accumulate small moves until a 1 % move is realized. Therefore the movements of this MV will be step-like (1 %).

[0158] - Opening of CO2bypass valve MV12. The MPC unit is configured try to minimize the valve position of the bypass over the CO2absorber and thus increase the gas flow to the CO2absorber. The main constraints are the upper limit of the gas flow to the CO2absorber and the differential pressure of the CO2absorber.

[0159] Additionally, data pertaining to one or more of oxygen flow, ethylene concentration, and loopgas temperature, preferably as values of second feedforward variables.

[0160] For the case of the controlled variable loopgas pressure CV24 and the manipulated variables purge flow MV09 and CH4flow MV08 a special consideration can be applied for an improved control. The loopgas pressure is defined as a ramp variable in the controller and as a consequence a setpoint can be defined for this variable. However the aim is no strict setpoint control for the loopgas pressure, but to keep the pressure in a zone around the setpoint: BASF SE 231731

[0161] - if the pressure is within the zone around the setpoint both the purge flow and the CH4flow will be kept at their respective lower limits and the pressure will float.

[0162] - If the pressure falls below the lower limit of the zone the CH4flow will start opening until the pressure is back in the zone. The purge flow will stay at the lower limit.

[0163] - If the pressure rises above the higher limit of the zone, the purge flow will start to open until the pressure is back within the zone around the setpoint and the CH4flow will remain at its lower limit.

[0164] Preferably, the gaps defined around the setpoint and the limits are all set to 0.005Mpag. These gaps should not be changed by the operator. The operator can only change the HI and LO limits of the loopgas pressure and the setpoint of the loopgas pressure to influence the zone around the setpoint.

[0165] The CH4flow and the purge flow are forced to the lower limit by putting internally the upper limit to the lower limit. This means that the operator cannot change the lower limit if the purge flow is forced to close or the CH4flow is forced to its lower limit since the lower limit cannot be higher than the higher operator limit. If the purge flow or the CH4flow are above their respective lower limits, then the lower limits can be increased (since the high limits are not set equal at that moment to the low limits).

[0166] The zone around the setpoint is defined by the setpoint value itself and by the settings of the operator HI and operator LO limits; around the setpoint a gap is defined and below the operator high and above the operator low limit also a gap is defined. The strictest of both gaps is taken as the lower limit and upper limit for the zone.

[0167] Optionally, upon determining that at least one critical variable (CRV) is not available for the recovery sub-controller unit and / or the reaction sub-controller unit, and / or or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range, the recovery sub-controller unit and / or the reaction sub-controller unit are configured to provide operation instructions for stopping operation of the linear model based predictive controller unit.

[0168] Fig. 5 shows a flow diagram of an exemplary method 200 for controlling operation of a chemical production plant in accordance with the invention. The computer implemented method 200 is suitable for controlling operation of a chemical production plant for producing BASF SE 231731 ethylene oxide (EO) in a production process. The method comprises, in a step 202, receiving, in particular at an recovery sub-controller unit 152 of a linear model based predictive controller unit 150, first current variable data VD1 indicative of a current operational state of the recovery unit 1 12, and, at a reaction sub-controller unit 154 of the linear model based predictive controller unit 150, second current variable data VD2 indicative of a current operational state of a reaction unit 102. The method also comprise, in a step 204, determining and providing, based on a predetermined first dynamic control sub-matrix DCM1 , respective setpoint values SP1 of one or more first manipulated variables MV1 associated to the recovery unit 1 12, and, based on a predetermined second dynamic control sub-matrix DCM2, respective setpoint values SP2 of one or more second manipulated variables MV2 associated to the reaction unit 102. The method further comprises, in a step 206, generating and providing first operation instructions OP1 for controlling operation of the recovery unit 1 12 in accordance with the determined setpoint values SP1 , and second operation instructions OP2 for controlling operation of the reaction unit 102 in accordance with the determined setpoint values SP2.

[0169] In an optional step 207, the method comprises providing operation instructions for stopping operation of the linear model based predictive controller unit, upon determining that at least one critical variable is not available for the recovery sub-controller unit and / or the reaction sub-controller unit or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range. The critical variables are a subset of the first and / or of the second controlled variables, and, preferably, a subset of the first and / or the second manipulated variables as critical variables, whose values are necessary to continue the control process.

[0170] In summary, the invention is directed to a chemical production plant for producing ethylene oxide, which comprises a reaction unit for providing a reaction stream comprising ethylene oxide and a recovery unit for recovering or extracting ethylene oxide. The chemical production plant further comprising a linear model based predictive controller unit that includes an recovery sub-controller unit and a reaction sub-controller unit, which based on respective current variable data and respective dynamic control sub-matrices, are configured to determine and provide respective setpoint values of one or more manipulated variables associated to the recovery unit or the reaction unit; and to generate and provide respective operation instructions for controlling operation of the recovery unit or the reaction unit in accordance with the determined setpoint values.

[0171] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. BASF SE 231731

[0172] A single unit or device may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0173] Procedure steps performed by one or several units or devices can be performed by any other number of units or devices. These procedures can be implemented as program code means of a computer program and / or as dedicated hardware.

[0174] A computer program product may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium, supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0175] Any units described herein may be processing units that are part of a classical computing system. Processing units may include a general-purpose processor and may also include a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or any other specialized circuit. Any memory may be a physical system memory, which may be volatile, non-volatile, or some combination of the two. The term “memory” may include any computer-readable storage media such as a non-volatile mass storage. If the computing system is distributed, the processing and / or memory capability may be distributed as well. The computing system may include multiple structures as “executable components”. The term “executable component” is a structure well understood in the field of computing as being a structure that can be software, hardware, or a combination thereof.

[0176] For instance, when implemented in software, one of ordinary skill in the art would understand that the structure of an executable component may include software objects, routines, methods, and so forth, that may be executed on the computing system. This may include both an executable component in the heap of a computing system, or on computer- readable storage media. The structure of the executable component may exist on a computer-readable medium such that, when interpreted by one or more processors of a computing system, e.g., by a processor thread, the computing system is caused to perform a function. Such structure may be computer readable directly by the processors, for instance, as is the case if the executable component were binary, or it may be structured to be interpretable and / or compiled, for instance, whether in a single stage or in multiple stages, so as to generate such binary that is directly interpretable by the processors. BASF SE 231731

[0177] In other instances, structures may be hard coded or hard wired logic gates, that are implemented exclusively or near-exclusively in hardware, such as within a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or any other specialized circuit. Accordingly, the term “executable component” is a term for a structure that is well understood by those of ordinary skill in the art of computing, whether implemented in software, hardware, or a combination. Any embodiments herein are described with reference to acts that are performed by one or more processing units of the computing system. If such acts are implemented in software, one or more processors direct the operation of the computing system in response to having executed computer-executable instructions that constitute an executable component. Computing system may also contain communication channels that allow the computing system to communicate with other computing systems over, for example, network.

[0178] A “network” is defined as one or more data links that enable the transport of electronic data between computing systems and / or modules and / or other electronic devices. When information is transferred or provided over a network or another communications connection, for example, either hardwired, wireless, or a combination of hardwired or wireless, to a computing system, the computing system properly views the connection as a transmission medium. Transmission media can include a network and / or data links which can be used to carry desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general-purpose or special-purpose computing system or combinations. While not all computing systems require a user interface, in some embodiments, the computing system includes a user interface system for use in interfacing with a user. User interfaces act as input or output mechanism to users for instance via displays.

[0179] Those skilled in the art will appreciate that at least parts of the invention may be practiced in network computing environments with many types of computing system configurations, including, personal computers, desktop computers, laptop computers, message processors, hand-held devices, multi-processor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, mobile telephones, PDAs, pagers, routers, switches, datacenters, wearables, such as glasses, and the like. The invention may also be practiced in distributed system environments where local and remote computing system, which are linked, for example, either by hardwired data links, wireless data links, or by a combination of hardwired and wireless data links, through a network, both perform tasks. In a distributed system environment, program modules may be located in both local and remote memory storage devices. | BASF SE | 231731

[0180] Those skilled in the art will also appreciate that at least parts of the invention may be practiced in a cloud computing environment. Cloud computing environments may be distributed, although this is not required. When distributed, cloud computing environments may be distributed internationally within an organization and / or have components possessed across multiple organizations. In this description and the following claims, “cloud computing” is defined as a model for enabling on-demand network access to a shared pool of configurable computing resources, e.g., networks, servers, storage, applications, and services. The definition of “cloud computing” is not limited to any of the other numerous advantages that can be obtained from such a model when deployed. The computing systems of the figures include various components or functional blocks that may implement the various embodiments disclosed herein as explained. The various components or functional blocks may be implemented on a local computing system or may be implemented on a distributed computing system that includes elements resident in the cloud or that implement aspects of cloud computing. The various components or functional blocks may be implemented as software, hardware, or a combination of software and hardware. The computing systems shown in the figures may include more or less than the components illustrated in the figures and some of the components may be combined as circumstances warrant.

[0181] Any reference signs in the claims should not be construed as limiting the scope.

Claims

BASF SE231731Claims:1 . Chemical production plant (100) for producing ethylene oxide (EO) in a production process, the chemical production plant comprising:- a reaction unit (102) comprising an ethylene input unit (104) for receiving ethylene, an oxygen input unit (106) for receiving oxygen and a reactor assembly (108) for generating and providing a reaction stream (1 10) comprising ethylene oxide (EO); and- a recovery unit (1 12) comprising an absorber unit (1 14) configured to receive the reaction stream (1 10) and a lean absorbent stream (1 16), and to provide a rich absorbent stream (1 18) comprising a higher ethylene oxide (EO) content than that of the lean absorbent stream (1 16); the recovery unit (1 12) further comprising a stripper unit (120) configured to receive the rich absorbent stream (1 18), to recover ethylene oxide (EO) from the rich absorbent stream (1 18) thereby generating the lean absorbent stream (1 16), to provide the recovered ethylene oxide (EO) via an ethylene oxide output unit (122) and to provide the generated lean absorbent stream (1 16) to the absorber unit (1 14); the chemical production plant (100) further comprising a linear model based predictive controller unit (150) that includes:- a recovery sub-controller unit (152) that is configured to receive first current variable data (VD1 ) indicative of a current operational state of the recovery unit (1 12), and based on a predetermined first dynamic control sub-matrix (DCM1 ), to determine and provide respective setpoint values (SP1 ) of one or more first manipulated variables (MV1 ) associated to the recovery unit (1 12); and to generate and provide first operation instructions (OP1 ) for controlling operation of the recovery unit (1 12) in accordance with the determined setpoint values (SP1 ); and- a reaction sub-controller unit (154) that is configured to receive second current variable data (VD2) indicative of a current operational state of the reaction unit (102), and based on a predetermined second dynamic control sub-matrix (DCM2), to determine and provide respective setpoint values (SP2) of one or more second manipulated variables (MV2) associated to the reaction unit (102); and to generate and provide second operation instructions (OP2) for controlling operation of the reaction unit (102) in accordance with the determined setpoint values (SP2).BASF SE2317312. The chemical production plant (100) of claim 1 , wherein the first current variable data (VD1 ) comprises:- first current manipulated variable data (MVD1 ) indicative of a current value of the respective first manipulated variables (MV1 ) associated to the recovery unit (1 12);- first current controlled variable data (CVD1 ) indicative of a current value of first controlled variables (CV1 ) that are associated to the recovery unit (1 12), wherein the current values of the first controlled variables (CV1 ) depend on the current values of one or more first manipulated variables (MV1 ) in accordance with the first dynamic control sub-matrix (DCM1 ); and preferably, first current feedforward variable data (FVD1 ) indicative of a current value of respective first feedforward variables (FV1 ) associated to the recovery unit (1 12) that influence the values of the first controlled variables (CV1 ) but are not adjustable by the recovery sub-controller unit (152).

3. The chemical production plant (100) of claim 1 or 2, wherein the second current variable data (VD2) comprises:- second current manipulated variable data (MVD2) indicative of a current value of the respective second manipulated variables (MV2) associated to the reaction unit (102);- second current controlled variable data (CVD2) indicative of a current value of second controlled variables (CV2) that are associated to the reaction unit (102), wherein the current values of the second controlled variables (CV2) depend on the current values of one or more second manipulated variables (MV2) in accordance with the second dynamic control sub-matrix (DCM2); and preferably, second current feedforward variable data (FVD2) indicative of a current value of respective second feedforward variables (FV2) associated to the reaction unit (102) that influence the values of the second controlled variables (CV2) but are not adjustable by the reaction sub-controller unit (154).

4. The chemical production plant (100) of any of the preceding claims, wherein the recovery sub-controller unit (152) and / or the reaction sub-controller unit (154) are further configured to receive constraint data (CD1 , CD2) indicative of allowed operational ranges forBASF SE231731one or more controlled variables (CV1 , CV2) and / or for one or more manipulated variables (MV1 , MV2).

5. The chemical production plant (100) of any of the preceding claims, wherein the reception of the first and second current variable data (VD1 , VD2) and the determination of the setpoint values (SP1 , SP2) of the first manipulated variables (MV1 ) and the second manipulates variables (MV2) is performed at regular time intervals.

6. The chemical production plant (100) of any of the preceding claims, wherein the first current variable data (VD1 ) includes: data pertaining to one or more of ethylene oxide loss (CV01 ), differential pressure in the absorber unit (CV02), differential pressure over a lean absorbent valve (CV03), temperature in a top section of the stripper unit (CV04), temperature in a middle section of the stripper unit (CV05), temperature in a bottom section of the stripper unit (CV06), ethylene oxide content in the bottom section of the stripper (CV07), inert limit (CV08), opening of stripper steam valve (CV09), differential pressure in the stripper unit (CV10), density of glycol (CV1 1 ) and level of demineralized water tank (CV12), preferably as values of first controlled variables (CV1 ); and / or data pertaining to one or more of lean absorbent flow (MV01 ), stripper temperature (MV02), demineralized water flow (MV03), residual absorber pressure (MV04) and glycol bleed flow (MV05), preferably as values of first manipulated variables (MV1 ); and, preferably, data pertaining to one or more of a lean absorbent temperature, opening of a process water valve, lean absorber to residual absorbent flow and an ethylene oxide flow to the stripper unit, preferably as values of first feedforward variables (FV1 ).

7. The chemical production plant (100) of any of the preceding claims, wherein the second current variable data (VD2) includes data pertaining to one or more of oxygen inlet (CV13), margin inlet (CV14, CV15), oxygen outlet (CV16, CV17), margin outlet (CV18, CV19), oxygen conversion (CV20, CV21 ), inlet temperature (CV22, CV23), loopgas pressure (CV24), pressure of a methane system (CV25), loopgas flow (CV26), gas flow to CO2(CV27), differential drum pressure (CV28), differential oxygen reactor outlet (CV29), opening of drum pressure valve (CV30), and dif-BASF SE231731ferential pressure of CO2absorber (CV31 ), preferably as values of second controlled variables (CV2); and / or data pertaining to one or more of drum pressure (MV06, MV07), methane flow (MV08), purge flow (MV09), pressure at a drum located between the reactor unit and the EO absorber unit (MV10), opening of guidevane for gas flow (MV1 1 ), and opening of CO2bypass valve (MV12), preferably as values of second manipulated variables (MV2); and, preferably, data pertaining to one or more of oxygen flow, ethylene concentration, and loopgas temperature, preferably as values of second feedforward variables (FV2).

8. The chemical production plant (100) of any of the preceding claims, wherein a subset of the first and / or of the second controlled variables (CV1 , CV2), and, preferably, a subset of the first and / or the second manipulated variables (MV1 , MV2) are critical variables (CRV), and wherein, upon determining that at least one critical variable (CRV) is not available for the recovery sub-controller unit and / or the reaction sub-controller unit, and / or or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range, the recovery sub-controller unit and / or the reaction sub-controller unit are configured to provide operation instructions for stopping operation of the linear model based predictive controller unit.

9. The chemical production plant (100) of claim 8, wherein the critical variables include one or more of ethylene oxide loss (CV01 ), differential pressure in the absorber unit (CV02), differential pressure over a lean absorbent valve (CV03), temperature in a top section of the stripper unit (CV04), temperature in a bottom section of the stripper unit (CV06), opening of stripper steam valve (CV09), differential pressure in the stripper unit (CV10), level of demineralized water tank (CV12), lean absorbent flow (MV01 ), residual absorber pressure (MV04), oxygen inlet (CV13), margin inlet (CV14, CV15), oxygen outlet (CV16, CV17), margin outlet (CV18, CV19), oxygen conversion (CV20, CV21 ), loopgas pressure (CV24), differential oxygen reactor outlet (CV29), drum pressure (MV06, MV07), methane flow (MV08), and purge flow (MV09).

10. Computer implemented method (200) for controlling operation of a chemical production plant (100) for producing ethylene oxide (EO) in a production process, the method comprising:BASF SE231731- receiving (202), at a recovery sub-controller unit (152) of a linear model based predictive controller unit (150), first current variable data (VD1 ) indicative of a current operational state of a recovery unit (1 12), and, at a reaction sub-controller unit (154) of the linear model based predictive controller unit (150), second current variable data (VD2) indicative of a current operational state of a reaction unit (102),- determining and providing (204), based on a predetermined first dynamic control submatrix (DCM1 ), respective setpoint values (SP1 ) of one or more first manipulated variables (MV1 ) associated to the recovery unit (1 12), and, based on a predetermined second dynamic control sub-matrix (DCM2), respective setpoint values (SP2) of one or more second manipulated variables (MV2) associated to the reaction unit (102); and- generating and providing (206) first operation instructions (OP1 ) for controlling operation of the recovery unit (1 12) in accordance with the determined setpoint values (SP1 ), and second operation instructions (OP2) for controlling operation of the reaction unit (102) in accordance with the determined setpoint values (SP2).1 1 . The method of claim 10 wherein- the first current variable data (VD1 ) comprises- first current manipulated variable data (MVD1 ) indicative of a current value of the respective first manipulated variables (MV1 ) associated to the recovery unit (1 12);- first current controlled variable data (CVD1 ) indicative of a current value of first controlled variables (CV1 ) that are associated to the recovery unit (1 12), wherein the current values of the first controlled variables (CV1 ) depend on the current values of one or more first manipulated variables (MV1 ) in accordance with the first dynamic control submatrix (DCM1 ); and,- preferably, first current feedforward variable data (FVD1 ) indicative of a current value of respective first feedforward variables (FV1 ) associated to the recovery unit (1 12) that influence the values of the first controlled variables (CV1 ) but are not adjustable by the recovery sub-controller unit (152); and / or wherein- the second current variable data (VD2) comprises:BASF SE231731- second current manipulated variable data (MVD2) indicative of a current value of the respective second manipulated variables (MV2) associated to the reaction unit (102);- second current controlled variable data (CVD2) indicative of a current value of second controlled variables (CV2) that associated to the reaction unit (102); and,- preferably, second current feedforward variable data (FVD2) indicative of a current value of respective second feedforward variables (FV2) associated to the reaction unit (102) that influence the values of the second controlled variables (CV2) but are not adjustable by the reaction sub-controller unit (154); and / or wherein- the method further comprises receiving (203), at the recovery sub-controller unit and / or at the reaction sub-controller unit, constraint data (CD) indicative of allowed operational ranges for one or more controlled variables (CV1 , CV2) and / or for one or more manipulated variables (MV1 , MV2), and determining the setpoint values further in dependence on the constraint data (CD).

12. The method of claim 10 or 1 1 , wherein receiving the first and second current variable data (CVD1 , CVD2), and determining the setpoint values (SP1 , SP2) of the first manipulated variables (MV1 ) and the second manipulated variables (MV2) is performed at regular time intervals.

13. The method of any of the preceding claims 10 to 12, comprising: defining a subset of the first and / or of the second controlled variables (CV1 , CV2), and, preferably, a subset of the first and / or the second manipulated variables (MV1 , MV2) as critical variables (CRV), and providing (207) operation instructions for stopping operation of the linear model based predictive controller unit, upon determining that at least one critical variable (CRV) is not available for the recovery sub-controller unit and / or the reaction sub-controller unit or upon determining that a value of at least one of the critical variables is outside a respective predetermined validity range.

14. Computer program comprising instructions that, when executed by a linear model based predictive controller unit (150) of a chemical production plant causes the linear modelbased predictive controller unit (150) to carry out the method of any of the preceding claims 10 to 13.

Images