EP2149022A2

EP2149022A2 - Method for controlling a cryogenic distillation unit

Info

Publication number: EP2149022A2
Application number: EP08805713A
Authority: EP
Inventors: Jean-françois RAUCH; Athanasios Kontopoulos; Jean-Marc Peyron; Thierry Roba
Original assignee: Air Liquide SA; LAir Liquide SA pour lEtude et lExploitation des Procedes Georges Claude
Current assignee: Air Liquide SA; LAir Liquide SA pour lEtude et lExploitation des Procedes Georges Claude
Priority date: 2007-05-11
Filing date: 2008-04-25
Publication date: 2010-02-03
Also published as: FR2916039A1; WO2008152264A2; FR2916039B1; JP2010528245A; CN101796360B; CN101796360A; US20100211221A1; WO2008152264A3

Abstract

In a method for controlling a cryogenic distillation separation apparatus, at least one manipulated variable is modified, each manipulated variable being modified using at least one controlled variable, whereby each controlled variable can be adjusted using a control method, and a predictive control method is used to control at least one set point of one of the controlled variables.

Description

Process for regulating a cryogenic distillation unit

The present invention relates to a method of controlling a cryogenic distillation unit, for example an air separation apparatus or a mixture separation apparatus having as main components hydrogen and carbon monoxide.

The control method according to the invention uses predictive and multivariable control (for example MVPC ("Multi-Variable Predictive Control") and possibly control by non-predictive methods, such as as the AFF (Advanced Feed Forward) strategy.

The approach is illustrated by examples, such as the rapid change of market and the optimization of the argon extraction yield.

The process of air distillation will not be presented in detail, being sufficiently explained in the literature, for example in "Oxygen Enhanced Combustion" Editions CRC, 1998, "Tieftemperaturtechnik" of Hausen and Linde, etc.

In a nutshell, this process is used to produce oxygen, nitrogen and argon (more rarely krypton and xenon) by compressing and then cooling (liquefying) and distilling ambient air.

In a conventional system the air is compressed and then separated using low and medium pressure columns (which are more and more frequently "superimposed" and which thermally communicate by an oxygen / nitrogen exchanger called vaporizer-condenser). In the medium pressure column, the nitrogen is separated from the air by creating oxygen-rich liquid at the bottom of the column and nitrogen-rich liquid and vapor at the top of the column. These products are extracted and at least some are fed separately to the low pressure column. Because of differences in relative volatility between argon, nitrogen and oxygen, substantially pure nitrogen is formed at the top of the column, substantially pure oxygen is formed at the bottom of the column and gas rich in argon in the middle of the column. The central argon-rich fraction, often referred to as raw argon, can be withdrawn from the low pressure column to feed an auxiliary column (argon column) for the purpose of producing argon. The raw argon is rectified into a rich oxygen reflux (which is subsequently sent to the low pressure column for condensing) and in a very rich argon stream (often called "argon mixture") that can be used as a product as such or purified later.

In a modern unit it is rare for the values of the setpoints of inflowing air inflows, nitrogen, oxygen and argon produced as well as those of intermediate flows (for example, flow rates of the high pressure column at the low pressure column) are fixed. Regulatory systems are used to simultaneously meet product quality specifications (grades) while producing the required quantities and, increasingly, safety and environmental constraints.

These control systems are often of the Advanced Feed Forward (AFF) type and in the more recent times of the MVPC (Multi Variable Predictive Control) type.

Both have advantages and disadvantages. The present invention provides a combined system that optimizes the use of one and the other.

According to the present invention, there is provided a method of controlling a cryogenic distillation separation apparatus in which at least one manipulated variable is modified, the manipulated variable or each manipulated variable being modified by means of at least one controlled variable , each controlled variable being adjustable by means of a regulation method characterized in that a predictive method of regulation is used to regulate at least one set point of a first controlled variable.

According to other aspects:

at least one set point of a first controlled variable regulated by the predictive method is used to calculate, by a non-predictive method, possibly of Advanced Feed Forward type, at least one set point of at least one second controlled variable.

at least one set point, derived from a set point of one of the controlled variables regulated by the predictive method, is used to compute by a non-predictive method, possibly of Advanced Feed Forward type, at least one setpoint of at least one second controlled variable. the set point is derived from a set point of one of the controlled variables regulated by the predictive method by filtering, possibly by filtering of the 'ramp' type.

the first controlled variable is a supply air flow rate for an apparatus for separating air by cryogenic distillation in a double column comprising a medium pressure column and a low pressure column and the second controlled variable is a liquid flow rate of reflux from the medium pressure column and / or for the low pressure column or a reflux liquid capacity level from the medium pressure column for the low pressure column.

the calculated value of reflux liquid setpoint from the medium pressure column to the capacity is treated by filtering of the 'Lead-lag' type, preferably of the 'inverse response' variant.

the calculated value of the reflux liquid set point from the capacity to the low pressure column is treated by 'lead-lag' filtering, preferably of the 'overshoot' variant.

the reflux liquid is enriched in nitrogen.

the method is a method of regulating an air separation apparatus comprising a medium pressure column, a low pressure column and an argon separation column and the first controlled variable is the oxygen content at a predetermined height of the low pressure column, where the argon content is preferably substantially maximum wherein i) the nitrogen content at the top of the argon separation column is measured and if the nitrogen content exceeds a first threshold, at least one an upper or lower limit for the first controlled variable and / or ii) the oxygen content of a high oxygen flow rate withdrawn from the low pressure column is measured and if the oxygen content falls below a second threshold, increases at least one upper or lower limit for the first controlled variable.

the at least one upper or lower limit of at least 0.1%, preferably at least 0.5%, is increased.

the at least one upper or lower limit is increased instantaneously. - either (a) once the nitrogen content has exceeded the first threshold, if the nitrogen content subsequently falls below a third threshold, lower than, equal to or greater than the first threshold, then the at least one upper limit is reduced or lower for the first controlled variable and / or ii) once the oxygen content has dropped below the second threshold, if then the oxygen content exceeds a fourth threshold, lower than, equal to or greater than the second threshold, then the oxygen content is reduced. at least one upper or lower limit for the first controlled variable.

the at least one upper or lower limit of at least 0.1%, preferably at least 0.2% is reduced.

the at least one upper or lower limit is reduced for a period of at least 10 minutes.

the first threshold is at least 0.2% nitrogen, preferably at least 0.3% and optionally the third threshold is equal to the first threshold.

The invention will be described in more detail with reference to the Figures.

FIGS. 1, 2 and 7 schematically show control methods according to the invention, FIGS. 3 to 6 show the effect of the cleating systems that can be exploited in the context of the invention, FIG. 8A shows a control method according to the invention. The invention in the context of the air separation apparatus of Figure 8B and Figures 9 and 10 are graphs showing variables regulated according to the method of the invention.

The invention consists of a combined process control system that provides the benefits of both AFF and MVPC systems.

The first step is to define the control matrix, ie MVs (Manipulated Variables), CVs (Controlled Variables) and DVs (Disturbances and / or Observable Deviations).

Using the knowledge of the process: the static behavior of the unit (thermodynamic equilibrium ...) but also the dynamic behavior (hydraulic flow and dynamic retention), we define the relationships between certain variables controlled by the SNCC (Digital Control System) Command) and other control matrix variables (DVs and MVs). Subsequent calculations may be performed from the values of the MVs and the results of these calculations are new setpoints, as shown in Figure 1. The MVPC controller receives disturbance values DV1, DV2 and CV1 values, CV2 of controlled variables. From these values, the MVPC controller calculates (using dynamic correlations as explained above as well as various ad hoc parameters) new setpoints (RSP), for the manipulated variables, MV1, MV2 and these new points. Setpoints are sent to controllers of different types (for example, to a FIC or "Flow Indicator and Control" or Indicator and Flow Control or to a LIC "Level Indicator and Control" or Indicator and Level Control). In this case, in this example, it is a flow control. Usually these relationships use one or more Manipulated Variables.

But in some cases, we can also use one or more Disturbances (DV1, DV2) and one or more Controlled Variables (CV1, CV2) to generate new setpoints (RSP or Remote Set Point, "remote setpoint" ) for a flow controller (FIC) and a level controller (LIC), combining with the values of some manipulated variables (MV1, MV2), as seen in Figure 2. The difference between Figure 1 and Figure 2 is that in the case of Figure 2 some CVs and DV participate in the calculation of the value of certain calculated instructions (RSP) which are transmitted to FIC, LIC type regulators, etc. without going through the MVPC.

In some cases, the values of Manipulated Variables are used directly. These being recalculated at each computation cycle of the multivariable predictive controller, the calculation of the setpoints is incremental.

Most often, the value coming from the controller passes through a filter to go from the discrete domain to the continuous domain.

This makes it possible to put slow filters (first-order for example) to have slowly varying setpoints (for systems with high inertia) as shown in Figure 3.

Other times we use filters that limit the variations as in Figure 4. Another type of filtering is of type "lead-lag" (or "advance / delay") to give a dynamic to the setpoint change.

We have lead-lag type "inverse response" (Figure 5): when the setpoint given by the controller increases, the signal begins to be reduced first, before increasing to the desired value.

Another type is that of over-overshoot: the filter amplifies the setpoint changes transiently (Figure 6).

It is not necessary to use only one filter, we can use a combination of several filters. As shown in Figure 7, a first filter is used to change the value of MV1, a second filter is used to change the value of MV2 and a third filter is used to change the value of the product setpoint by calculation.

The benefits are many:

• First, the size of the matrix of the multivariable controller is reduced (especially less Manipulated variables). The system is therefore easier to implement.

• Less time spent identifying models (correlations expressing the dynamic links between CVs and DVs and MVs) of the system (this time is directly proportional to the number of Manipulated Variables).

• Less communication between the DCS and the PC that contains and runs the MVPC software is the most common case).

• Less configuration (programming) in the SNCC.

• Fewer settings settings in the MVPC controller (faster commissioning).

• Greater robustness of the controller.

The system according to the invention makes it possible to optimize the production unit. Optimization variables are included in the matrix. The linear or quadratic optimization program makes it possible to find the optimum operating point of the unit by pushing the controlled variables against their constraints. But the system according to the invention also makes it possible to make very rapid changes in the market. Indeed, some of the control loops being predefined, this makes it possible to anticipate the load changes of the unit.

This system therefore makes it possible both to optimize and to make load variations between 0.1% / min (shift in pseudo static mode) up to more than 7% / min (very fast changeover).

The efficiency of the process according to the invention in the context of a rapid change of market (up to 7% of product flow / minute) will be demonstrated using Figures 8A and 8B.

In FIG. 8B, a double column comprises a medium pressure column MP and a low pressure column LP thermally connected to each other by a vaporizer-condenser. The apparatus produces low pressure oxygen in gaseous form OGBP in the bottom of the LP column.

Air pressure medium air is sent to the MP medium pressure column and AirTurb air is sent to the BP column.

Rich liquid is sent from the tank of the MP column to the LP column.

Fluid rich in LP nitrogen called lower lean liquid is sent to a capacity C and liquid of the capacity is sent to the LP column.

Higher lean liquid is sent from the MP column to the LP column.

The objective is to increase and / or very quickly reduce the air load of an air separation unit in order to adapt more rapidly to the consumption demand. It is understood that these load changes must comply with the safety instructions and quality specifications of the delivered product.

To maintain the purities in the specifications of an air separation unit which benefits from the system according to the invention, we must maintain the reflux as constant as possible in:

• Low pressure column (BP)

• the Medium Pressure (MP) column

In the case of very fast changes of market, the solution to this problem can not come solely from the control system. Indeed, during a fast change of gears, the gaseous flows in the column (both the MP and the BP) change more quickly than the liquid flow rates (which change with much more delay, given the liquid retentions, whether trays or lining inside the columns) . This creates a drastic change in reflux values in the column with, as an immediate consequence, a loss of grades and a cessation of production.

The solution according to the invention will be to exploit all the liquid capacities of the column, or even to install another one, which, managed by an effective control system, will guarantee sufficient reflux so that the purities are maintained also during the changes. Steps.

A brief overview of the installation is given in Figure 8.

An additional capacity is installed so that one can benefit from a volume of liquid necessary during a fast change of market. The useful volume of this capacity maybe is based on detailed calculations (dynamic modeling). Capacity C is filled with Low Oxygen Liquid (LP) from the MP column and the outgoing liquid is directed to the LP column at an appropriate location.

The principle of filling / emptying the capacity C is as follows: when the air flow (the appliance load) is at its highest value, the level of the capacity is at the lowest value (say 20 %) and when the airflow is at the lowest possible value, the liquid level setpoint of the capacity is as high as possible (eg at 40%, 50% or 80%).

This relatively simple principle must nevertheless be managed by an effective control system because the filling or emptying flow rates of the capacity must not be modified in a way that is purely proportional to the air flow. This is due to the fact that the dynamic impact of the change of the air flow and that of the LP on the reflux are not the same. These differences must therefore be managed by an adequate control system in order to maintain the most stable reflux possible. At the same time, we must maintain the level of capacity at the right value. This leads us, therefore, to three setpoints (Remote Set Point, to be calculated at any time (see Figure 8):

- RSP_1: Low Flow Liquid (LP) Flow Instruction from MP and to Capacity C

- RSP_2: LIC setpoint of the C capacity

- RSP_3: the LP flow setpoint of the capacity to the LP column In addition, to ensure the change of market, we must ensure an adequate change of the air flow as well as the flow rate of OGBP (low pressure gaseous oxygen) for:

- meet the demand for OGBP production as quickly as possible

- keep the content of the OGBP within the limits imposed

We then use a combination of different types of filters in combination with the AFF method and the MVPC (to exploit its multivariable and predictive variables management capabilities).

To summarize, in the case presented:

- the LP flow rates to and from the additional capacity are managed by an AFF method with appropriate use of various filters (see Figure 8). This helps to keep the ebb at the right value in the column

- Airflow and OGBP are managed by the MVPC. This guarantees the production of OGBP at the desired value and maintenance of the OGBP content.

Indeed, if we follow Figure 8A: the oxygen demand (GOX request) is translated by an adequate calculation (calcuM) into an Air flow demand (this is, also, justified by the fact that the device can share an air-supply network - as well as an oxygen-production network - with other devices)

- the MVPC, taking into account this request for OGBP, the possibilities at this time of the air compressor, the OGBP content, the value of the disturbance variables, etc., will propose new points of instructions for the 'Air (FAIR_1) and the OGBP the new setpoint for air, FAIR_1, has a "staircase" look because the MVPC requires time for its calculations, so it sends a setpoint (RSP) to PID every minute, or 30 seconds, etc. It would not be acceptable for the "AFF / Capacity CSR Management" system to have such a "split" entry. A "ramp" type filter is then used to "smooth" this set point before transmitting it to the flow management of the additional capacity. This gives us a new set point (FAIR_2)

this new instruction, through a calculation (calculation_2, for example of type ax + b) is translated into a value of flow rate of Poor Liquid (F LP) which would represent the flow of poor liquid in stationary state. The dynamic forces us to use this rate to: o calculate a setpoint (remote set point) LP from the MP to the capacity (RSP_1). This calculation requires the passage through a Lead-Lag type filter (inverse response) o calculate a setpoint (remote set point) (RSP_3) for the LP of the capacity to the BP column having passed through:

• an overshoot filter (or overtaking)

• a calculation (of type -ax + b, calculation_3) to which is added a correction of the LIC of the capacity

In this way we get the dynamic management of this event (fast change of market) by having usefully combined a management of AFF type with the MVPC. This is the principle of the present invention. Indeed, MVPC's predictive and multivariable intrinsic capabilities help to increase the speed of the operation within the limits of OGBP content.

At a site that has to respond very quickly to changes in oxygen consumption, we have implemented a MACCS system based on the principles mentioned above.

The purity of gaseous oxygen produced must, as a general rule, remain close to 95% in any case between 94% at the lowest (contract content) and 96.5% at the highest (for safety reasons) .

The evolution of different parameters is illustrated in Figures 9 and 10. The changes in market are made quickly but keeping the content of the OGBP within the desired limits.

The AFF part (with filters) controls all the part concerning the flow rates of the additional capacity and the MVPC the air flow and the OGBP flow.

Another use of the system according to the invention is the optimization of argon extracted from an air distillation apparatus (ASU).

Reference can be made to the brief description of an air distillation apparatus made above.

The raw argon stream (from the low pressure column to the argon column) contains a percentage of nitrogen. The presence of nitrogen creates many operational concerns when argon is distilled. Indeed, to extract a maximum of argon, we must maintain the "belly" argon (oxygen content at the location of the low pressure column where the argon flow is withdrawn) as low as possible. This comes from the fundamentals of distillation and is a well-known rule in the operation. On the other hand, a too low value of the argon belly results in an excessively high presence of nitrogen at the top of the argon distillation column which prevents this column from functioning correctly. These phenomena are eminently nonlinear. The result is a loss of pure product contents and an unintentional triggering of the operating device.

MVPC systems installed on air distillation columns have difficulty in accounting for this phenomenon because the models that reproduce the presence of nitrogen at the head of the argon column according to different parameters, are strongly linear and are difficult to manage with a "pure" MVPC approach.

In a case of an MVPC approach on a basic system (for example) to regulate the argon belly of a column we can build the following system:

Variables Manipulated, MV (whose setpoint is proposed by the MVPC system)

MV1: Air Flow

MV2: Low Pressure Oxygen Flow (OGBP)

Controlled Variables, CV (whose value must be maintained between 2 limits-high and low-as far as possible by the MVPC, by manipulating the MV variables)

CV1: The argon belly oxygen value at a predetermined height of the low pressure column (in%)

CV2: the objective value of the air flow (target air) which must be satisfied for production reasons

Disturbance Variables, DV (which the MVPC does not manipulate but whose influence on CV variables is determined by models):

DV1, DV2 ...: Measurement-setpoint deviation for Airflow, OGBP, etc. (the flows involved in MV variables). DVx, DVx + 1: possibly impact of the pressurization of the bottles of the purification at the head, flow OG average or high pressure, nitrogen flow gas of average or high pressure, ...

Obviously, this configuration is an example, different configurations between the MV, CV and DV can be considered to solve the same problem.

In the case of our Combined approach, we follow the following strategy:

1. Set a threshold for the presence of nitrogen at the top of the argon column. This can be typically of the order of 0.2% to 1% but it can be more or less important, this depends on the particular characteristics of each column. Call this threshold (A).

2. When the threshold (A) is exceeded, the values of the very low, low limit, high limit and very high limit, which are transmitted to the MVPC as limits within which the variable CV1 (argon belly) must be maintained, are all instantly increased by a predetermined value (let's call it V1) which depends on the process and which can typically be in the range of 0.2% up to 3% and more typically from 0.5% up to 1.5%. This value (V1) which moves all the limits upwards is called "Automatic Bias".

3. When the value of the nitrogen analysis at the head of the column then becomes lower than a threshold (B) which may be equal to the threshold (A) or (B) = (A) +/- (C), (C) being a value ensuring a hysteresis (typically of the order of 0.1% to 0.5% in the case under examination), then this time the V1 is removed from the values of the belly limits, preferably not instantaneously but rather with a ramp (Λ / 1 / min) to avoid a sudden return to the initial value of the argon belly set point.

This technique makes it possible to avoid untimely triggering of units which generate production losses, energy losses as well as the potential dangers of inadvertent tripping of the production unit and, at the same time, keeping an optimum argon belly set point. (very low) which makes it possible to optimize the extraction of argon.

The principles developed above are demonstrated and proven by the example presented in Figure 10. To aid the clarity of the exposition only the values of the very low limit and the very high limit are presented but the lower limit and the upper limit are also increased by the same value (V1).

In the case presented:

Bias activation threshold (A) = 0.3% nitrogen at the head of the argon column

Bias Deactivation Threshold (B) = (A) = O.3%

Value (V1) of the automatic bias: 1.5% which is automatically added to the argon belly limits transmitted to the MVPC

Ramp time to return to the initial value of the belly set point: 30 minutes

In addition, in the case presented, all the limits of the CV1 (argon belly) are calculated from a set of parameters such as the load of the apparatus, the flow of impure oxygen produced, etc.

It can be observed that the phenomenon of presence of nitrogen is strongly non-linear, hence the need to take into account the said phenomenon by this technique outside the MVPC.

It should also be noted that the activation of this automatic bias is not necessarily solely related to the presence of nitrogen at the head of the argon column but may be related to the presence of other phenomena such as a low threshold of oxygen content (eg the low pressure oxygen content produced by the low pressure column, etc.).

For some controlled variables whose timeout is greater than 15 minutes, a predictive method of regulation is used. For example a change in product flow of the impure argon column fed from the low pressure column has an impact on the oxygen content measured in the column whose dead time exceeds 15 minutes. The oxygen content of the impure argon column will therefore be regulated by a predictive method.

Claims

A method of regulating a cryogenic distillation separation apparatus in which at least one manipulated variable is modified, the manipulated variable or each of the manipulated variables being modified by means of at least one controlled variable, each controlled variable being adjustable to means of a method of regulation characterized in that one uses a predictive method of regulation to regulate at least a set point of a first controlled variable.

2. Method according to claim 1, in which at least one set point of a first controlled variable regulated by the predictive method is used to calculate, by a non-predictive method, possibly of the 'Advanced Feed Forward' type, at least one setpoint of at least one second controlled variable.

3. Method according to claim 1 wherein at least one setpoint, derived from a set point of one of the controlled variables regulated by the predictive method, is used to compute by a non-predictive method, possibly of type. Advanced Feed Forward, at least one set point of at least one second controlled variable.

4. Method according to claim 3 wherein the set point is derived from a set point of one of the controlled variables regulated by the predictive method by filtering, possibly by filtering type 'ramp'.

5. Method according to one of claims 2 to 4 wherein the first controlled variable is a supply air flow for an air separation apparatus by cryogenic distillation in a double column comprising a medium pressure column and a column low pressure and the second controlled variable is a reflux liquid flow from the medium pressure column and / or to the low pressure column or a reflux liquid capacity level (Capa) from the medium pressure column and intended for the low pressure column.

The method of claim 5 wherein the calculated reflux liquid setpoint value from the medium pressure column. to the capacity is treated by filtering type 'Lead-lag', preferably variant 'reverse response'.

7. The method of claim 5 or 6 wherein the computed liquid reflux setpoint value from the capacity to the low pressure column is treated by 'lead-lag' type filtering, preferably over-passed variant. .

8. Method according to one of claims 5 to 9 wherein the reflux liquid is enriched in nitrogen.

9. A method according to claim 1 for regulating an air separation apparatus comprising a medium pressure column, a low pressure column and an argon separation column and the first controlled variable is the oxygen content at a predetermined height. of the low pressure column, where the argon content is preferably maximum wherein i) the nitrogen content at the top of the argon separation column is measured and if the nitrogen content exceeds a first threshold, at least one an upper or lower limit for the first controlled variable and / or ii) the oxygen content of a high oxygen flow rate withdrawn from the low pressure column is measured and if the oxygen content falls below a second threshold, increases at least one upper or lower limit for the first controlled variable.

The method of claim 9 wherein the at least one upper or lower limit of at least 0.1%, preferably at least 0.5%, is increased.

The method of claim 9 or 10 wherein the at least one upper or lower limit is increased instantaneously.

The method of claim 9 or 10 wherein either

- once the nitrogen content has exceeded the first threshold, if the nitrogen content subsequently falls below a third threshold, lower than, equal to or greater than the first threshold, then the at least one upper or lower limit is reduced by the first controlled variable and / or once the oxygen content has passed below the second threshold, if the oxygen content then exceeds a fourth threshold, which is less than, equal to or greater than the second threshold, the at least one upper or lower limit for the first threshold is reduced; controlled variable.

The method of claim 12 wherein the at least one upper or lower limit of at least 0.1%, preferably at least 0.2%, is reduced.

14. The method according to one of claims 12 to 13 wherein the at least one upper or lower limit is reduced for a period of at least 10 minutes.

15. Method according to one of claims 9 to 14 wherein the first threshold is at least 0.2% nitrogen, preferably at least 0.3% and optionally the third threshold is equal to the first threshold. .