CN101547994A - Detection of catalyst losses in a fluid catalytic cracker for use in abnormal situation prevention - Google Patents

Abstract

The present invention discloses a method and system for detecting and/or predicting abnormal levels of catalyst loss in a fluid catalytic cracking unit. The method and system measures a differential pressure across portions of a fluid catalytic cracker, such as a reactor cyclone or a regenerator cyclone, and determines abnormal catalyst loss when the differential pressure changes significantly from a baseline differential pressure. The claimed method and system implements algorithms using computing devices to detect or predict an abnormal condition based on the change in a monitored differential pressure in a fluid catalytic cracking unit.

Description

Be used for the detection of the fluid catalytic cracker catalyst attrition of abnormal situation prevention

The application requires in the U.S. Provisional Patent Application No.60/848 of submission on September 29th, 2006,596 rights and interests, and the full content of this application is incorporated herein by reference.

Technical field

This patent relates in general to be implemented diagnosis and safeguards that more particularly, the mode that relates to the abnormal conditions in reduction or the prevention source mill provides diagnosis capability in source mill in the source mill.

Background technology

Fluid catalytic cracking is that modern refineries is cracked into high-molecular weight oil (hydrocarbon polymer) process commonly used than light constituent that comprises liquefied petroleum gas, gasoline, aviation fuel and diesel oil.Usually, the fluid catalytic cracking process is used the at first hydrocarbon polymer of decomposing macromolecular amount of catalyzer, uses at least one cyclone that the mixture separation that obtains is become collectable byproduct then.Employed catalytic specie can be recovered then to inject another reaction cycle.A problem that may take place in the fluid catalytic cracking process is that the catalyst attrition of reactor parts or revivifier parts may be too high.If do not correct, this catalyst attrition may cause going wrong in the subsequent disposal unit in fluid catalytic cracker downstream.

Summary of the invention

Abnormal rate of catalyst loss in desired method and system detection and/or the prediction fluid catalytic cracking unit.Parts that can the monitoring fluid cat cracker, for example pressure reduction of reactor cyclone or regenerator cyclone.The normal differential pressure noticeable change of the unitary parts of fluid catalytic cracking during normal running can indicate catalyst attrition to increase, and can indicate to break down in the fluid catalytic cracker and maybe need to safeguard.Desired method and system uses the computing equipment implementation algorithm, surveys or the predicted anomaly situation with the differential pressure based on the fluid catalytic cracking cyclone of being monitored.When detecting abnormal conditions, can generate warning to notify suitable entity.

Description of drawings

Fig. 1 illustrates the fluid catalytic cracking unit;

Fig. 2 illustrates the computing equipment that can be used to realize statistic processes monitoring (SPM) algorithm;

Fig. 3 illustrates the SPM module that can realize on computing equipment;

Fig. 4 illustrates the embodiment of abnormal operation detection (AOD) module of using regression model;

Fig. 5 illustrates to use and returns the process flow diagram flow chart of surveying catalyst attrition;

Fig. 6 illustrates the exemplary process factory that can realize abnormal situation prevention system; And

Fig. 7 illustrates the part of the abnormal situation prevention system of communicating by letter with various device shown in the source mill.

Embodiment

Generally speaking, Fig. 1 illustrates the FCC Fluidized Catalytic Cracker 10 that is used for high molecular weight oil is realized the fluid catalytic cracking process.The charging 12 that comprises high molecular weight oil can flow in the bottom of reactor 14, and reactor 14 is vertical or acclivitous pipe, is sometimes referred to as " riser tube ".Thereby high activated catalyst 16 can be imported into riser tube 14 contact chargings 12.Charging 12 can be preheated, and be sprayed on the bottom of riser tube 14 by the feed nozzle (not shown), and wherein charging 12 is at the very hot catalyzer of fluidizing of feed nozzle contact.Dispersion steam 18 can be used to spray the charging 12 by feed nozzle.When thermocatalyst contact charging 12, catalyzer makes charging 12 evaporations, and catalysis is carried out in the cracking reaction than light constituent of high molecular weight oil being resolved into such as liquefied petroleum gas (LPG) (LPG), gasoline and diesel oil.Catalyzer-hydrocarbon mixtures riser tube 14 of can upwards flowing through then, and finally flow into disengagement vessel 19.Catalyzer-hydrocarbon mixtures can be gathered in reactor cyclone 20, and the part of the hydrocarbon polymer in the mixture can be passed through this cyclone 20 and catalyst separating.Most of catalyzer can be exported from cyclone 20 and deposit in the disengagement vessel 19.The cyclone reactor effluent 22 that mainly comprises the hydrocarbon polymer that does not contain catalyzer can be transported to the main fractionator (not shown), further to be separated into the light cycle oil that uses in combustion gas, LPG, gasoline, diesel oil and the rocket engine fuel, heavy hydrocarbon gases etc.

When cracking catalyst moves up in riser tube 14, thus on catalyzer deposit coke reduce activity of such catalysts and optionally reaction make cracking catalyst quilt " consumption ".Emanated out in the cracked hydrocarbon container of the catalyzer that has used from disengagement vessel 19, and sent to stripping tower 24, in stripping tower 24, removed remaining residual carbon hydrogen compound in the catalyzer thereby stripped vapor can contact the catalyzer that has used.The catalyzer that is consumed can be introduced into fluid bed regenerator 28 subsequently, in fluid bed regenerator 28, use warm air 30 (or in some cases, the air oxygenation) burns deposits of coke, so that catalyzer is returned to active condition, and provide essential heat for next reaction cycle.The combustion of coke settling produces the stack gas that comprises carbonic acid gas and carbon monoxide.Can from the solid catalyst of revivifier 28 and solid-state coke mixture, isolate or filter out stack gas with regenerator cyclone 31." regenerated " catalyzer can return the bottom of riser tube 14, with recirculation.

Contingent problem is to follow the ring catalyst channels catalyst attrition may take place in the operation of FCC Fluidized Catalytic Cracker.Although may reckon with certain nominal catalyst loss in the fluid catalytic cracking process, bigger catalyst attrition may show that plant failure (for example revealing) maybe needs to safeguard and maintenance.In one embodiment, can by measure as shown in Figure 1 reactor cyclone 20 and the differential pressure (Δ P) of the one or both in the regenerator cyclone 31 survey catalyst attrition.For example, can obtain differential pressure between the whirlwind input terminus 32 of reactor cyclone 20 and the effluent output terminal 34 or between the whirlwind input terminus 36 of regenerator cyclone 31 and stack gas output terminal 38.In this embodiment, for normal running, differential pressure can keep approaching stationary value.If differential pressure significantly reduces from initial (normally) state, the unusual increase of catalyst attrition then may take place.This can indicate more than the catalyzer of normal amount and spill with the reactor effluent of reactor cyclone or with the stack gas of regenerator cyclone.

Survey abnormal catalyst loss

Can implement abnormal operation detection system described here, with prediction or survey catalyst attrition, thereby can adopt preventative measurement to reduce catalyst attrition in the fluid catalytic cracking unit.Abnormal operation detection system can realize in the existing processe Controlling System, perhaps be installed into the computing unit of independent operation.Usually, abnormal operation detection system can realize with hardware or the software that moves on computing equipment.Describe below and to realize to survey or to predict various types of algorithms of the catalyst attrition in the fluid catalytic cracker by abnormal operation detection system.

The statistic processes monitoring

A kind of algorithm that can be used for the catalyst attrition of definite fluid catalytic cracking unit is statistic processes monitoring (SPM) algorithm.SPM can be used to monitor the variable such as qualitative variables that is associated with process, and is detected notification operator when departing from its " statistics " standard at this qualitative variables.The SPM algorithm can calculate average and the standard deviation of process variable in non-overlapping sample window such as pressure reduction usually.

Fig. 2 illustrates in one embodiment, can be used to realize the computing equipment of SPM algorithm or SPM functional block.The parts of computing equipment 50 can include but not limited to processing unit 52, system memory 54 and various system components are connected to the system bus 56 of processing unit 52.Storer 54 can be to comprise volatibility and non-volatile media, removable and immovable medium by any usable medium of processing unit 52 visits.The user can be by the user input device 66 such as keyboard and pointing device to computing equipment 50 input commands and information.These and other input unit can be connected to processing unit 52 by user's input interface 60 that can be connected to system bus 56.The display equipment of monitor or other type also can be connected to treater 52 via user interface 60.Also can use other interface or bus structure.Particularly, can receive the input 62 of miscellaneous equipment (for example transmitter) by I/O (I/O) interface 58 at computing equipment 50 places, and the output 64 of computing equipment 120 can be provided to other equipment by I/O (I/O) interface 58. Interface 58 and 60 is connected to treater 52 with various device by system bus 56.

Fig. 3 illustrates statistic processes monitoring (SPM) module 70 that can realize on the computing equipment 50 of Fig. 2.Logical block 72 can receiving course signal set 74, and can calculate the statistical nature or the statistical parameter of this process signal set 74.These statistical parameters can calculate based on the moving window of first process variable data or based on the non-overlapping window of first process variable data.Statistical parameter that calculates or statistical nature can comprise for example average, standard deviation, the variance (S of process signal ²), rootmean-square (RMS), velocity of variation (ROC) and scope (Δ R).These statistical parameters can provide by following formula:

S=standard deviation=σ

$\mathrm{RMS}=\sqrt{\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{X}_i^2}$

$\mathrm{ROC}=r_i=\frac{x_i-x_{i-1}}{T}$

ΔR＝X _MAX-X _MIN

In above formula, N is the sum of the data point of sample phase, x _iAnd x _I-1Be two successive values of process signal, T is two timed intervals between the value.Further, X _MAXAnd X _MINBe respectively that process signal is in sampling or maximum value and the minimum value of training period.Can calculate separately or calculate these statistical parameters in the mode of arbitrary combination.In addition, should be appreciated that except the statistical parameter of clearly listing that the present invention also comprises any statistical parameter that can be implemented with the analytic process signal.The statistical parameter that calculates can be received by computing block 76, and computing block 76 is according to the regular operation that comprises in the regular piece 78.Rule piece 78 can be for example realizes in the part of the storer 54 (Fig. 2) of computing equipment 50, and can define the algorithm that is used to survey or calculate abnormal conditions, as discussed further below.

In another embodiment, can calculate and be updated periodically trained values by for example computing equipment 50.For example, in one embodiment, can generate trained values by statistical parameter logical block 72, statistical parameter logical block 72 generates or learns nominal or the normal statistics parameter in first operational phase, and first operational phase is the stage of process normal running normally.These nominal statistical parameters can be used as trained values then and are stored in the trained values block 80, for using (as described further below) future.This operation allows at specific loop and operating conditions dynamic adjustment of trained values 80.In this case, can in the at user option period, monitor based on the process dynamic response time statistical parameter (can be used for trained values).In one embodiment, the computing equipment such as computing equipment 50 can generate or receive trained values or be used for sending trained values to another process device.

In one embodiment, can realize surveying the algorithm of the catalyst attrition in the fluid catalytic cracker with the SPM piece that illustrates among Fig. 3 70, and be used for determining unusual condition by the input that receives such as the pressure reduction of the reactor of fluid catalytic cracker or regenerator cyclone.In this embodiment, SPM piece 70 can serve as abnormal operation detection (AOD) module.In this configuration, regular piece 78 can comprise the rule that is used for calculating based on the pressure reduction of the cyclone of being imported unusual condition.Computing block 76 can be programmed with output warning 82 when detecting unusual condition.Here, the pressure reduction that observes can be sampled at interval with routine, and is input to the SPM piece 70 of Fig. 3 as process signal 74.At learning phase, logical block 72 can be determined benchmark (baseline) average (μ) and the base standard poor (σ) of pressure reduction (Δ P).Can think that these parameters are expressions that process is in " normally " situation.Baseline mean and base standard difference can be used as trained values (promptly using piece 80) and are stored in the storer 54 then.At monitor stages, the SPM piece 70 of realizing this algorithm can obtain the currency of pressure reduction and calculate to have process average (x) and standard deviation (s) in the non-overlapping sample window of equal length with the employed sample window of learning phase.

Use the SPM algorithm by SPM piece 70,, and can export indication or alarm 82 if the difference of actual average or current average and baseline mean then can detect catalyst attrition at computing block 76 greater than certain threshold value.For example, if current average reaches greater than particular percentile than baseline mean is low:

$\stackrel{\&OverBar;}{x}<(1-\frac{\mathrm{\&alpha;}}{100})\&CenterDot;\mathrm{\&mu;}$

Wherein α is certain user-defined per-cent (for example 5%).This formula can be represented as the rule more than in the regular piece 78.In one embodiment, SPM piece 70 can comprise the input (for example detection threshold of being determined by the user) of detection threshold.In this embodiment, detection threshold can be stored as trained values.

The user that a shortcoming of aforesaid method may be to have the knowledge of this process may be necessary for α and determine suitable value.If there are a lot of various process variablees that threshold value need be set, then this requirement may be dull and consuming time.

In another embodiment, can threshold value be set based on variable in learning phase observation.For example, if x＜μ-3 is σ, then can detect catalyst attrition.In this case, can by trained values block 80 with the variable storage that observed in storer 54.Therefore, in this embodiment, can determine detection threshold automatically, thereby can reduce the amount of manual configuration.Should be noted in the discussion above that except three, can also according to the variable that observes or detected use other multiple arbitrarily at standard deviation.In addition, although can calculate variance variable (variance variable) automatically by the SPM module, this variable can be the configurable parameter of user as training variable (for example by user I/O 66) input.

Return and the residual error monitoring

If pressure differential deltap P only changes when the high catalyst loss takes place, then the SPM algorithm can be suitable for surveying catalyst attrition.Yet if pressure differential deltap P is owing to other factors changes (for example, when pressure differential deltap P changes owing to changing load or other intended procedure situation), the SPM algorithm may trigger false alarm.In one embodiment, according to unitary operating conditions of fluid catalytic cracking or operational stage, may generate the feature collection (for example average, standard deviation etc.) that is derived from SPM more than.For example, if the cracking unit with two different load operation, then computing block 76 can be programmed to realize at a regular collection of first loading condiction and to realize second regular collection at second loading condiction.In this embodiment, can use two SPM pieces.Can activate one or another SPM piece based on the process condition of the loading condiction that detects or other expection.

Although can use a plurality of SPM pieces for single condition changing (for example, in the time only may having two loads), when having the operational condition of a plurality of expections, a plurality of SPM pieces may be inefficient.In this case, can use the recurrence (for example, set up regression model, monitor residual error then) of certain form to survey catalyst attrition.

Usually, at learning phase, from cyclone pressure differential deltap P (y) with to cyclone pressure differential deltap P (x ₁, x ₂..., x _m) have an image data in the process variable of certain influence.Can set up model to predict y value as the function of x:

$\hat{y}=f(x_1,x_2,\&CenterDot;\&CenterDot;\&CenterDot;,x_m)$

This model can be any model from simple multiple linear regression model to the more complicated model such as neural network model, and described simple multiple linear regression model for example is:

f(x ₁，x ₂，…，x _m)＝a ₀+a ₁x ₁+a ₂x ₂+…+a _mx _m，

Coefficient is according to any known process, and for example common least square (OLS), principal component regression (PCR), inclined to one side least square (PLS), variable subset are selected (VSS), SVMs (SVM) etc.) calculate.As following further argumentation, in case model is set up at monitor stages, this model promptly can be used for calculating residual error (between actual value and the predictor poor).If residual error exceeds certain threshold value, then can detect abnormal conditions.

Fig. 4 illustrates the embodiment that can be used to realize to return with abnormal operation detection (AOD) module 90 of residual error policing algorithm.AOD module 90 can comprise a SPM piece 92 and the 2nd SPM piece 94 that is connected to model realization piece 96.The one SPM piece 92 can be operated in the mode that is similar to the SPM module 70 shown in Fig. 3.Equally, a SPM piece 92 receives first process variable and generates first statistic data according to first process variable.As discussed above, this operation can comprise the statistical nature data that generation calculates according to first process variable, for example mean data, intermediate value data, standard deviation data, velocity of variation data, range data etc.This data can be calculated based on the moving window of first process variable data or based on the non-overlapping window of first process variable data.As an example, a SPM piece 92 can use first nearest process variable sample and preceding 49 first process variable sample to generate mean data.In this example, can generate the variable average at each the first new process variable sample that receives by a SPM piece 92.As another example, a SPM piece 92 can use the non-overlapping period to generate mean data.In this example, can use five minutes window (or certain other suitable period), thereby will generate a variable average in per five minutes.In a similar fashion, the 2nd SPM piece 94 receives second process variable, and generates second statistic data in the mode that is similar to SPM piece 92 according to this second process variable.In one embodiment, can only use one (for example, only using piece 92) in SPM piece 92 or 94.In another embodiment, can not use SPM piece 92 or 94.

Model realize piece 96 can the fs receive the expression cyclone differential pressure Δ P dependent variable Y and express possibility and Δ P had the set of independent variable(s) X of the process variable of certain influence.As below will describing in more detail, model realizes can using a plurality of data sets by piece 96, and (X Y) generates regression model for the model Y (for example pressure differential deltap P) as the function of X (for example influencing the above independent variable(s) of Δ P).

Model realizes that piece 96 can comprise an above regression model, and each regression model can use function at any range of X, the stated limit of X and/or a plurality of scopes of X, dependent variable Y is modeled as the function of independent variable(s) X.For example, can use single X variable to predict Y variable under all normal running situations.In this case, can use any known simple regression method.In another embodiment, can set up different models at different scopes.For example, in extendible homing method, can set up regression model at a plurality of scopes of independent variable(s) X.This usual way is at U. S. application No.11/492, further describes in 467, and this application is incorporated herein by reference.

In one embodiment, regression model can comprise or use linear regression model (LRM).Usually, linear regression model (LRM) uses function f (X), g (X), h (X) to wait certain linear combination that commercial run is carried out modeling, and usually, suitable linear regression model (LRM) can comprise the function of first order (for example Y=m*X+b) of X or the second order function of X (Y=a*X for example ²+ b*X+c).Certainly, also can use the function of other type, for example more the polynomial expression of high-order, sinusoidal function, logarithmic function, exponential function, power function etc.

After model is by training, can realize that piece 96 generates the predictor (Y of dependent variable Y in second operational phase based on given independent variable(s) X input with model _P).Under the unitary situation of fluid catalytic cracking, Y _PCan represent the differential pressure Δ P that predicts, and Y can represent actual measured value or the current observed value of differential pressure Δ P.Model is realized prediction Δ P (or the Y of piece 96 _P) can be provided for deviation detector 98.Deviation detector 98 can receive Δ P (or the Y of prediction of the regression model of piece 96 _P) and dependent variable input Y (actual measured value or the current observed value of expression Δ P).As a rule, deviation detector 98 can compare the pressure differential deltap P of actual pressure differential Δ P and prediction, whether significantly deviates from the pressure differential deltap P of prediction to determine actual pressure differential Δ P.If actual pressure differential Δ P significantly deviates from the pressure differential deltap P of prediction, then can indicate the abnormal conditions catalyst attrition to take place, taking place or may take place in the near future.Therefore, deviation detector 98 can generate and depart from designator.In some embodiments, designator can be the warning or the alarm of indication abnormal catalyst loss.

Difference between the pressure differential deltap P of actual pressure differential Δ P and prediction can be called residual error.Deviation detector 98 can be configured to only generate alarm after meeting or exceeding certain threshold residual value.Can use any means in the various known method to set up threshold value to survey abnormal catalytic loss condition.With above-mentioned SPM model class seemingly, threshold value can be the particular percentile of the Y value for example predicted, also can be based on the variance (variance) of the residual error of using training data to calculate.Before the alarm that the generation factory personnel is seen, can use the alarm logic (two or more continuous measurements that for example, exceed threshold value) of arbitrary form.

Those skilled in the art will recognize that, can revise AOD module 90 in every way.For example process variable data is before being received by SPM piece 92 and 94, can be filtered, reduction etc.In another embodiment, can not use SPM piece 92 and 94.In addition, although the model that uses in the piece 96 is illustrated as the value Y with single independent variable(s) input X, single dependent variable input Y and single prediction _P, but the model in the piece 96 can comprise the regression model that a plurality of variable Y (for example differential pressure of two or more cyclones) is modeled as the function of a plurality of variable X.Model in the piece 96 can comprise multivariate linear regression (MLR) model, principal component regression (PCR) model, least square (PLS) model, ridge regression (RR) model, variable subset are selected (VSS) model, SVMs (SVM) model etc. partially.In one embodiment, can carry out modeling, for example the differential pressure Δ P of reactor cyclone 20 to two differential pressures ₁Differential pressure Δ P with regenerator cyclone 31 ₂By this way, independent variable(s) collection X closes the differential pressure Δ P that can represent to influence simultaneously reactor cyclone 20 ₁Differential pressure Δ P with regenerator cyclone 31 ₂Process characteristic.

Fig. 5 illustrates the process flow diagram flow chart of surveying or predicting the exemplary method of the abnormal catalyst loss in the fluid catalytic cracking unit.Method 100 can use the example AOD module 90 of Fig. 4 to realize.At piece 101 places, can training pattern realize piece, for example model block 96.For example, can use independent variable(s) X and dependent variable Y data set training pattern disposing this model, thereby prediction is as the Y of the function of X.This model can comprise for example a plurality of regression models, and each regression model is modeled as Y at the different range of X the function of X.

Then, at piece 102 places, the model of being trained uses the value of the independent variable(s) X of its reception to generate the predictor (Y of dependent variable Y _P).Next, at piece 103 places, with the actual value of Y and corresponding predictor Y _PCompare, to determine whether Y significantly deviates from Y _PFor example, deviation detector 98 can receive the output Y of model block 96 _P, and should export Y _PY compares with dependent variable.If determine that Y significantly deviates from Y _P, then can generate and depart from designator at piece 104 places.For example, in AOD module 90, deviation detector 98 can generate designator.This designator can be to indicate to detect for example warning or the alarm that significantly departs from, or the signal, sign, message etc. of other type arbitrarily.

As following with more detailed argumentation, at model by initial training and generated the predictor Y of dependent variable Y _PAfterwards, can repeatable block 101.For example, if the setting point in the process changes, perhaps process operation other constantly, training pattern again.

The program control system of using with the AOD module

The fluid catalytic cracking unit can be operated as the part in a lot of group interconnection devices or a device in source mill, thus the forming process line.Usually, can use the program control system shown in Fig. 6 and 7 to control and manage this device.

Specifically, can realize that wherein the exemplary process factory 210 of abnormal situation prevention system comprises by an above interconnection of telecommunication network several controls and maintenance system together referring to Fig. 6.Particularly, the source mill 210 of Fig. 6 comprises an above program control system 212 and 214.Program control system 212 can be traditional program control system, for example PROVOX or RS3 system, or other Controlling System arbitrarily, comprise the operator interface 212A that is connected to controller 212B and I/O (I/O) card 212C, I/O (I/O) card 212C is connected to the various field apparatus such as simulation and highway addressable remote transmitter (HART) field apparatus 215 again.Program control system 214 can be a distributed process control system, comprises an above operator interface 214A who is connected to an above distributed director 214B by the bus such as industry ethernet.Controller 214B can be the DeltaV that is for example sold by Emerson process management company in Texas Austin city ^TMThe controller or the controller of other desired type arbitrarily.Controller 214B is connected to an above field apparatus 216 by I/O equipment, for example HART or

The Fieldbus field apparatus, or other comprises for example use arbitrarily

,

, , any intelligence or non-smart field devices in AS-Interface and the CAN agreement.Usually, process controller can with plant network system communication so that the information about operation (for example field apparatus operation) under the management of process controller to be provided, and the plant network system that uses during from the operation of regulate process controller receives set point signal.Be known that field apparatus 216 control physical process parameters (for example as actuator) or can measure physical process parameter (for example as transmitter).Field apparatus can be communicated by letter with controller 214B, with the receiving course control signal or data about the physical process parameter are provided.This communication can be undertaken by the analog or digital signal.I/O equipment can be that field apparatus receives message from process controller perhaps from the message of field apparatus reception in order to communicate by letter with process controller.Operator interface 214A can store and implementation red-tape operati person available instrument 217,219, and with the operation of control process, described instrument 217,219 comprises for example Control and Optimization device, diagnostician, neural network, tuner etc.

Further, maintenance system can be connected to program control system 212 and 214 or be connected to wherein individual equipment implement to safeguard and monitor activities.For example, maintenance calculations machine 218 can be connected to controller 212B and/or be connected to equipment 215 by the communication link or the network (comprising wireless or the handheld device network) of any desired, communicating by letter, and reconfigure equipment 215 in some cases or equipment 215 is implemented maintenance activitys with equipment 215.Similarly, maintenance applications can be installed on the above user interface 214A who is associated with distributed process control system 214, and, comprise the maintenance and the monitoring function of the data gathering relevant with execution with the operational stage of equipment 216 by described user interface 214A execution.

As shown in Figure 6, computer system 274 can realize at least a portion of abnormal situation prevention system 235, and particularly, configuring application program 238 and abnormal operation detection system 242 can be stored and realize to computer system 274.In addition, computer system 274 can realize alert/alarm application 243.

Fig. 7 illustrates the part 250 of the exemplary process factory 210 of Fig. 6, to describe a kind of mode that abnormal situation prevention system 235 and/or alert/alarm application 243 can be communicated by letter with the various device in the part 250 of exemplary process factory 210.

Generally speaking, abnormal situation prevention system 235 can with the field apparatus 215,216 that is positioned at source mill 210 alternatively, controller 212B, 214B (shown in Figure 7) and abnormal operation detection system (not shown in Figure 6) and/or the abnormal operation detection system in the computer system 274 242 communication in other expectation equipment or the device arbitrarily, disposing each in these abnormal operation detection system, and when these abnormal operation detection system monitoring, receive information about the operation of these equipment or subsystem.Abnormal situation prevention system 235 can be connected to some computer at least in the factory 210 or each in the equipment by rigid line bus 245 in the mode that can communicate by letter, perhaps alternately, can communicate to connect by any other expectation that comprises for example wireless connections, uses the special use of OPC to connect, connects, be connected to some computer at least in the factory 210 or each in the equipment such as the intermittence that relies on handheld device image data etc.Equally, abnormal situation prevention system 235 can by LAN or such as the public connection of Internet, phone connection etc. (shown in Figure 6 connect 246 for Internet) obtain with source mill 210 in field apparatus and install relevant data and by for example data of third party service provider collection.Further, but abnormal situation prevention system 235 can be connected to computer/equipment in the factory 210 with various technology and/or the agreement of signalling methods by comprising for example Ethernet, Modbus, HTML, XML, proprietary technology/agreement etc.

The part 250 of the source mill 210 shown in Fig. 7 comprises the distributed process control system 254 with an above process controller 260, and process controller 260 is by can being that the I/O card or the equipment 268 and 270 of I/O I/O equipment of any desired type of accordance with any desired communication or controller protocol is connected to an above field apparatus 264 and 266.Field apparatus 264 is shown as the HART field apparatus, and field apparatus 266 is shown as

The Fieldbus field apparatus, but these field apparatus can use the communication protocol operation of other any desired.In addition, in the field apparatus 264 and 266 each can be the equipment of any type, for example transmitter, valve, transmitter, steady arm etc., and opening, proprietary or other communication or programming protocol that can accordance with any desired be to be understood that I/O equipment 268 should be compatible mutually with field apparatus 264 and 266 employed expecting contracts with 270.

In any case, can be by being connected to process controller 260 by communication link or bus 276 such as above user interface of deployment engineer, process control operator, maintenance personnel, factory management person, supervisor's etc. factory personnel visit or computer 272 and 274 (can be the Personal Computer, workstation etc. of any type), wherein communication link or bus 276 can use the rigid line of any desired or wireless communication configuration and the use communication protocol any desired such as Ethernet protocol or suitable to realize.In addition, database 278 can be connected to communication bus 276, with as collection or store configuration information and online process variable data, supplemental characteristic, status data and with source mill 210 in process controller 260 and the historical data base operation of other data of being associated of field apparatus 264 and 266.Therefore, database 278 can be operating as configuration database, comprise the current configuration of process configuration module with storage, and download and store into the control configuration information of process controller 260 and other field apparatus 264 and 266 o'clock storage process Controlling System 254 at the control configuration information of program control system 254.Similarly, database 278 can be stored historical abnormal situation prevention data, comprises by the statistic datas (for example training data) of field apparatus 264 in the source mill 210 and 266 collections, according to the statistic data of being determined by field apparatus 264 and 266 process variables of gathering and the data of other type.

Process controller 260, I/O equipment 268 and 270 and field apparatus 264 and 266 be usually located at and be dispersed throughout in the severe sometimes environment of plant, and workstation 272,274 and database 278 often be arranged in can be by easily watch-keeping cubicle, maintenance room or other not too severe environment of visit such as operator, maintenance personnel.

Generally speaking, an above controller application program using a plurality of control modules different, independent execution or piece to realize control strategy can be stored and carry out to process controller 260.In the control module each can be made up of usually said functional block, wherein each functional block is a part or the subroutine in the overhead control routine, and combine operation (by being called communicating by letter of link) with other functional block, to realize the process control loop in the source mill 210.Be well known that the functional block that can be used as the object in the Object oriented programming agreement is implemented one of input function, controlled function or output function usually.For example, input function can be associated with transmitter, transmitter or other process parameter measurement device.Controlled function can be associated with the control routine of implementing controls such as PID, fuzzy logic.Output function can be controlled the operation of some equipment such as valve, to implement some physical function in the source mill 250.Certainly, also exist such as the mixing of model predictive controller (MPC), optimizer etc. and the sophisticated functions piece of other type.Should be understood that, although Fieldbus agreement and DeltaV ^TMSystem protocol uses control module and the functional block with Object oriented programming contract design and realization, but control module also can use the control programming scheme of any desired that for example comprises order functional block, ladder logic etc. to design, and be not limited to the functions of use piece or arbitrarily other specific programming technique design.

As shown in Figure 7, maintenance service station 274 comprises treater 274A, storer 274B and display equipment 274C.Storer 274B stores abnormal situation prevention application program 235 and the alert/alarm application of discussing at Fig. 1 243 in the following manner, can realize on treater 274A that promptly these application programs are to provide information by indicating meter 274C (or any other display equipment such as printer) to the user.

Each field apparatus in above field apparatus 264 and 266 can comprise the storer (not shown), with storage such as the routine of the relevant statistical data collection of an above process variable that is used to realize detected and/or the routine the following routine that is used for abnormal operation detection that will describe with test set.In above field apparatus 264 and 266 each can also comprise the treater (not shown), and this treater is used to carry out such as the routine that realizes statistical data collection and/or is used for routine the routine of abnormal operation detection.Statistical data collection and/or abnormal operation detection need not realized by software.On the contrary, those of ordinary skills will appreciate that this system can be realized by the arbitrary combination of software, firmware and/or hardware in an above field apparatus and/or the miscellaneous equipment.

As shown in Figure 7, some in the field apparatus 264 and 266 (and might be all) can comprise abnormal operation detection piece 280 and 282.Although the piece of Fig. 7 280 and 282 is illustrated as being arranged in one of equipment 264 and one of equipment 266, but these modules or similarly module can be arranged in the field apparatus 264 and 266 of arbitrary number, perhaps be arranged in other such as controller 260, I/O equipment 268,270 or Fig. 6 shown in the equipment of any apparatus and so on.In addition, module or piece 280 and 282 are at the scene in equipment 264 and 266 the random subset.

Generally speaking, piece 280 and 282 or the daughter element of these pieces from their residing equipment and/or from the data of miscellaneous equipment collection such as process variable data.In addition, piece 280 and 282 or the daughter element of these pieces can For several reasons variable data be handled and these data is implemented and analyze.In other words, piece 280 and 282 can be represented above-mentioned AOD module 70 or 90.Therefore, piece 280 or 282 can comprise one group of above statistic processes monitoring (SPM) piece or unit, for example piece SPM1-SPM4.

Should be appreciated that opposite although piece 280 and 282 is shown as including the SPM piece in Fig. 7, the SPM piece can be and piece 280 and 282 isolating autonomous blocks, and can be arranged in and corresponding piece 280 or 282 identical equipment, also can be arranged in different equipment.SPM piece discussed herein can comprise known Foundation Fieldbus SPM piece or compare with known Foundation FieldbusSPM piece to have SPM piece different or additional capabilities.Term used herein " statistic processes monitoring (SPM) piece " be meant collection such as process variable data data and these data are implemented some statistical treatment to determine piece or the element such as any type of the statistical measures of average, standard deviation etc.Therefore, this term is intended to cover software, firmware, hardware and/or other element that can implement this function, and no matter whether these elements adopt the form of piece, program, routine or the element of functional block or other type, also no matter whether these elements meet Foundation Fieldbus agreement or some other agreement such as agreements such as Profibus, HART, CAN.If desired, the fundamental operation of piece 250 can be to small part such as U.S. Patent No. 6,017, implementing like that or realizing described in 143, and this patent is incorporated herein by reference.

Although should be appreciated that further piece 280 and 282 is shown as including the SPM piece in Fig. 7, SPM piece ability is not that piece 280 and 282 is necessary.For example, piece 280 and 282 abnormal operation detection routines can use the process variable data of not handled by the SPM piece to operate.As another example, the data that provided by one that is arranged in miscellaneous equipment above SPM piece can be provided piece 280 and 282 separately, and these data are operated.As an example again, process variable data can be by being not to be handled by the mode that a lot of typical SPM pieces provide.Only as an example, process variable data can be by finite impulse response (FIR) such as the wave filter of bandpass filter or certain other type (FIR) or infinite impulse response (IIR) filter filtering.As another example, can cut down process variable data, thereby it is remained in the specific scope.Certainly, can make amendment, so that this different or additional processing power to be provided to known SPM piece.

Be depicted as the piece 282 of the Fig. 7 that is associated with transmitter, can have the line sniffing of connection unit, whether this connection line sniffing element analysis is connected with the circuit in definite factory by the process variable data of transmitter collection.In addition, piece 282 can comprise an above SPM piece or unit, for example can gather process variable in the transmitter or other data and the data of being gathered are implemented the piece SPM1-SPM4 of more than one statistical computations with for example average of the data determining to be gathered, intermediate value, standard deviation etc.Comprise four SPM pieces separately although piece 280 and 282 is shown as, the SPM piece that can have other arbitrary number in the piece 280 and 282 is to gather and definite statistic data.

Realize the AOD module

Fig. 3 and 4 AOD module 70 and 90 can realize in the program control system shown in Fig. 6 and 7 respectively.For example, AOD module 70 and 90 can realize in all or part of equipment at the scene, and field apparatus can be connected to any one or both in reactor cyclone 20 and the regenerator cyclone 31 then.For example, if use AOD module 90, then the SPM piece 92 and 94 of AOD module 90 is realized in the equipment 266 at the scene, and model realizes that piece 96 and/or deviation detector 98 can realize in process controller 260 or workstation 274 (for example by detection application program 242) or certain miscellaneous equipment.Similarly, the procedure block of AOD module 70 all realizes in the equipment (for example 264 or 266) at the scene, perhaps is distributed between field apparatus and the process controller.In a specific implementations, AOD system 70 or 90 can be implemented as functional block, and is for example above-mentioned and realizing The functional block of using in the program control system of Fieldbus agreement.This functional block can comprise also can not comprise SPM piece 92 and 94.In another embodiment, at least one in AOD 70 and 90 the piece can be implemented as functional block.

Because can use the differential pressure of cyclone 20 and 31 to survey catalyst attrition, any field apparatus of describing among Fig. 6 and 7 with differential pressure pick-up can be used to obtain the observed value of differential pressure.Yet it may be favourable using the field apparatus (Rosemount3051S that for example, has abnormal situation prevention) with built-in signal processing.Particularly, since the process control field apparatus have to far faster than the access right of the data of the speed sampling of host computer system (for example gathering the workstation of observed value from field apparatus) by process controller, so the statistical nature that calculates in the field apparatus may be more accurate.As a result, AOD that realizes in the field apparatus and SPM module can be determined better statistical computation than the piece that is positioned at the device external that process variable data wherein gathered at the process variable data of being gathered usually.

Should be noted that Rosemount 3051

The Fieldbus transmitter has the advanced diagnostics block (ADB) that possesses the SPM ability.This SPM piece can have following ability: the baseline mean of learning process variable and standard deviation, process variable and current average and the standard deviation of study are compared, if and in these each change more than user's specified threshold value, then trigger the PlantWeb warning.Suppose that differential pressure Δ P can not change owing to process enters other normal operating area, then the SPM function in the field apparatus can be configured to that operation is possible to survey catalyst attrition as AOD module (for example the AOD module 70) based on the description here.

Alert/alarm application 243 can be used to manage and/or route by AOD module 280 that can comprise AOD module 70 and/or 90 and 282 warnings of creating.In this case, when detecting catalyst attrition, can provide significant warning to people or the group (for example, operator, slip-stick artist, maintenance personnel etc.) of being responsible for monitoring and attended operation.Can provide the help of directiveness to help the people to solve situation by user interface with (for example on the workstation 272 or 274 that is connected to program control system).Can comprise following guidance to the corrective action that the user presents in response to warning: a) increase the pressure in the revivifier; B) repair cyclone; And/or c) uses heavier catalyzer.

AOD module 70 and/or 90 can provide information to abnormal situation prevention system 235 by other system in warning application program 243 and/or the source mill.For example, depart from designator and can be provided for abnormal situation prevention system 235 and/or alert/alarm application 243, with to the operator notification unusual condition by what deviation detector 98 or computing block 76 generated.As another example, after the model of AOD module 90 realizes that the model of piece 96 has been trained, the parameter of model can be provided for other system in abnormal situation prevention system 235 and/or the source mill, thereby makes the operator can check model and/or model parameter can be stored in the database.As an example again, AOD module 70 or 90 can provide X, Y and/or Y to abnormal situation prevention system 235 _PValue, thus make the operator when departing from (for example detect) check these values.

In program control system, AOD module 70 or 90 (realizing by field apparatus or process controller) can be communicated by letter with configuring application program 238, disposes AOD module 70 or 90 to allow the user.For example, module 70 or an above piece of 90 can have user's configurable parameter, and these parameters can be modified by configuring application program 238.

Although text is listed the detailed description to numerous different embodiment, should be appreciated that the scope of law of this description is limited by the literal in the claim of listing at this patent.It is exemplary that this detailed description should be construed as merely, and do not describe each possible embodiment, because describing that each possible embodiment is non-can not be promptly unactual.The technology that can use the current techniques within the scope that still falls into these claims or develop after the applying date of this patent realizes numerous interchangeable embodiment.

Claims (25)

1, a kind of method of surveying the catalyst attrition in the fluid catalytic cracker comprises:

Measure the differential pressure of the cyclone in the fluid catalytic cracker;

In first operational phase of fluid catalytic cracker, determine the initial average differential pressure of cyclone;

In second operational phase of fluid catalytic cracker, the current average differential pressure of monitoring cyclone; And

If the initial mean deviation of the described cyclone of current mean deviation pressure ratio of described cyclone is forced down more than threshold value, then determine the abnormal catalyst loss incident.

2, method according to claim 1 comprises the initial average differential pressure of using the statistic processes policing algorithm to determine described cyclone.

3, method according to claim 2 is included in a part that realizes described statistic processes policing algorithm at least one in field apparatus or the process controller.

4, method according to claim 2 comprises that described threshold value is set to the per-cent of described initial average differential pressure.

5, method according to claim 1 comprise the standard deviation of using the statistic processes policing algorithm to determine the initial average differential pressure of described cyclone, and described threshold value is set to the multiple of the standard deviation of initial average differential pressure.

6, a kind of method of surveying the catalyst attrition in the fluid catalytic cracking unit comprises:

The differential pressure of cyclone in the monitoring fluid cat cracker;

The process parameter set of the differential pressure of the described cyclone of monitoring influence;

Process parameter set based on the monitoring of the differential pressure of differential pressure of being monitored and the described cyclone of influence gathered generates regression model at learning phase;

Use described regression model to calculate the differential pressure of prediction;

If the difference between the differential pressure of the prediction of the current differential pressure of cyclone and cyclone greater than threshold value, is then determined the abnormal catalyst loss incident.

7, method according to claim 6 comprises and uses simple regression to generate described regression model.

8, method according to claim 6 comprises and uses extendible recurrence to generate described regression model that described extendible recurrence provides a plurality of regression models at a plurality of scopes of process parameter set.

9, method according to claim 6, wherein at least one subclass of process parameter set is the statistical nature data that calculate in the equipment at the scene.

10, method according to claim 6 comprises in the reactor cyclone of measuring fluid catalytic cracker or the regenerator cyclone at least one differential pressure.

11, a kind of equipment of surveying the abnormal catalyst loss in the fluid catalytic cracking unit comprises:

Set of sensors is used for periodically measuring the pressure reduction of fluid catalytic cracking unit cyclone;

Logic module, determine a period intercycle the statistical parameter set of the pressure reduction that records;

Rule module, the store instruction set;

Training module, the storage process parameter sets;

Computing module gathers to determine the abnormal catalyst loss incident based on the process parameter in instruction set in the described rule module and the described training module, and wherein said computing module generates indication when the abnormal catalyst loss incident takes place.

12, equipment according to claim 11, wherein said logic module is calculated average and the standard deviation of pressure reduction in a period.

13, equipment according to claim 11, wherein said training module comprises with the statistical parameter that periodically records and gathers the corresponding first benchmark survey parameter sets, and the wherein said first benchmark survey parameter sets is to determine in the initial learn stage of equipment.

14, equipment according to claim 13, wherein computing block during greater than the average pressure reduction determined in the initial learn stage, is determined unusual catalyzer incident at the pressure reduction that records.

15, equipment according to claim 13, comprise: according to set of first process parameter and the described fluid catalytic cracking of second process parameter set operation unit, and wherein in described fluid catalytic cracking unit during according to the first process parameter set operation, the described first benchmark survey parameter sets is determined at learning phase, and during according to the second process parameter set operation, the second benchmark survey parameter sets is determined at learning phase in described fluid catalytic cracking unit.

16, equipment according to claim 15, wherein said computing block is in described fluid catalytic cracking unit during according to the first process parameter set operation, determine unusual catalyzer incident based on the described first benchmark survey parameter sets, and during according to the second process parameter set operation, determine unusual catalyzer incident based on the second benchmark survey parameter sets in described fluid catalytic cracking unit.

17, a kind of equipment of surveying the abnormal catalyst loss in the fluid catalytic cracking unit comprises:

First input is used for receiving the data about the pressure reduction of fluid catalytic cracking unit cyclone;

Second input is used to receive the data about the process parameter set that influences described pressure reduction;

Model is realized the unit, is used for calculating the pressure difference of prediction based on described process parameter set;

Deviation detector compares pressure difference and the actual pressure differential value of prediction, and at the pressure difference of prediction and the difference between the actual pressure differential value time generation signal that exceeds threshold value.

18, equipment according to claim 17, wherein said model uses simple regression.

19, equipment according to claim 17, at least one subclass of wherein said process parameter set are the statistical nature data that calculate in the equipment at the scene.

20, equipment according to claim 17, the differential pressure of the reactor cyclone of wherein said fluid catalytic cracker or at least one in the regenerator cyclone is measured.

21, a kind of system that surveys the abnormal catalyst loss in the fluid catalytic cracking unit comprises:

Program control system comprises workstation, process controller and a plurality of field apparatus, and wherein said workstation, process controller and described a plurality of field apparatus are connected to each other in the mode that can communicate by letter;

The fluid catalytic cracking unit has reactor cyclone and regenerator cyclone, and wherein at least one field apparatus is suitable for measuring the pressure reduction of described reactor cyclone or described regenerator cyclone;

Abnormal operation detection equipment is suitable for receiving the data about measured pressure reduction, and access needle is to the normal running value set of pressure reduction, and the difference between measured pressure reduction and normal running value set generates warning when exceeding threshold value.

22, system according to claim 21 further comprises alert management equipment, and it is suitable for receiving described warning from described abnormal operation equipment, and shows the indication of catalyst attrition.

23, system according to claim 21 further is included in the configuring application program that moves on the described workstation, and it is suitable for communicating by letter with described abnormal operation detection, and described normal running value set is provided.

24, system according to claim 21 realizes among wherein said abnormal operation detection equipment in described a plurality of field apparatus or described process controller.

25, system according to claim 21, wherein said abnormal operation detection equipment uses a kind of algorithm in statistic processes policing algorithm or the regression algorithm, calculates the normal running value set of pressure reduction in the initial training stage.

Images