ZA201005500B - Novel process for controlling and regulating units for the skeletal isomerization of c5 cuts - Google Patents

Description

Field of the invention ok

The field of the invention is that of advanced control of industrial units. In an advanced control process, using the vocabulary of the skilled person, the industrial unit to be regulated is represented by a model which allows the actions to be undertaken to be anticipated, thereby reaching a degree of delicacy in the corrective actions which cannot be achieved by regulation using simple proportional-integrative-derivative (PID) controllers.

In the present case, i.e. that of skeletal isomerization units, the model of the unit is used to predict the coke content in real time using a combustion model.

A skeletal isomerization unit comprises a reaction zone and a catalyst regeneration zone. The catalyst is composed of spherical beads with a diameter in the range 1 to 3 mm moving in a continuous manner from one zone to the other.

During the chemical reactions which occur in the reaction zone, the catalyst becomes charged with coke. That coke is chemically bonded to the catalyst and can only be removed effectively by burning. That coke also gives rise to a loss of activity of the catalyst.

A unit for the skeletal isomerization of C5 cuts has an optimum yield of isomerate which is linked to a certain coke content on the catalyst. The coke content is defined in the context of the present invention as the ratio of the quantity of coke burned during regeneration per unit of time over the throughput of the catalyst moving between the reaction zone and the regeneration zone.

This model is combined with an estimation of the throughput of moving catalyst which is based on a correlation between this throughput of catalyst and the pressure difference taken between two points in the line returning the catalyst from the regeneration zone to the reaction zone, termed the transport line. :

A linear multivariable predictive controller (hereinafter termed MVACQC) operates in a closed loop, i.e. starts by measuring all or part of the output . parameters and compares them with fixed objectives (setpoint values, upper and lower limits). The controller calculates and applies to the industrial unit the input : values required to attain or maintain the objectives for the setpoint values.

The MVAC controller can thus calculate the value of the catalyst throughput (DC) which allows the coke content to remain close to the optimum coke content (TCopt) in real time, while integrating various perturbations (PERT)

which may be linked to the operating conditions or to the throughput of the feed to be treated.

The present invention thus consists of a process for controlling and regulating units for the skeletal isomerization of C5 cuts which optimizes in real time the production of the isomerate by integrating all of the possible perturbations of said unit and essentially acting on the throughput of the catalyst.

Examination of the prior art

The prior art in the field of control and regulation of skeletal isomerization units is close to that concerning units for the catalytic reforming of gasolines, since both cases concern moving bed units. In this regard we can cite patent FR 2 837 113, which describes a method for control and regulation of regenerative reforming units, more particularly of the catalyst regeneration zone, by selecting a parameter, termed the characteristic parameter, and by acting on the throughput of the moving catalyst in order to maintain said characteristic parameter at a desired value or within a desired range.

The parameters considered as characteristics of the function of the regeneration zone are the coke content of the catalyst entering the regeneration zone or the burning capacity, defined as the product of the throughput of the moving catalyst and the coke content in the catalyst taken at the inlet to the regeneration zone.

The present invention differs from that of the cited patent essentially in : that the parameter which is maintained constant in the regeneration zone is an original parameter which is expressed as the throughput of burned coke in the regeneration zone divided by the throughput of the moving catalyst. This parameter, hereinafter termed the “coke .content”, is specific to skeletal isomerization units in the sense that the production of isomerate has an optimum directly linked to a particular value for this coke content, which value is not the same for all units but depends on the unit under consideration.

The present invention thus exploits this phenomenon of an optimum by seeking to maintain the coke content in real time at its optimum value, i.e. the value which corresponds to the maximum yield of isomerate.

In order to achieve this aim, it is necessary to know or to independently estimate the throughput of the moving catalyst and the coke content, this latter necessitating determination of the coke burned during catalyst regeneration.

Brief description of the figures

Figure 1 represents a typical curve showing the optimum isomerate production as a function of the variable “coke content”, i.e. the throughput of the coke burned in the regeneration zone over the throughput of the moving catalyst.

Figure 2 represents a layout of a skeletal isomerization unit showing the reaction zone and the regeneration zone linked by a pneumatic transport line for the catalyst.

Figure 3 is a flow chart of the sequence of operations carried out in 5 automatic steps, using a MVAC predictive controller, to carry out the control and regulation of the present invention.

Figure 4 shows the improvement in the yield of isomerate by application of the control and regulation process of the invention.

Brief description of the invention

The present invention can be defined as a process for advanced control and regulation applicable to units for the skeletal isomerization of C5 cuts with a view to producing a CS cut that is rich in paraffins and branched olefins hereinafter termed an isomerate, or more precisely a C5+ isomerate. The term “C5+ should be understood to designate an oil cut comprising hydrocarbon chains containing 5 or more than 5 carbon atoms.

Units for the skeletal isomerization of C5 cuts are generally units with moving beds where the reaction takes place, i.e. with a continuous gravitational flow of catalyst through which the feed to be treated passes in a radial manner.

The catalyst is present in the form of spherical particles approximately 1 to 3 mm in diameter. During the course of the reaction, coke is deposited on the catalyst due to unwanted reactions, in particular reaction effluent polymerization reactions.

This coke deposit is accompanied by a loss of activity of the catalyst and requires it to be regenerated; this is carried out in a regeneration zone positioned in series with respect to the reaction zone.

The reaction occurs essentially in accordance with 2 parallel mechanisms:

1- direct isomerization; : 2- dimerization to C10, then cracking of the C10 formed with the formation of coke. :

The coke provides the reaction with a certain selectivity, but deactivates : the catalyst by obscuring or blocking the pores.

Below the optimum, the catalyst is too active and an excess of cracking occurs, producing light species, which increases the pressure and is unfavourable to the selectivity.

Above the optimum, there is too much coke, the catalyst loses its activity, and the conversion drops.

Regeneration of the catalyst essentially consists of controlled combustion, in air or oxygen, of the coke present in the porosity of the catalyst and at its surface. For regeneration, the catalyst moves in a moving bed, i.e. under slow gravitational flow, from the top to the bottom, both in the reactor and in the regeneration zone.

The catalyst circuit may be described as follows:

Catalyst deriving from the regeneration zone is introduced to the head of the reactor in which it flows in a moving bed. It is collected from the bottom of the reactor by a set of extraction legs which join at a single volume (the lift pot) then is transported to the head of the regeneration zone by pneumatic transport using a transport gas.

The catalyst also flows as a moving bed inside the regeneration zone, where it is recovered by a set of extraction legs which join in a second lift pot before being re-introduced to the head of the reactor via a second pneumatic transport.

The time for the catalyst to move round the unit is of the order of 72 hours, which is an important value explaining the difficulty of regulating this type of unit which can be qualified as non-reactive, in the sense that it has a very high stabilization time.

The notion of stabilization time applied to the present invention can be : highlighted by an experiment consisting of causing a stepped variation in the pressure drop taken between two points on the catalyst transfer line and recording the time taken by the unit to stabilize at a new coke content.

a : The present invention can provide fine control and regulation despite a stabilization time of the order of 72 hours (it is of the order of a few hours for the majority of refining units), with an interval between two interventions of the MVAC controller which is usually in the range 20 to 30 minutes. 5 The ratio of the stabilization time of the unit to the intervention time as regards the throughput of the catalyst is of the order of 200, which is much higher than values which are normally encountered in processes for control and regulation of refining units, and guarantees complete stability of the unit which does not have the time to accommodate major drifts.

When the throughput of the catalyst is mentioned in the remainder of the text, this means the throughput for movement of said catalyst taken between the reaction zone and the regeneration zone.

The aim in operating such a unit is thus at all times to be as close to the optimum coke content corresponding to the maximum production of isomerate as possible. However, knowing this optimum at all times necessitates real time monitoring of the throughput of the catalyst on the one hand and of the quantity of coke deposited on the catalyst on the other hand.

The present invention provides an automatic method for monitoring these two parameters using a coke combustion model which can estimate the burned coke, and a correlation which allows the throughput of the catalyst to be estimated as a function of the pressure drop along the line for transporting said catalyst from the regeneration zone to the reaction zone.

In a variation of the present invention, the correlation allowing the throughput of the catalyst to be estimated is based on the notion of the time interval separating two successive transfers of catalyst to the regeneration zone.

This variation concerns units in which the regeneration zone functions sequentially. Rh

The unit for skeletal isomerization of a C5 cut comprises a reaction zone and a catalyst regeneration zone, the catalyst moving in a continuous loop from the reaction zone to the regeneration zone, in which the coke content on the catalyst (TC) in the reaction zone is maintained constant close to an optimum value, denoted TCopt, by an advanced control system employing a predictive controller denoted MVAC.

The method for control and regulation according to the present invention employs the following succession of steps:

Step 1: estimation of throughput of moving catalyst (DC);

Step 2: estimation of throughput of burned coke (CB);

Step 3: calculation of coke content (TC), the coke content being by definition the ratio of the throughput of burned coke to the throughput of catalyst;

Step 4: daily tuning of the coke content as a function of laboratory analyses (LAB);

Step 5: adjustment of the catalyst throughput in order to attain the optimum value for the coke content (TCopt) to within plus or minus 10%, said adjustment being carried out starting from an increment in the pressure drop IAP formulated by MVAC, and the frequency of the adjustments being of the order of 200 times the natural frequency for stabilization of the unit.

The series of the above steps is carried out automatically and in real time.

In a preferred variation of the present invention, the optimum value for the coke content is approached to within plus or minus 5%. In other words, the value for the coke content is thus maintained in real time close to said optimum value to within plus or minus 5%.

In another preferred variation of the invention, the time interval separating two successive adjustments of the throughput of the catalyst in accordance with step 5 is in the range 20 to 30 minutes.

Detailed description of the invention

The description will be made with the aid of the accompanying Figures 1, 2 and 3.

Figure 4 will be described in the example forming part of the application.

Figure 1 represents a typical curve showing the optimum production of isomerate as a function of the variable “coke content”, i.e. the throughput of burned coke in the regeneration zone over the throughput of moving catalyst.

This curve of experimental origin (the squares represent the experimental points) highlights the optimum isomerate production expressed as a percentage for a value for the coke content in the range 4.3 to 4.8. The variable “coke content” is dimensionless since it is a ratio of throughputs using the same units (for example kg/hour).

: oo 7

E_2010/ 05500

This curve shows the advantage of an advanced control and regulation device for units for the skeletal isomerization of C5 cuts. . In fact, the optimum isomerate yield is relatively flat since it varies between 52% and 54% for a corresponding variation in coke content in the range 3.8t05.3. Thus, it is essential to maintain the coke content at about the value corresponding to the optimum yield of isomerate (i.e. 53.7% on the curve of

Figure 1), i.e. in a range of between 4.3 and 4.8.

This range is substantially narrower than the “natural” variational range for the coke content, i.e. without the control and regulation device of the present invention.

Figure 2 represents a layout of a skeletal isomerization unit showing the reaction zone and the regeneration zone connected via the pneumatic catalyst transport line.

The catalyst deriving from the regeneration zone (II) is introduced into the reactor via the introduction pot 1. It moves in the reactor I where it encounters the feed 10 generally introduced in a radial manner and moving from the periphery towards the centre of the reactor. The effluents 11 are recovered in a central collector (not shown in Figure 2).

A skeletal isomerization unit employs catalyst and reaction fluid movements of the same type as that which is found in regenerative reforming units, a more complete description of which can be found in FR 2 837 113, for example.

The catalyst is recovered from the bottom in a transport pot 2 from which it is conveyed by pneumatic transport through the line 5 to the pot 3 located at the head of the regeneration zone (II). The catalyst is then regenerated in the regeneration zone (II) which essentially consists of combustion of the coke deposited on the catalyst after it has passed through the reactor (I).

The regeneration zone may include other steps such as reduction of the - catalyst with hydrogen before re-introducing it into the reactor.

The catalyst is returned to the head of the reactor (I) by pneumatic transport by means of the transport line 6. :

i 8

The pressure drop, delta P, which serves to evaluate the throughput of the moving catalyst, may be measured on the transport line 5 or on the transport line 6, as shown in Figure 2.

In a variation of the present invention, the throughput of the catalyst may be measured by the variation in the weight of the storage pot 3 which is thus * mounted on scales 7.

Figure 2 also shows the effluent sampling points 8 of the regeneration zone (II) which are necessary for calculating the thermal balance of said zone and for estimation of the throughput of the burned coke.

Figure 3 is a flow chart of the series of operations carried out automatically by the MVAC controller in 5 automatic steps, in order to effect the control and regulation of the present invention.

In this figure, the abbreviations have the following meanings:

DC designates the throughput of the catalyst;

CB designates the throughput of the burned coke;

TC designates the coke content;

TC* designates the coke content corrected after information originating from the laboratory; : TCopt designates the optimum coke content;

LAB designates experimental information regards the coke content originating from a laboratory which is usually the refinery’s laboratory;

PERT designates any type of perturbation affecting the unit, in particular the inlet feed throughput;

DP designates the pressure drop measured between two points on the catalyst transport line;

BCOMB designates the combustion balance produced from an analysis of fumes from the catalyst regeneration zone;

TC*-TCopt is the difference between the corrected coke content value and the optimum value;

MVAC designates the linear multivariable predictive controller;

IAP designates the increment on the pressure drop formulated by the

MVAGC,;. :

AP* designates the new value of AP after incrementation.

i SL 9

This series of operations was described in the brief description of the unit; it will now be supplemented as follows:

Concerning step 1: estimation of throughput of moving catalyst (DC)

The estimation of the throughput of moving catalyst is made from a measurement of the pressure drop taken between two points located on the transport line 5 or the transport line 6.

It concerns the production of an experimental curve linking the catalyst throughput (DC) to said pressure drop. In a pneumatic transport line, the pressure drop is directly linked to the mean concentration of catalyst moving in the line.

Knowing the speed at which the transport gas is moving, and taking any slip factor into account, the throughput of the moving catalyst can be deduced therefrom.

This type of correlation is well known to the skilled person and will not be developed any further below.

In general, for catalyst throughputs in the range 500 to 1500 kg/h, the variation corresponding to the pressure drop is practically linear. It is thus easy to provide a simple mathematical expression for this variation and of introducing it as a computational element into the MVAC controller .

Further, the measurement of the time for filling or emptying a storage pot 3 located at the head of the regeneration zone (II) also provides an evaluation of the mean throughput of the moving catalyst. The value estimated by correlation from the pressure drop and the value resulting from weighing the pot 3 should not differ too greatly. The two values are assumed to be in agreement if the difference between them is less than 15%, preferably less than 10%.

In the contrasting case, i.e. if the difference between the two values for the catalyst throughput is strictly more than 15%, the controller carries out an operation known as tuning which consists of correcting the value derived from the pressure drop so that it more closely approaches the value derived from weighing the storage pot. In practice, the value retained by the controller is then the value derived from weighing the storage pot to within plus or minus 5%.

If a difference of more than 10% is obtained between the value derived from the correlation of the pressure drop and that resulting from weighing, the tuning operation is not carried out and the controller keeps the value derived from the pressure drop correlation. :

Concerning step 2: estimation of the quantity of burned coke and thus of the coke content (CB)

The quantity of burned coke (CB) is estimated by a thermal balance about the regeneration zone (II) starting from the inlet gas throughput (generally air), the effluent throughput (CO, CO,, Ny) and its composition, this information set being denoted BCOMB in Figure 3.

This is a standard computation which is well known to the skilled person, which will not be described in more detail. This results in an evaluation of the burned coke throughput (CB) in the regeneration zone.

Concerning step 3: estimation of coke content (TC)

Step 3 requires no particular comments since it concerns producing the ratio of the result of step 2 (CB) over the result of step 1 (DOC).

Concerning step 4: daily tuning of coke content as a function of laboratory analyses (TC*)

Daily tuning of the coke content is carried out using analyses carried out in the laboratory (LAB), which provide an experimental value with a mean frequency of one value per day.

Two cases are possible: ¢ cither the value from the laboratory is close to the value obtained at the end of step 3, i.e. equal to said value to within plus or minus 5%, and the value derived from said step 3 is not modified; ¢ or the value from the laboratory differs from the value obtained at the end of step 3 by more than 5%, and a correction has to be made to the value derived from said step 3. The correction becomes effective at the end of a reliability test which is based on the notion of the statistical variation of the values for the coke content derived from the laboratory. The corrected value for the coke content is : denoted TC*.

These values are averaged and must remain within one standard deviation arising from a statistical treatment of a large number of values;

ERY e if the actual value for the coke content derived from the laboratory (LAB) does not fall within the mean of values already obtained, it - is not taken into account and the value for the coke content derived from step 3 is not corrected, * if the value for the coke content derived from the laboratory (LAB) falls within the normal statistical dispersion of the previously : obtained values, and also differs by more than 5% from the value derived from step 3, then the value for step 3 is replaced by the value for the coke content derived from the laboratory (LAB).

Concerning step 5: adjustment of catalyst throughput in order to attain the optimum value for the coke content to within plus or minus 10% (TCopt)

Step 5 necessitates an experimental curve such as the curve represented in

Figure 1, giving the yield of isomerate as a function of the coke content.

Two cases present themselves: e cither the value for the coke content obtained at the end of step 3 falls within the interval corresponding to the optimum for the isomerate yield, and there is thus no correction to be made - contingent upon the result of step 5; . or the value for the coke content obtained at the end of step 3 does not fall within the interval corresponding to the optimum of the isomerate yield, and a correction has to be made on the catalyst throughput to bring the value for the coke content back into the optimum range.

This step will be illustrated for actual values in the example below. ol

Example in accordance with the invention

In order to illustrate the present invention in an actual case, we shall take an example corresponding to a skeletal isomerization unit treating a feed constituted by a C5+ hydrocarbon cut with a distillation range in the range 80°C to 200°C.

The reaction zone was constituted by a reactor connected to the regeneration zone via a pneumatic transport line.

The regeneration zone was constituted by a reactor in which combustion of the burned coke deposited on the surface of the catalyst was carried out by injection of air.

The catalyst was constituted by particles approximately 2 mm in diameter.

The unit was equipped with a two-point pressure probe located on the transport line for the catalyst connecting the bottom of the treatment to the regeneration zone (line 6 in Figure 2).

The mean throughput of the moving catalyst was: 820 kg/hour;

The mean feed throughput was: 90 m*/hour;

The mean temperature in the reactor was: 420°C;

The composition of the feed is given in the table below:

Sw pe | We | @0

The mean time separating two significant fluctuations in the feed flow rate, : 15 i.e. more than 10% of the mean value for said throughput, was typically 6 hours.

Step 1: estimation of moving catalyst throughput (DC)

Measurements have shown that the mean throughput for the catalyst of : 820 kg/h corresponded to a mean pressure drop of 23 kPa (kPa designates 10° pascals). These values are given by way of indication as they depend on each umit, in particular on the exact shape of the transport line.

The static gain was calculated as the ratio of the variation in the catalyst throughput to the variation in the pressure drop in a given interval. From the static gain and assuming that the dead time in the present case was 0, the time constant for the process was determined, in the present case 10 minutes. Co

Step 2: estimation of the quantity of burned coke or throughput of burned coke (CB) a

This was essentially a thermal balance carried out about the regeneration zone starting from the quantity of air introduced and the quantity of combustion gas derived from the regeneration gas as well as their CO, CO; analysis, since the coke or carbon deposited on the catalyst entering the regeneration zone is found at the outlet in the form of CO and CO,.

This balance per se is not of an inventive nature and will not be developed further.

It resulted in the quantity of burned coke in kg/h, in this case 35.3 kg/h.

Step 3: computation of coke content (TC)

The coke content is by definition the ratio of the throughput of burned coke to the catalyst throughput. It was thus obtained as the ratio of the estimation made in step 2 to the estimation made in step 1; in the present case it was 35.3/820 =43x10%0r43asa significant figure.

It should be pointed out that it may be necessary here to change the variable for the catalyst throughput so that the values for the catalyst throughput and its inverse (1/catalyst throughput) are numerically quite close, so that the result for the ratio is numerically close to 1, in order to allow the linear multivariable predictive controller to approximate the inversion operation (1/catalyst throughput) by a static gain (of the order of -1.22) with a precision of + or -2%. ’

Step 4: daily tuning of coke content as a function of the analyses carried out in the laboratory (TC*)

If the value for the coke content derived from the laboratory is in the range plus or minus 5% with respect to the value calculated in step 3, then no significant correction needs to be made to the calculated value. If not, the estimation is corrected from the bias between the estimation (when the sample is taken by the laboratory) and the measurement of the laboratory analysis.

Step S: adjustment of catalyst throughput in order to reach the optimum value for the coke content to within plus or minus 10% (TCopt)

Step 5 necessitates an experimental curve such as the curve represented in

Figure 1, giving the isomerate yield as a function of the coke content (TC).

i 14

In the present example, since the result of the coke content from step 3 is 4.3, this coke content has to be brought back into the range 4.4 to 4.6. This necessitates a correction of the catalyst throughput which must change from 820 to 785 kg/h by acting on the pressure drop which changes from a value of 23 kPa t022.3kPa.

The stabilization time for the unit is of the order of 1 to 2 days in the absence of any regulating action. However, now there is an average of 3 adjustments per hour to the moving catalyst throughput. This thus means that regulation of the unit is carried out at a much higher frequency than the natural stabilization frequency of the unit, in a ratio of approximately 200 to 1. This very high frequency for the adjustments guarantees complete stability of the unit.

Figure 4 allows the effect of control and regulation of the unit of the present invention to be understood:

The curve in dashed lines corresponds to the dispersion of the coke content which would be produced in the absence of regulation. This dispersion of the values for the coke content (between 2.6 and 6.6) results in a mean value for the isomerate yield of 51.3%. The solid line curve corresponds to the dispersion of the coke content with the regulation of the invention in place. Compared with the dashed curve, the solid line curve is narrower about values in the range 4.1 to 4.9.

This results in a mean value for the isomerate yield of 53%, i.e. an increase of approximately 1.7 points, which is very significant for an industrial unit.

The regulation of the present invention thus very substantially improves the performances of the unit while providing greater stability. :

Claims (6)

i] Co lm CLAIMS

1. A method for advanced control of a unit for skeletal isomerization of a C5 cut comprising a reaction zone and a catalyst regeneration zone, the catalyst moving in a continuous loop from the reaction zone to the regeneration zone, in which the coke content on the catalyst (TC) in the reaction zone is maintained at close to an optimum value (TCopt), said method employing the following steps carried out automatically and in real time using a MVAC predictive controller, the socalled linear multivariable predictive controller (hereinafter termed MVAC) operating in a closed loop, i.e. measuring all or part of the output parameters and comparing them with fixed objectives (setpoint values), the predictive controller (MVAC) calculates and applies to the industrial unit the input values required to attain or maintain the objectives for the setpoint values (CV), Step 1: estimation of throughput of moving catalyst (DC); Step 2: estimation of throughput of burned coke (CB); Step 3: calculation of the coke content (TC), the coke content by definition being the ratio of the throughput of burned coke to the throughput of : catalyst; Step 4: daily tuning of the coke content as a function of analyses carried out in the laboratory (LAB); Step 5: adjustment of the throughput of catalyst in order to attain said optimum value for the coke content (TCopt) to within plus or minus 5 %, said adjustment being carried out starting from an increment in the pressure drop IAP formulated by the MVAC, at a frequency of two hundred times the natural frequency for stabilization of the unit.

2. A method for advanced control of a unit for the skeletal isomerization of a CS cut according to claim 1, in which the time interval separating two successive adjustments of the catalyst throughput in step 5 is in the range 20 to 30 minutes.

3. A method for advanced control of a unit for the skeletal isomerization of a CS cut according to claim 1, in which step 1 for estimation of the moving catalyst throughput is carried out from a measurement of the pressure drop between two points located on the line for transport of the catalyst from the outlet from the regeneration zone to the inlet to the reaction zone.

4, A method for advanced control of a unit for the skeletal isomerization of a C5 cut according to claim 1, in which step 1 for estimation of the moving catalyst throughput is carried out from a measurement of the time separating two successive transfers of catalyst from the storage pot to the regeneration zone.

5. A method for advanced control of a unit for the skeletal isomerization of a C5 cut according to claim 1, in which step 2 for estimation of the burned coke throughput is carried out from a combustion balance over the regeneration zone.

6. A method for advanced control of a unit for the skeletal isomerization of a C5 cut according to claim1, substantially as herein described and exemplified and/or described with reference to the example and/or described with reference to the accompanying drawings. Dated this oo day of August Stole; Patent A / Agent for the Applicant