Conventional Control Structure

Assuming a liquid-phase reaction, constant density, perfect control of the reactor holdup and temperature and perfect control of the recycle and product purities, the system is described at steady state by the following equations  

In Eqs. (4.1) to (4.6), k is the reaction rate constant. Fk and zk are the molar flow rate and reactant mole fraction in stream k. Note that the assumption of constant density implies that all streams have the same total molar concentration, which is 

At this point it is useful to introduce dimensionless variables. Thus, we choose as the reference value the flow rate of the feed stream at the nominal production rate, F and define the dimensionless flows  

the flow rate and composition of the reactor-outlet and the recycle flow rate are given by  

Note that the overall mass balance requires the equality of the feed and product flow rates, F0 = F4. Consequently, the Damkohler number accounts for the production rate (F ), reactor design (V) and reaction kinetics (k). 

---

These same notions can be extended to an entire plant in which several unit operations are connected together. The HDA process for hydrodealkylation of toluene to form benzene is a good example of where an eigenstructure can be found that provides a more easily and simply controlled plant. See Fig. 8.15. Assuming that the toluene feed rate to the unit is fixed, this plant has 22 valves that must be set. There are 11 inventory loops (levels and pressures), so they require 11 valves. One possible conventional control structure is shown in Fig. 8.15. 

The performance of the convention control structure, in which cooling water flow is manipulated to control temperature, is shown in Figure 3.52. The disturbance is the same increase in cooling water temperature. Feed flowrate is constant. The cooling water flowrate more than doubles to control reactor temperature, but the temperature is returned to the desired value in about 2 h. The peak deviation in temperature is less than 0.6 K. Controller settings are those given in Table 3.2 for the 95% conversion case with a 330 K reactor temperature (the integral time is 50 min). 

Figure 4.3 First-order reaction in a CSTR/separation/recycle system, (a) Conventional control structure, (b) Production rate F4 versus the reactor-inlet flow rate F,.

In this section we replace the CSTR by a plug-flow reactor and consider the conventional control structure. Section 4.5 presents the model equations. The energy balance equations can be discarded when the heat of reaction is negligible or when a control loop keeps constant reactor temperature manipulating, for example, the coolant flow rate. The model of the reactor/separation/recycle system can be solved analytically to obtain (the reader is encouraged to prove this) ... 

Figure 2.6 Conventional control structure with fixed reactor holdup,...

Conventional control structure. As shown in Fig. 2.6, the following control loops are chosen ... 

The process specifications on raw material speed through furnaces coils imposed the use of two or four parallel passes, e.g. the fumaees from the atmospherie distillation unit, vacuum distillation unit, catalytic reforming unit, coker unit, catalytic cracking unit. The conventional control structure of radiant section for a typical tubular furnace from the atmospheric distillation unit (output capacity 3.5 Mt/year) is presented in figure 1 [1]. Because the conventional temperature control system only controls one outlet temperature or in the best case the temperature of the mixing point, in current operations there are several situations [1, 2, 3] ... 

Fig. 1. The conventional control structure of a furnace from the atmospheric distillation unit.

The control levels introduced above are presented in the figure 3. following the structure of a RCS node. It can be observed that the node in any level complies with the RCS node. As a preliminary conclusion it can be said that the conventional control structure is RCS compliant, or can be considered as an implementation of it. So is there anything new under the sun , whafs the benefit of using RCS (or other type of cognitive architecture) for process control ... 

Let s Fo and V represent nominal feed flow rate and reaction volume. For the conventional control structure, Fig. 13.21-left displays the flow rate to separation vs. 

A sensitivity study demonstrates that systems with small reactor or slow reaction rate (Da < 2) are better controlled by the Luyben s structure. However, systems with large reactor or fast reaction rate (Da > 3) are better controlled by the conventional structure. To explain this fact, we recall that close to Da=l the system is sensitive to Da. For the conventional control structure, the plant Da number represents a disturbance, as it contains Fq. Because low sensitivity is required, designs with Da large perform better. For the Luyben control structure, the plant Da number is a manipulated variable through the reaction volume F. Consequently, disturbances can be rejected with small effort if Da is small. It should be also noted that high purity separation (zj = 1) improves the performance of the conventional control structure. It is worthy to mention that the design analysed by Luyben (1994) had Da= 1.12. Hence, poor performance of the conventional control structure is expected. 

Equations (6.7) and (6.8) can be combined to yield Eq. (6.11), which shows how reactor composition z must change as fresh feed flow rate Fq and fresh feed composition zo change when the conventional control structure is employed (i.e., reactor volume is constant). 

In the last three chapters, we have developed a number of conventional control structures dual-composition, single-end with RR, single-end with rellux-to-feed, tray temperature control, and so on. Structures with steam-to-feed ratios have also been demonstrated to reduce transient disturbances. Four out of the six control degrees of freedom (six available valves) are used to control the four variables of throughput, pressure, reflux-drum level, and base level. Throughput is normally controlled by the feed valve. In on-demand control structures, throughput is set by the flow rate of one of the product streams. Pressure is typically controlled by condenser heat removal. Base liquid level is normally controlled by bottoms flow rate. 

There are a host of conventional control structures, and the best choice depends on a number of factors, some of which we have already discussed the shape of the temperature profile and the sensitivity of the several flow ratios to feed composition. Economics obviously have an impact, mostly in terms of energy consumption versus control structure complexity. The control structure that minimizes energy is dual-composition control. But it is more complex than a simple single-end control structure. Therefore, we must evaluate how much money is lost by using a more simple structure. In areas of the world where energy is inexpensive, both the optimum design and the appropriate control structure are different than in areas with expensive energy. 

In this section, comparisons of the proposed unusual control structure with three more conventional control structures are presented. 

Reflux Ratio Structure. If we ignore the results of the analysis that shows that the R/F structure can handle feed-composition disturbances better than the RR structure, we can set up another conventional control structure in which reflux-drum level is controlled by reflux flow rate, Stage 3 temperature is controlled by reboiler heat input, and distillate flow rate is ratioed to reflux flow rate. This structure is commonly used in many distillation systems and would be expected to provide stable regulatory control but result in more deviation of product purities for feed-composition disturbances. 

The structure handles large disturbances in both feed composition and feed flow rate. Its performance is much better than those of other conventional control structures. 

Several complex distillation systems have been examined in this chapter. The complexity can arise in either the vapor-liquid equihbrium or in the configuration of multiple interconnected units. Conventional and not-so-conventional control structures have been applied. 

So the basic conventional control structure selected has reflux-to-feed ratio. This is implemented in Aspen Dynamics using a multiplier block (R/F) with one input being the molar flow rate of feed and the other input the specified reflux-to-feed molar ratio. Since Aspen Dynamics has the rather odd limitation of only being able to directly specify the mass flow rate of the reflux, a flow controller must be installed whose process variable signal is the reflux molar flow rate, and whose output signal is the reflux mass flow rate. This flow controller is put onto cascade with its set point signal coming from the R/F multiplier. 

CONVENTIONAL CONTROL STRUCTURE SELECTION 449 50% Feed increase TC with and w/o QR/F ratio... 

A conventional control structure for this process, which works for low to moderate activation energies, is shown in Figure 9. The flowrate Fob of gaseous fresh feed of B is manipulated to control system pressure. The flowrate Fqa of gaseous fi esh feed of A is ratioed to Fob and the ratio is reset by the composition controller, which maintains the composition of A in the circulating gas stream at yuA = 0.5. A bypass stream around the FEHE controls the furnace inlet temperature. Furnace firing controls reactor inlet temperature. Separator drum temperature is controlled by heat removal in the condenser (typically the cooling water valve is wide open to minimize drum temperature). Liquid product comes off on drum level control. 

High sensitivity of recycles with respect to fresh feed, known as snowball effect, can be avoided by designing a sufficiently large reactor. For example, in case of first-order reaction, the conventional control structure based on self-regulation works well if Da > 3. 

The design and control of a maximum-boiling azeotropic system has been studied in this chapter. Extractive distillation is shown to be capable of producing quite pure products. A conventional control structure is developed that provides effective disturbance rejection for both production rate and feed composition changes. Dual temperature control is required in the extractive column in order to handle feed composition disturbances. 

Thus there are two controlled variables, two sensors, and one manipulated variable, while the conventional control structure has one controlled variable, one sensor, and one manipulated variable. 

The control structure with feed tray manipulation ( coordinated control ) gives fast dynamics in the product composition, as can be seen in Figure 18.14 (solid lines) where the top and bottoms compositions return to setpoint in less than lOh. In contrast, the conventional control strucmre shows a little slower dynamic response with the product compositions taking more than 10 h to return to their setpoints. Of more importance, the coordinated control stmcmre results in a 21% energy savings compared to the conventional control structure, which can be seen from the smaller vapor rate in Figure 18.14. 

---