Operating Modes of the Integrated Controller

10.3.3.1 Control Mode 1

The ETBE purity and the isobutene conversion must both be controlled to specified targets in this mode. Ideally, this would be best achieved using direct measurements of the two control objectives in order to avoid the uncertainty that is associated with an inferential indication. However, this approach is impractical with linear controllers.

The limitations of using a direct measurement of the ETBE purity (or the conversion) have been discussed previously for both one-point and two-point controllers and mostly derive from the input multiplicity which is present between the controlled and manipulated variables. This is clearly shown in Figure 10.15 where the process gain between the bottoms composition and the reboiler duty can be both positive and negative at a constant reflux rate. This property is also observed if the reflux rate is used as the manipulated variable. The pairing with reflux rate also suffers from the increased physical distance between the point at which the purity is measured and the location of the final control element that leads to a less dynamically responsive system.

A technique for avoiding the input multiplicity was proposed in Chapter 9. Essentially, a temperature from near the middle of the stripping section (e.g. stage 7) is controlled instead of the reboiler temperature or composition. This results in the one-to-one relationship shown in Figure 10.16, which has much better characteristics for controllability. The selected temperature could either be used directly or within a cascade loop where the master controller measured the ETBE purity and updated the set-point of the temperature controller. This arrangement is more effective than using the analyser output directly but could still not be globally stable due to the input multiplicity that persists.

Boiler Distillation Still
Figure 10.15 - Effect of Reboiler Duty on the ETBE Purity
Figure 10.16 - Effect of Reboilcr Duty on Mid-Stripping Scction Temperature

The control of the isobutene conversion presents a more difficult measurement problem. Direct measurements would require a substantial amount of data and equipment. The use of several easily measured temperatures to predict the conversion was demonstrated in Section 10.1. This is valid and could be used to produce good control. However, a simpler model that uses only the maximum and minimum reactive section temperatures was used here. Figure 10.17 shows the pattern of the relationship between the reaction zone AT and the isobutene conversion for various operating conditions, with the ETBE purity control loop opened (i.e. fixed reboiler duty) and closed (i.e. fixed stripping section temperature).

The open-loop relationship includes a peak at around 4.0°C and is, therefore, bidirectional. However, temperature differentials below this value correspond to low values of the ETBE purity in the bottoms product and, therefore, are not likely operating points. Above the critical temperature differential, the relationship is smooth and mono tonic. The closed-loop relationship is smooth and monotonic (and nearly linear) for all operating points of practical interest. Thus, the inferential indication of the reactant conversion provided by the reaction zone temperature differential is suitable for closed-loop control. The reaction zone AT varies nearly linearly with the reflux rate for a constant reboiler duty so that stable and robust control of the conversion should be possible with this pairing.

□ Reboiler Duty = 8.66 kW A Reboiler Duty = 8.33 KW O Reboiter Duty = 8.00 KW ♦ Md-Stripping Sect. Tenp. = 115 C

□ Reboiler Duty = 8.66 kW A Reboiler Duty = 8.33 KW O Reboiter Duty = 8.00 KW ♦ Md-Stripping Sect. Tenp. = 115 C

Relative Volatility With Reboiler Duty

Reaction Zone Temperature Differential (C)

Figure 10.17 - I ink Between the Reaction Zone AT and Conversion

Reaction Zone Temperature Differential (C)

Figure 10.17 - I ink Between the Reaction Zone AT and Conversion

The two control loops (TC: reboiler duty-stripping section temperature; and, ATC reflux rate-reaction AT) interact strongly as indicated by the relative gain which has a value of 20-60 in around the desired operating point. This is high, but comparable to a high-purity, non-reactive distillation process. Although this interaction reduces the process controllability, satisfactory closed-loop performance was achieved (see Section 10.3.4).

This control mode should be used when: (1) there are specific limits on the allowable concentration of C4 and/or ethanol in the bottoms product; (2) there is a significant cost penalty associated with giveaway (i.e. there is an economic incentive to incorporate as much C4 or ethanol in the bottoms product without violating the specification limit); and (3) the benefit of producing additional ETBE from isobutene must be optimised with the cost of the additional energy consumption (via both the heating and cooling utilities).

10.3.3.2 Control Mode 2

In this mode, the ETBE purity in the bottoms product is controlled but the isobutene conversion is maximised to equipment constraints. The first control loop, TC, is unchanged from the first control mode. The reflux rate is the manipulated variable for the second loop but a new controlled variable must be found. Many different equipment constraints could be active but two common constraints (in a well designed column) are flooding and the reboiler duty. Composite variables could be built to approximate both of these values (neither is easy to measure directly) in order to control the column operation close to the appropriate limit. However, the reflux rate is nearly proportional to both these limits, regardless of the feed rate, and effective constraint management is possible by simply controlling the reflux rate directly.

This control mode is appropriate when: (1) there are limits on the bottoms product composition; (2) giveaway must be minimised concentration; and (3) the cost of additional energy consumption is always less than the value of producing additional ETBE.

10.3.3.3 Control Mode 3

The third control mode manipulates the reboiler duty and reflux rate in order to maximise the ETBE punty and to maximise the isobutene conversion. As demonstrated above, (manually) adjusting the reflux rate to close to the equipment constraints effectively maximises conversion. There is no comparable method of maximising ETBE purity because the process gain (between purity and reboiler duty) changes sign (see Figure 10.15). An optimum reboiler duty must be found.

A steady-state process model was used to find the optimal value of the mid-stripping section temperature for various feed rates and feed compositions. Surprisingly, the feed composition has a negligible effect on the optimisation results. This is fortunate since the composition is difficult to measure and less suitable for closed-loop control than flows and temperatures, etc. The simple model given by equation (10.10) was found to describe the relationship between the optimum temperature and the feed and reflux rates. This model was incorporated into a feed-forward controller (SPC) to update the temperature set-point on-line.

Optimum TC set-point = 121.2 + 9.2 (Reflux/Feed - 0.329) (10.10)

Although the performance of this control loop could be enhanced by dynamic compensation so that the set-point was updated at an appropriate time after a change in the feed or reflux rates, the optimal characteristics of a lead-lag unit vary strongly with the operating conditions. Therefore, only the steady-state model was implemented on the simulations.

This control mode is recommended when: (1) there is a cost penalty associated with any reduction in the purity of the ETBE product; and (2) the value of additional conversion is always higher than the required increase in energy consumption. Intuitively, this control mode would appear attractive but is probably less likely to be used in practice as specific targets which correspond to a global, plant-wide optimisation should yield a higher overall profitability.

The three different control modes can be implemented using only three control loops, as described above. The column operation can be changed from one mode to another by simply opening or closing particular loops, according to Table 10.5. A tick mark indicates that the control loop should be operated in automatic while a cross mark indicates that the loop should be operated in manual in order to realise the correct process objectives.

Table 10.5 - Operation of the Individual Control Loops for Each Control Mode

Control Mode 1

Control Mode 2

Control Mode 3

ETBE purity is...

controlled

controlled

maximised

Isobutene conversion is...

controlled

maximised

maximised

Control Loop

TC - temperature control of the

S

S

V

mid-stnppmg section

ATC - control of the reaction

S

X

X

section temperature difference

SPC - Set-point optimiser for TC

X

X

Was this article helpful?

0 0

Post a comment