Distillation control and optimization, Part 3

Maximum Column Loading
If both demand for product and availability of feedstock are unlimited, maximizing throughput maximizes profitability. In such installations, a valve position controller can increase the feed rate until the loading limit on the constraining equipment is reached. If the cooling capacity of the condenser sets the production rate limit, the feed rate can be manipulated by a valve position controller (VPC) on the back-pressure control valve (C in Figure 3, p.52) to keep it always nearly open (VPC set at 90%). The opening of this bypass valve also indicates condenser loading.

If the production is constrained by the reboiler, a similar cascade control configuration can be used by keeping the heating fluid valve nearly fully open (D in Figure 3). This strategy is particularly effective when a waste steam, which otherwise would be vented, is used as the heating media. The VPCs usually are integral-only controllers, and are set at around 90% of stem lift, which on an equal-percentage valve corresponds to about 70% of maximum flow. The integral time setting of the VPCs is slow, about 10 times the integral setting of the FRC. To eliminate reset windup, the VPCs always are provided with external feedback from the slave transmitter (FT).

If there are several constraints, a multiple constraint network is implemented. In that case, the most critical of several VPCs can be selected to set the feed flow.

Feed Temperature and Enthalpy Control
The thermal condition of the feed determines the amount of heat the reboiler has to provide, but constant feed temperature alone doesn’t keep that demand constant. For best separation efficiency, the feed should be preheated to its bubble point (when the lightest component in the feed starts to boil). If the composition of the feed varies, its bubble point also will vary. As the feed becomes lighter, some of it will vaporize, but this variation can be handled by subsequent controls.

The feed can be held at its bubble point by a temperature controller configured as the cascade master of the feed-flow controller (E in Figure 3). It is usually a three-mode controller, in which the main role of the derivative mode (D) is during start-up. Here, it initially provides a large correction, which helps get the unit up to stable operation quickly. The heat content of the bottom product also can be used to preheat the feed in an economizer (F in Figure 3). The goal usually is to maximize heat recovery, which is achieved when the economizer bypass flow is minimal. Therefore the VPC is usually set at about 10% of valve opening.

When preheating the feed is less expensive than adding the heat at the reboiler, or when the reboiler is the limiting constraint, column operation can be optimized by maximizing feed preheat. When the main constraint is the condenser, or if flooding occurs above the feed tray, preheaters shouldn’t be used. The primary effect of increasing feed enthalpy is to decrease vapor-liquid circulation below the feed tray relative to that above the feed tray.

If both an economizer and a preheater are used in a feed enthalpy control system, additional measurements will be needed. For example, if the cold feed    first passes through a bottoms-to-feed economizer and then through the steam preheater, one can calculate the steam required to keep the feed enthalpy constant:
 

This type of control can be provided by feed-forward controls and is usually implemented in software.

Reflux Controls
The key to stable column operation is to keep the internal reflux constant. Internal reflux controls compensate for changes in the temperature of the external reflux, which is usually caused by ambient conditions. A typical internal reflux control system and the equations that need to be solved in calculating the required external reflux rate are shown in A of Figure 4 below. This control system corrects for either an increase in overhead vapor temperature or a decrease in external reflux liquid temperature.

When the external reflux flow is cascade-controlled by the accumulator level, internal reflux control can be eliminated. The speed of response of this control system can be increased by reducing of the accumulator volume. To completely overcome the accumulator lag, the reflux rate (L) can be manipulated in direct proportion to the change in distillate rate (D), rather than by waiting for the response of a level controller. In some control systems (B in Figure 4 blow), the relative influence of the level controller is adjustable when K = 0, the reflux is adjusted by the level controller only. At other settings of K, the reflux flow is immediately altered to some extent, when a change in distillate flow occurs, and the level controller forces the balance of the change.

FIGURE 4: REFLUX FLOW AND ACCUMULATOR LAG

These are the controls required to keep the internal reflux flow of the column constant (top) and those needed to eliminate the effect of accumulator lag in controlling the external reflux to a column (bottom).

If K = 0.5, the reflux flow is changed to a new steady-state value,    and therefore the lead equals the lag, so the net effect is the elimination of dynamic contribution. If K = 1.0, the initial response is a first-order, lead-lag function. In this case, the reflux is greater than the new steady state, and the level controller eventually corrects that flow. The value of K affects the transient response only, but doesn’t change the steady-state flow. The greater the value of K, the faster the response. To maintain stability, K should not be set greater than 0.75.

This article is an excerpt from the ebook, “Distillation Control and Optimization” by