4. Feed Line

Piping and Instrumentation Diagram (P&ID): Feed Line
Click image to enlarge and see location of component.

Feed line for a distillation column includes feed storage, flow control, feed conditioning unit (typically a heat exchanger), and distribution manifold. All these components are essential for distillation operations because the feed flow rate, feed condition, and feed location are all sensible operation variables influencing distillation column operation, performance, and design (Green, 2019; Geankoplis, 2003; Coker, 2010). (Click on the Piping and Instrumentation Diagram (P & ID) on the right in order to see the location of this component within the distillation column.)

Figure 4.1 shows a process diagram for the feed line commonly used for industrial distillation columns. The liquid feed is pumped through a control valve, typically a pneumatic valve that is controlled by a flow controller (FC) with desired operation flow rate as setpoint and actual flow rate measured by a flow meter (FM) as input signal. The feed flow at the controlled operation flow rate flows through a heat exchanger to achieve desired feed condition, and is then introduced into an individual tray through a manifold and valves.

Figure 4.1. Schematic diagram of a typical feed line process unit for distillation column with (a) pump, (b) pneumatic valve, (c) flow meter, (d) heat exchanger, and (e) manifold and valves. | Description

Site Author(s) and University of Waterloo

Heat Exchanger for Feed Conditioning
Click image to enlarge

UWaterloo Chemical Engineering Virtual Learning. (2022, Feb 7). 4 Feed Stream [Video]. YouTube. https://www.youtube.com/watch?v=DyTtK7mhvek | Transcript

The feed condition or feed quality is generally characterized by the q value (Equations 4 and 5 on the Distillation Column component page). By definition, $q$ is the fraction of feed added to liquid stream in the column, and is generally evaluated by the enthalpy of the feed mixture at different conditions,

$$ q=\dfrac{H_V-H_F}{H_V-H_L} $$

where $ H_V $ is the enthalpy of the feed mixture as saturated vapour, $ H_L $ is the enthalpy as saturated liquid, and $ H_F $ is the enthalpy at feed condition. This definition allows one to conveniently change the feed condition by adjusting the feed temperature. For instance, if saturated liquid is needed for a distillation operation, then the feed temperature should be set at the bubble point of the feed mixture at the operation pressure. The bubble point can be easily calculated based on the VLE calculation on the Reboiler component page (Figure 1.3). This bubble point or boiling temperature can then be set to the heat exchanger for feed conditioning to ensure the feed is saturated liquid and $ q=1 $. Figure 4.2 shows various feed conditions and corresponding q values for a binary ethanol-water system at a feed ethanol mole fraction of 0.3. Evidently, the change in feed condition from liquid to vapour leads to more stages in the rectifying section, which is favourable for high purity of the more volatile component in the top product.

Figure 4.2. Effects of various feed conditions on the binary distillation of ethanol-water system. | Description

Site Author(s) and University of Waterloo

The feed condition also affects distillation column performance through the changed vapour and liquid flow rate in the column. As illustrated in Figure 4.3, the molar flow of a feed stream is split into two portions, one adding to vapour flow in the rectifying section and the other to liquid flow in the stripping section. The exact amount to each stream depends on feed condition or $q$ value. The changed flow rates in the column result in changed slopes of the operating lines and the column performance, as illustrated in Figure 2.5. As an example, a feed as saturated vapour adds all the feed to vapour stream, which decreases the slope of the rectifying operating line and leads to improved column efficiency and separation in the rectifying section. However, the drawback of the vapour feed condition is that it increases the condenser duty, hence operation cost, not to mention the increased control instability due to more fluctuating vapour flow.

Figure 4.3. Schematic representation of the effects of feed conditions on the liquid and vapour flow rates in rectifying and stripping sections of distillation column. All the symbols are the same as those in Equations 1~5 on the Distillation Column component page. | Description

Site Author(s) and University of Waterloo

Similar to feed condition, feed location can also pose a marked effect on distillation column performance and overall separation. As shown on the Distillation Column component page, for a distillation column with fixed number of trays, there exists an optimum feed tray location for best column performance. This can also be seen from Figure 4.2 that a feed tray closer to the bottom of column can result in more trays in rectifying section and increased purity of the more volatile component in the top product. The minor tradeoff of this design is that more subcooled liquid increases the reboiler duty and operation cost. In general, the optimum feed location can be found through rigorous process simulation and optimization. Alternatively, the empirical Kirkbride method (Geankoplis, 2003) provides a quick estimation of the optimum feed location,

$$ \dfrac{N_R}{N_S} = \left[ \dfrac{B}{D} \dfrac{x_{F,HK}}{x_{F,LK}} \left(\dfrac{1-x_{B,LK}}{1-x_{D,HK}}\right)^2 \right]^{0.206} $$

where $ N_R $ is the number of stages in rectifying section, $ N_s $ is the number of stages in stripping section. $ x_{F,HK} $ is the mole fraction of heavy key component or less volatile component in the feed, $ x_{F,LK} $ is the mole fraction of light key component or more volatile component in the feed, $ x_{B,LK} $ is the mole fraction of light key component or more volatile component in the bottom product, and $ x_{D,HK} $ is the mole fraction of heavy key component or less volatile component in the top product. This empirical equation relates the optimum feed location with distillation specification, and can be used with other short-cut methods to determine the optimum feed location for both binary and multicomponent systems.

References