Ordinary distillation

2.2.1 Thermodynamic fundamental

Fundamental knowledge of phase equilibrium is very important for understanding various separation pro-cesses. In particular, vapor– liquid equilibrium (VLE) is commonly encountered in distillation, and its calculable form plays a major role in establishing equilibrium stage (EQ) and nonequilibrium stage (NEQ) mathematical models of distillation columns [1]. Equilibrium is defined as a state that will be returned to its initial state after any short, small mechanical disturbance of external conditions. When equilibrium is reached for a p-phase,i-component system, the following criteria should be satisfied:

μ⁽¹⁾_i =μ⁽²⁾_i =μ⁽³⁾_i = . . . =μ^(p)_i (2.1)

T⁽¹⁾=T⁽²⁾=T⁽³⁾= . . . =T^(p) (2.2)

P⁽¹⁾=P⁽²⁾=P⁽³⁾= . . . =P^(p) (2.3)

which correspond to chemical potentialμi, thermal (i.e. temperatureT) and mechanical (i.e. total pressure P) equilibrium, respectively. For a vapor– liquid two phases system, Eq. (2.1) can be rewritten in the form of partial fugacity in vapor (V) and liquid (L) phases as

f^V_i =f^L_i (2.4)

The partial fugacity can be expressed in different forms by introducing fugacity coefficient φi in the mixture:

f^V_i =Py_iφ^V_i,f^L_i =Px_iφ^L_i (2.5) or activity coefficientγ_i:

f^V_i =y_iγiVf_iV⁰, f^L_i =x_iγiLf_i⁰_L (2.6) where the bar stands for a mixture fugacity, and the superscript “0” for pure component.

The following equations can be derived:

y_iφ^Vi =x_iφ^Li (2.7)

Py_iφ^V_i =x_iγ_iLf_iL⁰ (2.8)

Equilibrium ratio (or phase equilibrium constant) K_i is defined as the ratio of mole fractions of a component in the vapor and liquid phases at equilibrium, i.e.

K_i = y_i x_i = φ^L_i

φ^Vi

=γiLf_iL⁰

φ^Vi P (2.9)

where

φ^V_i =φ

T,P,y₁,y₂,. . . ,y_n−1

(2.10) φ^Li =φ

T,P,x₁,x₂,. . . ,x_n−1

(2.11)

γiL=γ

T,P,x₁,x₂,. . . ,x_n−1

(2.12)

f_i⁰_L=f (T,P) (2.13)

The ease of separation of a given mixture with the key componentsi andj is dependent on the relative volatility

α_ij =y_i/xi

y_j/xj

=K_i

K_j (2.14)

While at low pressure the relative volatility can be expressed in a more simplified form:

α_ij = K_i

K_j = γ_iP_i⁰

γjP_j⁰ (2.15)

wherexandyrepresent molar fractions in the liquid and vapor phases, respectively, andP_i⁰is the saturated vapor pressure of the pure componenti.

Where the relative volatility of the components to be separated is close to unity, a solvent is introduced to change the relative volatility as far away from unity as possible. Since the ratio of P_i⁰/P_j⁰ remains almost constant for a small temperature perturbation, the only way to modify the relative volatility is to change the ratioγ_i/γ_j. This ratio, in the presence of the solvent, is called selectivityS_ij:

S_ij = γi

γj

s

(2.16) Besides altering the relative volatility, the solvent should also be easily separated from the distillation products; that is, there should be a high boiling point difference between the solvent and the components or immiscibility with the components to be separated. Other criteria, such as corrosion, price, and source, should also be taken into consideration [2].

2.2.2 Distillation equipment

In most cases, either tray or packing (structured or random) columns are adopted in distillation. In general, vacuum applications are dominated by packing columns, while at the middle or even high pressure, it is better to choose tray columns. At the normal pressure, tray, structured or random packings can find suitable applications under different operating conditions. In order to decide which one is more suitable for biorefineries, some uniquely suitable special distillation equipment is introduced.

Where there is a high liquid-to-vapor flowrate ratio along the distillation column, the plate trays, espe-cially double overflow trays, are generally used as internal fittings. Both double overflow valve trays and double overflow slant-hole trays have been adopted. However, it is reported that, if double overflow valve trays are replaced by double overflow slant-hole trays in a column, the amount of feed to be treated can be increased to over 50% with a tray efficiency similar to or higher than that of the valve trays and an energy saving of 10% by decreasing the drop in pressure to about one-third of the original [3].

The slant-hole tray, as an excellent and extensively applied tray, has an opposite stagger arrangement of the slant-holes, which causes rational flowing of vapor and liquid phases and level blowing of vapor. It permits a high vapor speed, the reduction of mutual interference between vapors, a steady liquid level and high efficiency of trays. After combining the columns with multiple downcomer (MD) trays, a new type of multi-overflow compound slant-hole tray was proposed, which adopts the downcomer similar to MD trays. The number of downcomers used is normally two. The downcomer has a simple structure, liquid flows for a longer distance, it has a higher column capacity, and high efficiency of trays. The configuration

Figure 2.1 Configuration of multi-overflow (two flow) slant-hole tray. Reprinted from [4] c 2002, with permission from Elsevier

of the double overflow slant-hole trays is shown in Figure 2.1. In terms of the vapor and liquid load in the distillation column, the tray parameters are obtained by a computer program for tray design. The designed values should be within the range of normal operation conditions [4].

In the case of unclean feeding materials which are easy to polymerize or contain solid particles, the big-hole, flow-guided sieve trays are desirable because the holes with large diameters of 10–15 mm prevent the trays from blocking due to the formation of polymer or due to solid particles entering into the distillation column [5, 6]. Moreover, this type of tray is different from the ordinary sieve tray in that it has two modifications: one is that it opens a number of proper flow-guided sieves, which ensures that the gas and liquid flowing path is reasonable, and the other is the installation of bubble-promoting devices near the entrance to the liquid, which ensures that there is very small height difference in the liquid layer from the inlet to outlet on the tray, as shown in Figure 2.2.

This type of tray eliminates the gradient of liquid layer and causes the rational flowing of vapor and liquid phases on the tray, with the help of directed holes arranged for decreasing the radial mixing, and bubble promoter installed in the outlet of the downcomer.

Figure 2.3 (a) and (b) illustrate the gradient of liquid layer on the traditional tray and the improved gradient of liquid layer on the flow-guided sieve tray, respectively. Figure 2.4 shows the flowing direction of liquid phase on the flow-guided sieve tray, indicating that there is no liquid reflux on the tray.

In the case of clean feeding materials with high surface tension (e.g. dilute aqueous solutions and glycerol) or having a very strict separation requirements up to ppm level, new BeiHua (BH) structured packing has been proposed. In general, the corrugation angle in the conventional structured packing is designed to be 30^◦ or 45^◦ from the vertical position [7–9]. However, the BH packing has two kinds of transition, or wave-like, structures, which consist of central, upper and lower segments [10]. Each segment accounts for one-third of the total sheet height. The sheet corrugations vary in the order of 45^◦–30^◦–45^◦or 30^◦–45^◦–30^◦ as shown in Figure 2.5, and the different segments connect smoothly. The purpose of this

bubble-promoting devices

Figure 2.2 Structure of bubble-promoting device installed in the high-efficiency flow-guided sieve tray.

Reprinted from [1] c2005, with permission from Elsevier

(a) (b)

Figure 2.3 Gradient of liquid layers on the traditional tray (a) and on the flow-guided sieve tray (b). Reprinted from [1] c2005, with permission from Elsevier

geometry design is to spread the liquid film on the sheet surface as completely as possible and enhance the local gas–liquid contact area.

It was found that the corrugation angles with two transition structures (i.e. 30^◦–45^◦–30^◦ and 45^◦–30^◦–45^◦) is favorable when considering pressure drop and mass transfer coefficients together. As expected, a low ratio of packing height to diameter is favorable for increasing mass transfer coefficients, but leads to increasing pressure drop like common structured packings. It is necessary for us to identify the relationship between geometric configuration and performance of pressure drop and mass transfer coefficient so as to tailor the desirable packings.

2.2.3 Application in biorefineries

When conventional tray or packing columns cannot meet separation requirements in biorefineries, a process intensification approach to the existing distillation internals is needed. As the addition of extra streams and equipment to the original processes is not required, it is convenient to implement it in biorefinery plants.

The special distillation internals, as mentioned earlier, are not complicated in geometric structure and can

Figure 2.4 Flowing direction of liquid phase on the flow-guided sieve tray. Reprinted from [1] c2005, with permission from Elsevier

(a)

(b) a₁

a₁ a₂

Figure 2.5 Transition structure of BH packing (α₁= 45^◦ andα₂= 30^◦; orα₁= 30^◦andα₂= 45^◦). (a) Side view; (b) actual photography. Reprinted from [10] c2009, with permission from Elsevier

be manufactured easily and cheaply. The investment in technology is therefore small and may be used for solving the following separation problems encountered in biorefineries:

• Separation of aqueous organic solutions with low concentration, such as in biomass-to-ethanol biore-fineries, where the ethanol stream coming from the fermentor [11] is at a low concentration of about 5–10 wt%. In this case the ratio of liquid to vapor flowrates along the distillation column will be

high and thus the multi-overflow (double flow) slant-hole tray is suitable for the separation to obtain approximately 92.5 wt% ethanol.

• Separation of unclean feeding materials containing solid particles from pulp mills, forest products and wood wastes. In this case the big-hole flow-guided sieve tray is more suitable than other trays and packings because the latter will bring about a column jam.

• Separation of clean feeding materials with high surface tension or very strict separation requirements.

The high surface-tension materials include aqueous solutions (not containing undissolved fiber, germ and gluten), purification of biodiesel (fatty acid methyl ester) and byproduct glycerol [12, 13]. Mate-rials with very strict separation requirements include various wastewaters from biodiesel and other biorefinery production plants. Before discharging into the environment, the concentration of volatile organic compounds (VOCs) should be very low, usually at ppm level. In this case, a very large number of theoretical stages (up to a few hundred) are needed in order to achieve an environmentally friendly separation process. Addressing environmental requirements should therefore be an ongoing challenge in separation and purification in biorefineries.

In document Separation and Purification Technologies in Biorefineries (Pldal 61-66)