Design processes

Semi-batch processes are preferred to batch processes where possible. This point was made sbove (see Section 11.4). Continuous processes allow the inventory of reaction mass to be reduced, and represent an elegant way of reducing the severity of a process. 

Before designing a process scheme it is necessary to know the specification of the raw material input (or feedstock) and the specification of the enc/procfucf desired. Designing a process to convert fluids produced at a wellhead into oil and gas products fit for evacuation and storage is no different. The characteristics of the well stream or streams must be known and specifications for the products agreed. 

The quality and quantity of fluids produced at the wellhead is determined by hydrocarbon composition, reservoir character and the field development scheme. Whilst the first two are dictated by nature the latter can be manipulated within technological and market constraints. 

The main hydrocarbon properties which will influence process design are  

PVT characteristics - which describe whether a production stream will be in gas or liquid form at a particular temperature and pressure. 

Composition - which describes the proportion of hydrocarbon components (C, - Cj+) (which determine the fluid properties), and how many non-hydrocarbon substances (e.g. nitrogen, carbon dioxide and hydrogen sulphide) are present. 

The problem of process design in CRE typically stems from the requirement to produce a specified product at a particular rate (e.g., 1000 tonnes day -1 of NH,). The substance ) from which the product is made may be specified or may have to be chosen from more than one option. Process design then involves making decisions, as quantitatively as possible, about the type of reactor and its mode of operation (e.g., batch or continuous), its size (e.g., volume or amount of catalyst), and processing conditions (e.g., T, P, product distribution, if relevant). The criteria constraining these decisions 

The design problem usually fits into the spectrum ranging from (1) the rational design of a completely new reactor for a new process, to (2) the analysis of performance of an existing reactor for an existing process. A common situation, between these extremes, even for a new plant, is the modification of an existing type of reactor, the design of which has evolved over time. 

Process design involves making decisions about a series of matters, on as rational and quantitative a basis as possible, given the information available. The following is an illustrative list of such matters but not an exhaustive one the items are not all mutually exclusive. 

Most of these matters are subjects for exploration in the following chapters, but some are outside our scope. Some may be specified ab initio, and others may provide constraints in various ways. 

Data required include those specific to the situation in hand (design specifications) and those of a more general nature  

A process for the hydroformylation of 1-octene to nonanal was designed for an immobilised homogeneous catalyst. The production capacity was fixed at 100 kton of nonanal. Kinetic data reported for the rhodium catalyst complex of N-(3-trimethoxysilane-n-propyl)-4,5-bis(diphenylphosphino)-phenoxazine immobilised on silica, (2) was used as a starting point. Other process specifications are given in Table 3.8. 

TABLE 3.8. Process and reactor specifications for hydroformylation of 1-octene 

Reflux ratio R 0.5 2 Type Gas Liquid Solid Reactor  

For the supported catalyst it is expected that the ligand does not leach since it is chemically bonded to the carrier. In contrast, the rhodium metal bound to the ligand is subject to leaching due to the reversible nature of the complex formation. The amount will depend on the equilibrium between rhodium dissolved in the organic phase and that bound to the ligand. When an equilibrium concentration of 10 ppb Rh is attained, the yearly loss of Rh for a 100 kton production plant will be about 1 kg Rh per year. Compared to the reactor contents of rhodium (see Table 3.9, 70 kg Rh) this would result in a loss of 1.5% of the inventory per year, which would be acceptable. 

Application Typical membrane material Selectivity (a) Average pressure-normalized flux [10-6 cm3(STP)/ cm2 s cmHg] Module design commonly used 

The importance of pressure ratio in the separation of gas mixtures can be illustrated by considering the separation of a gas mixture with component concentrations of iiio and iijo at a feed pressure p0. A flow of component i across the membrane can only occur if the partial pressure of i on the feed side of the membrane (riiop0) is greater than the partial pressure of i on the permeate side of the membrane (niepe), that is, 

It follows that the maximum separation achieved by the membrane can be 

That is, the separation achieved can never exceed the pressure ratio p, no matter how selective the membrane  

The relationship between pressure ratio and membrane selectivity can be derived from the Fick s law expression for the fluxes of components i and j 

N THE FIRST CHAPTER we stated that chemical engineers create processes based on physical and chemical changes. In this chapter we develop designs for five chemical processes. The step-by-step examples will also introduce strategies for design and conventions for depicting a chemical process. 

Assuming that the annual production requirement has been established, one of the first problems faced in process design is to choose a process cycle so that material and energy balances on a time basis can be worked out and all of the necessary flow sheets prepared. Questions related to a 24-hr or 8 hr/day operation and production by sequential batch operation or on a steady-state continuous basis must be resolved. 

Continuous vs. Batch Processing. It is the usual rule in process design to choose continuous processing in preference to batch processing based purely on economic reasons. By operation on a continuous 24 hr/ day basis, smaller-size, less expensive processing equipment is used. Process operation is steady state and easier to control by automatic instrumentation than is batch operation. Thus, capital investment, fixed charges, and labor requirements are minimized. However, batch processing is not a completely obsolete method in the chemical industry. It is feasible in such cases as  

Small-volume output of relatively expensive products when sales demand is not steady, or when the same equipment can be used for several processes of this nature. 

When batch equipment from an abandoned process is available at a low transfer cost to the current project. 

When continuous process equipment has not been satisfactorily developed and batch-process equipment has been satisfactorily demonstrated this is a short-range solution when emergency deadlines must be met. 

In the following chapters, the background to all of these steps is given in detail. The reader should at all times consider each area in perspective with the integrated development.. Never forget the objective to. solve a process need with efficiency and economy in a reasonable time. 

Catalytic materials fall into welLdefined categories. Although we use broader classifications than those given in Chapter 1. the motivation is the same—to group according to common types of activity and to explain the catalytic behavior on the basis of common properties. 

Other rorms Molten salts ZnCl NajCO, 

Over 70% of known catalytic reactions involve some form of metallic component. Industrially, metals are used in catalytic reforming, hydrocracking, ammonia and methanol synthesis, indirect coal liquefaction, oxidation, and a vast number of organic hydrogenation and dehydrogenation processes. Academically, metals are favored for research since they are easily prepared in pure form and conveniently characterized. In fact, most of the fundamental information leading to conceptual theories in catalysis originated with studies on metal systems. 

The periodic table for transition metals is shown in Fig. 4.1. The periodic table is useful since catalytic behavior, like other chemical properties. 

Now that the various methodologies to overcome the limitations of biocatalysis have been discussed, a brief account of process design will give the necessary information to make these processes 

How much material do you need to supply to customer—100 g, 1 kg, 10 kg How much product can you obtain per unit of reaction volume (substrate/product solubility)  

How long does one batch take to run in process per unit of reaction volume How long do you have to make your delivery How large is the reaction vessel How long can the vessel be practically and safely operated  

The answers to these questions determine the nature of the process and the configuration of the final design. The basic strategy should be as follows, for both purified enzymes and whole cells  

AbSOrbBr. A column with 11 stages operates at 1 atm pressure at the top. A tray pressure drop of 0.2 psi is specified in order to satisfy the requirement that the specified pressure drop is greater than the pressme drop calculate from the hydraulics when exporting to Aspen Dynamics. The design feed gas is 13,100 kmol/h and is compressed to 1.136 atm and fed at the bottom of the absorber. The specified recovery of carbon dioxide is 90%, which corresponds to an absorber exit gas composition of 1.3 mol% CO2. 

The required flow rate of the lean solvent to the top of the absorber to achieve this recovery is 32,860 kmol/h. This flow rate is after the makeup water and makeup 

MEA have been added. The bottoms from the stripper has a concentration of 15.2 mol% MEA. The concentration of carbon dioxide in the stripper bottoms is not negligible (2.62 mol% CO2). 

Both fresh water and fresh MEA are necessary because of losses in the off-gas from the absorber and the vapor product stream from the reflux drum of the stripper. Small amounts of MEA are lost in the absorber off-gas (13.9 kmol/h) and in the vapor from the stripper reflux drum (3.17kmol/h). The water losses are quite signiflcant in the off-gas (2347kmol/h) and in the vapor from the stripper reflux dmm (1192 kmol/h). The water in the feed gas is 824.3 kmol/h. 

Stripper. Absorber bottoms at 322 K is preheated to 380 K in a heat exchanger using the hot stripper bottoms at 400 K and fed to the top of a stripping column with 10 stages and operating at 2 atm in the column and 1.5 atm in the reflux drum. Reboiler heat input is 54.12 MW to maintain a reflux-drum temperature of 363 K, as suggested by Desideri and Paolucci as a balance between stripper reboiler energy and water losses in the vapor from the reflux dmm. 

Industrial separation processes typically consist of various distillative and alternative separation steps that are coupled by material and eneigy streams. Such processes often have very complex stractures caused by the properties of the systems at hand and by the constraints set by cost and energy savings. In most cases, a rather empirical approach is used for process design. Novel developments concern a conceptual process design (e.g., Douglas 1988 Smith 1995 Blass 1997 Stichl-mair and Fair 1998 Seider et al. 1999 Doherty and Malone 2001 Mersmaim et al. 2005), which is based on the thermodynamic properties of the mixture at hand. 

Separation processes typically consist of hrmdreds of elements (pipes, vessels, heat exchangers, columns, pumps, compressors, engines, measuring and control devices, etc.) that are too complex for exact graphical depiction. Therefore, the process structures are depicted in a very abstract and geometrically nonsimilar manner in the form of flow sheets. Flow sheets show the process structure with symbols for the essential process elements according to international standards (e.g., DIN, ISO). 

The flow sheet of a process for the production of absolute alcohol by pressure swing distillation is shown in Fig. 11.0-1. The process consists, in essence, of two distillation columns, several heat exchangers, several vessels, and a complex network of pipelines (see Sect. 11.3.2). Also shown are measuring and control devices. According to international standards (e.g., DIN 19227, ISO 3511) these devices are illustrated by thin circles with a letter code indicating the function of the device, for instance, first letters F = flow, L = liquid level, P = pressure, T = temperature subsequent letters A = alarm, C = control, F = fraction, etc. 

Mersmann et al., Thermal Separation Technology Principles, Methods, Process Design, VDI-Buoh, DOI 10.1007/978-3-642-12525-6 ll, 

Pipe and instrument flow sheet of a process for the production of absolute alco- 

To choose the adequate bioreactor design for continuous PHA production, kinetics for both biomass and PHA production by the microbial strain should be considered. In the case of PHA production directly associated with microbial growth as it is found in Alcaligenes latus DSM 1122 on sucrose [128], or for Pseudomonasputida ATCC 29147 on fatty acids [97,98], a one-step continuous process using a continuous stirred tank reactor (CSTR) is a viable solution. 

The situation changes significantly in the case of strains like C. necator, where autocatalytic growth of biomass is followed by a phase of linear PHA production. In this case, biomass production should occur in the first step in a CSTR which is coupled to a subsequent plug flow reactor (PFR). The combination CSTR-PFR not only ensures higher productivity, but also minimizes the loss of substrates and co-substrates. Furthermore, product quality can be enhanced by the fact that the PFR features a narrow residence time distribution, leading to higher uniformity of cell populations. This should also have positive impacts on the distribution of the PHA molecular masses and the composition of polyesters [128]. 

Continuous processes are operated as so called chemostat , hence, the concentrations of as well substrates as products stay constant during the entire process duration. Such conditions of constant concentrations at constant process parameters (pH-value, temperature, dissolved oxygen concentration etc.) are known as steady state conditions in literature [128]. Besides the signiflcance for industrial process development, continuous studies in chemostats are also a precious tool for elucidating the relationships between cells and their environment. For example, the optimization of the composition of nutritional media can be accomplished this way. In addition, continuous processes enable the fine-tuned supply of growth inhibiting substrates as it is often the case in md-PHA production. 

The first studies on one-stage chemostat continuous PHA production were published by Ramsay and colleagues [130]. C. necator was cultivated on glucose at a dilution rate of 0.15 h and produced 5 g/L of biomass with a PHB content of 33%. Similar results were obtained with A. latus using sucrose at the same dilution rate biomass concentration and PHB content were 4 g/L and 40%, respectively. When grown on glucose and various concentrations of propionic acid up to 5 g/L, A. latus produced PHBHV with a content range of 3HV monomer in the PHA of up to 20% [130]. 

Several other studies were conducted to produce PHBHV with C. necator in one-stage continuous culture. This copolyester was produced from fructose and propionic acid with a maximum PHA productivity of 0.31 g L h and a 3HV content in the range of 11 to 79%. It turned out that the molecular mass increased with the dilution rate [129]. When propionate was employed as 3HV precursor, the continuous culture systems did not reach steady states when propionate concentration exceeded 7 g/L [131]. Zinn and colleagues [133] grew C. necator in 

Clearly the shape and operating capacity as defined by the breakthrough curve is very dependent upon such factors as  

In other areas of technology such as chromatographic separations, and hydrometallurgy which may often involve resin transfer (movement) operations, fluidized beds, and truly countercurrent contact of resin and liquid, the acquisition of design data is more often through a fluid dynamic and mass transfer approach to describe, and thereby scale up, column designs in order to realize commercial practices. A text of this nature is neither the vehicle for, nor can it do justice to, such an important topic and the reader is referred to the bibliography for a more detailed account of column dynamics. 

Helfferich, Ion Exchange , McGraw-Hill, London and New York, 1962. 

Helfferich and Yng-Long Hwang, Ion Exchange Kinetics , in Ion Exchangers , ed. K. Dorfner, Walter de Gruyter, Berlin and New York, 1991, Ch. 6.2, p. 1277. 

Helfferich, Ion Exchange Kinetics , in Ion Exchange (A Series of Advances) , ed. J. A. Marinsky, Edward Arnold (London), Marcel Dekker (New York), Vol. 1, 1966, p. 65. 

In the first place the safety technological bases have to be determined. This imphes the 

Furthermore, the possibilities of reactions have to be clarified by systematic experimental analyses. The ranges of the process parameters (temperature, pressure, material concentrations etc.) have to be identified, within which no undesired reactions can take place. On this basis the permissible ranges for plant operation are fixed and the consequences of leaving them are pointed out 

Due to the higher concentrations of Ca and Mg in those solar pond brines, brine purification (easing the task to achieve product purity) cannot be considered. The high consumptions of soda ash, calcined soda, and calcium chloride (or lime milk and hydrochloric acid), and the disposal costs for the filter cakes including the additional investment costs caimot reach feasibility. In this case and for this concept. 

Whereas the impurities leave the washing thicJcener with its overflow, the so treated solution reaciies the centrifugation already prepurified before the final separation takes place. Though this treatment cannot replace a complete brine purification that provides up to 99.9% NaCl, it avoids the chemical and other side costs and allows purities safely above the required 99.7% NaCl. 

In order to reach the desired particle size of 0.4 mm, decisions by experience have to be made. The salt retention times for the three FC-type crystallizers are set to 1.5 h each (span width for FCs 0.5-2 h), the tip speeds of the impeller pumps to 14 m/s, and the tip supersaturation to maximum 1 g/1. Longer retention times do not have significant effect on the particle size, whereas lower tip speeds of the impeller pumps and lower tip supersaturations again lead to higher specific energy inputs per crystallizer. 

The question of the salt retention time for the Oslo-type crystallizer can be left open until the start-up. The need to clarify the entire recirculation completely leads to a crystallizer diameter allowing any retention time between more than 10 h and lower than 20 h (easily achievable just by decreasing or increasing the crystal bed mass inside the crystallizer by simply opening or closing the salt withdrawal for a longer time). 

Measures against cycling are not foreseen as the customer desires to get various sieve cuts of granular salt for different market approaches. 

9 Affinity Chromatography and its Applications in Large-Scale Separations 

The intended use of the product has to be taken into account during the scale up and design of large-scale processes. For example, it may be necessary to show that any components of the adsorbent are not present in the final product if this is to be for human use. The affinity adsorbent, and other adsorbents used in the process, has to be controlled strictly to ensure that the product is sufficiently pure. This may preclude use of a particular adsorbent, particularly if the ligand leaches, due to the activation chemistry used in its manufacture. Alternatively it may be necessary to demonstrate that the ligand leakage into the product is of an acceptable level. 

---

The critical temperature of methane is 191°K. At 25°C, therefore, the reduced temperature is 1.56. If the dividing line is taken at T/T = 1.8, methane should be considered condensable at temperatures below (about) 70°C and noncondensable at higher temperatures. However, in process design calculations, it is often inconvenient to switch from one method of normalization to the other. In this monograph, since we consider only equilibria at low or moderate pressures in the region 200-600°K, we elect to consider methane as a noncondensable component. 

Null, H. R., "Phase Equilibrium in Process Design," John Wiley, New York (19 70). ... 

While many methods for parameter estimation have been proposed, experience has shown some to be more effective than others. Since most phenomenological models are nonlinear in their adjustable parameters, the best estimates of these parameters can be obtained from a formalized method which properly treats the statistical behavior of the errors associated with all experimental observations. For reliable process-design calculations, we require not only estimates of the parameters but also a measure of the errors in the parameters and an indication of the accuracy of the data. 

In many process-design calculations it is not necessary to fit the data to within the experimental uncertainty. Here, economics dictates that a minimum number of adjustable parameters be fitted to scarce data with the best accuracy possible. This compromise between "goodness of fit" and number of parameters requires some method of discriminating between models. One way is to compare the uncertainties in the calculated parameters. An alternative method consists of examination of the residuals for trends and excessive errors when plotted versus other system variables (Draper and Smith, 1966). A more useful quantity for comparison is obtained from the sum of the weighted squared residuals given by Equation (1). 

The most frequent application of phase-equilibrium calculations in chemical process design and analysis is probably in treatment of equilibrium separations. In these operations, often called flash processes, a feed stream (or several feed streams) enters a separation stage where it is split into two streams of different composition that are in equilibrium with each other. 

This text will attempt to develop an understanding of the concepts required at each stage during the creation of a chemical process design. 

Figure 1.2 Process design starts with the reactor. The reactor design dictates the separation and recycle problem. (From Smith and Linnhoff, Trans. IChemE, ChERD, 66 195, 1988 reproduced by permission of the Institution of Chemical Engineers.)...

The hierarchical nature of process design has been represented in different ways by different authors. A hierarchy of decisions and a process design ladder also have been suggested. 

The synthesis of the correct structure and the optimization of parameters in the design of the reaction and separation systems are often the single most important tasks of process design. Usually there are many options, and it is impossible to fully evaluate them unless a complete design is furnished for the outer layers of the onion. For example, it is not possible to assess which is better. 

In broad terms, there are two approaches to chemical process design ... 

Since process design starts with the reactor, the first decisions are those which lead to the choice of reactor. These decisions are among the most important in the whole design. Good reactor performance is of paramount importance in determining the economic viability of the overall design and fundamentally important to the environmental impact of the process. In addition to the desired products, reactors produce unwanted byproducts. These unwanted byproducts create environmental problems. As we shall discuss later in Chap. 10, the best solution to environmental problems is not elaborate treatment methods but not to produce waste in the first place. 

The decisions made in the reactor design are often the most important in the whole flowsheet. The design of the reactor usually interacts strongly with the rest of the flowsheet. Hence a return to the decisions made for the reactor must be made when the process design has progressed further and we have fully understood the consequences of those decisions. For the detailed sizing of the reactor, the reader is referred to the many excellent texts on reactor design. 

Myriantheos, C. M., Flexihility Targets for Batch Process Design, M.S. thesis. University of Massachusetts, Amherst, 1986. 

In preliminary process design, the primary consideration is contact by inhalation. This happens either through accidental release of toxic material to the atmosphere or the fugitive emissions caused by slow leakage from pipe flanges, valve glands, and pump and compressor seals. Tank filling causes emissions when the rise in liquid level causes vapor in the tank to be released to the atmosphere. 

To the process designer, life-cycle analysis is useful because focusing exclusively on waste minimization at some point in the life cycle sometimes creates problems elsewhere in the cycle. The designer can often obtain useful insights by changing the boundaries of the system under consideration so that they are wider than those of the process being designed. 

The design of the reactor usually interacts strongly with the rest of the flowsheet. Hence a return must be made to the reactor when the process design has progressed further. 

Evaluation of design options. Costs are required to evaluate process design options e.g., should unconverted raw material be recycled or disposed of ... 

In preliminary process design it mig] it be necessary to use marginal costs for steam, but the designer should keep in mind the dangers of using such costs. 

Ulrich, G. D., A Guide to Chemical Engineering Process Design and Economics, Wiley, New York, 1984. 

---