Separation Synthesis Learning Outcomes PDF

Document Details

PanoramicEnlightenment5497

Uploaded by PanoramicEnlightenment5497

National University of Singapore

IA Karimi

Tags

separation synthesis chemical engineering process engineering chemical plants

Summary

This document covers learning outcomes for separation synthesis, detailing the importance of separations in chemical plants, separation network synthesis, and common industrial separation methods with their underlying principles. It also analyzes various separation methods and heuristics for separating multicomponent mixtures.

Full Transcript

Separation Synthesis – Learning Outcomes Explain the importance of separations in a chemical plant. Define the problem of separation network synthesis. List common industrial separation methods with their underlying principles. Identify and assess various methods...

Separation Synthesis – Learning Outcomes Explain the importance of separations in a chemical plant. Define the problem of separation network synthesis. List common industrial separation methods with their underlying principles. Identify and assess various methods for separating a given multicomponent mixture. Select the best. List and rationalize heuristic rules for separation network synthesis. Use algorithms to sequence networks of distillation-type columns. Copyright © IA Karimi 1 cost of separation may outweigh the value of the product Separations can Make / Break a Process Separations separate components, but also ChEs from others. Form the backbone of most plants. Represent most of the equipment in a plant. 1-2 reactors, but several columns, drums, separators, exchangers, … Incur most CAPEX and OPEX (energy) costs. Control product costs, if feed materials are cheap. Introduce new materials. Offer multiple alternatives. Mostly responsible for plant complexity. Make process optimization critical. Pose huge simulation challenges, so fluid package is critical. Nonideal and hence often nonintuitive behaviors are the norm. Copyright © IA Karimi 2 Separation Network Synthesis: Problem? Given Mixture of 𝐼-species (𝐼 1) mixture (Say NG: C1, C2, C3, C4) P, T, z (overall mixture composition) Desired separation products (e.g. Gas = C1+C2 & NGL = C2+C3+C4) Targets: Purity, impurity, recovery, flow, T, P, … 99% recovery of C1, 97% pure C1, 99% recovery for all, 99% purity for all, … Develop “best” process (PFD) to get desired products & targets. Separation methods and agents (energy??, mass??, equipment??) Separation units and their network arrangement P, T, feeds, and products of each unit Target for each product (composition, purity, recovery, …) Define best. Mini/maximize what technoeconomic goal? Power? Water? Emissions? Heating? CAPEX? Margin? Total cost? Copyright © IA Karimi 3 Propose Easiest Ways for Following Two streams from a VL mixture? What made it possible? density, volatility, boiling point temperature Think of physical properties, external forces, separating agents, … solid liquid slurry Two streams from an SL slurry? What made it possible? filtration process which used the property of difference in particle size Three streams from a VLL mixture? What made it possible? separation through difference in density of components What is the general name for all above separations? It is a key step that most separation methods rely on. phase separation Copyright © IA Karimi 4 Easiest (Most Natural) Separation of All? electrostatic precipitator : pass the Method: ???? phase separation gas through electric field and attracts and removes all the solid Definition / Description: ??? particles and lets gas go through Separating a multi-???? phase mixture into individual phase ???? using some physical property, driving force, and/or separation agent. cyclone flash, decanter, filtration, sieve, separator, Feed states: VL, GS, LL, SL, SS, VLL cyclone separator: due to centrifugal force, solids collect Product states: Individual V, L, S phases and deposit on the side Physical properties: Density, buoyancy, size, immiscibility,... Separation agents: Gravity, centrifugal force, electric field, filter, sieve, membrane, … Favored operating conditions: ??? Examples: Flash, Settling, Decanting, Centrifugation, Filtration, Electrostatic separation, … Copyright © IA Karimi 5 Distillation (Ordinary, Vacuum, Cryogenic) Definition / Description: Separating a mixture based on differences in volatility (BP). usually pressure high What determines Ordinary vs Vacuum vs Cryogenic?operating below 0C above atmospheric usually high operating below temperatures pressure and temp temperature atmoshperic pressure Feed states: ??? for distillation, feed state has to be either vapour, liquid or vapour-liquid important Product states: ??? top product is vapour and bottom product is liquid properties as distillation relies on Physical properties: Volatility, density, buoyancy,... used as a driving phase separation Separation agents: Energy, G???, p???, … pressure force for allowing gas flow through the column from bottom Favored operating conditions: ?? gravity to top low pressure better (higher pressure diff across pressure = higher boiling point) Examples: Crude oil distillation each stage Ordinary: Pure water and methanol from water+methanol. separation is better in below distillation NOT lower pressure -> better separation Copyright © IA Karimi atmospheric pressure POSSIBLE beyond critical pressure because 6 conditions but not both liq and gas are feasible as it is $$$ indistinguishable Absorption Definition / Description: ?? removing a solute from a gas phase using a solvent Physical & Chemical: ?? Feed states: ? one feed HAS to be gas, other feed (solvent) has to be liquid solvent + solute is liquid product flow Product states: ?? gas and liquid (two product flows) gas product flow has the solute removed Physical properties: Solubility, reactivity, volatility, … Separation agents: Liquid solvent, G???, p???, … (Mass) Favored operating conditions: ??? MSA - mass separating agent low temperatures + Examples: ?? higher pressures better for absorption carbon capture Copyright © IA Karimi 7 Stripping Definition / Description: ?? removing a component from the liquid phase based on volatility Feed states: ? liquid feed with a gaseous mass separating agent (like air) Product states: ?? liquid and gas Physical properties: Volatility, … Separation agents: Mass (??, ??, ), G???, p???, … Favored operating conditions: ??? Examples: ?? higher temperatures + low pressures better for stripping petroleum refinery : removing volatile components Copyright © IA Karimi 8 Other Separation Methods & Principles Condensation Evaporation Freezing Sublimation Desublimation Drying Copyright © IA Karimi 9 Other Separation Methods Precipitation acid base neutralisation Leaching / Extraction removing component from solid phase Caffeine from beans into water or supercritical CO2. Vegetable oil into hexane LL extraction Caffeine from water into dimethyl chloride or ethyl acetate Small amounts of benzene from water into toluene Acetic acid from ethyl acetate into water. Supercritical extraction extraction in supercritical conditions Copyright © IA Karimi 10 Azeotropic Distillation Azeotrope: Constant boiling mixture in which both L-phase and V-phase have same composition. because both phase have the same composition Ordinary distillation cannot separate, so how? Homogeneous: One L-phase, one V-phase Water + Ethanol Heterogeneous: Two immiscible L-phases, one V-phase Water + Butanol Minimum boiling: Azeotrope BP < BPs of components Water + Ethanol Maximum boiling: Azeotrope BP > BPs of components Water + HCl hydrogen bonding mainly responsible for non-ideality 11 Copyright © IA Karimi Other Distillation Methods Extractive: Solvent as a mass separating agent. one of the component will prefer the solvent and break azeotrope Reactive: Two definitions exist. React away one species to eliminate azeotrope. Perform reaction and distillation together in one column. Pressure swing (PSD): Principle: Azeotrope composition changes with pressure. Change pressure to break or change azeotropic composition Water + Ethanol by lowering pressure Salt effect (salted): Principle: Ionic salt (mass separating agent) in an aqueous mixture changes volatility or azeotrope composition Water + Ethanol by adding NaCl, CaCl2, … Copyright © IA Karimi 12 Other Separation Methods using solid material for separating agent adsorb at high pressure and then lower Adsorption pressure to remove adsorbed component Pressure swing (PSA) adsorb at low temp then heat up so that component is removed Temperature swing (TSA) Vacuum swing (VSA) vacuum conditions needed to remove component from adsorber (e.g CO2) Chromatography gas feed GC: Gas chromatograph liquid feed HPLC: High pressure liquid chromatography solute removed from solution using crystallisation Crystallization Most pharmaceutical processes Copyright © IA Karimi 13 Other Separation Methods Membrane pervaporation: Membrane as a separating agent allow some substances to evaporate through membrane Membrane permeation: Membrane as a separating agent particles of different sized are separated based on their size by a membrane Reverse osmosis Example?? water purification Salt precipitation 𝑃𝑏 from aqueous solution of 𝑃𝑏 𝐴𝑔 by adding KCl Salt substitution Organics (proteins) from water by an inorganic (ionic) salt Ethanol from water by adding Na2SO4 Copyright © IA Karimi 14 How to Select & Sequence Separations? especially if the components Does mixture have multiple phases? are rich in composition in different phases VL, SL, LL, VLL, VLS? multi-phase means phase separation Heuristic S1: Phase separate a multi-phase mixture, if a relatively sharp separation exists, otherwise not. VL: Can it be sharp always? Method? degree of separation not very good and not guaranteed in VL SL: Can it be sharp always? Method? degree of separation is guaranteed LL: Can it be sharp always? Method? need to use multi stage phase separation VLL: Can it be sharp always? Method? SS: Can it be sharp always? Method? Copyright © IA Karimi 15 Separating a 1-Phase Mixture Will the easiest separation help? How much energy? Can we create multiple phases without much energy? Will the phases represent a sharp separation? Heuristic S2: Convert 1-phase mixture into multiple phases, and phase separate, if a sharp split can be achieved. V  VL: Method? change temperature - partial condensation V  VS: Method? desublimation or freezing L  SL: Method? crystallisation some liquids may be miscible but beyond L  LL: Method? a certain temp they become immiscible V  VS: Method? L  VL: Method? partial evaporation Copyright © IA Karimi 16 Separating a Gas/Vapor Mixture Heuristic S3: Avoid mass as a separating agent (Prefer energy or equipment (filter, membrane, …). Why? if mass used then another separation process is required to remove the separating agent Why? adding external material increases process complexity as we have to remove, recycle, etc Heuristic S4: Consider absorption, adsorption, cryogenic distillation, and membrane separation. Copyright © IA Karimi 17 Separating a Liquid Mixture Heuristic S5: Consider ordinary distillation, stripping, adsorption, extractive distillation, LL extraction, membrane permeation (ultrafiltration), membrane pervaporation, crystallization, reverse osmosis, chromatography, supercritical extraction, and ion exchange resin. Heuristic S6: If azeotropic, then consider azeotropic, pressure-swing, salt-effect/salted, and reactive distillation. Homogeneous: Entrainer dissolves all components fully. Heterogenous: Entrainer forms an immiscible liquid phase. Copyright © IA Karimi 18 Separating a Liquid Solution Heuristic S7: Crystallize via chilling, if solubility decreases rapidly with temperature. Keep temperature at most 1 or 2 F below saturation. Example? sugar made from sugar cane juice being chilled Heuristic S8: Crystallize via evaporation, if solubility is not temperature-sensitive. Example? salt separation from water Copyright © IA Karimi 19 Ease of Separation Assessing difficulty/ease of a separation is useful in developing separation networks. Recall most methods exploit phase separation via two phases. VL, VS, LL, and SL Species distribute across two phases in varying degrees. Relative preference of a species to one phase versus another can tell us something about ease of separation. Methanol + Water: Methanol will prefer V-phase in distillation. Oil seeds: Oil will prefer Hexane-phase in extraction. Benzene + Water: Benzene will prefer Toluene-phase in LL extraction. How to quantify this preference? Copyright © IA Karimi 20 Define Key Components Separate a multi-species mixture M = A+B+C+D+E via a method (e.g. distillation, LL extraction, …) What phases in distillation? Vapour phase and liquid phase What phases in LL extraction? F=Feed & MSA=Solvent MSA = mass separating agent Identify physical property behind separation method. Distillation? Volatility is the property behind distillation LL extraction? Arrange species in the order of that physical property. ABCDE with A most and E least volatile. ABCDE with A most and E least soluble in MSA phase. Copyright © IA Karimi 21 Define and Use Key Components Select two adjacent species as keys to break M into two phases/products: A&B, B&C, C&D, or D&E. C&D: C is LK (Light Key) & D is HK in distillation. Distillation products: V=A+B+C & L=D+E. Extraction products: MSA=A+B+C & F=D+E. C & D are distributing across L & MSA. Others are NOT. If we can quantify the relative preference of C & D for distributing across two phases, then that will tell us how easy it is to break M into ABC & DE. Copyright © IA Karimi 22 Separation Factor for Ease of Separation Let I & II denote phases. A & B denote keys. Assume we wish to know preference for I vs II. SF between A & B is defined as, 𝑆𝐹 / 𝜙 = fugacity coefficient of 𝑖 (A or B) in phase 𝑗 (I or II). measures how A will distribute in I vs II relative to B. High means A will distribute much more than B in I. Consider M = A+B+C+D+D+E. Higher 𝑆𝐹 means splitting M into A+B+C and D+E is easier. Copyright © IA Karimi 23 SF for Distillation & Extraction Distillation: V and L are two phases/products. for an ideal gas, fugacity of a component is equal to the mole fraction of the Assume V is an ideal gas, and L is an ideal solution. component in vapour phase Then, 𝜙 𝑦 and 𝜙 𝑥. 𝑆𝐹 , 𝑦 /𝑥 / 𝑦 /𝑥 𝛼 , is relative volatility. Relative volatility indicates ease of separation in ideal distillation. relative volatility only valid for ideal gas and ideal solution LL extraction: MSA and F are two phases. By nature immiscibility , these are non-deal solutions. Let S be solute, and L be liquid in the feed. Fugacity becomes activity coefficient 𝛾, and 𝑆𝐹 , / How will you get 𝛾? using Van Laar, Margules and Wilson’s equation Copyright © IA Karimi 24 Columns in Separation Networks Most separation units give two products from a single feed. Exception: Crude oil distillation gives multiple products. How many products from a distillation column with a partial condenser? If ordinary distillation can separate a binary feed, then we need only one column. For non-ideal solutions, ordinary distillation will Methanol + Water, Benzene + Toluene NOT work How many 2-product columns to separate an I-component mixture into I pure components using ordinary distillation? Copyright © IA Karimi 25 Columns for a Multicomponent Mixture Let M = A+B+C. How many ways to split M via keys? Column Network 1: Column 1: M  A (V or top) & BC (L or bottom) Column 2: Column Network 2: Sequence 1? M  AB (V or top) & C (L or bottom) Sequence 2? Two networks of exist for a ternary M. How many 2-product columns in each network? Copyright © IA Karimi 26 Networks for I-Component Mixture Number of networks increases rapidly. 𝑁 6 𝑁 𝐼 𝑁𝑁 2 𝐼 1 !/𝐼! 𝐼 1 ! 𝑁 14 𝑁 42 Each network uses (I-1) 2-product columns. How to generate them and decide the best network? Exhaustive enumeration is time consuming, so heuristics are common, and systematic optimization methods continue to evolve. One must define best. Copyright © IA Karimi 27 Heuristics for Separation Networks Heuristic S9: Remove thermally unstable, corrosive, or chemically reactive components earlier. Rationale? How does early vs late affect the impact of these properties? safety, for preserving quality Heuristic S10: Remove species one by one as distillate.  𝐼 1 species are top products, heaviest one is last bottom. CH4+C2H6+C3H6+C3H8  CH4 & C2H6+C3H6+C3H8 C2H6+C3H6+C3H8  C2H6 & C3H6+C3H8 C3H6+C3H8  C3H6 & C3H8 This is called direct sequence. Copyright © IA Karimi 28 Heuristics for Separation Networks Heuristic S11: Remove species one by one as bottoms.  𝐼 1 species are bottom products, lightest is last top product. CH4+C2H6+C3H6+C3H8  CH4+C2H6+C3H6 & C3H8 CH4+C2H6+C3H6  CH4+C2H6 & C3H6 CH4+C2H6  CH4 & C2H6 This is called indirect sequence. CO and CH3OH Heuristic S12: Perform easier separations earlier. Consider M = CO+H2+CH3OH+Water. Which split is easiest? Heuristic S13: Perform difficult separations later. Consider M = CH4+C2H6+C3H6+C3H8. Which is most difficult? Network 1: ?? most difficult separation is C3H6 and C3H8 as their boiling points Network 2: ?? are going to be the closest Copyright © IA Karimi 29 Heuristics for Separation Networks Heuristic S14: Remove largest (mol fraction) species first. A+B+C (10:10:80): ??; A+B+C (10:80:10): ?? Heuristic S15: Remove higher purity-target species last. How does purity-target affect separation difficulty? higher purity How does removing other species first affect difficulty? products require more difficult separation Heuristics often compete with and/or contradict each other. No clear priority or sequence of one over another. Difficult to judge best network without doing rigorous simulation and costing. Consider product gains/losses in defining best. Copyright © IA Karimi 30 Heuristics for Separation Networks Heuristic S16: It is easier to achieve higher purity in top versus bottom products due to the presence of high boiling impurities and solids in most industrial mixtures. Why? Which product gets evaporated? top product will just be one component evaporated but bottom product will contain impurities with high boiling point Heuristic S17: Direct sequence is widely used in the industry, but indirect is the least desirable. Can you explain using S16? Copyright © IA Karimi 31 Example for Separation Heuristics Hydrocarbon F = C3+iC4+nC4+iC5+nC5 Species flows are 45.4, 136.1, 226.8, 181.4, 317.5 kmol/h. Desired recoveries are 98% for each species. s ( ’s) for various splits or key pairs are: C3/iC4 = 3.6, iC4/nC4 = 1.5, nC4/iC5 = 2.8, iC5/nC5 = 1.35. S9 does not apply. S17 rules out indirect sequence. S15 (Highest purity last) does not apply. All are high-purity. Copyright © IA Karimi 32 Apply S10: Direct Sequence C3 Column 4 F iC4 iC4+nC4+ iC5+nC5 nC4 nC4+ iC5 Column 1 iC5+nC5 iC5 Column 2 +nC5 Column 3 nC5 Copyright © IA Karimi 33 Apply S12: Easiest Separation First C3 iC4 F iC4+nC4 Column 3 iC4+nC4+ nC4 iC5+nC5 Column 2 Column 1 iC5 iC5+nC5 Column 4 nC5 Copyright © IA Karimi 34 Example for Separation Heuristics S14 suggests largest species first. nC5 is the largest component, so F  C3+iC4+nC4+iC5 & nC5. However, iC5 / nC5 is the most difficult separation, which contradicts S13: Hardest separation last. Which of S10 vs S12 is the best? S12 gives the “best” sequence, but how do you prove? No easy / clear solution. Highly qualitative judgements. No choice, but faith! Copyright © IA Karimi 35 Most Important Heuristic Heuristics are not inviolable or infallible laws. You decide if it makes sense in a given situation. You can violate them, if situation warrants. The most important heuristic per Prof Towler is: Heuristic S18: Never use a heuristic, unless you can rationalize it and/or derive it. Heuristics do not give you clear unambiguous solutions. Highly qualitative suggestions. Other recent methods offer some quantitative bases. Copyright © IA Karimi 36 Marginal Vapor Rate Method Modi & Westerberg (1992) from CMU (USA) Needs rough simulation, but not complete column designs or costs. Premise 1: Adding a non-key species complicates separation and increases cost.  If B/C are keys, then AB / C is costlier than B / C. Premise 2: This increase (called marginal annualized cost) is well measured by marginal increase in molar vapor flow (MV) exiting the column top. Heuristic 19: Use network with minimum total MVR. Consider A and B as key components. Split A / B has a vapor flow of 𝑉 / kmol/h. Split A / BC has a vapor flow of 𝑉 / > 𝑉 / kmol/h. Marginal vapor flow 𝑀𝑉𝑅 / = 𝑉 / 𝑉/. Copyright © IA Karimi 37 Apply Marginal Vapor Rate Method F = C3-iC4-nC4-iC5-nC5. Apply S10: Easiest separation. Separate C3 first. Apply method to A+B+C+D. Recover at least 99.9% of each component in each column. Use Short Cut Distillation to get top vapor flows. Split Top R= Vapor MVR Split Top R= Vapor MVR Pressure 1.2 Rate Pressure 1.2 Rate (kPa) 𝑹𝒎𝒊𝒏 (kmol/h) (kPa) 𝑹𝒎𝒊𝒏 (kmol/h) A/B 680 10.7 1,594 0 B/CD 490 3.06 921 227 A/BC 680 11.9 1,757 163 AB/CD 560 2.11 1,129 435 A/BCD 680 13.2 1,934 340 C/D 210 13.5 2,632 0 B/C 490 2.06 694 0 BC/D 350 6.39 3,017 385 AB/C 560 1.55 925 231 ABC/D 430 4.96 3,245 613 Direct: A/BCD + B/CD + C/D = 340 + 227 = 567 kmol/h S12: AB/CD + A/B + C/D = 435 kmol/h, hence it is the best. Copyright © IA Karimi 38 Reflect on Heuristic MVR Method Is it easy to get top vapor flows? What do you need to do? What is the method’s primary goal? Does it compute that goal rigorously? goal is least cost but usually least cost will not be most optimal process Is that goal critical in practice? Is it most important in deciding the best? What is more important in practice? should consider cost AND revenue Given the work involved and its limitations, what is the best approach to get the best sequence? Copyright © IA Karimi 39

Use Quizgecko on...
Browser
Browser