+ All Categories
Home > Documents > 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING...

08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING...

Date post: 10-Feb-2018
Category:
Upload: akash-agrawal
View: 216 times
Download: 0 times
Share this document with a friend

of 76

Transcript
  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    1/76

    1

    Chapter 8

    Introduction to

    Non-ideal Reactors

    Our treatment on reactor design in the

    previous chapters is based on the assump-

    tion that the reactorsare ideal

    What are ideal reactors?

    For CSTRs,

    they must be well-mixed the concentration of the exit

    stream is equal to that of fluid in

    the reactor

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    2/76

    2

    For PFRs, theflowin the reactor must

    be in order, i.e. there are

    no overtaking () no backmixing ()

    With these idealbehaviours/character-istics, at steady state, the concentrationof

    the fluidflowing outof the reactor is

    uniform(or constant)

    In reality, however, no reactorsare

    ideal, i.e. theflow patternin the reactor

    (either CSTRs or PFRs) is non-ideal

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    3/76

    3

    The main reasonsof the non-ideal

    flow (or the deviation of the ideal flow) are:

    1) the residence time distribution(RTD)

    For ideal reactors, it is assumed

    that the residence time(or space time, )

    ofall substances(all molecules) in the

    reactors are identical, i.e. all molecules

    that enter the reactor at the same time

    must exit the reactor at the same residence

    time; no molecules stay in the reactor

    shorter or longer than the residence time

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    4/76

    4

    For non-ideal reactors(or real

    reactors), there may be some molecules

    whose flow rates are slower or faster than

    those of others

    Channellingof fluid causes some

    molecules to havefasterflow ratethan

    others, as illustrated below

    Channelling of Fluid

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    5/76

    5

    A stagnant(/) zone

    in the reactor, on the contrary, delaysthe

    flowof some molecules, causing such

    molecules to flow out of the reactor slower

    than others

    Stagnant zone in a CSTR

    Stagnant Zone

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    6/76

    6

    2) State of aggregationFor ideal reactors, all molecules

    are assumed to be negligibly small(when

    compared to the reactor); thus, they can

    move freely

    In reality, a group ofmolecules

    are combined together(i.e.aggregationof

    molecules) to form macrofluid molecules

    (as apposed to microfluid molecules as

    per ideal reactors)

    These macrofluid molecules move

    in package(or in aggregates), thus causing

    non-ideal flowbehaviour to occur

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    7/76

    7

    Microfluid

    Macrofluid

    3) Earliness or lateness of mixingFor an ideal CSTR, it is assumed

    that the reactor is well-mixed, meaningthat the concentrations at any position

    (point) of the CSTR are identical throughout

    the reactor

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    8/76

    8

    The mixing in the non-idealCSTR,

    on the contrary, is NOTideally perfect,

    causing the concentrationsin the CSTR

    vary from position to position

    For an ideal PFR, it is assumed

    that there are no overtaking and/or no

    backmixing (i.e. the flow is uniform or in

    order)

    In a real PFR, however, there may

    be some overtaking or backmixing, which

    results in early mixing or late mixing in the

    PFR as illustrated on the next Page

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    9/76

    9

    Uniform Mixing

    (Ideal Flow in a PFR)

    EarlyMixing

    LateMixing

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    10/76

    10

    RTD Function (or The Age Distribution

    of Fluid)

    As mentioned earlier, the residence

    timesof molecules in a real (actual)

    reactor are NOTidenticalas per the ideal

    reactor

    In other words, there is a distribution

    of the residence timesoffluidentering

    and leaving the reactor

    The distribution of these times for the

    stream of fluid leavingthe reactor is called

    the exit age distribution, E (we shall

    discuss about Ein detail later)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    11/76

    11

    In order to determine the distribution of the

    residence times (or the degree of the non-ideal

    of the flow), a small amount of substance

    (called a tracer) is injected into the reactor,

    and the concentrations of the tracer flowing

    out of the reactor are recorded

    A tracershould be

    easy to analyse for itsconcentration(using an ordinary

    equipment available in a common

    laboratory, e.g., pH meter, gas

    chromatograph: GC)

    stable(i.e. it is not decomposed easily) inactive(i.e. it does not react with any

    other species in the reactor)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    12/76

    12

    To introduce (or inject) tracerinto the

    reactor, it can be done by

    apulseexperiment a stepexperiment

    The Pulse Experiment

    In thepulseexperiment, a known

    amount of tracer is injected into the

    reactor onlyonceat any instant of time

    Assume that a tracer in the amount of

    M kg or kmol (kg-mol) is injected into thefluid entering the reactor (or vessel) with

    the volumetric flow rate ofv m3/s

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    13/76

    13

    The concentration vstime (which can

    be called time-resolved concentration) of

    the tracer leaving the reactor is monitored,

    and the plot between the concentration of

    the tracer ( )pulseC and time ( )t for an ideal

    reactor and a non-ideal reactor can beillustrated on the following Pages (Pages

    1415)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    14/76

    14

    For an ideal flow

    For the case of an idealflow, after a

    certain period of time (which is, in fact, the

    residence time), the tracershall leave the

    reactor at oncewith the concentration

    equalto that enteringthe reactor

    Cpulse

    tResidence time

    =out inC C

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    15/76

    15

    For a non-ideal flow

    For a non-idealflow pattern (or a non-

    ideal reactor), there is a distributionoftimesthat the tracer leaves the reactor

    from = 0t to t=

    Cpulse

    t

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    16/76

    16

    The area under the pulseC vs t curve for a

    non-ideal reactor from = 0t to t= is

    ( )0

    C t dt

    The unit of the integral is, for example,

    [ ]mol mol s

    sL L

    =

    or

    [ ]3 3kg kg ssm m

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    17/76

    17

    Multiplying the integral with the

    volumetric flow rate gives

    ( )0

    v C t dt

    and the unit of the resulting multiplication

    is, for example,

    L mol s

    s L

    = mol

    or

    3

    3m kg s kgs m

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    18/76

    18

    By considering the unit of the term

    ( )0

    v C t dt

    , it leads to the fact that this term

    is, in fact, the total amount of the tracer

    leaving the reactor, which, by employing

    the principle of the mass conservation, is

    equal to that entering the reactor

    Thus,

    ( )0

    M v C t dt

    = (8.1)

    which can be re-arranged to

    ( )

    = =

    0

    Area under

    the curve

    from 0

    M C t dtv

    (8.2)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    19/76

    19

    If the integration is performed from = 0t

    to t t= ,

    ( )0

    t

    C t dt

    we obtain the fact that

    ( )0

    t

    m v C t dt = (8.3)

    which can be re-arranged to

    ( )

    = ==

    0

    Area under

    the curvefrom 0

    tm

    C t dtvt

    (8.4)

    where m is the mass (or moles) of the

    tracer leaving the reactor between the time

    period of = 0t to t t= as shown graphically

    on the next Page

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    20/76

    20

    Cpulse

    t t=t

    ( )0

    Area under

    the curvefrom 0

    t

    mC t d t

    vt

    = =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    21/76

    21

    Dividing throughout Eq. 8.3:

    ( )0

    t

    m v C t dt =

    (8.3)

    withM

    vand re-arranging yields

    ( )0

    t

    v C t dtm

    M M

    v v

    =

    ( )0

    t

    v C t dt

    mvMM

    v

    =

    ( )0

    t

    C t dtm

    MM

    v

    =

    (8.5)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    22/76

    22

    It is defined that

    ( )

    ( )0

    0

    t

    t

    C t dt

    E t dtM

    v

    = (8.6)

    Thus, the meaning ofE (the exit age

    distribution) or ( )E t is, in fact,

    ( )( )C t

    E tM

    v

    = (8.7a)

    pulseCE

    M

    v

    = (8.7b)

    ( )

    pulse pulse

    0

    Area under

    the curve

    from 0

    C C

    EC t dt

    = =

    (8.7c)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    23/76

    23

    In other words, we can obtain Eat any

    instant of time by simply dividing the value

    of the concentration of a tracer ( )pulseC by

    M

    vor by ( )

    0

    C t dt

    or by the area under the

    curve from 0t= to t=

    From Eq. 8.6, when t= , we obtain the

    fact that

    ( )

    ( )0

    0

    C t dt

    E t dtM

    v

    =

    (8.8)

    but, from Eq. 8.2, ( )0

    MC t dt

    v

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    24/76

    24

    Hence, Eq. 8.8 becomes

    ( )0

    1

    M

    vE t dt M

    v

    = = (8.9)

    ( )E t is also called the RTD functionwhose graph are as shown below

    E

    t

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    25/76

    25

    It is evident that the E vs tplot is

    identical to that ofC vs t, as E is, in fact,

    impulseCE

    M

    v

    = (8.7b)

    Since the area under the curveof the C

    vs tplot from 0t= to t= or ( )0

    C t dt

    is

    M

    v(see Page 18), the area under the

    curveof the E vs tplot from 0t= to t=

    or ( )0

    E t dt

    or0

    Edt

    is 1, as illustrated

    mathematically on Page 24 (Eq. 8.9)

    Note that by dividing Cor pulseC with a

    constant (i.e.Mv ), it is called normalisation

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    26/76

    26

    We can employ the RTD function to

    calculate the average (or mean) residence

    time using the following relationship:

    ( )0

    t tE t dt

    = (8.10)

    or

    ( )

    ( )

    0

    0

    tC t dt

    t

    C t dt

    =

    (8.11)

    which can be proved mathematically as

    follows

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    27/76

    27

    From Eq. 8.8, we obtain the fact that

    ( )

    ( )0

    0

    C t dt

    E t dtM

    v

    = (8.8)

    but, from Eq. 8.2,

    ( )0

    M C t dtv

    = (8.2)

    Combining Eq. 8.8 with Eq. 8.2 results

    in

    ( )

    ( )

    ( )

    0

    0

    0

    C t dt

    E t dt

    C t dt

    =

    (8.12)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    28/76

    28

    Multiplying throughout Eq. 8.12 with t

    gives

    ( )

    ( )

    ( )

    0

    0

    0

    tC t dt

    tE t dt

    C t dt

    =

    (8.13)

    but, from Eq. 8.10, which is defined that

    ( )0

    t tE t dt

    = (8.10)

    Thus, Eq. 8.13 is, in fact,

    ( )

    ( )

    0

    0

    tC t dt

    t

    C t dt

    =

    (8.11)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    29/76

    29

    Example To determine the non-ideal

    behaviour of a reactor with a volume of

    180 L, a pulse experiment is carried out by

    injecting a tracer only once in the amount

    ofM = 1.2 kg into the inlet stream flowing

    into the reactor with volumetric flow rate of

    11 L/min

    The concentration vs timedata of the

    tracer in the outlet stream of the reactor

    within the period of 035 min are as shown

    in the Table on the next Page

    Determine

    a) the mean residence time( )t ofthe flow in this reactor

    b) the residence time in this reactorif the flow is ideal

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    30/76

    30

    Time [min] C[g/L]

    05

    101520253035

    03

    554210

    The plot ofC vs tof this pulse

    experiment is as illustrated on the next

    Page (Page 31)

    Determining the area under the curve

    (try doing it yourself) gives

    ( )0

    M

    C t dtv

    = (8.2)

    which can also be written in thepractical

    form as

    ( )iM C t tv

    = (8.14)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    31/76

    31

    0

    1

    2

    3

    4

    5

    6

    0 10 20 30 40

    Time [min]

    Concen

    tration

    oftracer[g/L]

    The meaning of Eq. 8.14 is that the

    integral ( )0

    C t dt

    can practically be obtained

    by summing the multiplication ofC (at any

    instant of time) and t (or 1n nt t+ )

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    32/76

    32

    The area under the curve of the graph

    on the previous Page (Page 31) is found to

    be 100 g min

    L ; thus,

    ( )0

    g min100

    L

    MC t dt

    v

    = =

    and

    g min100

    L L11

    min

    g min L 100 11L min

    M

    M

    =

    =

    1,100 g= = 1.1 kgM

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    33/76

    33

    It is given that the tracer is injected

    into the reactor in the amount of1.2 kg,

    but the data from the outlet streamyield

    the fact that the total amountof the

    tracer leavingthe reactor is 1.1 kg

    Although there is a discrepancy (

    ), but it seems to be acceptable, as

    the error is only1

    ~ 100 8.3%

    12

    =

    The mean or average residence time ( )t

    can be calculated using Eq. 8.11

    ( )

    ( )

    0

    0

    tC t dt

    t

    C t dt

    =

    (8.11)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    34/76

    34

    which can also be written in thepractical

    form as follows

    ( )( )

    i i

    i

    t C t t t

    C t t

    =

    (8.15)

    Note that the term( )i

    C t t

    (or the

    denominator: ) is, in fact, ( )0

    C t dt

    ,

    which is the area under the Cpulse vs tcurve

    and is equal tog min

    100L

    (see Page 32)

    The computation of the value of

    ( )0

    tC t dt

    or ( )( )i it C t t is illustrated

    in the Table on the next Page

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    35/76

    35

    t [min] iC [g/L] t ( ) i it C t 0 0

    5 3 5 535 = 75

    10 5 5 1055 = 25015 5 5 1555 = 37520 4 5 2045 = 40025 2 5 2525 = 25030 1 5 3015 = 15035 0 5 3505 = 0

    = 1,500

    Thus, by employing Eq. 8.15, we can

    calculate the mean residence time ( )t as

    follows

    ( )

    ( )2g min

    1,500 Lg min

    100L

    i i

    i

    t C t t t

    C t t

    =

    =

    15 mint =

    a)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    36/76

    36

    The residence timefor the ideal-flow

    pattern ( ) can simply be calculated using

    the following equation:

    180 L

    L11min

    V

    v =

    =

    16.4 min =

    b)

    It is evident that the mean residence

    time ( )t for the non-idealflow (or the

    actualcase) is shorterthan the ideal-flow

    residence time (15 min vs16.4 min)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    37/76

    37

    This implies that, on average, the

    molecules stay in the reactor shorter

    than the required residence time, which

    should lead to the fact that the conversion

    may be lower than expected

    The reasons that the mean residence

    time is lower than the ideal residence time

    may include the following:

    There is a channellingof the flow inthe reactor

    There is a stagnant zonein thereactor, which makes the volume( )V of the reactor smaller than the

    actual one; thus,V

    v = is lower

    than expected

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    38/76

    38

    The Step Experiment

    Another way of determining the non-

    idealflow pattern of the reactor is to carry

    out a step experiment

    In the stepexperiment, at any instant of

    time ( )or at 0t= , a tracer is introduced

    into the inlet stream flowing into the

    reactor (volume = V m3) with the volume-

    tric flow rate ofv m3/s

    The traceris injected continuouslyintothe reactor with the concentration

    (relative to the inlet stream) of maxC

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    39/76

    39

    The concentration of the tracer in the

    outlet stream leaving the reactor can be

    presented graphically as follows

    Let

    ( )

    ( )

    max

    C t

    F t C=

    (8.16)

    Cstep

    t

    Cmax

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    40/76

    40

    The concentration of the tracer at any

    instant of time, C or ( )C t , is defined asm

    V

    (on massbasis) or

    ( )m

    C tV

    = (8.17)

    where m is the total amount of mass

    leaving the reactor from 0t= to t t=

    Let concentration of maxC be

    max

    MC

    V= (8.18)

    Combining Eqs. 8.17 & 8.18 with Eq.8.16 yields

    ( )( )

    max

    mC t VF t

    MC

    V

    = = (8.19)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    41/76

    41

    For a constant-V system, Eq. 8.19

    becomes

    ( ) ( )

    max

    C t mF t

    C M= = (8.20)

    but, as we have learned previously (or from

    Eq. 8.3)

    ( )0

    t

    m v C t dt = (8.3)

    and Eq. 8.1

    ( )0

    M v C t dt

    = (8.1)

    which can be combined to form

    ( )

    ( )

    ( )0

    0

    0

    t

    tv C t dt

    mE t dt

    Mv C t dt

    = =

    (8.21)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    42/76

    42

    Combining Eqs. 8.20 & 8.21 together

    yields

    ( )( )

    ( )max 0

    tC t mF t E t dt

    C M= = =

    or

    ( ) ( )0

    t

    F t E t dt = (8.22)

    Eq. 8.22 is the relationshipbetween F

    (from a stepexperiment) and E(from a

    pulseexperiment)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    43/76

    43

    Eq. 8.22 can be written in the differential

    form as follows

    ( ) ( )dF t

    E tdt

    = (8.23)

    which implies mathematically that the

    value ofEor ( )E t can be obtained by

    differentiating the graph ofF vs tat any

    given time ( )t

    Note that

    max

    MC

    V=

    can also be written as

    max

    MmtC

    V v

    t

    = =

    (8.24)

    where m

    is the mass flow rate of the tracer

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    44/76

    44

    The shaded area on the stepC vs t curve:

    is

    max

    step

    0

    C

    tdC

    max

    max

    0

    Shaded

    area

    C

    tdC

    =

    Cstep

    t

    Cmax

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    45/76

    45

    Thus, the average residence time( )t for

    the stepexperiment can be calculated

    using the following equation:

    max max

    max

    step step

    0 0

    max

    step

    0

    C C

    C

    tdC tdC

    tC

    dC

    = =

    (8.25)

    Eq. 8.25 can be written in thepractical

    form as

    ( )

    max

    i it CtC

    = (8.26)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    46/76

    46

    Conversion in Non-ideal Flow Reactors

    In the previous chapters, our treatment

    on reactor design is based on the ideal

    model, in which the conversionis affected

    by

    thermodynamic constraint[maximum possible conversion

    (how faror how much) or K:

    equilibrium constant]

    rate (how fast) of the Rxn. (or k:rate constant)

    design equation of the reactor (ortype)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    47/76

    47

    For the non-idealreactors, in addition

    to those factors, the following additional

    factors must also be taken into

    consideration in determining/calculating

    the conversion of the system (or reactor)

    RTD of fluid in the reactor

    Micro- or macro-fluid behaviour Earliness & lateness of mixing

    As we have learned previously, these

    factors cause the reactor to deviate

    () from the ideal behaviours

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    48/76

    48

    We have just learned the effect of the

    RTDon the non-ideal behaviours of the

    reactor

    In the subsequent sections, lets

    consider the effects of the earliness & late-

    ness of mixingand the state of aggregation

    (or the macrofluidbehaviour)

    If the early mixingtakes place in the

    plug flow reactor (PFR) as shown below

    EarlyMixing

    the section in which the early mixing is

    occurring will behave as a CSTR

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    49/76

    49

    Thus, the concentration in this section

    will

    reduce substantially be uniform quickly

    The resulting lowconcentration of thereactant will then proceed its conversion

    along the length of the PFR

    On the contrary, if the late mixing

    occurs in the PFR:

    LateMixing

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    50/76

    50

    the concentration of the reactantwill decrease gradually along the

    length of the PFR during the first

    stage of the reaction

    but it will drop quickly after itenters the mixing zone, whichbehaves as a CSTR

    In either case (early or late) of mixing, if

    the order of the Rxn. ( )n is lowerthan unity( )1 : e.g., n = 0.5

    or

    0.5

    A A

    r kC = (8.27)

    the rate of Rxn. ( )Ar would be

    decreasing at the rate slowerthan AC (e.g.,

    when AC decreases by 2 folds, Ar would be

    decreasing at the rate of 0.52 = 1.44 folds)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    51/76

    51

    higherthan unity: e.g., n = 2 or2

    A Ar kC = (8.28)

    the rate of Rxn. ( )Ar would be

    decreasing at the ratefasterthan AC (e.g.,

    when AC decreases by 3 folds, Ar would be

    decreasing at the rate of 23 = 9 folds

    Hence, to obtain as high conversion as

    possible, when

    n< 1, we want the reactor to havean earlymixing, in order to enable

    AC to drop substantially during the

    early stage of the reaction n> 1: we desire to have a late

    mixing, to obtain as high rate of

    Rxn. during the early stage of the

    reaction as possible

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    52/76

    52

    The relationship between the

    concentration of A (reactant) at any instant

    of time ( )AC and its initial concentration

    ( )oA

    C for a PFR or a BR for the nth order

    Rxns. are as follows

    1st order Rxn.:

    ( )expo

    A

    A

    Ckt

    C= (8.29)

    2ndorder Rxn.:1

    1o o

    A

    A A

    C

    C kC t =

    +(8.30)

    nthorder Rxn.:

    ( )

    11 1

    1 1 oo

    nA n

    AA

    C

    n C kt C

    = + (8.31)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    53/76

    53

    For the non-idealPFR or BR, since the

    concentration of A does NOT leave the

    reactor at once (i.e. it leaves the PFR or BR

    with the different value from 0t= to t= ,

    as shown as a graph on Page 15), the

    average concentration ofA

    C ( )AC that

    leaves the reactor during the time interval

    of 0 must be used instead of the exact

    value of AC

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    54/76

    54

    The average (or mean) concentrationof A

    with reference tooA

    C o

    A

    A

    C

    C

    for the non-

    idealPFRs or BRs can be calculated using

    the following equation:

    0o o

    A A

    A A

    C C EdtC C

    =

    (8.32)

    and by employing the same principle, the

    average conversionof A ( )Ax can be

    computed using the following equation:

    0

    A Ax x Edt

    = (8.33)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    55/76

    55

    Example The reaction with the rate

    equation of A Ar kC = , where k = 0.307

    min-1, is taken place in a 180-L reactor

    with a volumetric flow rate of the inlet

    stream of 11 L/min

    Determine the conversion of A ( )Ax ifa) the reactor is the idealPFRb) the reactor is the non-idealPFR

    (use the data of the Example on

    Pages 2930 to determine the non-

    ideal behaviour of the reactor)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    56/76

    56

    For PFRs and 1st order Rxns.,

    ( )expo

    A

    A

    Ck

    C

    = (8.29)

    Assume that this is a constant-V system

    Thus,

    ( )1oA A A

    C C x= (1.24)

    Combining Eq. 1.24 with Eq. 8.29 and

    re-arranging results in

    ( )( )

    ( ) ( )

    1exp

    1 exp

    o

    o

    A A

    A

    A

    C xk

    C

    x k

    =

    =

    ( )1 expAx k= (8.34)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    57/76

    57

    For an ideal PFR,

    180 L16.4 min

    L11min

    V

    v

    = = =

    Hence, the conversion of A for the ideal

    PFR can be calculated, using Eq. 8.34, as

    follows

    ( ) ( )[ ]1 exp 0.307 16.4

    1 0.00651

    Ax =

    =

    0.993Ax =

    a)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    58/76

    58

    If the reactor is non-ideal, the average

    conversion ( )Ax can be computed using

    Eq. 8.33:

    0

    A Ax x Edt

    = (8.33)

    The value ofEcan be calculated using

    Eq. 8.7c

    ( )

    pulse pulse

    0

    Area under

    the curve

    from 0

    C CE

    C t dt

    = =

    (8.7c)

    From the previous Example, the area

    under the C vs tcurve or ( )0

    MC t dt

    v

    = was

    found to beg min

    100L

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    59/76

    59

    Thus, in this Example, the value ofEis,

    in fact,

    pulse

    100 100

    C CE= =

    The conversion at any given residence

    time can be computed using Eq. 8.34:

    ( )1 expAx k= (8.34)

    From the given data (on Pages 2930),

    we can compute the mean conversion of A

    ( )Ax using Table as shown on the next

    Page

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    60/76

    60

    t CE=

    C/100( )= 1 expAx k Ax E t

    0

    5

    101520253035

    0

    3

    554210

    0

    0.03

    0.050.050.040.020.01

    0

    1 exp[-(0.307)(5)]= 0.7850.9540.9900.9980.999

    11

    0.785

    0.03

    5= 0.11780.23850.24750.19960.09990.05

    0 = 0.953

    Note that the term0

    Ax Edt

    can be

    written in a practical form as follows

    ( )Ax E t (8.35)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    61/76

    61

    Thus, the average conversion of A

    ( )A

    x , or0

    A

    x Edt

    or ( )

    A

    x E t

    , for the

    non-ideal PFR is found to be 0.953

    b)

    Note that, in this Example, the

    conversion of the non-idealPFR is lower

    than that of the idealone (0.953 vs0.993)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    62/76

    62

    In the case ofCSTRs, whose design

    equation is

    oA A

    A

    C C

    r

    =

    (1.40)

    If the Rxn. is of 1st order:

    A Ar kC =

    by combining the rate equation (1st order)

    with the design equation (of a CSTR), we

    obtain

    oA A

    A

    C C

    kC

    = (8.36)

    Re-arranging Eq. 8.36 gives

    oA A

    A

    C Ck

    C

    =

    1oA

    A

    C

    k C =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    63/76

    63

    1oA

    A

    Ck

    C= +

    1

    1o

    A

    A

    C

    C k=

    +(8.37)

    Thus, for a non-idealCSTR and the 1st

    order Rxn., we obtain the following

    equation:

    0o o

    A A

    A A

    C C EdtC C

    =

    0

    1

    1o

    A

    A

    CEdtC k

    = + (8.38)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    64/76

    64

    Example Repeat the Example on Page 55

    but for ideal and non-idealCSTRs

    Since the volume of the reactor (CSTR),

    V, and the volumetric flow rate, v, of the

    inlet stream are still the same as per the

    previous Example, the residence time of

    the ideal CSTR is also the same; i.e.

    180 L16.4 min

    L11min

    = =

    Substituting 16.4 min = and 0.307k=

    -1min into Eq. 8.37 yields

    ( ) ( )

    1

    1 0.307 16.4o

    A

    A

    C

    C=

    +

    0.166o

    A

    A

    C

    C=

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    65/76

    65

    Re-arranging Eq. 1.24:

    ( )1oA A A

    C C x= (1.24)

    results in

    ( )1o

    AA

    A

    Cx

    C=

    1o

    AA

    A

    Cx

    C= (8.39)

    Thus,

    1 1 0.166 0.834o

    AA

    A

    Cx

    C= = =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    66/76

    66

    For the non-ideal CSTR,

    0o o

    A A

    A A

    C CEdt

    C C

    =

    (8.32)

    The above equation (Eq. 8.32) can be

    written for the 1st

    order Rxn. taken placein a CSTR as

    0

    1

    1o

    A

    A

    CEdt

    C k

    =

    + (8.38)

    Applying Eq. 8.39 for the non-ideal

    reactor (can be any type of reactor; BR,

    CSTR, or PFR) yields

    1o

    AA

    A

    Cx

    C= (8.40)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    67/76

    67

    Combining Eq. 8.40 with Eq. 8.38 gives

    0

    11

    1Ax Edt

    k

    =

    + (8.41)

    which can be written in a practical form as

    follows

    11 1

    Ax E tk

    = +

    (8.42)

    From the given data, we can compute

    the value ofo

    A

    A

    C

    Cfor a CSTR and 1st order

    Rxn. or

    1

    1 E tk

    + as illustrated

    in the Table on the next Page

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    68/76

    68

    t CE=

    C/100

    1

    1+k

    1

    1 + E tk

    0

    5

    1015202530

    35

    0

    3

    55421

    0

    0

    0.03

    0.050.050.040.020.01

    0

    1/[1+(0.307)(5)]= 0.3940.2460.1780.1400.1150.098

    0.085

    0.394 0.5 5= 0.05910.06150.04450.02800.01150.0049

    0 = 0.210

    Thus,0

    1

    1Edt

    k

    + or

    1

    1E t

    k

    +

    oro

    A

    A

    C

    Cis

    0.210o

    A

    A

    C

    C

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    69/76

    69

    Hence, the average conversion for the

    non-ideal CTSR in this Example can be

    calculated, using Eq. 8.40, as follows

    1 1 0.210

    0.790

    o

    AA

    A

    Cx

    C= =

    =

    b)

    The conversion of the non-idealCSTR

    is found to be lowerthan the idealone(0.834 vs0.790)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    70/76

    70

    In the case that the fluid in the reactor

    behaves as a macrofluid, the residence

    time distribution (RTD) of the tracer in the

    outlet stream for thepulseexperiment is

    as shown in the following Figure

    Cpulse

    t

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    71/76

    71

    The graph on the previous Page

    illustrates that the fluid are combined

    together to become a lump or a group of

    fluid

    This lump of fluid moves together from

    one place to another, and, eventually, it

    comes out of the reactor as a lump of fluid

    The example of the calculations

    concerning the non-ideal, macrofluid

    behaviour is illustrated in the following

    Example

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    72/76

    72

    Example The liquid A in the form of

    macrofluidhas an initial concentration

    ( )oAC of 2 mol/L is decomposed according

    to the rate equation of

    2 L; 0.5

    mol min

    A Ar kC k = =

    The RTD for the pulse experiment is as

    shown below

    Determine the conversion of this Rxn. if

    it is taken place in a BR

    Cpulse

    [mol/L]

    t[min]

    2

    1 3

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    73/76

    73

    Since this is a non-ideal, macrofluid

    flow pattern, we can employ the following

    relationships:

    0o o

    A A

    A A

    C CEdt

    C C

    =

    (8.32)

    and

    1o

    AA

    A

    Cx

    C= (8.40)

    For the 2nd order Rxn.,

    1

    1o o

    A

    A A

    C

    C kC t =

    +(8.30)

    Combining Eq. 8.30 with Eq. 8.32 yields

    0

    1

    1o o

    A

    A A

    CEdt

    C kC t

    = +

    (8.43)

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    74/76

    74

    We have learned that

    ( )

    pulse pulse

    0

    Area underthe curve

    from 0

    C CE

    C t dt

    = =

    (8.7c)

    In the Example, the area under thecurve from 0 - can be computed as

    follows

    ( ) ( )[ ]

    pulse

    Area under

    the curve

    2 mol/L 3 1 min

    C t

    =

    =

    Area under mol min4the curve L

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    75/76

    75

    Hence,

    pulse -1

    mol2

    L 0.5 minmol minArea under4

    Lthe curve

    CE= = =

    Substituting corresponding numericalvalues into Eq. 8.43 and integrating gives

    ( ) ( )

    ( )[ ]

    ( ) ( )[ ]

    ( )

    0

    0

    3

    1

    3

    1

    1

    1

    10.5

    1 0.5 2

    10.5

    1

    0.5 ln 1

    0.5 ln 1 3 ln 1 1

    0.5 ln4 ln2

    40.5ln2

    o o

    A

    A A

    CEdt

    C kC t

    dtt

    dtt

    t

    = +

    = +

    =+

    = +

    = + +

    =

    =

  • 7/22/2019 08 Introduction to Non-Ideal Reactors....by akash agrawal.itis good for every one: I THINK GIFTING IS GOOD TECH

    76/76

    0.5 ln2 0.347o

    A

    A

    C

    C= =

    but, from Eq. 8.40, 1o

    AA

    A

    CxC

    =

    Thus, the average conversion of A for

    non-ideal, macrofluid in this Example can

    be computed as follows

    1

    1 0.347

    0.653

    o

    AA

    A

    A

    Cx

    C

    x

    =

    =

    =

    Question: What is the conversion of A forthe ideal BR for the residence time of 3


Recommended