artigo catarina ferro - ulisboa...! 1!!! development!of!a!kinetic!model!for!catalytic!reforming!...

9
1 Development of a Kinetic Model for Catalytic Reforming Catarina Alexandra Semedo Barros Ferro Abstract The Catalytic Reforming is the main process in a refinery to produce gasoline with a high octane number. It is equally an excellent producer of hydrogen reused for hydrotreatment processes and BTX production (benzene, toluene and xylene) widely used in petrochemical industry. The aims of this work were monitoring an industrial unit of catalytic reforming and study a kinetic model developed by CEPSA Company for a year of cycle. The monitoring was performed in order to understand the functioning of an industrial unit. The unit diagram was studied in detail, the process variables were collected and subsequently represented in order to understand their variations. The kinetic selected for this model was Henningsen and Bundgaard – Nielson in which the reaction rates are represented by simple first order and the reaction rate constants shown by Arrhenius Law. Posteriorly simulations were realized to comprehend if the chosen kinetic would be the most appropriated. It was also tested the calculus change of RON and it was verified that the initial correlation produces better results than other correlations. The model developed by CEPSA Company presents consistent data with the real ones, noting that the chosen kinetic for the model was a good possibility. Keywords: Catalytic Reforming, Monitoring, Kinetic Model, Simulations, Octane Number 1. Introduction The first Catalytic Reforming industrial unit was developed in 1940 by UOP and is worldwide recognized as Platforming process. The goal of this process is to convert naphtha with low octane number into high octane number gasoline called reformate as a blending component of motor fuels. It is also a primary source of hydrogen and aromatics used in the petrochemical industry. The octane number represents the ability of a gasoline to resist knocking during combustion of the airgasoline mixture in the engine. This number is defined as a volume percentage of isooctane in blending with nheptane that equals the knocking performance of the gasoline. By definition the octane number of nheptane is zero and the octane number of isooctane is 100. Normally there are two ways to measure octane number, the research octane number (RON) and the motor octane number (MON).

Upload: others

Post on 28-Jan-2021

3 views

Category:

Documents


0 download

TRANSCRIPT

  •   1  

     

     

    Development  of  a  Kinetic  Model  for  Catalytic  Reforming  

    Catarina  Alexandra  Semedo  Barros  Ferro

    Abstract  

    The   Catalytic   Reforming   is   the  main   process   in   a   refinery   to   produce   gasoline  with   a   high   octane   number.   It   is  

    equally   an   excellent   producer   of   hydrogen   reused   for   hydrotreatment   processes   and   BTX   production   (benzene,  

    toluene  and  xylene)  widely  used  in  petrochemical  industry.    

    The   aims   of   this   work   were   monitoring   an   industrial   unit   of   catalytic   reforming   and   study   a   kinetic   model  

    developed  by  CEPSA  Company  for  a  year  of  cycle.    

    The  monitoring  was  performed  in  order  to  understand  the  functioning  of  an  industrial  unit.  The  unit  diagram  was  

    studied   in   detail,   the   process   variables  were   collected   and   subsequently   represented   in   order   to   understand   their  

    variations.  

    The   kinetic   selected   for   this   model   was   Henningsen   and   Bundgaard   –   Nielson   in   which   the   reaction   rates   are  

    represented  by  simple  first  order  and  the  reaction  rate  constants  shown  by  Arrhenius  Law.    

    Posteriorly  simulations  were  realized  to  comprehend  if  the  chosen  kinetic  would  be  the  most  appropriated.  It  was  

    also   tested   the   calculus   change   of   RON  and   it  was   verified   that   the   initial   correlation   produces   better   results   than  

    other  correlations.  

    The  model   developed   by   CEPSA   Company   presents   consistent   data  with   the   real   ones,   noting   that   the   chosen  

    kinetic  for  the  model  was  a  good  possibility.    

     

    Keywords:  Catalytic  Reforming,  Monitoring,  Kinetic  Model,  Simulations,  Octane  Number  

     

    1. Introduction  The   first   Catalytic   Reforming   industrial   unit   was  

    developed   in   1940   by   UOP   and   is   worldwide  

    recognized  as  Platforming  process.    

    The  goal  of  this  process  is  to  convert  naphtha  with  

    low  octane  number  into  high  octane  number  gasoline  

    called   reformate   as   a   blending   component   of   motor  

    fuels.   It   is   also   a   primary   source   of   hydrogen   and  

    aromatics  used  in  the  petrochemical  industry.    

     

    The   octane   number   represents   the   ability   of   a  

    gasoline   to   resist   knocking   during   combustion   of   the  

    air-‐gasoline   mixture   in   the   engine.   This   number   is  

    defined   as   a   volume   percentage   of   isooctane   in  

    blending   with   n-‐heptane   that   equals   the   knocking  

    performance  of  the  gasoline.  By  definition  the  octane  

    number  of  n-‐heptane   is   zero  and   the  octane  number  

    of  isooctane  is  100.    

    Normally   there   are   two   ways   to   measure   octane  

    number,   the   research   octane   number   (RON)   and   the  

    motor  octane  number  (MON).  

  •   2  

    2.  Catalytic  Reforming  Process  A  typical  feed  to  Catalytic  Reforming  is  a  mixture  of  

    medium   and   heavy   straight   run   naphtha   obtained  

    directly   from   the   atmospheric   crude   oil   distillation  

    column.   This   naphtha   normally   contains   40-‐70   wt.%  

    paraffins,   20-‐50   wt.%   naphthenes,   5-‐20   wt.%  

    aromatics  and  0-‐2  wt.%  olefins.  [1]  

    The   distribution   of   paraffins,   olefins,   naphthenes  

    and   aromatics   in   the   feed   to   catalytic   reforming  

    determines   the   richness   of   the   feedstock,   which   is  

    normally   rated   by   its   naphthenes   +   aromatics   or  

    naphthenes  +  2  aromatics  value.  A  high  concentration  

    of   aromatics  means   that   the   octane   number   is   quite  

    high.  The  naphthenes  are   transformed   into  aromatics  

    with   a   high   selectivity   and   a   high   octane   number   is  

    easily  achieved.  However  a  paraffinic   feedstock  has  a  

    low  octane  number  and  it  is  more  difficult  convert  into  

    naphthenes.  

    Presently  the  standard  operating  conditions  in  this  

    process   are   elevated   temperature   (450-‐520ºC)   and  

    moderate  pressure  (4-‐30  bar).  [1]  

    A   large   number   of   reactions   occur   in   catalytic  

    reforming   such  as  dehydrogenation  of  naphthenes   to  

    aromatics,   isomerization  of  paraffins  and  naphthenes,  

    dehydrocyclization   of   paraffins   and   hydrocraking   of  

    paraffins  and  naphthenes  to  lower  hydrocarbons.  

    All   reactions   are   desirable   except   hydrocraking,  

    which   occurs   at   high   temperature   and   consumed   a  

    high  amount  of  hydrogen.  

    2.1. Process  Description  

    An   usual   Catalytic   Reforming   unit   is   presented   in  

    Figure   1.   The   feed   is   initial   mixed   with   recycled  

    hydrogen   gas,   and   then   the   mixture   passes   through  

    the   effluent-‐to-‐feed   heat   exchanger.   The   charge   is  

    completely   vaporized   and   it   is   transported   to   the  

    reactor  section.  Effluent  from  the  last  reactor  is  cooled  

    by   the   effluent-‐to-‐feed   heat   exchanger   for  maximum  

    heat  recovery.    

    After   the   effluent   is   charged   to   the   separation  

    section,   where   the   liquid   and   gas   products   are  

    separated.    

    A   fraction   of   the   gas   from   the   separator   is  

    compressed  and  recycled  back  to  the  reactor  section.  

    The   separator   liquid   is   pumped   to   the   reformate  

    stabilizer  for  separate  the  desired  product,  reformate,  

    from  the  rest  of  hydrogen  and  light  hydrocarbons.    

     Figure  1  –  Catalytic  Reforming  flow  diagram  [2]  

    These  processes  are  commonly  classified  according  

    to  the  frequency  and  mode  of  catalyst  is  regenerated,  

    into   semiregenerative,   cyclic   regeneration   and  

    continuous   regeneration.   The   main   difference  

    between   the   three   types   of   processes   is   the   need   of  

    shutdown  of   the  unit   for   catalyst   regeneration   in   the  

    case   of   semiregenerative   process,   the   use   of   an  

    additional   swing   reactor   for   catalyst   regeneration   for  

    the   cyclic   process   and   catalyst   replacement   during  

    normal   operation   for   the   continuous   regeneration  

    type.  

    2.2. Catalyst  

    Catalytic  reforming  reactions  are  conducted  in  the  

    presence  of  heterogeneous  and  bifunctional  catalysts.  

    The  double  function  is  provided  by  the  acid  sites  of  the  

    support,  usually  alumina  (Al2O3),  and  the  metallic  sites,  

    platinum  (Pt)  dispersed  on  the  support.    

  •   3  

    The   addition   of   components   to   the   acid   function,  

    such  as  chloride  or  fluoride,  changes  the  strength  and  

    amount  of  support  acid  sites.  

    A   simplified   schematic   diagram   of   the   alumina  

    functionality  is  given  in  Figure  2.  

     Figure  2  -‐  Alumina  schematic  [3]  

    The  acid  function  catalyses  the  C-‐C  bond  reactions  

    such   as   isomerization,   dehydrocyclization   and  

    hydrocraking.   For   other   hand   the   dehydrogenation  

    and  hydrogenolysis  reactions,  C-‐H  bond,  are  catalysed  

    by  a  metallic  function.    

    Currently   catalysts   are   bimetallic   or  multimetallic,  

    the   platinum   has   remained   the   key   component   and  

    the  second  element  is  Re,  Sn,  Ge  or  Ir  which  interacts  

    with   platinum   and   offers   a   better   selectivity   and  

    stability.  

    2.3.  Kinetics  Model  

    Naphtha  is  a  very  complex  mixture  of  hydrocarbon  

    and   there  are  more   than  300   components  present   in  

    this  mixture.  [4]  Different  reactions  occur  through  the  

    process   such   as   dehydrogenation,   hydrocraking,  

    isomerization  and  dehydrocyclization.  

    Due   to   the   large   number   of   components   involved   in  

    the   reactions,   the   development   of   kinetic   models  

    becomes   a   very   complex   process.   To   simplify   this  

    problem  were   used   groups   of   compounds,   known   as  

    lumps,  which  organise   the   species  as  a   single  entities  

    when  they  present  affinity  (chemical,  structural,  etc.).  

    The   first   effort   to   model   a   reforming   system   has  

    been   made   by   Smith   in   1959   and   is   described   by  

    Figure   3.   His   model   contained   three   classes   of  

    hydrocarbons:   paraffins,   naphthenes   and   aromatics.  

    No  distinction  was  made  on  the  basis  of  the  number  of  

    carbons   atoms  within   each   class.   Hydrogen   and   light  

    gases  were  also  taken  into  account.  

    This   model   involves   five   pseudocomponents:  

    paraffins,   naphthenes,   aromatics,   light   gases   and  

    hydrogen  and  four  reactions.  [5]  

     

    Figure  3  -‐  Reaction  schemes  of  Smith  model. [4]  

    Krane  et  al.  also   in  1959  proposed  another  model  

    for   catalytic   reforming   reactions.   In   this  model   there  

    are   20   pseudocomponents,   containing   hydrocarbons  

    from   C6   to   C10.   It   is   also   recognized   the   difference  

    between   paraffins,   naphthenes   and   aromatics   within  

    each   carbon   number   group.   All   reactions   are  

    represented  by  a  pseudo-‐first  order  rate  equation  with  

    respect  to  hydrocarbons  concentration.    

    The   model   chosen   by   CEPSA   Company   was  

    developed   in   1970   by   Henningsen   e   Bundgaard   –  

    Nielson   (Figure   4)   and   it   was   considered   an  

    improvement  to  Krane’s  model.    

     Figure  4  -‐  Reaction  schemes  of  Henningsen  and  Bundgaard  –  

    Nielson  model. [6]  

    This   model   takes   into   account   the   differences   in  

    the  behaviour  of  cycloalkanes  with  five  and  six  carbon  

    atoms  in  the  ring.  [7]    

    The   reaction   rates   are   normally   represented   by  

    simple   first  order  with   respect   to  partial   pressures  of  

    hydrocarbons   and   the   pressure   drop   through   the  

    reactores  is  despised.  [7]  The  reaction  rates  constants  

    are   expressed   in   the   form   of   an   Arrhenius   law   to  

  •   4  

    account  for  the  influence  of  temperature  and  catalyst  

    deactivation  were  also  included  in  the  model.  [5]  

    A   heat   balance   was   added   into   the   system   of  

    equations.   This   was   a   considerable   improvement   on  

    the   previous   models   that   treated   catalytic   reforming  

    as  an  isothermal  system.  [6]  

    The   differential   equations   that   describe   the  

    reaction  rates  are  equations  (1)  to  (6).  [6]  [7]    

    𝑑C𝑑𝜏

    = 𝑘!(𝑃!" + 𝑃!")   (1)  

    𝑑𝑁𝑃𝑑𝜏

    = −(𝑘! + 𝑘! + 𝑘! + 𝑘!)𝑃!" + 𝑘!𝑃!"#

    + 𝑘!𝑃!"# + 𝑘!𝑃!"  (2)  

    𝑑𝐼𝑃𝑑𝜏

    = −(𝑘! + 𝑘! + 𝑘! + 𝑘!")𝑃!" + 𝑘!𝑃!"#

    + 𝑘!𝑃!" + 𝑘!!𝑃!"#  (3)  

    𝑑𝐴𝐶𝐻𝑑𝜏

    = −(𝑘! + 𝑘! + 𝑘!" + 𝑘!")𝑃!"# + 𝑘!𝑃!"

    + 𝑘!𝑃!" + 𝑘!"𝑃!"#  (4)  

    𝑑𝐴𝐶𝑃𝑑𝜏

    = −(𝑘! + 𝑘!! + 𝑘!")𝑃!"# + 𝑘!𝑃!"

    + 𝑘!"𝑃!" + 𝑘!"𝑃!"#  (5)  

    𝑑𝐴𝑅𝑑𝜏

    = 𝑘!"𝑃!"#   (6)  

    The   differential   equation   that   describes   the   heat  

    balance  is  equation  (7).  [6]  [7]  

     !"!"= − !

    !! !!!(𝑘!𝑃!"∆𝐻!"→! +

    𝑘!𝑃!"∆𝐻!"→! + 𝑘!𝑃!"#∆𝐻!"#→!" + 𝑘!𝑃!"∆𝐻!"→!"# +

    𝑘!𝑃!"#∆𝐻!"#→!" + 𝑘!𝑃!"∆𝐻!"→!"# + 𝑘!𝑃!"∆𝐻!"→!"# +

    𝑘!𝑃!"#∆𝐻!"#→!" + 𝑘!𝑃!"∆𝐻!"→!" + 𝑘!𝑃!"∆𝐻!"→!" +

    𝑘!"𝑃!"∆𝐻!"→!"# + 𝑘!!𝑃!"!∆𝐻!"#→!" +

    𝑘!"𝑃!"#∆𝐻!"#→!"# + 𝑘!"𝑃!"#∆𝐻!"#→!"# +

    𝑘!"𝑃!"#∆𝐻!"#→!")  

       3. Analysis  of  Kinetic  Model  

    For   a   better   understanding   of   the   model  

    developed   by   CEPSA   Company   was   made   a   detailed  

    study  and   in  Figure  5   is   shown  a   flowchart   to  explain  

    the  functioning  of  the  simulator.  

    The   reactors   used   in   this   type   of   processes   are  

    heterogeneous   because   there   is   a   presence   of   two  

    phases,   one   solid   and   other   vapour.   Reactors   are  

    tubular  fixed-‐bed  with  axial  or  radial  flow  and  work  in  

    the  adiabatic  regime.  

     

       

    Figure  5  –  Model  calculation  flowchart  

     

    Since   the   differential   equations   described   by  

    Henningsen   and   Bundgaard   –   Nielson   kinetic   are  

    complex   equations   is   necessary   use   a   numerical  

    method  to  solve  these  equations.  The  method  chosen  

    was   the   Euler  method   described   by   the   equation   (8)  

    with  unit  step  size.  

    𝑦 𝑡 + ∆𝑡 = 𝑦 𝑡 +𝑑𝑦𝑑𝑡  ∆𝑡   (8)  

     

    (7)  

  •   5  

    4. Experimental  section  In  this  chapter  is  described  whole  the  work  process  

    to  obtain  the  results.  

    4.1. Monitoring  of  Industrial  Unit  

    The   data   used   in   this   study   were   obtained   from  

    Excel   Tool,   called  Pl@nt@.  This   tool   gets   all   the   tags  

    related   to   the   industrial   unit   as   well   as   all   the  

    laboratory  analyses.  

    Figure   6   shows   the   process   diagram   of   Catalytic  

    Reforming  unit  designed  from  the  available  plans.  

     

     Figure  6  –  Process  Diagram  of  Catalytic  Reforming  unit.  

     

    Initially   data   were   extracted   for   six   months   of  

    operation   just   to   get   an   idea   of   how   the   refinery  

    operated  and  how  the  process  variables  were  changed  

    through  the  working  days.    

    After   this   first   approach   it   was   verified   that   for   a  

    better  utilization  of  the  simulator  it  would  be  better  to  

    use  more   data   to   have   a   greater   range   of   values   for  

    comparison.  To  accomplish  this  goal  it  were  extracted  

    further  data,  in  this  case  18  months  of  operation.    

     

     Figure  7  –  Recycle  Gas  Pressure  for  18  months  of  operation  

    In  Figure  7   it   is  obvious   that   this  variable  as  other  

    shows  discrepant   values   for   a   few  days   of   operation.  

    This   fact   is   justified  with  stops   in   the  unit   for  catalyst  

    regeneration  or  operational  problems.    

    Due   to   the   presence   of   these   shutdowns   and  

    consequently   missing   of   laboratory   analysis   data,   it  

    was   only   considered   the   days   of   operation  

    represented  by  the  circle.    

    4.2. Real  Data  Simulation  of  Industrial  Unit  

    Before   using   the   simulator   it   was   important   to  

    conduct  a  study  of  the  optimal  number  of  iterations  in  

    order   to   minimize   adjustment   function   selected.   For  

    this   study   several   iterations   were   performed   as   is  

    noticeable  in  Figure  8.  

     

     Figure  8  –  Adjustment  Function  and  Iteration  Time  as  

    function  of  number  of  iterations  

     

    As   show   in   Figure   8   the   adjustment   function  

    decreases  steeply  until  3000  iteration,  and  this  value  is  

    relatively   lower.   It   is   also   perceptible   that   from  5000  

    0  

    5  

    10  

    15  

    20  

    25  

    30  

    35  

    0   50   100   150   200   250   300   350   400   450   500   550  Recycle  Gas  Pressure  (atm

    )  

    Days  of  OperaJon  

    0  

    20  

    40  

    60  

    80  

    100  

    120  

    140  

    160  

    180  

    200  

    500  

    1000  

    1500  

    2000  

    2500  

    3000  

    3500  

    0   1000   2000   3000   4000   5000   6000   7000   8000   9000   10000  

    IteraJo

    n  Time  

    Adjustmen

    t  Fun

    cJon

     

    Number  of  IteraJons  

    Adjustment  Funcgon   Iteragon  Time  (mim)  

  •   6  

    iterations   the   value   of   adjustment   function   does   not  

    change  too  much  and  since  the  simulation  time  begins  

    to   be   excessive,   the   best   choice   is   5000   iterations   to  

    have   one  minimization   of   adjustment   function   and   a  

    simulation  time  reasonable.    

    Moreover   to   prove   the   choice   of   5000   iterations  

    five   trials  were   conducted  where   it   was   verified   that  

    results  were  consistent.  

     

    5. Results    This  chapter  presents  some  results  of  the  study.  

    5.1. Simulation  of  the  real  data  

    The  results  of  this  subchapter  are  associated  to  an  

    interval   of   425   days   of   operation   and   in   all   figures  

    there  are  points  which  don´t  converge  represented  by  

    the  yellow  colour.  

     Figure  9  -‐  Comparison  between  Real  and  calculated  RON  

    The  expression  used  to  calculate  the  model  RON  is  

    indicated   by   the   equation   (9).   This   RON   depends  

    exclusively   on   the   molar   percentage   of   aromatics  

    present  in  reformate.    

    𝑅𝑂𝑁 = 𝑚  𝑥!" + 𝑏   (9)  

     

    In   Figure   9   is   visible   that   RON   calculated   by   the  

    simulator  follows  the  same  trend  as  the  real  RON,  but  

    their  values  are  always  higher.  Due  to  this   fact   it  was  

    tested  a  new  correlation  for  the  calculated  RON.  

    The   chosen   correlation   is   described   by   equation  

    (10),   and   it   takes   into   account   the   contributions   of  

    each  hydrocarbon  family  presents  in  reformate.  [8]  

     

    𝑅𝑂𝑁 = 𝑥!" 𝑅𝑂𝑁 !" + 𝑥!" 𝑅𝑂𝑁 !" +

    𝑥! 𝑅𝑂𝑁 ! + 𝑥!" 𝑅𝑂𝑁 !"    

    (10)  

     

    Where   xNP,   xIP,xN,   xAR   are   molar   fraction   of   n-‐

    paraffins,   iso-‐paraffins,   naphthenes   and   aromatics  

    groups,  respectively.  

    For   each   hydrocarbon   family   the   value   of   RON   is  

    described  by  equation  (11).  

    𝑅𝑂𝑁 = 𝑎 + 𝑏 𝑇 + 𝑐 𝑇 ! + 𝑑 𝑇 ! + 𝑒 𝑇 !   (11)  

     

    Where  T=Tb/100  in  which  Tb  is  the  normal  boiling  

    point  given  by  an  average  value  between  initial  boiling  

    point   and   final   boiling   point   in   the   ASTM   D86.  

    Coefficients  a-‐e  are  given  in  table  1.    

     

    Table  1  –  Coefficients  for  equation  (11)  for  Estimation  of  RON  

     

    In  Figure  10  is  visible  that  the  correlation  described  

    by  equation  (10)  despite  follows  the  same  tendency  as  

    the   real   RON   is   not   appropriate   because   its   values  

    show  a  percentage  error   of   16  %  while   the   values  of  

    85  

    90  

    95  

    100  

    105  

    110  

    91   141   191   241   291   341   391   441   491   541  

    RON  

    Days  of  Operation  

    Real  RON   Model  RON  

    HC  Family   a   b   c   d   e  

    N-‐Paraffins   92,81   -‐70,97   -‐53   20   10  

    Iso-‐  Paraffins            2-‐

    Methylpentanes  

    95,93   -‐157,53   561   -‐600   200  

    3-‐  Methylpenta

    nes  92,07   57,63   -‐65   0   0  

    2,2-‐  Dimethylpen

    tanes  109,38   -‐38,83   -‐26   0   0  

    2,3-‐  Dimethylpen

    tanes  97,65   -‐20,8   58   -‐200   100  

    Naphthenes   -‐77,53   471,59   -‐418   100   0  

    Aromatics   145,66   -‐54,33   16,27   0   0  

  •   7  

    RON   calculated   by   equation   (9)   show   an   error   of   5%  

    compared  to  the  real  values.  

     Figure  10  -‐  Comparison  between  Real  RON  and  calculated  

    RON  by  equation  (9)  and  (10)  

    Due   to   these   results,   the   equation   (10)   was  

    modified   to   take   into   account   only   the   compounds  

    with   a   high   value   of   octane   number,   such   like   iso-‐

    paraffins   and   aromatics.   This   new   correlation   is  

    described  by  equation  (12)  and  the  calculation  method  

    is  the  same  described  previously.  

    𝑅𝑂𝑁 = 𝑥!" 𝑅𝑂𝑁 !" + 𝑥!" 𝑅𝑂𝑁 !"   (12)  

    Likewise   equation   (12)   is   not   appropriated   for  

    adjustment   the   real   data.   Therefore   the   correlation  

    described  by  equation  (9)  remains  the  best  option,  as  

    confirmed  in  Figure  11.  

     

     Figure  11  -‐  Comparison  between  Real  RON  and  calculated  

    RON  by  equation  (9),  (10)  and  (12)  

    From   Figure   12   it   is   observed   that   production   of  

    aromatics  present  consistent  values  with  experimental  

    results  with  percentage  error  of  1%  

     

     Figure  12  –  Comparison  between  the  percentage  real  and  

    calculated  of  aromatics  in  reformate  

    The   process   variable   most   frequently   used   by  

    refiners   to   control   reformer   operation   is   weighted  

    average  inlet  temperature  (WAIT).  This  variable   is  the  

    sum   of   the   inlet   temperature   to   each   reactor  

    multiplied   by   the   weight   percent   of   total   catalyst   in  

    each  reactor.    

     Figure  13  –  WAIT  

    It   is  observed   that   the  value  of  WAIT   in   Figure  13  

    increases   due   to   the   loss   activity   and   stability   of  

    catalyst.   This   loss   of   activity   happens   because   coke  

    deposition  in  both  acid  and  metal  sites.    

    This   increase   in   temperature   happens   to   allow  

    octane   number   keeps   constant   and   produces   the  

    desired  reformate.  

     

    75  

    80  

    85  

    90  

    95  

    100  

    105  

    110  

    186   196   206   216   226   236  

    RON  

    Days  of  OperaJon  

    Real  RON   RON  Equagon  9   RON  Equagon  10  

    75  

    80  

    85  

    90  

    95  

    100  

    105  

    110  

    186   196   206   216   226   236  

    RON  

    Days  of  OperaJon  

    Real  RON   RON  Equagon  9  

    RON  Equagon  10   RON  Equagon  12  

    50  

    55  

    60  

    65  

    70  

    75  

    80  

    85  

    90  

    91   141   191   241   291   341   391   441   491  

    Mol%  of  A

    romaJ

    cs  

    Days  of  OperaJon  

    Mol  %  Real  AR  in  Reformate  

    Mol  %    Model  AR  in  Reformate  

    502  504  506  508  510  512  514  516  518  520  522  

    91   141   191   241   291   341   391   441   491   541  

    WAIT  (ºC)  

    Days  of  Operation  

  •   8  

     Figure  14  –  Evolution  of  compounds  along  reforming  

    catalytic  

    Figure   14   shows   the   evolution   of   compounds  

    present   in   naphtha   along   the   reaction   system.   It   is  

    observed  that  molar  percentage  of  aromatics  through  

    the  reaction  system  increases  while  molar  percentage  

    of   naphthenes   and   paraffins   decreases,   as   expected,  

    because   aromatics   formation   by   dehydrogenation  

    reaction   requires   a   lower   amount  of  naphthenes   and  

    consequently   the   formation  of  naphthenes   leads   to  a  

    reduction  of  the  amount  of  paraffins.    

    It   is   also   noted   that   the   largest   decrease   in   the  

    naphthenes   percentage   occurs   in   the   first   reactor,  

    because  hydrogenation  reaction  is  the  fastest  and  the  

    most  endothermic  reaction  happens  mainly  in  the  first  

    reactor.  

    Furthermore   it   is   observed   that   the   molar  

    percentage   of   cracked   products   undergoes   a  

    significant   increase   in   the   second   and   third   reactor,  

    because   these   reactions   are   considered   the   slowest  

    reactions.    

     

    6. Conclusions  The   kinetic   model   developed   by   Cepsa   Company  

    was   applied   to   real   data   of   an   industrial   unit   of  

    reforming  catalytic.    

    In   relation   to   the   monitoring   carried   out   it   was  

    observed  the  behaviour  of  an  industrial  unit  compared  

    to  a  pilot  unit  and  how  to  change  any  variable  affects  

    the  whole  production  of  reformate.    

    A   study   of   the   optimal   number   of   iterations   was  

    done  to  ensure  that  the  adjustment  function  would  be  

    minimized   within   a   reasonable   simulation   time.   We  

    conclude  the  more  appropriate  number  of  iterations  is  

    5000.  

    The   correlation   used   initially   by   kinetic  model   for  

    estimate   of   RON   is   the   best   option   despite   having  

    higher  values  than  the  real.    

    The   real   values   were   compared   with   those  

    calculated  by  the  model.  The  values  for  the  aromatics,  

    n-‐paraffins   and   iso-‐paraffins   are   consistent   with  

    percentage  error  of  2%.  

    However   in   relation   to   naphthenes   there   is   a  

    greater  discrepancy   in   the  values.  This  difference  can  

    be   explained   by   the   fact   that   their   values   are   lower  

    and   the   possible   integration   errors   made   in  

    chromatography  equipment  (GC-‐Reformulizer).  

    On   the   other   hand   one   of   the   assumption   of  

    chosen  kinetic  indicates  there  is  a  distinction  between  

    the   behaviour   of   alkyl-‐cyclopentanes   and   alkyl-‐

    cyclohexanes.  However   this   assumption   in   the  model  

    was   not   considered,   feedstock   presents   only   alkyl-‐

    cyclohexanes   and   reactions  where  ACP   appears  were  

    not   taken   into   consideration.   This   fact   can   also   be   a  

    possible  reason  for  the  disparity  of  values.    

    The  difference  between  model  yield  and  real  one  is  

    5%  and  it  is  considered  acceptable.    

    In   a   following   work   the   model   used   should   be  

    complemented   using   more   reactions   to   describe   the  

    process  as  well  as   improve  the  way  of  calculating  the  

    RON.  

     

     

     

     

    0  5  

    10  15  20  25  30  35  40  45  50  

    Mol%  of  C

    ompo

    unds  

    EvoluJon  over  of  reacJon  system  

    mol%  nP   mol%  IP   mol%  N  

    mol%  AR   mol%  C  

  •   9  

    7. Nomenclature    ∆𝐻!→!  -‐  Heat  of  reaction    

    𝐶!  -‐  Heat  capacity  

    𝑘!  –  Reaction  rate  constant  

    𝑃!  -‐    Partial  pressure  of  the  component  i  

    ACH  –  Alkyl-‐cyclohexanes  

    ACP  –  Alkyl  –  cyclopentanes  

    AR  –  Aromatics  

    C  –  Hydrocraking  products  

    HC  -‐  Hydrocarbon  

    IP  –  iso-‐Paraffins    

    𝑛  -‐  Hydrogen/Hydrocarbon  ratio  

    N-‐Naphthenes  

    NP-‐  normal-‐Paraffins  

    WAIT  -‐  Weighted  Average  Inlet  Temperature  

    τ  –  Reaction  time  

     

    8. Bibliography  

    [1]  Antos,  George  J;  Aitani,  Abdullah  M;,  Catalytic  

    Naphtha  Reforming,  2nd  ed.  New  York:  Marcel  

    Dekker,  Inc.,  2004,  Revised  and  Expanded.  

    [2]  Lapinski,  Mark;  Baird,  Lance;  James,  Robert;,  

    "Chapter  4.1  UOP  Platforming  Process,"  in  

    Handbook  of  Petroleum  Refining  Processes,  3rd  

    ed.:  McGraw-‐Hill,  pp.  4.3-‐4.31.  

    [3]  Mark  Moser  and  Peter  R.  Pujadó,  "Chapter  5  -‐  

    Catalytic  Reforming,"  in  Handbook  of  Petroleum  

    Processing.  The  Netherlands:  Springer,  2006,  pp.  

    217-‐237.  

    [4]  Rahimpour,  Mohammad  Reza;  Jafari,  Mitra;  

    Iranshahi,  Davood;,  "Progress  in  catalytic  naphtha  

    reforming  process:  A  review,"  Elsevier,  no.  109,  pp.  

    79-‐93,  2013,  www.elsevier.com/locate/apenergy.  

    [5]  Jorge  Ancheyta,  "Chapter  4  -‐  Modeling  of  Catalytic  

    Reforming,"  in  Modeling  and  Simulation  of  

    Catalytic  Reactors  for  Petroleum  Refining.  New  

    Jersey:  John  Wiley  &  Sons,  Inc,  2011.  

    [6]  Raseev,  Serge;,  "Chapter  13  -‐  Catalytic  Reforming,"  

    in  Thermal  and  Catalytic  Processes  in  Petroleum  

    Refining.  New  York:  Marcel  Dekker,  Inc.,  2003,  pp.  

    771-‐786.  

    [7]  Henningsen,  J;  Bundgaard-‐Nielson,  M;,  "Catalytic  

    Reforming,"  Bristish  Chemical  Engeneering,  vol.  15,  

    no.  11,  pp.  1433-‐1436,  November  1970.  

    [8]  M.  R.  Riazi,  Characterization  and  Properties  of  

    Petroleum  Fractions,  1st  ed.,  AMERICAN  SOCIETY  

    FOR  TESTING  AND  MATERIALS,  Ed.  Philadelphia,  

    PA,  2005.