OIL

 &  GAS  PROCESS     WATER  INJECTION  

ENGINEERING   SYSTEMS  
 

 
 
 
Chapter  13  
 
Water  Injection  Treatment    
 
 
Having  worked  through  this  chapter  the  student  will  be  able  to:  
• Identify  the  sources  of  water  to  be  treated  for  oil  contamination    
• Discuss  the  role  of  microbiologically  induced  corrosion  and  its  impact  on  water  
treatment    
• Discuss  the  factors  affecting  and  techniques  to  mitigate  MIC    
• Use  solubility  data    
• Identify  the  issues  surrounding  dissolved  gases  in  water.    
• Discuss  the  role  of  ionisation,  pH  and  dosage  levels  in  chlorine  treatments.    
• Discuss  treatment  options  for  MIC,  dissolved  gases  and  suspended  solid.    
• Discuss   the   mechanisms   and   operational   aspects   of   solids   filtration   from  
water.    
• Describe  the  systems  used  to  add  and  remove  dissolved  gases    
• Discuss   the   issues   in   selecting   and   operating   packed   columns   for   de-­‐
oxygenation.    
• Estimate  the  size  of  packed  columns  for  de-­‐oxygenation    
• Identify   typical   flowschemes   for   treating   surface   and   produced   water   for   re-­‐  
injection    
 
 

© School of Engineering and Physical Science Page 1

OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 

Content
1   INTRODUCTION   3  
5.2   Removing  Suspended  Solids   19  
5.2.1   Theory  of  Filtration   21  
2   QUALITY   OF   WATER   FOR   5.2.2   Filter  Types   23  
FLOODING   4   5.2.3   Sand  and  Granular  Filters   24  
5.2.4   Filtration  Rate   26  
2.1   Suspended  Solids   4   5.2.5   Backwash  Issues   26  
5.2.6   Back-­‐flush  frequency   29  
2.2   Micro-­‐organisms  Issues   4   5.2.7   Design  Issues   29  

2.3   Soluble  Gases   5   5.3   Aeration   31  

2.4   Note  on  Units   5   5.4   Deaeration  by  Packed  Columns   32  

5.4.1   Design  Methods   34  
5.4.2   Maldistribution   53  
3   MICROBIOLOGICALLY   INDUCED  
CORROSION   6   5.5   Deareation  by  Chemical  Methods   60  

3.1   Sulphate   Reducing   Bacteria   (SRB)   –   5.6   Enhanced  Extraction  Systems   60  

Desulfovibrio  desulfuricans  (Anaerobic)   7  
6   PROCESSING  SCHEMES   62  
3.2   Iron  Bacteria  –  Gallionalla  &  Crenothrix  
(Aerobic)   8  
6.1   Generalised  Water  Treatment  System  63  
3.3   Clostridium  (Anaerobic)   8  
6.2   Produced  Water  Reinjection  on  the  ULA  
63  
(13)
Field  
3.4   Slime  Formers   8  
6.3   Closed   Water   Treatment   System   in  
3.5   Chemical   Treatments   for  
66  
(6)
Kanas  
Microorganisims   8  

3.6   Treatment  with  Biocides   9   7   REFERENCES   67  

4   DISSOLVED  GASSES   11  

4.1   Influence  of  Dissolved  Oxygen   11  

4.2   Solubility  of  Gases  in  Liquids   12  

4.3   Influence  of  CO2   13  

4.4   Influence  of  Hydrogen  Sulphide   13  

5   WATER  TREATMENTS   15  

5.1   Microbial  Treatment   15  

5.1.1   Chlorination   15  
5.1.2   Chlorine  Dosage  Amounts   17  
5.1.3   Choice  of  Chlorine  Agents   18  
5.1.4   Ozonation   18  

Page 2
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 

1 INTRODUCTION
Injection  of  water  into  the  oil  reservoir  is  perhaps  one  of  the  more  important  factors  
affected  the  longevity  of  an  oil  field,  maintaining  the  reservoir  pressure  and  ensuring  
the   best   possible   prospects   for   exploiting   the   hydrocarbon   reserves.   The   term  
“Enhanced   Oil   Recovery”   referred   to   technologies   to   maximise   production   from  
ageing   wells;   techniques   such   as   in-­‐situ   combustion,   surfactant   flooding   have   been  
research   for   many   years.   Today,   these   have   lost   their   favour   with   the   advent   of  
horizontal   and   directional   drilling   to   maximise   coverage   of   a   well.   However   water  
injection   remains   tried   and   tested.   Increasing   environmental   pressures   ensure   that  
how  water  is  processed  remains  a  significant  issue  for  oilfield  producers.    
 
The  processing  systems  used  to  treat  water  for  injection  can  be  thought  of  as  being  
divided  into  two  categories;  those  which  treat  produced  water  and  those  which  treat  
water   drawn   from   the   surface   environment   e.g.   sea   water   or   fresh   water.   It   is   useful  
therefore  to  define  four  types  of  water  and  their  disposal  routes:  
• Injection  water  –  water  taken  from  a  separate  source  to  be  injected  for  the  
purposes  of  oil  recovery  
• Injected   Produced   water   which   is   essentially   “bad   water”   injected   for   the  
primary  purpose  of  increasing  oil  recovery.  It  also  serves  as  a  route  to  dispose  
of  the  hydrocarbon  and  chemically  contaminated  water.  
• Disposed   Produced   Water   or   water   heavily   contaminated   by   chemical  
additives  which  for  whatever  reason  cannot  be  re-­‐injected  into  the  reservoir  
but  disposed  on  by  injection  into  its  own  disposal  well.  
• Surface   Disposed   Produced   Water   –   this   is   water   which   is   not   re-­‐injected   but  
disposed  of  into  the  environment.  
While   the   technology   employed   to   process   water   may   be   general,   the   details   of  
flowschemes   will   depend   on   the   chemical   make   up   of   the   water   and   the   quality  
requirements   of   the   water   when   it   reaches   its   disposal   point.   In   cases   where  
produced   water   is   insufficient   to   support   a   pressure   maintenance   campaign,   water  
should  be  drawn  from  sources  requiring  the  lowest  degree  of  processing.  
 
This   chapter   reviews   the   issues   with   respect   to   water   quality   for   injection;   namely  
presence   of   solids   via   natural   or   generated   by   corrosion,   disinfection   etc,   and  
discusses  aspects  of  the  technology  not  already  covered  elsewhere.  Technology  for  
treating  water  is  not  restricted  to  the  oil  industry  and  many  of  the  techniques  arise  
from  developments  in  the  civil  supply  of  drinking  water.  
   

© School of Engineering and Physical Science Page 3

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 

2 Quality of Water for Flooding

Water   quality   in   this   respect   refers   to   the   suitability   of   water   for   injection   into   an  
underground  reservoir  rather  than  quality  in  terms  of  consumption;  it  is  particularly  
important   for   injected   water   to   be   chemically   compatible   with   any   formation   or  
other  water  in  the  reservoir  as  chemical  reactions  may  result  in  the  precipitation  of  
solids   that   damage   the   reservoir.   The   presence   of   suspended   solids   is   the   major  
cause   of   injection   failure.   Suspended   solids   arise   two   principle   routes;   those  
associated   with   pick   up   and   carry   through   of   sand,   grit   etc,   and   those   created   by  
chemical   and   microbial   action.   The   degree   of   treatment   required   to   counter   these  
effects  depends  on  the  reservoir  porosity  and  pore  size;  low  porosity,  small  pore  side  
requires  injection  water  of  a  higher  quality  than  reservoirs  with  high  porosity,  large  
pore  size.  
 
In   dealing   with   water   treatment,   it   is   common   practice   to   discuss   separately   the  
issues   surrounding   removal   of   suspended   solids   and   treatments   to   tackle   micro-­‐
organisms  and  chemicals  present  even  though  the  two  are  interlinked  in  the  method  
by  which  they  damage  the  reservoir  i.e.  The  issues  surrounding  these  two  principle  
issues  with  water  quality  for  injection  are  similar  to  those  for  treating  any  sources  of  
water  for  human  consumption.    

2.1 Suspended Solids

Water   from   surface,   sea   or   produced   water   can   contain   solids,   which   will   be   in  
suspension   due   to   their   small   size.   Suspended   solids   found   in   water   can   be   due   to  
grains  of  sand,  corrosion  products  such  as  iron  sulphide  or  iron  oxide,  free  sulphur.  
Bacteria  may  be  present  contributing  both  to  the  suspended  solids  count  as  well  as  
generating   corrosion   products.   If   these   materials   enter   the   injection   well,   they   can  
cause   plugging   or   result   in   a   need   to   increase   injection   pressure;   suspended   solids   in  
could  plug  the  pore  spaces  in  the  formation  or  build  up  an  impenetrable  layer  on  the  
face  of  the  reservoir  rock  impeding  further  injection.  This  has  two  impacts  depending  
on   the   water   injection   route;   for   pressure   maintenance   via   injection   into   the   oil  
reservoir  production  rates  are  likely  to  fall,  and  for  water  disposal  wells  the  ability  to  
inject   would   reduce.   Both   of   these   impacts   affect   the   overall   capacity   of   the  
separation  process  and  hence  results  in  reduced  production  of  stabilised  crude.    
 
The   presence   and   chemical   nature   of   suspended   solids   is   best   determined   by  
laboratory  filtration  of  an  extracted  sample  using  cellulose  membranes.  In  a  typical  
test,   the   rate   at   which   the   dirty   water   filters,   at   a   fixed   pressure   through   the  
membrane,   should   be   recorded   as   this   provides   some   measure   of   the   water’s  
quality.      

2.2 Micro-organisms Issues

Surface   water   extracted   from   natural   water   sources   e.g.   sea   or   rivers   will   probably  
contain   bacteria   which   can   cause   corrosion   of   the   processing   equipment;   injection  
pumps,  pipelines,  vessels  etc,  the  by-­‐product  of  this  corrosion  are  suspended  solids  
which  can  be  carried  into  the  reservoir.  Bacteria  have  two  effects;  one  related  to  the  
Page 4
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
chemical   action   producing   corrosive   environments   and   second   related   to   the  
deposition  of  bacteria  itself  within  the  rock  structure.    

2.3 Soluble Gases

The   issue   surrounding   soluble   gases   found   in   waters,   is   related   to   corrosion   and  
hence   the   formation   of   corrosive   products   manifesting   themselves   as   suspended  
solids.   Oxygen,   CO2   and   H2S   are   soluble   gases,   which   have   a   corrosive   impact   and  
hence  require  some  degree  of  treatment.  In  some  instances,  oxygen  wil  be  added  to  
water   to   enhance   the   quality.   Aeration   of   water,   a   common   practice   in   plants  
processing   water   for   drinking,   is   used   to   both   remove   some   volatile   organic  
compounds  and  compounds  which  interfere  in  further  processing  operations;  in  the  
case   of   drinking   water,   aeration   is   used   to   remove   H2S   prior   to   chlorination  
treatment  systems  used  to  control  micro-­‐organisms.    

2.4 Note on Units

Treating  water  is  not  restricted  to  one  branch  of  engineering  and  hence  there  are  a  
multitude  of  ways  to  express  composition  and  related  quantities;  ppm  and  mg/litre  
are  perhaps  the  more  important  units  of  measurements  :  
 
Parts  per  Million  (ppm)  
This  is  a  common  measurement  of  concentration  which  can  be  expressed  normally  
by  weight  but  can  be  given  on  a  volume  basis:  
1 mg of solute
1 ppm =  
1000 g of solution
Milligrams  per  litre  (mg/litre)  
This   is   the   normal   unit   of   concentration   which   often   gets   wrongly   confused   with  
ppm,   although   for   dilute   water   solutions   the   two   values   are   almost   numerically  
identical.   For   example,   let   “ms”   be   the   mass   in   grams   of   solute   dissolved   in   “m”   in  
grams   of   solution.   If   ρ   is   then   density   of   the   solution   in   g/litre,   then   the  
concentration  of  solute  is:  
1000 xms m C
C=  (mg/litre),  so  that   s = x10 −3  
m/ρ m ρ
ms ⎛ mg ⎞ ⎛ C ⎞ C
The   ppm   value   is   x106 ⎜⎜ ⎟⎟ ,   so   that   ppm = 10 6 ⎜⎜ ⎟⎟ = 1000 .   If   the  
m ⎝ 1000 g ⎠ ⎝ 1000 ρ ⎠ ρ
density  of  water  is  1000  g/litre,  then:  
C
ppm = ,  C  is  the  concentration  in  mg/litre  
sg
where   sg  is  the  specific  gravity  with  respect  to  water,  for  the  liquid  phase.  
   

© School of Engineering and Physical Science Page 5

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
3 Microbiologically Induced Corrosion
Microbiologically  Induced  Corrosion  (MIC)  is  a  recognised  issue  in  terms  of  corrosion;  
MIC   results   in   pitting,   crevice   corrosion,   selective   dealloying,   stress   corrosion  
cracking   and   under   deposit   corrosion.   Bacteria   present   in   either   surface   water   or  
other   waters   may   be   the   ultimate   cause   of   corrosion   products   plugging   lines   and  
pores  within  the  reservoir.  Elimination  of  the  bacteria  should  only  be  done  once  an  
economic  assessment  concludes  the  costs  of  treatment  are  lower  when  set  against  
the  cost  in  handling  resulting  corrosion  and  reservoir  plugging;  the  mere  presence  of  
bacteria   does   not   automatically   imply   operational   problems   although   prudence  
would   indicate   designs   should   provide   for   future   installation   of   necessary   counter  
measures.  
 
Bacteria   can   grow   in   many   environments   adapting   as   life   does   to   new   environments  
over   long   periods   of   time.   For   oilfield   operations,   the   more   important   bacteria   are  
classed   in   terms   of   their   requirements   for   oxygen;   the   obligate   aerobes   are   those  
while  require  oxygen,  obligate  anaerobes  do  not.  In  addition  to  the  oxygen  criteria,  
the  majority  of  bacteria  will  grow  better  in  warmer  weather  conditions  rather  than  
cooler  conditions.  
Microbial   life   can   only   be   sustained   provided   that   key   elements   are   present   at   the  
same  time  namely:  
• Energy  source:  usually  light  or  chemical    
• Carbon   sources:   CO2   or   organic   substances   that   can   be   utilised   with   the  
lowest  energy  required.  
The  range  of  conditions  under  which  micro-­‐organisms  exit  is  vast:  
• Temperature  from  –5°C  to  over  120°C  
• pH:  from  0  to  13  
• Pressure:  to  1000  bar  (barophilic  bacteria)  
• Salinity:  can  grow  in  ultrapure  water,  to  almost  saturated  water  
Chemical   biocides   are   the   major   route   to   combating   MIC   and   typical   figures   for  
bicodes  are  shown  in  Table  1.  
 
Biocide   Dosage  Level   Contact  Time  
(mg/liter)   (hrs)  
Chlorine   50  –  100   1  –  2  
Ozone   10-­‐50   <1  
Chlorine  dioxide   50-­‐100   1-­‐2  
Hydrogen  peroxide   10%  (v/v)   2-­‐3  
Iodine   100-­‐200   1-­‐2  
Quaternary  ammonium  cmpds.   300-­‐1000   2-­‐3  
Formaldehyde   1-­‐2%  (v/v)   2-­‐3  
Anionic  &  nonionic  surfactants   300-­‐500   3-­‐4  
Table  1:  Typical  biocide  treatment  and  loadings.  After  Mittelman  (4)  

Page 6
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

3.1 Sulphate Reducing Bacteria (SRB) – Desulfovibrio desulfuricans

(Anaerobic)
 
These   are   perhaps   the   most   damaging   of   all   bacteria   in   terms   of   corrosion   as   the  
mechanism   for   them   to   grow   is   via   reduction   of   the   sulphate   ion   by   hydrogen,  
usually  obtained  as  part  of  the  electrolysis  action  of  a  metal  ion  in  water:  For  iron,  
the  corrosion  mechanism  in  water  occurs  as  a  result  of  the  reactions:  
=
At  the  anode   4Fe ←⎯→ 4Fe 2+ + 8e1−  
At  the  cathode     8H1+ + 8e1− ←⎯→ 8H  
Anode  corrosion  product   ( )
3Fe 2+ + 6 OH1− ←⎯→ 3Fe(OH)2  
The  mechanism  of  iron  corrosion  as  described  is  accelerated  in  the  presence  of  SRB:  
=
Cathodic  depolarisation  by  SRB     SO 24− + 8H ←⎯→ S 2− + 4H2O  
=
Dissociation  of  water   8H2O ←⎯→ 8OH1− + 8H1+    
Reversible  reaction   2H1+ + S2 − ←⎯→ H2S  
=
Anode  corrosion  product   Fe 2 + + S2 − ←⎯→ FeS  
 
Figure  1  highlights  the  pathways  for  the  enhanced  corrosion  of  iron  by  SRB  
 

 
Figure  1  Chemical  pathway  for  SRB  action  on  iron  
 
From   the   chemical   reactions,   the   presence   of   SRB’s   is   commonly   associated   with   the  
formation  of  hydrogen  sulphide  or  black  ferrous  sulphide  if  the  iron  is  present.    
In  order  for  the  SRB  to  act  in  this  way,  the  following  conditions  must  apply:  
• Sulphates  in  the  water  –  these  could  be  natural  salts  or  deliberately  added  for  
another  purpose.  
• An  external  energy  source  –  free  electrons  must  be  present.  
• Temperature   should   normally   be   less   than   65°C(2),   although   some   sources  
claim   SRB’s   grow   best   between   25°C   to   35°C.   Recently,     thermophilic   SRB  

© School of Engineering and Physical Science Page 7

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
have   been   detected   in   North   Sea   produced   waters,   where   the   reservoir  
temperatures  vary  from  60°C  to  130°C.    

3.2 Iron Bacteria – Gallionalla & Crenothrix (Aerobic)

There  are  branches  of  microorganisms,  which  oxidize  iron  from  Fe2+  to  Fe3+.  The  Fe3+  
ion   formed   by   the   reaction   can   attract   chloride   ions   and   produce   ferric   chloride  
deposits,  which  can  in  turn  attack  austenitic  stainless  steel.  Depositions  on  the  metal  
surface   are   usually   brown   or   red   brown   mounds.   In   another   mode,   these   bacteria  
are  thought  to  assist  in  the  formation  of  ferric  hydroxide  which  consumes  dissolved  
oxygen   in   the   surrounding   water   eventually   leading   to   suitable   conditions   for  
anaerobic  bacteria  such  as  SRB  to  thrive.  

3.3 Clostridium (Anaerobic)

This   bacteria   is   known   from   experience   to   be   associated   with   corrosion   in   oilfield  
waters.   Unfortunately   this   species   forms   spores   that   have   shown   resistance   to  
biocides.    

3.4 Slime Formers

The   class   of   bacteria   belonging   to   several   genera;   Pseudomonas,   Bacillus   are   two  
examples  of  organisms  which  form  slime.  Slime  is  generally  a  dense  mass  of  material  
deposited   on   part   of   the   metal   surface   which   becomes   impervious   to   oxygen   and  
hence  form  anaerobic  cells  where  SRB  can  be  active.  Slime  deposition  is  more  likely  
in  cases  where  water  velocities  are  low.  

3.5 Chemical Treatments for Microorganisims

A  wide  range  of  chemical  treatments  exist  to  control  the  growth  of  microorganisms  
although  some  organisms  show  signs  of  adapting  to  new  environments  and  repeated  
application   of   the   same   chemical   may   become   ineffective.   Some   chemicals   will   kill  
the   organisim   (ie.   Bactericides)   whilst   others   will   interrupt   the   growth  
(bacteriostats).   The   term   biocides   and   biostats   strictly   refer   to   chemicals   which   kill  
and   retard   growth   on   other   forms   of   life;   in   this   chapter   the   term   biocide   will   be  
used   as   this   correctly   suggests   that   any   chemical   added   to   water   can   harm   the  
environment   if   the   water   reaches   surface   waters.   Inorganic   compounds   e.g   chlorine,  
chromates,   mercury,   silver   can   be   used   as   biocides   but   the   vast   majority   can   be  
classed  as  organic;  Table  2  shows  some  typical  biocides.  
 
 
 
 
 
 
 
 
 
 
 
 

Page 8
 
 OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 
 
 
 
 
Types   of   Examples   Compound  Formula   State   Concentration  
Chemicals   (ppm)  
Chromium   Sodium  Chromate   Na2CrO4   Solid   500  
Mercury   Mercury  Chloride   HgCl2   Solid   50  –  300  
Silver   Silver  Nitrate   AgNO3   Solid   0.05  
Amine   Primary   Amine   (R-­‐NH3)+(CH3CO-­‐O)-­‐   Solid   10  –  40  
Acetate  
Damine   Trimethylene  diamine   C12H25-­‐NH(CH2)3NH2   Liquid   5  –  25  
Quaternary   Alkyl   RN(CH3)3Cl   Liquid   25  -­‐  100  
Ammonium   Trimethylquaternary  
ammonium  chloride  
Imidazolines   1-­‐phenyl-­‐4,   4-­‐ C6H5NCHNC(CH3)2CH2   Solid   75  -­‐  100  
dimethyl  imidazoline  
Chlorinated   Sodium   NaOC6HCl4   Solid   12  –  50  
Phenols   tetrachlorphenate  
Aldehydes   Acrolein   CH2CHCHO   Liquid   10  –  85  
Peroxygens   Peracetic  acid   CH3COO2H   Liquid   >  10  
Table  2  Examples  of  biocides(1)  
 
Proprietary  biocides  will  be  complex  mixtures  and  may  include  corrosion  inhibitors.  
In  applying  biocides,  the  water  cannot  be  discharged  into  the  environment  without  
treatment  stages  to  remove  the  biocides.  Additional  safety  procedures  will  need  to  
be  put  in  place  to  maintain  stocks  on  site  as  to  prepare  the  compound  for  injection  
into  the  water  carrying  pipelines  and  processing  units.  

3.6 Treatment with Biocides

Potentially,   biocides   can   be   added   at   numerous   points   in   the   treatment   process.  
Biocides   are   preferred   over   biostats   as   this   will   reduce   the   chance   of   organisims  
building  up  resistance  to  the  chemical.  It  is  advisable  to  evaluate  the  effectiveness  of  
biocides   in   the   laboratory   before   application   although   injection   systems   can   be  
allowed  for  during  the  design  phase.  In  addition  to  being  effective  at  the  right  cost,  
the   chemical   should   be   compatible   with   other   chemicals   already   present   or  
potentially   added   to   the   water   stream.   One   of   the   benefits   of   an   integrated  
approach   to   water   handling   will   be   to   account   for   all   the   processing   stages   and  
hence   the   chemicals   added   to   the   water   as   they   pass   through   either   the   crude   oil  
treatment   process   (in   the   case   of   produced   water)   or   other   water   treatment  
processes.  
Key  issues  with  introducing  biocides  include:  
• slime  and  sludge  deposits  adhering  to  the  inner  walls  of  equipment  should  be  
removed.  Acid  washing,  mechanical  scraping  followed  by  detergent  washing  
and   flushing   are   possible   methods   although   washing   should   be   diverted,  

© School of Engineering and Physical Science Page 9

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
stored   and   treated   before   disposal.   It   will   be   easier   to   treat   a   clean   system  
rather  then  retrofit  into  an  existing  facility.  
• there  is  a  free  flowing  stream  of  diluent  water  –  no  stagnant  pockets  where  
diffusion  of  biocides  is  slow  or  prevented.  Injection  of  biocodes  should  not  be  
into   any   stagnant   or   slow   flowing   water   streams.   It   is   preferable   to   install  
static   mixers   or   inject   upstream   of   chokes   and   into   the   impeller   region   of  
centrifugal  pumps  to  ensure  rapid  mixing.  
• Biocides   will   not   act   instantaneously.   Table   1   illustrates   that   several   hrs  
contact   is   normally   required   for   biocide.   Continual   dosing   a   system   as   a  
constant   addition   rate   is   preferable   over   spasmodic   slug   injection   although  
slug   injection  can   ensure   a   sufficient   level   of   kill   as   the   biocide   concentration  
will  be  higher  in  the  initial  period  of  the  slug  treatment.  A  hold  up  tank  with  
sufficient  residence  time  may  be  required  downstream  of  the  injection  point.  
Biocides   should   be   injected   where   they   are   needed   in   order   to   protect   equipment  
down   stream;   ideally   the   injection   should   take   place   as   close   as   possible   to   the  
source   of   the   water.   It   may   not   be   practical   to   inject   upstream   of   the   primary  
separators,   prior   to   the   water   treatment   systems   would   be   a   possibility.   Figure2  
illustrates  one  position  for  the  biocide  injection  point,  directly  after  the  main  intake  
pump  for  a  system  using  surface  water  as  its  source  for  injection.  In  this  illustration,  
water  for  injection  is  drawn  from  a  sea  water  supply  which  requires  treatment  stages  
to  reduce  any  solids  and  adjust  the  pH.  
 

 
Figure  2  illustration  of  a  sea  water  treatment  process  
 
   
 
 
 

Page 10
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
4 Dissolved Gasses
The   corrosion   behaviour   of   water   varies   considerably   with   the   composition   of  
dissolved  components  and  it  is  often  the  dissolved  gases;  oxygen,  carbon  dioxide  and  
hydrogen  sulphide,  that  are  more  important.  Removing  dissolved  gases  depends  on  
the   solubility   of   that   gas   at   the   operating   temperature   and   pressure.   The   common  
method   to   remove   dissolved   gases   is   to   contact   the   water   with   another   gas   which  
ahs  low  solubility  but  an  affinity  for  the  dissolved  component.  

4.1 Influence of Dissolved Oxygen

Dissolved  oxygen  is  probably  causes  more  corrosion  issues  than  the  vast  majority  of  
soluble  gases.  The  following  equations  describe  the  role  of  dissolved  oxygen  in  water  
on  the  corrosion  of  iron:  
Fe + 2H + ←⎯→ Fe 2+ + 2H0  
1
2H0 + O 2 ←⎯→ H2 O  
2
1
2Fe2+ + O 2 + H2 O ←⎯→ 2Fe3+ + 2OH−  
2
 
Pure   water   can   contain   from   14   to   6   ppm   dissolved   oxygen   depending   on   the  
temperature.   This   value   reduces   slightly   due   to   the   presence   of   chloride   compounds  
(Figure  3)  

 
Figure  3  Solubility  of  oxygen  in  water,  with  and  without  the  presence  of  chloride  
 
 
In  terms  of  corrosive  power,  corrosion  in  steel  is  usually  proportional  to  the  oxygen  
content   (wt   loss   of   steel   =   dissolved   O2   content   (cm3/liquid))   up   to   about   5.5  
cm3/litre   (6).   It   is   generally   taken   that   an   O2   content   of   10   ppb   (parts   per   billion)   is  

© School of Engineering and Physical Science Page 11

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
sufficient   to   cause   corrosion   issues   in   down   stream   equipment   which   will   lead   to  
corrosion  products  in,  water  injection  lines.  
 

4.2 Solubility of Gases in Liquids

For  some  systems,  particulalr  those  up  to  100  kPa,  the  thermodynamic  equilibrium  
of  a  gas  with  water  usually  follows  Henry’s  Law:  
P̂ = HC  
where   P̂  is  the  partial  pressure  of  the  gas  solute  in  the  gas  phase  
  C  is  the  concentration  of  the  dissolved  gas  in  solution    
  H  is  the  proportionality  constant,  known  as  Henry’s  law  constant  
Note  that  the  units  of  H  vary  depending  on  the  literature  source,  and  may  be  written  
in   terms   of   1/H.   Table   3   shows   values   of   Henry’s   law   constant   for   a   range   of  
components(7)  with  respect  to  temperature.  
 
Temp   Air   N2   O2   H2S   O3  
           
  x10-­‐4   x10-­‐4   x10-­‐4   x10-­‐2   x10-­‐3  
           
0   4.32   5.29   2.55   2.68   1.94  
5   4.88   5.97   2.91   3.15   2.18  
10   5.49   6.68   3.27   3.67   2.48  
15   6.07   7.38   3.64   4.23   2.88  
20   6.64   8.04   4.01   4.83   3.76  
25   7.2   8.65   4.38   5.45   4.57  
30   7.71   9.24   4.75   6.09   5.98  
35   8.23   9.85   5.07   6.76   8.18  
40   8.7   10.4   5.35   7.45   12  
45   9.11   10.9   5.63   8.14    
50   9.45   11.3   5.88   8.84   27.4  
Table  3a  Henry’s  law  constant  (atm/(mol/mol))  for  common  gases  in  water(7)  
 
Henry’s   law   is   an   approximation   and   as   shown   in   the   table,   varies   with   temperature.  
The  lower  the  value,  the  greater  the  solubility.  
 
Additional  values,  based  on  data  from  Henley  &  Seader(11)  
Temperature   CO   CO2   Cl2   SO2   NH3  
0°C   3.372   0.0703   0.0224   0.00160   0.000212  
10°C   4.390   0.1092   0.0380   0.00228   0.000242  
20°C   5.973   0.1552   0.0589   0.00324   0.000276  
30°C   6.816   0.1851   0.0703   0.00048   0.000329  
Table  3  b  Henry’s  law  constants  for  selected  gases  (x10-­‐4  atm/(mol/mol))  
 
Example  
Estimate  the  solubility  (in  ppm)  of  Oxygen  in  water  at  10°C,  at  1  atm  pressure.  
 

Page 12
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 
 
Solution  
From   the   table,   Henry’s   law   constant   for   O2   at   10°C   is   3.27x104   atm/(mol/mol)   (note  
the  convention  applied  in  the  tables)  
Assuming   the   molar   composition   of   oxygen   in   air   is   20.9%,   the   partial   pressure   is  
therefore   P̂ = 0.209x1 = 0.209 atm  
0.209 atm
Mole  fraction  of  O2  in  solution  at  10°C  is   x = 4
= 6.39x10 −6  
3.27x10 (atm / (mol / mol))
If   the   average   molar   mass   of   the   solution   is   18   kg/kmol,   and   for   oxygen   is   32  
kg/kmol,  the  mass  fraction  is:  
⎛ 32 ⎞ -­‐6  
6.39x10 −6 ⎜ ⎟ =11.36x10 g  O2/g  solution  =  11.36  ppm,  taking  the  solution  density  
⎝ 18 ⎠
as  1000  g/litre  
 
Compared  to  Figure  3,  this  value  is  very  close  to  the  oxygen  content  in  fresh  water.  
Henry’s   law   constants   vary   from   source   to   source   and   usually   show   a   temperature  
dependency.   The   solubility   of   gases   in   liquids   depends   on   the   composition   of   the  
water   used   and   therefore   quoted   values   in   literature   should   be   used   with   caution.  
Experimentally  determined  solubility  values  using  the  water  from  the  process  should  
be  used  if  possible.  

4.3 Influence of CO2

Looking  at  the  reactions  of  CO2  and  O2  in  water:  
2CO 2 + 2H 2O + 2e − ⎯⎯ → 2HCO −3 + H 2  
2O 2 + 4H2 O + 8e − ⎯⎯→ 8OH−  
 
suggests  that  as  the  oxygen  reaction  requires  4  times  the  number  of  electrons,  that  
O2   is   four   times   as   corrosive   as   CO2.   However,   there   appears   to   be   a   synergistic  
effect   when   oxygen   and   carbon   dioxide   is   present:   for   a   solution   with   10   cm3/litre  
CO2,  corrosion  rate  is  8  mm/year,  a  solution  with  0.67  cm3/litre  O2,  corrosion  rate  is  
4   mm/year,   whereas   as   solution   containing   both   gases   gives   a   corrosion   rate   17  
mm/year(6).  Due  to  potential  synergies  between  dissolved  gases,  it  will  sometimes  be  
difficult   to   extract   the   influence   of   one   particular   gas   without   ensuring   the   other  
gases   and   indeed   other   dissolved   components   are   taken   out   of   solution   or  
accounted  for  in  some  way.    

4.4 Influence of Hydrogen Sulphide

Like   CO2,   H2S   is   corrosive   only   in   the   presence   of   moisture.   H2S   is   highly   soluble   in  
water   and   forms   a   weak   acidic   solution.   Under   an   oxygen   free   environment,  
dissolved   H2S   will   attack   iron   and   some   alloys.   When   dissolved   H2S   comes   in   contact  
with  dissolved  oxygen,  the  two  can  react  liberating  free  sulphur:  
1
H 2 S + O 2 ⎯⎯ → H 2O + S  
2

© School of Engineering and Physical Science Page 13

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
In  terms  of  corrosion,  mild  steel  is  prone  to  H2S  attach  up  to  about  150  to  200  ppm  
dissolved  H2S  where  the  corrosion  rate  is  about  24-­‐28  mm/year.  The  corrosion  rate  
remains   relatively   constant   for   200   –   400   ppm,   then   drops   slowly   to   about   10  
mm/year  at  about  1600  ppm.  Corrosion  rates  tend  to  be  enhanced  when  there  are  
other  corrosive  elements  in  solution.  
   

Page 14
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 

5 Water Treatments
Preventing   MIC,   removing   dissolved   gases   and   removing   suspended   solids   are   the  
three  main  operations  required  in  any  water  treatment  process,  for  injection  water  
and  incidentally  for  producing  clean  drinking  water.    

5.1 Microbial Treatment

Killing   micro-­‐organisms   within   water   is   usually   accomplished   through   the   addition   of  
a  chemical  biocide  with  chlorine  and  ozone  being  the  two  widely  accepted  routes.    

5.1.1 Chlorination
Chlorination  is  widely  used  to  sterilize  drinking  water  and  is  proven  technology;  the  
difference   when   applied   to   the   oilfield   industry   is   the   level   of   chlorination   necessary  
to   control   microorganisms   in   oilfield   water.   Remember   that   while   chlorine   has  
proven  economic  route  for  water  disinfection,  other  materials  may  be  more  lethal.  
Table  4  shows  typical  properties  of  the  three  common  disinfecting  agents  
 
Name   Symbol   Molar   Solubility   Boiling   Melting Latent  Heat   Normal  
Mass   in  water   Point   Point Vaporisation   Availability  
at  1  atm,   (°C)   (°C)   (cal/g)  
25°C  
(g/litre)  
Compressed  
Chlorine   Cl2   70.91   7.29   -­‐34.5   -­‐101   68.7  
liquid  
Manufactured  
on  site  due  to  
Chlorine  
ClO2   67.45   8   10.9   -­‐59   96.6   dangers  in  
dioxide  
long  term  
storage  
Manufactures  
Ozone   O3   48   0.006   -­‐112   -­‐192   54   on  site  since  it  
decomposes  
(3)
Table  4  Physical  properties  of  common  disinfecting  agents  
 
For   drinking   water   disinfection,   chlorine   has   a   proven   record   although   ozone   is   a  
more   powerful   disinfectant.   The   theory   behind   the   action   of   chlorination,   which  
applies   to   other   chemicals,   is   that   enzymes   within   the   bacteria   cell   are   oxidized.  
Chemically,   when   chlorine   is   added   to   water,   it   hydrolyses   to   form   the   hypochlorous  
ion  and  dissociated  hydrochloric  acid  and  it  the  formation  of  these  species,  which  is  
important:  
Cl2 + H 2O ←⎯ → HOCl + H + + Cl−  
The  hydrolysis  constant  for  this  reaction  is  approximated  by:  
1190.7
pK H = −0.579 +  
T
 

© School of Engineering and Physical Science Page 15

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
where   T  is  the  absolute  (Kelvin)  temperature  
  KH  is  the  hydrolysis  constant,  where   pK H = − log10 K H  
 
The  reaction  rate  is  fast  although  temperature  dependant;  needing  a  few  tenths  of  a  
second  for  complete  hydrolysis  at  18°C,  and  a  few  seconds  at  0°C.    
 
Being  a  weak  acid,  hypochlorous  acid  dissociates  as:  
HOCl←⎯ → H + + OCl−  
 
Values  of  the  ionisation  constant  KI  for  this  reaction  vary  depending  on  the  source  of  
data,  but  generally  KI  varies  as:  
HOCl  Ionization  Constant  Table  
Temperature  (oC)   0   5   10   15   20   25   30  
-­‐8
Ki  x  10  (moles/liter)   1.488   1.753   2.032   2.320   2.621   2.898   3.175  
 
From  this  data,  the  percentage  of  HOCl  present  at  different  temperatures  and  pH’s  is  
approximated  by:  
−1
⎛ K ⎞
%HOCl = 100 ⎜ 1+ +I ⎟  
⎝ ⎡⎣ H ⎤⎦ ⎠
where   [H+]  is  the  hydrogen  ion  concentration  
Thus,  for  a  temperature  of  25°C,  and  a  pH  of  8,  then  percentage  HOCl  is:  
 
−1
⎛ 2.898x10 −8 ⎞
100 ⎜ 1+
⎝ 10 −8 ⎟⎠ =25.65%  
 
A  convenient  correlation  for  the  pKI  value  is  found  by  fitting  the  data  to  the  absolute  
temperature  as:  
945.1619
pK I = 4.358 +  
T
The  importance  of  pK    values  is  highlighted  by  the  data  in  Table  5  which  shows  which  
of   the   chlorine   ionisation   products   are   useful   in   disinfecting   water;   ClO2   and   HOCl  
forms  are  the  most  lethal.  The  degree  of  effectiveness  is  related  to  the  ease  at  which  
ions   migrate   through   the   bacteria   cell   walls.   Note   that   ozone   if   the   more   lethal  
compared  to  chlorine  disinfectants.  
 
Disinfectant   Bacteria   Viruses   Spores  
O3   500   5   2  
ClO2   10   1.5   0.6  
HOCl   20   1   0.05  
-­‐
OCl   0.2   0.02   0.0005  
Table  5  Specific  lethality  of  disinfectants(3)  
 
 

Page 16
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
Therefore,  it  is  preferable  to  ensure  a  high  concentration  of  HOCl  rather  than  allow  
ionisation   into   OCl-­‐   to   take   place.   Graphing   the   correlation   for   pKI   against   pH   for  
several   temperatures   shows   that   pH   is   the   main   influencing   factor   on   the  
concentration  of  HOCl  (Figure  4).  Generally,  HOCl  dominates  at  pH  values  less  than  7.    
 
 
  100
 
  80
 
Percentage HOCl

  0 °C
  60
  15 °C
  40 30 °C
 
 
20
 
 
  0
  5 6 7 8 9 10 11
  pH
 
 
 
Figure  4  Effect  of  pH  on  the  concentration  of  HOCl  in  solution  at  3  temperatures  
 

5.1.2 Chlorine Dosage Amounts

The  amount  of  chlorine  added  will  depend  on  the  chlorine  demand  of  the  water,  the  
contact  time,  temperature  and  water  pH.    Because  of  its  reactivity,  chlorine  will  react  
with  elements  dissolved  into  the  water  and  it  will  only  be  the  residual  chlorine  left  
after   these   reactions   that   is   free   for   affecting   bacteria   e.g.   there   are   a   series   of  
complex  reactions  with  ammonia  and  its  derivatives  –  
Monochloramine   NH +4 + HOCl ! HN 2Cl + H 2O + H +  
Dichloramine   NH 2Cl + HOCl ! NHCl2 + H 2O  
Trichloramine   NHCl2 + HOCl ! NCl3 + H 2O  
Nitrogen   2NH +4 + 3HOCl ! N 2 + 5H + + 3Cl− + 3H 2O  
Nitrate   NH +4 + 4HOCl ! NO −3 + H 2O + 6H + + 4Cl−  
 
The  formation  of  chloramines  is  an  issue  for  chlorine  use  in  drinking  water  systems  
and   the   reason   why   water   treatment   plants   particularly   in   the   US   are   looking   into  
ozone.  European  water  processes  have  used  ozone  for  some  time.  
 
Practical   experience   recommends   that   9   to   10   mg   Cl2/mg   NH3   be   allowed   for   in  
design   (3)   and  that  a  residual  chlorine  concentration  of  0.2  ppm  is  often  sufficient  to  
control   micro   organisms   in   oilfield   waters(1).   This   value   increases   to   0.4   ppm   if   the  

© School of Engineering and Physical Science Page 17

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
water   pH   is   between   8   to   9,   and   0.8   ppm   for   a   pH   above   9   due   to   the   reduced  
generation  of  HOCl  as  pH  increases.  Note  that  1  ppm  residual  chlorine  can  corrode  
iron.   Should   chlorine   react   with   elements   not   normally   present   in   fresh   or   seawater,  
the   chlorine   levels   may   have   to   be   increased   which   affects   the   economics   of   the  
treatment  process.    
 

5.1.3 Choice of Chlorine Agents

There  are  several  ways  to  introduce  chlorine  into  water  systems;  direct  injection  of  
chlorine,  application  of  chlorine  dioxide.    
 
• Liquid   Chlorine   -­‐   As   indicated   in   Table   2,   free   chlorine   can   be   supplied   as   a  
compressed  liquid.  Chlorine  is  particularly  hazardous  not  only  because  it  is  a  
good   disinfectant,   but   from   risks   associated   with   asphyxiation.   In   cold  
climates,   any   water   injected   with   liquid   chlorine   should   be   heated   as   there   is  
a  possibility  of  forming  chlorine  ice  at  water  temperatures  below  10°C.  Note  
that  introduction  of  chlorine  gas  tends  to  lower  the  pH  of  water  and  hence  
favour  the  formation  of  HOCl.  Electrolysis  of  brines  can  also  be  used  to  yield  
chlorine.   Depending   on   efficiency   or   the   process   expect   to   consume   1.2   to  
2.2  kW-­‐hrs  per  1lb  of  chlorine.  Hydrogen  produced  will  need  to  be  separated,  
vented  and  disposed  of.  
 
• Chlorine   dioxide   –   Although   the   lethality   of   ClO2   is   less   than   HOCl,   this  
material   does   have   several   advantages   in   that   it   is   not   susceptible   to  
variations  in  pH  and  does  not  tend  to  form  chloramines.  On  site  production  
of  ClO2  accomplished  by  addition  of  chlorine  to  sodium  chlorite  in  solution:  
Cl2 + 2NaClO 2 ! 2ClO 2 + 2NaCl  
Chlorine   dioxide   solutions   in   the   range   6   to   10   g/litre   are   normally   prepared   and  
diluted  to  1  g/litre  for  storage  and  subsequent  use.  

5.1.4 Ozonation
Ozone,   which   has   been   used   in   European   water   treatment   plants,   does   not   suffer  
from  the  same  problems  as  with  chlorine  in  terms  of  producing  potentially  harmful  
by-­‐products,   and   is   significantly   more   lethal   than   chlorine   to   microorganisms.   Ozone  
is  a  highly  reactive  gas  formed  by  electrical  discharges  in  the  presence  of  oxygen.  It  is  
more  soluble  in  water  than  oxygen  and  obeys  Henry’s  law  in  that  the  concentration  
in  water  is  a  linear  function  of  the  partial  pressure  in  the  gas  phase.  This  relation  can  
be  written  in  terms  of  Henry’s  law  constant  or  as  a  solubility  constant(3):  
x = K s P̂O2  
where   x  is  the  ozone  concentration  in  water    
  Ks  is  the  solubility  constant  
  P̂O2  is  the  partial  pressure  of  ozone  
 
The  solubility  constant  is  expressed  as  a  function  of  temperature  as:  

Page 18
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
1.29x10 6
Ks =  
T − 3720.5
where   T  is  the  absolute  temperature  
 
The  chemical  pathways  by  which  ozone  acts,  is  complex  but  essentially  there  are  two  
routes;  direct  action  and  auto-­‐decomposition.  Direct  action  is  the  preferred  route  for  
eliminating  microorganisims  and  required  a  high  residual  ozone  concentration.  Auto-­‐
decomposition   involves   hydroxyl   radicals,   which   act   on   organic   matter   but   are  
removed   by   carbonate   and   bicarbonate   ions.   pH   also   plays   a   role   in   determining  
which  pathways  are  active.  A  crude  estimation  of  the  residual  ozone  concentration  
due  to  auto-­‐decomposition(3)  is  given  by:  
[O3 ]t = 10 − At  
[O3 ]o
where   [O 3 ]t   is  the  residual  concentration  at  any  time  t  (t=0  is  the  residual  at  time  
0)  and A = 10(0.636pH − 6.97 ) .  This  expression  is  plotted  in  Figure  5  and  shows  not  only  the  
half-­‐life   but   the   dependence   of   this   system   on   pH.   In   order   to   maintain   a   high  
residual  ozone  concentration,  pH  should  be  low.  
 
  1
 
 
0.8
 
 
Residual Ratio

  0.6

  Decreasing pH
  0.4 from 9 to 5

 
  0.2
 
 
0
  0 20 40 60 80 100
  Time (mins)
 
Figure  5  Residual  oxone  concentration  with  time  and  pH  
   

5.2 Removing Suspended Solids

Filtration   through   a   fixed   bed   media   remains   the   key   technology   for   removing  
suspended  solids  from  water.  The  basic  principle  is  to  pass  the  dirty  water  through  a  
filter  media,  the  solid  particles  becoming  trapped  within  the  filter  media  and  clean  
water  passing  through.  As  more  and  more  solids  become  trapped,  the  pressure  drop  
across   the   media   bed   increases   and   there   will   come   a   limiting   point   when   the   bed  
has   to   be   cleaned   or   replaced.   Figure6   shows   two   representations   of   the   action   of  
filters;  a  classical  filter  paper/cloth  system  and  one  where  the  filter  bed  is  made  up  
of  solid  particles.  
 
© School of Engineering and Physical Science Page 19
 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 

 
Figure  6a  Schematic  representation  of  cloth  or  fibre  filtration  process.  Solids  
deposited  on  the  media  form  a  cake  which  captures  more  solid  particles.  
 
 

 
Figure  6b  Schematic  representation  of  a  fixed/solid  bed  filtration  system  where  
particle  becomes  trapped  within  the  bed.  
 
 
 
Page 20
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
The  pressure  necessary  to  force  liquid  through  the  bed  comes  from  either  a  gravity  
feed  system,  under  pressure  produced  by  pumps,  or  pulled  through  under  vacuum.    
Filtration  systems  using  the  mechanism  in  figure  6b  are  widely  used  in  the  chemical  
industry;  the  plate  and  frame  filter  press,  bag  filter  and  rotary  drum  filter,  are  typical  
examples  of  filter  systems  which  require  a  layer  of  the  suspended  solid  to  grow.  In  
these   systems,   it   is   the   collected   solids   which   are   needed,   not   the   liquid   effluent.  
Fixed   bed   filters   contain   a   volume   of   solids   which   trap   the   suspended   solids   –   in   this  
respect,   the   cake   refers   to   the   entire   bed   of   solid   material.   Although   the   two  
mechanisms   appear   to   differ,   the   fundamental   operating   modes   and   underlying  
theory  is  the  same.  
Filters  can  operate  in  two  modes:  
• Constant   flowrate   –   where   the   pressure   will   gradually   increase   up   to   an  
optimum  limit  
• Constant   ΔP   –   where   the   flowrate   of   water   through   the   filter   will   gradually  
reduce  again  to  a  minimum  limit.  
 

5.2.1 Theory of Filtration

Theoretical   analysis   of   filtration   results   in   equations   for   estimating   the   area   required  
for  filtration  and  the  duration  of  the  filtration  step.  Under  operating  conditions,  the  
total   filtration   time   will   consist   of   two   parts;   that   due   to   the   filtration   process   and  
that  due  to  the  time  required  to  clean  and  removed  the  collected  solids.    
 
For   streamlined   flow   through   a   bed   of   solids,   the   pressure   drop   across   the   bed   is  
related  to  the  velocity  and  properties  of  the  bed  by  the  Kozeny  equations(5):  
1 e3 1 ΔP
u=  
K S (1− e ) µ l
2 2

where   K   is   the   Kozeny   constant,   usually   has   the   value   5   but   depends   on  
properties  of  the  bed.  
  e   is  the  fractional  voidage  –  volume  of  the  bed  not  occupied  by  solid  
material  
  S   is  the  specific  surface  area  of  the  solid  particles  and  is  the  surface  
area   of   a   particle   divided   by   its   volume.   For   spheres,  
π d2 6
S = 3 =  
πd / 6 d
  µ, ΔP, l   are   the   viscosity   of   fluid,   pressure   drop   and   length/height   of   the  
bed  respectively.  
 
K,   S,   e   are   properties   of   the   bed   and   if   the   bed   properties   remain   constant,   then  
these   terms   are   constant.   This   is   the   case   where   the   build   up   of   solids   within   the  
filter  media  does  not  appreciably  alter  the  voidage  and  hence  the  flow  through  the  
bed.  Therefore,  at  any  instant,  the  instantaneous  velocity  of  liquid  through  the  bed  is  
given  by:  
 

© School of Engineering and Physical Science Page 21

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
1 dV ΔP
u= =  
A dt rc µl
−1
⎛ e3 ⎞
where   rc = ⎜ ⎟  is  the  specific  resistance  of  the  filter  cake.  
⎝ KS (1− e ) ⎠
2 2

  V   is  the  volume  of  filtrate  (dirty  water)  

  A   is  the  cross  sectional  area  of  the  filter  
 
There  is  a  relationship  between  the  depth  (l)  of  cake  and  the  volume  of  filtrate  (V)  in  
Al
the  bed  and  the  volume  of  solids  deposited  as:   ν = .  This  gives:  
V
dV A 2 ΔP
=  
dt rc µν V
a) For  filters  under  constant  pressure  filtration,  which  is  a  more  frequent  mode  
of  operating,  integration  of  the  above  gives:  
A 2 ΔP
V t

∫ VdV = ∫
Vi ti
rc µν
dt  

1 2 A 2 ΔP
2
⎡⎣V − Vi2 ⎤⎦ =
rc µν
( t − t i )  
Thus,  there  is  a  linear  relationship  between  V2  and  t.  To  understand  the  process  
further,  this  expression  can  be  rearranged  as:  
1 A 2 ΔP
2
(V − Vi )(V + Vi ) = r µν ( t − t i )  
c

t − ti rc µν
=
V − Vi 2A 2 ΔP
(V + Vi )  
t − ti r µν 2r µν
= c 2 ( V − Vi ) + c2 Vi  
V − Vi 2A ΔP 2A ΔP
For   a   fixed   bed   filter,   the   term   on   the   extreme   right   represents   the   properties   of  
the  bed  prior  to  operation.  
 
b)   For   filters   under   constant   pressure,   operating   at   steady   state,   the   relationship  
between  the  volume  of  filtrate  processed  and  the  time  taken  is  given  by:  
t µα rcV µrs L s
= +  
V 2A 2 ΔP AΔP
 
 
where   V   volume  of  filtrate  
  t     filtration  time,  V  is  processed  
vs
  α   volume  of  cake  per  unit  volume  of  filtrate  ,  given  by   α =  
vc − vs
  vs   is  the  volume  fraction  of  solids  in  the  slurry  
  vc    volume  fraction  of  solids  in  the  filter  cake  
  A   filtering  area  

Page 22
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
  ΔP   total  pressure  drop  across  the  filter  bed  
  Ls    is  the  height  of  the  fixed  bed  
  rc   specific  resistance  of  the  cake  
  rs   specific  resistance  of  the  bed  
 
 
The   majority   of   the   variables   in   the   above   expression   are   determined  
experimentally;  a  plot  of  t/V  vs  V  should  provide  sufficient  information  to  calculate  rc  
and  rs  for  a  given  bed.  
 
The  value  of  rc  changes  with  filtration  pressure  and  is  usually  represented  by  a  power  
law  (6)  
rc = kΔP c  
where    k  and  c  are  constants,  found  by  analysing  filtration  curves.  
 

5.2.2 Filter Types

Material  making  up  the  filter  bed  ranges  from  beds  of  particles  such  as  sand,  glass  
beads  etc,  to  rigid  porous  plates,  paper,  cloths.  Only  a  few  of  the  many  types  have  
been   proven   in   the   oil   industry.   The   majority   of   filters   capture   and   retain   the  
separated   solid   particles   within   or   on   the   filter   media;   rotary   drum   and   belt   filters  
are   exceptions   where   the   filter   cake   builds   up   on   the   filter   surface   to   be  
automatically   removed   by   scraping   the   solids   away   from   the   filter   cloth.   In   cases  
where  solids  are  retained,  periodically  the  filter  must  be  withdrawn  from  service  in  
order  for  build  up  solids  to  be  removed  and  disposed  of.  Backwashing  is  the  normal  
route  to  clean  the  filter;  feeding  wash  water  in  the  reverse  direction  to  the  normal  
flow  of  dirty  water.  To  be  effective,  the  backwashing  period  should  be  quicker  than  
the  filtration  period.    
 
For  oilfield  operations,  filtration  equipment  fall  into  5  categories:  
 
Granular  Media  Filter   These   are   packed   beds   loaded   with   fine   particulate  
matter,   which   are   re-­‐generated   by   back-­‐flushing   to  
remove   collected   solids.   When   packed,   sand   usually  
have  25%  voidage.  
Fibre  Filters   A  fabric  version  which  operates  in  a  similar  fashion  to  
granular   systems.   Suspended   solids   will   be   captures  
by  the  fabric  strands  as  dirty  water  is  passed  through  
the   filter.   Synthetic   fibres   are   commonly   used   over  
natural  fibres.  
Diatomaceous   earth   (DE)   Diatomaceous   earth,   usually   called   diatomite,   is  
filters   nearly  pure  silica  with  a  density  lighter  than  sand  and  
usually  contains  70%  voidage  when  packed.    
Disposable  cartridge  filters   These  are  usually  filter  cloths  wound  into  form  cartri  
Acid   cleanable   cartridge   These   are   fairly   lightweight   units   producing   clean  
filters   water  from  20  micron  to  1  micron,  depending  on  the  

© School of Engineering and Physical Science Page 23

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
cartridge.  
 

5.2.3 Sand and Granular Filters

Sand   in   combination   with   gravel   is   perhaps   the   more   common   oilfield   filtration  
systems  due  to  the  low  cost  of  the  media  and  being  easy  to  clean.  Figure7  illustrates  
a  typical  gravel  filter  systems  where  the  dirty  water  is  fed  under  pressure  to  the  top  
of   the   packed   bed.   As   the   liquid   flows   down   through   the   bed   to   reach   the   outlet,  
suspended  solids  are  captured  within  the  bed.    

 
Figure  7  Example  illustration  of  Natco’s  Hiperfiler  
 
Filter  Beds  can  consist  of  different  sized  particles  either  arranged  with  sizes  from  fine  
to  coarse,  coarse  to  fine  or  randomly  packed.  Solids  with  a  narrow  size  distribution  
can  be  used  in  filters  to  give  uniform  size  characteristics.  
 
In  systems  where  the  filter  bed  is  arranged  from  fine  to  coarse,  dirty  water  is  fed  first  
through   the   fine   size   band   and   passing   out   through   the   coarse   band.   Figure7  
illustrated   three   types   of   material   making   up   the   entire   filter   bed.   Where   the   filter  
bed  consists  of  the  same  material  but  of  different  sizes,  backwashing  in  the  up-­‐flow  
direction   causes   smaller   particles   to   rise   to   the   top   of   the   bed.   Switching   to   the  
filtration   cycle,   the   redistribution   of   bed   material   results   in   very   rapid   solid-­‐liquid  
separation  with  the  majority  of  the  suspended  solids  collecting  in  the  top  section  of  
the   filter.   Attempts   to   ensure   more   of   the   bed   have   been   tried;   in   this   case   the   filter  
bed  is  made  up  of  materials  with  differing  densities,  with  the  coarser  lighter  material  
on  top  of  a  denser  finer  material.  On  the  backwash  cycle,  the  differencing  densities  
ensure  the  bed  size  distribution  is  retained;  coarse  to  fine.  
 

Page 24
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
Fixed  bed  systems  come  in  two  varieties;  single  flow  systems  where  the  filtrate  is  fed  
either   through   the   bottom   or   the   top.   The   split   flow   or   dual   flow   system   has   feed  
water   through   the   bottom   or   top   and   the   clean   water   extracted   from   the   centre  
portion;  effectively  two  filter  beds  within  the  one  pressure  vessel.  Figure8  shows  a  
schematic  of  these  types  of  filter  beds.  
 

   
Figure  8  Schematic  of  the  single  and  dual  flow  filter  systems  
 
Table  6  shows  examples  of  size  distributions  for  the  filter  media  in  sand  filters.  
 
Sieve   US  Sieve   Percent  Passing  sieve  size  
Opening   no  
(mm)     Fine   Medium   Coarse  
  0.26  –  0.45  mm   0.45  –  0.55  mm   0.55  –  2.0  mm  
1.19   16   94  –  100   84  –  99   68  –  93  
0.84   20   71  –  97   49  –  84   30  –  71  
0.59   30   31  –  73   14  –  39   6  –  31  
0.42   40   6  –  25   3  -­‐  6   0  –  1  
0.30   50   0  -­‐  3   0  -­‐  1   0  
(6)
Table  6  Typical  specifications  for  filter  sands.  
 
In  general,  filter  sand  is  classed  as  material  with  mean  size  less  than  2mm  and  should  
not  contain  clay,  dirt  or  organic  matter.  Gravel  used  in  filters  should  consist  of  hard  
rounded  stones  with  average  specific  gravity  2.5.  
 
In   terms   of   bed   depth,   single   media   fine-­‐coarse   bed   filters   are   typically   12   to   36  
inches.  Mixed  media  filters  have  been  tried  where  the  initial  distribution  is(3):  
• 6  to  12  inches  silica  sand  (sg  2.55  to  2.65,  size  0.4  to  0.55  mm)  
• 20  to  27  inches  anthracite  (sg  1.5  to  1.75,  size  0.8  to  1.1  mm)  
Upflow  coarse-­‐fine  filters  used  Russia,  have  been  reported  to  be  relatively  deep  –  6.5  
to  8.5  ft  of  sand  with  media  size  0.5  to  2  mm.    Bioflow  filters  where  feed  is  delivered  

© School of Engineering and Physical Science Page 25

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
to  the  top  and  bottom,  with  the  clean  water  outlet  in  the  middle,  are  typically  5  to  
5.5  ft  deep  with  the  intermediate  point  1.5  to  2ft  below  the  top  surface.    
 

5.2.4 Filtration Rate

Filtration  rate  through  a  sand  filter  depends  on:  
• Quality   of   the   water   required   –   generally   filter   units   are   sized   according   to  
their  design  capacity.  Increasing  the  flow  through  the  filter  is  normally  said  to  
reduce   the   quality   but   the   operational   problems   may   relate   more   to   a  
reducing  in  filterability  due  to  higher  deposition  of  solids.  
• Quality   of   feed   water   –   filtration   is   designed   to   remove   suspended   solids  
which   may   come   from   chemical   reactions   of   components   in   the   water.  
Chemical   action   should   therefore   be   elimination   to   avoid   chemically  
produced   solids   after   the   filtration   step.   Oil   present   in   the   feed   to   filters  
needs  to  be  removed  prior  to  filtration;  oil  combined  with  finely  suspended  
solids   forms   a   gum   like   composite   which   is   difficult   to   remove   from   sand  
grains  in  the  filter  by  standard  backwashing.  
• Turbidity   of   the   feed   water   –   when   treating   surface   waters,  
flocculating/coagulating  agents  can  be  used  to  increase  the  size  of  suspended  
solid  flocs.  Turbidity  is  a  measure  of  the  optical  clarity  of  water  samples  and  
hence  related  to  the  concentration  of  suspended  solids.  
Table  7  provides  some  indication  of  the  filtration  capacities  from  US  systems  
Media   Average  rate   Turbidity   Turbidity  of   Average   Terminal  
of  feed   filtrate   Run  Time   head  loss  
  (gpm/ft2)       (hrs)   ft  
Single   2   4.9   0.28   45   8  
Mixed   2   4.9   0.24   63   8  
Single   4   6.3   0.35   19   8  
Mixed   4   6.3   0.29   24   8  
Single   6   4.5   0.45   10   6.8  
Mixed   6   4.5   0.3   15   6.8  
Single   2   2   0.1   42   3.0  
Dual   3.5   2   0.1   60   2.2  
Dual   6.87   -­‐   0.06   51   -­‐  
Mixed   3  -­‐  6   0.5  –  10   0.1  –  0.5   15  –  80   -­‐  
Single   2   1   0.06   40   8  
Mixed   5  -­‐  6   1   0.05   23   8  
Mixed   5   -­‐   0.3   8  –  12   -­‐  
Dual   3.3   1  –  3   0.07   80   -­‐  
(3)
Table  7  Plant  scale  filtration  performance  results  
 

5.2.5 Backwash Issues

If   the   a   slow   filter   is   used   where   the   suspended   solids   build   up   on   the   upper   layer   of  
the   filter   bed,   simple   manual   removal   and   replacing   can   be   sufficient   in   cleaning   the  

Page 26
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
filter;   for   pressure   filters   this   will   not   be   a   practical   option   as   de-­‐pressurising   the  
vessel   will   call   for   additional   down   time   associated   with   safety   checks   when   the  
vessel   is   re-­‐pressurised.   For   rapid   filters   where   solids   are   captured   deeper   into   the  
bed,   this   option   is   not   practical   and   washing   by   back-­‐flushing   with   clean   water  
becomes   a   viable   method.   The   general   principle   is   that   back-­‐flush   water   should  
cause  the  filter  bed  to  expand  and  the  agitation  between  solid  particles  will  release  
any  deposited  solids.  
 
Effective  backwashing  depends  on  a  number  of  factors(6)  
• Good   distribution   of   wash   water   –   the   distributor   should   produce   uniform  
flow  to  the  bed  and  avoid  channelling  to  ensure  the  whole  bed  is  expanded.  
Attention   should   be   paid   to   the   arrangements   of   orifices   from   flow  
distributors;  uneven  flow  is  possible  due  to  the  gradual  loss  in  pressure  from  
the  feed  point  to  the  tip  of  the  distributor.  Orifices  should  be  relatively  close  
together  enabling  good  coverage  of  the  filter  bed.  
• Back-­‐flush  water  velocity  should  not  be  too  high  to  cause  a  mixed  sized  bed  
to  be  disturbed;  supports  and  guards  may  be  needed  to  hold  parts  of  the  bed  
in  place  during  back-­‐flush  and  normal  filtering  modes.  Wash  velocity  should  
however   be   sufficiently   high   to   cause   the   filter   bed   to   expand   by   50%   and  
hence  cause  sufficient  agitation  to  remove  the  deposited  solids.  
• Viscosity  of  the  wash  water  may  influence  the  back-­‐flush  operation  –  a  higher  
temperature   may   require   a   higher   flowrate   since   the   viscosity   of   water   is  
lower.   This   becomes   an   issue   if   winter   and   summer   temperatures   vary  
significantly.  
• On  the  surface  of  downflow  filters,  pockets  of  dense  flocculated  material  may  
gather   as   a   result   of   the   filtration   cycle   which   may   not   be   adequately  
removed   by   a   normal   back-­‐flush   operation.   In   this   event,   washing   jets   angled  
to   agitate   the   upper   surface   of   the   bed   may   be   used   to   break   up   these  
clusters.    
• Air   scouring   is   technique   introduced   to   reduced   the   volume   of   back-­‐flush  
water   required.   Air   or   other   suitable   gas   is   passed   through   the   filter   bed   to  
lift,   expand   and   agitate   the   bed   media.   Scouring   can   take   place   for   several  
minutes  for  a  flowrate  of  3  to  5  ft3/min  per  ft2  of  filter  area.  
• The   water   used   for   cleaning   should   be   treatable   to   collect   the   removed  
solids.  One  of  the  major  operational  issues  with  these  filters  will  be  the  cost  
assonated  in  supplying  and  treating  backwash  water.  
 
Figure   9a   shows   an   example   of   the   filter-­‐wash   cycle   where   the   bed   is   fluidised   by  
clean   up   liquid,   the   waste   extracted   and   the   media   left   to   settle   before   the   next  
filtration  cycle.  
 

© School of Engineering and Physical Science Page 27

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 

 
Figure  9a  Backflush  cleaning  operation  
 

Page 28
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

 
Figure  9b  shows  an  illustration  of  a  back-­‐flush  with  air  scouring  cycle.  

5.2.6 Back-flush frequency

One  of  the  more  important  features  of  any  filtration  system  is  the  effectiveness  of  
the   back-­‐flushing   operation.   When   to   backflush   can   be   estimated   from   a  
performance   plot   relating   the   volume   of   filtrate   processed   against   the   time,   for   a  
constant   pressure   drop   system.   Usually,   once   the   linear   relationship   between   time  
and  volume  begins  to  fail,  is  an  indication  the  filter  should  be  cleaned;  the  equations  
developed   earlier   show   a   linear   relationship   between   time   and   volume   of   filtrate  
should   exist   for   beds   which   are   functioning   correctly.   In   terms   of   back-­‐flush  
throughput,  usually  the  volume  of  clean  water  is  usually  1  to  5%  of  the  dirty  water  
flowrate.  
 
Backwashing  implies  that  solids  filtration  is  a  semi-­‐batch  operation  and  therefore  to  
maintain  constant  operation,  a  number  of  filter  beds  would  normally  be  used;  some  
beds  would  be  in  backflush  mode  while  others  are  operating  in  filtration  mode.  
 

5.2.7 Design Issues

Fixed   bed   filter   systems   are   usually   propriety   to   vendors   due   to   the   arrangements   of  
solids  used  in  the  filterbed,  particular  features  of  the  feed  and  off  take  nozzles  and  
the  mode  of  backflush.  Figures  10  illustrate  the  size  of  filters  supplied  by  Natco.  
 

© School of Engineering and Physical Science Page 29

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 

 
Figure  10  Capacities  for  single  and  dual  flow  solid  filter  systems  
 
 
Figure   11   shows   typical   installation   of   a   solids   filtration   system,   with   a   number   of  
filtration  units  mounted  on  a  skid  arrangement.    

Page 30
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

 
Figure  11  Skid  mounted  filtration  units  

5.3 Aeration
The   process   known   as   aeration   is   usually   taken   to   mean   the   addition   of   oxygen   to  
water;  there  are  several  reasons  why  this  may  be  wanted:  
a) To   replace   or   increase   the   oxygen   levels   prior   to   water   discharge   into   the  
environment  
b) For  the  specific  purpose  to  oxidize  hydrogen  sulphide  or  to  affect  the  removal  
of  iron  and  manganese.  
Equipment   for   adding   oxygen   taken   from   air   into   water   fall   into   two   categories;  
those  which  spray  water  into  an  air  stream,  and  those  which  bubble  air  into  a  tank  or  
pipeline   transporting   water.   Spray   aerators   are   quite   efficient   but   require   a   large  
surface   area   and   cannot   operate   effectively   if   freezing   conditions   are   likely.   In   this  
unit,  a  number  of  jets  will  be  positions  and  angled  to  ensure  the  trajectory  of  water  
drops  gives  the  maximum  contact  time  with  the  air.  
 
In  the  second  class  of  unit,  compressed  air  is  forced  though  sparger  systems;  either  
perforated   pipes   or   porous   pipes,   located   at   the   base   of   the   aeration   tank.   These  
tanks   can   be   9   to   15   ft   deep   and   10   to   30   ft   wide(3)   .   Table   8   shows   some   typical  
capacities   from   water   treatment   works   and   demonstrates   the   range   of   capacities,  
residence  times  and  air  supply  typically  used:  
 
 
 
 
 
 
 
 
© School of Engineering and Physical Science Page 31
 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 
Usage   Capacity   Aeration   Air  Supply   Power  Required  
(mgd)   Period   (ft3/gal)   (kW/mgd)  
(min)  
Gas  removal   4   14   0.07   0.48  
Iron  &  odor  removal   0.75   18   0.16   1.87  
Odor  removal   1   30   0.1   1.46  
  0.4   8.5   0.15   0.75  
  104   1.76   0.035   0.9  
  100   15   0.16   1.49  
  48   0.16   0.01    
  11.6   0.21   0.01    
(3)
Table  8  Capacities  of  water  aeration  systems  (mgp  =  million  gallons  per  day)  
 
The  rate  of  mass  transfer  of  a  component  carried  by  an  inert  gas  into  water  depends  
on   the   diffusion   rates   across   the   interface   and   hence   on   the   surface   area   between  
gas   and   liquid   phases.   Spargers   and   devices   to   manufacture   small   bubbles   of   gas   are  
one   method   used   to   enhance   the   mass   transfer   process.   Another   technique   for   a  
contacting  gas  and  liquid  is  to  spread  the  liquid  on  an  inert  surface  and  hence  expose  
the  liquid  to  the  gas.    
 
Transfer   of   material   can   take   place   in   the   opposite   direction;   it   all   depends   on   the  
solubility  of  the  component  in  the  water  phase.  For  oilfield  operations,  aeration  will  
increase  the  oxygen  content  of  water  and  would  normally  be  used  onto  to  remove  
other  dissolved  materials.  Oxygen  should  be  removed  as  early  as  possible  to  prevent  
corrosion  and  hence  formation  of  solids  which  could  block  the  reservoir.    
Sparging   and   spray   systems   are   useful   for   contacting   water   with   air,   but   packed  
columns  are  more  useful  for  contacting  liquids  with  a  specific  gas  where  the  gas  has  
to  be  contained  downstream.    
 

5.4 Deaeration by Packed Columns

Packed  tower  absorption/stripping  columns  are  the  principle  unit  operation  used  to  
contact  gas  and  liquid  phases  for  the  express  purpose  of  transferring  a  solute  from  
one   phase   to   the   other;   absorption   is   used   to   describe   transfer   from   gas   to   liquid,  
stripping   is   used   for   transferring   solute   from   liquid   to   gas.   To   specify   a   particular  
column,  the  role  of  the  designer  is  to:  
a) Determine  the  height  of  packing  required  
b) Set  the  flowrate  of  stripping/absorption  gas/liquid  
c) Specify  the  type  of  packing  
d) Find  the  column  diameter  and  overall  pressure  drop.  
In   processing   water   for   injection,   the   packed   column   would   be   used   in   a   stripping  
operation,  removing  dissolved  gas  e.g.  oxygen.  For  drinking  water  treatment,  it  may  
be  desirable  to  remove  sulphide  and/or  volatile  organic  compounds,  or  add  oxygen.  

Page 32
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
One   of   the   more   common   ways   to   operate   deaeration   systems   is   under   reduced  
pressure  –  in  what  is  commonly  referred  to  as  vacuum  deaeration.  In  this  case  the  
solubility  of  dissolved  gases  is  reduced  without  the  need  to  increase  the  temperature  
of   the   water   hence   an   overall   reduction   in   energy   costs.   Several   vendors   supply  
units,   which   operate   either   totally   under   vacuum   or   where   part   of   the   tower   is  
under,  vacuum,  the  other  part  under  higher  pressure.  As  with  all  reduced  pressure  
systems,  maintaining  an  effective  of  air  from  the  environment  remains  an  issue.  
Other   issues   with   using   packed   columns   for   removing   oxygen   (although   the   term  
deaeration   refers   to   removal   of   air,   it   is   usual   to   mean   removal   of   oxygen   from  
water),  include:  
• The   stripping   gas   used   should   be   one   which   can   be   easily   recovered   or  
disposed  of;  it  would  be  common  to  use  a  gas  source  which  is  in  ready  supply  
and  can  easily  be  disposed  of.    
• Water   soluble   gases   present   in   the   stripping   gas   will   become   dissolved   in   the  
water.   A   good   example   of   this   is   natural   gas;   which   is   commonly   used   as   a  
stripping  gas;  containing  quantities  of  carbon  dioxide;  CO2  is  more  soluble  in  
water  than  oxygen  (see  Henry’s  law  constants).  
• Deaeration  towers  should  be  uses  as  early  as  possible  to  eliminate  corrosion  
of  down  stream  processing  equipment.  Using  materials  resistant  to  corrosion  
e.g.   plastic   lined   pipe   provides   more   control   over   where   deaeration   towers  
are  placed.  
• The  solubility  of  gases  in  water  is  higher  at  lower  temperatures  and  columns  
should   be   designed   with   the   coldest   temperature   likely   to   occur.   A   rise   in  
water  temperature  would  them  benefit  oxygen  removal.  
• Solubility   of   oxygen   is   also   affected   by   external   pressure   due   to   the   lower  
boiling  point  of  water.  In  some  cases,  columns  cane  be  run  under  vacuum.  
• If   the   liquid   rate   increases,   all   other   variables   remaining   constant,   then   the  
column  efficiency  will  reduce  –  less  oxygen  being  removed.  
• If   the   gas   rate   increases,   all   other   variables   remaining   constant,   then   the  
liquid   is   likely   to   be   held   up   within   the   packing   and   the   column   is   said   to  
approach  a  flooding  condition.  
• Packed  columns  can  become  clogged  by  fine  particulate  material  present  in  
the  water  feed.  Filters  ahead  of  deaeration  columns  are  advisable.  
A  packed  column  absorber/stripper  consists  of  a  tower  filled  with  packing  material.  
Liquid  enters  at  the  top  of  the  tower  and  is  passed  through  a  distributor  so  that  the  
liquid  coats  the  surface  of  the  packing.  At  the  base  of  the  tower,  liquid  is  collected  
and  fed  by  pipeline  away  from  the  unit.  Gas  enters  the  bottom  of  the  tower  through  
a   distributor   and   rises   counter   current   to   the   liquid   eventually   being   collected   at   the  
top   of   the   tower   and   passed   to   further   processing   stages.   Packed   tower   absorbers  
therefore  operate  continually.    

© School of Engineering and Physical Science Page 33

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
5.4.1 Design Methods
To  specify  the  size  and  hence  estimate  the  cost  of  a  packed  column,  there  are  three  
stages  to  follow:  
1. Determine  the  overall  height  of  packing  to  use  for  a  given  liquid  and  gas  rate,  
which  is  a  function  of  the  mass  transfer  characteristics  of  the  system.  
2. Find   the   diameter   for   a   given   liquid   and   gas   flowrate   which   avoids   flooding  
the  column  and  poor  contact  between  the  two  phases.  This  is  the  Hydraulic  
constraint  of  the  column.  
3. Determine   the   height   of   a   transfer   unit   and   hence   the   overall   height   of   the  
tower.  
 
1.  Overall  Height  of  Packing  
It  is  convenient  to  describe  the  column  in  terms  of  the  gas  and  liquid  phase  although  
the   theory   can   be   applied   to   contact   between   any   immiscible   phases.   The   packing   is  
effectively   divided   into   a   number   of   stages   –   called   “transfer   units”   where   we  
assume  the  rate  of  mass  transfer  is  uniform.  An  estimate  of  the  number  of  transfer  
units  is  made  from  knowledge  of  the  driving  force  causing  mass  to  transfer  from  one  
phase  into  the  other.  Thus,  this  section  will  derive  the  expressions  for  the  number  of  
transfer  units.  
  Gas Phase Liquid Phase
 
  Mole ratio of solute in gas Mole ratio of solute in liquid
  phase out of column phase into column
  Y2 X2
 
  Carrier Gas Molar Flowate Carrier Liquid Molar Flowate
 
 
G m' L'm
 
 
  Height of
  Packing
  Z
 
 
 
 
  Mole ratio of solute in gas Mole ratio of solute in liquid
phase into column
  phase out of column
  Y1 X1
 
 
 
For  dilute  systems,  which  is  typical  of  operations  for  injection  water  treatment,  the  
overall  height  of  packing  can  be  estimated  from  the  following  equations:  

Page 34
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
Overall  height  is  given  by  :   Z=HOGxNOG  
Y1
dY
Where   NOG    is  the  number  of  transfer  units,   NOG =
Y2
∫ (Y
e − Y)
 

Gm'
  HOG     is  the  height  of  a  transfer  unit,   HOG =  
aK G
  Z   is  the  overall  height  of  packing  required  
 
    '
Gm   is   the   molar   flowrate   of   carrier   gas   kmols/hr-­‐m2   of   tower   csa  
V
'
Gm = ,  where  V  is  the  molar  flowrate  of  carrier  gas,  A  is  the  tower  
A
cross  sectional  area.  
  L   is   the   molar   flowrate   of   carrier   liquid   kmols/hr-­‐m2   of   tower   csa  
'
m
L
L'm = ,  where  L  is  the  molar  flowrate  of  carrier  gas,  A  is  the  tower  
A
cross  sectional  area.  
  KG   is  the  overall  gas  phase  mass  transfer  coefficient  (kmols/s-­‐m2-­‐mole  
ratio)  
  a   is  the  specific  surface  area  of  packing  (surface  area  per  unit  volume)  
  Y   is  the  mole  ratio  of  solute  in  the  gas  phase  
  Ye   is   the   mole   ratio   of   solute   in   the   gas   phase   which   would   be   in  
equilibrium  with  that  in  the  liquid  phase.  
  Y1,Y2   are  in  inlet  and  outlet  mole  ratios  of  solute  in  the  gas  phase  into  and  
out  of  the  tower.  
  X1,  X2   are  the  inlet  and  outlet  mole  ratios  of  solute  in  the  liquid  phase  into  
and  out  of  the  tower.  
 
In  essence,  the  tower  is  assumed  to  be  made  from  of  a  number  of  equilibrium  stages  
commonly   referred   to   as   transfer   units,   where   the   rate   of   mass   transfer   of   solute  
from   liquid   to   gas   is   constant   and   the   molar   flux   represented   by   equations   for   the  
form:  
N = K G ( Ye − Y)  
 
The   term   (Ye-­‐Y)   represents   the   driving   force   for   mass   transfer   in   that   particular  
element;   Y   is   the   concentration   (e.g.   mole   ratio,   partial   pressure   etc)   on   the   gas  
phase,  Ye  is  the  concentration/mole  ratio  is  the  gas  had  reached  its  equilibrium  value  
with  the  liquid  phase.  The  integral  term  in  the  number  of  transfer  units  expression  
indicates  that  the  driving  force  varies  from  the  top  of  the  tower  to  the  base  of  the  
tower.   Note   that   the   transfer   can   be   written   in   terms   of   solute   from   the   gas   into  
liquid.  It  can  also  be  given  in  terms  of  overall  mass  transfer  coefficients  for  the  liquid  
phase  e.g.  
 
N = K L ( X − X e )  
 
 

© School of Engineering and Physical Science Page 35

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 
  Gas Phase Liquid Phase
 
Direction of Transfer
 
 
 
Bulk Phase
 
concentration
 
  X
Interface Interface
  concentration concentration
  Yi Xi
 
 
At steady state in terms of
  Bulk Phase interface concentrations
  concentration
N=kL(X-Xi)=kG(Yi-Y)
  Y
  or in terms of equilibrium
  concentrations
  N=KoL(X-Xe)=KoG(Ye-Y)
  Two stagnant films on either
  side of interface
 
 
Figure  12  Model  of  the  Whitman  two  film  mass  transfer  across  a  phase  boundary,  
where  resistance  to  mass  transfer  is  contained  in  two  stagnant  films.  
 
The   use   of   NOG   for   the   number   of   transfer   units   indicates   that   the   mass   transfer  
coefficient  in  written  in  terms  of  overall  gas  phase  units.  Should  data  be  available  in  
terms  of  overall  liquid  phase  units,  then  the  symbol  NOL  is  likely  to  be  used  and  the  
equations  listed  above  changed  accordingly.  
 
The  principle  behind  mass  transfer  is  outlined  in  Figure  12  -­‐  the  Whitman  two  film  
theory   models   the   mass   transfer   process   as   transport   across   two   stagnant   films  
where   the   resistance   to   transfer   lies.   Depending   on   the   solubility,   resistance   to  
transfer   may   be   significant   in   one   of   the   films   and   often   this   dictates   which   set   of  
integral  sizing  equations  to  use  e.g  it  is  known  that  HCl  is  highly  soluble  in  water  and  
that  for  a  mixture  of  air  and  HCl,  as  soon  as  an  HCl  molecule  reaches  the  air/water  
interface  ,  it  will  quickly  be  absorbed  by  the  water.  In  this  case  the  resistance  to  mass  
transfer  lies  in  the  gas  layer  and  not  on  the  liquid  side.  
 
Depending   of   the   slope   of   the   equilibrium   line   between   gas   and   liquid   phase,   the  
number   of   transfer   units   integral   can   be   solved   in   several   ways.   If   the   equilibrium  
line  is  linear  and  the  flowrates  of  carrier  gas  and  liquid  do  not  vary,  then  the  number  
of   transfer   units   can   be   estimated   from   a   relatively   simple   integrated   formula.   For  
cases  where  the  equilibrium  line  is  non-­‐linear,  alternative  rigours  techniques  should  
be  used.  This  would  be  the  case  if  there  was  an  appreciable  temperature  difference  
across  the  column  which  would  affect  the  equilibrium  conditions.    

Page 36
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 
Assume  the  equilibrium  relationship  is  linear  and  of  the  form  Y=mX,  where  Y  refers  
to  the  mole  ratio  of  solute  in  the  gas  phase,  and  X  to  the  liquid  phase.  
 

 
 
 
By  a  simple  material  balance  round  the  lower  section  of  the  tower,  the  relationship  
between  solute  mole  ratio  in  the  gas  phase  to  that  in  the  liquid  phase  is  written  as  :    
G 'm
X = ' ( Y − Y1 ) + X1  
Lm
Since   the   two   molar   flowrates   are   presented   as   a   ratio,   it   is   not   necessary   at   this  
stage  to  know  the  cross  sectional  area  of  the  tower.  
'
mGm
Multiplying  by  m  gives  :   mX = (Y − Y1 ) + mX1  
L'm
'
mGm
Hence  :   Ye = (Y − Y1 ) + mX1  
L'm
Replacing  for  Ye  in  the  integral  expression  gives  :  
Y2
dY
NoG   =   ∫Y ⎛ ⎛ mG' ⎞ ⎞
 
⎜⎝ ⎜⎝ L' ( Y − Y1 ) + mX1 ⎟⎠ − Y⎟⎠
1 m

m
Y1
dY
ð NoG   =   ∫⎛ mG '
mG 'm ⎞
 
⎜⎝ Y − L' Y + L' Y1 − mX1 ⎟⎠
Y2 m

m m

© School of Engineering and Physical Science Page 37

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
Y1
dY
ð NoG   =   ∫ ⎛⎛
Y2 ⎜ ⎜ mG' ⎞ mG' ⎞
 
1 − ' m ⎟ Y + ' m Y1 − mX1 ⎟
⎜⎜ L m ⎟⎠ Lm ⎟
⎝⎝ ⎠

This  is  of  the  form    

Y2
⎡ ⎤
⎢ ⎥
⎢ 1 ⎡⎛ mG' ⎞ mG ' ⎤⎥
ð NoG   =   ⎢ In⎢⎜1 − ' ⎟ Y + ' Y1 − mX1 ⎥ ⎥  
m m
⎛ ' ⎞ ⎜ L m ⎟⎠
⎟ ⎢⎣⎝ ⎥⎦ ⎥
⎢⎜ mG m Lm
− 1
⎢ ⎜ L' ⎟ ⎥
Y1 ⎣ ⎝ m ⎠ ⎦

1
ð NoG   = x  
⎛ mGm' ⎞
⎜ − 1⎟
⎜ L' ⎟
⎝ m ⎠

⎡ ⎡⎛ mG ' ⎞ mGm ' ⎤ ⎡⎛ mGm '

⎞ mGm ' ⎤⎤
⎢In⎢⎜1 − ' m ⎟ Y2 + Y − mX ⎥ − In ⎢ ⎜ 1 − ⎟ Y1 + Y − mX 1 ⎥ ⎥  
⎢⎣ ⎣⎢⎜⎝ ⎟ ⎜ ⎟
1 1 1
Lm ⎠ L'm ⎦⎥ ⎣⎢⎝ L'm ⎠ L'm ⎦⎥ ⎥⎦
⎡ ⎛ mG m '
⎞ mG m' ⎤
⎢ ⎜⎜1 − ' ⎟ Y2 +
⎟ '
Y1 − mX 1 ⎥
⎢ Lm Lm ⎥
x In ⎢ ⎝ ⎠
1
NoG   = ⎥  
⎛ mGm' ⎞ Y1 − mX 1
⎜ − 1⎟ ⎢ ⎥
⎜ L' ⎟ ⎢ ⎥
⎝ m ⎠ ⎣ ⎦
1 ⎡⎛ mG 'm ⎞ ( Y2 − mX1 ) mG 'm ⎤
ð NoG   = x In ⎢⎜ 1− ' ⎟ + ' ⎥  
⎛ mGm
⎜
' ⎞
− 1⎟ ⎣ ⎝ L m ⎠ ( Y1 − mX1 ) Lm ⎦
⎜ L' ⎟
⎝ m ⎠
A  useful  method  when  the  equilibrium  line  is  non-­‐linear,  is  the  log-­‐mean  method  of  
determining   the   number   of   transfer   units.   The   derivation   which   follows,   uses   the  
liquid  side  equations  to  derive  the  number  of  transfer  units.  
 
The  overall  height  of  packing  in  this  case  is  given  by:    
 
Overall  height  is  given  by  :   Z=HOLxNOL  
X1
dX
Where   NOL    is  the  number  of  transfer  units,   NOL = ∫ (X − X
X2 e )
 

L'm
  HOL     is  the  height  of  a  transfer  unit,   H OL =  
aK L
(Note  the  difference  in  the  terminology)  
 
Replacing  the  integral  by  its  log-­‐mean  average  gives:  

N OL =
( X2 − X1 )  
( X − Xe )LM

Page 38
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
where   (X − X e )LM =
(X 2 − X 2e ) − (X1 − X1e )  
⎛ X − X 2e ⎞
ln⎜⎜ 2 ⎟
⎟
⎝ X 1 − X 1e ⎠
The  equilibrium  values  are  derived  from  :   Xe = Y / m  
Thus:  
⎛ Y2 ⎞ ⎛ Y1 ⎞
⎜⎝ X 2 − ⎟⎠ − ⎜⎝ X1 − ⎟⎠
( X − Xe )LM = m m
 
⎛ Y 2 ⎞
X2 −
ln ⎜ ⎜ m⎟
Y1 ⎟
⎜⎝ X1 − ⎟⎠
m
 
The   size   of   the   mole   fractions   being   used   implies   that   NOG   and   NOL   will   have   very  
different   numerical   values.   Therefore,   the   height   of   a   transfer   unit   varies  
considerably   depending   on   the   approach   used;   overall   gas   phase   or   liquid   phase  
equations.  
 
Example  
Determine   the   number   of   transfer   units   in   terms   of   both   gas   and   liquid   phase  
equations  for  the  following  duty.  
 
A  packed  column  is  used  to  treat  6  m3/hr  of  water  containing  0.008  mole%  oxygen  
using   380   m3/hr   of   stripping   gas   which   is   oxygen   free.   The   column   is   designed   to  
reduce   the   oxygen   content   to   8x10-­‐7   mole%   and   operates   at   1   atm   absolute  
pressure.  Take  Henry’s  law  constant  to  be  3.3x104  mol/atm,  density  of  water  as  1000  
kg/m3,  density  of  gas  as  0.8  kg/m3,  average  molar  mass  of  gas  as  16  kg/kmol  
 
Solution  
The  information  given  and  any  additional  assumed  values  are:  

Liquid properties Gas Properties

3 3
m m
Volumetric flowrate Qwater 6. Qgas 380.
hr hr

kg kg
Mass Density ρ water 1000. ρ gas 0.7.
3 3
m m

kg kg
Molar mass MMwater 18. MMgas 16.
kmol kmol
 
 
 
 
 
 
 

© School of Engineering and Physical Science Page 39

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 
 
 
Step  1:  Find  the  molar  flowrates  of  the  gas  and  water  streams  
 
kg
Mass flowrate of water is Gwater Qwater.ρ water Gwater = 1.667
s
Gwater kmol
Molar flowrate of water is L L = 0.093
MM water s

kg
Mass flowrate of gas is Ggas Qgas.ρ gas Ggas = 0.074
s
0.008
x = 8 .10
5
Mole fraction of oxygen in water stream
Ggasto be treated is x2
. 3 kmol 2
Molar flowrate of gas is G G = 4.618 10100
MM gas s
8 .10
7
x1 = 8 .10
9
Mole fraction of oxygen in treated water is to be x1
100

Mole fraction of oxygen in the stripping gas y1 0

Step   2:   Use   the   operating   line   expression   to   find   the   oxygen   in   the   stripping   gas  
leaving  the  column  
L.
y2 = 1.604.10
3
y2 x2 x1 y1
G
The   small   values   for   mole   compositions   imply   that   either   mole   fractions   or   mole  
ratios  can  be  used  in  this  balance.  
 
Step  3:  Find  the  number  of  transfer  units  by  the  simplified  integral  equation:  
 
x1 = 8 .10 x2 = 8 .10 h .x1 = 2.64.10 h .x2 = 2.64
9 5 4

G
y 2 = 1.604.10 h. = 1.646.10
3 3
y1 = 0
L

Number of transfer units is then

1 G . y2 h .x1 G
Nog .ln 1 h. h. Nog = 5.599.10
3

.G L y1 h .x1 L
h 1
L

y1
x2
1 .ln 1 L . h L
Nol Nol = 9.215
L h .G y1 h .G
1 x1
h .G h
 
It   is   clear   from   this   result   that   application   of   these   sizing   equations   must   be  
consistent  in  subsequent  parts  of  the  calculation.  
 

Page 40
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
An   alternative   to   the   simplified   integral   equation   above,   the   log-­‐mean   method   can  
often  be  a  quicker  method  to  apply:  
 
Step  3:  Find  the  number  of  transfer  units  by  log-­‐mean  method:  
As  in  the  previous  step,  this  calculation  shows  both  results  from  the  gas  and  liquid  
phase  sides  of  the  calculation.  
 
y1
x1 = 8 .10
9
y1 = 0 xe1 xe1 = 0
h
y2
x2 = 8 .10 y 2 = 1.604.10 xe2 = 4.86.10
5 3 8
xe2
h

x2 x1
Nol Nol = 9.215
x2 xe2 x1 xe1
x2 xe2
ln
x1 xe1
 
Repeat  the  calculation  in  terms  of  gas  phase  units  
x1 = 8 .10 h .x1 ye1 = 2.64.10
9 4
y1 = 0 ye1

y 2 = 1.604.10 x2 = 8 .10 h .x2

3 5
ye2 ye2 = 2.64

y2 y1
Nog = 5.599.10
3
Nog
ye2 y2 ye1 y1
ye2 y2
ln
ye1 y1
 
It   should   be   no   surprise   that   the   two   methods,   simplified   integral   and   log-­‐mean  
produce  the  same  numerical  results.  
 
It  is  useful  to  consider  the  meaning  of  the  driving  force  for  mass  transfer  at  the  top  
and   bottom   of   the   column.   Figure   13   shows   graphically   the   relationships   between  
solute  mole  ratio  in  the  gas  and  liquid  phases,  where  solute  is  transferring  from  the  
liquid  phase  into  the  gas  phase.  

© School of Engineering and Physical Science Page 41

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 

 
Figure  13a  Operating  and  equilibrium  lines  for  a  gas-­‐liquid  stripping  column  
 
 
Figure  13a  shows  the  standard  situation  with  the  driving  force  for  mass  transfer  at  
the   top   and   bottom   of   the   tower   is   given   by   (Y2e − Y2 )   and   (Y1e − Y1 ),   of    
respectively.   Similar   diagrams   can   be   constructed   using   driving   force   expressions  
described  in  terms  of  the  liquid  phase.  For  any  particular  operation,  which  involves  
removing  material  from  the  water  phase,  Y2e  and  Y1e  are  fixed  due  to  the  equilibrium  
line,  since  X2   and  X1  are  fixed  by  the  requirements  for  the  tower.   L'm  will  be  fixed  as  
will   Y1;   note   it   is   normal   for   the   stripping   gas   not   to   contain   any   of   the   solute  
removed  from  the  liquid  however  in  cases  where  the  stripping  gas  is  processes  and  
recycled  this  will  not  always  be  the  case.  The  only  variable  to  alter  and  hence  find  a  
value  for  is   G m' ,  the  flowrate  of  stripping  gas.  Looking  at  the  previous  diagram,  if  the  
gas   flowrate   were   reduced   the   slope   of   the   operating   line   would   be   increased   hence  
(Y2e − Y2 )  would  reduce.  If   (Y2e − Y2 )  became  zero,  this  would  mean  there  was  no  
mass  transfer  taking  place  at  the  top  of  the  tower.    
 

Page 42
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

 
Figure  13b  Operating  and  equilibrium  lines  for  a  gas-­‐liquid  stripping  column  at  
minimum  gas  rate  conditions  
 
If  the  rate  was  reduced  further,  then  the  operating  line  expression  implies  transfer  
will  be  in  the  opposite  direction.  Thus,  there  is  an  optimum  combination  of  flowrates  
for  a  particular  absorption  column.  For  absorption  of  solute  from  gas  to  liquid,  the  
operational  diagram  has  the  equilibrium  and  operating  lines  in  reverse  positions:  

 
Figure  14  Operating  and  equilibrium  lines  for  a  gas-­‐liquid  absorber  
 
 
In  this  case,  Y1,  X1,   G m' ,  X2  are  fixed  which  means  the  liquid  rate  can  vary.  If  the  liquid  
rate  reduced  the  slope  of  the  operating  line  reduces  and  there  is  a  chance  the  driving  
force  at  the  bottom  of  the  tower  reduced  to  zero.    
 

© School of Engineering and Physical Science Page 43

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 
It   is   generally   accepted   for   absorption   columns   that   the   actual   liquid   flowrate   be   1.3  
to1.5  x  the  minimum  rate.  This  suggests  that  for  stripping  columns:  
 
( ) ( )
G 'm actual = 1.3 to 1.5x G 'm min  
 
While   this   value   is   useful   for   a   first   approximation,   the   physical   operation   of   the  
column  requires  that:  
• Liquid  must  wet  the  packing  surface  allowing  the  gas  to  come  in  contact  with  
the   maximum   available   surface   area.   At   high   liquid   to   gas   ratios,   this  
condition  will  not  be  met  and  the  packing  is  said  to  become  flooded.  
• The  pressure  drop  across  the  packing  must  be  relatively  low.  
Selection   of   the   stripping   gas   flowrate   in   reality   is   determined   by   the   hydraulic  
operating  limits  of  the  column.  
 
2.  Hydraulic  Characteristics.  
Flooding   is   an   observable   feature   of   liquid   and   gas   flow   in   packed   columns.   If   the  
liquid   is   held   up   within   the   packing,   the   voids   between   packing   become   filled   with  
liquid  and  the  gas  is  then  dispersed  within  the  liquid.  At  this  point,  the  packing  is  not  
utilised  for  mass  transfer  and  hence  the  efficiency  for  the  process  will  depend  on  the  
gas  bubble  size  rather  than  the  type  and  nature  of  the  packing.  The  increased  hold  
up  of  liquid  within  the  column  will  lead  to  increased  pressure  drop  for  the  gas  flow.  
Sizing  procedures  for  packed  columns  therefore  involve  checking  the  ratio  of  gas  to  
liquid  flowrate  with  a  given  diameter  against  the  flooding  capacity  of  the  column  and  
its  packing.    
 
The   Generalised   Pressure   Drop   Correlation   (GPDC)   was   developed   out   of  
observations   that   for   two   phase   flow   through   towers   filled   with   random   packings,  
flooding  occurred  at  a  column  pressure  drop  of  1.7  –  3  5  mm  of  water/cm  of  packing  
(2   –   3.5   in   of   H2O/ft   of   packing)(9).   Figure   15,   adapted   from   Coulson   &   Richardson  
Volume   6(8),   illustrates   the   GPDC   chart   used   in   sizing   packed   columns.   The   chart  
shows   the   normal   operating   ranges   for   gas   and   liquid   flows   for   various   levels   of  
flooding  and  pressure  drop  across  a  unit  of  packing.  
 

Page 44
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

 
Figure  15  Generalised  Pressure  Drop  Correlation  for  Packed  Columns(8)  
 
 
0.1
⎛µ ⎞
13.1( Vw ) Fp ⎜ l ⎟
2

⎝ ρl ⎠
Note   the   term   K 4 =     is   the   “Flow   Capacity   Factor”,   plotted  
ρv ( ρl − ρv )
Lw ρv
against  the  “Flow  factor”   FLV =  ,  for  various  value  of  pressure  drop.    
Vw ρ l
Where   Vw   =  gas  mass  flowrate  per  unit  cross  sectional  area  (kg/m2-­‐s)  
  Lw   =  liquid  mass  flowrate  per  unit  csa  (kg/m2-­‐s)  
Fp = the “Packing Factor” which is a function of the type and size of
packing used.
ρ, µ = the density and viscosity of the liquid and vapour phases.

The  procedure  to  use  this  type  of  chart  is  as  follows(8):  

© School of Engineering and Physical Science Page 45

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
Step   1:   Calculate   the   flow   factor   for   the   given   liquid   and   vapour   rates.   Note   that   the  
flow  factor  does  not  require  the  column  cross  sectional  area  and  the  ratio  of  the  two  
mass  flowrates  equals  the  ratio  of  the  two  “flowrate  per  unit  area”.  
 
Step  2:  Select  an  appropriate  pressure  drop  per  unit  height  of  packing  e.g.  absorbers  
and   stripping   columns   usually   operate   with   pressure   drops   15   to   50   mm/m   of  
packing  and  a  lower  value  is  preferred.    
 
Step  3:  Find  the  flow  capacity  factor  K4,  and  hence  calculate  the  gas  mass  flowrate  
per  unit  area  Vw.    
 
Step  4:  Calculate  the  column  diameter  and  hence  cross  sectional  area.  Select  a  size  
which  is  suitable  for  construction.  
 
Step   5:   Calculate   the   percentage   flooding   by   first   finding   the   K4   value   at   flooding,  
then  using  the  relation:  
K 4 at design pressure drop
% flooding = 100  
K 4 at flooding
In  addition  to  the  flooding  limits,  there  are  recommended  ratios  of  column  diameter  
to   packing   size   and   for   the   typical   heights   of   packing   sections,   to   ensure   adequate  
distribution  of  liquid  e.g(10).  
 
  Raschig  Rings   Berl  &  Intalox   Pall  Rings  
Saddles  
Ratio   column   diameter   to   30   15   10  to  15  
packing  size  
Recommended   heights   of   6  m  or  2.5-­‐3x   6  m  or  5–8x   6  m  or  5-­‐10x  
packing   sections   for   column   column  diameter   column  diameter  
redistribution   diameter  
 
Updates   to   the   GPDC   graph   appear   in   the   literature   from   time   to   time   with  
corrections   for   systems   other   than   air   and   water.   It   is   not   unusual   to   find   the   flow  
factor   ordinate   extend   to   10   which   indicates   an   upper   limit   to   the   liquid/gas   ratio  
which  are  commonly  found  in  packed  columns.    
 
3.  Height  of  a  Transfer  Unit  
Finding   the   overall   height   of   a   transfer   unit   can   be   troublesome;   there   is   no   exact  
method  for  new  situations  and  if  at  all  possible,  data  collected  from  existing  towers  
should   be   used.   The   height   of   a   transfer   unit   depends   on   the   gas   &   liquid   molar  
velocities  (kmols/m2-­‐s),  the  mass  transfer  coefficient  KG  and  the  specific  surface  area  
of  the  packing  and  hence  depends  on  the  nature  of  the  packing  used,  the  operating  
conditions   and   distribution   of   liquid   and   gas   in   the   packing   section.   Several   methods  
can  be  used  to  find  the  value  of  HTU:  
 
• Correlations  in  equation  or  graphical  format  of  KGa,  KLa  values  against  gas  and  
liquid   flowrates   for   specific   chemical   systems.   Some   vendors   will   show  

Page 46
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
correlations   for   their   packings   for   well-­‐known   chemical   systems.   Dimensional  
similarity  techniques  can  be  used  to  convert  these  into  other  systems.  
• Recommended  values  –  typical  HTU  values  for  random  packings  range  from  
0.3  to  0.1  m  for  packings  25mm  to  50mm  in  size(8)  
• Empirical   equations   –   several   methods   have   been   published   outlining  
techniques  to  predict  HL  and  HG  values  (heights  of  packing  in  terms  of  liquid  
and   gas   phase   resistance   terms).   Conversion   to   the   HTU   value   in   terms   of  
overall  gas  phase  units  is  achieved  be  realising  that:  
G 'm
H OG = H G + m H L  
L' m
L' m
H OL = H G + H L  
mG 'm
where  m  is  the  slop  of  the  equilibrium  line  
 
As  an  illustration  of  empirical  equations,  the  example  outlined  below  has  been  taken  
from   Coulson   &   Richardson   Vol   6(8)   which   is   based   on   Cornell’s   method   to   predict  
HTU  values  for  the  gas  and  liquid  films.  Two  empirical  expressions  form  the  basis  of  
the  technique:  
 
The  two  film  HTU  values,  one  for  gas  the  other  for  the  liquid  film  are  give  as:  
1.11 0.33
0.5 ⎛ D ⎞ ⎛ Z ⎞
0.011ψ n ( Sc )V ⎜
⎝ 0.305 ⎟⎠ ⎜⎝ ⎟
3.05 ⎠
HG =  
( L w f1f2 f3 )0.5
0.15
⎛ Z ⎞
H L = 0.305ϕ h ( Sc )L K 3 ⎜
0.5
 
⎝ 3.05 ⎟⎠
where   HG   HTU  for  the  gas  film  (m)  
HL   HTU  for  the  liquid  film  (m)  
  ψ n   factor  from  Figure16a  
  φ h   factor  from  Figure16b  
µv µ
  Sc   Schmidt  number,   Sc v = for  the  gas  and   Sc L = L for  the  liquid  
ρ vDv ρ L DL
  D     column  diameter  
  Z     height  of  packing.  It  is  usual  to  estimate  this  value  first  then  iterate  until  
the  specified  and  calculated  height  agree.  
  K3   factor  related  to  the  %’age  flooding  from  Figure16c  
  f1,  f2  and  f2  are  ratios  of  the  viscosity,  density  and  surface  tension  of  the  liquid  
phase  in  the  application  to  that  of  water.    
 

© School of Engineering and Physical Science Page 47

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 

 
Figure  16a  Factor   ψ h  (C&R  Vol  6)  (8)  
 

 
Figure  16b  Factor   φ h  (C&R  Vol  6)  (8)  
 

Page 48
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 

 
Figure  16c  K3  factor  (C&R  Vol  6)  (8)  
 
 
The  graphs  used  to  find  the  constants  for  the  two  HG  and  HL  expressions  depend  on  
the   type   of   packing;   the   graphs   given   here   apply   to   Berl   saddles   which   will   give  
conservative   estimates   should   packings   with   higher   mass   transfer   abilities   e.g.   Pall  
rings  or  Interlox  Saddles  be  used.  
 
Several   vendors   offer   both   structured   and   random   packings   for   their   gas-­‐liquid  
contacting   columns.   Some   of   these   include:   Sluzer,   Kkoch-­‐Glitsch   and   Jaeger.  
Structured   packings   offer   significantly   lower   pressure   drops   than   random   packings.  
Figure17  illustrates  several  forms  of  structured  packings  
 

   
Figure  17  Examples  of  Metal  Structure  packings  for  gas-­‐liquid  contacting  columns  
   
With   methods   similar   to   Cornell’s   technique,   it   is   necessary   to   find   the   diffusion  
coefficient  in  liquid  and  gas  phases  of  the  solute  transferring,  along  with  the  other  
physical   properties.   Liquid   diffusivity   can   be   predicted   by   the   Wilke-­‐Chang  
correlation(8):  
© School of Engineering and Physical Science Page 49
 
 OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
1.173x10 −13 (ϕ M ) T
0.5

DL =  
µVm0.6
where   DL   liquid  diffusivity  (m2/s)  
  φ   association  factor  for  the  solvent  
    =2.6  for  water  (could  be  2.26)  
    =  1.9  for  methanol  
    =  1.5  for  ethanol  
    =  1.0  for  un-­‐associated  solvents  
  µ   Molar  mass  of  the  solvent  
  M   viscosity  of  the  solvent  (mNs/m2)  
  T   absolute  temperature  (K)  
  Vm   molar  volume  of  solute  at  its  boiling  point  (m3/kmol)  
 
Molar  volumes  of  selected  solutes  are  given  below(8)  
Molecule   Air   Cl2   CO2   NH3   H2O   NO   O2   SO2  
Molar  Volume  
0.0299   0.0484   0.0340   0.0258   0.0189   0.0236   0.0256   0.0448  
(m3/kmol)  
 
Example  
Calculate  the  diffusivity  of  oxygen  in  water  at  25°C.  
 
Solution  
Molar  volume  of  the  solute  is  0.0256  m3/kmol,  molar  mass  is  18  kg/kmol.  Viscosity  of  
the  solvent  phase  (water)  depends  on  the  temperature  but  may  be  estimated  from:  
µ = −0.4767 ln(T ) + 2.4187  
where   T  is  the  centigrade  temperature  
Note:  correlation  estimated  from  data  at  1atm,  over  range  10  –  70°C  
Viscosity  of  the  solvent  phase  is  then:   µ = 0.884 cP  
1.173x10 −13 ( 2.6x18 ) ( 25 + 273)
0.5

Diffusivity  is  therefore:   D L = =2.44x10-­‐9  m2/s  

0.884x0.0256 0.6
Due   to   the   effect   of   temperature   on   viscosity,   diffusion   of   solute   may   be   a   strong  
function  of  temperature  e.g  using  the  procedure  in  the  example,  diffusivity  of  water  
varies  from  1.55x10-­‐5  to  4.22x10-­‐5  cm2/s  for  a  temperature  10°C  to  50°C  
For  gases,  estimates  can  be  made  from  the  following:  
0.5
−7 ⎛ 1 1 ⎞
1.013x10 T 1.75
+
⎜⎝ M M ⎟⎠
DV = a b
 
1/3 2
⎡⎛ ⎞
1/3
⎛ ⎞ ⎤
P ⎢ ⎜ ∑ vi ⎟ + ⎜ ∑ vi ⎟ ⎥
⎢⎣⎝ a ⎠ ⎝ b ⎠ ⎥⎦
This   uses   a   contribution   method   for   the   molar   volumes   of   the   continuous   phase  
(phase  b)  and  the  diffusing  phase  (phase  a).    
where   DV     =  the  diffusivity  (m2/s)  
  T     =  absolute  temperature  (K)  
  Ma,  Mb   =  molar  masses  of  phases  a  and  b  
  P   =  total  pressure  (bar)  
Page 50
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
  ∑ v i   =  the  sum  of  diffusive  volume  coefficients  for  phase  n  
n
Diffusive  volume  coefficients  for  selected  gases  are:  
Molecule   N2   O2   Air   CO2   N2O   NH3   Cl2   SO2  
Diffusive  volume   17.9   16.6   20.1   26.9   35.9   14.9   37.7   41.1  
 
For   hydrocarbons,   the   molar   volume   contributions   come   from   the   structure   of   the  
molecule:  
Molecule   C   H   O   N   Cl   C   Rings  
Diffusive  volume   16.5   1.98   5.48   5.69   19.5   17.0   -­‐20  
 
Example  
Estimate  the  diffusivity  of  oxygen  in  air  at  0°C,  1  bar  pressure  
Molar  masses  are  Ma=32  kg/kmol,  Mb=29  kg/kmol.  Molar  volume  contributions  are  :  
∑ v i = 16.6 ,  and   ∑ v i = 20.1  
a b
0 .5
⎛ 1 1 ⎞
1.013 x10 −7 (0 + 273 )1.75 ⎜ + ⎟
⎝ 32 29 ⎠
DV = =1.715x10-­‐5  m2/s  
[
1 (16 .6 ) + (20 .1)
1/ 3 1/ 3 2
]
Note  at  10°C,  this  method  gives  1.826x10-­‐5  m2/s  
 
Example  
Estimate  the  diffusivity  of  oxygen  in  methane  at  10°C,  1  bar  pressure  
Solution  
Molar  masses  are  Ma=32  kg/kmol,  Mb=16  kg/kmol.  Molar  volume  contributions  are  :  
∑ v i = 16.6 ,  and   ∑ v i = 16.5 + 4x1.98 = 24.42  
a b
0 .5
⎛ 1 1⎞
1.013 x10 −7 (10 + 273 )1.75 ⎜ + ⎟
DV = ⎝ 32 16 ⎠ =2.037x10-­‐5  m2/s  
[
1 (16 .6 ) + (24 .42 )
1/ 3 1/ 3 2
]
 
Alternatively,   values   may   be   taken   from   published   tables   e.g.   Table   9   lists   selected  
diffusivities  of  compounds  in  selected  gases  and  liquids.  
 
 
 
 
 
 
 
 
 
 
 
 

© School of Engineering and Physical Science Page 51

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
In  Gases  at  0°C   cm2/s    
In  Liquids  at  25°C   cm2/sec  x105  
Acetone  in  air     0.109    
Acetic  Acid  in  water   1.24  
Benzene  in  air     0.077    
Ammonia  in  water   2  
CO2  in  air     0.138    
Benzene   (50   mole%)   in   2.47  
heptane  
CO2  in  Oxygen   0.139     CO2  in  water   1.96  
CO2  in  Methane   0.153     CO2  in  ethanol   4.0  
O2  in  air     0.178     Chlorine  in  water   1.44  
      Hydrogen  Sulphide  in  water   1.61  
      Nitrogen  in  water   1.9  
      Methanol  in  water   1.6  
      Oxygen  in  water     2.5  
      SO2  in  water   1.7  
      Urea  in  water   1.37  
Table  9:  Diffusivities  of  selected  components  
 
Example  
Determine   the   number   of   transfer   units   in   terms   of   both   gas   and   liquid   phase  
equations  and  the  corresponding  height  of  packing,  for  the  following  duty.  
A  packed  column  is  used  to  treat  6  m3/hr  of  water  containing  0.008  mole%  oxygen  
using   380   m3/hr   of   stripping   gas   which   is   oxygen   free.   The   column   is   designed   to  
reduce   the   oxygen   content   to   8x10-­‐7   mole%   and   operates   at   1   atm   absolute  
pressure.  Take  Henry’s  law  constant  to  be  3.3x104  mol/atm,  density  of  water  as  1000  
kg/m3,  density  of  gas  as  0.8  kg/m3,  average  molar  mass  of  gas  as  16  kg/kmol.  Take  
the  height  of  a  transfer  unit  as:  HL=0.216  m,  and  HG=0.3  m  
 
Solution  
Following  from  the  previous  example,  find  the  molar  flowrates,  the  oxygen  in  the  gas  
leaving  the  column  and  then  the  number  of  transfer  units  
Liquid properties Gas Properties
3 3
m m
Volumetric flowrate Qwater 6. Qgas 380.
hr hr

kg kg
Mass Density ρ water 1000. ρ gas 0.7.
3 3
m m

kg kg
Molar mass MMwater 18. MMgas 16.
kmol kmol
 
 
 
 
 
 
 

Page 52
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
0.008
x2 = 8 .10
5
Mole fraction of oxygen in water stream to be treated is x2
100

8 .10
7
x1 = 8 .10
9
Mole fraction of oxygen in treated water is to be x1
100

Mole fraction of oxygen in the stripping gas y1 0

 
 
 
The  height  of  a  transfer  unit  and  height  of  packing  is  therefore:  
L .
Hol Hg Hl Hol = 0.216 m
y1
h .G9
x1 = 8 .10 y1 = 0 xe1 xe1 = 0
h
Z [Link] Z = 1.992 m y2
x2 = 8 .10 y 2 = 1.604.10 xe2 = 4.86.10
5 3 8
xe2
h
L.
y2 = 1.604.10
3
y2 x2 x1 y1
G
x2 x1
Nol Nol = 9.215
x2 xe2 x1 xe1
x2 xe2
ln
x1 xe1
In   terms   of   gas   phase   units,   the   calculation   proceeds   from   the   number   of   transfer  
units  as:  
x1 = 8 .10 h .x1 ye1 = 2.64.10
9 4
y1 = 0 ye1

y 2 = 1.604.10 x2 = 8 .10 h .x2

3 5
ye2 ye2 = 2.64

y2 y1
Nog = 5.599.10
3
Nog
ye2 y2 ye1 y1
ye2 y2
ln
ye1 y1
 
The  height  of  a  transfer  unit  and  height  of  packing  is  therefore:  
G
Hog Hg h . .Hl Hog = 355.809m
L

Z Hog .Nog Z = 1.992 m

 
The  two  routes  give  the  same  result,  as  would  be  expected.  
 

5.4.2 Maldistribution
In   terms   of   operation   of   packed   columns   along   with   gas/liquid   ratio   another   issue  
which   concerns   the   performance   of   packed   columns   is   the   distribution   of   liquid   as   it  

© School of Engineering and Physical Science Page 53

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
travels  down  through  the  packing.  As  liquid  travels  over  the  surface  of  the  packing,  
preferential   flow   towards   the   side   walls   can   take   place   and   hence   by-­‐pass  
subsequent  sections  of  packing.  Re–distributing  flow  is  therefore  useful  to  ensure  all  
the  packing  is  used.  Packed  columns  are  therefore  divided  into  a  number  of  sections,  
with   liquid   collection   and   redistribution   between   sections.   Figure18   illustrates   a  
tower  with  two  packing  sections.  In  addition  to  re-­‐distribution,  many  vendors  supply  
proprietary  distribution  systems.    
 
 

 
Figure  18:  Schematic  representation  of  two  packing  sections  in  a  packed  column  gas-­‐
liquid  contacting  column  
 
Example  
To  conclude  the  discussion  on  gas  stripping,  the  following  illustration  highlights  some  
of   the   issues   involved.   In   the   following   example,   oxygen   is   stripped   from   water   by  
contacting  with  natural  gas  under  the  assumptions  that:  
• Temperature  is  constant  at  10°C  allowing  plastic  pall  rings  to  be  used.  
• Henry’s  law  applies  
• Oxygen  is  the  only  component  undergoing  transfer.  
 
The  data  for  this  problem  was  presented  earlier  and  is  provided  for  completeness.    
 
Example  
Calculate  the  number  of  transfer  units  required  to  reduce  the  oxygen  concentration  
of   1000   bpd   water   from   saturation   conditions   to   10   ppb   using   natural   gas   as   the  
stripping  media.  Assume  the  column  operates  at  atmospheric  pressure  and  10°C.    

Page 54
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
Step   1:   Convert   the   units   into   SI,   assuming   that   the   gas   flowrate   is   given   at  
operational  conditions:  
3
m
Water flowrate is Qwater [Link] Qwater = 6.621
hr
Inlet oxygen mass flowrate is p2 [Link]
Exit oxygen concentration is p1 [Link]

Physical properties at operating conditions

kg
Water density ρ water 1000. water viscosity µ water [Link]
3
m
kg
Water molar mass is Mwtl 18.
kmol
 
kg
Mass flowrate of carrier water in the tower is Lw Qwater.ρ water Lw = 1.839
s
Lw kmol
Molar rate of carrier water is L L = 0.102
Mwtl s
 
Gas  properties  are  
Gas density can be found from the ideal gas law as

Operating pressure P 1 .atm operating temperature T 293.K

kg
Gas molar mass is Mwtg 16.
kmol

Mwtg .P kg lb
so the density is ρ gas ρ gas = 0.666 ρ gas = 0.042
Rc.T m
3 3
ft
 
 
Step  2:  Find  the  concentration  of  solute  in  terms  of  mole  ratios  
Many  of  the  equations  for  absorption  columns  are  derived  in  terms  of  mole  fractions  
and   there   is   often   issues   in   determining   the   mean   flowrate   of   liquid   and   gas.   For  
dilute  systems,  there  should  not  be  a  considerable  change  in  bulk  flow  from  the  top  
to  bottom  of  the  column.  Mole  ratios  offer  a  useful  alterative.  

© School of Engineering and Physical Science Page 55

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
kg
Molar mass of solute 32.
Mwts
kmol
5 kg
Mass flowrate of solute in the water feed is Mass p 2 .Lw Mass = 2.207.10
s
Mass kmol
Molar = 6.897.10
7
Molar rate of solute into the tower is Molar
Mwts s
kg
p 1 .Lw .10 6.10
Molar 8
Massratio
Mole flowrate of solute
of solute in theinwater
to water, feed outlet
is is Mass
X2 Mass = 1.839
X2 = 6.75
L s

Mass kmol
Molar = 5.747.10
10
Molar rate of solute into the tower is Molar
Mwts s

Molar
X1 = 5.625.10
9
Mole ratio of solute to water in feed is X1
L
 
 
Step  3:  Determine  the  minimum  gas  flowrate  
Henry’s   law   coefficient   for   oxygen   in   water   at   the   operating   temperature   is  
H = 3.27x10 4 atm / mole fraction,  hence:  
yPT = 3.27x10 4 x  
For  the  operating  pressure  of  1  atm,  the  gradient  of  the  equilibrium  line  is:  
32700
h=  
1
 
The   minimum   gas   flowrate   condition   occurs   when   the   gas   leaving   the   column   is   in  
equilibrium   with   the   incoming   water   stream.   The   equilibrium   mole   fraction   of  
oxygen  in  the  methane  is:  
X2 = 6.75.10
6
Mole fraction of solute in the liquid into the tower is

Mole ratio of solute in gas phase in equilirum is Y2 h .X2 Y2 = 0.221

 
X2 X1 kmol
L. Gmin = 3.122.10
6
Minimum gas molar rate is therefore Gmin
Y2 Y1 s
As  a  first  try,  estimate  the  actual  gas  flowrate  as  1.3x  the  minimum  flowrate:  
6 kmol
G [Link] G = 4.059.10
s
5 kg
Mass flowrate of gas is therefore Gw [Link] Gw = 6.494.10
s

Lw . ρ gas
Calculate the flow factor Flv Flv = 730.632
Gw ρ
water
 
The   flow   factor   value   suggests   the   gas   rate   is   too   low   for   this   application   and   should  
be  increased.  

Page 56
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
kg
Set the gas flowrate Gw 0.1.
s

Lw . ρ gas
Calculate the flow factor Flv Flv = 0.474
Gw ρ
water
 
Step   4:   Find   the   oxygen   mole   ratio   in   the   gas   stream   leaving   the   column   for   the  
specified  gas  flowrate:  
Gw 3 kmol
Molar rate of carrier gas is G G = 6.25.10
Mwtg s
L
Ratio of molar rate of liquid to gas = 16.348
G

Mole ratio of oxyen leaving the tower in the gas phase is then
L.
Y2 = 1.103.10
4
Y2 X2 X1 Y1
G
 
The   equilibrium   line-­‐operating   line   plot   for   this   problem   is   shown   below;   the  
operating   line   at   1.3   x   the   minimum   gas   flowrate   is   also   shown   to   illustrate   the   issue  
with  driving  forces  at  the  top  and  bottom  of  the  tower.  As  the  gas  fowrate  increases,  
the   slope   of   the   operating   line   decreases   and   in   this   particular   example,   runs   almost  
in   line   with   the   x   axis.   The   driving   force   for   mass   transfer   at   the   top   of   the   column   is  
therefore  very  large  and  the  number  of  transfer  units  will  be  small.  
 
For  absorption  problems,  the  operating  line  will  be  above  the  equilibrium  line.  
 
0.250
 
  Y2e
 
0.200
 
Equilibrium line
Vapour Phase mole ratio

 
 
  0.150

 
 
  0.100
  1.3xGmin
 
  0.050
 
  Chosen operating line X2
  0.000
  0.00 0.02 0.04 0.06 0.08
  Liquid Phase mole ratio x10 4

 
 
 
© School of Engineering and Physical Science Page 57
 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
Step  5:  Find  the  number  of  transfer  units  
Although   the   equation   derived   earlier   applies   to   the   number   of   transfer   units   in  
terms   of   overall   gas   phase   units,   it   would   be   normal   to   define   this   expression   in  
terms  of  liquid  phase  units.  Numerically  the  two  numbers  will  be  different.  
G
X1 = 5.625.10 X2 = 6.75.10 h .X1 = 1.839.10 h. = 2 .10
9 6 4 3
L

Y2 = 1.103.10 h .X2 = 0.221

4
Y1 = 0

Number of transfer units is therefore

1 G . Y2 h .X1 G
Nog .ln 1 h. h. Nog = 3.546.10
3

.G L Y1 h .X1 L
h 1
L
 
Step  6:  Estimate  the  column  diameter  
320.m
1
Select the packing factor for 16mm plastic pall rings Fp

kg kg
Mass rates of liquid and gas are Lw = 1.839 Gw = 0.1
s s

Lw . ρ gas
Flow factor is = 0.474
Gw ρ
water

Design for 20mm water pressure drop, so the K4 factor is K4 0.5

 
The   gas   rate   in   this   example   has   been   selected   specifically   to   place   the   operating  
conditions  within  the  middle  of  the  flooding  curves.    
Now  check  the  flooding  limit  
At flooding, the K4 value is K4_flood 1.3

K4 .100
Percentage flooding is flood flood = 62.017
K4_flood
 
 
Now  find  the  mass  flow  velocity  (note  α=13.1)  

K4.ρ gas. ρ water ρ gas kg

Vw Vw = 0.554
m .s
0.1 2
µ water
α .Fp .
ρ water

Page 58
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
4 .A 1
Column diameter is D D = 0.479 m
π

Set the column diameter to D 0.5.m

π. 2 2
Column area is A2 D A 2 = 0.196 m
4  
 
 
Check  that  the  packing  size  is  consistent  with  the  diameter  of  the  column:  
 

Check the flooding values with the new diameter

A1
flood flood . flood = 56.991
A2
 
 

Set the packing size as d 16 mm

D
Column diameter to packing size = 31.25
d
Typical   flooding   limits   are   between   50%   to   75%,   hence   this   column   configuration  
seems  acceptable.  
 
Step  7:  Find  the  height  of  a  transfer  unit  
The  diffusion  limits,  calculated  in  a  previous  example,  are:  
Diffusivity  is  therefore:   D L =1.55x10-­‐9  m2/s,   D v = 2.037x10-­‐5  m2/s  

1.077.10 .cP
2
Gas viscosity at operating temperature is µ gas

µ water
Schmidt number for water is Scwater Scwater = 851.613
ρ [Link]
µ gas
Schmidt number for gas phase Scgas Scgas = 0.794
ρ [Link]
 
 
Lw kg
Liquid mass velocity is La La = 9.367
A2 m .s
2

Gw kg
Gas mass velocity Ga Ga = 0.509
A2 m .s
2

 
Select  the  appropriate  factors  from  the  supplied  performance  charts:  
K3=  0.9,   ψn =60,   φ h =0.03.  
Guess  that  the  overall  height  of  packing  will  be  Z=1.5m,  then:  
© School of Engineering and Physical Science Page 59
 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
and  for  the  liquid  side:  
1.110.15
Z 0.15
αα.φ.ψn
h . . Sc
0.5 . D
Hl Scwater
0.5 . K3. . Z Hl = 0.216 m
gas 3.05
0.305 3.05
Hg Hg = 0.299 m
( La.1 .1 .1 )
0.5

The  height  of  a  transfer  unit  is  then:  

G
Hog Hg h . .Hl Hog = 432.458m
L
Hence  the  height  of  packing  is:  
Z Hog .Nog Z = 1.534m
 

5.5 Deareation by Chemical Methods

Another  way  to  remove  dissolved  oxygen  and  other  gases  is  to  chemically  changes  
them  into  something  else  which  can  easily  be  removed  or  becomes  inactive  in  terms  
of  corrosion.  Such  methods  are  useful  for  removing  small  amounts  of  dissolved  gases  
from  large  volumes  of  water.  
Sodium  Sulphite  is  perhaps  the  most  common  method:  
2Na 2 SO 3 + O 2 ⎯⎯→ 2Na 2 SO 4  
Stoichiometrically,   7.9   ppm   of   sodium   sulphite   is   required   to   react   with   1   ppm  
dissolved   oxygen,   and   a   typical   dosage   is   therefore   10   mg   Na2SO4   per   mg   of  
dissolved  oxygen.    
 
At   normal   atmospheric   temperatures,   the   rate   of   the   sulphite-­‐oxygen   reaction   is  
relatively   low   and   sufficient   time   is   needed   for   the   reaction   to   be   assumed   to   be  
complete.  A  catalyst  (e.g  Cobalt  or  Cupric  ions)  can  be  added  to  reduce  the  reaction  
time;  0.001ppm  Cobalt  ion  will  significantly  reduce  the  sulphite-­‐oxygen  reaction  time  
(6).    
 
As  this  treatment  method  relies  on  a  chemical  reaction,  other  dissolved  components  
either  present  or  added  for  other  purposes  can  influence  the  reaction  rate  and  the  
mass  of  sulphite  needed;  in  systems  requiring  Chlorine,  sulphite  process  is  done  at  
an  earlier  stage,  or  further  up  steam  before  chlorine  is  added.  
 
Hydrazine   has   also   been   used   as   an   oxygen   scavenger   for   treating   boiler   feed   water.    
At   pressures   above   650   psi,   the   water   temperature   is   sufficient   high   as   to  
decompose   sulphite   into   corrosive   SO2.   Hydrazine   is   seen   as   a   viable   alternative   in  
these  instances.  The  principle  reaction  is:  
N2H4 + O 2 ⎯⎯→N2 + 2H2 O  

5.6 Enhanced Extraction Systems

De-­‐oxygenation   towers   tend   to   be   large   and   heavy   and   several   vendors   have  
developed   and   support   enhanced   processing   systems.   One   such   system   is   the  
MINOX   “Serpentine”   de-­‐oxygenation   system,   using   nitrogen   as   the   stripping   gas,  
which   is   cleaned   and   re-­‐used.   Water   and   nitrogen   are   passed   through   two   static  

Page 60
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
mixers   operating   in   essentially   co-­‐current   mode.   Figure19   shows   a   representation   of  
this   process   where   there   are   two   static   mixers   and   gas/liquid   separation   tanks.  
Oxygen  removed  from  the  water  and  collected  in  the  nitrogen  is  removed  from  the  
nitrogen  by  catalytic  oxidation  of  methanol  at  180°C  to  230°C.  Reaction  products  are  
returned  to  the  water  stream  and  the  process  operates  without  gas  release  into  the  
atmosphere.   airborne   release   of   with   reaction   products   fed   back   into   the   water   to  
prevent   A   small   portion   of   the   nitrogen   is   absorbed   by   the   water,   as   will   products  
from   the   methanol   oxidation,   and   hence   a   small   makeup   stream   is   needed,   which  
may  come  from  fresh  air  or  from  nitrogen  generators.    
 

 
Figure  19:  Schematic  flowscheme  for  the  MINOX  process.  
 
 
 
MINOX   was   developed   during   1985-­‐1990   and   installed   on   several   North   Sea  
platforms.   The   units   which   can   be   skid   mounted,   have   been   built   to   process   100,000  
to   400,000   bwpd.   Nitrogen   can   be   supplied   either   from   the   air,   or   by   dedicated  
nitrogen   gas   generators.   Generators   using   membrane   separation   techniques   have  
been  developed  and  are  supplied  by  a  number  of  vendors  as  an  alternative  to  cold  
separation   of   air.   Chlorine   can   be   added   upstream   of   this   unit   without   damaging   the  
catalyst.  
   

© School of Engineering and Physical Science Page 61

 
 OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
 

6 Processing Schemes
Processing   schemes   to   treat   water   from   either   production   separators,   extracted  
from  rivers  and  that  extracted  from  sea  water  depend  on  the  ultimate  destination,  
the   chemical   and   microbial   composition   of   the   raw   water   and   that   present   in   the  
target   destination.   In   addition,   provision   for   testing   new   techniques   would   be   a  
sensible  step  in  new  designs  as  advances  in  technology  are  best  evaluated  on  field  
waters   under   field   conditions.   In   the   preceding   sections,   there   has   been   no  
discussion   regarding   placing   sampling   points   for   both   chemical   analysis   and   these  
would  normally  be  present  at  a  number  of  points  e.g  pH  probes,  sample  extraction  
ports  and  for  corrosion  monitoring.  Provision  for  accessing  the  water  stream  before  
and   after   major   treatment   units   will   provide   supplementary   data   should   operational  
problems  arise.  
 
The   preceding   sections   have   outlined   the   major   issues   and   treatment   options   for  
processing   water   for   either   injection   or   disposal.   The   costs   for   various   individual  
treatment   options   varies   and   considerable   debate   is   ongoing   to   the   degree   and  
direction   produced   water   treatment   will   take;   Figure20   depicts   relative   costs   for  
some  of  the  treatment  options  covered  earlier;  
 
7
 
6 Membranes
 
 
 Total 5
 Cost Centrifuge
 
($MM/yr) 4
  Filter/coalescer
Floatation
  3
 
  2 Reinjection
 
 
1 Hydrocyclones
 
 
 
  0 10 20 30 40 50
  Discharged oil in water content (ppm)
 
 
Figure  20  Various  options  and  costs  for  water  treatment(12)  
 
This   section   reviews   some   of   the   processing   schemes   which   have   been   used   in  
various  locations  to  treat  water  from  various  sources.  

Page 62
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
6.1 Generalised Water Treatment System
Figure21  depicts  a  typical  water  treatment  system,  comprising  of  two  filters;  coarse  
filtration  followed  by  fine  filtration,  chemical  injection  and  finally  deoxygenation  in  a  
packed  column.  The  two  filter  stations  were  a  common  feature  of  water  treatment  
systems   and   some   operators   have   found   one   degree   of   filtration   is   adequate   for  
their  uses.  The  backwash  system  has  been  included  as  a  reminder  that  infrastructure  
for  providing  sufficient  backflush  volume  is  needed.  
 
 
 
 
  De-oxygenation Column

 
  Injection

  Gas & Oxygen

 
 
  Cooler
 
  Source
Coarse Fine

  Water
Filter Filter

 
Stripping Gas
 
 
  Backwash
  Tank
 
 
  Injection Pumps
Backwash pumps
 
 
 
Figure  21  Schematic  representation  of  a  generalised  water  processing  scheme  

6.2 Produced Water Reinjection on the ULA Field(13)

The  ULA  field  located  in  the  Norwegian  sector  of  the  North  Sea,  280  km  South  West  
from   Stravanger   was   discovered   in   1979   and   oil   production   started   in   October   1986.  
The   well   system   now   consists   of   9   production   wells   and   6   injection   wells.   Injecting  
seawater   started   in   1988   with   the   treatment   plant   consisting   of   coarse   filtration,  
followed   by   fine   filtration   and   then   deoxygenation.   Original   specifications   for   the  
injected   seawater   called   for   filtration   to   remove   98%   of   particles   larger   than   2  
microns.  Increasing  costs  of  filtration  led  to  a  review,  which  in  1990  resulted  in  by-­‐
passing  the  fine  filter  with  no  appreciable  effects  on  the  reservoir.  
 
Table  10  shows  measurements  of  solids  upstream  and  downstream  of  the  fine  filter  
and   indicates   that   fine   filtration   is   effective   in   removing   solids   larger   than   2   microns.  
The  data  also  indicates  that  small  particles  less  than  2  microns  can  be  generated  in  
© School of Engineering and Physical Science Page 63
 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
the  processing  operations  and  pipelines  after  the  fine  filter  presumably  as  a  result  of  
corrosion:  
 
 
Particulates  in  water   Upstream   Downstream   Wellhead   Downhole  
–  no  of  particles  per   fine  filters   fine  filters   injection   well  
500  µl   header  
>  2  micron   6653   58   555   17000  
>  4  micron   900   2   25   3327  
>  8  micron   73   0   0   418  
>  16  micron   5   0   0   54  
Suspended  solids   3.5   0.1   0.2   6  
(mg/l)  
Table  10:  Examples  of  fine  filtration  on  sea  water  
 
Water   production   started   in   1992   and   the   original   treatment   route   was   processing  
via   gas   flotation.   In   1993   the   capacity   of   the   flotation   system   was   exceeded   and  
became   the   limiting   factor   in   oil   production.   Hydrocyclones   were   subsequently  
installed   which   reduced   the   oil-­‐in-­‐water   content   to   20-­‐30   mg//litre.   This   made  
produced   water   suitable   for   direct   re-­‐injection.   The   current   scheme   is   shown   in  
Figure22:  
Gas to
  De-emulsifier processing
 
 
  Separators Oil to separators
  Oil
  Hydrocyclones
 
  De gassing
  Production Wells

 
Hypochlorite Antifoam Produced water cooler
 
  Filter
  Biocide, O2 Discharge to sea
scavenger
 
Water injection pumps
 
 
  Residence
  drum
Sea water
 
Deoxygenation tower Injection wells
 
 
 
Figure  22  Flow  diagram  for  the  water  treatment  facility  for  the  ULA  field  
 
 

Page 64
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
The  produced  water  cooler  was  designed  to  reduce  the  water  temperature  to  55°C  
in  order  to  prevent  cavitation  in  the  injection  pumps.  Once  the  hydrocyclones  were  
installed,   upstream   modifications   were   made   in   order   to   overcome   problems   with  
variability   in   oil   concentrations   affecting   the   performance   of   the   cyclone   units;  
optimising   the   use   of   de-­‐emulsifiers,   improvements   to   the   oil/water   separator  
internals.  Without  the  flexibility  to  inject  chemicals,  make  modifications  to  internals  
etc,  optimisation  would  have  been  difficult.  
 
Table   11   shows   an   extract   from   the   analysis   of   sea   water   and   produced   water   which  
indicates  the  impact  of  adding  produced  water  to  sea  water.  
 
Water  Quality  Parameter   Sea  Water   Mixed  Produced  
Water/Sea  Water  
(35%/65%)  
Total  Solids  (mg/l)   Non  detected   22  
Suspended  solids  (mg/l)   4.2   6  
No  particules  per  500  mico  liters   15000   155000  
Hydrocarbons  (mg/l)   0   15  
Temperature  (C)   18   31  
Calcium  (mg  Ca/l)   4120   2330  
Barium  (mg  Ba/l)   0   5.3  
Strontium  (mg  Sr/l)   7.7   96  
Iron  (mg  Fe/l)   0.3   10.7  
Chlorine  (mg  Cl/l)   19850   28300  
Sulphate  (mg  SO4/l)   2790   1900  
Oxygen  (µg  O2/l)   80   15  
pH   8.1   7.3  
Sulphate  reducing  bacteria  (no/ml)   0   4.5  
General  anerobic  bacteria  (no/ml)   45   200  
Table  11  Examples  of  compositions  of  sea  water  and  injected  water.  
Injection   rates   are   typically   300,000   m3/day   with   the   sea   water   intake   of   15,000  
m3/day.  
 
 
 
 
 
 
 
 
 
 
 
 

© School of Engineering and Physical Science Page 65

 
OIL  &  GAS  PROCESS     CHAPTER  13  
ENGINEERING  
 
6.3 Closed Water Treatment System in Kanas(6)
Figure23  shows  the  flowscheme  employed  to  treat  2500  bpd  of  general  supply  water  
and  6500  bpd  of  produced  water  for  re-­‐injection.  Analysis  of  the  two  water  sources  
indicated  no  additional  treatment  other  than  filtration  and  chemical  injection.  
  Gas
 
 
  Separation Train
 
  Produced Oil
 
  Chemical Injection
 
Produced water
 
 
  Sand Bed Filter
 
 
 
  Injection Pumps
  Injection wells
 
 
Figure  23  Filtration  treatment  system  
 
This  is  a  relatively  straight  forward  scheme  which  uses  oil  free  water  passed  directly  
through  a  sand-­‐gravel  filter  bed.  
   

Page 66
 
OIL  &  GAS  PROCESS     WATER  INJECTION  
ENGINEERING   SYSTEMS  
 
 

7 References
1. [Link]   “Produced   Water   Management   From   An   Offshore   Operators  
Perspective”  Produced  Water  Workshop,  26th-­‐27th  March  2003  
2. Environmental  Engineering  News,  1998,  No  1.  Sintef  Applied  Chemistry,  Norway  
3. American  Society  of  Civil  Engineers,  “Water  Treatment  Plant  Design  2nd  Edition”,  
McGraw  Hill,  1990  
4. Mittelman,   M.W.   “Biological   Fouling   of   Purified-­‐Water   Systems:   Part   3,  
Treatment”,   Microcontamination   4(1),   pp.   30-­‐40,   70   (January   1986).   (See  
[Link])  
5. [Link],   [Link]   “Chemical   Engineering   Volume   2”   Pergamon  
International  Press  
6. [Link]   “Introdcution   to   Oilfield   Water   Technology”   1979,   National  
Association  of  Corrosion  Engineers.  
7. [Link],  [Link]  “Perry’s  Chemical  Engineers  Handbook”,  McGraw  Hill  
8. [Link],   [Link]   “Chemical   Engineering   Volume   6”   Pergamon  
International  Press  
9. [Link],   “Reconsider   Packed-­‐Tower   Pressure   Drop   Correlations”,   Chemical  
Engineering  Progress,  Jan  1992,  (65-­‐72)  
10. [Link],   [Link]   “Process   Pant   Design”,   Heinemann   Educational   Books,  
1981  
11. [Link],   [Link]   “Equilibrium   Stage   Separation   Operations   in   Chemical  
Engineering”,John  Wiley  
12. [Link],   [Link],   [Link]   “The   CHARM   model   used   in   Environmental   Risk  
Management   of   produced   Water   on   ULA”,   Produced   Water   2,   Environmental  
Science  Research  Vol  52.  Plenum  Press  1995  
13. [Link],   [Link],   [Link],   [Link]   “Produced   Wtare   Reinjection   (PWRI)  
Experiences  from  the  field”,  Produced  Water  2,  Environmental  Science  Research  
Vol  52.  Plenum  Press  1995  
 
 
 

© School of Engineering and Physical Science Page 67