Final Report Port Authority

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 21

lOMoARcPSD|1717990

Final Report - Port Authority

Coastal Dynamics II (Technische Universiteit Delft)

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

 
Coastal  dynamics  II  
Report  exercise  2  |  Port  Authority  
 
 
 

08-­‐Jun-­‐15  
TU  Delft  
 

   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

Contents  
1.  Introduction  .............................................................................................................................................  2  
1.1  Background  information  ....................................................................................................................  2  
1.2  Brief  summary  of  earlier  research  (UNIBEST/ASMITA)  ......................................................................  2  
1.3  Problem  description  (port  authority)  .................................................................................................  3  
1.4  Objective  and  approach  .....................................................................................................................  3  
2.  Model  setup  .............................................................................................................................................  4  
2.1  Introduction  of  Delft3D  ......................................................................................................................  4  
2.2  Grid  and  bathymetry  ..........................................................................................................................  4  
2.3  Boundary  conditions  ..........................................................................................................................  6  
2.4  Other  model  settings  ..........................................................................................................................  6  
3.  Model  results  ...........................................................................................................................................  7  
3.1  Introduction  proposed  solutions  ........................................................................................................  7  
3.1.1  Breakwater  extension  at  channel  side  ........................................................................................  7  
3.1.2  Breakwater  extension  at  river  side  ..............................................................................................  7  
3.1.3  Sand  trap  .....................................................................................................................................  8  
3.1.4  Multiple  groynes  along  the  breakwater  ......................................................................................  8  
3.2  Results  base  case  ................................................................................................................................  9  
3.3  Results  scenarios  ..............................................................................................................................  10  
3.3.1  Two  week  simulation  period  .....................................................................................................  10  
3.3.2  One  year  simulation  period  .......................................................................................................  12  
4.  Final  solution  ..........................................................................................................................................  14  
5.  Discussion  and  conclusions  ....................................................................................................................  15  
5.1  Discussion  .........................................................................................................................................  15  
5.1.1  Single  layer  ................................................................................................................................  15  
5.1.2  Wave  conditions  ........................................................................................................................  15  
5.1.3  Sensitivity  ..................................................................................................................................  15  
5.2  Conclusions  .......................................................................................................................................  16  
5.3  Recommendations  ...........................................................................................................................  17  
Sources  .......................................................................................................................................................  18  
Appendix  ....................................................................................................................................................  19  
 

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

1.  Introduction      
 

This  report  refers  on  the  Carrara  Coast  in  Tuscany  (Italy).  In  this  introductory  chapter,  brief  background  
information  is  provided  as  well  as  a  brief  summary  of  earlier  research  already  done  with  software  
UNIBEST.  Moreover,  there  is  a  problem  description  of  our  stakeholder,  which  is  the  Port  Authority  of  
Carrara  and  finally  the  objective  and  the  approach  followed  to  solve  the  problem  are  explained.  

1.1  Background  information  


The  Carrara  Coast  is  located  in  Tuscany  in  the  Northwestern  part  of  Italy.  The  city  of  Carrara  is  located  
on  a  sandy  coast,  which  is  part  of  a  coastal  cell  with  a  length  of  about  50  km.  Tuscany  area  and  more  
specifically  Carrara  beach  is  a  very  popular  holiday  destination,  which  also  means  that  the  coast  
characteristics  are  important  for  keeping  this  source  of  income  as  high  as  possible.  The  main  problem  of  
the  beach  concludes  to  the  fact  that  erosion  is  dominant  along  the  northern  part  of  the  beach  and  it  is  
travelling  towards  the  southern  part.  By  this  erosional  wave,  all  stakeholders  of  the  area  are  negatively  
influenced,  like  restaurant  owners  have  to  face  beach  erosion  in  front  of  their  restaurants,  port  
authority  has  to  dredge  a  lot  in  order  to  keep  approach  channel  open  and  the  port  operational  and  the  
municipality  of  Carrara  has  to  confront  relative  sand  loss  and  swimming  safety.  
The  causes  of  this  erosion  are  briefly  a  dam  constructed  upstream  in  the  Magra  River,  which  was  the  
main  sediment  supplier  in  the  area  and  the  construction  of  the  port  of  Carrara  constructed.  In  order  to  
explore  further  which  caused  the  problem  of  erosion,  a  research  with  UNIBEST  was  conducted  and  the  
results  are  presented  in  paragraph  1.2.  
In  the  past,  a  series  of  groynes  was  constructed  privately  and  in  a  rather  independent  and  so  ineffective  
way,  in  front  of  restaurants,  which  just  pushed  the  problem  downwards.  With  a  more  integrated  
approach,  many  potential  solutions  have  been  looked  for.  Depending  on  the  stakeholder,  several  
measures  like  groynes,  emerged/submerged  breakwaters,  nourishment,  creation  of  surf  resort  in  the  
area  downstream  of  the  port  and  a  tidal  basin  upstream  of  the  port  are  applied.  

1.2  Brief  summary  of  earlier  research  (UNIBEST/ASMITA)  


Our  historical  overview  starts  in  1878  until  now.  During  this  period,  there  were  several  human  
interventions  which  had  a  great  impact  on  the  coastline.  In  1878  the  revetment  was  built  (length  about  
1.66  km  -­‐  boulevard  near  the  sea)  and  after  35  years  (in  1913)  a  dam  was  constructed  in  the  Magra  River  
which  decreased  dramatically  the  sediment  supply  of  the  river  in  the  coastal  system.  After  8  years  (in  
1921)  the  construction  of  the  Port  of  Carrara  started  and  was  finished  after  4  years  in  1925.  
Our  earlier  research  aimed  to  explore  the  severity  of  each  human  intervention  on  the  erosion  problem  
of  the  coast.  Several  scenarios  with  no  human  interventions,  with  only  one  and  with  a  combination  
between  them  were  run  and  helped  us  come  to  a  conclusion  about  the  main  cause  (if  only  one)  and  the  
factors  that  made  this  problem  so  important.  
Our  earlier  research  with  UNIBEST  came  to  the  conclusion  that  human  interventions  had  a  lot  of  impact  
on  the  coast  over  the  years.  According  to  the  several  scenarios  considered,  we  can  conclude  that:  
1.   The  port  mainly  has  a  local  effect  on  the  coast.  It  causes  accretion  on  one  side  and  extra  erosion  
on  the  other  side.  This  erosion  is  limited  to  a  small  stretch  of  coast.  
2.   The  damming  of  the  river  has  a  large  impact  on  the  coast.  The  effects  are  not  local,  but  affect  
the  entire  coast.  The  damming  causes  a  deficit  of  sediment,  which  causes  less  total  accretion  on  
the  southern  part  of  the  coast  (compared  to  the  situation  without  damming)  and  more  erosion  
on  the  northernt  side  of  the  coast.    

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

3.   The  revetment  mainly  has  a  local  effect  on  the  coast.  It  prevents  further  erosion  nearby  the  river  
mouth  and  just  after  the  port.  
4.   A  tidal  basin,  if  constructed,  can  have  two  effects.  These  effects  depend  on  the  fact  whether  we  
consider  the  basin  as  a  source  or  sink  term.  This  however  depends  on  the  dimensions  of  the  
basin  and  the  possible  effect  of  sea  level  rise,  which  are  now  unknown.  When  the  basin  is  a  sink  
term,  it  will  mainly  result  in  erosion  at  the  end  of  the  coastline.  Even  though  the  UNIBEST  model  
results  are  not  referring  to  the  local  area  near  the  basin,  changes  can  be  expected  there  as  well  
due  to  secondary  effects.    

All  these  interventions  had  a  large  impact,  but  overall  it  can  be  said  that  the  damming  of  the  river  had  
the  biggest  global  influence.  The  construction  of  the  port  had  the  largest  local  impact.  

1.3  Problem  description  (port  authority)  


Apart  from  the  general  erosion  problem  of  the  whole  Tuscany  Coast,  Port  Authority  of  Carrara  Port  has  
to  face  one  additional  problem.  Each  year  a  several  amount  of  sediment  is  trapped  in  the  entrance  
navigation  channel  of  the  port,  which  leads  to  both  cost  of  dredging  and  disturbance  of  the  operation  of  
the  port.  Moreover,  once  the  sediment  enters  the  boundaries  of  the  port,  it  is  characterized  as  
“polluted”,  making  its  deposition  a  very  hard  and  quite  expensive  task.  
Because  of  the  great  importance  of  the  Port  for  the  regional  economics,  Port  Authority  wants  to  keep  it  
open  all  around  the  year,  maintaining  its  operational  depth  by  dredging  and  trying  to  avoid  settled  
sediment  reaching  the  entrance  of  the  port  in  order  to  avoid  to  be  polluted  and  in  this  way  to  be  able  to  
dump  it  in  a  more  easy  way.  
According  to  the  data  provided  the  majority  of  the  dredged  material  is  sand  and  only  15%  is  silt  and  the  
volumes  that  are  dredged  are  not  constant  during  the  last  years,  but  in  a  large  perspective  it  seems  to  
have  been  increased  in  the  last  decades.  

1.4  Objective  and  approach  


The  Port  Authority  of  Carrara  is  strongly  intended  to  fund  a  research  program  in  order  to  solve  the  
problem  of  sedimentation  in  the  navigation  channel  of  the  port.  The  goal  of  the  research  is  to  evaluate  
variants  to  limit  the  sedimentation  rate  in  the  port  entrance  channel.  This  can  be  accomplished  by  
finding  a  way  to  minimise  or  even  diminish  the  yearly  dredging  cost  of  the  operational  maintenance  of  
the  port.  To  do  so,  several  causes  of  the  sedimentation  and  relevant  dredging  volumes  must  be  revealed  
and  consequently  several  ways-­‐mechanisms  to  counteract  the  sedimentation  problem.  
The  approach  followed  towards  this  goal  includes  four  different  scenarios:  

•   Breakwater  extension  at  channel  side  


•   Breakwater  extension  at  river  side  
•   Sand  trap  
•   Multiple  groynes  along  the  breakwater  

   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

2.  Model  setup      

2.1  Introduction  of  Delft3D    


Delft3D  is  a  modelling  package  developed  by  Deltares  in  the  eighties  which  is  still  constantly  updated.  It  
is   an   integrated   software   package   for   the   simulation   of   2D   and   3D   flow,   sediment   transport,  
morphology,  waves  and  the  interaction  between  these  processes  in  time  and  space.  Delft3D  is  used  for  
modelling   of   coasts   and   rivers   but   also   for   the   simulation   of   manmade   features   such   as   harbours   and  
locks.  
 
Delft3D  consists  of  several  modules  combined  through  a  single  user  interface.  The  two  most  important  
modules  used  in  this  case  are  the  FLOW  and  WAVE  module  which  influence  each  other  directly.  
 
Delft3D-­‐FLOW  solves  the  Navier-­‐Stokes  equations  for  an  incompressible  fluid,  under  the  shallow  
water  and  the  Boussinesq  assumptions.  The  set  of  partial  differential  equations  in  combination  with  an  
appropriate   set   of   initial   and   boundary   conditions   is   solved   on   a   finite   difference   grid.   In   this   case   the  
Navier-­‐Stokes  equations  are  integrated  over  depth,  meaning  a  2D  instead  of  a  3D  approach  is  used.  
 
Delft3D-­‐WAVE   uses   the   SWAN   model   to   simulate   the   evolution   of   random,   short-­‐crested   wind-­‐
generated  waves.  The  SWAN  model  is  based  on  the  discrete  spectral  action  balance  equation  and  is  fully  
spectral  (in  all  directions  and  frequencies).  This  implies  that  short-­‐crested  random  wave  fields  
propagating  simultaneously  from  widely  different  directions  can  be  used.  

2.2  Grid  and  bathymetry    


In  Delft3D  a  different  grid  is  used  for  each  module.  In  our  case  we  work  with  the  flow  and  wave  module  
so  we  use  2  different  grids.  The  grids  are  composed  of  rectangular  grid  cells.  The  grid  cells  vary  in  size.  
The  more  grid  cells  you  have  at  a  location  the  higher  the  resolution  and  the  more  accurate  the  info  is.  As  
seen   in   the   picture   below   the   grid   cells   get   smaller   near   the   port,   because   of   that   we   have   more   grid  
cells  over  there  and  thus  the  information  is  more  detailed  which  is  necessary  to  make  valid  predictions.    
 
The  wave  grid  is  made  larger  than  the  flow  grid  because  the  influence  of  refracting  waves  from  outside  
the  flow  domain  also  needs  to  be  taken  into  account.  Otherwise  there  is  a  chance  that  we  get  a  wrong  
estimation  of  the  flow  and  the  sediment  transport.  As  mentioned  in  the  previous  paragraph  we  use  only  
one  grid  layer  (2D  model),  this  means  that  the  sediment  concentration  is  represented  by  one  value  over  
the  entire  water  column.  In  reality  this  concentration  is  high  near  the  bottom  and  low  near  the  water  
surface.  So  by  using  one  layer  we  underestimate  the  value  at  the  bottom  and  overestimate  it  in  the  top  
of  the  water  column.  
 
 

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

 
Figure  1:  Grid  (wave  grid  is  the  larger  grid.  On  top  of  it,  the  smaller  flow  grid  is  visible)  

 
Two   aspects   that   are   important   for   the   bathymetry   are   the   depth   and   the   sediment   availability.   The  
depth   profile   for   the   Carrera   was   already   implemented   in   the   model.   The   sediment   availability   gives  
information   about   how   much   sediment   is   available   to   be   transported   by   the   flow.   In   the   model   the  
standard   value   is   10.   Meaning   there   is   a   layer   of   10   meter   of   sediment   available   for   transport.   The  
breakwaters  are  of  course  fixed  structures,  meaning  the  value  for  the  sediment  availability  over  there  is  
zero.  This  results  in  the  picture  as  seen  below,  with  the  green  areas  representing  the  breakwaters.    
   

 
Figure  2:  Bathymetry  of  the  area  around  the  harbor.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

2.3  Boundary  conditions  


The   two   different   (FLOW   and   WAVE)   modules   have   their   own   boundary   conditions.   The   specified  
boundaries  are  SW,  NW  and  SE.  
 
FLOW  
The  SW  boundary  induces  the  tide  on  the  system.  The  NW  and  SE  boundaries  are  Neumann  boundaries.  
The  key  characteristic  of  Neumann  boundaries  is  that  it  limits  the  reflection  of  waves.  
 
WAVE  
The  SW  and  NW  boundaries  are  the  boundaries  where  the  waves  enter  the  domain.  These  boundaries  
are   divided   into   sections   and   each   section   has   its   own   wave   condition.   This   is   done   because   the  
boundaries   are   quite   long   and   thus   the   conditions   can   vary   significantly   along   the   boundary.   The   SW  
boundary   has   4   sections   and   the   NW   boundary   has   3   sections.   The   SE   boundary   is   different   than   the  
others,  here  the  waves  leave  the  domain.  Here  there  is  no  division  in  sections  but  just  one  uniform  value  
for  the  entire  boundary.  This  is  done  because  the  waves  leaving  the  domain  don’t  have  a  big  influence  
on  the  system.  This  is  not  an  essential  boundary  but  necessary  to  build  in  safety  in  the  model  so  that  it  is  
guaranteed  that  the  waves  leave  the  system  at  that  boundary.  
 

2.4  Other  model  settings  


Delft3D  has  a  large  number  of  settings  that  change  the  behavior  of  the  model.  Important  settings  which  
were  essential  in  our  modeling  exercise  are:  
 
•   Dredging  and  dumping:  adds  a  sediment  source  and  sink  to  the  model.  
•   Sediment  properties:  such  as  roughness,  Dn50,  density.  
•   Morphological   scale   factor:   this   factor   represents   the   difference   in   scale   between   the  
hydrological  and  morphological  timescale.  
•   Directional  space:  all  the  directions  in  which  waves  can  come  into  the  domain.  
 

   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

3.  Model  results      

3.1  Introduction  proposed  solutions  


To  avoid  sedimentation  in  the  approach  channel  as  much  as  possible,  four  solutions  are  proposed.  

1.   Breakwater  extension  at  channel  side  


2.   Breakwater  extension  at  river  side  
3.   Sand  trap  
4.   Multiple  groynes  along  the  breakwater  

It  is  expected  that  these  variants  will  either  block  the  sediment  transport  completely,  or  let  most  of  the  
sediment  accrete  before  it  reaches  the  approach  channel.  

3.1.1  Breakwater  extension  at  channel  side  


As  can  been  seen  in  figure  (3)  ,  this  variant  consists  of  an  extension  of  the  breakwater  at  the  approach  
channel  side.  It  is  expected  that  this  variant  will  trap  sediment  and  therefore  avoid  sedimentation  of  the  
approach  channel.  After  time,  it  is  expected  that  a  bypass  will  be  created,  but  that  a  lot  of  sediment  goes  
offshore   due   to   cross   shore   transport.   The   lack   of   sediment   however,   might   cause   erosion   on   the  
downstream  side  of  the  port.  

 
Figure  3:  Variant  1,  breakwater  extension  at  the  channel  side  of  the  breakwater.  

3.1.2  Breakwater  extension  at  river  side  


This   variant   is   similar   to   the   previous   one,   however   the   extension   is   now   at   the   upstream   side   of   the  
breakwater.   The   sediment   will   now   be   trapped   before   the   port   which   will   also   decrease   the  
sedimentation   of   the   approach   channel.   The   sediment   in   front   of   the   breakwater,   might   distribute  
however,  which  could  cause  stability  problems  of  the  breakwater.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

 
Figure  4:  Variant  2,  breakwater  extension  at  the  river  side  of  the  breakwater.  

3.1.3  Sand  trap  


In  contrast  to  the  last  variants,  this  proposed  solution  is  a  soft  measure.  The  sand  trap  is  basically  a  big  
hole  in  the  sediment  transport  area.  This  sudden  increase  in  depth  will  decrease  the  flow  velocity  and  
therefore   cause   sedimentation   in   the   hole.   This   should   lead   to   a   decrease   in   sedimentation   of   the  
approach  channel.  

 
Figure  5:  Variant  3,  sand  trap.  A  hole  close  by  the  breakwater,  more  to  the  river  side.  

3.1.4  Multiple  groynes  along  the  breakwater  


This   variant   is   again   a   hard   measure.   Instead   of   using   one   long   breakwater   extension,   which   is   quite  
costly  to  construct  because  of  the  depth  of  the  water  there,  three   shorter  groynes  are  used  alongside  
the  breakwater.  After  time,  it  is  expected  that  a  bypass  will  be  created,  but  that  a  lot  of  sediment  goes  
offshore  due  to  cross  shore  transport.    It  is  expected  that  all  three  of  these  groynes  will  trap  sediment.  
On  the  downstream  side  however,  erosion  might  occur.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

 
Figure  6:  Variant  4,  multiple  groynes  along  the  breakwater.  

3.2  Results  base  case  


To   be   able   to   see   how   effective   the   different   variants   are,   the   base   case   has   to   be   modeled   as   well.  
Below   are   the   results   of   a   2   week   and   a   3   month   morphological   simulation   time.   It   is   visible   that   the  
approach  channel  of  the  harbor  quickly  fills  with  sediment  and  is  almost  completely  filled  after  3  months  
already.  

 
Figure  7:  2  week  result  of  the  reference  case.     Figure  8:  3  month  result  of  the  reference  case.  

When  figure  (8)  and  figure  (9)  are  compared,  it  can  be  concluded  that  the  approach  channel  does  not  fill  
a  lot  more  after  3  months.  The  sediment  is  mostly  just  spreading  out  over  a  bigger  area.    

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

 
Figure  9:  One  year  result  of  the  reference  case.  

3.3  Results  scenarios  


In  this  chapter,  the  results  of  the  scenarios  are  shown  and  explained.  Firstly,  all  four  scenarios  have  been  
simulated  over  a  period  of  two  weeks  in  morphological  time.  Based  on  the  results  of  these  simulations,  
variant   (1)   and   (4)   has   been   simulated   over   3   months   morphological   time,   to   estimate   the   long   term  
results.    

3.3.1  Two  week  simulation  period  


For   these   simulations,   a   hydrological   simulation   time   of   9   hours   and   a   morphological   factor   of   40   are  
used.  This  leads  to  a  two  week  morphological  simulation  time.    

3.3.1.1  Breakwater  extension  at  channel  side  


After  two  weeks,  this  variant  is  successful  in  trapping  sediment  and  therefore  causing  no  sedimentation  
in  the  approach  channel.  There  is  little  erosion  at  the  downstream  side,  but  not  yet  extensive.  

 
Figure  10:  Two  week  simulation  of  breakwater  extension  at  the  channel  side.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

3.3.1.2  Breakwater  extension  at  river  side  


This  variant  is  also  effective  in  trapping  sediment  before  the  harbor,  but  the  sediment  at  the  breakwater  
started  to  redistribute.  This  leads  to  still  a  little  sedimentation  in  the  approach  channel  and  erosion  in  
front   of   the   breakwater.   Because   erosion   at   the   breakwater   is   unwanted   since   it   will   cause   stability  
problems   of   the   breakwater,   this   variant   is   not   a   good   measure   to   decrease   sedimentation   in   the  
approach  channel.  

 
Figure  11:  Two  week  simulation  of  breakwater  extension  at  river  side.  

3.3.1.3  Sand  trap  


The   sand   trap   is   not   that   effective   in   trapping   sediment   as   can   be   seen   in   figure   12   .   This   is   probably  
because   the   simulation   is   a   2DH   simulation   and   therefore   the   sediment   distribution   over   the   vertical  
deviates   too   much   from   reality.     In   reality,   the   concentrations   at   the   bottom   are   higher   than   at   the  
surface,  which  would  result  in  a  more  effective  sandtrap.  However,  it  would  also  mean  that  the  sandtrap  
would  fill  quicker.  Therefore,  this  measure  is  not  very  effective,  but  only  delays  the  time  a  little  bit  after  
which  dredging  is  needed.  

 
Figure  12:  Two  week  simulation  of  the  sand  trap.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

3.3.1.4  Multiple  groynes  along  the  breakwater  


This   variant   has   a   similar   effect   as   variant   (1).   The   sediment   gets   trapped   between   the   groynes   and  
therefore  no  sediment  is  available  for  sedimentation  in  the  approach  channel.  In  this  case  there  is  also  
little  erosion  at  the  downstream  side.  

 
Figure  13:  Two  week  simulation  of  multiple  groynes  along  the  breakwater.  

3.3.2  One  year  simulation  period  


For   these   simulations,   a   hydrological   simulation   time   of   6   days   and   a   morphological   factor   of   60   are  
used.   This   leads   to   an   one   year   morphological   simulation   time.   As   stated   above,   two   results   are  
regarded  as  good  measures;  variant  (1)  and  (4).  Therefore  only  these  variants  will  used  to  make  a  long  
term  estimation.  

3.3.2.1  Breakwater  extension  at  channel  side  


To  make  a  good  long  term  estimation,  the  model  is  improved  a  little,  which  resulted  in  figure  (14).  This  
renewed  model  was  then  used  for  the  long  term  estimation  as  visible  in  figure  (15).    

 
Figure  14:  Improved  model  for  variant  (1)     Figure  15:  Result  after  a  one  year  simulation  period.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

After  one  year,  there  is  a  lot  of  accretion  at  the  extension.  There  is  also  a  bypass  of  sediment  into  the  
approach  channel  so  here  we  also  observe  a  large  amount  of  sedimentation.    

3.3.2.2  Multiple  groynes  along  the  breakwater  


For  this  variant,  the  model  is  also  improved  a  little,  to  have  a  better  long  term  estimation.  This  resulted  
in  the  new  model  as  can  be  seen  in  figure  (17)  .  This  renewed  model  was  then  used  for  the  long  term  
estimation  as  visible  in  figure  (18)  .  

 
Figure  16:  Improved  model  for  multiple  groynes.                  Figure  17:  Result  of  improved  model  of  multiple  groynes.  

A  large  amount  of  accretion  occurs  between  the  groynes.  Again  there  is  a  bypass  of  sediment  into  the  
approach  channel.  

   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

4.  Final  solution        
In  chapter  3  variant  1  (Extended  Breakwater)  and  4  (Groynes)  were  chosen  to  be  the  most  promising.  
After  comparing  the  figures  for  an  one  year  simulation  the  differences  between  the  two  variant  results  
were  minimal.  Variant  4  traps  a  larger  amount  of  sediment  between  its  groynes  then  variant  1  with  just  
one  long  groyne.  So  it  takes  longer  for  the  approach  channel  to  accrete  for  variant  4.  But  after  a  while  
the  bypass  is  in  both  cases  so  large  that  in  both  cases  the  approach  channel  accretes  drastically.  
Dredging  is  thus  needed  in  both  cases.  

However  when  considering  the  structures  themselves  the  variant  1  is  the  most  easy  to  construct  and  the  
most  realistic  option.  Constructing  three  small  groynes  will  be  much  more  expensive  and  labour  
intensive.  So  variant  1  is  chosen  as  the  final  variant.  The  graphs  below  depict  the  accretion/erosion  for  
the  reference  situation  and  variant  1  and  4  after  one  year.  

 
Figure  18:  Accretion/erosion  of  the  reference  situation,  extended  breakwater  and  groynes  after  1  year.  

 
The  structures  also  result  in  a  changing  current  pattern  in  the  entrance  channel.  The  flow  velocities  for  
variant  1  increase  in  time  because  of  the  bypass  of  the  current  after  the  extension  accretes.  The  
velocities  are  shown  in  the  graph  below.  The  velocities  are  directed  perpendicular  to  the  entrance  
channel,  this  can  have  a  negative  effect  for  shipping.  
 

 
Figure  19:  Flow  vectors  for  the  Extended  Breakwater  after  3  months  and  1  year.  

   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

5.  Discussion  and  conclusions      


 

5.1  Discussion    
In  this  section  model  uncertainties  and  possible  improvements  are  discussed.  

5.1.1  Single  layer  


A  2D  model    is  used  since  the  concentration  is  integrated  over  the  depth.  Meaning  we  work  in  a  single  
layer   model.   As   mentioned   before   this   overestimates   the   sediment   transport   in   the   upper   part   of   the  
water  column  and  underestimates  it  near  the  bed.  As  a  result  of  this  some  simulated  solutions,  like  the  
sand  trap,  will  produce  less  accurate  output.  This  can  be  simply  improved  by  adding  multiple  layers,  this  
will  however  increase  the  calculation  time.  

5.1.2  Wave  conditions  


In  the  model  only  a  limited  amount  of  wave  conditions  are  used.  The  162  wave  conditions  have  been  
reduced  to  only  6  conditions.  This  effects  the  accuracy  of  the  model  results.  This  could  be  improved  by  
adding   more   wave   conditions   but   that   makes   it   computationally   much   more   intensive   and   thus   takes  
more  time.  
 
The  given  amount  of  dredged  material  (10.000  m3/year)  is  an  averaged  value  over  a  certain  amount  of  
years.  But  the  conditions  can  strongly  vary  per  year,  mainly  the  storm  conditions.  Therefore  the  amount  
of  10.000  m3/year  can  over-­‐  and  underestimate  the  real  amount  that  needs  to  be  dredged  significantly.  
 

5.1.3  Sensitivity  
The  sensitivity  represents  the  robustness  of  the  model.  A  large  sensitivity  means  that  small  changes  in  
the  input  parameters  can  change  the  output  significantly.  In  this  section  we  will  research  the  sensitivity  
of  the  model  by  changing  different  input  parameters.  
 
Dn50  
The  first  parameter  we  changed  is  the  nominal  diameter  of  the  sediment.  At  first  it  had  a  value  of  200  
µm  and  its  increased  to  500  µm  and  a  value  of  1000  µm.  The  bigger  the  diameter  of  the  sediment  the  
larger   the   velocities   that   are   needed   to   bring   the   sediment   into   suspension.     In   the   figures   below   we  
observe  that  a  change  in  nominal  diameter  strongly  impacts  the  results.  The  accretion  in  the  approach  
channel  is  much  smaller  for  the  bigger  diameters.  The  conclusion  is  that  the  model  is  very  sensitive  to  
changes   in   the   nominal   diameter.   This   is   probably   because   it   is   all   based   on   just   one   value   for   this  
important  parameter.    

 
Figure  20:  Erosion/Accretion  plots  for  a  changing  Dn50,  starting  with  200µm  up  to  1000µm  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

Wave  height  
The  second  input  parameter  we  change  is  the  significant  wave  height.  We  start  with  the  values  as  given  
in  the  model  and  add  a  meter  of  wave  height  in  each  step.  The  bigger  the  significant  wave  height,  the  
more   sediment   is   brought   into   suspension   and   the   sediment   transport   is   thus   larger.   Just   as   in   the  
previous  case  we  observe  that  a  change  in  the  significant  wave  height  strongly  impacts  the  results.  The  
accretion  is  much  larger  for  the  situation  with  larger  wave  heights.  So  again  it  can  be  concluded  that  the  
model  is  very  sensitive  to  changes  in  wave  height.  
 

 
  Figure  21:  Erosion/Accretion  plots  for  a  change  in  Wave  Height,  from  reference  to  largest  wave  height.  

Wave  direction  
The  third  input  parameter  that  is  varied  is  the  wave  direction.  Again  we  start  with  the  given  values  and  
increase  it  first  with  10  and  then  with  30  degrees  in  an  anti-­‐clockwise  direction.  Due  to  this  increase  in  
degrees   the   angle   of   incidence   of   the   waves   relative   to   the   coast   becomes   smaller.   Because   of   this  
decrease   in   angle   of   incidence   of   the   waves   the   sediment   transport   decreases.   As   seen   in   the   figures  
below  this  small  change  in  angle  creates  large  variations  in  the  output.  Again  it  can  be  concluded  that  
the  model  is  very  sensitive  to  this  change  in  input.  
 

 
Figure  22:  Erosion/Accretion  plots  for  a  change  in  Wave  direction,  from  reference  to  plus  10  and  30  degrees.  

5.2  Conclusions  
The  Delft3D  model  has  a  lot  of  uncertainties  in  the  results.  This  is  because  the  large  influence  of  the  
input  parameters.  Changing  these  parameters(Dn50,  waveheight  &  wave  direction)  will  result  in  large  
change  in  the  output  of  the  model.  The  Dn50  is  the  most  sensitive  because  this  is  represented  in  the  
input  as  just  one  value.  Wave  height  and  wave  direction  are  also  very  sensitive  input  parameters.  But  
these  parameters  are  less  sensitive  than  the  Dn50.  This  is  because  the  wave  height  and  wave  direction  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

are  represented  in  the  boundary  conditions  with  several  sections.  You  can  therefore  add  more  
information  about  these  parameters  and  by  that  making  them  less  sensitive.    

The  two  variants  (1&4)  are  very  similar  in  results.  Both  the  variants  do  not  really  solve  the  sedimentation  
of  the  approach  channel.  It  will  fill  up  eventually  but  it  takes  a  bit  more  time  in  variant    (4).  So  in  the  end  
a  dredging  program  is  needed  annually  for  both  variants  to  provide  a  sufficient  depth  in  the  approach  
channel.  Variant  (1),  the  extended  breakwater,  is  chosen  as    the  best  solution.  This  is  for  practical  
reasons,  it  is  easier  to  build  one  breakwater  instead  of  three  short  ones.    

5.3  Recommendations  
In  this  sections  some  recommendations  are  given  to  improve  the  model  results:  
•   Multiple  layers  in  water  column;  see  paragraph  5.2.1.  
•   Runtime:  a  longer  runtime  so  that  also  long-­‐term  results  are  available.  
•   More   detailed   input   parameters:   such   as   a   more   specific   Dn50   at   different   locations   and   more  
wave  conditions.  See  paragraph  5.2.2.  
More  detailed  look  at  the  different  variants  for  example  a  better  specification  of  the  dimensions.  

And  last  but  not  least,  as  stated  in  the  conclusion  and  discussion,  there  are  too  many  uncertainties  in  the  
model  that  need  more  research.  There  is  more  time  necessary  to  determine  the  best  dimensions  for  the  
extension   so   that   the   negative   side   effects   (accretion   on   the   channel   side   and   increase   of   velocities  
perpendicular  to  the  entrance  channel)  of  this  solution  are  eliminated  as  much  as  possible.      

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

Sources  
 
[1]   http://oss.deltares.nl/documents/183920/185723/Delft3D-­‐FLOW_User_Manual.pdf  
   

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)


lOMoARcPSD|1717990

Appendix  
 

Long  term  estimation,  3  months  vs.  1  year  


As  seen  in  the  report,  1  year  simulations  are  used  to  look  at  the  effectiveness  of  the  variants.  However,  
it  was  first  thought  that  a  3  month  simulation  was  good  enough  for  a  long  term  estimation.  

 
Figure  23:  (left)  3  month  simulation,  (right)  1  year  simulation.  

Figure  (21)  covers  the  results  of  variant  (1).  It  is  clearly  visible  that  a  3  month  estimation  differs  a  lot  
from  a  1  year  estimation.  This  is  also  the  case  for  the  other  variants.  Therefore  it  can  be  concluded  that  
a  3  month  estimation  can  not  be  used  for  a  long  term  estimation.  

Distributing prohibited | Downloaded by mazen alqadi (mazen.a.alqadi@gmail.com)

You might also like