Water quality and aquatic ecology modelling suite

D-Water Quality

Sediment Water Interaction

User Manual
Sediment Water Interaction

User Manual

D-Water Quality

Version: 3.0
SVN Revision: 52614

October 1, 2017
Sediment Water Interaction, User Manual

Published and printed by:

Deltares telephone: +31 88 335 82 73
Boussinesqweg 1 fax: +31 88 335 85 82
2629 HV Delft e-mail: info@deltares.nl
P.O. 177 www: https://www.deltares.nl
2600 MH Delft
The Netherlands

For sales contact: For support contact:

telephone: +31 88 335 81 88 telephone: +31 88 335 81 00
fax: +31 88 335 81 11 fax: +31 88 335 81 11
e-mail: sales@deltaressystems.nl e-mail: support@deltaressystems.nl
www: http://www.deltaressystems.nl www: http://www.deltaressystems.nl

Copyright © 2017 Deltares

All rights reserved. No part of this document may be reproduced in any form by print, photo
print, photo copy, microfilm or any other means, without written permission from the publisher:
Deltares.
 Contents

Contents
Preface 1

1 Introduction 3

2 Modelling of sediment-water interaction with D-Water Quality 5

2.1 Modelling concepts and development status . . . . . . . . . . . . . . . . . 5
2.2 Mass transport in the sediment . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Input specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.1 The sediment grid . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.2 Sediment boundaries . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.3 Sediment layer thickness . . . . . . . . . . . . . . . . . . . . . . . 12
2.3.4 Porosity and initial sediment composition . . . . . . . . . . . . . . . 12
2.3.5 Dispersive transport . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3.6 Advective transport . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3.7 Process coefficients . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 Output specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

References 15

Deltares iii
Sediment Water Interaction, User Manual

iv Deltares
Preface
This additional manual deals with the input for new functionality that is not supported in the
user interfaces, in particular the concepts of and the additional input for the modelling of
sediment quality and sediment-water interaction. The use of the new functionality requires
editing of D-Water Quality’s ascii input files. A full description of the D-Water Quality input file
can be found in D-WAQ Input file (2013). This document describes what to add to these files
for multilayer sediment quality and sediment-water interaction.

Deltares 1 of 18
Sediment Water Interaction, User Manual

2 of 18 Deltares
1 Introduction
This manual constitutes an addition to the main user manuals for D-Water Quality, which can
be used stand-alone but is also included in the modeling suits Sobek and Delft3D. The main
user manuals describe the creation of input via the dedicated graphical user interfaces. The
present additional manual deals with the input for new functionalities that are not supported
by the user interfaces, in particular for the modelling of sediment quality and sediment-water
interaction.

The use of the new functionalities requires editing of D-Water Quality’s ascii input files. The
information allows the user to edit basic input files that have been created with Delft3D’s or
Sobek’s user interface. A full description of the input file can be found in D-WAQ Input file
(2013). It is advisable to start reading this manual first to understand the principles of the
input file. A warning should be given here: the input file is generally the output file of the
GUI. This means that your modifications generally do not enter into the GUI again and any
modified ASCII file will generally be overwritten by the GUI if you do not make the precaution
to store it in a safe place and/or under a safe name.

chapter 2 of this manual is devoted to sediment-water interaction modeling with D-Water Qual-
ity. Consecutively, the modeling concept for sediment-water interaction, the formulations for
mass transport in the sediment and the sediment specific input are described.

For other model documentation on the structure, concepts, basic formulations and numerical
methods the user is referred to the D-WAQ UM (2013) and D-WAQ Input file (2013). The same
holds for guidance as to running the model and error messages. Moreover, the description of
simulated substances, process formulations, and input and output parameters is not included
in this manual. The D-Water Quality user is referred to the D-WAQ TRM (2013), for a full
account of the contents of D-Water Quality’s extensive processes library.

Deltares 3 of 18
Sediment Water Interaction, User Manual

4 of 18 Deltares
 2 Modelling of sediment-water interaction with D-Water Quality

2.1 Modelling concepts and development status

Originally D-Water Quality contained two alternative options for sediment-water interaction
modelling, the so-called S1/S2 approach and the SWITCH module.

The S1/S2 approach concerns simplified modelling: The active sediment layer is not modelled
as such, but is represented by one or two sets of settled substances that are decomposed,
which leads to return fluxes of nutrients, a consumption flux of oxygen and burial fluxes. The
S1/S2 approach is still available and is described in D-Water Quality’s Technical Reference
Manual.

SWITCH was a rather complex process routine that simulates the vertical concentrations gra-
dients of organic matter, oxygen, nitrate and nutrients (N, P, Si) taking into account 4 sediment
layers, and that generates similar fluxes as well as the denitrification flux and the phosphate
adsorption and precipitation fluxes. SWITCH was fixed with respect to the substances and
layers, and was therefore not generic and not very flexible. SWITCH is no longer available in
D-Water Quality.

In 2002 a start was made to extend D-Water Quality with the functionalities and process
formulations needed to simulate sediment-water interaction and sediment diagenesis in a
generic and comprehensive way. To fully and optimally take into account sediment-water
interaction in water quality modelling D-Water Quality now facilitates the simulation of both
water layers and sediment layers. The sediment and water quality processes are identical
as to substances and kinetic formulations. The process formulations are generic. However,
the local chemical conditions such as the dissolved oxygen concentration determine how
processes turn out. D-Water Quality has been extended with:

 a computational grid for the active sediment layer (in addition to the computational grid for
the overlying water), including a "deep" sediment boundary for the lower sediment layer
deep;
 keywords to define this grid in D-Water Quality’s input;
 a process routine for vertical mass transport (included in D-Water Quality’s processes
library), that also delivers the exchange fluxes between sediment and overlying water, and
between the lower sediment layer and the "deep" sediment boundary;
 input parameters to define the number and thickness of sediment layers (cells in the com-
putational grid), the porosity of the sediment, the seepage velocity (Vseep), the bioturba-
tion dispersion coefficient for solid substances (TurCoef) and the bio-irrigation dispersion
coefficient for dissolved substances (DifCoef);
 a number of substances (sulphate, dissolved sulphide, particulate sulphide, apatite-phosphate,
vivianite-phosphate, methane) and redox processes that are particularly relevant for sedi-
ment diagenesis (included in D-Water Quality’s processes library);
 input parameters for these processes, such as kinetic coefficients;
 exchange fluxes between microphytobenthos biomass (GEMMPB) and benthic grazer
biomass (CONSBL) modules and substances in the sediment;
 exchange fluxes for settling and resuspension between the lower water layer and the upper
sediment layer;
 a numerical integration scheme for explicit integration in horizontal directions and implicit
integration in the vertical direction (extended scheme 12).

Settling and resuspension fluxes and required input were already available in D-Water Qual-
ity. The initial composition of the sediment and the lower boundary concentrations for the
sediment are specified in D-Water Quality’s input in exactly the same way as for water com-

Deltares 5 of 18
 Sediment Water Interaction, User Manual

partments. The same holds for simulation output.

D-Water Quality as based on the comprehensive configuration Delft3D-ECO for eutrophication

modelling has now been tested extensively for cases with settling only, in which sediment layer
thickness and sediment porosity are constant over time. A correct application was also made
for organic micro-pollutants. Further testing needs to be done for resuspension, transient layer
thickness, transient porosity and heavy metals. Transient porosity as based on consolidation
formulations has not yet been implemented.

2.2 Mass transport in the sediment

At present only vertical mass transport is considered for bottom sediment. Vertical mass trans-
port between the sediment layers, across the sediment-water interface and across the lower
sediment boundary is calculated in D-Water Quality by a set of process routines integrated
in D-Water Quality’s processes library (AdvTra, DspTtr, TraSe2_X or TrSe2_X). Processes
TraSe2_X (or TrSe2_X depending on the length of name X) are substance specific where X
indicates the name of a simulated substance. Some of these names have been shorted to
avoid the process name being longer than 10 characters (e.g. APAT and SAL in stead of AP-
ATP and Salinity). Like all processes in D-Water Quality these processes need to be activated
in section 6 of D-Water Quality’s input by means of Active_Y, where Y contains the full name
of the process. An example is:

CONSTANTS ’ACTIVE_AdvTra’ DATA 1.00000e+000

CONSTANTS ’ACTIVE_DspTra’ DATA 1.00000e+000
CONSTANTS ’ACTIVE_TraSe2_AAP’ DATA 1.00000e+000

Vertical mass transport processes in the sediment may include burial, digging, seepage, bio-
turbation, bio-irrigation, flow induced dispersion and molecular diffusion. Burial, digging and
seepage (upwelling or downwelling) are forms of advection. Bioturbation of solid substances
and bio-irrigation of dissolved substances are forms of dispersion. All dispersive processes
are formulated as diffusion.

The burial of particulate substances results from net settling at the sediment-water interface,
digging results from net resuspension at this interface. The advection of particulate sub-
stances by burial or digging follows from:

vp × f p × C
F advp = (2.1)
(1 − ϕ)
where:
C bulk concentration of a substance [g m−3 ]
F adv p particulate advection flux [g m−2 d−1 ]
fp particulate fraction of a substance [-]
vp volumetric burial or digging velocity [m d−1 ]
ϕ porosity [-]

Fraction f p is equal to 1.0, except for micro-pollutants. The model calculates fp for these
substances as it depends on adsorption.

For fixed porosities and fixed layer thickness burial and digging imply transport fluxes across
all the interfaces of the sediment layers. This includes the interface of the lower active sed-
iment layer in the model and the deeper inactive sediment (boundary condition). The burial
and digging velocities vp are calculated in the model from the settling and resuspension fluxes

6 of 18 Deltares
 Modelling of sediment-water interaction with D-Water Quality

in such a way that constant porosity in and constant thickness (volume) of each sediment layer
is maintained. This uses the following definition of porosity:
n
X f pi × Ci
ϕ=1− (2.2)
i=1
ρi
where:
ϕ porosity [-]
Ci bulk concentration of a substance [g m−3 ]
f pi particulate fraction of a substance [-]
ρi density of a solid matter component [g m−2 d−1 ]
i index of a solid matter component [-]
n number of solid matter components [-]

For transient layer thickness or for transient porosity the volumetric burial or digging velocity
vp is the sum of an imposed velocity (input parameter, see Sections 2.3.3 and 2.3.6) and an
additional velocity to maintain maximal layer thickness or minimal layer thickness. In the case
of fixed porosity the additional velocity also serves to maintain porosity at its imposed value
(see subsection 2.3.3). The additional velocity is calculated within the model.

Porosity ϕ is the input parameter Porinp (see subsection 2.3.4). Porinp can be used for fixed
porosity (constant) as well as transient porosity (time series).

The settling fluxes for the water column are calculated according to the settling process de-
scribed in D-Water Quality’s Technical Reference Manual using input settling velocities. Re-
suspension can be simulated according to Partheniades-Krone formulations or De Boer for-
mulations (input option parameter SwErosion is 0.0 or 1.0 respectively).

Seepage can be upwelling or downwelling (infiltration). It affects only the dissolved sub-
stances. Seepage implies transport fluxes across the sediment-water interface, the interfaces
of the sediment layers, and the interface of the active sediment and the deeper inactive sedi-
ment. The seepage advection flux is:
vd × f d × C
F advd = (2.3)
ϕ
where:
C bulk concentration of a substance [g m−3 ]
F adv d dissolved advection flux [g m−2 d−1 ]
fd dissolved fraction of a substance [-]
vd volumetric seepage velocity [m d−1 ]
ϕ porosity [-]

The fraction f d is equal to 1.0, except for micro-pollutants (calculated by the model as 1 −
f p). The seepage velocity is the input parameter Vseep (see subsection 2.3.6), which has a
negative value in the case of downwelling.

Bioturbation by benthic organisms causes the dispersion of particulate substances. The per-
tinent dispersion flux is approximated with:

(f p1 × C1 / (1 − ϕ1 ) − f p2 × C2 / (1 − ϕ2 ))
F disp = max (1 − ϕ1 , 1 − ϕ2 ) ×Dp×
(L1 + L2 )
(2.4)
where:

Deltares 7 of 18
 Sediment Water Interaction, User Manual

C bulk concentration of a substance [g m−3 ]

Dp particulate dispersion coefficient [m2 d−1 ]
F disp particulate dispersion flux [g m−2 d−1 ]
fp particulate fraction of a substance [-]
L dispersion length [m]
ϕ porosity [-]

indexes 1 and 2 refer to two adjacent sediment layers (grid cells)

The dispersion coefficient Dp is the input parameter TurCoef (see subsection 2.3.5). Each
dispersion length L is the half thickness of the sediment layer concerned. The bioturbation
flux is zero at the sediment-water interface.

Benthic organisms also cause bio-irrigation, the dispersion of dissolved substances. Water
flow over the sediment causes micro-turbulence in the upper pore water, which is another
source of dispersion. The overall dispersion coefficient includes the effects of bio-irrigation,
flow and molecular diffusion. The dispersion of dissolved substances implies transport fluxes
across the sediment-water interface. These fluxes include the so-called return fluxes of nutri-
ents to the water column and the sediment oxygen consumption flux. The dispersion flux of a
solute follows from:

(f d1 × C1 /ϕ1 − f d2 × C2 /ϕ2 )
F disd = min (ϕ1 , ϕ2 ) × Dd × (2.5)
(L1 + L2 )
where:
C bulk concentration of a substance [g m−3 ]
Dd diffusion or dispersion coefficient [m2 d−1 ]
F disd dissolved dispersion flux [g m−2 d−1 ]
fd dissolved fraction of a substance [-]
L dispersion length [m]
ϕ porosity [-]

indexes 1 and 2 refer to two adjacent sediment layers (grid cells)

The dispersion coefficient Dd is the input parameter DifCoef (see subsection 2.3.5). Each
dispersion length L is the half thickness of the sediment layer concerned. For the sediment-
water interface L1 in the lower water layer is input parameter Diflen. The bio-irrigation flux
is zero at the interface of the lower active sediment layer and the deeper sediment (lower
boundary).

2.3 Input specification

2.3.1 The sediment grid

Exchanging substances with the water body, the sediment bed is modelled with an additional
computational grid. Each sediment layer is a flat relatively thin cell in this grid. Together
the sediment layers form a sediment segment, which is in fact a sediment column. Each
sediment segment is connected to a cell in the computational grid for the water body (water

8 of 18 Deltares
 Modelling of sediment-water interaction with D-Water Quality

grid), namely the lower water layer of a water segment, which is in fact a water column. The
set of sediment segments constitutes the sediment grid. This sediment grid must be defined
using the MULTIGRID keyword just after the number of water segments in block 3 of the
input file. The MULTIGRID part of the input consists of the input items described below as
based on an input example. The MULTIGRID part of the input file must end with the keyword
END_MULTIGRID.

The keyword ZMODEL indicates that the underlying hydrodynamic model uses a z -layer ap-
proach (fixed thickness). If omitted, a σ -layer or 2D model approach is assumed, for which
layer thickness is relative to water depth.

The keyword NOLAY is to be followed by the number of water layers in the (aggregated) water
grid.

Several sediment grid files can be defined in the input file, but the first one specified is the
one that defines the computational grid for the sediment in the D-Water Quality simulation.
The other grid files can be used for the specification of input items. The grid can be divided
in zones. A zone comprises a number of sediment segments that have similar properties.
Zones can be different as to the number of sediment layers and input items such as initial
composition, temperature forcing, and process coefficients. Sediment segments as well as
zones are identified by sequence numbers. The sediment segments must be sequentially
numbered from 1 to M (the number of segments). The zones must be sequentially numbered
from 1 to N (the number of zones).

The keyword BOTTOMGRID is used to identify the sediment grid and the various grid files. It
is to be followed by the name of the sediment grid or a sediment zone. This combination is
(optionally) followed by the keyword REFERENCEGRID and the name of the reference grid.
This set of keywords can be repeated several times to further specify the sediment grid and
the zones:

 The first keyword BOTTOMGRID is used to identify the sediment grid that defines the
computational grid in the D-Water Quality simulation. It is followed by the name of the
sediment grid. This combination is followed by the keyword REFERENCEGRID and the
name of the reference grid. The reference grid is the water grid for which the sediment
grid is specified. By default this is the (aggregated) grid of the water body in the model,
and has the prescribed name ‘Base grid’.
 After this the specification of the sediment grid follows, which consists of a table with the
numbers of the sediment segments for each water segment in the reference grid. The
segment numbers must be assigned in the order of the water grid cells. That means that
the first record is connected to water grid cell 1, and the last record is connected to water
grid cell L (the number of water grid cells in a water layer). Water grid cells without a
sediment bed must be assigned a zero in this table.
 The second keyword BOTTOMGRID defines the zoning of the sediment segments. It is
followed by the name of the sediment grid to which zoning applies. The reference grid is
usually the entire sediment grid. After this the specification of the zones follows, which
consists of a table with the numbers of the zones for each sediment segment. The zone
numbers must be allocated in the order of the sediment segments. That means that the
first record is connected to segment 1, and the last record is connected to segment M.
 The third keyword BOTTOMGRID can be used to create a sediment zone that comprises
all sediment segments, which can be conveniently called ‘sediment-default’. This zone
can be used to allocate the same input item values to all segments. It is followed by a
reference grid specification and value 1, which refers to the ‘Base grid’.

The keyword NOBOTTOMLAY is used to specify the number of layers in the sediment. The

Deltares 9 of 18
Sediment Water Interaction, User Manual

keyword must be followed by an additional keyword indicating for which segments the number
of layers are specified:

 A possible keyword is ALL, which indicates that the number of layers given concerns all
sediment segments. The keyword DEFAULT indicates that the default number of layers is
applied to all sediment segments.
 If the sediment grid has zones, then the number of layers can be defined for each of the
zones. This requires the keyword INPUTGRID placed immediately after NOBOTTOMLAY,
followed by the name of the grid that has the zones and the number of the layers in each
of the zones. The numbers of layers must be given in the order of the sequence numbers
of the zones (number of layers for zone 1, number for zone 2, number for zone 3, etc.).

Zoning of the water body can be achieved by defining sub-grids with the keyword SUBGRID.
The definition of a water sub-grid is analogous to the definition of a sediment zone sub-grid. A
water sub-grid (zone) is only used to structure the input items such as the initial composition.
The same value of an input item can be allocated to all cells in a sub-grid (zone). The keyword
SUBGRID is followed by a name identifying the sub-grid (zone). This set is optionally followed
by the keyword REFERENCEGRID with the name of the reference grid, on which the sub-grid
is defined. The reference grid is by default the model grid which is called the ‘Base grid’, but
it could also be a previously defined sub-grid. After the name the specification of the sub-grid
follows which consists of a table with zone sequence numbers for each grid cell in the water
grid. The zones must be allocated in the order of the grid cells in the water grid. That means
that the first record is connected to grid cell 1. The last record is connected to grid cell L (the
number of cells in a water layer).

A typical example of the specification of the sediment grid using included files is as follows:

#2 ; delimiter for the second block

;
; third block of model input (grid layout)
1212 ; number of grid cells in the water body (all water layers)
;
MULTIGRID
; z layer model with 4 layers
ZMODEL
NOLAY 4
; specification of the computational sediment grid
BOTTOMGRID ’sediment-grid’
REFERENCEGRID ’Base grid’
INCLUDE ’../model-input/grid/sediment-ag02.dwq’
; allocation of the segments to zones
BOTTOMGRID ’sediment-zone’
REFERENCEGRID ’sediment-grid’
INCLUDE ’../model-input/grid/sediment-zone-ag02.dwq’
; specification of a default zone
BOTTOMGRID ’sediment-default’
REFERENCEGRID ’sediment-zone’
INCLUDE ’../model-input/grid/sediment-default.dwq’
; all sediments segments have 7 layers, this example has 1 zone
NOBOTTOMLAY INPUTGRID sediment-default 7
; allocation of the water grid cells to one zone for input purposes
SUBGRID ’water-zone’
INCLUDE ’../model-input/grid/water-zone-ag02.dwq’
END_MULTIGRID

In the resulting overall computational grid as created by DELWAQ1 the sediment segments are
placed after the water grid cells, and sediment boundaries are added for each sediment seg-

10 of 18 Deltares
 Modelling of sediment-water interaction with D-Water Quality

ment. If the number of sediment layers is the same for all sediment segments, the numbering
of the grid cells for K water layers and P sediment layers proceeds as follows:

 The grid cells in the top water layer run from 1 to L. The grid cells in the second water layer
are numbered from 1+L to 2*L (etc.). The grid cells in the lower water layer are numbered
from 1+(K-1)*L to K*L. The grid cells in a water segment (column) are numbered n, n+L,
n+2*L, ..., n+(K-1)*L.
 The grid cells in the top sediment layer run from K*L+1 to K*L+P*M. The grid cells in the
second sediment layer are numbered from K*L+1+M to K*L+2*M (etc.)., The grid cells in
the lower sediment layer are numbered from K*L+1+(P-1)*M to K*L+P*M. The grid cells in
a sediment segment (column) are numbered n+K*L, n+K*L+M, n+K*L+2*M, ..., n+K*L+(P-
1)*M.

Notice that M is equal to L when each water segment (column) has a sediment segment
(column).

2.3.2 Sediment boundaries

"Deep" sediment boundaries are added by DELWAQ1. If the water grid has B boundaries,
DELWAQ1 will number the sediment boundaries conform the sediment segment numbering
from B+1 to B+M. The first boundary belongs to sediment grid cell K*L+1+(P-1)*M of grid seg-
ment 1. The last sediment boundary belongs to sediment grid cell K*L+P*M of grid segment
M.

The definition of the names sediment boundaries follows the rules for water boundaries as
described in D-WAQ Input file (2013). In the input file it must appear in block 5 immediately
after the names of all water boundaries. Names are user determined. An example for M is 10
looks as follows:

#4 ; delimiter for the fourth block

;
; specification of names of water boundaries
;

;
; specification of names of sediment boundaries
;
’sedbnd0001’ ’sedbnd0001’ ’sedimentbound’
’sedbnd0002’ ’sedbnd0002’ ’sedimentbound’
’sedbnd0003’ ’sedbnd0003’ ’sedimentbound’
’sedbnd0004’ ’sedbnd0004’ ’sedimentbound’
’sedbnd0005’ ’sedbnd0005’ ’sedimentbound’
’sedbnd0006’ ’sedbnd0006’ ’sedimentbound’
’sedbnd0007’ ’sedbnd0007’ ’sedimentbound’
’sedbnd0008’ ’sedbnd0008’ ’sedimentbound’
’sedbnd0009’ ’sedbnd0009’ ’sedimentbound’
’sedbnd0010’ ’sedbnd0010’ ’sedimentbound’

Concentrations must be allocated to the simulated substances for the sediment boundaries.
This allocation proceeds in the same way as for water boundaries, starting with keyword
ITEM followed by the name of an individual boundary(e.g. ’sedbnd0001’), or the name of
a boundary zone (e.g.’sedimentbound’). The concentrations of the particulate substances
should be approximately equal to the concentrations of the same substances in the lower
sediment layer.

Deltares 11 of 18
 Sediment Water Interaction, User Manual

2.3.3 Sediment layer thickness

The thicknesses of the sediment layers are to be defined in block 7 of D-Water Quality’s
input. Fixed layer thickness applies when input option parameter SWSediment is equal to
zero (default value). If SWSediment is equal to 1.0, layer thickness will be transient (variable).

For layer thickness three parameters can be defined. FixTh is used to quantify fixed layer
thicknesses. MaxTh and MinTh specify the maximal and minimal layer thickness in the case of
transient layer thickness. Layer thickness is to be specified as a parameter for each sediment
layer in a sediment zone, or for each sediment layer in a sediment segment when zoning is not
applied (see also subsection 2.3.1). The unit is meter. The specification for a zone requires
the keyword INPUTGRID followed by the name of a zone (sub-grid). The sequence of the
various keywords and the data for thickness follows from the following example:

PARAMETERS ’fixth’ INPUTGRID ’sediment-zone’

DATA
0.001 ; layer 1
0.002 ; layer 2
0.003 ; layer 3
0.006 ; layer 4
0.018 ; layer 5
0.07 ; layer 6
0.1 ; layer 7

In the case of steep concentration gradients the top layer should be 0.001 m thick. This is a
feasible minimal value in order not to get unacceptably large computational burden (simulation
run time) due to a very small timestep for integration. With such a top layer the timestep will
mostly be between 2 and 5 minutes. The thickness of the other layers should increase with
depth, in order to not have too many layers, which will also cause unacceptable computational
burden. The overall thickness of the active sediment layer is 0.1–0.2 meter in fresh water
systems, and 0.3–0.4 m in saline and brackish water systems.

2.3.4 Porosity and initial sediment composition

Like any input parameter in block 7 of the input the porosity Porinp (m3 water/m3 bulk sediment)
can be provided as a constant, a parameter, a time function, or a segment function. It is rec-
ommended to use a constant or a parameter, for D-Water Quality has not yet been fully tested
for the porosity varying over time. In the case of a parameter, the porosity is quanitified for
each layer (grid cell) in each sediment zone or each sediment segment. See subsection 2.3.3
for a typical example of parameter specification.

The porosity is made equal to the input value if Porinp is larger than 10−4 . A smaller input
value can be used to inform the model that porosity will be transient (variable over time) as
simulated by the model. The initial porosity is then derived from the initial composition (im-
posed concentrations of particulate substances). However, the option for transient computed
porosity would require consolidation formulations, which have not been implemented yet.

Realistic values for the porosity in a sediment bed are in between 0.35 (pure sand) and 0.9
(peat or peaty silt). A silt sediment has a porosity of approximately 0.7.

2.3.5 Dispersive transport

The dispersion coefficients TurCoef for particulate matter and Difcoef for solutes need to be
specified in block 7 as a constant, a parameter, a time function or a segment function. It
is recommended to specify these coefficients as segment functions (both space and time

12 of 18 Deltares
 Modelling of sediment-water interaction with D-Water Quality

dependent) for each sediment zone or for each sediment segment. The bioturbation and
bio-irrigation coefficients vary seasonally, dependent on the activity of benthic organisms. A
typical example for input for a segment function is:

SEG_FUNCTIONS
DIFCOEF
INPUTGRID ‘sediment-zone’
DATA
;
; Layer 1 Layer 2 Layer 3 Layer 4 Layer 5 Layer 6 La
;
2004/01/01-00:00:00 3.00E-04 3.00E-04 3.00E-04 3.00E-04 3.00E-04 1.00E-04 5.0
2006/12/31-00:00:00 3.00E-04 3.00E-04 3.00E-04 3.00E-04 3.00E-04 1.00E-04 5.0

TurCoef affects mass transport across all sediment interfaces, except for the sediment-water
interface. The first given value concerns the interface between the top sediment layer and the
second layer. A representative summer value for TurCoef near the sediment-water interface
for a shallow freshwater system is 2.0 × 10−6 m2 /day. The winter value can be 10 % of
the summer value. Bioturbation can be significantly faster in marine sediments. TurCoef
decreases exponentially with depth, and is practically equal to zero at depths below 0.1 m in
freshwater systems, and below 0.4 m in marine water systems.

DifCoef affects mass transport across all sediment interfaces, except for the lower sediment
boundary. Any value given for this interface will be ignored. The first given value concerns
the sediment-water interface. A representative summer value for DifCoef near the sediment-
water interface for a shallow freshwater system is 5.0 × 10−4 m2 /day. This value is the
sum of bio-irrigation, flow induced dispersion and molecular diffusion. The winter value can
be 20% of the summer value. Bio-irrigation can be significantly faster in marine sediments.
DifCoef decreases exponentially with depth, and is practically equal to the molecular diffusion
coefficient corrected for tortuosity (ϕ2 ) at depths below 0.1 m in freshwater systems, and
below 0.4 m in marine water systems. A representative value for the corrected molecular
diffusion coefficient is 0.25 × 10−4 m2 /day.

The dispersion length at the water side of the sediment-water interface Diflen can usually be
provided as a constant value between 0.0005 and 0.001 m.

2.3.6 Advective transport

The seepage velocity Vseep (m/day) is the only advection parameter needed in input block
7 when layer thickness and porosity are both fixed. Burial and digging are calculated within
the model. A positive value ofVseep will cause an upward transport, a negative value a
downward transport. Vseep can be specified as as a constant, a parameter (example in
subsection 2.3.4), a time function or a segment function (example in subsection 2.3.5). In the
cases of a parameter and a segment function Vseep can be specified for each sediment zone
or for each sediment segment. An typical example of a time function is:

FUNCTIONS
VSEEP
BLOCK
DATA
;
2009/01/01-00:00:00 0.01
2010/12/31-00:00:00 0.01

Deltares 13 of 18
 Sediment Water Interaction, User Manual

In the case of transient layer thickness the volumetric burial and digging velocity needs to be
provided as input parameter VburDM (m/day).

2.3.7 Process coefficients

All coefficients of water and sediment quality processes are specified in D-Water Quality’s
Technical Reference Manual and in updated process descriptions. All coefficients are speci-
fied in block 7 of the input as a constant, a parameter, a time function or a segment function.
Values can be different for water grid cells and sediment grid cells. Some parameters can
even vary among sediment grid cells. This holds in particular for:

 RcNit20 , the nitrification rate at 20 ◦ C

 fr_FeIM1 , the fraction of reactive iron in sediment fraction IM1
 fr_FeIM2 , the fraction of reactive iron in sediment fraction IM2
 fr_FeIM3 , the fraction of reactive iron in sediment fraction IM3
 fr_Feox , the fraction of oxidised iron in total iron under reducing conditions
 KadsP_20 , the adsorption equilibrium constant for phosphate
 CoxDenInh , the inhibition concentration of oxygen for denitrification
 pH_min, the minimal pH
 pH_max, the maximal pH

In the cases of a parameter and a segment function a coefficient can be specified for each
sediment zone or for each sediment segment. A typical example of a parameter that is differ-
ent for the water column and the sediment is:

; water grid cells sediment grid cells

PARAMETERS RcNit20 ALL DATA 820*0.1 126*40.0

The number of sediment grid cells is the sum of the numbers of layers in each of the sediment
segments.

2.4 Output specification

All output items are essentially the same for water grid cells and sediment grid cells, implying
that (additional) output parameters need to be specified only once. Poros is the name of the
output parameter for porosity. Like for water grid cells, the output concentrations for sediment
grid cells are bulk concentrations in g/m3 . The concentrations of solutes in pore water can be
derived from the bulk concentrations by dividing bulk concentrations by porosity.

All output of sediment grid cells are added to the standard map output. Sediment segments
can be added as monitoring point or area for monitoring, history and balance output. A
monitoring point or area can be excluded from the balance output by adding the keyword
NO_BALANCE directly after the name of the monitoring point or area.

Process fluxes in the balance output are dealt with in the same way for water and sediment
compartments. All parts of the sediment water interaction are lumped and reported in the bal-
ance as transport, because the net flux is calculated by the D-Water Quality transport solver.
If the user wants output for the individual sediment-water exchange fluxes, he is advised to
include both the water and the sediment compartment as a monitoring point or area, in com-
bination with the balance option “not lump internal transport”.

14 of 18 Deltares
References
D-WAQ Input file, 2013. D-Water Quality Description of Input file. Deltares, 2.00 ed.

D-WAQ TRM, 2013. D-Water Quality Technical Reference Manual. Deltares, 4.00 ed.

D-WAQ UM, 2013. D-Water Quality User Manual. Deltares, 4.02 ed.

Deltares 15 of 18
Sediment Water Interaction, User Manual

16 of 18 Deltares
 Photo by: Mathilde Matthijsse, www.zeelandonderwater.nl

PO Box 177 +31 (0)88 335 81 88

2600 MH Delft sales@deltaressystems.nl
Rotterdamseweg 185 www.deltaressystems.nl
2629 HD Delft
The Netherlands