February 24, 1999
EPA-SAB-EC-99-011
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, SW
Washington, DC 20460
Subject: An SAB Report: Review of the D-CORMIX Model
Dear Ms. Browner:
In response to a request from the Office of Science and Technology (OST) in EPA's
Office of Water (OW), the Science Advisory Board (SAB) convened the D-CORMIX Review
Subcommittee to conduct an external peer review of the D-CORMIX model. The Subcommittee
met in public session on August 25-26, 1998 in Washington, DC and reviewed a number of
technical aspects as well as implementation issues with the D-CORMIX model for mixing zone
analysis. The Subcommittee members were given several written documents and a PC-version of
the D-CORMIX model to review prior to the meeting, and then had the opportunity to interact
with EPA staff, EPA Contractor staff, and staff of the US Army Corps of Engineers during the
public meeting.
There were four charge questions related to the technical aspects and one charge question
related to the implementation of the model with regard to use of an allocated impact zone (AIZ)
for negatively buoyant wastewater discharge. These are summarized as follows:
a) Is D-CORMIX an appropriate mixing zone model to use for continuous
dredged material discharge mixing zone analysis? D-CORMIX has the
potential for the mixing zone analysis of dredged material discharges, assuming
additional validation as suggested in Section 3.4 in the attached report is
completed. Nevertheless, the appropriateness of D-CORMIX for intended
applications will depend on the level of detail and sophistication required for the
analysis. For those applications for which suspended solids are the only
contaminant of interest, D-CORMIX might be said to be partially validated.
However, if intended applications for predicting mixing zones include the need to
predict dissolved phase contaminant distributions that originate from chemicals
originally bound to the sediment, then D-CORMIX must be considered to be, at
present, unvalidated. Nonetheless, it is important that an inability of D-CORMIX
to handle one set of processes, such as contaminant interactions and fate, does not
-------
invalidate it as an appropriate model for cases in which mixing and sedimentation
are the dominant processes.
b) Does the model accurately capture the physics of negatively buoyant surface
plumes, in particular, behavior of the density current and particle settling
associated with dredged disposal plumes? D-CORMIX shows mixed results
with respect to the predicted distribution of suspended solids and associated
changes in density in comparison with several laboratory data sets. D-CORMIX
appears to warrant additional validation as outlined in Section 3.4 in the attached
report. One of the Subcommittee members had access to several laboratory data
sets and simulated these using both the CORMIX and D-CORMIX models
(properly accounting for the issue associated with positive or negative buoyancy as
required by the particular model). A comparison of the simulations led to the
following conclusions: (1) in most situations, the models performed relatively well
in predicting dilutions and plume dimensions observed in the experiments; (2) in
cases where vertical density gradients were present in the ambient fluid of
sufficient magnitude to trap the plume within the water column, a number of
deviations were noted; and (3) comparison of data in which plumes impinge
directly on a boundary indicates that the model conservatively underestimates the
dilution in the initial layer spreading along the surface.
c) What are the essential differences between the D-CORMIX and CD-FATE
models and which is preferable as a mixing zone model for continuous
dredged material discharge? D-CORMIX focuses on the movement and
dilution of a mixture of solids and liquids initially as a sinking jet or plume and then
as a density current once the spreading elevation is attained (either the bottom for
no or weak density stratification or internally at a neutrally buoyant level for a
water column with significant stratification). It accounts for particle settling
removing a portion of the sediment from the density current but does not account
for any of this material to re-enter the water column. CD-FATE calculates a
suspended sediment contribution due to two additional effects; a component
stripped from the descending plume and a second component entrained from the
material deposited on the bottom. These predictions are specified by empirical
models. CD-FATE does not directly compute the sediment distribution in a
bottom density current, but the sediment component entrained from the bottom
deposits resembles the density current to some extent.
d) Does the SAB approve of our outline for laboratory validation? What
further suggestions can be offered? As noted in Section 3.4 below, the
Subcommittee recommends a staged validation process. An initial objective in the
design of any validation study must be to decide whether or not the processes in
D-CORMIX are to be validated or whether it is the ability of D-CORMIX to be
able to represent physically important processes that is being investigated. Even if
-------
additional capabilities were developed for D-CORMIX to properly model chemical
and physio-chemical processes, it is possible for a validation effort to fail if the
mixing/sedimentation processes are not represented properly in the current version
of the model. In light of this, it appears to be more fruitful to approach the
validation effort in a staged manner.
e) What factors should be considered in developing an AIZ that will not
adversely impact the integrity of the aquatic ecosystem? How should the
AIZ be sized, especially in relation to distance from the bottom (substrate),
and portion of water column encompassed? The Subcommittee supports the
continued use of mixing zones as allocated impact zones (AIZs). In addition, we
support Agency efforts to define and model AIZs as they may exist under different
conditions. However, we note that current practices at the state level often result
in designation of AIZs that are not well defined and which do not consider
(measure/estimate) risk to aquatic species or humans. Therefore, the
Subcommittee recommends EPA continue to develop models that can assist in
setting risk-based AIZs.
The Subcommittee appreciates the opportunity to provide advice to the Agency on the
technical, validation, and implementation aspects of the D-CORMIX model. We look forward to
the response of the Assistant Administrator for the Office of Water to the advice contained in this
report.
Sincerely,
/signed/ /signed/
Dr. Joan Daisey, Chair Dr. Ishwar Murarka, Chair
Science Advisory Board D-CORMIX Review Subcommittee
Science Advisory Board
-------
NOTICE
This report has been written as part of the activities of the Science Advisory Board, a
public advisory group providing extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency. The Board is structured to provide
balanced, expert assessment of scientific matters related to problems facing the Agency. This
report has not been reviewed for approval by the Agency and, hence, the contents of this report
do not necessarily represent the views and policies of the Environmental Protection Agency, nor
of other agencies in the Executive Branch of the Federal government, nor does mention of trade
names or commercial products constitute a recommendation for use.
-------
ABSTRACT
The US EPAs Science Advisory Board (SAB) convened the D-CORMIX Review
Subcommittee to conduct an external peer review of the Agency's D-CORMIX model. The
Subcommittee met in public session on August 25-26, 1998 in Washington, DC and reviewed a
number a technical aspects as well as implementation issues with the D-CORMIX model for
mixing zone analysis.
The charge to the Subcommittee is summarized as follows: a) Is D-CORMIX an
appropriate mixing zone model to use for continuous dredged material discharge mixing zone
analysis?; b) Does the model accurately capture the physics of negatively buoyant surface plumes,
in particular, behavior of the density current and particle settling associated with dredged disposal
plumes?; c) What are the essential differences between the D-CORMIX and CD-FATE models
and which is preferable as a mixing zone model for continuous dredged material discharge?); d)
Does the SAB approve of our outline for laboratory validation? What further suggestions can be
offered?; and e) What factors should be considered in developing an AIZ that will not adversely
impact the integrity of the aquatic ecosystem? How should the AIZ be sized, especially in relation
to distance from the bottom (substrate), and portion of water column encompassed?
In its report, the Subcommittee provided responses to the above questions, addressed
several concerns over the actual model itself, and made suggestions for improvements in
validation.
Keywords: D-CORMIX, CD-FATE, mixing zone, water quality, dredge discharge
-------
U.S. Environmental Protection Agency
Science Advisory Board
Cormix Review Subcommittee
CHAIR
Dr. Ishwar Murarka, Chief Scientist and President, ISH Inc., Cupertino, CA (Chair Emeritus of
the Environmental Engineering Committee of the SAB)
SUBCOMMITTEE
Dr. Eric Adams, Senior Research Engineer & Lecturer, MIT, Department of Civil &
Environmental Engineering, Cambridge, MA
Dr. William J. Adams, Director, Environmental Science, Kennecott Utah Copper Corporation,
Magna, UT (Member of the Ecological Processes and Effects Committee of the SAB)
Dr. Alan Maki, Environmental Advisor, Exxon Company, USA, Houston, TX (Member of the
Executive Committee of the SAB and Member of the Ecological Processes and Effects
Committee of the SAB)
Dr. Keith Stolzenbach, Professor, Department of Civil and Environmental Engineering,
University of California, Los Angeles, CA
Dr. Thomas Theis, Professor and Chair, Department of Civil & Environmental Engineering,
Clarkson University, Potsdam, NY
Dr. Steven J. Wright, Professor, Civil & Environmental Engineering, University of Michigan,
Ann Arbor, MI
Note: All Panelists are Consultants to the Science Advisory Board unless otherwise noted
SCIENCE ADVISORY BOARD STAFF
Mr. Robert Flaak, Designated Federal Officer, Environmental Protection Agency, Science
Advisory Board (1400) 401 M Street, SW, Washington, DC 20460
Ms. Wanda R. Fields, Management Assistant, Environmental Protection Agency, Science
Advisory Board (1400), 401 M Street, SW, Washington, DC 20460
in
-------
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 5
2.1 Background 5
2.2 Charge to the Subcommittee 6
3. RESPONSE TO THE CHARGE 8
3.1 Charge Question #1 - Is D-CORMIX an appropriate water quality model to
use for continuous dredged material discharge mixing zone analysis? 8
3.1.1 Definitions 8
3.1.2 Appropriateness of D-CORMIX as a Mixing Model 9
3.1.3 Appropriateness of D-CORMIX as a Sedimentation Model 9
3.1.4 Appropriateness of D-CORMIX as a Model of Non-Conservative
Reacting Constituents 10
3.1.5 Summary 11
3.2 Charge Question #2 - Does the model accurately capture the physics of
negatively buoyant surface plumes, in particular, behavior of the density
current and particle settling associated with dredged disposal plumes? 12
3.2.1 Potential Deficiencies in original CORMIX Model that are
incorporated into D-CORMIX 12
3.2.2 Restrictions on the Situations that D-CORMIX can Analyze 12
3.2.3 Possible Deficiencies in D-CORMIX 13
3.3 Charge Question #3 - What are the essential differences between the
D-CORMIX and CD-FATE models and which is preferable as a mixing zone
model for continuous dredged material discharge? 14
3.4 Charge Question #4 - Does the SAB approve of our outline for laboratory
validation? What further suggestions can be offered? 16
3.4.1 Validation of Mixing Processes 16
3.4.2 Validation of Sedimentation Processes 17
3.4.3 Validation for Non-Conservative Reacting Constituents 17
3.4.4 Need for Field Validation 18
3.5 Charge Question #5 - What factors should be considered in developing an
allocated impact zone (AIZ) that will not adversely impact the integrity of
the aquatic ecosystem? How should the AIZ be sized, especially in relation
to distance from the bottom (substrate), and portion of water column
encompassed? 18
Appendix A - Potential Deficiencies in original CORMIX Model that are incorporated
into D-CORMIX A - 1
REFERENCES CITED R - 1
IV
-------
1. EXECUTIVE SUMMARY
The Science Advisory Board (SAB) convened the CORMIX Review Subcommittee to
conduct an external peer review of the US EPA Office of Water's D-CORMIX model. The
Subcommittee met in public session in Washington DC on August 25-26, 1998. At the meeting
the Subcommitttee discussed the model with staff of both EPA and the US Army Corps of
Engineers. The Agency requested responses from the Subcommittee to four charge questions
related to the technical aspects and one charge question related to the implementation of the
model with regard to use of an allocated impact zone (AIZ) for negatively buoyant wastewater
discharge. These charge questions and the Subcommittee's summary responses follow. More
detailed responses to these questions can be found in Chapter 3.
a) Is D-CORMIX an appropriate mixing zone model to use for continuous dredged
material discharge mixing zone analysis? [see section 3.1 for further details]
The appropriateness of D-CORMIX for intended applications will depend
on the level of detail and sophistication required for the analysis. Limited
validation studies have been undertaken thus far with mixed results with respect to
the predicted distribution of suspended solids and associated changes in density in
comparison with several laboratory and field data sets. Neither the hindered
settling option nor contaminant decay/growth features have been tested. For those
applications for which suspended solids are the only contaminant of interest, D-
CORMIX might be said to be partially validated. However if intended applications
for predicting mixing zones include the need to predict dissolved phase
contaminant distributions that originate from chemicals originally bound to the
sediment, then D-CORMIX must be considered to be, at present, unvalidated.
Nonetheless, it is important that an inability of D-CORMIX to handle one set of
processes, such as contaminant interactions and fate, not invalidate it as an
appropriate model for cases in which mixing and sedimentation are the dominant
processes.
D-CORMIX has the potential for the mixing zone analysis of dredged
material discharges, assuming additional validation as suggested in Section 3.4
below is completed. In its current form, it does not account for any physico-
chemical interactions between the sediment and liquid phases. Due to short transit
times through the mixing zone, this may not be an important limitation for most
applications. It may be appropriate to define specific applications or classes of
applications where this simplification is not appropriate and to clearly identify
these conditions for potential users. D-CORMIX is a modification of CORMIX
that considers the presence of suspended sediment as a density-producing agent
which can subsequently settle out of a density current flowing along the bottom of
the receiving fluid. The deposition is handled by specification of a settling velocity
-------
which is dependent on sediment size. Whether particle settling within the density
current is the only important physical process distinguishing dredged material
plumes from dissolved phase ones and whether the simple settling computation can
be used to estimate sediment removal from the plume needs to be subject to a
more rigorous validation effort than has been conducted to date. In addition, the
nature of above water discharges and their contribution to a splash component of
sediment flux to the environment needs to be further investigated to determine the
significance to local turbidity within the mixing zone.
b) Does the model accurately capture the physics of negatively buoyant surface
plumes, in particular, behavior of the density current and particle settling
associated with dredged disposal plumes? [see section 3.2 for further details]
D-CORMIX shows mixed results with respect to the predicted distribution
of suspended solids and associated changes in density in comparison with several
laboratory data sets. D-CORMIX appears to warrant additional validation as
outlined in Section 3.4 below. One of the Subcommittee members had access to
several laboratory data sets and simulated these using both the CORMIX and D-
CORMIX models (properly accounting for the issue associated with positive or
negative buoyancy as required by the particular model). A comparison of the
simulations led to the following conclusions: (1) in most situations, the models
performed relatively well in predicting dilutions and plume dimensions observed in
the experiments; (2) in cases where vertical density gradients were present in the
ambient fluid of sufficient magnitude to trap the plume within the water column, a
number of deviations were noted; and (3) comparison of data in which plumes
impinge directly on a boundary indicates that the model conservatively
underestimates the dilution in the initial layer spreading along the surface.
In addition to the above mentioned issues, there are additional issues which
have not been adequately explored. Concerns were expressed with regards to a
number of issues that fall into this category such as the magnitudes of the various
entrainment coefficients used in the density current model, involving frontal,
vertical and lateral entrainment. An additional concern is that the density current is
essentially a box model and thus predicts an average or "top-hat" concentration. If
more detailed information on plume concentrations is required, information on the
concentration distributions is required. This may be especially critical in the case
where sediment particle settling will significantly distort the sediment profiles
within the density current.
c) What are the essential differences between the D-CORMIX and CD-FATE models
and which is preferable as a mixing zone model for continuous dredged material
discharge? [see section 3.3 for further details]
-------
D-CORMIX is a model which predicts the movement and dilution of a
discharge starting with the source conditions and progressing through a near-field
jet-mixing region, a plume-mixing region, a density current region, and a far-field
ambient mixing region for a variety of possible configurations. CD-FATE is a
model which also predicts the movement and dilution of a discharge but with
several important differences from the D-CORMIX formulation. CD-FATE
begins with the source conditions and utilizes the formulation in the original model
CORMIX to describe the descending plume. However, it also assumes that an
empirical fraction of the source sediment is stripped from the descending plume
and enters the water column subject to subsequent ambient water column transport
and mixing processes. Once the sinking plume reaches the bottom, the sediment
load is assumed to settle directly to the bottom as opposed to being transported as
a bottom density current. However, CD-FATE does allow for entrainment from
the bottom deposits and an additional component of sediment is thereby
introduced to the bottom of the water column to be subsequently transported and
diluted by the ambient flow. CD-FATE as a model appears to be less generally
formulated than D-CORMIX in that it does not handle the movement and dilution
of a suspended sediment density current that may be important in some situations.
However, the component of the model that accounts for entrainment from the
bottom deposits does introduce a suspended sediment plume near the bottom in a
fashion somewhat analogous to the density current. The processes assumed to
model the bottom plume are significantly different in D-CORMIX and CD-FATE
and no rigorous attempt has been made to delineate differences in the predictions
of the two models for typical applications. It is, therefore, not possible to assess
the relative validity of the two models without further investigation.
In summary, D-CORMIX focuses on the movement and dilution of a
mixture of solids and liquids as a jet or plume and as a density current and does not
account for the distribution within the water column of a component stripped from
the sinking plume or entrained from bottom deposits. CD-FATE calculates a
solids concentration in the water column resulting from an ad hoc specified amount
of solids stripped from the descending jet/plume and also calculates suspended
sediment distributions associated with material entrained from bottom deposits. It
does not predict the transport and dilution of suspended material moving along the
bottom as a density current except that the material entrained from the bottom may
be considered to be somewhat analogous to a density current. However, this
material is not assumed to move under its own weight but is simply transported
and mixed by the ambient flow.
d) Does the SAB approve of our outline for laboratory validation? What further
suggestions can be offered? [see section 3.4 for further details]
-------
As noted in Section 3.4 below, the Subcommittee recommends a staged
validation process. An initial objective in the design of any validation study must
be to decide whether or not the processes in D-CORMIX are to be validated or
whether it is the ability of D-CORMIX to be able to represent physically important
processes that is being investigated. For example, if near field turbidity within the
water column is deemed to be important in a specific mixing zone application, it is
likely that the splash from an above water discharge could provide a significant
contribution to turbidity. However, since this process is not currently modeled by
D-CORMIX, any validation effort that relies on observations of near field turbidity
is likely to be unsuccessful. On the other hand, if the validation effort includes
attempts to properly characterize processes that ought to be included in D-
CORMIX, the investigation of near field turbidity may be a fruitful exercise.
A similar consideration arises with respect to chemical and physio-chemical
processes since D-CORMIX currently does not represent these in any meaningful
fashion. It may therefore be pointless to even attempt to validate the current
model in systems with reactive constituents. However, if an attempt is made to
determine the potential significance of reactions to mixing zone processes, a
second dilemma is present. Even if additional capabilities were developed for D-
CORMIX to properly model chemical and physio-chemical processes, it is possible
for a validation effort to fail if the mixing/sedimentation processes are not
represented properly in the current version of the model. In light of this
consideration, it appears to be more fruitful to approach the validation effort in a
staged manner.
e) What factors should be considered in developing an AIZ that will not adversely
impact the integrity of the aquatic ecosystem? How should the AIZ be sized,
especially in relation to distance from the bottom (substrate), and portion of water
column encompassed? [see section 3.5 for further details]
Continued use of mixing zones as allocated impact zones (AIZs) is
supported. Additionally, the Subcommittee supports Agency efforts to define and
model AIZs as they may exist under different conditions. Current practices at the
state level often results in designation of AIZs that are not well defined and which
do not consider (measure/estimate) risk to aquatic species or humans.
-------
2. INTRODUCTION
2.1 Background
Understanding the fate of dredged material discharged at open water sites is essential in
order to predict potential effects of released contaminants on aquatic life and human health.
Mathematical models of the physical processes determining the fate of the discharged material can
be used to provide an estimate of concentrations in the receiving water as well as the initial
deposition pattern of material on the bottom. Any evaluation of potential water column biological
effects has to consider the effects of mixing and resulting dredged material/chemical correction.
As defined by EPA (EPA, 1991) the mixing zone (or dilution zone) is defined as a limited area or
volume of water where initial dilution of a discharge takes place; and where numeric water quality
criteria can be exceeded but acutely toxic conditions are prevented from occurring. (See section
3.1.1 a) for a further discussion of 'mixing').
The draft Inland Testing Manual (EPA, 1993) for the evaluation of dredged material under
CWA Section 404, which was previously reviewed by the Science Advisory Board (SAB, 1994),
contains a mathematical model for evaluating the mixing of instantaneous discharges from barges
and hopper dredges. Following the SAB's review, EPA has conducted an evaluation of possible
modeling approaches for continuous dredge pipeline disposal. During that evaluation, EPA
determined that continuous pipeline dredge operations were primarily conducted in relatively
shallow waters, where boundary interaction between the disposal plume and the bottom would
strongly influence the hydrodynamics of the mixing process. After a careful review of available
EPA and U.S. Army Corps of Engineers initial mixing methodologies, EPA selected CORMIX for
further development. CORMIX was chosen because it was deemed by EPA to be the only easily
applied methodology which: a) accounts for plume boundary interaction in shallow waters; and b)
simulates the resulting plume density current after boundary interaction.
D-CORMIX, the model which is presently being examined in this SAB review, predicts
the initial dilution and mixing zone of a typical continuous dredge outfall operation such as a
pipeline discharge. EPA anticipates that D-CORMIX, when fully validated, will be an important
tool to evaluate potential exceedences of water quality standards due to continuous dredged
material or other negatively buoyant discharges.
The CORMIX family of models consists of CORMIX1 (Doneker and Jirka, 1989) for
mixing of a submerged single port discharge, CORMIX2 (Akar and Jirka, 1991), for multiport
discharges, CORMIX3 (Jones et al., 1996), for discharge from channels, and most recently D-
CORMIX, for mixing of continuous dredge disposal. The first three models were developed to
predict mixing for buoyant plumes with little or no suspended solids present, such as for example
the discharge of a treated wastewater to the ocean. D-CORMIX was developed to model mixing
for negatively buoyant plumes which may contain high solids concentrations, such as occur during
dredge disposal operations. It incorporates Stokes settling for up to five particle sizes, although
-------
optional use of hindered settling is possible. It also contains a first order decay/growth function
for modeling contaminant reactions. The CORMIX models are structured within an expert
system framework which is used to facilitate data entry, and select the type of mixing regime
using a reasoning approach which mimics the way an engineer would go about solving the
problem. Once the regime is identified, the model solves appropriate equations. The solution
generates tabular and graphical output. The CORMIX models operate under a DOS
environment, although a Windows-based version of CORMIX is being developed.
CORMIX1, CORMIX2, and CORMIX3 have, overtime, established themselves as useful
tools for defining mixing zones and the concentration distribution of conservative contaminants.
These models are widely distributed and used throughout the engineering community. The
CORMIX modeling approach has been applied to address a wide range of discharge situations,
ranging from complex multiport diffusers to small toxic discharges and large cooling water
outflows. A strength of the CORMIX approach is that it contains no adjustable parameters that
must be estimated or found indirectly; input data consist of values measured directly for the
system of interest.
D-CORMIX appears to have been developed in response to initiatives by the Corps of
Engineers to model dredge disposal problems. The Corps approach is based on the concept that
negatively buoyant surface discharges can be modeled as the mirror image of positively buoyant
subsurface discharges. To this end, the CD-FATE program, ADDAMS (Automated Dredging
and Disposal Alternatives Management System), was written (ADDAMS, 1994). CD-FATE
incorporates the CORMIX positive buoyant models into a larger framework which contains pre-
and postprocessing modules for facilitating interpretation of input and output data. D-CORMIX
takes a similar approach, but as noted incorporates additional mechanisms related to particle
settling, contaminant reactions, and the movement of density currents over sloping bottoms.
2.2 Charge to the Subcommittee
a) Technical aspects of D-CORMIX:
1) Is D-CORMIX an appropriate water quality model to use for continuous
dredged material discharge mixing zone analysis? (Charge Question #1)
2) Does the model accurately capture the physics of negatively buoyant
surface plumes, in particular, behavior of the density current and particle
settling associated with dredged disposal plumes? (Charge Question #2)
3) Is D-CORMIX, a model based on conservation of mass, momentum and
energy principles that provides continuous simulation of near-field,
intermediate-field, and far-field physical processes, preferable to models
which make empirical assumptions on the amount of suspended materials
available for transport (e.g., CD-FATE)? (Note - the Subcommittee revised
-------
this charge question to the following: What are the essential differences
between the D-CORMIX and CD-FATE models and which is preferable as
a mixing zone model for continuous dredged material discharge?) (Charge
Question #3)
4) Does the SAB approve of our outline for laboratory validation? What
further suggestions can be offered? (Charge Question #4)
b) Implementation of model with regard to use of an allocated impact zone for
negatively buoyant wastewater discharges (not dredged material discharges):
1) EPA's current policy addresses mixing zones as allocated impact zones
(AIZs) where certain numeric water quality criteria may be exceeded as
long as: there is no lethality to organisms passing through the mixing zone;
there are no significant risks to human health; and the integrity of the water
body as a whole is not impaired as a result of these exceedences. These
AIZs, if disproportionately large, could potentially adversely impact the
integrity of the aquatic ecosystem and have unanticipated ecological
consequences.
(a) D-CORMIX predicts the 3-dimensional characteristics and mixing
behavior of a negatively buoyant discharge plume. What factors
should be considered in developing an AIZ that will not adversely
impact the integrity of the aquatic ecosystem? How should the AIZ
be sized, especially in relation to distance from the bottom
(substrate), and portion of water column encompassed? (Charge
Question #5) We realize that this will be influenced by site-specific
ambient characteristics, but seek general (or specific, if appropriate)
guidance.
-------
3. RESPONSE TO THE CHARGE
3.1 Charge Question #1 - Is D-CORMIX an appropriate water quality model to use for
continuous dredged material discharge mixing zone analysis?
Before commenting on the appropriateness of D-CORMIX, it is important to identify
three very general categories of processes that affect the constituents in a dredged material
discharge:
3.1.1 Definitions
a) Mixing: the physical mixing (or dilution) of the discharged material (water,
sediment, and all constituents of interest) with water from the receiving water body
will generally tend to reduce the concentrations of suspended sediment and
constituents in the discharge jet or plume. In many applications, mixing will be the
most important mechanism for concentration reduction as the jet or plume falls (or
rises) within the water column and reaches an equilibrium position (often the
surface or bottom), and, in some cases, as the discharge is transported as a density
current or dispersed by ambient turbulence. The time scale for this mixing to
occur within the limits of typical mixing zones is usually on the order of seconds to
minutes.
b) Sedimentation: suspended sediment in the discharge will affect the density of the
discharge and, as a result, will influence the movement of the discharge within the
receiving water. In addition, suspended sediment in the discharge will settle
downward under the action of gravity. This settling may cause the sediment
ultimately to move differently from the liquid phase of the discharge, resulting in
phase separation in the jet, plume, or density current region.
c) Chemical and physio-chemical reactions: both the constituents and the suspended
solids in the discharge may undergo transformations of a chemical or physio-
chemical nature. Constituents in the discharge may react with each other or with
constituents in the receiving water. Dissolved constituents may become associated
with solid material and constituents initially associated with solids may become
dissolved. Finally, suspended solids may interact with each other, resulting in a
change in the size distribution of solid material. Although some important
reactions may occur within seconds or minutes, many of these reactions occur on a
time scale of tens of minutes to hours, or even days.
The appropriateness of D-CORMIX will depend upon which of these processes are
important in a given mixing zone analysis. It is likely that, by definition, mixing will always be
important. For dredged material the presence of sediment will almost always affect the density,
-------
but sedimentation and related phase separation may or may not be important. In many mixing
zone analyses the effect of reactions is neglected by assuming that the mixing time is small
compared to the reaction time. Because of these potential differences in importance, the
Subcommittee responses to this and subsequent questions will, where appropriate, address the
capability of D-CORMIX to handle each of these three processes separately. It is important that
an inability of D-CORMIX to handle one set of processes, such as reactions, not invalidate it as
an appropriate model for cases in which mixing and sedimentation are the dominant processes.
3.1.2 Appropriateness of D-CORMIX as a Mixing Model
D-CORMIX is a derivative of CORMIX which was developed for analyzing the
characteristics of the mixing zone for positively (upwards) buoyant, dissolved phase continuous
discharges under a variety of different ambient and discharge configurations. CORMIX is,
however, restricted to certain configurations and simulations may not be performed for various
scenarios mainly involving highly irregular receiving water geometries. The structure of
CORMIX involves the use of a number of different simulation modules, depending on the
combination of discharge and ambient conditions, and this feature allows for simulation of a wider
variety of scenarios than other currently available models. There are separate models (CORMIX
1, 2, and 3) for the simulation of single port, multiport, and surface discharges respectively.
Although CORMIX allows the specification of a first order coefficient for pollutant
removal/generation, in general, the model presently does not consider chemical/biological
interactions and should be considered to be most applicable for the simulation of a conservative
pollutant.
3.1.3 Appropriateness of D-CORMIX as a Sedimentation Model
D-CORMIX involves the re-configuration of CORMIX from a buoyant dissolved phase
discharge to describe a negatively (downwards) buoyant particle laden discharge in order to
simulate the condition associated with typical dredged material discharge. Analogues to
CORMIX I and III have been developed. A number of specific modifications to CORMIX have
been implemented. The modification from positive to negative buoyancy is trivial and no
discussion of this is required. A further modification is required because of this change since a
positively buoyant discharge will tend to rise up to a horizontal water surface and subsequently
spread along it. In contrast, a negatively buoyant discharge will sink down to and spread along a
bottom. Since the bottom topography can generally be quite complex, this provides an additional
degree of difficulty in specifying a discharge condition. D-CORMIX restricts simulations to a
constant slope in one direction although it does allow for two different zones with different
slopes. The effects of any deviation of the actual bathymetry from this idealized geometry cannot
be described with the current version of D-CORMIX. An additional modification to D-CORMIX
was to describe the effect of separate phase sediment particles contained within the flow. During
the descent phase of the discharged plume, relative motion (sedimentation) of the sediment is not
considered but settling is allowed once the plume reaches the terminal level and begins to spread
as a density current. The model considers sediment to be composed of five discrete size classes
-------
(clay, silt, etc.) with prescribed settling velocities. Deposition of the sediment is computed on the
basis of these settling velocities and the removal of this sediment is used to establish a reduction in
buoyancy within the density current. Finally, actual dredged materials may involve discharges
from above water pipelines, often with deflectors in front of the discharge to spread the material
over a wider surface area than would occur with a simple submerged discharge. D-CORMIX
considers a number of different above water discharge configurations to establish an "equivalent"
discharge diameter and angle to be used in subsequent submerged jet analyses.
During the Subcommittee meeting, concerns were expressed regarding the computational
capabilities of D-CORMIX. Two specific items included: a) the stripping of a small fraction of
the discharge (including suspended sediment) from the sinking plume which is subsequently
transported with the prevailing ambient current; and b) phase separation within the density current
portion of the discharge involving zone settling which is observed at high suspended solids
concentrations. Whether or not this second phenomenon has been observed in open water dredge
spoil disposal was unclear, but it is the basis for the design and operation of confined disposal
facilities as well as a number of industrial processes. The stripping process may be a function of
above water discharge (splash from deflected discharges, for example) and possibly low levels of
temperature stratification (leading to trapping of low momentum splash near the water surface).
Although D-CORMIX does not consider either one of these two processes, there are possibilities
for future modifications to the model that could incorporate these or other processes.
With specific regards to the two processes described above, the nature of the required
changes would be substantially different. The "splash" or stripping into the water column is not a
natural feature of the flow physics of D-CORMIX. The nature of this process is apparently
poorly understood at present and is probably strongly correlated with the nature of the discharge
(above water, type of deflector, etc.). With regards to dissolved phase contaminant
concentrations, splash is probably unimportant, but it may provide a significant contribution to
local turbidity, which may be important in some mixing zone applications. It is difficult to
visualize how the incorporation of the splash component into D-CORMIX could be achieved in a
mechanistic fashion comparable with the remainder of the model, but it is conceivable that an ad-
hoc formulation could be developed from observations at field sites and incorporated into the
model. The remaining process of phase separation could probably be incorporated into relevant
simulation modules of D-CORMIX if it is assessed to be an important factor in sediment plume
flows.
3.1.4 Appropriateness of D-CORMIX as a Model of Non-Conservative
Reacting Constituents
The extension of CORMIX to negative buoyancy, high solids, nonconservative reactant
problems raises several fundamental issues which will ultimately impact the level of resources
needed for development and the way in which experimental validation studies will be carried out.
First, one of the principal strengths of CORMIX, the lack of any adjustable coefficients, is
compromised. Parameters describing particle interactions (agglomeration, shearing) and
10
-------
contaminant reactivity must be estimated from ancillary experiments, or adapted from the
literature. Second, although the hydrodynamic portion of CORMIX is relatively sophisticated,
the contaminant and particle behavior portions are less so. Five particle size classes are unlikely
to sufficiently capture the interactions that grow and shrink particle masses, and there is
essentially no consideration of how contaminants distribute themselves between the solid and
solution phases, and how this is in turn affected by particle size changes. Third, these factors will
significantly affect the conduct of laboratory and/or field validation experiments. It may not be
sufficient to measure merely the spatial distribution of a few size fractions of suspended solids and
sample densities if reactive contaminant distribution is sought as the guiding factor in defining the
mixing zone. Rather, it may become necessary to measure a specific contaminant(s), to establish
a greater number of particle size classifications, and to obtain direct evidence of the nature of
particle aggregation states. Such determinations would be in addition to separate experiments
related to contaminant partitioning and other reactions.
3.1.5 Summary
The appropriateness of D-CORMIX for intended applications will depend on the level of
detail and sophistication required for the analysis. Limited validation studies have been
undertaken thus far with mixed results with respect to the predicted distribution of suspended
solids and associated changes in density in comparison with several laboratory and a field data set.
Neither the hindered settling option nor contaminant decay/growth features have been tested. For
those applications for which suspended solids are the only contaminant of interest, D-CORMIX
might be said to be partially validated. However if intended applications for predicting mixing
zones include the need to predict dissolved phase contaminant distributions that originate from
chemicals originally bound to the sediment, then D-CORMIX must be considered to be, at
present, unvalidated. It is important to recognize, however, that an inability of D-CORMIX to
describe one set of processes, such as reactions, does not invalidate it as an appropriate model for
cases in which mixing and sedimentation are the dominant processes.
In summary, D-CORMIX appears to have the potential for the mixing zone analysis of
dredged material discharges. In its current form, it does not account for any physico-chemical
interactions between the sediment and liquid phases. Due to short transit times through the
mixing zone, this may not be an important limitation for most applications. It may be appropriate
to define specific applications or classes of applications where this simplification is not appropriate
and to clearly identify these conditions for potential users. D-CORMIX is a modification of
CORMIX that considers the presence of suspended sediment as a density-producing agent which
can subsequently settle out of a density current flowing along the bottom of the receiving fluid.
The deposition is handled by specification of a settling velocity which is dependent on sediment
size. Whether particle settling within the density current is the only important physical process
distinguishing dredged material plumes from dissolved phase ones and whether the simple settling
computation can be used to estimate sediment removal from the plume needs to be subject to a
more rigorous validation effort than has been conducted to date. In addition, the nature of above
water discharges and there contribution to a splash component of sediment flux to the
11
-------
environment needs to be further investigated to determine the significance to local turbidity within
the mixing zone.
3.2 Charge Question #2 - Does the model accurately capture the physics of negatively
buoyant surface plumes, in particular, behavior of the density current and particle
settling associated with dredged disposal plumes?
3.2.1 Potential Deficiencies in original CORMIX Model that are incorporated
into D-CORMIX
The Subcommittee expressed concerns with regards to a number of issues that fall into
this category such as the magnitudes of the various entrainment coefficients used in the density
current model, involving frontal, vertical and lateral entrainment. An additional concern is that
the density current is essentially a box model and thus predicts an average or "top-hat"
concentration. If more detailed information on plume concentrations is required, information on
the concentration distributions is required. This may be especially critical in the case where
sediment particle settling will significant distort the sediment profiles within the density current.
One Subcommittee member used some personally available data sets to simulate these two
models. Information concerning this individual analysis is included in Appendix A.
3.2.2 Restrictions on the Situations that D-CORMIX can Analyze
There are a number of physical processes within a sediment plume that may potentially
affect the distribution of dissolved phase constituents or suspended solids. Many of these will be
restricted to specific classes of problems and may not be general restrictions as to the use of D-
CORMIX. Further assessment may be required to determine the significance to applications
typically encountered within the mixing zone of a dredged material disposal.
a) D-CORMIX makes fairly restrictive specifications on the specification of the
bathymetry. Actual applications may involve bottom irregularities that have a
significant impact on the density current motion.
b) D-CORMIX does not allow for flocculation of fine-grained sediments to occur
within the mixing zone and therefore cannot account for possible changes in
settling velocities.
c) D-CORMIX does not allow for resuspension/bottom scour as a source of sediment
to the density current.
d) Except in a fairly simplistic way, D-CORMIX does not allow for the far-field
build-up to influence the plume dynamics. An allowance is made for the possible
re-entrainment of material from a previous tidal cycle but only insofar as it may
12
-------
impact the dilution. The build-up of previously discharged fluid may also influence
the dynamics of the density current as well.
e) D-CORMIX does not provide estimates of suspended solids from top to bottom of
the water column, i.e., represents the density current plume with a top-hat profile
within the water column, and the plume is represented with a gaussian vertical
profile in the passive diffusion region.
3.2.3 Possible Deficiencies in D-CORMIX
a) There is a body of knowledge that indicates whenever suspended sediment
concentrations exceed something on the order of 10 g/1, sediment interactions
become more complex than can be described by a simple settling model.
Discharge concentrations of suspended sediment are generally above this limit and
the possibility exist that these processes may become operative in at least some
cases. Even in cases in which the sinking plume dilutes the flow below the
threshold level, settling within the density current will redistribute sediment near
the bottom of the plume; the bulk density computed by CORMIX would not be
capable of detecting this situation. The assumptions used by D-CORMIX to
represent the density current behavior need to be further validated.
b) There are some fundamental differences between dissolved phase plumes and two-
phase plumes such as air bubble plumes and presumably sediment plumes. These
differences become more pronounced in stratified receiving fluids.
c) The source description for D-CORMIX simulations (e.g., the CORJET module)
requires the specification of a discharge diameter and angle. D-CORMIX allows
the general description of above water discharges and converts them into the
required input. The accuracy of such a conversion has not been demonstrated at
present. In particular, the use of a deflector plate may spread the discharge over
an area comparable with the water depth and the application of a plume model to
this discharge configuration is questionable. In addition, these above water
discharges may contribute a "splash" component which could have a significant
contribution on near field turbidity which is ignored by the present formulation of
D-CORMIX.
d) The preliminary validation efforts on D-CORMIX indicated that simulations
became more reliable on steeper slopes and less satisfactory on smaller slopes.
Flows on shallow slopes are more influenced by far-field influence and may be
more difficult to simulate accurately even when only dissolved phase plumes are
considered. The existing validation effort is insufficient to conclude that the model
adequately simulates the density current, in particular on mild slopes.
13
-------
3.3 Charge Question #3 - What are the essential differences between the D-CORMIX
and CD-FATE models and which is preferable as a mixing zone model for
continuous dredged material discharge?
The answer to this question will consist of a brief description of the essential differences
between the two models followed by a discussion of their relative merits as mixing zone models
for continuous dredged material discharge. Additional characteristics of D-CORMIX are
discussed in the answers to Charge Questions #1 and #2 (Sections 3.1 and 3.2 of this report).
D-CORMIX is a model which predicts the movement and dilution of a discharge starting
with the source conditions and progressing through a near-field jet-mixing region, a plume-mixing
region, a density current region, and a far-field ambient mixing region for a variety of possible
configurations. As discussed in the answers to Charge Questions #1 and #2, the algorithms used
by D-CORMIX in calculating the movement and dilution are based upon a body of theoretical and
empirical information applicable to these configurations and are not adapted to site specific
conditions or fit to a particular set of field observations.
In predicting the movement and dilution of the discharged material, D-CORMIX treats the
liquid and solids in the discharge as combined phases that follow the same trajectory and
experience the same dilution. The only exception to this is that, beyond the near-field jet mixing
region, the solid material is removed according to specified settling velocities. Remaining solids
and liquids are still assumed to move and dilute together. Furthermore, the model assumes that
settled solids are deposited on the bottom immediately. Consequently, the present form of D-
CORMIX will never predict that solid or liquid material leaves the original mixture and enters the
water column. More specifically, D-CORMIX will not predict that solid material is 'stripped' off
the jet or plume as it descends through the water column and will not predict that liquid leaves the
bottom layer as a result of consolidation of the solids. The solids concentrations predicted by D-
CORMIX can only become more dilute and can not increase as would be the case if consolidation
were considered.
CD-FATE is a model which also predicts the movement and dilution of a discharge
starting with the source conditions, a jet/plume mixing region and a far-field mixing region for a
variety of different discharge configurations. CD-FATE relies on CORMIX whenever possible in
predicting the characteristics of these regions, but makes different assumptions about how these
regions are linked together. These assumptions are based on the general experience of the Army
Corps of Engineers with open water and confined dredged material disposal operations. The
resulting CD-FATE model has a more limited empirical basis compared to D-CORMIX, but is
more closely linked to observations of actual dredged material disposal. [Note: the
Subcommittee's evaluation of CD-FATE is based upon material contained in Attachments 12, 14,
and 15 (COE, 1994a; 1994b; 1995) of the information provided to the Subcommittee in advance,
as well as a verbal summary provided by Dr. Paul Schroeder of the US Army Corps of Engineers
at the Subcommittee meeting. The Subcommittee notes that the description of CD-FATE in the
Attachments is very sketchy and, with respect to some of the aspects of CD-FATE described by
14
-------
Dr. Schroeder, incomplete. The Subcommittee recommends that, if CD-FATE is to be routinely
used for mixing zone analysis, better documentation should be prepared.]
In predicting the movement and dilution of the discharged material, CD-FATE uses a
linked set of CORMIX calculations, each of which assumes that the liquid and solid material in
the discharge move and dilute in a similar manner. However, CD-FATE assumes that a fraction
of the suspended sediment will be lost to the water column as the jet or plume descends to the
bottom. This fraction is specified in an ad-hoc basis as an upper limit on that observed at actual
dredged disposal sites. This solid material is assumed to remain in suspension and is accounted
for in predicting the turbidity in the water column downstream from the discharge, but above any
density current which may form. The calculation of solids concentration in the water column
resulting from the stripping of material uses a CORMIX module starting from a surface source of
solids. Solid material that is not stripped from the descending jet or plume and all of the original
liquid in the discharge ultimately reach the bottom and are assumed to form an unmoving layer of
consolidating solids from which liquid is ejected into the overlying water column at a rate
determined by specified consolidation rates. The subsequent transport and dilution of the ejected
liquid phase in the water column is predicted by a CORMIX module whose source conditions are
related to the estimated size of the consolidating layer. The concentration of the solids in the
consolidating layer is predicted to be greater than that of the material reaching the bottom.
In summary, D-CORMIX focuses on the movement and dilution of a mixture of solids and
liquids as a jet or plume and as a density current and does not account for the distribution within
the water column of a component stripped from the sinking plume or entrained from bottom
deposits. CD-FATE calculates a solids concentration in the water column resulting from an ad
hoc specified amount of solids stripped from the descending jet/plume and also calculates
suspended sediment distributions associated with material entrained from bottom deposits. It
does not predict the transport and dilution of suspended material moving along the bottom as a
density current except that the material entrained from the bottom may be considered to be
somewhat analogous to a density current. However, this material is not assumed to move under
its own weight but is simply transported and mixed by the ambient flow.
The adequacy with which D-CORMIX represents physical processes associated with
discharges containing high solids concentrations has been identified as a concern in the answers to
Charge Questions #1 and #2, and as an area in need of experimental study in the answer to
Charge Question #4 (Section 3.4 of this report). If the processes of solids stripping and
consolidation are found to be important for such discharges, D-CORMIX as presently formulated
would be incapable of predicting accurately the behavior of the liquid and solid material stripped
out of the plume and the consolidation of the suspended sediment into a bottom deposit with the
attendant ejection of water from the deposit. These aspects may be important if regulatory focus
is on the water column above the density current.
CD-FATE as a model appears to be less generally formulated than D-CORMIX in that it
does not handle the movement and dilution of a suspended sediment density current that may be
15
-------
important in some situations. For this reason, CD-FATE may not be an appropriate model if
regulatory focus is on the concentrations within the density current region. CD-FATE does
however, account for suspended sediment components stripped from the descending plume and
entrained from bottom deposits, and does make predictions of the results of these processes on
water column sediment concentrations. CD-FATE, if validated, may be an appropriate model if
the regulatory focus is on the water column outside of the density current.
3.4 Charge Question #4 - Does the SAB approve of our outline for laboratory
validation? What further suggestions can be offered?
D-CORMIX appears to warrant additional validation as suggested by the staged approach
outlined below. An initial objective in the design of any validation study must be to decide
whether or not the processes in D-CORMIX are to be validated or whether it is the ability of D-
CORMIX to be able to represent physically important processes that is being investigated. For
example, if near field turbidity within the water column is deemed to be important in a specific
mixing zone application, it is likely that the splash from an above water discharge could provide a
significant contribution to turbidity. However, since this process is not currently modeled by D-
CORMIX, any validation effort that relies on observations of near field turbidity is likely to be
unsuccessful. On the other hand, if the validation effort includes attempts to properly characterize
processes that ought to be included in D-CORMIX, the investigation of near field turbidity may
be a fruitful exercise.
A similar issue arises with respect to chemical and physio-chemical processes since D-
CORMIX currently does not represent these in any meaningful fashion. It may therefore be
pointless to even attempt to validate the current model in systems with reactive constituents.
However, if an attempt is made to determine the potential significance of reactions to mixing zone
processes, a second dilemma is present. Even if changes were made to D-CORMIX to properly
model chemical and physio-chemical processes, it is possible for a validation effort to fail if the
mixing/sedimentation processes are not represented properly in the current version of the model.
In light of this consideration, it appears to be more fruitful to approach the validation effort in a
staged manner. Such an effort could be structured along the lines of the effort listed below.
3.4.1 Validation of Mixing Processes
Jet/plume mixing processes are reasonably well understood and major validation efforts
are probably not required on these. However, a well designed study could also focus on these to
some extent. One of the more important changes in D-CORMIX relative to CORMIX is the
extension to above water surface discharges. A laboratory study that investigates the near field
mixing associated with above water discharges including discharges impinging on deflector plates
would allow the assessment of the below surface mixing associated with some generic
configurations. An additional advantage is that these same experiments could be used to assess
the splash problem and the nature of its contribution to near field turbidity. Finally, if these
discharges are allowed to impinge on a bottom (probably a sloping one but this would not be an
16
-------
absolute necessity), the nature of bottom impingement contributions to mixing that are not
presently incorporated into CORMIX or D-CORMIX could also be investigated. It would also
be fruitful to examine typical discharge configurations and determine by application of the current
model whether or not ambient water column stratification is likely to be of any consequence in
these circumstances. If so, then it would also be worthwhile to conduct a separate laboratory
study to investigate this effect.
3.4.2 Validation of Sedimentation Processes
Efforts at multiple levels may be the best approach to this problem. At a laboratory scale,
a careful investigation of the dynamics of density currents spreading on a slopping bottom should
be undertaken. These experiments should be at as large a scale as is feasible in order to minimize
Reynolds number effects. They should also incorporate sediment as the density producing agent
and at concentrations that are realistic of dredged disposal application . The discharge
configuration should be simple in order to minimize confounding issues addressed above.
Different bottom slopes from "mild" to "steep" should be considered and should be of sufficient
length that a significant amount of sediment deposition will occur. A key issue is to determine
whether the current sedimentation model of D-CORMIX is adequate or whether phase separation
processes are important.
One possible investigation at the field scale, but with more controlled ambient conditions,
would be the investigation of flows in confined disposal facilities. Initially, a review of existing
data sets could determine whether or not there is existing information that could be utilized in a
model validation effort. If no such data sets are available, a properly designed experiment in an
existing facility might be arranged. Although such an approach would necessarily not include all
processes in place at open water disposal sites, the advantages of more control over system
variables will allow a more detailed set of observations against which to compare D-CORMIX
predictions.
3.4.3 Validation for Non-Conservative Reacting Constituents
The Subcommittee suggests that some consideration be given at this stage of model
development to the specific uses intended for D-CORMIX. Are there certain classes of
contaminants that are generally found in dredge materials for which water quality standards are
well established (e.g., PAHs, PCBs, heavy metals)? Are there certain common physical and
chemical features of dredge particles that will tend to define particle-particle interactions and
particle-contaminant interactions (size distributions, organic matter content, pH, etc.)? If these
questions are answerable, it may be possible to build a degree of sophistication into D-CORMIX
for reactive contaminants and particle interactions which is appropriate for a large number of
intended uses, but not so complex that much of the value of the CORMIX approach is lost.
Validation studies should address issues related to selected cases of contaminant
partitioning, consistent with the important sediment, water, and contaminant characteristics
17
-------
identified above. In so doing, it is not necessary that these studies be exhaustive, but should
include enough experimental data to provide confidence that the model adequately captures the
physics and chemistry that are relevant to evaluating the distribution of contaminants within the
mixing zone. In general, validation studies can be designed to measure various particulate
distribution patterns, but incorporate contaminants into the source and fate aspects of the
experiments. It is anticipated that the inclusion of actual contaminants in validation studies will
entail some degree of modification to D-CORMIX. However, it should be possible to accomplish
this with a limited number of added "modules", each containing process and reaction-level
mechanisms relevant to the contaminant/sediment/water classes of interest. In this way, the 'rule-
based' classification approach of D-CORMIX can be used to include contaminant partitioning
behavior classes.
3.4.4 Need for Field Validation
It must be remembered that without adequate calibration during development and
subsequent real-world validation, all models remain as cartoons of reality subject to speculation
and challenge. Validation against laboratory systems is an important step but it does not carry the
full validation that comparison with a real-world receiving water ecosystem would afford. To be
effective, any model must be calibrated during its development using real-world data to ensure
that all inputs are realistic and based on actual performance of natural systems. Once the model is
assembled, its outputs must then be validated against real-world ecosystem behavior to ensure the
accuracy of its calculated outputs. A small-scale field program with intensive monitoring of
sediment behavior and plume dynamics is needed to validate the D-CORMIX model.
3.5 Charge Question #5 - What factors should be considered in developing an allocated
impact zone (AIZ) that will not adversely impact the integrity of the aquatic
ecosystem? How should the AIZ be sized, especially in relation to distance from the
bottom (substrate), and portion of water column encompassed?
CORMIX model objectives clearly state the desire to model the behavior of dissolved and
particulate components within discharge plumes. However there is no provision to model the all
important sorption kinetics of organics or other dissolved chemicals which are essential to
properly define mixing zones. There is little value in considering merely the physical transport of
particles if we have no means to consider the sorption kinetics and effects on bioavailability of the
sorbed chemicals on the particle surface. Provisions must be incorporated to model this key
aspect of environmental partitioning before results of CORMIX models can be extrapolated to
real-world ecosystems.
Continued use of mixing zones as allocated impact zones (AIZs) is supported.
Additionally, the Subcommittee supports Agency efforts to define and model AIZs as they may
exist under different conditions. Current practices at the state level often results in designation of
AIZs that are not well defined and which do not consider (measure/estimate) risk to aquatic
species or humans.
18
-------
Looking to the future, the Subcommittee recommends that Agency guidance on mixing
zones be developed using a risk based approach consistent with EPA risk assessment guidelines.
Additionally, the development and application of the CORMIX models for the delineation of
mixing zones needs to consider how the resultant data would ultimately be used to reach or
support conclusions from a human or ecological risk assessment. Currently the program reads as
though "we will expend all these resources to develop a model and then we'll give it to the risk
assessors and see if they can use it." Obviously this is not a productive sequence and the modelers
would be well served to incorporate risk assessment experts into the model development phase to
maximize the ultimate application of the results. Additionally, we recommend that the following
be considered in developing AIZs/AIZ guidance:
a) dynamics of site conditions;
b) organisms of interest at a given site with emphasis on keystone species,
commercial species, and endangered species; and
c) physical/chemical properties (sorption, hydrolysis, volatilization, etc.) of the
chemical under review.
Generalizations on AIZs are difficult because site specific conditions often dominate.
Some generalizations include the following: the AIZ should be relatively small to avoid population
effects, and the AIZ should not limit migratory species (zone of passage should be protected).
Generalizations about the AIZ in relation to the distance above the bottom is difficult. Important
considerations include aerial extent, species present, and depth and width of the negatively
buoyant zone. The particular species that are potentially effected by a negatively buoyant effluent
should be considered. As a general rule, organisms with a short reproductive cycles (amphipods,
oligochaetes, polychaetes, etc.) will recolonize an area much more rapidly than those with longer
reproductive cycles (clams, mussels, bottom fish) and therefore, would require less protection.
Additionally, consideration should be given to the potential for avoidance of the chemical(s) in the
negatively buoyant effluent by mobile organisms such as bottom fish
19
-------
Appendix A - Potential Deficiencies in original CORMIX Model that are incorporated
into D-CORMIX
One of the Subcommittee members had access to several laboratory data sets (See end of
this appendix for references) and simulated these using both the CORMIX and D-CORMIX
models (properly accounting for the issue associated with positive or negative buoyancy as
required by the particular model). A comparison of the simulations led to the following
conclusions:
a) In most situations, the models performed relatively well in predicting dilutions and
plume dimensions observed in the experiments. In general, the models appeared to
under-predict the observed dilution, but the differences were not generally
significant.
b) In cases where vertical density gradients were present in the ambient fluid of
sufficient magnitude to trap the plume within the water column, a number of
discrepancies were noted. The models generally predicted dilution within the
jet/plume region fairly well with a tendency to slightly under-predict the dilution as
noted above. However, the plume then spreads as an internal density current.
Early versions of CORMIX tended to severely over-estimate the collapse of the
density current plume resulting in an layer that was much thinner and wider than
observed in experiments. Subsequent versions of CORMIX rectified this problem
to some extent by introducing wind and bottom shear induced mixing. The
subsequent dilution reduced the lateral spreading forces and therefore the plume
collapse. However, the laboratory experiments involved stagnant ambient fluids
with no wind or shear induced turbulence and inclusion of these mechanisms to
improve model agreement is not appropriate (Wright, 1977). It appears that the
magnitude of lateral spreading forces has been over-estimated in the models and
therefore CORMIX or D-CORMIX may not provide accurate simulations for
these cases.
c) Comparison of data in which plumes impinge directly on a boundary indicates that
the model underestimates the dilution in the initial layer spreading along the
surface. The reasons for this are clear; the study of Wright et al. (1991) indicate a
significant dilution associated with the initial phase of the horizontal spread of the
impinging flow while CORMIX and D-CORMIX do not generally consider any
mixing within this region. This is more significant in the case of weak ambient
current and/or shallow water depths where the sinking plume impinges nearly
normal to the bounding surface. Differences between model predictions and
experimental observations can be several hundred percent in this cases. The
significance of these deviations within the context of an entire mixing zone analysis
has yet to be determined.
A- 1
-------
Appendix A - Data Source References:
LTI, and S. J. Wright. 1993. Recommendations of Specific Models to Evaluate Mixing Zone
Impacts of Produced Water Discharges to the Western Gulf of Mexico Outer Continental
Shelf. Produced for USEPA Office of Wastewater Enforcement and Compliance by
Limno-Tech, Inc (LTI) and Steven J. Wright, under subcontract to SAIC, Contract No.
68-C8-0066. April 1993.
Wright. SJ. 1977. Effects of Ambient Cross/lows and Density Stratification on the Characteristic
Behavior of Round Buoyant Jets. W.M. Keck Lab Report, KH-R-36, California Institute
of Technology, May, 1977.
A-2
-------
REFERENCES CITED
ADDAMS. 1994. Automated Dredging and Disposal Alternatives Management System,
(ADDAMS) CDF ATE Module (Diskette).
Akar, P. J. and G.H. Jirka. 1991. CORMIX 2: An Expert System for Hydrodynamic Mixing Zone
Analysis of Conventional and Toxic Multipart Dijfuser Discharges. EPA 600/3-91/073.
DeFrees Hydraulics Laboratory, School of Civil and Environmental Engineering, Cornell
University, Ithaca, NY (Prepared for US EPA, Office of Research and Development).
COE. 1994a. DROPMIX Users Manual Prepared by Don Chase, Department of Civil &
Environmental Engineering, University of Dayton, Dayton, OH for the U.S. Army Corps
of Engineers, Waterways Experiment Station, Vicksburg, MS.
COE. 1994b. Mixing Zone Simulation Model for Dredge Overflow and Discharge into Inland
and Coastal Waters. Prepared by HA VIS Environmental, Fort Collins, CO for the U.S.
Army Corps of Engineers, Waterways Experiment Station, Vicksburg, MS.
COE. 1995. Contaminant Dispersion Around Open Water Pipeline Disposal into Inland and
Coastal Waters. Prepared by HA VIS Environmental, Fort Collins, CO for the U.S. Army
Corps of Engineers, Waterways Experiment Station, Vicksburg, MS.
Doneker, R.L., and G.H. Jirka. 1991. An Expert System for Hydrodynamic Mixing Zone
Analysis of Conventional and Toxic Submerged Single Port Discharges (CORMIX 1).
EPA 600/3-90/012. DeFrees Hydraulics Laboratory, School of Civil and Environmental
Engineering, Cornell University, Ithaca, NY (Prepared for US EPA Office of Research
and Development).
EPA. 1991. Environmental Protection Agency Procedures for Approving State Water Quality
Standards. Code of Federal Regulations 40 CFR 131.35(d)(8).
EPA. 1993. Evaluation of Dredged Material Proposed for Discharge in Inland and Near-
Coastal Waters: Testing Manual (draft). Office of Water, U.S. Environmental Protection
Agency, Washington, DC. May 1993.
Jirka, G.H., Doneker, R.L., and S.W. Hinton. 1996. User's Manual for CORMIX: A
Hydrodynamic Zone Model and Decision Support System for Pollutant Discharges into
Surface Waters. DeFrees Hydraulics Laboratory, School of Civil and Environmental
Engineering, Cornell University, Ithaca, NY (Prepared for US EPA Office of Science and
Technology, Office of Water).
R- 1
-------
Jones, G.R., Nash, J.D. and G. Jirka. 1996. CORMIX3: An Expert System for Mixing Zone
Analysis and Prediction of Buoyant Surface Discharges. DeFrees Hydraulics Laboratory,
School of Civil and Environmental Engineering, Cornell University, Ithaca, NY (Prepared
for US EPA Office of Science and Technology, Office of Water).
SAB. 1994. An SAB Report: Evaluation of a Testing manual for Dredged Material Proposed
for Discharge in Inland and Near-Coastal Waters. EPA-SAB-EPEC-94-007. Science
Advisory Board, U.S. Environmental Protection Agency, Washington, DC. February
1994.
Wright, S.J., Roberts, P.J.W., Yan Zhongmin, and N.E. Bradley. 1991. Surface Dilution of
Round Buoyant Jets. Journal of Hydraulic Research, Vol. 29, No. 1. pp 67-90.
R-2
-------
DISTRIBUTION LIST
Deputy Administrator
Assistant Administrators
EPA Regional Administrators
Director, Office of Science Policy, ORD
Director, Office of Science and Technology
EPA Laboratory Directors
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
Library of Congress
National Technical Information Service
Congressional Research Service
-------
United States Science Advisory EPA-SAB-EC-99-011
Environmental Board (1400) February 1999
Protection Agency Washington, DC www.epa.gov/sab
&EPA AN SAB REPORT: REVIEW
OF THE D-CORMIX MODEL
PREPARED BY THE
D-CORMIX REVIEW
SUBCOMMITTEE OF THE
SCIENCE ADVISORY BOARD
------- |