RESULTS OF AN ADAPTIVE ENVIRONMENTAL ASSESSMENT MODELING WORKSHOP
CONCERNING POTENTIAL IMPACTS OF DRILLING MUDS
AND CUTTINGS ON THE MARINE ENVIRONMENT
Gregor T. Auble
Austin K. Andrews
Ri chard A. Elli son
David B. Hami1 ton
Richard A. Johnson
Janes E. Roel1e
U.S. Fish and Wildlife Service
Fort Collins, Colorado
and
David R. Marmorek
Environmental and Social Systems Analysts Ltd.
Vancouver, B.C. , Canada
Western Energy and Land Use Team
Office of Biological Services
U.S. Fish and Wildlife Service
2625 Redwing Road
Fort Collins, Colorado 80526
This workshop was held Sept. 14-18, 1981
in conjunction with the Gulf Breeze, Florida
Laboratory of the U.S. Environmental
Protection Agency as part of the
Federal Interagency Energy/Environment
Research and Development Program
Office of Research and Development
U.S. Environmental Protection Agency.
-------
DISCLAIMER
This publication has not been subjected to review by the
mental Protection Agency or U.S. Fish and Wildlife Service and
does not necessarily reflect the views of these agencies
endorsement should be inferred. Mention of trade names
products does not constitute recommendation for use.
U.S. Environ'
therefore
No official
or commercial
-------
EXECUTIVE SUMMARY
Drilling fluids or "muds" are essential components of modern drilling
operations. They provide integrity for the well bore, a medium for removal of
formation cuttings, and lubrication and cooling of the drill bit and pipe.
The modeling workshop described in this report was conducted September 14-18,
1981 in Gulf Breeze, Florida to consider potential impacts of discharged
drilling muds and cuttings on the marine environment. The broad goals of the
workshop were synthesis of information on fate and effects, identification of
general relationships between drilling fluids and the marine environment, and
identification of site-specific variables likely to determine impacts of
drilling muds and cuttings in various marine sites.
The workshop was structured around construction of a model simulating
fate and effects of discharges from a single rig into open water areas of the
Gulf of Mexico, and discussion of factors that might produce different fate
and effects in enclosed areas such as bays and estuaries. The simulation
model was composed of four connected submodels. A Discharge/Fate submodel
dealt with the discharge characteristics of the rig and the subsequent fate of
discharged material. Three effects submodels then calculated biological
responses at distances away from the rig for the water column, soft bottom
benthos (assuming the rig was located over a soft bottom environment), and
hard bottom benthos (assuming the rig was located over a hard bottom environ-
ment). The model focused on direct linkages between the discharge and various
organisms rather than on how the marine ecosystem itself is interconnected.
Behavior of the simulation model indicated relatively localized effects
of drilling muds and cuttings discharged from a single platform into open
water areas. Water column fate and effects were dominated by rapid dilution.
Effects from deposition of spent mud and cuttings were spatially limited with
relatively rapid recovery, especially in soft bottom benthic communities which
were conceptualized as being adapted to frequent storms. This behavior was
generated by the set of assumptions about linkages and functional relationships
used to construct the model. Areas of uncertainty included methods for extra-
polating 96-hr LC50 results to exposures of varying lengths and concentrations;
recovery rates of benthic communities; responses to various depths and rates
of burial; fate and effects of the plume in relationship to stratification
layers; and long-term and sub-lethal effects of slightly elevated concentra-
tions of discharged materials. Evaluation of the assumptions of the Soft
Bottom Submodel suggest that the assumptions used may have been relatively
liberal estimates of resiliency of these communities.
Discussion of "closed" water bodies such as bays and estuaries indicated
several reasons to expect different and more complex fate and effects behavior
in these areas. These factors included different species and communities
(such as aquatic macrophytes and oyster beds), more complex circulation and
stratification patterns, and potentially more active resuspension processes.
-------
Much of the possible difference in behavior in these areas centers around the
extent to which they are "closed" or in the relative residence times of water
and sediments in these areas as they determine the long-term dispersion of
discharged material. Despite the complexity and variability of these areas, a
large body of knowledge (such as that concerning fate and physical effects of
dredge spoil) that could be effectively employed in analysis of potential fate
and physical effects in enclosed areas was identified.
-------
CONTENTS
Page
EXECUTIVE SUMMARY iii
FIGURES vi
TABLES " ix
GLOSSARY . x
ACKNOWLEDGEMENTS xi
INTRODUCTION 1
Objectives 1
The Adaptive Environmental Assessment Methodology 2
BOUNDING THE WORKSHOP MODEL .... 4
Actions . 4
Indicators .. 5
Space . . 6
Time 6
Submodel Definitions 7
Submodel Interactions 8
SUBMODEL STRUCTURES . . . 10
Di scharge/Fate Submodel 10
Water Column Effects Submodel 17
Soft Bottom Effects Submodel 28
Hard Bottom Effects Submodel 35
SYSTEM MODEL 42
Structure 42
Behavior 42
WORKSHOP RESULTS 49
Communication 49
Information Integration 53
Information Gaps 54
Factors Determining Fate and Effects 56
LITERATURE CITED 60
APPENDIX 62
Comments on the Soft Bottom Submodel by D. F. Boesch 62
-------
FIGURES
Number
1 Workshop looking outward matrix of information
transfers between submodel s
2 Idealized drilling platform discharge ........................ 11
3 Depth of spent mud solids and cuttings under
various conditions ......................................... 16
4 Top view of assumed development and movement of the
upper plume at several times during discharge and
post-di scharge phases ........................................ 18
5 Top view of upper plume slices used in Water Column
submodel calculations ........................................ 19
6 Concentration gradient of soluble phase in discharge
plume at 1000 bbl/hr discharge rate .......................... 19
7 Generalized toxicity curve used to calculate survival
rates in the water column at different concentrations
relative to 96-hr 1C concentration ......................... 21
8 Survival rate in post-discharge phase versus position
in pi ume [[[ 23
9 Total loss of plankton in post-discharge phase versus
position in plume ............................................ 24
10 Concentration gradient of soluble phase in discharge
plume at 30 bbl/hr discharge rate ............................ 24
11 Selective and nonselective toxicity curves ................... 26
12 Sensitivity of total plankton mortality of exposed
population to assumed 96-hr LCcn ............................. 27
13 Epifaunal tissue concentration of chromium (above
background) as a function of fraction whole mud in
sediment [[[ 29
-------
FIGURES (continued)
Number Page
15 Monthly survival rate of infauna and epifauna as a
function of the toxicity of the fraction whole mud
in sediment 31
16 Monthly survival rates of soft bottom fauna as a
function of sediment removal by storms .... 32
17 Fraction of soft bottom fauna exhibiting sublethal response
as a function of fraction whole mud in sediment 32
18 First month recolonization response of microbes to change
in fraction whole mud 33
19 First month recolonization response of meiofauna to change
in fraction whole mud 33
20 Fraction of first month potential recolonization increment
realized due to toxicity of residual mud 34
21 Coral monthly survival rate as a function of depth of
drilling mud solids and cuttings 37
22 Coral monthly survival rate as a function of solids
concentration for a 3-hr exposure 38
23 Potential coral annual growth rate 39
24 Reduction in coral growth rate as a function of
solids concentration for a 24-hr exposure 41
25 Scenario I: depth of deposited spent muds and
cuttings at three distances from platform 43
26 Scenario I: fraction whole mud at three distances
from platform 43
27 Scenario I: concentration of fine grained particulates
in the upper plume' at three distances from platform 44
28 Scenario I: coral biomass at three distances
from platform 45
29 Scenario I: microbial biomass at two distances
from platform 45
VI 1
-------
FIGURES (continued)
Number Page
30 Scenario I: macro-infaunal biomass at two distances
from platform 46
31 Scenario I: epifaunal tissue concentrations of
chromium at three distances from platform 46
32 Scenario III: fraction whole mud at three distances
from platform 48
33 Scenario III: coral biomass at three distances from
platform 48
VI 1 1
-------
TABLES
Mumper Page
1 List of actions developed at workshop ... ... 4
2 List of indicators developed at workshop . .-, 5
3 Upper plume characteristics at various discharge rates 15
4 Deposition radius and total deposition of drilling muds
andcuttings... ... 16
5 Effects of assumotions on population variability in
sensitivity and selectivity of toxicant. .. . .. 26
6 Derivation of a density-dependent coral growth '-ate,
assuming a hemispheric growth form ..... ... 40
IX
-------
GLOSSARY
"closed" water body - a salt or brackish water area, such as a bay or an
estuary, where exchange of water with surrounding areas is restricted.
dispersion ratio - suspended solids in discharge divided by suspended.sol ids
in upper piume .
drilling cuttings - formation solids generated by drilling.
drilling muds - fraction of the drilling mixture that is not formation
cuttings; includes drilling fluid additives, formation water, and
compounds generated under down-hole conditions.
epifauna - organisms larger than meiofauna living on the substrate surface.
fraction whole mud - fraction of a sediment sample composed of discharged
drilling muds, calculated as: [Ba] in sediment/[B'a] in drilling muds.
infauna - organisms larger than bacteria living beneath the substrate surface.
lower plume - plume containing discharged drilling cuttings and mud solids.
macrofauna - general term referring to infauna and epifauna.
meiofauna - microscopic (exclusive of bacteria) and small macroscopic metazoan
fauna inhabiting the substrate surface; includes nematodes, ostracods,
copepods, tubellarians, gastrotrichs, ol igochaetes, etc. (after Pennak
1964).
96-hr ECrQ - concentration of substance at which 50% of exposed population
exhibits an effect from a 96-hr exposure.
96-hr LC^Q - concentration of substance that produces a 50% mortality in
exposed population from a 96-hr exposure.
pycnocline - plane separating two layers of different density.
upper plume - plume containing discharged soluble components and suspended
solids (fine-grained particulates).
-------
ACKNOWLEDGEMENTS
The workshop was attended by participants whose professional expertise,
hard work, and interest in the fates and effects of discnargea drilling muds
and cuttings contributed greatly to the success of the workshop. They include:
NAME
Robert C. Ayers, Jr
James M. Barkuloo
Donald Baumgartner
Deborah Bli zzard
Donald F. Boesch
Jim Cimato
Lester Dauterive
Thomas W. Duke
ADDRESS
Exxon Production Research Co.
Driving & Completion Division
P.O. Box 2189
Houston, TX 77001
Asst. Coastal Ecosystems
Activities Leader
1612 June Ave.
Panama City, FL 32401
U.S. EPA
Marine Sciences Center
Newport, OR 97365
Office of Planning &
Budgeting Coordination
Office of the Governor
404 Carlton Bldg.
Tallahassee, FL 32301
Louisiana Universities
Marine Consortium
Star Route 541
Chauvin, LA 70344
Bureau of Land Management
(RM-543)
Washington, D.C. 20240
U.S. Geological Survey
Gulf of Mexi co Region
P.O. Box 7944
Metairie, LA 70010
U.S. EPA
Environmental Research Lab.
Gulf Breeze, FL 32561
PHONE
FTS 8-713-965-4344
COM (713) 965-4344
FTS 8-946-5215
COM (904) 769-5430
FTS 8-423-4111
(Opr. Asst.)
COM (503) 867-4040
FTS 8-904-488-5551
COM (904) 488-5551
FTS 8-682-2611
(Opr. Asst.)
COM (504) 594-7552
FTS 8-343-7744
COM (202) 343-7744
FTS 8-680-9011
COM (505) 837-4720
Ext. 205
FTS 8-686-9011
COM (904) 932-5311
-------
R. Warren Flint
Alina S. Froelich
B.J. Gallaway
Thomas Gilbert
Charles Hill
Michelle Miller
Douglas Lipka
H.R. (Rob) Moseley, Jr.
Jerry Neff
Gary Petrazzuolo
Eric Powel
University of Texas
Marine Science Institute
Port Aransas Marine Lab.
Port Aransas, TX 78373
Florida State University
Dept. of Oceanography
Tallahassee, FL 32306
LGL Ecological Research Assoc
1410 Cavitt St.
Bryan, TX 77801
New England Aquarium
Central Wharf
Boston, MA 02110
Bureau of Land Management
500 Camp St.
Suite 841
New Orleans, LA 70130
U.S. EPA (EN-336)
401 M St. SW
Washington, D.C. 20460
U.S. EPA
Room 3821 WSM (RO-682)
401 M St. SW
Washington, D.C. 20460
Environmental Engineering
Magcobar Group
Dresser Industries, Inc.
P.O. Box 6504
DCOB RM-551
Houston, TX 77005
Battelle
New England Marine Research
Laboratory
397 Washington St.
Duxbury, MA 20332
Tech. Resource, Inc.
10215 Fernwood Dr.
Suite 408
Be'thesda, MD 20034
Texas A&M University
Dept. of Oceanography
College Station, TX 77843
FTS 8-729-4011
COM (512) 749-6775
FTS 8-904-644-5839
COM (904) 644-5839
FTS 8-713-775-2000
COM (713) 775-2000
FTS 8-617-742-8830
COM (617) 742-8830
FTS 8-682-6541
COM (504) 589-6541
FTS 8-426-7010
COM (202) 426-7010
FTS 8-426-0260
COM (202) 426-0260
FTS 8-713-972-5674
COM (713) 972-5674
FTS 8-617-934-5682
COM (617) 934-5682
FTS 8-202-493-5300
8-202-426-7035
COM (301) 426-7035
(301) 493-5300
FTS 8-713-845-3441
COM (713) 845-3441
XI 1
-------
Jonn Proni
W Lawrence Puah
K. Ranga Rao
James Ray
Robert F Shokes
John Trefry
Richard Walentowicz
David White
NOAA
Atlantic Cceanographic &
Meteorological Labs
Ocean Acoustics Lab
4301 Rickenbacker Causeway
Miami , FL 33149
NOAA
6010 Executive 2-vd.
Rockville, MD 20852
University of West Florida
Dept. of Biology
Pensacola, FL 32504
Shell Oil Co.
One Shel1 Plaza
P.O. Box 4320
Houston, TX 77210
Div of Environmental
Chemistry and Geochemistry
Science Applications, Inc.
P.O. Box 2351
La Jolla, CA 92038
Florida Institute of Technology
Dept. of Oceanography
Melbourne, FL 32901
U.S. EPA (WH-585)
Ocean Programs Branch
Washington, D.C. 20460
Florida State University
Nuclear Research Bldg. 310
Tallahassee, FL 32306
FTS 3-350-1319
COM (305) 361-4319
Ext. 336
FTS 8-443-8963
COM (301) 443-8963
FTS 8-904-476-9500
(Opr. Asst.)
COM (904) 476-9500
FTS 3-713-241-3060
COM (713) 241-3060
FTS 8-714-454-3811
COM (714) 454-3811
Ext. 2532
FTS 3-325-2211
(Dor. Asst.)
COM (305) 723-3701
FTS 3-245-3154
COM (202) 245-3154
FTS 8-904-644-5027
COM (904) 644-5027
The workshop was conducted under a Federal Intsragency Energy/Environment
Agreement (EPA-81-D-X0581) between the U.S. Environmental Protection Agency
and the U.S. Fish and Wildlife Service. Project officers were Dr. William
Krohn (FdS) and Dr. Thomas Duke (EPA).
We wish to extend special thanks
coordination of the workshop. We also
to a myriad of organizational details.
to Dr. T Duke for his sponsorship and
thank Mrs. P. Wells for her attention
Although the above individuals deserve the credit for the accompli
of the workshop described in this report, the authors take responsibili
any errors the report may contain.
shments
responsibi1i ty for
XT
-------
INTRODUCTION
Increased oil and gas exploration/production at offshore sites has
generated concern over potential environmental impacts of marine disposal of
spent drilling muds and cuttings. This concern has resulted in a broad array
of publicly and privately sponsored research beginning in, the mid-19701s.
Drilling fluids or "muds" are essential to provide integrity for the
well bore, a medium for removal of drill cuttings, and lubrication and cooling
of the drill bit and drill pipe. Study of the environmental effects of drill-
ing muds and cuttings disposal has been particularly difficult for three
primary reasons. First, the composition of a drilling mud is tailored to
expected or actual down-hole conditions. This means that in addition to the
typical base of bentonite or barite, various chemical agents are added as pH
modifiers, biocides, corrosion inhibitors, defoamers, emulsifiers, flocculating
agents, surfactants, thinners, particle dispersers, and mud weighting agents.
Second, many of the chemical ingredients and materials accumulated from cutting
through the various formations may undergo change when exposed to bore temper-
atures and pressures or to each other (especially in deep wells typical of
offshore drilling activities). The resulting complexity of discharged
materials is reflected in the wide range of concentrations over which effects
are observed. Finally, the fate of discharged drilling muds and cutting is
extremely hard to predict because localized discharges are subject to highly
variable hydrologic conditions.
Although the bulk of drilling muds and cuttings constituents is relatively
inert, discharge of this material may constitute a significant perturbation of
the physical environment. In addition, some mud additives (e.g., lignosulf-
onates and formaldehydes) and components of formation cuttings (e.g., heavy
metals and petroleum hydrocarbons) have been a source of concern because of
toxicity and potential for accumulation and movement through food chains.
OBJECTIVES
To focus available information on these complex, interdisciplinary prob-
lems an Adaptive Environmental Assessment modeling workshop was held with the
broad goals of information synthesis, identification of general relationships
between drilling fluids and the marine environment, and identification of
site-specific variables likely to determine the impacts of drilling muds and
cuttings on the marine environment. The workshop was sponsored by the U.S.
Environmental Protection Agency in conjunction with its research program and
regulatory responsibility in the area of environmental effects of drilling
muds and cuttings discharges into the marine environment. Specific objectives
were:
-------
(a) provide a forum for effective communication between scientists and
administrators working with fate and effects of drilling fluids
disposal;
(b) begin construction of a simulation model to capture the physical and
biological dynamics of drilling fluids disposal in the marine
environment;
(c) identify gaps in information on fate and effects of drilling fluids
discharged into the marine environment; and
(d) initiate identification of factors determining fate and effects,
which will eventually result in guidelines to assist in permit
formulation.
The workshop was held September 14-18, 1981 in Gulf Breeze, Florida. It
was facilitated by the staff of the Adaptive Environmental Assessment Group of
the Western Energy and Land Use Team, U.S. Fish and Wildlife Service and
attended by participants representing Federal, State, and private exoertise on
the fate, effects, and regulation of drilling muds and cutting discharge.
This report is a synthesis of workshop activities and results.
THE ADAPTIVE ENVIRONMENTAL ASSESSMENT METHODOLOGY
The Adaptive Environmental Assessment methodology was developed by
environmental scientists and systems analysts at the University of British
Columbia and the International Institute for Applied Systems Analysis in
Austria. The approach is organized around a series of 3- to 5-day workshops
that define information needs and promote a common understanding of the issues.
These workshops are followed by periods of information collection, analysis,
and synthesis. The workshops are attended by groups of participants, drawn
from key agencies and interests, who collectively represent a range of
scientific expertise, management responsibility, and decisionmaking authority.
These individuals are not only involved in the workshops, but undertake some
of the key tasks of information collection, analysis, and guidance between
workshops.
The focus of AEA workshops is the construction and refinement of a quanti-
tative, dynamic simulation model of the system under study. Early in a par-
ticular application, the process of building a model is usually of greater
benefit than the model itself. Development of a simulation model enables
participants to veiw their expertise in the context of the whole system,
thereby promoting interdisciplinary communication and understanding. Simula-
tion models require explicit information; in building a model, participants
must thus be precise about their assumptions. Conceptual uncertainties about
system behavior are exposed.objectively, and questions that must be addressed
in order to understand system responses to resource development projects are
identi fied.
A modeling workshop thus provides a good beginning to an environmental
analysis. Scientists and policymakers from government agencies, as well as
-------
affected private interests are given an opportunity to participate in and
contribute to an integrated assessment process. A large part of the value of
such a workshop is that it provides a neutral structure or framework, for
focused communication among this set of participants.
-------
BOUNDING THE DRILLING FLUIDS WORKSHOP MODEL
The analysis of fate and effects of marine discharge of muds and cuttings
began by explicitly simplifying the system. Since any simplification of a
real system is an abstraction and therefore incomplete, the representation of
the system must be detailed enough to address most concerns while remaining
understandable to the participants. The process of simplification, or bound-
ing, was accomplished in the workshop by describing management alternatives
(actions), identifying performance measures used to evaluate the effects of
those actions (indicators), and defining a reasonable spatial and temporal
framework.
ACTIONS
Actions, or human interventions, identified at the drilling fluids work-
shop are listed in Table 1. As one would expect, all of the actions pertained
to operations at the drilling site since there is no practical means of
altering the fate of the materials once they have been released into the
marine environment. Therefore, the general issue addressed at the workshop
was the potential environmental effects of various modes of drilling
di scharge.
Table 1. List of actions developed at workshop.
Alter discharge depth
Alter discharge rate
Dilute prior to discharge
Alter spatial configuration of discharge (i.e., spread out)
Alter mud composition (i.e., light*, medium, heavy*)
Locate the drilling rig over either a soft bottom or a hard bottom
Dispose on land*
Treat drilling fluids before discharge*
*Not explicitly addressed in the model.
-------
INDICATORS
Indicators are defined as those variables used to evaluate the performance
or health of a system. They are the links between the simulation model and
participants' perceptions of the system. Therefore, it is important to compile
a comprehensive set of indicators to represent the concerns of all interests.
Indicators identified at the workshop are listed in Table 2. For purposes
of clarity, they have been grouped according to the submodel responsible for
producing them. Many of the indicators were judged to be of secondary import-
ance. Others could not be included within the time constraints of the
workshop.
Table 2. List of indicators developed at the workshop.
Model component
Indicator
Discharge/Fate
concentration of suspended solids, barium,
and chromium in discharged muds and the
resulting plume
depth and area of deposited muds and cuttings
pH of discharge and plume*
salinity of discharge and plume*
DO in plume*
light transmittance in plume*
drilling costs*
Water Column Effects
zooplankton mortality rate within
primary and secondary production*
recruitment to benthos
pi ume
Soft Bottom and
Hard Bottom Effects
population size for coral, microbes,
meiofauna, and macrofauna (infauna,
epi fauna)
bioaccumulation by benthic organisms
coral growth rate
species diversity*
respiration*
reproduction*
di sease*
nutritional status*
material transfer*
organism behavior*
fishery yield*
Not explicitly addressed in model
-------
SPACE
For purposes of simulation modeling, two aspects of space are usually
defined. First, the boundaries of the total area represented in the model,
and second, the degree of resolution or number of smaller subunits considered
within the overall boundaries must be specified.
It was decided that a specific geograohic location was inappropriate for
this model. The model was structured to represent the discharge plume from a
hypothetical drilling rig in an "open" water environment in the Gulf of Mexico.
Three effects submodels then calculated biological responses for the water
column, soft bottom benthos (assuming the rig was located over a soft bottom
community), and hard bottom benthos (assuming the rig was located over a hard
bottom community). Two spatial resolutions were aefined within the plume.
The Water Column submodel used a set of plume slices each representing 1 min
of discharge (see Water Column suomodel discussion), while the Hard and Soft
Bottom submoaels represented environmental effects in l-m: areas at five
distance down current from the discharge (1, 50, 100, 500, and 1500 m)
The workshop simulation model was developed for "open" water environments.
Participants felt that modeling fate and effects of discharged drilling muds
and cuttings in more enclosed water environments, such as bays and estuaries,
would require an effort devoted more completely to those environments.
However, because of their importance a subgroup was convened to discuss fate
and effects in these areas. This group's objective was to identify factors
determining fate and effects in more "closed" water environments, focusing on
variables that might produce different behavior from that expected in "open"
water environments or that might produce differences among various "closed"
water environments.
TIME
There are two aspects of time that must be considered in a simulation
model: the time horizon or length of time for which model predictions are
desired, and the time step or interval used to calculate changes in variables
throughout the length of the simulation.
For example, in a simulation of human population a time horizon of 50
years might be appropriate, indicating that the model would track population
size over a 50-year period. An annual time step might be chosen, in which
case, annual birth and death rates might be utilized to calculate new values
of the population size each year. In contrast, the U.S. Census Bureau's
approach to tracking population size has been to utilize a time step of 10
years, updating the value of population size by enumeration every 10 years.
A time horizon of 20-30 years was selected for this model. The partici-
pants chose this time horizon so that effects on slow growing corals and their
recovery could be simulated. The incremental time step proved to be more
troublesome because relevant processes operate at very different time scales.
For example, plankton in the water column and microbes in the sediments respond
-------
to perturbations in a matter of minutes to hours while response times of
organisms such as corals or crabs may be months to years. Because of this
disparity, a monthly interval was selected as a reasonable compromise given
the degree of knowledge about population dynamics of the indicator organisms
and the amount of time available to model these dynamics. The exception to
this decision was the 1-minute time step used to represent plankton dynamics.
SUBMODEL DEFINITIONS
The marine system defined by the actions, indicators, spatial scale, and
temporal framework described above was divided into four subsystems. The
criteria for useful division of a model into submodels at a workshop are:
(a) minimizing information transfers between submodels (each subgroup
considers a relatively isolated part of the whole system);
(b) allocating participant expertise efficiently (each submodel repre-
sents the concerns and expertise of a set of participants); and
(c) partitioning the workload equally among facilitators so that partic-
ipants have an opportunity to incorporate an appropriate amount of
depth in their area of expertise.
After considerable discussion the following major components (submodels)
were selected for the model:
(a) Discharge/Fate - discharge characteristics of oil and gas explora-
tion rigs and production platforms and the subsequent fate of the
discharge materials;
(b) Water Column Effects - dynamics of zooplankton and larval forms of
benthic organisms within the upper plume;
(c) Soft Bottom Effects - effects of exposure to drilling muds and
cuttings on microbes, meiofauna, and infaunal and epifaunal repre-
sentatives of macrofauna; and
(d) Hard Bottom Effects - responses of coral to exposure to drilling muds
and cutti ngs.
As previously noted, an additional subgroup explicitly considered how the
fate and effects of drilling muds and cuttings might differ in more "closed"
water bodies such as bays and estuaries. This group did not attempt to build
a simulation model treating the components of these systems in the detail that
open water systems were being addressed. Instead they focused on identifying
the variables or factors that would determine differences in fate and effects
between these environments 'and those for which a simulation model was being
developed. The results of these discussions are incorporated in the concluding
section of this report.
-------
SUBMODEL INTERACTIONS
Following submodel definition, workshop participants defined the linkages
or information transfers between the submodels. These are deoicted in a
looking outward matrix (Fig 1) in which submodels are arrayed as both row and
column headings. For each element of the matrix, participants identified what
they needed to know from other submodels in order to meet their responsibil-
ities for quantifying indicators and for providing needed information to other
elements of the matrix (i.e., other submodels). In other words, each subgroup
was asked to "look outward" to other subgrouos for needed information. Note
that this is a qualitatively different question than the more common one of
what information can be provided, rather than what i-nformation is needed.
Identification of the information transfers in a looking outward matrix
is valuable in several ways. First, the exercise promotes interdisciplinary
communication and broadens participants' understanding of the system. Second,
the looking outward matrix lays the foundation for building a simulation
model Submodel construction quantifies how the information recuested in the
matrix affects the variables of a particular submodel. If sufficient informa-
tion exists, such relationships can usually be formulated. If not, an informa-
tion gap or research need is identified. Third, the resulting simulation
model can be used to test the sensitivity of the information transfers.
Sensitive transfers can be noted for further, more detailed, investigation.
The looking outward matrix constructed during the workshop contains
relatively few entries (Fig 1). This reflects a focus on direct linkages
between the discharge (Discharge/Fate submodel) and various organisms rather
than on how the marine ecosystem itself is interconnected (e.g , how corals
are dependent on plankton or how pelagic fish are dependent on bentnic fauna).
-------
figure 1. Workshop looking outward matrix of
Information transfers between submodels.
KISCIIAIIGE /FA It
WAI 1:11 COM/MM CTItCIS
SOI T UO1 IIHJI 1:1-1 tOTS
IIAIIIl UOI IOM LI I I'CIS
DlacliLifuu cliui licluilsllcn
I. f'liiniu unU cunceuli ulton
yraillunla ul 10 uiliiuloa.
1 liunr. and 2 liuma lur
lylilcul ilNchui uu
- SullllllOfll U4J|JUbtllull
I. Uaplli al (jlvtin Ulaluncu:
2. CuiiijiuallluM (% wl.olo
niuij) ul ylvon dluluncua
% fujticlluii lit honlltlc
I UCdllllllDIlt IfOIII WulUI
2. Maximum cuiu:i:itli ii
ul utwuil ill:.lnni:ii£i
3. Dtii ulluu ul ly|>!i:nl
2. Coc,,,>ual!lun ('«. whulu
litutl) ul u'v^''t onllilc
rticiiilliiiciil liuni wtilui
uuli
'^JJ^TTv^J','/-'-'~'~-'^r.7 '<'<"-"
^'&/t'-;r'?'/'/.-^'.£':'
^iMrn^l-
'"^''/sW/-''/^ ''-''''.
'^^2y^^:-
-------
SUBMODEL STRUCTURES
DISCHARGE/FATE SUBMODEL
Responsibi1i ties
The Discharge/Fate submodel was responsible for determining characteris-
tics of drilling muds and cuttings discharge plumes and fates of various
materials in those plumes. Specific indicators of interest included discharge
frequency and duration and associated plume size, dispersion ratios, concentra-
tions of soluble and solids fractions at different distances from the platform,
and both the depth of sediment added and fraction whole mud in sediments at
different distances. Actions of interest included variations in discharge
rate and amount, predilution, and shunting. The submodel could also respond
to differences in site characteristics such as current velocity, water depth,
depth of neutral buoyancy (an approximation for density stratification), and
storm frequency and severity.
Structure
Quantitative prediction of the fate of ocean discharged drilling materials
generally requires extremely complex mathematical models. This complexity
arises from temporal and spatial variation in current velocity and density
stratification, the highly variable composition of drilling muds, and the
chemical and physical interactions of mud components following discharge. A
number of complex mathematical ocean discharge models have been developed over
the last 10 years (e.g., Koh and Chang 1973; Teeter and Baumgartner 1979;
Brandsma et al. 1980; Houghton et al. 1980). Time during the workshop did not
permit such a complex treatment of plume dynamics; therefore, a more empirical
approach was taken.
Drilling rigs typically have continuous discharges of solids at low rates
(1-10 bbl/hr) while actually drilling, and periodic bulk discharges at higher
rates (100-1,000 bbl/hr). The continuous discharges primarily contain cuttings
that are separated from the mud before it is reused, while the bulk discharges
contain some cuttings but are primarily spent muds that have lost their
efficiency. These discharges were conceptualized as separating into two
components (Fig. 2); an upper plume containing the liquid fractions of the mud
as well as solids such as fine-grained silts and clays, and a lower plume
containing cuttings and most of the other discharged solids. Since the contin-
uous, low-rate discharge is primarily cuttings, only lower plume dynamics were
modeled for the continuous discharge. Both upper and lower plume dynamics
were modeled for bulk discharges.
10
-------
figure 2. Idealized drilling platform discharge.
I
Upper plume
Depth ot
neutral
buoyancy
Lower plume
Upper Plume
The upper plume of liquid mud fractions and fine-grained materials was
conceptualized as 20 m thick, spreading at an angle of 53° for the first
10 min of transport time (a function of current velocity), and maintaining a
constant width subsequently. The plume was assumed to be at a depth of neutral
buoyancy, specified for each model run. While plume characteristics can vary
greatly, these assumptions seemed reasonable based on observations and measure-
ments by subgroup participants on plumes in the Gulf of Mexico, Southern
California, and the Mid-Atlantic. They represent an empirical alternative to
the complex mathematics required to model explicitly the convective descent,
dynamic collapse, and passive diffusion phases of a plume. -In actuality, the
plume would be spread out in a wider, thinner layer following its dynamic
collapse phase, but the assumptions above yielded reasonable plume character-
istics for purposes of this model.
Plume volume (m3), dilution and concentration of the soluble fraction
(mg/1 or ppm), and dispersion ratio and concentration of the solids fraction
(mg/1) were calculated at distances of 1, 50, 100, 500, and 1,500 m from the
drilling rig. Soluble fraction dilution occurs by entrainment of seawater
into the plume and was calculated as the volume of liquid discharged divided
by the plume volume at each distance. Soluble fraction concentration was
calculated as initial concentration divided by the dilution factor. Dispersion
of the solids fraction occurs through entrainment of seawater as well as
particulate settling. The dispersion ratio (suspended solids in discharge/
suspended solids in plume) was calculated from a multiple regression using
transport time and the inverse of discharge rate as independent variables:
11
-------
OISPR = 104'4495 * (1/DSCHR)0'35674 * (TT)L1001 (1)
where DISPR = dispersion ratio
OSCHR = discharge rate (bbl/hr)
TT = transport time (min)
This regression was based on measured dispersion ratios from wells in the Gulf
of Mexico, Tanner Bank, and the mid-Atlantic summarized in Petrazzuolo
(Table 6-4, 1981). The squared correlation coef f ici ent..(R2) for this regres-
sion was 0.74.
Lower Plume
It was assumed that, over a sufficiently long time period, solids from
the lower plume (cuttings and spent muds) would be deposited evenly over a
circular area around the platform. In actuality, solids from individual
discharges are deposited primarily in one direction away from the platform by
prevailing currents. As currents change through the life of a platform, a
starburst depositional pattern is often produced with greater sediment depths
nearer the platform. An attempt was made to incorporate varying sediment
depths based on Petrazzuolo's (1981) empirical model of fraction of whole muds
in surface sediments; however, an adequate formulation could not be derived in
the time available. Although discussed during the workshop, time also did not
permit incorporation of horizontal spreading of the descending plume near the
sea bed or resuspensive spreading of deposited materials in this first cut
model. The approximation of even deposition over a circular area, therefore,
did not completely reflect the spatial variability of deposition or severity
of impact. The circular area and depth of added sediment were, however,
useful indicators for comparing scenarios and for use by other submodels. The
radius of this circle was calculated as:
RADIUS = tangent (ANGLE) * (DEPTH) (2)
where RADIUS = radius of deposition (m)
ANGLE = angle of drift
DEPTH = depth from discharge to bottom (m)
12
-------
The angle of drift was calculated as:
ANGLE = arc tangent (CURR/PSR) (3)
where ANGLE = angle of drift
CURR = current velocity
PSR = particle settling rate
A portion of this circle could be specified to receive the total deposition,
thus simulating situations where currents are predominantly in one direction.
Depth of added sediment (cm) was calculated by dividing the total volume of
discharged solids by the area covered. These added solids can also change the
sediment particle size distribution which may in turn affect indigenous benthic
organisms and recruitment of benthic organisms. Particle size effects were
not incorporated in the workshop model.
The fraction of a sediment sample that is whole drilling muds has been
used as an indicator of toxicity to benthic organisms. This is usually
measured by sediment barium concentrations. In the Discharge/Fate submodel,
excess barium added to the sediment from each well at different distances from
a drilling rig was calculated from the following empirical relationship
modified from Petrazzuolo (1981):
EBAR = (50,000 * e"'003 * DIST)/(10 + DIST'5) (4)
where EBAR = excess barium (mg/kg)
DIST = distance from rig (m)
The fraction of whole muds was calculated as excess barium in the sediment at
each distance divided by the concentration of barium in whole muds.
Sediments near drilling rigs are also affected by periodic severe storms
that can displace the upper 1 cm or more of sediment and thereby eventually
eliminate any indication of drilling solids deposition. In the Discharge/
Fate submodel , the average time between such major storms and the amount of
sediment displaced could be specified to represent different geographical
locations. The effect of these storms in the submodel was to reduce added
sediment and associated excess barium.
13
-------
Behavior
The Discharge/Fate submodel could be parameterized to simulate either an
exploratory rig or a production platform. It was assumed that an exploratory
rig would drill a single well over a 3-month period with bulk discharges of
600 bbl every 3 days and a total solids discharge of 2,250 metric tons. A
production platform was assumed to drill 20 wells consecutively at 6 weeks per
well with similar bulk discharge characteristics but a total of only 1,500
metric tons of solids per well. For purposes of model runs, characteristics
of a 13 Ib/gal mud were assumed.
A production platform scenario was run to demonstrate behavior of the
Discharge/Fate submodel. Assuming a 10 m/min current and a total discharge of
600 bbl, upper plume characteristics were calculated for discharge rates of
30, 100, 275, 500, 750, and 1,000 bbl/hr. Results are shown in Table 3.
Assuming the same current, an 80-m depth from the discharge to the bottom, and
no periodic severe storms; sediment buildup was calculated as 17 cm over a
circular area of radius 154 m (Fig. 3). Figure 3 also shows the effects of
periodic storms; which occur on the average every 6-months and remove either
1 cm or 2 cm of sediment per storm. Assuming no periodic severe storms,
sediment buildup characteristics for water columns 20 m and 80 m deep with
currents of 1, 5, and 10 m/min are presented in Table 4.
Limitations
The Discharge/Fate submodel provided reasonable plume characteristics for
use by other submodels. However, the lack of explicit mathematical treatment
of detailed physical and chemical plume dynamics, spatial and temporal vari-
ability in currents, and density stratification precluded addressing certain
important questions. For example, plume constituents may become concentrated
at stratification layers where certain life stages of some species are found.
This possible concentration and its effects on organisms could not be explored
with the submodel structure described above. Another question that was not
addressed concerns potential integrated or cumulative effects of multiple
platforms in close proximity. Another topic of discussion was the effect of
shunting. The purpose of shunting discharges to the bottom is to limit the
area impacted by cuttings and solids and to keep the liquid fractions and
fine silts and clays below the pycnocline. Shunting in the submodel did limit
bottom areas impacted, but assumed the liquid and fine-grained fractions would
rise to the specified level of neutral buoyancy and therefore potentially
still affect the pycnocline. The fate of these shunted upper plume components
under actual discharge conditions (staying approximately at shunted depth vs.
rising to pycnocline) was discussed at the workshop but not resolved. To
address questions such as the ones posed above, a much more detailed,
mechanistic modeling approach would be required.
14
-------
Table 3. Upper plume characteristics at various discharge rates.
(xEy = x * 10y)
Oischaryif rale Plume volume (m |
(bbl/hr) 1m 5Sn jOOin " MOm "IBOOin
30 ').9 2.5E4 I.OE5
lOfl 9.9 2.SF4 I.OE5
275 9.9 2.5E4 I.OF6
500 9.9 2.5F4 1.015
750 9.9 2.5E4 1.UE5
1.01)0 9.9 2. 511 1.0F5
9.0E5
9.0E5
9.0E5
B.9E5
H 6E5
6.2F5
2.9E6
2.9E6
2.5E6
1.3E6
0.6E5
6.2E5
Ullutlon factor for soluble fraction
1m 50in lOOin SOIJiii IGOIIiii
I-6E3 7.9E4 1.6E5 2.0E5 3. US
4.7E2 2.414 4. 7E4 B.5E4 9.2f4
1.7E2 H.6E3 1.7E4 3 IE1 3 3H
9.5EI 4.7E3 9.5E3 I.7F4 I.UI4
6..3E1 3.2F3 6.3E3 1.114 1.114
4.7EI 2.4E.3 4.7E3 B.2E3 fl 2E3
(xly = x . Id')
Dispersion ratio for solids fraction
lin fill."
6.1.12 4.')I4
4.317 3.2(4
3.1112 2.2E4
2.4E2 1.IIM
2.6C2 1 6f4
1.9E2 1.4f4
lOlhu 5011m I'jlllhn
1.115 6.?f5 2.116
6.114 4.1115 1 41 f.
4.Bf4 2.1)1!. 9.1fb
3.9I4 2.2I5 7.6f5
3.3(4 2.015 6.Cf5
3.0E4 I.B15 5.915
-------
figure 3. Depth of spent mud solids and cuttings
under various conditions.
-------
WATER COLUMN EFFECTS SUBMODEL
Responsibi11 ties
The prime indicators of water column "health" were considered to be
primary and secondary production. The Water Column Effects submodel focused,
however, on estimating the proportion of planktonic animals within the plume
that might be killed by a single discharge of drilling fluids. Subgroup
participants felt that this would be a sensitive and tractable indicator of
water column effects, given the spatial and temporal scales of the discharge
from a single rig. Zooplankton mortality in the plume was used to estimate
the percentage loss in monthly recruitment of larval forms to the benthos,
considering the number of discharges per month and duration of each discharge.
Zooplankton mortality was calculated separately for the discharge and post-
discharge phases of the plume. Development and movement of the plume during
these two phases is depicted in Figure 4.
Structure
Mortality during discharge phase. The form of the plume assumed by the
Discharge/Fate submodel was divided into slices, each representing 1 minute's
discharge (Fig. 5). Since the plume was assumed to remain at constant width
after 10 minutes, organisms were entrained only within slices 1 to 10. The
submodel considered only the "area" of organisms entrained, since plume depth
was assumed to be constant. Zooplankton populations were thus represented by
areas (m2), which could be converted to more conventional measures of number
of individuals or biomass by utilizing the constant depth of the plume and a
site-specific estimate of Zooplankton density. The area entrained within a
given slice "i" was simply the area of slice i minus the area of slice (i-1).
It was assumed that animals entrained at a given point within the plume (i.e.,
somewhere in slices 1 through 10) would be carried with the current and thus
exposed to a declining concentration gradient (Fig. 6). The duration of a
subpopulation's exposure to this gradient during the discharge phase depended
upon which slice entrained it and how long the discharge continued after the
subpopulation was entrained. For example, within a 36-minute discharge, there
were 315 different subpopulations with different exposure "schedules".
The concentrations of solubles in the slices (calculated as outlined in
the Discharge/Fate submodel description and shown for a test run in Fig. 6)
were used to compute an average concentration (c ) for the period of exposure
(t) of each subpopulation:
s+(d-t)
4 c(Q
cst (d-t+1) (5)
17
-------
ruqure ^. Tea vvsu OT cssurnec ceveLcpmerv end
<- ' i . i , ^
movement or uccer pLume c^ savercL tumes duru
Ml ,''.',,, ,
cuschcrae cnc 3CST,~GL-scncrce cncses.
DISCHARGE
PHASE
POST-DISCHARGE
PHASE
o
-------
Figure 5. Top view of upper plume slices used In
water column submodel calculations.
Figure 6. Concentration gradient of soluble phase
In discharge plume at 1000 bbl/hr discharge rate.
s
Q_
Q_
en
en
n:
Q_
[jj
CO
O
CO
2500n
2000-
1500-
1000-
500-
) 10
1
20
1
30
1
40
PLUME SLICE NUMBER
19
-------
where c = mean exposure concentration for organism entering slice "s" at
time "t" (where t=l is first minute of discharge)
c(i) = concentration of solubles (ppm) in slice "i
d = total duration of discharge (min)
Following Petranuolo (1981), an LCso value appropriate to each subpopula-
tion's "t" minutes of exposure was estimated for the discharge phase by
converting the 96-hr LC50 according to:
,
if nn u if *t 96 hr * 60 Jiin\2 (6)
t-minute LCcn = 96-hr LCC~ *( * ; v
50 50 t mm nr
Using equation 6, 60 "toxicity curves" were constructed for 1 to 50-min
exposures, assuming that the general sigmoid shape of Figure 7 aoolied in all
cases. A 96-hr LCso of 50 ppm was assumed in the right hand side of
equation 6. This is a conservative value for 96-hr LC 5,3 since a va'ue of
100 ppm whole mud is reported as the 96-hr LC 50 for the most sensitive species
tested (Petrazzuolo 1981) The survival rate of each subpopul ation was then
calculated and used to compute the total loss of plankton, expressed as a top
view area of plume (m2), during the discharge phase (TIQP):
n
TLDP = I A.(1-SD.)
J J
where A. = area of subpopul ation j (m-)
j = subpopul ation index
n = total number of subpopul ations (= 10d-45)
d = total duration of discharge (min)
SD = survival of subpopul ati on j in discharge phase
20
-------
Flqure 7. Generalized toxlcltij curve used to
calculate survival rates In tne water column
at different concentrations.
1
UJ
t-i
en
or:
C£
ID
cn
0.75-
0.50H
0.25-
50 100 150
PERCENT OF LC
50
200 250 300
CONCENTRnTION
350
Mortality following discharge. Although observed concentrations from the
upper plume gradually decline over several hours following the discharge, the
model assumed that exposure during the post-discharge period could be repre-
sented by an exposure at the concentration found at the end of the discharge;
rather than a series of decreasing concentrations resulting from continued
dilution. The plume was thus assumed to remain the same size throughout the
post-discharge period (Fig. 4). This assumption was necessary since the
dynamics of the upper plume during the dynamic collapse phase were not
explicitly represented in the Discharge/Fate submodel. For discharges less
than 1 hour in length, the post-discharge period was assumed to be 2 hours
long. Exposures during discharges greater than 1 hour duration were (for ease
21
-------
of computation) divided into a discharge phase 1 hour long (at declining
concentrations) and a "post-discharge phase" with exposure in the "post-
discharge phase" fixed at the concentrations existing after 1 hour. This
simplifying assumption is reasonable due to the slow rate of change in concen-
trations after 1 hour. A survival rate for the post-discharge period was
computed for each subarea (or subpopul ati on) , following the same procedure as
outlined for the discharge period, but substituting t=120 minutes or more in
equation 5.
Total plume mortality and potential monthly benthic recruitment losses.
Total plume mortality rate (TPM) over the two phases was calculated as:
n
TPM = I A.*(1-SD.)*(1-SPD.)/(FPA) (8)
= J J J
where SPD . = survival of subpopul ati on j, during the post-discharge phase
J
FPA = final area of plume (m2)
j, n, A., and SD . are as defined in equation 7
\J >J
The relative loss in monthly recruitment to the benthos (RLOSS) was then
computed (on a scale from 0 to 1) by:
RLOSS - NDIS - DDIS * TPM , .
KLUii - 3o(days/month) * 24(hrs/day) * K { '
where NDIS = number of discharges per month
DDIS = duration of discharge and post-discharge phases (hr)
TPM = total plume mortality rate as defined in equation 8
death of water column
K =
depth of plume
Equation 9 illustrates that even with 100% mortality in the plume, the monthly
reduction in potential benthic recruits would be very small. Assuming 10
discharges per month, each lasting 2.5 hours (discharge + post-discharge
phase) and causing 100% mortality, benthic recruitment would be reduced by
only 1 7% in a 40-m water column:
22
-------
RLOSS =
10 * 2.5 * 1.0
30 * 24 * 2
= 0.017
(10)
Behavior
Normal discharge rate. The water column concentrations of solubles for a
36-minute, 1000 bbl/hr discharge at a 10 m/sec current velocity were as shown
in Figure 6. Although concentrations of solubles in the post-discharge phase
were generally lower than in the discharge phase, survival rates in the post-
discharge phase were also generally lower. Lower survivorship in the post-
discharge phase was due to longer exposure times. The total mortality under
these conditions over the two phases was 8.2%, with 96% of this mortality
occurring in the post-discharge phase.
Although survival rates during
the slices nearest the rig, as shown
occurred in slice 5 (Fig. 9). This
both the survival rates (a function
the post-discharge phase were lowest in
in Figure 8, the highest plankton losses
is because the total losses depend upon
of concentration and exposure time) and
the size of the exposed population (area of plankton) that are in a given
slice. As one moves away from the rig, these variables change at different
rates, producing the largest total losses in slice 5.
Decreased discharge rate. When the discharge rate was reduced from
1,000 bbl/hr to 30 bbl/hr the water column concentrations dropped from the
levels shown in Figure 3 to those in Figure 10. Total plankton mortality per
discharge fell from 8.2% to 0.003%.
Figure 8. Survival In post-dlacharge phase versus
position In plume.
1 -i
0.75-
faJ
E-H
a:
CC 0.50-
cc
=3
CT,
0.25-
r
10
r
20
T
30
T~
40
PLUME SLICE NUMBER
23
-------
figure 9. Total loss of plankton In post-discharge
phase versus position In plume.
400-1
300-
CD
CO
O
_) 200-
UJ
az
en
_j
S 100~
0
n
i
0
-j
tk^^
10 20 30 10
PLUME SLICE NUMBER
figure 10. Concentration gradient of soluble phase
In discharge plume at 30 bbl/hr discharge rate.
80-1
a. 60
a.
UJ
in
LJ
_1
no
o
to
10-
20-
10 20 30
PLUME SLICE NUMBER
24
-------
Limitations
The Water Column Effects submodel consists of a collection of hypotheses
about exposure and effects of drilling muds. These hypotheses need to be
stated explicitly and criticized to reveal the uncertainties associated with
model predictions and the priorities for information needs. This section of
the report challenges the basic hypotheses of the Water Column Effects submodel
as a means of discussing the major difficulties with effects prediction.
It was clear at the workshop that 96-hr toxicity tests at constant concen-
trations do not accurately simulate the exposures experienced by organisms in
the field. Equation 6, used to convert 96-hr LC50 values to shorter periods
of exposure, assumes that the LC50 for a shorter period-should be increased_by
a factor equal to the square root of the relative exposure time (e.g.,/96/l
for a 1-hour exposure). For example, equation 6 predicts that a population
exposed to a toxicant far 1 hour rather than 96 hours would require a concen-
tration equal to about ten times the 96-hr LC to kill 50% of the exposed
population. Estimates of mortality in the plume itself are quite sensitive to
the assumptions used to apply 96-hr tests to other time periods. Although
assumptions used in the LC50 extrapolation produce large differences in
mortality within the plume, they do not have a large overall effect on a
variable such as benthic recruitment because the bulk discharge plumes occur a
relatively small fraction any month as indicated in equations 9 and 10.
A second problem with assessment of plankton survival is the assumption
that survival through an exponentially decreasing series of concentrations
over the discharge can be estimated by survival at the mean concentration over
this period. An alternative approach to this problem would be to use only
1-min toxicity curves, and use the product of the respective survival rates to
estimate survival over the whole discharge period. This method potentially
runs into other conceptual difficulties, namely, the issues of variability
(within a subpopulation) in individual organisms' sensitivity to the toxicant,
and selection for tolerant individuals over the duration of the plume.
To clarify this conceptual problem, consider a series of two exposures
(of equal duration) at 50 ppm and 100 ppm to an initial population of 100
individuals. Survival using the "non-selective" toxicity curve shown in
Figure 11 for both exposures would yield 5 individuals at the end of the
second test (0.5 survivorship in first exposure * 0.1 survivorship in second
exposure * 100 = 5) However, if one assumes that the first exposure removes
the 50 most sensitive organisms, then the toxicity curve for the remaining 50
individuals might be as shown in the "selective" curve of Figure 11. Under
this toxicity curve, exposures of 50 ppm or less have no effect, because the
population receiving such exposures consists of the more tolerant individuals
from the original population. The second exposure of 100 ppm would only cause
20% of these hardy organisms to die (Fig. 11), leaving 40 individuals at the
end of the second exposure.. Table 5 summarizes these calculations. Though
consideration of selection for toxicant resistance is probably unnecessary for
very short exposures, it may be important if longer term survival is to be
considered as the result of a large number of such exposures as might be the
case in more "closed" water bodies or multiple platform fields.
25
-------
figure 11. Selective and non-selective Loxlclty
curves.
Legend
NON^SELECTWE
SELECTIVE
UJ
fr-i
or
cc:
cc:
r?
CO
0.75-
0.50-
0.25-
CONCENTRflTION (pprn
Table 5. Effects of assumptions on population variability in
sensitivity and selectivity of toxicant.
Time
Number of animals remaining
Start of 1st exposure
End of 1st exposure
End of 2nd exposure
No selection
(using Fig. 11
"Non-selective"
curve in both
exposures)
100
50
5
With selection
(using Fig. 11
"Selective" curve
in second ex-
posure)
100
50
40
26
-------
It would be interesting to do some 2-hr toxicity tests with exponentially
decreasing concentrations of drilling muds, using organisms that have pre-
viously been extensively tested at constant concentrations.
The assumption that the 96-hr LC50 of 50 ppm is representative of most
zooplankton seems unduly conservative. Measured EC5g values of 50 ppm were
attained for scallop larvae using relatively toxic Mobile Bay muds, but values
are as high as 50,000 ppm for low density muds (Tom Gilbert, see
ACKNOWLEDGEMENTS Section). Similar ranges in toxicity have been found for
grass shrimp larvae and copepods. When the assumed 96-hr LC was varied in
the model, the total plankton mortality under normal discharge (1000 bbl/hr,
36 min) decreased according to Figure 12. At 96-hr LCso values greater than
930 ppm there was zero mortality.
The assumption that concentrations in the plume remain constant during
the post-discharge phase and return to background levels after 2 hours may
have led to either an overestimation or underestimation of post-discharge
phase mortality. The direction of error depends upon the extent to which the
real world decreases in concentrations over those two hours compensate for the
fact that parts of the plume may remain above background concentrations for
1onger than 2 hours .
Figure 12. Sensitivity of total plankton mortality
of exposed population to assumed 96~hr LCC
a:
E-.
en
a
az
E-i
o
i-l
IQ-i
8-
50
100 200 300 -10Q 500
flSSUMED 96-HOUR LC (ppm)
600
27
-------
The estimation of water column effects depends on the assumptions used to
represent plume dynamics. The behavior of the plume in relation to water
column stratification (the pycocnline and more subtle stratification layers of
particulates) is especially important. A relative concentration of discharged
material, and perhaps of the biota (such as larval stages), in these zones
might lead to greater effects than those indicated by the assumptions used
here.
Notwithstanding the above uncertainities about zooplankton mortality
within the upper plume, the relatively rapid return of water column concentra-
tions to background levels suggests that the impact of a single drilling rig
on benthic recruitment in the open ocean is likely to be negligible. The
impact might be more serious with multiple drilling rigs, in enclosed areas,
or in situations where a species is present in the water column for a very
short time (e.g., as a larval stage) or in a restricted location (such as a
particular stratification layer) that coincides with high concentrations of
discharged materials.
SOFT BOTTOM EFFECTS SUBMODEL
Responsibi1i ti es
This submodel had a deceptively simple set of responsibilities. The
first was to represent population levels, expressed as g/m2 or numoers/m2, of
microbial, meiofaunal, infaunal, and epifaunal components of a hypothetical
benthic community. The second responsibility was to produce an index of
bioaccumulation levels and sublethal effects due to exposure to sediments
containing a fraction of deposited spent mud and cuttings.
Addressing these responsibilities required considerable simplification of
complex biological processes. However, subgroup members, after much agonizing,
decided that the general behavior of the separate components of a generic soft
bottom benthic community could be reasonably represented although such a model
would be highly deficient in explicit representation of interactions between
benthic components.
One of the consequences of construction of such a general conceptutal
model was that specific examples could not always be used to define the
responses of the hypothesized community. For example, recol onizaton by infauna
and epifauna or redevelopment of the oxygenated zone were generalizations
developed from the collective input of the subgroup participants. If another
type of community had been hypothesized, it may have been equally valid to use
the results of specific experiments (i.e., Boesch and Rosenberg 1981 or
Cantelmo et al. 1979) to derive appropriate response behavior The main
point is not how accurately the submodel portrays a particular site, but what
has been learned about the information needed if a credible predictive model
of soft bottom benthos is to'be constructed.
28
-------
Structure
The subgroup members first emphasized that the basic assumption under-
lying this particular submodel is that the soft bottom ecosystem represented
by the model is dominated by storm events. Therefore, the resulting community
is composed of "invader" species. The lack of stability in the substrate
structure means short recovery/colonization times often characterized by
overcompensation (increases) in the biomass of microbes and meiofauna. If a
different type of benthic community (i.e., one from a stable substrate) had
been modeled, the above characteristics would certainly be very different.
The submodel dealt with bioaccumulation, survival, and sublethal responses of
four indicator groups (expressed as g biomass/m2 or -numbers/m2).
Bioaccumulation of chromium depended on exposure to deposited sediments
expressed as fraction whole mud (Fig. 13). Tissue buildup continued until all
of the drilling mud was^removed by storm events. Estimates of tissue concen-
tration of chromium (in oysters), in this case ppm above background, were
derived by the subgroup members based on work by McCulloch et al . (1980). The
subgroup was presented with the dilemma of how to deal with the ability of the
organism to flush excess chromium from its system while accounting for exposure
on a monthly time step. Oyster flushing rate was considered sufficient to
reduce tissue concentrations to ambient levels in less than one month. There-
fore, the subgroup concensus for the modeling approach was that if more than
four drilling fluid discharges occurred in a month, the tissue concentration
of chromium would be that which would be expected from exposure to the sediment
input during that month (i.e ., additional exposure).
Figure 13. Eplfaunal tissue concentration of
chromium (above background) as a function of
fraction whole mud In sediment.
4Q-,
E
Q_
Q_
£-<
or
cc:
UJ
o
^.
o
o
LJ 10-
cn
en
30-
20-
0.00 0.02 0.04 0.06 0.08
FRACTION HHOLE MUD IN SEDIMENT
29
-------
Mortality of soft bottom organ-isms was caused by burial with spent mud
and cuttings (Fig. 14), by toxicity of the spent mud (cig. 15), and by removal
of deposited sediments by storms (Fig. 16). Burial survival rates were esti-
mated using data collected by U.S. Army Corps of Engineers (CE) in the Great
Lakes. Toxicity estimates due to exposure to barium concentration in the
sediments were derived by Petrazzuolo (1981). When interpreting the responses
of the soft bottom community it must be kept in mind that use of the
Petrazzuolo (1981) toxicity responses assumes that the community represented
in the submodel is not qualitatively or quantitatively different from those
used by Petrazzaolo to derive the toxicity responses. Storm events only
affected deposited sediments. Population changes due to storm intensity were
indexed according to the amount of sediment removed by each storm. Sublethal
effects were derived from Petrazzuolo (1981) and expressed as the percentage
of organisms showing altered physiological indicators in response to various
fractions of whole muds (Fig. 17) Although it was recognized that sublethal
effects will, in part, govern such things as recovery rates and population
levels, the functional relationships were unknown and therefore not incorpoi
ated into the submodel. Therefore sublethal effects stand as an unconnected
i ndicator.
Population recovery, or colonization, was affected by the depth of
deposited sediments, the fraction whole mud, and the time required for re-
establishment of the layer of oxygenated sediment. In the cases of microbes
and meiofauna, population response due to addition or removal of sediments
resulted in considerable overshoot in the populations in the month of the
disturbance (Figs. 18 and 19) before settling back to original bicmass levels
after 2 months. Recovery to original population levels was modified by the
time required for re-establishment of the oxygenated layer. The rate at which
the oxygenated layer was reformed depended on the degree of disturbance, which
was estimated by the ratio of the post-disturbance population to the pre-
disturbance population. Therefore, the original 3-cm oxygenated layer was
re-established according to the formula:
OXYGENATED LAYER (cm) = M + (Pd/P$)(K-M) (11)
where M = minimum depth of reoxygenated sediment (cm) regardless of degree
of disturbance (set at 1.0 cm for all model runs)
P, = population size after disturbance (note: this may be a partially
recovered population)
P = population size before disturbance
K = maximum depth (cm) of undisturbed oxygenated layer (set at 3.0 cm
for al1 model runs)
30
-------
Figure 14. Monthly survival rata of soft bottom
fauna as a function of aedlment depth.
0.75-
01
a:
0= 0.50-
cc
r3
en
0.25-
100
150
50 100 150 200 250
SEDIMENT DEPTH (cm
300
350
Figure 15. Monthly survival rate of Infauna and
epufauna as a function of the toxlclty of the
fraction whole mud In the sediment.
1 -i
0.75-
£-
or
a:
ac 0.50-
cc:
ID
CD
0.25-
0.00 0.02 0.04 0.06 0.08 0.10
FRflCTION WHOLE MUD IN SEDIMENT
31
-------
Figure 16. Monthly survival rates of soft bottom
fauna as a function of sediment removal by storms
1 -i
0.75-
LJ
t-i
a:
en
en 0.50-
cn
0.25-
l
SEDIMENT REMOVED (cm!
Figure 17. Fraction of soft bottom fauna
exhibiting sublethal response as a function of
fraction whole mud In sediment.
1 -i
a
LJ
E
O
UJ
L_
Q_
or
or
-z.
n
or
o
z
o
u
01
0.75-
0.50-
0.25-
0.0
10.0
12.0
FRF1CTION WHOLE MUD IN SEDIMENT *10~
32
-------
Figure 18. First month recolonlzatlon response of
microbes to change In fraction whole mud.
2.5-1
0.000 Q.OQ2 0.004
FRflCTION
0.006 0.008
WHOLE MUD
0.010
Figure 19. First month recolonlzatlon response of
meuofauna to change In fraction whole mud.
en
en
x:
o
ii
CQ
_J
en
^:
Ii
CD
ii
a:
o
u.
o
o
en
or
L..
2.2-1
1.8-
1.6-
1.2-
0.00 0.02 0.04 0.06 0.08
FRflCTION WHOLE MUD
0.10
33
-------
The potential increment of recovery of the infaunal and epifauna! compon-
ents of the macrofaunal group was decremented by residual toxicity of the
sediments as represented by the fraction whole mud (Fig. 20).
Limitations
Comments by Donald Boesch, a workshop participant, on the approaches
taken in the Soft Bottom Submodel are attached as an appendix.
One major deficiency in the submodel was that there was no interdependency
between the fauna! groups. While such interactions obviously exist, the
relative importance of their omission on the qualitative behavior of the
submodel was unknown. A second major problem was the necessity for using
short-term toxicity information to predict effects of longer-term chronic
exposure. This probably resulted in an overestimate of survival of soft
bottom organisms. Finally, there were no vertebrates included in the submodel
because of a lack of information.
Use of burial survival rates based on experience in the Great Lakes
probably represents extreme tolerance to burial. While this is not inconsis-
tent with expected behavior in a storm dominated system, it indicates how
model behavior would be altered by using different assumptions of community
composition. There was some evidence that population recovery times may be as
much as six times longer than those currently incorporated in the model
Figure 20. Fraction of first month potential
recolonlzatlon Increment realized due to toxicity
of residual drilling muds.
l.l-i
0.5
0.000 0.005 0.010 0-015
FRflCTION WHOLE MUD
0.020
34
-------
(Fredette 1980; Tagatz et al. 1980; Boesch and Rosenberg 1981; Shaffner et al.
1981). Although there was insufficient time at the workshop to investigate
the effects of this assumption on model behavior, examination of variation in
natural community recovery rates and the factors influencing that variation is
an important area of further investigation in predicting effects of drilling
muds and cuttings discharges on these communities which was not fully addressed
due to time constraints at the workshop.
HARD BOTTOM EFFECTS SUBMODEL
Responsibi1i ties
The Hard Bottom Effects subgroup was responsible for representing the
potential impacts of various discharge patterns of drilling muds and cuttings
on the dynamics of a "typical" hard bottom community in the Gulf of Mexico.
Basic information available from other subgroups included sediment depth,
concentrations of various constituents in the sediments, and concentrations of
drilling muds in the water column (both solid and soluble fractions). The
task of the Hard Bottom Effects subgroup was to formulate mathematical expres-
sions describing how hard bottom organisms might respond to these discharges
as reflected in indicators such as biomass, growth rate, mortality rate, and
recruitment.
Structure
In an attempt to simplify the task into something manageable in the time
available, the subgroup made several initial assumptions.
(1) Whi1e-other organisms (e.g., sponges or gorgonians) may well dominate
a typical hard bottom community, corals were used as an indicator.
This decision was necessitated by the lack of data on the toxicity
of drilling muds to other hard bottom organisms.
(2) Corals were considered in a nonreef situation to reduce complications
caused by the dynamics of a plume striking an irregularity in the
ocean bottom.
(3) Coral dynamics were represented in biomass units of grams carbon per
square meter (gC/m2).
(4) Coral biomass was represented only at discrete distances (0, 50,
100, 500, and 1500 m) down current from a drilling rig.
(5) The drilling rig was located on the hard bottom community.
(6) Uncertainties were, insofar as possible, resolved in favor of a
worst case assumption.
35
-------
Initial suogroup discussion highlighted four major potential impacts of
aril ling fluids on nonreef corals: direct mortality due to sediment deposition;
direct mortality due to plume toxicity; reduced growth due to plume toxicity;
and reduced recruitment due to burial of appropriate substrate by sediments.
Several other possible mechanisms were discussed at length and, for trie pur-
poses of the modeling exercise, ignored on the basis of having lower potential
for significant effects than the four listed above. For example, there was
considerable discussion concerning the possibility that light attenuation by a
discharge plume passing over, but not in contact with, corals would signifi-
cantly reduce photosynthetic activity. Such a mechanism was eventually
discarded on the basis that plumes would simply not be present for a signifi-
cant fraction of the daylight hours, and that photosynthesis recovers rapidly
following periods of reduced light. Possible growth rate reductions due to
temperature variations were ignored for similar reasons. In addition, larval
mortality due to plume toxicity was discussed as a factor having potential for
reducing recruitment of new corals. In the context of the spatial and temporal
scales of the model, however, this factor was judged to be relatively insignif-
icant for organisms (such as coral) with planktonic larval forms, since the
moving water mass would likely replace the larval community in a matter of
hours. The significance of this factor for organisms having nonplanktonic
larval forms may deserve further attention.
Biomass dynamics of coral were thus conceptualized in the framework of
the following equation:
Ct+1 = Ct - S*Ct - P*Ct + G + R (II)
where C = coral bicmass (gC/M2)
S = mortality due to burial (%)
P = mortality due to toxicity of plume (%)
G = growth (gC/m2)
R = recruitment (gC/m2)
The following sections discuss model formulations for each of these
factors.
Sediment deposition. Sediment depths at each of the five distances
downcurrent from the simulated rig were calculated by the Discharge/Fate
submodel. Corals were assumed to be uniformly covered with sediment of those
depths, and resulting survival reductions were computed using the relationship
shown in Figure 21. Data values for Figure 21 were extrapolated by subgroup
36
-------
Figure 21. Coral monthly survival rate as
function of depth of muds and cuttings.
0.75-
LU
£-1
ac
cc:
en 0.50-
(J-)
0.25-
3.5
DEPTH (cm)
members from information given by Thompson (1980). Lack, of information pre-
vented consideration of other aspects of sediment deposition, such as growth
rate reduction and recovery following incomplete burial, and effects of
repeated intermittent burials followed by flushing or clearing.
Plume toxicity. Coral survival was further reduced due to toxicity of
the plume. Maximum concentrations reached at each of the five locations
downcurrent during any single discharge were generated by the Discharge/Fate
submodel. These maximum concentrations were modified by a multiplier (nomin-
ally set at 0.5) designed to reduce the maxima to average concentrations to
which corals might be exposed over the course of a discharge.
Survival rates were calculated for these average concentrations using
duration of discharge, number of discharges per month (both supplied by the
Discharge/Fate submodel), and unpublished toxicity data contributed by Eric
Powell (see ACKNOWLEDGEMENTS section). Powell found that Acropora cervicorm's
suffered no mortality and no obvious zooxanthel1ae loss during a24-hour
exposure to 100 ppm whole drilling mud, and total zooxanthel1ae loss after a
24-hour exposure to 500 ppm. It was assumed, based on visual and biochemical
data, that the corals exposed to 500 ppm drilling mud were dying and would
have suffered 100% mortality. These experiments used a Mobile Bay drilling
mud judged by Conxlin et al. (1980) to be more toxic than most to PIaemonetes
pug 10. An LC of 300 ppm was therefore arbitrarily assumed for purposes of
the workshop model. Concentrations likely to produce 0, 50, and 100% mortality
for discharges for durations other than 24 hours were calculated using equa-
tions of the following form (after Petrazzuolo 1981):
37
-------
3-hr LC
50
= 24-hr LC,Q
(24/3)^
The results of these calculations for a 3-hour discharge are depicted as
survival rates in Figure 22. A new curve was calculated for each simulation
using the duration of discharge provided by the Discharge/Fate submodel
Survival rate for a particular average concentration was then interpolated
from the curve and applied repetitively for as many discharges as occurred
duri ng the month.
Growth. Growth of the coral remaining after mortality due to sediment
deposition and plume effects was treated using a density-dependent potential
growth rate and a proportion of the potential growth rate realized due to
plume concentrations. Formulation of the growth rate as a density-dependent
function prevented unlimited exponential growth of corals in the model.
Figure 22. Coral monthly survival rate as a
funcllon of solids concentration for a 3~hr
exposure.
500 1000
CONCENTRATION (pprn)
1500
38
-------
The density-dependent potential growth rate (Fig. 23) was derived in the
following manner. An estimate of the biomass of the coral Montastrea annularis
in gC/m2 of tissue was obtained from unpublished data contributed by Alina
Froelich (see ACKNOWLEDGEMENTS section). She found an average of about 65 ug
atoms N/cm2 of tissue. Assuming a carbon/nitrogen ratio of approximately 7,
and adjusting for the molecular weight of carbon, this translates to about
54.6 gC/m2 of tissue. An annual linear growth rate of 5 cm was assumed and,
using a hemisphere as an approximation of the growth form of this coral,
annual increases in surface area were computed for corals ranging from 5 to
50 cm radius (Table 6). Increases in surface area were converted to gC added
annually by multiplying by 54.6, and expressed as a percentage of the biomass
present at the start of the year. The resulting values, plotted as a function
of biomass present, are shown in Figure 23. Monthly growth rates were obtained
simply by dividing values interpolated from Figure 23 by 12.
Figure 23. Potential coral annual growth rale
Biomass In gC/m2.
600-
50 100
BIOMPiSS PRESENT
150
39
-------
Table 6. Derivation of a density-dependent coral growth
rate, assuming a hemispheric growth form.
Radius of
hemisphere (cm)
Surface area(m2)
3i omass(gC/'m2
Start of
year
0
5
10
20
40
50
>50
End of
year
Start of
year
End of
year
Start of
year
Enc o*
year
Biomass
added
(gC/nr)
O2
10
15
25
45
55
0
0.
0
1
1
.016
.063
.251
005
.571
0.
0,
0.
1
1
.063
, 141
.393
.272
.901
0
3.
13.
54
85
150
.86
.43
.72
.37
.78
.OO2
2
7
21.
59
103.
.43"
72
.44
. ^5
.78
2.
4
7
14
1 *"*
lo
,57
29
72
.58
.00
Growth
rate
/0/\
(°)
5002
300
125
56
27
21
O2
'Assuming annual growth of 5 cm.
2Arbitrarily assigned value.
Potential monthly growth rates were treated as maxima and reduced accord-
ing to drilling fluids concentrations produced by the Discharge/Fate suomodel .
A concentration/growth response curve was derived from unouolisned data con-
tributed by Eric Powell for 24-hour exposures of Acroccra ce^v: ccrni s to
various concentrations of whole drilling mud (Fig. 24) The mud and corals
used were the same as those mentioned earlier in the discussion of mortality
data. Growth rate reductions for exposures of durations otner than 24-nour
were simply calculated as proportions of the 24-hour reduction; that is, a
12-hour exposure to a given concentration resulted in half the growth rate
reduction caused by a 24-hour exposure. Multiple exposures 'in any month
resulted in continued reduction of the growth rate. Recovery of the growth
rate was allowed only in months without discharge. In the first such month,
recovery halfway to the maximum was allowed. A second consecutive month
without discharge resulted in complete restoration of the maximum growth rate.
These assumptions concerning growth rate recovery and reductions in growth
rate for exposure durations less than 24 hours were arbitrary, there being
little or no information of this kind available for corals.
Recruitment. Recruitment of new coral was allowed only at times and
locations where: (a) no larger coral was present; and (b) sediment depth was
zero. This aspect of coral biomass dynamics was included only
the potential for recovery following episodic events (such
generated by the Discharge/Fate submodel) that
substrate suitable for establishment of corals
rates were unavailable for this situation. Spat
designated as 0.05 gC/m2 for locations meeting
and reduced by the percent reduction in benthic
Water Column Effects submodel.
to illustrate
events suc as the storms
remove sediment and expose
. Data on larval settlement
set was therefore arbitrarily
the conditions listed above,
recruitment calculated in the
40
-------
Figure 24. Reduction In coral growth rate
as a function of solids concentration for a
24~hr exposure.
120 -i r
100-
E-i
O
u
a:
UJ
cc
in
t-i
2
Q
80-
Legend
EXPERIMENTS.
EXTRflPOLflTED
600
800
1000
1200
CONCENTRflTION (pprn)
41
-------
SYSTEM MODEL
STRUCTURE
For each monthly iteration, the Discharge/Fate submodel calculated upper
plume characteristics and deposition of drilling muds and cuttings. The Water
Column "Effects submodel next calculated impacts of the upper plume on zooplank-
ton and benthic recruitment. Information from these two submodels was then
used by the Soft Bottom and Hard Bottom Effects submodels to calculate
potential benthic impacts.
BEHAVIOR
In the following section we present sample output generated with the
workshop model. The output is organized into four scenarios which differ in
water depth and discharge rate. The baseline scenario represents a production
platform in 80 m of water, sequentially drilling a total of 20 wells at 6 weeks
per well, with bulk discharges of 600 bbl every 3 days at a rate of 1,000
bbl/hr, and a total discharge of 1,500 metric tons of solids per well. Each
model run represents 20 years with drilling initiated halfway through the
first year and ending in year 3. Results from this scenario are presented in
some detail to establish baseline conditions. Discussion of subsequent
scenarios focuses on those variables that show large differences from the
baseline scenario.
The scenario results are presented in terms of absolute quantities (depth
of added sediment, coral density). In so doing, we run the risk of inputing
greater accuracy to this initial model than is justified. We present the
results in this form not because we necessarily believe them to be entirely
accurate, but rather in the hope of promoting constructive discussion. Models
cannot be validated; like hypotheses, they can only be invalidated. Only by
subjecting the model and its results to criticism can we establish the limits
of its credibility. In comparing scenarios, it should therefore be remembered
that qualitative changes and general trends probably have greater meaning than
actual numbers. The numbers are included only as points of reference and
discussion.
Scenario I
Under baseline drilling and discharge conditions, drilling muds and
cuttings built up to a maximum depth of 15 cm over a circular area of radius
154 m (Fig. 25). This 'added sediment was completely dispersed by periodic
storms 6 years after drilling stopped. The fraction whole mud in the sediment
at various distances from the platform showed a similar temporal pattern
(Fig. 26). At 50 m from the platform, the maximum fraction whole muds was
42
-------
Figure 25. Scenario Is depth of deposited spent
mua and cuttings at three distances from platform,
LOCflTION
TIME (years)
figure 26. Scenario I: fraction whole mud at three
-distances from platform.
UJ
o
0
II
u
0.20n
0.15-
0.10-
0.05-
0.00
LOCRTION
50 m
IQQj* _ _
500 m
5 10 15
TIME (years)
20
43
-------
0.12, which also decreased to zero 6 years after drilling stopped. Figure 27
shows the concentration of fine-grained participates in the upper plume at 50,
100, and 500 m from the platform.
With high rate of discharge and relatively deep water, coral is not
subjected to toxic materials in the soluble fraction of the upper plume.
However, all coral at 50 and 100 m was smothered by cuttings and spent mud
and they had not fully recovered by the end of the 20-year model run (Fig. 23).
At 50 and 100 m, microbes and meiofauna showed very little response because
the stimulation to the population from deposition of new substrate was only
slightly overridden by the toxicity of the deposited materials (Fig. 29).
Macro-infauna showed severe reductions in their populations during the period
of drilling and continued population oscillations until-all of the deposited
cuttings and spent muds had been removed by storm action (Fig 30). Epifaunal
tissue concentrations of chromium, above background, were less than 2 ppm at
50 and 100 m, and 0 ppm -at 500 m (Fig 31). Change in recruitment to the soft
bottom communities was insignificant as 99.98% of the organisms survived.
Figure 27. Scenario I: concentration of fine
grained partlculates In the upper plume at
three distances from platform.
bU -
£
D_
CL
40-
Z
O
1 I
t 1
cc
or
t-l
IF
LJ 20-
o
CD
t
]
" ^"^
LQCflTION
50 m
IQOjD _ _
500 m
III'
5 10 15 20
TIME (years)
44
-------
Figure 28. Scenario I: coral blomass at three
distances from platform.
ISO-,
J= 100-
o
en
en
en
en
z:
O 50-
»«
CD
LOCflTION
50 n FIND 100
500 a
10
T~
15
20
TIME (years)
Figure 29. Scenario ]: mlcroblal blomass at two
distances from platform.
~ 150
M£
O
CD
cn
en
or
100-
50-
LOCnTION
50 n>
500
~T~
10
r~
15
TIME (years)
"~i
20
-------
Figure 30. Scenario I : macro~lnfaunal bLomass at
two distances from platform.
0.15-1
0.10-
o
en
en
tn
or
2Z
O
*1
in
0.05-
Q.OO
TIME (years)
Figure 31. Scenario I: eplfaunal tissue
concentrations of chromium (above background.
three distances from platform.
at
£
O_
a.
fr-
ee
cc:
E-t
-Zi
LJ
U
~ZL
0
O
1.5-
1 -
0.5-
]
I
-------
Scenario II
In the second scenario, the rate of bulk discharge was decreased from
1,000 to 30 bbl/hr. All other drilling and discharge characteristics remained
the same. Decreasing the discharge rate affected concentrations in the upper
plume but had no affect on the lower plume. Therefore, added sediment and
fraction whole mud were the same as the baseline scenario. The lower discharge
rate resulted in concentrations of fine-grained particulates at 50, 100, and
500 m from the platform that were approximately 1/3 of baseline levels.
With the exception of survival rate in the water column (essentially
100%), the biological response was identical to that seen in Scenario I. This
was due to the time step selected for the model runs. For example, despite
the fact that the discharge rate was much lower, the total amount of material
discharged during a month was the same.
Scenario III
The third scenario had the same discharge characteristics as the base-
line, but it was assumed that drilling occurred in only 20 m of water. Upper
plume characteristics were unchanged from the baseline scenario because
discharge characteristics were identical. The shallower water depth resulted
in greater maximum sediment build up (34 cm) over a much smaller area (33-m
radius). The fraction whole muds was therefore higher than in the baseline
scenario and dispersed much slower (Fig. 32).
Corals had a very different response than in the baseline scenario,
Scenario I, (Fig. 33) with the dominant effect in this scenario due to the
toxicity of the solids fraction of the upper plume rather than burial. There-
fore, colonization can begin as soon as drilling is completed without having
to wait for sediment removal from the substrate. This resulted in total
recovery of the coral after about 16 years. The reduction in organisms avail-
able for recruitment to the soft and hard bottom communities was somewhat
greater but still relatively insignificant (99.81% survival). There were no
effects on the soft bottom community at any of the distances chosen for display
because there was no sediment buildup. Note that the soft bottom submodel did
not respond to the toxicity of sediments (i.e., fraction whole mud) in the
absence of a change in sediment depth.
Scenario IV
Scenario 4 assumed a 20-m water depth and a 30-bbl/hr discharge rate.
Upper plume characteristics were the same as Scenario II (also 30 bbl/hr)
while added sediment characteristics were the same as Scenario III (also 20-m
depth).
Coral response at 50 and 500 m was identical to that of Scenario III.
The difference (at 100 m) b'etween these two sets of discharge conditions is
that the lower discharge rate allowed sufficient dispersion of the toxic
portion of the plume so that there was no coral mortality at 100 m. The rest
of the biological behavior was the same as that of Scenario III.
47
-------
figure 32. Scenario III: fraction whole mud at
three distances from platform.
UJ
_j
o
in
12:
z:
o
0.20-1
0.15-
0.10-
O
Cc! 0.05-
0.00
i
10
r
15
20
TIME (years)
figure 33. Scenario III: coral blamass at three
distances from platform.
ISQ-i
100-
o
en
CO
CO
a:
r:
o
it
CD
50-
i
10
15
20
TIME (years)
48
-------
WORKSHOP RESULTS
COMMUNICATION
The workshop was effective in providing a forum for communication among
the participants on the somewhat controversial topic of, the fate and effects
of discharged drilling muds and cuttings. In large part this was due 'to the
interest, expertise, and openness of individual participants. It resulted in
broadened individual perspectives of the issue, exchange of data and insight,
and plans for future cooperative activity. These aspects are difficult to
document for any workshop; however, they are extremely valuable to the extent
that the participants represent a community that will continue to be involved
with the issue of marine discharge of drilling muds and cuttings.
Construction of a simulation model focused discussion on a number of
critical areas. Some of the most useful discussions concerned composition of
discharged materials and linkages between the processes influencing fates and
the processes determining effects. Several examples of these discussions are
presented below.
One subgroup concentrated on identification of factors that might produce
differences in the fate and effects of drilling muds and cuttings discharged
into more "closed" bodies of water, such as bays and estuaries. Results of
these discussions are highlighted in this section on communication and also
formed much of the basis of a later section summarizing general factors deter-
mining fate and effects of marine drilling discharges.
Composi ti on
Uncertainty about the composition of discharged drilling muds and cuttings
has complicated analysis of their fate and effects in the marine environment.
However, they are not unknown substances. The vast majority (by weight) of
the material is relatively inert and only a small fraction of the many com-
pounds available as additives are actually used at a given site. It is also
possible to identify muds that are characteristic of a mud type representing
the probable combination of materials that would be used in a majority of
similar sites.
Discussion centered around the extent to which it was possible to define
the composition of the material as it is discharged. The two initial sides to
thi s question were:
49
-------
(1) Drilling muds are closely controlled mixtures tailored to meet
specific performance criteria. Materials used at a particular
location can be exactly specified, and in fact are specified in the
drilling log. The substrate that produces the cuttings can be
defined. The composition of the material being discharged is, in
principle, absolutely predictable and is, in fact, measurable with
respect to elemental composition.
(2) There is much variation in materials added and the composition of
cuttings at different locations and over time and depth at one
location. This uncertainty is aggravated by the complexity of
possible reactions among components and in the breakdown of compo-
nents, variations in temperature and pressure within the "reaction
vessel" (drilling apparatus and mud circulation system), and the
possibility that the mixture is not at equilibrium. In combination,
these factors make it practically impossible to specify the composi-
tion of material as it is discharged at the level of chemical resolu-
tion appropriate for investigation of chemical toxicity
There was some resolution of this question through the perspective of
drilling muds and cuttings as a dynamic chemical system. There was then
better acceptance of the levels at which this system could be specified and
the levels at which uncertainty exists. It was possible to phase meaningful
statements about the muds and cuttings system from a toxicity standpoint. One
example was the statement that the bulk of the toxic materials seemed to
settle out in a relatively unavailable form, bound to clays and fine sediments;
whereas a large part of the toxicity of the discharge seemed to be associated
with materials in the more available soluble phase. The actual availability
and toxicity of particulate and bound materials in "relatively unavailable
forms" remains uncertain, especially with respect to long-term behavior in the
bottom sediments. The question of composition was resolved in the model
itself by specifying a 13-bbl/gal mud with toxicological properties expressed
in terms of the ppm or fraction of this whole mud present. Considerable
concern remained, however, over how various environmental fractions of the
discharge, such as solubles in the upper plume, corresponded to various
fractions utilized in laboratory experiments.
Fate and Effects
Expected exposure levels. The modeling effort provided a logical struc-
ture for discussing expected concentrations over time at various distances
from the rig. This discussion and the results of simulation runs were
effective in indicating to biologists involved with toxicity testing the
approximate levels of environmental concentration that might be expected in
the field.
Toxicity evaluation. A considerable amount of discussion occurred among
the group as a whole and wi'thin subgroups on the relationship between the
results of defined toxicity tests such as a 96-hr LC 30 and the effects of
time-varying field concentrations on individuals and populations. The problem
was basically how to convert results from fixed length and concentration to
exposures of variable concentration over much shorter and longer times.
50
-------
Suggestions included using the integral of concentration over time or a time-
averaged concentration with an algorithm to account for differing lengths of
exposure. The types of functional relationships utilized, in fact, differed
somewhat among submodels, reflecting both uncertainty about the correct form
and perhaps organismal differences in the relationship.
Short of extremely complex and expensive toxicity tests, there seemed to
be no highly accurate way of connecting predictions of variable field concen-
trations to results of laboratory toxicity tests. Utilization of laboratory
toxicity results in the workshop model was more in the mode of indicating
where toxicity problems might be encountered, rather than quantitative accuracy
in prediction of effects.
Worst case sediment deposition. The Discharge/Fate submodel required
that some assumptions be made concerning patterns of sediment deposition.
There was some uncertainty about what constituted a "worst case" assumption
about the pattern of sediment deposition from the platform. A given quantity
of mud solids and cuttings deposited in a deep layer over a small area would
kill a high proportion of the benthic organisms in that area, whereas a
shallow layer over a larger area would kill a smaller proportion of the
benthic organisms in a larger area. The "worst case" pattern or maximum
number of benthic organisms killed thus depends on organisms' responses to
sediment depostion. This response most likely has a strong threshold component
with organisms able to survive a certain depth of burial depending on the
natural sedimentation regime to which they are adapted. The workshop did not
resolve a clear "worst case" pattern and, in fact, the "worst case" is likely
to be species-specific, since it is critically dependent on organisms' ability
to to!erate burial.
Shunting. Shunting, or discharging at some greater depth than the surface
(e.g., below the pycnocline), is considered as a management action to produce
the following results:
(a) reduce the visible plume;
(b) entrap discharge in nephloid layer minimizing impacts above discharge
depth;
(c) avoid a potential buildup of material as the discharge encounters a
diffusion barrier at the pycnocline; and
(d) minimize the area of deposition for material settling out (i.e., mud
sol ids and cutti ngs).
There were questions raised about some of the benefits of shunting,
despite its value in routing the plume away from features that rise above the
bottom. As noted above, it was not clear that minimizing the area of sediment
deposition minimized the total impact on benthic organisms. There was also
uncertainty about the behavior and importance of the plume encountering the
pycnocline. It is also possible that shunted soluble material might rise
above the depth of discharge, possibly encountering the pycnocline, as the
upper plume moved to a level of neutral bouyancy.
51
-------
"Closed" water bodies. Fate and effects of discharged drilling muds and
cuttings in closed water bodies such as bays and estuaries was felt to be
considerably different and more complex than that in more "open" water envi-
ronments such as those treated in the simulation model Many of the critical
variables producing these differences have been incorporated in the following
section (Factors Determining Fate and Effects). Conceptual models of fate and
effects in these areas would contain components similar to those utilized in
the open water simulation model with several modifications.
(1) Additional communities, such as aquatic macrophytes and oyster beds,
would have to be treated.
(2) The importance of "closed" water bodies as...food production and
rearing areas would necessitate more detailed incorporation of
population-level processes and trophic interactions.
(3) Many of the processes represented in the open water simulation
model, such as plume dynamics, sediment deoosition, and sediment
redistribution, would require fundamentally different mathematical
treatment due to shallower water and more complex circulation and
stratification patterns.
(4) The importance of resuspension in shallower water and slower long-
term dispersion would necessitate more detailed consideration of
long-term effects of slightly elevated concentrations.
In addition to these considerations complicating the extension of open
water analyses to closed water environments, participants felt that a general
analysis or model was less appropriate for closed water environments because
of the large amount of variability among these areas in factors strongly
influencing fate and effects (such as circulation and salinity patterns,
community composition, and natural sedimentation regimes).
Much of the possible difference in behavior centers 'around the extent to
which these areas are "closed" or the relative residence time or amount of net
exchange in water and sediment between these areas and surrounding areas.
This is a critical factor in determining long term dispersion of discharged
material. It was suggested that indices expresssing residence or turnover
time of water and material in the surficial sediments might be useful in eval-
uating differences in fate and effects in "closed" areas, and that such indices
might be calculated from information on freshwater inputs, circulation pattern,
volume of the basin, and natural sediment loadings.
Although enclosed areas were considered more complex and variable than
those treated in the model, a large base of knowledge and understanding does
exist for many well-studied bays and estuaries. Information and models
developed by the U.S. Army Corps of Engineers with respect to fate and effects
of dredge spoil disposal were identified as being particularly relevant to
discharge of drilling muds and cuttings.
52
-------
INFORMATION INTEGRATION
A simulation model is a structure for representing the net result of a
series of statements about how the system operates. A number of assumptions
are often necessary to integrate more well-established individual relation-
ships and linkages. Some set of assumptions (often unstated and relatively
crude) is used in any integrated statement or management criteria on the fate
and effects of drilling muds and cuttings into the marine environment. The
value of a simulation model is that it forces an explicit statement of what
assumptions are being used.
The simulation model developed at the workshop' for open water environments
in the Gulf of Mexico indicated relatively localized effects of drilling muds
and cuttings discharge (see SYSTEM MODEL section). Water column fate and
effects were dominated by relatively rapid dilution. Deposition of spent mud
solids and cuttings was. localized spatially with relatively rapid recovery
especially in soft bottom benthic communities.
This is the behavior generated by the set of assumptions about linkages
and functional relationships used to construct the model. There are two
general ways in which such a model can be inadequate. The first is that
linkages and processes included in the model may have been poorly represented.
Areas of uncertainty in the workshop model included the relationship between
time-varying exposures and 96-hr LC so results, recovery rates of benthic
communities, and responses of organisms to various depths and rates of burial.
The second area is that important aspects of the system may not have been
included in the model. Many potential linkages and processes are excluded
from a simulation model because they are judged to be of secondary importance,
such as the effect of light attenuation from the plume passing over corals on
annual coral growth. Others are excluded because they are unknown or not
currently tractable within the modeling constraints. They could very well be
critical in the behavior of the real world system. Some of the interactions
and processes not incoporated in the model included density stratification and
possible dispersion barriers it might create, long-term effects of slightly
elevated concentrations, potential food chain transfers, and the interactions
that might occur among discharges from multiple platforms. Some of these
limitations could be partially addressed through model refinements. Some,
however, reflect lack of current understanding rather than lack of ability to
integrate existing information.
53
-------
INFORMATION GAPS
A number of information gaps were identified at the workshop in the
process of constructing the simulation model and in discussing factors deter-
mining fate and effects in enclosed areas. These represent areas of uncer-
tainty where additional information would be desirable. This does not neces-
sarily mean that no work has been done in these areas. It may merely indicate
that participants were not fully aware of the relevance of completed work or
that additional analysis needs to be undertaken to interpret that; work more
fully in terms of its relevance to fate and effects of discharged drilling
muds and cutti ngs .
These information gaps are detailed throughout the report in the descrip-
tions of the simulation model and discussions of enclosed areas, such as bays
and estuaries. The following list is a summary of the more important of these
areas of uncertainty identified at the workshop:
(1) The extension of 96-hr LC -Q results (or any fixed-concentration,
fixed-interval toxicity test) to other exposure times at other,
perhaps time-varying, concentrations was a central problem in esti-
mating effects on field populations from predictions of environmental
fate. The relatively simple algorithms utilized for this extension
involve considerable extrapolation and interpolation from observed
cases.
(2) The relationship between variation in composition of discharged
drilling fluids and cuttings (variation in additives, different
sites, and across time and depth at one site) and variation in
toxicity does not seem to be well-established. Current research
(Thomas Duke, ACKNOWLEDGEMENTS section) is addressing this question
through a series of standardized tests on a large set of drilling
fluid samples.
(3) There seems to have been little explicit consideration of indirect
or community-level effects (such as accumulation of materials through
food chains, indirect effects on a secondary species through direct
effects on a competing, predator, or food-source species). Detailed
prediction of effects at this level may, in fact, be beyond the
state of the art with respect to analysis methods and knowledge of
the relevant marine systems. It may be possible, however, to
strengthen the value of toxicity tests on individual species and
life history stages by more consideration of the position and
importance of these species in the communities of which they are a
part. One example in terms of life history stages is the possible
importance of effects on benthic larval stages of benthic organisms
54
-------
on these populations, which may be more severe than the generally
very small effects on recruitment due to effects on planktonic
larval stages.
(4) Variations in the rate of recovery of disturbed benthic communities,
sensitivity of these communities and their recovery rates to altered
particle size distributions, and sensitivity to depth and rate of
burial are all areas where additional quantitative information was
needed in model construction. These areas are amenable to experi-
mental investigation and it may be possible to make considerable
progress through synthesis of existing information. Recovery rates
for corals after exposure to drilling fluids are now being inves-
tigated (Eric Powell, ACKNOWLEDGEMENTS section).
(5) Little information was available on long-term effects of slightly
elevated concentrations and sub-lethal effects (such as growth rate
depression) in general.
(6) Information on hard bottom community effects seemed to be concen-
trated on several species of coral. A broader set of species and
hopefully community-level indicators would be especially desirable
for these areas.
(7) There was considerable uncertainty about behavior of the plume at
water stratification layers and possible effects of a potential
higher concentration of discharged materials in areas where organisms
might also tend to be concentrated.
(8) The interaction among discharges from multiple platforms is not
explicitly treated by current plume models, including the workshop
simul atiorr model This interaction, if important, would require a
much more complex mathematical treatment to address integrated or
cumulative effects in densely utilized lease area.
(9) A resolution of the relative advantages of shunting at different
depths'would be very useful from a management perspective. Questions
were raised at the workshop about the benefits of some of these
alternatives. Clear resolution will depend on better understanding
of the movement of the upper plume from various density points
(including efficiency of entrapment in nephloid layer) and as it
encounters the pycnocline, effects at the pycnocline, and the optimum
pattern of sediment deposition. The optimum patter of deposition
may be dependent on avoiding impacts on features rising above the
surface, such as coral reefs, as well as minimizing impacts on
benthic communi ties.
(10) Finally, there seemed to be a major need to synthesize information
concerning fate and effects to be expectad in enclosed areas. A
number of factors limit the applicability of open water results to
these areas. The potential sensitivity of these areas argues for
more detailed consideration of fate and effects. Several factors
were identified that could support such an effort. A number of bays
55
-------
and estuaries have been extensively studied. Many of the toxicity
tests have, in fact, been conducted on estuarine organisms. In
addition, models and a relatively large body of data are available
on the fate and effects of dredge spoil in enclosed water bodies
which should have considerable relevance to fate and effects of
dri11 ing discharges.
FACTORS DETERMINING FATE AND EFFECTS
Discharge of drilling muds and cuttings into a marine ecosystem is a
perturbation of that system. A number of factors interact to determine the
fate and effects of any particular drilling mud and cuttings discharge, and
thus need to be considered as a whole in a scientific evaluation of the
system's response to the perturbation and in management decisions concerning
an acceptable level of perturbation.
The workshop addressed identification of these variables and their inter-
actions through two complementary activities. The first approach was construc-
tion of a simulation model of the fate and effects of drilling muds and cuttings
discharged into several types of open water environments in the Gulf of Mexico.
This activity identified a set of important variables and their interactions
for each environment. Discussions were also held to identify features of
"closed" water environments, such as bays and estuaries, which would need to
be considered in evaluating fate and effects in these areas.
The factors identified at the workshop are discussed below in terms of
three broad categories; discharge characteristics, physical/chemical character-
istics, and biota. The list represents a guideline of variables that need to
be considered in evaluating and/or regulating the discharge of drilling muds
and cuttings at any particular site. The list is an attempt to synthesize
discussions of the workshop participants as to what should be considered. It
is not intended to substitute for detailed synthesis of the scientific litera-
ture as it relates to these variables, nor does it imply that all variables
need to be given equally detailed consideration in all management decisions
concerning discharge of drilling muds and cuttings.
Pi scharge
Composition. The drilling muds and cuttings discharge is itself a complex
and dynamic chemical system varying across different drilling locations and
over time and drilling depth at a particular location. Mud components are
adjusted to meet local conditions as they occur. Composition can be defined
in terms of materials added and in terms of elements and major compounds for
the actual discharged mixture. Precise composition and activity of discharged
material (in terms of the exact chemical compounds and chemical associations
resulting from breakdown of added components, reactions occurring in the well
at elevated temperatures and pressures, and complexing and sorption processes)
are more elusive.
56
-------
Aspects of drilling muds and cuttings composition that most directly
determine differences in fate and effects following discharge into the marine
environment are density, particle size distribution, and toxicity. Density
and particle size distribution are important determinants of the transport of
various fractions of the discharge. Particle size distribution of deposited
material in relation to the particle size distribution of existing sediments
can influence the recovery time and composition of benthic communities.
Generalization about the toxicity of drilling muds and cuttings discharges
is difficult due to their variability and complexity. Approaches have included
toxicity measurements, such as the 96-hr LC5Q, utilizing "typical" whole mud
samples or fractions of such samples, as well as' toxicity measurements of
individual additives, such as biocides. Although a laj:ge proportion of the
material (by weight) is relatively inert, little progress has been made in
multivariate approaches for isolating the compositional determinants of varia-
tions in toxicity. A series of reasonable "worst case" extrapolations from
defined toxicity measurements appears to be the only currently feasible
approach.
In addition to the relatively short-term, high-concentration toxicity
associated with the immediate discharge plume, possible long-term effects of
slightly elevated concentrations of stable constituents, such as heavy metals,
were raised as a point of concern at the workshop. These potential effects
were not incorporated into the simulation model primarily due to lack of
quantitative information. It was pointed out that they might be expected to
be more important in a "closed" water body such as a bay where long-term
dispersion of discharged material would be less rapid.
Delivery. The rate and amount of discharge are principle .parameters
determining the extent and dynamics of the discharge plume. Predilution of
the discharge was discussed as a management action that would ameliorate toxic
effects, especially in the immediate vicinity of the discharge point, by
reducing concentrations.
Location and configuration of the discharge port or ports is also an
important factor in determining discharge plume behavior. Discharge from a
series of ports could reduce maximum concentrations by distributing the dis-
charge over a wider area. The location of the discharge port in the water
column in relation to the total depth and stratification layers in the water
column can strongly affect the resulting discharge plume. Shunting, by locat-
ing the discharge port below a stratification layer, has been proposed to
avoid impacts to features above the discharge depth (e.g., coral reefs,
pychnocline) by entrapping the discharge in a deeper layer. Shunting should
also tend to localize the area of cuttings and mud solids deposition and
minimize aesthetic impact by reducing the visible plume.
Location of the discharge port close to the bottom sediments, as would be
unavoidable in a shallow-water environment, produces a fundamentally different
plume behavior. Unless overt action is taken to redirect the discharge, plume
dynamics in these situations involve a "rebound" component as the discharge
hits the bottom and require basically different mathematical treatments than
those utilized to represent the dynamics in deeper water situations.
57
-------
Physical/Chemical Environment
Salinity and temperature regimes. Salinity and temperature are important
factors for several reasons. Stratification of the water column affects plume
dynamics and resuspension from bottom sediments, which can be especially
important in shallow water areas. Salinity can significantly influence floccu-
lation of drilling fluids and solids with resulting effects on the proportions
of various components that remain suspended in the upper plume. In addition,
temperature and salinity are important determinants of the biota and its
sensitivity, especially in areas such as near-shore environments where there
are strong temperature and salinity gradients.
Depth. Water column depth and its relation to depth of the discharge
port is a parameter of the representation of plume behavior used in the work-
shop simulation model for open water environments. Discussion of how more
"closed" water environments might differ suggested that some qualitatively
different types of behavior would be expected in the shallower water columns
generally associated with such environments. Depth would be a very important
varaible in such systems through its influence on circulation within the
system, expected short-term dilution of the discharge, stratification of the
water column, and resuspension from sediments.
Water movement. Current velocity and direction are two of the primary
parameters governing short-term dilution and direction of discharge. Long-
term dispersion of the dissolved or suspended fraction and movement of
deposited sediments are also critically dependent on the intensity and pattern
of water movement. Turnover time or exchange rate for water in "closed" water
bodies was identified as an important factor distinguishing these environments.
Long-term dispersion of discharged materials would be reduced to the extent
that these bodies of water were "closed" Effects of wind and wave action on
resuspension of deposited material would also be expected to be higher in
these generally shallower areas.
Sedimentation regime. The nominal or natural Sedimentation regime is
another site-specific factor determining the effects of sediments introduced
by drilling solids discharge. Higher natural sedimentation rates result in a
relatively lower level of perturbation from additional sediment. Differences
in particle size distribution between drilling mud solids and cuttings and
naturally occurring sediments could increase the perturbation since particle
size distribution is an important determinant of benthic community composition.
Benthic communities might thus recolonize at a different rate and recover to
an altered state that could be maintained for as long as particle size distri-
butions remained different.
Frequency and severity of storms play a major role in redistributing
sediments. The long-term fate of sediments added to a particular area would
be influenced by these events much as natural sediments are. Drilling mud
solids and cuttings might thus be expected to accumulate in certain areas as a
result of bottom topography, water movement patterns and velocities, and storm
events. To the extent that these factors influence the movement of natural
58
-------
sediments in the same manner, there is reason to expect that this will result
in a net dilution of drilling materials with natural sediments in comparison
to the initial area of deposition.
Biota
Composition and sensitivity of the biota in a particular area determine,
in large part, effects of a given drilling fluids and solids discharge into
that area. Laboratory toxicity tests such as 96-hr LC50 experiments can
provide indicators of sensitivity, especially with respect to short-term
effects in the immediate discharge area. As discussed earlier, direct connec-
tion of this information to population level effects from various discharge
scenarios is complicated by temporal variation in actual field concentrations.
In addition to toxicity, sensitivity to burial mortality, growth reduction due
to sediment deposition, and recolonization rates of benthic communities are
important factors in assessing effects of a given discharge.
Discussions at the workshop indicated several areas of special concern
where significant, and possibly larger than expected, effects might occur.
These included oyster bed, coral reef, and submergent or emergent aquatic
macrophyte communities. Concern was also expressed about possible effects on
endangered species and critical life history stages. If sensitive species or
life stages of species concentrate in portions of the environment, such as the
pycnocline, where discharged material also tends to concentrate, it might lead
to greater effects than would be predicted based on assumptions of more even
exposure.
Little information was available at the workshop that quantitatively
addressed the potential long-term effects of relatively low environmental
concentrations that might result from drilling mud and cuttings discharge.
The possibility of indirect effects through trophic interactions was identified
in cases of a depression of primary production affecting higher trophic levels,
potential for bioaccumulation and transport of toxic materials by rooted
aquatic macrophytes, and possible transfer of introduced materials such as
heavy metals through a food web with resulting increase in effective dose for
certain species over what would be estimated based on general environmental
concentrations. It is unlikely that effects in these areas will ever be
completely predictable in the general case, due to the variety and complexity
of drilling mud and cuttings discharges and of the marine environments into
which they might be discharged. They thus represent a responsibility for
continued attention and monitoring especially in conjunction with discharge
operations in those areas in close proximity to sensitive and "important"
biological communities.
59
-------
LITERATURE CITED
Boesch, D. F. , and R. Rosenberg. 1981. Response to stress in marine benthic
communities. Pages 179-200 j_n G. W. Barrett and R. Rosenberg, eds.
Stress effects on natural ecosystems. John Wiley and Sons, NY.
Brandsma, M. G., L. R. Davis, R. C. Ayers, and T. C. Saver, Jr 1980. A
computer model to predict the short-term fate of drilling discharges in
the marine environment. Pages 5S8-610 j_n Research on environmental fate
and effects of drilling fluids and cuttings. Lake Buena Vista, Florida,
January 1980.
Cantalmo, F. R., M. E. Tagatz, and K. R. Rao. 1979. Effects of barite on
meiofauna in a flow-through experimental system. Mar. Environ. Res.
1:301-309.
Conxlin, P. J., D. G. Doughtie, and K. R. Rao. 1980. Effects of barite and
used drilling needs on crustaceaus with particular reference to the grass
shrimp Palaemonetes pugi o. Pages 912-943 in Research on environmental
fate and effects of drilling fluids and cuttings. Lake Buena Vista,
Florida. January 1980.
Fredette, T. J. 1980. Macrobenthic colonization of muddy and sandy substrates
in the York River, Virginia. M.A. Thesis, College of William and Mary.
62 pp.
Houghton, J. P., R. P. Britch, R. C. Miller, A. K. Runchal, and C. P. Falls.
1980. Drilling fluid dispersion studies at the lower Cook Inlet C.O.S.T.
well. Pages 285-308 J_n Research on environmental fate and effects of
drilling fluids and cutt"ings. Lake Buena Vista, Florida, January 1980.
Koh, R. C. Y., and Y C. Chang. 1973. Mathematical model for barged ocean
disposal wastes. EPA Grant No. 16070 FBY, EPA Pacific Northwest Environ-
mental Research Laboratory, EPA-660/2-73-029.
McCulloch, W. L., J. M. Neff, and R. S. Carr. 1980. Bioavai1abi 1 i ty of
selected metals from used offshore drilling muds to the clam Rangia
cuneata and the oyster Crossostrea gigas. Pages 964-983 _i_n Research on
environmental fate and effects of drilling fluids and cuttings. Lake
Buena Vista, Florida, January 1980.
Pennak, R. W. 1964. Collegiate Dictionary of Zoology. Ronald Press Co., NY.
583 pp.
60
-------
Petrazzuolo, G. 1981. Preliminary report: An environmental assessment of
drilling fluids and cuttings released onto the Outer Continental Shelf.
U.S. Environmental Protection Agency, Ocean Programs Branch, Office of
Water and Waste Management and NPDES Technical Support Branch, Office of
Water Enforcement and Permits.
Schaffner, L. C., D. F. Boesch, and M. A. Bowen. 1981. Macrobenthos coloniza-
tion. Pages. 6,1-6,47 j_n Experimental colonization of crude oil contam-
inated sediments by benthos on the Middle Atlantic continental shelf.
Final report to U.S. Bureau of Land Management (Contract AA551-CT878-32).
Tagatz, M. E., J. M. Ivey, H. K. Lehman, M. Tobia,-and J. L. Ogelsby. 1980.
Effects of drilling mud on development of experimental estuarine macro-
benthic communities. Pages 847-865 j_n research on environmental fate and
effects of drilling fluids and cuttings. Lake Buena Vista, Florida,
January 1980.
Tester, A. M., and 0. J. Baumgartner. 1979. Prediction of initial mixing for
municipal ocean discharges. EPA Corvallis Environmental Research Lab.
CERL Publication 043.
Thompson, J. H., Jr. 1980. Responses of selected scleractinian corals to
drilling fluids used in the marine environment. Ph.D. Dissertation.
Texas A&M University, College Station, Texas. 129 p.
61
-------
APPENDIX
Workshop participants raised a number of important and valid points in
their comments on this workshop report. These points included concern over
the extrapolation of fixed-length, fixed-exposure toxicity tests to field
conditions; observations that shunting has been successful in routing plumes
away from coral reefs; identification of the importance of considering fate
and effects in "closed" water bodies; and concern over consideration of dis-
charged material at density stratification layers where sensitive organisms
might also concentrate. In addition, Donald Boesch provided a detailed
critique of the Soft Bottom Effects Submodel. Although the submodels developed
in a 1-week workshop are often of limited value in themselves, the structured
modeling approach does provide a well-focused framework for discussing the
relevant mechanisms and relationships. In this spirit, Dr. Boesch1s comments
are included here as an appendix to the report.
COMMENTS ON THE SOFT BOTTOM EFFECTS SUBMODEL
Donald F. Boesch
Louisiana Universities Marine Consortium
Chauvin, LA 70344
Comparison of the Water Column Effects Submodel and the Soft Bottom
Effects Submodel illustrates the strengths and weaknesses of the adaptive
environmental assessment approach. The physics of dispersion of contaminants
in the water column is better known than that of deposited particulate
material. Bioassay procedures, although not without limitations, more reason-
ably simulate the conditions of exposure of pelagic organisms to contaminants
than those experienced by benthic organisms exposed to a complex sediment
medium. Consequently, the water column fate and effects submodels are more
richly supplied with observations which allow for development of models with
variable parameters. This permits the heuristic use of sensitivity analysis,
thus identifying which factors might realistically influence the effects
predicted and which processes deserve further research.
The contributors to the Soft Bottom Effects Submodel were evidently
deterred because a lack of data or sound conceptual framework in which to
consider variable conditions and used a rather narrow set of assumptions, most
of which are relatively liberal, in the sense of diminishing the extent of
expected effects. This -is unfortunate because the majority of drilling fluid
solids are deposited on the seabed rather rapidly, the benthic organisms are
62
-------
exposed to them for longer periods of time relative to pelagic organisms, and
it is only with the soft bottom benthos that effects of drilling fluid
discharge have been detected in nature.
The Lower Plume Submodel is based on unrealistic assumptions concerning
the settling of particles as individuals, whereas actual observations indicate
a negatively bouyant jet with horizontal spreading near the seabed. Also
resuspensive or bed load spreading are not dealt with except as a source of
dilution. The spurious nature of this model is illustrated by the prediction
of confinement of particle accretion to extremely small radii (as little as
3.4 m in 20 m of water with a 1 m/min current; Table 4) and the counter-
intuitive prediction that deposited muds are dispersed much more slowly in
waters 20 m deep than in waters 80 m deep.
The assumptions of the Soft Bottom Effect Submodel regarding the life
history characteristics ("invader" species) of constituent organisms and their
resistence to burial restrict the potential relevance of this model to, at
best, a few extreme environments. Continental shelf benthic communities,
particularly those on the outer shelf, include many "equilibrium" species
which have long generation times and are slow to recruit. Also, the assump-
tion of 50% survival following burial by more than a meter of sediment is
probably in error by an order of magnitude or two for continental margin
macrobenthos, although relevant data do not exist. In' environments character-
ized by a low rate of sediment flux (resuspension plus net deposition), such
as the continental slope, tolerance to burial is probably very low.
Additional problems relate to the use of Petrazzuolo1s (1981) model for
predicting toxicity effects on soft bottom benthos. Petrazzuolo used two
approaches: Type I Analysis based on published LC50 values with an application
factor of 0.01, and Type II Analysis based on the relationship of sediment
barium concentration to community development in laboratory experiments
conducted on the Florida Gulf coast. It is unclear which of Petrazzuolo1s
analyses were applied, although there are serious limitations to the applica-
tion of either First, the LC50 data represent aqueous concentrations in the
sediment medium in which the benthos lives. Petrazzuolo's analysis is based
on a tenuous inference that "benthic impacts of drilling fluids were thought
likely to correspond to dispersions of these fluids in the water column." In
fact, both field data and the upper plume and lower plume submodels contradict
this assumption. Petrazzuolo's Type II Analysis is based exclusively on a
series of experiments conducted at Gulf Breeze assessing the effects of drill-
ing fluids and barite on community development in aquaria through which sea
water was pumped. Benthic colonists of laboratory aquaria represent species
predisposed for rapid recruitment rather than natural communities.
As in the case of tolerance to burial, the mortality induced by storms is
also likely to vary for different habitats. Natural communities are, however,
adapted to storms and other -sediment disturbances which are normal features of
their environments. Although severe storms undoubtedly cause heavy mortal-
ities, many continental shelf communities (e.g., Middl e Atlantic Bight) undergo
resuspension or erosion of 1 cm or more of sediment with greater than 25%
survi val (Fig. 16).
63
-------
The variable to which the predictions of effects is most sensitive is
perhaps the recovery time or resilience of benthos. The justification for the
model predicting enhanced populations of meiobenthos one month after additions
of drilling fluids is not supported. As indicated above the estimated macro-
fauna! colonization rates are based on experiments in laboratory aquaria
through which estuarine water flows and are unrealistically rapid for conti-
nental shelf macrobenthos. Data now exist to show that "recovery" of macro-
benthos following its annihilation ranges from weeks to several years depending
on the habitat and the adaptation of its community and populations to disturb-
ance (Boesch and Rosenberg 1981). Model predictions incorporating a range of
colonization rates could easily have been included in this assessment.
64
-------
0261
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
2.
Workshop Summary
3. RECIPIENT'S ACCESSION NO.
. TITLE AND SUBTITLE
RESULTS OF AN ADAPTIVE EIWIIIONMENTAL ASSESSMENT MODELING
WORKSHOP CONCERNING POTENTIAL IMPACTS OF DRILLING MUDS
AND OJTTINGS ON THE MARINE ENVIRONMENT
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
G.T. Auble, A. K. Andrews, R.A. Ellison, D.B. Hamilton,
R.A. Johnson, J.E. Roelle and D.R. Marmorek
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Fish and Wildlife Service
Western Energy and Land Use Team
Office of Biological Services
Fort Collins, Colorado 80526
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA-81-D-X0581
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Environmental Research Laboratory
Office of Research and Development
Gulf Breeze. FL 32561
13. TYPE OF REPORT AND PERIOD COVERED
. SPONSORING AGENCY CODE
J600-9
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This publication summarizes findings of a -workshop held September 14-18, 1981,
under a Federal Interagency Energy /Environment Agreement (EPA-81-D-X058D between
the U.S. Environmental Protection Agency and the U.S. Fish and Wildlife Service. The
U.S. EPA Environmental Research Laboratory, Gulf Breeze, Florida, was host for
the sessions held on Pensacola Beach, FL. Discussions focused on information
pertaining to fate and effects, identification of general relationships between
drilling mud fluids and the marine environment, and identification of site-specific
variables likely to determine impacts of drilling muds and cuttings in various
marine sites. The workshop was structured around the construction of a model
simulating fate and effects of discharges from a single rig into open waters of
the Gulf of Mexico. Factors that might produce different fate and effects in enclosed
areas such as bays and estuaries also were discussed. Considerable knowledge
(such as that concerning fate and physical effects of dredge spoil) that could
be effectively employed in analysis of potential fate and physical effects in
enclosed areas was identified.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
18. DISTRIBUTION STATEMEN1
Release to public
19. SECURITY CLASS (This Report/
Unclassified
;i. NO. OF PAGES
64
20. SECURITY CLASS (This page I
Unclassified
22. PRICE
EPA
Form 2220-1 (Rev. 4-77) PREVIOUS EDITION i s OBSOLETE
------- |