P16S i^n
AQUACULTURE WASTEWATER TREATMENT:
WASTEWATER CHARACTERIZATION
AND DEVELOPMENT OF APPROPRIATE
TREATMENT TECHNOLOGIES FOR THE
ONTARIO TROUT PRODUCTION INDUSTRY
DECEMBER 1990
Environment
Environnement
Ontario
MOE AOUACULTURE COMMITTEE RESPONSE TO
"AOUACULTURE WASTEWATER TREATMENT; WASTEWATER
CHARACTERIZATION AND DEVELOPMENT OF APPROPRIATE TREATMENT
TECHNOLOGIES FOR THE ONTARIO TROUT PRODUCTION INDUSTRY"
The above referenced report submitted by Canadian
Aquaculture Systems provides results of an extensive field
study which examined fish production practices, culture
facilities design, suspended solids settling behaviour, and
critical design parameters of existing effluent treatment
facilities, in an effort to identify those factors which
most significantly affect effluent quality and management.
The consultant's conclusions and recommendations address
suitable treatment technology, design requirements,
management and operating principles to be applied at
aquaculture facilities in order to minimize the loadings of
suspended solids and total phosphorus in effluent
discharged to a receiving water.
The conclusions and recommendations contained in Chapter
8.0 of the report are examined below and the Committee's
position is provided.
8.1 Ontario Aquaculture Operations
The Committee concurs with the conclusions and
recommendations listed in this section as a whole. It
should be noted, that while feed-related factors were found
to be subordinate with respect to pollution impact (8.1.7),
they should not be discounted for their potential in the
overall management of an aquaculture operation with the
goal of minimizing effluent contaminant discharges.
Several feed-related factors, such as reduced phosphorus
content, were not included in the study and may play an
important role in reducing contaminant discharges.
.../2
-2-
8.2 Effluent Treatment Design
8.2.1
Overflow rate was recommended as the basis for effective
facility design. Specific design limits are listed. The
Committee concurs with this recommendation and would add
that other design considerations, such as flow through
velocity and inlet and outlet design, which were addressed
in the body of the report are also of importance in
ensuring a treatment facility design that will be effective
in reducing contaminant discharges.
The report also notes that with the recommended design
limits that Total Phosphorus (TP) may marginally exceed the
0.10 mg/1 compliance limit. While this may be occurring
with several existing operations, the Committee believes
that through measures such as improved feed formulations,
which have reduced phosphorus content, and with improved
overall facility management practices, that the TP
compliance limit will be achieved consistently.
8.2.2 and 8.2.3
The Committee concurs with the recommendations.
8,3 Best Management Strategies
The Committee concurs with the conclusions and
recommendations listed in this section as a whole. The
conclusion that vacuuming of settled wastes is the most
effective removal method (8.3.4) should also identify the
importance of a regular schedule of cleaning in ensuring
the overall performance of the treatment system. The
appropriate cleanout frequency will be as required to
satisfy the effluent limits, but can be expected to be
approximately twice weekly. It should be noted that the
dollar values listed in 8.3.6 are in 1990 dollars.
./3
-3-
8.4 Enforcement of Ministry Guidelines
8.4.1
The current compliance limit contained in the guideline for
Total Phosphorus (TP) is 0.10 mg/1 except where the
background level in a surface water supply exceeds
0.10 mg/1 in which case TP shall not exceed the background
concentration. Maintaining an absolute limit for
groundwater supply systems reflects the Committee's
position of not allowing the export of a poor quality water
supply to other receiving waters. For surface water supply
systems some of the background TP will be removed as a
result of the treatment system, thus facilities will be
able to operate and achieve the no net increase above
background criterion. The position of no net increase when
background concentrations exceed 0.10 mg/1 is in keeping
with Policy 2 from the Publication "Water Management -
Goals, Policies, Objectives and Implementation Procedures
of the Ministry of the Environment", May 1984.
The Committee concurs that the solubility of phosphorus can
affect the overall performance of the treatment system, and
thus advocates a regular and frequent cleaning program.
Additionally, the Committee concurs that the reduction of
excess phosphorus in commercial feeds could reduce
contaminant discharges.
8.4.2
The Committee does not concur. Environmental impacts due
to aquaculture operations have been documented. As well,
the potential for impacts in sensitive headwater areas is
strong. Additional documentation of phosphorus loadings
will be provided through the reporting requirements
contained in Certificates of Approval.
8.4.3
The Committee concurs that research is required into the
discharge of supernatant from manure storage lagoons, but
would add that specific attention should be focused on the
potential for impacts on groundwater quality.
.../4
-4-
8.4.4
The Committee concurs that a co-operative effort will best
protect the environment. Consultation with industry
organizations and other provincial agencies has been
ongoing. The Committee does not concur with the
recommendation to place the interim guidelines in abeyance,
as this would allow uncontrolled development within the
industry and would place the Ministry in a reactive
position in dealing with environmental impacts. Through
the use of the interim guidelines the Ministry is
attempting to be proactive and prevent adverse impacts from
occurring.
ISBN 0-7729-7314-8
AQUACULTURE WASTEWATER TREATMENT:
WASTEWATER CHARACTERIZATION AND DEVELOPMENT OF APPROPRIATE
TREATMENT TECHNOLOGIES FOR THE ONTARIO TROUT PRODUCTION INDUSTRY
Report prepared by:
Daniel Stechey and Yves Trudell
Canadian Aquaculture Systems
Report prepared for:
Environmental Services
Water Resources
Ministry of the Environment
DECEMBER 1990
RECYCUBLf
Cette publication technique
n'est disponible qu'en anglais
Copyright: Queen's Printer for Ontario, 1990
This publication may be reproduced for non-commercial purposes
with appropriate attribution
FIBS 1319
log 90-2309-041
EXECUTIVE SUMMARY
The environmental impact of intensive fish culture has long been discussed,
however, implementation of economically feasible and effective technologies for
wastewater treatment remains minimal on an industry-wide scale. The design of
wastewater treatment units requires a fundamental understanding of all aspects
of fish culture and wastewater engineering. This extensive field study examines
fish production practices, culture facilities design, suspended solids settling
behaviour, and critical design parameters of existing effluent treatment
facilities, in an effort to identify those factors which most significantly
affect effluent quality and management. The design of culture units and specific
culture/management practices are the principal factors governing effluent water
quality prior to treatment. Most interestingly, for a given volumetric flow of
water, no difference was observed in the carrying capacity (kg fish/Lps) of
different facility designs; however, water quality was found to vary quite
significantly, suggesting that certain designs and/or management practices are
much more polluting. Sedimentation practices (i.e. gravity settling) tend to be
most widely applicable in intensive salmonid aquaculture; they require no energy
input or specialized operating skills, are relatively inexpensive to install and
operate, and can be easily incorporated into both new and existing facilities.
A detailed overview of sedimentation design for fish culture units is included.
ACKNOWLEDGEMENTS
This study was developed and implemented through the co-operative efforts
of the private sector, and the provincial and federal governments. Financial
support for the investigation was provided by The Ontario Ministry of the
Environment and The National Research Council of Canada.
Mr. Stew Thornley was responsible for co-ordinating the support of the
Ontario Ministry of the Environment. His efforts led to the use of the OME
laboratories in London for analysis of water quality samples collected during the
field investigation. Mr. Dan Lynch (NRC) arranged for Industrial Research
Assistance Program (IRAP-H) funding to support the field monitoring program. Mr.
John McFarlan (OTFA) was responsible for aligning the co-operation of the Ontario
Trout Farmers' Association and those individual aquaculturists who are not
members of the association.
The comments and constructive criticism offered by the members of the
Ministry of the Environment Aquaculture Committee have enhanced the quality and
presentation of the final report. Mr. Keith Somers (OME, Water Resources Branch,
Toronto) provided invaluable assistance with statistical analyses. Many thanks
are extended to the private and government aquaculturists throughout Ontario who
permitted us to visit their operations to obtain the water quality samples and
information which we required.
Qanadian
tIquaculture
TABLE OF CONTENTS
EXECUTIVE SUMMARY i
ACKNOWLEDGEMENTS i
TABLE OF CONTENTS ii
1.0 INTRODUCTION 1
1.1 Aquaculture in Ontario 1
1.2 Environmental Impact of Intensive Salmonid Aquaculture .... 1
1.2.1 Suspended Solids Production in Intensive Salmonid
Culture 3
1.2.2 Phosphorus Production in Intensive Salmonid Culture . . 3
1.3 Project Purpose and Objectives 5
2.0 TECHNOLOGIES FOR AQUACULTURE EFFLUENT TREATMENT 9
3.0 SEDIMENTATION TECHNOLOGIES FOR AQUACULTURE WASTEWATER TREATMENT . . 10
3.1 Fundamental Concepts of Sedimentation 10
3.2 Factors Influencing the Design of Sedimentation Basins .... 14
3.2.1 Tank Velocity, Turbulence & Scour 14
3.2.2 Short-Circuiting i Tank Stability 15
3.2.3 Inlet & Outlet Design Considerations 17
3.2.4 Principal Design Considerations 20
4.0 CHARACTERIZATION OF AQUACULTURE OPERATIONS & EFFLUENTS 25
4.1 Purpose & Objectives 25
4.2 Field Survey & Water Quality Sampling 26
4.2.1 Predictor Variables 27
4.2.2 Response Variables 28
4.2.3 Scope and Limitations 29
4.3 Data Analysis & Results 30
4.4 Chapter Summary and Principal Findings 38
5.0 STATUS OF AQUACULTURE EFFLUENT TREATMENT IN ONTARIO 50
5.1 Aquaculture Settling Facilities in Ontario 50
5.2 El^fectiveness of Existing Treatment Operations 51
5.3 Chapter Summary and Principal Findings 53
5.0 IMPLEMENTATION OF AQUACULTURE EFFLUENT TREATMENT 57
5.1 Commercial Fish Production Facilities in Ontario 57
6.2 Design of Effluent Treatment Facilities for
Intensive Salmonid Aquaculture 58
5.3 Projected Production of Phosphorus and Solids from Intensive
Fish Culture Facilities 62
6.4 Projected Performance of Effluent Treatment Facilities .... 63
6.5 Management Strategies for Effective Effluent Quality Control . 68
5.5.1 Feed and Feeding Practices 68
5.5.2 Rearing Unit Design 69
6.5.3 Solids Settling Unit Design 59
5.5.4 Solids Removal and Disposal 70
n
Qanadian
tXquaculture
gYSTEMS
6.6 Economics of Effluent Treatment 71
7.0 INTERIM GUIDELINES & THE CERTIFICATE OF APPROVAL 78
8.0 PROJECT SUMMARY 81
8.1 Ontario Aquaculture Operations 81
8.2 Effluent Treatment Design 81
8.3 Best Management Strategies 82
8.4 Enforcement of Ministry Guidelines 82
9.0 LITERATURE CITED 84
APPENDIX I -- Field Survey Questionnaire 89
APPENDIX II -- Water Quality Data from Participating Fish Farms .... 91
APPENDIX III -- Spearman's Correlation Matrix for Participating Fish
Farms 94
APPENDIX IV — Effluent Treatment Facilities Data 98
APPENDIX V -- Ministry of the Environment Interim Guidelines .... 100
in
1.0 INTRODUCTION
1.1 Aquaculture in Ontario
Aquaculture began in Canada over 100 years ago with the controlled
propagation of a variety of sport fish species for stock enhancement programs
(MacCrimmon 1984). Such operations were predominantly hatchery operations
designed to supplement the recruitment and survival of fry. It was not until the
1950s, with the culture of oysters and trout in British Columbia, that commercial
production of fish for human consumption began.
In Ontario, commercial aquaculture commenced in 1962 and has grown at a
rapid rate ever since. Moccia and Castledine (1987) have outlined this
development in three major phases. The "Novice" phase (1962 - 1975) was
characterized by low productivity. From 1975 to 1981, the "Skills Development"
phase witnessed a large increase in the number of farms, the establishment of a
domestic supply network, and the formation of the Ontario Trout Farmers'
Association, Since the early 1980s, the "Industrialization" phase has seen the
intensification of culture practices and the beginning of the Ontario Trout
Producers' Co-operative. The latter was instrumental in opening up distribution
channels into high-volume markets and, consequently, it induced many more farmers
and entrepreneurs to establish fish farming operations.
Since the beginning of commercial aquaculture in the province. Rainbow Trout
has been the only species cultured in economically significant quantities.
Although Brook (Speckled) Charr and both Smallmouth and Largemouth Bass are also
farm-raised, these species are generally cultured for stock enhancement purposes
and are not cultured in large quantities in feedlot-type operations. Presently,
there are about 227 private, community-run, and government aquaculture facilities
in Ontario (LIMA 1988). In 1988, the 120 commercial farms throughout the province
produced approximately 1,830 tonnes (4 million pounds) of trout, and production
is expected to double by 1995 (Moccia and Bevan 1989). Currently, the industry
remains somewhat fragmented, with the majority of operations being only small -to
medium-size and serving to supplement the income of traditional farmers (Stechey
et al . 1987). Moccia and Bevan (1989) found that 25% of commercial farms
produced less that 2 tonnes in 1988, 50% produced less than 9 tonnes, and 75%
less than 23 tonnes.
1.2 Environmental Impact of Intensive Salmonid Aquaculture
In many types of commercial aquaculture, production practices must be highly
intensive for an operation to achieve economic feasibility. In such systems, it
is not unusual for the available water supply to be re-used in multi-pass
facilities to attain the necessary production goals. Moreover, in aquaculture,
there is a saying: "Fish don't pollute, but feed does!" As with all animals,
feed is only partially utilized by the fish; the remainder is discarded into the
water as soluble metabolic by-products and solid waste. Consequently, intensive
culture practices can lead to substantial deterioration of water quality. Unlike
the terrestrial farming of livestock, which often creates off-site pollution
problems, unutilized feed and metabolic waste products excreted by the fish foul
Aquaculture
*S~^ STEMS
the water and, therefore, the system requires continuous flushing to maintain an
acceptable culture environment (Wheaton 1977).
Nearly 20 years ago, Liao (1970a) stated that "The potential problem of
pollutants discharged from a salmonid fish hatchery has long been overlooked.
There is almost no literature on the subject." Today, however, the problem is
well recognized and the issue of effluent quality has been a concern from an
environmental perspective for many years. Aquacultural wastewater treatment is
no longer viewed as an option in fish farming; it is necessary for the protection
of the receiving water course and the assurance that water quality objectives are
attained.
Nevertheless, while it is commonly accepted that rapid and gentle removal
of the solid waste component of discharged effluent is essential, years of
discussion and research have failed to produce appropriate technologies for
effective and economical wastewater treatment at aquaculture facilities. In
fact, Liao's statement, with minor modification, still applies today:
The potential problem of pollutants discharged from a salmonid fish
hatchery has long been discussed. There remains no rationalized
solution on the subject.
Moreover, this problem is not limited to New Brunswick smolt production
operations. In Aquaculture Magazine, Hopkins and Manci (1989) state that,
industry-wide,
"pollution-related problems in fish culture are on the rise. As the
industry grows and expands, these problems will continue to intensify
unless economically and scientifically sound solutions are discovered
and applied."
Liao (1970a) classifies aquacultural waste products into three categories:
(1) chemicals and drugs employed for cleaning and/or control of
diseases, parasites and aquatic weeds;
(2) bacteria and parasites which may be harboured in the facility and
subsequently released; and,
(3) fish fecal matter and residual feed, and soluble metabolic products.
While the first two categories of pollutants are important to understanding
the full impact of aquacultural discharges, for the most part they tend to occur
relatively infrequently. Nevertheless, these issues must also be examined by the
aquaculture industry -- ideally before they lead to problems similar to current
concerns with respect to solids and nutrient discharges. In contrast, however,
the discharge of fish fecal matter and residual feed is an ongoing occurrence at
commercial, hobby, and R&D fish culture facilities. Moreover, since most
salmonid aquaculture facilities are located on sites in the proximity of high-
quality water resources, the solid waste component presents a serious pollution
problem, especially when farm effluents are discharged to sensitive cold-water
receiving bodies. Conventional loadings of seven common fish farm pollutants are
presented in Exhibits 1.1 and 1.2.
1.2.1 Suspended Solids Production in Intensive Salraonid Culture
Aquaculture-related pollution problems manifest themselves in various ways.
The accumulation of solid matter on stream bottoms can be especially detrimental
to aquatic environments. In fact, the Michigan DNR (1973) concluded that the
total suspended solids load was the most significant problem caused by hatchery
discharges. Since fecal and feed solids are predominantly organic matter, their
oxidation depletes the dissolved oxygen concentration in the water and results
in the release of dissolved nutrients. Moreover, these solid wastes support a
diverse population of micro-organisms which also contribute to oxygen depletion,
and which may be deleterious to other aquatic species. If left to accumulate to
depths where bottom layers become anaerobic, noxious gases and other toxic
compounds can form. Samuelsen et al . (1988) found that the gases generated from
fish farm sediment deposits consisted primarily of methane (70-90%), carbon
dioxide (10-30%), and hydrogen sulphide (1-2%), all of which are toxic to fishes
and other aquatic organisms.
Relative to phosphorus, waste solids production in intensive fish culture
is less complex to model, and solids are substantially simpler to remove from
discharged effluent. Due to periodic changes in feed formulations based upon
feed ingredient costs, however, waste solids production is not necessarily
predictable. Nevertheless, industry average production rates have been
established for intensive trout culture which are sufficiently accurate for
modelling.
Merican and Phillips (1985) report no significant correlation (P>0.05)
between TSS output and fish size, and conclude that feeding rate, rather than
fish size, is the primary determinant of solid waste production. Depending on
feed composition and digestibility (efficiency of utilization), waste solids
production varies between approximately 200 kg and 1000 kg per metric tonne of
fish produced (Westers 1989). A number of researchers suggest that 300 grams of
waste solids per kilogram of feed added to the system is an acceptable industry
average (Willoughby et al . 1972; Mudrak and Stark 1981; Zeigler 1988). This
production rate is confirmed by Cho et al . (1985), who found waste solids
production to range between 200 to 400 grams per kilogram feed. Westers (1989)
indicates that levels below 200 gTSS/kg feed are unrealistic in commercial
culture settings and that levels above 450 gTSS/kg feed are infrequent since they
tend to be indicative of poor feed utilization and/or feed wastage. Moreover,
since feed is the largest single cost factor in commercial trout culture, the
latter quickly becomes uneconomical. For modelling purposes, a solids production
rate of 300 grams TSS per kg feed is conventionally applied (Exhibit 1.1).
1.2.2 Phosphorus Production in Intensive Salmonid Culture
The addition of phosphorus to aquatic habitats can present serious, long-
term pollution problems. Fortunately, from a biological productivity
perspective, inorganic orthophosphate is the only form of phosphorus of
r* AMADIAf!
7\quaculture
SVS'EMS
significance (Wetzel 1975), and the phosphorus component in fish feed which
ultimately ends up in the receiving body of water is principally organic
phosphorus bound within the solid waste particles (Persson 1988). As such, it
is biologically unavailable in the water column.
Nevertheless, chemical and biological reactions can act to liberate
phosphorus from bound particles, thereby enhancing primary productivity and
eutrophication. In the water column, an exchange equilibrium is established
between phosphorus in sediments and the water. Under aerobic conditions, this
equilibrium is largely uni-directional toward the sediments. Once the sediment
layer becomes sufficiently thick (deep), such that diffusion of oxygen to lower
layers becomes limiting, anaerobic conditions develop. In the absence of oxygen,
the equilibrium shifts toward the water column due to changes in redox
conditions, resulting in the release of biologically available phosphorus
compounds. Phosphorus mobilizing bacteria are also present, especially in highly
organic sediments such as fish culture wastes, and serve to increase the
pollution impact of phosphorus, although their impact is generally about l/20th
that of anaerobic chemical reactions (Wetzel 1975; Sondergaard 1988).
This biologically available fraction of phosphorus (and to a lessor extent,
nitrites and nitrates) in fish farm discharges, contributes to the enhancement
of primary productivity (i.e. increased populations of micro- and macro-algae),
thereby promoting eutrophication of the receiving body of water (Trojanowski et
al. 1985; Persson 1988). In Ontario, the Ministry of the Environment (1982)
observed that the aquatic environment was substantially impaired in streams
downstream from discharges from trout farms as compared to sites immediately
upstream from the farms. Detrimental effects were principally related to
increases in the total number of benthic organisms, especially Chironomidae
(midge) and Simuliidae (black flies), the stimulation of luxuriant plant and
algal growths, and the accumulation of organic solids on stream bottoms. The
Michigan Department of Natural Resources (1973) studied water quality below 9
salmonid fish culture facilities in the state and discovered downstream increases
in the concentrations of biochemical oxygen demand (BOD), suspended solids,
organic nitrogen, ammonia nitrogen, soluble ortho-phosphate, and total
phosphorus. Liao (1970a) also reports similar upstream-downstream changes at
aquaculture effluent discharge sites in Washington state.
Salmonid fishes have 3.8 to 4.5 grams of phosphorus per kilogram body
tissue; the accepted average being 4.0 grams P/kg tissue or 0.4%. Nevertheless,
the dietary requirement for phosphorus is generally 0.7 to 0.8 percent feed
weight, in spite of research results which indicate that experimental diets
having phosphorus concentrations as low as 0.4% produce no detrimental effects
to production and growth (Ketola 1985; Eskelinen 1986; Wiesmann et al . 1988).
Many commercial diets, however, have phosphorus levels ranging between 1.1% and
1.8%. This excess phosphorus ultimately ends up in the receiving body of water
since it cannot be utilized by the fish.
Merican and Phillips (1985) found that an average of 65.9% of dietary
phosphorus is discharged in the solid fecal waste. Not all of this discharged
phosphorus remains bound, however. EIFAC (1972) reports that up to 32% of the
particulate phosphorus is solubilized within 2 days of discharge. Alternatively,
field data show that, on average, 30% to 60% of all phosphorus discharged from
salmonid fish culture operations is permanently retained in the sediment fraction
of the solid waste (Persson 1988). Variability is due to the feed brand and
specific feed ingredients. For instance, phytin, a phosphorus compound in plant
proteins, is poorly absorbed by fish, yet it is frequently included in fish
feeds. Other sources of phosphorus, such as defluorinated rock phosphate, are
better utilized by salmonid fishes and are less polluting (Ketola 1985). Based
on usual ingredients, commercial diets tend to contain excessive quantities of
phosphorus which are not required for optimum performance (Wiesmann et al . 1988).
The figures presented in Exhibit 1.3 simplify the utilization of phosphorus
by salmonids to illustrate the potential pollution impact of excess dietary
phosphorus. The numbers reveal that effluent phosphorus levels increase
substantially with increased dietary phosphorus since fish have a limited
phosphorus uptake capacity, which does not change with dietary level. Notice
that a change in dietary phosphorus concentrations from 0.7% to 1.6% increases
total discharged phosphorus four-fold.
Westers (1989) suggests that the average discharge of phosphorus from
salmonid culture operations ranges between 1.5 and 25 kg P per metric tonne of
fish produced. Castledine (1986) provides factors used to determine waste
phosphorus production in intensive rainbow trout culture facilities. He
indicates that, for every kilogram of feed added to the system, 5.4 grams of
settleable and suspended phosphorus and 2.2 grams of dissolved phosphorus are
discharged in the production effluent; a total of 7.6 grams of phosphorus per
kilogram feed. The proportion of TP to DP is 71:29. Cho (1990) and Ketola
(1989) suggest that 9 grams of phosphorus per kilogram of feed may be more a
appropriate factor, and that the breakdown between solid and soluble forms
approximates 57:33. Cho further adds that, at a minimum, 25% of all phosphorus
in the effluent will be dissolved in solution. Castledine' s factors are most
frequently applied for effluent modelling (Exhibit 1.1).
Therefore, the production of pollutants in salmonid aquaculture is directly
related to the input of feed to the culture system and can be estimated using the
accepted feed factors as described by Castledine (1986), and presented in Exhibit
1.1. It should be noted, however, that feed quality, daily operations (i.e.
feeding, cleaning, etc.), culture practices (i.e. stocking densities, water flow
rates, general management techniques, etc.), and seasonal variation in standing
crop biomass create fluctuations in the actual amount of wastes produced.
Although suspended solids and nutrient concentrations in salmonid fish culture
effluent are relatively dilute (see Exhibit 1.2), long-term cumulative problems
may develop in receiving waters due to the high flow rates required by salmonid
fishes (Liao 1970a) .
1.3 Project Purpose and Objectives
Consequently, in view of the wastes generated in salmonid fish culture,
aquacultural effluents can potentially violate several of the Provincial Water
Quality Objectives (Ministry of the Environment 1984) and/or specific interim
guidelines established for the Ontario aquaculture industry:
/Iquaculture
pVSTEMS
1) Average effluent concentrations exceeding 10 mg total suspended solids
per litre and/or 0.10 mg total phosphorus per litre are above interim
compliance limits and are, therefore, deemed to be detrimental to
»"eceiving bodies of water.
2) The mass of solids produced may settle to form objectionable deposits
in the receiving water course.
3) Dissolved phosphorus and nitrogen compounds may contribute to
undesirable aquatic life or result in the dominance of nuisance
species.
4) To a lesser extent, floating bacterial and/or algal mats and scum may
become a nuisance.
5) If poorly designed settling ponds are used (as is currently the case at
several operations), long detention times may cause warming of the
discharged effluent during summer months, which may create a barrier to
the migration of fish and aquatic life in the mixing zone of the
receiving body of water.
Therefore, an effective means for reducing the solids content of the
discharged effluent is necessary; however, the high volumetric flow rate, and the
relatively dilute concentration of pollutants in the effluent, impose a unique
constraint on effluent treatment. Westers (1989) reports that the concentration
of phosphorus discharged from tertiary sewage treatment lagoons may be as much
as ten times greater than that in untreated fish hatchery effluent. The total
volumetric flow from salmonid hatcheries compounds the impact, however. Solbe
(1988) suggests that the impact of untreated effluent from 1000 kilograms of fish
is equivalent to the daily discharge of treated sewage from a city of 300,000
people. Consequently, treatment technologies must be highly efficient, yet they
must also be economically feasible to install and operate (Liao 1970b).
The uniqueness of most fish culture operations has meant that standardized
effluent treatment technologies have not been developed. Rather, treatment
facilities tend to be custom-made for individual operations (Lindqvist 1986).
Although custom design may continue to be the norm in this industry, with its
diversity of facility designs and culture practices, a need nevertheless exists
for the development of standardized design protocol.
Consequently, the purpose of this investigation has been to conduct applied
research to identify standardized and cost-effective techniques for the design
and implementation of appropriate wastewater treatment technologies and operating
practices for both new and existing fish culture facilities in the Province of
Ontario. Within the scope of this investigation, the following objectives have
been targeted:
1) Characterization of fish farming wastes on an industry-wide scale.
2) Evaluation of existing effluent treatment facilities.
3) Identification of primary constraints relating to the high-volume, low-
concentration nature of fish farm effluents and identification of
appropriate technologies for aquaculture wastewater treatment.
4) Assessment of the economics and cost structuring of aquaculture wastewater
treatment facilities, including capital and operating expenses.
5) Assessment of the suitability of the current "Interim Environmental
Guidelines for Salmonid Aquaculture Facilities in Ontario."
Exhibit 1.1:
Accepted factors used to determine waste production from
salmonid fish culture facilities based upon feed consumption
Pollutant Factor (x feed weight)^
Total Settleable & Suspended Solids .300'
Settleable & Suspended Phosphorus .0054
Dissolved Phosphorus .0022
Total Phosphorus .0075
Settleable & Suspended Nitrogen .0064
Dissolved Nitrogen .0317
Ammonia .0383
Source: Castledine (1986)
^ Factors will vary with changes in dietary efficiency.
These factors assume dry feed (10% moisture) with an average
digestibility of 80% and a Food Conversion Ratio = 1.2
2 Westers (1989)
r*Ar!AOlAri
Aquaculture
Qystems
Exhibit 1.2:
Average increase in pollutant concentrations at salmonid
fish culture facilities during normal operations. All
units in mg/L.
Pollutant
Averaae
; Increase
Ranae
BOD
5.36
0.12 - 36.5
Amnion i a
0.532
0.00 - 2.55
Nitrite
1.676
0.045 - 3.1
Phosphorus
0.077
0.01 - 0.26
Suspended Solids
7.0
0.0 - 55
Settleable Solids
3.5
0.0 - 35
Source: Liao (1970a)
Exhibit 1.3:
The fate of dietary phosphorus based upon % phosphorus in feed, and assuming that
fish utilize 0.4% phosphorus by body weight. This oversimplifies the issue,
however, it is intended to illustrate the fate of excess dietary phosphorus.
% P in Feed kg P/1000 kg kg P Retained % P Retained kg P Discharged
Feed by Fish by Fish in Effluent
0.7 7.0 4.0 57.0 3.0
1.1 11.0 4.0 36.0 7.0
1.6 16.0 4.0 25.0 12.0
2.0 TECHNOLOGIES FOR AQUACULTURE EFFLUENT TREATMENT
Due to the high volume of process water associated with intensive salrnonid
fish culture ooerations, and the concomitant low concentration of particulate and
dissolved pollutants, effluent treatment practices are generally limited to
primary treatment technologies; that is, the removal of particulate matter
(Persson 1988). Secondary (i.e. biological filtration, aerated lagoons, carbon
absorption) and tertiary '(i -e- chemical coagulation, nitrification-denitrific-
ation, ion exchange, etc.) treatment technologies have received only limited use
in aquaculture, and tend to be limited to highly-specialized applications.
The removal of solid material from wastewater streams and the selection of
appropriate treatment technologies is governed by (1) particle size, (2) soecific
gravity, and (3) strength, or shear resistance, of the particulate matter. An
initial requirement, therefore, in selecting an appropriate sol ids separation
technique for any application, is identification of the properties of the
material to be separated.
Wastewater solids are routinely classified into four categories, based upon
particle size: soluble (<0.001 nm) ; colloidal (0.001 - 1.0 \im) ; suora-col loidal
(1.0 - 100 Jim); and settleable (>100 ^m) . The total suspended solids (TSS)
component of a wastewater stream is determined by filtration of a sample of the
wastewater through a 0.45 ^m membrane filter, and measurement of the dry weight
of the sample in mg/L. Settleable solids are further defined as that portion of
the TSS which will settle, in quiescent conditions, within one hour due to
gravity alone (Sundstrom and Klei 1979).
Numerous technologies, most of which have originated in municipal and/or
industrial wastewater treatment applications, have been applied in aquaculture
for solids control. The UMA Engineering Ltd. report (1988) describes the use of
several of these methods in fish farming, including gravitational separation,
screening, and fi Itration, among others. Consequently, simi lar descriptions wi 1 1
not be presented here. Alternative technologies, including peat bed filtration,
flocculation throuah addition of coagulants, and nutrient removal via plant
biomass uptake, are" described by Parjala (1984). Stechey (1987) discusses the
use of static screen, vibrating screen, centrifugal screen, micro-screen, and
compact sedimentation units.
Peat bed filtration is conducted in earthen ponds having a 30 - 50 cm thick
layer of peat overlaid on a gravel percolation bed. While the specific
absorption of peat varies with the species of peat, its level of humification,
and various physical and chemical properties, a rule of thumb states that up to
200 grams of phosphorus can be removed per cubic meter of peat. In spite of its
capacity for phosphorus absorption, commercial-scale aoplication is limited to
only the smallest farms due to total flow limitations. Similarly, coagulation
also tends to be inefficient for aquacultural application, again due to the high
volume, low concentration nature of the wastewater stream. Moreover, the cost
of chemical coagulants would be relatively high and large contact chambers would
be required (Parjala 1984).
static screens offer the advantages of high volumetric capacity, compact
size and portability, relatively long screen life, production of a relatively
concentrated solids fraction, and the addition of dissolved oxygen to the
wastewater stream. In aquacultural applications, however, the specific gravity
of the solids to be clarified is often too low to adequately permit their
collection and discharge from the units. Moreover, most high-volume static
screens require a pressurized feed source, resulting in an added energy component
(Dorr-Oliver 1982). As well, on-site testing with low-head gravity-feed screens
has revealed that undigested plant materials (presumably fibre from feed
ingredients) and the relatively sticky fish fecal material cause blinding of the
screen, resulting in the need for frequent cleaning (Stechey 1987).
Consequently, it is generally agreed that, for efficient and effective
removal of solids from fish farm effluents, most existing filtration and
screening technologies tend to be impractical from either a functional and/or
economical approach. Notwithstanding that some commercially available treatment
units have an application in specific aquaculture operations, sedimentation
practices (i.e. gravity separation) tend to be most widely applicable in
commercial salmonid culture. Sedimentation basins require no energy input, are
relatively inexpensive to install and operate, require no specialized operational
skills, and can be easily incorporated into both new and existing facilities.
Moreover, they tend to produce acceptable effluent quality.
3.0 SEDIMENTATION TECHNOLOGIES FOR AQUACULTURE WASTEWATER TREATMENT
3.1 Fundamental Concepts of Sedimentation
When a liquid containing particulate matter is placed into a relatively
quiescent state, those particles having a specific gravity greater than that of
the liquid begin to settle. This principle has been applied for generations as
a means of clarifying wastewater streams. Depending upon particle size and
density, and the physical characteristics of the waste and the wastewater medium
(water), the sedimentation process can be defined as either discrete (Class I),
hindered or flocculent (Class II), zone (Class III), or compression settling
(Class IV). These processes are defined in Exhibit 3.1.
A rule of thumb suggests that, at total suspended solids concentrations less
than approximately 500 mg/L (0.05% solids), sedimentation is typically discrete
(Camp 1946). Discrete settling particles maintain their size and shape in a
dilute suspension, and have little tendency to adhere upon collision. Each
discrete particle has a constant settling velocity, which is independent of that
of other particles. Flocculation tends not to occur due to the remote likelihood
of chance collisions between particles in such dilute concentrations. Therefore,
in salmonid fish culture effluents, it is reasonable to assume that settling is
discrete.
10
Exhibit 3.1:
Suspension classes and their settling characteristics.
Class I Suspensions - Discrete
Definition: Dilute suspensions of non-flocculent particles that
settle out of suspension unhindered by the presence of
other particles.
Behaviour: Sedimentation rate is governed only by the surface area
of the sedimentation basin.
Class II Suspensions - Flocculent
Definition: Dilute suspensions of flocculent particles which, upon
collision, coalesce and settle at rates greater than did
the parent particles.
Behaviour: Sedimentation rate governed by (1) surface area of
sedimentation basin, (2) concentrations of suspensions,
and (3) depth of basin.
Class III Suspensions - Zone
Definition: Suspensions of high concentration in which particles
settle in a fixed position relative to each other in a
regime referred to as hindered or zone settling.
Particles may be flocculent or non-flocculent.
Behaviour: Sedimentation rate governed by (1) surface area of
sedimentation basin, (2) concentration of suspension, (3)
solids loading, and (4) depth of basin.
Class IV Suspensions - Compression
Definition: As particles accumulate in layers on the bottom of the
basin, each layer of solids provides support for those
layers above and the solids undergo compressive stress
which compacts the layers further.
Behaviour: Sedimentation has already been achieved prior to this
stage; concentration of the settled solids is governed by
(1) the sludge depth, and (2) the retention time in this
compressive zone.
Source: Adapted from Sundstrom and Klei (1979); Rich (1980J
11
Under the force of gravity, a particle accelerates downward through the
water column until the gravitational force comes into equilibrium with the
opposing frictional fluid drag force. When these opposing forces become
balanced, the particle settles at a uniform velocity, which is referred to as the
terminal settling velocity (Vs) .
All continuous flow settling basins are comprised of four zones according
to function (Exhibit 3.2). The inlet zone serves to uniformly distribute the
suspension over the entire cross-section of the basin. Sedimentation occurs in
the settling zone and, upon removal from the water column, the solids accumulate
in the sludge zone. The clarified liquid is generally collected over the entire
cross-section of the basin at the outlet zone and is discharged.
Inlet
zone
Settling
^ zone
u
u
Outlet
zone
u
Sludge zone
,
Exhibit 3.2: Four principal zones of a rectangular
continuous flow sedimentation basin.
Sedimentation theory is most easily defined using an "ideal basin," for
which the following assumptions apply:
(1) the direction of fluid flow is horizontal and the fluid velocity is
constant at all points in the settling region;
(2) the concentration of suspended particles of each size is uniformly
distributed over the vertical cross-section of the tank at the inlet
end; and,
(3) all particles which settle to the basin floor remain permanently
removed from the suspension.
In a rectangular ideal basin, the settling paths of all discrete particles
are straight lines defined by the vector sum of the horizontal velocity component
(U) due to fluid flow across the basin and the vertical settling velocity (Vs)
of the particle due to gravity (Exhibit 3.3).
All particles with the same settling velocity will move in parallel paths.
Under ideal conditions, a particle which starts at the top of the inlet zone and
settles to the tank floor at the junction of the outlet zone during its
theoretical detention period represents the tank discharge per unit surface area.
This is defined as the tank overflow rate, which represents the unit volume of
water flow per unit time divided by the unit area of the settling basin (Vo =
Q/A); it is the average upflow velocity of the basin.
12
^
^\- u
Q
H
h
Exhibit 3.3: Settling paths of discrete particles
in a rectangular sedimentation basin.
Any particle with a settling velocity (Vs) greater than the overflow rate
(Vo) will settle out of suspension. Other particles, for which Vs < Vo, will be
removed in the ratio Vs/Vo, depending upon their vertical position in the tank
at the inlet (Hazen 1904). This fractional removal (Fx) is defined by the
equation:
Fx -
Vo
Vs
QIA
The settling velocity analysis of a suspension may be presented as a
cumulative distribution curve (CDC), with the vertical axis (Cr) representing the
fraction of particles (f^) in the suspension with settling velocities less than
the corresponding settling velocity on the horizontal axis. [Note: This concept
is applied in the settling curves for those farms included in the study; see
Exhibit 4.7]. The fraction of particles removed completely then is equal to 1 -
f^; (i.e. those with Vs > Vo) . The fractional removal of those particles with
Vs < Vo is defined by the integral:
.f -I*
Vo
The overall removal (F) of particles in suspension in an ideal basin is
defined by the sum of the terms for settling velocities greater than and less
than Vo, the tank overflow rate:
''-(1-/0) + -^ o^•^'*4r
Vo
Therefore, the removal of suspended matter by sedimentation in an ideal
basin is a function of the basin surface area and is independent of tank depth.
Moreover, removal is a function of the overflow rate (Vo) and, for a given
discharge, is independent of retention time.
13
uACULTli.RE
3.2 Factors Influencing the Design of Sedimentation Basins
Sedimentation processes in actual basins are much more complex than the
process described for an ideal basin. In an actual tank the behaviour of the
settling particles is altered by turbulence and scour, which tend to re-suspend
particles; and by velocity gradients, density currents and short-circuiting,
which give rise to mixing and reduced removal efficiency.
3.2.1 Tank Velocity, Turbulence & Scour
In an ideal basin, the sedimentation of discrete particles is a function of
overflow rate, and is independent of tank depth and retention time. In theory
then, economics would suggest the use of infinitely shallow basins. Real basins,
however, are restricted to minimum tank depths based upon the maximum allowable
tank velocity (Uc) that can be maintained without excessive scour. In any basin
of fixed dimensions, and for a given overflow rate, the tank velocity varies
inversely with tank depth.
Unlike ideal basin theory, tank velocity is not uniform over the entire
cross-section of the basin. Due to boundary layer drag forces, horizontal fluid
flow is quickest in the centre of the tank near the surface, and slowest near the
tank walls and floor. Even in the best designed settling basins, the Reynold's
number (Re) will approximate 2 x 10^ to 2 x 10"*, however, laminar flow conditions
exist at Re < 1500 to 2000. Therefore, flow in settling basins is typically in
transition or turbulent, which results in the rapid and continuous mixing of the
fluid.
Since most particles remain undisturbed once they settle to the floor of the
basin, the effect of turbulence is limited to a delay in the sedimentation
process. This can be compensated for by lengthening the sedimentation basin in
accordance with the turbulent flow characteristics. In general, since an
increase in tank velocity results in increased turbulence, faster flowing tanks
must be longer, or of greater diameter, than slower flowing tanks to comparably
clarify the same wastewater stream.
Therefore, to achieve the efficiency of removal projected by theoretical
basin design, it is necessary to compensate for turbulence by increasing the
surface area of the settling basin. Compensation can be readily achieved by
applying the concepts presented by Fair and Geyer (1958). Using ratios based
upon sedimentation theory, a table of scale-up factors was compiled to enable
designers to compensate for turbulence by adjusting the overflow rate in the
principal design equations (Exhibit 3.4).
For example, assume that one is working with a settling basin of "good"
design (i.e. proper inlets, outlets, flows, area, etc.), and that a design
overflow rate of 40 m^/m^/d (0.046 cm/s) has been applied and, furthermore, that
this overflow rate is projected to yield 75% solids removal. The data in Exhibit
3.4 indicate that, in reality, the performance overflow rate is equal to 68
m^/m^/d (0.079 cm/s); i.e. 40 x 1.7. Consequently, the performance efficiency
will be less than the design efficiency; the extent of the reduction being
dependent on the shape of the solids settling curve. The net result is a
14
reduced, but more accurate calculation of the removal efficiency of the basin.
When applying these scaling factors, however, the characterization of basin
performance necessitates judgmental assessment, based upon the net sum of all
design considerations for both new and existing basins.
Exhibit 3.4
Ratios of required-to-theoretical surface areas of settling basins and
of actual -to-theoretical settling velocities of particles at different
removal efficiencies.
Basin Performance Removal Efficiency (%)
70 75 80 85 90 95
Very Good
Good
Poor
Very Poor
1.3
1.5
1.8
2.2
2.7
3.5
1.4
1.7
2.0
2.5
3.2
4.8
1.7
2.0
2.5
3.2
4.4
6.9
2.3
3.0
4.0
5.9
10
19
Source: Adapted from Fair and Geyer (1958)
The re-suspension of settled particles by shear forces generated along the
sludge surface by the horizontal flow of water through the basin is termed scour.
Scour reduces sedimentation efficiency by stirring up the settled particles and
reintroducing them to the water column. Camp (1946) defined the critical flow
velocity (Uc) necessary to initiate scour of settled particles:
Uc
(y) (ff) (s-1) (^
where: Uc = critical scour velocity (m/s)
B = constant = 0.06
f = friction factor (approx. 0.024]
g = gravitational acceleration (m/s")
s = specific gravity of particles (approx. 1.005)
D = mean particle diameter (m)
From this equation, it is evident that the critical flow velocity for
commencement of scour is independent of tank size and depth. The governing
factors are the friction factor and the mean size and specific gravity of the
particles. For aquacultural waste solids ranging from 0.25 to 1.5 mm in
diameter, approximate scour velocities equal 1.6 to 3.8 cm/s. This is in
15
Qanadian
7\quaculture
3ystems
agreement with Parjala (1984) who recommends a maximum horizontal fluid flow rate
of 2 - 4 cm/s in aquacultural settling basins. Conversely, Boersen and Westers
(1986) concluded that horizontal flow velocities between 10 and 40 cm/s are
required to prevent fecal and waste feed solids from settling in production
tanks.
Dobbins (1944) determined that since solids removal is not independent of
tank depth as indicated in ideal settling theory. The effect of depth is,
nevertheless, quite small. On average, a 50% reduction in depth will result in
approximately only a 5% decrease in removal of discrete particles. A 50%
decrease in volume due to reduced depth, however, can be compensated for by a 5%
increase in volume via increased surface area (Camp 1946). Thus, as economy
dictates, the tank depth should still be maintained as shallow as possible,
provided that scour-related problems do not arise.
3.2.2 Short-Circuiting & Tank Stability
Fluid flow through an ideal basin approximates plug flow, which means that
there is no variation of lateral velocity over the cross-section of the tank.
That is, a cross-section of water entering the settling zone from the inlet zone
travels as a solid body throughout the length of the tank. Displacement is
steady and uniform, and each unit volume of fluid is detained for the theoretical
retention time of the basin, R, = Vol/Q.
In reality, however, a portion of the flow exits from the tank after a
shorter interval than the theoretical retention time while some of the water is
retained longer. Short-circuiting is inherent to all tanks due to varying tank
velocities in different stream paths caused by thermal-, density-, eddy- and
wind-induced currents.
When the velocity of water entering the basin is too great, eddy currents
are established by the inertia of the flowing water and mixing results. In large
basins with long retention times, non-uniform heating of the water by the sun can
result in vertical convection currents. Density currents typically occur when
fluids of differing densities are combined in a quiescent tank or pond. The
heavier fluid settles to the bottom of the basin and flows quicker than the
lighter, top fluid. Fluid density differences may result from differences in
temperature (solar heating), salinity, or suspended solids concentration. In
most settling basins, however, fluid flow is generally sufficiently turbulent to
disrupt density currents.
Wind shear over large tanks and ponds can create substantial surface
currents which cause serious short-circuiting. The surface drift current is an
almost constant ratio of the wind speed at 10 metres and this ratio varies
minimally between researchers, and with Reynold's number (Plate 1970). A value
of 1% to 3% of the 10 metre sustained wind speed is typical. Therefore, a wind
speed of only 180 cm/s (approx. 6.5 km/hr) can induce a surface current with a
speed of approximately 3.6 cm/s, which is of similar magnitude to conventional
tank velocities.
16
Short-circuiting can also result from the presence of dead zones in a
settling basin, which most commonly result from improper design of inlet and
outlet structures (see Section 3.2.3). Dead zones are regions within a basin in
which the liquid plays little or no part in the displacement through the tank.
In fact, this occurrence can be readily observed on inappropriately designed
basins during winter months. Where water flow is slowest (usually around pond
edges or behind baffles and booms) the surface of the basin freezes over. Where
flow is swift (such as along the direct line between point source inlet and
outlet pipes), open water remains. This results in an effective tank volume, and
area, which is less than actual tank dimensions. Consequently, a percentage of
the influent stream passes through the basin in a shorter time than the
theoretical detention time while some water remains in the basin for a longer
time.
Thus, the net effect of short-circuiting on sedimentation processes is
reduced efficiency. The influent flow rate to any settling basin is tuned to the
basin overflow rate (Vo) and is relatively constant; however, short-circuiting
streams of quicker flow have an overflow rate which is greater than the design
rate. As a result, any particles in that stream having a settling velocity less
than the stream overflow rate, but greater than the basin overflow rate, will not
settle as they otherwise would have. Conversely, slow moving or stagnant areas
must also exist, which further reduce the effectiveness of the settling process.
Control of short-circuiting, therefore, is dependent upon dissipation of inlet
velocity, protection of basins from wind shear and uneven heating, and reduction
of density currents. Foremost, however, proper inlet and outlet design is
critical for minimizing short-circuiting in settling basins.
3.2.3 Inlet & Outlet Design Considerations
As discussed, improper inlet and/or outlet design can result in short-
circuiting and reduced sedimentation efficiency. Placing a single pipe at each
end of a basin is the simplest design for inlet/outlet structures and,
unfortunately, it is perhaps the design most frequently used in aquaculture
today. This type of point-source inflow and outflow produces extremely poor flow
characteristics for sedimentation. Short-circuiting is severe as the influent
stream flows directly through the basin, at high velocity, toward the discharge
pipe. Consequently, a large portion of the fluid passes through the basin in a
matter of minutes, although the theoretical detention period may be a matter of
hours. To be effective, the design of inlet and outlet structures must address
the specific function and constraints of each process.
Inlet Design. When designing inlet structures, Ingersoll et al . (1956)
suggest that the following factors must be considered:
(1) The influent stream should be introduced evenly across the entire
cross-section of the settling zone.
(2) All flow through the settling zone should begin in even, horizontal
paths.
17
' ANADIAN
Vguaculture
Systems
(3) The influent velocity to the settling zone should be slow enough to
prevent excessive turbulence and mixing.
The influent to a settling basin usually flows through a pipe or channel
which is smaller in cross-section than the basin. The flow, therefore, must
diverge upon entering the inlet zone, often producing separation and turbulence
in the process. A number of inlet structures have been designed to distribute
the influent flow to a settling basin so that the above-listed factors are
satisfied. Giles (1943) presents a thorough historical review of such inlet
designs. Those inlet designs which are most inappropriate, and which should not
be used in settling basins, are illustrated in Exhibit 3.5.
Screen-type, or orifice-plate, baffles have also been applied to achieve
complete flow dispersion over the entire width and depth of settling tanks.
Fouling of the screen, however, often prohibits the use of such baffle systems
with many waste streams. Envirotech Corp. developed and marketed the Modular
Energy Dissipating Feedwell (MEDF) , a honeycomb of small tubes mounted across the
basin inlet. The honeycomb creates a laminar flow pattern with uniform
velocities across the inlet (US EPA 1975). Although effective in distributing
the inlet flow, serious plugging of the tubes led to discontinued use of the
design.
Submerged weirs, extending across the entire width of a basin, have been
found to provide an optimal balance between practicality and flow characteristics
for sedimentation. The first submerged weir (Exhibit 3.6) was introduced by
Hubbel (1934). Hubbel's design utilizes an upward, diverging flow to the weir
crest which greatly reduces flow velocity. The submergence of the weir serves
to reduce velocity even further. As well, this weir features a wide, goose-neck
crest, which is effective in utilizing any residual approach velocity to
distribute the fluid over the vertical depth of the basin. This type of weir is
highly effective in distributing the flow over the entire cross-section of the
basin with a minimal inlet velocity.
A simplified yet effective version of Hubbel's weir is now commonly used to
introduce wastewater streams into sedimentation basins (Exhibit 3.6). Flow
enters the settling zone from a header tank at the water level of the settling
basin. To smooth the flow, the weir crest, like Hubbell's goose-neck, is
typically 30 to 60 centimetres wide and has chamfered edges. The top of the weir
should be submerged approximately 15% of the basin depth.
Outlet Design. The discharge of water from settling basins has changed
little over the years. Traditionally, an outlet weir extends across the width
of the basin at the end opposite the inlet. In circular tanks, the outlet weir
extends around the outer periphery; the inlet is at the centre. Clarified
effluent overflows into the weir trough and is discharged. The weir rate (volume
of water discharged per unit length of weir per unit time) governs the length of
the outlet weir. It is critical that the weir edge be level to assure a uniform
discharge rate across the entire weir length.
For weirs which are long in relation to the flow (i.e. having a low weir
rate), a saw-toothed or V-notch edge is necessary for uniform discharge along the
18
•veir length (Fair and Geyer 1958). Such low-flow systems, however, are virtually
non-existent in intensive salmonid culture operations. Conventional municipal
wastewater treatment design standards recommend that weir rates not exceed 186 -
248 nr/d per metre length of weir (ASCE 1959; US EPA 1975). -"or aquacultural
operations, where the solids component of the effluent tends to be heavier and
more viscous than conventional municipal wastes, a weir rate of 372 m-7d/m has
been recommended for design purposes (Mudrak 1981).
Caution must be exercised in designing outlet weirs due to the updraft
effect generated by such structures, as depicted by the flow net diagram in
Exhibit 3.7. As the water approaches the outlet, the horizontal flow velocity
increases due to a reduction in cross sectional area and the lifting velocities
of stream lines flowing toward the surface weir permit the escape of solids.
This is quite evident to the eye at overflow standpipe discharges in tanks,
raceways and settling basins. As a result, a certain portion of the tank volume
becomes ineffective for settling since suspended particles entering this section
become entrained in the effluent flow. The ineffective volume is defined by a
concave arc, which extends from a point at the surface of the basin upstream of
the weir to a point below the weir, but not necessarily penetrating the full
depth of the basin.
Mohlman et al . (1946) developed the following equation to calculate this
ineffective volume in a settling basin:
Ineffieettve Voiame - k ( ^)
where: k = constant = 0_.134
q = weir rate (m^/d/m)
Vs = settling velocity of particles (m/d)
This equation shows that the magnitude of the entrainment effect is
proportional to a ratio of the weir rate to settling velocity. Most importantly,
however, this is a quadratic equation, meaning that should the weir rate double,
the ineffective volume will increase four-fold! Consequently, for every unit
increase in the weir rate (q) , the corresponding decrease in settling efficiency
is disproportionately greater.
Ingersoll et al . (1956), however, contest the value of k, the constant in
the above equation, stating that the only proper way to determine the magnitude
of entrainment is to characterize the flow net in each specific application.
Fortunately, for tanks which are long relative to their width (as is generally
the case in aquaculture) , the entrainment effect is confined to an upstream
distance equivalent to only about 1 to 2 multiples of the basin depth. Since the
updraft effect of an outlet weir reduces the effective length of a settling
basin, proper design requires that settling basins be lengthened proportionately
to compensate for this reduced efficiency. An additional length requirement of
1.5 to 2.0 times tank depth is appropriate in most cases. Weir length should be
maximized in all applications.
19
(^ArjADIAfJ
ApUACULfURE
S 'STEMS
3.2.4 Principal Design Considerations
In summary then, for the removal of solids from aquacultural effluents,
sedimentation basins should be designed with the following factors in mind:
(1) Overflow rate (Vo) , that is, the settling velocity of the smallest
particles to be theoretically 100% removed, must be the foundation
of basin design.
(2) Detention periods are, of themselves, immaterial; in spite of current
industry guidelines which still state retention times (Daley 1989).
The detention periods result from the design and are not a basis for
design.
(3) Basin depth should be as shallow as possible, providing that scour
does not interfere with performance.
(4) Turbulence should be compensated for in basin design equations.
(5) Inlet and outlet structures should be specifically engineered in
accordance with the characteristics of the basin dimensions and water
flow rates.
20
PLAN
LONGITUDINAL
U
/
w
\
h
u
H
Exhibit 3.5:
Examples of inlet structures which are inappropriate for effective
introduction of wastewater into settling basins, a. Straight pipe,
b. Downward elbow, c. Straight pipe with target baffle.
21
ANADIAN
QUACULTURE
YSTEMS
PLAN
LONGITUDINAL
\^
3
aooa □
MM
U
1 1
h
Exhibit 3.5: (cont'd)
Examples of inlet structures which are inappropriate for effective
introduction of wastewater into settling basins, d. Reversing-flow.
e. Multiple channels with target baffles, f. Free-fall weir.
22
A
=- U
/l\/l\
Exhibit 3.6:
Inlet structures designed for effective introduction of wastewater
streams into settling basins. A. Hubbel's submerged goose-neck weir.
B. Submerged wide-top weir with chamfered edges.
23
%
ANADIAN
QUACULTURE
YSTEMS
FLOW NET
PARTICLE SETTLING PATHS
Exhibit 3.7:
Two-dimensional representation of outlet weir effects on the fluid flow net
and on particle settling paths in a rectangular settling basin. The
updraft effect increases the required basin length for settling of particle
A from a to a'. Particle B should settle at point b, however, it becomes
entrained in the effluent and is discharged.
Source: Adapted from Ingersoll et al . (1956)
24
4.0 CHARACTERIZATION OF AQUACULTURE OPERATIONS & EFFLUENTS
4.1 Purpose & Objectives
Proper engineering design is essential if aquacultural wastewater treatment
facilities are to be effective. Too often, facility design is based upon
conventional municipal wastewater treatment design, and the specific
characteristics of fish hatchery effluents are not considered. Waste
characteristics, however, are the most important factor governing engineering
design. Mudrak and Stark (1981) report that the solids component of aquaculture
effluents is denser, faster settling, and tends to form a heavier, more viscous
sludge than domestic municipal wastes. When these unique characteristics are not
considered in facility design, the resultant treatment unit is typically
oversized, expensive, and inefficient.
Therefore, settling analyses of the solids component of the wastewater
stream must be an integral part of the design process. Surprisingly, however,
in the intensive salmonid aquaculture industry, a comprehensive compilation of
effluent solids sedimentation data have yet to be compiled and, consequently, the
relatively poor performance of effluent treatment operations at such facilities
should come as no surprise.
Since a wide variety of particle shapes and sizes exist in most suspensions,
calculation of the settling velocity of discrete particles based upon Stokes'
Law, or other theoretical equations, is impractical. By conducting a settling
analysis of the suspension to be clarified, however, the overall gravitational
removal of particles from suspension can be determined for any given overflow
rate. The procedure for analysis of suspensions has been somewhat standardized,
and is described by several authors (Camp 1946; Fitch 1957; Al-Layla et al .
1980) .
Geometric models, although useful in evaluation of hydraulic short-
circuiting, cannot be used to measure settling characteristics. Instead, a
settling column having an inside diameter greater than 13 cm is required;
narrower columns are not reliable due to boundary layer resistance introduced by
column walls. To avoid extrapolation of data, the depth of the column should be
greater than the depth of a full-scale settling basin in the desired application.
Sample ports, where an aliquot of the suspension can be drawn off, are spaced
along the side of the column.
Three additional components of fish culture processes must also be assessed
in the quest for appropriate aquaculture wastewater treatment systems: (1)
analysis of pertinent water quality parameters throughout the culture facility;
(2) the design and layout of culture facilities; and (3) standard management
practices. The field portion of this investigation was designed to collect these
data from commercial, hobby, and provincial fish culture facilities throughout
Ontario.
By simultaneously collecting production, facility design, and water quality
data, answers to several key questions can be obtained:
25
Qanadian
t^quaculture
3 ''STEMS
Q. WHAT is the impact of intensive fish culture operations on water quality?
Q. WHICH production and facility design factors contribute to the
deterioration of water quality?
Q. HOW do the suspended solids behave under quiescent settling conditions?
Q. WHAT are the values of critical design components at existing effluent
treatment facilities on fish farms and how are they related to the
operating efficiency of these facilities?
Several previous studies have addressed the first question, however, it is
the inter-relationship of answers to all four questions that makes this
investigation unique and, furthermore, that will provide practical solutions for
effluent management at fish culture facilities. Such answers will identify those
variables which facilitate the segmentation of fish culture operations into
discernible groups, based upon similarities in design and operating
characteristics of the production unit. Past research has already shown that
different farms can have substantial differences in solid and dissolved waste
output (Ont. Min. of Environment 1982; Hilton and Slinger 1984; Merican and
Phillips 1985). The compilation of a broad data base will enable the design of
effective effluent treatment facilities based upon the characteristic traits of
a given farm and the associated production features of the group to which that
farm belongs.
Protection of the receiving water course must be a prime consideration in
the design, construction, and operation of any aquaculture facility.
Standardization of design criteria for the various types of culture facilities,
and the development of proper management practices, will enhance the responsible
use of aquatic resources by aquaculturists. Moreover, standardization of design
principles will provide confidence to decision-makers in their evaluation of
aquacultural wastewater treatment facilities, thereby permitting the
identification of the underlying cause(s) of non-compliance; i.e. whether the
problem hinges on a design factor or poor management practices.
4.2 Field Survey & Water Quality Sampling
A field monitoring program was developed during the summer of 1989 through
the cooperative efforts of The Ontario Trout Farmers' Association (OTFA) , the
Ontario Ministry of the Environment (OME) , the National Research Council of
Canada (NRC), and the Faculty of Engineering Science at The University of Western
Ontario (UWO) . The survey goal was to schedule site visits to approximately 60
to 90 commercial, community, government, and private fish farms across Ontario.
For practicality, Southern Ontario was the principal study area. Moreover, this
region has the largest concentration of fish farms in the province. Over the
course of the field season, site visits were scheduled to fish farms which were
progressively further from the operating base in London, Ontario.
26
A sumrner student from the Faculty of Engineering Science at UWO was hired
and trained to conduct the field portion of this investigation, which included:
- contacting fish farm operators and scheduling site visits;
- collecting production and operations data from respondent farm
managers;
- conducting on-site sedimentation tests at fish culture
facilities;
- collecting representative water chemistry samples and delivering
samples to the OME laboratory for nutrient analyses;
- conducting suspended solids filtration analyses in the
engineering laboratories at UWO; and
- maintaining project records.
Data collection was organized into three principal categories:
A. PREDICTOR VARIABLES
(1) Production and Management Variables
B. RESPONSE VARIABLES
(2) Water Chemistry Variables
(3) Solids Settling Variables
Predictor variables consist of those production and management practices,
and specific facilities design features, which characterize the fish culture
operation. These data were collected to illuminate potentially causative
agent(s) of deteriorating water quality. Response variables are those parameters
of primary concern from a water quality and effluent management perspective. A
standard data sheet was compiled to facilitate the collection of similar data
from all operations included in the study (see Appendix I).
4.2.1 Predictor Variables
Predictor variables were divided into three groups to facilitate data
management.
Operations data provide information on the size of the operation in terms
of annual production capacity, the brand and type of feed used, and the source
and total volume of water used.
Since many fish culture operations utilize different culture units (i.e.
circular tanks, raceways, troughs, ponds, etc.), an on-site decision was made by
the investigator to conduct detailed sampling within a particular culture unit.
Sampled Unit data provide information on the type and dimensions of the culture
unit sampled, the total weight (or number) and size of fish in the unit, the
method of feeding the fish (i.e. hand, demand feeder, automatic feeder) and the
temperature and volumetric flow of water to the unit. As well, if water re-use
was being utilized, the percentage of the total flow which was comprised of re-
used water was also recorded.
27
O A'lAOlAM
"aGUACULTUFIE
According to the findings of LIMA Engineering Ltd, (1988), 26% of all fish
culture operations in the province utilize some form of effluent treatment. At
those operations where treatment facilities were in operation at the time of the
site visit, data were collected to permit an evaluation of the design parameters
and operating efficiency of the treatment unit. Effluent Treatment data provide
information on the treatment technology in place, the dimensions of the treatment
unit, the hydraulic loading rate of the unit, and the design of inlet and outlet
structures.
For most farms, sketches of the overall farm layout and process water flow
paths were compiled to assist with data compilation and interpretation.
4.2.2 Response Variables
Response variables are divided into two distinct groups: water chemistry
variables and solids settling variables.
At each fish culture operation. Water Chemistry Variables were collected
in standard 1 litre sample bottles. The location and number of these samples
varied somewhat with the nature of the culture facility; however, the influent
source water and the final farm discharge were sampled at all operations.
Samples were collected at the end of the "sampled production unit" and prior to
and following effluent treatment, if the latter facilities were in operation.
The water quality parameters that were analyzed included total Kjeldahl
nitrogen (TKN) , total ammonia (NH3) , nitrite (NO,), nitrate (NO3) , total
phosphorus (TP) , dissolved phosphorus (DP), and total suspended solids (TSS) .
These parameters were targeted as being of most concern at a meeting of OME and
OTFA representatives during the early planning phases of the investigation.
Except for TSS, all water quality analyses were conducted by the OME at their
analytical laboratory in London, Ontario. To ensure the confidentiality of
individual fish culture operations included in the study, all samples were
identified by a "Farm Number" only. Moreover, only the principal researchers
have a record of the farm numbers and farm names.
To facilitate the segmentation of farms into discernible groups, and to
enable the definition of principal effluent treatment design criteria, Solids
Settling Data were also collected. A settling column measuring 220 cm high by
20.4 cm inside diameter was used to conduct on-site settling tests. Three
sampling ports, located at distances of 120, 160, and 200 cm from the top of the
column, were used to draw off samples for solids content analysis. Based upon
past experience, two shallower ports (40 and 80 cm) were not used during this
investigation.
Settling tests were conducted by filling the column to capacity with
approximately 45 litres of water from the production/treatment unit. At most
facilities, it was necessary to re-suspend solids within the unit to gather
samples having a sufficient concentration of solids for the analysis. Once in
the column, the suspension was gently agitated by introducing compressed air
through a sparger at the bottom of the column. Agitation serves to create a
28
gentle rolling action in the column, thereby creating a uniformly mixed
suspension from which to draw initial samples from each sampling port.
Follc-.ving collection of the initial samples, the air supply was
discontinued and the column contents -vere permitted to settle under quiescent
conditions. Samples were collected at timed intervals, which spanned a one-hour
period. Past experience by the principal investigator nas indicated that longer
time periods (up to three hours) do not improve the results. All samples were
vacuum filtered through Type 11306 Sartorius Membrane Filters, having a pore size
of 0.45 \i\n, to determine the total suspended solids concentration.
The sedimentation analyses provide information pertaining to the shape of
the cumulative solids settling curve, which was discussed in Section 3.1. The
slope and intercept of the curve(s) provide information pertaining to the value
of Cr at various design overflow rates (settling velocities), and are thus vital
to the design of treatment facilities.
Settling data were not collected at all farms. Pond operations were
largely excluded from this portion of the investigation due to the complexities
introduced by suspended organic and inorganic matter naturally present in pones.
Additionally, settling tests were not conducted at farms where solids removal
(via vacuuming, flushing, etc.) had not been conducted within the past seven
days. Parjala (1984) suggests that solids be less than one week old to obtain
reliable results. Generally, most operators had cleaned solids from their
facilities within the four days prior to the site visit.
4.2.3 Scope and Limitations
This field investigation was designed to enable the collection of data from
a number of fish culture operations to provide a sufficiently broad data base to
permit the segmentation of operations into discernible groups based upon
production and management characteristics. Moreover, it was designed to maximize
the co-operative efforts of all parties involved. To conduct a survey of this
magnitude, over a geographic region the size of Ontario, necessitates that the
time spent at individual sites be utilized efficiently.
Although composite sampl ing and/or multiple sampl ing programs offer certain
advantages, the time required for such activities would greatly limit the total
number of operations which could be visited, given the size of the geographic
region surveyed in this study. Instead, grab samples were collected at each
farm. Similarly, the utilization of filtered effluent samples and/or re-
suspended solids for settling tests also permitted a larger number of sites to
be visited during the sampling period. Nevertheless, additional insight may be
provided through collection of solids in sediment traps for a 1 to 4 day period
prior to conducting the sedimentation test. To dispel any argument regarding the
solids sampling procedures applied in the field study, statistical comparison of
settling data from re-suspended and filtered samples revealed no difference in
settling behaviour. Similarly, no such differences were identified by CAS in
studies conducted in New Brunswick.
29
Qanacian
^^uaculture
Despite these limitations, the study did achieve its objective of obtaining
data from a number of fish culture operations; a data set which may, in fact, be
unparalleled in the industry. Additionally, the nature of the data collected at
all operations is essentially identical, chereby, incorporating uniformity into
the investigation and subsequent interpretation of the results.
4.3 Data Analysis & Results
Between May 14 and September 2, 1989, a total of 82 fish culture operations
were visited throughout Southern Ontario. From this sample, full data sets,
including settling data, water chemistry data, and production and management
data, were obtained from 50 farms. Partial data sets were obtained from the
remaining 32 farms. The latter are predominantly pond culture operations and,
therefore, settling data were not collected. The following presentation of data
analyses and results is largely based upon those 50 farms with complete data
sets. Data from all farms visited are presented in Appendix II.
In total, 25 continuous variables (12 predictor variables; 5 solids
settling variables; and 7 water chemistry variables) and 4 class, or non-
continuous, variables (farm type; feed brand; feed type; and feeding method) were
recorded. These variables are defined in Exhibit 4.1.
Prior to data analyses, all continuous data were transformed to natural
logarithms to conform with assumptions of the various statistical methods. For
each observation (z) the transformation equation is as follows:
transformed value of z = In ((z+1) - minimum z value)
By subtracting the minimum observed value from each variable, the data are
adjusted for differences in measurement scales and units; adding 1 eliminates the
potential for negative values associated with the logarithmic transformation of
numbers less than 1.0.
Since one of the underlying goals of this investigation is to reveal cause-
and-effect relationships between fish culture practices and deterioration of
water quality, Pearson's (parametric) and Spearman' s (non-parametric) correlation
analyses were conducted on the entire data set to examine the nature of the
relationships between variables. Key relationships between predictor variables
and response variables were scrutinized. As suspected, however, very few
statistically significant relationships were revealed using this statistical
approach, suggesting that considerable variability exists among intensive
salmonid culture operations. As a result, data for the various farms were
examined to determine whether different types of farming operations were evident.
Due to the physical layout and construction of intensive fish culture
operations in Ontario, it is relatively simple to separate farms on an a priori
basis. That is, flow-through raceway facilities are distinct from oval raceways,
and similarly from flow-through circular tanks, ponds, and intensive
recirculation units. Consequently, data were re-organized into the following
five groups prior to subsequent analyses.
30
Group 1 Standard raceway operations in which water enters at one
end, flows through the length of the unit, and is
discharged at the opposite end.
Group 2 Oval raceway operations resembling modified Burrows
ponds in which two parallel raceways share a common
dividing wall, which stops short of either end of the
unit. Water circulates around the unit such that the
flow on each side of the central wall is in opoosite
directions. High-volume low-head pumps are commonly
used to maintain circulation within the tank. Overflow
drains are usually located at one or both ends.
Group 3 Circular tank operations consisting of concrete, steel,
or fiberglass tanks having a circular flow of water from
the outer edge toward a central drain.
Group 4 Pond culture operations in which fish are raised in
earthen excavations, generally square to rectangular in
shape and having a water inlet and outlet at opposite
reaches of the pond.
Group 5 In all groupings, it is usually difficult to categorize
certain individuals. Rather than attempting to force
decisions which may compromise results. Group 5 was
created to include any farms which did not fit into
Groups 1 through 4. (Others)
Using this a priori approach, the 50 farms were re-classified as follows:
Group (n)
1 21
2 11
3 12
4 3
5 3
Having re-classified the data into separate types of facilities based on
configuration, it was necessary to determine whether this classification strategy
was statistically meaningful, based upon the observed variables. Using
multivariate analysis of variance (MANOVA) , the five groups were, in fact, found
to be significantly different.
Multivariate Analysis of Variance
Test Criteria & F Approximation
Statistic Value F Num DF Den DF Prob.
Wilks' Lambda 0.007388 2.1007 100 85.5 0.0003
31
Qanadian
TlQUACULTURE
3ystems
From a univariate perspective using standard analysis of variance (ANOVA) ,
significant differences between groups were identified in both production
variables and water chemistry variables; no differences were found in the solids
settling data. Differences in intensity, production area, standing crop,
density, water flow, water reuse, and hydraulic retention time are responsible
for distinguishing at least one of the five groups. Similarly, groups were found
to differ with respect to all seven water chemistry variables (i.e. pre-treated
effluent quality) .
To confirm this finding, and to visually present the magnitude of the
difference between the five groups, a canonical discriminant analysis (CDA) was
completed. CDA uses linear combinations of the observed variables in a
multivariate approach to construct composite variables, which are comprised of
components of all 25 observed variables. Canonical variables, or "axes," serve
as the basis for classifying farms into one of the groups to confirm group
membership and thereby identify unusual individual farms that fail to fit into
the a priori grouping. One canonical axis is developed to maximally separate
each pair of groups based upon the degree of variation inherent in the original
data (for 5 groups, 4 canonical axes are generated). The analysis identifies
variables which separate groups in combination as opposed to separating them on
a "one variable at a time" basis (Campbell and Atchley 1981).
The CDA did indeed support the finding of the MANOVA, indicating that
significant differences (P<0.025) exist between the groups. The CDA results
(Exhibit 4.2) illustrate that differences are significant along the first two
canonical axes, which account for 79.6% of the variance between groups. The
third and fourth axes, which account for the remaining 20.4% of the total
variance, were not significant (P>0.320).
The five groups are plotted on the resultant canonical axes 1 and 2 to
visually display the five groups on X-Y co-ordinates (Exhibit 4.3). This graph
illustrates that considerable overlap exists between Groups 1 and 3, standard
raceways and circular tanks. The degree of overlap appears to be greater along
axis 2 than on axis 1. Oval raceways (Group 2) are significantly removed from
Groups 1 and 3 along axis 2, and from Group 3 (circular tanks) along axis 1.
Groups 4 and 5 (ponds and "others") are separated from all other Groups, and each
other, along axis 1. The implications of this pattern is revealed by examining
the canonical coefficients presented in Exhibit 4.4.
Standardized Canonical Coefficients are used to derive the actual data
points which are plotted in Exhibit 4.3; they are the raw canonical coefficients,
standardized to a standard deviation equal to unity. The total, within, and
pooled canonical structures, however, reveal the correlations between the
original data and the canonical axes. More importantly, each structure presents
a different aspect of the inter-relationship between and within groups.
The total canonical structure assumes that no groups exist. Generally,
this summary reveals few strong correlations, however, significant correlations
are indicative of important variables across all five groups. This analysis
suggests that along canonical axis 1, the axis which maximally separates Groups
5 (others), 3 (circular tanks), and 2 (oval raceways), segmentation is
principally due to differences in intensity of production, total water flow,
32
water reuse, hydraulic retention time, and nitrate concentration. Along axis 2,
which separates oval raceways from all other groups, principal segmentation
variables include production area, standing crop biomass, total water use, and
water reuse from a- production perspective. Moreover, from a water chemistry
perspective, axis 2 reveals differences with respect to ammonia, total Kjeldahl
nitrogen, nitrite, total phosphorus, and dissolved phosphorus (Exhibit 4.4).
The between canonical structure compares the five groups collectively,
using the mean, or centroid, of each group as the basis for comparison rather
than the individual observations within each group (i.e. n=5 versus n=50) . This
approach identifies those variables which act to distinguish groups and is
similar to MANOVA in that multi-variable interaction is used to reveal
differences. The high correlation coefficients for the between canonical
structure along both axes in Exhibit 4.4 suggest that nearly all of the measured
variables contribute to group segmentation. In fact, density is the only
variable which is not highly correlated on either canonical axis, suggesting that
the between-group differences are not attributed to this variable.
Lastly, the within canonical structure questions the degree of variability
for each variable within each group and across groups. That is, it considers
whether the role of one variable in one group is similar to the relationship of
that same variable in other groups. High correlations suggest that a
relationship exists within groups and across groups. The data in Exhibit 4.4
show no strong correlations along axis 1; however, area, total water flow, water
reuse, ammonia, total Kjeldahl nitrogen, total phosphorus, and dissolved
phosphorus are correlated along axis 2. For these variables, the correlation
that exists within each group is similar to the correlation that exists between
groups.
Groups 4 and 5, that is pond production and "other" facilities, are of
limited significance to the objective of this investigation. The use of ponds
continues to decline in the industry due to their inherent inefficiency from
management, harvesting, and product quality perspectives. Moreover, this group
is largely unrepresented in the data since settling tests were not conducted at
most facilities. "Other" facilities are also of little meaning since, from a
categorical viewpoint, these farms are essentially misfits. Each of these three
operations did not readily fit into one of the other four groups, and each was
placed into Group 5 for different reasons. As such, there are no representative
characteristics of Group 5 farms in contrast to the farms in other groups.
Consequently, Groups 4 and 5 were removed from the data set for subsequent
analyses. The small size and overall irrelevance of these groups only introduces
unnecessary variability into the analyses which can potentially misdirect the
interpretation of results.
Again, significant differences between Groups 1, 2, and 3 were confirmed
by MANOVA and ANOVA. Furthermore, the impact of standard raceways, oval
raceways, and circular tank production facilities are more pertinent to the
investigation.
33
OArJADIAN
TXquaculture
3 VST EMS
Multivariate Analysis of Variance
Test Criteria & F Approximation
Statistic
Value
F
Num DF
Den DF
Prob.
Wilks' Lambda
0.047601
2.4367
50
34
0.0037
ANOVA results, including Student-Newman-Keuls multiple range testing, are
presented in Exhibit 4.5. Among the predictor variables, significant differences
(P<0.007) between groups are associated with production area, standing crop,
water flow, water reuse, and hydraulic retention time. With respect to
production area and water reuse, the mean values for oval raceway operations are
significantly greater from both standard raceway and circular tank operations
(P<0.0010). The latter two groups are not different with respect to these
variables. Hydraulic retention time was found to be significantly greater in
both oval raceways and circular tanks than in standard raceway operations
(P=0.0064). The total volumetric water supply to each of these three groups was
found to be different (P=0.0001; oval raceways > standard raceways > circular
tanks) .
Five of the seven water chemistry variables were found to differ
significantly between the 3 groups (Exhibit 4.5). Only the concentration of
nitrate and total suspended solids produced in fish culture facilities did not
differ (P>0.4197). Interestingly, however, for the other five variables (i.e.
TKN, NH,, NO3, TP and DP) oval raceway operations were found to be different from
both standard raceways and circular tanks (P<0.0295). In all cases, the
concentration of these five nutrients is greatest in oval raceways. Although the
mean values for circular tank operations are consistently larger than the mean
values for standard raceways, these differences are not significantly different
(P>0.05).
Canonical discriminant analysis was used to further define the relationship
between these three groups and, again, the CDA results confirm those of the
MANOVA and ANOVA analyses. Since three groups are being compared, two canonical
axes are generated. Only axis 1 is significant (P=0.0037), accounting for 76.8%
of the variability between the 3 groups (Exhibit 4.6).
These CDA results are graphically presented in Exhibit 4.7. It is apparent
that considerable overlap remains between Groups I and 3 (standard raceways and
circular tanks) along axis 2, the axis which presents no significant difference
(P=0.2250). Along axis 1, however, it is evident that Group 2, oval raceways,
is significantly different from Groups 1 and 3.
Examination of the total, between, and pooled within canonical structures
for this analysis lends insight into the source of these differences (Exhibit
4.8). The total canonical structure indicates that strong correlations exist
across several variables, including the port slopes, production area, standing
crop, total water flow, water reuse, and five of the seven water chemistry
variables, suggesting that these variables contribute strongly to group
segmentation. The between canonical structure suggests that group differences
are not related to fish density. Additionally, however, the concentration of
total suspended solids produced during fish production is also eliminated as a
34
source of group difference. Lastly, the pooled within structure re-confirms that
total water use, water reuse, and nitrate production display patterns within each
group which are similar (Exhibit 4.8).
Having identified that the fundamental design of fish production facilities
is indeed a basis for separating the groups, we then examined individual groups
to further quantify and qualify the impact of fish production practices and
system design on the quality and treatment of discharged effluent. Correlation
analysis is a tool for revealing relationships between the observed variables.
Both Pearson's (parametric) and Spearman's (non-parametric) correlation analyses
were used to resolve patterns in the transformed data sets for Groups 1,2, and
3 individually.
Differences between Pearson's and Spearman's analyses revealed that some
of the relationships were not linear. Consequently, Spearman's correlation
analysis was considered to be more appropriate. This technique compares data on
a rank order basis and provides a more conservative interpretation of results.
Tables of Spearman correlation coefficients and the corresponding statistical
probabilities are attached in Appendix III.
A number of the predictor variables (i.e. production area, depth, flow,
density, etc.) were highly correlated; this was to be expected since many of
these variables are inter-related by design. For example, the carrying capacity
of an operation (standing crop) is generally related to the available supply of
water. Similarly, the solids settling data and water chemistry data tend to be
highly correlated with their own kinds. More importantly, however, we must
understand the relationships between the predictor variables and both solids
settling variables and water chemistry variables to properly assess the impact
of aquaculture operations on water quality, and thereby facilitate the design of
effective effluent treatment units.
With respect to solids settling data, few statistically significant
correlations exist in any of the three groups (Appendix III). Furthermore, among
those relationships that are significant, little meaning can be inferred because
the correlated variables have little relevance from a design or management
perspective. This suggests that, individually, none of the predictor variables
are correlated with solids settling behaviour.
In comparison, however, a number of meaningful relationships have been
identified between production and water chemistry variables. In standard
raceways, oval raceways, and circular tanks, increases in intensity and fresh
water intensity are significantly correlated (P<0.097) with an increase in the
concentrations of TKN, TP, and/or DP in the effluent (Exhibit 4.9). In circular
tanks, these same two production variables are also positively correlated with
the effluent concentration of NHj. The depth of production facilities and the
weight of the standing crop of fish are correlated with the concentration of TKN
in effluent waters in standard raceways and circular tanks. Increases in crop
and density are highly correlated (P<0.048) with increases in effluent TP in
circular tanks as well. Weak positive correlations (P<0.099) exist between
average fish size and TKN and NO^ in standard raceways and circular tanks. There
is also evidence suggesting a positive correlation between average fish size and
TP concentration in circular tanks.
35
QArjADiAri
7\quaculture
gVSTEMS
Once again, similarities between standard raceways and circular tanks, and
dissimilarities between these two groups and oval raceways, are revealed (Exhibit
4.9). Depth and average fish size correlate with on water quality in both
standard raceways dnd circular tanks, yet they do not correlate in oval raceway
operations. Additionally, total flow and water reuse are not correlated in
standard raceways and circular tanks, but they display positive correlations in
oval raceways. The production variables for which no meaningful correlations
were identified across all three groups include hydraulic retention time, annual
production, area, and temperature. Conversely, intensity and fresh water
intensity are most highly correlated across all groups.
The objectives of this investigation clearly suggest a need for the
establishment of design criteria to facilitate solids removal from fish culture
process water. The data analyses, however, have not yet characterized solids
settling patterns. As discussed in Section 3.1, settling data can be plotted to
yield a cumulative distribution curve, which is a principal tool in the design
of sedimentation basins. Such data were collected at 50 farms during the field
sampling portion of this study. The data from the 44 operations comprising
Groups 1, 2, and 3 have been analyzed using regression analysis, and compared
using an analysis of covariance (ANCOVA) .
When all of the transformed settling curve data from each farm are pooled
within the three respective groups, significant regressions are obtained
(P<0.001). The un-transformed regression lines (settling curves) for standard
raceways, oval raceways, and circular tanks are plotted in Exhibit 4.10. The
data obtained at each farm from the three sampling ports on the settling column
fit a single regression line, with an overall average coefficient of
determination (r) greater than 63.8%. This confirms that settling is indeed
discrete in intensive salmonid fish culture. The regression equations for the
pooled data set for each group are presented below:
Group Equation t P n r^
Std. Raceways In Cr = 0.09889 + 0.80730 In Vo 21.77 P<0.001 270 0.638
Oval Raceways In Cr = 0.31979 + 0.72947 In Vo 16.80 P<0.001 159 0.640
Circular Tanks In Cr = 0.06397 + 0.72781 In Vo 16.65 P<0.001 151 0.648
ANCOVA was used to identify similarities and differences betv/een these
three regressions. An F^^ test confirmed homogeneity of variance prior to
conducting the ANCOVA. The analysis indicates that the slopes of all three lines
are not significantly different (F=0.6485; P>0.05); however, the intercepts are
different (F=5.5180; P<0.01). Interestingly, it is the intercept of the
cumulative distribution curve for oval raceways which is significantly different
from the intercepts of standard raceways and circular tanks. The latter two are
not significantly different, thereby further emphasizing the similarities between
these two groups.
36
ANCOVA's were also used to compare farms within each group and to determine
the degree of within-group variation. For all three groups, clusters of farms
were found to have regression equations which were different from other subsets
within the group. The number of clusters, and the degree of overlap between
clusters, indicates that these within-group differences are subtle, and are
likely relatively unimportant in comparison with the between-group differences.
Feed brand, feed type, and feeding method were also explored as possible sources
of between group differences but no pattern was evident.
The class variable "feed brand" was also examined as a possible contributor
to differences in effluent water quality. Effluent water chemistry data were re-
grouped by the brand of feed used at the culture facility. The groups consisted
of (1) users of Martins Feed\- (2) users of Zeigler's Feed"; and (3) users of
a combination of these two brands. No other feed brands were identified in the
survey.
MANOVA was used to assess the impact of feed brand on effluent water
quality. No significant difference was found between groups.
Multivariate Analysis of Variance
Test Criteria & F Approximation
Statistic
Value
F
Num DP
Den DF
Prob.
Wilks' Lambda
0.82044
0.60926
14
82
0.8500
Furthermore, univariate F-tests (ANOVA) revealed no significant difference
(P>0.406) in effluent concentrations of ammonia, TKN, nitrite, nitrate, total
phosphorus, dissolved phosphorus, or total suspended solids between farms using
either feed brand.
Similarly, MANOVA and ANOVA were used to determine the effect of feed brand
on solids settling characteristics. As before, no significant differences were
observed using either univariate or multivariate techniques (P>0.232).
Multivariate Analysis of Variance
Test Criteria & F Approximation
Statistic Value F Num DF Den DF Prob.
Wilks' Lambda 0.86624 0.79783 6 31 0.5790
As a result, there is no evidence of differences in effluent water quality
or solids settling characteristics associated with feed brand.
^ Martin Feed Mills Ltd., Elmira, Ontario.
- Zeigler Bros. , Inc., Gardners, PA.
37
O iNACIAN
^QUACULTURE
3VSTEMS
4.4 Chapter Sumnary and Principal Findings
4.4.1 The five a priori groups of farms (i.e. standard raceways, oval raceways,
circular tanks, ponds, and others) display significant differences with
respect to design, operating, and untreated effluent water quality
perspectives. Most notably, while standard raceways and circular tanks
appear to be similar across several production and operational aspects,
oval raceways differ significantly from these other two groups. Oval
raceways have larger production areas, and they re-use substantially more
water. Moreover, the quality of untreated effluent from oval raceways has
higher concentrations of TKN, ammonia, nitrite, total phosphorus, and
dissolved phosphorus than do the other two facility designs. Again,
standard raceways and circular tanks were found to be statistically
similar. No differences were found with respect to effluent suspended
solids and nitrate concentrations across all three groups. Water use did
differ significantly across all groups, however, with the greatest amount
of water being used by oval raceways, followed by standard raceways, and
circular tanks.
4.2.2 Strong correlations exist between production parameters and untreated
effluent water quality. Most notably, increased intensity (i.e. carrying
capacity) and fresh water intensity are related to increased
concentrations of dissolved nutrients (TKN, TP, and/or DP) in all three
groups of production facilities. Fish density is not correlated with any
of the water quality variables in standard raceways, however, increased
density is significantly correlated with ammonia concentrations in oval
raceways and with total phosphorus concentrations in circular tanks.
Average fish size and production depth were also related to water quality
(NO,, TKN, TP, DP) in standard raceways and circular tanks, but not in oval
raceways. This further substantiates the similarity between standard
raceways and circular tanks and the differences between these two groups
and oval raceways. Since water re-use is principally practiced only in
oval raceways, it is understandable that only this group displays
correlations between water re-use and effluent water quality.
4.3.3 The retention time of water in the production facility, total production
area, total water use, and annual production capacity of a farm are not
related to any aspect of effluent water quality. This suggests that all
farms, big and small, have the potential to pollute receiving bodies of
water. Temperature is not correlated with effluent water quality in all
three groups.
4.4.4 Solids settling behaviour differs moderately within individual groups,
however, the between-group differences are highly significant and
inherently more meaningful. While the production effluent from standard
raceways and circular tanks displays statistically similar solids settling
curves, the curve for oval raceways is shifted to the left. This
indicates that the suspended solids fraction in oval raceway production
effluents settles significantly slower than that in the other production
systems.
38
Exhibit 4.1:
Definition and units of measure for the 25 continuous and 4 class variables
observed and recorded at the time of the site visit. The abbreviated name used
in the results section is also presented.
Variable Name Units Abbr.
Predictor Variables - Continuous
Intensity (kg/Lps) INTEN
Annual Production (kg) PROD
Production Area (m^) AREA
Production Depth (m) DEPTH
Standing Crop (kg) CROP
Density (kg/m^) DENS
Water Flow
Water Reuse
Fresh Water
Intensity
Temperature
(Lps)
(%)
(kg/Lps;
(°C)
FLOW
REUSE
FWI
TEMP
Definition
The approx. biomass of fish within
the observed production unit per unit
water flow.
The approx. total weight of fish
produced at the facility each year.
The surface area of the sampled
production unit.
Average depth of the sampled
production unit.
The approx. weight of fish within the
sampled production unit.
The approx. weight of fish per unit
volume in sampled production unit.
Total water flow through sampled
production unit.
The percentage of total flow (Q)
reused within the production unit.
The approx. weight of fish within the
sampled production unit per unit
volume source water added.
Water temperature.
Average Fish Size (g) SIZE
Hydraulic Retention HRT
Time (min)
The average weight of an individual
fish within the sampled unit.
The theoretical retention time of
water within the production unit
based on volumetric flow of source
water.
39
'ANADIAN
Xquaculture
Systems
Exhibit 4.1: (cont'd)
Definition and units of measure for the 25 continuous and 4 class variables
observed and recorded at the time of the site visit. The abbreviated name used
in the results section is also presented.
Variable Name Units Abbr.
Predictor Variables - Class
Farm Type — GROUP
Feed Brand
Feed Type
Feeding Method
FEED
TYPE
METH
Definition
Standard raceways, oval raceways,
circular tanks, ponds, other.
Martins, Zeigler's, combination of
Martins & Zeigler's, other.
Sinking, floating, combination.
Hand, demand, automatic, other.
Response Variables - Settling Data
Slope - Port 1 -- ml
Slope - Port 2
Slope - Port 3
Slope - CDC
Intercept - CDC
Coefficient of
Determination
- CDC
m2
m3
mCDC
INTPT
Slope of regression line for [TSS]
versus time for 120 cm port samples.
Slope of regression line for [TSS]
versus time for 160 cm port samples.
Slope of regression line for [TSS]
versus time for 200 cm port samples.
Slope of regression line for the
cumulative distribution curve for
sedimentation data.
Intercept of regression line for the
cumulative distribution curve for
sedimentation data.
Coefficient of determination of the
regression equation for the cumulative
distribution curve for sedimentation
data.
40
Exhibit 4.1: (cont'd)
Definition and units of measure for the 25 continuous and 4 class variables
observed and recorded at the time of the site visit. The abbreviated name used
in the results section is also presented.
Variable Name Units Abbr.
Response Variables - Water Chemistry
Definition
Total Kjeldahl
Nitrogen
Total Ammonia
Nitrite
Nitrate
(mg/L)
(mg/L)
(mg/L)
(mg/L)
Total Phosphorus (mg/L)
Dissolved Phosph. (mg/L)
Total Suspended (mg/L)
Solids
TKN
NH3
NO,
N03
TP
DP
TSS
KSO^ / H,S0^ Digestion - Colorimetric
@ 530 nm
Undigested - Colorimetric (3 630 nm
Undigested - Colorimetric @ 520 nm
Undigested - Colorimetric (3 520 nm
KSO4 / H^SO^ Digestion - Colorimetric
(a 880 nm
Undigested - Colorimetric @ 880 nm
Filtration through 0.45 \i\n membrane
filter paper.
Exhibit 4.2:
Canonical discriminant analysis - statistical results for segmentation of five
a priori groups of salmonid fish culture facilities.
CANONICAL
AXIS
CANONICAL
CORRELATION
EIGENVALUE
PROPORTION
CUMULATIVE
PROPORTION
1
2
3
4
0.921
0.880
0.761
0.697
5.629
3.423
1.376
0.943
0.495
0.301
0.121
0.083
0.495
0.796
0.917
1.000
CANONICAL
AXIS
LIKELIHOOD ;
RATIO
'\PPROXIMATE
F VALUE
NUM D.
F. DEN D.F.
PROBABILITY
1
2
3
4
0.007
0.049
0.217
0.515
2.101
1.613
1.149
1.029
100
72
46
22
85.8
66.6
46.0
24.0
0.0003'
0.0248*
0.3204
0.4707
41
%
i.NADIAN
OUACULTURE
^YSTEMS
0)
>
<
< <
5 = to « oc
lU ^ O Hi
o <z z I
< > < O I-
K O H- Q. O
• ■ O <1 •«
<n
CO
z
<
3 -r-
cn.rl
.>>>,
c
U
OJ
■
-o
+J
l/»
1/1
■(->
X
CM
fO
(Q <U
m <U <U
ifl
TS C
^ o
?] o u
_U CO IB
"i <"
O . u
O fO 4)
1/1
1/1 ^M-
>i^ O
*« "^ O
"3 ? TJ
.. -u
c
(U
U I u,
U H <U
en "O
c ,_
n3
(N
(J ^o
(T3 C
+J U 1/1 •>-
.r- .P- 0)
.a c 01 (U
■r- O 1/1 3
.C C 13 ■—
X fO I — fO
LU O U >
< o
o "
ro o "5r ^^ CM U3 CO
CM -^ "^ l~^ m CTi CNJ
-^ --^ ^ o o o o
«:r-HOU3CMO^a3CM«roLn
om-^cMOcnroO'-HO-^
"^ o^ u^ .— I tn o^ CO
-^ CM r«» -^ O Ln r^
ro m -^ -^ CM m o
ooooooo ooooooooooo
o o o o o o o
I — t/1
o
o
'— I CTi '^ ro tn U3 LT)
^ CO on CM "^ o on
o o o o o o ^
CMCMLncocMCO'=rr^O«=r'a-
iou3CMco--<cococncMOLn
OOOOO— <CMOO'-^— <
ro 1^ 0^ CM r^ CO "^
m CM r^ on CO o Lf)
O O O CM O -^ O
ooooooo ooooooooooo
O O O CJ o o o
•a- CM on '-^ o r~~ LD
o^ r^ CM o^ CO Ln CM
CO lO (T> «a- m CO -^
cocM<— lo^pno^o^iopococo
^•mCTiior^uncMtOLD^ooi
OO^cor^CMP^r>~CMCOCMro
rg- !Ti O "-^ O ro O
LD Ln f>«. m fo — < CTi
O O^ f~v ro O^ O^ CO
OOOOOOO
00(000000000 OOOOOOO
^^ CTi to CM un ^^ .-H
CO CTi Ln O r^ U3 CM
PO VO CM CM to O CTi
Q0OtDrv>,-<rrr».Ln-Hir)^^
tOrn>— i<3-cOCMiOOl<— iCTir^
aOCMCMOOOLntOr^CMCO'3"
O •— I ^^ O^ O "^ o^
ro o ui CO "^ IT! >a-
— < -^ «r CO ^^ m po
ooooooo ooooooooooo ooooooo
-— CM
<— I .— < .— I Ln p^ r^ r^
tn CM C7i «3- "^ CJi «^
CM CM CM ^H o '^ O
o^Ln«3-CMroo^o^^j'^-^^
OCOtOroOLfioocO'-HGOr-v
OUOCM^— ^LDLDOCMOCM
<* LD 0> CO LD -H "S-
to CO m a> CO f-< irj
LD Lf) OO -H IT) to <— '
OOOOOOO OOOOOOOOOOO OOOOOOO
— i~ ^M
CM o ("^ r^ ro ^ to
O -H r»~. IT) -H -M CM
-^ CM o O -^ O ro
r^LD-Htoo^cTi^-Hocoro
LnootocacMto-H<^Lnun.-^
^H,-HO'-<Ofr)LncMOOsifo
m o^ CO CO CM to -H
r>. tn CO O CO CM m
o o -^ un o CM -H
ooooooo ooooooooooo
ooooooo
CTi CT> ^» 00 O^ ^^ ^
to •— I o^ -H o^ •— I ^
m CM O ^H -H CM CTl
•^tOmc^-^CMtO«^■CTltOO
Ln^nr-^oor^tncotna^ooo
Oro<— ii — ommpotocMPO
on O «5" CTi CO r>«. CM
00 to tn — < CM ro to
tn 00 CM rr ^^ o P-)
-.- (U
TJ -i-
i- O
ns -^
T3 i*-
C 4-
,_4
IT3 0)
z
■t-> O
«*
00 <_>
C_)
ooooooo ^^O'^O'^CMOOOOO
^^ 00 CM '^ in (Ti CT>
to 00 --< •— 1 LT) o^ m
-H O ^H to O to CTi
tOP^tOCT>"^— itTiO^tnCMt^
o^Ln«a---itor^o^CTiocM'— I
^r-^iriLD^rcMLDtO'^tOLn
ooooooo
in ro LD to o^ o^ o
o oo tn to CM tn to
-H -^ o ro un m ^
0<3;3C300-^ OOOPO— <ro-HOCMC3-H O-^O-^OOO
O
o
a Q. h-
c_; uj I— I
O c_) <^
uj q; Q z
Q- UJ f) LU
O I— ^ H-
CM on _i zoi z
E e on —• s_ — .
z Q. — I—
3= O
Q<ci— o: uj Lu en ■z.
Oi-ua.c_) 3ooq:ljj
q:q:uj oooi— uji—
Q-<tQt3 —ILUI — cihsjuj
z>-u_QC«a:3'— q:
— I z z — ' t— a: t/n
^QQ^OOLlJUJCi.OO^Q;
zoo«a:zi— ►— sujooQ
^cKQCi— lij<:<^luq:>— .>-
<tQ.Q.OOQ33l— Li_Li_^
a. a. •
oo oo Q.
o o oo
:e =: =3
uj ex a. oo
Q ArJADlAN
tXquaculture
3ystems
>i "^ a-
.^
i.
£
<U
C J
•^
■<-i
T)
i/i
fO
I.
f-
£
p..
fO
(U
>
2
s-
</>
o
><
0)
<u
(9
ae.
i
"3
i.
(U o
■a
•::^
c
•F- '•■^
IT3
to
u =
S-'*- s-
o
<u
*+-
2!^
(/)
c
42 -a
0)
3 >> .
E
(J^ u
„ c <u
o
■C IQ -M
1/1 tj "^
IT) OJ
•^.J:: s
• ^_
'+-4-
T ja
■ _ 0)
fO
■o c O
■M +J
— Ol s-
•f-»
C -1- 3
^ <:
O i/l O
•— >
e on
J= o
-— -M
X z
iC O H-
UJ <
1/1 c o
Q. iQ
3 >
O O
>»
3
1^ 00 en «^ o lo
«r o ro 00 <Ti O
"5r Ol 00 CM i£) en
CM -^ — < CM 00 «3-
o o o o o o
— ' ^ ^ ^ o o
VO 00 O^ O 1^ "^
lO r«^ r». r^ O ro
O O O O O O
poor^oooo-^'— i^^^Ln«3"Ln
ooc\jocM»-<o«ro<roi-n
ooofno^^ino"5ro«a-ocT>
o o Lf) <n 00 ^^ f~^
^-. PO CTi U3 »-H o m
O O CM ^ O O -H
o o o <n o o "a-
oooooooooooo ooooooo
P>. vo m O 1*^ CM O
— Hr^rocooocMCM<n'^(ncocM
OCMOCMOO"3-Oi^OCMO
o 1^ 00 «r T 1^ '^
I — «r LT) o o 00 «r
CM CM — < O CM r^ O
o o o o o o
oo t^ -H un — < CO
r~» o <n r^ f*l "O
un in uD r>^ O r^
o o o o o o -^
oooooooooooo OOOOOOO
omo^voootn.— oo<n<n>3"ia
CO 'r^CTi • '-^ -vo • • •
•a-m •m<ovo •^oo'^oo
VOO ^Hom^ •lOvO^^
CM CM O CM CM
fo on ^r CO 'S' o
<n o ^o r^ 00 «n
<r in o- to ^H CO
o o o o o o
r^ r»N. o CM r>. CO
lO O^ O r«» ro ^H
m un kO vo o 00
o o o o o o
^^ CM m
•«-< -M -M
S_ S- S- <_)
O O O O CJ
a. a. Q. o o
0)
<_) CJ
1 1 1 1 o
^
1 C_)
<e
<U 0) 0) OJ
•^
a. Q. a. Q. • 1
u
O O O O -M
ror^OvoocMiomm^^r^O
to .o^ • "O • ^ • ' •
CMOO -CMOia •iniOCMr^O
mio>-^c0CMOr^CM •r^.ro^^
mro CMOO»-<omro
•— tfO<nr«vOOoOr^mm'-Hi— I
IT) 'OO-^ • -CM 'CM • • •
^to • ^T ao ^^ -ooocMm— »
-^oooocM«a-0(n ' o m i—
»y^^ ^^ rotO'^OCM— <
2 ur>.--
.— . ^^ O Q. S
Ol ^^ 1/1 r— _1 -—^
J>i C3V-^ CLU- — .
. — . ;/l _J Ol OJ - — .
m a.-^ o\= -ii EC-)
Q. e a; — I Ol — - — "r-^— '
•M S-
N ■
<u
<u
(/I </1 <•! «/1 1-^ ^
Ol—
(U a.c-; ^si: i/^
-al-(U ' OS-i3 — - 3
o<:qoi -c-— -MO) ■!->■»->
1- c>ii/iLj--i-Q:i-a)n3
Q.CC-'-+J''- i/l OJQCt-
---a-.-u_5_c:s--t-> (u
• ■a-oci/i oi (u <D f^ i- a.
C O O lO C Ol-M 4-> ■!-> 3 "O E
ci-i--i->a)><T]c<Q I >^a)
<:Q.Q.ooa<si— 'Su-^i—
in f^ lO o 00 CM CO
o^ in r^ ro in po ^r
«a- r^ O m -^ O — ^
^ o o o o O "^
CM lO o m ^ — < lO
o^ ro r»» p«» m CM rv.
— < 00 ro CO ro F-^ O
PO —H O O O O (*«•
•— I CM 00 •-' r>. r^ f^-
r». lO lO tD r^ CM O
CO ro O f^ O O O
O O O O O O r^
_l —J — I 3 3
. ^~^~-^ ;_ 1- 1/1
.C Ol Ol Ol O O "O
TS E E E J= -C •--
■a ^— «— '«.^ a. Q.1—
f— i/i CO O
OJ fO 0) <U O O OO
•^•r- ^_l +J .C -C
i<: c — (13 Q. a. •
o i- i- a.
■«-> E -•-> •I-' ■«-> ui ;/!
O E — •-- O ■- 3
h- <C Z Z I— Q VI
Exhibit 4.5:
Canonical discriminant analysis - statistical results for segmentation of three
a priori groups of salmonid fish culture facilities.
CANONICAL
AXIS
CANONICAL
CORRELATION
EIGENVALUE
PROPORTION
CUMULATIVE
PROPORTION
1
2
0.929
0.809
6.263
1.893
0.768
0.232
0.753
1.000
CANONICAL
AXIS
LIKELIHOOD
RATIO
APPROXIMATE
F VALUE
NUM D.
F. DEN D.F.
PROBABILITY
1
2
0.048
0.346
2.437
1.420
50
24
34
18
0.0037*
0.2250
45
9:
QUACULTURE
^YSTEMS
RACEWAYS
OVAL RACEWAYS
TANKS
■
■ ■ ■
■
■
• ■ O
ja
■ j
■
• 1
I
-
•
•
•
•
••
1 °
1 o
•
•
•
•
1 •
1 o
•
•
1 °
•
9
• • o ° o
' o
o
-
•
1 o
1 °
o
CO
OJ -o
(U
V
s-
+J
J=
0)
-M
-:»:
u
o-
ra
c
1-
o
CO
E
CM
00
(O
<
CNJ
en c
0) -I- lO
-a +-> •.-
•r- 00 X
•r- rn
n3 (^ O
■a
(/I i/i O)
(U •.- ■!->
X X =
(O fd 3
o
r- >, O
(O ■ — o
a c ns
•r- o
c <u
o . u
C on =
fO rtj fO
O r- >
0) 0)
I. (/I
•• 3 ^
t'1 ■!-> O
c 2 cn
n3 O (T3
-M +J C
C _ ID
= 5-
i- o
CM
CM
CM
§5
Z
CM
<
CO
u
CM
r-- T3
O
>4-
-a
«3- —
O
c
03
•f—
■M U
l/l
•f— •^-
(U
I/)
ja c
lo
(U
•r- O
i/i
3
x: c
-13
^—
X TJ
^—
TJ
LU C_)
U
>
j= <:
-^ u"/ — ■ i^ — r~. >5-
:^.4 "5- o T cv <3- m
>c f^ "o ^r o lO c^
— c; ?C o — > o ro
r^ r^ Ln — — cc
— — r^ -o r-^ .-^
— =T CNj "TT C^ ro u^
OOCOOCJO ooooooooooo
o o o o o o o
a a
<
= i.
'.^
T} 3
t_) -J
U
C 3
<U s-
<U ■!-«
S VI
1— <
ro vO O ".O "S- (O
— O ro in "^ oo
i>o O (^~ po r<-) CSJ
r-^ c; '>o o «3" po Lrt
^ u". O «5- C^ O O
r^ ro >c- —H C> un CO
O O O O O O O
ro r^ O "a- lT) o^ csj
O^ r^ O 00 ro CO r«^
O^ O^ — i^ CTi CI o^
OOOOOOOOOOO ooooooo
vOOP-10CMCS;COiOOJCM-^
r>^CTrOO(~~"^SD-^'~OuOO
LDO^O^C^Oa^C^^OCCX^O
O f*^ m en un ^r "S-
m r»^ CO 00 O ^ c^
o o^ oc c^ -^ 00 1-n
OOOOOOO
OOOOOOOOOOO OOOOOOO
r~> fn CM 00 O o^ i~~
m r^ O ro ro r»^ lT!
o o o ^^l o o o
LTiroCNjcO'^ro — Ln«a--^v^
OO^O'^oCi'^'S-mcMCSjcsj
— 0^-c\jonc\i— "O^-com
--^ 00 O CM O — in
00 CM c^ "^ tn — vc
— O CM — CM -^ <3-
OOOOOOO
OOOOOOOOOOO OOOOOOO
o
^O ro m <3- cT! r>» lO
I^- C^ O C^ CO 00 o
CM CM m — O — ' CM
«3--^CMm«a-i~.rorO«:rvOCvj
Ol — POI^CMi^C0'3"COi-O00
OLr)CM"^0'Or~«0'— '5r'3"
CO 'O o m p^ o a^
(^ 1^ C> ^O CM O O^
ro O «3" »0 O — " CM
OOOOOOO
OOOOOOOOOOO
OOOOOOO
1- u
(B —
■a >*-
C M-
■M O
rs. m — . lO CO CM o
CM LTi m CM CM — 1^
a> -^ CM — O — <Ti
O '-< O — ' o -- —
O^ CO «3- m O CM CM
r^ — — 00 CO CM CM
—" O CM «3- m o o
O O O O O O Osj
CT^O^^OO^OOLD-— 'VO"— '
ocor-»ooro^Lr)ocri«3-Ln
CMCMOr^"a-C0r^<^O«3"O
O — OCMOOOOOO —
vOO"=r"=J"cn«!rmroc^CMr>»
-^— HCOOOOl^OCTimi-DC^
CMrM00'-"CM«3---«OLr)'TLn
-'OOOO — OO — oo
CM CO 1-n o o- un CO
P>» CM r>« CM <yi -— O
ro <3- m o "sr m CM
o o o -< o o o
1^ vD CM ^^ — ' CM lD
CM Ln r>~ •— ' O^ kO lo
on rn "^^ rn ^r ^^ ^r
OOOOOOO
O - = O
cj Q<:i— c: LULucr z
oljjo-o 3ooq;uj
(_)h— >- Qiaiuj ooot— Luk—
OQ. I— Cl.«a:QCJ3 IUJ^— eCrvjUJ
c_)uj — z>-Li_a:<t3 — ai
luqiqz <:--o — a:cs:LLj =
Q-Lucii." OOOZV1 — -i-a.<yi = Q;
o>— — zoo<i:z — — S-U1/1Q
— ■ CM m _i zaj ^ ;zor:a:i— uj«s:<*:-uq: — >-
= = =!•!>— S-— <:2.Q.'vio33t— ij-ij-:c
o a. a. •
uj o o oo
-5 = = O
cC :^ ljj Lij o. o. uo
z — — «a: -J • — I
o <c a; a: <t oo <:
s: ►— ■— — f— oo ►—
z: o — — o -; o
<: ►— z z I— Q t—
QA;JAOiAN
/XpUACULTURE
Exhibit 4.9:
Spearman correlation coefficients and statistical probabilities of finding such
correlations by chance between predictor and response variables among three
groups of intensive salmonid fish culture operations.
Variables
Corr. Coef.
P
Variables
Corr. Coef.
P
STANDARD RACEWAYS
INTEN - TKN
0.462
0.035
INTEN - TP
0.372
0.097
INTEN - DP
0.584
0.005
DEPTH - TKN
0.396
0.076
CROP - TKN
0.371
0.097
FWI - TKN
0.436
0.048
FWI - DP
0.587
0.005
SIZE - TKN
0.414
0.062
SIZE - N02
0.430
0.052
OVAL RACEWAYS
INTEN - TKN
0.627
0.039
INTEN - TP
0.718
0.013
INTEN - DP
0.773
0.005
CROP - NHS
0.582
0.060
CROP - TKN
0.600
0.051
CROP - TP
0.564
0.071
CROP - DP
0.791
0.004
DENS - NH3
0.545
0.083
FLOW - N03
0.733
0.010
REUSE - N03
0.590
0.056
FWI - TKN
0.609
0.047
FWI - TP
0.527
0.096
FWI - TP
0.836
0.001
CIRCULAR TANKS
INTEN - NHS
0.804
0.002
INTEN - TKN
0.664
0.019
INTEN - TP
0.776
0.003
DEPTH - TKN
0.790
0.002
DEPTH - TP
0.580
0.048
DEPTH - TSS
0.524
0.080
CROP- TKN
0.545
0.067
CROP - TP
0.727
0.007
DENS - TP
0.552
0.063
FWI - NHS
0.734
0.007
FWI - TKN
0.671
0.017
FWI - TP
0.776
0.003
SIZE - TKN
0.510
0.090
SIZE - N02
0.497
0.099
SIZE - TP
-0.508
0.092
SIZE - DP
0.601
0.039
48
• A c
1/1 Q. (C
in
•r- 1_
r>;
facilit
fferent
1). Fo
IT)
= S=?
CO
.2 >-?
_
-i r^^-
in
en
ut pro
ignifi(
R and
in
E
o
in
>
5! = ^
■* OJ U
-^
o
itensi
es an
e of
o
^ Q. 1/1
"^ O O
in
-1
g l/l 4->
CO
liJ
O
>
to ^_>
o
T3 .
— U1 '^
in
z
■ — ^1/1
04
-J
^^ >,
1-
V i- —
C r— C
H
OJ 3 ro
UJ
i/l t. ••-
in
(/)
3 -^ "^
es for
ys; T =
is sign
3.1.
in
^ 5o
o
■^ <U oj o
o u > .X
k ^ *- +j
Q- s- 3 u
CO
(O
CVi
O
II ■>-'
1/1
o
'
Q.
a>
o iJ
*
>
U
a;
«a-
.. s-
>
CT
1/1 0)
i-
+j
C
>,-M
3
•r-
• ^
« c
U
J3
■ —
2 •'-
UJ uo I- ^-1 -M
%'
ANADIAN
QUACULTURE
YSTEMS
5.0 STATUS OF AQUACULTURE EFFLUENT TREATMENT IN ONTARIO
The 44 farms included in Groups 1, 2, and 3 comprise a substantial
proportion of the commercial trout producers in Ontario. Of these operators,
more than half utilize some means of effluent treatment prior to discharging
process water from their facilities. Sedimentation technologies are applied
exclusively at these facilities; the use of technologically advanced filtration
equipment for effluent treatment has not been witnessed on a commercial scale in
the province.
5.1 Aquaculture Settling Facilities in Ontario
Four principal gravitational solids removal strategies are used: (1) in-
raceway settling basins; (2) external, rectangular basins; (3) external ponds;
and (4) external circular basins.
In-raceway settling consists of allocating a portion of the raceway unit
for settling, usually at the downstream end. Screens are typically used to keep
fish out of the settling area; however, it is not uncommon for a few fish to be
present in the quiescent zone. In some instances, overflow or underflow weirs
are used to introduce water to the settling zone. In virtually all cases, flow
is introduced across the entire width of the raceway. Few operators utilize
full-width discharge weirs, however, as overflow discharge standpipes tend to be
more common.
Due to the physical layout of some fish culture facilities, the major
stream of process water is not treated within the culture unit but rather it is
subsequently treated in a separate sedimentation basin. Such units must be large
enough to accommodate the entire farm process flow. In these instances,
rectangular concrete basins, earthen ponds, and/or circular basins are used.
External rectangular basins are often much like raceways, however, they are
frequently wider and shorter. Flow is usually introduced through a modified pipe
inlet, not dissimilar from those inlet structures depicted in Exhibit 3.5. The
use of submerged, wide-top weirs is limited. Outlet structures vary between
full -width overflow weirs and vertical standpipes.
Earthen pond settling basins were the accepted effluent treatment process
of the past for fish culture facilities. While most ponds have been designed
with a substantial surface area for settling, poor inlet and outlet design,
excessive depth, and the inability to clean solids from the bottom render them
inefficient in many applications. These design flaws generally stem from out-
dated criteria mandating a specific retention capacity within the pond.
Moreover, point source inlet and outlet structures are the norm, thereby
exacerbating the inherent inefficiency of ponds.
In recent years, the use of circular settling basins has evolved in
aquaculture. These basins are typically constructed of galvanized steel silo
walls, which are laid into a concrete pad. A single pipe inlet conventionally
rises from below ground at the centre of the tank. A distribution baffle,
constructed of a perforated steel or plastic barrel, is usually placed over the
50
inlet pipe, extending from the floor to the surface of the basin. This baffle
serves to reduce turbulence and distribute the flow radially across the tank.
The outlet from the basin most commonly consists of perforated drainage tile
which is moored at the basin surface around the perimeter of the tank. In spite
of the use of this effective design, some operators utilize circular basins with
a peripheral inlet and a central drain. In the latter cases, internal baffles
are often used to control the flow of water within the basin. This inward flow
approach is comparatively inefficient.
5.2 Effectiveness of Existing Treatment Operations
During the field survey component of this investigation, specific design
and operations data were collected from existing effluent treatment facilities
at fish culture operations. Only those facilities from which it was possible to
obtain water samples from the treatment influent and effluent flows, and where
the dimensions and total water flow through the treatment unit could be
quantified, were included in this portion of the study. In all, 25 treatment
units, at 24 farms, were sampled. The 17 observed variables are defined in
Exhibit 5.1.
^ priori segmentation was used to classify the different treatment methods
prior to statistical analysis. Since earthen settling ponds act as a nutrient
sink (Westers 1989), and since they introduce substantial variability due to the
inherent biological activity within, pond treatment units have been eliminated
from this analysis. Moreover, data were collected from only two earthen settling
ponds during the field investigation. MANOVA was used to identify differences
between the three remaining groups: i.e. in-raceway settling units; external,
rectangular basins; and external, circular basins. The analysis revealed a
significant difference between groups (P=0.051).
Multivariate Analysis of Variance
Test Criteria & F Approximation
Statistic
Value
F
Nm DF
Den DF
Prob.
Wilks' Lambda
0.00986
2.6671
34
10
0.051
Using univariate ANOVA, coupled with Student-Newman-Keuls multiple range
tests, between-group differences were found to be associated with hydraulic
loading, inlet area, outlet weir rate, and NH3 removal. For the most part,
however, no differences in removal efficiency of the seven water chemistry
variables were observed between the three different types of treatment facilities
(Exhibit 5.2).
The data indicate that the removal efficiency of TSS from aquaculture
effluents is poor, averaging only 29.5% in in-raceway settling facilities, 31.7%
in external, rectangular basins, and 15.5% in external circular basins.
Similarly, removal of total phosphorus is also poor, averaging only 0.8% in in-
raceway treatment units and 38.3% in external rectangular basins. For dissolved
phosphorus, effluent concentrations were greater than influent concentrations in
all treatment designs (Exhibit 5.2), confirming that a portion of particulate
51
Q*
%
ANAOIAN
QUACULTURE
YSTEMS
phosphorus quickly becomes solubilized within the system. Complete data sets for
all treatment units are presented in Appendix IV.
Examination of the design and operating parameters for these effluent
treatment facilities provides several explanations for the overall poor level of
performance. In spite of various recommendations which suggest that overflow
rates be in the range of 40 to 80 m^/m^/d for aquaculture clarifiers (note that
this is not necessarily a recommendation of this report; see Section 7), average
overflow rates were found to surpass this range in 17 of 26 cases. ^ Values
ranging from 85.9 m^/nr/d in external rectangular basins to 344.3 m^/wr/d in
external circular basins were observed (Exhibit 5.2). Notice, too, that external
rectangular basins have higher average removal efficiencies for TSS and TP than
do in-raceway and external circular basins; the difference, however, is not
statistically significant. This difference is likely a reflection of overflow
rate, although certainly not exclusively. Outlet weir rates which exceed design
recommendations by approximately 2 to 6 fold were observed at 19 of the 26
treatment facilities (Exhibit 5.2).
Since all three groups were found to display similar treatment
efficiencies, design and operations variables were correlated with water
chemistry variables in search of meaningful relationships within the total data
set. Correlation coefficients and statistical probabilities are presented in
Exhibit 5.3. The cross-sectional area of the inlet zone and the total surface
area of the settling basin hold the most meaningful correlations with respect to
treatment efficiency. Correlation coefficients are greater between Vo and
settling performance than between outlet weir rate and settling performance,
suggesting that overflow rate is the more critical of these two design factors.
Basin surface area is highly correlated with both TSS removal (P=.069) and
TKN removal (P=.072). Negative correlation coefficients indicate that an
increase in surface area is concomitant with enhanced removal of TSS and TKN.
Recall from Exhibit 5.1 that the water chemistry data assessed in this analysis
consist of ratios of treatment effluent to treatment influent — the lower the
number, the lower the effluent concentration relative to the influent
concentration. The cross sectional area of the inlet zone is also correlated
with TSS removal (P=.075), and with nitrite removal (P=0.001). Similarly,
negative correlations imply that removal of these components increases as cross
sectional area of the inlet zone increases, suggesting reduced inlet energy and
turbulence. Conversely, however, removal of dissolved phosphorus is enhanced
(P=.015) when inlet area is reduced (Exhibit 5.3), perhaps due to greater short-
circuiting, which reduces retention time for a portion of the flow and thus
decreases contact time for solubilization of nutrients within the settling basin.
Three variables (TSS, TKN, TP) are negatively correlated with hydraulic
retention time (P<.092), suggesting that increased retention time improves
settling efficiency. This is most certainly a reflection of the impact of basin
surface area on settling and not due to retention time. Retention time is the
quotient of the volumetric flow rate divided by basin volume and, therefore, an
increase in volume leads to an increased retention time. Increased volume,
however, is generally attributed to increased area since, in aquaculture, basin
depths are generally relatively constant at 0.7 to 1.2 metres. Moreover, this
52
explanation is substantiated by the fact that improved removal of the same
parameters is correlated with basin area (Exhibit 5.3).
Interestingly, basin depth was found to be highly correlated with enhanced
removal of ammonia (P=.017) and TKN (P=.002). No correlation exists, however,
between depth and total suspended solids removal (P=.186), supporting the concept
that particulate matter in aquaculture effluents exhibits discrete settling
behaviour. Lastly, a significant negative correlation (P=.007) exists between
average fish size and total suspended solids removal, suggesting that solids
removal is facilitated in production units rearing larger fish (Exhibit 5.3).
5.3 Chapter Sumnary - Principal Findings
5.3.1 All three different treatment facility designs display similar treatment
efficiencies. Moreover, all performed relatively poorly with respect to
TSS and TP removal. Effluent concentrations of DP increase as water
passes through the treatment facility in all three treatment designs.
5.3.2 Poor performance is likely related to a combination of operational and
design flaws; namely, excessive overflow rates and outlet weir rates, and
improper inlet and outlet structure designs. Given high design overflow
rates and for the most part only poor to good treatment facilities, the
applicable scaling factor for determination of an appropriate performance
overflow rate becomes quite large. Consequently, poor treatment
efficiency is expected.
5.3.3 TSS removal is highly correlated with settling unit surface area. Since
depth is relatively uniform across treatment facilities, this suggests
that TSS removal is related to overflow rate; i.e. as area increases and
depth is held constant, overflow rate decreases. While this finding is
not supported by the correlation data for TSS and overflow rate, it was
found to be highly significant in data collected at land-based New
Brunswick salmon smolt production operations.
5.3.4 TSS removal is highly correlated with average fish size, suggesting that
solids settling is enhanced in operations raising large fish. The smaller
feed and fecal pellets associated with small fish reduce settling
velocities and, therefore, make effluent treatment via sedimentation
somewhat more difficult. Given similar hydraulic flows and production
intensities, a system producing small fish will require a larger settling
facility (area) to achieve the same level of performance as an operation
raising larger fish.
53
Qanadian
TXquaculture
Exhibit 5.1:
Definition and units of measure for 17 effluent treatment variables observed and
recorded during site visits. The abbreviated name used in the results section
is also presented.
Variab'<e Name Units Abbr.
Design/Operational Variables
Settling Area (m^) AREA
Hydraulic Load (m^/d) Q
Overflow Rate
Inlet Area
(m^/m^/d) Vo
(m^) INLET
Outlet Weir Rate (m^/m/d)
Hydr. Ret'n Time (min)
Depth (m)
Slope of CDC
Intercept of CDC
Fish Size
(g)
OWR
HRT
DEPTH
mCDC
INTPT
SIZE
Definition
Total surface area of the settling
unit.
Total water flow through treatment
unit on a daily basis.
Calculated; Vo = Q/AREA.
Cross sectional area of inlet zone of
the settling unit.
Daily volumetric flow of water per
unit length of outlet weir.
Theoretical retention time of water
in settling unit.
Average depth of settling unit.
Slope of cumulative distribution
solids settling curve.
Intercept of cumulative distribution
solids settling curve.
Average weight of individual fish in
prod'n unit feeding settling unit.
Water Chemistry Variables
Tot. Susp. Solids — TSS
Total Ammonia -- NH3
Tot. Kjeldahl N — TKN
Nitrite — N02
Nitrate — N03
Total Phosphorus — TP
Diss. Phosphorus -- DP
These data are dimensionless. They
are ratios of the effluent concen-
tration of the parameter to the
influent concentration. A number
less than 1.0 indicates a net loss of
that parameter via settling.
54
Exhibit 5.2:
ANOVA table of mean design/operational factors and treatment efficiency for in-
raceway, external rectangular, and external circular settling basins at intensive
salmonid culture facilities. Values having the same superscript are not
significantly different.
Variable In-raceway
Ext. Rect.
Ext. Circ.
F Ratio
Prob.
(n=ll)
(n=5)
(n=8)
Desian/Ooerational Vari
ables
Basin Area (m^)
27.4
29.9
28.9
0.880
0.430
Hydraulic Load (m^/d)
Overflow Rate^ (irP/m^/d)
6251"
1486""
1060"
3.393
0.053
269.5
85.9
344.3
2.307
0.124
Inlet Area (m^)
3.17"
0.96"^
0.30"^
19.93
0.000
Out. Weir Rate (m^/m/d)
2062"
1081""
945"
3.308
0.056
Hydr. Ret'n Time (min)
9.75
27.7
42.1
1.357
0.279
Basin Depth (m)
0.93
0.72
1.08
2.289
0.126
Slope CDC
.6210
.7851
.8253
2.269
0.128
Intercept CDC
.0057
.0871
.1639
0.452
0.643
Average Fish Size (g)
216
242
159
1.643
0.217
Effluent Treatment Efficiency Vari
ables (see
Note)
Tot. Susp. Solids
0.705
0.683
0.845
0.176
0.839
Ammonia
0.995""
2.097"
0.870"
2.892
0.078
Tot. Kjeldahl N
0.894
0.999
0.816
0.500
0.514
Nitrite
0.954
1.229
1.800
1.632
0.219
Nitrate
0.981
1.753
1.872
1.349
0.281
Total Phosphorus
0.992
0.617
1.016
1.044
0,370
Diss. Phosphorus
1.569
1.808
1.067
1.319
0.289
N.B. Effluent treatment efficiency variables are dimensionless values; they
represent the ratio of effluent to influent concentrations. A number less
than 1.0 signifies removal of that contaminant from the effluent stream.
55
Qa'4A0IAN
/XpUACULTURE
gvSTEMS
Exhibit 5.3:
Matrix of correlation coefficients and statistical probabilities (P) between
sedimentation unit design/operations factors and treatment efficiency with
respect to seven water quality parameters (n=24) .
^^\
Water Chemistry
^^^ Parameters
Design ^^v^^
& Operating ^^^
Parameters
\
TSS
NH3
TKN
N02
N03
TP
DP
AREA
-.3112
P= .069
-.1870
P= .191
P=
.3073
.072
P=
.0345
.436
P=
.1692
.215
P=
.2615
.109
P=
.0886
.340
Q
-.2246
P= .146
-.1347
P= .265
P=
.1303
.272
P=
.1063
.311
P=
.0674
.377
P=
.1110
.303
P=
.0375
.431
Vo
.2438
P= .125
.2026
P= .171
P=
.2408
.129
P=
.1627
.224
P=
.0945
.330
P=
.2491
.120
P=
.0553
.399
INLET
-.3025
P= .075
.1717
P= .211
P=
.0044
.492
P=
.5813
.001
P=
.2463
.123
P=
.0627
.385
P=
.4442
.015
OWR
-.0155
P= .471
.0455
P= .416
P=
.1285
.275
P=
.1299
.273
P=
.2345
.135
P=
.0175
.468
P=
.0847
.347
HRT
-.2809
P= .092
-.2651
P= .105
P=
.3710
.037
P=
.1944
.181
P=
.1544
.236
P=
.2823
.091
P=
.0908
.336
DEPTH
-.1910
P= .186
-.4327
P= .017
P=
.5575
.002
P=
.1045
.313
P=
.0941
.331
P=
.1969
.178
P=
.2542
.115
mCDC
-.0107
P= .480
.0525
P= .404
P=
.1121
.301
P=
.0615
.388
P=
.1283
.275
P=
.0854
.346
P=
.0267
.451
INTPT
-.2077
P= .165
-.1389
P= .259
P=
.0572
.395
P=
.0817
.352
P=
.2653
.105
P=
.1444
.250
P=
.1276
.276
SIZE
-.4960
P= .007
-.0916
P= .335
P=
.2252
.145
P=
.0007
.499
P=
.0064
.488
P=
.2553
.114
P=
.1752
.206
N.B. A negative correlation coefficient implies improved treatment efficiency
due to an increase in the corresponding design/operations variable.
Recall that the treatment data are ratios of effluent-/ influent
concentrations.
56
6.0 IMPLEMENTATION OF AQUACULTURE EFFLUENT TREATMENT
5.1 Coimercial Fish Production Facilities in Ontario
Across Ontario, three principal fish culture production units exist, and
the decision to construct one over the others is generally based upon the
ind-ividual preference of the aquaculturist. Circular tanks are often selected
for their self-cleaning qualities; standard raceways for their ease of
management; oval raceways for their reduced concrete content and the perceived
benefits of extended water use. Other decision factors include familiarity <vith
a given system; i.e. fish farmers often construct a facility wnich is similar to
a neighbour' s.
Throughout the investigation oval raceway production facilities have
consistently stood out as unique production facilities, different from standard
raceway and circular tank operations with respect to several production and water
quality parameters. Oval raceways generally incorporate a larger production
area, utilize more water reuse, and produce an effluent of poorer quality than
the other production units (Exhibit 4.5). With respect to solids settling
behaviour, oval raceways produce a cumulative distribution sedimentation curve
which has a significantly greater elevation (Exhibit 4.10), implying that solids
removal via settling is more difficult in these facilities than in standard
raceways and circular tank operations. From a productivity perspective, however,
oval raceway systems seem to offer no additional benefits. Most importantly,
these units do not offer extended water use. That is, productivity, in terns of
carrying capacity of fish per unit water flow (kg/Lps), is statistically similar
(P=.4415) in standard raceways, oval raceways, and circular tanks (Exhibit 4.5).
This finding suggests that, for a given volumetric flow of water, a
relatively constant amount of fish can be produced. Naturally, this amount will
differ due to the experience of the culturist, and will increase if additional
factors are introduced, such as liquid oxygen injection and/or recirculation with
biofiltration for ammonia removal. For the most part, however, productivity is
not influenced by the use of standard raceways, oval raceways, or circular tanks.
Consequently, any tangible benefits related to using oval raceway culture units
must derive from other factors, which were not observed and recorded within the
scope of this investigation. Such factors may include: faster growth, improved
fish health, ease of management, reduced capital and/or operating costs, and so
forth.
In all culture facilities examined, culture intensity (kg fish/Lps),
standing crop (kg fish), and average fish size (g) are significantly correlated
with many aspects of effluent quality (Exhibit 4.9). Intuitively, this is
logical since an increase in any one of these factors necessarily translates to
an increase in feed ration and, consequently, poorer water quality. Feed input
is responsible for deterioration of water quality. Therefore, feeding strategies
must be considered as an important management tool for effluent quality control.
While feed presents the source of the pollution problem, specific feed-related
factors (i.e. feed brand and type, and feeding method) were not found to impact
upon the quality of effluent or effectiveness of treatment. While this may not
be entirely plausible, it does suggest that feed-related factors are subordinate
to the polluting influence of facility design and culture management practice.
57
Qa.naoian
^^uaculture
^YSTEMS
6.2 Design of Effluent Treatment Facilities for
Intensive Salmonid Aquaculture
Fo'^ most commercial -scale intensive salmonid aquaculture operations, sol ids
removal via gravity settling remains the best available technology. Moreover,
since total phosphorus removal is significantly correlated with total suspended
solids removal (Mudrak 1981; Stechey 1988; Appendix III), gravity settling is
also recommended for reduction of effluent concentrations of phosphorus.
Specific phosphorus removal technologies (i.e. addition of metal -salts, chemical
polymers, and/or lime to promote phosphorus precipitation; biological uptake into
cell tissue) are not warranted in aquaculture due to the dilute nature of the
wastewater stream and the relative expense of these processes.
Overflow Rate. As stated in Section 3.2.4, overflow rate (Vo) must be the
foundation of design for solids settling units. Based upon the cumulative
distribution curves for solids settling in standard raceway, oval raceway, and
circular tank production facilities, relationships between overflow rate and
solids removal efficiency can be projected. These relationships are graphically
presented in Exhibit 6.1. The LIMA Engineering^ study (1988) recommends design
overflow rates ranging between 40 and 80 m^/m-/d (0.046 - 0.092 cm/s). This
means that, for every Imperial gallon of water flowing through the treatment unit
each minute (IGPM), 1.75 to 3.50 square feet of settling basin surface area must
be provided. Theoretically, for overflow rates in this range, TSS removal
efficiency varies between 86% and 94% (Exhibit 6.1). In similarly designed
treatment units, however, removal efficiency is better for standard raceway and
circular tank production operations due to differences in the settling behaviour
of suspended particulate material in oval raceway effluent.
Recall, however, that turbulence must be accounted for in design by
applying an appropriate scaling factor, as presented in Exhibit 3.4. In reality
then, solids removal in the range of 86% to 95% is unlikely. For theoretical
solids removal efficiencies in this range, and assuming that basin design and
performance is "very good," the data in Exhibit 3.4 recommend that a scaling
factor between 2.2 and 3.6 be applied. The net result will be a reduced
efficiency rating for the design overflow rate. This is best explained using the
following example.
Let's assume that an in-raceway settling basin has been designed using
proper inlet and outlet structures, a design overflow rate equal to 40 m^/m^/d
(0.046 cm/s) based upon the actual surface area of the settling basin and the
hydraulic flow, and that the design is "very good." In theory, the curve in
Exhibit 6.1 suggests that TSS removal efficiency will equal 94.5%. From Exhibit
3.4, we determine that a scaling factor of 3.6 must be used to adjust this
performance projection. Consequently, the performance overflew rate is more
appropriately equal to 0.166 cm/s (0.046 x 3.6) or 143 m^/m-/d. Going back to
the plot in Exhibit 6.1, we see that, in reality, the projected efficiency of
this system is closer to 86%.
58
100
90
S
^
TSS
REMOVAL
{%)
8
of '""^
^
70
60
50
^
N
0 .1 .2 .3 .4 .5
OVERFLOW RATE
( cm/s)
Exhibit 6.1:
59
Qa.N'ADIAN
Taouaculture
S ■''STEMS
Therefore, when designing settling basins for aquaculture operations, an
iterative approach is necessary:
(1) The surface area required for settling is calculated based upon the
design overflow rate (see Section 6.4) and the hydraulic loading
rate of the operation.
(2) The theoretical efficiency of this design is determined using the
data plotted in Exhibit 6.1 for the appropriate fish culture
facility.
(3) A scaling factor must be selected from Exhibit 3.4, based upon this
design. This scaling factor is used to adjust the design overflow
rate to enable a realistic projection of basin efficiency.
(4) If the effectiveness of solids removal is too inefficient to meet
effluent discharge criteria, then the process must be repeated from
Step 1 using a lower overflow rate.
At existing operations, the design overflow rate can be calculated by
dividing the area of the settling basin into the hydraulic flow rate of the
basin. Then, using Exhibit 6.1, a theoretical removal rate can be determined for
the treatment facility. Next, using Exhibit 3.4, an appropriate scaling factor
can be selected to calculate the performance overflow rate, and to project a more
realistic level of efficiency. Experiential judgement will be required to
classify the overall level of basin performance such that an appropriate scaling
factor from Exhibit 3.4 can be selected and applied-. If the projected level of
efficiency is insufficient to meet discharge criteria, the area of the settling
basin will need to be increased (or the hydraulic load decreased) to improve
performance.
In all settling basins, the average horizontal fluid flow velocity must be
maintained below 2 to 4 cm/s to ensure that scour problems do not arise.
Inlet Structures. Proper inlet design is critical for minimizing inlet
velocity, turbulence, and hydraulic short-circuiting. The specific design of the
inlet structure, however, will depend upon the type of treatment facility being
used. In all cases, the inlet should be designed such that fish from the
production area of the culture facility do not enter the settling basin.
For in-raceway settling basins, there is no need for an inlet structure,
per se. Instead, simply screening off a section of the raceway at the downstream
end provides for effective introduction of water to the quiescent settling zone.
It is imperative, however, that all air diffusers be kept upstream of the fish
retaining screen at a distance of approximately 2 times the depth of the water
column. This is necessary since the rising air bubbles create large convective
cells having a circular mixing motion over this region. The screen material
(usually semi-rigid plastic) must be sufficiently small to restrict the passage
of fish into the settling zone.
60
If external rectangular basins are used, it is important to distribute the
inlet flow across the entire width of the basin, and to achieve this with minimal
energy input. A submerged, wide-top weir with chamfered edges, which extends
across the full width of the settling basin, is most appropriate. The degree of
submergence should be approximately 15% of the basin depth; i.e. in a 1 metre
deep basin, the weir should be 85 cm tall. Since some turbulence is still
introduced with this inlet structure, the settling zone should be lengthened to
account for any reduction in efficiency due to the inlet. A typical length
increment is equal to the basin depth.
For external circular basins, process water should enter the basin at the
centre, ideally from below the basin floor. The inlet pipe should be enclosed
within a concentric perforated baffle, designed to reduce inlet velocity and
distribute the flow radially through the full depth of the basin. The inlet
turbulence can be accounted for by adding 0.25 times the basin depth to the basin
diameter. Moreover, since circular tanks come in standard sizes, the designer
of a circular settling basin will be wise to use the standard tank size which is
larger than that recommended in design calculations, as opposed to a tank
slightly smaller than required.
Outlet Structures. For all settling basins, an overflowing outlet weir is
most appropriate for gently drawing off clarified effluent from the treatment
unit, this structure must be designed such that the clear water is skimmed off
the top of the basin; hence the overflow design. To minimize the entrainment of
fine particulate material in the rising discharge flow, it is important to
maximize the length of the outlet weir.
For in-raceway and rectangular settling basins, at a minimum the weir must
extend across the entire width of the basin. The outlet weir rate (m^/m/d or
IGPM/foot weir length) can be measured and used as an indicator of the
entrainment effect. If the weir rate is greater than 400 m^/m/d (18.5 IGPM/ft) ,
then the settling zone should be lengthened by 1.5 times the basin depth in
compensation. For weir rates between 350 and 400 m^/m/d (16.3 to 18.5 IGPM/ft),
the additional length requirement is 1.3 times basin depth, and for weir rates
below 350 m^/m/d (16.3 IGPM/ft), the added length need only be equal to the basin
depth. Finally, in circular settling basins, the overflow weir should extend
around the periphery of the basin. Since the circumference of circular tanks is
large in relation to the width of similarly sized rectangular basins,
compensation for outlet entrainment effects is generally not required in circular
basins.
In most existing effluent treatment facilities, overflow standpipes are
used for outlet structures. These can be readily upgraded at minimal expense.
Two alternatives exist. Placement of an overflow weir constructed of damboards
or stoplogs immediately upstream of the standpipe will provide the necessary
weir. To ensure an overflow action, the damboards should be used to control tank
water level and the standpipe should be cut approximately 6 to 10 cm shorter.
Alternatively, a T-fitting can be placed over top of the existing stand pipe.
From each end of the T, a halved pipe is cemented into place, extending from one
side of the basin to the other. This configuration essentially provides a
double-width weir, since water now overflows on both the upstream and downstream
61
^c
ANADIAN
QUACULTURE
^YSTEMS
sides of the halved pipe. Effluent is discharged through the standpipe as
before. In adapting overflow weirs, however, care must be taken to ensure that
the weir is level, thereby ensuring that water flow is uniform across the entire
weir length.
6.3 Projected Production of Phosphorus and Solids from Intensive
Fish Culture Facilities
Operations data collected during the field study have provided a
quantitative understanding of standing crop biomass and fish production
intensities based upon water use. Production intensity represents the weight of
fish in the tank per unit water flow (kg fish/litre per second of source water
flow). Standing crop is the average total weight of fish in the production
unit(s). These field data can be applied, in conjunction with the waste
production factors defined in Section 6.4, to project average daily
concentrations of TSS, TP, and DP in untreated fish production effluent.
Moreover, projections can be generated for the three groups of production
facilities.
For each group, the average standing crop (kg fish) was used as a starting
point. A feeding rate of 1.2% body weight per day was assumed, based on
production data which show that the average size of individual fish ranged
between 204 and 340 grams, and was not significantly different (P>0.05) between
groups. The mean fresh water intensity, and the 95% confidence limits around
this mean, were incorporated into the model as an indication of production
intensities. From these data, an average influent flow rate was calculated, and
then, using the waste production factors for TSS, TP, and DP, average daily
concentrations of these pollutants in the production effluent were projected.
These projected values are compared with observed values in Exhibit 6.2.
For all three groups, variability exists between observed and expected
concentrations of pollutants in the effluent. Among the three pollutants, this
variability is greater for total and dissolved phosphorus than for solids
concentrations. In some cases, projected values are two- to three-fold greater
than observed values. Solids modelling is reasonably accurate for standard
raceways and circular tanks, but it over-estimates the concentration of effluent
solids in oval raceways. Conversely, phosphorus modelling is reasonably accurate
for oval raceways, but is over-estimated in standard raceways and circular tanks
(Exhibit 6.2).
Oval raceways produce the highest expected concentration of TSS, TP, and
DP among the three groups while standard, flow-through raceways produce the
lowest concentrations. Indeed, observed data confirm that TP and DP
concentrations are significantly lower (P<0.002) in standard raceway and circular
tank operations (Exhibit 4.5). Circular tank operations were observed to produce
the greatest concentration of TSS, however, field observations reveal no
significant difference between the three groups (P=0.4197).
Possible factors which contribute to the variability between predicted and
observed values may include the use of "fresh water intensity" as a model input.
In group comparisons, FWI was not found to differ between groups (P=0.4415),
62
however, individual groups means were incorporated into the model. As well, for
individual pollutants, the reliability of the waste production factors may be
suspect, especially for phosphorus. Solids modelling is comparatively simple;
the only critical factor is the overall digestibility of the diet -- a component
which is well quantified by feed manufacturers. Total and dissolved phosphorus
production, however, is largely influenced by the quality and quantity of
phosphorus in the diet, which is subject to change due to variations in feed
ingredients. Moreover, the range of production factors for phosphorus compounds
in published literature is greater than that for total suspended solids, thereby
affirming the complexity of nutrient load modelling. Lastly, the assumption of
a feed rate of 1.2% body weight per day is valid for modelling, however, one
cannot assume that this feed rate was applied in all fish culture facilities on
the specific days when field samples were collected.
Assuming, however, that these modelled waste production figures present
best available data, then these data can be applied in conjunction with effluent
treatment efficiency figures to generate numbers reflecting the anticipated
quality of final discharged effluent from aquaculture operations. The data
provided in Exhibit 5.2 offer a margin for error since, for nearly all pollutants
and fish culture facilities, the projected values are greater than those values
observed during field monitoring.
6.4 Projected Performance of Effluent Treatment Facilities
When the nature and composition of aquacultural wastes are considered in
design, aquacultural treatment systems can be particularly effective in
clarification of discharged effluent. Moreover, since the removal of total
phosphorus is positively correlated with removal of suspended solids, a reduction
in the latter necessarily translates into a reduction in the former.
Operational data from existing fish hatchery treatment units in the United
States have been compiled (Exhibit 6.3). Settleable solids (SS) removal in
excess of 90% is readily achieved, and greater than 80% removal of total
suspended solids (TSS) is reported (Mudrak 1981; McLaughlin 1981). Note that the
design overflow rates for these facilities range between 21.4 and 120.0 m^/m^/d
(0.025 to 0.139 cm/s). Total phosphorus removal via sedimentation in concrete
clarifiers is reported for three fish culture stations in Pennsylvania. Oswayo
FCS, Tionesta FCS, and Big Spring FCS report average reductions of total
phosphorus in discharged effluent equalling 52%, 58%, and 77%, respectively
(Mudrak and Stark 1981; Mudrak 1981).
Regression analysis of the data presented by Mudrak (1981) for total
suspended solids and total phosphorus removal provides the following
relationship:
% TP Removal = 0.838 x % TSS Removal - 4.579 ■r = 0.936 (n=9)
Using the TSS and TP production rates in Ontario fish culture operations
(see Exhibit 6.2), and applying the removal efficiencies for TSS in properly
designed clarifiers (see Exhibits 3.4 and 5.1), the likely concentration of
discharged pollutants from commercial fish culture operations can be projected
(Exhibit 6.4).
63
QAfJADIAN
tXquaculture
3YSTEMS
Exhibit 6.2:
Projected and observed concentrations of TSS, TP, and DP in untreated fish
culture effluent assuming a daily feed ration of 1.2% body weight. Data for
standing crop and fresh water intensity are from field observations.
STANDARD RACEWAYS
Standing Crop (kg)
Feeding Rate (% body weight/day)
Daily Feed Ration (kg)
Fresh Water Intensity (kg/Lps)
Average Flow Rate (Lps)
TSS Prod'n @ 300 g/kg Feed (kg)
Projected Avg. Daily [TSS] (mg/L)
Observed [TSS] (mg/L)
TP Prod'n @ 7
Projected Av
Observed [TP
.6 g/kg Feed (kg)
, Daily [TP] (mg/L)
(mg/L)
DP Prod'n @ 2.2 g/kg Feed (kg)
Projected Avg. Daily [DP] (mg/L)
Observed [DP] (mg/L)
Lower C.L.
Mean
10417
1.2
125.0
Upper C.L
99.1
105.1
202.3
51.5
305.5
34.1
37.50
4.13
4.20
37.50
8.43
7.01
37.50
12.73
9.82
0.950
0.105
0.053
0.950
0.214
0.077
0.950
0.322
0.102
0.275
0.030
0.012
0.275
0.062
0.027
0.275
0.093
0.041
OVAL RACEWAYS
Standing Crop (kg)
Feeding Rate (% body weight/day)
Daily Feed Ration (kg)
Fresh Water Intensity (kg/Lps)
Average Flow Rate (Lps)
TSS Prod'n @ 300 g/kg Feed (kg)
Projected Avg. Daily [TSS] (mg/L)
Observed [TSS] (mg/L)
TP Prod'n (a 7.6 g/kg Feed (kg)
Projected Avg. Daily [TP] (mg/L)
Observed [TP] (mg/L)
DP Prod'n (a 2.2 g/kg Feed (kg)
Projected Avg. Daily [DP] (mg/L)
Observed [DP] (mg/L)
8216
1.2
98.6
196.9
41.7
372.1
22.1
547.4
15.0
29.58
8.20
4.18
29.58
15.50
7.08
29.58
22.81
9.97
0.749
0.208
0.138
0.749
0.393
0.354
0.749
0.578
0.570
0.217
0.060
0.072
0.217
0.114
0.121
0.217
0.167
0.170
64
Exhibit 6.2: (cont'd)
Projected and observed concentrations of TSS and TP in untreated fish culture
effluent assuming a daily feed ration of 1.2% body weight. Data for standing
crop and fresh water intensity are from field observations.
CIRCULAR TANKS
Standing Crop (kg)
Feeding Rate (% body weight/day)
Daily Feed Ration (kg)
Fresh Water Intensity (kg/Lps)
Average Flow Rate (Lps)
TSS Prod'n (3 300 g/kg Feed (kg)
Projected Avg. Daily [TSS] (mg/L)
Observed [TSS] (mg/L)
TP Prod'n @ 7.6 g/kg Feed (kg)
Projected Avg. Daily [TP] (mg/L)
Observed [TP] (mg/L)
DP Prod'n (3 2.2 g/kg Feed (kg)
Projected Avg. Daily [DP] (mg/L)
Observed [DP] (mg/L)
Lower C.L.
Mean
995
1.2
12.0
Upper C.l
130.5
7.6
265.9
3.7
403.2
2.5
3.59
5.44
4.73
3.59
11.12
11.15
3.59
16.80
17.56
0.091
0.138
0.039
0.091
0.282
0.158
0.091
0.425
0.277
0.026
0.040
0.012
0.025
0.082
0.032
0.025
0.123
0.052
Exhibit 6.3:
Efficiency of solids sedimentation for removal of settleable solids (SS) , total
suspended solids (TSS), and total phosphorus (TP) from intensive trout culture
effluent.
Site
Desi
gn Overflow
%
Removal
Source
Rate
im'/w/d)
SS
TSS
TP
Lamar NFH
120.0
90.0
na
na
McLaugnlin 1981
Jordan River NFH
58.7
79.2
55.8
na
"
Jones Hole NFH
57.5
91.2
90.0
na
"
Oswayo FCS
69.7
97.8
87.3
55.6
Mudrak 1981
Tionesta FCS
40.5
99.1
88.9
68.2
"
Big Spring FCS
21.4
99.0
88.0
75.7
65
Qanadian
TXquaculture
3ystems
Exhibit 6.4:
Projected performance of settling basins at intensive rainbow trout culture
operations in Ontario at three different overflow rates and assuming "very good"
basin design.
DESIGN PARAMETERS
Scale-up Factor^
Design Overflow Rate-
m^/m-/d cm/s
Performance Overflow Rate''
m^/m^/d cm/s
2.2
2.2
2.2
43.2
0.050
64.8
0.075
86.4
0.100
95.0
0.110
142.6
0.155
190.1
0.220
PERFORMANCE PARAMETERS - STANDARD RACEWAYS
Untreated
mgTSS/L
Effluent"*
mgTP/L
Design 0\
m^/nr/d
/erflow Rate
cm/s
% Removal
TSS TP
Effluent
mgTSS/L
Conc'n
mgTP/L
8.43
8.43
8.43
0.214
0.214
0.214
43.2
64.8
86.4
0.050
0.075
0.100
89.5
85.5
81.5
70.3
67.0
63.6
0.89
1.22
1.56
0.064
0.071
0.078
PERFORMANCE PARAMETERS
- OVAL RACEWAYS
Untreated
mgTSS/L
Effluent"*
mgTP/L
Design Q\
m^/nr/d
/erflow Rate
cm/s
% Removal
TSS TP
Effluent
mgTSS/L
Conc'n
mgTP/L
15.50
15.50
15.50
0.393
0.393
0.393
43.2
64.8
86.4
0.050
0.075
0.100
84.5
78.5
73.5
56.1
61.1
56.9
2.40
3.33
4.11
0.133
0.153
0.169
PERFORMANCE PARAMETERS - CIRCULAR TANKS
Untreated Effluent"*
mgTSS/L mgTP/L
11.12
11.12
11.12
0.282
0.282
0.282
Desig^n Overflow Rate
m^/m-/d cm/s
43.2
64.8
86.4
0.050
0.075
0.100
% Removal
TSS TP
87.5 68.6
83.0 64.9
79.0 61.5
Effluent Conc'n
mgTSS/L mgTP/L
1.39
1.89
2.34
0.089
0.099
0.109
Scale-up Factor = 2.2; i.e. theoretical removal efficiency of approx. 85%
in basins with "very good" design (see Exhibit 3.4).
Design Vo based on hydraulic load and required surface area.
Performance Vo based on scaling factor turbulence adjustment (see Exhibit
6.1).
Projected values (see Exhibit 6.2).
66
It is evident that the effluent concentration of total phosphorus is the
limiting factor governing the necessary level of efficiency for aquaculture
wastewater treatment facilities. Over the^range of design overflow rates
examined in Exhibit ^A (i.e. 43.2 to 86.4 m^/m^/d) , an average standard raceway,
oval raceway, or circular tank operation will meet the 10 mg/L discharge
criterion for TSS. Over this same range of overflow rates, however, only
standard raceway operation? will achieve the 0.10 mg/L non-compliance limit for
total phosphorus discharge. In fact, an average standard raceway operation will
achieve compliance limits at design overflow rates as high as 167 m^/wr/d (0.193
cm/s), producing daily average discharge concentrations of TSS and TP equal to
2.6 and 0.100 mg/L, respectively. Circular tank operations reach a ceiling at
a design overflow rate of approximately 65 m^/m^/d, producing a daily average
discharge concentration equal to 0.099 mgTP/L.
Oval raceway facilities, once again, produce the poorest quality effluent
and, therefore, require comparatively larger treatment facilities. The data
presented in Exhibit 6.4 suggest that an average oval raceway operation will not
produce a final effluent within compliance limits for TP at design overflow rates
as low as 43.2 m^/m^/d (0.05 cm/s). However, if one considers the observed
concentrations of TP in untreated effluent at oval raceway operations and,
furthermore, if the single largest observation is removed from the data set, the
mean concentration of TP in oval raceway effluent falls to 0.279 mg/L. Assuming
that this is representative of oval raceway operations, 64.2% of the TP must be
removed to achieve the compliance limit. This requires that 82.2% of the TSS be
removed, which corresponds to a design overflow rate of 49.3 m^/m^/d (0.057
cm/s) .
In view of these findings, design overflow rates below 167 m^/m^/d (0.193
cm/s) are recommended for standard raceway operations and 64.8 m^/m^/d (0.075
cm/s) for circular tank fish culture operations. This value should be reduced
to 40.0 m^/m^/d (0.046 cm/s) for oval raceway production units. At these rates,
the effluent concentration of total phosphorus is projected to be marginally
below the non-compliance limit of 0.10 mgTP/L. These data are presented in the
following table, along with conventional flow and area terminology for ease of
implementation. Note that these data assume the use of groundwater and/or
springwater sources having a negligible influent concentration of total
phosphorus. The maximum design overflow rates for fish fanns utilizing surface
waters will be much greater, ultimately depending upon the background
concentration of total phosphorus.
Culture System
Standard Raceway
Oval Raceway
Circular Tank
Design Vo
(m^/nr/d)
167
40
65
Settling Rate
(cm/s)
0.193
0.046
0.075
Req'd Area
(ftVlGPM)
0.42
1.76
1.08
67
Qanadian
tXquaculture
3ystep,/is
Since these values represent industry-wide recommendations for average
facilities, specific overflow rates may necessitate fine tuning at individual
aquaculture facil ities to maximize performance. Routine water qual ity assessment
for TSS and TP will be required to permit treatment unit adjustment. In all
cases, to achieve optimum performance, it is imperative to compensate for inlet
and outlet structure effects by increasing the length of the settling basin as
required, over and above the area requirement define^ by hydraulic loading and
overflow rate. Refer to Section 6.2 for additional length requirements for
specific settling basin designs.
6.5 Hanagement Strategies for Effective Effluent Quality Control
Management strategies must address the issue of fish farms as sources of
aquatic pollution on four principal fronts: feed and feeding practices, rearing
unit design, solids settling unit design, and solids removal and disposal. All
four of these factors are necessarily inter-related. A chain is only as strong
as its weakest link and, therefore, pollution must be challenged on all fronts
if abatement is to be effective.
6.5.1 Feed and Feeding Practices
Feed is ultimately the source of pollution in aquaculture. From the feed
manufacturers' perspective, R&D efforts must be focused upon the development of
commercial feeds which have reduced phosphorus content and higher digestibility.
Indeed, on-going research in these areas at nearly every major feed manufacturer
and fish nutrition laboratory in North America, Europe and Japan is steadily
improving the overall quality of feed.
It is too convenient, however, to blame feed manufacturers for feed-related
pollution. The fish culturist can make considerable inroads toward reducing the
pollution impact of feeds through the use of wise feeding practices. Moreover,
such practices often also produce a financial benefit. Rough handling and
improper storage of feeds promote pellet erosion, resulting in an increased level
of fines which cannot be utilized by the fish. Proper storage and handling,
therefore, can reduce the pollution impact of feeds and result in less feed
wastage. Prior to adding feeds to feed hoppers or troughs, screening the pellets
will remove the fines, thus preventing their addition to the water. Feeding
practices can also be managed to reduce pollution and improve fish production.
A carefully monitored feeding program will enable the fish culturist to select
the optimum pellet size for the size of fish in the facility. Feeding of
oversized and/or undersized pellets promotes waste, inefficiency due to poorer
feed conversion, and pollution. Educational symposia on feeding strategies and
programs, sponsored by feed manufacturers and/or university extension offices,
would be highly beneficial.
68
6.5.2 Rearing Unit Design
The results of this study clearly indicate that different designs of fish
P'-oduction facilities have different pollution impact. Selection of those
designs which facilitate solids removal and minimize the release of particulate
and soluble pollutants should be emphasized in the development of new aquaculture
ventures.
In all culture facilities, tank design should facilitate the rapid and
gentle removal of waste feed and particulate fecal matter from the water column.
The longer these materials remain in suspension, the more they are subject to
agitation and, consequently, the smaller they become. These smaller particles
are then more difficult to remove from suspension and require substantially
larger treatment units to achieve the desired level of efficacy. Additionally,
leaching of soluble nutrients is enhanced as particle size decreases.
When designing and constructing culture tanks, it should be borne in mind
that production area does not limit the amount of fish that can be produced; the
volumetric flow of water does. In Ontario, a standing crop of 245.8 + 226.1 kg
fish per Lps of water (41.1 + 37.7 Ibs/IGPM) has been observed as an industry-
wide average production intensity. This average intensity is not different
between the three main groups of fish culture operations examined. The
production area of these groups averages 242.8 + 358.6 m^. Moreover, average
area is significantly different between the three principal designs (Exhibit
4.5). Clearly then, some culture units are larger than they need be.
Given a constant water flow rate and tank depth, increased area means that
the rate of water movement through the tank is reduced, and slower flow results
in more solids settling within the rearing unit where it is not desired. Solids
settling within the culture unit hinders solids removal and, hence, increases the
solubilization of nutrients. Shallow tanks (generally less than 76 cm or 30" of
water) are recommended to provide good horizontal flow and to promote self-
cleaning. In shallower tanks, fish density is higher and thus fish activity,
coupled with aeration, generally keeps particulate matter suspended within the
water column so that it is carried to the settling unit.
Many existing aquaculture operations can improve solids removal from the
culture units by reducing the level of water in the units. In fact, at one
highly-intensive facility in the province, the shallowing of raceways from 152
cm to 76 cm greatly reduced solids accumulation in the production unit and
facilitated daily maintenance procedures, with no reduction in productivity.
Standard raceway and circular tank culture systems should be selected in future
design considerations due to the quality of discharged effluent observed at these
facilities.
5.5.3 Solids Settling Unit Design
The specific requirements for effective solids settling unit design are
presented on a micro level in Section 6.2 and will not be repeated here. From
a macro level, however, one key issue warrants discussion; that of treatment unit
size. In virtually all applications, a number of small treatment units designed
59
%
ANADiAN
QUACULTURE
YSTEMS
to treat a portion of the total process flow is recommended, as opposed to one
large treatment unit sized to treat the entire process flow. Small units, such
as in-raceway settling basins, are easier to manage and clean. It is more
realistic to schedule cleaning of ten smaller basins, say 15 m^ each, than one
large basin that is 150 m^. In each case, the total settling area is the same,
however, cleaning of one small basin as part of a daily routine can be scheduled
more readily than can cleaning the entire large basin. Hence, the smaller units
are more manageable.
6.5.4 Solids Removal and Disposal
Sedimentation of solid matter from fish culture effluent is only the first
step of the effluent treatment process. Once settled to the floor of the
settling basin, the accumulated manure must be effectively removed and disposed.
Solids Removal. Vacuuming of settled manure, much the same way that
swimming pools are vacuumed, still offers the best method for solids removal.
Vacuum heads designed specifically for fish culture are commercially available,
or can be easily home-made.
The frequency of vacuuming must also be addressed since, over time,
accumulated solids may become anaerobic at lower layers, resulting in the release
of dissolved nutrients. Moreover, Mudrak (1981) reports that aged manure becomes
quite viscous and removal is impaired. Before either of these events occur,
however, the depth of the manure will certainly increase to a point where basin
efficiency is greatly reduced due to lost volume.
Aerobic solubilization of dissolved nutrients (especially phosphorus) is
most rapid during the first 2 to 3 days after the solid pellet is discharged into
the water, and comparatively little aerobic solubilization occurs beyond day 4
(Zeigler 1988; Parjala 1984). Consequently, unless the solids had been vacuumed
from the settling basins and production facilities within 2 to 3 days prior to
site visits, the data collected during this investigation represent steady-state
concentrations of dissolved and total phosphorus. As such, so long as effluent
compliance criteria are achieved in treated wastewaters, the frequency of solids
removal from settling facilities is somewhat flexible. From the fish culturists'
perspective, relatively frequent cleaning (1-2 week intervals) is easiest since
the depth of accumulated manure remains manageable within this time period.
Mudrak (1981) suggests a maximum of six weeks between cleanings, however, an
interval not longer than two weeks will facilitate removal and should effectively
maintain appropriate effluent quality.
Solids Disposal. There is a Chinese philosophy which states: "substances
that by others may be considered wastes are only resources out of place." Such
is the case with settled fish manure, which has similar properties to low grade
fertilizer. Typical nutrient concentrations in trout culture manure are
presented in the following table.
70
Nutrient Composition of Settled Manure (% dry weight)
Mudrak & Stark Will at & Jakobsen
Nitrogen 1.4% to 4.9% 3.30%
Phosphorus 1.5% to 3.0% 1.03%
Potassium 0.2% to 0.7% 0.03%
Organic Carbon NA 25%
pH NA 6.31
Consequently, land application of fish manure remains the best solution for
disposal. Since the manure is highly liquid (generally less than 5% solids), it
can be spread directly onto agricultural land using liquid manure spreaders
and/or irrigation pumps. Naturally, this solution is only feasible during the
agricultural growing season. Throughout the winter months, it is necessary to
store the vacuumed manure in a suitable holding facility.
The application of settled fish culture wastes over agricultural land has
been successfully demonstrated. While heavy manure loadings can be detrimental
to cover crops and can cause severe odour problems, properly managed loadings can
be beneficial to vegetative growth. Trout manure has been found to increase the
availability and uptake of both nitrogen and phosphorus, and enhances the dry
matter yield of crop harvests. No significant increases in the concentration of
potassium (K), sulphur (S) , magnesium (Mg) , copper (Cu) , zinc (Zn), aluminum
(Al , boron (B), iron (Fe), sodium (Na) , calcium (Ca) , cadmium (Cd) , chromium
(Cr) , lead (Pb) , nickel (Ni), or molybdenum (Mo) were detected in cover crops
(Mudrak and Stark 1981; Willet and Jakobsen 1986).
The concentration of nitrogen in settled trout manure is the limiting
factor governing land application rates. Pennsylvania experience has shown that,
in general, manure can be safely applied over agricultural land. For mixed grass
crops, which have a nitrogen uptake capacity of about 224 kg/ha (200 lbs/Ac),
manure application rates should approximate 120 m^ per hectare (10,600 Imp.
gal /Ac), assuming a 5% solids concentration.
6.6 Economics of Effluent Treatment
Internationally, aquaculture tends to be most profitable in regions where
environmental regulatory controls are lax or non-existent (Pruder 1939). While
any recommendations put forth in this study must necessarily be effective in
controlling effluent concentrations of suspended solids and phosphorus, the
installation and operation of any wastewater treatment facility must be
economically feasible for the industry to adopt. Effluent treatment does not
increase productivity and, therefore, the expense associated with effluent
treatment must be minimized. Consequently, least-cost treatment technologies,
based upon effectiveness, simplicity and practicality, must be applied. From a
71
QAriAOlAN
tXquaculture
3YSTEMS
cost/benefit perspective, Parjala (1984) reports that sedimentation technologies
present the most realistic methodology for clarifying aquaculture effluents.
Within the context of aquaculture in Ontario, Stechey and Kantor (1988)
examined the underlying cost structure of intensive trout farming. These data,
which are similar to cost structure data for intensive trout culture presented
by Lewis (1979) and Castledine (1986), have been used as the basis for this
economic assessment. Note, however, that this analysis is not intended to be a
comprehensive assessment of the economics of trout culture. Rather, it is
presented to offer comparative assessment of practical effluent treatment
operations for standard raceway, oval raceway, and circular tank production
facilities.
The following assumptions have been made in conjunction with this economic
study.
1) Analyses have been conducted for 3 hypothetical farms, each having an
annual production capacity of 45,000 kg (100,000 lbs). One farm has
standard raceways, one has oval raceways, and the third has circular
tanks. The average standing crop at these operations equals one-third of
the annual production output (15,000 kg).
2) Water flow for each facility is calculated using the observed production
intensities for the three classes of farms (see Exhibit 4.5). The number
and size of rearing units is calculated accordingly.
3) Moccia and Sevan (1989) indicate that the average farm gate price received
by farmers in 1988 was $5.06/kg whole fish. Dupont (1990) suggests that
current figures reflect a reduction in the farm gate price to
approximately $4.41/kg whole fish ($2.00/lb). This latter price is used
in the analysis.
4) In all cases, it has been assumed that the farm is already in production
and that effluent treatment facilities are to be retrofitted into the
facility. The most appropriate design has been selected for each
operation.
The rationale applied to reliably forecast productivity, culture system and
effluent treatment design, manure storage requirements, and the general
necessities of the treatment facility are presented in Exhibit 6.5. For the
standard raceway operation, in-raceway settling is most appropriate. In standard
raceways, it is necessary to design the treatment units such that they are
capable of handling the entire process flow through each raceway, including any
water reuse. External circular clarifiers are most appropriate for oval and
circular tank operations. With circular production tanks, the opportunity does
not exist to treat the effluent within the culture tank and, therefore, treatment
must be in a separate basin. Note, however, that two smaller units are used
rather than one large unit. Due to the substantial volume of reuse water in oval
raceway operations, it is more practical to treat only the discharged effluent
(equivalent to the influent flow) in external basins rather than treating the
entire process flow with in-raceway settling zones. Again, two units are used.
72
The requirements for inlets and outlets, a vacuuming system, labour, electricity,
and water quality sampling are also described in Exhibit 6.5.
In Exhibit 6.5, a dollar figure has been affixed to each compcnent of
effluent treatment. These have been broken down into fixed (capital) costs and
variable (operating) costs to facilitate the analysis. A 10 year straight-line
depreciation schedule has been applied on capital expenditures to derive ar.
annual expense. The total annual operating expense is lowest for effluent
treatment in circular tank production facilities (S3583). Treatment operations
are marginal ly more expensive in standard raceway operations (S3897/yr) , and most
expensive in oval raceway facilities (S4392/yr). The largest component of the
total expense is the annual operating cost, which is approximately three times
greater than the amortized portion of the total fixed cost (Exhibit 5.6).
For the hypothetical intensive trout culture operations, gross revenue has
been projected at $198,450; i.e. 45,000 kg (3 $4.41/kg. All variable and fixed
costs have been allocated based upon the expenses allocated to each account, as
presented by Stechey and Kantor (1988). Without effluent treatment, earnings
before interest, tax, and depreciation (EBIT) are projected at 21.5% of sales.
Variable and fixed costs account for 71.5% and 28.5% of total costs, respectively
(Exhibit 6.7).
As expected, incorporation of effluent treatment causes the bottom line to
worsen in all three production facilities. From a cost perspective, effluent
treatment accounts for 2.25% to 2.74% of total production costs. Comparatively,
LIMA Engineering (1988) estimates effluent treatment costs between 5% and 10% of
total production costs. Furthermore, the LIMA study presents data which suggest
an increase in production costs ranging from SO. 07 to SI. 74 per kilogram of fish
produced. This analysis indicates that, on a per kilogram basis, effluent
treatment will increase total production costs by 50.08 in circular tank
production operations, SO. 09 in standard raceway operations, and SO. 10 in oval
raceway facilities. Total production costs increase from S3.46/kg to more than
S3.54/kg. Consequently, profit margins decline by approximately $0.08 to $0.10
per kilogram (SO. 036 to 50.045 per lb) (Exhibit 5.7).
Note, however, that these data reflect projected treatment costs in a very
large commercial operation. In Ontario, many fish farms are small ventures which
serve to supplement the income of traditional fanners. Moreover, since the
annual operating cost of effluent treatment presents the largest cost component
of effluent treatment, and since these will change little with the size of the
facility (water qual ity monitoring costs will not change), then the costs become
much more significant in smaller-scale operations.
73
ANADIAN
QUACULTURE
YSTEMS
Exhibit 6.5:
Rationale for economic analysis of effluent treatment operations in intensive
salmonid aquaculture.
PRODUCTIVITY
Annual Production (kg)
Stdg. Crop (1/3 Prod'n)
Intensity (kg/Lps)^
Hydraulic Flow Rate (Lps)
CULTURE SYSTEM
Rearing Unit Dimensions
No. of Units
Unit Flow Rate (Lps/IGPM)
EFFLUENT TREATMENT
Design
No. of Units
Hydr. Load/Unit (m^/d)
Design Vo (m^/m^/d)
Unit Dimensions
Standard
Raceways
45,000
15,000
198.7
75.5
30.5x2.4x0.76
8
9.4/125
Oval
Raceways
45,000
15,000
125.3
119.7
Circular
Tanks
45,000
15,000
266.9
55.2
61.0x4.8x0.76 5.8 dia. x 0.76
2 15
119.7/1580 3.7/49
In- raceway
Ext. Circular
Ext. Circular
8
2
2
812
1741
1457
86.4
49.2
64.8
5.75x2.44x0.76
7.3 dia. x 0.76
5.8 dia. x 0.71
MANURE STORAGE LAGOON
Volume (0.864m^/dxl50 d)
129.6
129.6
129.6
REQUIREMENTS
Structure
N/R
2 tanks
15m^ cone, pad
2 tanks
lOm^ cone, pad
Inlet
15m^ screen
30 m 25cm pipe
30 m 15cm pipe
57m framing
2 baffles
2 baffles
Outlet
176m damboards
46m 10cm Big 0
37 m 10cm Big 0
Vacuum Pump
1 HP
1 HP
1 HP
Vacuum Head &
Hose
1 ea
1 ea
1 ea
Discharge Pipe
45 m
45 m
45 m
Labour (vacuuming)
4 hrs/wk
4 hrs/wk
3 hrs/wk
Electrical
3 kwh/wk
3 kwh/wk
2.2 kwh/wk
Sampling (source/effl .)
2-TP 2-TSS/mo
2-TP 2-TSS/mo
2-TP 2-TSS/mo
For standard raceways, total water use is considered since treatment is
in-raceway; for ovals and circulars, treatment volumes equal the flow of
source water, net of reuse, since treatment is external to fish culture.
74
Exhibit 5.5:
Annual capital and operating expenses for effluent treatment facilities
incorporated into existing intensive salmonid aquaculture operations.
Standard
Oval
Circular
Racewavs
Raceways
Tanks
($)'
(S)
(S)
CAPITAL COSTS
STRUCTURE^
Tanks
N/R
1750
1300
Concrete ($130/nf)
N/R
1950
1300
Preparation
N/R
400
300
INLET
Screens
250
N/R
N/R
Frames
70
N/R
N/R
Piping
N/R
1100
700
Baffles
N/R
50
50
OUTLET
Damboards
600
N/R
N/R
Big 0 Pipe
N/R
70
55
Piping
N/R
550
350
VACUUM SYSTEM
Pump
200
200
200
Vac. Head
100
100
100
Discharge Pipe
140
140
140
MANURE STORAGE LAGOON
Excavation
2500
2600
2600
Fencing
1740
1740
1740
TOTAL CAPITAL COST
5700
10650
8835
ANNUAL FIXED EXPENSE"
570
1065
884
OPERATING COSTS
LABOUR ((3 S12/hr)
LABORATORY ANALYSES
2-TP, 2-TSS per mo. + deliv
ELECTRICITY (3 $.073/kwh
ANNUAL OPERATING EXPENSE
TOTAL ANNUAL EXPENSE
2496
2496
1872
816
816
816
15
3327
15
3327
11
2699
3897
4392
3583
Fish culturists will install treatment units themselves-
10 yr. depreciation schedule.
75
p'anadian
TaQuaculture
Systems
. — C
u a;
■I- X
5 2r
(J 2
03 OJ
^ TS -r-
i = =
r= <U
4- C
i I-
■^ ^3
?
mS-
-a
C "U
cu
1
•r- O.
•r-
-M O
r—
ae
u
o.
p!
<U (U
u
-— u
(O
4- 3
1-
Hi +j
cn
S_ r—
.
3
<T3
«
1/1 (J
C (13
1/1
o
O 3
•■- cr
UJ
0)
+J fO
•^"
ir
o
4-
UJ
-?
tl
<
O "^
(O
>-
M
a
3
^ "^
u
lO
E -^
o
3
c ^
U
o c
u
O 13
•1—
(^J(^Jaoa^O^-OOm^r^
i-ooooaoaof\jr\j
oaino^J^Tnaooo
inr-r-OaOinO(Mr-
ar\j>oaoir»0'OroKt>ooo
V»V »^V fOvv
■Or-r- T-T-OO >rfO
(^i-ooo>r^-•omr-^o^-
a3irioaOin-*o«-r^
o t) a w a
o W ro d fo o S
a d d d d d <-'
s s
rvj r\i l>- in r- so (>
^ ** in m ^ Ki t—
(M -o 00 .- <d d od
SN-mfy-rooo oQru h-
vrMOO^I^r- -opo o
«-f\i35coma soK oo
^ O K> nj Ki
«- >0 p in oo O
T- r- O^ rg o (\j
f\i in NO r- -o a <\j
f- a to
3 :?
□1 >. <« a.
— tJ c I. o
Z f
v u. ->
? s
*-•
8
X
(0
■
■
§
c
«l
3
tl
a
^
^
^
u
a
5
L.
4-*
o
Ui
Q. W) ^ L. UJ
v>ua9M3U "o MO. oe
<-iU ti irt f> tJ U «-• K-
oai-wu S— o w
Ot-f-r-O .311- CJ^.* 3
3 5 _..,-♦. _/ UJ
•atAO.'-'i-O'*--' -• m 3
lic4>«->«-b'»-a a)- -I
x-^a:30a.ujw vii-i il
•,- o o ca u.
>tin-i<NMC>».r-Q(^>»
rijf>j<3ao^«-OOv»i?>
r-'dddddddd(\ii-^
vp(vj^N.<-rg>fogh>
(>rgomo^in(\jo>*
inN.T-a«-o^inOOr-
nO»-»- r-«-oo «-r-^
ior~T-in<h-*in>oQO
ooinaaam^do-d
o !Q Ki o S o S
:?5:
o o o o a o o
to o
0> 00 p> o O O M
>0 in N- m CO S m
8
(\j sO 00 r- oo O oo
s
^ a m ckj m ^
IS
j:;2SiQS83
5S
(M in -o T- -o O tvj
ffs]
5 5
" 3 «
c <-• -< a.
o « a t-i a
r- a UI
41 u. -<
1. X tJ • r-
o (3
ujauc^'-i'^-a
vii-uoioxaiq
Z UU.UJXUU.^
«> £ H- -> t.
I. 4-> '4- a w
I- O UJ w c
a
UI
f-
(.
UI
<J
na
UI
3
t
ii
a.
V
i/>
4-*
*j
b
O
a
1.
4-*
V
0
I.
u
3
u.
o
3
a
s
HI
C
i^
4-*
■♦-
t
J:
a
a
t-
o
CL
4-<
o
o
ID
M IJ M-t OS
1- . —
. — c
3 S-
i- -*-)
•<- X
(J a;
1 =
>i -
(O CO
5 >>
'J ?
(O
;/1
■i fO -1-
^ -^ y
^ = (U
-a ■ ^
.— 1/1 4->
■M 1— O
OJ -M ••-
U 0) U
X <u
« 5-
O 5-
c <u <u
• r— ^-'~
"-> S"
<U (U ^
■'^ S- I — i-
1/1(1
+-> = In />
c o 3 s-
o— S-o;
.- s </i^
.a o 3
•>- c J^ u
f o c s-
X U 15 -1-
fNJCMOOOvOr-OO'J^QO
«-^cidoooooo<Ni^
l> fO o ^o «- (M «-
O <M ro O fO O O
d d d d d d «-
ro fo c^ ^ O ir\ m
O *T u^ i/% "O Lr\ ^
(\4 ,o od «- oo o 00
T- -o O O O ry 00
r^ f<i po t- o t\j (\i
ir» O ^O ro *o O >r
odir>OOOm^O»-l^
p N- u^ OJ ^ -J' ry
5^ ~T OO 00 ^ 00 Kl
<- fM *o ^ flO 00 ro
^ O fO ru ro tn
O V
*- m
I. IS
a. u.
C 4J U) 1-
O <u 1) t-* n>
u c 4-> — ' a.
t.u.UJZ<->u._ll-OUJ«; c
^ O i/^ (O ir* ^
,- O rv O^ >>• oo
01 t' -w
o o «-
■
C/l
V
4-*
«l
4.1
c
X
o
«-*
V
V
o
T
o
r
«i
a
<0
a:
-1
a.
UJ
fvjrvjooo-OT-OOin
<-^ddddddddpa
u^ oo 00 o^ O
O to a ^ «- ry r-
O (vj to o ro o O
d d d d d d r^
trt 00
to d
^OOu^O^tOtOfMO
»i-t>^r^ddi>irid(\J
<\1 O ■* ir\
^ ^ in ir\
»» -O fM Q
>o 05 ^ o
>*aor^oo«->ooor^OQ
«-^<-oor*toto?-(NjmQ
Oc^J-ooOlnO^OlOr<^^ooo
XXX XXX pOxx
-0«-r- »-r-0O xTtO
(\l■ooOxt^-^o^n«-^o^-
od^nddou^xIO«-^-
rg xT tn to
xT 00 00 xj-
r\j NO x» oo
x» O to (M fO
r- vO Q i/> 00 x^ to
r" V ^ fM 0^ ^/^ On
(\j in nO «- nO O <vj
.- 3 u>
O 4) (0 4-1 (Q
c a> ♦- w
— (-• c I- o
— ' ID U )— u C
— < 4) O
o o) >N u) a. -u a.
o c i- — ' Q. u X .
«J3-D<»*JSS0CU— '
(/) 4^ oa
(O 41 Ol ^ <U
j3 aj j: t- — '
t/»r-4JOio.ciO(a(-*-''*-4J4-'
^ s
Z C u. UJ
u. _1 >— O liJ 4^ C
. , c^ tf) ~ *
(rt C « •'- u.
O (0 (- I-* V
3 (0 _. — *^
•O tf» Q-<- •*- O
4> C 4> *-• **- <-
X N4 (K 3 O O.
■q tfi
O C3 u.
AMADIAN
QUACULTURE
YSTEMS
7.0 INTERIM GUIDELINES & THE CERTIFICATE OF APPROVAL
In Ontario, the Municipal Industrial Strategy for Abatement (MISA) presents
a significant effort directed toward the reduction of water pollution from all
municipal and industrial users of water. This initiative presents a
conscientious policy targeted at restricting the concentration of pollutants in
discharged effluents in a manner which is in line with the implementation of
BATEA (see below). As technologies improve, the restrictions will become
tighter.
The two principal bases for abatement are (1) the best available pollution
control technology which is economically achievable (BATEA) and, (2) the impact
of the discharged pollutant(s) on the receiving body of water. In some cases,
the latter may be more restrictive than the former due to the inability of the
receiver to assimilate the pollution load. Although MISA is being implemented
on a sector-by-sector basis, and in spite of the fact that there is no sector
classification for aquaculture, the underlying philosophy of MISA is currently
being adopted in all industrial and municipal sectors. In aquaculture, BATEA is
generally applied, however, in some regions, where receiving bodies consist of
pristine streams and rivers, the more restrictive water quality track may be
applied.
By the power and authority granted under the Ontario Water Resources Act.
the Ontario Ministry of the Environment has established guidelines (see Appendix
V) for the design, operation, and management of fish culture operations in the
province. Within the context of returning process water to receiving streams,
these guidelines dictate that fish farmers must:
1) treat all process water, to a level of cleanliness as stated by the
Ministry, prior to discharge;
2) provide for appropriate storage and disposal of collected fish manure;
3) collect and submit regular (monthly) samples of influent and effluent
water for TSS and TP analysis;
4) maintain a log of daily operations at the farm; and.
5) report the findings of the water quality analyses and the data log to the
Ministry on a monthly and annual basis.
The Ministry has taken the position that efficient and effective solids
management is to be the focus of these guidelines. By reducing the concentration
of total suspended solids in fish farm effluent, concentrations of other
pollutants (namely total phosphorus, BOD, TKN) are also reduced. Moreover,
solids management is perhaps the only aspect of effluent quality control which
is realistically achievable in intensive salmonid aquaculture.
Current guidelines indicate that treatment facilities at fish culture
operations be designed, constructed, and operated with the intention of routinely
producing an effluent stream having a maximum concentration of 5 mg TSS/L above
the background level of TSS in the source water. An absolute maximum
78
concentration of 0.05 mg TP/L has been established for design purposes, however,
at sites using a surface water supply where background concentrations of TP are
greater than 0.05 mg/L, design criteria require that no increase over background
levels be incurred.
The operator(s) of a fish culture facility will be considered to be in non-
compliance with Ministry guidelines on any occasion when the concentration of TSS
in the final farm discharge is observed to be greater than 10 mg/L above the
concentration of TSS in the source water. Similarly, at no time can the
concentration of TP in the final farm effluent exceed 0.10 mg TP/L.
Notwithstanding this, at sites using surface water where the source water
concentration of TP exceeds 0.10 mg/L, farm effluent concentrations of TP cannot
exceed background levels.
These phosphorus discharge criteria may be severely limiting to future and
existing fish culture operations in the province. The requirement that effluent
concentrations of TP not exceed source water concentrations implies that removal
of total phosphorus from discharged culture water must be 100% efficient at those
culture sites where the source water already contains total phosphorus
concentrations greater than 0.05 mg/L. Since this is impossible, such fish
culture operations would be in non-compliance with Ministry regulations. The
only recourse for aquaculturists remains to undertake a study to demonstrate the
assimilative capacity of the receiving water course and the impact of the fish
culture operation.
The requirement for a manure storage lagoon (Sections 2.5 and B.l)'' for
retention during the winter is not contested. This structure, however, need not
be lined in all cases. Certainly, at sites containing highly porous soils, or
where the ground structure is unable to retain the vacuumed manure, a lined pond
is required. Several sites have soils of sufficiently low permeability that the
expense of a lined pond can be eliminated. Moreover, the manure itself serves
to seal pond bottoms. Consequently, the necessity for a lined pond should be
assessed on a site-specific basis.
On the same subject, research is required to assess the impact of overflow
drains from manure storage lagoons, which feed into subsurface absorption fields.
Such drainage facilities would greatly reduce the size and expense associated
with manure storage by permitting clarified supernatant to be removed from the
lagoon and may thus enhance fish farmer participation in effluent quality control
programs.
The requirement for routine sampling of influent and effluent process water
(Sections 2.7 & 3.3) raises considerable concern on two fronts: costs and
usefulness of data. The Ministry's viewpoint is that economics must be
considered in any business venture and that water quality monitoring is only a
portion of the overall cost structure of commercial fish production. While this
is true, the specific case of commercial aquaculture must be considered. In
Ontario, the aquaculture industry remains fragmented, with 25% of commercial
operations producing less than 2,000 kilograms of fish per year (Moccia and Sevan
^ Sections of the Interim Guidelines (see Appendix V)
79
%
ANADIAN
QUACULTURE
YSTEMS
1989). Consequently, the financial burden of monthly effluent monitoring will
be significant on these smaller producers. For example, sampling costs of $816
for an operation producing only 2000 kg of fish account for 9.25% of total
operating revenue as opposed to only 0.4% of income for the large-scale producer.
Furthermore, the purpose of collecting water quality samples is to compile
data on fish culture operations and to monitor effluent quality. Since the
responsibility of collecting the samples lies with the aquaculturist,
considerable variability will be introduced into the resultant data set.
Consequently, to be effective, an OME-directed or a co-operative monitoring
program will be required. This may entail a shared-cost approach and the
education of aquaculturists with respect to proper sampling procedures.
In concurrence with the DMA Engineering report (1988), the maintenance of
a daily record of weather events, power outages, etc. (Section 2.8) will likely
offer little to the interpretation of effluent quality data. It will, however,
impact upon the aquaculturists' perception of enforced effluent treatment, likely
creating greater discontent and reducing co-operative participation. Instead,
only routine fish culture records of feeding rates, fish transfers, etc., which
are more related to effluent quality, should be maintained.
On an annual basis, each aquaculture operator is required to submit a
report to the District Officer which summarizes all sample results, log entries,
and abatement procedures at the operation (Section 2.9). It is likely, however,
that these reports will be of little use to Ministry personnel since, in many
instances, fish fanners are not skilled in the interpretation of water quality
results. Moreover, given the same data, ten aquaculturists will likely provide
ten distinct interpretations. Such reports will lack the consistency and
validity required to make them beneficial. Consequently, on an annual basis, the
Ministry should consider in-house interpretation and reporting of monthly log
records submitted by individual aquaculturists or, alternatively, funds should
be allocated to contract this task to outside organizations. The task of
interpreting and reporting annual effluent quality records should not be
delegated to aquaculturists. To facilitate the information gathering process,
the Ministry should consider compiling and distributing standard "reporting
sheets" which direct the aquaculturist to provide specific information.
Lastly, from a design and monitoring perspective, the Ministry's two-level
approach (i.e. 5.0 ppm TSS for design and 10 ppm for monitoring and enforcement)
appears appropriate. Upon identification of a farm which does not meet the
established compliance standards, every effort should be made to determine the
cause of non-compliance. Sufficient direction is provided within this report to
enable Ministry personnel to assess the design and operation of production and
effluent treatment facilities and make constructive recommendations. Foremost,
every effort should be made to control the effluent concentration of total
suspended solids, since this remains the only practical course for effluent
quality control in commercial aquaculture. At the farm level, phosphorus control
can only be practically achieved via effective solids control. The specific
design and non-compliance limits for total suspended solids and total phosphorus
should be carefully considered by the Ministry's Aquaculture Committee. The
information provided in Section 6.4 (especially Exhibits 6.2 through 6.4) will
facilitate the Committee's task.
80
8.0 PROJECT SUMMARY
8.1 Ontario Aquaculture Operations
8.1.1 The three principal commercial fish culture facility designs utilized in
Ontario have been found to produce more or less the same amount of fish
per unit water flow through the system.
8.1.2 Generally, standard raceways and circular tanks are similar with respect
to production parameters and effluent quality. Oval raceways are
distinctively different, producing poorer quality effluent; particularly,
TP, DP, TKN, NH,, and NO^ have been observed to be discharged in greater
concentrations.
8.1.3 In all three designs, individual production and management practices are
not correlated with solids settling behaviour. Cumulative design curves
of sedimentation behaviour are different on a group basis, however.
Suspended particulate matter from oval raceway production facilities
displays slower settling rates than particles from standard raceway and
circular tank production facilities. The latter two groups are not
statistically different.
8.1.4 Settling behaviour is discrete in all three facilities.
8.1.6 In all three groups, production intensity is positively correlated with
enhanced concentrations of dissolved nutrients (TP, DP, NH3, TKN) in the
discharged effluent.
8.1.7 Feed-related factors (feed brand, feed type, feeding method) appear to be
subordinate to facilities design and culture practices with respect to
pollution impact.
8.2 Effluent Treatment Design
8.2.1 Overflow rate must be the basis for effective treatment facility design.
Specific design parameters are laid out in Sections 6.2 and 6.3. In all
cases, the required level of treatment efficiency will be governed by the
effluent concentration of total phosphorus since solids criteria are
generally easily attained, even at relatively high overflow rates. In
many cases, phosphorus criteria will be difficult to meet on a continuous
basics. To achieve compliance limits, the following overflow rates (in
w^/w/d) should be applied: Standard Raceways - 167; Oval Raceways - 40;
Circular Tanks - 65. These values will provide average effluent
concentrations of TSS considerably below the compliance limits; however,
TP concentrations will approach or marginally exceed the 0.10 mg/L
compliance limit. Further reduction of effluent phosphorus concentrations
will likely be achieved only from improved feed formulations, which have
a reduced phosphorus content (see Section 8.3).
81
Qanadian
TVouaculture
3vSTEMS
.2.2 Generally, the use of several smaller treatment units, each designed to
clarify a portion of the total system flow, is more appropriate than
having one large treatment unit to clarify the entire system flow.
,2.3 For maximum effectiveness, the design of effluent treatment operations
should be conducted or approved by qualified personnel having an
understanding of aquaculture operations and the related effluent
characteristics and constraints.
8.3 Best Management Strategies
8.3.1 The development and commercialization of low-phosphorus feeds is required
to facilitate the reduction of phosphorus from aquaculture operations.
Many commercial feeds currently contain excess dietary phosphorus.
8.3.2 At the farm level, improved feed handling and storage, and feed screening
to minimize the introduction of "fines" to the water column, will also
reduce the pollution impact of commercial aquaculture. As well, properly
managed feed programs can reduce feed waste and pollution by providing
feeds of optimum pellet size to the fish.
8.3.3 Fish culture facilities should be designed to maximize fish production and
promote the removal of solids wastes from the culture medium. To this
end, shallow tanks (<75 cm) are recommended. Tank flows should be
engineered to prevent solid waste accumulation in production zones, and to
facilitate solids removal in settling zones. Moreover, standard raceway
and/or circular tank facilities are most appropriate.
8.3.4 Vacuuming of settled manure from the bottom of clarifiers remains the most
effective method for waste removal.
8.3.5 Due to the highly liquid state of settled fish manure and its low-grade
fertilizer properties, application of settled manure on agricultural land
remains the most appropriate method for disposal of solid wastes from fish
culture.
8.3.6 Assuming that treatment facilities are retrofitted into existing fish
culture operations, production costs can be expected to increase by
approximately $0.08 to $0.10 per kilogram of fish produced. Total
effluent treatment costs are greatest in oval raceway operations and
lowest in circular tank operations.
8.4 Enforcement of Ministry Guidelines
8.4.1 If the current total phosphorus guidelines are enforced, it is likely that
several farms will not achieved compliance limits due to the concentration
of total phosphorus in their source water. As well, it should be noted
that the best available technologies which are economically achievable
(i.e. sedimentation and/or filtration) may not be sufficient to reduce
phosphorus levels to compliance limits due to the solubility of phosphorus
82
in fish farming wastes and the excess concentrations of phosphorus in many
commercial feeds.
.4.2 Before the interim guidelines are finalized and enforced, nore definitive
research is required with resoect to phosphorus loading from aquaculture
operations.
.4.3 Research is required relating to ihe discharge of clarified supernatant
from manure storage lagoons into subsurface absorption fields.
.4.4 The Ministry should meet with the membership of the Ontario Trout Farmers'
Association to discuss the final draft of the interim guidelines prior to
their implementation and enforcement. This will enable both parties to
offer their concerns and it will promote industry-government co-operation.
A co-operative effort will best protect the environment.
83
%
ANADIAN
QUACULTURE
VSTEMS
9.0 LITERATURE CITED
Al-Layld, M.A., S. Ahmad, and E.J. Middlebrooks (1980). Handbook of Wastewater
Collection and Treatment. Principles and Practice, Chapter 13:
Sedimentation, p. 177-209. Garland STPM Press, N.Y.
American Society of Civil Engineers (1959). Sewage Treatment Plant Design.
Manual of Engineering Practice No. 35, 375 p.
Castledine, A.J. (1986). Aquaculture in Ontario. Ont, Min. Nat. Res., Ont.
Min. Agric. and Food, Ont. Min. Environ. Publ., Queen's Printer for Ont.
Camp, T.R. (1946). Sedimentation and the design of settling tanks. Trans. ASCE
111:895-958.
Campbell, N.A. and W.R. Atchley (1981). The geometry of canonical variate
analysis. Syst. Zool . 30(3) :268-280.
Cho, C.Y. (1990). Personal communication. Fish Nutrition Laboratory, Dept. of
Nutritional Sciences, University of Guelph, Guelph, Ontario NIG 2W1.
Cho, C.Y., C.B. Cowey, and T. Watanabe (1985). Fin fish nutrition in Asia.
Methodological approaches to research and development. Intern. Dev. Res.
Ctr., Ottawa, Ont.
Daley, W.J. (1989). Waste water: current and potential treatment methods and
uses. Presented, AFS Program, Aquaculture '89, Feb. 12-16, 1989, Los
Angeles, CA. 3 p.
Dobbins, W.E. (1944). Effect of turbulence on sedimentation. Trans. ASCE
119:529-640.
Dorr-Oliver, Inc. (1982). DSM screens for the food industry. Bulletin DSM 5.
10 p.
Dupont, D. (1990). Personal communication. Dept. Agricult. Econ. & Business,
Univ. of Guelph, Guelph, Ontario NIG 2W1.
Eskelinen, P. (1986). The phosphorus balance of rainbow trout. Vesihal lituksen
monistesarja 241:33-42. Can. Transl. Fish. Aquat. Sci. No. 5274. 13 p.
Fair, G.M. and J.C. Geyer (1958). Elements of Water Supply and Wastewater
Disposal. John Wiley and Sons, N.Y. 615 p.
Fitch, E.B. (1957). The significance of detention in sedimentation. Sewage and
Ind'l Wastes 29(10) :1123-1133.
Gandemer, J. (1978). Wind shelters. Proc. 3rd Coll. Indust. Aerodyn. Aachen,
79-108.
84
Giles, J.H.L. (1943). Inlet and outlet design for sedimentation tanks. Sewage
Works Journal 15(4) :609-514.
Hagen L.J. and E.L. Skidmore (1971). Turbulent velocity fluctuations ?nd
vertical flow as affected by windbreak porosity. Trans. ASAE 14(4) :534-
637.
Hazen, A. (1904). On sedimentation. Trans. ASCE 53:45-88.
Hilton, J.W. and S.J. Slinger (1984). Trout farm effluent study. Contract
Investigation for The Ont. Min. of the Environment. 26 p.
Hopkins, T.A. and W.E. Manci (1989). Feed conversion, waste and sustainable
aquaculture: the fate of the feed. Aquacuiture Magazine 15(2):30-36.
Hubbel , G.E. (1934). Experiments with inlet devices for sedimentation tanks.
Sewage Works Journal 6(4) : 774-783.
Ingersoll, A.C., J.E. McKee, and N.H. Brooks (1955). Fundamental concepts of
rectangular settling tanks. Trans. ASCE 121:1179-1218.
Ketola, H.G. (1985). Mineral nutrition: effects of phosphorus in trout and
salmon feeds on water pollution. In: Cowey, C.B., A.M. Mackie, and J.G.
Bell (eds) . Nutrition and feeding in fish. Academic Press, London, p.
465-473.
Ketola, H.G. (1989). Personal communication. Tunison Laboratory of Fish
Nutrition, U.S. Fish & Wildlife Service, 28 Gracie Road, Cortland, N.Y.
13045.
Lewis, M.R. (1979). Fish farming in Great Britain - an economic survey with
special reference to rainbow trout. Univ. of Reading, Dept. Agric. Econ.
& Mgmt., Misc. Study No. 67. 74 p.
Liao, P.B. (1970a). Pollution potential of salmonid fish hatcheries. Water and
Sewage Works 117(8) :291-297.
Liao, P.B. (1970b). Salmonid hatchery wastewater treatment. Water and Sewage
Works 117(12) :439-443.
Lindqvist, O.V. (1986). Fish farming in Finland. Vesihal 1 ituksen monistesarja
241:5-8. Can. Transl. Fish. Aquat. Sci . No. 5273, 7 p.
MacCrimmon, H.R. (1984). An overview of aquaculture in central Canada. In:
Pritchard, G.I. [Ed.], Proceedings of the National Aquaculture Conference -
strategies for aquaculture development in Canada. Can. Spec. Publ . Fish.
Aquat. Sci. 75:42-55.
Mayo, R.D., P.B. Liao, and W.G. Williams (1972). A study for development of
fish hatchery water treatment systems. Walla Walla District Corp of
Engineers, Walla Walla, Washington.
85
Qanadian
t^quaculture
gYSTEMS
McLaughlin, T.W. (1981). Hatchery effluent treatment, US Fish & Wildlife
Service, p. 157-173, In: Allen, J.L. and E.C. Kinney (eds) . Proc.
Bio-Eng'g. Symp. for Fish Culture, FCS Publ . 1, Fish Culture Section,
American Fisheries Society.
Merican, Z.O. and M.J. Phillips (1985). Solid waste production from rainbow
trout, Salmo gairdneri Richard^ion, cage culture. Aquaculture & Fisheries
Management 1:55-69.
Michigan Department of Natural Resources (1973). Michigan salmonid hatchery
water quality evaluation. Mich. D.N.R., Lansing, MI. 234 p.
Moccia, R.D. (1988). Personal Communication. Aquaculture Extension Centre,
Univ. of Guelph, Guelph, Ontario.
Moccia, R.D. and D.J. Bevan (1989). AQUASTATS. Ontario aquacultural trout
production in 1988 with an historical perspective of the industry's
development. Ont. Min. Agric. & Food, AGDEX 485/42. Ord. No. 89-113. 4 p.
Moccia, R.D. and A.J. Castledine (1987). Perspectives on aquaculture
development in Ontario. Proc. Ann. Meet. Aquacult. Assoc. Can. 1:98-99.
Mohlman, F.W., H.A. Thomas, Jr., G.M. Fair, R.E. Fuhrman, J.J. Gilbert, R.E.
Heacox, J.C. Norgaard, and C.C. Ruchhoft (1946). Sewage treatment at
military installations. Sewage & Industrial Wastes 18(5) :791-1030.
Mudrak, V.A. (1981). Guidelines for economical commercial fish hatchery
wastewater treatment systems. p. 174-182, In: Allen, J.L. and E.C.
Kinney (eds). Proc. Bio-Eng'g. Symp. for Fish Culture, FCS Publ. 1, Fish
Culture Section, American Fisheries Society.
Mudrak, V.A. and K.R. Stark (1981). Guidelines for economical commercial
hatchery wastewater treatment systems. U.S. Dept. of Commerce, N.O.A.A.,
N.M.F.S. Project No: Commercial 3-242-R, Penn. Fish Commission. 345 p.
Ontario Ministry of the Environment (1982). Biological surveys of five trout
farming operations in southern Ontario. Water Resources Assessment Unit,
Southwest Region and West Central Region. 25 p.
Ontario Ministry of the Environment (1984). Water management - Goals, policies,
objectives and implementation procedures of the Ministry of the
Environment. 70 p.
Parjala, E. (1984). Reduction of the loading effects of fish culture - a
literature review. Vestialous 5:11-16. Can. Transl. Fish. Aquat. Sci . No.
5252, 1986. 17 p.
Persson, G. (1988). Environmental impact by nutrient emissions from salmonid
fish culture. IN: G. Balvay (Ed.), Eutrophication and Lake Restoration.
Water Quality and Biological Impacts. Thonon-les-Bains.
86
Plate, E.J. (1970). Water surface velocities induced by wind shear. ASCE J.
Eng. Mech. Div., Trans. ASCE 96:295-312.
Pruder, G.D. (1989). Aquaculture effluent discharge in the United States:
issues and conflict. Presented, Fish Fanning Expo III, Dec. 8-12, 1989,
New Orleans, LA.
Rich, L.G. (1980). Solids removal and treatment systems. Chapter 5, In: Low-
Maintenance, Mechanically Simple Wastewater Treatment Systems. McGraw-Hill
Publ. Co., N.Y.
Samuelsen, 0.8., A. Ervik, and E. Solheim (1988). A qualitative^ and
quantitative analysis of the sediment gas and diethylether extract of the
sediment from salmon farms. Aquaculture 74:277-285.
Solbe, J. (1988). Water Quality. p. 59-86, In: Laird, L. and T. Needham
(eds), Salmon and Trout Farming. John Wiley & Sons, N.Y. 271 p.
Sondergaard, M. (1988). Seasonal variations in the loosely sorbed phosphorus
fraction of the sediment of a shallow and hypereutrophic lake. Environ.
Geo!. Water Sci . 11(1) :115-121.
Stechey, D. (1987). Alternatives to sedimentation practice in clarification of
aquacultural effluents. Industrial Research Report for Terlyn Industries,
Ltd., Pickering, Ontario. 23 p. + App.
Stechey, D. (1988). Factors influencing the design of effluent quality control
facilities for commercial aquaculture. Proc. Ann. Meet. Aquacult. Assoc.
Can., AAC Bulletin 88(4) :208-210.
Stechey, O.P.M., J. Dobrowolsky, and J. Renaud (1987). Growth opportunities for
commercial trout aquaculture in Ontario: an industry analysis.
(manuscript) .
Stechey, D. and J. Kantor (1988). Ontario trout farming: a cost comparison of
various alternatives. Proc. Ann. Meet. Aquacult. Assoc. Can., AAC
Bulletin 88(4) :161-153.
Sundstrom, D.W. and H.E. Klei (1979). Wastewater Treatment. Prentice-Hall,
Inc., Englewood Cliffs, N.J. 444 p.
Thomson, D.E. (1986). Determination of the effects of fish size and feed pellet
size on the settling characteristics of Rainbow Trout culture cleaning
wastes. M.Sc. Thesis, Univ. of British Columbia.
Trojanowski, J., C. Trojanowska, and H. Ratajczyk (1985). Primary production in
lakes with cage trout culture. Pol. Arch. Hydrobiol. 32(2) : 113-129.
UMA Engineering Ltd. (1988). Wastewater Treatment in Aquaculture Facilities.
Ontario Ministry of the Environment, Queen's Printer for Ontario.
87
%
NADIAN
QUACULTURE
YSTEMS
U.S. Environmental Protection Agency (1975). Process Design Manual for
Suspended Solids Removal. EPA 625/l-75-003a.
Weismann, 0., H. Scheid, and E. Pfeffer (1988). Water pollution with phosphorus
of dietary origin by intensively fed rainbow trout {Salmo gairdneri Rich.).
Aquaculture 69:263-270.
Westers, H. (1989). Water quality impacts of aquaculture. A review.
(Manuscript) . 34 p.
Wetzel, R.G. (1975). Limnology. W.B. Saunders Company, Philadelphia.
Wheaton, F.W. (1977). Aquacultural Engineering. Wiley-Interscience, John Wiley
& Sons, N.Y. 708 p.p.
Willet, I.R. and P. Jakobsen (1986). Fertilizing properties of trout farm
waste. Agric. Wastes 17:7-13.
Willoughby, H., H.N. Larson, and J.T. Bowen (1972). The pollution effects of
fish hatcheries. Am. Fishes U.S. Trout News 17(3):6-20. Cited in:
Westers (1989).
Zeigler, T. (1988). Effluent management of trout farms. Presented, 1988 Ann.
Conv. Ont. Trout Farmers' Assoc. 25 p.
88
APPENDIX I
Field Survey Questionnaire
89
^-
ANADIAN
QUACULTURE
YSTEMS
AQUACULTURE WASTEWATER TREATMENT
DATA SHEET
Farm No:
Date:
Fish Farm:
\kg)
(#)
Operator:
Address:
Species:
Foodfish (lbs
Fingerlings
[] Martin's
[] Floating
[] Wen(s)
OPERATIONS:
Annual Prod'n:
Feed Brand:
n EWOS n other:
Water Source:
Volume (IGPM\U
[] Sinking
[] Spring (s) [] Surface
[] Gravity
SAMPLED UNIT:
[] Tanks
[] Raceways
Prod'n Area (ft\m) :
Prod'n Area (ft\m) :
Dia.
Deep
W
D
Carrying Capacity in unit (lbs\kg\# fish):
Approximate Size of Fish: (cm\in\lbs\g) :
Feed Method: [] Demand [] Hand [] Automatic [] Other
Water Supply: Q (IGPM\LPM) [] Flow-Thru
[] Re-use: %
Temp (C\F):
EFFLUENT TREATMENT:
Method: [] In-Raceway Settling [] Ext. Settling [] Other
Dimensions:
Hydr. Load Rate:
Flow Pattern (IN/OUT):
Comments:
HOE SAMPLE LOCATIONS:
Sample 1:
Sample 3:
Sample 2:
Sample 4;
90
r-1 OO OO -O ~£ Q
3 o o o o o o
3 '"' o o o o o
It r-i ^ oo r<- eo p^
Lu a- ir» ry in f^ ^;
t- o «- «- O O
« """ d d d d d
<
p— u> ro z r- «*
O O O O
I d d d d
^ r- oo r>.
oo»-oooooa
ddddddooooo
in < O ^ f^ ir\ CO
O zrog
53
a P 3 := ^ S S S 3 ?i S
SryOO»-oco
3:?
i d d d o
r-OK^'OvOO^O
PU (M
O r-
ddddddooooo
dddddoooooo
o a
d d
Ooo<Mrvjinp~2;^5-oo^JCg;C2SrC3!SN:
^(\j(\iinind-*od^dod>ra3>od'-rv.ro»--nj
(Ni
r\j
O O
oo O
r-. ao
ddddddooooo
ooo«-f^<^J'^~'"'£;2'^!i5
?:o^no-oioo«-ooo-too
K!io»-(\j(MnjOrnO-T^
rt 00 o o <•
O o
■O !0
O r-
d d
§ 8 8 S o 8 S S S S S J: 8 8 5 8 8 fe. S 8
ddddddddddddddddoaao
o -*
•o in
r- o
d d
»- t^
dddddddoooo
Otor>-ooOi^Ooginogoo
~Jr-0(\l'^0«-«-mf\ir-
ddddddooooo
s8ofyooooruoor-;OOoaor-oo
ddddddddddddddddoooo
5S5SS85fCS:S3;S3f^§S!C3!?SSS ?^?^5!S8 8?8
a oo
iot\ji>o^^oo«»a
^_ u
•J' rg
O O
d d
o o o o o o
r^ -o
00 -^
Kl O
d d
ro «- 00 •* -O lO ~J
— - — «- O
d d
1*1 nO >o
^ in
O fV4
d d
3 3
ry >!■
88
d d
S8
I CO ro m N- r- nj
I ?- «- o Q oj >T
1 O O O O O O
idddddddddooooo
■ O ro Q t^- O (M ro
■ OJ ro O to QJ r- r-
, O O O «- O o o
; S 5 ?i S :§ Si ? § g 2i fe S i 2
i o o a rv» o o ^ o o '- r- o a ^
Idddddddddooooo
in h-
vi- d
ooO'-^«4'^";'^~<'S"JS"!Z*
i(vii^r^»J'^fMOinu-iOOrv.»-_
!^d«-r-ddr-^t\isr>roinc\i
r-^roro-Joo-or-r-H
0?-«-»-0 0r-'00<-
dddddooooo
.0(N-or^ror--Omrop
itnr-Ji-3-~r-o(Mininrs-
;o^-(MtMO»-ro-o^f\i
• dddddoo«-oo
njinojiOvOr«-OM
p^oo^«~r*-r-ropo
to<)«-d«--Ofor--rom
fc:8
fsj •** CO ro «- O
f\i r- »- ^ o tM
o o o o o o
d d d d d a
.0-00nj0«---i----
•<friio>or^^mona
_ . O ro «- O m
iro-viQi^ggS
0 ^ oo rvi
1 S S 8 8 :? S S o a o 5 5 g 3 g,
Idddddddddddooooo
d d o o o o
o ro
d d
fVi Z O O ^ >T
>0 >o O in ■^
^o (M ro o »- g;
1 8 8 o 8 8
d d d d d d
in o ro -o og O
088080
d d d d o o
S^ z o (NJ 5 o
(\i d ro rii o
o o
d d
o Q
ni o
vI-inoooo^OroiM-4-PU-5inO-0-0()J
SSjSriifM^inr-rur-^r-KIgg^
T-t-r-ooruO*— ^"00<-^oo<-^
ddddddddddddddoo
^OvO'NiOOOro-^
^r^rooorginr^fM
T-r-00<-roO^
dddddooo
° [;; |C ? :;; :§ S ?; 8 S 8
r-or^fMriiO^-ror«-«-ro
r^ddddddd«-oo
ro ^ i> m
T-in>^oo^O^^Qo0^f\i<
inror^ro-o«-inP>-mOrooO<>z
lnd^d«»<-■<^dooodro^-Jln^-o
ro O 00 (Ni ^
>» r»- (M f^ 00
^ ro ^o rvj O r- >T
d m ro -o «* r^ <M
O O
00 <-
8S.
d d
r>- ro
~4- O
rij oi
> 00
18
d d
CO o
r- ro
(\i ro
585SS8S§8838§8S8
ddddddddddooooo
<-T•000-0^-0>»;»l>;qrOr--ln^-
^ 8 3 8 8 8 8 S 5 8 o 8 3 5 8 8
ddddddddddooooo
8St8S^^SSS3?5S
d«-di>-i^'-dd>o(M~»'0
, . , , na ro ru r-
I g O Q Cy Q (NJ
■ «- sO f\l
""000
ddddoooo
000
ddddoooo
nj nj ^ o ro
^ O ro
d d ivj
*Or-«-roin^>rroco^orij
o CO t— ro o
■ O <— '^J O r- oj
nj>if\iOOOO^
(\j(Njt\j(\ininjrurvjf\jf\j(M
ln»»oOln>^•>^•tM^o-or«-
>Oin^Oror^inininojn-r*-r^nj
%
ANADIAN
QUACULTURE
VSTEMS
ryoooooooooaoo
gzgggg^S
oooSoooooooo
3>_i ooooooooooooo
o o o o
oooooooooooo
K>«-»-o«-Or-o«-*-t-r-o So oooo of\i o«-ooaooooooo
ooooooooooooo
>»(\IP-r->OOOOOOOOOr~in
tno^orvjOO'^CNJON-u^inr-
o o o o
do a o o d ad dddddddddddd
0^-<u-l^-Ml/^<0>r<oo^-^-IMO^oo-ooo^l^>
>Tr>-Z00(\J00OZ-*«-ZK1(MOOr-
sO >0 Kl «- ro e-
Z (- r- O"' rO O fO
-» O m O O <M o
Oror>-ot>OO^o>ic\j
o.ii
I— ■ o-
~ s
m>OT— *-(>^»0^<— _
^or^jg^CT^Ot^J>»'OOoo^oof^J
t\JOOOOO~~ ___
a o o o o o o
- — -- orvjgggg^g^t-«-^ogggggg^gr-fi
ooooooooo»-aooaooooooooo
ooooooooooooo
oooooooooooooooooooooooo
OoO'j'-^ob(\ir^mi_ ,_
r-»-Or-0«-0»-»-T-»-0
o^^^p^^-ao^o>r«-^r^'Oaoo^aO(^Jcooo^fMfoao
(Mtr»ryfMOKi^cvjf^oOf\j*-h*aot-^Ou^roy^Oro^
oooooaooo(NO»-«-oooooooooo
ooooooooooooo
(^Jf-^-o^->T^-^ooor^t>flOl^\
t-N-ON-ou-rruroor-o^-r-
ooooooooooooooooooooooo
^»fOO^^-l/^0^-oo^-^-f^J^Or-oO'0^*u^(^i^o
Nj- irt f\j f\j T- >T ao
>T o >r o (M o
Ocxjr^o»-'0^-oo
JO(\jroio-OOu><\jro^O> Kleo»->»ln••*^O^u^Oln^OrOfOOO^-f^Jr-oo^-(^J•0'0
OOOOOOOOOOOO
rvj r- (\J (M
g z g g g g = S - = 5 g
»-ggf\]~ogKit>-
S8SSS88S
ooooooooooooo
«->4-(M».-o»-ir>»-o>»OrONj-
>»^^.^ogmoo^^•(^J^.Kl
(\jO»-0«-0»-«-«-r-P^r-0
OOOO
^. «- fo »-
^ m ro fO
- - z-or^zm>o^g^
OOOO 0<M «-t-ooo
gooopooooooo
0- GO
1- o
(\i C GO o >*• o
<- I- o g «- fy
o o r\j o o o
ooooooooooooo
1
-om^o»-o^•^nu^^oo«->t
Ofv««t^-moo^•oO'0•Ofl/^
OO OOOO OO OOOOOOOOOOOO
^^-<^n^.Kl^<0>T<^o^-^^<MOm^r^^ooo^o^n
>or>-z»-ruooflOZ>Tt-ZP>-ryoo«-Kim-oo»--Or-
O0ro>»0^oj0ino0t«i
ry r»i <- K> «- t-
^- 1/^ oj f\J
^ m r* o ^ o
>0 ^ (NJ Ki ro
r»3(\ig^mgg-oi>-~»o>rg grjzggg«-g^r-»-^gggggggoggMi
ooooooooooooo So SSSoSo^ocoooSSSSooSoi
-!>-> ooooooooooooo
ooooooooooooooooooooo
3 r-i
a <n
o «
ir t—
a. uj
Nr-Otnininoo-*-or-tnir<«a-
-»0(MO»-»-<ooooor>-rut«-'0
r^ -4-
5S
•«aot^coo-»»-m-ooo>roo(Mcoooory»-GO»*
ZOOfO-^cviN-ooryinf^cOf-^Qr-ro^rurj-Ot-
«-SoOOO(\JOru»-OOOOOOtVOOOO
ooooooooooooo
>00-.»<Mr^Ki-j'>oo^Or^u->
o o
O r-
OOOOOOOOOOOOOOOOOOOOO
«fnO-*■0u^O^-a0(^J^.(^J■*t^«-^n■C^-^/^»-■0O
z^rtc-^^-«->^•o^•J•(\lOln^-o^or^l»1«-o^-^«-
^-rooo>o^m^Of^JO^o^-^^l
0^i/>-4'r»->j'mo^m'Omroo^
p^ %r ^o fO
ggggggggSgggg gs=5ggg=gg=sSSSSS8SSSSS
a:i_i ooooooooooooo
OOOO
OOOOOOOOOOOO
OOOO
fn(Mm>o^«-mrO'OKiirtf\iO ryoo<«— oOlr»ao<^-^o<ooa^>fO«-oo^-<— ^^J^o^^O^
gOi^ag^gg^gggfo r-vr2T-u^g»-zggzog«j-ggggr-ir\i-orn
SooSoQOoSSSo OO ooSo S3 oSoSSSSooooo
do dddddddddddd
Sro<ogLn-ogaoooo-o<3oO'-
ooz^ooiAO>oryr-»Ja^fNjNO
d<M vddinodorOinfNJ^rr-to
OOOOOOOOOOOOO o
ooogor--»ry>rgr>-o«-g g
mt^0^r\j«»r^'00-»»»>oS O
(Mr-0<Mdrvjr-ddrOfO«-d di^
■* «t QO p-
fO Kl sO ^
oj d «- -*
ro ro m fO fO fO fO
ro ro fO m
^«»>T>»"*>f>»'~r>r>t>>-^>Tvr>»'>r>r>r>rNT-»sr^>r
^>oor\j<-inao^O(MO>>rrviO
f\(T-mK>»»ry»-(MT-<MIO(Mr-
(\if\iP.>oooroo«-o^--*r>-ro«-oo«-N-in
>»'f*>>o^'Oir»»o>rN.u>in«-
oo fvj o ru ^
>0 P~ >r ^- 00 ft -o
o.^i
ZQ. 2«-cz<-c c;3ot-ooo
= 1-1 d d d s (6 cddo'odd
q:— 1 d d ^ ^ :^ csoddod
<
UJ CO
a. S»-o<-a ooo«-«-ao
zi-i ddddd ddddddd
_j cotrtaoQiri o^-^?o^/^aOl/^
u-a. Or-0~TO «-00«-r-0«-
1-1 ddddd ddddddd
z
a:
ikc/i oo.*OLrtN> tncoroOO'OO
«
^ ~Try-Oir\>T ki»-(MOO«-0
1-1 «- (\i t- »- (M
UIO. OOO.-NO 0Q0»-000
-ii-i ddddd ddddddd
u.
zo. ooo-oo «-oo^noo«-
oi-i ddddd ddddddd
o
(/J N-K-N-OP"- -0r0O(M«*O(\i
V- </>
Z 1-1 «- •- r-
-JQ. OOOtMO ooo»-oco
-ii-i ddddd ddddddd
ui ooi/^tooin or^-oo~TSJf^
ZO. OCO-OO r-00K1»-0<-
— i-i ddddd ddddddd
u
31—1 'OfflN-Qiri >*rurga*'0(M«
Qi/5 OBroOP^ •OD^TrMiri'Org
O t/9
OC^- ^>0'0r-*^ r«-*->T<\}»4'r-Lri
a. 1-1 »- <M «- (\i
rginfo^u^ flOr-<— (Mpn*-<—
2^ Scoco SSoSoSS
u a
asi-i ooooo ooooooo
o
^a. aScSo ooooooo
zi-1 ddddd ddddddd
lA-crt f\j«->Ow^o o-^'oo^tno
•-**— F^lt-^Or*! OOOOfNi-O^
a.
O ^ "1 >o N-
O O ^ O^ pn «- rg
^» -O K» <0 QO 00
^.
?^
3
.-vj
>
=
o
=
=
O
^
■o
in
J^
,^
o
:\j
»—
o
"•
'^
**
"^ ■
Ai
•n
3^
_
S^ 1
■o
p^
~r\
23
Jl
c
^
-^
ro 1
(M
_
^
-M
OJ
o
^"
o
C
O
o
o
o
=
o
o
>
3-
p-
■o
-O
o
o
!M
o
o
o
O
O
>0
a
^
m
o
-o
o
u-1
>0
>
•^
r>.
nO
>0
~o
CO
o
o
o
O
o
o
o
=
o
o
OO
■o
in
^
s
^
a
o
c
o
=
=
o
<M
tM
^
=9
>o
03
in
c>
•o
>
■J-
^
ro
O
~o
Q
ro
fO
(M
o
o
o
O
o
O
o
?;
-
s
s
nj
nj
^
o
o
o
o
O
^_
~T
o
o-
O
33
■o
33
OO
^
.n
f*^
-O
^
"
"
■•B
3
0)
c
«
V
o
a:
3
a.
•n
-0
a
c
ac
T)
3
r
u
«
a
^
■0
trt
o
w
UJ
>-
M
•o
>r
lO
z
2
z
z
z
<
<
<
<
z
Z
z
z
z
Qanaoian
>\quaculture
Systems
APPENDIX III
Spearman's Correlation Matrix for Participating Fish Farms
94
IS =§ ss
t 2 t
i I
i
1 |i |. p is p p si is |i *i 11 ii ii |i == al si p 2I
% §§ io §3 11 is li
SS S «j root ^. S
O ^ O ^ O O
§ i I 3 5 a
13
O
1/1
oJ^tN^oS-lTvOcoO-S^ <r -^ so o S SrM ooryatNSJS;:?---
0<Oc\i-OCT"OjaCT^rooO ao;^r-vOO roro (MsOia-o~j>-0'0
OfNJOr^o'r^fNJo'oOC! t-doOO ^O Ot-«-«-0»-C^
«-^ddr\jdr^r^«-^dd«-^ <-«-d^r- «-d 5-r-«-cory.-c:
co«-p~r-r— oc^^-oooo<^ oooinr>.oo O-o ooooct-"<i^>o
di-dt-ddddddo «-r-ddo »-o o»-'-acc!3^
for»cof^Ow^«-r--c-unvO aM>^CT*'0 o<— ^a--0'>o^^2;
r-r-»-^«— r-ddd«-dd «->>'rodo oo o<-«-oocoo
fNjOOOroOflO>Tiritn^ por^Nfoo^* (>oo f\j-coOP^>OQ
fviOO«-f*io^^Ou^pONOC- r-r^r^>Tfy oo-o roLnpor'-c>j>oa-'0
«-^dd<-ddr^dddd -r-oodo oo «-r-t-ac:oc:o
On-OKi-OforgfOOiriO OfO-OMO ^fo roroinC;ri;^^^
un^LnfOOfOLAr-r-rOf»*- v0(>fM>J'O «-fO ^ •^>GO^co
■O i> -o
>- -o ry _ . __
CO oj ~o r- r— O o
Ki r\i o
-» O pn _ _ _ _ _
f\I^NO>JOu^<-fMU->CM
dddddcddaad
0^rgOi/>*o>C^N-r^rgO
.>?-fO?-oco-oc\i,^-ir«'0
xj-soco-or^oo>rr*^N-?or--
ddddddddddd
»■ 3
OinmOOoooOM
r^d<-^dddd^
-O r^ qg P~ ^C
oa p^ O ro PO
vO ru (M PO T-
ru (M «— O fO
d d d d d
O sO QO 00 LTt
r^ d d d d
-o ^ o -o o
00 ^ o> r^ -o
d d d d d
<o o
po ro
^■o-o(^rMOO>J
oooooooo
I I I
P-O^r-POO-O-Oiri
-003>^PO^(Mm
OoO^POOr-23^
dd«-dd^do
ry-oaoc\!r\jO'--^
in^-^(\jmopoco
>»• f^ 5- <- Q f^
PO (> in oo O ~»
O Q O O in
m O ? PO -*
S O S i> f-
vl- o
(NJ (\J
^inOOOOOQ
Oin.*in>Tt\iinO
u^ fv -o u^ CO in
CO r». p>- oo T- vt
>r «- O >- m
PO (\J
V m m ^ O
i> P- ^0 r^ vO
m m
in >»• O ^ m
PO a
in -o
**-OOOpo0^n.
^ PO ^-. r-
O^OCOpOOOO<-
■Opocoo>T«»-o«-inr-r~
«-vOOfMO^inojO(Minnj
inmpooO'-T^po-oor^tv
O nj r-- O in « i>
PO »- »» 00 !^J O in
ry o in <M ^ p^ oo
r- (\J O ^
(M >J PO PO
t\lO(^Jpo^-^noo«-^(^J'0
njinin»-pO(\j<->4'^(M
-J- >3- O «- O
o r~ >J- o ^o
d d PO d d
o ro
d d
t\if\j!\jinpo^cNrM
moor^oooo
dddr-^dddd
p^ininOcofy>T>OoOPO>»
PO d PO m ps! (M 5-
po ^ o «» r" — ~
fNj ry r\j »-
popopo^o^po«o
p^ <- r^ -o m
vi PO 00 rvj r^
oj oo 00 «- «-
CO m O m i> PO
OO^POP*-000-OfVJinOO r-
fM^inopo«-fMcooor-<- p^
pop*^Oijpo(\i>oar-OPO-o in
O0(\i«- intvjr^NO «-f
in PO ru
vo o> i> r-
«- O PO IM
^0 (M
PO O
•O O
>jNOP'*<-njrupo-o
OOOOO-OOfMCO
^ooo^oo>»■oo-J■o^^o
•jpop^porudodin-o-OPO
PO o
p^ in
.n ru
forgr^p«-r^**ruO^
»-f\jf\i«-<\j«-fu(\jr-inin
poin-ooOCT'POvj'r^inoni
ru (M ry f\j (M
rg rvj ^ m r»» 00
popo NT>r>*^^>*>f*^
rvjpopopopopo^oj
^ocoij-^ino^o
ru nj rg PO *T in 1
Qanadian
^|\quaculture
Systems
a- o Q r^
-o r^ g fo
in o o ^^
«- O ro d
nj o Q 00
o o o »-
0> ro (\l >0
o^ »- O >r
d a rvi d
CO >o r* ^
o o> >0 <-
oo <o t- fo
.-^ d ~» <-^
r^ >»• o 00
«- 00 tr» Pu
>0 PO (\i fO
d d ^ d
83
O O (\J o
«* t^ «» Kl
in ^• ^■ rvi
o r\j ru -o
a* ro ro r\j
d d r^ d
in
f\J
oo o
1^ tj >0 m
d d <-^ d
<o f\J o >o
T- vn ?— ^
rvj m »-
(^ <J «- o
S:*-
o
(M
m o^
oO S
d
d
d d
O r-
tO (M
d
d
r^ d
IC
8S~
c>
*-
S"
o
r*
CO (N
i
N.
s
Ki
'^
m «-
vn
00
^0 1*1
in i>^
»» 00
in
(M
5
CM f\4
OO <o
t- -»
o
Kl O
c n)
fo"
«» «-
(^ O O CT>
o r^ p^ i\i
AJ in (M sO
<\1 OO OO PO
^ a r^ <D
^ ^• in CO
o »- o> N.
O m 00 -*
d d «-^ d
r«- o OO o
wi m OO oj
»- r- -o in
r^ N. Kl 00
00 r\j »- O
O nj fO »-
d d d d
»- r^ -o »-
in r- ^ «-
OO «» ^ o>
P^ vn o t-
d d r^ o
rvj r- Q P^
r>- ^ 5 «-
d d d d
m 00 O PO
^0 00 O ry
si >d r>^ -»
(\i m (M
a o in nj
o o
^ t>lj
00 PO
s
d d
'-
d
N. PO
o
00
d d
^
d
§§
1
§
•^ ^
«-
d
OO »T
fO
PO
1
d d
^
d
00 o
5 S
O
1
d d
'-
d
in r-
o
»o
OO
Sj
d d
d
d
^i
00
2;
CO
s
p«- o ^- m
-o -o
CO i> a ro
<- d od IN
O ^ O' ^
r~ ^ r\i \/\
oo ro o rvj
d d 'T^ (D
c a o r^
r- O -O (\J
oo Ni r>- ^
d d «-^ d
^.
f".
in OO
§
-*
m
fO (\l
*"
*"
~o
^
o -o
(>•
o
^ (M
d
d
PO r^
i>
"*!
r«. o>
*-v OO
rg
S:8
in
II
c
-S
PO
q 00
«»
PO
r- >0
b t-
(NJ
«
3
gio
*"
r~ «"
a <>
<-^
o' r^
*! rvj »- >o (\j
o o o o
CM r- PO r-
00 ^ O CO
o fM r^ ru
r>- >0 o 00
«» in m o
d d d d
^ in rj o
§vg >j ro
00 PO r^
CO ru »-
r^ d r-^ d
>o >o ^O o
PO O ^0 PO
to lO lO o
nj ^ o OO
r>. (\j (\j >i
O in ^o in
P^ PO O PO
r~ r^ 00 O
PO r^ 00 -o
ru O ro r-
d d d d
(\i in o r^
<M o -* fy
nj «- pu
>T -O 04 00
>0 ro o^ (M
fo -o o r-
— »- r- m >r
o in nj fO
■5 ^0 ^0 P«- O
^ 00 in (^
«-
Oil
1^
fO
d
d
d
d
00
vO
(M
00
-J
0
ro
d
d
r^
d
00 00 ru >o
O >f in ro
00 O O -T
O ^ in ^o
ru d O fo
00 CO o 00
>r d ~J- ry
^ PO PO in
O (\4 in ^.
fO O ^- in
d d r-^ d
FO in O m
>»• fO d «-
•»• «- t^ o^
PO 00 m
00 Q ^O >T ^
II ■« l> nO O
c a »- i> in
b
10 o> f^ 1^ py
3 00 d m' OS
b z
I < _ .,
e lij '^ <
< z X o
(M ry r\j ry (M
V UU UJ LU LU
& a. a. a. a.
w « Z X o
X UJ » <
Hi s x: s v>
I I I I I
PO PO PO ro PO
V uj uj oi uj
& a. a. Q. a.
< z X a
>»• N» >»• >!■
UJ UJ UJ UJ
. a. a. a. CL
APPENDIX V
Ministry of the Environment Interim Guidelines
100
INTERIM ENVIRONMENTAL GUIDELINES FOR SALMONID*
AQUACULTURE FACILITIES IN ONTARIO
1.0 BACKGROUND
The Ministry of the Environment has a legislated mandate for
the management of surface and groundwater quality and
quantity throughout the province. It has been established
that fish culture operations have the potential to impact on
all four of these areas. The Ministry's concerns/ therefore/
relate to both the talcing of water from surface or
groundwater sources and fair sharing among all users/ and the
discharge of wastewaters to receiving streams or back to the
groundwater aquifer in a condition ensuring acceptability for
the greatest variety of uses.
Competition among users for adequate water supplies for
drinking/ irrigation/ recreation/ aquaculture and so on/ can
place demands on some aquifers and streams resulting in
complaints of interference among users. For this reason/ the
Ontario Water Resources Act/ S"BCtion 20, requires a Permit to
Take Water for most significant water uses and protects
established uses from interference by more recent ones. A
guideline relating to the requirements for a hydrogeologic
study for the taking of groundwater for aquacultural
purposes/ is under preparation and will become a part of this
publication. Surface water taking will be assessed on a
site-by-site basis.
Wastewater discharges containing fish manure and uneaten
feeds can have unacceptable water quality impacts. Waste
solids tend to settle within or downstream from aquaculture
facilities/ smothering valuable habitat of fish and other
organisms. There they decompose to consume oxygen or release
nutrients which stimulate weed and algae growth/ and
generally interfere with other uses of the water. Therefore/
this Ministry has made efficient and effective solids
management the focus of these guidelines.
The water quality impacts of discharges are controlled by
approved treatment systems authorized and required by Section
24 of the Ontario Water Resources Act. The following
guidelines are intended to provide direction in the design of
land-based facilities. They will provide uniformity of
application to farms across the province/ however it is
recognized that extreme situations may necessitate individual
effluent quality compliance limits be established, based on
*Salmonid - Includes the rainbow and speckled trout, Atlantic
salmon, and various Pacific salmon species. Does
not include warm water species such as basses,
walleye/ bait fish/ etc.
- 2 -
local receiving water capability. Where more stringent
compliance limits may be necessary to protect a sensitive
receiver, it will be the responsibility of the Ministry to
define and substantiate the appropriate limits.
Guidelines for cage culture operations are under development
and will be included in this publication when available.
These guidelines pertain only to Ministry of the Environment
mandates. Federal, municipal and other provincial aaencies
may have additional approval
contacted directly.
2.0 REQUIREMSNTS
requirements and should be
2.3
to
2.1 A Permit to Take Water is reauired bv law prior
ln^nnn° ,°^ ^^^ facility for all takings in excess of
50,000 litres per day (Ontario Water Resources Act,
Section 20) . '
2.2 A Certificate of Approval (Sewage) is required by la
prior to the initiation' of the
treatment works (Ontario
Section 24) .
construction of any
Water Resources Act,
Treatment facilities will be designed, constructed and
operated with the intention to routinely achieve a
concentration in the final effluent, of 5 mg/1 suspended
soliGs above the background level measured in the source
water.
2.4 Treatment facilities will be designed, constructed and
operated with the intention to routinely achieve a
f?n!?"^f^'''°" °^ °*°^ ""^/^ °^ ^°^^^^ phosphorus in the
rmal effluent, except as follows:
a)
Where background levels in a surface water supply
exceed 0.05 mg/1, the total phosphorus Desian
Criterion will be
concentration.
no increase over the background
2.5 Compliance Limits are established
a)
b)
Suspended solids in the final effluent shall not
exceed background (as measured on a given day in
the source water) plus 10 mg/1.
Total phosphorus in the final effluent shall not
exceed 0.1 mg/1 except where the backaround levol
m a surface water supply exceeds 0.1 mc/1 in which
case total phosphorus in the final effluent ==hall
not exceed the background concentration as measured
on a given day in the source water.
- 3 -
Any exceedence of Compliance Limits shall be considered
a violation of the terms and conditions of the
Certificate of Approval (Sewage).
2.6 Solids removal facilities are required prior to
discharge to the watercourse. Solids will be regularly
removed to an off-line solids retention structure prior
to land application in accordance with the appended
"Criteria for Land Application of Aquaculture Cleanout
Wastes" .
2.7 A program of regular sampling and analysis at the
expense of the operator is required. Analytical results
shall be forwarded monthly to the District Officer along
with a complete explanation of the circumstances of any
exceedence of the Compliance Limits and of abatement
measures taken to prevent a recurrence.
2.8 A log is required/ which documents daily operations and
other occurrences which may affect effluent quality
(fish transfers or harvest/ cleaning/ major mortalities/
major weather events/ power outages/ etc.). The log
book shall be available for inspection by Ministry
personnel.
2.9 A copy of the log and a report of all sample results and
abatement action taken must be sent annually to the
District Officer/ Ministry of the Environment.
2.10 Where cleanout waste will be removed from the property
for land application elsewhere/ Certificates of Approval
for an Organic Waste Management System and Site are
required. Land applications will be as outlined in
"Criteria for Land Application of Aquaculture Cleanout
Wastes". (See Appendix).
3.0 RATIONALE
3.1 Solids Manaaement
Proper design and management of an aquaculture facility will
result in early settlement of solids in specific areas
throughout the system and their frequent removal. Suspended
and settled solids decompose quite rapidly in water to
soluble forms which are not readily removed by treatment.
consequently/ solids should be handled gently and be
completely removed from the rearing and settling facilities
to an off-line retention structure with no outlet/ on a
frequent basis. Vacuuming is one effective method of
cleaning which minimizes both the disturbance of settled
solids and the volume of cleanout wastewater requiring
storage in the retention structure.
The appropriate cleanout frequency will be as required to
satisfy the effluent limits/ but is expected to be
- 4 -
approximately twice weekly. To enable and facilitate
cleaning and to prevent contamination of groundwater, the
Ministry is suggesting that all rearing, settling and waste
retention facilities be lined with a firm, non-permeable
surface, such as concrete or metal. The use of unlined penes
is strongly discouraged due to difficulties in cleaning and
resultant poor effluent quality.
Cleanout waste utilization methods should be in agreement
with the general principles of the appended "Criteria for
Land Application of Aquaculture Cleanout Wastes".
Alternative disposal methods will be evaluated on a
case-by-case basis. Transport of the cleanout wastes for
utilization off the proprietor's property requires an
additional Certificate of Approval for an Organic Waste
Management System, application for which may be made through
the Ministry's District Office.
It is the opinion of the Ministry that by effective solids
management, phosphorus and other waste components will also
be controlled.
3.2 Physical Facilities
Water use at aquaculture facilities generally falls into one
of two categories: flow through (single pass) or
recirculating. A variety of rearing systems are in use in
Ontario, generally including some combination of rectangular
raceways, circular tanks or ponds. While the Ministry is not
prepared to dictate requirements for rearing structures,
their design should be developed with effluent requirements
in mind. As mentioned in Solids Management (3.1 above),
settlement of solids should be promoted in specific areas
(e.g. screened sumps in raceways, level controls on circular
tanks, pre-settling before recirculating, etc.) throughout
the system and these areas should be thoroughly cleaned on a
frequent basis.
All water from the rearing facilities should pass through a
solids "removal facility to achieve effluent limits prior to
discharge. The solids removal facility must be engineered to
ensure proper performance (good distribution of flow, minimal
turbulence, minimal resuspension of settled material, sized
and proportioned to suit effluent flow rate, ease of
cleanout, etc.). In this regard, the Ministry suggests the
system be designed by a consultant competent in aquaculture
desian.
The retention structure shall be of sufficient size to
contain all cleanout waste from all settlement areas
including the final settling facility throughout the period
of frost (approximately December 1 to March 31) during which
time no land application is allowed. The latter requirement
is due to the fact that, if applied to land in winter, the
wastes would remain on the surface of the frozen ground and
- 5 -
could, run off to the nearest stream at times of snowmelt or
rain. During the frost-free period of the year
(approximately April 1 to November 30) the cleanout wastes
may be applied to croplands at appropriate times.
3.3 Monitoring Program
A program to monitor compliance of the operation of the
aquaculture facility with conditions of the Certificate of
Approval (Sewage) will in general require samples from the
source water and the final effluent.
In most cases where no background contamination of the source
water exists/ its monitoring can be reduced or discontinued
once stable quality is documented. Where effluent limits are
set relative to background levels/ monitoring of the source
water would be needed as a basis to establish the effluent
limit. The Ministry's ultimate interest is in environmental
protection.
Most facilities with year-round production and harvest should
be sampled monthly. Operations with distinct peaks of
production should be sampled weekly during the periods of
peak production.
Results of the first year of monitoring for all facilities
will be reviewed and the methods/ frequency and location of
sampling and testing/ as well as reporting requirements may
be subject to alteration as required by the District
Officer.
The Ministry may inspect facilities from time to time to
verify adherence to all conditions of the Certificate of
Approval.
4.0 APPLICATION AND AMENDMENT OF GUIDELINES
These guidelines have been developed by the Ministry of the
Environment Aquaculture Committee on the basis of direct
Ministry of the Environment experience and knowledge of the
available literature on wastewater treatment in aquaculture
and in consultation with representatives of the a.quaculture
industry in Ontario. As research continues and additional
experience is gained/ revisions to these guidelines may be
made from time to time. Comments on/ or formal proposals to
amend/ these guidelines should be addressed directly to:
Director
Water Resources Branch
Attention: MOE Aquaculture Committee
Ministry of the Environment
1 St. Clair Ave. W.
Toronto/ Ontario
M4V 1K6
25 April/ 1989
Appendix I: Criteria* for Land Application of Aquaculture
Clean-Out Wastes
1.0 Under the Environmental Protection Act and the Ontario
Water Resources Act ail spreading sites must be
certified by the Ministry of the Environment. Prior to
site certification and use, factors such as site
location, land a.nd soil characteristics and proposed
site management methods must be assessed to minimize
potential hazards to surface watercourses, groundwater,
wells and residences.
2.0 Abdication rates not more than 1.3 cm depth or 130
cubic metres per hectare of clean-out wastes may be
applied at one time. There may be no subsequent
application until the preceding application has dried.
3 .0 Separation Distances
3.1 Surface Watercourses
For the ourpose of these Guidelines, a surface
watercourse is defined as a natural or established
surface watercourse or an open municipal drain along
which water flows on a regular basis.
The minimum distance between the spreading site and a
surface watercourse should normally be determined from
Table 1 which takes into account land slope and soil
permeability. If clean-out wastes are incorporated into
the soil at the time of application, it may be applied
closer to a watercourse than indicated in the Table.
However, it should not be applied within 10 metres of
any watercourses or body of water .
tlinistry of the Environment staff will advise on ^
separation distances from bodies of water or drainage
channels other than surface watercourses as defined
above .
3.2 Groundwater
The ground water table should not be less than 0.9
meters below the soil surface at the time of clean-out
waste application.
The criteria listed in this appendix have been derived
from the publication "Guidelines for Sewage Sludge
Utilization on Agricultural Lands", Ontario Ministries of
Aariculture and Food, Environment, and Health, January,
1986.
- 2 -
3.3 Bedrock
Clean-out waste application should not normally be
allowed where the soil overlying bedrock is less than
1.5 metres thick. Under special circumstances, and
based on site-specific information which demonstrates
that the risk of ground and surface water contamination
Is minimal, sites with a lesser thickness of soil may be
used .
3 . 4 Residence
According to Ontario Regulation 309, the minimum
separation distances from residences in residential
areas and individual residences not in residential areas
shall be 1500 feet (450 metres) and 300 feet (90 metres)
respectively. If the Regulation should, in the future,
permit it and where there is little cause for concern,
these distances may be reduced. In no case should they
be less than 50 metres and 25 metres respectively.
3.5 Water Wells
According to Ontario Regulation 309, the minimum
separation distance from any water well shall be 300
feet (90 metres). If the Regulation should, in the
future, permit it and where there is little cause for
concern, this distance may be reduced. In no case
should it be less than 15 metres.
4.0 Snow Covered and Frozen Ground
To minimize runoff, clean-out wastes should not be
spread on frozen or ice covered soil. The period of
frost is approximately December 1 to March 31.
5.0 Record Keeping
Records are to be kept of the location of all fields
receiving clean-out wastes, the volume of clean-out
wastes applied to each field and the date of
application .
RIVER/ SYS . 6/0742RS
- 3
Table 1
Minimum Pi stances to '.-'at er courses
1
tlaximum
Sustained
Soil *
Distance
Slope
Perneab: 1 i tv
f "'et r es )
Rapid to Moderately
0 to 3%
• Rapid
60
!loderate to Slow
120
Rapid to Moderately
] 20
Rapid
3 to 6%
Moderate to Slov;
240
Rapid to Moderately
180
Rapid
6 to 9%
Moderate to Siov;
Not Permi t ted
Greater than
All permeabilities
Not Permitted
9%
* Soil permeability shall be determined in accordance
with the requirements of the ministry of Agriculture
and Food's publication entitled "Drainage Guide
for Ontario". The type of soil should be determined
with the aid of County Soil tiaps , which are available
from that Ministry.
NOTES: 1. Spreading should be suspended when run-off
i s
expected .