Abboah-Afari Ebenezer 201108 Ms

ALTERNATIVE PHYSICAL TREATMENT METHOD FOR POULTRY
PROCESSING WASTEWATER USING MEMBRANE FILTRATION
by
EBENEZER ABBOAH-AFARI
(Under the Direction of Brian H. Kiepper)
ABSTRACT
Regulatory environmental permits governing wastewater disposal have become
stricter for poultry processing plants that discharge effluent directly into the environment
or pre-treat their wastewater stream prior to discharge to a municipal sewer systems.
Therefore, more effective means of on-site wastewater treatment is a constant demand.
The viability of membrane filtration as an on-site treatment method for pre- and post-
DAF (dissolved air flotation) poultry processing wastewater (PPW) was investigated. A
performance index (Pmi) was used to indicate which membrane was the most effective
and suitable for the membrane filtration of pre-and post-DAF PPW. For the membranes
tested, the 0.3µm PVDF membrane was determined to be the most effective in treating
pre-DAF PPW by producing a 115 Lm-2h-1 permeate flux, while the 100,000 MWCO
Ultrafilic membrane was determined to be the most effective in treating post-DAF PPW
by producing a 224 Lm-2h-1 permeate flux. COD, TS and TSS reductions are evaluated
and reported.
INDEX WORDS: Poultry processing, wastewater treatment, membrane filtration,
permeate flux, COD, TS, TSS

by
B.S., Kwame Nkrumah University of Science and Technology, Ghana, 2007
A Thesis Submitted to the Graduate Faculty of The University of Georgia in Partial
Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
ATHENS, GEORGIA
2011
© 2011
Ebenezer Abboah-Afari
All Rights Reserved

by
Major Professor: Brian H. Kiepper

Committee: Jenna Jambeck
Ke Li
Ernest W. Tollner
Electronic Version Approved:
Maureen Grasso
Dean of the Graduate School
The University of Georgia
August 2011
DEDICATION
To my family and friends, thank you for the love and support.
iv
ACKNOWLEDGEMENTS
I would like to thank everyone whose through their endeavors has brought this
project to light. First, my sincere gratitude goes to my major professor, Dr. Brian Kiepper
for his undying help, leadership and support for these past years. I would also like to
thank rest of my committee members Dr. Jenna Jambeck, Dr. Ke Li and Dr. E.W. Tollner
for their advice and support. I appreciate all of the help, advice and support given to me
by Herbert Ssegane, Husain Plumber, Dan Geller and the faculty, staff and graduate
students in Bio. & Ag. Engineering. A special thank you to Benedicta Antwiwaa-Wiafe,
Elsie Kitcher, Sasu-Boakye and my Athens family. Finally, to Him who has remained
faithful, Glory be to His name.
v
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS .................................................................................................v
LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES .............................................................................................................x
CHAPTER
1 INTRODUCTION .............................................................................................1
2 LITERATURE REVIEW ..................................................................................4
Poultry Processing Wastewater (PPW) ........................................................5
Characteristics of Poultry Processing Wastewater ......................................6
Poultry Processing Wastewater Treatment ................................................12
Thesis Goal and Objectives .......................................................................29
3 EFFECTS OF MEMBRANE FILTRATION ON PRE-DAF (DISSOLVED
AIR FLOTATION) POULTRY PROCESSING WASTEWATER ................31
Abstract ......................................................................................................32
Introduction ................................................................................................33
Materials and Methods ...............................................................................35
Results and Discussion ..............................................................................39
Conclusions ................................................................................................52
References ..................................................................................................55
vi
4 EFFECTS OF MEMBRANE FILTRATION ON POST-DAF (DISSOLVED
AIR FLOTATION) POULTRY PROCESSING WASTEWATER ................57
Abstract ......................................................................................................58
Introduction ................................................................................................59
Materials and Methods ...............................................................................62
Results and Discussion ..............................................................................67
Conclusions ................................................................................................77
References ..................................................................................................80
5 CONCLUSIONS AND FUTURE DIRECTIONS...........................................83
Conclusions ................................................................................................83
Future Directions .......................................................................................87
REFERENCES ..................................................................................................................88
vii
LIST OF TABLES
Page
Table 2.1: Summary of the various constituents of organics, inorganics, particulates and
nutrients in poultry processing wastewater (PPW) ................................................11
Table 2.2: Wastewater Screen Classifications by Open Space Sizes ................................14
Table 3.1: Pair-wise comparison of the criteria (i.e., COD, Permeate flux, TS) for
membrane filtration of pre-DAF poultry processing wastewater ..........................39
Table 3.2: Mean permeate flux (Lm-2h-1±SEM), COD (mg/L±SEM) and TS (mg/L±SEM)
values for 6 membranes filtering pre-dissolved air flotation (DAF) poultry
processing wastewater for 120 min........................................................................40
Table 3.3: Mean permeate flux (Lm-2h-1±SEM), COD (mg/L±SEM) and TS (mg/L±SEM)
values at 2 pressure levels (50 and 80psi) for pre-dissolved air flotation (DAF)
poultry processing wastewater for 120 min ...........................................................40
Table 3.4: Best performing membrane for treatment of pre-DAF poultry processing
wastewater based on maximum permeate flux, and reduction of COD and TS over
120 min of operation based on significant difference (P<0.05)............................53
Table 4.1: Pair-wise comparison of the criteria (i.e., COD, Permeate flux, TS) for
membrane filtration of post-DAF poultry processing wastewater .........................66
Table 4.2: Mean permeate flux (Lm-2h-1), COD (mg/L) TS (mg/L) and TSS (mg/L)
values for 6 membranes filtering post-dissolved air flotation (DAF) poultry
processing wastewater for 60 min..........................................................................68
viii
Table 4.3: Best performing membrane for treatment of post-DAF poultry processing
60 min of operation based on significant difference (P<0.05)..............................78
Table 5.1: Best performing membrane for treatment of pre-DAF poultry processing
Table 5.2: Best performing membrane for treatment of post-DAF poultry processing
ix
LIST OF FIGURES
Page
Figure 3.1: SpinTek Static Test Cell (STC) Membrane Filtration System ........................36
Figure 3.2: Pre-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at 10
minute intervals for 3 membrane filters at 50 psi operating pressure ....................41
Figure 3.3: Pre-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at 10
minute intervals for 3 membrane filters at 80 psi operating pressure ....................42
Figure 3.4: Pre-DAF poultry processing wastewater mean permeate flux values (Lm-2h-1)
for 3 membranes. ...................................................................................................43
Figure 3.5: Pre-DAF poultry processing wastewater mean permeate flux values (Lm-2h-1)
at 50 and 80 psi ......................................................................................................44
Figure 3.6: Pre-DAF poultry processing wastewater mean permeate COD concentrations
(mg/L) for 3 membranes ........................................................................................46
Figure 3.7: Pre-DAF poultry processing wastewater mean permeate COD concentrations
(mg/L) at 50 and 80 psi ..........................................................................................46
Figure 3.8: Pre-DAF poultry processing wastewater mean permeate TS concentrations
(mg/L) for 3 membranes ........................................................................................48
(mg/L) at 50 and 80 psi ..........................................................................................50
(mg/L) for 100,000MWCO Ultrafilic membrane at 50 and 80 psi ........................50
x
(mg/L) for 0.10µm Polysulfone membrane at 50 and 80 psi .................................51
(mg/L) for a 0.30µm PVDF membrane at 50 and 80 psi .......................................51
Figure 4.1: SpinTek Static Test Cell (STC) Membrane Filtration System ........................63
Figure 4.2: Post-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at 10
min intervals for 6 membrane filters over 60 min .................................................69
Figure 4.3: Mean permeate flux (Lm-2h-1) values for 6 membranes filtering post-DAF
poultry processing wastewater over 60 min...........................................................70
Figure 4.4: Post-DAF poultry processing wastewater permeate COD concentrations
(mg/L) values at 10 minute intervals for 6 membrane filters over 60 min ............71
Figure 4.5: Mean permeate COD concentrations (mg/L) for 6 membranes filtering post-
dissolved air flotation (DAF) poultry processing wastewater for 60 min .............73
Figure 4.6: Mean permeate TS (mg/L) for 6 membranes filtering post-DAF poultry
processing wastewater for 60 min ........................................................................74
Figure 4.7: Mean permeate TSS (mg/L) for 6 membranes filtering post-DAF poultry
processing wastewater for 60 min..........................................................................76
xi
CHAPTER 1
INTRODUCTION
U.S. poultry processing plants are constantly faced with challenges resulting from
the generation of high-strength wastewater and slaughter byproduct recovery and
handling (APHA, 2005; Kiepper et al., 2008). Due to constant environmental pressures
the poultry industry is relentlessly searching for better wastewater treatment practices.
Since the inception of the Clean Water Act in 1972, regulations governing wastewater
disposal have increased and become stricter on industrial and domestic wastewater
treatment plants that discharge effluent directly into the environment (i.e., direct
dischargers) (Del Nery et al., 2007; Kiepper et al., 2001; USEPA, 2002). Alternatively,
poultry processing plants that discharge their effluent to local sewage systems (i.e.,
indirect dischargers) often must pay substantial surcharges fees to local municipalities to
cover costs associated with treatment of high-strength wastewater discharges. Therefore
the need for alternative, more effective means of wastewater treatment is a constant
challenge for U.S. poultry processors (Avula et al., 2009).
Currently, more than 80% of poultry processors use a combination of mechanical
screens followed by dissolved air flotation (DAF) to treat their wastewater on-site
(Kiepper, 2003; Lo et al., 2005). Although DAF is effective, the aggressive aeration of
the fine poultry processing wastewater particulates causes excessive oxidative damage
and bacterial degradation of fat and protein components of the recovered byproduct (Del
Nery et al., 2007; Park et al., 2001; Viitasaari et al., 1995). Also, various chemicals (e.g.,
1
metal salts, cationic and anionic polymers) are used in combination with the physical
DAF process which improves treatment, but further adulterates and greatly reduces the
value of the recovered DAF skimmings as a nutrient source in rendered animal byproduct
feed ingredients. Thus, utilization and disposal of DAF skimmings is a major expense
and inconvenience to the poultry industry (Kiepper, 2003).
There is great potential in exploring advanced physical separation systems
(APSS) (e.g., microscreening and membrane filtration) on both pre- and post-DAF
poultry processing wastewater (PPW) (Avula et al., 2009). The benefits of improved pre-
DAF PPW treatment include a substantial increase in the volume of unadulterated
byproduct (i.e., offal) recovered for rendering while reducing a corresponding volume of
lower value DAF skimmings, substantial chemical cost savings, and a potential reduction
in energy consumption. The benefits of improved post-DAF PPW treatment include
reduced cost of operation of biological treatment systems and sludge volume production
in direct discharge systems, and the potential to significantly reduce surcharge fees paid
by indirect dischargers.
To examine the potential for use of APSS in pre- and post-DAF PPW treatment
on-site at poultry processing plants, two membrane filtration experiments were
completed. The first experiment (i.e., Chapter 3) involved the use of membrane filtration
in the treatment of pre-DAF PPW as a potential enhancement or replacement for existing
DAF technology. Flat sheet membranes within the microfiltration (i.e., 0.1-10µm) and
ultrafiltration (i.e., 0.01-0.1µm) ranges that use a semi-permeable surface under pressure
to separate colloids and high molecular weight materials in solution were tested on pre-
DAF PPW. The 3 membranes sizes/materials tested were 0.30µm Polyvinylidene
2
fluoride (PVDF), 0.10µm Polysulfone (PS), and 100,000MWCO (molecular weight cut-
off) Ultrafilic at two operating pressures (i.e., 50 psi and 80 psi) on pre-DAF PPW pre-
sieved (i.e., microscreened) to 106µm. Effluent permeate flux (Lm-2h-1), and
concentration (mg/L) of chemical oxygen demand (COD) and total solids (TS) were
measured and analyzed.
The second experiment (i.e., Chapter 4) involved the use of membrane filtration
in treating post-DAF PPW. Flat sheet membranes within the microfiltration and
ultrafiltration ranges were tested on post-DAF PPW to measure the potential for
surcharge fee reduction. A total of six membranes (i.e., 0.3µm PVDF, 0.1µm
Polysulfone, 100,000MWCO PVDF, 100,000MWCO Ultrafilic, 30,000MWCO PVDF
and 30,000MWCO Polysulfone) were tested at 50 psi on post-DAF PPW pre-sieved to
106µm. Effluent permeate flux (Lm-2h-1), and concentration (mg/L) of COD, TS and
total suspended solid (TSS) were measured and analyzed.
The results of both experiments were enumerated and comparisons were made
with the results found (i.e., Chapter 5). General limitations of membrane materials,
membrane pore sizes and transmembrane pressures are identified.
3
CHAP TER 2
LITERATURE REVIEW
The term poultry refers to a group of avian species that are raised for the
production of meat and eggs. Poultry includes chicken, turkey, geese, quail, pigeons,
pheasants, and ostrich (Perry et al., 1999). However for purpose of this thesis, the term
poultry is used synonymously with young chickens (i.e., broilers) as defined by the
United States Department of Agriculture (USDA), which account for over 95% of the 9
billion poultry processed in the U.S. on an annual basis (USDA, 2010).
Poultry processing plants utilize relatively large volumes of potable water to clean
both products and processing equipment. In the U.S., commercial broiler slaughter plants
use 19 to 38 L (5 to 10 gal) of potable water per bird processed (Kiepper, 2003; Northcutt
and Jones, 2004). The live weight of poultry plays a critical role in determining the
amount portable water utilized in processing. For example large male turkeys can weigh
18 kg (40 lbs) and require potable water volumes in the range of 130 to 150 L (35 to 40
gal) per bird for processing (Avula et al., 2009).
One common problem facing the U.S. poultry industry is that increasing levels of
the water use, due to increased production levels, that results in a corresponding increase
in high-strength wastewater generation (APHA, 2005). Most of the water used in poultry
processing is for scalding, defeathering, evisceration, washing of equipment and
carcasses, and overall plant sanitation. Also, water is used by poultry processors as a
transport medium for byproducts generated during processing (e.g., offal including
4
feathers, heads, and viscera) (Northcutt and Jones, 2004; Thornton and O’Keefe, 2002;
Veerkamp, 1999).
Since the inception of the Clean Water Act in 1972, environmental regulatory
permits governing wastewater disposal have consistently become stricter on industrial
wastewater treatment plants that discharge effluent directly into the environment or to
local municipal sewage systems. Poultry processing plants are classified as either
‘indirect’ or ‘direct’ wastewater effluent dischargers. Indirect dischargers pre-treat their
wastewater stream on-site prior to discharge to local municipal sewerage collection
systems commonly known as publicly owned treatment works (POTWs). Indirect
dischargers are subject to both regulatory permit limits as well as surcharge fees set by
local environmental authorities to recover added costs associated with the treatment of
high-strength wastewater. Direct dischargers must fully treat their wastewater stream
prior to releasing the treated effluent straight into the environment under federal or state
environmental regulatory permit.
Poultry Processing Wastewater (PPW)
Poultry processing wastewater (PPW) consists of spent potable water and the
proteins, fats and carbohydrates generated from the meat, blood, skin and feathers
removed from the bird during processing. PPW also contains grit and other inorganic
particulates (Fonkwe et al., 2001). The term PPW is typically used to define the combine
wastewater stream generated during poultry slaughter after all processing operations and
following offal screening which removes large solids from the PPW stream (e.g, feathers,
heads and viscera). Prior to discharge of PPW, poultry processors are required to remove
the majority of the soluble and particulate organic and inorganic material in order to
5
comply with environmental regulations. With an estimated 2 to 5% of total carcass
proteins lost to the wastewater stream during processing, PPW contains predominantly
(i.e., ~35%) protein, resulting in a high-strength effluent with a substantially higher
biochemical oxygen demand (BOD) and chemical oxygen demand (COD) concentration
(mg/L) than domestic wastewater (Zhang, 1997). Thus PPW requires intensive treatment
before it can be discharged (Wesley, 1985).
Characteristics of Poultry Processing Wastewater
PPW consists of various constituents in the form of organics, particulates and
inorganic nutrients. BOD and COD are the two most common analytical tests used to
assess the organic ‘strength’ of PPW. The series of analytical tests used to establish the
concentration of particulates in PPW include total solids (TS), total suspended solids
(TSS), total volatile solids (TVS), total dissolved solids (TDS) and total fixed solids
(TFS). Inorganic nutrients of concern such as nitrogen (N) and phosphorus (P) can be
found in substantial quantities in PPW. Also, pathogens (e.g., bacteria) and colloidal
particles (e.g., fats, oil and grease) are found in PPW (APHA, 2005). The following is a
glossary of common PPW terms:
BOD (biochemical oxygen demand) – measures the amount of oxygen used by
aerobic microorganisms to stabilize the biodegradable organic matter in a wastewater
sample (Metcalf and Eddy Inc., 2003). The BOD test is also known as BOD5 since the
test is conducted over a period of 5 days (USEPA, 2002). BOD involves calculating the
change in the concentration of dissolved oxygen (DO) in a wastewater sample incubated
at 20oC over a 5-day period (APHA, 2005). BOD is the traditional test used to track the
6
organic concentration of pollutants though the wastewater treatment process prior to
discharge into a receiving water body (USEPA, 1974).
COD (chemical oxygen demand) – measures the amount of oxygen required for
oxidation of the organic matter in wastewater using a standard reagent (APHA, 2005). A
measured sample with a known excess of potassium dichromate (K2Cr2O7) is mixed after
the addition of sulfuric acid (H2SO4) and silver sulfate (Ag2SO4). The organic matter is
oxidized over a 2 hour heating period at 150oC. Once samples are cooled, all samples,
blanks and standards are measured against a reference solution (reagent blank) at a
620nm wavelength. COD is a useful control parameter for oxidation operations, and
given a consistent waste stream, a stable empirical ratio can be established between COD
and BOD. Thus COD can be an effective predictor of BOD. COD is used as a more
accurate, cheaper and less time consuming means of determining the effectiveness of
wastewater treatment (Nielsen, 1989).
Colloidal particles – are solids in a liquid that do not settle due to gravity (i.e., low
specific gravity) and as a result causes them to float or suspend in wastewater (APHA,
2005). These immiscible particles blinds screens, clogs pipes and pumping systems. They
can cause decrease in the efficiency of aerobic reactors by reducing oxygen transfer. Oils
and greases (O&G) are examples of common PPW colloidal particles.
DO (dissolved oxygen) – is the concentration (mg/L) of oxygen in solution in
water (APHA, 2005).
Nutrients - are elements required to support living organisms. Poultry processing
plants can contribute significant loadings of nutrients of concern (e.g., nitrogen and
phosphorus) into PPW. PPW typically contains organic nitrogen, phosphorus and
7
dissolved ammonia gas (Contreras et al., 2000; CSUS, 1993; Metcalf and Eddy Inc.,
2003; Welch and Lindell, 1992). Nutrients are important in aquatic biosystems, but
excess amounts accelerate eutrophication and is detrimental to water quality. The
presence of ammonia in PPW results from the breakdown of proteins and other
nitrogenous compounds which can be toxic to aquatic life. About 2 to 5% of the total
carcass protein can be found in PPW (Grant, 1980). Mead (1989) reported the bulk of
solids in PPW to be nitrogen, which can be recovered and recycled for feed for farm
animals and pets. Nitrate (NO3) is another pollutant of concern in water and is formed as
a result of oxidation of ammonia. In high levels in natural water bodies, nitrate brings
about excess algal growth and accelerated eutrophication (Eremektar et al., 1999; Welch
and Lindell, 1992). Along with the nutrient present in poultry tissues, other factors that
increase the pollutant load of nutrients in PPW can be attributed to cleaning agents and
disinfectants used in cleaning the equipment and birds, grit or soil particles from the birds
and transportation equipment, undigested feeds from the carcasses, and ingredients used
in further processing operations (Kiepper, 2009; Mead, 1989; Merka, 2001). Excess
nitrate in water is toxic to both microorganisms and public health. Excess phosphorus in
water also adversely affects quality (Eremektar et al., 1999).
TFS (total fixed solids) – measures the concentration (mg/L) of inorganic solids in
a wastewater sample. TFS are the remaining solids in the form of ash following the
combustion of organic solids in a TVS test (APHA, 2005). It can be calculated by
subtracting the TVS from TS.
TDS (total dissolved solids) – is the concentration (mg/L) of soluble inorganic
and organic matter that is in solution in a wastewater sample (APHA, 2005). TDS is
8
determined by measuring the solids remaining in the filtrate that passes through the glass
fiber filter in the TSS test (USEPA, 1974).
TKN (total Kjeldahl nitrogen) – measures the concentration (mg/L) of organic
nitrogen and ammonia combined in a wastewater sample through acid digestion and
distillation (Merka, 1989).
TS (total solids) – measures the concentration (mg/L) of the total amount of
inorganic and organic matter recovered after a wastewater sample has been evaporated to
dryness (APHA, 2005). A measured volume of wastewater is poured in a pre-weighed
crucible and dried to dryness at 105oC to a constant weight over a period of 24 hours
(APHA, 2005).
TSS (total suspended solids) – is the concentration (mg/L) of insoluble inorganic
and organic matter that is suspended in a wastewater sample (APHA, 2005; Nielsen,
1989). TSS is determined by passing a measured volume of wastewater through a
standard glass fiber filter (i.e., < 2.0µm) and drying the filter to a constant weight at 103
to105oC (USEPA, 1974).
TVS (total volatile solids) – is the concentration (mg/L) of organic solids in a
wastewater sample. A measured volume of wastewater is evaporated and the residue
combusted at 550oC (APHA, 2005). The loss in weight is equal to the organic solids
which volatilized during combustion. It can be calculated by subtracting the total fixed
solids (TFS) from the total solids (TS).
To be able to meet environmental regulatory permits, most of the organics,
particulates and nutrients must be removed from PPW before discharge (Kiepper et al.,
2001; Mead, 1989; Merka, 2001). From the early 1960s to late 1970s, researchers
9
reported average BOD readings from US broiler processing plants ranging from 400 to
1300 mg/L (Camp and Willoughby, 1968; Carawan et al., 1974; Chen et al., 1976.; Glide,
1968; Lillard, 1978; Nemerow, 1969; Singh et al., 1973; Teletzke, 1961; USEPA, 1975;
Whitehead, 1979; Woodard et al., 1972; Woodard et al., 1977). Hamza et al. (1978)
reported an Egyptian poultry processing plant generating wastewater with a BOD
concentration of 2341 mg/L. As poultry processing plant line speeds have increased,
there have been corresponding increases in BOD and COD levels of PPW. Merka (1989)
reported an average BOD and COD concentrations of PPW to be 2178 mg/L and 3772
mg/L, respectively from a broiler processing plant. Avula et al., (2009), Del Nery et al.(
2007), Eremektar et al. (1999), Kiepper (2009), Lo et al. (2005), Merka (2001), Rusten et
al. (1998) and Zhang (1997), all reported increasing BOD and COD concentrations in
PPW in recent decades.
TS and TSS in PPW show similar increasing trends over the years. Camp and
Willoughby (1968) reported average TS and TSS concentrations of 650 mg/L and 196
mg/L, respectively in PPW. USEPA (1975) reported average TS and TSS concentration
ranges of 600 to 1000 mg/L and 200 to 700 mg/L, respectively in a broiler processing
plant survey. Merka (1989) reported average TSS and TVS concentrations of 1446 mg/L
and 1745 mg/L. Rusten et al. (1998) reported an average TSS concentration of 1360
mg/L for PPW. Kiepper et al. (2009) reported TS, TSS and TVS concentration ranges of
2000 to 2700 mg/L, 880 to 1250 mg/L, and 1600 to 2200 mg/L, respectively, in PPW.
Yordanov (2010) reported an average TSS concentration range of 2280 to 2446 mg/L in
PPW.
10
The concentration of O&G collidial partciles in PPW was reported by Rusten et
al. in 1998 at an average O&G concentration of 970 mg/L. Yordanov (2010) reported an
average O&G concentration range of 289 to 389 mg/L in PPW.
Nutrient concentration levels in PPW have documented by Eremektar et al. (1999)
who reported average phosphorus concentrations of 48, 16 18 and 40 mg/L for four
different samples of PPW. Rusten et al. (1998) reported an average phosphorus
concentration range from 14.1 to 18.5 mg/L. Kiepper et al. (2009) reported TKN
concentration range of 120 to 250 mg/L in PPW.
The increasing trend these constituents is a result of increase in production
poultry over the years, thus a corresponding increase with load on wastewater.
Table 2.1. Summary of the various constituents of organics, inorganics, particulates and
nutrients in poultry processing wastewater (PPW)
Discharge Range of Concentration Average Concentration

Characteristics (mg/L) (mg/L)
BOD 200-2341 1271
COD 1300-3772 2536
TS 600-2700 1650
TSS 200-2446 1323
TVS 1600-2200 1900
TKN 120-250 185
O&G 100-970 535
Phosphorus 14-50 32
The composition of wastewater from the poultry varies from one plant to another
depending on the type of systems, the methods of operations, treatments and the
processing loads (Zhang, 1997).
11
Poultry Processing Wastewater Treatment
Most of the water used in poultry processing is utilized for scalding, evisceration,
washing of equipment and birds, and overall plant sanitation (Northcutt and Jones, 2004;
Thornton and O’Keefe, 2002). These wastewaters cannot be simply discharged into
water bodies like rivers and lakes because of the high content of organic and inorganic
matters, and microorganisms they contain (Mead, 1989). Various treatment procedures
can be used, ranging from simple mechanical screening to advanced biological treatment
systems, to treat PPW. PPW primary treatment involves the separation of offal (i.e.,
inedible poultry byproducts) from the wastewater stream through physical separation
(e.g., mechanical screens) driven by force or gravity (Pankratz, 1995).
Secondary treatment consists of advanced physical, chemical and biological
processes that remove finer particulates of the poultry processing byproducts contained in
PPW. Although effective, secondary treatment changes and adulterates the byproducts
recovered from the wastewater stream, thus reducing their value and uses (Pierson and
Pavlostathis, 2000).
Physical Treatment of PPW
Primary treatment of PPW traditionally involves the use of mechanical screens to
physically remove coarse solids (i.e., offal) from the wastewater stream. Depending on
the type of processing wastewater and desired removal efficiencies, other methods of
physical treatment are available and have been documented by researchers. Examples
include using air stripping to remove volatile compounds, sedimentation and sand
filtration (Torrens, 2001). Barrett (1977) and Chen et al. (2002) documented electrolysis
and electroflocculation to be effective in removing fine suspended solid particles in
12
piggery wastewater. Electro-coagulation has also proven to be an effective method for
treatment of animal processing wastewater. Kobya et al. (2006) found electro-
coagulation to be an effective method for the treatment of PPW. Kobya et al. found that
by means of combined usage of iron and aluminum as anode materials at appropriate
treatment conditions, high levels of COD and O&G could be removed from PPW.
Screening
In most poultry processing plants, physical screening is the typical primary
treatment method utilized. Mechanical screen are the most energy efficient and least
expensive form of primary treatment in reducing solids in PPW. Currently, primary
treatment of PPW at U.S. poultry processing plants consists of physical screening in the
Fine (1500 to 6000µm) and Very Fine (200 to 1499µm) ranges as defined by the Water
Environment Federation (Kiepper, 2003). Screening recovers offal (i.e., feathers,
viscera, meat particles) removing the large solid particulates that might hinder operation
of downstream treatment processes (Arundel, 1995; Pankratz, 1995).
Screening is defined as the physical removal of particulate matter from a waste
stream by the insertion of a perforated surface that retains particles larger that the surface
openings and allows the flow through of smaller particles and water (Arundel, 1995).
Screen opening size and flow rate capacity are the most important criteria used to select a
screen. However many other factors must be considered including plant hydraulics,
operational costs, size constraints, debris handing systems, operator qualifications and
engineering design (Pankratz, 1995). Common problems associated with screens include
mechanical failures and blinding due either to the overloading of the screen or to under
sizing of screen gaps. Blinding is defined as the overloading of a screen resulting in the
13
coating over of the open spaces by solids thus preventing the pass-through of water
(AWWA, 1977). Screens are classified by the size of the open spaces that allow the
passage of water. Screens are classified as coarse, fine, very fine and micro. Table 2.2
shows the range of space openings for each screen size (WEF, 1998).
Table 2.2. Wastewater screen classifications by open space sizes
Type Inches (in) Millimeters (mm) Microns (µm)

Coarse > 0.25 > 6.0 > 6000
Fine 0.059 - 0.25 1.5 - 6.0 1500 - 6000
Very Fine 0.008 - 0.058 0.2 - 1.49 200 - 1499
Micro 3.9 x 10-8 - 1.2 x 10-2 0.001 - 0.19 1 - 199
In most poultry slaughter applications screens are used both in parallel and in
series. Two larger gap size opening ‘primary’ screens are typically used in parallel to
remove feathers from one flume, while another screen handles solids removal from a
separate viscera flume. These primary screens typically range in gap opening size from
1500 to 3500µm (0.059 to 0.138 in) and are designed to remove large offal pieces.
Effluent flow from the primary screens is usually combined and then flows through a
‘secondary’ screen. Commercial secondary rotary screens are commonly available with
minimum gap size openings of 500µm (0.020 in) and 250µm (0.010 in). Traditionally,
any particulate matter remaining in the wastewater then flows to advanced chemical
and/or biological treatment operations for removal prior to discharge.
Membrane filtration
Membrane filtration involves a pressure driven process that uses a semi-
permeable membrane to separate particulate matter from wastewater (Strathmann, 1979).
14
The membrane serves as a selective barrier that allows the passage of water and fine solid
particulates (i.e., permeate) while retaining and concentrating the larger solids (i.e.,
retentate) (de Morais Coutinho et al., 2008). The nature (i.e., texture, structure and
material) of the membrane controls which components will permeate and which will be
retained, since they are selectively separated according to their molar masses or particle
size (Cheryan, 1998).
Ultrafiltration (UF) and microfiltration (MF) are membrane-based, typically
pressure driven membrane filtration processes widely used to simultaneously purify,
separate and concentrate colloids and high molecular weight materials in water (Cheryan,
1998; Lo et al., 1996). UF membranes have a pore size in the range of 0.001 to 0.1µm,
whereas MF membranes have a pore size range of 0.1 to 1.0µm.
UF and MF membranes have found applications in removing particulates, bacteria
and pyrogens in water, as well as in recovering valuable ingredients and byproducts in
the food, chemical, and pharmaceutical industries (Cheryan and Rajagopalan, 1998).
However, limited efforts have been made to recover the byproducts and nutrients from
PPW beyond traditional offal screening (El Boushy and van der Poel, 1994; Grant, 1976;
Shih and Kozink, 1980). The idea of using UF and MF to recover valuable byproducts
from PPW is not new, however since treatment of PPW is usually driven by
environmental regulations and not economics, only limited studies on its potential have
been conducted (Avula et al., 2009). Shih and Kozink (1980) reported that the
combination of UF and dehydration could produce a byproduct that contains 30 to 35%
protein and 24 to 45% fat from PPW. Unfortunately, due to the presence of both fat and
protein in the feed stream, and the single-stage use of a standalone 30,000MWCO UF
15
membrane unit, the effectiveness of the process was greatly hindered due to severe
fouling of the membrane, leaving considerable doubt to its applied full-scale application.
The importance of pretreatment of PPW to remove larger particulates, especially
fat, on the performance of membrane-based processes was confirmed by Zhang et al.
(1997) in their attempt to investigate the feasibility of recycling chiller water using
membranes made from different polymeric materials. The work helped establish the
feasibility of recovering protein from PPW using UF by establishing the optimization of
processing parameters crucial to overall UF performance. From a processing standpoint,
if the optimal filtration conditions are identified, it is highly applicable that a multi-stage
MF and UF system could be operated at its highest flux for an extended period of time,
minimizing the adverse effects caused by fouled membranes (le Roux and Belyea, 1999;
Yushina and Hasegawa, 1994). The importance is that: (1) if the majority of fat in PPW
is recovered by primary (physical) treatment steps, a MF and/or UF unit inserted directly
following the primary treatment could dramatically reduce the severity of membrane
fouling; (2) although fouling of membrane inevitably remains as the primary concern in
the recovery of valuable byproducts from PPW, means to restore MF and UF
performance (e.g., backwashing) after fouling could be employed in order to make this
process cost-effective; and (3) while reducing the organic and nutrient loads of
downstream effluent, the recovered protein is expected to retain its quality and
functionality because the non-thermal MF and UF operations prevents protein from
undergoing thermal denaturation (Avula et al., 2009).
Permeate flux, defined as the permeate flow per unit area per unit time, is the
major operating parameter monitored during membrane filtration. Permeate flux is
16
typically reported in the unit of liters per square meter of membrane per hour (Lm-2 h-1).
The gradual decline in permeate flux is observed during membrane filtration and it is
indicative of membrane fouling when all other operating parameters (e.g., pressure,
velocity, temperature, concentration) are held constant over an extended period of time
(Avula et al., 2009). In 2005 Lo et al., published work on recovery of protein from PPW
using UF. In this study the volumetric flux declined rapidly from 264 to 140 Lm -2 h-1 in
the first 20 minutes and continued to drop until the end of the filtration process. The
crude protein (CP) concentration of the retentate increased from 80 to 273 mg/L, yielding
a concentration ratio of 3.4. This concentration ratio, which was less than the theoretical
value of 5.0, is determined by the change in liquid volume after UF, indicates loss of
proteins that fouled the polysulfone membrane operating at 25 °C and pH 7.0 S.U.
This initial decrease of flux is typical while using MF and UF to process
proteinaceous materials. This is because protein, like fat, does not have a fixed
conformation but rather is in a dynamic state where conformation and activity are a
compromise between flexibility of structure and stability of the molecules (Campbell et
al., 1993). Pressure, and the resulting shear forces, has been identified as one of the
major energy inputs required to destabilize such structures in membrane filtration
systems (Durchschlag et al., 1996).
Permeate flux is affected by both the temperature and the pH of PPW. In general,
the temperature of PPW increases during membrane filtration due to hydraulic
turbulence, friction with the membrane, and heat transmitted from the pump. Typically, a
cooling coil is used to maintain the operation around 40°C in order to prevent any heat
damage to the membrane (Cheryan and Rajagopalan, 1998). It is also well recognized
17
that pH of a solution impacts the conformation of protein molecules. When PPW is
acidified with hydrochloric acid from its regular pH of ~6.6 to pH 3.0, it becomes less
colloidal and permeate flux significantly increases from 20 to 24 Lm-2 h-1, up from 14 to
16 Lm-2 h-1 (Shih and Kozink, 1980).
Also, the electrical properties of protein molecules play an important role in
determining their interactions with the membrane surface (Martínez et al., 2000).
Therefore to assess the effect of pH on permeate flux, the isoelectric point (pI) of the
protein recovered from PPW needs to be determined. In work published by Martinez et
al. (2000), the pI of the recovered protein was found to be at pH 4.6, which was
substantially different from the optimal pH value, which was determined at pH 6.74. UF
of PPW at pH 6.74 is optimal because it avoids protein coagulation that could result in
adverse effects on membrane performance (Lo et al., 2005).
Increases in volumetric feed flow rate are expected to also increase permeate flux
in MF and UF systems. However, due to rejection of solutes in a membrane separation
process, the excessive increase of flow rate often leads to the problem of the bulk flow
carrying away the majority of solution without permitting adequate retention time,
causing a poor tangential flow for water removal (Chung et al., 2000). The average flux
(Jave) throughout the operation of the membrane filtration unit can be estimated using the
following equation (Lo et al., 2005):
Jave = 0.33 Ji + 0.67 Jr
where Ji is the initial flux and Jr is the fouled flux, with respective leading constants
corresponding to the ratio of average time span during a typical MF/UF batch. Under a
18
transmembrane pressure of 96.5 kPa, Lo et al. found the optimum combination of pH and
flow rate for PPW was found to be 6.74 and 683mL min−1, respectively.
Membrane characteristics
According to Wagner (2001), membranes can be classified as symmetrical or
asymmetrical. Asymmetrical membranes are defined by the internal structure of the
membrane and the pores of these materials are usually larger when located further from
the filter surface. Conversely, symmetrical membranes have uniform pore sizes in their
cross-sections that are identical on both sides of the filter surface.
The main performance characteristics of the any membrane are directly related to
pore size (i.e., diameter of openings), solvent density and viscosity, membrane thickness,
solvent permeability and porosity. Other factors such as temperature, operating pressure,
pH and permeate flow rate also play an essential role in the filtration process (Cheryan
and Rajagopalan, 1998; Ostergaard, 1989).
Types of membranes
Membrane filtration processes such as microfiltration (MF), ultrafiltration (UF),
nanofiltration (NF) and reverse osmosis (RO) are used for wastewater treatment (Cheryan
and Rajagopalan, 1998). Microfiltration is the physical separation of particles from 0.1 to
10μm under pressure through a semi-permeable membrane. It is used for clarification,
sterilization, and to detect or analyze bacteria and other organisms and particulates in
wastewater. Nakao (1994) reported MF uses at pressures lower than 0.2 MPa and
separation of molecules between 0.025 and 10μm in treating wastewater.
Ultrafiltration is physical separation of large molecules according to their
molecular weight under pressure through a semi-permeable membrane (Zeman and
19
Zydney, 1996) . UF uses pressures lower than 10MPa and separates particles between
MWCO values of 1 to 3000Da molar masses (Nakao, 1994). Membranes in the MF and
UF ranges are expressed in microns (µm).
Nanofiltration is also a pressure driven process that separate minute molecular
particles using a semi-permeable membrane. It is a fairly new development in the range
of membrane separation processes. It uses pressures between 1 and 4MPa and separates
particles between MWCO values of 350 to 1000Da molar masses (Nakao, 1994).
Reverse Osmosis (RO) is the tightest possible membrane process in liquid-liquid
separation through a semi-permeable membrane under pressure. It uses pressures
between 4 and 10 MPa and separates particles below MWCO value of 350 Da molar
masses (Nakao, 1994).
Membrane materials are made up of polymers consisting of either organic or
inorganic materials. Organic materials include cellulose acetate (CA), polyamide (PA),
polysulfone (PS), polyvinylidene difluoride (PVDF) and ultrafilic. Inorganic materials
include metals and ceramics (Cheryan and Rajagopalan, 1998; Cuperus and Nijhuis,
1993; Luque et al., 2008; Nakao, 1994)
Cellulose acetate membrane materials operate effectively at a pH range of 3 to 8
S.U. and a maximum temperature of 50oC. It is used for UF, RO and NF applications
(Almas, 1985; Cheryan, 1998). The disadvantage of this membrane material is that they
are vulnerable to disinfectants and can be eaten by microorganisms (Cheryan, 1998;
Wagner, 2001)
Polysulfone membrane materials have been in use since 1975 for UF and MF. The
advantage of this material is its high temperature and pH resistance, however they have
20
low resistance to mechanical compacting (Porter, 1990; Wagner, 2001). PVDF
membrane materials have high resistance to hydrocarbons and oxidizing environments
(Wagner, 2001).
Ceramic membrane materials are inorganic and are made up of alumina oxide or
zirconium oxide placed on graphite material surfaces. Ceramics are resistant to high
temperature (over 400oC) and pH (14 S.U.), and support high pressure levels. The
downside of these materials is high cost compared to organic membranes (Cuperus and
Nijhuis, 1993; Wagner, 2001).
There are also hybrids or combinations of several membranes with different pore
size to obtain optimum results in the filtration process. Aider et al. (2008) used
continuous electrophoresis with porous membrane (CEPM) to separate organic
molecules. This technique can be coupled with dialysis and electrolysis (Bazinet et al.,
1998; Langevin and Bazinet, 2011; Nagarale et al., 2006; Tanaka, 2006). An external
direct electric field is the driving force for the separation process. Under an electric
current effect, the ions (i.e., cations or anions) are transported from one solution to
another through a single or multiple semi-permeable membranes. One commonly known
form of the electro-separation method is conventional electrodialysis, mainly used in the
bio-food industry, biotechnology and nutraceutical industries (Aider et al., 2008).
Factors affecting membrane processes
One major factor affecting membrane performance is the type of filtration flow.
The two major types of filtration flow are cross flow and dead-end flow. In cross flow,
flow is parallel to the membrane surface whereas in dead-end flow, flow is perpendicular
to the membrane surface. The advantage of cross flow is that it does not cause build up or
21
cake formation on the membrane surface and therefore does not suffer as dramatically as
dead-end flow from reduced permeate flux overtime (Almas, 1985; Zeman and Zydney,
1996).
During the initial stage of membrane filtration, there is a gradual fall in the
permeate flow rate which is mainly attributed to concentration gradient (i.e.,
concentration polarization) which may lead to gel-layer formation. The gel-layer acts as a
‘second membrane’ if pressure and the solute concentration in the flow of feed are
significantly high. Attempts have been made to explain mass transference through the
gel-layer. (Cheryan and Rajagopalan, 1998; Mulder, 1991; Mulder, 1995; Porter, 1990;
Roesink et al., 1991; Singh et al., 1998). In the second stage, the flow rate decreases
slowly due to the interactions between the membrane pores and the solute, a process
known as mechanical fouling. In the final stage, the flow rate is relatively stationary due
to interactions between the molecules of the solute and molecules of the membrane
leading to adhesion, a process known as physico-chemical interaction (Dresch et al.,
1999; Khayat et al., 1997; Marshall and Daufin, 1995). Susanto and Ulbricht (2005;
2009) reported that fouling of a membrane is dependent on the chemical constitution of
membrane structure, on the interaction between the solute and membrane and on the
interaction between the molecules of the solute. In other words, fouling develops by the
adsorption of solute onto the membrane surface, which is determined by the interaction
between the solute and membrane.
Blocking of the membrane pores is a result of size and form of the solute in
relation to the membrane pore size distribution. Giorno et al. (1998) and Todisco et al.
(1996) reported that during internal blockage, chemical species are adsorbed on the inside
22
of the membrane pores reducing the permeate flux. In partial blockage, the pores of the
membrane are partial sealed by the solute particles causing reduction in permeate flux.
Complete blockage of the membrane pores occurs when the particles on the surface of the
membrane are larger than the membrane pore causing no permeate flow.
Membrane cleaning processes
Membrane cleaning processes are essential in filtration since it is essential in
developing a membrane’s potential technical and economic viability (de Morais Coutinho
et al., 2008). To obtain the optimum conditions, the cleaning process is dependent on
several factors such pressure, temperature, flow rate, feed concentration, duration of the
cleaning and concentration factor. By developing the proper combination of these
operational parameters, maximum permeate flow rate and minimal fouling can be
obtained (Ebrahim, 1994; Mulder, 1995; Peng and Tremblay, 2008; Smith et al., 2006).
Membrane cleaning can be grouped into four types: physical, chemical, physico-
chemical and biological. The most widely used membrane cleaning procedure is the
chemical method. Cleaning agents such as alkalis, acids, enzymes, surfactants,
sequestering agents, disinfectants and formulated agents have been described in
membrane research (Chakrabarty et al., 2010; Cheryan and Rajagopalan, 1998; Hyun,
1997; Smith et al., 2006). Chemical cleaning has several advantages. As an example,
alkaline cleaners exhibit no cloud point and therefore cleaning can be accomplished at a
high temperature of 82oC. However, there are also several disadvantages to chemical
cleaning methods. These include high costs and handling safety problems (Peng and
Tremblay, 2008). Ang et al. (2006) investigated the chemical and physical aspects of
cleaning organic-fouled RO membranes. The results showed that using a metal chelating
23
agent like disodium ethylenediaminetetraacetate – EDTA Na2 and an anionic surfactant
like sodium dodecyl sulfate – SDS was effective and efficient in reducing the attractive
interactive forces induced by the divalent cations - Ca2+ ions between the two fouling-
causing agents- alginate and natural organic material (Suwannee River natural organic
matter - SRNOM).
Physical methods of cleaning include air sparging (Takizawa et al., 1996), sponge
ball cleaning (Yanagi and Mori, 1980), ultrasonication (Kobayashi et al., 2003; Li et al.,
2002; Muthukumaran et al., 2007), vibration (Ebrahim, 1994), and backflushing (Kim
and Chang, 1991; Levesley and Hoare, 1999; Nakatsuka et al., 1996; Smith et al., 2006).
Physical cleaning is dependent on mechanical action to remove foulant particles from the
membrane surface. Physical cleaning methods have disadvantages such as difficulty in
operation and handling of mechanical cleaning equipment. However, it is often the most
preferred method because of its environmental friendliness in that it does not involve the
use of chemicals or bioactive agents and does not have chemical supply, storage and
handling problems (Peng and Tremblay, 2008).
Backflushing is the physical process of cleaning filters by reversing the flow of
fluid through the system. It can be an effective tool in many filtration applications,
potentially extending system life and reducing cost (Peng and Tremblay, 2008; Smith et
al., 2006). Peng et al. (2005) investigated the use of synthetic bilge water using a
cascaded membrane system consisting of a backflushed coalescing MF membrane as a
pre-treatment for UF of bilge water. Smith et al. (2006) reported the effect of
backflushing in maintaining the flow rate of membrane filtration system to be effective
and efficient in removing the majority of the particles responsible for reversible fouling
24
of a membrane and increasing permeate flow rate while reducing the working pressure of
the system. In physico-chemical cleaning methods there are chemical additions to the
physical process to enhance cleaning effectiveness and efficiency (Ebrahim, 1994;
Maartens et al., 2002).
In the every membrane cleaning process, two main steps of rinsing and cleaning
are completed. Time, water and cleaning agent consumption play essential roles in these
two steps (Heinemann et al., 1988; Kulozik, 1994). The cleaning effect of the rinsing is
provided by mechanical action and the power of water. Rinsing is considered an essential
step to reduce cleaning agent consumption and to restore permeate flux after cleaning.
Cabero et al. (1999) examined the influence of rinsing cleaning procedures in the removal
of whey proteins from ceramic membranes.
Biological methods involve the use of cleaning mixtures containing bioagents to
enhance the cleaning effectiveness (Peng and Tremblay, 2008). Using enzymes as
cleaning agents are also considered biological cleaning (Allie et al., 2003; Argüello et al.,
2003; Petrus et al., 2008).
Life Cycle Assessment of Membrane filtration
Life cycle assessment (LCA) is a method to evaluate environmental impacts
associated with all the stages of a product's life from the beginning to ending stages (i.e.
from raw material stage process through the recycling or disposal stage) (USEPA, 2006).
In 1998, Meijers et al. performed a LCA to membrane filtration process used in the
production of potable water. In their studies, they found the operational stage was
responsible for the majority of environmental problems and the energy consumed in this
stage to obtain optimal filtration pressures was of major importance. Also, the use of
25
chemicals such as H2SO4 for the cleaning in place proved to cause significant
environmental problems. The economic feasibility of incorporating membrane filtration
unit in the poultry processing plant indicates positive impacts. Vedavyasan (2007)
investigated that even though the capital costs of microfiltration and ultrafiltration are
currently 0-25% higher than conventional wastewater treatment systems, and as such
their life cycle costs are comparable with conventional treatments.
Physical - Chemical Treatment of PPW
Harper et al. (1988) and WEF (1998) reported that the most popular form of
physical-chemical treatment in the U.S. poultry processing plants is dissolved air
flotation (DAF). The process of DAF is used to remove the majority of the free O&G and
protein from PPW. The removal is achieved by dissolving air in wastewater under
pressure and then releasing the pressurized air to form fine bubbles at atmospheric
pressure that attach to the suspended fat and protein particles and float then to the surface
where they are collected (Kiuru, 2001; Wang et al., 2005). The most important feature of
the DAF process is the minute bubble formation (Cassell et al., 1975). DAF systems
using chemicals (e.g., alum, ferric sulfate, sodium aluminate and ferrous sulfate) have
been shown to remove more than 90% of O&G and solids, and 70% of the BOD load
(Rusten et al., 1998) in PPW. Bough et al. (1975) reduced the COD in PPW of a broiler
processing plant by 86%. Chemicals enhance the coagulation and flocculation processes
in DAF, which are used to separate the suspended solids portion from wastewater
(Karpati and Szabo, 1984; Rusten et al., 1998; Travers and Lovett, 1985; Woodard et al.,
1972; Woodard et al., 1977). Although, the majority of fat is removed during the DAF
26
treatment, residual fat ranging from 50 to 60 mg/L is found in the post-DAF PPW (Avula
et al., 2009).
Poultry processing plants using DAF suffer from the following disadvantages: (1)
other than specific rendering operations utilizing the degraded nutrients produced from
the DAF process (El Boushy and van der Poel, 1994), only limited efforts have been
made to reclaim DAF skimmings (Grant, 1976; Shih and Kozink, 1980); (2) use of
chemicals (e.g., metal salts, polymers) in the DAF process makes the precipitated protein
useless for many applications (Whittemore, 1994); and (3) potentially valuable
constituents in PPW are degraded during the secondary (biological) treatment to
biosolids, which in itself presents a disposal problem (Barik et al., 1991; Janosz Rajczyk,
1993).
Biological Treatment of PPW
The most widely utilized biological treatment systems for PPW can be classified
as either aerobic (e.g., activated sludge, trickling filters), anaerobic (e.g., covered
lagoons, sealed reactors) or facultative (i.e., a combination of aerobic and anaerobic
treatment) (Del Nery et al., 2007; Nemerov and Dasgupta, 1991). All biological
wastewater treatment processes involve the use of microorganisms (e.g., bacteria) to
convert both inorganic and organic materials into a stable biomass and treated effluent
(Del Nery et al., 2001)
Aerobic treatment processes such as activated sludge systems are limited by high
levels of energy needed for aeration and high sludge volume production (del Pozo et al.,
2000; Nguyen and Shieh, 2000). Anaerobic treatment processes have the ability to treat
higher strength waste streams as compared to aerobic system, but often required
27
subsequent aerobic treatment to meet environmental regulatory permit limits for effluent
discharge.
Biological treatment systems that have been used to effectively treat PPW include
sequential batch reactors (SBR), moving bed bioreactors (MBBR), membrane bioreactors
(MBR), as well as conventional activated sludge plants (CASP). All these biological
treatment systems can achieve more than 90% organic matter removal in PPW.
Sequential batch reactors (SBR) are designed to reduce the organic and nutrient
loadings in medium strength wastewater to comply with surface water discharge limits or
as a pre-treatment stage to tertiary recycling plants. Oxygen is passed in a form of
bubbles through the wastewater to reduce BOD and COD before release into POTWs.
The advantages of an SBR include a reduced footprint compared to CASP and since a
portion of the biomass is maintained in the reactor after decanting, no sludge recycling
system is required. However, the requirement of a settling tank is an important
disadvantage due to the lack of physical filtration (Andreottola et al., 2001; Bernet et al.,
2000)
Moving bed bioreactors (MBBR) used an activated sludge aeration system where
the sludge is collected on carriers (biofilters). These carriers made of plastic have an
internal surface for water, air and bacteria contact. The biofiltering process is used for the
removal of organic substances, nitrification and denitrification (Pastorelli et al., 1997).
This process is quite costly compared to CASP due to increased operation and
maintenance (Andreotolla et al., 2000).
The membrane bioreactors (MBR) process involves an activated sludge system
that uses membranes for solute separation in place of secondary clarifiers. The membrane
28
component uses low pressure MF or UF membranes. The cost of building, maintaining
and operating an MBR is usually high compared to CASP. The disadvantage of this
system includes the fouling of the membranes (Bernal et al., 2002).
Conventional activated sludge plants (CASP) involve atmospheric air being
introduced to a mixture of primary treated or screened wastewater combined with
organisms to develop a biological flocculation which reduces the organic content of the
wastewater. The major disadvantages of this process are high energy costs and poor
settling of sludge caused by the problematic filamentous microbes. In the presence of
inadequate oxygen and iron, the filamentous microbes flourish and cause the biomass to
not settle in the clarifier. Also, CASP are poor removers of nitrate and phosphate
compounds (Woolard and Irvine, 1995).
The anaerobic treatment process occurs in the absence of free-oxygen (Metcalf
and Eddy Inc., 2003). Digestion of organic materials by microorganisms under these
conditions results in the production of a gaseous byproduct. This resulting gas mixture,
mostly methane and carbon dioxide, with smaller amounts of hydrogen, hydrogen sulfide,
and ammonia, is referred to as ‘biogas’. Despite the no free-oxygen and low energy
requirement during digestion, further treatment with an aerobic treatment process may be
needed to meet discharge requirements (Nguyen and Shieh, 2000).
Thesis Goal and Objectives
The goal of this study was to evaluate the effect of membrane filtration on pre- and post-
DAF PPW.
The specific objectives were;
1. To evaluate the performance of different membrane pore sizes in
Microfiltration (0.1-10µm) and Ultrafiltration (0.01 - 0.1 µm) pore size
ranges,
29
2. To evaluate the performance of three membrane materials; polysulfone (PS),
polyvinylidenedifluoride (PVDF) and Ultrafilic,
3. To determine optimum conditions for membrane filtration with respect to
flow rate, temperature and operating pressure, and
4. To measure the effect of membrane filtration on chemical oxygen demand
(COD), total solids (TS) and total suspension solids (TSS) in PPW.
30
CHAPTER 3
EFFECTS OF MEMBRANE FILTRATION ON
PRE- DAF (DISSOLVED AIR FLOTATION) POULTRY PROCESSING
WASTEWATER1
_________________________
1
Abboah-Afari, E., and B.H. Kiepper. To be submitted to the ASABE Journal Applied
Engineering in Agriculture
31
Abstract
An experiment was conducted to measure the effects of membrane filtration on
pre-dissolved air flotation (DAF) poultry processing wastewater (PPW). Bulk grab
samples of raw PPW were collected after primary (1500–3000µm) and secondary
(500µm) offal screens, but before additional wastewater treatment at a commercial
broiler processing plant. 4L subsamples of raw PPW were pre-sieved (500/106µm) and
then filtered using one of three membranes (0.3µm polyvinylidene difluoride (PVDF),
0.1µm Polysulfone, or 100,000 MWCO Ultrafilic) at one of two pressures (50 or 80 psi)
for three 2-hour trials. The resulting 3x2 treatments with 3 repetitions totaling 18 trials
were completed over 6 wks, 3 trials taking place one day per wk. Mean chemical oxygen
demand (COD) and total solids (TS) concentrations (mg/L) were determined for PPW
samples prior to each trial. Mean permeate flux was calculated, and membrane permeate
PPW samples were collected at 10 minute intervals during each trial for subsequent COD
and TS analyses. Results showed that COD and TS for raw PPW samples (sieved at
500µm) were 5263 mg/L and 3355 mg/L, respectively. COD and TS for pre-sieved PPW
samples (sieved to 106µm) were 4390 and 2991 mg/L, respectively, which represented
16.6% and 10.8% reductions, respectively. The maximum mean permeate flux of 115
Lm-2h-1 and TS reduction of 34% was obtained by the 0.3µm PVDF membrane, while the
maximum mean COD reduction of 89% was achieved by the 100,000MWCO Ultrafilic
membrane. Based on a calculated performance index (Pmi), the 0.3µm PVDF membrane
was determined to be the most effective membrane for pre-DAF PPW treatment.
Keywords: Poultry processing, wastewater, membrane filtration, permeate flux, COD,TS
32
Introduction
In 2009, the U.S. poultry industry slaughtered 8.5 billion broilers (USDA, 2010),
using an average of 26L (7 gal) of potable water per bird. Thus, the processing of
broilers in the U.S. in 2009 generated approximately 2270 million L (60 billion gal) of
high strength poultry processing wastewater (PPW) requiring treatment. Current
wastewater treatment methods provide viable ways of reducing pollutants such as
organics, solids and nutrients in PPW, but often neglect the fact that advanced wastewater
treatment alters and adulterates potential valuable poultry processing byproducts (e.g.,
protein, oil and grease) producing low-value chemical skimmings and biological sludges
(Ockerman and Hansen, 2000; Wessels, 1972). Previous analysis of solids recovered
from PPW using microscreening have been reported to average 63.5% fat, 17.5% protein,
4.8% crude fiber and 1.5% ash on a dry weight basis (Kiepper et al., 2008).
Dissolved air flotation (DAF) is the most popular physical/chemical PPW
treatment method currently utilized by U.S. poultry processors with approximately 80%
of slaughter plants employing the technology. Although DAF is effective at PPW
treatment, the aggressive aeration of the fine particulates causes excessive oxidative
damage and bacterial degradation of fat and protein components. Also, various
combinations of chemicals are added to enhance the coagulation and flocculation of the
particulates. The combination of this physical and chemical treatment greatly reduces the
value of DAF skimmings as a nutrient source in rendered animal byproduct feed
ingredients. Moreover, utilization and disposal of DAF skimmings is a major expense and
an inconvenience to the poultry industry. Thus, the U.S. poultry processing industry
33
would benefit greatly by PPW treatment technology aimed at recovering unadulterated
byproducts prior to DAF treatment (i.e., pre-DAF PPW).
The benefits for pre-DAF PPW treatment include a substantial increase in the
volume of unadulterated byproduct (i.e., offal) recovered for rendering with a
corresponding reduction in the volume of lower value DAF skimmings produced,
substantial chemical cost savings, and reduced energy consumption (Avula et al., 2009).
One emerging technology of interest in treatment of high-strength wastewater is
membrane filtration. Even though membrane filtration technology has been studied in
various wastewater studies, only limited work has been published on the treatment of
PPW. In 2005, Lo et al. successfully recovered concentrated protein from PPW using
ultrafiltration (UF). The study was successful, however the work was focused on
recovery of a potential high-value byproduct as a retentate rather than the effect
membrane filtration had on PPW treatment efficiency.
In membrane filtration, protein and fat exhibit similar characteristics in terms of
flexible and stable structures. It is possible that a membrane surface can be operated at its
highest flux for a longer time, minimizing the conditions caused by fouled membrane
(Cheryan and Rajagopalan, 1998; Yushina and Hasegawa, 1994). Increases in volumetric
feed flow rate are expected to increase flux in the membrane filtration system which
reduces trapping or retention efficiency. An experiment was conducted to quantify the
efficiency of membrane filtration in reducing COD and TS concentration (mg/L) and
increase the permeate volume (i.e., flux) in pre-DAF PPW at optimum conditions.
34
Materials and Methods
Grab samples of approximately 40L of pre-DAF PPW were collected in large
plastic containers after the primary (1,500-3,000µm) and secondary (500µm) physical
mechanical screens at a north Georgia broiler processing plant during normal slaughter
operations and transported immediately to the laboratory. Within 1 hr of collection, 4L
subsamples of PPW were poured through a 500µm sieve and designated as ‘raw PPW’.
Each 4L raw PPW sample was then poured through a 106µm sieve and designated ‘pre-
sieved PPW’ in preparation for membrane filtration trials. The pre-sieved PPW
subsamples ranged in temperature from 26-30oC (79-86oF) and pH from 6.0-6.2 S.U.
The experimental trials were carried out using a Spintek STC bench-scale
membrane filtration system (Spintek, 2008). As shown in Figure 3.1, this system
simulates a full-scale membrane filtration system by circulating raw wastewater from a
feed tank, past a sample (i.e., coupon) of flat-sheet test membrane in a cross-current
configuration under pressure. A variable speed pump and back pressure control valve
can be manipulated to maintain constant pressure on the membrane. The wastewater
stream circulates back through the feed tank as a concentrate, while the treated water
passes through the membrane as permeate.
The experiment was designed in a 3 x 2 factorial configuration with 3 membrane
sizes/materials (100,000MWCO Ultrafilic, 0.10µm Polysulfone and 0.30µm
polyvinylidene difluoride (PVDF)), operated at 2 pressure levels (50 and 80 psi) for a
total of 6 treatments. The 6 treatments were conducted once per wk for 3 wks for a total
of 18 trials. Each test membrane had an effective surface area of 0.005m2.
35
Figure 3.1. SpinTek Static Test Cell (STC) Membrane Filtration System (Spintek, 2008)
Each trial consisted of a 4L pre-sieved subsample of PPW being placed into the
feed tank of the STC system fitted with one of the 3 membranes and operated for 2 hours
at one of the 2 pressure levels. A total of 16 wastewater samples were collected in 1L
glass jars during each trial. Representative raw PPW and pre-sieve PPW samples were
collected prior to each membrane filtration trial (i.e., samples 1 and 2). During each trial,
membrane filtration permeate samples were collected every 10 minutes (i.e., samples 3-
12) with the volume of effluent (mL) noted for subsequent permeate flux calculations.
Finally, 1L concentrate samples were collected after 1 and 2 hr (i.e., samples 15 and 16).
The pH of all samples was adjusted to < 2.0 S.U. using H2SO4 as a preservative and
stored at 4oC prior to analysis.
36
Analytical Methods
All PPW samples were analyzed for COD (chemical oxygen demand method
5220D) and TS (total solids method 2540B) (APHA, 2005). COD was used to determine
the concentration of organic materials in each PPW sample, while the TS test was used to
determine the concentration of solids present in each PPW sample.
Statistical Analysis
Data were subjected to statistical analysis by the SAS JMP 8.0.2 program (SAS
Institute, 2009). Data from the 3 x 2 treatments with 3 replications were analyzed by
factorial ANOVA with membrane at 3 levels and pressure at 2 levels as the main effects.
Data was first run as a factorial ANOVA with a main effects (i.e., membrane and
pressure) interaction term. If the interaction was not significant (P>0.05), ANOVA was
re-run without the interaction term and each factor was analyzed independently. Means
were separated using the Tukey-HSD procedure (SAS Institute, 2009). Difference in
means were regarded as significant if P<0.05.
Data Analysis
Permeate Flux
Permeate flux is the membrane effluent flow per unit area per unit time. This is
the most common calculation used to determine membrane efficiency (Cheryan and
Rajagopalan, 1998). In this experiment, permeate flux was calculated and expressed as
liters per square meter of membrane per hour (Lm-2h-1). The permeate volume collected
during each 10 min interval of trial runs was recorded in mL. The total volume (mL)
collected for the 10 min interval was divided by 10 to produce a mL/min. This was
37
converted to L/hr and then divided by 0.005 m2 (i.e., the surface area of each test
membrane coupon) to produce the reported permeate flux value (Lm-2h-1).
COD and TS Percentage Reduction
Data were analyzed for percentage reduction in concentration of COD and TS. This
was achieved by subtracting the end point mean concentration (i.e., pre-sieved or
permeate PPW) from the corresponding start point mean concentration (i.e., raw or pre-
sieved PPW), and dividing the difference by the start point mean concentration. The
result was then multiply by 100%.
Membrane Efficiency
The most effective and suitable membrane for pre-DAF PPW membrane filtration
was generated using weighted linear aggregation (WLA) of the three parameters of COD,
permeate flux and TS. WLA is a weighted average in which the decision maker assigns
the weights of relative importance to each criterion (Eastman et al., 1995). The relative
weights of each criterion were computed using a pair-wise comparison matrix (Saaty,
1977). As shown in Table 3.1, the value of each matrix cell represents the relative
importance of the row criterion against the column criterion within a range from 1 to 9. A
consistency ratio (CR) was used to determine the probability that the matrix ratings were
randomly generated which should be less than 0.1. Each relative importance weight was
evaluated using the CR. The relative weights are computed by corresponding the
maximum eigen value with the relative values of the eigen vector.
Consistency ratio is expressed by
where λmax is the maximum eigen value of the matrix and n is the total number of criteria.
38
Table 3.1. Pair-wise comparison of the criteria (i.e., COD, Permeate flux, TS) for
membrane filtration of pre-DAF poultry processing wastewater
COD Permeate Flux TS Weight

COD 1 3/2 2 0.46
Permeate Flux 2/3 1 3/2 0.32
TS 1/2 2/3 1 0.22
Consistency ratio (CR) = 0.00077
A performance index was used to indicate which of the membranes was the most
effective and suitable for the membrane filtration of pre-DAF PPW. Performance index
is expressed by , where Pmi =
Performance index, wcod and Ricod = weight and rank of COD, wf and Rif = weight and
rank of flux, wts and Rits = weight and rank of TS. Ranks were developed in descending
order based on their significant differences (i.e. from the best rank (4) to the least rank
(1)).
Results and Discussion
Permeate Flux
Permeate flux (Lm-2h-1) values calculated at 10 minute intervals and averaged for
the 3 repetitions at each membrane size over the course of the 2 hour STC system trials at
50 and 80 psi are summarized in Table 3.2 and 3.3, and graphically shown in Figures 3.2
and 3.3, respectively. These permeate flux values were in the expected range and are
relatively similar to the results reported by Lo et al. (2005).
39
Table 3.2. Mean permeate flux (Lm-2h-1±SEM), COD (mg/L±SEM) and TS
(mg/L±SEM) values for 6 membranes filtering pre-dissolved air flotation (DAF) poultry
processing wastewater for 120 min
Membrane size/material Mean Permeate Mean Permeate Mean Permeate

Flux COD TS
-2 -1
(Lm h ±SEM) (mg/L±SEM) (mg/L±SEM)
(P<0.0001) (P<0.0001) (P<0.0001)
100,000MWCO Ultrafilic 109a±3 465b±13 2324b±61
0.10µm Polysulfone 91ab±3 869a±40 2686a±79

0.30µm PVDF 115a±5 530b±16 1978c±51
a,b,
- differing superscripts within a column indicates statistically significant difference
(P<0.05)
Table 3.3. Mean permeate flux (Lm-2h-1±SEM), COD (mg/L±SEM) and TS

(mg/L±SEM) values at 2 pressure levels (50 and 80psi) for pre-dissolved air flotation
(DAF) poultry processing wastewater for 120 min
Pressure Mean Permeate Mean Permeate Mean Permeate

(psi) Flux COD TS
(Lm-2h-1±SEM) (mg/L±SEM) (mg/L±SEM)
(P<0.0001) (P<0.0001) (P=0.3007)
50 96b±4 682a±33 484±81
80 114a±3 560b±37 359±60
a,b,
(P<0.05)
Visual inspection of Figures 3.2 and 3.3 show that while the 100,000MWCO
Ultrafilic and 0.30µm PVDF membranes flux values produced a flux curve that was
downward sweeping and similar results reported by Lo et al. (2005), the 0.10µm
Polysulfone membrane did not. At both pressures, the 0.10µm Polysulfone membrane
produced flux curves that were relatively flat with only slight increases over the 2-hour
trial time period.
40
Figures 3.2 and 3.3 also showed that the differences between the 3 membranes in
flux curves seen over time at the 50 psi level were negated at the 80 psi level. Finally, all
trials showed a steady state permeate flux after 2 hours with no visible trend towards
blinding for of membranes, with the exception of the 0.10µm Polysulfone membrane at
80 psi which indicates the start of a flux decline at 110 minutes.
230
210
100,000 MWCO Ultrafilic
190
0.1 µm Polysulfone
FLux (Lm-2h-1)
170
0.3 µm PVDF
150
130
110
90
70
50
10 20 30 40 50 60 70 80 90 100 110 120
Elapsed Time (minutes)
Figure 3.2. Pre-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at 10
minute intervals for 3 membrane filters at 50 psi operating pressure
41
210
190
170 100,000 MWCO Ultrafilic
0.1 µm Polysulfone
FLux (Lm-2h-1)
150 0.3 µm PVDF

130
110
90
70
50
10 20 30 40 50 60 70 80 90 100 110 120
Elapsed Time (minutes)
Figure 3.3. Pre-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at 10
minute intervals for 3 membrane filters at 80 psi operating pressure
A mean permeate flux value was calculated for each membrane by averaging the
twelve (12) 10-minute interval flux values during each trial. Statistical analysis of mean
permeate flux results showed no interaction between membrane and pressure (P=0.5293).
Therefore, the main effects were analyzed independently. The total mean permeate values
at the 3 membrane sizes and the 2 pressure settings are summarized in Figures 3.4 and
3.5.
As shown in Table 3.2 and graphically represented in Figure 3.4, the mean
permeate flux values of the 0.30µm PVDF (115 Lm-2h-1) and 100,000MWCO Ultrafilic
(109 Lm-2h-1) membranes were not significantly different, however the mean permeate
flux of the 0.10µm Polysulfone (91 Lm-2h-1) membrane was significantly lower. As
expected, the membrane with the largest nominal gap openings (i.e., 0.30µm PVDF) had
the largest total mean flux values, however it was unexpected that the membrane with the
42
next largest nominal gap openings (i.e., 0.10µm Polysulfone) would have the lowest
mean flux values. These results would indicate that the membrane material (i.e.,
Polysulfone) played a significant role in reducing the separation efficiency of the 0.10µm
membrane.
115a
109a
91b
Figure 3.4. Pre-DAF poultry processing wastewater mean permeate flux values (Lm-2h-1)
for 3 membranes
43
114a
96b
Figure 3.5. Pre-DAF poultry processing wastewater mean permeate flux values (Lm-2h-1)
for 3 membranes operated at 50 and 80 psi
From the results shown in Table 3.3 and Figure 3.5, the mean permeate flux at 80
psi (114 Lm-2h-1) was significantly greater than at 50 psi (96 Lm-2h-1). From these results
it can be concluded that for the optimum efficiency and high permeate flux, either the
0.3µm PVDF or 100,000MWCO Ultrafilic membrane at 80 psi operating pressure should
be selected.
COD Concentration
The concentration (mg/L) of organic matter in each PPW sample was measured
using COD. COD concentration data were also used to calculate a COD removal
efficiency value, as a percentage (%), for pre-sieved (i.e., 106µm) and membrane
permeate PPW samples. The membrane producing the lowest mean COD concentrations
and highest COD removal percentages was deemed the most effective.
44
The mean COD concentrations (mg/L) for raw PPW and pre-sieved PPW samples
for the 18 membrane filtration trials were 5263 and 4390 mg/L, respectively. Thus the
106µm pre-sieving reduced the COD concentration on average by 16.6%.
Statistical analysis of mean COD results showed no interaction between
membrane and pressure (P=0.7439). Therefore, the main effects were analyzed
independently. The mean COD values at the 3 membrane sizes and the 2 pressure settings
are summarized in Table 3.2 and 3.3, and graphically represented in Figures 3.6 and 3.7,
respectively.
From results shown in Table 3.2 and Figure 3.6, expectantly, the 100,000MWCO
Ultrafilic membrane (i.e., the membrane with the smallest nominal gap openings)
produced the permeate with the lowest mean COD concentration (465 mg/L), but this
value was not significantly different from the 0.30µm PVDF membrane (530 mg/L). The
465 mg/L effluent COD concentration represents an 89% removal efficiency by the
100,000MWCO Ultrafilic membrane as compared with the pre-sieved sample of mean
COD concentration of 4390 mg/L.
45
868a
530b
465b
Figure 3.6. Pre-DAF poultry processing wastewater mean permeate COD concentrations
(mg/L) for 3 membranes
682a
560b
Figure 3.7. Pre-DAF poultry processing wastewater mean permeate COD concentrations
(mg/L) at 50 and 80 psi
46
Unexpectantly, the 0.10µm Polysulfone membrane produced the highest mean
COD concentration (868 mg/L), which was significantly higher than both the 0.30µm
PVDF and 100,000MWCO Ultrafilic membranes. The 868 mg/L effluent COD
concentration represented an 80% removal efficiency as compared with the pre-sieved
PPW sample mean COD concentration of 4390 mg/L. With a mean of 530 mg/L of
effluent COD concentration, the 0.30µm PVDF had an 88% removal efficiency, while the
465 mg/L effluent concentration of the 100,000MWCO Ultrafilic membrane represented
an 89% reduction. Like the permeate flux results, the COD concentration results showed
that the 0.10µm Polysulfone membrane was the least efficient producing the highest
mean COD values.
Results in Table 3.3 and Figure 3.7 showed that the mean permeate COD
concentration at 80 psi (560 mg/L) was significantly better than at 50 psi (682 mg/L). As
with permeate flux, from these results it can be concluded that, for the lowest COD
permeate concentration and corresponding highest removal efficiencies either the
100,000MWCO Ultrafilic or 0.3µm PVDF membrane at 80 psi operating pressure should
be selected.
TS Concentration
The concentration of solids in each PPW sample was measured using TS. TS
concentration data were also used to calculate a TS removal efficiency value for pre-
sieved (i.e., 106µm) and membrane permeate PPW samples. The membrane producing
the lowest mean TS concentrations and highest TS removal percentages was deemed the
most effective.
47
The mean TS concentrations (mg/L) for raw PPW and pre-sieved PPW samples
106µm pre-sieve reduced the TS concentration on average by 10.8%.
Statistical analysis of mean TS results showed significant interaction between the
membrane and pressure main effects (P=0.0366). Therefore, the main effects were
analyzed simultaneously. The mean TS values at the 3 membrane sizes and the 2 pressure
settings are summarized in Table 3.2 and 3.3, and Figures 3.8 and 3.9, while each
membrane was analyzed at the 2 pressure levels, as shown in Figures 3.10, 3.11 and 3.12
to determine which membrane/pressure combinations performed best at TS reduction.
2685a
2324b
1978c
Figure 3.8. Pre-DAF poultry processing wastewater mean permeate TS concentrations

(mg/L) for 3 membranes
48
Surprisingly as shown in Figure 3.8, the 0.30µm PVDF membrane (i.e., the
membrane with the largest nominal gap openings) produced the permeate with the
significantly lowest mean TS concentration (1978 mg/L) as compared with the two other
smaller gap opening membranes. This resulting mean permeate TS concentration
represents a removal efficiency of 34% as calculated against the pre-sieved PPW sample
mean TS concentration of 2991 mg/L.
The 100,000MWCO Ultrafilic membrane produced a permeate with a mean TS
concentration (2324 mg/L) which was significantly lower than the 0.10µm Polysulfone
membrane (2685 mg/L). The 100,000MWCO Ultrafilic membrane effluent TS
concentration represented a 22% TS removal efficiency, while the 0.10µm Polysulfone
membrane had a 10% TS removal efficiency. Like the permeate flux and COD results,
the TS mean concentration results showed that the 0.10µm Polysulfone membrane was
the least efficient, producing the highest mean TS values. Figure 3.9 shows that the mean
permeate TS concentration at 50 psi (2381 mg/L) and 80 psi (2277 mg/L) were not
significantly different (P=0.1481).
49
2381 2277

(mg/L) at 50 and 80 psi
2321 2326

(mg/L) for 100,000MWCO Ultrafilic membrane at 50 and 80 psi
50
2871a
2500b

(mg/L) for 0.10µm Polysulfone membrane at 50 and 80 psi
1953 2004

(mg/L) for a 0.30µm PVDF membrane at 50 and 80 psi
51
The mean permeate TS concentrations for each membrane at both pressures are
shown in Figures 3.10, 3.11 and 3.12. Results represented in Figures 3.10 and 3.12
showed that there was no significant difference in mean TS concentration for the
100,000MWCO Ultrafilic (P=0.9676) and 0.30µm PVDF membranes (P=0.6246) at
either pressure. The mean permeate TS concentrations for 100,000MWCO Ultrafilic
membrane at 50 and 80 psi were 2321 mg/L and 2326 mg/L, respectively. The mean
permeate TS concentrations for the 0.30µm PVDF membrane at 50 and 80 psi were 1953
mg/L and 2004 mg/L, respectively. Thus, operating pressure did not significantly affect
the TS removal efficiency of either the 100,000MWCO Ultrafilic or 0.30µm PVDF
membranes.
However, the mean permeate TS concentration for the 0.10µm Polysulfone
membrane at 80 psi (2500 mg/L) was significantly (P=0.0152) lower than that at 50 psi
(2871 mg/L). These results suggest that operating the 0.10µm Polysulfone membrane at
the higher 80 psi pressure will significantly improve TS removal versus the lower 50 psi.
Conclusions
Using the previously described performance index (Pmi), the most effective
membrane for treatment of pre-DAF PPW was determined. Based on calculated Pmi
values as shown in Table 3.4, the 0.3µm PVDF membrane (i.e., the membrane with the
largest nominal gap openings) was determined the most effective (Pmi = 2.54 on a 3.0
scale) in regards to increase in permeate flux, and COD and TS reduction. The
100,000MWCO Ultrafilic membrane (Pmi = 2.32) was the second most effective
membrane, while the 0.1µm Polysulfone membrane (Pmi = 1.32) was the least effective
52
at pre-DAF PPW treatment. The maximum mean permeate flux of 115 Lm-2h-1 and TS
reduction of 34% was obtained by the 0.3µm PVDF membrane, while the maximum
mean COD reduction of 89% was achieved by the 100,000MWCO Ultrafilic membrane.
Table 3.4. Best performing membrane for treatment of pre-DAF poultry processing
wastewater based on maximum permeate flux, and reduction of COD and TS over 120
min of operation based on significant difference (P<0.05)
COD Permeate TS Performance Index

Flux (Pmi)
1 0.3µm PVDF 2 3 3 2.54
2 100,000 MWCO 2 3 2 2.32
Ultrafilic
3 0.1µm Polysulfone 1 2 1 1.32
The experiment determined that pre-screening of pre-DAF PPW to 106µm prior
to membrane filtration reduced COD by 16.6% and TS by 10.8% making pre-screening a
potentially valuable tool in membrane filtration PPW treatment. Membrane filtration of
pre-DAF PPW reduced the COD concentration by 80 to 89%, and TS concentration by
10 to 34%. Thus experiment results indicate that membrane filtration in the MF and UF
ranges will be substantially more effective at reducing the organic load of pre-DAF PPW
as opposed to solids removal.
Mean permeate flux values (Lm-2h-1) calculated in this experiment can be used to
determine the surface area of membrane (m2) required to treat a known volume of PPW
in a full scale membrane filtration system. As an example, a typical broiler processing
plant slaughtering 250,000 birds per day (bpd) and utilizing 26 L per bird will produce
6,500,000 L of PPW. If a membrane filtration system was designed to operate 24 hours
per day, then 270,833 L of PPW would need to be treated each hour. Results of this
53
experiment showed the top performing membrane in terms of permeate flux was the
0.3µm PVDF membrane which averaged 115 Lm-2h-1. Thus, by dividing 270,833 Lh-1 by
115 Lm-2h-1 it can be determined that 2355 m2 would be needed to filter the total PPW
stream using 0.3µm PVDF membranes.
54
References
APHA. (2005) Standard Methods for the Examination of Water and Wastewater 21st ed.
Am. Public Health Assoc., Washington, DC.
Avula R.Y., Nelson H.M., Singh R.K. (2009) Recycling of poultry process wastewater by
ultrafiltration. Innovative Food Science & Emerging Technologies 10:1-8.
Cheryan M., Rajagopalan N. (1998) Membrane processing of oily streams. Wastewater
treatment and waste reduction. Journal of Membrane Science 151:13-28.
Eastman J.R., Jin W., Kyem P.A.K., Toledano J. (1995) Raster procedures for multi-
criteria/multi-objective decisions. Photogram. Eng. Rem. Sen.:539–547.
Kiepper B.H., Merka W.C., Fletcher D.L. (2008) Proximate composition of poultry
processing wastewater particulate matter from broiler slaughter plants. Pou.Sci.
87:1633-1636.
Ockerman H.W., Hansen C.L. (2000) Animal ByProduct Processing and Utilization.
Technomic Publishing Company, Lancaster, Pennsylvania.
Saaty T.L. (1977) A scaling method for priorities in hierarchical structures. Journal of
Mathematical Psychology 15:234-281.
SAS Institute. (2009) SAS JMP 8.0.2 Statistical Program. SAS Institute. Cary, N.C.
Spintek. (2008) Static Test Cell (STC) Membrane Filtration System. Retrieved from
http://www.spintek.com/ stc.htm.
USDA. (2010) Poultry Slaughter. 2009 Annual Report. U.S. Department of Agriculture.
National Agricultural Statistics Service. Pou 2-1(10).
Wessels J.P.H. (1972) A study of the protein quality of different feather meals. Poultry
Sci. 51:537.
55
Yushina Y., Hasegawa J. (1994) Process performance comparison of membrane
introduced anaerobic-digestion using food-industry waste-water, Elsevier Science
Bv. pp. 413-421.
56
CHAPTER 4
EFFECTS OF MEMBRANE FILTRATION ON
POST- DAF (DISSOLVED AIR FLOTATION) POULTRY PROCESSING
WASTEWATER1
______________________
1
Abboah-Afari, E., and B.H. Kiepper. To be submitted to the ASABE Journal Applied
Engineering in Agriculture
57
Abstract
An experiment was conducted to measure the effects of membrane filtration on
post-dissolved air flotation (DAF) poultry processing wastewater (PPW). Grab samples
were collected and passed through a 500µm sieve and 106µm sieve to prepare the
samples for membrane filtration. 4L presieved subsamples were filtered through one of
six membranes (30,000MWCO Polysulfone, 30,000MWCO PVDF, 100,000MWCO
Ultrafilic, 100,000MWCO PVDF, 0.10µm Polysulfone and 0.30µm PVDF) operated at
one pressure level (50 psi) with membrane pore size and membrane material as the main
effects. The six treatments were conducted for once per wk for 3 wks for a total of 18
trials. The mean permeate flux values were calculated by averaging the six 10-minute
interval flux values from each STC system one hour run. Permeate samples were
analyzed for concentration (mg/L) of COD, TS and TSS. Results showed that mean
COD, TS and TSS for raw post-DAF PPW samples (sieved to 500µm) were 635, 1706
and 49 mg/L, respectively. Mean COD, TS and TSS for pre-sieved post DAF PPW
samples (sieved to 106µm) were 624, 1604 and 48 mg/L, respectively. Thus, pre-sieving
reduced COD by 1.7%, TS by 5.9% and TSS by 2.0% prior to membrane filtration. All
trials ended with permeate flux in a steady state indicating good resistance to blinding by
the membranes. The maximum mean permeate flux value of 224 Lm-2h-1 and lowest
mean TS value was obtained by the 0.3µm PVDF membrane. The lowest COD
concentration, and corresponding highest reduction percentage, was obtained by the
100,000MWCO Ultrafilic membrane (373 mg/L, 40%), which had the highest
performance index (Pmi = 3.56 on a 4.0 scale) of the 6 membranes tested.
Keywords: Poultry processing, wastewater, membrane filtration, permeate flux, COD,
TS, TSS
58
Introduction
In 2010, the U.S. Department of Agriculture reported that approximately 8.8
billion broilers were slaughtered in the U.S. with an average live weight per bird of 2.6
kg (5.7 lbs) and a total live weight of 22.7 million metric tons (50.1 billion lbs) (USDA,
2011). It has been reported that U.S. broiler processing plants use 26 L (7 gal) of water
during the processing of poultry and the cleaning of plants and equipment (Northcutt and
Jones, 2004; Veerkamp, 1999). Treatment and potential reuse of the resulting poultry
processing wastewater (PPW) could benefit poultry processing plants by reducing water
demand, effluent volume and energy consumption (Avula et al., 2009), while meeting
discharge limits set forth in environmental regulatory permits.
A typical poultry processing plant’s wastewater treatment system consists of
mechanical physical screens (designed to remove inedible offal), DAF (dissolved air
flotation) units, and biological treatment provided either onsite (i.e., direct discharge
system) or by a public utility (i.e., indirect discharge or pretreatment system). The first
treatment process, mechanical physical screening, recovers offal which is a valuable raw
material for the rendering industry and is a profit center for poultry processing plants.
However, the subsequent treatment processes of PPW such as DAF suffer several
disadvantages including effluent containing residual fat in the range of 50 to 60mg/L and
adulterated byproducts (e.g., DAF skimmings, biosolids) that have diminished value (Lo et
al., 2005), all these costs at the expense of the processor.
Most poultry processors in the U.S. use DAF for treatment of their wastewater
(Harper et al., 1988). The concentration or strength of treated wastewater varies from one
plant to another depending on the standards set, methods of operations, type of systems and
59
processing loads (Vidal et al., 2000). The process of DAF is used to remove the majority of
the free oil and grease (O&G) and particulate protein from PPW. The removal is achieved
by dissolving air in the wastewater under pressure and then releasing the pressurized air to
form fine microbubbles at atmospheric pressure that adhere to the suspended fats and
protein, and float them to the surface where they are collected (Kiuru, 2001). DAF
performance is achieved by the number and the size of the air bubbles formed by the
pressure relief of the supersaturated air in the wastewater stream (Cassell et al., 1975;
Nielsen, 1989). DAF systems using chemicals (e.g. alum, ferric sulfate, sodium aluminate,
and ferrous sulfate) can achieve chemical oxygen demand (COD) reductions ranging from
32 to 90% and are capable of removing large amounts of nutrients (Mittal, 2006).
Chemicals enhance the coagulation and flocculation processes, which are used to separate
the suspended solids from the wastewater.
Using membrane filtration in post-DAF PPW treatment could help further reduce
fats, bacteria and other particulate matters (Avula et al., 2009), reduce the cost of operation
of biological treatment systems, and lower sludge volume production in direct discharge
systems. Membrane filtration could also significantly reduce the amount of surcharge fees
paid by indirect or pretreatment dischargers to local utilities for the treatment of their high-
strength wastewater streams. Significant reduction in concentration of chemical oxygen
demand (COD) and total solids (TS) of PPW by ultrafiltration (UF) was reported by Lo et
al. (2005).
There are several factors that influence the effectiveness and efficiency of
filtration membranes in treating wastewater. Increase in feed concentration and pressure
alters the viscosity, density and diffusivity of the feed solution, thus causing a decrease in
60
permeate flow rate (i.e., flux) (Manjula and Subramanian, 2006). An increase in
temperature results in a decrease in fluid viscosity and increase in molecular diffusivity
through the membrane (Cheryan and Rajagopalan, 1998; Marenchino, (2006).; Nakao,
1994; Zak and Pawlak, 2006; Zhang and Song, 2000).
Membrane material also affects membrane performance. For example,
Polysulfone has exceptional temperature (up to 75oC) and pH (1 to 13 S.U.) resistance,
but has been known to perform poorly in the treatment of protein-based food processing
wastewater streams. There should be balance between membrane performance with
respect to permeate flow rate and retention of the desired solute and membrane
characteristics such as tendency for fouling, cost and life span (Cheryan and Rajagopalan,
1998).
Many variables, such as composition of membrane, method of membrane
manufacturing, shape and configuration of the fluid molecules, their interactions with each
other and the membrane surface, fluid dynamics of the membrane unit, pressure,
temperature and velocity of the mixture, influence the filtration process (Del Nery et al.,
2007). Two of the main factors that control the efficiency of membrane separations are size
exclusion and chemical interaction between membrane surfaces and permeate.
An experiment was conducted to investigate the efficiency of the treatment of post-
DAF PPW using membrane filtration. Permeate flux, COD, TS and total suspension solids
(TSS) were measured and analyzed. The performance of 6 membranes (4 pore sizes and 3
membrane materials) within the microfiltration (MF) and UF ranges were evaluated at 50
psi.
61
Materials and Methods
Sample Collection and Preparation
Grab samples of approximately 40L of post-DAF PPW were collected in large
plastic containers after primary screening and pretreatment in a DAF unit at a north
Georgia broiler processing plant during normal slaughter operations and transported
immediately to the laboratory. The DAF process has been shown to remove more than
90% of fats and solids in PPW (Grant, 1980; Nielsen, 1989). Within 1 hr of collection
and in preparation for each experimental trial, 4L subsamples of PPW were poured from
the collection containers through a 500µm sieve and designated as ‘raw PPW’. Each 4L
raw PPW sample was then poured through a 106µm sieve and designated ‘pre-sieved
PPW’. Pre-sieved subsamples ranged in temperature from 18 to 34oC (64 to 93oF) and
pH ranged from 5.0 to 5.6 S.U.
Each experimental trail was conducted using one of 6 membranes (30,000MWCO
Polysulfone, 30,000MWCO PVDF, 100,000MWCO Ultrafilic, 100,000MWCO PVDF,
0.10µm Polysulfone and 0.30µm PVDF) within the microfiltration (MF) and
ultrafiltration (UF) ranges, operated at one pressure level (50 psi) with membrane
material and membrane pore size as the statistical main effects. The 6 membranes (i.e.,
treatments) were tested once per week for 3 weeks for a total of 18 trials. The experiment
was carried out using a Spintek STC bench-scale membrane filtration system as shown in
Figure 4.1.
62
Figure 4.1. SpinTek Static Test Cell (STC) Membrane Filtration System (Spintek, 2008)
This system simulates a full-scale membrane filtration system by circulating raw
wastewater from a feed tank, past a sample (i.e., coupon) of flat-sheet test membrane in a
cross-current configuration under pressure. A variable speed pump with a maximum
speed of 1207 rpm and back pressure control valve can be manipulated to maintain
constant pressure on the membrane. For this experiment the pressure remained constant
at 50 psi. The wastewater stream circulates back through the feed tank as a retentate (i.e.,
concentrate), while the treated water passes through the membrane as permeate. The flow
through the STC system during the experiment ranged from 2.3 to 2.6 L/min. Each test
membrane coupon had an effective surface area of 0.005m2.
Each trial consisted of a 4L pre-sieved subsample of PPW being placed into the
feed tank of the STC Membrane Filtration System fitted with one of the 6 membranes
(previously rinsed with deionized water) and operated for 60 min at 50 psi. The 6 trials
63
were randomly conducted each week for 3 weeks. A total of 9 wastewater samples were
collected in 1L glass jars during each trial. Representative raw PPW (500µm) and pre-
sieve (106µm) PPW samples were collected prior to each membrane filtration trial
(Samples 1 and 2). During each trial, membrane filtration permeate samples were
collected every 10 minutes (Samples 3 – 8) with the volume (mL) of effluent noted for
subsequent permeate flux calculations. Finally, a sample of recycled concentrate was
collected after 60 min (Sample 9). The pH of all samples was adjusted to < 2.0 S.U. using
H2SO4 as a preservative and stored at 4oC prior to analysis.
Analytical Methods
The post-DAF PPW samples were analyzed for COD (chemical oxygen demand
method 5220D), TS (total solids method 2540B) and TSS (total suspended solids method
2540D) concentration (mg/L) (APHA, 2005). A COD test was run on each 10 min
interval permeate sample collected and was used to determine the organic ‘strength’ of
post-DAF PPW. A TS test was also run on each 10 min interval permeate sample
collected and was used to determine the concentration (mg/L) of total solids present in
each sample. Due to the small amount of permeate sample available following COD and
TS analysis, a composite of the 10 minute interval permeate samples for each trail was
prepared and a TSS test was run on each composite sample to determine the
concentration (mg/L) of suspended solids present. TS can be defined in terms of
particulate size as the sum of TSS and total dissolved solids (TDS) as represented in the
equation: TS = TSS + TDS (APHA, 2005).
64
Statistical Analysis
Data were subjected to statistical analysis by the SAS JMP 8.0.2 program (SAS
Institute, 2009). Data from the 6 membranes (30,000MWCO Polysulfone, 30,000MWCO
PVDF, 100,000MWCO Ultrafilic, 100,000MWCO PVDF, 0.10µm Polysulfone and
0.30µm PVDF) with 3 replications were analyzed by one-way ANOVA procedures for
completely randomized design. Means were separated using Tukey-HSD procedure to
assess the significance (P<0.05) of the data (SAS Institute, 2009).
Data Analysis
Permeate Flux
Permeate flux is defined as the membrane effluent flow per unit area per unit time
and is the most common calculation used to determine membrane efficiency (Cheryan
and Rajagopalan, 1998). In this experiment, permeate flux was calculated and expressed
as liters per square meter of membrane per hour (Lm-2h-1). The permeate volume
collected during each 10 min interval of trial runs was recorded in mL. The total volume
(mL) collected for the 10 min interval was divided by 10 to produce a mL/min value.
The mL/min value was then divided by 1000 to produce a L/min value. The L/min value
was then multiplied by 60 to produce a L/hr value. Finally, the L/hr value was divided by
0.005 m2 (i.e., the surface area of each test membrane coupon) to produce the reported
permeate flux value (Lm-2h-1).
COD, TS and TSS Percentage Reduction
Percentage reduction in concentrations (mg/L) of analytical tests was calculated
by subtracting the end-point (i.e., pre-sieved or permeate) values from the corresponding
65
start-point (i.e., raw or pre-sieved) values, and dividing the difference by the start-point
value. The result was then multiply by 100%.
Membrane Efficiency
The most effective and suitable membrane for post-DAF PPW treatment was
generated using weighted linear aggregation (WLA) of 3 of the parameters tested (i.e.,
COD, permeate flux and TS). TSS results were not utilized in the WLA due to non-
significance statistically. The WLA is a weighted average in which the decision maker
assigns the weights of relative importance to each criterion (Eastman et al., 1995). The
relative weights of each criterion was computed using a pair-wise comparison matrix
(Saaty, 1977). As shown in Table 4.1, the value of each matrix cell represents the relative
importance of the row criterion against the column criterion with a range from 1 to 9. A
consistency ratio (CR) was used to determine the probability that the matrix ratings were
randomly generated, which should be less than 0.1. Each relative importance weight was
evaluated using the CR. The relative weights are computed by corresponding the
maximum eigen value with the relative values of the eigen vector.
Consistency ratio is expressed by
where λmax is the maximum eigen value of the matrix and n is the total number of criteria.
Table 4.1. Pair-wise comparison of the criteria (i.e., COD, Permeate flux, TS) for
membrane filtration of post-DAF poultry processing wastewater
COD Permeate Flux TS Weight

COD 1 3/2 2 0.46
Permeate Flux 2/3 1 3/2 0.32
TS 1/2 2/3 1 0.22
Consistency ratio (CR) = 0.00077
66
A performance index was used to indicate which of the membrane was the most
effective and suitable for the membrane filtration of post-DAF PPW.
Performance index is expressed by
where Pmi = Performance index, wcod and = weight and rank of COD, and =
weight and rank of flux, wts and = weight and rank of TS. Ranks were developed in
descending order based on their significant differences (i.e. from the best rank (4) to the
least rank (1)).
Results and Discussion
Permeate Flux
The mean permeate flux values for each trail were calculated by averaging the six
10-minute interval flux values from each STC system run for a period of one hour. The
resulting means from the 3 repetitions were then averaged to determine an overall
permeate flux mean for each membrane. The overall mean permeate flux (Lm-2h-1) values
for the 6 membranes at 50 psi operating pressure are summarized in Table 4.2 and
graphically shown in Figures 4.2 and 4.3.
Visual inspection of Figure 4.2 showed that all membranes flux values produced a
downward sweeping flux curve similar to Lo et al. (2005), with a much steeper curve
from the 0.30µm PVDF membrane versus the other membranes which remained
relatively flat throughout the run.
As shown in Table 4.2 and graphically represented in Figure 4.3, the 0.30µm
PVDF (224 Lm-2h1), 100,000MWCO PVDF (149 Lm-2h-1) and 0.1µm Polysulfone (146
Lm-2h-1) membranes were not significantly different from each other, nor were the
30,000MWCO Polysulfone (114 Lm-2h-1), 100,000MWCO Ultrafilic (100 Lm-2h-1) and
67
30,000MWCO PVDF (84 Lm-2h-1) membranes. However, the mean permeate flux of the
0.3µm PVDF membrane was significantly higher than the 30,000MWCO Polysulfone,
100,000MWCO Ultrafilic and 30,000MWCO PVDF membranes.
Table 4.2. Mean permeate flux (Lm-2h-1±SEM), COD (mg/L±SEM), TS (mg/L±SEM)

and TSS (mg/L±SEM) values for 6 membranes filtering post-dissolved air flotation
(DAF) poultry processing wastewater for 60 min
Membrane size/material Mean Permeate Mean Mean Mean

Flux Permeate Permeate Permeate
(Lm-2h-1±SEM) COD TS TSS
(P<0.0001) (mg/L±SEM) (mg/L±SEM) (mg/L±SEM)
(P<0.0001) (P<0.0001) (P=0.1984)
30,000MWCO 114b±6 388cd±8 ab
2292 ±209 21±12
Polysulfone
30,000MWCO PVDF 84b±2 397cd±13 2863a±172 29±5
100,000MWCO Ultrafilic 100b±1 373d±11 2332 ab±116 8±2
100,000MWCO PVDF 149ab±7 483bc±32 1925 bc±208 12±4
0.10µm Polysulfone 146ab±4 748a±42 2020bc±55 24±3
0.30µm PVDF 224a±59 519b±12 1498c±133 13±6
a,b,c,d
(P<0.05)
As expected, the membrane with the largest nominal gap openings (i.e., 0.30µm
PVDF) had the largest mean flux values, while the 2 membranes with the smallest
nominal gap openings (i.e., 30,000MWCO) and one of the next smallest nominal gap
opening membranes (i.e., 100,000MWCO) had the significantly lowest mean flux values.
These results indicate that membrane pore size (i.e., 0.30µm versus 100,000 MWCO and
30,000MWCO) plays a critical role in determining the permeate flow rate of membranes.
Results also showed that the membrane materials of the 100,000MWCO membranes (i.e.,
68
PVDF and Ultrafilic) and the 30,000MWCO membranes (i.e., Polysulfone and PVDF)
did not play a significant role in mean permeate flux. From the results shown in Table
4.2, it can be concluded that, for the optimum efficiency and high permeate flux, either
the 0.3µm PVDF, 100,000MWCO PVDF or 0.10µm Polysulfone membrane should be
selected.
600
30,000 MWCO
Polysulfone
500
30,000 MWCO
PVDF
100,000 MWCO
400 Ultrafilic
FLux (Lm-2h-1)
100,000 MWCO
PVDF
300
0.1µm Polysulfone
0.3µm PVDF
200
100
0
10 20 30 40 50 60
Elapsed time (minutes)
Figure 4.2. Post-DAF poultry processing wastewater permeate flux (Lm-2h-1) values at
10 min intervals for 6 membrane filters over 60 min
69
a
ab ab
b b
b
Figure 4.3. Mean permeate flux (Lm-2h-1) values for 6 membranes filtering post-DAF
poultry processing wastewater over 60 min
COD Concentration
The organic strength of the post-DAF PPW samples was measured using COD.
Mean permeate COD concentrations (mg/L) were determined and analyzed for the 6
membranes. COD concentration (mg/L) data were also used to calculate a COD removal
efficiency (%) values for the pre-sieved and membrane permeate PPW samples.
Statistical analysis showed that there were significant differences in the COD
concentration means (P<0.0001). The COD results are summarized in Table 4.2 and
graphically represented in Figures 4.4 and 4.5. The membrane producing the lowest mean
COD concentrations and highest COD removal percentages was deemed the most
effective.
70
The mean COD concentrations (mg/L) for raw PPW (i.e., sieved at 500µm) and
pre-sieved PPW samples (i.e., sieved at 106µm) for the 18 membrane filtration trials
were 635 and 624 mg/L, respectively. Thus the 106µm pre-sieving reduced the COD
concentration on average by 2%.
1000 30,000 MWCO

Polysulfone
900
Permeate COD concentration (mg/L)
30,000 MWCO
800 PVDF
700
100,000 MWCO
600 Ultrafilic
500 100,000 MWCO

PVDF
400
0.1µm
300 Polysulfone
200 0.3µm PVDF

100
0
10 20 30 40 50 60
Elapsed time (minutes)
Figure 4.4. Post-DAF poultry processing wastewater mean permeate COD concentrations
(mg/L) values at 10 min intervals for 6 membrane filters over 60 min
Visual inspection of Figure 4.4 showed that all of the membrane mean COD
values produced a slight downward sweeping curve. The downward curve is indicative
of improving COD removal over the course of the filtration run. As shown in Table 4.2
and Figure 4.5, the 0.1µm Polysulfone membrane had the highest COD permeate
concentration (748 mg/L), which was significantly higher than all the other membranes.
This established the 0.1µm Polysulfone membrane as the least efficient at COD reduction
71
and the only membrane with a negative removal efficiency (i.e.,-20%), indicating that the
Polysulfone membrane material degenerated in some way when in contact with PPW,
thus adding COD to the permeate stream. There was no significance difference between
the 0.30µm PVDF membrane (519 mg/L) and the 100,000MWCO PVDF (483 mg/L),
which represented removal efficiencies of 17% and 23%, respectively. There was also no
significant difference between the 100,000MWCO PVDF, 30,000MWCO PVDF
(397mg/L) and 30,000MWCO Polysulfone (388 mg/L) membranes, which had removal
efficiencies of 23%, 36% and 38%, respectively. Finally, the 100,000MWCO Ultrafilic
(373 mg/L) membrane produced the lowest COD mean concentrations, although not
significantly different from the 30,000MWCO PVDF or the 30,000MWCO Polysulfone
membranes. The 373 mg/L permeate COD concentration produced by the
100,000MWCO Ultrafilic membrane resulted in the highest COD removal efficiency of
40% making it the most effective COD reduction membrane.
Interestingly, the 30,000MWCO PVDF (36% COD reduction) and 30,000MWCO
Polysulfone (38% COD reduction) membranes were not significantly different, indicating
that membrane material did not influence the COD removal efficiency at the smaller gap
opening size. This result is surprising since the Polysulfone membrane material seems to
have a strong detrimental effect on the removal efficiency of the 0.1µm membrane which
produced a negative % reduction. However in the case of the 100,000MWCO
membranes, membrane material seems to play a significant role in that the Ultrafilic
membrane material (40% COD reduction) significantly outperformed the PVDF
membrane material (23% COD reduction). From the results in Table 4.2, it can be
concluded that, based on the lowest mean COD mean concentration and highest COD
72
mean removal percentages, either the 100,000MWCO Ultrafilic, 30,000MWCO PVDF or
30,000MWCO Polysulfone membranes membrane should be selected.
b
bc
cd cd
d
Figure 4.5. Mean permeate COD concentrations (mg/L) for 6 membranes filtering post-
dissolved air flotation (DAF) poultry processing wastewater for 60 min
TS Concentration
The total amount of solids contained in the PPW membrane permeate samples
was measured using TS. TS mean concentration (mg/L) data for each membrane was
determined and analyzed. These data were also used to calculate a TS removal efficiency
(%) value for pre-sieved and membrane permeate PPW samples. The membrane
producing the lowest mean TS concentrations and highest TS removal percentages was
deemed the most effective. Statistical analysis showed that there were significant
73
differences in the TS concentration permeate means (P<0.0001). The TS results are
summarized in Table 4.2 and Figure 4.6.
ab ab
bc bc
Figure 4.6. Mean permeate TS (mg/L) for 6 membranes filtering post-DAF poultry
The mean TS concentrations (mg/L) for raw PPW and pre-sieved PPW samples
106µm pre-sieve reduced the TS concentration of the post-DAF PPW on average by 6%.
It should be noted that only one membrane (i.e., 0.3µm PVDF) produced a positive
removal efficiency for TS of 7%. The 5 remaining membranes all produced mean TS
permeate concentrations which were higher than the initial trial pre-sieved PPW samples.
This would indicate that the membranes themselves, in contact with PPW, release
particulates into the permeate stream causing TS concentrations to rise.
74
As shown in Figure 4.6, the 30,000MWCO PVDF (2862 mg/L), 100,000MWCO
Ultrafilic (2332 mg/L) and 30,000MWCO Polysulfone (2292 mg/L) membranes were not
significantly different from each other, and were the least efficient membranes at
reducing TS. The 0.1µm Polysulfone (2020 mg/L), 100,000MWCO PVDF (1925 mg/L),
and 0.3µm PVDF (1498 mg/L) membranes were also not significantly different, and were
the most efficient at reducing TS in post-DAF PPW. The only significant difference was
between the top performing 0.3µm PVDF membrane and the bottom performing 0.1µm
Polysulfone, 100,000MWCO PVDF and the 30,000MWCO PVDF membranes. With a
mean permeate TS concentration of 1498 mg/L, the 0.3µm PVDF had the highest TS
concentration removal efficiency of 7%.
As with the permeate COD concentration results, the 30,000MWCO PVDF and
30,000MWCO Polysulfone membranes were not significantly different in TS reduction
in post-DAF PPW, indicating again that membrane material did not influence the TS
removal efficiency at the smallest nominal gap size. The same was true of the
100,000MWCO Ultrafilic and 100,000MWCO PVDF membranes which were not
significantly different. Like the permeate flux results, it can be concluded from the results
in Table 4.2 that, for the lowest mean TS concentrations and highest TS removal
percentages, either the 0.3µm PVDF, 100,000MWCO PVDF or 0.10µm Polysulfone
membrane should be selected.
TSS Concentration
The mean concentration of suspended particulates in each PPW sample for the 6
membranes was measured using TSS. The TSS values were determined by averaging
TSS values from each trial after a period of one hour for the 3 repetitions. The membrane
75
producing the lowest TSS concentrations and highest TSS removal percentages was
deemed the most effective. The TSS results are summarized in Table 4.2 and graphically
represented in Figure 4.7.
29
24
21
13 12
7
Figure 4.7. Mean permeate TSS (mg/L) for 6 membranes filtering post-DAF poultry
The mean TSS concentrations (mg/L) for raw PPW and pre-sieved PPW samples
for the 18 membrane filtration trials were 49 and 48 mg/L, respectively. Thus the 106µm
pre-sieve reduced the TSS concentration on average by 2%. There was no significant
difference (P=0.1984) in the performance of any of the membranes in terms of TSS
reduction. These results indicate that TSS concentration reduction would not be a viable
criterion for membrane selection.
76
Conclusions
The maximum mean permeate flux value of 224 Lm-2h-1 and lowest mean TS
value was obtained by the 0.3µm PVDF membrane (1498 mg/L, 7%). The lowest COD
concentration, and corresponding highest reduction percentage, was obtained by the
100,000MWCO Ultrafilic membrane (373 mg/L, 40%).
Using the previously described performance index (Pmi), the most effective
membrane for treatment of post-DAF PPW was determined. Based on calculated Pmi
values as shown in Table 4.3, the 100,000MWCO Ultrafilic membrane was determined
to be the most effective (Pmi = 3.56 on a 4.0 scale) in regards to maximizing permeate
flux, and COD and TS reduction. Interestingly the 100,000MWCO PVDF membrane
(Pmi = 2.22) was the least effective membrane indicating that membrane material played
a critical role in membrane performance at the 100,000MWCO level. The second most
effective membrane was the 30,000MWCO PVDF membrane (Pmi = 3.24), followed by
the other 30,000MWCO membrane made of Polysulfone (Pmi = 3.02) indicating that
membrane material does not make a significant difference at the 30,000MWCO level in
treatment of post-DAF PPW. The 0.3µm PVDF membrane (Pmi = 2.86) placed fourth in
performance index value, followed by the 0.1µm Polysulfone membrane (Pm i = 2.40) in
fifth place.
77
Table 4.3. Performance index values for 6 membranes treating post-DAF poultry
processing wastewater based on maximum permeate flux, and reduction of COD and TS
over 60 min of operation based on significant difference (P<0.05)
COD Permeate TS Performance

Flux Index (Pmi)
1 100,000 MWCO 4 4 2 3.56
Ultrafilic
2 30,000 MWCO 4 3 2 3.24
PVDF
3 30,000 MWCO 4 3 1 3.02
Polysulfone
4 0.3µm PVDF 2 4 3 2.86
6 100,000MWCO 2 3 3 2.22
PVDF
The permeate flux values (Lm-2h-1) calculated in this experiment can be used to
determine the surface area of membrane (m2) required to treat a known volume of PPW
in a full scale membrane filtration system. As an example, a typical broiler processing
plant slaughtering 250,000 birds per day (bpd) and utilizing 26 L per bird will produce
6,500,000 L of PPW. If a membrane filtration system was designed to operate 24 hours
per day, then 270,833 L of PPW would need to be treated each hour. Results of this
experiment showed the top performing membrane in terms of flux was the 0.3µm PVDF
membrane which averaged 224 Lm-2h-1. Thus, by dividing 270,833 Lh-1 by 224 Lm-2h-1 it
can be determined that 1209 m2 of membrane would be needed to filter the total PPW
stream. Use of the 100,000MWCO Ultrafilic membrane would require 2708 m2 of
membrane.
The experiment determined that pre-sieving of post-DAF PPW to 106µm prior to
membrane filtration reduced COD by 2%, TS by 6%, and TSS by 2% making pre-
78
screening of post-DAF PPW most likely unnecessary in preparation for membrane
filtration. Membrane filtration of post-DAF PPW reduced the COD concentration by as
much as 40%, and TS concentration by as much as 7%. These results indicate that
membrane filtration of post-DAF PPW will be effective at reducing the organic content,
but not the total solids content of post-DAF PPW.
79
References
Cassell E.A., Kaufman K.M., Matuevic E. (1975) The effects of bubble size on
microflotation. Water Research 9:1017-1024.
Del Nery V., de Nardi I.R., Damianovic M., Pozzi E., Amorim A.K.B., Zaiat M. (2007)
Long-term operating performance of a poultry slaughterhouse wastewater
treatment plant. Resources Conservation and Recycling 50:102-114.
Grant P.E. (1980) Treatment of fatty effluents. In: Herzka A and Booth RG (eds.). Food
Industry Wastes: Disposal and Recovery. Applied Science Publishers, UK.
Harper S.R., Valentine G.E., C. C. Ross. (1988) Feasibility of Packed-Bed Anaerobic
Treatment of Poultry Processing Wastewater. Proceedings of the 42nd Industrial
Waste Conference. Purdue University, West Lafayette, Indiana, May 12-14.
Lewis Publishers, Chelsea, Michigan.
Kiuru H.J. (2001) Development of dissolved air flotation technology from the first
generation to the newest (third) one (DAF in turbulent flow conditions), I W a
Publishing. pp. 1-7.
80
Lo Y.M., Cao D., Argin-Soysal S., Wang J., Hahm T.S. (2005) Recovery of protein from
poultry processing, wastewater using membrane ultrafiltration. Bioresource
Technology 96:687-698.
Manjula S., Subramanian R. (2006) Membrane technology in degumming, dewaxing,
deacidifying, and decolorizing edible oils. Critical Reviews in Food Science and
Nutrition 46:569-592.
Marenchino R., Pagliero, C., & Mattea, M. 200, 562–564. ((2006).) Vegetable oil
degumming using inorganic membranes. Desalination 200:562-564.
Mittal G.S. (2006) Treatment of wastewater from abattoirs before land application--a
review. Bioresource Technology 97:1119-1135.
Nakao S. (1994) Determination of pore-size and pore-size distribution. Filtration
membranes. Journal of Membrane Science 96:131-165.
Nielsen V.C. (1989) In: Mead, G.C. (Ed.), Processing of poultry. Elsevier Science
Publishers, New York, NY, pp. 361–412.
Northcutt J.K., Jones D.R. (2004) A survey of water use and common industry practices
in commercial broiler processing facilities. . Journal of Applied Poul-try Research
13:48-54.
81
USDA. (2011) Poultry Slaughter – 2009 Annual Report. U.S. Department of Agriculture.
Veerkamp C. (1999) Challenges for water management in processing. Poultry processing
worldwide 7:20.
Vidal G., Carvalho A., Méndez R., Lema J.M. (2000) Influence of the content in fats and
proteins on the anaerobic biodegradability of dairy wastewaters. Bioresource
Zak S., Pawlak Z. (2006) The use of flotation in fat recovery and the pretreatment of
wastewaters from animal fat production. Polish Journal of Environmental Studies
15:507-515.
Zhang M.M., Song L.F. (2000) Pressure-dependent permeate flux in ultra- and
microfiltration. Journal of Environmental Engineering-Asce 126:667-674.
82
CHAPTER 5
CONCLUSIONS AND FUTURE DIRECTIONS
Membrane filtration has become a popular wastewater treatment method because
of its ability to remove minute particles such as fats, protein and pathogens, and is now
cost competitive with traditional treatment methods (Lo et al., 2005). The objectives of
this study were to evaluate the effects of membrane filtration on pre- and post-DAF PPW.
To accomplish this, two membrane filtration experiments were carried out as described in
Chapters 3 and 4. The pre- and post-DAF PPW samples were pre-sieved to 106µm, to
reduce fouling of the test membranes. Membrane filtration was carried out using the
Spintek STC bench-scale system.
Conclusions
The first experiment involved membrane filtration of pre-DAF PPW with
emphasis on 3 membrane pore sizes (0.30µm PVDF, 0.10µm Polysulfone and 100,000
MWCO Ultrafilic) and 2 transmembrane pressures (50 and 80psi) as main effects. The
Spintek system was operated at a flow rate range of 2.4 to 3.0 L/min and pump speed
range of 750 to 1000 rpm. The second experiment involved membrane filtration of post-
DAF PPW with emphasis on 4 membrane pore sizes (0.3µm, 0.1µm, 100,000MWCO and
30,000MWCO) and 3 membrane materials (Polysulfone, PVDF and Ultrafilic) as main
effects. The Spintek system was operated at 50 psi, at flow rate range of 2.3 to 2.6 L/min
and pump speed 1207 rpm.
83
The concentration of organics present in the PPW was measured by using COD in
Experiments 1 and 2, while solids content was determined using TS in Experiments 1 and
2, and TSS in Experiment 2. Membrane efficiency was determined by permeate flux,
which was expressed in terms of Lm-2h-1. Both experiments were carried out using the
same general procedures. Differences between the two experiments included statistical
analysis, time of trial runs, and main effects. While the first experiment was designed as a
3x2 factorial design and thus was analyzed using a factorial analysis of variance
(ANOVA) with an interaction term, the second experiment was analyzed using a one-
way ANOVA since 4 membrane sizes and 3 membrane materials were utilized in a non-
factorial arrangement. Trials in the first experiment were conducted for 2 hr, however
results from the first experiment showed that permeate flux values generally stabilized
after one hour of filtration. Thus, trials in the second experiment were conducted for 1 hr
which allowed for the doubling in the number of trials conducted in a single day. Finally,
the first experiment focused on the main effects of membrane pore size and
transmembrane pressure. Analysis for results in experiment 1 created the question of the
impact of membrane material on treatment of PPW. Thus, the second experiment
focused on the main effects of membrane pore size and membrane material.
The most effective and suitable membrane for both pre-DAF and post-DAF PPW
membrane filtration based on the 3 parameters of COD, permeate flux and TS was
generated using weighted linear aggregation (WLA) as described by Eastman et al.
(1995) and Saaty (1977). A performance index was generated to indicate which of the
membranes was the most efficient for the membrane filtration of pre- and post-DAF PPW
as shown in Tables 5.1 (Experiment 1) and 5.2 (Experiment 2). From these results,
84
membrane filtration on pre-DAF PPW using 0.30µm PVDF membrane was the most
suitable in terms of permeate flux, COD, and TS reduction and its performance was
comparable to the existing DAF pre-treatment, whereas in post-DAF PPW pre-treatment,
the 100,000MWCO Ultrafilic membrane was determined to be the most effective.
Table 5.1. Best performing membrane for treatment of pre-DAF poultry processing
wastewater based on maximum permeate flux, and reduction of COD and TS over 60 min
of operation based on significant difference (P<0.05)
Flux Index (Pmi)
1 0.3µm PVDF 2 3 3 2.54
2 100,000 MWCO 2 3 2 2.32

Ultrafilic
Table 5.2. Performance index values for 6 membranes treating post-DAF poultry
processing wastewater based on maximum permeate flux, and reduction of COD and TS
over 60 min of operation based on significant difference (P<0.05)
Flux Index (Pmi)
1 100,000 MWCO 4 4 2 3.56
Ultrafilic
2 30,000 MWCO 4 3 2 3.24
PVDF
3 30,000 MWCO 4 3 1 3.02
Polysulfone
4 0.3µm PVDF 2 4 3 2.86
6 100,000MWCO 2 3 3 2.22
PVDF
85
Permeate flux (Lm-2h-1) was used to determine effluent flow efficacy. Logically,
as the concentration of a PPW sample decreases, permeate flux increases. This can be
seen in the results of the 2 experiments. When the 0.3µm PVDF membrane was utilized
to treat the higher concentration pre-DAF PPW, an average permeate flux of 115 Lm -2h-1
was achieved. Conversely, when the same 0.3µm PVDF membrane was used to treat the
less concentrated post-DAF PPW, an average permeate flux of 224 Lm-2h-1 was achieved,
which represented a 95% increase in the effluent flow rate.
Permeate flux results from the two experiments can be used to determine the
square meters (m2) of membrane that would be required by a full-scale PPW treatment
systems at the poultry processing plant, in that the mean permeate flux plays an essential
role in the calculation of the required membrane surface area. For example, using a
typical poultry processing plant that slaughters 250,000 birds per day (bpd) and uses the
average 26 L of portable water per bird (Lpb), it can be calculated that such a plant would
generate 6.5 million L of PPW per day. The assumption is also made that an automated
membrane filtration system operating 24 hr a day is utilized. In the Conclusion section of
Experiment 1 (Chapter 3), it was calculated that 2355 m2 of the top flux producing 0.3µm
PVDF membrane would be required to treat 6.5 million L of pre-DAF PPW. In
comparison, the Conclusion section of Experiment 2 (Chapter 4) reports that only 1209
m2 of 0.3µm PVDF membrane would be required to treat 6.5 million L of post-DAF
PPW, effectively reducing the required surface area of membrane by 49%. From the
above deductions, a post-DAF PPW on-site treatment system less would require
approximately half the membrane surface area required for treatment of a pre-DAF PPW
stream of the same volume. Thus from an initial capital investment standpoint,
86
membrane filtration of post-DAF PPW would be preferable over pre-DAF PPW.
However, because chemical addition plays a critical role in DAF treatment, the operating
cost savings of reducing chemical demand in DAF systems by employing pre-DAF PPW
membrane filtration could play a substantial economic role.
Future directions
Future research in membrane filtration of PPW should consider exploring the use
of membrane (materials and pore sizes) combinations in treating PPW. Further analytics
should be done to look at membrane filtration effectiveness on other critical wastewater
parameters such as TKN and other nutrients. In addition, future experiments should
explore the potential for maximizing pre-screening (e.g., tertiary microscreening) prior to
membrane filtration of PPW. The ability of membrane filtration to reduce pathogens
(e.g., bacteria) should be further studied. Exploring the different membrane cleaning
processes in membrane filtration that works best on pre- and post-DAF PPW will help in
better filtration. The life cycle assessment (LCA) of pre- and post-poultry processing
wastewater membrane filtration unit in an operating plant should be addressed.
87
REFERENCES
Aider M., de Halleux D., Bazinet L. (2008) Potential of continuous electrophoresis
without and with porous membranes (CEPM) in the bio-food industry: review.
Trends in Food Science & Technology 19:351-362.
Allie Z., Jacobs E.P., Maartens A., Swart P. (2003) Enzymatic cleaning of ultrafiltration
membranes fouled by abattoir effluent. Journal of Membrane Science 218:107-
116.
Almas K.A. (1985) Applications of crossflow membrane technology in the fishing
industry. Desalination 53:167-180.
Andreotolla G., Foladori P., Ragazzi M., Tatano F. (2000) Experimental Comparison
between MBBR and Activated Sludge System for the Treatment of municipal
Wastewater. Water Science and Technology 41:375-382.
Andreottola G., Foladori P., Ragazzi M. (2001) On-line control of a SBR system for
nitrogen removal from industrial wastewater. Wat. Sci. Tech. 43:93-100.
Argüello M.A., Álvarez S., Riera F.A., Álvarez R. (2003) Enzymatic cleaning of
inorganic ultrafiltration membranes used for whey protein fractionation. Journal
of Membrane Science 216:121-134.
Arundel J. (1995) Sewage and industrial effluent treatment: a practical guide. Blackwell
Science. Oxford, England.
88
AWWA. (1977) Wastewater Treatment Plant Design: A Manual of Practice. American
Water Works Association. Lancaster Press Inc., Lancaster, Pennsylvannia.
Barik S., Forgacs T., Isbister J. (1991) Bioconversion of chicken wastes to value-added
products. Bioresource Technology 36:229-234.
Barrett F. (1977) Farm effluent; electrical disposal methods. Effluent and water treatment
Journal 11:207-209.
Bazinet L., Lamarche F., Ippersiel D. (1998) Bipolar membrane electrodialysis:
Applications of electrodialysis in the food industry. Trends in Food Science and
Technology:107–113.
Bernal R., van Gottberg A., Mack B.P. ( 2002) Using Membrane Bioreactors for
Wastewater Treatment for Small Communities. Proceedings, WEFTEC02 on CD-
ROM. Session 7. Water Environment Federation Annual Technical Exhibition
and Conference, Chicago, IL, September -28-October 2.
Bernet N., Delgenes N., Akunna J.C., Delgenes J.P., Moletta R. (2000) Combined
anaerobic-aerobic SBR for the treatment of piggery wastewater. Water Research
34:611-619.
Bough W.A., Shewfelt A.L., Salter W.L. (1975) Use of chitosan for the reduction and
recovery of solids in poultry processing waste effluents. Poult Sci 54:992–1000.
Camp W.J., Willoughby E. (1968) Extended aeration purifies effluent. Food Eng. 40:72–
74.
89
Campbell M.J., Walter R.P., McLoughlin R., Knowles C.J. (1993) Effect of temperature
on protein conformation and activity during ultrafiltration. Journal of Membrane
Science 78:35-43.
Carawan R.E., Crosswhite W.M., Macon J.A., Hawkins. B.K. (1974) Waste and Waste
Management in Poultry Processing. . Publ. EPA-660/2-74-031. US Environ. Prot.
Agency, Washington, DC.
Cassell E.A., Kaufman K.M., Matuevic E. (1975) The effects of bubble size on
microflotation. Water Research 9:1017-1024.
Chakrabarty B., Ghoshal A.K., Purkait M.K. (2010) Cross-flow ultrafiltration of stable
oil-in-water emulsion using polysulfone membranes. Chemical Engineering
Journal 165:447-456.
Chávez P.C., Castillo L R., Dendooven L., Escamilla-Silva E.M. (2005) Poultry slaughter
wastewater treatment with an up-flow anaerobic sludge blanket (UASB) reactor.
Bioresource Technology 96:1730-1736.
Chen T.C., Hioll J.E., Hayes. R.L. (1976.) Quality characteristics of raw and treated
effluents from Mississippi poultry processing plants. . Poult. Sci. 55:2390–2395.
Chen X., Chen G., Yue P.L. (2002) Novel electrode system for electroflotation
wastewater. Environ. Sci. Technol. 36, pp. 778–783.
Cheryan M. (1998) Ultrafiltration and microfiltration handbook. Technomic Publishing,
Lancaster, PA.
90
Chung T.-S., Qin J.-J., Gu J. (2000) Effect of shear rate within the spinneret on
morphology, separation performance and mechanical properties of ultrafiltration
polyethersulfone hollow fiber membranes. Chemical Engineering Science
55:1077-1091.
Contreras E.M., Giannuzzi L., Zaritzky N.E. (2000) Growth kinetics of the filamentous
microorganism Sphaerotilus natans in a model system of a food industry
wastewater. Water Research 34:4455-4463.
CSUS. (1993) Operation of Wastewater Treatment Plants. Volume 2. Fourth Edition.
Office of Water Programs. Department of Civil Engineering. California State
University, Sacramento. Hornet Foundation, Inc., Sacramento, California.
Cuperus F.P., Nijhuis H.H. (1993) Applications of membrane technology to food
processing. Trends in Food Science & Technology 4:277-282.
De Morais Coutinho C., Chiu M.C., Basso R.C., Ribeiro A.P.B., Gonçalves L.A.G.,
Viotto L.A. (2008) State of art of the application of membrane technology to
vegetable oils: A review. Food Research International 42:536-550.
Del Nery V., Damianovic M.H.Z., Barros F.G. (2001) The use of upflow anaerobic
sludge blanket reactors in the treatment of poultry slaughterhouse wastewater.
Water Sci. Technol. 44:83–88.
Del Nery V., de Nardi I.R., Damianovic M., Pozzi E., Amorim A.K.B., Zaiat M. (2007)
Long-term operating performance of a poultry slaughterhouse wastewater
treatment plant. Resources Conservation and Recycling 50:102-114.
Del Pozo R., Diez V., Beltrán S. (2000) Anaerobic pre-treatment of slaughterhouse
wastewater using fixed-film reactors. Bioresource Technology 71:143-149.
91
Durchschlag H., Zipper P., Purr G., Jaenicke R. (1996) Comparative studies of structural
properties and conformational changes of proteins by analytical
ultracentrifugation and other techniques. Colloid & Polymer Science
274:117-137.
Dyrset N., Selmer-Olsen E., Havrevoll O., Ratnaweera H., Storro I., Birkeland S.E.
(1998) Feed supplement recovered from dairy wastewater by biological and
chemical pretreatment. Journal of Chemical Technology and Biotechnology
73:175-182.
Ebrahim S. (1994) Cleaning and regeneration of membranes in desalination and
wastewater applications: State-of-the-art. Desalination 96:225-238.
El Boushy A.R.Y., van der Poel A.F.B. (1994) Poultry feed from waste. . Chapman and
Hall, London.
Eremektar G., Ubay Çokgör E., Övez S., Germirli Babuna F., Orhon D. (1999) Biological
treatability of poultry processing plant effluent -- A case study. Water Science and
Fairbank W.C., Brambell B.L., Bulletin. (1968) Dairy Manure Liquid–Solid Separation.
Bulletin No. AXT-271. University of California Agricultural Extension Service,
Sacramento, CA.
Fonkwe L.G., Singh R.K., Lee J.H. (2001) Utilization of poultry processing wastes.
Journal of Food Science Nutrition, 6, 257−262.
92
Fresenius W.K.E., Quentin W.W., Scheneider W. (1988) Water analyses. A practical
guide to physicochemical and microbiological water examination and quality
assurance. Spinger-Verlag, Berlin.
Glide L.C. (1968) Cannery and poultry waste treatment. . Water Sewage Works 15:338–
342.
Grant P.E. (1980) Treatment of fatty effluents. In: Herzka A and Booth RG (eds.). Food
Industry Wastes: Disposal and Recovery. Applied Science Publishers, UK.
Grant R.A. (1976) Protein recovery from meat, poultry and fish processing plants. In:
Birch, G.G., Parker, K.J., Worgan, J.T. (Eds.) Food from waste. Applied Science,
London.
Harper S.R., Valentine G.E., C. C. Ross. (1988) Feasibility of Packed-Bed Anaerobic
Treatment of Poultry Processing Wastewater. Proceedings of the 42nd Industrial
Waste Conference. Purdue University, West Lafayette, Indiana, May 12-14.
Lewis Publishers, Chelsea, Michigan.
Heinemann P., Howell J.A., Bryan R.A. (1988) Microfiltration of protein solutions: effect
of fouling on rejection. Desalination 68:243-250.
Hyun S.H.a.K., G.T. (1997) Synthesis of ceramic microfiltration membrane for oil/water
separation. Sepax. Sci. Technology 32:2927-2943.
Janosz Rajczyk M. (1993) Fermentation of food industry wastewater. Water Research
27:1257-1262.
Karpati A., Szabo L. (1984) Suitable pretreatment of sewage resulting in pollution drop
in meat processing. In: J. Hollo, Editor, Food Industries and the Environment Int.
Symp.Budapest, Hungary, , Elsevier, Amsterdam 367–376.
93
Kiepper B.H. (2009) Effects of tertiary microsieving on the composition of poultry
processing wastewater. Journal of Applied Poultry Research 18:716-724
Kiepper B.H. ( 2003) Characterization of poultry processing operations, wastewater
generation, and wastewater treatment using mail survey and nutrient discharge
monitoring methods. . M.S. Thesis. University of Georgia, Athens, GA.
Kiepper B.H., Merka W.C., Fletcher D.L. (2008) Effects of vibratory microscreening on
proximate composition and recovery of poultry processing wastewater particulate
matter. Bioresource Technology 99:8593-8597.
Kiepper B.H., Merka W.C., Reynold A.E., Sellers J. (2001) Profile and Production
Process Determination of Phosphorus and Nitrogen Discharges from Poultry
Processing Plants. USPOULTRY Final Report: Project No. 558. U.S. Poultry and
Egg Association, Tucker, Georgia.
Kim B.S., Chang H.N. (1991) Effects of periodic backflushing on ultrafiltration
performance.
Kiuru H.J. (2001) Development of dissolved air flotation technology from the first
generation to the newest (third) one (DAF in turbulent flow conditions), I W a
Publishing. pp. 1-7.
Kobayashi T., Kobayashi T., Hosaka Y., Fujii N. (2003) Ultrasound-enhanced
membrane-cleaning processes applied water treatments: influence of sonic
frequency on filtration treatments. Ultrasonics 41:185-190.
Kobya M., Can O.T., Bayramoglu M. (2003) Treatment of textile wastewaters by
electrocoagulation using iron and aluminum electrodes. Journal of Hazardous
Materials 100:163-178.
94
Kobya M., Senturk E., Bayramoglu M. (2006) Treatment of poultry slaughterhouse
wastewaters by electrocoagulation. Journal of Hazardous Materials 133:172-176.
Kulozik U. (1994) Water rinsing in reverse osmosis and ultrafiltration, in: Fouling and
cleaning in pressure driven membrane processes, FIL-IDF, Supplement to B-Doc
250:190–209.
Langevin M.-E., Bazinet L. (2011) Ion-exchange membrane fouling by peptides: A
phenomenon governed by electrostatic interactions. Journal of Membrane Science
369:359-366.
le Roux L.D., Belyea R.L. (1999) Effects of ultrafiltration membrane concentration and
drying temperature on nutritional value of biosolids from a milk processing plant.
Bioresource Technology 70:17-21.
Levesley J.A., Hoare M. (1999) The effect of high frequency backflushing on the
microfiltration of yeast homogenate suspensions for the recovery of soluble
proteins. Journal of Membrane Science 158:29-39.
Li J., Sanderson R.D., Jacobs E.P. (2002) Ultrasonic cleaning of nylon microfiltration
membranes fouled by Kraft paper mill effluent. Journal of Membrane Science
205:247-257.
Lillard H.S. (1978) Improving quality of bird chiller water for recycling by diatomaceous
earth filtration and chlorination. Journal of Food Science 43:1528-1531.
Lin S.H., Peng C.F. (1994) Treatment of textile wastewater by electrochemical method.
Water Research 28:277-282.
95
Lo Y.-M., Yang S.-T., Min D.B. (1996) Kinetic and feasibility studies of ultrafiltration of
viscous xanthan gum fermentation broth. Journal of Membrane Science 117:237-
249.
Lo Y.M., Cao D., Argin-Soysal S., Wang J., Hahm T.S. (2005) Recovery of protein from
poultry processing, wastewater using membrane ultrafiltration. Bioresource
Luque S., Gómez D., Álvarez J.R. (2008) Industrial Applications of Porous Ceramic
Membranes (Pressure[hyphen (true graphic)]Driven Processes), in: M. Reyes and
M. Miguel (Eds.), Membrane Science and Technology, Elsevier. pp. 177-216.
Maartens A., Jacobs E.P., Swart P. (2002) UF of pulp and paper effluent: membrane
fouling-prevention and cleaning. Journal of Membrane Science 209:81-92.
Manjula S., Subramanian R. (2006) Membrane technology in degumming,
dewaxing, deacidifying, and decolorizing edible oils. Critical Reviews in Food
Science and Nutrition 46:569-592.
Manjula S., Subramanian R. (2006) Membrane technology in degumming, dewaxing,
deacidifying, and decolorizing edible oils. Critical Reviews in Food Science and
Nutrition 46:569-592.
Marenchino R., Pagliero, C., & Mattea, M. 200, 562–564. ((2006).) Vegetable oil
degumming using inorganic membranes. Desalination 200:562-564.
Marshall A.D., Munro P.A., Tragardh G. (1997) Influence of permeate flux on fouling
during the microfiltration of beta-lactoglobulin solutions under cross-flow
conditions. Journal of Membrane Science 130:23-30.
96
Martínez F., Martín A., Prádanos P., Calvo J.I., Palacio L., Hernández A. (2000) Protein
Adsorption and Deposition onto Microfiltration Membranes: The Role of Solute-
Solid Interactions. Journal of Colloid and Interface Science 221:254-261.
Mead G.C. (1989) Processing of Poultry. Elsevier Applied Science, London, England.
Merka W.C. (1989) Characteristics of wastewater. Broiler Ind. 52:20–27.
Merka W.C. (2001) Processing water and wastewater. Poultry Meat Processing. A. R.
Sams, ed. CRC Press, Boca Raton, FL:301–310.
Metcalf and Eddy Inc. (2003) Wastewater Engineering. Fourth Edition. McGraw-Hill,
New York, New York : 984-1026.
Mittal G.S. (2006) Treatment of wastewater from abattoirs before land application--a
review. Bioresource Technology 97:1119-1135.
Morgan J.M., Juang D., Sung S. (1988) Anaerobic filter-sequencing batch reactor
treatment of chicken processing wastewater. Proceedings 1988 food processing
waste conference, Atlanta, GA (1988).
Mulder M.H.V. (1991) Basic principles in membrane technology. Dordrecht: Kluwer
Academic Publishers.
Mulder M.H.V. (1995) Polarization phenomena and membrane fouling. In R. D. Noble &
S. A. Stern (Eds.), Membrane separations technology: Principles and applications
Amsterdan: Elsevier.:45–84.
Muthukumaran S., Kentish S.E., Stevens G.W., Ashokkumar M., Mawson R. (2007) The
application of ultrasound to dairy ultrafiltration: The influence of operating
conditions. Journal of Food Engineering 81:364-373.
97
Nagarale R.K., Gohil G.S., Shahi V.K. (2006) Recent developments on ion-exchange
membranes and electro-membrane processes. Advances in Colloid and Interface
Science 119:97-130.
Nakao S. (1994) Determination of pore-size and pore-size distribution .39 filtration
membranes. Journal of Membrane Science 96:131-165.
Nakatsuka S., Nakate I., Miyano T. (1996) Drinking water treatment by using
ultrafiltration hollow fiber membranes. Desalination 106:55-61.
Nemerov N.L., Dasgupta A. (1991) Industrial and Hazardous Waste Treatment. Van
Nostrand. Reinhold, New York, New York.
Nemerow N.L. (1969) Baffled biological basins for treating poultry plant wastes. J.
Water Pollut. Control Fed. 41:1602–1608.
Nguyen V.T., Shieh W.K. (2000) Secondary Treatment. In: Wastewater Treatment. D. H.
F. Liu and B. G. Liptak (Editors). Lewis Publishers, New York, New York. :210-
232.
Nielsen V.C. (1989) In: Mead, G.C. (Ed.), Processing of poultry. Elsevier Science
Publishers, New York, NY, pp. 361–412.
Northcutt J.K., Jones D.R. (2004) A survey of water use and common industry practices
in commercial broiler processing facilities. . Journal of Applied Poul-try Research
13:48-54.
Ockerman H.W., Hansen C.L. (2000) Animal ByProduct Processing and Utilization.
Technomic Publishing Company, Lancaster, Pennsylvania.
98
Ostergaard B. (1989) Applications of membrane processing in the dairy industry. In D.
MacCarthy (Ed.), Concentration and drying of foods Oxford: Elsevier Applied
Science Publishers:133-145.
Pankratz T.M. (1995) Screening Equipment Handbook. 2nd ed. Technomic Pub. Co,
Landcaster, PA.
Park E., Enander R., Barnett S.M., Lee C. (2001) Pollution prevention and biochemical
oxygen demand reduction in a squid processing facility. Journal of Cleaner
Production 9:341-349.
Pastorelli G., Andreottola G., Canziani R., Darriulat C., Frangipane E.d.F., Rozzi A.
(1997) Organic carbon and nitrogen removal in moving-bed biofilm reactors.
Water Science and Technology 35 91-99.
Peng H., Tremblay A.Y. (2008) Membrane regeneration and filtration modeling in
treating oily wastewaters. Journal of Membrane Science 324:59-66.
Perry J., Banker D., Green. R. (1999) Poultry Production in the United States. Agriculture
Information Bulletin No. (AIB748). Economic Research Service, U.S Department
of Agriculture, Washington D.C.
Petrus H.B., Li H., Chen V., Norazman N. (2008) Enzymatic cleaning of ultrafiltration
membranes fouled by protein mixture solutions. Journal of Membrane Science
325:783-792.
Porter M.C. (1990) Handbook of industrial membrane technology. . New Jersey: Noyes
Publications.
Roesink H.D.W., Beerlage M.A.M., Potman W., Vandenboomgaard T., Mulder M.H.V.,
Smolders C.A. (1991) Characterization of new membrane materials by means of
99
fouling experiments - adsorption of bsa on polyetherimide polyvinylpyrrolidone
membranes. Colloids and Surfaces 55:231-243.
Rusten B., Siljudalen J.G., Wien A., Eidem D. (1998) Biological pretreatment of poultry
processing wastewater. Water Science and Technology 38:19-28.
Shih J.C.H., Kozink M.B. (1980) Ultrafiltration treatment of poultry-processing waste-
water and recovery of a nutritional byproduct. Poultry Science 59:247-252.
Singh R. (2004) Vibratory separators still make the grade for screening dry bulk powders.
Filtration & Separation 41:20-21.
Singh S., Khulbe K.C., Matsuura T., Ramamurthy P. (1998) Membrane characterization
by solute transport and atomic force microscopy. Journal of Membrane Science
142:111-127.
Singh S.P., Wesley R.L., Budd. E.A. (1973) Characteristics of poultry processing
effluents. Poult. Sci. 52:1478-1481.
Smith P.J., Vigneswaran S., Ngo H.H., Ben-Aim R., Nguyen H. (2006) A new approach
to backwash initiation in membrane systems. Journal of Membrane Science
278:381-389.
Strathmann H. (1979) Structure and function of synthetic membranes. Chemiker-Zeitung
103:211-219.
100
Takizawa S., Fujita K., Soo K.H. (1996) Membrane fouling decrease by microfiltration
with ozone scrubbing. Desalination 106:423-426.
Tanaka Y. (2006) Irreversible thermodynamics and overall mass transport in ion-
exchange membrane electrodialysis. Journal of Membrane Science 281:517-531.
Teletzke G.H. (1961) Chicken for BBQ, wastes for aerobic digestion. Wastes Eng.
32:135–139.
Thornton G., O’Keefe T. ( 2002) Poultry processing: Washing troubles away. Poult. USA
3(8):40, 42, 44.
Torrens S.M. (2001) Bowling Balls or Molecules: The roles separation technologies play.
Pollution Engineering 33(6):10-11.
Travers S.M., Lovett D.A. (1985) Pressure flotation of abattoir wastewaters using carbon
dioxide. Water Research 19:1479-1482.
USDA. (2010) Poultry Slaughter. 2009 Annual Report. U.S. Department of Agriculture."
USDA. (2011) Poultry Slaughter – 2009 Annual Report. U.S. Department of Agriculture.
USEPA. (1974) US EPA, Development document for effluent limitations: Guidelines and
new source standards for basic fertilizer chemicals,. EPA-440/1-001/A (1974)
Washington, DC.
USEPA. (1975) Development Document for Proposed Effluent Limitations: Guidelines
and New Source Performance Standards for the Poultry Segment of the Meat
Product and Rendering Process Point Source Category. Publ. EPA 440/1-75-
1031b. US Environ. Prot. Agency, Washington, DC.
101
USEPA. (2002) Development document for the proposed effluent limitations guidelines
and standards for the meat and poultry products industry point source category.
(40 CFR 432). EPA-821-B-01-007, Washington, DC, USA.
USEPA. (2006). Life Cycle Assessment: Inventory Guidelines and Principles,
EPA/600/R-92/245, Washington, DC, USA.
Vedavyasan, C. V. (2007). Pretreatment trends—an overview. Desalination, 203,
296−299.
Veerkamp C. (1999) Challenges for water management in processing. Poultry processing
worldwide 7:20.
Vidal G., Carvalho A., Méndez R., Lema J.M. (2000) Influence of the content in fats and
proteins on the anaerobic biodegradability of dairy wastewaters. Bioresource
Viitasaari M., Jokela P., Heinänen J. (1995) Dissolved air flotation in the treatment of
industrial wastewaters with a special emphasis on forest and foodstuff industries.
Water Science and Technology 31:299-313.
Wagner J. (2001) Practical tips and hints, in: Membrane Filtration Handbook, Chemical
Engineering. 2nd edn. Revision 2, Osmonics, Minnetonka, MN:7–12.
Wang L.K., Fahey E.M., Wu Z. (2005) Dissolved Air Flotation, in: L. K. Wang, et al.
(Eds.), Physicochemical Treatment Processes, Humana Press. pp. 431-500.
WEF. (1998) Design of Municipal Wastewater Treatment Plants. . Volume 2 & 3. 4th
Edition. Manual of Practice No. 8. Water Environment Federation, Alexandria,
VA.
102
Welch E.B., Lindell T. (1992) Ecological Effects of Wastewater: Applied Limnology and
Pollutant Effects. Second Edition. Chapman & Hall, New York, New York.
Wesley R.L. (1985) Water re-use and conservation in poultry processing. Science (64),
476–478.
Wessels J.P.H. (1972) A study of the protein quality of different feather meals. Poultry
Sci. 51:537.
Whitehead W.K. (1979) Activated-sludge process for pretreating poultry-processing
wastewater. Transactions of the Asae 22:211-214.
Whittemore C.T. (1994) Food from animals: Environmental issues and implications. In:
J.M. Dalzell, Editor, Food industry and the environment, Blackie Academic
Professional, London 1–14.
Woodard F.E., Sproul O.J., Hall M.W. (1972) Treatment of poultry processing wastes
using DAF and chemical coagulants. Journal of water pollution control federation
44:1909-1915.
Woodard F.E., Hall M.W., Sproul O.J., Ghosh M.M. (1977) New concepts in treatment
of poultry-processing wastes. Water Research 11:873-877.
Woolard C.R., Irvine R.L. (1995) Treatment of hypersaline wastewater in the sequencing
batch reactor. Water Research 29:1159-1168.
WPCF. (1969) Sludge De-watering. WPCF Manual of Practice No. 20. Water Polution
Control Federation Arlington, Virginia.
Xu L.J., Sheldon B.W., Larick D.K., Carawan R.E. (2002) Recovery and utilization of
useful byproducts from egg processing wastewater by electrocoagulation. Poultry
Science 81:785-792.
103
Yanagi C., Mori K. (1980) Advanced reverse osmosis process with automatic sponge ball
cleaning for the reclamation of municipal sewage. Desalination 32:391-398.
Yushina Y., Hasegawa J. (1994) Process performance comparison of membrane
introduced anaerobic-digestion using food-industry waste-water, Elsevier Science
Bv. pp. 413-421.
Zak S., Pawlak Z. (2006) The use of flotation in fat recovery and the pretreatment of
wastewaters from animal fat production. Polish Journal of Environmental Studies
15:507-515
Zeman L.J., Zydney A.L. (1996) Microfiltration and Ultrafiltration: Principles and
Applications. New York: Marcel Dekker, Inc.
Zhang S.Q., Kutowy, O., Jumar, A., Malcolm, I. ( 1997) A laboratory study of poultry
abattoir wastewater treatment by membrane technology. Can. Agric. Eng. 39:99-
105.
104

Abboah-Afari Ebenezer 201108 Ms

Uploaded by

Copyright:

Available Formats

Abboah-Afari Ebenezer 201108 Ms

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Abboah-Afari Ebenezer 201108 Ms

Uploaded by

Copyright:

Available Formats

ALTERNATIVE PHYSICAL TREATMENT METHOD FOR POULTRY

PROCESSING WASTEWATER USING MEMBRANE FILTRATION

(Under the Direction of Brian H. Kiepper)

Regulatory environmental permits governing wastewater disposal have become

or pre-treat their wastewater stream prior to discharge to a municipal sewer systems.

Therefore, more effective means of on-site wastewater treatment is a constant demand.

INDEX WORDS: Poultry processing, wastewater treatment, membrane filtration,

permeate flux, COD, TS, TSS

PROCESSING WASTEWATER USING MEMBRANE FILTRATION

B.S., Kwame Nkrumah University of Science and Technology, Ghana, 2007

A Thesis Submitted to the Graduate Faculty of The University of Georgia in Partial

Fulfillment of the Requirements for the Degree

All Rights Reserved

PROCESSING WASTEWATER USING MEMBRANE FILTRATION

Major Professor: Brian H. Kiepper

Electronic Version Approved:

faithful, Glory be to His name.

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES .............................................................................................................x

2 LITERATURE REVIEW ..................................................................................4

Poultry Processing Wastewater (PPW) ........................................................5

Characteristics of Poultry Processing Wastewater ......................................6

Poultry Processing Wastewater Treatment ................................................12

Thesis Goal and Objectives .......................................................................29

3 EFFECTS OF MEMBRANE FILTRATION ON PRE-DAF (DISSOLVED

AIR FLOTATION) POULTRY PROCESSING WASTEWATER ................31

Materials and Methods ...............................................................................35

Results and Discussion ..............................................................................39

AIR FLOTATION) POULTRY PROCESSING WASTEWATER ................57

Materials and Methods ...............................................................................62

Results and Discussion ..............................................................................67

5 CONCLUSIONS AND FUTURE DIRECTIONS...........................................83

Future Directions .......................................................................................87

nutrients in poultry processing wastewater (PPW) ................................................11

Table 2.2: Wastewater Screen Classifications by Open Space Sizes ................................14

membrane filtration of pre-DAF poultry processing wastewater ..........................39

values for 6 membranes filtering pre-dissolved air flotation (DAF) poultry

processing wastewater for 120 min........................................................................40

poultry processing wastewater for 120 min ...........................................................40

120 min of operation based on significant difference (P<0.05)............................53

membrane filtration of post-DAF poultry processing wastewater .........................66

values for 6 membranes filtering post-dissolved air flotation (DAF) poultry

processing wastewater for 60 min..........................................................................68

60 min of operation based on significant difference (P<0.05)..............................78

60 min of operation based on significant difference (P<0.05)..............................85

60 min of operation based on significant difference (P<0.05)..............................85

minute intervals for 3 membrane filters at 50 psi operating pressure ....................41

minute intervals for 3 membrane filters at 80 psi operating pressure ....................42

for 3 membranes. ...................................................................................................43

at 50 and 80 psi ......................................................................................................44

(mg/L) for 3 membranes ........................................................................................46

(mg/L) at 50 and 80 psi ..........................................................................................46

Figure 3.8: Pre-DAF poultry processing wastewater mean permeate TS concentrations

(mg/L) for 3 membranes ........................................................................................48

Figure 3.9: Pre-DAF poultry processing wastewater mean permeate TS concentrations

(mg/L) at 50 and 80 psi ..........................................................................................50

Figure 3.10: Pre-DAF poultry processing wastewater mean permeate TS concentrations

(mg/L) for 100,000MWCO Ultrafilic membrane at 50 and 80 psi ........................50

(mg/L) for 0.10µm Polysulfone membrane at 50 and 80 psi .................................51