U.S. ENVIRONMENTAL PROTECTION AGENCY
EFFLUENT VARIABILITY IN
THE MEAT-PACKING AND POULTRY
PROCESSING INDUSTRIES
by
James F. Scaief
PNERL Worki rig Paper
Number 16
PACIFIC NORTHWEST ENVIRONMENTAL RESEARCH LABORATORY
An Associate Laboratory of
National Environmental Research Center—Corvallis
-------
EFFLUENT VARIABILITY IN
THE MEAT-PACKING AMD POULTRY
PROCESSING INDUSTRIES
by
James F. Scaief
PNERL Working Paper
Number 16
INDUSTRIAL WASTES BRANCH
Pacific Northwest Environmental Research Laboratory
National Environmental Research Center-Corvallis
U.S. Environmental Protection Agency
Corvallis, Oregon 97330
June 1975
-------
EFFLUENT VARIABILITY IN THE MEAT-PACKING
AND POULTRY PROCESSING INDUSTRIES
by
James F. Scaief*
INTRODUCTION
Increased industrialization has resulted in higher waste loadings
on receiving bodies of water and has caused regulatory agencies to
require increasingly more stringent effluent standards. These standards,
in addition to five-day biochemical oxygen demand (BOD^) and total
suspended solids (TSS), include parameters such as ammonia-nitrogen
(NH-j-N), oil and grease (O&G), fecal coliform, and pH.
Tables 1 and 2 summarize the guidelines established for the red
meat industry and recommended for the poultry processing industry.
For many meat-packing and poultry processing plants to meet these
new guidelines, much needs to be accomplished in controlling their
effluent variability. As stated by Ford (1), an inherent variability is
attributed to the treatment system, the characteristics of the raw waste
load, and geographical and climatological conditions. Therefore, to
minimize the variability one needs to take a look at each factor contrib-
uting to this variability and determine what control measures are
possible. Analyses of in-plant processes will be necessary and in some
cases newer processes may have to be introduced. Provisions might be
made to minimize storm water runoff entering the system. The existing
* EPA, Pacific Northwest Environmental Research Laboratory; Corvallis,
Oregon.
-------
Table 1. RED MEAT EFFLUENT STANDARDS
Federal Register, Vol. 39, No. 41, 2/28/74
Maximum average of daily values for a 30 consecutive day period
(kg/kkg LWK)
ro
Plant
Subcategory
Slaughterhouse:
I. Simple
II. Complex
7/1/77
Best Practical Treatment
BOD,
.12
.21
TSS
.20
.25
0&G
.06
.08
New Sources
nh3-n
.17
.24
7/1/83
Best Available Treatment
BODr
.03
.04
TSS
.05
.07
Packinghouse:
III. Low-Processing .17
IV. High-Processing.24
.24
.31
.08
.13
.24
.40
.04
.08
.06
.10
pH limit is 6.0-9.0
Fecal coliform maximum limit is 400 mpn/100 ml
Maximum for any 1 day is 2X the average.
1983: 0&G, Maximum = 10 mg/1
NH^-N, Maximum = 8 mg/1
-------
quidelines and the probable inclusion of nutrients in effluent standards
will, in many cases, require the industries to install further treatment
beyond existing secondary treatment. If an industry elects to discharge
to a municipal system, pretreatment might be required.
Table 2.
POULTRY PROCESSING: RECOMMENDED
EFFLUENT
LIMITATIONS
EPA
Contract No. 68-01-0593
Average Limitation
(kg/kkg LWK)
7/1/77
7/1/83
Plant.
Best
Practical Treatment
Best Available Treatment
Subcategory
b^5
TSS O&G
TSS
O&G
Chickens
.46
.62 .20
.30
.34
.20
Turkey
.36
.57 .14
.21
.24
.14
Fowl
.61
.72 .15
.23
.27
.15
Ducks
.77
.90 .26
.39
.46
.26
Fecal colifortn maximum limit is 400 mpn/100 ml.
NH^-N: New Source and 1983 limit is 4 mg/1.
1983: TKN = 4 mg/1, TP = 2 mg/1, N03/N02-N = 5 mg/1.
Through analysis of data from existing outstanding meat packing,
slaughtering and poultry processing plants, it is hoped to identify
those combinations of in-plant controls and subsequent wastewater
treatment processes which are most efficient in meeting discharge
regulations.
3
-------
Primary objectives:
1. Determine in-plant controls which produce least magnitude and
variability in treated effluent loads.
2. Determine treatment system or systems used to produce least
magnitude and variability in effluent loads.
3. Determine what percentage of time plant discharge meets estab-
lished effluent limits.
4. Determine causes when limits are exceeded.
5. Determine the level beyond which a treatment plant ceases to
provide effective treatment.
Information needed:
Records on:
1. Production
2. Effluent flow
3. Effluent compositions
a. Untreated
b. Treated
4. In-plant process description
5. Treatment plant design
4
-------
RESULTS
Data Search
Data was obtained for this project from individual state pollution
control agencies, the U.S. Department of Agriculture, North Star Research
Institute, and private industry in addition to EPA.
Initially it was decided that in order to have a meaningful statis-
tical analysis, at least 40 data points for the various wastewater
quality parameters would be needed. This alone limited the number of
plants for which data would be available.
Due to the lack of a regular reporting system in many states, only
those plants that were believed to be producing a poor quality effluent
were monitored regularly. This study is concerned with the more exemplary
plants, and consequently there is little information available. Origin-
ally it was intended to evaluate the treatment systems with respect to
biochemical oxygen demand, chemical oxygen demand, suspended solids, oil
and grease, ammonia nitrogen, and phosphorus. It was found that data on
parameters other than BOD^ and SS was very limited, and as a result,
comparison of the different types of treatment systems with respect to
the other parameters was not possible.
Meat: Raw Wastewater
Table 3 shows the mean water use as well as raw BOD^ and SS wastewater
loads from meat-packing plants. The two plants showing the lowest mean
values for BOD^ and SS are 8A and 9A with values of 10.2 and 8.1 kg
BOD^/kkg LWK and 10.0 and 8.4 kg SS/kkg LWK, respectively. Another
plant, 1A, even though the mean BODg and SS loads are not quite as low,
does exhibit a variability that is minimal as can be seen by the
5
-------
Table 3. RAW WASTEWATER CHARACTERISTICS
OY
Effluent
B0D5d SSa
Guidelines Std. . Std. Water
Plant Category Mean Max Dev. N Mean Max Dev. N Use
1A II 10.7 18 3.8 45/26 12.8 24 4.5 46/26 4.38 (525)
3A II 23.7 33 4.9 34/12 17.7 27 3.1 35/12 15.60 (1870)
4A III 12.0 23 7.8 49/37 13.0 27 13 48/37 8.29 (993)
7A II 18.0 45 10.6 33/17 14.4 36 10 30/17 9.27 (1111)
8A II 10.2 21 5.3 91/24 10.0 21 6.4 91/24 7.89 (946)
9A II 8.1 15 3.1 42/11 8.4 16 7.5 53/11 8.44 (1012)
aValues in kg/kkg LWK
- number of observations/number of months in observation period
cWater use in rn^/kkg (gal/1000 lb)
-------
standard deviations. At the other end of the scale are plants 3A and 7A
with raw wastewater loads of 23.7 and 18.0 kg BOD^/kkg LWK and 17.7 and
14.4 kg SS/kkg LWK, respectively. Figures 1 and 2 show probability
curves developed to graphically show the variability of the effluent.
It is difficult to pinpoint the reasons for the lower wasteloads
without conducting an in-plant survey; therefore, conclusions have to be
drawn based on the processing plant raw wasteload. For the above
plants, one feature that might explain the differences is water use. In
Table 3, plants 1A, 8A, and 9A, the three plants with the lowest water
use for category II, are also the ones mentioned above exhibiting the
least raw wastewater load. Plant operations that could contribute
toward this lower water use are better housekeeping practices and making
use of dry cleanup prior to washdown. Another significant factor in
water use reduction is the installation of high pressure nozzles on
water outlets.
Meat: Final Effluent
Results of different treatment systems employed in the meat-packing
industry are presented in Table 4. All make use of a biological secondary
system, either an anaerobic-aerobic lagoon system or an anaerobic contact
process followed by aerobic lagoons.
As can be seen from the aforementioned table and Figure 3, comparing
the results with the effluent limitations presented previously shows
that only one of the plants could not meet the 1977 mean BOD^ limits for
its respective category and that the maximum value was exceeded by five
of the six. Figures 4 and 5 show probability curves developed to graphic-
ally show the BODjj variability of the effluent.
7
-------
<40.000
cc
3
30.000
^1
*
20.000
o
a
a
TH
a
¥
10.000
Q
-------
VO
42.000
2 34.000
3
J 26.000
O
*
o
o
o
19.000
N
o
\£ ia.aaa
Q
le load
-------
Table 4. FINAL EFFLUENT CHARACTERISTICS
Plant
1A
3A
4A
7A
8A
9A
Effluent
Guidelines
Category
BOD,
SSC
II
I
I
I
Mean Max
Std.
Dev.
Mean Max
.11 .23 .06 29/14
16 .55 .10 32/36
33 1.45 .26 49/29
14 .46 .10 92/24
21 .62 .15 42/11
Std.
Dev,
.13 .58 .13 43/27 .33 2.31 .37 46/27
.30 1.15 .22 29/14
.67 1.95 .47 28/36
34 1.52 .34 46/29
.43 1.55 .32 92/24
,44 1.56
31 53/11
aValues in kg/kkg LWK
= number of observations/number of months in observation period
-------
Q I
—I cr>
IT) ^
Q ^
O \
CD cn
KEY:
guidelines
avg.
actual avg
£
-J
cn
V
Cn
Q
<
O
-J
|
•95
lv*J
I
2
ll
2t£
4 A 7 A
PLANT
j
81
/AV
8 A
J
9 A
Figure 3. HOD^ and SS actual vs. 1977 guidelines 30-day average and
daily maximum.
-------
3
-J
(3
*
o
a
a
tH
N
(2>
*
Q
_ <1
r : O
-J
a
o
CD
I-
Z
uj
D
U
lL
U.
u
_l
N=.H
9ft: MEAN=.2l
Figure 4.
10 20 30 40 50 BO 70 00 90
PR0BABIl.IT 1 OF VALUES lE CORRESPONDING lOAD)
Log-Probabi1ity representation of meat-packing plant final effluent GOD^ waste load,
category II plants.
-------
*
3
U
O
*
o
CD
CD
O
¥
Q
O
-J
O
o
GQ
Ui
3
J
Lj.
U.
U
_J
-------
From Figures 3, 6 and 7 it can be seen that of the plants under
consideration in this evaluation, none were able to meet either t.ho 1977
mean or maximum suspended solids limits. Plants 1A, 3A, and 7A were
closest to meeting the 0.25 kg/kkg LWK average limit, but none were near
the established maximum value.
It should be noted that the average reported for each of the plants
is a long-term average and not the maximum average for a 30-day period
as specified in the guidelines. Because past monitoring policies were
not. established to meet the present system of effluent limitations, it
was believed that reporting the data as a maximum average for a 30-day
period did not truly reflect the capabilities of the treatment system.
This was considered, but in some cases with infrequent sampling, the
maximum average for a 30-day period approached the maximum daily value.
In other cases, with more frequent sampling, the maximum average for a
30-day period was in fact less than the long-term average.
The treatment system employed by plant 3A produced the lowest mean
and maximum BOD^ and suspended solids loads of the six meat-packing
plants. This system consisted of an anaerobic contact process and
anaerobic-aerobic stabilization ponds. As mentioned previously, all
others made use of an anaerobic-aerobic lagoon system.
Since there is a lack of diversity and number in the type of
treatment systems employed, no attempt is made to judge one system type
better than another. Of the ones represented here, the ability of one
to perform more effectively than the other is assumed then to be based
on the operation of the plant.
An attempt was made to determine the cause of the occasions when
the maximum allowed limits were exceeded. This was done using what
information was available in regard to weather conditions and plant
14
-------
-J
a
o
a
is:
o
O
-j
u_
Ll
UJ
-J
Cateaor.
)1 3 n t
-------
1.500
CJ> 1.000
o
o
a
X X
.500
MQXTMI IM 1 TMTT
X*
MFftM I TMTT.
R5Q=.BB7
4R: MEP1N=.5B
I—I
.100
1Q 20 30 40 50 GO 70 80 30
PROBABILITYC% OF VALUES LE CORRESPONDING LOAD)
Figure 7. Log-Prcbabi1ity representation of neat-packing plant final effluent SS waste load,
Category III plant.
-------
processes. The extent of the data varied widely from plant to plant.
In some cases, when data permitted, an attempt was made to determine if
the treatment plant was operating under normal loading or if some in-
plant upset occurred which might affect the treatment system. In regard
to weather data, extremes in temperature and periods of high rainfall
were variables in which data was more readily available that might aid
in drawing some conclusion as to the operation of the treatment plant.
Beginning with plant 1A and making use of the basic data, events of
high effluent load, both B0D5 and suspended solids, are presented when
available in Table 5 with an explanation of the prevailing operating
conditions.
From looking at those conditions, it is assumed that extremes in
weather affected the performance of the lagoon systems used in the
plants. For plants 1A, 7A, and 8A all periods of ineffective operation
coincided with either unusually cold weather or periods of high rainfall
which produced an effluent flow greater than normal. These unusual
operating conditions account for the occasions when the established 1977
effluent limitations were exceeded.
Table 6 shows the frequency of occurrence of values less than or
equal to the 1977 mean and maximum effluent limitations based on the
predicted curves of Figures 4-7. Plants 1A, 3A, and 7A are the only
ones that might be capable of meeting the mean or 30-day average limita-
tion for suspended solids. Of these plants the maximum limitation can
only be met up to 85 percent of the time. The other 15 percent would be
out of compliance.
17
-------
Table 5. HIGH EFFLUENT LOAD OPERATING CONDITIONS
Effluent
Parameter
Date
Condition
B0D£
SS
2/10/72
2/24/72
3/9/72
Period of sub-zero temperature,
snow, and treatment plant
overloaded
Temperature near or below
freezing, no overload
BOD,
SS
5/17/71
4/24/74
5/17/71
4/17/72
BOD^ 70 mg/1 (20 mg/l'inf1uent
to final laqoon), flow and
production normal.
B0D& 50 mg/1, 2nd highest
reported between 3/71 and
4/74
SS 250 mg/1 (200 mg/1 >inf1uent
to final lagoon)
SS 240 mg/1 (110 mg/1 Mnf 1 uent
to final lagoon)
BODr
SS
1/25/73
7/13/73
5/24/74
6/20/72
10/24/72
4/23/74
Final flow-normal
Rain, final flow>nornial
2" rain in previous 2 days,
high flow, SS 133 mg/1
2 1/2" rain, high flow,
SS .140 mg/1.
B0Dc
SS
5/11/72
11/8/72
11/15/72
5/4/72
6/15/73
Rain, final flow>normal
flow>normal
M H
1" rain, flow>normal
flow>normal, SS 136 mg/1
18
-------
Table 6. FREQUENCY OF OCCURRENCE OF VALUES < 1977 EFFLUENT LIMITATIONS
(From predicted curves,
in percent)
B0D5
SS
Plant
Mean
Max
Mean
Max
1A
82
93
56
82
3A
89
100
57
82
4A
68
96
25
55
7A
38
78
54
82
8A
80
95
44
71
9A
64
87
28
68
Poultry:
Raw Wastewater
Table 7 shows the mean raw BOD^ and SS wastewater loads from
poultry processing plants. Raw wastewater quality data was very limited,
with only two of the plants having BOD^ data and one with SS data. The
two, both broiler processing plants, were quite different in terms of
raw waste load, 6.4 vs. 36.7 kg/1000 kg LWK. Figures 8 and 9 show the
variability of the raw waste load on a probability basis. One factor
that would account for part of the variation in the load between the two
plants is that plant 2B does not engage in further processing or rendering.
However, the low wastewater load of plant 2B should merit further attention.
Shown also in Table 7 is the water use by the four poultry processing
plants in this study. Carawan, et. al., (2) concluded that lower water
use in a poultry processing plant reduces the raw wastewater load.
Plant 2B instituted a water conservation program in which the use dropped
from approximately 42 to 26 liters per bird. Water reduction measures
19
-------
Table 7. RAW WASTEWATER CHARACTERISTICS
BOD,a SSa
Effluent 5
Guidelines Std. . Std. Water
Plant Category Mean Max Dev. N Mean Max Dev. N Use
2B Broiler 6.4 15 2.2 62/8 12 41 7.0 61/8 26.0 (6.9)
3B Duck -- -- — — 26.7 (7.06)
5B Broiler — — -- — -- -- -- -- 67.4 (17.8)
6B Broiler 36.7 52 11.8 10/5 - -- -- -- 42.8 (11.3)
aValues in kg/kkg LWK.
^N=number of observations/number of months in observation period
cWater use in liters/bird (gal/bird)
-------
40.000
s~\
*
3 3Q.000
-J
3
^ 2D.000
O
CD
O
O
V-/
o
%
Log-Probability represe
+
:xxx
26: MEPN= 6.4 RS0=.952
BB: MEAN=36.7 RSQ=.943
— j—--.—— -f - f -4 —- --4-—
30 43 50 EO 7G BO
;ES L£ CORRESPONDING
poultry procession plant raw SOD- waste load.
30
LOAD)
-------
a
a
a
x *
xx
5.000
2B: MEAN=6.4
10 2fl 30 40 5C SO. 7Q 80 90
PROBABILITY of' v'Al_U£3 l£ corresponding load;
Figure 9. Log-Probability representation of poultry processing plant raw SS waste load.
-------
consisted of an employee awareness program, daily inspections, elimina-
tion of piping leaks and the use of hoses where possible, and automatic
shut off valves on essential hoses. Supply lines were equipped with
valves to allow for flow regulation. Dry cleanup was practiced prior to
washdown with a portable high pressure cleaning system. Lastly, chill
vat water was recycled for reuse in the scalder. It is assumed, based
on Carawan's earlier conclusions, that the water conservation measures
undertaken by plant 2B are largely responsible for the lower raw waste
load of the plant.
As an additional note plant 3B, the other plant with low water use,
instituted water conservation measures. These consisted of a water
pressure reduction from 414 to 138 kN/m (60-20 psi), installation of
spray type hand washers, use of high pressure-low water systems in
clean-up, and cutoff in flow trough system to stop water flow when lines
are not running.
Poultry: Final Effluent
Table 8 summarizes the results of treatment systems employed in the
poultry processing industry. Biological secondary systems used are
aerated lagoons, anaerobic-aerobic lagoon system, and activated sludge.
Due to the guidelines regarding effluent limitations not being
finalized at the time of this writing, the results obtained by the
plants in this study are compared to the 1977 values recommended by the
contractor to EPA. These are not to be taken as a limit, but only a
value from which to compare the results obtained here. As with the
meat-packing plants, the averages presented are the long-term averages
and not the maximum average for a 30-day period.
23
-------
Table 8. FINAL EFFLUENT CHARACTERISTICS
BOD5a SSa
Std. , Std.
Plant Category Mean Ma_x Dev. N Mean Max Dev ¦ N
26 Broiler .40 1.7 .35 102/11 1.0 3.5 .79 102/11
3B Duck .18 .56 .11 46/27 .32 1.35 .23 47/27
^ 5B Broiler .67 1.2 .27 36/5 1.61 4.8 1.33 28/5
6B Broiler .29 1.2 .24 68/8
aValues in kg/kkg LWK
= number of observations/number of months in observation period.
-------
Of the four plants presented, two of the recommended categories are
represented. These are the duck processor and the broiler processor
with one of the broiler processors (6B) engaged in further processing
and rendering.
Table 8 and Figure 10 show that two of the broiler processors and
the duck processor would meet the average BOD^ limits. The maximum
daily value, of the plants within the recommended average varies in the
range of 1.8 - 4.2 times the mean obtained by these plants. Based on
the recommended averages of 0.46 and 0.77 kg BOD^/kkg LWK for plants 2B
and 3B, respectively, the maximum would be 3.7 and 0.73 times the
average. Plant 3B is a case where the values are so low as to make the
maximum less than the recommended average. Plant 6B, with the maximum
4.1 times its average or 1.7 times the recommended average is a special
case since this plant also handles further processing and rendering
operations. Though it would meet the more stringent B0D5 limit for a
normal broiler processing plant, its mean value of 0.29 kg/kkg LWK would
readily be in compliance with the 0.72 kg/kkg LWK recommended limit
based on an adjustment factor for the ancillary operations. Figures 11,
12, and 13 show the probability curves developed to graphically show the
BOD^ variability of the effluent for these plants.
Table 8 also shows the suspended solids levels obtained by all the
plants except 6B and these are compared to the recommended limits in
Figure 10. Only one of these (3B) was able to meet the recommended
limits and it did so readily, with the maximum of 1.35 kg/kkg LWK being
greater than the recommended average of 0.90 kg/kkg LWK by a factor of
1.5. Figures 14 and 15 graphically show the suspended solids variability
by the use of probability plots.
The treatment system employed by plant 3B consists of in-plant
screening followed by three aerobic lagoons in series. This plant also
made use of extensive water conservation practices as mentioned earlier.
UBRATtf protector.
11 S En»«OnBie . i U.. i< •ll-"
25 2C»9* S'TS-IJJ®
-------
o 5
< 5
O -1
I en
in
Q *
O ^
00 2
_J
o>
CP
Q
<
O
_J
CO
CO
1-2
80
¦40
0-0
5-0
i
4-0
3-0
2-0
1-0
0-0
i
KEY:
guidelines
avg. ~*—
actual avgS
^actual
max.
2 B 3 B 5 B
PLANT
6 B
f-'iqure 10. BODc, and SS actual mean and maximum vs. 1977 recommended guidelines
mean. 26
-------
*
ro
*
3
-J
t!>
X.
o
a
a
O
\-j
a
c
o
-J
Q
O
CD
F—
2
Ui
D
-J
Ljl
Ll
LlI
_J
-------
.700
.E00
=S
.500
-J
.400
U>
X
.300
o
o
o
tH
.200
O
¥
Q
-------
3
-J
U>
¥
o
o
o
tH
s
¥
O
a
o
a
o
CD
UJ
D
-J
Li.
U.
LU
_J
<1
2
l'.j
l.QQQ
.010
.500
.100
.050
MFftM I TMTT
,xxx
>©*,
jnorr
/rvvvy
xxx:
GB: MEAN=.29 RSQ=.959
10 2G 30 4Q 50 60 70 BO 90
PROBABILITY c% of VALUES LE CORRESPONDING LOAD)
Figure 13.
Log-Probabi1ity representation of Doultry processing plant final effluent E0D- waste load,
broiler category plus further processing and rendering.
-------
co
o
« .CQQ
2.003
/-% 2.0(30
3
-1 1.000
o
o
a
a
Ci
N-/
Q
C
o
_l
U1
U)
LU
Z>
-J
Lu
Li.
LU
a
.500
.LOO
.050
.010
.005 ..
.001
Figure 14.
MPWTM1IM I TMTT
MFftM I IMTT
2Bj MERN=1.0
58: MEDN=l.e
RSQ=. 932
RSG=. 353
10 20 30 40
PROBABILITY i% OF VALUES LE
Log-Probabi1ity representation of poultry orocessino plant
broiler category.
50 GO 70 90 30
CORRESPONDING LOAD)
inal ef'ljert SS waste load.
-------
Ca>
X.
3
_J
o
a
o
N
o
is:
Q
-J
u.
Ll
LJ
a
2
1.500
1.400
1.300
1.200
1.100
1.000
.500
¦ LOO
MFftN I TMTT
3B: MEAN=.32 R5G=.99Q
PROBABILITY
10 20 30
(5£ OF VALUES
50 BO 7Q 90 90
CORRESPONDING LOAD?
Figure 15. Log-Probabi1ity representation of poultry processing plant final effluent SS waste load,
duck category.
-------
Plant 6B, producing a low BOD^ load, also is capable of effective
suspended solids treatment. Sufficient data was not available to
properly evaluate it, but it has been able to produce an effluent
suspended solids of 5 mg/1. This system consists of a grease trap, wet
well, holding tank, chemical mixing, flotation, activated sludge, two-
parallel microstrainers, and a chlorine contact tank.
As with the meat-packing plants an attempt was made to determine
the cause for the extremely high values. Information on weather condi-
tions and plant processes is not as complete as meat-packing plants
and there is adequate data on only one plant, 3B. Table 9 presents the
periods of high effluent-loads with an explanation of the prevailing
operating conditions. In all cases of high effluent BOD^ or SS loads
there also existed an unusual weather condition, either heavy rainfall
or high wind. This plant has since made provisions to eliminate rainfall
runoff from its treatment system.
Table 9. HIGH EFFLUENT LOAD OPERATING CONDITIONS
PI ant
3B
Effluent
Parameter
Date
8/9/72
6/25/73
9/5/73
Condi tion
bod5
Heavy rain, high runoff
Heavy rain, muddy outlet
High discharge, affected
by summer rainfall
SS
8/9/72
8/28/72
10/9/72
Heavy rain, high runoff
High wind, SS > 100 mg/1
Heavy rain
32
-------
Table 10 shows the frequency of occurrence of values • the 1977
mean effluent limitations based on the predicted curves. Plants 2B, 3B,
and 6B are capable of meeting the mean BOD^ effluent limitation with at
least 72 percent of their values being lower than the recommended mean.
For 3B and 6B, this value increases to 92 and 100 percent, respectively.
With respect to suspended solids, only 3B falls within the recommended
average greater than 50 percent of. the time, in this case 100 percent.
It is noted again that there was not sufficient data to evaluate plant
6B for SS.
Table 10. FREQUENCY OF OCCURRENCE OF VALUES < RECOMMENDED
1977 EFFLUENT LIMITATIONS
(From predicted curves, in percent)
Plant Mean BQD^ Mean SS
2B 65 35
3B 100 100
5B 25 25
6B 100
Winter-Summer Variation of Treated Effluent
Both the meat-packing and poultry processing plants were tested for
any winter versus summer variation. The summer being classified as May
through October. The means and variances were tested for equality at a
confidence level of 95 percent. Table 11 shows the results of the
statistical tests performed on the means and variances. No clear
pattern could be detected for the variation. Plants 1A and 4A, which
both meet the 1977 BODg limits but not the suspended solids limits, were
33
-------
the only meat-packing plants meeting the test for the equality of the
means. Plant 4A also meets the criteria for equality of the variances
at a 95 percent confidence level. All other meat-packing plants fail
the test except for 3A which did not have sufficient data for comparison.
Table 11. RESULTS OF STATISTICAL TESTS FOR WINTER-SUMMER
EQUALITY OF MEANS AND VARIANCES
(Tests conducted at a 95% confidence level; "1" implies meeting
the test, "0" does not; first two digits relate to BOD^ mean
and variance, second two are for SS.)
Plant
1A
4A
7A
8A
9A
2B
6B
Plants 2B and 6B, the two poultry processors tested for any winter-
summer variation, both met the tests for the equality of the means and
variances with respect to any BODg or suspended solids variation except
for 6B which had no suspended solids data. With respect to ability to
meet the 1977 recommended effluent limitations, plants 2B and 6B readily
meet the BOD^ guidelines, but 2B is not able to meet the suspended
solids limits.
Results
1010
1111
0000
0101
0100
1111
11--
34
-------
Summary and Conclusions
The results on the raw waste load tend to support conclusions
reached by Carawan (2) that lower water use reduces the raw waste load.
Plants that instituted a water use reduction program also had a correspond-
ing wastewater load reduction.
On a long-term basis, the meat-packers appear to be capable of
meeting the 1977 limitations on BOD^, but for suspended solids the
maximum daily value is critical. Of the six plants in this study, the
most effective one exceeded the maximum suspended solids limitation 15
percent of the time.
Based on the recoimiended guidelines for the poultry processing
industry, there are plants in existence that can readily meet the 1977
BOD^ and suspended solids limitations. The 1983 BODg average limits are
being met by two of the plants, a duck processor and the broiler or
chicken processor that engages in further processing and rendering. The
duck processor is also able to meet the 1983 maximum BOD^ and average SS
limitations assuming the maximum limit will be established as 2X the
average.
For both the meat-packer and poultry processor, periods of high
effluent load coincided with periods of abnormal weather conditions. In
the case of the meat-packing plants, where sufficient data was available,
all occasions of exceeding the maximum limitation were accompanied by
severe weather conditions.
Of the plants evaluated in this study, there is no conclusive
evidence as to the variability of the effluent with respect to winter-
summer. Overall performance of a treatment plant appeared to have no
effect on the variability with respect to the seasons.
35
-------
Due to lack of diversity and number, it was not possible to compara-
tively evaluate the effectiveness of the treatment systems. For the
meat-packing industry, both treatment systems of an anaerobic contact
system followed by aerobic stabilization ponds and anaerobic-aerobic
lagoon systems were similarly effective. For the poultry processing
industry, effective treatment was provided by three different biological
secondary systems; aerated lagoons, anaerobic-aerobic lagoons, and
activated sludge. The wastewater management systems employing the
aerated lagoons (three in series) and the activated sludge were more
readily able to meet the recommended effluent limitations. The latter
system consisted of a grease trap, wet well, holding tank, chemical
mixing, flotation, activated sludge, two-parallel tnicrostrainers, and a
chlorine contact tank.
Inferences drawn from the plots of the effluent load versus frequency
for the meat-packing and poultry plants are that in order to stay within
the maximum limitation, an average less than the 30-day average limit
will be necessary. This is due to the fact that the maximum value of
the effluent load is greater than twice the average load. Some plants,
as in the case of BODg for 6B, even though the maximum value is greater
than 2X the mean, both fall within the average and maximum limits due to
the low average value. Since a plant such as this is weighted more
heavily in the lower range of values, it is to be expected that the
maximum would be greater than 2X the average due to the range of effluent?
loads inherent to the system.
Relating the results to the enforcement of the established guidelines
would lead one to question the enforceability of the maximum limit.
Using the system of plant 1A, the maximum allowed BOD^ value was exceeded
only one time in forty-three samples. For monthly or weekly sampling
this would result in a probability of sampling on a day that the maximum
36
-------
-9 -5
was exceeded of only 2x10 and 5x10 respectively. See appendix for
calculation. In any plant that exceeded this limit frequently enough to
cause enforcement action, the high results would increase the mean such
that enforcement of these extreme values would be covered in the mean.
This leads to a system proposed for one by Popel (3) in which a
combined standard would be established with a set limit and an allowed
probability of exceeding it. To illustrate this concept, the results of
plant 6B are plotted again in Figure 16 as in Figure 13, except this
time using an arithmetic scale. This more effectively shows the point
at which the treatment efficiency falls off. The combined standard
would set the limit at this point, or the 90 percent level in this case.
This would allow for the few occasions of the less effective plant
operation that might be due to weather conditions, but in actuality
would closely represent the true performance of the treatment system for
a major percentage of the time.
37
-------
*
3
J
O
o
o
a
(2>
Q
XX
SQCX*fi?XXX
xxx%xxx~
^x^xxx''
.-5cXXxxa
.Utb L_t. CQK^.-b-J V- - \ J
OQn ¦>.
Figure 16. Arithmetic representation of plant 6B final effluent BOD^ load and frequency of occurrence.
-------
ACKNOWLEDGEMENTS
This paper was a result of an in-house project and two Corvallis
FPA personnel deserve special acknowledgement. Judy Carkin for her role
in the computer analysis and Jack L. Witherow for his suggestions made
during the course of the project.
In addition to private industry, several individuals were instrumental
in providing data for the project. These are:
Jeffery D. Denit E. E. Erickson
Effluent Guidelines Division-EPA North Star Research
Institute
Jim Chittenden
National Independent Meat Packers Assn.
Don Dencker
American Meat Institute
Dr. Arnold Giesemann
U.S.D.A-APHIS
John Schmidt
Pennsylvania Dept. of
Environmental Resources
Brian J. Holmes
Virginia State Water Control Board
Russell C. Felt
Minnesota Pollution Control
Agency
Philip R. O'Leary
Wisconsin Dept. of Natural Resources
Tom Wallin
Illinois EPA
Richard Rankin
Iowa Dept. of Environmental Quality
-------
REFERENCES
1. Ford, Davis L. Factors Affecting Variability From Wastewater
Treatment Plants. Proceedings, International Conference on Effluent
Variability from Wastewater Treatment Process and Its Control.
December 2-4, 1974. New Orleans, Louisiana. Sponsored by the
International Association on Water Pollution Research, Tulane
University, and Vanderbilt University.
2. Carawan, Roy E., W. M. Crosswhite, J. A. Macon, and B. K. Hawkins.
Water and Waste Management in Poultry Processing. EPA-660/2-74-
031, May 1974.
3. Popel, H. J. A Concept for Realistic Effluent Standards. Proceed-
ings, International Conference on Effluent Variability from Waste-
water Treatment Processes and Its Control. December 2-4, 1974.
New Orleans, Louisiana. Sponsored by the International Association
on Water Pollution Research, Tulane University, and Vanderbilt
University.
4. Burr, I. W. Engineering Statistics and Quality Control. McGraw-
Hill Book Company, New York, 1953.
40
-------
appendix
Explanation of Abbreviations Used
BOD^: 5-day Biochemical Oxygen Demand
TSS: Total Suspended Solids
SS: Suspended Solids, same as TSS
O&G: Oil and Grease
NH^-N: Ammonia Nitrogen
N03/N02-N: Nitrate-Nitrite Nitrogen
TKN: Total Kjeldahl Nitrogen
TP: Total Phosphorus
mpn: Most probable number
kg/kkg LWK: Kilograms per 1000 kilograms live weight killed
rn: Meter
1: 1i ter
2
kN/m : Kilonewton per square meter
psi: pounds per square inch
inch
LE: less than or equal to
Equations Used for Calculation of Adjustment Factors for Plant 6B
(Poultry)
For 1977 Discharge limitations:
Rendering
B0D5: (.15 kg BODg/kkg RM) (kkg RM,/kkg LWK)
SS: (.17 kg SS/kkg RM) (kkg RM/kkg LWK)
41
-------
Further Processing
BOD5 (.30 kg BOD5/kkg FP) (kkg FP/kkg LWK)
SS: (.35 kg SS/kkg FP) (kkg FP/kkg LWK)
For 1983 Discharge Limitations
Rendering
BOD5: (.07 kg BODg/kkg RM) (kkg RM,/kkg LWK)
SS: (.10 kg SS/kkg RM) (kkg RM/kkg LWK)
Further Processing
BOD5: (.15 kg BODg/kkg FP) (kkg FP/kkg LWK)
SS: (.18 kg SS/kkg FP) (kkg FP/kkg LWK)
where
RM = amount of raw materials rendered on site
FP = amount of further processing done
BODg (adjusted) = BOD^ (recommended for subcategory) + BOD^ (rendering)
+ BOD5 (further processing)
SS (adjusted) = calculate same as for BODg
42
-------
Calculation of Probability of Sampling the Occasion of Maximum Effluent
Load
From Burr (4)
C(n,r) = P(n,r)/P(r,r)
= n!/r!(n-r)!
where
C(n,r) is the desired number of combinations of n different things
r at a time,
P(n,r) is the number of permutations of n different objectives
taken r at a time, and,
P(r,r) is the number of different orders in which all of r objects
may be drawn.
For the case of monthly sampling, or 12 samples in 260 working days and
a violation that occurs 1 in 43 (6 in 260) working days.
C(254,6) = number of 12-sample monitoring schedules that contain
all 6 violations in a 260-day sampling period.
C(260,12) = number of 12-sample monitoring schedules in 260-day
sampling period, and
P(6 violations) = C(254,6)/C(260,12)
Where P(6 violations) is the probability of sampling the six viola-
tions.
C(254»6) - 2541/6!(248!)
C(260,12) = 2601/12!(248)!
43
-------
P(6 violations) = 2.28 x 10"^
Similarly, for the case of weekly sampling, or f>2 samples in ;'60
working days.- ,,
C(2M .46) - ?i54!/46! (208)!
C(260,b2) - 2601/52! (208)!
and,
P(6 violations) = 5.03 x 10"^
44
------- |