PROPOSED DEVELOPMENT DOCUMENT
FOR
NEW SOURCE PERFORMANCE STANDARDS
FOR THE
PHARMACEUTICAL MANUFACTURING
POINT SOURCE CATEGORY
WILLIAM D. RUCKELSHAUS
ADMINISTRATOR
JEFFERY D.'-DENIT
DIRECTOR, EFFLUENT GUIDELINES DIVISION
ROBERT W. DELLINGER
ACTING CHIEF, WOOD PRODUCTS & FIBERS BRANCH
FRANK H. HUND, Ph.D.
PROJECT OFFICER
WENDY D. SMITH
ASSISTANT PROJECT OFFICER
SEPTEMBER 1983
EFFLUENT GUIDELINES DIVISION
.OFFICE OF WATER
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
-------
TABLE OF CONTENTS
SECTION
CONCLUSIONS
II
III
IV
V
GENERAL
NEW SOURCE PERFORMANCE STANDARDS (NSPS)
INTRODUCTION
PURPOSE AND AUTHORITY
SCOPE OF THIS RULEMAKING
SUMMARY OF METHODOLOGY
DESCRIPTION OF THE INDUSTRY
INTRODUCTION
SUBCATEGORIZATION
EXISTING END-OF-PIPE TREATMENT AT PHARMACEUTICAL
PLANTS
WASTE CHARACTERIZATION
INTRODUCTION
WASTE CHARACTERIZATION
DEVELOPMENT OF MODEL PLANT RAW WASTE
CHARACTERISTICS
Subcategory A and C Plant Group
Subcategory B and D Plant Group
Summary
DEVELOPMENT OF CONTROL AND TREATMENT OPTIONS
INTRODUCTION
CONTROL AND TREATMENT OPTIONS
NSPS Option A
NSPS Option B
EFFLUENT VARIABILITY ANALYSIS
Introduction
Daily Variability Factors
Thirty-Day Average Variability Factors
DEVELOPMENT OF VARIABILITY FACTORS USED
IN DEVELOPMENT OF PROPOSED NSPS
Advanced Biological Treatment
Filtration
Summary
PAGE
3
3
5
7
7
11
1 1
11
18
18
18
21
21
23
26
28
28
28
29
32
36
36
36
-------
TABLE OF CONTENTS (Continued)
SECTION
VI
VII
VIII
IX
COST, ENERGY, AND NON-WATER QUALITY
ASPECTS
INTRODUCTION
METHODOLOGY FOR DEVELOPMENT OF COSTS
Introduction
Model Mill Approach
Cost Estimating Criteria
Costs for Implementation of NSPS
Options
ENERGY AND NON-WATER QUALITY IMPACTS
Energy Requirements
Solid Waste Generation
Air Pollution and Noise Potential
EFFLUENT REDUCTION ATTAINABLE THROUGH THE
APPLICATION OF NEW SOURCE PERFORMANCE
STANDARDS
GENERAL
IDENTIFICATION OF THE TECHNOLOGY BASIS
OF PROPOSED NSPS
PROPOSED NSPS
RATIONALE FOR THE SELECTION OF THE
TECHNOLOGY BASIS OF PROPOSED NSPS
METHODOLOGY USED FOR DEVELOPMENT OF
PROPOSED NSPS
COST OF APPLICATION AND EFFLUENT
REDUCTION BENEFITS
NON-WATER QUALITY ENVIRONMENTAL IMPACTS
REFERENCES
ACKNOWLEDGEMENTS
PAGE
39
39
39
39
39
41
46
46
46
52
55
55
55
55
55
57
57
59
61
-------
NUMBER
LIST OF TABLES
TITLE
PAGE
Section I
1-1
Proposed Conventional Pollutant NSPS for the
Pharmaceutical Manufacturing Point Source
Category
Section III
III-2
Summary of Method of Discharge at
Pharmaceutical Plants
In-Place Treatment Technology at Direct
Discharging Pharmaceutical Plants
Section IV
IV-1
IV-2
IV-3
IV-4
IV-5
Section V
V-l
V-2
V-3
Raw Waste and Final Effluent Characteristics
of Direct Discharging Pharmaceutical Plants
Raw Waste Characteristics of Subcategory
A and C Best Performers
Raw Waste Characteristics of Subcategory
B and D Best Performers Employing Biological
Treatment
Raw Waste Characteristics of Subcategory
B and D Best Performers Employing Biological
Treatment and Effluent Filtration
New Source Model Plant Raw Waste
Characteristics
Conventional Pollutant Removal at Plant 12161
Through the Application of Effluent Filtration
Technology
Final Effluent Characteristics of Best
Performing Subcategory A and C Pharmaceutical
Plants Employing Advanced Biological Treatment
Final Effluent Characteristics of Best
Performing Subcategory B and D Pharmaceutical
Plants Employing Advanced Biological Treatment
12
15
16
17
19
22
24
25
-------
LIST OF TABLES (Continued)
NUMBER TITLE
V-4 Final Effluent Characteristics of Subcategory
B and D Pharmaceutical Plants Employing
Advanced Biological Treatment and Effluent
Filtration
V-5 Individual Variability Factors for Specific
Subcategory A and C Pharmaceutical Plants
Employing Advanced Biological Treatment
V-6 Individual Variability Factors for
Specific Subcategory B and D Pharmaceutical
Plants Employing Advanced Biological Treatment
V-7 Individual Variability Factors for Specific
Pharmaceutical Plants Employing Advanced
Biological Treatment and Effluent Filtration
V-8 Individual Variability Factors for Specific
Pharmaceutical Plants Employing Advanced
Biological Treatment and Effluent Filtration
Section VI
VI-1
VI-2
VI-3
VI-4
VI-5
VI-6
Cost Estimating Criteria
Design Basis of the Treatment Systems
Expected To Be Employed at New Source
Pharmaceutical Industry Direct Dischargers To
Meet Baseline Effluent Levels
Design Basis of the Treatment Systems
Expected To Be Employed To Meet NSPS
Option A Effluent Levels
Design Basis of the Filtration Systems
Expected To Be Employed To Meet NSPS
Option B Effluent Levels
Model Plant Costs Associated with Meeting
Baseline, NSPS Option A, and NSPS Option B
BOD5. and TSS Final Effluent Concentrations
Summary of Costs for Treatment System
Components for the 1.2 MGD Subcategory
A and C Model New Source Plant
PAGE
27
33
34
35
37
40
42
44
47
48
49
-------
LIST OF TABLES (Continued)
NUMBER TITLE
VI-7 Energy Use at New Source Pharmaceutical Plants
To Attain NSPS Option A and NSPS Option B
Effluent Levels
VI-8 Solid Waste Generation at New Source
Pharmaceutical Plants To Attain NSPS Option A and
NSPS Option B Effluent Levels
Section VII
VII-1 Proposed Conventional Pollutant NSPS for the
Pharmaceutical Manufacturing Point Source
Category
PAGE
51
53
56
-------
-------
SECTION I
CONCLUSIONS
GENERAL
The Environmental Protection Agency (EPA) is proposing regulations
that would limit the discharge of five-day biochemical oxygen demand
(BOD5.) and total suspended solids (TSS) into waters of the United
States by new sources in four subcategories of the pharmaceutical
manufacturing point source category. This document addresses new
source performance standards (NSPS) for conventional pollutants
required under the Clean Water Act.
NEW SOURCE PERFORMANCE STANDARDS (NSPS)
The technology basis of proposed NSPS for control of BOD5. and TSS is
advanced biological treatment (i.e., biological treatment with longer
detention time than considered as the basis of best practicable
control technology currently available (BPT)) in combination with
effluent filtration. Proposed NSPS are shown in Table 1-1 .
-------
TABLE 1-1
PROPOSED CONVENTIONAL POLLUTANT NSPS
FOR THE
PHARMACEUTICAL MANUFACTURING POINT SOURCE CATEGORY
Pollutant
Subcategory
A-Fermentation
B-Extr"ction
C-Chenrical Synthesis
D-Mixing/Compounding
and Formulation
Maximum
30-Day Average
76.8 mg/1
11.2 mg/1
76.8 mg/1
BOD5
Daily
Maximum
115.0 mg/1
40.7 mg/1
115.0 mg/1
TSS
11.2 mg/1 40.7 mg/1
Maximum
30-Day Average
193.0 mg/1
26.5 mg/1
193.0 mg/1
26.5 mg/1
Daily
Maximum
491.0 mg/1
58.9 mg/1
491.0 mg/1
58.9 mg/1
-------
SECTION II
INTRODUCTION
PURPOSE AND AUTHORITY
The Federal Water Pollution Control Act Amendments of 1972 (P.L.
92-500; the Act) established a comprehensive program to "restore and
maintain the chemical, physical, and biological integrity of the
Nation's waters" (see Section 101(a)). New industrial direct
dischargers were required to comply with new source performance
standards (NSPS), established under authority of Section 306, based on
the best available demonstrated technology.
Although Section 402(a)(l) of the 1972 Act authorized the setting of
requirements for direct dischargers on a case-by-case basis in the
absence of regulations, Congress intended that, for the most part,
control requirements would be based on regulations promulgated by the
Administrator of EPA. Sections 304(c) and 306 of the Act required
promulgation of regulations for NSPS. Section 501(a) of the Act
authorized the Administrator to prescribe any additional regulations
"necessary to carry out his functions" under the Act.
As a result of the Settlement Agreement in Natural Resources Defense
Council, Inc, v. Train, 8 ERC 2120 (D.D.C. 1976), modified, 12 ERC
1833 (D.D.C. 1979), modified by Orders dated October 26, 1982, and
August 2, 1983, the Clean Water Act was amended in 1977 to strengthen
the Agency's toxic pollutant control programs. The Settlement
Agreement did not impact NSPS for conventional pollutants.
SCOPE OF THIS RULEMAKING
On November 26, 1982, EPA proposed regulations applicable to the
pharmaceutical manufacturing point source category (47 FR 53584). At
that time, EPA (a) proposed to modify the existing BPT TSS effluent
limitations for three subcategories (subcategory B—extraction
products, subcategory D—mixing/compounding and formulation, and
subcategory E-—research), (b) proposed BPT TSS effluent limitations
for two subcategories (subcategory A—fermentation products/ and
subcategory C—chemical synthesis products, (c) proposed to modify the
existing BPT effluent limitations for BOD5_ and COD for subcategories
A, B, C, D, and E, (d) proposed BPT and BAT effluent limitations,
NSPS, PSES, and PSNS for cyanide to apply uniformly to subcategories
A, B, C, and D, (e) proposed BAT limitations and NSPS for chemical
oxygen demand (COD) to apply uniformly to subcategories A, B, C, arid
D, (f) proposed BCT effluent limitations for BOD5_, TSS, and pH to
apply uniformly to subcategories A, B, C, and D, and (g) proposed NSPS
for BOD5_, TSS, and pH to apply uniformly to subcategories A, B, C, and
D, based on the application of advanced biological treatment (i.e.,
biological treatment systems with longer detention times than those
considered as the basis of effluent limitations reflecting the best
practicable control technology currently available (BPT)).
-------
Simultaneously with publication of this proposed development document,
the Agency is promulgating regulations covering most aspects of the
November 1982 proposal. In brief, EPA is promulgating BPT effluent
limitations for TSS for subcategories A and C and is modifying
existing BPT BOD5., COD, and TSS effluent limitations for subcategories
B, D, and E. The Agency is also establishing BPT and BAT effluent
limitations guidelines, NSPS, PSES, and PSNS controlling cyanide
discharges from pharmaceutical plants in subcategories A, B, C, and D.
EPA has not addressed best conventional pollutant control technology
(BCT) because the BCT methodology has not yet been issued. The BCT
methodology and BCT limitations for the pharmaceutical industry will
be published at a later date. EPA also has not promulgated final BAT
effluent limitations and NSPS for COD because the Agency needs more
information on the identity of pollutants that contribute to COD and
on applicable COD removal technologies.
The remaining issue to be addressed is NSPS for conventional
pollutants. In commenting on the November 1982 proposal, the industry
complained that new sources in subcategories A and C could not meet
the proposed NSPS because the Agency's proposed subcategorization
scheme was incorrect and because the data base used to develop
proposed NSPS contained too many low raw waste load (subcategory D)
facilities. They also contended that percent reduction-based
standards are more appropriate than concentration-based standards
because of the wide variation in the raw waste characteristics of
pharmaceutical plant discharges.
The Agency's review of the data used to develop the November 1982
proposed NSPS indicated that subcategory D plants did indeed dominate
the data base. EPA analyzed all available data, including new data
submitted with comments, and found that fermentation (subcategory A)
and chemical synthesis (subcategory C) plants have higher conventional
pollutant raw waste loads than extraction (subcategory B) and
formulation (subcategory D) plants. (See Section IV of the
Development Document for Effluent Guidelines, New Source Performance
Standards, and Pretreatment Standards for the Pharmaceutical
Manufacturing Point Source Category (U.S. EPA, September 1983),
hereafter, "final development document). (1) Additionally, the Agency
was aware that permitting authorities and the regulated industry were
familiar with the original subcategorization scheme and the format of
the Code of Federal Regulations. Therefore, as explained more fully
in the final development document, EPA decided to maintain the
original BPT subcategorization scheme.
After proposal, EPA identified four pharmaceutical plants which added
effluent filtration systems to advanced biological treatment systems.
Conventional pollutant discharges from these plants are significantly
lower than from plants where only advanced biological treatment is
employed. Consequently, the Agency believes that the addition of
effluent filtration to advanced biological treatment is a technology
option which must be considered in establishing NSPS for conventional
pollutants in this industry.
-------
The public had not yet had an opportunity to provide comments on
Agency estimates of the costs of the addition of effluent filtration
or on the additional effluent reduction benefits of filtration
technology when applied at new source pharmaceutical plants.
Therefore, EPA determined that it would be appropriate to propose
rather than promulgate NSPS for conventional pollutants based on this
model treatment technology. After reviewing all available data, as
explained in Section VII, EPA determined that effluent filtration is
the appropriate technology basis of NSPS and decided to propose NSPS
based on the combination of advanced biological treatment and effluent
filtration.
The Agency continues to believe that concentration-based standards are
appropriate as the basis for NSPS in the pharmaceutical industry.
Available data on the application of advanced biological treatment and
effluent filtration indicate that industry is capable of designing and
operating end-of-pipe treatment systems that will achieve the
concentration-basedstandards specified in Sections I and VII of this
document.
SUMMARY OF METHODOLOGY
EPA's implementation of the Act required a complex development
program, described in detail in the Proposed Development Document for
Effluent Limitations Guidelines and Standards for the Pharmaceutical
Point Source Category (U.S. EPA, November 1982), hereafter, proposed
development document.(2) First, EPA studied the pharmaceutical
industry to determine the impact of raw material usage, final products
manufactured, process equipment, size and age of manufacturing
facilities, water use, and other factors on the level of conventional
pollutants discharged from plants in this industry. This required the
identification of raw waste and final effluent characteristics,
including the sources and volumes of water used, the manufacturing
processes employed, and the sources of pollutants and wastewaters
within the facility.
EPA then identified all subcategories for which NSPS should be
proposed. The Agency characterized the raw waste conventional
pollutant discharges from plants in these subcategories. Next, EPA
identified several distinct control and treatment technologies which
are in use or capable of being used to control conventional pollutants
in pharmaceutical industry wastewaters. The Agency compiled and
analyzed historical and newly-generated data on effluent quality
resulting from the application of these technologies. The long-term
performance, operational limitations, and reliability of each of the
treatment and control technologies were also identified. In addition,
EPA considered the non-water quality environmental impacts of these
technologies, including impacts on air quality, solid waste
generation, and energy requirements.
The Agency then estimated the costs for each control and treatment
technology from unit cost curves developed by standard engineering
analysis as applied to the specific pharmaceutical industry wastewater
-------
characteristics. EPA derived unit process costs from model plant
characteristics (flow, pollutant raw waste loads) applied to each
treatment process unit cost curve (i.e., primary clarification,
activated sludge, filtration). These unit process costs were combined
to yield the total installed equipment cost at each treatment level.
Total capital costs were then derived from the installed equipment
costs. After confirming the reasonableness of these cost estimates,
the Agency evaluated the economic impacts of these costs. The
economic analysis is the subject of another document: Economic
Analysis of Effluent Standards and Limitations for the Pharmaceutical
Industry (U.S. EPA, September 1983). (3)
Upon consideration of these factors, EPA identified the combination of
control and treatment technologies that reflect the best available
demonstrated technology (NSPS). The proposed regulations, however, do
not require installation of any particular combination of
technologies. Rather, they require achievement of effluent
limitations representative of the proper application of these
technologies or equivalent technologies.
-------
SECTION III
DESCRIPTION OF THE INDUSTRY
INTRODUCTION
Pharmaceutical plants manufacture biological products, medicinal
chemicals, botanical products, and other pharmaceutical products. EPA
identified 466 operating facilities involved in the manufacture of
pharamceutical products. Most of the pharmaceutical industry is
located in the eastern half of the United States. The most prevalent
manufacturing operation in the industry is the formulating, mixing,
and compounding operation; batch-type production is the most common
type of manufacturing technique for this industry.
The wastewaters produced and discharged by the pharmaceutical industry
are very diverse. Plant size, products, processes, and materials to
which wastewater is exposed vary greatly. Additionally, the ratio of
finished product to the quantity of raw materials, solvents, and other
processing materials is generally very low. A detailed discussion of
the pharmaceutical industry is included in Section III of the final
development document and in Section III of the proposed development
document.(1)(2)
SUBCATEGORIZATION
As described in Section II of this document, the Agency is maintaining
the original BPT subcategorization scheme, under which the
pharmaceutical manufacturing industry was segmented into the following
five subcategories:
Subcategory A: Fermentation Products
Subcategory B: Extraction Products
Subcategory C: Chemical Synthesis Products
Subcategory D: Mixing/Compounding and Formulation
Subcategory E: Research
A detailed description of the manufacturing processes and raw
materials used in each of these subcategories is presented in Sections
III and IV of the proposed development document and in the final
development document.
EPA is not proposing NSPS for the research Subcategory (Subcategory E)
because pharmaceutical research does not involve production, nor does
research generate wastewater in appreciable quantities on a regular
basis. Additionally, pharmaceutical research is not mentioned in the
-------
Settlement Agreement. For these reasons, EPA focused its
the four production subcategories.
studies on
As discussed in Section II of this document, commenters on the
November 1982 proposed NSPS contended that different standards should
apply to high raw waste load plants in subcategories A and C than to
low raw waste load plants in subcategories B and D. Some commenters
submitted new data to support their contentions. EPA added these new
data to the existing data base.
The Agency's analyses of the most recent data, including the new data
submitted with comments, indicate that the subcategorization scheme
for this industry should separate fermentation and chemical synthesis
plants (subcategory A and C plants) from extraction and formulation
plants (subcategory B and D plants). Specifically, EPA's analyses
show that usually the influent and effluent conventional pollutant
concentrations and discharge flows of subcategory A and C plants are
similar. The Agency also found that these characteristics for
subcategory B and D plants are similar. However, EPA found that the
characteristics of the subcategory A and C plant group are not similar
to the corresponding characteristics of the subcategory B and D plant
group. Because conventional pollutant raw waste characteristics are
similar for subcategory A and C plants, the Agency believes that
conventional pollutant NSPS for those plants should be identical. For
the same reason, conventional pollutant NSPS for subcategory B plants
should be identical to those for subcategory D plants.
EXISTING END-OF-PIPE TREATMENT AT PHARMACEUTICAL PLANTS
Table III-l presents information on the methods of wastewater
discharge employed at the 466 pharmaceutical manufacturing plants in
the Agency's data base. At 12 percent of the plants, wastewater is
treated on-site in a treatment system operated by plant personnel and
discharged directly to waters of the United States. At 59 percent of
the pharmaceutical facilities, wastewater is discharged to a publicly
owned treatment works (POTW). At 29 percent of the pharmaceutical
plants, wastewater is not generated or all of the wastewater that is
generated is not discharged to navigable waters.
Table II1-2 presents information on the types of treatment currently
in-place at direct discharging pharmaceutical plants. Seventy-five
percent of the direct discharging plants in the industry utilize
biological treatment, and 16 percent of the direct discharging plants
employ filtration systems in addition to biological treatment.
-------
TABLE III-l
SUMMARY OF METHOD
OF DISCHARGE AT
PHARMACEUTICAL PLANTS
Method of Discharge
Direct Dischargers
Indirect Dischargers
Zero Dischargers
No. of Plants
55
277
134
Total Plants
466
TABLE III-2
IN-PLACE TREATMENT TECHNOLOGY AT
DIRECT DISCHARGING PHARMACEUTICAL PLANTS
Treatment Technology
Biological Treatment
Biological Treatment Plus Filtration
Physical Chemical
Unknown
No. of Plants
38
8
3
2
Total Plants 51*
* 4 direct discharging plants primarily produce products other than
Pharmaceuticals and, therefore, have not been included in the data
base.
-------
-------
SECTION IV
WASTE CHARACTERIZATION
INTRODUCTION
The Agency conducted an extensive data gathering effort and developed
qualitative and quantitative information on the characteristics of the
wastewaters discharged by the pharmaceutical industry. This section
summarizes available information on the characteristics of raw waste
and final effluent discharges from direct discharging pharmaceutical
plants. Only conventional pollutant data are presented in this
document.
WASTE CHARACTERIZATION
Table IV-1 presents
effluent BOD5_ and
plants.
a summary of
TSS data for
available raw waste and final
direct discharging pharmaceutical
DEVELOPMENT OF MODEL PLANT RAW WASTE CHARACTERISTICS
As shown in Table IV-1, EPA was able to determine applicable long-term
average BPT BOD5_ and TSS effluent levels for 27 of the 51 direct
discharging plants in the Agency's data base. The Agency identified
best performing plants by comparing actual effluent levels of BOD5_ and
TSS discharged from pharmaceutical plants to the long-term average
BOD5_ and TSS levels that form the basis of BPT effluent limitations.
EPA defined best performers as those plants that meet both the BPT
BOD5_ and TSS effluent levels.
Plants 11111, 12022, 12026, 12036, 12132, 12161, 12236, 33333, and
55555 are best performing subcategory A and C plants employing
biological treatment; plant 12161 also employs effluent filtration in
combination with biological treatment to effect a further removal of
BOD5_ and TSS. As explained in the footnotes on Table IV-1, at
present, sufficient data are not available for plants 11111, 33333,
and 55555 to characterize properly their final effluent BOD5_ and TSS
concentrations. Plants 12015, 12053/12117, 12317, 12459, 12463, and
44444 are best performing subcategory B and D plants. Plants 12015,
12117, 12459, and 12463 employ biological treatment; plants 12053,
12317, and 44444 employ effluent filtration technology in combination
with biological treatment.
Tables IV-2, IV-3, and IV-4 present raw waste characteristics of
subcategory A and C best performers for which sufficient data are
available to characterize properly their final effluent
characteristics, of subcategory B and D best performers employing
biological treatment, and of subcategory B and D best performers
employing effluent filtration, respectively.
11
-------
co cr> t-. r— cr> o to o CM r— r—
CO
co
o *j- co co r~ CM
i — co co r** i —
CM
i en co to co CM
CM i — i — en
CO
o CM en ft- to in o
to o CM in •— o
i — i — in i — co to
• CM i— O CO O CM O
en co to en eo co o
en en tr— i
"tf" o
co to en i-~ o
to to i — o to
O
co
to t— enr— ir> co
i en «3- ro P~. CM o
co to CM vo i— in
co to cvi ,—
i— in
• •
i co i i in
tO CM
CM
o co i—
r— CO CO
in CM
O
CO CO
CO CM O i— CM Ol
i co «i- r-~ r~ to 01
CM ,— to in to CM
CM CO i—
C Ol to
O •* CO
CO ,—
>
•—«
IU
00
CO
in
§
CC"
r~ o> i co
co in co
c co
to to O CO O O
CM r— O O CM CTl
CM ,— tO If! to CM
CM CO r—
i o in i—
in r— CM
r- tO •*
O CM to
• • •
O Or-
8CT^ tC
«* CO
CO r-
CO
II
CMOi—p~p—coi—CMtooini—o«a-«d-toOi—Tj-r—oiin
oootccc' ~
^
___ __ _ __ ___ QQ £\J fSJ
Or^- OCOr— ^^-OCOmOOCOr— OOC*4i— OC^OO^
OOOOOf— Or-COCJOOOOOOCMOOOCSr-^
to
CO
o
r— O i — i — i — r— I — r— CMOr— I — CO CO i— i —
CO CO
•t M
i — i — i — CM CM f— CM
cj a a co i
a
o
tn
: co o o a a o o
M M M fl
-------
co
CO
o
in
o
in
o o
1 O CO
in i —
o o co
co to CM
r— CM co
000
i en o
CO
o
c
r-~ co to
tO CM 01
i— co
in
o o o
en r~ to
in «*•,—
o o
1 CO «*•
co r-.
in
1 CO
• en
t
IU
IT
< O
cotnr~-oooooo
- in,— i r-» o icor»in«*oiniococo
- r— i— CO i— r— CM CM r— «*"
. «C 00
I— 00
to r— ir> in
o o
CO IT)
CM
i en
10
CM
to o m to
CM O CM CM
CO
in
CM
CM
CO
ir
Q
in crt
vc CM
in o o i in to in in
r- O CM r- IT- r-
«*•
•
CO
CM
I
OO
OO
O »3- O O 10 CO
r~ r-^ o c co r~
^o «i- o CM in co
CM «3- cr>
i i i i «*• i i
in
o.
o
CM «*
in en
wo CM
$
i o O
. O CM
cn in CM o
i
-------
CM CTI CM
I i— IO
CM
LU
IT
•a: o
o o in
• • •
i— CO CO
S
o
,— CO «3-
.— CO CTI
CO CM
CO
I
in co en
i— co «i-
i— CO CTt
CO CM
in
01
o
o
ra
4-1
l»-
o
. -o 0)
•o o +J
O -r- IO
•r- C_ E
C- 01 -i-
O) O. X
a. o
co
o s ai
0)
o
•u
o
e
^- *^ r— 'r-
Q 1 ,_ CM r- «3-
f-3 O ,— O r-^
»-^| O O O C
4->
cz
(1)
s
ra
o
£
r— r— OJ I—
£?
o
O)
0)
ra
£
3
in
•a
01
t_
O
>r_
U-
,,j
ai
^_
<4-
LU
C CZ
0) 0)
E E
0.4-1 4-1
01 ra ro
o£
C
ro
E
o co ^J- in
^^ co "ct" in
co co «3- in
o co •* in
CM co ^ in
4-1
in
01
•o
o
o
4-1
cz
01
E
+3
ra
01
c_
I—
c
ai
§
ra
(U
L.
(—
z:
c
I— H-
re ro
U U
o o
*o "o
CO (5
II II
1 «
,
£
ta-
il
r- CM CO
V)
CZ
o
4-1
ra
4-1
2
•a
c:
ra
. c
o
4-1
ro
•r-
>
01
L.
.O
s
ai
i/i
%
ra
cc
II
3
C£
CZ
ro
JZ^
4-1
in
in
a)
52
01
tzn o
ro o
u
CU T3
> (U
C.
Eo
Q.
(U
0
o
t
o
a.
ai
u
re
a
n
•X
C-
4->
0
U
in
CZ
o
•r-
re
ai
CiXI
0 01
CZ OJ
ra O-
jzr
4-> J=
in c
in o
0 40
O 3
O
•a aj
t- o. O U
4J
ra ra
O 4-1
.£> ra
3 "O
o
a z:
II II
a.
0) ra
r.
re 01
4-> >
ra O
c, u
n
*
14
-------
TABLE IV-2
RAW WASTE CHARACTERISTICS OF SUBCATEGORY A AND C BEST PERFORMERS
Raw Waste Characteristics
Plant
12022
12026
12036
12132
12161
12236
Average
Subcategory
A, C
C
A
A, C
A, C, D*
C
Flow (M6D)
1.448
0.161
1.092
0.981
1.925
1.007
1.078
BODR (mg/1)
2142
3670
1571
3000
1362
1652
2233
TSS (mg/1 )
N.A.
88
1059
1150
422
N.A.
680
N.A. = Not available
*About 2 percent of the total wastewater discharge flow results from
formulation operations.
15
-------
TABLE IV-3
RAW WASTE CHARACTERISTICS OF SUBCATEGORY B AND D
BEST PERFORMERS EMPLOYING BIOLOGICAL
TREATMENT
Raw Waste Characteristics
Plant
12015
12117
12459
12463
Average
Subcategory
D
B, D
D
B, D
Flow (MGD)
0.101
0.101
0.049
0.056
0.077
BODR (mg/1)
233
35
70
102
118*
TSS (mg/1)
124
N.A.
59
N.A.
103*
N.A. = Not available
*Flow-weighted average
-------
TABLE IV-4
RAW WASTE CHARACTERISTICS OF SUBCATEGORY B AND D
BEST PERFORMERS EMPLOYING BIOLOGICAL
TREATMENT AND EFFLUENT FILTRATION
Raw Waste Characteristics
Plant
12053
12317
44444
Average
Subcategory
D
D
D
Flow (MGD)
2.50
0.74
0.016
1.085
BODR (mg/1)
299
1004
333
459*
TSS (mg/1)
383
41.4
270
305*
*Flow-weighted average
17
-------
Subcateqory A and C Plant Group
As shown in Table IV-2, plant 12161, the only best performing plant in
the subcategory A and C group employing filtration technology, has
relatively low raw waste BOD5_ concentrations compared to the other
subcategory A and C best performers. Rather than base model plant raw
waste characteristics solely on this plant, EPA averaged the BOD5_ and
TSS raw waste concentrations for all six best performers to develop
NSPS model plant raw waste characteristics. These are shown in Table
IV-2. The BOD5_ and TSS raw waste concentrations are 2230 mg/1 and 680
mg/1, respectively.
Subcategory B and D Plant Group
By comparing Tables IV-3 and IV-4, it is apparent that flow-weighted
average BOD5_ and TSS raw waste concentrations at best performing
subcategory B and D plants employing biological treatment are
considerably lower than for best performers employing biological
treatment and effluent filtration. EPA averaged the BOD5_ and TSS raw
waste concentrations for the best performing plants employing
filtration in combination with biological treatment to develop the
model new source subcategory B and D plant. This ensures that the
entire range of raw waste BOD5_ concentrations that exist within the B
and D subcategories are represented by the model plant. The model new
source subcategory B and D BOD5_ and TSS raw waste concentrations are
459 mg/1 and 305 mg/1, respectively, as shown on Table IV-4.
Summary
Table IV-5 presents the BOD5_ and TSS raw waste characteristics for the
new source model plants representative of the subcategory A and C and
the subcategory B and D plant groups. Estimates of the cost of the
application of conventional pollutant control options and of the
non-water quality implications of these options are based, in part, on
these raw waste characteristics.
18
-------
TABLE IV-5
NEW SOURCE MODEL PLANT
RAW WASTE CHARACTERISTICS
Subcategory A and C
Plant Group
Subcategory B and D
Plant Group
Raw Waste Characteristics (mg/1)
BODc TSS
2233
459
680
305
19
-------
-------
SECTION V
DEVELOPMENT OF CONTROL AND TREATMENT OPTIONS
INTRODUCTION
The basis for new source performance standards (NSPS) under Section
306 of the Act is the best available demonstrated technology. At new
plants, the opportunity exists to design the best and most efficient
pharmaceutical manufacturing and wastewater treatment technologies.
Therefore, Congress directed EPA to consider the best demonstrated
process changes, in-plant controls, and end-of-process treatment
technologies that reduce pollution to the maximum extent feasible. As
a result, limitations for NSPS should represent the most stringent
numerical values attainable through the application of demonstrated
control technology for all pollutants (conventional, nonconventional,
and toxic).
As explained in Section II, after proposal, EPA identified four
pharmaceutical plants which have added filtration systems to advanced
biological treatment systems to control further the discharge of the
conventional pollutants BOD5_ and TSS. As shown in Table IV-1,
conventional pollutant discharges from these plants are significantly
lower than from plants where only advanced biological treatment is
employed. Consequently, the Agency believes that the addition of
filtration to advanced biological treatment is a technology option
which must be considered in establishing NSPS for conventional
pollutants in this industry. Therefore, in addition to the technology
option that formed the basis of the November 1982 proposed NSPS (i.e.,
advanced biological treatment), EPA considered a second option —
effluent filtration in combination with advanced biological
treatment — as a possible basis for NSPS controlling conventional
pollutant discharges from new source pharmaceutical plants.
CONTROL AND TREATMENT OPTIONS
In Section IV, data are presented on the actual conventional pollutant
removals that are being acheived at individual direct discharging
pharmaceutical plants. In addition to these data, EPA received data
for one pharmaceutical plant, plant 12161, that can be used to
estimate the BOD5 and TSS removal that occurs through the application
of filtration technology subsequent to advanced biological treatment.
These data are summarized in Table V-l. EPA's analysis of these data
indicates that about 5.5 percent BOD5_ removal and 29.2 percent TSS
removal is achieved at plant 12161 through the application of effluent
filtration technology. EPA relied on the data presented in Table IV-1
and in Table V-l to determine the conventional pollutant removal
capabilities of the two technology options considered for control of
BOD5_ and TSS at new source direct discharging pharmaceutical plants.
[NOTE: In the preamble to the 1983 proposed NSPS for BOD5 and TSS, EPA
is requesting additional data on the conventional pollutant removal
21
-------
TABLE V-l
CONVENTIONAL POLLUTANT REMOVAL AT PLANT 12161
THROUGH THE APPLICATION OF EFFLUENT FILTRATION TECHNOLOGY
No. of
Observations
BODt; TSS
Long-Term Average
Effluent Characteristics (mg/1)
BODc TSS
Long-term Average Biological
Treatment Effluent (mg/1)1 90 157
Long-term Average Filtration
Effluent (mg/1) 191 319
Pollutant Removal Through
Application of Filtration
26. 082
24. 643
5.5%
37. 072
26. 253
29.2%
lEstimated filtration influent, based on final effluent values prior to
installation of the filtration system.
2Data are for the period 1/1/80 to 7/31/80; raw waste BOD5 was 1279 mg/1
during that timeframe.
3Data are for the period 8/1/80 to 12/31/81; raw waste BOD5 was 1402 mg/1
during that timeframe.
22
-------
capability of filtration technology when applied to pharmaceutical
effluents from biological treatment systems. EPA intends to use these
data to confirm the accuracy of the conventional pollutant removals
shown in Table V-1.]
NSPS Option A
Base NSPS controlling BOD5_ and TSS on the performance of the best
plants employing advanced biological treatment. This option is
identical to the technology option selected for the November 1982
proposal. This would require that specific concentration-based limits
be met. Standards for extraction (subcategory B) and formulation
(subcategory D) plants would be identical. Standards for fermentation
plants (subcategory A) would be the same as those for chemical
synthesis plants (subcategory C).
Tables V-2 and V-3 present long-term average final effluent BOD5_ and
TSS concentrations discharged from best performing pharmaceutical
plants employing advanced biological treatment in the subcategory A
and C plant group and in the subcategory B and D plant group,
respectively. For the subcategory A and C plant group, EPA expects
the application of NSPS Option A to attain long-term average BOD5_ and
TSS discharge levels of 70.1 and 130.1 mg/1, respectively. These
values are the weighted averages of the individual plant data
presented in Table V-2, weighted based on the number of data points
available for each plant.
As explained in Section IV, the best performing subcategory B and D
plants employing advanced biological treatment have significantly
lower raw waste BOD5_ concentrations than the subcategory B and D
plants employing effluent filtration in addition to advanced
biological treatment. Because the subcategory B and D plants
employing advanced biological treatment are not representative of the
entire range of raw waste BOD5_ concentrations that exist in
subcategories B and D, EPA did not base its assessment of the removal
capability of NSPS Option A on the data in Table V-3. Rather, EPA
determined the attainable long-term average BOD5_ and TSS effluent
concentrations achieved at subcategory B and D plants (BOD5 = 7.85
mg/1 and TSS = 9.80 mg/1, based on the median level attained at
subcategory B and D plants employing filtration in addition to
advanced biological treatment; see Table V-4) and adjusted these
concentrations based on the BOD5_ and TSS removal that occurs through
the application of filtration at plant 12161. This calculation yields
long-term average effluent concentrations at subcategory B and D
plants for NSPS Option A of:
BOD5 - (7.85 mg/1)/(1-0.055) = 8.31 mg/1
TSS « (9.80 mg/1)/(1-0.292) = 13.84 mg/1
EPA estimated conventional pollutant removals for new source plants
having conventional pollutant raw waste concentrations equal to those
for the model plants developed in Section IV. EPA estimates that a
23
-------
TABLE V-2
FINAL EFFLUENT CHARACTERISTICS OF BEST
PERFORMING SUBCATE60RY A AND C
PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT
No. of
Observations
Plant
12022
12026
12036
12132
12161
12236
Subcategory
A, C
C
A
A, C
A, C, D*
C
BODs
392
44
366
200
249
105
TSS
395
53
364
204
355
105
Long-Term Average
Effluent Characteristics (mg/1 )
BODR
110.24
108.14
33.04
68.58
19.78
155.60
TSS
84.85
283.68
78.12
452.92
31.55 ^"
108.25
NSPS Option A Long-Term
Average Effluent Characteristics:
70.1** 130.0**
*About 2 percent of the total wastewater discharge flow results from
formulation operations.
**Weighted average based on number of observations for each parameter at
each plant.
24
-------
TABLE V-3
FINAL EFFLUENT CHARACTERISTICS OF BEST
PERFORMING SUBCATEGORY B AND D
PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT
Plant
Subcategory
No. of
Observations
BODc TSS
.Long-Term Average
Effluent Characteristics (mg/1)
• BODc TSS
12015
12117
12459
124631
D
B, D
D
B, D
46
39
51
NA
195
51
47
NA
9.70
1.94
3.82
5.70
10.76
16.00
16.74
9.60
NSPS Option A Long-Term
Average Effluent Characteristics:
8.312 13.842
1-Only long-term average effluent concentrations are available for this plant,
not individual data points.
^Based on adjusting effluent concentrations shown on Table V-4 for NSPS Option
B by the BODs and TSS removal that occurs at plant 12161. (See Table V-l.)
25
-------
model new source A or C plant discharging 1.2 million gallons of
wastewater per day (MGD), in complying with NSPS Option A, would
remove 1.47 million pounds of BOD5_ and TSS per year beyond that
removed in complying with BPT. The Agency estimates that a model new
source B or D plant discharging 0.050 MGD, in complying with NSPS
Option A, would remove about 16,000 pounds of BOD5_ and TSS per year
beyond that removed in complying with BPT.
NSPS Option B
Base NSPS controlling BOD5_ and TSS on the performance of the best
plants employing advanced biological treatment in combination with
effluent filtration. This would require that specific
concentration-based limits be met. As with Option A, standards for
subcategory B and D plants would be identical, and standards for
subcategory A would be the same as those for subcategory C.
Table V-4 presents long-term average final effluent BOD5_ and TSS
concentrations discharged from pharmaceutical plants in the
subcategory B and D plant group employing filtration in combination
with advanced biological treatment. For the subcategory B and D plant
group^ EPA expects the application of NSPS Option B to attain
long-term average BOD5_ and TSS discharge levels of 7.85 mg/1 and 9.80
mg/1, respectively. These values are the median long-term averages
for the three subcategory B and D plants employing filtration
technology. As shown on Table V-4, individual daily data are
available for only one plant.
As shown in Table IV-1, at plant 12161, the only subcategory A and C
plant employing both advanced biological treatment and filtration,
BOD5_ raw waste concentrations are in the low end of the range for all
subcategory A and C plants. For this reason, EPA did not base its
assessment of the removal capability of NSPS Option B solely on the
effluent levels attained at plant 12161. Rather, EPA adjusted the
attainable long-term average BOD5_ and TSS effluent concentrations
achieved at subcategory A and C plants through installation of
advanced biological treatment (BOD5 = 70.1 mg/1 and TSS = 130.0 mg/1;
see Table V-2) based on the BOD5_ and TSS removal that occurs through
the application of filtration at plant 12161. This calculation yields
long-term average effluent concentrations for NSPS Option B of:
BOD5. « (70.1 mg/1) (1-0.055) =66.2 mg/1
TSS - (130.0 mg/1)(1-0.292) =92.1 mg/1
EPA estimated conventional pollutant removals for new source plants
having conventional pollutant raw waste concentrations equal to those
for the model plants developed in Section IV. EPA estimates that a
model new source A or C plant discharging 1.2 MGD, in complying with
NSPS Option B, would remove 1.63 million pounds of BOD5_ and TSS per
year beyond that removed in complying with BPT. The Agency estimates
that a model new source B or D plant discharging 0.050 MGD, in
26
-------
TABLE V-4
FINAL EFFLUENT CHARACTERISTICS OF SUBCATEGORY B AND D
PHARMACEUTICAL PLANTS EMPLOYING ADVANCED BIOLOGICAL
TREATMENT AND EFFLUENT FILTRATION
No. of
Observations
Plant Subcategory BODc; TSS
120531 D N.A. N.A.
12317 D 52 262
44444! D N.A. N.A.
NSPS Option B Long-Term
Average Effluent Concentrations:
Long-Term Average
Effluent Characteristics (mg/1 )
BODt; TSS
8.00 2.00
7.85 9.84
3.00 9.80
7.85 . 9.80
N.A. - Not Available
long-term average effluent concentrations are available for this plant,
not individual data points.
27
-------
complying with NSPS Option B, would remove about 17,000 pounds of BOD5.
and TSS per year beyond that removed in complying with BPT.
EFFLUENT VARIABILITY ANALYSIS
Introduction
The quantity of conventional pollutants discharged from wastewater
treatment systems varies daily. EPA accounts for this variability in
deriving standards limiting the amount of a pollutant that may be
discharged. The statistical procedures used by EPA to analyze the
variability of conventional pollutant discharges from the
pharmaceutical industry are described below.
Daily Variability Factors
The daily variability factor is defined as the ratio of the estimated
99th percentile of the distribution of daily pollutant values to the
estimated mean value of the distribution. For a specific pollutant
discharged from a facility, EPA estimated the mean and 99th percentile
from all daily effluent values which were not deleted on the basis of
being erroneous or descriptive of aberrant performance.
In developing daily variability factors, the Agency considered both
parametric (e.g., normal, lognormal) and nonparametric estimation
procedures. In the course of examining the various parametric
approaches and the data, it became apparent that no individual
parametric distributional assumption would apply to all
plant/pollutant data sets. For that reason, the Agency relied on a
nonparametric procedure when enough daily data were available to apply
the procedure and on a 2-parameter lognormal distribution when the
amount of data was not sufficient to utilize the nonparametric
procedure. Nonparametric procedures do not require satisfying
assumptions on the form of the probability distribution of the
underlying data. The specific nonparametric procedure has been used
previously by the Agency to determine daily variability factors for
other industries (e.g., BPT pesticide industry regulations). The
lognormal distribution has also been used with effluent discharge
data, because such data are generally skewed to a few large values and
are bounded in the lower concentration range by zero. This dual
approach provides a consistent methodology which minimizes the number
of statistical assumptions required to analyze the data while
utilizing as much plant data as possible for the treatment
technologies of interest.
The nonparametric procedure estimates the 99th percentile from a set
of daily discharge measurements by determining the smallest ordered
discharge value in that set of values which is greater than or equal
to the population 99th percentile with probability at least 0.5 (i.e.,
for a specified value of n, determine the smallest ordered value Xtj.)
such that P[X(i> > 99th percentile] n .
1 ~ I (n) ('99)
i-j i
28
-------
The smallest ordered discharge value, satisfying this criterion, was
determined by nonparametric methods (see, e.g., J.D. Gibbons,
Nonparametric Statistical Inference, McGraw-Hill, 1971 (4)). An
estimate chosen in this .manner is sometimes referred to as a 50
percent reliable estimate, or 50 percent tolerance level, for the 99th
percentile and is interpreted as the value below which 99 percent of
the values of a future sample of size n will fall with probability
criterion of at least 0.5. Therefore, the nonparametric procedure was
applied only for plant/pollutant data sets with 69 or more
observations. The arithmetic average of a facility's daily effluent
values was used for the denominator of the nonparametric daily
variability factor.
For plant/pollutant data sets with less than 69 daily observations, a
2-parameter lognormal distribution was used to estimate the 99th
percentile and long-term average of the daily variability factor. The
2-parameter lognormal distribution is the probability distribution
whose natural logarithm has a normal distribution, and is
characterized by parameters v and * relative to its logarithmic
distribution. If Yi. = In Xj^, i = 1, . . ., n, then the estimates of the
parameters are £ = 7 (sample mean of the natural logarithms), and
A
n
= [.I (Yi -y)V(n-l)].
The daily variability factor is then calculated as
VF = -
_ e
A
E(X)
where Z = 2.326, the standard normal 99th percentile and
2
n
n^(n-H)
t
2!
is used to determine a minimum variance unbiased estimate of E(X) .
Thirty-Day Average Variability Factors
A 30-day average variability factor (VF30) is defined as the ratio of
the estimated 99th percentile of the distribution of 30-day averages
of daily pollutant values to the estimated long-term mean value. A
30-day average is the arithmetic mean of 30 daily measurements; the
sets of measurements used in determining each monthly average are
assumed to be distinct. The long-term mean is the long-term
arithmetic mean of 30-day averages and is the same as the long-term
mean estimated from the daily pollutant values.
-------
EPA developed the 30-day average variability factors on the basis of a
statistical result known as the Central Limit Theorem (CLT). The
theorem states that, under general and nonrestrictive assumptions, the
distribution of a sum of a number of random variables, say n, is
approximated by the normal distribution. The approximation improves
as the number of terms in the sum increases. The CLT is quite general
in that no particular distributional form is assumed for the
distribution of the individual values. Thus, this approach is also
nonparametric. In most applications (as in determining 30-day
variability factors), the theorem is used to approximate the
distribution of the average of n observations of a random variable.
The result is important because it makes it possible to compute
approximate probability statements about the average in a wide range
of cases. For instance, it is poss-ible to compute a value below which
a specified percentage (e.g., 95 or 99 percent) of the averages on n
observations are likely to fall. Most textbooks state that 25 or 30
observations are sufficient for the approximation to be valid
although, in many cases, 10 or 15 are adequate. In applying the
theorem to the determination of 30-day limitations, one approximates
the distribution of the average of 30 observations drawn from the
distribution of daily measurements.
Various forms of this theorem exist and are applicable for different
situations. A key assumption in the most familiar version of the
Central Limit Theorem is that the individual measurements are
independent. That is, it is assumed that measurements made on
successive days, or any fixed number of days apart, are statistically
independent or not related. This assumption of independence is rarely
satisfied in an absolute sense in effluent data. In many cases,
however, the assumption is satisfied to a degree sufficient to yield a
suitable result. Because many of the facilities used to determine
variability factors were known to have substantial retention periods,
such effluent data can be expected to exhibit some evidence of
dependency in the daily data. The Central Limit Theorem can still be
used to develop 30-day average variability factors in the case of
dependent data. However, some of the necessary calculations must be
modified to account for the dependency, and more samples (i.e., larger
n) may be required for the approximation to be adequate. In the case
of positive dependence (the usual situation with effluent data), the
modification will result in a larger estimate of the variance of the
mean of 30 observations than would be obtained if independence is
assumed. This in turn results in a larger 30-day average variability
factor than would be obtained if independence is assumed.
-------
The technical details of adjusting the variance for the case of data
dependency are presented below. As stated above, the Central Limit
Theorem will still hold for dependent observations with the
modification that the variance must be adjusted to reflect the
dependence among individual daily mearurements. The covariance
between daily measurements is one way to express this dependence; the
most straightforward approach to effect the necessary modification is
to estimate the variance directly including all the appropriate
covariance terms. The variance estimate is based on the following:
Let X,, X2, ..., Xn denote n random variables each with mean i> and
variance )(*«) where k = |i - jj, i * j and />k is the correlation
between measurements k units apart. . Correlation is-another-measure of
dependence and is related to covariance. Regardless of the
distribution of the Xi^, the mean and variance of the average _ n
Xn= 'I
are: i=l
mean
and
n-1
[n + 2V (n - k) Pkl.
k=l
In the case that Xi. and Xj. are independent, the correlation and
covariance between them are zero. Therefore, --var (Xn) =
[n + 0] = j£ which is the well known expression for the variance of
n ""--•' : ' • -•-.-;:.•: -;;:-: •; ; .-.- . -•' •„• . ' -:
a mean of n independent observations.
Given a set of N measurements on the variable X, denoted by X1V"X2,
..., XN, the mean and variance of the average of n dependent
observations of X, denoted by Xn, are estimated by
31
-------
and
n-l
k=l
(n - k)rk]
respectively, where
and
-> N >^2
S2 = I (Xj - u )_
i=l N-l
rk ^ estimate of pk, the correlation between measurements that are k
units apart (k < n)
N-k
~
(Xj. -
In order to estimate the variance of Xn, there must be a sufficient
number of measurements to estimate the n - 1 correlations. In the
case of an average of 30 observations, there are 29 (lag) correlations
that must be estimated. Thirty-day variability factors (incorporating
dependence) were estimated for a plant/pollutant data set only if two
or more pairs were available to estimate each of the necessary 29
correlations. If sufficient data were not available to estimate these
correlations, then the Central Limit Theorem was utilized assuming
independence. A Thus, the 30-day variability factor was calculated as
VF30 =
with Z
where V(X30) was estimated as described above,
^*
2.^26, the standard normal 99th percentile.
DEVELOPMENT OF VARIABILITY FACTORS USED IN DEVELOPMENT OF PROPOSED
NSPS
Tables V-5, V-6, and V-7 present estimates of individual variability
factors for specific pharmaceutical plants, based on the results
obtained from the above described analyses. EPA determined individual
variability factors for best performing pharmaceutical plants
employing (1) advanced biological treatment and (2) advanced
biological treatment plus effluent filtration.
* See Wilks, S.S.,Mathematical Statistics, Wiley & Sons, 1963, p. 552.
32
-------
TABLE V-5
INDIVIDUAL VARIABILITY FACTORS FOR
SPECIFIC SUBCATE60RY A AND C PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT
Plant
12022
12026
12036
12132
12161
12236
Subcategory
A, C
C
A
A, C
A, C, D*
.C
No. of
Observations
BOD 5 TSS
392
44
366
200
249
105
395
53
364
204
355
105
Variability Factors
Daily Maximum Maximum 30-day Average
BODj TSS BOD fi TSS
4.90
4.96
4.33
5.05
3.22
2.69
3.09
3.03
7.98
6.98
6.97
3.88
2.59
1.41
1.78
1.64
1.56
1.22
1.94
1.21
1.71
2.04
2.19
1.32
Weighted Average
Variability Factors
4.29
5.82
1.90
1.89
*About 2 percent of the total wastewater discharge flow results from
formulation operations.
33
-------
TABLE V-6
INDIVIDUAL VARIABILITY FACTORS FOR
SPECIFIC SUBCATEGORY B AND D PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT
Plant Subcategory
12015 D
12117 B, D
12459 D
No. of Variability Factors
Observations Daily Maximum Maximum 30-day Average
BODR TSS BODR TSS BODR TSS
46 195 5.09 5.58 1.43 1.71
39 51 6.37 5.87 1.30 1.34
51 47 6.52 5.36 1.30 1.52
Weighted Average
Variability Factors
5.99
5.60
1.34
1.62
34
-------
Plant
TABLE V-7
INDIVIDUAL VARIABILITY FACTORS FOR
SPECIFIC PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT AND
EFFLUENT FILTRATION
Subcategory
No. of
Observations
BODs TSS
Variability Factors
Daily Maximum Maximum 30-day Average
BODc TSS BODc TSS
12317
12161
D 52 262 5.19 6.01 1.43 2.70
A, C, D* 191 319 1.73 5.33 1.16 2.09
*About 2 percent of the total wastewater discharge flow results from
formulation operations.
35
-------
Advanced Biological Treatment
Tables V-5 and V-6 present variability factors for plants in the
subcategory A and C and subcategory B and D plant groups where
advanced biological treatment is employed. EPA computed variability
factors for use in developing effluent limits for NSPS Option A by
taking weighted averages of the daily maximum and maximum 30-day
average variability factors, weighting the individual factors based on
the number of daily observations available for each plant. The
weighted average variability factors for each plant group are shown on
Tables V-5 and V-6.
[NOTE: In the preamble to the proposed NSPS, EPA is requesting
additional information on the performance of biological treatment
systems in treating pharmaceutical wastes. The Agency intends to use
any new data received with comments on the proposed rules to review
its analysis of the variability associated with advanced biological
treatment systems in treating pharmaceutical industry wastewaters.]
Filtration
Table V-7 presents variability factors for pharmaceutical plants where
advanced biological treatment and effluent filtration are employed.
As shown, EPA received sufficient data to compute individual
variability factors for only two plants, one representative of the
subcategory A and C plant group and one representative of the
subcategory B and D plant group. EPA determined variability factors
for use in developing effluent limits for NSPS Option B for
subcategories A and C based on the variability factors characteristic
of plant 12161 and for subcategories B and D based on the variability
factors characteristic of plant 12317.
[NOTE: In the preamble to the proposed NSPS, EPA is requesting
additional information on the performance of effluent filtration in
further removing BOD5_ and TSS from pharmaceutical effluents. The
Agency intends to use any new data submitted with comments on the
proposed rules to review its analysis of the removal capability and
variability of effluent filtration when used in combination with
advanced biological treatment.]
Summary
Table V-8 presents the variability factors used by the Agency in
developing effluent limits consistent with NSPS Options A and B,
described previously in this section.
36
-------
TABLE V-8
INDIVIDUAL VARIABILITY FACTORS FOR
SPECIFIC PHARMACEUTICAL PLANTS EMPLOYING
ADVANCED BIOLOGICAL TREATMENT AND
EFFLUENT FILTRATION
NSPS Option A*
Daily Maximum Maximum 30-day Avg.
Subcategory BODe; TSS BOD^ TSS
NSPS Option B2
Daily Maximum Maximum 30-day Avg.
BODc TSS BODc TSS
A and C
B and D
4.29
5.99
5.82
5.60
1.90
1.34
1.89
1.62
1.73
5.19
5.33
6.01
1.16
1.43
2.09
2.70
^Advanced Biological Treatment.
2Advanced Biological Treatment Plus Filtration.
37
-------
-------
SECTION VI
COST, ENERGY, AND NON-WATER QUALITY ASPECTS
INTRODUCTION
Previous sections describe the respective NSPS control options that
were considered as the basis for proposed rules. This section
summarizes the cost, energy, and other non-water quality impacts
(including implementation requirements, air pollution, noise
pollution, and solid waste) of the various treatment options as
required by Section 306(b) of the Clean Water Act.
METHODOLOGY FOR DEVELOPMENT OF COSTS
Introduction
This section describes how estimates of the costs of implementation of
the various technology options were developed. The actual cost of
implementing these technology options can vary at each individual
facility, depending on the design and operation of the production
facilities and on local conditions. EPA developed treatment costs
that are representative of costs anticipated to be incurred at new
source direct discharging plants in the pharmaceutical industry. The
methodology for development of costs is summarized below.
Model Plant Approach
EPA estimated the costs of implementation of two NSPS technology
options in order to determine the economic impact that would result
from application of each technology option at new source direct
discharging pharmaceutical plants. EPA based its cost estimates on
the model plant raw waste characteristics presented in Section IV.
EPA selected model plant sizes that are representative of the
anticipated sizes of new plants in the pharmaceutical industry. For
the subcategory A and C plant group, EPA developed .costs for three
process flow rates: a large-sized plant discharging 1.2 MGD, a medium-
sized plant discharging 0.5 MGD, and a small-sized plant discharging
0.1 MGD. For the subcategory B and D plant group, EPA developed costs
for a medium-sized plant discharging 0.05 MGD.
Cost Estimating Criteria
In order to develop cost estimates for the technology options under
consideration as the basis for proposed NSPS controlling conventional
pollutants, criteria were developed relating to capital, operating,
and energy costs. These criteria are presented in Table VI-1. EPA's
estimates are pre-engineering cost estimates and are expected to have
a variability consistent with this type of estimate, on the order of
plus or minus 30 percent.
39
-------
TABLE VI-1
COST ESTIMATING CRITERIA
1. Capital costs are as of 1982:
2. Miscellaneous Construction Costs:
Piping:
Electrical:
Instrumentation:
Site Preparation:
ENR = 3825
20% of installed equipment cost^
14% of installed equipment cost^
8% of installed equipment cost^
6% of installed equipment cost^
3. Engineering and Contingencies are 30% of total installed costs,
including installed equipment, piping, electrical, instrumentation,
and site preparation costs.
4. Annual fixed costs are 22% of capital expenditures.
5. Operation/Maintenance Costs:
Labor:
Maintenance:
Sludge Disposal:
Electricity:
Chemicals:
hydrated lime:
sulfuric acid (66°):
anhydrous ammonia:
phosphoric acid (80%):
chlorine gas:
polymer:
$24,000/man-year including taxes and
fringe benefits2
3% of total capital costs-*
$8.64/cubic yard (non-hazardous)4
$0.046/ki1owatt-hou r5
$51/ton6
$85/ton6
$392/ton4
$618/ton4
$441/ton6
$2.54/1b6
Development Document for Interim Final Effluent Limitations Guidelines
and Proposed New Source Performance Standards for the Pharmaceutical
Manufacturing Point Source Category, U.S. EPA, Washington, D.C.,
December 1976. fFJ"
"National Survey of Professional, Administrative, Technical, and
Clerical Pay, March 1981," U.S. Department of Labor, September 1981. (7)
Proposed Development Document for Effluent Limitations Guidelines and
Standards for the Pharmaceutical Point Source Category, U.S. EPA,
Washington, D.C., November 1982. (2)
Vendor and Supplier Quotations to Environmental Science and Engineering,
Inc., Gainesville, Florida, 1982 and 1983. (8)
"Electric Utility Company Monthly Statement," March 1980 - Forward:
Federal Energy Regulatory Commission, Form 5, as cited in Monthly
Energy Review, U. S. Department of Energy, Energy Information
Administration, DOE/EIA-0035 (81/12), December 1981. (9)
Innovative and Alternative Technology Assessment Manual, U.S. EPA,
Office of Water Program Operations, Washington, D. C., February 1980. (10)
40
-------
All costs presented are in terms of 1982 dollars. Since construction
costs escalate, these estimates may be adjusted through the use of
appropriate cost indices. The most accepted and widely-used cost
index in the engineering field is the Engineering News Record (ENR)
construction cost index. The ENR Index for cost data presented in
this document is 3,825.
Costs for Implementation of NSPS Options
EPA estimated the costs associated with baseline conditions {i.e., a
new source must comply with BPT conventional pollutant limits) and
with two technology options capable of controlling conventional
pollutant discharges from new direct discharging plants in the
subcategory A and C and the subcategory B and D plant groups of the
pharmaceutical industry. To develop the cost estimates, the Agency
primarily relied on information contained in Section VIII of the
Development Document for Interim Final Effluent Limitations Guidelines
and Proposed New Source Performance Standards for the Pharmaceutical
Manufacturing Point Source Category (U.S. EPA, December 1976M6) and
on information contained in Supplement A of the BPT rulemaking record.
EPA first estimated total capital costs using the methodology
described in the 1976 Development Document and Supplement A. These
costs were in terms of May 1976 dollars. Next, the Agency updated
these costs, first to September 1980 dollars and then to 1982 dollars
using the ENR index. These estimates were then adjusted to reflect
solids dewatering based on the application of horizontal belt filters
rather than vacuum filters. Horizontal belt filter costs were derived
from information contained in Innovative and Alternative Technology
Assessment Manual, EPA-430/9-78-009, U.S. EPA, Office of Water
Programs Operations, Washington, D.C., February 1980. (10) EPA
updated unit costs of chemicals, labor, energy, and sludge disposal to
September 1980 dollars and then adjusted the unit costs to 1982
dollars using the ENR index. The Agency used these unit costs (shown
in Table VI-1) to estimate operating and maintenance costs associated
with compliance with baseline conditions and with two technology
options capable of further reducing conventional pollutant discharges
from new source direct discharging pharmaceutical plants.
Baseline; In the absence of nationally applicable NSPS, new source
direct discharging pharmaceutical plants must, at a minimum, attain
BPT limits for BOD5. and TSS. Therefore, BPT is the baseline
condition. Design criteria for the baseline end-of-pipe biological
treatment systems for the subcategory A and C and the subcategory B
and D plant groups are presented in Table VI-2.
NSPS Option A. Base NSPS for BOD5_ and TSS on the performance of the
best plants with advanced biological treatment. Design criteria for
the end-of-pipe biological treatment systems for the subcategory A and
C and the subcategory B and D plant groups are presented in Table
VI-3.
41
-------
TABLE VI-2
DESIGN BASIS OF THE TREATMENT SYSTEMS
EXPECTED TO BE EMPLOYED AT NEW SOURCE
PHARMACEUTICAL INDUSTRY DIRECT DISCHARGERS
TO MEET BASELINE EFFLUENT LEVELS
Wastewater Pumping:
Design flow: Average daily flow
Basis for power cost: 15 m. total dynamic head
Flow Equalization (Subcategory B-D only):
Detention time: 48 hrs; concrete basins for volumes less than
52 cu. m., earthen basins for volumes greater
than 52 cu. m.
Aeration design requirement: 0.77 hp per cu. m.
Diversion Basin (Subcategory A-C only):
Detention time: 48 hrs
Neutralization (Subcategory A-C only):
Detention time: 20 min
Chemical dosage: lime = 4.3 kg/cu. m., acid = 15.3 kg/cu. m.
Flocculator - Clarifiers:
Type: Primary, secondary, and final for Subcategory A-C; secondary
for Subcategory B-D
Overflow rate: 24 cu. m./d/sq. m.
Sidewater depth: 2.1 to 4.0 m.
Activated Sludge Basin:
Number of basins: 2 minimum
Hydraulic detention time: 4 days for Subcategory A-C;
1.06 days for Subcategory B-D
Nutrient feed: BOD applied:N:P = 100:5:1
Aeration design requirements: 1 kg 02/kg BOD5 removed
16 kg 02/aerator h.p./d
Sludge Thickener (Subcategory A-C only):
Sludge loading rate: 29.3 kg/sq. m./day
42
-------
TABLE VI-2
(continued)
DESIGN BASIS OF THE TREATMENT SYSTEMS
EXPECTED TO BE EMPLOYED AT NEW SOURCE
PHARMACEUTICAL INDUSTRY DIRECT DISCHARGERS
TO MEET BASELINE EFFLUENT LEVELS
Aerobic Digester:
Detention time: 20 days
Aerator design requirements: 1.6 kg 02/kg VSS destroyed
0.044 h.p./cu. m.
Solids Dewatering:
Type: Horizontal belt-filter press
Loading rate: 7.1 kg/sq. m./d
Chemical dosage: 3 kg of polymer/kkg of solids
Trickling Filter (Subcategory A-C only):
Loading rate: 0.5 cu. m./sq. m./d
Depth: 3.7 m.
Polishing Ponds (Subcategory A-C only):
Detention time: 2 days
Solids removal: Pumping from multiple bottom draw-offs
Effluent Chiorination:
Detention time: 30 min.
Chemical dosage: 0.1 kg/cu. m.
Primary/Biological Sludge Transportation and Disposal:
Hauling distance: 64 km
Sludge content: Primary and biological digested sludge at 100 kg/cu. m.
Sludge disposal: Sanitary landfill, off-site
NOTE: Subcategory A-C model treatment system based on Subcategory C
model system in 1976 development document. (6)
Subcategory B-D model treatment system based on Subcategory D
model system in 1976 development document. (6)
43
-------
TABLE VI-3
DESIGN BASIS OF THE TREATMENT SYSTEMS
EXPECTED TO BE EMPLOYED TO MEET
NSPS OPTION A EFFLUENT LEVELS
Wastewater Pumping:
Design flow: Average daily flow
Basis for power cost: 15m. total dynamic head
Flow Equalization (Subcategory B-D only):
Detention time: 48 hrs; concrete basins for volumes less than
52 cu. m.j earthen basins for volumes greater
than 52 cu. m.
Aeration design requirement: 0.77 hp per cu. m.
Diversion Basin (Subcategory A-C only):
Detention time: 48 hrs
Neutralization (Subcategory A-C only):
Detention time: 20 min
Chemical dosage: lime = 4.3 kg/cu. m., acid = 15.3 kg/cu. m.
Flocculator - Clarifiers:
Type: Primary, secondary, and final for Subcategory A-C; secondary
for Subcategory B-D
Overflow rate: 24 cu. m./d/sq. m.
Sidewater depth: 2.1 to 4.0 m.
Activated Sludge Basin:
Number of basins: 2 minimum
Hydraulic detention time: 5 days for Subcategory A-C;
1.33 days for Subcategory B-D
Nutrient feed: BOD applied:N:P = 100:5:1
Aeration design requirements: 1 kg 0£/kg BODs removed
16 kg 02/aerator h.p./d
Sludge Thickener (Subcategory A-C only):
Sludge loading rate: 29.3 kg/sq. m./day
44
-------
TABLE VI-3
DESIGN BASIS OF THE TREATMENT SYSTEMS
EXPECTED TO BE EMPLOYED TO MEET
NSPS OPTION A EFFLUENT LEVELS
(continued)
Aerobic Digester:
Detention time: 20 days
Aerator design requirements: 1.6 kg 02/kg VSS destroyed
0.044 h.p./cu. m.
Solids Dewatering:
Type: Horizontal belt-filter press
Loading Rate: 7.1 kg/sq. m./d
Chemical dosage: 3 kg of polymer/kkg of solids
Trickling Filter (Subcategory A-C only):
Loading rate: 0.5 cu. m./sq. m./d
Depth: 3.7 m.
Polishing Ponds (Subcategory A-C only):
Detention time: 2 days
Solids removal: Pumping from multiple bottom draw-offs
Effluent Chi on'nation:
Detention time: 30 min
Chemical dosage: 0.1 kg/cu. m.
Primary/Biological Sludge Transportation and Disposal:
Haul distance: 64 km
Sludge content: Primary and biological digested sludge at
100 kg/cu. m.
Sludge Disposal: Samitary landfill, off-site
NOTE: Subcategory A-C model treatment system based on Subcategory C
model system in 1976 development document. (6)
Subcategory B-D model treatment system based on Subcategory D
model system in 1976 development document. (6)
45
-------
NSPS Option B. Base NSPS for BOD5_ and TSS on the performance of the
best plants employing advanced biological treatment and effluent
filtration (i.e., Option A plus effluent filtration). Design criteria
for the biological treatment systems are the same as for NSPS Option
A. Design criteria for the end-of-pipe filtration systems are shown
in Table VI-4.
Table VI-5 presents capital, operating and maintenance, and total
annual costs of implementation of baseline treatment and
implementation of NSPS Options A and B at model new source
pharmaceutical plants. Table VI-6 presents a summary of detailed cost
estimates for each component of the treatment systems expected to be
used at a new source subcategory A or C plant to comply with baseline
conditions or with NSPS Options A or B.
ENERGY AND NON-WATER QUALITY IMPACTS
Energy Requirements
The implementation of the control and treatment options considered as
the basis of these proposed rules are expected to affect energy demand
at new source pharmaceutical plants. Table VI-7 summarizes Agency
estimates of total energy used at new source direct discharging plants
for the baseline case and after the application of each specific NSPS
option. Total energy is presented in terms of equivalent barrels of
No. 6 fuel oil; purchased electrical energy (kwh) required was
converted to heat energy (BTU) at a conversion of 10,500 BTU/kwh,
which reflects the average efficiency of electrical power generation.
To allow a comparison with overall pharmaceutical industry energy use,
EPA estimated the average total energy consumed by the pharmaceutical
industry based on information in the 1980 Annual Survey of
Manufactures, Fuels and Electric Energy Consumed, Industry Groups and
Industries, MB0(AS)-4.1, U.S. Department of Commerce, Bureau of the
Census. (11) This estimate includes purchased fuels and electrical
energy, distillate and residual fuel oil, and energy generated less
that sold. Based on the survey information, the total energy consumed
by the pharmaceutical industry is equivalent to about 28.8 billion
kilowatt-hours. This is equivalent to about 51.5 million barrels of
No. 6 fuel oil.
Solid Waste Generation
The implementation of the control and treatment options considered as
the basis of proposed rules is expected to result in increased
generation of wastewater treatment sludges. Wastewater treatment
facilities produce both primary and biological sludges that are
usually dewatered prior to disposal. The amount of wastewater
treatment sludge generated depends on a number of conditions
including: 1) raw waste characteristics; 2) the existence, efficiency,
and/or type of primary treatment; 3) the type of biological treatment
system employed; and 4) the existence, efficiency, and/or type of
secondary clarification.
46
-------
TABLE VI-4
DESIGN BASIS OF THE FILTRATION SYSTEM
EXPECTED TO BE EMPLOYED TO MEET
NSPS OPTION B EFFLUENT LEVELS
Filt rat i on:
Type:
Hydraulic Loading:
Backwash Rate:
Multimedia
0.122 cu. m./min/sq. m.
0.813 cu. m./min/sq. m.
for 10 minutes
47
-------
TABLE VI-5
MODEL PLANT COSTS ASSOCIATED WITH
MEETING BASELINE, NSPS OPTION A, AND NSPS
OPTION B BODs AND TSS FINAL EFFLUENT CONCENTRATIONS
Subcategory
Subcategory A and C
1.2 MGD
Baseline
NSPS Option A
NSPS Option B
0.5 MGD
Baseline
NSPS Option A
NSPS Option B
0.1 MGD
Baseline
NSPS Option A
NSPS Option B
Subcategory B and D
0.05 MGD
Baseline
NSPS Option A
NSPS Option B
Costs (Millions of 1982 Dollars)
Capital 0 & M Total Annual
12.846
14.115
15.147
7.678
8.036
8.806
3.387
3.480
3.797
2.026
2.069
2.297
1.496
1.572
1.624
0.651
0.741
0.773
0.244
0.254
0.266
0.121
0.123
0.132
4.322
4.678
4.957
2.340
2.507
2.711
0.989
1.020
1.102
0.567
0.579
0.637
48
-------
TABLE VI-6
SUMMARY OF COSTS FOR TREATMENT
SYSTEM COMPONENTS FOR THE 1.2 M6D
SUBCATEGORY A AND C MODEL NEW SOURCE PLANT
Treatment Component
Baseline
COST ESTIMATES
($1000)
NSPS Option A
NSPS Option B
CAPITAL
1 ) Low lift pump station
2) Neutralization tanks
3) Primary Floc-Clarifier
4) Secondary Floc-Clarifier
5) Final Floc-Clarifier
6) Aeration Basin
7) Sludge Thickener
8) Digester
9) Digester Aerators
10) Aeration Basin Aerators
11) Belt Press
12) Trickling Filter
13) Diversion Basin
14) Polishing Pond
15) Polymer Feed
16) Chi ori nation Facilities
17) Lime Feed
18) H2S04 Feed
19) Primary Sludge Pumps
20) Sludge Transfer
21) Nutrient Addition
22) Recycle Pumps
23) Control Building
24) Flow Measurement
25) Multimedia Filter
Subtotal
Misc. Construction
Engr. & Contingencies
Total (May 1976 Dollars)
Total (1982 Dollars)
125
44
350
350
350
550
98
290
212
468
72
467
47
49
22
65
151
35
16
12
79
13
177
25
NA
4,067
1,952
1 ,806
7,825
12,846
125
44
350
,350
350
670
117
340
340
539
78
467
47
49
24
65
151
35
21
13
79
13
177
25
NA
4,469
2,145
1,984
8,598
14,115
125
44
350
350
350
670
118
350
294
520
79
467
47
49
24
65
151
35
21
13
79
13
177
25
380
4,796
2,302
2,129
9,227
15,147
49
-------
TABLE VI-6
(Continued)
Component
TOTAL ANNUAL
Energy
Labor
Maintenance (@ 0.03)
Sludge Hauling
Chemicals
Capital Recovery (P 22%)
Total (September 1980 Dollars)
Total (1982 Dollars)
Baseline
319
146
334
169
330
2,452
3,750
4,322
COST ESTIMATES
($1000)
NSPS Option A
340
146
367
177
334
2,695
4,059
4,678
NSPS Option B
341
146
394
188
340
2,892
4,301
4,957
50
-------
TABLE VI-7
ENERGY USE AT NEW SOURCE PHARMACEUTICAL
PLANTS TO ATTAIN NSPS OPTION A AND
NSPS OPTION B EFFLUENT LEVELS
Flow
Subcategory (MGD)
Subcategory A and C 1.2
Baseline
NSPS Option A
NSPS Option B
Subcategory B and D 0.05
Baseline
NSPS Option A
NSPS Option B
Energy Requirements for
Wastewater Treatment
(bbl of oil/yr)
13,314
14,190
14,232
313
351
380
Energy Increase
Over Baseline
Wastewater Treatment
Energy Requirements
6.6 %
6.9 %
12.1 %
21.4 %
51
-------
Table VI-8 summarizes Agency estimates of wastewater sludge generation
for the baseline case and after the application of each NSPS
technology option.
Air Pollution and Noise Potential
The technologies under consideration are not a significant source of
noise potential or air pollution. EPA anticipates that implementation
of the control and treatment options under consideration will have no
direct impact on air pollution or noise pollution.
52
-------
TABLE VI-8
SOLID WASTE GENERATION AT NEW SOURCE
PHARMACEUTICAL PLANTS TO ATTAIN NSPS OPTION A
AND NSPS OPTION B EFFLUENT LEVELS
Flow
Subcategory (MGD)
Subcategory A and C 1.2
Baseline
NSPS Option A
, NSPS Option B
Subcategory B and D 0.05
Baseline
NSPS Option A
NSPS Option B
Wastewater Sludge
Generation
(million Ibs/yr)
5.140
5.337
5.343
0.069
0.071
0.071
Sludge Increase
Over Baseline
Wastewater Sludge
Generation
3.8
3.9
2.8
2.8
53
-------
-------
SECTION VII
EFFLUENT REDUCTION ATTAINABLE THROUGH THE
APPLICATION OF NEW SOURCE PERFORMANCE STANDARDS
GENERAL
The basis for new source performance standards (NSPS) under Section
306 of the Act is the best available demonstrated technology. At new
plants, the opportunity exists to design the best and most efficient
production processes and wastewater treatment facilities. Therefore,
Congress directed EPA to consider the best demonstrated process
changes, in-plant controls, and end-of-pipe treatment technologies
that reduce pollution to the maximum extent feasible. As a result,
limitations for NSPS should represent the most stringent numerical
values attainable through the application of demonstrated control
technology for all pollutants (conventional, nonconventional, and
toxic).
IDENTIFICATION OF THE TECHNOLOGY BASIS OF PROPOSED NSPS
The technology basis selected for control of BOD5_ and TSS under
proposed NSPS is advanced biological treatment (i.e., biological
treatment with longer detention time than that considered as the basis
of effluent limitations reflecting the best practicable control
technology currently available (BPT)) in combination with effluent
filtration.
PROPOSED NSPS
Table VII-1 presents proposed NSPS for the conventional pollutants
BODS and TSS at pharmaceutical manufacturing, facilities.
RATIONALE FOR THE SELECTION OF THE TECHNOLOGY BASIS OF PROPOSED NSPS
As discussed in Section V, EPA identified two options that could form
the basis of NSPS controlling the discharge of conventional pollutants
from pharmaceutical manufacturing facilities. EPA based proposed NSPS
on the application of biological treatment and effluent filtration
because filtration is an available, demonstrated technology in this
industry that results in additional conventional pollutant removal
beyond that attained by the application of advanced biological
treatment only.
METHODOLOGY USED FOR DEVELOPMENT OF PROPOSED NSPS
For subcategories B and D, EPA determined attainable long-term average
BOD5. and TSS effluent concentrations resulting from the application of
advanced biological treatment and effluent filtration by analyzing
effluent data from three subcategory B and D plants employing this
combination of end-of-pipe treatment'technologies.
55
-------
TABLE VII-1
PROPOSED CONVENTIONAL POLLUTANT NSPS
FOR THE
PHARMACEUTICAL MANUFACTURING POINT SOURCE CATEGORY
Pollutant
Su beat e gory
A-Fermentation
B-Extraction
C-Chemical Synthesis
D-Mixing/Compounding
and Formulation
Maximum
30-Day Average
76.8 mg/1
11.2 mg/1
76.8 mg/1
BOD5
Daily
Maximum
115.0 mg/1
40.7 mg/1
115.0 mg/1
TSS
11.2 mg/1 40.7 mg/1
Maximum
30-Day Average
193.0 mg/1
26.5 mg/1
193.0 mg/1
26.5 mg/1
Daily
Maximum
491.0 mg/1
58.9 mg/1
491.0 mg/1
58.9 mg/1
56
-------
For subcategories A and C, EPA identified six plants where advanced
biological treatment is employed. The Agency analyzed effluent data
for these six plants and determined attainable long-term average BOD5_
and TSS effluent concentrations resulting from the application of
advanced biological treatment. EPA also determined the BODS^ and TSS
removal capability of effluent filtration based on available data from
the one subcategory A/C plant employing the combination of advanced
biological treatment and effluent filtration. This plant has
relatively low raw waste concentrations compared to other fermentation
and chemical synthesis plants. Rather than propose NSPS based on data
for this one plant, EPA computed long-term average BOD5_ and TSS
effluent concentrations by reducing the attainable long-term average
BODJ5 and TSS effluent concentrations for advanced biological treatment
by the percentage removals of BODJ5 and TSS that occur at the one plant
employing both advanced biological treatment and effluent filtration.
For all four subcategories, EPA calculated maximum 30-day average and
daily maximum limitations by multiplying attainable long-term average
BOD5. and TSS effluent concentrations by appropriate variability
factors, as discussed in Section V of this document.
COST OF APPLICATION AND EFFLUENT REDUCTION BENEFITS
EPA estimates that a model new source subcategory A or C plant
discharging 1.2 million gallons of wastewater per day (MGD), in
complying with proposed NSPS, would remove 1.63 million pounds per
year of BODS^ and TSS beyond that removed in complying with BPT
effluent limitations. The incremental capital and total annual costs
beyond BPT would be $2.30 and $0.64 million, respectively (1982
dollars). EPA estimates that a model new source subcategory B or D
plant discharging 0.050 MGD of wastewater, in complying with NSPS,
would remove about 17,000 pounds per year of BODS^ and TSS beyond BPT.
The incremental capital and total annual costs beyond BPT would be
$271,000 and $70,000, respectively (1982 dollars).
NON-WATER QUALITY ENVIRONMENTAL IMPACTS
Sections 304(b) and 306 of the Act require EPA to consider the non-
water quality environmental impacts (including energy requirements) of
certain regulations. In conformance with these provisions, EPA
considered the effect of these regulations on air pollution, solid
waste generation, and energy consumption, as summarized below.
Implementation of proposed NSPS would not substantially increase air
pollution, energy use, or solid waste generation. The proposed
regulations are not expected to cause any significant air pollution
problems. EPA estimates that compliance with proposed NSPS for
conventional pollutants will increase energy use by less than one
percent at subcategory A or C and subcategory B or D plants.
EPA estimates that, to comply with proposed NSPS, the incremental
solid waste generated at a model new source fermentation (subcategory
A) or chemical synthesis (subcategory C) plant discharging 1.2 MGD of
57
-------
wastewater and a model extraction (subcategory B) or formulation
(subcategory D) plant discharging 0.050 MGD of wastewater will be
approximately 200,000 and 2,100 additional pounds per year of
wastewater treatment sludge, respectively, beyond that generated in
meeting BPT effluent limitations. This is equal to an incremental
increase of about 3.9 percent for subcategory A or C plants and about
3.0 percent for subcategory B or D plants over that generated to meet
BPT effluent limitations. The solid wastes generated through
wastewater treatment at pharmaceutical plants have not been listed as
hazardous in regulations promulgated by the Agency under Subtitle C of
the Resource Conservation and Recovery Act (RCRA) (see 45 FR 33066;
May 19, 1980). Accordingly, it does not appear likely that the
wastewater sludges generated by new source pharmaceutical plants under
the proposed NSPS will be subject to the comprehensive RCRA program
establishing requirements for persons handling, transporting,
treating, storing, and disposing of hazardous wastes. The Agency's
estimates of the costs of this regulation include the cost of handling
these sludges as a non-hazardous waste.
58
-------
SECTION VIII
REFERENCES
1. Development Document for Effluent Guidelines, New Source
Performance Standards, and Pretreatment Standards for the
Pharmaceutical Manufacturing Point Source Category, U.S. EPA,
Washington, D.C., September 1983.
2. Proposed Development Document for Effluent Limitations Guidelines
and Standards for the Pharmaceutical Point Source Category, U.S.
EPA, Washington, D.C., November 1982.
3. Economic Analysis of_ Effluent Standards and Limitations for the
Pharmaceutical Industry, U.S. EPA, Washington, D.C., September
1983.
4. Gibbons, J. D., Nonparametric Statistical Inference, McGraw-Hill,
1971.
5. Wilks, S. S., Mathematical Statistics, Wiley & Sons, 1963.
6. Development Document for Interim Final Effluent Limitations
Guidelines and Proposed New Source Performance Standards for the
Pharmaceutical Manufacturing Point Source Category, U.S. EPA,
Washington, D.C., December 1976.
7. "National Survey of Professional, Administrative, Technical, and
Clerical Pay, March 1981," U.S. Department of Labor,
September 1981 .
8. Vendor and Supplier Quotations to Environmental Science and
Engineering, Inc., Gainesville, Florida, 1982 and 1983.
9. "Electric Utility Company Monthly Statement," March 1980 Forward:
Federal Energy Regulatory Commission, Form 5, as cited in Monthly
Energy Review, U.S. Department of Energy, Energy Information
Administration, DOE/EIA-0035 (81/12), December 1981.
10. Innovative and Alternative Technology Assessment Manual,
11 .
EPA-430/9-78-009,
February 1980.-
U.S. EPA, Office of Water Program Operations,
1980 Annual Survey of_ Manufactures, Fuels and Electric Energy
Consumed, Industry Groups and Industries, M80(AS)-4.1, U.S.
Department of Commerce, Bureau of the Census.
59
-------
-------
SECTION IX
ACKNOWLEDGEMENTS
The U.S. Environmental Protection Agency wishes to acknowledge the
contributions to this project by Environmental Science and
Engineering, Inc., of Gainesville, Florida. The key contributors were
John Crane, Bevin Beaudet, Susan Albrecht, Russell Bowen, Leonard
Carter, and Margaret Farrell. We also wish to thank the following
personnel of the E.G. Jordan Co., of Portland, Maine, for their
assistance: Willard Warren, Conrad Bernier, Robert Steeves, Michael
Crawford, and Neal Jannelle.
The assistance of PEDCo, of Cincinnati, Ohio, is also acknowledged for
their technical input in this project. The efforts of The Research
Corporation of New England (TRC) in developing and maintaining an open
literature data base are also acknowledged.
We wish to acknowledge the plant managers, engineers, and other
representatives of the pharmaceutical industry without whose
cooperation and assistance in site visitions and information
gathering, the completion of this project would have been greatly
hindered. We also thank the environmental committees of the
Pharmaceutical Manufacturers Association for their assistance.
Appreciation is expressed to those at EPA Headquarters who contributed
to the completion of this project, including: Louis DuPuis, Russ
Roegner, and Joseph Yance, Office of Analysis and Evaluation, Office
of Water Regulations and Standards; Alexander McBride and Richard
Healy, Monitoring and Data Support Division, Office of Water
Regulations and Standards, Susan Lepow and Catherine Winer, Office of
General Counsel; Mahesh Podar, Office of Policy and Resource
Management; and Bruce Newton/ Office of Water Enforcement.
Within the Effluent Guidelines Division, Joseph Vitalis, Gregory
Aveni, Glenda Colvin, Kointheir Ok, Carol Swann, Pearl Smith, and
Glenda Nesby made significant contributions to this project.
The assistance of all personnel at EPA Regional Offices and State
environmental departments who participated in the data gathering
efforts is also greatly appreciated.
-------
------- |