United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-79-019
June 1979
Air
Refinery Waste Disposal
Screening Study
-------
EPA-450/3-79-019
Refinery Waste Disposal
Screening Study
by
Ms. Juanita Galloway
Radian Corporation
Suite 125, Culpeper Building
7923 Jones Branch Drive
McLean, Virginia 22102
Contract No. 68-02-2608
Project No. 61
EPA Project Officer: Kerri C. Brothers
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1979
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35), U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina 27711; or for a nominal fee,
from the National Technical Information Service, 5285 Port Royal Road,
Springfield,Virginia 22161.
This report was furnished to the Environmental Protection Agency by the
Radian Corporation, Suite 125, Culpeper Building, 7923 Jones Branch
Drive, McLean, Virginia 22102, in fulfillment of Contract No. 68-02-2608.
The contents of this report are reproduced herein as received from the
Radian Corporation. The opinions, findings, and conclusions expressed
are those of the author and not necessarily those of the Environmental Protec-
tion Agency. Mention of company or product names is not to be considered
as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-79-019
ii
-------
TABLE OF CONTENTS
List of Figures iv
List of Tables v
1.0 INTRODUCTION 1
1.1 Approach 1
1.1.1 Quantify and Characterize Waste ... 1
1.1.2 Waste Control Technology 2
1.1.3 Listing of Solid Waste Operations . 2
1.2 Summary of Results 3
2.0 CONCLUSIONS AND RECOMMENDATIONS 5
2.1 Conclusions 5
2.2 Recommendations 5
3.0 SOLID WASTE GENERATION 6
3.1 Data Collection 6
3.2 Data Compilation 6
3.3 Data Analysis 13
3.4 Comparison with Previous Studies 18
3.5 Extrapolation of Data 21
4.0 HYDROCARBON EMISSIONS 25
4.1 Data Analysis 25
4.1.1 Phase Composition of Waste Streams. 25
4.1.2 Hydrocarbon Content of Refinery
Solid Waste 25
4.1.3 Estimation of VOC Emissions 28
4.2 Data Limitations 30
4.3 Factors Controlling Volatile Emissions ... 30
4.4 Supplementary Data Collection Program .... 34
4.5 VOC Emissions from Other Sources 35
5.0 SOLID WASTE MANAGEMENT PRACTICES 37
5.1 Source Control 37
5.2 Treatment 38
5.3 Final Disposal 39
5.4 Process Descriptions 41
Bibliography 43
11
-------
TABLE OF CONTENTS (Continued)
Appendix I
Appendix II
Appendix III
Appendix IV
Appendix V
Evaluation of Solid Waste Processing and
Disposal Techniques
Page
1-1
Description of Refinery Solid Waste
Streams II-1
Estimation of the Parameters of the Log-
normal Distribution III-l
Selection of Refineries for Supplementary
Sampling Program IV-1
Supplementary Sampling Program V-l
iii
-------
LIST OF FIGURES
Number Page
3-1 Histogram of Solid Waste Generation Data for
API Separator 14
3-2 Histogram of the Logarithms of Solid Waste
Generation Data for API Separator Sludge 16
4-1 Organic Emission Versus Volatility of
Hydrocarbon Phase 29
5-1 Matrix Summary of Control Technologies and
Refinery Waste Streams 42
Appendix
IV-1
Block Flow Diagram for a Representative U.S.
Refinery IV-3
iv
-------
LIST OF TABLES
Number
Page
3-1 Companies Receiving EPA-QAQPS Data Request Forms
and Refineries for Which Forms were to be
Completed 7
3-2 Solid Waste Streams Included in the EPA-OAQPS
Survey 9
3-3 Solid Waste Generation Data 12
3-4 Solid Waste Generation Factors 17
3-5 Comparison of Jacobs and EPA-OAQPS Generation
Factors 20
3-6 Comparison of API and EPA-OAQPS Generation
Factors 22
3-7 Annual Solid Waste Generation of the Entire U.S.
Petroleum Refining Industry 23
4-1 EPA-OAQPS Survey Summary of Reported Phase Data
Water, Solid, and Hydrocarbon Content of
Refinery Solid Waste Streams 26
4-2 EPA-OAQPS Survey Total Hydorcarbons in Refinery
Solid Wastes 27
4-3 Volatile Components of Waste Streams in EPA-
OAQPS Survey 31
4-4 Factors Affecting VOE From Solid Wastes 33
Appendix
IV
1 Comparison of "Representative" Refinery Product
Slate with Total Actual U.S. Production IV-4
2 Refinery Process Unit Capacities "Representa-
tive" Compared to Average of U.S. Refineries ... IV-5
3 Subcategorization of the Petroleum Refining
Industry Reflecting Significant Differences in
Wastewater Characteristics IV-8
4 Refineries Simulated by Cluster Models IV-11
5 PAD Districts and Bureau of Mines Refining
Districts IV-12
6 Summary of Major Refinery Processing Units for
1973 IV-14
-------
LIST OF TABLES (Continued)
Number Page
Appendix
V
1 Solid Waste Streams Number of Samples and
Phases to be Analyzed V-2
2 Refinery Solid Waste Constituents to be
Analyzed V-3
3 Refinery Solid Waste Stream Analyses Currently
Used Methods and Costs V-4
4 Solid Waste Sampling and Analysis Comparison:
Precision Versus Cost V-6
vi
-------
1.0 INTRODUCTION
The Emission Standards and Engineering Division of EPA's
Office of Air Quality Planning and Standards (OAQPS) is currently
investigating the need to control volatile organic carbon (VOC)
emissions from the handling and disposal of wastes from the U.S.
Petroleum Refining Industry. Radian Corporation was contracted
to assist OAQPS in determining the quantities, characteristics,
and fate of solid wastes produced by the refining industry and
to estimate the national VOC atmospheric emissions resulting from
refinery solid waste treatment and disposal. As defined by this
program, solid wastes include all refinery waste streams which
are not controlled by existing wastewater and air emissions regu-
lations.
Radian's objectives in this program were as follows:
• Estimate the quantity and characteristics
of major refinery solid waste streams.
• Compare AP-42 emission factors against the
refinery solid waste data base to determine
if the data base warranted revisions to the
AP-42 factors.
• Describe and evaluate both proven and poten-
tial solid waste control technologies
utilizable by the refining industry.
• Develop a comprehensive listing of operators
involved in the organic wastes industry.
1.1 Approach
1.1.1 Quantify and Characterize Wastes
Data submitted to the OAQPS by eight petroleum refining
companies, in response to an April 1978 request served as the
primary data base for this study. The industry responses were
screened for obvious errors, keypunched and validated. They
-------
were filed on an automated data base on the Statistical Analyses
System to facilitate computation. Statistical models were built
for the solid waste streams generation rates and the most precise
estimate of the mean generation rates. Using the models, esti-
mates of the national waste generation factors and confidence
intervals for the estimates were determined.
1.1.2 Waste Control Technology
Existing and planned techniques used by the refining
industry were identified and described. Identification of these
processes required a survey of the literature, contacts with in-
dustry representatives, and review of industry responses to the
1978 EPA survey of eight refining companies. The technology
search covered methods used for recovery, treatment, and disposal
of refinery solid wastes.
After the solid waste technologies were identified, ad-
vantages and disadvantages of each technique were defined in Ap-
pendix I. The discussion of each technology covered the following
areas:
• Development status
• Performance
• Secondary pollutants
• Cost and energy requirements
1.1.3 Listing of Solid Waste Operations
A listing was developed of names, addresses and tele-
phone numbers of operators involved in the collection, disposal,
treatment and reclamation of organic wastes. These firms accepted
wastes from petroleum refineries, re-refiners, waste solvent re-
claimers, chemical processing plants and degreasing operations.
-------
The list was compiled through use of recent EPA surveys, contracts
with refinery management, state and local agencies, representa-
tives from the disposal industry, and trade associations. The
search focused on geographical areas where the refining industry
is most concentrated. The list was submitted as a separate re-
port during the course of this study.
1.2 Summary of Results
The major finding of this study was that volatile
organic emissions resulting from handling of refinery solid wastes
could not be accurately estimated from the available data. However,
it can be estimated from the OAQPS data base that the total hydro-
carbon content of solid waste streams from United States refineries
is approximately 190,000 tons/year (dry weight). About 75 percent
of the hydrocarbons in the solid wastes are contained in four
waste streams: Dissolved Air Floatation Skimmings, Slop Oil
Emulsion Solids, API Separator Sludge, and Biosludge.
The release of hydrocarbons from the solid wastes is
primarily dependent on volatility. For example, if as much as
10 percent of the total hydrocarbon content were volatile, emis-
sions from refinery solid waste could approach 19,000 tons/yr.
It may be expected that solid waste hydrocarbon volatility would
be low, due to degradation or loss of volatiles in upstream
refining process units and wastewater treatment facilities.
Data on hydrocarbon volatility and on many other factors
which may affect VOC emissions from solid wastes were not obtained
in the present study. Radian did, however, identify and describe
the data gaps and develop a data collection program that would
provide adequate information to estimate VOC emissions. A better
characterization of refinery solid wastes would also aid the
-------
design of treatment, disposal, and resource recovery systems, and
could identify sources of hazardous compounds. A detailed dis-
cussion of this data collection system is provided in the
Appendices.
During the conduct of this study, Radian determined
that in addition to the petroleum refining industry, other indus-
tries could contribute significant quantities of solid organic
wastes. These industries include: petroleum re-refiners, solvent
reclamation, and degreasing. It was recommended that these indus-
tries be included in a comprehensive assessment of organic solid
waste and VOC emissions from the waste streams.
-------
2.0 CONCLUSIONS AND RECOMMENDATIONS
2.1 Conclusions
1. The solid waste generation data collected in the
EPA-OAQPS survey are an adequate basis for only order-of-magnitude
estimates of solid waste generation by the petroleum refining
industry.
2. The existing data on the volatile organic carbon
content of refinery solid waste is not adequate for accurately
estimating the volatile emissions from the solid waste during
processing and disposal.
3. Industries other than the petroleum refining indus-
try generate solid wastes that contain significant amounts of
volatile organic carbons which could be emitted during solid
waste processing and disposal.
2.2 Recommendations
1. A well-designed sampling program to obtain the
data required to develop accurate estimates of refinery solid
waste generation and the volatile emissions from that solid waste
should be initiated.
2. Studies of the volatile emissions from solid waste
generated by industries other than the petroleum refining indus-
try should be initiated.
-------
3.0 SOLID WASTE GENERATION
Estimates of the quantities of solid waste generated by
the petroleum refining industry are needed as a basis for estimat-
ing the atmospheric emissions from refinery solid waste handling
and disposal. EPA-OAQPS recently conducted a survey in which one
of the major objectives was to collect solid waste generation rate
data. The collection, compilation, and analysis of this data are
described in this section. The resulting solid waste generation
factors are compared to generation factors developed in previous
studies. The generation factors are then used to estimate annual
solid waste generation by the petroleum refining industry.
3.1 Data Collection
In April 1978, EPA-OAQPS initiated a survey to gather
information needed to quantify the atmospheric emissions from re-
finery solid waste handling and disposal. Data request forms
were mailed to the companies listed in Table 3-1 with instructions
to complete one form for each refinery listed under the company
name in the table. The data request form contained twenty waste
streams for which quantity, composition, frequency of generation,
storage, transport, and disposal information was sought. The
twenty waste streams of interest are listed in Table 3-2. A
description of most of the waste streams is provided in Appendix
I. Satisfactory responses were received from all of the companies
by September 1978.
3.2 Data Compilation
The solid waste generation rate data were read from the
survey forms and recorded by waste stream. During the compilation
process, it was found that a substantial amount of double account-
ing was occurring. The double accounting resulted from the
-------
TABLE 3-1
COMPANIES RECEIVING EPA-OAQPS DATA REQUEST FORMS
AND REFINERIES FOR WHICH FORMS WERE TO BE COMPLETED
Refinery
Company and Refinery Number
Ashland Petroleum Company (ASH)
Catlettsburg, KY 1
Canton, OH 2
Buffalo, NY 3
St. Paul, MN 4
Continental Oil Company (CONT)
Lake Charles, LA 5
Ponca City, OK 6
Billings, MT 7
Getty Refining & Marketing Company (GETT)
El Dorado, KS 8
Delaware City, DE 9
Marathon Oil Company (MARA)
Texas City, TX 10
Detroit, MI 11
Robinson, IL 12
Garyville, LA 13
Mobil Oil Corporation (MOB)
Paulsboro, NJ 14
Femdale, WA 15
Joliet. IL 16
Beaumont, TX 17
Augusta, KS 18
Buffalo, NY 19
Torrance, CA 20
Phillips Petroleum Company (PHILL)
Kansas City, KS 21
Borger, TX 22
Sweeney, TX 23
-------
TABLE 3-1
COMPANIES RECEIVING EPA-OAQPS DATA REQUEST FORMS
AND REFINERIES FOR WHICH FORMS WERE TO BE COMPLETED (Continued)
Refinery
Company and Refinery Number
Texaco Inc. (TEX)
Lockport, IL 24
Lawrenceville, IL 25
Westville, NJ 26
Anacortes, WA 27
Port Arthur, TX 28
St. James Parish, LA 29
Wilmington, CA 30
Union Oil Company (UNION)
Beaumont, TX 31
Chicago, IL 32
Los Angeles, CA 33
Rodeo, CA 34
8
-------
TABLE 3-2
SOLID WASTE STREAMS INCLUDED IN THE EPA-OAQPS SURVEY
Stream
Stream Number
Slop Oil Emulsion Solids 1
Silt from Storm Water Runoff 2
Exchanger Bundle Cleaning Sludge 3
API Separator Sludge 4
Nonleaded Gasoline Tank Bottoms 5
Crude Tank Bottoms 6
Other Storage Tank Bottoms 7
Leaded Gasoline Tank Bottoms 8
Dissolved Air Flotation Skimmings 9
Kerosene Filter Clays 10
Other Filter Clays 11
HF Alkylation Sludge 12
Waste Bio-Sludge 13
Once-Through Cooling Water Sludge 14
FCC Catalyst 15
Coke Fines 16
Spent Amines 17
Salts from Regeneration 18
Ship and Barge Ballast 19
Other 20
-------
reporting of generation rate data for waste streams that were
subsequently discharged to the sewer system. To illustrate,
suppose a refinery reported generating exchanger bundle cleaning
sludge. The generation rate reported would be recorded under the
exchanger bundle cleaning sludge waste stream. Also, suppose
that this sludge was discharged to the sewer system. As a result,
exchanger bundle cleaning sludge would appear in the wastewater
treatment sludges; thus, the generation rate for exchanger bundle
cleaning sludge would also be included in the generation rate
reported for the wastewater treatment sludge waste streams.
An apparent solution to the double accounting problem
was to continue to record the generation rates of all waste
streams at the point of generation and then to subtract the gen-
eration rates of those waste streams discharged to the sewer
system from the reported generation rates of the wastewater
treatment sludges. Unfortunately, this approach was not feasible
because the fate of a waste stream after it enters a sewer system
cannot be determined. For example, discharging exchanger bundle
cleaning sludge to a refinery sewer system could result in the
sludge reacting with other wastes, being removed from the waste-
water in any or all of several wastewater treatment units, and/or
passing through the treatment process. As a result, the genera-
tion rate of a waste stream that is discharged to the sewer system
could not be subtracted from the generation rates of wastewater
treatment sludges because the disposition of the stream among the
sludges was not known.
Another solution to the double accounting problem was
to record the generation rates of those waste streams discharged
to the sewer system as zero at their point of generation. The
disadvantage of this approach is that the true generation rates
for the waste streams that are discharged to the sewer system
would be under-estimated. Under-estimation of the true generation
10
-------
rate of certain waste streams was considered to be preferable to
over-estimation of the total solid waste generation rate for all
the waste streams; therefore, this approach to eliminating double
accounting was used.
The resulting compilation of solid waste generation
rates by waste stream for all of the refineries surveyed was con-
verted to a dry weight basis and then to a unit capacity basis.
In order to convert the data to a dry weight basis, the composi-
tion data from the survey on moisture content of the waste streams
was compiled. In some cases, data on moisutre content were not
reported where generation rate data had been reported. The only
other available source of refinery solid waste stream moisture
content at the time of this analysis was a study done by Jacobs
Engineering Company (JA-216); therefore, the moisture contents
contained in this report were used where moisture data from the
survey were missing.
Since some refineries in the OAQPS survey did not pro-
vide adequate capacity data, all capacities used in the solid
waste calculations were obtained from the 1977 Annual Refining
Survey published in the Oil and Gas Journal (CA-673). Capacities
on a Barrels Per Calendar Day (BPCD) basis, reflecting actual
volume of oil processed, were used to calculate the most accurate
solid waste generation factors. Then, to permit direct compari-
son of the OAQPS data with the API and Jacobs Engineering studies,
all calculations performed on a BPCD basis were converted to a
Barrels Per Stream Day basis (BPCD = 0.9 x BPSD). The resulting
solid waste generation rate data are presented by solid waste
stream and refinery in Table 3-3.
Three potential biases in the survey were identified
as a result of studying the data during the compilation process.
First, the twenty waste streams were not adequately described.
11
-------
TABLE 3-3
SOLID WASTE GENERATION DATA
(Dry Weight)
Metric Tons/Yr.
1000 Barrels/Stream Day
Company
(Ute
Table 3-1)
ASH
ASH
ASH
ASH
CONT
COHT
CONT
BET
CETT
MARA
MARA
MARA
MARA
MOB
HOB
HOB
HOB
HOB
MOB
HOB
PHILL
PIIILL
PHILL
TEX
TEX
TEX
TEX
TEX
TEX
TEX
UNION
UNION
UNION
UNION
Refinery No.
(See
Table 3-1)
1
2
3
A
5
6
7
a
9
10
11
12
11
14
IS
16
1?
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Streaa Number (See Table 3-2)
1
p. 5.2
0-58
0,60
0.31
1.13
0.58
UK
(2)
HR
?.?A
HR
0.28
4.-1Q
(1)
UK
0.01
UK
2.16
(1)
(1)
UK
0
UK
UK
(2)
(2)
NR
0,51
.L21_.
JLOi
0,01
.Ul_
0.02
(2)
UK
JLflL
NR
Q
UK
(2)
UK
it)
12L_
NR
NR
UK
D
UK
UK
UK
o,qi
0.17
4
4.55
2.26
2.56
JL13_
0.37
0.4^8
0.21
12,90
0.03
2.54.
1-25
JL_6i_
6.72
5.}6
(K95
0.59
UK
0.16
1,12
a)
27.72
0.66
55.92
59.34
(2)
0,32
NR
(2)
0.25
SLAi
0.21
1.67
6.22
1.6
5
0,01
0.0|
0.01
0.01
UK
(2)
0
0.22
0.66
(2)
(?)
(2)
0.08
UK
<2)
0
UK
UK
12)
(i)
UK
UK
UK
(2)
(2)
(2)
(2)
0.01
UK
UK
fl_
UK
UK
0.02
6
IUQ1
0.01
Q.flL
Q
(2)
L..24.
0
0
Ai22_
O.p*
1-66
0
aoi.
(2)
UK
0
UK
0.18
2
0
5.44
0
6.09
_fl_
0.23
0.79
(2)
NR
7.60
NR
Q
0
0.2.4.
0
0,64
0
(21
0
<1)
(2)
1.30
0-01
0
NR
0
0
0
Q
0
0
0
0
0
13
0.17
Q.ll
0
UK
2.19
0
4.61
9.32
10.80
7.66
NR
(1-08
0
7.64
19.64
0.03
(2)
0
0
Q
0.19
0.34
UK
1.24
(2)
0.51
NR
(2)
1.97
0
11.25
9.54
0
0.76
14
0
0
P
P
0
0
0
0
NR
0
P. 04.
0
0
q
P
P
0
0
1.85
0
0
0
0
0.56
NR
0
4.04
UK
0
Q_
0
P
P. 17
0.9
>5
5.95
2.09
4.P2
4.65
0
0.74
1-9*
0.02
3.61
4.05
NR
3.18
0
P
0
4. SO
1.18
tt
8.37
(1)
4.30
34.92
14.28
9.25
(2)
1.21
(2)
0
(2)
5.16
6.75
3.87
8.0Q
0
16
O.O2
0.02
0.02
0.02
6.35
UK
P
12.56
NR
P
NR
1.62
P
UK
P
UK
UK
P
P
UK
0
0
UK
2.19
NR
0
P
0
P
UK
0
UK
P
7.6
"
<2)
-Iil_
(?)
(2)
UK
(2)
(2)
(2)
NR
0
(2)
(2)
0
(2)
<2)
P
0
(2)
NR
0
0
UK
UK
P
(2)
P
NR
0
P
0
(2)
UK
0,01
0
18
O
0
P
0
UK
p
9
p
NR
(2)
NR
(2)
P
(2)
(2)
p
JUL_
P
P
p
0
UK
UK
0
NR
(2)
(2)
(2)
-------
As a result, there appeared to be a lack of uniformity in the
interpretation of the nomenclature used to describe the waste
streams which probably caused certain quantities of solid waste
to be attributed to the wrong waste streams. Second, the allow-
ance of best estimates in the absence of actual measurements was
apparently used by many respondents as authority to provide esti-
mates which were not based on any sampling. Third, most likely a
completely random or a stratified random sample was not made. A
bias can result if a random selection of companies and refineries
within each company is not made. These biases could be a source
of error in the estimates of the mean solid waste generation rates
for all refineries.
3.3 Data Analysis
The solid waste generation rate data in Table 3-3 was
examined to identify any generation rates which seemed unreason-
able based on knowledge of typical refinery operations. The
generation rate of 12.56 metric tons/year per 1000 barrels/
stream day (MTPY/TBPSD) reported by Refinery 8 for Stream 16
was considered unreasonably large, and the 0.02 MTPY/TBPSD rate
reported by Refinery 8 for Stream 15 was considered unreasonably
low. As a result, these two generation rates were eliminated
from the data base.
The remaining generation rate data in Table 3-3 were
keypunched, verified, and placed in a data file. A histogram of
the generation rate data for each waste stream was developed. As
is the case with much process and environmental data, the histo-
grams revealed that the data are skewed. A typical example is
contained in Figure 3-1. This is a histogram of the number of
13
-------
FIGURE 3-1
HISTOGRAM OF SOLID WASTE GENERATION RATE DATA FOR API SEPARATOR
S~*
U
oo o
•* 0
SI
fr
o
%^
41
«J
S
C
0
iH
«J
2
0)
0
**
Id
41
0.
»j
X
«
O
u
o
u
u
FREQ.
0-33 ,1**************************************************************** ig
X*H
X
a
T>
i
M
U
n
10
tH
2
3.5 ^********************
6-9 ,(****************
12-15 y****
27-30 ^****
30-33 /********
„ ..4...__i,_ MA«_ .1. i J 4^ | ^_
— — — | 'i1 — — , — ^^* ( — —- |i_-»^^ — ^^ • — — — 1 — — — T-
!•>?/. c£iaa
, . 5
4
1
1
2
i 4. 1, . 4, J-. . . | ._ . |
••• 1 — — * — — — . — — — 1 — — — 1 — — — 1 — — 1
in 11 10 14 i/. ic it
CUM.
FREQ.
16
21
25
26
27
29
PERCENT
55
17
13
3
3
6
.17
.24
.79
.45
.45
.90
CUM.
PERCENT
55
72
86
89
93
100
.17
.41
.21
.66
.10
.00
Is
NUMBER OF REFINERIES WHOSE GENERATION RATE
FALLS IN THE INTERVAL
-------
refineries reporting various generation rates for API Separator
Sludge. To illustrate, sixteen refineries reported generation
rates between 0 and 3 MTPY/TBPSD, five refineries reported 3-6
MTPY/TBPSD, four refineries reported 6-9 MTPY/TBPSD, one refinery
reported 12-15 MTPY/TBPSD, one refinery reported 27-30 MTPY/TBPSD,
and two refineries reported 30-33 MTPY/TBPSD.
As can be seen, several large generation rates have
been reported relative to the bulk of the data. For this type of
data, the arithmetic average is an inefficient estimation of the
population average. The median, although being insensitive to a
few large values, is also inappropriate as an estimate of the
population mean. Since the data are not symmetric, the popula-
tion mean and median are different. The sample median is an
under-estimate of the population mean. For example, the mean for
the API separator sludge is 6.8 MTPY/TBPSD while the median is
0.96 MTPY/TBPSD.
In order to efficiently estimate the population mean,
a lognormal model was fit to the non-zero generation rate data
for each waste stream. This model is appropriate when the natural
logarithms of the data are approximately normally distributed.
The percentage of non-generating refineries was modeled by a bi-
nomial distribution. Figure 3-2 is a histogram of the logs of the
API separator sludge data. The most efficient unbiased estimates
of the population mean and variance have been given by Finney
(FI-163). These estimators have been used in computing an esti-
mate of the population mean or the generation factor for each
waste stream. The mathematical details about the lognormal dis-
tribution and the estimates are contained in Appendix II. The
generation factors computed using the estimate of the population
mean for the lognormal model are presented in the first column of
Table 3-4. A generation factor is not shown for Stream 18 because
no non-zero generation rates were reported for this stream.
15
-------
FIGURE 3-2
HISTOGRAM OF THE LOGARITHMS OF SOLID WASTE
GENERATION RATE DATA FOR API SEPARATOR SLUDGE
>,
•3
e
cd
oj-3.5
£-2.0
^2 ~1>5
0) 0) i n
4-1 t-4 — *• • "1
cd l-t
« 3 -0.5
.O
go 0.0
•HO
M in
0) V< 1.0
J 8. 1.5
~« 2.0
bO 0)
1 3.5
o
** 4.0
u
/**********
jt**********
^-.v*****************************
jt****************************************
jl****************************************
jt******************************
^***A******
U 0, J.J, j. j-i J,A i,i,Jrii1.i,af,i.
^ «« » AwWWW«rw**fWWyf»W A A A A' w A A W*C w W w'/f "A'A w« Ifc » A' W « w
,4********************
)t********************
ji**********
^**********
/******************** ' '
— ...... — -i 1 .—i -_ 1-
FREQ.
1
1
3
4
4
3
1
/,
H
2
2
1
1
2
CUM.
FREQ.
1
2
5
9
13
16
17
23
25
26
27
29
PERCENT
3.45
3.45
10.34
13.79
13.79
10.34
3.45
U7Q
. /7
6.90
6.90
3.45
3.45
6.90
CUM.
PERCENT
3.45
6.90
17.24
31.03
44.83
55.17
58.62
79 Al
/ fc . H 1.
79.31
86.21
89.66
93.10
100.00
123
Number of Refineries
-------
TABLE 3-4
SOLID WASTE GENERATION FACTORS
(Dry Weight)
Waste Stream
(See Table 3-2)
,
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
1 D
IB
19
20
Generation Factor fur
Generating Refineries
Metric tons/year
1000 barrels/stream Joy
4.44
12.8
0.083
3.19
0.083
1.21
2.43
0.30
6.09
0.36
5.8
6.7
8.95
0.67
8.3
4.05
0.006
3.96
0.42
Estimate of
Percentage
of Generating
Refineries
76
62
47
91
62
68
50
82
71
38
44
29
62
21
68
50
15
29
32
95Z Confidence
Interval on estimate
of Percentage of
Generating Refineries
58 -
43 -
29 -
76 -
43 -
49 -
32 -
65 -
52 -
22 -
27 -
15 -
43 -
8 -
49 -
32 -
4 -
15 -
17 -
89
78
65
98
78
83
68
93
85
56
62
47
78
38
83
68
31
47
51
Generation Factor
951 Confidence Interval
for Generation Factor
for All Refineries for All Refineries
Metric tons/year Metric tuns/year
1000 barrels/stream day 1000 barrels/strean day
3.37
7.94
0.039
2.90
0.051
0.82
1.22
0.25
4.32
0.14
2.55
1.94
5.55
0.14
5.64
2.03
0.001
1.15
0.13
0.55
0.93
0.005
1.22
0.0045
0.14
0.013
0.061
1.4
0.031
0.24
0.12
1.27
0
2.1
o.uy
0
0
0.008
- 16.3
-52.4
- 0.21
- 5.34
- 4.1
- 3.2
- 8.5
- 0.79
- 9.9
- 0.46
- 21.5
- 23.9
- 20.5
- 2.7
- 7.0
- 46.3
- 0.02
- 59,600
- 1.32
-------
An estimate of the population mean for each waste stream
including not only the refineries reporting non-zero generation
rates but also the refineries reporting zero generation rates was
desired; therefore, the non-generating refineries had to be taken
into account. The percentage of generating refineries in the
total refinery population was estimated from the data in Table 3-3.
Non-generating refineries for a particular waste stream were de-
fined as those refineries reporting a zero generation rate, the
waste stream discharging the sewer system, and not reporting any
generation rate for that waste stream. The estimates of the per-
centage of generating refineries were multiplied by the gen-
eration factors for the generating refineries to compute the gen-
eration factors for all refineries. The total estimation proce-
dure is contained in Appendix III. Table 3-4 contains the genera-
tion factors for all refineries and the estimates of the percent
of generating refineries and 95 percent confidence intervals
for each. The confidence interval for the generation factor for
Stream 19 is unreasonably large because of the extreme variability
in the few generation rates reported; therefore, Stream 19 was
dropped from further consideration. The confidence intervals are
based on the variability in the data and do not allow for biases
mentioned in Section 3.2. Therefore, the confidence intervals
may be an under-estimate of the uncertainty in the generation
factors.
3.4 Comparison with Previous Studies
Two previous studies have estimated solid waste genera-
tion for the petroleum refining industry. The first study (JA-216)
was conducted by Jacobs Engineering Company in 1974 for EPA-Office
of Solid Waste. Jacobs sampled waste streams at sixteen refiner-
ies and used the resulting data to calculate generation factors
and national solid waste generation by waste stream. The second
18
-------
study (EN-818) was conducted by API in 1976. API sent data re-
quest forms to all the operating U.S. refineries. The responses
received represented 57 percent of the refining capacity at that
time. API used the data from the responses to estimate national
solid waste generation by waste stream.
In comparing the results of these studies, it would be
helpful to have a measure of the uncertainty in the estimated
generation factors for each study. The confidence intervals in
Table 3-4 serve that purpose for this study. However, no error
bounds for the generation factors in either the Jacobs or API
reports were given. Therefore, the confidence intervals in Table
3-4 will be used as a basis of comparison, even though they are
actually too narrow for this purpose since they account for the
uncertainty in the generation factors in this study only.
The comparison will be made by determining if the gen-
eration factors from the Jacobs and API reports fall within the
error bounds of the generation factors developed in this study.
If they do, then there is evidence that the differences in gen-
eration factors are due to the expected differences in refineries.
If they do not fall within the error bounds, then possibly other
factors or biases are causing the differences.
4
Jacobs reported solid waste generation factors based on
medians; therefore, the generation factors in Table 3-4 are not
comparable to these factors. In order to check the reasonable-
ness of the factors from this study against Jacobs' factors, the
medians from this study, not including the refineries not gener-
ating, were compared with the medians from the Jacobs report.
The refineries not generating were excluded because Jacobs'
medians included only generating refineries. Table 3-5 contains
these medians and 95 percent confidence intervals for medians
19
-------
RADIAN
TABLE 3-5
COMPARISON OF JACOBS AND EPA-OAQPS GENERATION FACTORS
Stream
Number
(See
Table 3-2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Median Solid Waste Generation
Factors (Dry metric tons/yr.
per 1000 barrels (Stream day)
Jacobs
1.38
2.00
0.33
2.2
46.0
0.17
-
0.75
3.64
0.93
215.6
21.7
0.8
12.6
5.45
0.28
EPA-OAQPS
0.58
1.90
0.02
0.96
0.01
0.13
0.19
0.06
1.42
0.22
0.21
0.64
2.20
0.50
4.04
1.60
95% Confidence
Intervals for
EPA-OAQPS Medians
0.19-3.13
0.09-4.52
0.006-0.51
0.45-2.6
0.01-0.22
0.04-1.2
0.004-0.51
0.01-0.28
0.7-11.6
0.06-0.57
0.02-1.04
0.03-5.44
0.17-9.5
0.04-1.85
3.18-5.95
0.02-6.34
20
-------
from this study. Five of the fifteen medians from the Jacobs
report, Streams 5, 10, 11, 12, and 14 do not fall with the con-
fidence intervals. The median for Stream 10 is only slightly
larger than the upper confidence bound; so, if the uncertainty
associated with the median were accounted for, it would probably
fall within the interval. The medians for the other four streams
are quite large and are in definite disagreement with the results
of this study. The medians for these streams are based on only
two or three samples; therefore, the possibility exists that the
medians are biased.
Table 3-6 contains the mean generation factors for the
API study and this study and the confidence intervals for the
factors from this study. The API generation factors were not
presented explicitly in the API report; therefore, the factors
had to be calculated by dividing the total solid waste generation
for each waste stream by the capacity of the refining industry as
cited in the report. The API generation factors for Streams 13
and 14 are the only factors that do not fall within the EPA-OAQPS
confidence intervals. The factors for the streams are only
slightly outside the confidence bounds. If the uncertainty
associated with the factors was accounted for, they would prob-
ably fall within the intervals.
3.5 Extrapolation of Data
In order to obtain an estimate of the annual solid waste
generation of the entire U.S. petroleum refining industry, the
generation factors for all refineries in Table 3-4 were multiplied
by the crude capacity of all operating refineries as given in the
Annual Refining Survey published in the Oil and Gas Journal (CA-
673). The resulting estimates of total generation by waste stream
were then summed to obtain a total for all waste streams. The
confidence intervals for the generation factors were also multi-
plied by refinery capacity to obtain confidence intervals for the
21
-------
TABLE 3-6
COMPARISON OF API AND EPA-OAQPS GENERATION FACTORS
Stream Number
(See Table 3-2)
Mean Solid Waste
Generation Factors
(Dry metric tons/year
per 1000 barrels/stream day)
95% Confidence
Intervals for
EPA-OAQPS Means
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
API
-
1.46
-
4.56
-
-
-
-
2.25
-
-
0.39
1.12
2.76
3.60
-
-
-
-
_
EPA-OAQPS
3.37
7.94
0.039
2.90
0.051
0.82
1.22
0.25
4.32
0.14
2.55
1.94
5.55
0.14
5.64
2.03
0.001
-
-
0.13
0.55
0.93
0.005
1.22
0.0045
0.14
0.013
0.061
1.4
0.031
0.24
0.12
1.27
0
2.1
0.05
0
0.008
- 16.3
- 52.4
- 0.21
- 5.34
- 4.1
- 3.2
- 8.5
- 0.79
- 9.9
- 0.46
- 21.5
- 23.9
- 20.5
- 2.7
- 7.0
- 46.3
- 0.02
-
-
-1.32
22
-------
TABLE 3-7
ANNUAL SOLID WASTE GENERATION
OF THE ENTIRE U.S. PETROLEUM REFINING INDUSTRY
957o Confidence Interval
Waste Stream
(See Table 3-2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Total Generation
(metric tons/year)
59,380
139,900
690
51,090
900
14,450
21,500
4,400
76,110
2,470
44,930
34,180
97,790
2,470
99,370
35,770
18
—
—
2.290
for Total Generation
(metric tons/year)
9,690 -
16,390 -
90 -
21,500 -
80 -
2,470 -
230 -
1,070 -
24,670 -
550 -
4,230 -
2,100 -
22,380 -
0 -
37,000 -
880 -
0 -
-
-
140 -
28,720
923,200
3,700
94,090
72,200
56,460
149,800
13,920
174,400
8,100
378,800
421,100
361,200
47,570
123,300
815,800
350
23,300
Total
687,708
143,470 - 3,695,950
23
-------
estimates of total generation by waste stream. The lower bounds
and the upper bounds of the confidence intervals were then summed
to obtain a confidence interval for the total generation for all
waste streams. These figures are presented in Table 3-7.
The estimate of total annual solid waste generation for
the refining industry developed by this study is approximately
twice the total reported by API and three times the total reported
by Jacobs. However, the API total of 357,000 metric tons/year
and the Jacobs total of 240,000 metric tons/year both fall within
the confidence interval for the total from this study. Since the
API and Jacobs totals fall within the confidence limits, there is
evidence that the differences in the totals are due to the ex-
pected differences in refineries.
24
-------
4.0 HYDROCARBON EMISSIONS
One objective of the EPA-OAQPS study was to estimate
the national VOC atmospheric emissions resulting from refinery
solid waste treatment and disposal. Petroleum refineries are a
known source of VOC emissions. Large quantities of organics flow
through refineries, so that process units, storage units, waste
treatment systems, and fugitive sources (leaks) may possibly emit
volatile organics. Volatile hydrocarbons may also become part of
a solid waste stream, along with hydrocarbons of lower volatility.
Exposure of the solid waste to the atmosphere, which occurs during
common handling, transport, and disposal methods, may allow some
or all of the volatiles to evaporate.
4.1 Data Analysis
4.1.1 Phase Composition of Waste Streams
The generation rate and gross phase composition (solid,
aqueous, hydrocarbon) for twenty major refinery solid waste streams
were obtained in the EPA-OAQPS survey. The phase composition
data is summarized in Table 4-1, along with values from the API
study for comparison. Overall, the two studies are in close
agreement.
4.1.2 Hydrocarbon Content of Refinery Solid Waste
The generation rate data and the phase composition
data may be combined to determine the hydrocarbon content of all
United States refinery solid wastes. Hydrocarbon content, on a
dry weight basis, was calculated for each stream, and multiplied
by the appropriate generation rate to obtain the hydrocarbon
generation rate for each stream; the result is presented in Table
4-2. As shown in the table, hydrocarbons in DAF Skimmings, Slop
25
-------
TABLE 4-1
EPA - OAQPS SURVEY
SUMMARY OF REPORTED PHASE DATA
WATER, SOLID, AND HYDROCARBON CONTENT OF REFINERY SOLID WASTE STREAMS
(Compositions in Weight Percent)
o\
Slop
Oil
Sample Phase Emulsion
Solids
Water
Range 0-94
Mean 34
No. of Responses 17
Solids
Range 3-90
Hean 27
No. of Responses 13
Hydrocarbons
Range 1-80
Hean 35
No. of Responses 14
Kerosene
Filter
Exchanger
Storm Bundle
Runoff Cleaning
Silt Sludge
15-94 0-96
44 50
12 8
6-100 3-100
52(61) 28
14 9
0-35 4-17
5(4) 4
13 8
Neutral-
Other Iced IIF Waste
Filter Alky. Blo-
API
Separator
Sludge
0-98
59
26
0-60
18(22)
25
0.2-90
19(13)
28
Once-
Through
Cooling
^"U'J^JJlf!8-* Clays Clays Sludge Sludge Sludge
Water
~Rnnf-e 0-50
Hean 16
No. of Responses 3
Solids
Range 50-95
Henn "(70)
No. of Responses 3
Hydrocarbons
Range 0—13
Hean 5(11)
No. of ReHoonnca
0-98 39-98 0-99
24 53 81
10 5 20
1-100 2-100 1-79
57(73) 45 11(9)
10 5 19
0-48
-------
TABLE 4-2
EPA - OAQPS SURVEY
TOTAL HYDROCARBONS* IN
REFINERY SOLID WASTES
(Total U. S. Refineries)
Hydrocarbons
in Waste Streams
Solid Waste Stream
Dissolved Air Flotation
Slop Oil Emulsion Solids
API Separator Sludge
Biosludge
Storm Runoff Silt
Other Filter Clays
Crude Tank Bottoms
Other Tank Bottoms
HF Alklation Sludge
Coke Fines
Leaded Tank Bottoms
Non-Leaded Tank Bottoms
Kerosene Filter Clays
Exchanger Bundle Cleaning Sludge
Spent Amines
FCC Catalyst Fines
Once-through Cooling Water
Sludge
TOTALS:
ric Tons /Year)
49,470
31,470
23,500
20,540
12,590
8,090
8,090
7,310
2,050
5,370
480
290
150
60
5
o**
0
169,465
Percent of Total
29.2
18.6
13.9
12.1
7.4
4.8
4.8
4.3
1.2
3.2
0,3
0.2
0.1
0.04
0
0
0
100.14
Notes:
* Hydrocarbons were defined as the portion of the waste stream present in the
liquid organic phase (distinct from the aqueous and solid phases).
** Reported Composition <1% hydrocarbon.
27
-------
Oil Emulsion Solids, API Separator Sludge, and Biosludge
comprise about 75 percent of total refinery solid waste
hydrocarbons.
The total hydrocarbon content of refinery solid waste
was found to be approximately 190,000 tons/yr. Although the
potential exists for hydrocarbon recovery from solid wastes,
recovery practices vary greatly among the refineries and many
wastes are not treated before disposal. Of the four streams
which are major hydrocarbon sources, only DAF Skimmings are
widely treated for hydrocarbon recovery (EN-821). Therefore, a
major part of the hydrocarbon phase of the wastes is disposed.
4.1,3 Estimation of VOC Emissions
Only a fraction of the total hydrocarbon content will
be released to the atmosphere; the amount is primarily dependent
on volatility. One way to express the sensitivity of total hydro-
carbon emission to volatility is shown in Figure 4-1. If the
volatile organic fraction of solid waste hydrocarbons was known,
then an upper bound for VOC emissions could be determined by the
figure. For instance, assuming that 10 percent of the total
hydrocarbons are volatile, refinery solid waste VOC emissions
could approach 19,000 tons/yr. It may be expected that only a
few percent of the hydrocarbons are volatile, due to upstream
processing in the refinery. Actual emissions would be less than
the upper bound, depending on waste source, treatment and dis-
posal, and mass transport effects. Volatility, however, is the
most important indication of VOC emission potential.
28
-------
200 -r
150
Maximum
Solid Waste
Volatile
Organic
Emissions
10 Ton/yr
100
10 20 30 AQ 50 60 70 80 20 100
% VOLATILITY OF
TOTAL HYDROCARBONS
FIGURE 4-1
ORGANIC EMISSION VERSUS
VOLATILITY OF HYDROCARBON PHASE
29
-------
4.2 Data Limitations
The EPA-OAQPS survey did not produce a sufficient data
base to accurately estimate volatile organic emissions. Although
the requested data included VOC and benzene content for each
waste stream, little usable information was received. The number
of observations of volatile components reported for each waste
stream is presented in Table 4-3. Some of the observations were
not usable because specific values were not reported. The number
of usable, quantitative observations reported for each waste
stream is also shown in Table 4-3. The 35 quantitative observa-
tions reported to OAQPS are inadequate to develop hydrocarbon
and benzene emission factors and subsequently accurate estimates
of total hydrocarbon and benzene emissions from refinery solid
waste.
Missing information and estimated values limit the
accuracy of estimates based on the OAQPS data. In the survey
responses, lack of standard definitions and testing procedures
for volatile organics were cited as justification for not sub-
mitting the requested information. Best estimates, in the absence
of test results, were allowed in the survey, with a requested
accuracy of 20 percent. Many of the reported values are esti-
mates, since analytical test results were not available to the
companies. Also, the wide variation in reported compositions may
indicate that some waste samples were fresh from refinery sources,
while other samples had been air dried for several days or had
undergone other treatment.
4.3 Factors Controlling Volatile Emissions
An accurate estimate of VOC emissions from solid wastes
cannot be simply based on hydrocarbon volatility, since the
emission of volatiles is a complex mass transport process. Solid
30
-------
TABLE 4-3
VOLATILE COMPONENTS OF WASTE STREAMS REPORTED IN EPA-OAQPS SURVEY
Volatile Organic Carbon Total Light Hydrocarbons
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Waste Streams
Slop Oil Emulsion Solids
Silt from Stormwater
Runoff
Exchanger Bundle Clean-
ing Sludge
API Separator Sludge
Non-Leaded Gasoline Tank
Bottoms
Crude Tank Bottoms
Other Storage Tank Bottoms
Leaded Gasoline Tank
Bottoms
Dissolved Air Flotation
Skimmings
Kerosene Filter Clays
Other Filter Clays
HF Alkylation Sludge
Waste Bio-Sludge
Once-Through Cooling
Water Sludge
FCC Catalyst
Coke Fines
Spent Amines
Salts from Regeneration
Ship and Barge Ballast
Other
Number
of Data
Points
1
1
0
1
0
0
0
'
0
1
0
1
2
1
0
1
0
0
0
0
1
Number of
Useable*1)
Data Points
0
1
0
0
0
0
0
•
0
0
0
0
2
1
0
1
0
0
0
0
1
Number
of Data
Points
0
0
0
0
0
1
1
2
0
0
0
0
0
0
0
0
0
0
0
0
Number of
Useable(1)
Data Points
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Number
of Data
Points
6
6
4
11
4
6
7
7
10
1
1
4
6
0
5
6
4
0
2
5
Number of
Useable(1)
Data Points
0
1
0
3
4
2
1
4
6
0
0
2
3
0
1
0
0
0
0
2
(1)
Useable data points were defined as analyses reported as a specific value for the constituent.
-------
waste composition is only one of several physical and chemical
factors which affect VOC emissions; some of these are listed in
Table 4-4.
Factors which determine the exposure of solid wastes
to the air may have an important effect on VOC emission. Long
exposure time and large exposed surface area would give the most
opportunity for volatile evaporation. However, evaporation may
be so rapid that all VOC emissions take place in preliminary
dewatering or transportation, so that an actual disposal method
is not significant. Also, evaporation of volatiles in process
sewer lines may cause VOC emissions at that point. Since waste
handling procedures vary widely among refineries, it is difficult
to estimate at which point the volatiles may be emitted from
solid wastes. But, if volatiles are present in the fresh waste
stream, there is ample opportunity for air emissions during
treatment and disposal.
Mass transport of VOC from the solid waste depends on
both chemical composition and physical properties. The con-
centration of volatiles in the hydrocarbon phase is the driving
force behind VOC mass transport. The vapor pressure and molecu-
lar weight of each chemical species affect the evaporation rate,
as do oil-water phenomena and interphase binding. The VOC emis-
sion rate may be low even with a rapid evaporation r.ate, since
transport of volatiles in the sludge may be the controlling
factor. For instance, rapid evaporation may form a crust or
impervious layer at the surface which prevents further VOC emis-
sion from the bulk of the waste. No information on the occurrence
of these phenomena in refinery solid waste is believed to exist.
32
-------
TABLE 4-4
FACTORS AFFECTING VOE FROM SOLID WASTES
1) Disposal Site Construction
covered or open
sand bed drainage
land farm or landfill
dilution in lagoon
2) Composition .
phase composition (overall)
hydrocarbon phase composition (voLatiles)
concentrations in each phase
3) Physical Transport in Sludge
crusting - impervious layer
cracking
volatile diffusion
4) Exposure Time
drying - open air
tilling and soil aeration - land farm
5) Transportation and Pre-Treatment Handling
wastewater process units
open dump truck or tank car transport
spreading method
6) Weather
temperature
wind
sun
rain
humidity
7) Waste Deployment
thick or thin layer
packed or unpacked
density
8) Mixing of Waste Streams
9) Oil-Water Phenomena
surface oil layer
emulsification
10) Inter-Phase Binding
hydrocarbon-water solubility
hydrocarbon-solid adsorption
33
-------
4.4 Supplementary Data Collection Program
The most important data missing from the OAQPS survey
results are the amounts of volatile hydrocarbons present in each
stream. Standard test procedures specifically designed for VOC
in solid wastes can be developed and applied to obtain the missing
information. This would permit estimation of the maximum poten-
tial VOC emissions. To estimate actual VOC emissions, in-lab sim-
ulation of waste disposal methods are necessary to measure multi-
phase mass transport of volatile organics.
A supplemnetary sampling program will provide the data
required for an accurate estimate of VOC emissions from refinery
solid wastes. The more complete characterization of solid waste
streams obtained in such a program would have many valuable uses.
These would include:
• Better estimation of solid waste compositions
and generation rates through statistically de-
signed survey methods
• Provision of required data for VOC emissions
estimate
• Identification of toxic and hazardous compounds
• Better data base for design of solid waste treat-
ment and disposal methods
• Identification of valuable constituents for
possible resource recovery
A suggested sampling program is presented in Appendices
IV and V which addresses the following factors:
• Refinery complexity, crude slate and location
• Solid waste stream characteristics
• Number of samples required
• Analytical parameters
• Development of reproducible sampling/sample
preservation techniques.
34
-------
Appendix IV discusses alternative schemes for selecting
a representative sample of the entire U.S. refining industry. A
modification to the refinery sampling program developed by Arthur
D. Little, Inc. (ADL) is presented. The suggested program util-
izes the Bureau of Mines Petroleum Administration for Defense (PAD)
refinery classes to serve as the basis for a stratified sampling
approach to collect the required refinery waste samples. The
stratified sample design eliminates bias from the sample data
and therefore permits more accurate estimate of waste volumes and
characteristics.
Appendix V describes the design of a supplementary
sampling program that could be pursued. Table 4 in this Appendix
shows the cost and the precision of the estimate of constituent
content for various sampling programs. From the table, given a
desired level of precision for the estimate, a sampling program
and its associated cost can be found.
4.5 VOC Emissions from Other Industries
The petroleum refining industry is only one of many
potential sources of solid wastes containing organic solvents.
In order to make a realistic estimate of national atmospheric
emission rates of VOCs from organic solid wastes, other indus-
tries must be considered.
Solid wastes containing volatile organic compounds are
generated by industry as the result of both primary production
activities and hydrocarbon recovery and reprocessing. Many in
dustrial production processes have been studied from the stand-
point of organic emissions from vapor-emitting point and fugitive
sources. However, liquid and solid solvent-contaminated waste
streams and the resultant organic emissions have not been studied
35
-------
adequately. Reprocessing and recovery industries closely related
to the refining industry, such as the petroleum re-refining and
solvent reclamation facilities have process streams with charac-
teristics common to refineries. Solid waste streams from these
industries may also contain VOCs similar to refining wastes. How-
ever, it is beyond the scope of this report to analyze in detail
the waste streams generated by other industries.
36
-------
5.0 SOLID WASTE MANAGEMENT PRACTICES
Numerous solid waste streams (which in this report in
elude sludges) are generated by refinery processes. The quanti-
ties and compositions of these streams have been identified pre-
viously in this report. This section will focus on:
• Source control
• Treatment
• Final disposal
5.1 Source Control
Source control can be defined as the reduction or con-
tainment of plant emissions at their point of generation. Instead
of continually investing in more extensive pollution control equip-
ment to meet new regulations, refiners are making significant moves
to reduce wastes generated. By continuously implementing new
source control methods, refiners anticipate the constantly chang-
ing regulations and minimize the effort and costs of compliance.
One approach to source control is the addition or alteration of
equipment. For instance, installation of tank mixers in crude
and other storage tanks minimizes the deposition of bottoms
sludge.
Another approach is through operational or procedural
changes. Two examples are:
• Shutdown planning. The amounts and qualities
of oily wastes are predetermined and handled
appropriately in the event of an equipment
shutdown.
• Internal refinery permit system. An oily
stream can be released to the oily drains
only with the approval of the Environmental
Department.
37
-------
Although source control has greater potential applica-
tion to refinery wastewaters, some utility in reducing solid waste
emissions is also possible. One major area currently undergoing
extensive study is the reduction of waste biosludge by lessening
the load on the wastewater treatment system. For example, water
reuse is a common strategy for reducing wastewaters.
With the increased promulgation of environmental regu-
lations and rising costs of pollution control, in many cases the
most cost-effective approach to management of refinery sludges
will be to locate and reduce emissions at the source.
5.2 Treatment
Up until the last few years, most refiners considered
solid waste disposal as a low priority item and a profit drain.
Consequently, convenience rather than environmental concern dic-
tated use of sludge pits and lagoons, usually located at the back
of refinery property. Little attention was given to leachate,
air emissions, or ultimate disposition of pollutants contained in
these pits. If solid wastes were consigned to a contractor, the
material was considered properly disposed of as soon as it left
the refinery.
The abundance of environmental regulations has forced
refiners to critically evaluate their pollution abatement systems.
One important aspect of these systems is the management of solid
waste. The process descriptions which follow evaluate the existing
and planned solid waste management techniques on a cost/performance
basis. Many costs are updated to December 1978; cost data given
in graphical form are not updated but rather the cost index values
are supplied to permit the reader to bring the data up to date.
38
-------
The process descriptions in Appendix I may be divided
into three sections:
(1) Proven treatment technologies currently
in use in refineries:
Vacuum filtration
Pressure filtration
Centrifugation
Dewatering lagoons
Aerobic digestion
Anaerobic digestion
Chemical fixation
(2) Proven treatment technologies potentially
usable in refineries:
• Sand drying beds
• Wet-air oxidation
• Composting
(3) Unproven treatment technologies with
respect to refinery sludges that may have
potential utility:
• BEST/solvent extraction process
Thermal drying as a treatment technology is not applic-
able to the refining industry. A high potential exists for odor
problems and evaporation of hydrocarbons. Drying is most amenable
to recovery of a valuable chemical of solid substance for resale.
Few refinery sludges, if any, meet these criteria.
5.3
Final Disposal
Much indecision on final disposal techniques currently
exists in the refinery industry. Most of this is due to pending
regulations, primarily the Resource Conservation Recovery Act (RCRA)
regulations for environmentally acceptable disposal means, and the
potential classification of many refinery wastes as hazardous. The
current technologies that this subsection of the report includes
are:
39
-------
• Incineration
• Landfarming
• Landfilling
• Ponding
• Disposal wells
• Ocean disposal
Only landfarming can truly be considered an ultimate
disposal technique because the wastes are degraded into matter
that becomes a part of the soil. Although anaerobic degradation
theoretically occurs in landfills, some of those that have been
recently reopened showed no significant deterioration of the oily
wastes. Ponding, disposal wells and ocean disposal are merely
storage techniques with no significant waste degradation. Incin-
eration generates air pollutants and ash residues that need dis-
posal.
Ocean disposal, deep well injection, and ponding are
being phased out due to their adverse environmental impact. The
use of landfarming, however, is increasing significantly where
land is available. The possibility of farming organic wastes
from retired refinery sludge pits after recovering and recycling
oil is being seriously considered.
Contracting wastes to local municipalities or private
disposal firms is prevalent in the petroleum refining industry.
Various proposed regulations will affect this practice which may
increase on-site disposal activities. To write effective stan-
dards for air emissions from solid waste disposal practices, the
refiners' intentions with respect to on-site versus contracted
disposal must be estimated.
40
-------
5.4 Process Descriptions
In Appendix I, both existing and planned techniques in
treatment and final disposal are identified, described, and evalu-
ated according to their development status, performance, and cost
and energy requirements. Particular attention is devoted to any
secondary pollutants generated by processing of solid wastes,
e.g., air emissions from incineration of oily wastes. Figure 5-1
depicts a matrix summary of the application of proven and unproven
technologies to the various waste streams.
41
-------
API Separator Sludge
Nonleaded Gasoline Tank Bottom
Crude Tank Bottoms
Other Storaee Tank Bottoms
DAF Skimmines
Waste Blosludee
Slop Oil Emulsion Solids
Silt from Storm Water Runoff
Once-Through Coolina Water S Indue
Kerosene Filter Clava
Other Filter Clavs
Exchanger Bundle Cleanlna Sludge
Neutralized HF Alkvlatlon Sludge
FCC Catalyst Fines
Spent Amines
Leaded Gasoline Tank Bottoms
Coke Fines
2
4
4
4
2
1
2
4
2
4
4
4
4
4
4
1
4
4
4
1
I
1
4
2
4
4
4
4
4
4
2
4
4
4
2
I
2
4
2
4
4
4
4
4
4
1
1
I
1
1
2
1
1
I
4
1
1
1
2
4
4
4
4
4
4
1
4
4
4
4
4
4
2
4
4
4
4
4
4
4
1
4
4
4
4
4
4
2
4
4
1
1
I
I
2
1
2
1
2
1
1
2
2
I
;z
i
2
4
2
2
2
4
2
4
2
2
4
4
2
2,
4.
2.
2
4
2
4
4
4
2
2
2
4
4
4
4
4
4
4
4
4
4
2
?
2
2
2
2
2
4
4
4
4
4
4
4
2
2
4
3
•}
•}
3
3
3
3
4
4
4
4
4
4
4
4
3
4
1
I
1
I
I
1
1
1
I
1
1
I
I
1,
1
1
I
1
1
1
\
\
1
1
1
1
1,
t
1
1
I
1
1
1
1
4
4.
4,
I
1
1
4
4
4
4
4
4,
2
1
4
4
1
7
]
1
1
l
l
\
1
4
4,
2.
1
1
1
\
1
4
4.
4
4,
2
2
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2
4
4
4
2
2
4
4
2
4
4
2
Key; 1 Proven technology that is currently being applied to this waste stream.
2 Proven technology that la currently being applied to industrial and/or municipal waste streams,
but is not currently being applied to this waste stream.
3 Unproven technology that might be applicable to this waste stream.
4 Proven or unproven technology that is not applicable to this waste stream.
Note; Each waste stream has been considered individually, not in combination with others. This limits
the applicable technology, but the streams must still be considered individually because it is
impossible to consider all the possible combinations.
The results listed above in Figure 5-1 include both current and/or past application to the
particular waste stream.
The consideration of the application of these technologies includes wastes processed on-site and
those consigned to a contractor.
The results listed in the matrix were not based upon the frequency of application of a technology
to the given waste streams.
FIGURE 5-1
MATRIX SUMMARY OF CONTROL TECHNOLOGIES
AND REFINERY WASTE STREAMS
42
-------
BIBLIOGRAPHY
AI-038 Aitchison, J., "On the Distribution of a Positive
Random Variable Having a Discrete Probability Mass at
the Origin," American Statistical Association Journal,
50, (9), 1955, 901-908.
AM-065 American Petroleum Inst., Div. Statistics & Economics,
Annual Statistical Review, U.S. Petroleum Industry
Statistics 1956-197T:Washington, D.C., 1973.
AM-046 Ambler, Charles M., "Centrifuge Selection," Chem. Eng.
78(4), 55 (1971).
AM-062 American Petroleum Inst., Div. of Refining, Manual
on Disposal of Refinery Wastes. Volume on Liquid Wastes,
First edition.Washington, D.C., 1969.
AS-070 Ashbrook, Simon and Hartley, "Sludge Dewatering with
Klampress," brochure. Houston, TX, undated.
BE-499 Benjes, Henry H., Jr., Cost Estimating Manual—Combined
Sewer Overflow Storage and Treatment, final report.
EPA Contract No. 68-02-2186, EPA 600/2-76-286.
Cincinnati, OH, EPA, Municipal Environmental Research
Laboratory, Systems & Engineering Evaluation Branch,
Wastewater Research Division, Dec. 1976.
BI-043 Bitler, John A. and L. John Minnick, "Lime-Sulfur
Dioxide Scrubbing System and Technology for Utili-
zation of Underflow Sludge," Presented at the Nat'l.
Lime Assoc. Annual Convention, 70th, Hot Spring, VA,
1972.
BO-196 Bombaugh, K. J., et. al., Sampling and Analytical
Strategies for Compounds in Petroleum Refinery Streams.
EPA Contract No. 68-02-1882. Radian Corp., Austin,
TX, Sept. 15, 1975.
BO-328 Bellinger, John C. and Jeffrey L. Shumaker, Control of
Volatile Organic Emissions from Solvent Metal Cleaning,
EPA 450/2-77-022, OAQPS 1.2-079, Research Triangle Park,
NC, Environmental Protection Agency, Office of Air
Quality Planning and Standards, Office of Air and Waste
Management, November 1977.
CA-099 Cantrell, Ailleen, "Annual Refining Survey," Oil and
Gas Journal, April 2, 1973.
43
-------
BIBLIOGRAPHY (Continued)
CA-236 Cantrell, Ailleen, "Annual Refining Survey," Oil and
Gas Journal. April 7, 1975.
CA-279 Cantrell, Ailleen, "Annual Refining Survey," Oil and
Gas Journal, April 1, 1974.
CA-339 Cantrell, Ailleen, "Annual Refining Survey," Oil and
Gas Journal, March 29, 1976.
CA-673 Cantrell, Ailleen, "Annual Refining Survey," Oil and
Gas Journal, March 20, 1978.
CH-196 Cheremisinoff, Paul N. and Richard A. Young, eds.,
Pollution Engineering Practice Handbook. Ann Arbor,
MI, Ann Arbor Science, 1975.
CH-353 Cheremisinoff, Paul N. , "Biological Wastewater Treat-
ment," Pollution Engineering 8(9). 32-8, 1976.
CH-447 Cheremisinoff, P. N., "Sludge Handling and Disposal,"
Poll. Eng. 8(1) 22, 1976.
CL-039 Clark, John W., Warren Viessman, Jr., and Mark J.
Hammer, Water Supply and Pollution Control, 2nd ed.
Scranton, PA., International Textbook Co., 1971.
CO-698 Colley, J. D., et. al., Assessment of Technology for
Control of Toxic Effluents from the Electric Utility
Industry. Revised,DCN 78-200-187-09-08, EPA Con-
tract No. 68-02-2608, Work Assignment 9. Austin, TX,
Radian Corp., Dec. 1977.
CR-154 Cross, Frank L., Jr., "Control and Treatment of Waste-
water," Poll. Eng. 9(1). 99-105, 1977.
CU-082 Gulp/Wesner/Gulp, Estimated Costs of the B.E.S.T.
System and Other Methods for Treatment and Use or
Disposal of Municipal Wastewater Sludges.Jan. 1977.
CU-087 Gulp, R. L., Handbook of Advanced Wastewater Treatment,
2nd Edition, New York, Van Nostrand Reinhold, 1978.
DR-048 Drew, J. W., "Design for Solvent Recovery," Chem. Eng.
Prog. 71(2). 92-99, February, 1975.
. 44
-------
BIBLIOGRAPHY (Continued)
EC-008 Eckenf elder, W. Wesley, Jr. , Industrial Water Pollution
Control. N.Y. , McGraw-Hill, 1976.
EN-295 Environmental Protection Agency, Process Design Manual
for Upgrading Existing Wastewater Treatment Plants.
Oct. 1974.
EN-325 Environmental Protection Agency, Technology Transfer,
Process Design Manual for Sludge Treatment and Dis-
posal. EPA 625/1-74-006, October 1974.
EN-407 Environmental Protection Agency, Development Document
for Effluent Limitations Guidelines and New Source
Performance Standards for the Petroleum Refining Point
Source Category. EPA Contract No. 68-01-0598, Washington,
D.C. , April, 1974.
EN-602 Environmental Protection Agency, Office of Solid Waste
Management Programs , Decision-Makers Guide in Solid
Waste Management. SW-500. 1976.
EN-605 Environmental Protection Agency, Office of Solid Waste
Management Programs, Disposal of Hazardous Wastes,
Report to Congress. SW-115.1974.
EN-818 Engineering Science, Inc., The 1976 API Refinery Solid
Waste Survey, Pt. 1. Solid WTste Quantities, Draft
Final Report.Austin, TX, Jan. 1978.
EN-819 Environmental Protection Agency, Effluent Limitations
Guidelines (BATEA) New Source Performance StandardsT"
and Pretreatment Standards for the Petroleum Refining
Point Source Category, Washington, D.C., March. 19787
EN-820 Environmental Protection Agency, (OAQPS), Refinery Waste
Disposal - Responses to April 11, 1978, Section 114
EN-820 Environmental Protection Agency, (OAQPS), Refinery Waste
Disposal - Responses to April 11, 1978, Section 114
Information Request, 1978.
EN-821 Engineering-Science, Inc., The 1976 API Refinery Solid
Waste Survey. Pt. 2, Solids'Management Practices"
April 1978, Austin, TX.
45
-------
BIBLIOGRAPHY (Continued)
FA-178 Environmental Protection Agency (OSW), Information
About Hazardous Waste Management Facilities, Publica-
tion SW-I45, Prepared by D. Farb and S. D. Ward,
February 1975, Cincinnati, OH.
FE-055 Ferrel, James F., "Sludge Incineration," Poll. Eng.
1973 (March) 36.
FI-163 Finney, D. J., "On the Distribution of a Variate Whose
Logarithm is Normally Distributed," Journal of the
Royal Statistical Society, Series B, 7 (1941), 155-161.
GO-279 Goldstein, D. J. and David Yung, Water Conservation and
Pollution Control in Coal Conversion Processes, Final
Report, EPA Contract No. 68-02-2207, EPA 660/7-77-065.
Cambridge, MA., Water Purification Associates, June 1977.
HU-236 Hulswitt, Charles, (Astro Mettalurgical Corp., Wooster,
OH), Private Communication will Bill Arnold, 7 Feb.
1978.
JA-216 Jacobs Engineering Co.,"Alternatives for Hazardous
Waste Management in the Petroleum Refining Industry,"
Draft Final Report, EPA Contract No. 68-01-4156.
Pasadena, CA, April 1977.
JO-273 Jones, Jerry L., et. al., "Municipal Sludge Disposal
Economics," Env. Sci. Tech. 11(10), 968 (1977)
KN-053 Knowlton, H. G., "Source Control in Petroleum Refineries,"
Presented to NPRA Annual Meeting, San Antonia, TX,
March, 1978.
KN-054 Knowlton, H. E. and J. E. Rucker, "An Overview of
Petroleum Industry Use of Landfarming," Presented
at the 71st Annual AIChE Meeting, Miama Beach, FL,
Nov. 1978.
KN-055 Knowlton, H. E. and J. E. Rucker, "Refinery Oily Waste
Treatment and Disposal Techniques and Their Potential
Application to Oil Spill Debris," Presented at the
ASTM Symposium on Oil Spill Material Disposal, Denver,
CO, Nov. 1978.
LE-380 Levin, J. E. and F. Scofield, "An Assessment of the
Solvent Reclaiming Industry," Published in Preprints
46
-------
BIBLIOGRAPHY (Continued)
170th ACS Meeting. Division of Organic Coatings &
Plastics, 35(2),416-418, Wapora. Inc.. Washington.
D.C., 1975.
LE-381 Lewis, Robert S. , "Sludge Farming of Refinery Waste
as Practiced at Exxon's Bayway Refinery and Chemical
Plant," Presented at the Nat'l Conference on Disposal
of Residues on Land, St. Louis, Sept. 1976.
LI-120 Liptak, Bela G., Environmental Engineers' Handbook,
vols. 2 and 3. Radnor, PA., Chilton, 1974.
LI-150 (Arthur, D.) Little, Inc., The Impact of Producing Low
Sulfur. Unleaded Motor Gasoline on the Petroleum Refin-
ing Industry, Draft Report, Vol. 1. Study Results and
Planning Assumptions. EPA Contract No. 68-02-1332.
Task Order No. S, Cambridge, MA, Dec. 1975.
LI-151 (Arthur D.) Little, Inc., The Impact of Lead Additive
Regulations on the Petroleum Refining Industry, Draft
Report, Vol. 1, Study Results and Planning Assumptions.
EPA Contract No. 68-02-1332, Task Order No. 7, Cambridge,
MA, Dec. 1975.
LI-152 (Arthur D.) Little, Inc., Draft Copy: The Impact of SO
Emissions Control on the Petroleum Refining Industry.
EPA Contract No. 68-02-1332, Task Order No. 1, December,
1975.
MA-232 Marks, R. H., "Wastewater Treatment," Special Report,
Power, June 1967.
MC-389 McCabe, Warren L. and Julian C. Smith, Unit Operations
of Chemical Engineering, Third Edition.New York,
McGraw-Hill, 1976.
ME-095 Metcalf and Eddy, Inc., Wastewater Engineering--Col-
lection. Treatment. Disposal.McGraw-Hill Series in
Water Resources and Environmental Egineering. New
York, McGraw-Hill, 1972.
ME-218 Meserole, N. P., et. al., Review and Assessment of the
Existing Data Base Regarding Flue Gas Cleaning Wastes,
Draft Report.DCN 77-200-169-06, EPRI Research Projec
786-2, Task 1. Austin, TX, Radian Corp., Sept. 1977.
47
-------
BIBLIOGRAPHY (Continued)
ME-292 Meyer, William R. , "Solvent Broke," Cincinnati, OH,
Vulcan, Cincinnati, Inc., undated.
ME-294
Metcalf and Eddy, Inc., Wastewater Engineering: Treatment
Disposal, Reuse. 2nd Edition, New York, McGraw Hill
1979.
PA-193 Patterson, W. L. and R. F. Banker, Estimating Costs
and Manpower Requirements for Conventional Wastewater
Treatment Facilities"! Kansas City, MO., Black & Veatch
Oct. 1971.
PA-402 Patterson, R. L., "Difficulties Involved in the Estima-
tion of a Population Mean Using Transformed Sample
Data:, Technometrics 8. No. 3, (1966), 535-537.
PE-030 Perry, John H., Chemical Engineers1 Handbook, 4th Ed.
New York, McGraw-Hill, 19637
PE-065 Peters, Max S. and Klaus D. Timmerhaus, Plant Design
and Economics for Chemical Engineers, 2nd Edition.
McGraw-Hill Chemical Engineering Series, New York,
McGraw-Hill, 1968.
PE-224 Peeler, J. P. K. and K. L. Piggott, "Combined Gas Turbine-
Steam Cycle for Power Generation, Part 2, Theoretical
Study of a Coal-Fired System," Inst. Eng. Aust. Mech.
Chem. Eng., Trans. MC8(2), 130-35 (1972).
PE-277 Perry, John H., Chemical Engineers Handbook, 5th Edition,
New York, McGraw-Hill, 19737
PO-079 Porter, H. F., J. E. Flood, & F. W. Rennie, "Filter
Selection," Chem. Eng. 78(4), 39, 1971.
RA-596 Radian Corporation, Final Report; Control Techniques
for Volatile Organic Emissions from Stationary Sources,
EPA Contract No 7 200-187-2308, Austin, TX, 1978.
RA-729 Radian Corporation, Environmental Control Selection
Methodology for a Coal Conversion Demonstration
Facility. Final Report. DCN 78-200-151-06-06, DOE
Contract No. EX-76-C-01-2314, Austin, TX, Oct. 1978.
RO-423 Rosenberg, D. S., et. al., Assessment of Hazardous
Waste Practices in the Petroleum Refining Industry,
Final Report.EPA Contract No. 68-01-2288, PB 259
097, Pasadena, CA, Jacobs Engineering Co., June 1976.
48
-------
BIBLIOGRAPHY (Continued)
SE-066 Selmeczi, Joseph G. and R. Gordon Knight, "Properties
of Power Plant Waste Sludges," Paper #B-8, Presented
at 3rd Int'l. Ash Utilization Symposium, Pittsburgh,
PA, March 1973.
SH-418 Shaw, N. R., "Vapor Adsorption Technology for Recovery
of Chlorinated Hydrocarbons and Other Solvents," Pre-
pared for 68th Annual Meeting of the Air Pollution
Control Assoc., Boston, MA, June, 1975.
SL-034 Slack, A. V. and J. M. Potts, "Disposal and Use of By-
Products from Flue Gas Desulfurization Processes.
Introduction and Overview." Presented at the Flue
Gas Desulfurization Symposium, New Orleans, May 1973.
SL-084 Slimak, K. M., Landfill Disposal Systems. McLean, VA.,
TRW Inc., 1977.
ST-351 Stanford Research Institute, 1976 Directory of Chemical
Producers, U.S.A., Menlo Park, California, 1976.
ST-667 Straus, M. A., Hazardous Waste Management Facilities
in the United States - 1977. Publication SW-146.3,
Environmental Protection Agency, Cincinnati, OH, 1977.
TI-062 Tierney, D. R. and T. W. Hughes, Source Assessment;
Reclaiming of Waste Solvents, State of the Art,lPA
Contract No.68-02-1874, Monsanto Research Corporation,
Dayton, OH, April 1978.
US-209 U.S. Bureau of Mines, .Crude Petroleum, Petroleum
Products, and Natural Gas Liquids, December, 1974,
Mineral Industry Surveys, Petroleum Statement Monthly,
Washington, D.C., April, 1975.
VA-203 Van Ingen, R. E. and B. N. Bastian, "Factors Affecting
Waste Disposal Alternatives," Paper No. AM-76-31.
Presented at the 1976 NPRA Annual Meeting, San Antonio,
TX, March 1976.
VI-068 Viraraghanam, D. and R. C. Landine, "Economic Evaluation
of Four Sludge Dewatering Devices," Poll. Eng. 1978
(Sept.), 55-57.
WA-412 Wachter, R. A. et. al., Study to Determine Need for
Standards of Performance for New Sources of Waste
Solvents and Solvent Reclaiming, EPA Contract No^. 68-
49
-------
BIBLIOGRAPHY (Continued)
02-1411, Task 15, Monsanto Research Corporation,
Dayton, Ohio, February 1977.
WA-413 Water Pollution Control Federation, Technical Practice
Committee, Operation of Wastewater Treatment Plants,
A Manual of Practice, No. 2., Oct. 1975.
WI-197 Wisniewski, Ralph H., Ultimate Disposal Methods of
Refinery Sludges, Pittsburgh, PA, CHEMFIX, Inc.,
undated.
WI-331 Witt, P. A. Jr., "Disposal of Solid Waste," Chem. Eng.
78(22), 62-78, 1971.
YA-089 Yaverbaum, Lee, Fluidized Bed Combustion of Coal and
Waste Materials.Park Ridge, NJ, Noyes Data Corp.,
1977.
50
-------
APPENDIX I
EVALUATIONS OF SOLID WASTE PROCESSING AND DISPOSAL TECHNIQUES
TABLE OF CONTENTS
Page
Treatment
Vacuum Filtration 1-2
Pressure Filtration 1-7
Centrifugation 1-16
Dewatering Lagoons 1-24
Aerobic Digestion 1-29
Anaerobic Digestion 1-35
Chemical Fixation 1-43
Sand Drying Beds 1-47
Wet-Air Oxidation 1-53
Composting 1-57
BEST/Solvent Extraction 1-62
Final Disposal
Landfilling 1-66
Landfarming 1-71
Incineration 1-75
Disposal Lagoons 1-86
Disposal Wells 1-88
Ocean Disposal 1-90
1-1
-------
Process/Equipment: Vacuum
Filtration
Application:
S
Si
Sludge Dewatering
- Raw or Digested Biological
Treatment Sludges
- Slop Oil Emulsions
- Separator Sludges
Process Description; Vacuum filtration involves separation of
waste solids from water at the surface of a porous filter medium
under vacuum. Solids accumulate on the filter's surface while
water passes through as filtrate. The filter cake is removed as
the dewatered product. Although there are many variations to the
basic vacuum filter, it operates by drawing a vacuum on the dis-
charge side of a filter medium made of cloth, steel, tightly
wound coil springs, etc. Often a precoat layer is applied to the
filter drum to prevent oily solids residue from clogging the pores
of the filter medium. The vacuum provides the driving force for
filtrate flow through the filter medium. Solids in the feed are
physically strained from the filtrate. To increase capacity and
improve collection efficiency, precoats, flocculants, or coagulants
are frequently used. Figure 1 shows a rotary vacuum filter sys-
tem.
Development Status: Vacuum filters have been used by refineries
for many years to dewater sludges before further processing and/or
final disposal.
Performance:
Control Effectiveness; Vacuum filtration is most frequently
used to dewater sludge containing 5 to 10 percent solids.
Solids losses typically range from 2-10 percent. The
dewatered sludge contains 15-35 wt percent solids, with 25
percent being typical.
Process Reliability:
Complexity - Skilled labor is required.
Start-up and Shutdown - Can be performed easily.
Equipment - Vacuum system, filtration unit, sludge
pump,flocculant feeder and mixer.
Materials of Construction - Materials required for the
aforementioned equipment.
Raw Materials - Flocculants or coagulants may be needed.
1-2
-------
u>
AIR AND FILTRATE
CLOTH CAULKING
STRIPS
AUTOMATIC VALVE
DRUM
FILTRATE PIPING
CAKE SCRAPER
AIR BLOW.BACK LINE
SLURRY FEED
^SLURRY AGITATOR
VAT
EN-325
I
FIGURE 1. CUTAWAY VIEW OF A ROTARY DRUM VACUUM FILTER
-------
RADIAN
Secondary Pollutants; The secondary pollutants from vacuum
filtration are the filtrate and the dewatered sludge (filter
cake). The filtrate is sent to a separator where oil suit-
able for distillation or cracking is recovered and the water
layer recycled to the treatment plant. The dewatered sludge
requires disposal such as landfill, landfarm, or incineration.
Applicability; This sludge processing method has wide applica-
tion in dewatering refinery waste sludges. It is most suit-
able for low-volume streams where a high quality effluent is
desired.
Process Advantages:
• Concentration of solids to reduce disposal cost
• Compact equipment
Process Disadvantages:
• Skilled labor required
• Use of coagulants or flocculants
Cost and Energy Requirements:
Capital Costs; Capital costs for vacuum filters vary sig-
nificantly depending on the requirements of the specific
application and the filtering characteristics of the solids.
Costs are primarily based on the filter area and Figure 1
gives an estimate of the installed cost as a function of
the area.
An estimate of the total capital costs versus plant capacity
in tons of dry solids recovered per day is provided in
Figure 2. These costs are based on a filtration rate of
4 Ib dry solids/ft2 hr.
Operating Costs; An estimate of the current operating costs
is also shown in Figure 2. These costs include labor for
operations and maintenance, chemicals for sludge condition-
ing prior to filtering, and maintenance materials.
1-4
-------
3lO»
I t (I 1 | l_t J | ! t ' . t 1 t I '
10 10» 10*
Tllt«r Surfae* Ar««. Squ»r« FMC
10*
July 1977 cost
index - 204.5
December 1978
cost index «
225
CU-OS2
FIGURE 1. VACUUM FILTRATION CAPITAL COSTS (JULY 1977)
a.
u ,
FIGURE 2.
10*
rir
10«?
3
3
n
10 «,
W
S
(A
o
10 10'
PlJBC Capacity. Ten* Or* Solid*/Day
10'
July 1977 cost
index - 204.5
December 1978
cost index -
CU-082
VACUUM FILTKATIOn CAPITAL AND O&M COSTS (JULY 1977)
1-5
-------
RADIAN
CORPORATION
Energy Requirements: Energy is required for sludge pump,
flocculant feeder and mixer, vacuum pump, etc. The reported
annual requirements were 340,000 kwh for one case* and
230,000 kwh for another case.**
Manufacturers Dorr-Oliver, Eimco, References: AM-046, EN-295,
or Developer: Komline-Sanderson, MA-232, PO-079, CU-082,
Envirex EN-325, CU-087, AM-042,
AM-062
*10xl03 gal/day wastewater treatment plant design parameters for
vacuum filter are feed solids, 5 percent, yield 2.8 lb/hr/ft2,
cake moisture 75 percent, and lime dose 4 percent by weight.
**10xl03 gal/day wastewater treatment plant, design parameters for
vacuum filter are feed solid 12 percent, yield 8 Ib/hr ft*,
cake solids 24 percent, polyelectrolyte dose 10 Ib/ton of dry
solids.
-------
Process/Equipment: Pressure Filtration Application;
Sludge Dewatering
Process Description: Filtration may be defined as the removal of
solid particles from a fluid by passing the fluid through a fil-
tering medium, or septum, on which the solids are deposited.
The two major categories of filtration apparatus are pressure and
vacuum, which may be further classified by continuous or batch
operations. This description includes continuous and batch pres-
sure filters. The major types that find use in processing refin-
ery wastes are:
• Filter presses (plate-and-frame)
• Leaf filters
• Horizontal belt filters
The first two are discontinuous filters; the third is a continu-
ous filter designed to compete with rotary vacuum filters. Fig-
ures 1 through 3 illustrate different views of the three differ-
ent systems.
The plate-and-frame filter press, as the name implies, contains
alternate solid plates, the faces of which are studded, grooved,
or perforated to permit drainage, and hollow frames, in which the
cake collects during filtration. The plates, most often rectangu-
lar, are covered with a fabric filter, and are hung on parallel
bars. Hydraulic rams or powered screws are used to hold the plates
together. Slurry is admitted to each compartment under pressures
of 60 to 225 psi; liquor passes through the canvas and out a dis-
charge pipe, leaving a wet cake of solids behind. Typical filtra-
tion times range from 1 to 3 hours, making a total cycle time of
2 to 5 hours. The most common problem with filter presses is the
clogging or blinding of the pores in the filter cloth.
The horizontal leaf filter permits higher operating pressures and
less labor requirements than the plate-and-frame press. However,
its capital costs are much higher for a given quantity of filtra-
tion capacity. Leaves covered with fabric filter are vertically
hung and inserted in a closed tank. Slurry at high pressure is
admitted to the tank and allowed to filter through the noncontigu-
ous leaves. Filtrate is drained off through a discharge tube and,
after sufficient time for full caking, the tank opened and dewatered
product removed. Batch and cycle times are similar to the plate-
and-frame system.
In response to the high capital costs and labor expenses associ-
ated with the above discontinuous filters, many refiners have
introduced the horizontal belt gravity filter press to dewater
1-7
-------
FIXED END
1
•
Q
Q
Jo
OPI HATING HANDLE,
\ m M
eucnic
CLOSING GIAI
7
(a) Side View
surooi IN
RITIATt DRAIN MOU5
(b) Cutaway View
FIGURE 1
FILTER PRESS
EN-325
1-8
-------
Filtrate dischor
manifold
MC-389
(a) Isometric View
'..'K>O>—-etoth
(b) Cutaway View
FIGURE 2
HORIZONTAL LEAF FILTERS
PE-277
1-9
-------
Pr««» Belt
Draining
Zone
Press
Zone
Shear
Zone
Drive Rollers
Scraper
FIGURE 3
A SCHEMATIC OF THE GRAVITY BELT FILTER PRESS
EN-325
I-10
-------
sludges. This filter consists of two continuous belts mounted on
horizontal guides and rollers. Direction and speed of the two
belts are identical. Three zones of varying pressures effect
dewatering. In the initial draining zone, gravity is the driving
force. Static pressure is applied in the pressure zone and, in
the shear zone, the belt guides cause the belts to be displaced
relative to each other; this generates shear forces which accom-
plish final dewatering. A scraper or knife blade removes dried
cake. Based on suspended solids of 3 to 6 wt. percent in the
feed, this friable cake contains 20 to 35 wt. percent solids, with
30 wt. percent being typical.
Development Status: Pressure filtration is a common unit opera-
tion in use for many years. Horizontal belt filters are widely
used in Europe; these systems were introduced to the United
States in 1970.
Performance:
Control Effectiveness; Data for the different systems is
presented in the following table:
Type of System
Filter press
Horizontal leaf
Horizontal
gravity belt
Type of
Municipal Sludge
Raw primary
Digested
EAS
Raw primary
Digested or EAS
Suspended
Solids in
Feed (wt Z)
5-10
6-10
<5
Cake
Solids
(wt %)
50
50
50
3-8 50
(Not Available)
Raw primary
Digested
Digested with
EAS
3-6
4-7
10
29-35
39-44
38
Sources: EN-325, PE-277, AS-070
Process Reliability:
Complexity - Skilled labor is required.
Start-up and Shutdown - Again, skilled labor is manda-
tory for batchwise operation.
Equipment - Pumps, filtration apparatus, handling equip-
ment for filter aids.
1-11
-------
Materials of Construction - No special materials re-
quired.
Raw Materials - Filter aids such as flocculants or co-
agulants may be required. Polymers are the most widely
used chemical additive for belt filter presses; FeCl3
and CaO are the usual additives for plate-and-frame
presses.
Secondary Pollutants; Filtrate is sent to an oil/water
separation unit before recycling to the wastewater treatment
facility. Dewatered filter cake must be disposed of in a
landfill or another environmentally acceptable manner.
Applicability: The use of pressure filters is applicable to
most refinery sludges. Slurries of fines are also amenable
to this treatment technique.
Process Advantages:
• Concentration of solids to reduce disposal cost.
• Compact equipment/low floor space requirements.
• No complex equipment or processing.
• Eliminates high land requirements.
• High solids capture.
Process Disadvantages:
Batch filters:
• Skilled labor required.
• Filtrate and cake require further treatment.
• High maintenance usually required.
• Pretreatment chemical conditioning often required.
Horizontal continuous gravity belt filters:
• Lower solids cake than batch filters.
• Filtrate and cake require further treatment.
• Pretreatment chemical conditioning required.
1-12
-------
Costs and Energy Requirements;
Capital Costs; Capital costs vary widely depending on the
filtration characteristics of the sludge or slurry, the
materials of construction required, the amount of auxiliary
equipment required, and the filter capacity. Tables 1 and 2
show purchased cost data for batch filters.
Operating Costs; Table 3 compares capital and operating
costs for a proposed municipal sludge dewatering system.
Energy Requirements; For above system:
Filter Press Belt Filter Press
Power 14 kwh/ton dry solids 7 hp
Labor 20 hr/wk 7 hr/wk
Chemicals .13 ton/ton dry solids 5 Ib/ton dry solids
Manufacturers Batch filters: Numerous. References: MC-389,
or Developer: Belt filter presses: PE-065, PE-277, EN-325,
Ashbrook Simon Hartley, AS-070, VI-068, AM-042,
Tait-Andritz, Parkson CU-082
Corp., Ralph B. Carter
Co.
1-13
-------
TABLE 1
PURCHASED COST OF PLATE-AND-FRAME FILTER PRESS
(DECEMBER 1978 DOLLARS)
Type of Filter
and Material
Plate- and-Frame
CF-8M stainless steel
PVC-coated iron
Cast iron
Filter Area, Ft2
10
$1,650
$1,550
20
$2,950
$2,360
$1,150
40
$5,300
$3,630
$1,800
60
$7,450
$4,600
$2,320
80
$9,400
$5,450
$2,800
100
$11,500
$ 6,300
$ 3,200
Source: PE-065.
TABLE 2
PURCHASED COST OF HORIZONTAL LEAF FILTERS
(DECEMBER 1978 DOLLARS)
Type of Filter
and Material
Horizontal Leaf Filter
304 stainless steel
Mild steel
Filter Area, Ft2
2
$9,900
$4,500
5
$16,000
$ 7,780
10
$21,800
$11,000
20
$28,800
$15,500
30
$34,000
$18,400
40
$37,800
$20,800
Source: PE-065.
1-14
-------
TABLE 3
COMPARISON OF COSTS FOR TWO FILTER SYSTEMS
Capital Cost
Equipment1
Installation2
Total
Annual O&M Cost
Operator Wages3
Power1*
Chemicals5
Maintenance & Repair6
Sludge Transport
Total
Amortized Capital &
Interest Per Year8
Total Annual Cost
Total Cost Per Ton of
Dry Solids9
Filter Press
117,650
23.120
140,770
8,400
500
39,600
3,800
9.700
62,000
22,900
84,900
80.1
Belt Filter Press
101,000
10.000
111,000
2,900
400
10,700
3,300
12,800
30,100
18,000
48,100
45.4
Basis:
NOTES:
5,800 Ib dry solids treated per day.
1 - Costs include chemical mix tank, chemical feed and sludge pumps
and accessories; costs also include taxes and freight.
2 - Installation cost at 20 percent of equipment for filter press;
12 percent of equipment for belt filter press.
3 - Wages - $8/hr.
" - Power @ 3c/kwh.
5 - Dry FeCl3 @ $505/ton and dry CaO @ $60/ton for filter press;
polymer @ $2/lb of polymer for belt filter press.
6 - M&R @ 4 percent of equipment cost.
7 - Cake solids - 33 percent for filter press; 25 percent for hori-
zontal belt filter with transport cost @ .15/ton-mile.
8 - Amortization of 10 years at 10 percent interest - CRF « .16275.
9 - Dry solids - 1060 ton/year.
Source: VI-068.
1-15
-------
RADIAN
Process/Equipment: Centrifuge Application;
Sludge Dewatering
Process Description: Centrifugation is a mechanical process using
centrifugal force to speed up the sedimentation rate of sludge
solid particles. A sludge is pumped into the centrifuge which is
essentially a rotating drum where centrifugal forces act to hold
the solids against the walls while the water stream is removed.
The dewatered sludge can be removed from the centrifuge either
as a continuous stream or in a batch manner. There are four
types of centrifuges:
• Countercurrent solid bowl conveyor centrifuge.
• Continuous concurrent solid bowl conveyor centrifuge.
• Basket centrifuge.
• Disc centrifuge.
Figures 1 through 4 show the various types of centrifuges.
Development Status: Centrifuges of various types have been em-
ployed for solid-liquid separation processes in refineries for
many decades.
Performance:
Control Effectiveness; Sludges can be dewatered to 20 to
40 percent solids.^?ith batch operation and long residence
times, this may be increased slightly. The effectiveness in
dewatering sludge is highly dependent on the physical proper-
ties of the sludge, especially the particle morphology. In-
dustrial sludges such as limestone scrubber sludges can be
dewatered up to 50-60 weight percent solids. Table 1 pre-
sents typical data which can be expected with the solid bowl
configuration of various organic sludges.
Process Reliability:
Complexity - Moderately simple process; however, oper-
ates with variable success depending on sludge proper-
ties. Skilled labor is generally required.
Start-uo and Shutdown - Simple.
1-16
-------
.COVER
8!
1
DIFFERENTIAL SPEED
GEAR BOX
MAIN DRIVE SHEAVE
L-::H"-:^P-.
LJ^1
^BEARING
'ROTATING
CONVEYOR
PIPES
(SLUDGE AND
CHEMICAL)
BASE NOT SHOWN
CENTRATt
DISCHARGE
SlUDGE CAKE
DISCHARGE
t
EN-325
FIGURE 1. CONTINUOUS COUNTERCURRENT SOLID BOWL CONVEYOR DISCHARGE CENTRIFUGE
-------
OIAIN IliO
"
CASt _ ftOC rOOt INIflNAl
SOI IDS DISCHARGE
AND f IOW4
PlliOW HOCK AND
MAIN SCARING
COnvtvOM
Of Ai DRIVE
SKIMMER
C CONItOl
IO8OUI OVitlOAO
SWIICH
ft to
INUI
HOC
INI II
ffflUfNI
OISCHARGC
DRIVEN
SltfAVt
fCIO CONVfYOft
riff
SP1ASH
COMPARIMtNl
IRUNNION
SfAlS
SOUOS
OlSCHARGf
• ASt
EN-325
FIGURE 2. CROSS SECTION OF CONCURRENT FLOW SOLID BOWL CENTRIFUGE
-------
FEED
POLYMER-,
SKIMMINGS
CAKE
CAKE
EN-325
FIGURE 3. SCHEMATIC DIAGRAM OF A BASKET CENTRIFUGE
1-19
-------
RADIAN
FEED
^ EFFLUENT
DISCHARGE
SLUDGE
DISCHARGE
EN-325
FIGURE 4. DISC TYPE CENTRIFUGE
1-20
-------
RADIAN
TABLE 1
TYPICAL SOLID BOWL CENTRIFUGE PERFORMANCE
Wastewater Sludge Type
Raw or Digested Primary
Raw or Digested Primary, Plus
Trickling Filter Humus
Raw or Digested Primary, Plus
Activated Sludge
Sludge Cake Characteristics
Solids
Solids Recovery Chemical
Percent Percent Addition
28-35
20-30
15-30
70-90
80-95
80-95
no
yes
yes
EN-325
1-21
-------
RADIAN
Equipment - Equipment required includes centrifuge,
sludge pump, and flocculant feeding systems. The high-
speed equipment is relatively expensive but eliminates
the need for bulky equipment and floor space. Proper
balancing of the rotating bowls and durable bearings is
essential in order to minimize relatively high mainten-
ance costs.
Materials of Construction - No special materials of con-
struetion are usually required. When a sludge contains
abrasives, special materials may be required for internal
parts of the machine.
Raw Materials - Flocculants are almost always needed to
improve solid capture.
Secondary Pollutants: The concentrated solids or sludges
represent a secondary waste that must be disposed. The
centrate is generally sent back to the treatment plant.
Applicability: The principal application of this treatment
method in petroleum refineries is the separation of solids
from an oil/water/solids slurry.
Process Advantages:
• Solids concentration.
• Compact equipment.
• High throughput.
Process Disadvantages:
• Skilled labor.
• Energy costs.
• Maintenance costs.
1-22
-------
Costs and Energy Requirements:
Costs: Listed below are capital and operating and mainten-
ance costs for one centrifuge which is designed to handle
40 gpm of 5 wt percent municipal sludge. The sludge is de-
watered to 40 wt percent solids. A capacity factor of 80
percent was assumed.
Installed Capital Costs
Operating and Maintenance
Costs
$313,600 (December 1978)
$135,000/year (December 1978)
Energy Requirements: The major energy consumer for the
above centrifuge is a 30 h.p. electric motor.
Manufacturers Numerous
or Developer;'
References; AM-042, AM-046,
EN-295, MA-232, PA-193,
PE-030, RO-423, SE-066,
SL-034, EN-325, CR-154
1-23
-------
Process/Equipment; Dewatering Lagoons Application:
(Evaporation Ponds) Sludge Dewatering
Process Description; This process is similar to the evaporation
pond process described in the ultimate disposal subsection of
this report, except that solids in a dewatering lagoon are peri-
odically dredged from the lagoon. This technique is also similar
to the sand drying bed — the only differences being the lack of
filtration media and underdrains. Lagooning is included in this
subsection because it is the most common refinery practice for
sludge dewatering. Both solar energy and gravity are used to de-
water the sludge. The solar energy flux on the lagoon surface
causes water to evaporate; this concentrates the suspended solids.
Often, though, dewatering due to gravity settling is more effec-
tive in separating the two phases. After sufficient accumulation
of the solids occurs at the lagoon bottom, they must be dredged
and further processed.
The design of dewatering lagoons is affected by the net evapora-
tion rate, i.e., the gross evaporation rate minus rainfall. Geo-
logical characteristics of the area are also a major, factor. If
the soil is highly permeable, the lagoon must be lined with clay
or a synthetic liner. In some areas, state or local regulations
may prohibit dewatering lagoons. Lagoons can be constructed by
excavating a hole, typically 8 to 15 feet deep; enclosing an area
with dikes; building a dam or a combination of the above.
Development Status: Dewatering lagoons have been used for many
decades by the petroleum refining industry for the disposal of
refinery sludges.
Performance:
Control Effectiveness: All solid waste streams that require
dewatering can potentially be dewatered by lagoons, but land
availability and evaporation rate may limit high flow rate
streams from being discharged to a lagoon. The major limi-
tations to the control effectiveness of drying lagoons even
in arid regions is the problem of seepage of toxic compon-
ents into groundwater supplies. Even with the most care-
fully designed lined dewatering lagoon, there is always the
possibility of rupture of the lining and subsequent leakage.
There is also the question of the effectiveness of lagoon
liners. Chemical reactions of the contents of the lagoon
with the clay liners has been observed which increases liner
permeability and thus the possibility of groundwater contam-
ination. The volatile organic compounds are evaporated.
1-24
-------
Process Reliability:
Complexity - Simple process.
Start-up and Shutdown - Simple.
Equipment - No special or unique process equipment
required.
Materials of Construction - Clay or synthetic liners
may be needed.
Raw Materials - No raw materials required.
Secondary Pollutants; These ponds are seldom designed for
environmental compatability. Only if adequate lining and
continued monitoring are provided does this technique pro-
vide sufficient water and land protection. Even then,
evaporation of volatile organic compounds occurs. These
air emissions, coupled with a history of ineffective leach-
ate control, have elicited heavy regulation from the govern-
ment which curtails the use of dewatering lagoons.
Applicability: Have been demonstrated in refineries and
chemical process industries.
Costs and Energy Requirements;
Costs: In Figures 1 and 2, capital and operating costs of
dewatering lagoons are shown as a function of water rate to
the lagoon with the effective net evaporation rate as a para-
meter. The effective net evaporation rate is less than the
net evaporation rate due to the reduction of the water vapor
pressure that is caused by the dissolved salts in the water
in dewatering lagoons. The capital cost figure is based on
a 4 foot high dike with a 27 foot dike base, and includes
liner and land costs. The cost of excavation can range from
$2 to $6/cu yd depending upon the type of land to be exca-
vated. Liner costs can range from $0.20 to $0.40/ft2 depend-
ing upon the type and thickness of liner used. Land costs
can range anywhere from $200/acre to $10,000/acre or more.
Energy Requirements: The major energy requirement for de-
watering lagoons is for pumping wastewater to the lagoon.
1-25
-------
RADIAN
CORPORATION
0.1
SYSTEM CAPACITY (MILLION GALS/OAY)
I 60 INCHES
EFFECTIVE NET EVAPORATION RATE
-------
CORPORATION
tfr
I
o
0)
x
(0
o
o
i
P
Ul
0.1-
0.01
III
0.1
10
SYSTEM CAPACITY (MILLION GALS/DAY)
100
CO-698
I 60 INCHES
EFFECTIVE NET EVAPORATION RATE «/YEAR)' II 30 INCHES
III 20 INCHES
FIGURE 2. DEWATERING LAGOON OPERATING COSTS
1-27
-------
The energy requirement for pumping 1000 gallons a distance
of 1000 feet is about 0.5 kwhr (based upon a 5 ft/sec flow
velocity).
Manufacturers Numerous. References: AM-042,
or Developer? CO-698, EN-818, RO-423
1-28
-------
Process/Equipment: Aerobic Digestion Application:
Waste Biological Sludge
Volume Reduction and
Stabilization
Process Description; Aerobic digestion is similar to the acti-
vated sludge process. The primary difference is a change in
respiration type due to extended detention times in the diges-
ters. As the available dissolved organics are depleted, the
bacteria begin to consume themselves (endogenous respiration)
to obtain the energy required in cell metabolism. The funda-
mental concept is that biological self-destruction occurs at
a faster rate than microbial growth.
Figure 1 shows a schematic of an aerobic digestion system. Waste
biological sludge is admitted to the tank where it undergoes
aeration and mixing. Bacteria consume the organic nutrients
present before metabolizing themselves. Supernatant is drawn
off, clarified, and recycled to the head of the treatment plant.
Thickened sludge is drained and dewatered or sent immediately
to a landfarming or landfilling site.
The important process parameters in conventional air aerobic
digestion are:
• Oxygen requirements
• Detention time
• Sludge age
• Temperature
• Biodegradable volatile solids level
• pH
High-rate volatile solids reduction occurs in the first 10-12
days, typically with 35 percent to 45 percent of the volatile
suspended solids (VSS) removed. Fifteen days of detention time
are commonly provided for stabilization of excess biological
sludges. Loadings will normally vary from .1 to .2 Ib VSS/ft3
day. If the influent sludge has a VSS content of greater than
3 percent, an oxygen-rich stream must be provided which has a
higher oxygen fraction than atmospheric air. Figure 2 depicts
a typical continuous flow aerobic digester.
Development Status; Aerobic digestion has been used for many
years in municipal wastewater treatment systems. It has also
been extensively used in petroleum refinery excess activated
sludge stabilization.
1-29
-------
PRIMARY SLUDGE
i
u>
o
EXCESS
ACTIVATE DOR
TRICKLING FILTER
SLUDGE
CLEAR
OXIDIZED
OVERFLOW
TO PLANT
SETTLED SLUDGE RETURNED TO AERODIGESTER
FIGURE 1
EN-325
A SCHEMATIC OF AN AEROBIC DIGESTION SYSTEM
-------
AIR PLUG VALVE
SUPERNATANT
DRAW OFF
x
CONTROL BAFFLE
AIR LINES
DECANT
HAMBER
4" RETURN SLUDGE
AIRLIFT PUMP
WASTE-*.
SLUDGE
DRAW-OFF
EDUCTOR TUBE
EN-325
FIGURE 2
A TYPICAL CIRCULAR (CONTINUOUS FLOW) AEROBIC DIGESTER
1-31
-------
Performance:
Control Effectiveness; Total VSS removal ranges from 40
to 50 wt percent at fifteen days detention time. The
recycled supernatant typically contains 500 mg/5, of
BODS, 2600 mg/2, of COD, and suspended solids of 3400 mg/£
The sludge is essentially pathogen-free.
Process Reliability; Aerobic digestion is a proven
process; yet some questions still exist regarding design
parameters, process kinetics, and economics.
Complexity - Similar to most biotreatment techniques,
this process is complex because temperature, pH, and
oxygen transfer must be closely controlled. Required
analytical monitoring adds to the complexity.
Start-up and Shutdown; Seeding of this system and
initial loading and draining of sludge must be care-
fully undertaken.
Equipment/Materials of Construction; Circular con-
crete or steel tanks are most often used.
Raw Materials: Seed sludge and air/oxygen; auxiliary
heat may be required during cold months.
Secondary Pollutants; The supernatant contains low amounts
of BOD, total P, and kjeldahl N; it is high in suspended
solids. Sludge must be disposed of in a landfarm or land-
fill. Odors may require control in high wind areas.
Applicability; Aerobic digestion has been applied by a
significant number of refineries in treatment of waste
biosludges.
Process Advantages:
• Low capital costs
• No generation of significant odors
• Reduces pathogens to low levels
• Low BOD concentrations in supernatant
• Production of an easily dewatered, biologically
stable end product
1-32
-------
100
10 100
DRY SOLIDS (TON/DAY)
NOTES:
1. Minnwpolis. Mar.. 1972. ENR Construction Con Imtoc of 1827.
2. Amortization at 7% for 20 yaars.
3. Influent tludo* of 38% primary and 62% waste activated riudgt with a
•aollds. content of 33%.
4. 20 day volumetric displacement time.
b. Souce: EPA Cost and Manpower Report and Stanley Consultants.
FIGURE 3
.01
1000
March 1972
Cost Index - 136.2
December 1978
Cost Index « 225.0
CAPITAL COSTS OF AN AEROBIC DIGESTER
1-33
-------
• Recovery of fertilizer values
• Fewer operational problems than anaerobic
digestion.
Process Disadvantages:
• High operating costs
• Less developed design parameters and kinetics
• Close process control required
• No useful by-products, e.g., methane
Costs and Energy Requirements;
Capital Costs; Figure 3 depicts the capital costs.
Operating Costs; These vary widely depending primarily
on the nature of the influent sludge. A value of $3.90/yr
Ib BOD removed was given for a small Pennsylvania municipal-
ity.
Energy Requirements; Sludge pumps and heaters are required.
Aerator costs range from 20-40 kw/1000 m3 of tank volume.
Manufacturer Numerous. References; EN-325,
or Developers; ME-294, WA-413
1-34
-------
Proces s/Equipment: Anaerobic Application;
Digestors Waste Biosludge
Process Description; Anaerobic digestion is used to stabilize
sludge, particularly in domestic applications. In the absence
of oxygen, the faculative and anaerobic bacteria continue to grow
and digest the organic matter. This digestion consists of two
stages:
1) Hydrolysis of high-molecular-weight organics
and conversion to organic acids.
2) Gasification of the organic acids to methane
and carbon dioxide.
The standard rate and high rate systems are the two main digestion
processes employed. Schematics of the processes as well as their
operating criteria are given in Figure 1.
In practice, four types of systems have evolved from the two basic
digestion modes. They are:
1) Standard Rate Digestion-One Stage
2) High Rate Digestion-One Stage
3) Two-Stage Digestion (Figure 2)
4) Anaerobic Contact Process (Figure 3)
Generally, the organic loading for an anaerobic system is 2-3 times
greater than that for aerobic systems. However, overloading of
the anaerobic system can drop the pH to toxic levels. The pH
control is therefore very critical and should be maintained in the
range from 6.2 - 7.5 in the anaerobic digester.
Development Status; This process is commercially proven, and is
available through numerous wastewater engineering contractors.
Performance:
Control Effectiveness: Anaerobic digesters typically reduce
organic solids 50 percent by weight.
1-35
-------
SLUDGE IN LET ~
GAS OUTLET
SCUM LAYER
H^y////////////.
SUPERNATANT
ACTIVELY
DIGESTING SLUDGE
STABILIZED
SLUDGE
[SUPERNATANT
\ REMOVAL
SLUDGE OUTLET
(A)
STANDARD RATE DIGESTION
1. UNH GATED
Z DETENTION TIME 30 - 60 DAYS
3. LOADING 0.03 - 0.10 Ib. VSS/cu. ft./day
4. INTERMITTENT FEEDING AND WITHDRAWAL
5. STRATIFICATION
A
C
T
1
V
E
^
Z
o
N
E
^
— SLUDGE
— INLET
SLUDGE OUTLET
(B)
HIGH RATE DIGESTION
1. HEATED TO 8S° - 9S° F
2. DETENTION TIME 15 DAYS OR LESS
3. LOADING 0.10 - O.SO Ib. VSS/cu. ft/day
4. CONTINUOUS OR INTERMITTENT FEEDING
AND WITHDRAWAL
5. HOMOGENEITY
FIGURE 1. STANDARD RATE AND HIGH RATE DIGESTION
-------
GAS
RELEASE
SLUDGE
INLET I
I
Co
ZONE OF
ACTIVELY
DIGESTING
SLUDGE
I
!Z
GAS
RELEASE
MIXED
LIQUOR
SUPERNATANT
REMOVAL
SLUDGE DRAWOFF^
DIGESTED SLUDGE
TO FURTHER PROCESSING
EN-325
FIGURE 2. TWO-STAGE ANEROBIC DIGESTION
-------
I
J*
oo
GAS
RELEASE
SLUDGE
INLET _
ZONE OF
ACTIVELY
DIGESTING
SLUDGE
SLUDGE RETURN
GAS
RELEASE
MIXED
LIQUOR
SUPERNATANT
REMOVAL
SLUDGE
DRAWOFF
SUPERNATANT
DIGESTED SLUDGE
EN-325
FIGURE 3. ANAEROBIC CONTACT DIGESTION
-------
Process Reliability;
Complexity - This process is complex because specific
temperature and alkalinity levels need to be maintained.
The analytical monitoring required for this system also
adds to its complexity.
Start-up and Shutdown - Seeding of this system from a
well operating sewage sludge digestor and careful addi-
tion and withdrawal of sludge required for start-up.
This can be a difficult procedure.
Equipment - The basic anaerobic digestion process
unit consists of a circular tank with a cover, either
fixed or floating. The equipment is readily available.
Materials of Construction - No unusual materials are
required tor tills process.
Raw Materials - None other than seed sludge.
Secondary Pollutants; The digestor supernatant is high in
suspended, solids, BOD, and ammonia nitrogen. This waste-
water is normally recycled to the head of the treatment
plant. Evolved methane gas requires proper venting or re-
covery .
Applicability: Anaerobic digestion has not been applied to
waste biosludges from refineries. Further study is required
to determine its applicability.
Process Advantages:
• Methane Recovery.
• Solids stabilization and enhancement of dewatering
properties.
• Small process area required.
• Volatile solids destruction.
• Use of digested sludge as fertilizer or soil
conditioner.
1-39
-------
Process Disadvantages:
• Temperature control required.
• Alkalinity control required.
• Skilled operation required.
• Excess volatile acids generation can cause process
upsets.
• Odors.
• Explosion hazard.
Costs and Energy Requirements:
Capital Costs: Figure 5 shows the relationship of capital
costs to digested sludge volume.
Operating Costs; Figure 4 shows the relationship of annual
operating costs to digested sludge volume.
Energy Requirements; Energy is required for the pump and
digester heater.No specifics on these energy requirements
are available.
Manufacturers Numerous References; BI-043, AM-062,
or Developer: CH-353, CU-082, CR-154,
EN-325
1-40
-------
100.000
9
8
7
6
9
4
3
2
i
j
o
o
z
10,000
9
8
7
6
5
1,000
1
6
3
4
100
10
I I
3 4 36789
100
3 4 5 6 789
1.000
SLUDGE VOLUME. 1,000 CUBIC FEET
July 1977 cost index = 204.5
December 1978 cose index » 225
2 3 4 3 S 739
10.000
CU-082
FIGURE 4 OPERATING COSTS VS. SLUDGE VOLUME
July, 1977 Dollars*
^Projected from 1976.
1-41
-------
10,000
IB
i
!Z
o
o
o
1,000
4J
(0
o
o
nt
4J
•H
fr
o
100
10
L
L
100 1000
Sludge Volume (Cubic Feet) x 1000
July 1077 cost index = 204.5
December 1978 cost index = 225
10,000
CU-082
FIGURE 5 CAPITAL COSTS VS. SLUDGE VOLUME
JULY. 1977 dollars-
"Trojected from 1976.
-------
Process/Equipment: Chemical Fixation- Application:
Chemfix® Stabilization of
Hazardous Sludges
Process Description: In the Chemfix® process, sludge is pumped
into a mobile van 40 feet long and 8 feet wide. This unit con-
tains the chemical storage, metering and mixing equipment for
operating the process at flow rates of 300-500 gallons per minute.
The process is continuous and occurs at ambient temperatures and
pressure. Process control is maintained by automatic equipment
which meters the required ratios of sodium silicate and Portland
cement into the water as it flows through the van. The treated
sludge then requires 24-72 hours set-up time under normal condi-
tions before it can be excavated and trucked to landfill. The
Chemfix® process chemistry consists of three reactions: first, the
diffusion-controlled precipitation of nearly all heavy polyvalent
metal ions with silicate ions supplied from soluble sodium sili-
cate; second, a gel-forming reaction between the setting agent
(Portland cement) and the silicate which absorbs large quantities
of both water and oily solids through physical and chemical fixa-
tion; and finally, reaction of the cement with the waste and/or
water further increasing the solids content. The process is not
designed to reduce the permeability of the waste, but to bind the
constituents together chemically. This process supplies adequate
leachate quality while providing structurally stable properties.
The volume of chemicals added to the waste is less than 10 percent.
Development Status: The Chemfix® process has been tested by a
broad range of industries such as refineries, chemical companies,
electronic companies, municipalities with sewage and landfill
problems, steel plants, automobile assembly plant, power plants
and others. Chemfix® has also participated in the EPA scrubber
sludge disposal program underway at TVA's Shawnee plant. In this
pilot plant demonstration, Chemfix® treated the clarifier underflow
stream from a limestone-based turbulent contact abosrber. The
clarifier and underflow averaged 38 wt percent which contained 38
wt percent fly ash on a dry basis during the filling period Decem-
ber 7-8, 1974. At least one Texas refinery has used this process
to treat API separator bottoms and crude tank bottoms. By 1974,
20 million gallons of refinery wastes had been treated by this
process.
The current status of this technology is not clear. The patent
rights have recently been sold to National Environmental Controls,
Inc., Metair, Los Angeles, and their marketing plans have not been
defined.
1-43
-------
Performance:
Control Effectiveness: Sludge with solid concentrations from
35-53 percent by weight have been stabilized successfully.
The Chemfix® product has unconfined compressive strengths
ranging from 24 to 154 psi, bulk density 1.4 g/cm3, and per-
meability ranging from 5x10"6 to 5x10"6 cm/sec. (Materials
with coefficients of 10"6 cm/sec and lower are considered
impervious for civil engineering purposes.)
Process Reliability;
Complexity - Process control is maintained automatically
and is relatively simple.
Start-up and Shutdown - Can be performed easily.
Equipment - Chemfix® mobile van which contains the
chemical storage, metering and mixing equipment.
Material for Construction - No unusual materials.
Raw Materials - Sodium silicate and Portland cement are
required for Chemfix® process. Lime addition may be
needed if pH of the sludge is below 4.
Secondary Pollutants: Potential of contaminating groundwater
and surface water by the leachate of Chemfixed product still
exists.
Applicability: This process is used at some refineries. It
has been proposed as a method of permanent stabilization of
hazardous waste of old refinery lagoons and pits. It may
also be a viable technique for stabilizing hazardous tetra-
ethyl lead wastes. Further research is needed to study the
applicability.
Process Advantages:
• Improve leachate quality.
• High continuous throughput rate.
• Small volume increase due to chemical additives.
• Ability to react with complex waste mixtures.
• Non-toxicity of the solid material.
1-44
-------
Process/Equipment: Chemical Fixation- Application:
Chemfix® Stabilization of
Hazardous Sludges
Process Description: In the Chemfix® process, sludge is pumped
into a mobile van 40 feet long and 8 feet wide. This unit con-
tains the chemical storage, metering and mixing equipment for
operating the process at flow rates of 300-500 gallons per minute.
The process is continuous and occurs at ambient temperatures and
pressure. Process control is maintained by automatic equipment
which meters the required ratios of sodium silicate and Portland
cement into the water as it flows through the van. The treated
sludge then requires 24-72 hours set-up time under normal condi-
tions before it can be excavated and trucked to landfill. The
Chemfix® process chemistry consists of three reactions: first, the
diffusion-controlled precipitation of nearly all heavy polyvalent
metal ions with silicate ions supplied from soluble sodium sili-
cate; second, a gel-forming reaction between the setting agent
(Portland cement) and the silicate which absorbs large quantities
of both water and oily solids through physical and chemical fixa-
tion; and finally, reaction of the cement with the waste and/or
water further increasing the solids content. The process is not
designed to reduce the permeability of the waste, but to bind the
constituents together chemically. This process supplies adequate
leachate quality while providing structurally stable properties.
The volume of chemicals added to the waste is less than 10 percent.
Development Status: The Chemfix® process has been tested by a
broad range of industries such as refineries, chemical companies,
electronic companies, municipalities with sewage and landfill
problems, steel plants, automobile assembly plant, power plants
and others. Chemfix® has also participated in the EPA scrubber
sludge disposal program underway at TVA's Shawnee plant. In this
pilot plant demonstration, Chemfix® treated the clarifier underflow
stream from a limestone-based turbulent contact abosrber. The
clarifier and underflow averaged 38 wt percent which contained 38
wt percent fly ash on a dry basis during the filling period Decem-
ber 7-8, 1974. At least one Texas refinery has used this process
to treat API separator bottoms and crude tank bottoms. By 1974,
20 million gallons of refinery wastes had been treated by this
process.
The current status of this technology is not clear. The patent
rights have recently been sold to National Environmental Controls,
Inc., Metair, Los Angeles, and their marketing plans have not been
defined.
1-43
-------
Performance:
Control Effectiveness; Sludge with solid concentrations from
35-53 percent by weight have been stabilized successfully.
The Chemfix® product has unconfined compressive strengths
ranging from 24 to 154 psi, bulk density 1.4 g/cm3, and per-
meability ranging from 5x10"6 to 5x10"6 cm/sec. (Materials
with coefficients of 10"6 cm/sec and lower are considered
impervious for civil engineering purposes.)
Process Reliability:
Complexity - Process control is maintained automatically
and is relatively simple.
Start-up and Shutdown - Can be performed easily.
Equipment - Chemfix® mobile van which contains the
chemical storage, metering and mixing equipment.
Material for Construction - No unusual materials.
Raw Materials - Sodium silicate and Portland cement are
required for Chemfix® process. Lime addition may be
needed if pH of the sludge is below 4.
Secondary Pollutants; Potential of contaminating groundwater
and surface water by the leachate of Chemfixed product still
exists.
Applicability: This process is used at some refineries. It
has been proposed as a method of permanent stabilization of
hazardous waste of old refinery lagoons and pits. It may
also be a viable technique for stabilizing hazardous tetra-
ethyl lead wastes. Further research is needed to study the
applicability.
Process Advantages:
• Improve leachate quality.
• High continuous throughput rate.
• Small volume increase due to chemical additives.
• Ability to react with complex waste mixtures.
• Non-toxicity of the solid material.
1-44
-------
Process Disadvantages:
Limited data on leachability.
pH adjustment.
Undefined current status.
Costs and Energy Requirements;
Costs: From available data, no breakdown on capital costs
and operating costs are supplied. Therefore, total costs
are presented. Figure 1 demonstrates the relationship of
total disposal costs for the Chemfix® process and the weight
percent solids in sludge. These costs include operating
costs, and dewatering equipment capital costs.
Energy Requirements; Energy is required for sludge pumping,
mixing ana hauling of the solids to disposal site. No speci-
fics on requirements are available.
Manufacturers National Environmental References: ME-218,
or Developer: Controls, Inc. RO-423, WI-197
1-45
-------
34
32
30
•T* 28
a
^ 26
Ed
o
g 24
CO
C
o
22
20
181-
o
u
nJ 16
-------
Process/Equipment; Sand Drying Beds Application:
Sludge Dewatering of Low-
Oil Waste Streams, e.g.,
Well-Digested Biosludge
from Biotreatment Facili-
ties
Process Description: The most widely used sludge dewatering
method in the U.S. is drying of the sludge on open or covered
sandbeds. Nevertheless, most refinery wastes are oily sludges
and not amenable to this treatment technique (see Applicability
below). A drying bed consists of a holding tank and filtration
media made of sand and gravel. Drying beds usually consist of a
layer of sand 4 to 9 inches thick placed over 8 to 18 inches of
graded gravel or stone. The sand typically has an effective size
of 0.3 to 1.2 mm and a uniformity coefficient of less than 5.0.
Gravel is normally graded from 1/8 to 1.0 inches. In many indus-
trial applications, the gravel is supported by a heavy wedgewire
system. Drying beds have underdrains that are spaced from 8 to
20 feet apart. Underdrain piping is often vitrified clay laid
with open joints, has a minimum diameter of 4 inches, and has a
minimum slope of about 1 percent. Water is removed by drainage
or evaporation. Collected filtrate is usually returned to the
treatment plant.
Air drying is normally restricted to well-digested sludge, because
raw sewage sludge is odorous, attracts insects, and does not dry
satisfactorily when applied at reasonable depths. The design and
use of drying beds are affected by many parameters, including
weather conditions, sludge characteristics, land values and prox-
imity of residences, and use of sludge conditioning aids. Climat-
ic conditions are most important. Factors such as the amount and
rate of precipitation, percentage of sunshine, air temperature,
relative humidity, and wind velocity determine the effectiveness
of air drying. The nature and moisture content of the sludge dis-
charged to the beds also influences the drying process. It is
important that wastewater sludge be we11-digested for optimum dry-
ing. In well-digested sludge, entrained gases tend to float the'
sludge solids and leave a layer of relatively clear liquid, which
can readily drain through the sand.
Dewatered sludge is removed either mechanically or manually. Fig-
ure 1 shows one type of sandbed which has been employed widely in
industry. The filtration support is a heavy wire mesh often re-
ferred to as a wedgewire system.
Development Status: Over 6,000 wastewater treatment plants use
this method.Although they are especially popular in small plants,
drying beds are also used by 38 percent of the cities serving
1-47
-------
I
£-
00
CONTROLLED DIFFERENTIAL HEAD IN VENT
BY RESTRICTING RATE OF DRAINAGE
1
VI
*ta^i
ENT
1
^ PARTITIO
1* 1 XV>C.A*\. I?W f%*
WV V
WEDGEWIRE SEPTUM,
OUTLET VALVE TO CONTROL
RATE OF DRAINAGE""
EN-325
FIGURE 1. CROSS SECTION OF A WEDGEWIRE DRYING BED
-------
populations of over 100,000. Furthermore, sandbed drying is the
most common technique employed in Europe.
Performance:
Control Effectiveness; Typical performance data show rapid
dewatering during the first one to two days while water re-
moval by drainage predominates. This is followed by two to
five weeks of slow dewatering principally by evaporation.
In good weather, a cake of 45 percent solids may be achieved
in six weeks with a well-digested sludge. As with many de-
watering techniques, the performance of a bed may be improved
by proper chemical conditioning and the dewatering time may
be reduced by 50 percent or more.
Typical wedgewire drying bed performance data with the ap-
plication of chemical conditions is shown in the following
table:
TYPICAL PERFORMANCE DATA FOR WEDGEWIRE DRYING BEDS
Sludge Type
Primary
Trickling Filter Humus
Digested Primary + EAS*
Fresh EAS*
Fresh EAS*
Thickened EAS*
Feed Solids
(Percent)
8.5
2.9
3.0
0.7
1.1
2.5
Cake Solids
(Percent)
25
8.8
10.0
6.2
9.9
8.1
Time
Interval
14 days
20 hours
12 days
12 hours
8 days
41 hours
Solids
Capture (%)
99
85
86
94
87
100
EAS • Excess Activated Sludge
EN-325
Process Reliability:
Complexity - Little operator attention and skill are
required.
Start-UP and Shutdown - Can be performed easily.
Equipment - Sandbeds and sludge removal equipment.
Materials of Construction - Sand and gravel, concrete,
etc.No unusual material is required for construction.
1-49
-------
RADIAN
Raw Materials - Sludge conditioning chemicals (e.g.,
flocculants and coagulants) are sometimes required.
Secondary Pollutants: Three types of secondary pollutants
result from sandbed operations:
• Dewatered sludge which requires disposal.
• Odors evolved if treated sludge is not well-digested.
• Collected drainage water which is generally sent
back to treatment plant.
Applicability: Some refinery sludges contain significant
amounts of oil and oily solids, and are only slightly de-
watered by sand filtration. The principal application of
gravity filtration lies in dewatering well-digested sludge
from sanitary sewage treatment, activated sludge and trick-
ling filter solids from wastewater treatment, and solids
slurries that are essentially oil-free.
Process Advantages;
• Skilled labor is not required.
• Operator attention is minimal.
• Low capital costs.
Process Disadvantages:
• Land area required.
• Weather problems for open sandbeds; e.g., freezing,
odor, insects, etc.
Costs and Energy Requirements:
Capital Costs: Figure 2 shows construction costs of sand dry-
ing beds for various bed sizes. The curve includes excavation,
piping for sludge distribution, sand and gravel drainage beds,
and underdrain collection piping.
Operating Costs and Energy Requirements: Recent operating
costs were not located; however,these costs are very small
compared to other sludge dewatering methods. Energy is only
required for mechanical sludge removal if employed.
-------
RADIAN
CO
g
,-4
CO
H
CO
O
z
Q
BS
H
CO
Z
8
l.OOO.OQC
»
T
1
1
!
t
s
\
j
s~
s
>
s~~
r
/
/
/
-
^i
l
/
/
/
t
/
J
/
/
/
4
«
I
t
T
2 J » i«7i» 2 14 »«7»« 2 14 9*71*
10.000 100.000
ABEA, FT2
July 1977 cost index - 204.5
December 1978 cost index - 225
CU-087
FIGURE 2 SAND DRYING BEDS, CONSTRUCTION COSTS, JULY 1977*
^Projected from 1975.
1-51
-------
RADIAN
Manufacturers Numerous. References: AM-042,
or Developer? CR-154, CU-087, EN-325
-------
Process/Equipment; Wet-Air Oxidation
Application:
Sludge Conditioning
Process Description: Wet-air oxidation, also known as wet incin-
eration, wet combustion, or wet oxidation, is a type of thermal
reduction process which is used to partially oxidize volatile
organic solids present in untreated sludge and hence condition
the material prior to dewatering. Wet-air oxidation occurs at
elevated temperatures and pressures. A diagram of a typical flow
scheme is shown below.
Sludge is preheated by exchange with the reaction products and
then mixed with stoichiometric amounts of air or oxygen. The mix-
ture enters the reactor at 350-600°F depending upon the particular
application and under sufficient pressure (300 to 3000 psig) to
keep the water contained in the sludge in the liquid state. The
high temperature and free oxygen serve to break down the heavy
organic molecules by chemical oxidation and hydrolysis. The
amount of oxygen required is determined by the COD of the waste
stream; excess oxygen has little effect on the rate. The oxida-
tion process itself provides much of the necessary heat in the
reactor. The process may become thermally self-sustaining for
wastes in excess of 10,000 ppm COD. A mixture of gases (primar-
ily C02 and N2), water, and solids leave the reactor and are used
to preheat the incoming waste before being separated in a cyclone.
The liquid and stabilized solids are separated by vacuum filtra-
tion, centrifugation, or drying beds, or in dewatering lagoons.
The supernatant from dewatering is recycled to the wastewater
treatment plant. This high strength liquor is a major disadvan-
tage of the process because of its substantial organic load on
the biotreatment facility.
Reactor
Sludge
Sludge storage Heat exchangers
tank
Exhaust
gases
Separator
Effluent
suspension*
CL-039
Compressor
WET-AIR OXIDATION SYSTEM
1-53
-------
Development Status; Wet-air oxidation is a proven process devel-
oped through 30 years of municipal sludge treating and industrial
applications.
Performance:
Control Effectiveness: This technique improves the dewater-
ing properties of sludge quite dramatically. The BOD and COD
content of the stabilized sludge is greatly reduced; however,
these oxygen demands are concentrated in the recycle liquors.
The BOD content may be as high as 40-50 percent of that of
the unprocessed sludge; the COD content typically ranges from
7000 to 10,000 mg/Jl.
Process Reliability: This process is reliable but is more
complex than most treating processes. It is operated at high
pressures and temperatures.
Complexity - Fairly complex. Standard concepts and
equipment are used.
Start-up and Shutdown - Relatively easy.
Materials of Construction - The process employs high
pressures and moderate temperatures. Corrosion resis-
tant materials are required.
Equipment - Air compressor, water or sludge pump, heat
exchangers, and reaction vessel designed for high pres-
sure operation.
Raw Materials - Compressed air or oxygen, and steam or
fuel for preheating reactants. Some pH adjustment
chemicals may also be required.
Secondary Pollutants: The reactor exhaust gases may contain
odorous substances that may require scrubbing, incineration,
or catalytic oxidation before venting. The liquid effluent
contains very high COD, BOD, and ammonia concentrations.
Applicability; One refiner reported development of a skid or
truck-mounted wet-air oxidation unit; the mobility of the
equipment permits emergency solid waste management during
shutdown.
1-54
-------
Costs and Energy Requirements:
Capital Costs: The costs graphically presented in Figure 1
have been updated to July 1977. The assumed operating condi-
tions are 1500 psig and 550°F (HU-236).
Operating Costs: The bulk of the operating cost is for power
for the pumps "and compressors. Updated costs reported by two
municipal facilities were $27/ton and $37/ton. These values
do not include the cost of handling the oxidized solids or
recycled liquors.
Energy Requirements:
• Power: 25 kw/gpm using air; 5 kw/gpm using oxygen.
• Maintenance: less than 3 percent of capital costs
per year.
• Some energy may be required for heating the reactor.
However, for waste streams with over 10,000 ppm COD
the reaction becomes thermally self-sustaining and
little heat input is required.
Manufacturers Several. References: CL-039,
or Developer? EC-008, EN-325, GO-279,
HU-236, ME-095, ME-294
1-55
-------
I10
1.1
10
10* 10*
CAFAcrrr. cm
July 1977 cost index - 204.5
December 1978 cost index - 225
FIGURE 1 APPROXIMATE CAPITAL COSTS FOR
WET OXIDATION TREATMENT SYSTEMS
1-56
-------
Process/Equipment; Composting Application;
Waste Biological Sludges
with Oily Wastes
Process Description: Composting is the process of biochemical
degradation of organic material under controlled conditions to
a stable end product. Most composting operations consist of
three steps:
1) Preparation of the wastes
2) Decomposition/airing of the prepared wastes
3) Preparation and marketing of the product
The addition of a bulking material, such as wood chips, is re-
quired to adjust the moisture content of the material to be com-
posted. A wet weight of 45 to 65 wt percent water is sufficient.
Decomposition occurs by aerobic bacterial degradation with oxygen
supplied by turning/mixing in windrow composters and forced
ventilation in mechanical aeration systems. The reaction is
exothermic and the wastes reach the thermophilic temperature
range (140-160 F). Virtually complete pathogen kill is effected
at these temperatures. The period of digestion is normally
about six weeks for windrows and several days for mechanical
aeration systems. Important process parameters include:
• Moisture content
• Temperature
• Oxygen transfer rate
• CO 2 emission rate
• pH
• Porosity
• C/N ratio
After sufficient digestion, measured by temperature and C02 emis-
sion rate, the compost is cured to insure the necessary stabiliza-
tion before screening and final disposal or sale. Figure 1 illus-
trates a general flow diagram with operating procedures for the
two major composting systems.
Development Status: This technology has been used for many decades
in Europe as a refuse and sludge treatment technique. In the U.S.,
its use has been largely confined to municipal refuse disposal.
One refinery, however, has conducted a full-scale test and reported
successful results.
1-57
-------
CoopoBtlog Syatea
1. Windrow
2. Nechonlcal
Aeration
Sludge
Bulking
Bulking
. Recycle
M.rk.t or
""""" * Mining * Aeiatlon '* Oiylug —•••—# Cuilug * Scieenlug -. - >»
Tlae In Each Stage
Flled-% d«y 32-42 imjm 2 day* 20 day* % day
Placed In a Mecb- 21 day* 2 day* 30 day* % day
anlcally aerated
»n«-l| day
FIGURE 1
oo
FLOW DIAGRAM AND TIMES REQUIRED FOR COMPOSTING IN TWO DIFFERENT SYSTEMS
-------
Performance:
Control Effectiveness: Approximately 20 to 30 percent of
the volatile solids are converted to carbon dioxide and
water. The compost is free of pathogens and spores. It has
been found that 20 to 50 percent less landfill space is
required for composted material.
Process Reliability: Not proven as a process on refinery
sludges.
Complexity; Skilled labor is required.
Start-up and Shutdown: Initial seeding and proper
moisture content can make start-up difficult. Care-
ful monitoring is required throughout.
Equipment: Composting bins for some mechanical
aeration systems; blowers or agitation equipment
for aeration are required as well as screening
equipment.
Materials of Construction: No special materials
are required.
Raw Materials: Wood chips or other bulking material;
chemical conditioners are sometimes required. Initial
seed compost is necessary.
Secondary Pollutants; Composted material must be disposed
of in a landfill or by other final disposal methods. No
odor problems are generated from properly managed composting
systems. Air emissions from windrow piles has yet to be in-
vestigated.
Applicability; Preliminary testing indicates that composting
could be applied to mixtures of biological solids and oily
sludges. Further evaluation should be conducted to deter-
mine the flexibility of composting systems to accomodate
different feed sludges.
1-59
-------
Process Advantages:
• Possibility of year-round operation.
• No groundwater monitoring required.
• Solid waste is reduced in volume and weight and
this extends the life of land disposal facilities.
• Degrades waste rather than stores it.
• Composted material is sanitary and humuslike.
• Can be used as a soil conditioner.
• Possible revenue from product.
Process Disadvantages;
• Insufficient testing on refinery sludges.
• Economics are currently unclear.
• Skilled operation and monitoring required.
Costs and Energy Requirements;
Costs: Capital and operating costs are highly variable for
composting depending on the type of waste, the type of corn-
poster, and the capacity. A range of $2 to $20/ton have
been reported. Figure 2 shows some typical cost data. Rev-
enues from the sale of composted product range from $2.00 to
$4.00 per ton of material fed.
Energy Requirements: Energy is required for feed and mixing
augers and the blowers for mechanical aeration units. Trac-
tors or other turning/agitation equipment is necessary for
windrow operation. No specific data are available on energy
requirements.
Manufacturers Numerous. References: EN-325,
or Developer? KN-055, LI-120, ME-294
1-60
-------
GROSS COST/TON
(SCALE RIGHT)
NET COST/TON
(SCALE RIGHT)
INITIAL CAPITAL COST
1000 DOLLARS/INSTALLED TON
(SCALE LEFT)
50
150
200
250
300
PLANT CAPACITY (TON/DAY)
350
EN-325
NOTES:
1. Plant capacity is normally ont or two shift* par day to achieve plant capacity.
2. Gross cost trend is the owning and operating facilities without any credits.
3. Nat cost trend is for owning and operating facilities considering sales of compost
and salvaged materials.
4. All costs consider compost digested sludge with refuse.
6. Source: Composting of Municipal Solid Wastes in the United States. US
Environmental Protection Agency (1971).
1971 Cost Index = 132.2
Dec. 1978 Cost Index = 225.0
FIGURE 2
CAPITAL AND OPERATING COSTS FOR A COMPOSTING SYSTEM
1-61
-------
Proces s / Equipment ; B.E.S.T. Solvent Application: _
Extraction Process Sludge Dewatering
Process Description: A process flow diagram is shown in Figure 1.
Sludge is continuously metered to the system and mixed with recir-
culated solvent in a 6:1 ratio. The mixture is centrifuged and
the solvent-wet cake discharged. The centrate is water, oil, and
solvent. The oil-depleted cake enters a continuous belt dryer
where, at 240°F to 350°F it is dried to a 95 wt percent solids
cake. The evaporated solvent and water are condensed and recir-
culated. The centrate is separated by heating in a decanter into
a solvent/oil fraction and a water fraction. Water is pumped to
a still where any residual solvent is steam stripped and returned
to the system. The separated solvent/oil stream is cooled to 15°F
and split; one stream is recycled and the other fed to a solvent
still where the oil is removed by steam stripping. Bottoms from
the solvent still are separated and the oil recovered; water is
returned to the wastewater treatment plant or disposed of. Due
to its variation of solubility with temperature, triethylamine is
usually the solvent selected.
De ve 1 opmen t S t a tus ; The B.E.S.T. solvent extraction procedure is
being commercialized in the pulp and paper industry. However,
such a technique is only at the pilot plant level for refinery
waste disposal.
Performance ;
Control Effectiveness : Oil is recovered, and a very dry
cake (95 percent solid) is discharged. Pathogen kill is
complete in the dryer. Much flexibility in the feed sludge
composition is permissible; the final product is very con-
sistent. Dissolved solids are not removed in this process;
they appear in the effluent from the water still.
Process Reliability;
Complexity - This is a complicated process with
skilled labor or sophisticated automatic control
required.
Start-up
lar-problems .
Start-up and Shutdown - This presents no particu
b
1-62
-------
i
ON
10
HEAT /
EXCHANGER
MIX
JUNCTlONv
HEAT 5°'F
EXCH ANGER v
SLUDGE
if
r?
¥:*:«
IS
ii
•* • •
• § • «
t • • •
• • • *
» • * •
rii'i
IM
—T
< •*
' : ,
l^j
20*F OIL
fu**r+ A ifa
» 60*F 4O
*'*'*! * •••••*• •*•*• • j
;— \~,'.:y.'-s
zi
S
/£
•F,
vwvJ
^LJ^uJ
V
HEAT
EXCHANGER
f:
50
>T
T • •
* •
* *
• * •
•
|Xv
LVENT
ILL
•$
I
DE
rizi
ts=s
I
•••••••••*•••*
»••••••••••*»*•
••*§•••?••••••
•••••••»*••*•*•
•••»•**•*•••••
"
^HEAT
EXCHANGER
^ LIQUID/
•: /SOLID
SEPARATOR
40*F yDRYER
BANTER
1 140«F
STILL
S PATHOGEN
FREE DRY
^^^
i=S^
^^=
rg^^g
^^
grrrssr
^- i'^L-
1
SOLIDS ^T
'........M k STERILE
r ^^^ r CLEAR WATER
SOLVENT ^ SOLVENTS & WATER
V/ATER ES3 SOLVENTS & SLUDGE
FIGURE 1
FLOW DIAGRAM OF B.E.S.T. PROCESS
JA-216
-------
Equipment - The following units are required:
dryer, centrifuge, refrigeration system, heat
exchangers, water still, solvent still, and
sludge and centrate pumps.
Materials of Construction - Stainless steel con-
struction is mandated in some of the heat exchang-
ers and pumps; carbon steel will suffice elsewhere,
Raw Materials - Solvent and possibly alkaline addi-
tives .
Secondary Pollutants: Cake needs disposal. Air monitoring
on dryer may be required. Water from the water still is
usually recycled to the treatment plant.
Applicability; This process could potentially be applied
to refinery waste sludges.
Process Advantages;
• Very high quality, pathogen free, dry cake.
• Low floor space.
• No phase change required to remove water; hence,
energy savings are substantial.
Process Disadvantages:
• Complex process.
• High capital costs.
• Sophisticated automatic control required.
Costs and Energy Requirements;
Capital Costs: For a fully installed and operating system,
capital costs are estimated at $2,440,000. Without build-
ing or foundations, the cost is $730,000. This is based on
wastes from a hypothetical 200,000 bbl/day Gulf Coast refin-
ery generating 80 tons/day of sludge.
1-64
-------
Operating Costs: For a hypothetical 200,000 bbl/day refin-
ery generating 80 tons of oil/water/sludge per day, total
operating costs are estimated $26/ton. This does not in-
clude amortization of the capital.
Energy Requirements;
Thermal Energy 15,500 Btu/lb solids
Electrical Energy .40 kwh/lb solids
Manufacturers Resources Conservation References: JA-216
or Developer: Company
1-65
-------
Process/Equipment: Landfill Application:
Ultimate disposal, Usually
of Dewatered Sludges
Process Description; A landfill currently serves as a final dis-
posal technique tor solids and sludges. Because of its lack of
significant biodegradation, this classification is being called
into question. A landfill site is more properly termed a storage
technique. Two types are commonly encountered, sanitary landfill
and secure landfill. The sanitary landfill is an engineered
method of disposal which involves spreading and compacting the
waste into cells, and covering it each day with earth in a manner
that poses no threat to the public health or the environment.
Oily wastes are attacked by soil microorganisms and partially
decomposed. The main problems in a sanitary landfill are the
production of leachate that may contaminate the groundwater
and the accumulation of gas that may catch fire or cause explo-
sion. A basic configuration of a sanitary landfill is shown in
Figure LA.
A secure landfill (used synonomously with California Class I land-
fill) is similar to a sanitary landfill in nature. However, a
secure landfill has the following characteristics:
* Use of sealant to minimize pollution.
• Reroute of surface water at landfill site if needed.
• Water table is below the landfill site.
• Provisions of leachate collection and treatment,
monitoring wells and venting, as needed.
• Segregation of incompatible wastes.
Figure IB demonstrates one possible design for a secure landfill.
Pollution potential still exists for secure landfills for a pro-
longed period of time.
The important operating parameters for landfills are:
• Hydraulics of surface and subsurface waters.
• Soil permeability
• Site geology
• Properties of input sludges
• Leachate quality
1-66
-------
RADIAN
SL-084
FIGURE 1A. GENERAL CONFIGURATION OF A SANITARY LANDFILL
Leachate Collection
Monitoring
Well
Uacer Table
SL-084
Monitoring
Veil
FIGURE IB. ONE POSSIBILITY OF A SECURE LANDFILL
1-67
-------
RADIAN
Development Status: This technique has widespread use for disposal
of refinery and municipal wastes. Its continued employment is
highly contingent upon future regulations, especially those per-
taining to leachate quality.
Performance:
Control Effectiveness; A landfill can be an effective ulti-
mate disposal means of refinery wastes when pollution pre-
ventive practices are exercised. Choice between sanitary
landfill and secure landfill depends on the properties of
solid wastes. The effectiveness is also contingent upon
site-specific geological and climatological conditions.
Process Reliability;
Complexity - Configuration and operation of a landfill
itself is simple. However, leachate collection treat-
ment and monitoring systems when available in a secure
landfill can add operational complexity.
Start-up and Shutdown - Can be performed easily.
Equipment - Trucks, bulldozers, front-end loaders, vent-
ing equipment, monitoring equipment, etc.
Materials of Construction - Soil, landfill sealants
sucn as clay, and synthetic materials.
Raw Materials - Soil and leachate treatment chemicals.
Secondary Pollutants: Contamination of groundwater and sur-
tace water may occur. The pollutants of some concern in
petroleum refinery wastes are heavy metals. Gases evolved at
landfill site can be a safety hazard.
Applicability: Landfilling is currently the most widely
used metnod for ultimate disposal of many petroleum refinery
wastes. The geology of the site location is a major deter-
minant of applicability. For example, the high water table
of the Gulf States presents problems in finding a suitable
landfill site to the many refineries located there.
T-Afl
-------
RADIAN
Process Advantages:
• Economical ultimate disposal of solids.
• Possibility of reclaiming landfill site when
completed.
• Design data readily available.
Process Disadvantages:
• High degree of maintenance required especially
when leachate collection, treatment, and monitor-
ing are performed.
• Possibilities of groundwater pollution and safety
hazard caused by evolved gases.
• High land requirements.
Costs and Energy Requirements:
Capital Costs: The magnitude of capital cost depends on the
landfill size and the complexity of the landfill. From EPA
published data, sanitary landfill costs less than $3/ton.*
Table 1 summarizes the capital costs for a sanitary landfill,
Capital costs for secure landfill are highly site specific.
Costs depend on size, leachate collection and monitoring
systems, type of sealant, etc. No generalized conclusion
can be drawn for secure landfills.
erating Costs; Operating cost depends mainly on the cost
ot labor and equipment maintenance, the method of preparation,
and the efficiency of the operation. Operating cost reported
by EPA for sanitary landfill ranges from $1.50/ton to $9 ton.**
Energy Requirements; The energy requirements are for hauling
the solid material to the disposal site, which is highly site
specific, and for operating the landfill equipment.
Manufacturers Not applicable
or Developer:
References: LI-120, EN-605,
JO-273, SL-084, WI-331,
PE-224, EN-602, O-055,
VA-203
^Projected from 1975.
**Projected from 1976.
1-69
-------
RADIAN
TABLE 1
CAPITAL COSTS FOR A SANITARY LANDFILL
(Dec. 1978 Dollars)(l)
Site 1
Item (50 TPD). $/ton
Planning and
Design 0.14
Site Development 0.07
Facilities .03<2)
Equipment 2.03
Total 2.28
EN-602
lProjected from 1975.
^'Includes fencing.
1-70
-------
RADIAN
• Soil temperature
• Local evaporation rate
• Hydraulics of surface and sub-surface waters
• Metals accumulation
• Leachate quality and quantity
• Soil nutrients
• Site grading and runoff
Low pH (below 7.0) may cause leaching of metals. Runoff must
often times be collected and routed into the refinery wastewater
system. If soil nutrients (nitrates and phosphates) are low,
these must be added to the soil. Periodic sampling of leachate
using monitor wells or core samples from the site is necessary;
further, draft regulations promulgated under RCRA may require
leachate collection systems in some instances. In general, heavy
metals are in lower concentration in refinery sludges than in
municipal sludges. Host refiners and API recommend that crops
not be grown on landfarms.
Development Status; Landfarming has been used in the petroleum
industry for 25 years. Because of its economic and ecological
advantages, landfarming is becoming the most common method of
acceptably disposing of oily wastes.
Performance;
Control Effectiveness; Landfarming is an effective sludge
disposal method when pollution preventive practices are exer-
cised. As in landfilling, site-specific conditions, pri-
marily climatological, determine the control effectiveness.
Research by API is currently being conducted on polynuclear
aromatics degradation in landfarms; air emissions are also
being investigated.
Process Reliability:
Complexity - Configuration and operation of land
spreading are simple. However, site monitoring re-
quires analytical capability.
Start-up and Shutdown - Can be performed easily.
Equipment - Trucks, tilling, and monitoring equipment.
1-72
-------
RADIAN
Materials of Construction - None.
Raw Materials - Nutrients and pH adjustment chemicals
when required.
Secondary Pollutants: Contamination of groundwater from soil
seepage and surface waters from runoff may occur. Collection
and treatment of runoff may be necessary. Odors may be pro-
duced when site is not properly operated.
Applicability: The use of land spreading as a final disposal
technique for refinery wastes is rapidly increasing. It*is
suitable for disposal of almost any oily wastes generated by
a refinery. An operating landfarm has processed the follow-
ing sludges:
• Tank cleanings, including crude and distillate tank
bottoms and slop oil emulsion tank cleanings.
• API separator bottoms .
• Desalter sludges.
• Waste biological solids.
• Filter clays (on an emergency basis).
• Oil spill clean-up debris.
Finally, landfarming has been used extensively as the final
clean-up step in the renovation of oily pits.
Process Advantages:
• It is an economical and relatively foolproof method
when a suitable site is available.
• Can be conducted under a wide range of soil and
environmental conditions.
• Improves the physical and chemical properties of
soil by increasing organic matter and nitrogen con-
tent, porosity and moisture-holding capacity.
• Oily fractions are substantially degraded; in a
landfill, only minor degradation of oily materials
has been found to occur.
-------
RADIAN
Process Disadvantages;
• Site monitoring is required.
• Odor and residual hydrocarbon problems when not
properly operated.
• Large land area required.
• Collection and treatment of runoff water may be
necessary.
• Long-term trace metal accumulation.
Costs and Energy Requirements:
Costs: The total cost of land spreading is reported to be
$3 to $18 per ton of waste sludge. Land costs form the
largest portion of these cost figures.
Energy Requirements; Energy is required for the operation
of tilling equipment, but the requirement varies widely.
Manufacturers Not applicable.
or Developer;'
References; WI-331, LE-381
KN-054, KN-055, VA-203,
Personal Communications
with Juanita Galloway of
Chevron, El Segundo, CA.
1-74
-------
Process/Equipment: Incineration
Application:
Separator Bottoms,
DAF Float, Waste Biosludge,
Slop Oil Emulsion Solids
Process Description; Incineration is a controlled process for
oxidizing solids, liquids and sometimes gaseous combustible wastes.
Incineration products include the particulate matter carried by the
gas stream, incinerator residue, grate siftings, and process water.
Incinerator residue consists of noncombustible materials and re-
quires further land disposal. Because of these secondary waste
streams, incineration is not really an ultimate disposal technique
but is included here for its extensive volume reduction and chem-
ical alteration capabilities. Four types of incinerators are cur-
rently in common use: multiple hearth furnaces, fluidized bed
combustion, vortex (cyclone) furnaces, and rotary kilns. Figures
1-4 show their general configurations. Most sludges will not com-
bust easily because of their water content. Therefore, dewatering,
preheating, and auxiliary fuel are needed at times. Table 1 sum-
marizes some characteristics of the aforementioned incineration
methods.
A fluidized bed, due to detention time, stable combustion tempera-
ture, and high heat efficiency, is the most promising incineration
technology for refinery sludges. The heating value of the sludge
must exceed 29,000 Btu per gallon to permit a self-sustaining re-
action. This often necessitates dewatering the sludge to 30 per-
cent to 40 percent solids, depending on the oil content, and sup-
plying auxiliary fuel to the combustion chamber. Energy effi-
ciency is aided when an air preheater is included in the config-
uration. Bed plugging remains the major difficulty in this tech-
nology.
Development Status: Conventional municipal-scale incinerators have
been in use since 1920. As of May 1972, 193 of these plants x*ere
operating. Smaller incinerators with capacities of 5 to 12 tons
per day (compared with 50 to 300 tons for most conventional munici-
pal incinerators) are being used in commercial and industrial
facilities, as well as communities. Fluidized bed incineration in
the refinery industry began in 1970 with a joint AMOCO-EPA pilot
unit. Successful testing warranted construction by AMOCO of a
larger incinerator and other refiners have followed suit with test
fluidized bed units.
1-75
-------
SLUDGE INLET
AIR IN
AIR OUT
FLUOZED MEDIA
GAS DISTRIBUTION
PLATE
LI-120
(a) FLUIDIZED BED
Hot go In
isoof.
REACTOR
^
"""^ (+»+ I'M if
J-J
AIR
PREHEATER
EN-325
(b) FLUIDIZED BED WITH AIR PREHEATER
FIGURE 1. CONFIGURATIONS OF FLUIDIZED BED INCINERATORS
-. -re
-------
RADIAN
GAS
OUT
RABBLE ARM
RABBLE TEETH
CH-196
FIGURE
2. CROSS SECTIONAL VIEW OF MULTIPLE HEARTH FURNACE
1-77
-------
RADIAN
Annular space filled
vnith *ir under
pressure for tuyere*
Baffle shelU, N
Air tuyerw *C.
Tuytf• air shell
and plenum ——*
/ Refractory wall .
Combustion air
'to tuyeres
_. Refractory
•"* cooling air
'ir- Combustion air
3um«r nozzle
^'l 6 .3 q b b iy^vSx. ^Gas burner ring
Cooling air porn
cnt in refractory slob
Gas flow N Cooling air
(forced draft)
- Tuyere air shell
• Qjffle shell
WI-331
FIGURE 3. VORTEX FURNACE
1-78
-------
RADIAN
Tonpamlon
Ghamatrand
gnicnibMr
Wl-331
FIGURE 4. ROTARY KILN
1-79
-------
TABLE 1
CHARACTERISTICS OF VARIOUS TYPES OF INCINERATORS
Type Characteris tics
Multiple Hearth Built in wide range of sizes, 500-2,500 lb/
hr. dry solids; air cooling and auxiliary
fuel are required; not very heat efficient.
Vortex (Cyclone) Designed for small capacities (500 Ib/hr.
or less); lightweight and compact; low de-
tention time (approximately 10 sec.).
Rotary Kiln Built in wide range of sizes, 40 Ib/hr. to
2,400 Ib/hr.; designed for continuous oper-
ation; low investment cost.
Fluidized Bed Silica sand is the fluidizing media, excess
air is required for combustion; excellent
mixing of air, sludge, and hot sand; highly
heat efficient; no odor problem.
-------
Performance:
Control Effectiveness: Incineration generally can reduce 80
to 90 percent of the total volume of solid waste, and 98 to
99 percent (by weight) of the combustible portion.
Process Reliability:
Complexity - Fairly complex, skilled labor is required.
Start-uo and Shutdown - Conventional incinerators are
operated continuously. Start-up requires auxiliary fuel
for small industrial size incinerators. Shutdown is
easy while start-up is more complicated.
Equipment - Incinerators; and pollution control equip-
ment that could be extensive for areas (air quality
control regions) with stringent air emission require-
ments.
Materials of Construction - Materials needed for con-
structing incinerator.
Raw Materials - Auxiliary fuel and sand (for fluidized
bed).
Secondary Pollutants; Incineration products include the par-
ticulate matter carried by the gas stream, incinerator resi-
due, grate siftings, and process water. Incinerator residue
consists of noncombustible materials and requires further land
disposal. Pollutants such as particulates, adsorbed heavy
metals, N0x, and SO in the exit gas may present pollution and
odor problems. Mechanical collectors, and/or scrubbers may be
required. After-burners on incinerator stack are commonly
used to eliminate noxious odors and combustible particulates.
Also, conventional incinerators may produce wastewater from
wet scrubbing of off-gases which requires treatment.
Process Advantages:
• Land area required is. small_which is a definite
advantage when land is scarce and/or expensive.
• The use of the thermal energy indigenous to oily
wastes.
• Independent of geological and climatological conditions
1-81
-------
RADIAN
• Solid waste is reduced in weight and volume, and
this extends the useful life of the available land
disposal facilities.
• Where waste must be hauled long distances to a land-
fill, incineration may be economically advantageous
if close-in locations can be found for the inciner-
ator.
• Incineration is adaptable to energy recovery pro-
cesses such as steam generation, and to recovery of
minerals from the residue.
Process Disadvantages:
• Necessity for disposal of incineration residues.
• Potential for air pollutants, particularly heavy
metals.
• Uneconomical for streams where much of the oil
has been recovered.
• The process requires large capital expenditures and
high operating costs.
• Skilled labor is required to properly operate and
maintain the facility.
Costs and Energy Requirements:
Costs: Published costs have wide variations due to the dif-
ferent types, capacities, waste treatment facilities, sludge
dewatering, etc. Figure 5 shows the allocation of incinera-
tion costs. Table 2 summarizes some of the available pub-
lished costs.
Energy Requirements; Auxiliary fuel is needed for start-up
maintenance, afterburner, and sometimes during incinerations,
Two factors affect the amount of auxiliary fuel required for
incineration: the heat value of the sludge, and the tempera-
ture required for complete combustion.
1-82
-------
Labor
cost
Utilities
Receiving
and
storage
»
— *
Repaira
and
maintenance
^m
1
Vok
redui
»
Jmt
Men »—'—— s"
treatment
operating
cost
External
cost
WI-331
FIGURE 5
ALLOCATION OF COSTS FOR THE INCINERATION OF SOLID WASTES
« ,Q
-------
RADIAN
Manufacturers Numerous
or Developer;
References; FE-055, LI-120
CU-082, WI-331, EN-602,
CH-447, PE-224, CH-196,
AM-042, RO-423, YA-089,
EN-325
1-84
-------
TABLE 2
SUMMARY OF COSTS OF INCINERATORS
Type of Incinerator
Multiple Hearth
Vortex
Capital Coat
(Dec 197B Dollars)
$0-$16/Ton
$193,000
Operating Coat
(Dec 1978 Dollars)
$7-$ll/Ton
NA
Coat Baala
oo
Cn
Rotary Kiln
Fluldlzed-bed
$800-$1700/lb/hr.
$390,000
NA
$30-$45/ton dry
solids
Conventional
Incinerators
$5.6 million
$10-$18/Ton
25-year payment; 500-
2500 Ib/hr. capacity;
cost scaled up from
1973.
200 Ib/hr. capacity;
size 4* x 6* x 6';
excluding dewatering;
costs scaled up from
1973.
40-2400 Ib/hr. capac-
ity; coat scaled up
from 1973.
500 Ib/hr capacity;
capital coats include
centrifuge and incinera-
tion, capital coat
scaled up from 1968,
operating cost from
1973.
EPA data; 250 ton/day
capacity; 20 year life;
6% interest; capital
cost scaled up from 1971;
operating coat scaled
up from 1976.
NA - Not Available.
-------
Process/Equipment; Disposal Lagoons Application:
Ultimate Disposal
Process Description: This process description is similar to the
description of dewatering lagoons given in the treatment subsec-
tion of this report. The major difference is that solids de-
posited in this type of lagoon are never removed. Typically,
little if any environmental control is provided. Before govern-
ment regulation spurred refiners to action, disposal lagoons were
the solid waste management practice.
Disposal lagoons can only loosely be termed a control technology.
Some evaooration does occur; this includes volatile organics and
other compounds. Use of a pond liner would inhibit the leaching
rate, however, these are seldom used. Lagoons usually consist of
an excavated hole and sometimes are diked and/or dammed.
Development Status: These lagoons have been used for many years
in the petroleum refining industry.
Performance:
Control Effectiveness: The major limitation to the control
effectiveness of a disposal lagoon is the inability to con-
tain either leachate or air emissions.
Process Reliability:
Complexity - A very simple process.
Start-up and Shutdown - Simple.
Equipment - No special process equipment is required.
Materials of Construction - No materials are required.
Raw Materials - No raw materials are required.
Secondary Pollutants: These ponds are seldom designed for
environmental compatibility. Air emissions, coupled with a
history of ineffective leachate control, have elicited heavy
regulation from the government and now virtually prohibit
these ponds as a final disposal technique.
1-86
-------
Applicability: The previously mentioned regulations will
eventually eliminate lagoons as a disposal technique.
Costs and Energy Requirements:
Costs; These costs are similar to those of a dewatering
lagoon, given in the treatment subsection of this report.
Capital and operating costs are extremely low in comparison
with more environmentally acceptable technologies.
Energy Requirements: The major energy requirement for dis-
posal lagoons is for pumping wastewater to the lagoon. The
energy requirement for pumping 1000 gallons a distance of
1000 feet is about 0.5 kwhr (based upon a 5 ft/sec flow
velocity).
Manufac turers Numerous. References: AM-042,
or Developer? CO-698, EN-818, RO-423
1-87
-------
Process: Disposal Wells Application:
Ultimate Disposal, Usua1ly
of Noxious Wastes
Process Description: Deep well disposal involves injecting re-
finery wastes into permeable underground formations. After suit-
able formations are located (typically from geological data for
the area) a well is drilled and perforated into the desired for-
mation. Wastes are pumped or injected into the well. Sufficient
pressure is produced either by pumps or by the weight of the
water column above the injection zone. Two major design para-
meters are the location and the depth selection. Because of the
possible contamination of potable aquifers, many areas are not suit-
able for this method of disposal. In addition, seismically active
areas are not suitable for deep well injection due to strains
forced on the unstable formations by pressurizations of the in-
jection zone. One operation conducted by a refinery injected
wastes into three disposal wells discharging into unusable saline
subterranean formations between 2100 and 4000 feet in depth.
Development Status: The technique was first developed in the oil
and gas extraction industry. Few refineries currently use this
methodology. Approximately 100 industrial waste injection wells
presently operate in the United States.
Performance:
Control Effectiveness: Not applicable.
Process Reliability; Disposal wells have been demonstrated
to be reliable for disposal of several wastewaters and oily
wastes.
Complexity - Simple.
Start-up and Shutdown - Easy.
Equipment - Pumps and wastewater storage facilities
above ground. Casing for well to protect shallow
aquifers.
Materials of Construction - Not applicable.
Raw Materials - None.
1-88
-------
Secondary Pollutants; The pollutants in the wastes may
pollute water in nearby aquifers.
Applicability: Can be used for petroleum refinery noxious
wastes depending on Federal state and local regulations.
Its current application is minimal.
Costs and Energy Requirements: All of the cost and energy require-
ments associated with deep well disposal are highly site specific.
They will depend on the availability of land with the correct
geological make-up and the distance from the source of the wastes
to the well. The cost of the well depends on whether an abandoned
well is available. If a well has to be drilled, the cost is pro-
portional to the depth of the well, within limits, but costs
escalate very rapidly for depths of more than a few hundred feet.
Manufacturer Numerous. References; LI-120,
or Developer; EN-605, RO-423,
MC-377
1-89
-------
Process/Equipment; Ocean Disposal Application:
Ultimate Disposal
Process Description: It was reported in 1974 that 600,000 tons
of refinery wastes were being dumped at 120 ocean disposal sites.
This method is declining as required permits become more difficult
to obtain. The Marine Protection Act of 1972 (PL 92-532) trans-
ferred regulatory control to EPA. Ocean disposal of certain
hazardous wastes is prohibited, and the list is growing.
Barges load sludges and other wastes in their holds and travel to
a selected site before dumping the waste. The site selection is
determined by distance from land, ocean currents, marine life, and
ocean depth. Another method is to sequester waste in 55-gallon
drums before disposal.
The principal advantage of ocean disposal is its very low cost
(about $2/ton). Its inherent disadvantages are much more numer-
ous. These include the effect on marine life, the concentration
of wastes at any one disposal site, slow breakdown, and formation
of surface oil slicks. Solids blanket the ocean floor and dis-
rupt or halt the natural benthic life, with drastic effects on
marine ecology. Although ocean disposal will continue for some
time to come, present trends indicate it will gradually be elim-
inated.
Performance:
Secondary Pollutants: Hazardous wastes or solids can
drastically alter marine habitats as indicated above.
Applicability: This technique is feasible only for
coastal refineries.
Costs and Energy Requirements:
Costs: For large dumping, $2/ton is the approximate current
cost. Containerized disposal costs range from $15 to $50/
ton.
Energy Requirements; Energy is required for barge fuel.
1-90
-------
Manufacturers Not applicable. References; RO-423, WI-197
or Developer:
1-91
-------
APPENDIX II
DESCRIPTION OF REFINERY SOLID WASTE STREAMS
Following are descriptions of individual solid waste
streams. For the purposes of this study, a "solid waste" is de-
fined as any "refinery waste materials other than trash, garbage,
and materials regulated by air emission and wastewater discharge
regulations" (EN-818). The streams described were shown by the
OAQPS, API and Jacobs studies to contribute significant volumes
to the total solid waste load generated by the refining industry.
Slop Oil Emulsion Solids
Oils skimmed from the API separators and DAF units are
usually pumped to an oil recovery tank where the mixture is separ-
ated into three phases: oil, water and emulsions. The oil is
blended for reprocessing and the water is recycled to the API
separator. The emulsion layer must be disposed of as a solid, or
it may be further treated to break the emulsion.
Storm Water Silt
Silt which collects in the stormwater settling basins
in some refineries is periodically removed, dewatered, and dis-
posed. The quantity of silt is usually a function of the amount
of rainfall and of refinery paved area, rather than of process
complexity.
Exchanger Bundle Cleaning Sludge
Heat exchanger bundles are periodically cleaned during
plant shutdown. Scale and sediment removed with cleaning solutions
II-1
-------
or by high pressure water are collected in sumps. The sludge is
then flushed to special sumps or process sewers, shoveled, or
vacuumed out and disposed of.
API Separator Sludge
Solids settle in the API separator during primary waste-
water treatment for removal of free oils. These sludges are peri-
odically removed using manual labor and vacuum trucks. API separ-
ators serve as a collection point for the oily water sewers. The
bottoms, therefore, contain a mixture of all sewered wastes such
as storage tank draws, de-salter wastes, laboratory wastes, sample
line purges along with miscellaneous chemical leaks and spills.
Leaded or Nonleaded Product and Intermediate Tank
Bottoms
Solids also settle to the bottom of distillate and resi-
dual storage tanks. The characteristics of these deposited sludges
will vary according to the product stored in the tank. These tanks
are cleaned when there is a change in service, product specifica-
tions cannot be met, or repairs are required.
Factors known to impact the volume and composition of
the various tank sludges include:
• Crude type
• Distillate cut
• Type and quantity of chemical additives
(e.g., lead)
• Recovered oil processing methods
• Use of tank mixers
• Process unit and tank metallurgy
• Product cut treatment employed upstream of tank
• Processes used in producing gasoline blend com-
ponents and other distillate products
II-2
-------
Crude Tank Bottoms
Solids in crude oil accumulate at the bottom of the
crude oil storage tanks. These tanks are cleaned periodically to
remove the sediment. The cleaning frequency is a function of the
amount of sediment in the crude oil sent to distillation units
and the reduction in storage capacity. Contaminants in crude oil
tank sludges vary with the characteristics of the crude oil and
the shipping and handling methods used prior to receipt at the
refinery. The waste sludges, in general, consist of a mixture
of iron, rust, clay, sand, water, sediment, oil, and wax.
Dissolved Air Flotation (DAF) Float
Following treatment in an API separator, the remaining
oil and solids in the wastewater are removed by DAF units in many
refineries. The process takes place in a vessel where the finely
divided solids and oil particles are brought to the surface and
skimmed off for disposal. The release of air dissolved in the
wastewater causes the solid and oil particles to rise.
Kerosene Filter Clays and Other Filter Clays
Treatment with fixed bed clay is used to remove color
bodies, chemical treatment residues and traces of moisture from
product streams such as gasoline, kerosene, jet fuel, and light
fuel oils. Clay is also used to treat lube oils by mixing the
clay with the oil and subsequently removing it by filtration.
HF Alkylation Sludges
Alkylation sludges are produced by both the sulfuric
acid and the hydrofluoric acid alkylation process. In most re-
fineries spent acid from these processes will accumulate in a
II-3
-------
storage tank where sludge will settle. These sludges, containing
polymerized hydrocarbons, tank scale and sulfuric acid, are removed
when the tank is cleaned or repaired. The acid sludge is usually
neutralized prior to disposal.
Waste Bio-Sludge
In the process of biological treatment of refinery aque-
ous waste streams, excess bio-sludge is created which must be
wasted. The waste bio-sludge has a very high water content (80 to
99 percent); therefore, it is dewatered prior to disposal. The
waste generation rate is dependent upon the type of biological
treatment process, operating conditions, desired level of removal,
and the raw waste load.
Once-Through Cooling Water Sludge
In the past, large volumes of water were pumped from
readily available sources and passed through primary settling
sumps or tanks prior to usage for once-through cooling. In re-
fineries where, once-through water is still used, silt must peri-
odically be removed from the settling vessel.
Fluid Catalytic Cracker Catalyst Fines
Fluid catalytic cracker (FCC) catalyst is continuously
regenerated by burning off the coke formed on the catalyst during
the cracking process. The flue gas from the regenerator passes
through a series of cyclones that recover most of the catalyst.
This recovered catalyst is then returned to the reactor vessel.
Because of current and future air pollution regulations, more re-
fineries have installed electrostatic precipitators or an equiva-
lent device to remove any catalyst fines which would otherwise be
released to the atmosphere with the regenerator flue gas. These
II-4
-------
catalyst fines are disposed or in some cases sold. They are gen-
erated on a continuous basis, but are generally disposed inter-
mittently.
Coke Fines
Coke which is produced in the course of various refinery
operations, such as fluid coking and delayed coking, is sold as
solid industrial fuel. Coke fines are generated intermittently,
and their quantity is a function of handling techniques. A cer-
tain amount of spillage and consequent contamination with dirt
results in the course of loading operations onto trucks and rail-
road cars. The contaminated coke is disposed.
II-5
-------
APPENDIX III
r
ESTIMATION OF THE PARAMETERS OF THE LOGNORMAL DISTRIBUTION
Because of the high degree of skewness in the distribu-
tion of generation rates from refineries, conventional statistics
are inadequate for efficient estimation of generation factors and
their variances. In addition to the skewness, a large percentage
of the refineries studied did not generate waste from all streams.
These sources affect the generation factor and therefore must be
considered in developing estimates for these factors.
Statistical Distribution Models for Generation Rates
A lognormal distribution was used to model the distri-
bution of generating sources. This distribution has the property
that when the original data are transformed by taking natural log-
arithms, the transformed data will follow a normal distribution.
The lognormal distribution is often appropriate when the standard
error of an individual value is proportional to the magnitude of
the value. The form of the lognormal distribution is as follows:
x - u)
f(x) = C^L 2a2 J for 0>x>»
xa/2ir
= 0 for x £ 0
2
Mean = exply 4-
Variance = exp[2y + 2a2] - exp[2y + a2]
In order to develop estimates for generation factors,
the non-generating sources also had to be modeled. A mixed dis-
tribution, specifically a lognormal distribution with a discrete
probability mass at zero, was used for this purpose. Letting p
equal the fraction of nongenerating sources in the population,
this distribution has the following form:
III-l
-------
f (x) = - - for
p for x = 0
0 for x < 0
Mean = (1-p) exp
' l_r 4J
Variance = (1-p) [exp(2y + a2)] [exp(a2) - (1-p)]
Efficient estimates of the mean and variance of the population
modeled by this mixed distribution have been developed [Finney
(1941), Aitchison (1955)]. These estimates are as follows:
The best, unbiased estimator of the population mean
generation rate is
GR =
and the best, unbiased estimator of the population variance of
the emission rates is
v
where n = number of refineries
r = number of refineries not generating waste for the
particular stream
m =» n-r = number of generating sources
g(t) = infinite series
= i + (n-Dt , (m-1)3 t2 (m-1)5 t3
L m tn'21 (m+1) m33! (m+1) (m+3) ' ' '
x = average of the logarithm of generating sources
n-r
= E In (generation rates)/(n-r)
1
S2 _ variance of the logarithm of generating sources
n~r — t
= Z [in (generating rates) - x]2/(n-r-l).
1
III-2
-------
The mean and variance formulas hold whenever there is more than
one generating source (n-r>l) . When only one generating source
is identified, the following estimates are appropriate:
Mean = — and variance = ^— ,
n n
where x, is the single measured generation rate. If no generation
is found (r=n) , then the best estimate for both the mean and vari-
ance is zero .
Computer programs were developed during a previous
study on refinery fugitive emissions for these estimators and the
estimator for the mean was used for all generation factors pre-
sented in this publication. Finney (1941) showed that this esti-
mator is more than twice as efficient as the arithmetic mean for
data distributed similarly to the generation rates from the re-
porting refineries .
Conf idence Intervals for Percent Sources Generating
and for Emission Factors
Confidence intervals for the percent of generating
sources were computed using the Binomial Distribution. The Bi-
nomial is used to model data when a random sample is selected
and each item is classsified into one of two categories (generat-
ing or non- generating here) . Exact confidence limits (level 1-a)
for the estimate of percent generating can be obtained by itera-
tion solving for P, in
I (^Pi^1-^)11'1 - T for the lower limit and for P in
i=kW L L L u
.* i)PuL(1~Pu)n~i = I for the upper limit
i=o \ /
III-3
-------
where n = number of refineries and k = number of generating sources
Tables of these solutions, available for most cases, were used to
develop 95 percent confidence intervals reported in this publica-
tion and for computing 97.5 percent confidence intervals which were
used in developing confidence intervals for generation factors.
97.5 percent was selected so that 95 percent confidence intervals
for generation factors would result when the estimated percent
generating was combined with the estimated mean generation rate
(0.975 x 0.975 « 0.95).
Patterson (1966) described how confidence intervals for
the mean from a lognormal distribution can be computed using esti-
mators developed by Finney (1941). 97.5 percent confidence inter-
vals were computed for the average, y, of the transformed data,
y = In (generation rate), using
CL = lower limit - y - 2.24
'U
upper limit = y + 2.24
sV(n-r)
: :
s2/(n-r)
and
where s2 is the variance of the transformed data and n-r is the
number of generating sources. Then, following Patterson's argu-
ments , confidence intervals for the mean generation rate can be
computed using:
lower limit =* exp
upper limit = exp
g(s2/2)
g(s2/2)
and
where g(t) is the series given above.
To obtain 95 percent confidence limits for the genera-
tion factors, the confidence limits for the percent generating
and for the mean generation rate were combined as follows:
III-4
-------
lower 95 percent limit for emission factor = P, (C,)
upper 95 percent limit for emission factor = ^n^lP
These confidence intervals are conservative in the sense that 95
percent is a lower bound for the confidence coefficient for the
intervals.
III-5
-------
APPENDIX IV
SELECTION OF REFINERIES FOR SUPPLEMENTARY SAMPLING PROGRAM
In order to develop a data base which is representative
of refinery solid waste streams' generation rates and volatile
hydrocarbon content, it is ncessary to conduct a technical/and
statistically valid sampling program.
In developing a sample base for the simulation of the
entire refining industry, some compromise must be made between
the number of refineries sampled and the accuracy of the sample
base. In March 1978, there were 285 refineries operating in the
U.S. (CA-679) . The U.S. refining industry could be most precisely
and accurately represented by extensively sampling 285 refineries.
Obtaining and analyzing such a large set of samples would be very
expensive. Reducing the number of refineries sampled, reduces the
costs of sampling and analyses of refinery solid waste streams,
but the reduction in the number of samples also reduces the ability
to accurately define the refining industry solid waste generation
rates and volatile hydrocarbon content. Three sampling schemes
utilized for modeling the U.S. petroleum refining industry are
discussed below.
Single Representative Refinery
It is often customary to propose a "typical" or "repre-
sentative" refinery model for purpose of illustration, discussion,
simulation or sampling.
An example of this type of representative refinery has
been defined for process analysis and stream characterization pur-
puses (BO-196). The flow diagram of this plant is presented in
IV-1
-------
Figure 1. The product slate for this refinery is given in Table 1,
and the capacities of the various individual processing units
within the refinery are shown in Table 2. This model refinery
was developed under the following assumptions:
• The refinery capacity was 15,900 m3/day
(100,000 barrels/day).
• The process units shown on the flow diagram
are those in common use in the refining
indus try.
• The capacities of the process units rela-
tive to the crude feedstock agree with the
average 1974 capacities of all U.S. refin-
eries, as shown in Table 2.
• The refinery product distribution is con-
sistent with that of the entire refining
industry in 1974, as shown in Table 1.
• The crude feedstock was a weighted composite
of crudes from the major oil fields supply-
ing petroleum to domestic refineries in 1974.
In theory, the entire U.S. refining industry could be
simulated by sampling a number of these representative model
refineries. Practically, however, there are a number of serious
drawbacks to this simulation methods, and only a very superficial
and imprecise analysis of nationwide refinery operations can be
accomplished with this technique. Some of the more obvious de-
ficiencies of this method are listed below.
• The types of refinery process units are
fixed. There are, however, many refineries
which utilize fewer process modules, and a
number which employ more than the repre-
sentative model refinery.
• The configuration of theprocess units within
the representative refinery is also fixed.
Although the configuration of the representa-
tive refienry can be developed, there may not be
an actual refinery with these characteristics.
For sampling, this approach has limited appli-
cations.
IV-2
-------
r~~ ™ p"i
*.*•• MM**"""!*! *°**** I
r
JIMISfJtsI l_iii
w« .«!*.»
IljUflt
a tMMMUKl I-'
\ ll.t «'*t
•I.UI tf/fer
•j f*T» KOMIt J
Itiimiu inn*
It.it* lt"«f
it • ••»**
tltM M>4
'fit wJ.r"
1
I —i •* .n
J Jil^^g>| >,TO ~ YM
•Tl IOMIIKI »• (- —
LJ.IUXUII. J MIM.
|--Si-£MS ••'&»».
itSnGIb" mim^
1 [m «•»*»
Ul Mk
II ••<«•»
I.IM «•>
I.Ill «•/«.»
I.t4l •'(*»
^flJ
rtn
.—•— .^ — .SSft* CfitfU9..
•M«*7 I.MI •'(<•»
| I *-UU - Itl.Mt >,««.,
•»•••
• Ml »'i
FIGURE 1. BLOCK FLOW DIAGRAM FOR A REPRESENTATIVE U.S. REFINERY
Source: BO-196
-------
TABLE 1
COMPARISON OF "REPRESENTATIVE" REFINERY PRODUCT
SLATE WITH TOTAL ACTUAL U.S. PRODUCTION
(1974 Data Base)
Volume % of Total Refinery Products
Representative
Refinery Total U.S.
Product Production Production
Gasoline 50.3 49.0
Kerosine 1.2 1.2
Jet Fuel,
Naphtha Type 1.5 1.5
Kerosine Type 5.0 4.9
Distillate Fuel Oil 20.4 20.4
Asphalt 3.4 3.4
Residual Fuel Oil 8.2 8.2
Marketable Coke 1.4 1.3
LPG 2.4 2.4
Petrochemical Feedstocks 2.8 2.8
Other (Fuels, miscellaneous) 3.4 4.9
100.0 100.0
Source: BO-196, US-209.
IV-4
-------
TABLE 2
REFINERY PROCESS UNIT CAPACITIES
"REPRESENTATIVE" COMPARED TO AVERAGE OF U.S. REFINERIES
(1974 Data Base)
Volume Percent of Crude Feedstock
Representative Average of
Unit Refinery U.S. Refineries
Reformer 24.6 27.3
Fluid Cat Cracker 28.9 33.9
Hydrocracker 5.6 6.9
Coking 1.4 1.7
Asphalt 3.6 5.4
Isomerization 0.8 1.0
Alkylation 5.6 6.8
Naphtha HDS 20.8 •: ' 25.2
Distillate HDS 11.3 13.7
Gas Oil HDS 3.5 4.2
Resid. Oil HDS 0.04 0.05
Source: BO-196, CA-236.
IV-5
-------
• While the representative refinery can effi-
ciently process the selected composite
crude oil, it is not suitable for refining
many of the individual crudes processed by
refineries across the U.S.
• The necessary specification of the indivi-
dual processing units and the associated
range of operating conditions further limits
the utility of this model refinery for repre-
senting the entire industry.
• A model refinery may be representative of the
variables considered in choosing the model
refinery but may not be representative of the
variables which are to be studied, e.g., solid
waste generation.
• The model refinery represents an average fig-
ure for refineries. Sampling model refineries
gives no information about how individual re-
fineries differ from this average.
In actuality, the only major variables that can be manipulated
within the representative refinery concept are the size of the
refinery (in terms of crude feedstock) and the process operating
conditions (within the operating constraints of each individual
process). For the purposes of developing quantitative and analy-
tical data describing refinery solid waste streams, the single
representative model refinery concept is clearly unsuitable. The
inflexibility of sampling one or even a group of model refiner-
ies prevents detailed analyses of the different types of refiner-
ies, refinery process units, solid waste characteristics and
quantities and refinery pollution control technology.
EPA/Refinery Classification and Proposed Model Refineries
One possible means of modeling the refining industry is
to categorize the nation's petroleum refineries according to a
logical set of basic characteristics, and to develop one typical
representative model refinery for each of the categories. The
total refining industry could then be simulated by sampling this
IV-6
-------
group of model refineries. Solid waste generation rates and com-
positions representative of the entire industry could then be
determined.
The EPA has developed refinery subcategories, shown in
Table 3, which are reflective of the wastewater loading with
respect to refinery type, processing units, and operating severity
(EN-407). A representative model of each of the five refinery
types could be developed and sampled as an average of all the re-
fineries in each classification. The processing unit charge
capacities and production capacities of each model refinery would
correspond to the average capacities (expressed as a percent of
total crude feed) of each of the various refinery types. It would
be possible to simulate the process unit capacities and product
output of the U.S. refining industry with selected groups of model
refineries. An extensive amount of work would be required, how-
ever, to calibrate the sampling data obtained from the models such
that they would best simulate the refineries within each category.
For example, it would be difficult to arbitrarily choose
specific types of processing units and associated operating con-
ditions that would accurately represent actual operations in the
different types of refineries. It would be necessary to obtain
from the refining companies a considerable amount of data concern-
ing the operation of the individual processing units within the
specific refineries. Companies are normally very reluctant to
divulge this type of proprietary information.
IV-7
-------
TABLE 3
SUBCATEGORIZATION OF THE PETROLEUM REFINING INDUSTRY
REFLECTING SIGNIFICANT DIFFERENCES IN WASTEWATER CHARACTERISTICS
Subcategory
Topping
Cracking
Petrochemical
Lube
Integrated
Basic Refinery Operations Included
Topping and catalytic reforming whether or hot the facility
includes any other process in addition to topping and
catalytic process.
This subcategory is not applicable to facilities which
include thermal processes (coking, visbreaking, etc.) or
catalytic cracking.
Topping and cracking, whether or not the facility includes
any processes in addition to topping and cracking, unless
specified in one of the subcategories listed below.
Topping, cracking and petrochemical operations, whether or
not the facility includes any process in addition to topping,
cracking and petrochemical operations,* except lube oil
manufacturing operations.
Topping, cracking and lube oil manufacturing processes,
whether or not the facility includes any process in addi-
tion to topping, cracking and lube oil manufacturing pro-
cesses, except petrochemical operations.*
Topping, cracking, lube oil manufacturing processes, and
petrochemical operations, whether or not the facility in-
cludes any processes in addition to topping, cracking,
lube oil manufacturing processes and petrochemical opera-
tions.*
*The term "petrochemical operations" shall mean the production of second
generation petrochemicals (i.e., alcohols, ketones,-cumene, styrene, etc.)
or first generation petrochemicals and isomerization products (i.e., BTX,
olefins, cyclohexane, etc.) when 15 percent or more of refinery produc-
tion is as first generation petrochemicals and isomerization products.
Source: EN-407.
IV-8
-------
In summary, sampling five model refineries instead of
one would provide more flexibility and inherently greater accuracy
in defining production rates and analytical characteristics of
refinery solid waste streams. Still, an extensive base of re-
finery operating data would be needed to establish the validity
of the model refinery samples as representative of the total
industry.
Refinery Cluster Models
In some recent work done for the EPA, Arthur D. Little,
Inc. (ADL), has defined the impact of several existing or pro-
posed environmental regulations on the U.S. refining industry
(LI-150, LI-151, LI-152). The development of a method for pro-
viding a detailed simulation of the U.S. refining industry con-
stituted an important part of these studies. The acquisition and
collation of an extensive refinery data base, which was used in
calibrating the refinery models, was accomplished.
By sampling the ADL cluster model refineries, solid
waste generation rates and analytical data from eighteen refiner-
ies could be used to accurately project generation rates and com-
positions of solid waste streams from the total U.S. refining
industry.
Model Development
In the method developed by ADL, the refining industry,
as it existed in 1973, was simulated by six individual refinery
models. Each of the models, called "cluster" models, lies in a
different geographical area of the U.S. and consists of a group
of three existing operating refineries. In terms of crude oil
type, process configurations, operating conditions, and product
IV-9
-------
distributions, each group consists of refineries that are typical
of the refining industry. Crude capacities of the model refiner-
ies ranged from 54,000 to 640,000 barrels/day; and each cluster
model represented the average operation of the three refineries
selected to comprise each group. Representatives of the EPA and
members of a task force from the American Petroleum Institute
(API) and the National Petroleum Refineries Association (NPRA)
assisted ADL in the selection of the six geographical areas and
the three refineries within each area.
The two most important criteria observed in the selec-
tion of representative regions and refineries were: (1) each
selected cluster model was to represent, as nearly as possible,
a typical and realistic refinery operation with respect to type,
size of processing units, and operating flexibility; and (2) the
crude slate, processing configuration, and product slates for
each model were to be representative of the variations peculiar
to each geographical region.
The six cluster models and the refineries comprising
each cluster are listed in Table 4. The capacities of the
refineries are shown for 1973, the calibration year, and as of
1 January 1976 and March 1978.
The Bureau of Mines has grouped refining operations
into geographical refining districts which correspond with dis-
tricts designated as PAD (Petroleum Administration for. Defense)
districts. These districts are briefly described in Table 5.
The number of cluster models selected to characterize each PAD
district reflects the refining capacity and the variations in the
type of available crude found in each of the districts. For ex-
ample, sufficient refining capacity and enough crude with common
characteristics are available in PAD District I to permit char-
acterization of that district with one cluster model, the East
IV-10
-------
TABLE 4
REFINERIES SIMULATED BY CLUSTER MODELS
PAD
District
I
II
II
III
III
Cluster
Identifi-
cation
East
Coast
Large
Midwest
Small
Mid-
continent
Texas
Gulf
LA Gulf
Refineries Simulated
Arco - Philadelphia, PA
Sun Oil - Marcus Hook, PA
Exxon - Linden, NJ
Mobil - Joliet, IL
Union - Lemont, IL
Arco - East Chicago, IN
Skelly - El Dorado, KS
Gulf Oil - Toledo, OH
Champlin - Enid, OK
Exxon - Bay town, TX
Gulf Oil - Port Arthur, TX
Mobil - Beaumont, TX
Gulf Oil - Alliance, LA
Shell Oil - Norco, LA
Cities Service - Lake
Charles, LA
1973
Capacity
MBPCD
160.0
163.0
255.0
160.0
140.0
135.0
67.0
48.8
48.0
350.0
312.1
335.0
174.0
240.0
240.0
1976
Capacity
MBPCD
185.0
165.0
265.0
175.0
150.0
126.0
78.7
50.3
53.8
390.0
312.1
325.0
180.4
240.0
268.0
1978
Capacity
MBPCD
185.0
165.0
290.0
180.0
151.0
126.0
81.0
120.0
54.0
640.0
335.0
325.0
196.0
230.0
268.0
West Mobil - Torrance, CA 123.5 123.5 123.0
Coast Arco - Carson, CA 165.0 181.5 180.0
Chevron - El Segundo, CA 220.0 230.0 405.0
Source: LI-152.
IV-11
-------
TABLE 5
PAD DISTRICTS AND BUREAU OF MINES
REFINING DISTRICTS
PAD District Refining District
I East Coast & Appalachian No. 1
II Appalachian No. 2 & Indiana-Illinois-
Kentucky & Oklahoma-Kansas-Missouri &
Minnesota-Wisconsin-North Dakota-South
Dakota
III Texas Inland & Texas Gulf Coast &
Louisiana Gulf Coast & North Louisiana-
Arkansas & New Mexico
IV Rocky Mountains
V West Coast
IV-12
-------
Coast cluster. On the other hand, about 40 percent of the total
U.S. refining capacity is contained in PAD District III. Two
cluster models were used to simulate the refineries in this dis-
trict because of its importance, and because two types of refinery
configurations and crude slates were identified.
PAD District II, containing about 28 percent of domestic
refining capacity, was also simulated with two refinery cluster
models, since two distinctly characteristic types of refineries
could be identified within the region. One was a large Midwest
type refinery processing over 100,000 barrels/day of high sulfur
crudes. The other type is typified by the small Midcontinent re-
finery in which 50,000-100,000 barrels/day of low sulfur crudes
are refined.
PAD District V was characterized by a single refinery
cluster model, the West Coast cluster. Since less than 5 percent
of the country's refining capacity is located within PAD District
IV (Rocky Mountains), it was not represented by a cluster model.
Four of the five EPA refinery types or categories (which
are very similar to the five API refinery classes) are represented
in the 18 refineries which make up the cluster models. The Type
A (topping) refineries are not represented. Although there are
92 topping refineries in the U.S., they process only about 8 per-
cent of the total crude. The omission of this type of refinery
from the cluster models should not cause any problems in a solid
waste study unless the Type A refineries are found to be sources
of inordinate amounts of solid waste.
A summary of the major refinery processing units and
capacities in each cluster model is presented in Table 6.
These are compared with the average processing unit capacities
IV-13
-------
TABLE 6
SUMMARY OF MAJOR REFINERY PROCESSING UNITS FOR 1973
(Percentage of Crude Capacity)
Processing Unit
Catalytic Reforming
Catalytic Cracking
Hydrocracking
Alky lat ion
Delayed Coking
Texas
Gulf
Cluster
23
33
6
6
6
LA
Gulf
Cluster
18
4)
4
11
9
PAD III
Average
24
. 31
5
6
8
Large
Midwest
Cluster
21
35
0
9
10
Mid-
Continent
Cluster
27
39
0
9
8
PAD II
Average
22
34
4
7
9
East
Coast
Cluster
23
34
5
4
0
PAD I
Average
21
33
3
4
7
West
Coast
Cluster
24
29
16
4
24
PAD V
Average
23
24
15
5
20
-------
(1973/1974) in refineries located within the same PAD districts
as the cluster models. In general, the capacities of the cluster
model(s) characterizing each PAD district agree with or bracket
the average capacities of the respective PAD districts.
There are a few cases with an above-average deviation
between the 1974 model values and the average district capacities
In PAD District I, the East Coast cluster contains no coking pro-
cesses, even though coking capacity exists within the district.
Coking operations were purposely omitted in order to have one
cluster model without a coking unit, since many refineries do
not carry out coking operations.
The capacities of the cluster model catalytic cracking
units in PAD District II are higher than the average district
capacities for this process. However, the cluster model refin-
eries contain no hydrocracking units, so the composite of crack-
ing operations compares well with the district average.
Utility of the ADL "Cluster" Model Refineries
The cluster model method defines eighteen refineries
which would provide a valid sample base for simulation of the
total U.S. refining industry. The advantage of sampling solid
waste streams from the eighteen ADL cluster refineries over the
two potential sampling programs discussed in Sections in the pre-
ceding sections are discussed below.
• The cluster models are based on individual
existing and operating refineries. Thus,
assumptions and conclusions based on work
done with the models could be verified
with existing data or through testing in
these refineries.
• An extensive data base from these refin-
eries and their operation already exists,
and much of it has been gathered and eval-
uated by EPA and ADL.
IV-15
-------
• The methodology for calibrating these
models and scaling them up has been devel-
oped. The procedures for altering or up-
I dating the models are available.
• The cluster models were developed in co-
operation with API and industrial repre-
sentatives. Thus, they should be familiar
with this simulation method.
• From the results of scaling up these models,
it appears that they are accurate in simu-
lating the U.S. refining industry.
• Eighteen operating refineries were used in
developing the cluster models. Eleven
different companies are represented. Thus,
a significant pool of existing refineries
is identified.
• The refinery cluster models are all from
different geographical areas of the country,
but each represents a site of significant
refinery operations. Environmental assess-
ments based on the cluster models would be
developed under different, yet realistic
atmospheric conditions.
• Each of the cluster models has, as its
feed, different composite crude oils. Thus,
the effect of various crude oils on refin-
ery operation and solid waste character-
istics can be evaluated.
• The changing characteristics of the crude
feedstock and the product slates have been
projected through 1985. Future operating
and solid waste characteristics can be
evaluated if desired.
• The processing configurations of each of
the refinery cluster models is different.
Thus, the characteristics of many differ-
ent processes can be evaluated while using
only the six cluster models.
There are some deficiencies and omissions in the ADL
cluster model simulation method.
• There are no topping refineries (EPA Type A
refineries) represented in the clusters.
There are over 90 of these refineries in
the U.S. Most of them are small, however,
IV-16
-------
and the total refining capacity of these
refineries amounts to only 8 percent of
the total U.S. capacity.
• Small refineries of less than 50,000 bar-
rels/day capacity are not represented in
the cluster models. Over half the refin-
eries in the U.S. fall in this category.
They refine only about one-quarter of the
total U.S. refinery crude input, however.
• Production facilities for petrochemicals
other than BTX are not included in any of
the cluster models. If petrochemical
manufacturing facilities are to be in-
cluded xtithin the scope of a planned
study, they will either have to be incor-
porated into the existing models, or a
new representative model or models will
have to be developed.
• Prepration and execution of a detailed
sampling program for eighteen refineries
could be very expensive.
In order to determine whether the models as developed
in 1974-75 are valid at the time of a planned study the current
capacities of refineries and refinery process units should be
compared to the base capacities. A recalibration of the models
might be required if there are significant changes in capacity.
Refinement of Cluster Model Site Selection
The number of cluster models and the refineries selected
by ADL to represent the five PAD classes reflect engineering judg-
ment on: (1) what factors represent a typical and realistic re-
finery and (2) which crude slates, processing configuration and
product slates for refineries selected within each model were
typical of the variations peculiar to each PAD. The extent to
IV-17
-------
which these model refineries agree with actual industry averages
cannot be determined from the data collected for them. There-
fore, no error limits or confidence intervals for the true gen-
eration factors can be derived.
Selection of refinery solid waste sample sites by the
ADL approach could bias the data generated. Factors which control
waste generation rates and characteristics have not been well de-
fined. Even if the model refineries are well chosen, information
gathered from sampling them would pertain only to the industry
average. If guidelines were to be set at something other than
the average, e.g., the 75th percentile, then the data collected
from the model refineries would be of limited use. An alterna-
tive method can be utilized for developing a data base to be
statistically analyzed.
The ADL approach does improve on the single representa-
tive refinery method by taking into account the different refinery
configurations in setting up the PAD clusters. These clusters
could serve as the basis for a stratified sampling approach to
gathering the necessary information. Using this approach, each
refinery would be classified into one of the six clusters and
then a random sample from each of the classes would be drawn to
determine the refineries to be sampled. This is a common method
for increasing the precision of estimates made from survey data
(see Hansen, Hurwitz and Madow).
Using this approach, poor selection of the clusters
would only decrease the efficiency of the estimates made from
the data and not result in an uncorrectable bias which would be
the base with the model refinery approach. In addition, con-
fidence intervals from the true average of all refineries and a
characterization of the distribution of refineries around this
average could be obtained.
IV-18
-------
APPENDIX V
SUPPLEMENTARY SAMPLING PROGRAMS
INTRODUCTION
This appendix addresses major areas to be addressed
when conducting a refinery solid waste sampling program:
• Streams to be sampled
• Analytical methods
• Cost of sampling and analyses
• Sample collection guidelines
Streams
The major refinery solid waste streams, the number of
samples required to characterize the stream and the phases to be
analyzed are listed in Table 1. The individual constituents to
be analyzed for are listed in Table 2.
Analytical Methods
The primary objective of the proposed analytical pro-
gram is to determine whether potentially hazardous VOC constituents
are present in refinery waste streams and to measure their levels
of concentration. Determination of trace element contaminants of
the waste streams is a second objective. Applicable methods for
analyses of the various constituents and rough cost estimates are
listed in Table 3.
Cost Comparisons
Using the PAD refinery classes the population of re-
fineries can be stratified into six categories. A random sample
V-l
-------
TABLE 1
SOLID WASTE STREAMS
NUMBER OF SAMPLES AND PHASES TO BE ANALYZED
Number of Analyses Per Refinery
Hydrocarbon Aqueous Solid
Phase Phase Phase
Refinery Background Samples
Intake Waters 2 2
Background Soil 3
Crude Oil 3 —3
Solid Wastes Samples
Once-Through Cooling Water
Sludge 1
Cooling Tower Sludges 2
Alkylation Sludge 1 1
Waste FCC Catalyst 1
Treating Clays 2 2
Tank Bottoms 4 44
Storm Water Silt 1
API Separator Bottoms 2 22
Air Flotation Float 2 2 2
Waste Biological Solids 1 11
Slop Oil Emulsion Solids 2 22
Exchanger Bundle Cleaning
Solids 1 11
Coke Fines 1 1
Boiler Feedwater Lime
Sludge — — 1
Treatment Residue Samples
Sludge Dewatering 2 2
Sludge Farming 2 22
Incineration 2 22
Total 23 20 36
V-2
-------
TABLE 2
REFINERY SOLID WASTE CONSTITUENTS TO BE ANALYZED
Hydrocarbon Phase Aqueous Phase
Arsenic Arsenic
Benzene Benzene
Benzo-a-pyrene Benzo-a-pyrene
Cadmium Cadmium
Chromium (Total) Chromium (Total)
Copper Chromium (Hexavalent)
Cyanide Copper
Fluoride Cyanide
Lead Fluoride
Mercury Lead
Nickel Mercury
Phenol Nickel
PNA (Priority) Phenol
Selenium PNA (Priority)
TOC-VOC Selenium
Vanadium TOC-VOC
Zinc Vanadium
Zinc
Solid Phase Total Sample
Arsenic Phase Portions
Benzene Hydrocarbon Phase %
Benzo-a-pyrene 0 Aqueous Phase %
Cadmium . Solid Phase 7o
Chromium (Total) Loss %
Copper
Cyanide
Fluoride
Lead
Mercury
Nickel
Phenol
PNA (Priority)
Selenium
TOC-VOC
Vanadium
Zinc
V-3
-------
TABLE 3
REFINERY SOLID WASTE STREAM ANALYSES
CURRENTLY USED METHODS AND COSTS
Analysis
Arsenic
Benzene
Benzo-a-pyrene
Cadmium
Chromium (Total)
Chromium (Hexavalent)
Copper
Cyanide
Fluoride
Lead
Mercury
Nickel
Phenol
PNA (Priority)
Selenium
TOG
Vanadium
VOC
Zinc
Method
AA-Graphite Furnace
Gas Chromatography/Mass Spectrometry
Gas Chromatography/Mass Spectrometry
AA-Graphite Furnace
AA-Graphite Furnace
AA-Flame
AA-Graphite Furnace
Distillation, Colorimetry
Specific Ion Electrode
AA-Graphite Furnace
AA-Cold Vapor
AA-Graphite Furnace
Distillation, Colorimetry
Gas Chromatography/Mass Spectrometry
AA-Graphite Furnace
Non-Dispersive Infrared Analysis
AA-Graphite Furnace
Purge/Trap-GC
AA-Graphite Furnace
Cost Per
Sample ($)
30
300
150*
30
30
40
50
50
20
50
50
40
67
100**
40
40
45
100***
40
*BAP only.
**In addition to BAP.
***Estimated.
V-4
-------
could be selected from each category to form the total sample. If
the categories actually do have different average generation rates,
then a more precise estimate of the population mean will be ob-
tained than with a simple random sample. Even if the categories
do not have different average generation rates, an unbiased
estimate of the population mean can still be obtained using the
stratified sample.
The number of refineries to be sampled from each category
depends on the resources available and the precision with which
it is desired to estimate the population generation factor.
After consulting with refinery personnel and Radian
engineers experienced in refinery sampling projects, sampling
plans for selecting 50, 80 and 110 samples from a refinery were
developed and are contained in Table 4. Cost estimates for the
three sample plans for 6, 12, 24 and 36 refineries were made and
are contained in Table 4.
In order to select a sampling plan i.e., the number of
refineries to be sampled and the number of samples per refinery,
information is needed on how well the population generation factor
is estimated using a particular plan. The variance of a sample
for a particular stream taken from a randomly chosen refinery is:
02 = °2B + V
where: a2 is the total variance of the sample
a2* is the component of the variance due
to differences between refineries.
a* is the component of variance associated
with the stream variation within the
refinery.
V-5
-------
TABLE 4
SOLID WASTE SAMPLING AND ANALYSIS COMPARISON
PRECISION VS. COST
Number of
Refineries
6
6
6
12
12
12
24
24
24
36
36
36
Samples Per
Refinery
50
80
110
50
80
110
50
80
110
50
80
110
(g/GF)xlOQ7,
72.4
69.3
66.1
51.2
49
46.7
36.2
34.7
33.0
29.6
28.3
27.0
(a/GF)xlOO%
72.4
66.1
59.1
51.2
46.7
41.8
36.2
33.0
29.6
29.6
27.0
24.1
Cost ($M)
.54
.82
1.13
1.08
1.69
2.18
2.19
3.31
4.65
3.20
4.93
6.85
V-6
-------
Using the relative magnitudes of a2fi and o^2 the proper balance
between the number of refineries and the number of samples per
refinery can be established. For example if a2,, is large com-
pared to Oy2 than it is more efficient to place emphasis on
visting a large number of refineries rather than sample exten-
sively from a few refineries.
If N refineries are sampled and n samples selected for
a particular stream, then the variance of the generation factor
is:
a 2 . a2B , V
GF = TT + ffiT
This assumes no gain in precision from the stratification of the
population. If the categories do have different generation rates
then the variance of the generation rate will be smaller than that
given by the equation above. For purposes of this analysis pre-
cision will be described by the coefficient of variation, (app/GF)
x 10070 that is, the standard deviation of the generation factor
expressed as a percent of the generation factor.
Since all refineries in the OAQPS study reported generat-
ing solid wastes from the API separator and since the generation
factor for this stream and its associated variance were typical
of other high generating streams, this stream was chosen as a
basis for the cost comparisons. In this study the generation
factor for API separator sludge was estimated to be 3.19 metric
tons/year per 1000 barrels/stream day; and its variance was 32.
It was assumed that within refinery variation a,,2 in this study
was comparable to that which would be obtained from 50 samples.
Using these values the precision of each of the 12 sampling plans
considered was calculated for two cases o2fi = 2aw2 and a2« = aw2.
These values along with the costs are given in Table 4.
V-7
-------
One conclusion which is obvious from looking at Table 4
is that the plans calling for 110 samples per refinery are not
economical. At least for the two cases considered a plan
with 50 samples can be found which has better precision and costs
the same or less. For example, considering the case a2fi = 2aw2,
the plan calling for 12 refineries and 110 samples per refinery
has precision 46.7 and costs $2.18m while the plan with 24 re-
fineries and 50 samples has precision 36.2 and costs approximately
the same, $2.19m.
The coefficients of variation can be related to how well
the generation factor will be estimated for a particular plan.
The largest discrepancy between the estimated and true generation
factor which could reasonably be expected is approximately twice
the entry in the table. For example, considering the case
a2B = ay2 and the plan of 12 refineries and 80 samples, the esti-
mated generation factor could reasonably be expected to be within
93.4% of the true generation factor.
Sample Collection
The following guidelines were developed after reviewing
the sample collection procedures utilized by the API and Jacobs
Engineering. Also, refinery personnel involved in the previous
sampling programs were contacted to discuss problems associated
with obtaining representative refinery waste samples.
1. Many of the samples required for this proposed sur-
vey have three distinct phases (hydrocarbon, aqueous, and solid).
While obtaining the sample, care should be taken to preserve the
ratio of the three phases in the sample taken.
2. Samples should be collected in sufficient volume to
minimize sampling errors and to provide enough sample for analyt-
ical aliquots to be supplied to two or more laboratories. Four-
fifths of a gallon in a new, clean one-gallon can is suggested
V-8
-------
for most of the samples. Samples from tank or vacuum trucks
should be no less than four gallons in a five-gallon crimped top
pail to be representative.
3. Plastic sample containers should not be used for
field samples.
4. As a minimum, samples should be labeled to include
the following information:
Date
Refinery
Unit
Person sampling
Identification number
Any special sampling conditions.
5. The discharge from sludge removal or conveyance
pumps, such as the pumps for the bottoms in an API separator or
for the float in an air flotation unit, should be sampled as any
refinery flowing stream would be sampled.
6. Wastes should be sampled as near as possible to the
points of origin in the refinery, and care should be taken to
prevent evaporation of volatiles from the sample. For some
streams, additional samples may be desired after transportation,
dewatering, etc.
7. In the case of liquid or semi-solid wastes, numerous
small grab samples taken over a period of time will yield a more
representative sample.
8. Field samples should be separated on the day col-
lected into equivalent analytical aliquots. The field sample
should be made a homogeneous as possible by the use of a paint
shaker, if available, or by vigorous Stirling. Aliquots should
be stored in glass. These aliquots should be refrigerated until
the analyses are complete.
9. On each sample date, refinery personnel assisting
in the sampling should be consulted to ensure that the particular
process is running normally and that the waste is as representa-
tive as possible.
Phase data summarized in Table 4-1 may assist the samp-
ling team in determining the tools and procedures required for
obtaining representative samples.
V-9
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-79-019
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Refinery Waste Disposal Screening Study
5. REPORT DATE
May 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Juanita Galloway
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
Suite 125, Culpeper Building
7923 Jones Branch Drive
McLean, Virginia 22102
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2608, Task 61
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning Standards
U.S. Environmental Protection Agency
Emission Standards and Engineering Division
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA 200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Published data and data supplied to the OAQPS in response to a 1978
request for information were used to estimate the quantity and phase
characteristics of major refinery solid waste streams. A supplementary
sampling program was described which will provide a data base for development
of an AP-42 factor for VOC emissions from refinery solid wastes. Solid waste
control technologies utilized by the refining industry were analyzed and
described. A listing was developed of operators involved in the organic
waste industry.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
Refinery Waste
Waste Disposal
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
187
20. SECURITY CLASS (Thispage)
Unclassifipd
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall cany a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant numbu under which report wu prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented'at conference of.
To be published in. Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or leaf factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and an sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COS ATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COS ATI Subject Category List. Since the ma-
jority of documents are raultidiscipiinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. ft 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Pom. 2220-1 (R«x. 4-77) (R«v*rsa>
-------
United Stales
Environmental Protection
Agency
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park NC 27711
Official Business
Penalty for Private Use
S300
Publication No. EPA-450/3-79-019
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
If your address is incorrect, please change on the above label;
tear off; and return to the above address.
If you do not desire to continue receiving this technical report
series. CHECK HERE D ; tear off label; and return it to the
above address.
------- |