PB88-125893
Case Studies of Hazardous Waste
Treatment to Remove Volatile Organics
Volume 1
Research Triangle Inst.
Research Triangle Park, NC
Prepared for
Environmental Protection Agency, Cincinnati, OH
Nov 87
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS
-------
PBdd-125693
EPA/600/2-87/094a
November 1987
CASE STUDIES OF HAZARDOUS WASTE TREATMENT
TO REMOVE VOLATILE ORGANICS: VOLUME I
Prepared by
C. Allen, M. Branscome, C. Northeim, K. Leese, and S. Harkins
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
Contract No. 68-02-3992
Task 50
Prepared for
Benjamin L. Blaney
Proiect Officer
Thermal Destruction Branch
Alternative Technologies Division
HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
19.S CURITy CLASS (This Repo,,j
UNCLASSIFIED
20 SECURITY CLASS (This pagcj
UNCLASSIFIED
21 NO OF PAGES
‘/7
TECHNICAL REPORT DATA
(Please read I wijcnons on the reve,ie before
co’nplenng)
2
1. U i
5 REP DATE
Waste Treatment to Remove November 1987
I 6 PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO
Northeim, K. Leese, and
ADDRESS 10. PROGRAM ELEMENT NO
11 CONTRACT/GRANTNO
27709 68-02-39g2, Task 50
ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
Research Laboratory
Development 1 4.SPONSO ENCY CODE
Agency EPA-600/ 12
for treatment of refinery wastes in a pilot-scale thin-
of volatiles from industrial wastewater for two steam
of semivolatiles from water by steam Stripping followed by
This report provides data on removal efficiency, air
treatment costs, and process limitations. Details on
procedures, quality assurance, and process data are contained
II).
KEY WORDS AND DOCUMENT ANALYSIS
b IDENTIFIERS/OPEN ENOED
TERMS
C COSATI Ficld/Goup
-- -- - -
RELEASE TO PUBLIC
EPA Form 2220 -I (p.73)
122 PRIC
i
-------
NOTICE
The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under Contract 68-02-3992, Task
No. 50, to Research Triangle Institute. It has been subject to the Agency 1 s
peer and administrative review, and has been approved for publication as an
EPA document. The use or mention of trade names or commercial products does
not constitute an endorsement or recommendation for use.
11
-------
FOREWORD
Today’s rapidly developing and changing technologies and industrial
products and practices frequently carry with them Increased generation of
solid and hazardous wastes. These materials, If improperly dealt with, can
threaten both public health and the environment. Abandoned waste sites and
accidental releases of toxic and hazardous substances to the environment also
have important environmental and public health implications. The Hazardous
Waste Engineering Research Laboratory assists in providing an authoritative
and defensible engineering basis for assessing and solving these problems.
Its products support the policies, programs, and regulations of the Environ-
mental Protection Agency, the permitting and other responsibilities of State
and local governments and the needs of both large and small and businesses in
handling their wastes responsibly and economically.
This report presents the results of field assessments of three waste
treatment techniques that have the potential for use in control of emissions
of volatile organic compounds from hazardous waste facilities by removing
those compounds from the waste streams. Those treatment techniques are thin-
film evaporation, steam stripping, and steam stripping with carbon adsorption.
The report is Intended for use by government agencies which are considering
ways to reduce eniisslons from hazardous waste facilities and by facility
operators and managers who wish to do the same. For additional information,
please contact the Alternative Technologies Division of the Hazardous Waste
Engineering Research Laboratory.
Thonias R. Hauser. Director
Hazardous Waste Engineering Research Laboratory
111
-------
ABSTRACT
Three treatment processes were investigated for the removal of volatile
organic (VO) compounds from hazardous waste: thin-film evaporation, steam
stripping, and steam stripping with carbon adsorption. The data collected
included the VO removal effectiveness, air emissions from the process, cost,
and process limitations.
The thin—film evaporator (TFE) study was a pilot—scale evaluation of the
TFE for removal of VO from petroleum refinery wastes. The study was performed
under different controlled conditions at three temperatures, three flow rates,
and under both vacuum and atmospheric pressure. The removal of volatile com-
pounds was greater than 99 percent, and the removal of semivolatiles ranged
from 10 to 75 percent depending upon the processing conditions. When the
system was operated under vacuum, some carryover of the feed resulted in a
condensate that was a milky—white emulsion, which would require additional
treatment to separate the oils and water. Vent rates from the condenser were
found to depend on the type of waste and the quantity of light hydrocarbons,
which are difficult to condense. The cost estimates for the TFE plus land
treatment of the residuals yielded costs that were comparable to or less than
the cost of land treating the original waste without pretreatment to remove
Vo.
Two full-scale steam strippers used to treat Industrial wastewater
containing about 6,000 ppm of purgeable VO were tested. The tray column
stripper processed about 850 L/min of water that contained primarily ethylene
dichloride and chloroform. Total VO removal averaged about 99.8 percent with
an average concentration of 9.7 ppm In the bottoms. The condenser removed
about 99 percent of the VO from the vapor and yielded a vent rate of about 20
Mg/yr. The packed column steam stripper processed about 42 L/min of water
that contained primarily methylene chloride and chloroform. Total VO removal
averaged 99.999 percent with an average concentration of less than 37 ppb in
the bottoms. The condenser removed about 91 percent of the organic vapors and
yielded a condenser vent rate of 11 Mg/yr. Emissions from the solids decanter
and storage tank were estimated as 46 Mg/yr. The tray column stripper proc-
essed water containIng 1.4 gIL of filterable solids compared to 0.01 g/L for
the packed column. Costs for the small unit were confidential; the costs for
the larger steam stripper were about $0.89/1,000 L treated.
The steam stripping/carbon adsorption unit was used to remove semivola-
tiles from water, which contained nitrobenzene, 2-nitrotoluene, and 4—
nitrotoluene. Steam stripping reduced the concentration from 634 ppm to 48
ppm, a reduction of 92 percent. Liquid-phase carbon adsorption decreased the
concentration In the bottoms to below detection limits (0.8 ppm) and yielded
an overall removal efficiency of greater than 99.6 percent. Maximum air emis-
sions were estimated as 35 kg/yr. The total cost of treatment was estimated
as $8.90/1,000 L treated.
This report is submitted In partial fulfillment of EPA Contract Number
68-02-3992, Task 50, by Research Triangle Institute under the sponsorship of
the U.S. Environmental Protection Agency. This report covers a period from
April 1986 to May 1987, and work was completed as of June 1987.
iv
-------
TABLE OF CONTENTS VOLUME I
Notice
Foreword
Abstract
LIST OF FIGURES v i i
LIST OF TABLES vi ii
SECTION
1 INTRODUCTION . 1—1
Background 1—1
Purpose of the Program .... 1—1
Procedures 1—2
Scope of the Report .. 1—2
2 CONCLUSIONS 2—1
General Conclusions ••••....... .•••••• 2—1
Thin—film evaporator (TFE) Conclusions 2-1
Plant I Steam Stripping Conclusions 2—4
Plant H Steam Stripping Conclusions 2-5
Plant G Steam Stripping/Carbon Adsorption
Conclusions 2-6
3 RESULTS OF THE THIN—FILM EVAPORATOR TESTS 3-1
DescriptionofP llotFacility.. .. .. 3—1
Tests Performed with Thin—film Evaporators 3-4
Process Residuals . . . . . . . . . 3—33
Bottoms Sludge 3_33
Organic Condensate 3—33
Aqueous Condensate 3—34
Cost of Thin—film Evaporation 3—34
4 FIELD TEST RESULTS: STEAM STRIPPING .... 4-1
Results forPlant l 4—1
Process Description for Plant I .. ....... 4—1
VO Removal From Water . . . . . . . . . . . . . . . . 4—3
A r Emissions •.••••• .••• 4—7
Results for Plant H . 4—21
Process Description for Plant H 4-21
Removal of VO From Water ... 4—23
Condenser Efficiency . . . . . . 4—31
Process Costs and Limitations 4—31
V
-------
TABLE OF CONTENTS VOLUME I (Continued)
5 FIELD TEST RESULTS: STEAM STRIPPING/CARBON
ADSORPTION •....... 5 —i
Site and Process Description for Plant G 5-1
Results for Plant G •.............. . 5—5
ProcessStreamComposition ... 5—5
Removal Efficiencies of the Steam Stripper-Carbon
Adsorber . . . • • • • • , • • 5—6
Process Limitations 5—6
Process Residuals 5—11
Air Emissions •..........,.. . 5—11
Liquid and Solid Residuals 5—14
Process Cost •••••. ......... ... ... . ., . 5—14
6 PROCESS LIMITATIONS AND COMPARISONS 6-1
Process Limitations for Thin—film Evaporation . 6—1
Process Limitations for Steam Stripping 6—1
Comparisons 6—2
7 REFERENCES 7—1
TABLE OF CONTENTS VOLUME II
APPENDIX A SAMPLING AND DATA COLLECTION PROCEDURES . A-i
APPENDIX B SUMMARY OF ADDITIONAL MEASUREMENT B-i
APPENDIX C PROCESS DATA C-i
APPENDIXD ANALTYICALPROCEDURES D-1
APPENDIX E QUALITY ASSURANCE.......................,......... E—1
vi
-------
LIST OF FIGURES
NUMBER PAGE
3—1 Configurationoftestequlpment 3—2
3—2 Condensate flow rates as a function of feed rate at
150 C 39
3—3 Bottoms flow rate as a function of feed rate at 150 C 3-10
3—4 Condensate flow rates as a function of feed rate at
320 ec 3—11
3-5 Bottoms flow rate as a function of feed rate at 320 °C 3-13
3—6 Condensate flow rates as a function of feed rate at
230 c •............ 3—14
3-7 Bottoms flow rate as a function of feed rate at 230 °C 3-15
4—1 Simplifiedschemat1cofsamp1jngpojnts... •..... 4—2
4-2 SchematIc of steam stripper and sampling locations 4-22
5—1 Plantlayout(nottoscale)................ 5—2
5-2 Flow diagram of continuous steam stripping and carbon
absorption unit •••s•I•I•••III..... 5—3
vii
-------
LIST OF TABLES
NUMBER PAGE
2-1 Summary of Steam Stripper Performance Purgeable Volatile
Organics •..... .. ••••.......... 2—2
2-2 Summary of Steam Stripper and Carbon Adsorber Performance
for Semi vol ati 1 es . . . . . . . . . . . • • • . . . . . 2—3
2—3 Summary of Thin-Film Evaporator Results for Two
Temperatures 2—3
3—1 Process Equipment Specifications . 3—3
3—2 Test Matrix for Thin—Film Evaporator Tests .. 3—5
3—3 Run Process Conditions and Flow Rates . 3—7
3—4 Condensate Organic/Aqueous Splits 3—8
3-5 Reduction in Headspace Volatile Organic Concentrations,
From onsiteGCAnalysisofiieadspace.. 3—18
3-6 Reduction in Headspace Volatile Organic Concentrations,
BacharachlLv snjfferResults......... 3—19
3-7 Volatile Analysis of Liquid and Sludge Process Samples
(CompuChem Data) . 3—20
3-8 Semivolatile Analysis of Liquid and Sludge Process
Samples (CompuChem Data) 3—22
3-9 Metals Analysis of Liquid and Sludge Process Samples
(CompuChem Data) • • • • • 3—23
3-10 Performance of Thin-Film Evaporator, Volatile and
SemivolatileCompounds,Run#5 . ..... .. 3—24
3-11 Performance of Thin-Film Evaporator, Volatile and
SemivolatileCompounds,Run#7 3—25
3-12 Performance of Thin-Film Evaporator, Volatile and
Semivolatile Compounds, Run #8 .. . 3—26
3-13 Performance of Thin-Film Evaporator, Volatile and
Sem lvolatileCompounds,Run#1O. . . .. 3—27
viii
-------
LIST OF TABLES (Continued)
SECTION PAGE
3-14 Measurements of Vent Gas Flow Rate and Bacharach TLV
S ni ffer Measurements . . . . . . . . . . . . . . . . . 3—28
3—15 Vent Gas Concentrations . . . . . 3—30
3-16 GC/MS Analysis of Gas Samples, Vent Gas, and Headspace
C lEA) 3—31
3-17 Oil, Water, and Solids Analysis of Process Streams for
Runs 5, 7, 8, and 1O.............................,.. . .... 3—32
3—18 TFE Cost Estimation 3—35
4—1 Stripper Feed (In) and Bottoms (Out) Concentrations for
First Test Day (ppm) .......... •••• .•••........... 4—4
4-2 Stripper Feed (In) and Bottoms (Out) Concentrations for
Second Test Day (ppm) ........ 4_5
4-3 Summary of Feed and Bottoms Concentrations (ppm) 4-6
4-4 Removal Efficiency from Water for First Test Day
(Percent) ••••iit•••••s•i. ......................... 4—8
4-5 Removal Efficiency from Water for Second Test Day
(Percent) 49
4—6 Results of Vapor Analyses 4—10
4-7 Comparison of Vapor Concentrations at Primary (S8) and
Secondary (S9) Condenser Vents (Volume Percent) 4—11
4-8 Vapor Flow Rate from Primary Condenser (S8) ............... 4—13
4-9 Flow Rate Measurements from Secondary Condenser Vent 4-14
4-10 Mass Flow Rates into and from the Primary Condenser (S8)
andCondenserEfficiency . 4—15
4—11 Comparison of Measured and Predicted Efficiencies for
Primary Consenser (S8) . . . . . . . •1• • • • . • • • • • . . . . . . 4—16
4—12 Estimates of Emissions from Storage Tank (Sb) .... 4—18
4—13 Estimates of Emissions from Solids Decanters (S12) 4—19
4—14 Summary of Vapor Emissions 4—20
ix
-------
LIST OF TABLES (Continued)
SECTION PAGE
4-15 Stripper Feed (In) and Bottoms (Out) Concentrations for
F I rst Test Day (ppm) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4—24
4-16 Stripper Feed (In) and Bottoms (Out) Concentrations for
Second Test Day (ppm) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4—25
4-17 Removal Efficiencies from Water for First Test Day
(Percent) 4—27
4-18 Removal Efficiencies from Water for Second Test Day
(Percent) • •••....... 4—28
4—19 Results forChioroform 4—29
4-20 Headspace Results for Feed and Bottoms (mg/L at 25 C) .... 4—30
4—21 Organic Loading on the Condenser (g/s) ................ . . .. 4—32
4-22 Condenser Vent Rates (g/s) and Condenser Removal
Efficiency (Percent) .. 4._33
4—23 Process Instrumentation 4—34
4-24 Cost Estimate for the Steam Stripper .............. ... 4—35
5—1 Process Stream Characterization ...............••.••••••• • 5—4
5-2 Measured Concentration of Organics in Process Streams 5-7
5—3 Process Streamcharacterjzation .............. 5—8
5-4 Component Mass Balance Around Steam Stripper and
Condensate Tank (Streams F, B, C, 0, S) ....... 5-9
5—5 Characteristics of Measured Organics ................... . .. 5-10
5—6 Removal Efficiencies (Percent) 5—10
5—7 Vent Gas Measurements, Sample Location 6—VOC 5—12
5—8 Estimation of Maximum Condensate Tank Vent Emissions 5-13
6—1 Comparison of Steam Usage Rates 6—3
6—2 Comparison of Performance . 6—4
6—3 Cost Compari son . . . . 6—7
6-4 Summary of Average Condenser Vent Rates and Efficiencies .. 6-8
x
-------
SECTION 1
INTRODUCTION
BACKGROUND
The Environmental Protection Agency (EPA) Office of Air Quality Planning
and Standards (OAQPS) is developing regulations under the 1976 Resource
Conservation and Recovery Act (RCR.A) and its 1984 amendments to control air
emissions from hazardous waste treatment, storage, and disposal facilities
(TSDF). The purpose of the air emissions regulations is to protect human
health and the environment from emissions of volatile compounds and
particulate matter.
Sources of volatile organic (VO emissions include storage tanks, treat-
ment processes, surface lagoons, landfills, land treatment, and drum storage
and handling facilities. Approximately 5,000 TSDF locations exist in the
United States where one or more of these activities is in progress. Most of
these sites are part of industrial facilities, and the rest are commercial
facilities that accept wastes from offsite.
Research has concentrated on the characterization of uncontrolled emis-
sions from these sources by using field measurements and by determining the
reliability of emission models. Recent investigations have identified a
number of options for controlling VO emissions from TSDF. These include
restricting the VO concentrations of wastes going to sources where emission
rates would be high, i.e., the M pretreatnient of waste to remove volatiles,
and the use of In-situ (i.e., add—on) control techniques at the TSDF.
Pretreatment is In current use at several TSDF. In general, it Is
attractive because It can be used by either the waste generator or the TSDF
operator to remove volatiles from the waste before they can be emitted into
the air. Pretreatment may be a cost-effective control technique for TSDF
emission sources with large surface areas, such as land treatment facilities
and lagoons. Fbrdisposal surface impoundments (e.g., evaporation ponds) and
aeration tankcand lagoons, pretreatment appears to be an important option
because these TSDF processes rely on transfer of water or oxygen between the
waste and the atmosphere as part of the disposal or treatment process, making
process covers unattractive.
PURPOSE OF THE PROGRAM
The purpose of these Investigations was to collect data for the support of
regulations that consider waste pretreatment as an alternative for the control
of volatile air emissions from TSDF. To the extent possible, these data were
collected from processes that were treating hazardous wastes or that were
* For the purposes of this report, the term “volatile organic (VO)” refers
to the combination of purgeable organics (eg, as detected in water using
EPA Method 624) and extractable organics (eg, as detected in water using
EPA Method 625). The terms “volatiles” and “semi-volatiles” are used
at some points to refer to purgeable organics and extractable organics,
respectively.
1—1
-------
treating wastes with physical characteristics similar to hazardous wastes in
order to permit a comparison of pretreatment to other emission controls.
Field data collected on several waste treatment techniques helped deter-
mine (1) how efficiently they remove volatiles from hazardous waste streams,
(2) what the removal costs are, (3) how the byproducts from the pretreatment
technologies are collected and disposed of, and (4) what limitations (in terms
of waste types, volatile concentrations, etc.) are placed on the use of such
treatment techniques.
PROCEDURES
The processes selected for evaluation included a pilot-scale thin-film
evaporator used to treat refinery sludge, two steam strippers used
to remove purgeable organic compounds from industrial wastewater,
and one steam stripper used In combination with liquid-phase carbon adsorp-
tion to remove semivolatile organic compounds from wastewater. Preliminary
site visits were conducted to observe and discuss the process operation and to
collect information on process limitations, costs, operating conditions, and
potential sampling points. During the preliminary visits, samples were taken
for screening to identify the organic compounds and concentrations in the
waste streams.
Detailed sampling and analysis plans, which also included the quality
assurance plan, were written for each site. These plans provided details on
the proposed sampling and analytical approaches, sampling points and number of
samples, and the quality assurance/quality control (QA/QC) procedures and
goals. Liquid samples taken for VO analyses were collected in vials with no
headspace and Included samples of the waste before and after treatment.
Samples of process residuals, such as sludge, recovered organics, and air
emissions, were also taken. Vapor samples were collected in evacuated and
electropolished stainless steel containers. These samples were analyzed for
organic compounds by gas chromatography (GC) and gas chromatography/mass
spectroscopy (GC/MS) procedures. Additional details on the sampling proce-
dures are given In Appendix A, and details of the analytical procedures are
given In Appendix 0. QA results are discussed in Appendix E.
The data were analyzed to determine the process removal efficiency for the
specific organic compounds found in the waste. Removal efficiencies were
calculated for each constituent and for total VO. Measurements of vent flow
rates and vapor concentrations were used to calculate or estimate air emis-
sions from the process. Process residuals were characterized in terms of
quantity and VO content. In addition, cost data were collected and evaluated
to provide an assessment of the total cost of the process and the cost—
effectiveness for VO removal.
SCOPE OF THE REPORT
This report documents the use of three different processes at four sites
to remove VO from waste streams. The first process was a thin—film evaporator
that was used in a pilot-scale study to evaluate the treatment of refinery
sludges. The results of this study are given In Section 3. Steam
1-2
-------
strippers used in the chemical industry to remove purgeable organics were
evaluated at two sites. These results are discussed in Section 4. The third
process was a continuous steam stripper that also used liquid—phase carbon
adsorption to remove semivolatile organic compounds from aqueous wastes. The
results of this field test are given In Section 5.
Conclusions for all four sites are given in Section 2. A summary of proc-
ess limitations and comparisons is provided In Section 6. AddItional details
are provided In the appendixes for sampling procedures, analytical procedures,
process data, and QA/QC data.
1—3
-------
SECTION 2
CONCLUS IONS
GENERAL CONCLUSIONS
1. Each of the processes Investigated successfully removed VO from the
wastes. The thin—film evaporator (TFE) removed over 99 percent of the VO
from petroleum refinery sludge, the two chemical industry steam strippers
removed 99.8 to 99.999 percent of the VO (purgeable organic compounds)
from the wastewater, and the steam stripper/carbon adsorber removed 99.6
percent of the VO (primarily semivolatiles) from aqueous wastes. The
results are summarized in Tables 2-1, 2—2, and 2-3.
2. Table 2-1 shows that steam strippers can reduce purgeable volatile organ-
ics from 6,000 ppm to ppm levels (<9.8 ppm) at Plant H or to ppb levels
(<37 ppb) at Plant I. Semivolatiles may be reduced by steam stripping
from over 600 ppm to about 48 ppm as shown in Table 2-2. Carbon adsorp-
tion of these semivolatiles reduced concentrations to below detection
limits (<0.8 ppm). The TFE generally reduced volatile compounds by over
99 percent. Semivolatiles such as naphthalene and methylnaphthalene were
removed efficiently (85-97 percent) at the higher temperature runs (Table
2-3).
3. The applicability of each of these processes depends in part on the solids
content of the wastes. The TFE can handle sludges that contain high—
boiling oils (17 to 25 percent oil) and solids (2 to 3 percent solids).
The steam stripper tests showed that solids may need to be removed prior
to steam stripping to 0.01 g/L (as done at Plant I) or the operator may
experience fouling and frequent cleaning (as seen at Plant H with 1.4
gIL). Solids removal prior to steam stripping generates a sludge contain-
ing VO that may be a troublesome disposal problem.
4. The various processes that generate air emissions are preliminary treat-
ment tanks (e.g., solids decanters), feed and storage tanks, condensate
collection and storage tanks, and process vents (e.g., condenser vents).
Condenser efficiencies for volatile organics ranged from 91 percent (cool-
ing tower water at 21 C) to 99 percent for a condenser cooled with
refrigerated glycol (2 C).
THIN-FILM EVAPORATOR (TFE) CONCLUSIONS
1. TFEs are able to process nonhoniogenous feed streams such as oily refinery
sludges. The major process limitations are that the feed and bottoms
product must be pumpable and the feed should not foam excessively during
processing.
2—1
-------
TABLE 2-1.
SUMMARY OF STEAM STRIPPER PERFORMANCE FOR PURGEABLE
VOLATILE ORGANICS
Compound In (ppm) Out (ppm) Percent reduction
PLANT H
1,2—01 chi oroethane
Chloroform
1, 1-Dichloroethane
1, 2—Di chi oroethene
Vinyl chloride
1, 1,2-Trichioroethane
Other volatiles
Methylene chloride
Chloroform
Carbon tetrachioride
Chloromethane
Other volatiles
5,630
271
11
8.9
8.4
7.5
14
5,950
4,490
1,270
55
33
11
5,860
9.6
<0.01
<0.01
<0.01
<0.01
<0.01
<9.8
0.011
0.006
<0.005
<0.005
<0.005
<0.037
99.998
96.5
>99.9
>99.9
>99.9
>99.9
>99.9
>99 .8
99. 999
99. ggg
>99.99
>99.98
>99.95
>99. 999
0.097
Total
PLANT I
Total
2-2
-------
TABLE 2-2. SUMMARY OF STEAM STRIPPER AND CARBON ADSORBER PERFORMANCE
FOR SEMIVOLATILES (PLANT G)
Nltrobenzene 2—Nitrotoluene
4—nitrotoluene Total
Concentrations (ppm)
To stripper
505 78
51 634
From stripper
41 2.4
4.4 48
From adsorber
<0.8 <0.8
<0.8 <2.4
Percent reduction
Stripper
92 97
91 92
Adsorber
>98 >67
>82 >95
Overall
>99.8 >98.9
>98.4 >99.6
TABLE 2-3. SUMMARY
OF THIN-FILM EVAPORATOR RESULTS
FOR TWO TEMPERATURES
Compound
Percent
In Outa reduction
Percent
outb reduction
Toluene
2,800 5.8—6.1 99.8
2.7-4.6 99.8—99.9
2—Methyl naphthal ene
790 320—660 16—59
99—120 85—87
Naphthalene
765 160—520 32—79
24-46 94—97
m-Xylene
280 1.3—3.8 99-99.5
0.7-0.9 99.7-99.8
o,p-Xylene
280 1.4—4.4 98-99.5
0.7-0.9 99.7-99.8
Benzene
230 <0.01-1.0 99.6->99.9
<0.01-0.6 99.7->99.9
Ethylbenzene
180 0.7—2.1 99-99.6
0.4-0.6 99.7—99.8
Styrene
160 0.8-2.5 98-99.5
1.2-1.6 99-99.3
aFrom Runs 5 and 7 at 150 C.
bFrom Runs 8 and 10 at 320 C.
2-3
-------
2. The TFE was found to have very high removal efficiencies of VO compounds
from the waste sludges that were tested. In each of the three methods
used to assess the reduction of volatiles, the removal efficiencies for VO
compounds were greater than 99 percent.
3. The removal efficiency for VO was greatest when the TEE was operated at
the highest temperature (320 •C). VO removal at this temperature general-
ly exceeded 99 percent, with no clear trends relative to changes in feed
rate.
4. The percent of semivolatiles removed from the feed ranged from 10 to 75
depending on the TEE operating conditions.
5. There were difficulties when the system was operated at high temperature
(320 C) under vacuum, as some carryover of feed Into the condensate was
observed. The condensate from the vacuum runs was a milky-white emulsion
that would require additional treatment to separate the oils.
6. The gas flow rates and total VO emissions from the TEE condenser were
highly dependent on the waste being processed. The first waste, an emul-
sion tank sludge, showed only minimal (less than 250 mL/min) flows from
the condenser, and the second waste, oily tank bottoms, showed much higher
(0.75 to 10 Llmin) vent gas flow rates. All of the condenser vent gas
concentrations were greater than 10,000 ppm (reported as hexane). The
high VO concentrations in the vent gas were due to the vapor pressure of
light hydrocarbons at the cooling water temperatures. A glycol-cooled
condenser, a two-stage condenser (first stage cooling water, second stage
chilled glycol), an incinerator, or some other appropriate control device
could be used to reduce these emissions. The condenser and vent gas con-
trol system should be designed specifically for the waste to be treated
because different wastes may contain different quantities of noncon-
densible or difficult-to—condense compounds.
7. The approximate capital and operating costs of TEEs when used to process
petroleum waste sludges using various operational modes range from com-
parable to less than the cost of conventional land treatment. The cost of
TFE sludge treatment was either $27.60, $40.60, $97.40 or $128/Mg depend-
ing on the mode of operation as compared to a cost of $110/Mg for land
treatment. The process does not eliminate land treatment and the cost
analysis assumes that the sludge from the TEE is disposed of by land
treatment.
PLANT I STREAM STRIPPING CONCLUSIONS
1. The steam stripper reduced the total VO concentration by over 5 orders of
magnitude from a feed concentration of roughly 6,000 ppm (0.6 percent) to
less than 0.037 ppm. The removal of total VO was approximately 99.999
percent.
2. The primary condenser removed about 91 percent of the total VO in the
vapors. Efficiencies for Individual constituents ranged from 89 percent
for chlorornethane to 94 percent for chloroform.
2-4
-------
3. The secondary vent condenser (with cooling tower water) did not appear to
provide measurable control or condensation of VO. Theoretical calcula-
tions indicate that using refrigerated glycol cooling on the secondary
condenser may improve its control efficiency for total VO to 68 percent.
4. The major air emission sources for the process are the solids decanters,
storage tank, and noncondensibles from the steam stripper. Emissions were
estimated as 2.7 grams (g) per L of water treated. For an average treat-
ment rate of 41.6 liters per minute (L/min) or 11 gallons per minute
(gal/mm) for 75 percent of the year, annual emissions are estimated as 44
megagrams per year (Mg/yr) or 1.4 grams per second (g/s). These annual
emission estimates assume a constant feed concentration of 6,000 ppm.
5. The vapor flow rate from the primary condenser when It was vented directly
to the atmosphere was measured as 57 L/min. The emissions were measured
as 2.4 gIL of water treated or 39 Mg/yr (1.2 g/s) for operation for
75 percent of the year. When the primary condenser was vented to the
secondary condenser, the flow rate from the secondary condenser was
measured at 11 to 13 L/min with an emission rate of 0.5 gIL of water
treated or 8.2 Mg/yr (0.26 g/s) for operation for 75 percent of the year.
The difference in measured flow rates (57 versus 11 to 13 L/rnin) suggests
that some flow was not sampled because of leaks In the overhead system,
overflow pipes, or tank vents. The estimates of annual emissions are
based on the conditions during the test with an average feed concentration
of 5,000 ppm.
6. Cost data supplied by the company were classified as Confidential Business
Information (CBI) and are not included in this report.
PLANT H STEAM STRIPPING CONCLUSIONS
1. The steam stripper reduced the total VO concentration by approximately 3
orders of magnitude from a feed concentration of roughly 6,000 ppm (0.6
percent) to an average of 9.7 ppm. Removal of the major constituent (1,2-
dlchloroethane) consistently exceeded 99.99 percent.
2. The cost• effectjveness of the steam stripping operation was approximately
$220/Mg of VO removed.
3. The overhead condenser removed 99+ percent of the total VO in the overhead
vapors. The condenser efficiency was much lower for specific individual
compounds present at low parts per million levels in the stripper
influent.
4. The flow rate from the condenser vent ranged from 1.9 to 4.2 L/s (4.0 to
8.8 ft 3 /mln). The condenser vent on this steam stripper was routed to an
incinerator. A similar system vented to the atmosphere could emit 12 to
51 Mg/yr of VO.
5. Steam usage for this steam stripper appeared to be optimized because it
was lower than values observed for other steam strippers and was also
lower than values given in design manuals.
2-5
-------
6. The removal of all constituents was consistently high except for chloro-
form. The variations in chloroform removal appear to be related to foul-
ing from the accumulation of solids. Suspended solids concentrations in
the stripper influent were on the order of 0.1 percent.
7. The major operational problem experienced with this steam stripper is the
fouling of the heat exchanger and column trays. Solids removal prior to
the steam stripper may provide a more consistent operation. The results
Indicate that a steam stripper can be operated for wastewater containing
0.1 percent solids if the operator is willing to backflush and clean the
system periodically. However, if the solids are removed prior to steam
stripping, the resulting sludge may be a troublesome disposal problem and
an additional source of VU. Consequently, the company has chosen to Incur
the additional cost of cleaning the existing system periodically instead
of installing equipment for the removal, treatment, and disposal of
solids.
PLANT G STEAM STRIPPING/CARBON ADSORPTION CONCLUSIONS
1. Semivolatlle organic compounds can be removed from wastewater using steam
stripping and carbon adsorption. Removal efficiencies of 92 percent were
observed for the steam stripper, and the carbon adsorber removed more than
95 percent of the organics fed to it. The removal efficiency of the
combined steam stripper-carbon adsorber was greater than 99.6 percent.
2. Air emissions from the condenser vent were very low, and the gas flow from
the vent could not be measured reliably. Concentrations of VU in the vent
stream varied widely, with total VO (as ppm hexane) between 10 and 2,000
ppm. The maximum air emissions were estimated to be 4.0 g/h.
3. Carbon was added to the adsorbers In a pulse feed mode an average of 1.5
times per day, with a carbon addition of 908 to 1,360 kg/charge. The
carbon was regenerated offsite and was the major cost of the process. The
organic concentrations of the wastewater fed to the adsorber were rela-
tively low (47.8 ppm), and utilization of the carbon was correspondingly
low (0.021 kg organics removed/kg carbon used). The total annualized cost
of the steam stripper-carbon adsorption system was $14.30/kg organics
removed.
4. The high normalized operating costs of the system resulted primarily from
the low feed concentrations (634 ppm organics) and the high removal effi-
ciency (>99.6 percent) of the steam stripper carbon adsorber. On a water-
processed basis, the total annualized cost was $0.0089/kg water treated
(or 0.89 /L).
5. ApproxImately 78 percent of the steam used in the steam stripper was con-
densed into the water being stripped, and 22 percent was condensed with
the stripped organics. This condensation into the stripped liquid pro-
duces a varying gas/liquid (GIL) ratio within the column: 55 m’ /m 3 at the
base and 24 m 3 /& at the top. The heat exchanger, used to heat the feed
with the bottoms from the stripper column, reduced the steam requirements
for the column.
2-6
-------
6. Principal variables influencing
feed rate and steam rate of the
reported as less than 1 percent
fouling as the only maintenance
the effectiveness of the process were
process. Downtime of the process was
of operating time, with heat exchanger
problem.
the
2-7
-------
SECTION 3
RESULTS OF THE THIN-FILM EVAPORATOR TESTS
DESCRIPTION OF PILOT FACILITY
The tests were conducted at the pilot facility of a manufacturer of TFEs
(Luwa Corporation, Charlotte, NC). This facility contains a variety of
evaporators produced by the company and is used to test potential applications
of their equipment for clients. Consequently, the equipment used for the
tests was configured specifically for our applications. This pilot facility
was used to evaluate the effectiveness of a TFE for the removal of VO from
petroleum waste sludges. Samples of tank bottoms sludge were obtained from an
oil company refinery, tested in the equipment, and then returned to the
refinery for normal disposal.
Figure 3—1 shows the equipment used in the pilot—scale tests. The 380—L
(100-gal) feed tank was agitated with both an axial mixer and continuous re—
circulation of the feed liquid through a centrifugal pump. A Moyno positive
displacement pump was used to pump the feed sludge through the preheater and
into the top of the TFE. The sludge was continuously spread over the heated
surface of the TEE as it progressed down through the TEE to a collection pot
at the base. Materials evaporating in the TEE passed through an empty
entrainment separator (demister) and were condensed in a condenser that was
cooled with cooling tower water. Condensate flowed from the condenser and was
collected directly into liquid sample jars or a flask used for measuring con-
densate flows. Any uncondensed vapors flowed from the condenser through a wet
testmeter for flow measurement.
The TEE was heated by hot oil, although steam could be used for lower
temperatures. Entrainment separators are used frequently with TFEs to remove
entrained liquids from the vapors flowing to the condenser. The entrainment
separator was empty (but heated) during the testing, and very little materia’
condensed In the entrainment separator. The feed lines from the preheater to
the TEE were heated with low—pressure steam, as were the vapor lines from the
TFE through the entrainment separator and to the condenser. Table 3—1 is a
complete list of the process equipment used in the tests.
In actual operation, TEEs require relatively little maintenance and oper-
ator supervision. Feed material Is usually charged to a feed tank, the equip-
ment is preheated, and processing is started. Condensate and bottoms are
collected separately and emptied periodically by an operator. During the
pilot tests, three Luwa personnel operated the equipment, measured flow rates,
and collected samples. This was necessary because of the relatively rapid
changes In operating conditions and the extensive sampling of the process.
3—1
-------
Steam
Vent
To Vacuum
Recirculation
Pump
Process Streams:
F Feed
B Bottoms
C Condensate
V Vent
Belt Drive (1300 rpm)
100 Gallon
Feed Tank
Condenser
(Water)
Entrainment
Separator
(no demister p
1
Heated
Oil
Moyno
Feed
Pump
Aim.
Steam
Evacuated Pot
with 5 gallon bucket
Figure 3 -1. Configuration of test equipment.
-------
TABLE 3-1. PROCESS EQUIPMENT SPECIFICATIONS
Luwa thin-film evaporator type LN-0028, 3.02 ft 2
thermal surface area heated with oil
Upper seal — water
Lower bearing - roller
1.5 hp, 440 Vt three phase
fications
100-gal jacketed feed tank with agitator and recyle
pump
MOYNO positive-displacement pump with variable speed
drive
15 ft 2 shell-and—tube type heat exchanger heated with
controlled steam
15 ft 2 shell-and-tube type heat exchanger cooled with
tower water
Evacuated pot containing 5-gal bucket
4,000 mL flask
Stainless steel, type 304 and 316
Test Unit:
Lubrication:
Rotor drive:
Auxiliary eguioment speci
Feed from:
Feed pump:
Feed preheater:
Condenser:
Bottoms receiver:
Distillate receiver:
Material of
constant:
3—3
-------
The major maintenance on field units Is periodic replacement of the feed
pump and bottoms pump. The frequency of required replacement is highly
dependent on the materials processed, with abrasive feedstocks wearing out the
pumps more rapidly than nonabrasive feeds. In this test progressive failure
of the Moyno feed pump was noted and determined to be due to polymers in its
rotor and cavity absorbing hydrocarbons In the feed sludge, reducing the
pump’s effectiveness.
TESTS PERFORMED WITH THIN-FILM EVAPORATORS
A total of 22 tests were performed with the TEE, using two different
wastes, three temperatures, three flow rates, and under both atmospheric and
vacuum operation. The feed rates and temperatures were chosen to operate the
TFE over its normal range of operation and to demonstrate the removal of VO
from the feed sludges. The testing occurred over 1 week at the Luwa Corpora-
tion pilot-test facility in Charlotte, NC (September 8 to 11, 1986).
Both waste sludges tested were provided by the same oil company and were
shipped to the Luwa test facility in 55—gal drums. Tests 1 through 18 used
five 55—gal drums of the first waste, an emulsion tank sludge, while tests 19
through 22 were performed on the second waste sludge, oily tank bottoms. The
process testing, sampling, and analysis concentrated on the tests using the
first waste while the second waste was used to gather additional process data
and demonstrate process operation on a second waste sludge. During the first
day of testing (September 8, 1986), the first two barrels of the first waste
were charged to the feed tank. Although the sludge was well mtxed prior to
charging and agitated in the feed tank, it separated into two layers. The
water-rich layer at the base of the feed tank was pumped Into the TEE, pro-
ducing very little bottoms and more water than expected in the condensate.
Once this problem was discovered, an additional recycle pump was installed on
the feed tank to better mix the sludge and prevent it from splitting into two
phases. The remaining sludge in the tank was used for the first four shake-
down runs on September 9, 1986, which were not used as part of the test plan.
This allowed practice samples to be taken and potential problems to be solved
prior to adding a new drum of waste to the feed tank for use in the actual
test runs.
The basic test plan (and run numbers for the tests) is shown in Table 3-2.
This test plan studied the two major variables affecting TFE performance--the
temperature of the heating jacket and the feed rate. The indicated flows and
temperatures were the nominal process parameters during the test, and the
actual measured parameters varied somewhat from these values. Shakedown runs
1, 2, 3, and 4 were performed with the remainder of the first two barrels of
waste that experienced a phase split on September 8, 1986. Test runs 5, 6,
and 7 were a series of tests at a constant heating jacket temperature (150 °C)
at three different feed rates. Runs 8, 9, and 10 were conducted at similar
flow rates to runs 5, 6, and 7 but were at a much higher heating jacket tem-
perature (320 °C). Runs 12, 13, 14, 15, and 16 were conducted at an inter-
mediate temperature (230 C) and were limited to the two lower flow rates
because of the progressive failure of the Moyno feed pump. This pump was
equipped with a polymeric-based rotor and stator, which absorbed minor amounts
3-4
-------
TABLE 3-2. TEST MATRIX FOR THIN-FILM EVAPORATOR TESTS
Tests with feed No. 1, emulsion tank bottoms:
Temperature (C) = 150 230 310
Flow rate (lb/h )
70 RUN 5 RUN 14 RUN 8
100 RUN 6 RUN 15 RUN 9
RUN 16
150 RUN 7 RUN 10
Vacuum runs:
Run 11: 320 C, 23 lb/h
Run 17: 150 C, 57 lb/h
Run 18: 100 C, 57 lb/h
Tests with feed No. 2, oily tank bottoms:
Temperature (C) = 150 310
Flow rate (lb/h )
45 RUN 21 RUN 19
80 RUN 22 RUN 20
3—5
-------
of feed hydrocarbons, degrading pump performance. Runs 12 and 13 were shake-
down tests at the intermediate temperature.
Three vacuum runs were performed with the first waste sludge during runs
11, 17, and 18. These runs were to examine the effect of vacuum operation on
the removal of VO from the feed waste. One run (#11) was at high temperature
and a very low feed rate. This run was to demonstrate the large volume reduc-
tion of bottoms that result when the system Is operated at high temperatures
and under vacuum. The two low—temperature vacuum runs (17 and 18) were to
discover whether the process could achieve substantial removal of volatiles
and water at lower temperatures (vacuum reduces the boiling point of water,
decreasing the operating temperature of the TFE). The final four runs (19,
20, 21, and 22) were performed with the second waste. These tests were
principally to demonstrate the operation of the TFE with a second waste sample
and were not extensively sampled and analyzed during the project.
Four of the tests (5, 7, 8, and 10) were selected for extensive sampling
and analysis of process streams. These four runs allowed the process to be
examined with both high and low feed rates and at both high and low heating
temperatures. They represent the range of reasonable operating conditions for
the TFE processing waste sludges for the removal of volatiles, water, and oils
from the sludge.
The complete operating process conditions, flow rates, and aqueous!
organic splits of the condensate are shown In Table 3—3. Table 3-4 summarizes
the organic/aqueous splits that occurred in the collected samples. Relevant
features of the process results are described below:
1. The three runs at 150 C (5, 6, and 7) show only a slight decrease in
condensate flow (16.4 to 14.6 lb/h) as the feed rate was increased
from 72 to 154 lb/h. The temperature of the bottoms from the TFE
remained constant near 100 C, indicating that water was still
evaporating from the bottoms as it left the evaporator. This shows
that at low temperature the process was heat transfer limited, with an
almost constant amount of material being evaporated from the feed
regardless of feed rate. Figure 3—2 shows the condensate flow rate as
a function of the feed rate. The condensate flow rate decreased
slightly as the feed rate was increased because of the additional
heating of the feed sludge from about 70 to 100 ‘C. The organic con-
densate flow rate Increased proportionally with the Increases in feed
rate, showing that the organics more volatile than water were being
evaporated from the feed and recovered as organic condensate. The
aqueous flow rate shows exactly the opposite relationship-—as the feed
rate increased, the aqueous condensate flow rate decreased. Fig-
ure 3—3 shows the bottoms flow rate for the same set of data. The
bottoms flow rate from the process increased proportional with the
feed rate, with the bottoms containing unevaporated water, oils, and
solids.
2. The three runs at 320 ‘C (8, 9, and 10) showed a substantially differ-
ent variance of condensate flow rate with feed rate. Figure 3—4 shows
that the condensate flow rates for these runs were much higher than
for the lower temperature runs, and the bottoms temperatures were 216,
3-6
-------
TABLE 3-3 RUN PROCESS CONDITIONS AND FLOW RATES
Cond.n ..t.
Bot-
tom.
Op.rat-
I ng
TFE
H. .tin
F..d .trasm
Bottom,
Con—
den-
Or
unIc friction
Agu .ouu friction
Oil r.—
W.t.r is-
pro-
duc.d
pr..
Flo.
Floe
Viper
cit.
Floe
Pl c .
Flee
co .r.d,
coo.r.d
3
LIJWA •ur.
Run I ev If
jn.
C
O *.,
C
‘r • r.t.
C Sb/h
I 7 ep
C
r.t•
lb/h
limp
C
limp.
C
rsL .
lb/h
of con-
d.n..t.
rst.• D.n.Sty
lb/h g/.4.
of con-
d.ns.t .
nt.
lb/h
D.n.Ity,
g/.4.
3
of f..d
3
of feed
of
s.d
5
ATM
160
150
62 75 0
90
65 2
05
26
*6 4
0 100
I 74 0 0d4
0.094
*4 06
0.99?
2.4
20 5
11 1
6
ATM
*60
*60
10 *07 6
102
93
80
20
*4 0
6.543
2.52 0 6*2
0.067
*2.00
0 994
2 0
I I 0
06.3
7
ATM
166
*60
14 *63.7
*00
*39
00
26
*4 •
e sea
a.., 0.196
0.6*?
1*93
0.996
* 7
7 0
90 4
6
ATM
329
322
1) 08.6
*04
0.6
60
24
60
0.11)
6.70 0.840
0.081
63.22
0.993
9.9
77.7
*2 4
0
ATM
338
326
60 1*3 I
210
I I 7
11
28
102
0. 521
*2 9* 0.050
0.07)
09 69
1.606
*1.4
18 4
*0 3
so
ATM
310
US
00 *43 4
200
*0 o
12
20
*24.0
0.50)
*2.61 0.034
0.09?
5* 5.93
0.993
9.0
76. 5
*3 0
I I
45
139
334
42 23 2
*42
I
62
29
22.2
I 693
*1.11 0 980
0.401
9.03
0.993
50.7
30 9
4 3
I)
ATM
234
220
30 02
*04
22
10
38
60
0.001
4.86 0 854
0.919
65*6
0907
6.9
013
260
$4
ATM
236
228
12 682
124
10.2
72
38
68
0083
4 19 0038
0.9*?
63.21
0.900
10
780
150
15
AIM
234
228
08 94
*00
26 4
72
30
66 6
0 804
6 50 0.875
0.910
60 00
0.909
5 0
63 0
30 2
16
ATM
234
228
08 111 5
100
48
13
38
65 6
0 071
6 62 0,03*
0.92)
00.40
0 991
4 6
64 2
4* 3
I l
45
*50
*40
50 61 2
66
34 0
66
32
22 4
0 67?
5.12 0.83*
0.92)
20 60
0 99*
3 6
30 2
60 8
I I
46
*62
*00
42 67 0
62
40 2
12
38
57 6
0.044
6 17 0 0*0
0.058
*8 0)
0 094
*3
20 1
69 6
59
AIM
34*
337
04 63 0
196
0
66
10
46 I
0.424
*9 32 0 9*4
0 616
20 28
1.000
38 0
49 0
$1 0
20
AIM
148
334
03 00
2*6
*2.5
02
30
81 2
0 362
24 30 0.020
0 638
42.90
0 998
30.4
63 6
*8 0
21
ATM
*55
149
88 41 3
100
20 4
92
34
54 9
0.303
4.52 0.708
0 691
10 30
6 997
*0.9
26 1
63 9
22
AIM
15*
149
82 77.6
*00
64 2
16
34
23 3
0.217
8.46 0.176
0.72)
*0.86
0.093
8.3
2* 1
69 9
c i . . )
-------
TABLE 3-4 CONDENSATE ORGANIC/AQUEOUS SPLITS
LUWA
Aqueous
Organic
Organic/aqueous
run
No.
Cond.nsste
samples
Volume,
nt.
Weight,
g
Density,
g/irL.
Volume,
mL
Weight,
9
Density,
g/iriL
split
weight S organic
5 LUWA—84,46 1,850 1,845.0 0.997 294 195.3 0.864 10.8
8 LUWA—49 810 805.1 0.994 168 134.8 0.812 14.3
7 LUWA—62,63 1,624 1,518.8 0.996 427 340.0 0.798 18.3
8 LUWA—71,72 1,665 1,649.2 0.991 250 210.0 0.840 11.3
9 LUWA-88 806 806.0 1.000 136 118.7 0.858 12.7
10 LUWA—89,90 1,718 1,705.7 0.993 235 196.1 0.834 10.3
11 LUWA-101 365 362.3 0.993 650 528.0 0.980 59.3
13 LUWA-112 885 873.6 0.987 90 76.9 0.864 8.1
14 LUWA—liS 890 881.5 0.990 95 79.4 0.836 8.3
15 LUWA—118 885 876.4 0.989 98 80.2 0.836 8.4
16 LUWA—121 890 881.6 0.991 88 73.1 0.831 7.7
c c 17 LUWA—124 890 881.6 0.991 88 73.1 0.831 7.7
18 LUWA-127 1,000 994.1 0.994 66 46.7 0.816 4.4
19 LUWA—133 610 509.8 1.000 410 374.8 0.914 42.4
20 LUWA-136 680 579.1 0.998 400 328.1 0.820 36.2
21 LUWA—139 636 632.8 0.997 360 276.7 0.788 30.3
22 LUWA—142 655 660.2 0.993 320 248.8 0.778 27.7
-------
20 -
19 -
18 -
17 -
16 -
15 -
14 -
13-
I
N 12-
-J 11-
‘ -I
LI) 10-
9-
8-
0
-J 7-
tj
6-
5-
4-
3-
2-
1—
0—
50 70 90 110 130 150
FEEDRATE (LBS/HR)
+ TOTAL 0 ORGANIC AQUEOUS
Figure 3-2. Condensate flow rates as a function of feed rate at 150 °C.
-------
160 -
150 -
140 -
130 -
120 -
110 -
100-
N
90-
U) 80-
( 4
70-
40 -
30 -
20-
10-
0 — I I I I I - I
50 70 90 110 130 150
FEEDRATE (LBs/HR)
Figure 3-3. Bottoms flow rate as a function of feed rate at 150 °C.
-------
FEEDRATE (LBS/I-IR)
+ TOTAL 0 ORGANIC A AQLiEOiJS
FIgure 3-4. Condensate flow rates as a function of feed rate at 320 °C.
I
N
Ill
( F
0
‘if
130
120
110
100
90
80
70
60
50
40
30
20
10
0
50 70 90 110 130 150
-------
200, and 142 °C, with bottoms temperatures decreasing as flow rate
increased. These temperatures show that nearly all the water con-
tained in the feed was evaporated and that much of the heavier oils in
the sludges were also evaporated. As the feed rate for this run
increased, the total condensate flow rate and aqueous flow rates
increased while the organic condensate flow rate showed an increase
between 69 and 114 lb/h of feed and only a minor change between 114
and 143 lb/h feed. This is explained by examining the bottoms flow
rate shown In Figure 3—5. The bottoms flow rate Increased slightly
between the two lower feed rates but jumped substantially between the
two higher flow rates. There was only a 16° drop In bottoms tempera-
ture between the lowest and middle feed rates, while the bottoms tem-
perature dropped 58° between the middle and highest flows. Conse-
quently, the bottoms at the highest flow rate contain additional oils
that vaporized at the higher temperature achieved in the lower flow
rate runs.
3. The process results occurring at the Intermediate temperature resemble
the results at 150 °C. These tests were at 230 C (runs 14, 15, and
16) and at flow rates between 68 and 112 lb/h. The highest flow rate
obtained was 112 lb/h at the time of the tests, due to feed pump
failure. Figure 3—6 shows the condensate flows for these runs.
Although the condensate flow rate did Increase with Increases in feed
rate, the increases were not proportional to flow rate Increases. The
bottoms temperatures for the four runs were 124, 104, 100, and 100 °C,
with the bottoms temperature decreasing as feed rate Increased. The
bottoms temperature was above 100 C only at the lowest flow rate (68
lb/h). For this run, almost all of the water would be evaporated from
the feed, along with some of the oils boiling between 100 and 124 °C.
As the flow rates increased, there was a higher rate of water feed
Into the IFE and the bottoms temperature dropped to 100 C. The bot-
toms flows increased as feed rate Increased (Figure 3—7), but more
bottoms were produced as a percent of feed as the feed rate was In-
creased, due to more water passing through the TFE and exiting with
the bottoms.
4. The vacuum runs showed that some potential problems will likely occur
if water/organic sludges are processed by TFE’s under vacuum. The
condensates for all of the vacuum runs formed a milky white emulsion
of organics and water. This contrasted substantially with the clean
splitting organic/aqueous layers that occurred during all of the
atmospheric pressure runs. Such an emulsion would be relatively dif-
ficult to separate and would present processing problems. A simple
solution to the formation of these emulsions would be to process the
waste sludge with two passes through TFEs. The first TFE would remove
the water and VO from the waste, producing easily handled condensate
and oil fractions. The second TFE would operate at high temperature
and under vacuum. The oils heavier than water would be removed as a
single condensate, with a relatively low water content. The amount of
bottoms for ultimate disposal would be only a small fraction of the
original waste sludge, and would contain solids and the heaviest oils
of the waste.
3—12
-------
20 -
19 -
18 -
17 -
16 -
15 -
14-
0 %
a:
I
N 12-
11-
i 1 j 10-
9-
7-
6-
5-
4-
3-
2-
1—
0- U I I I U
50 70 90 110 130 150
FEEDRATE (LBs/HR)
Figure 3-5. Bottoms flow rate as a function of feed rate at 320 °C.
-------
80 -
70 -
30-
20 -
10 -
A - .
0-
60 80 100 120 140 160
FEEDRATE (LBS/HR)
+ TOTAL 0 ORGANIC i uEuus
FIgure 3-6. Condensate flow rates as a function of feed rate at 230 °C.
-------
60
FEEDRATE (L9s/HR)
Figuie 3-7. Bottoms flow rate as a function of feed rate at 230 °C.
I
N
U1 :
0
50
40
30
20
10
0
50
70 90 110 130
150
-------
The vacuum run at 320 C (run 11) had substantial carryover of feed
entering the IFE. The feed vaporized so rapidly at this temperature
that raw feed was carried into the condenser and condensate. There
was also substantial loss of volatiles through the condenser vent and
through the vacuum system. The measured feed rate for this run (23.2
lb/h) was determined from the bottoms and condensate collected, so it
is highly suspect. The density of the collected organic layer (0.96
g/mL) shows that ft contained substantial water when it was separated
from the aqueous layer. The listed percentage of oil recovered for
this run refers to this oil/water emulsion, rather than a cleanly
separated oil phase that occurred in the other runs.
This would be a very unrealistic set of conditions to operate the TEE
on this type of waste sludge, although the very large reduction in the
bottoms requiring disposal indicates that a two-stage TEE process
would be appropriate where large reductions in the amount of waste for
disposal are desired. The bottoms from the process were 4.3 percent
of the measured feed rate and were a viscous high carbon oil.
The low-temperature vacuum runs (17 and 18) showed that water and
volatiles could be recovered from the wastes at relatively low tem-
peratures, but the emulsions formed in the condensate would present
some difficulty. The bottoms temperature for these two runs were 55
and 52 C, and the TEE was heated at 150 and 100 C.
5. The second waste tested had a much higher oil content than the first.
The two runs at 150 C were heat transfer limited, with a bottoms
temperature of 100 C. The runs at 340 •C achieved much higher con-
densate rates and bottoms temperatures, removing substantially more
oil and all the water from the waste. These results confirmed the
process data gathered on the first feed and demonstrated the operation
of the TFE on a second waste sample.
Three different methods were used to determine the effectiveness of the
process for the removal of volatiles from the wastes. Two of these involved
analytical methods applied at the test site during testing, and the third used
samples collected and analyzed by a contract laboratory. These three methods
all showed that the TFE process was very effective in the removal of volatiles
from the wastes tested under essentially all of the conditions used for the
tests. The measured percent removal was, In general, greater than 99 percent
for all of the individual VO identified in the waste and also greater than
99 percent for the total VO in the wastes.
A portable GC was used to determine the concentration of VO in the head-
space above half—filled sample bottles containing feed waste and bottoms
product. Because both the feed samples and bottoms contain a high percentage
of oil, the headspace concentrations of VO are proportional to the VO concen-
trations within the samples. By measuring the headspace concentrations (and
the reduction In concentration after processing) the percentage of waste VO
removal can be determined. This also has the advantage of determining the
vapor concentrations above the raw feed and waste bottoms. The peaks were all
detected with a flame—ionization detector (FID) and Integrated with a portable
3—16
-------
GC integrator. The analytical system was calibrated with both a Cl to C6
hydrocarbon standard (methane, ethane, propane, butane, pentane, and hexane)
and a benzene standard. Toluene was injected to determine the retention time
of that compound. Peaks were Identified by both retention time on the GC
column and an analysis of feed vapor headspace by GC/MS conducted later.
Table 3—5 shows the results of this analysis and the calculated percent
reductions In headspace concentrations above the samples. In virtually all
the samples, the reduction In headspace concentration was greater than 99 per-
cent. The only exception to this occurred with the very light hydrocarbon
propane. This is probably because these light hydrocarbons would tend to be
much more in the vapor phase than the heavier hydrocarbons that are liquids at
room temperature.
After the headspace of the samples was analyzed by GC analysis, a total
hydrocarbon analyzer (Bacharach TLV) was used to measure the total hydrocar-
bons in the headspace above these samples. The results of this analysis are
displayed in Table 3-6. Unfortunately, the Bacharach TLV has a maximum range
of 10,000 ppm of hydrocarbons (measured as hexane). This limit was exceeded
by the headspace of the feed samples, but the headspace of the bottoms samples
was well within the range of the instrument. The calculated reductions in
headspace concentration based on these data used 10,000 ppm hexane as the
headspace concentration of volatiles in the feed waste. The actual concentra-
tion was greater than 10,000 ppm, thus the percent reduction figures are
minimum reductions in concentration with the actual headspace concentration
reductions expected to be higher than the percentages listed. Even with this
limitation, the minimum reductions in volatiles were all 99+ percent (except
runs 7 and 11, which were 98.8 and 98 percent, respectively). These results
confirmed the headspace concentrations measured with the GC analysis in
Table 3-5. The results of duplicate and triplicate samples gave very con-
sistent analyses, as did the GC measurements of headspace concentration. It
Is Interesting to note that the headspace concentrations for Runs 5, 6, and 7
(TFE temperature of 150 C) increased with increasing flow rate (80, 98, and
120 ppm, respectively), and runs 8, 9, and 10 showed the reverse relationship
to feed rate (78, 32, and 28 ppm, respectively). The high—temperature results
may be due to some cracking of hydrocarbons at low flow rates at high tempera-
tures. These trends were not observed with the GC measurements of headspace
concentration.
The third measurement of volatile reduction Involved the analysis of feed
and bottoms samples for volatiles in the collected samples. Table 3-7 com-
piles the analytical results for VO in feed, bottoms, and condensate samples.
These samples were from runs 5, 7, 8, and 10. One sample from each feed drum
was analyzed along with the bottoms from each run. Only one sample of aqueous
condensate was analyzed because the concentrations of organic compounds in
this condensate phase were expected to be very small relative to the organic
condensate samples. This was confirmed in the analysis of the sample. The
analysis of feed sample Luwa-188 for volatiles appeared to be in error, per-
haps because the sample split into two phases and the analyst used an aqueous
sample from the feed. The surrogate recovery on this sample was very good,
indicating that any analytical error occurred in the preparation of the sample
rather than the analysis.
3—17
-------
TABLE 3-5 REDUCTION IN HEADSPACE VOLATILE ORGANIC CONCENTRATIONS.
FROM ONSITE GC ANALYSIS OF HEADSPACE
Ave feed
Run $5
bottoms
Run $8
Run $7
bottoms
Run $8
bottoms
Run $9
Run $10
bottom,
Run $11
bottoms
Run $13
Run $14
Run $16
Run $18
Run $17
Run $18
(n=8),
eve,
bottom.,
cv.,
cv.,
bottom.,
cv.,
cv.
bottom.,
bottoms,
bottom.,
bottom.,
bottoms,
bottom.,
pg/L
/4g/L
pg/L
1 U 9 /L
(Sg/L
g/L
1 tL 9 /L
pg/I
g/L
fAg/L
1 Ug/L
pg/I
palL
pg/L
Propane 121 1.6 3.8 2.3 2.8 0.9 1.4 0.9 18.9 14.0 20.7 8.3 11.2 6.9
Butane 1,085 4.4 12.4 7.1 8.3 2.3 2.9 801.. 9.9 4.2 12.2 11.2 8.9 4.0
2-Pjethylbutsne 1,138 0.4 1.3 0.8 0.1 801 BOL 0.8 0.3 801 601 0.7 0.1 601
Pontan. 1,029 0.9 2.5 1.0 1.8 601 BOL 0.5 0.8 801 0.6 1.4 0.5 801..
2-Ilothylpentan. 421 BDL 601. 8DL BDL BDL BOL 0.2 801 601. 601 601 801 801
Benzen. 1.518 2.3 2.6 1.7 6.0 801 80L 1.2 801 BOL 0.8 801 801 801
H .xsn. 432 0 3 601 BDL 0.8 BOL 601. 0.3 801. BDL 601 601 801 BDL
Toluone 1,143 10 8 4.9 6.3 5.1 801. 601 9.4 0.7 BDL 801 BDL BDL 801.
% Reduction in h..dspac. concentrations
(where a component wa, not detected in the bottom. h.sdspace,
a value of 0.2 g/L was used)
Propane 98.72% 98.88% 98.10% 97.89% 99.28% 98.80% 99.26% 88.03% 88.43% 82.89% 93.14% 90.74% 95.12%
Butane 99.69% 98.84% 99.33% 99.41% 99.78% 99.13% 99.98% 99.07% 99.81% 98.85% 98.95% 99.16% 99.62%
2-Methylbutan . 99.97% 99.89% 99.95% 99.99% 99.98% 99.98% 99.95% 99.91% 99.98% 99.98% 99.94% 99.99% 99.98%
Pentan, 99.91% 99.78% 99.90% 99.84% 99.98% 99.98% 99.95% 99.92% 99.98% 99.96% 99.88% 99.95% 99.98%
2-Uethylpont.no 99.95% 99.95% 99.95% 99.95% 99.95% 99.95% 99.95% 99.95% 99.96% 99.96% 99.95% 99.95% 99.95%
Benz.ne 99.85% 99.84% 99.89% 99.87% 99.99% 99.99% 99.92% 99.99% 99.99% 99.95% 99.99% 99.99% 99.99%
Hexan. 99.93% 99.95% 99.95% 99.88% 99.95% 99.95% 99.93% 99.96% 99.95% 99.95% 99.95% 99.96% 99.95%
Toluen. 99.05% 99.57% 99.64% 99.65% 99.98% 99.98% 99.18% 99.94% 99.98% 99.98% 99.98% 99.98% 99.98%
Total Average 99.6% 89.3% 99.6% 99.5% 99.9% 99.8% 99.8% 98.1% 98.6% 97.7% 99.0% 98.7% 99.3%
BOL r Below dat.ct.on Umit (0.1 Lg/L)
-------
TABLE 3-6. REDUCTION IN HEADSPACE VOLATILE
ORGANIC CONCENTRATIONS, BACHARACH TLV
SNIFFER RESULTS
Sample
No.
Run
No.
Headspace
concentration
(ppm hexane)a
% Reduction in
headspace
coricentrationb
LUWA-43 Feed, 3rd drum >10,000
LUWA-44 Feed, 3rd drum >10,000
LUWA-81 Feed, 4th Drum >10,000
LUWA-82 Feed, 4th Drum >10,000
LUWA-1O6 Feed, 5th Drum >10,000
LUWA—107 Feed, 5th Drum >10,000
LUWA-47 5 82
LUWA-65 5 78
LUWA-66 5 81
Run 5 average 80 99.2
LUWA-50 6 98 99.0
LUWA-56 7 120
LUWA-57 7 120
Run 7 average 120 98.8
LUWA-75 8 83
LUWA-76 8 72
Run 8 average 78 99.2
LUWA-87 9 32 99.7
LUWA-93 10 32
LUWA-94 10 23
Run 10 average 28 99.7
LUWA-103 11 200 98.0
LUWA—ilO 12 69 99.3
LUWA-113 13 60 99.4
LUWA-116 14 20 99.8
LUWA-119 15 52 99.5
LUWA-122 16 55 99.5
LUWA-125 17 40 99.6
LUWA-128 18 100 99.0
aMeasurements taken at 25 C.
bThjs calculation used the maximum range of the Bacharach
TLV sniffer, 10,000 ppm, this percent removal is a minimum
percent reduction in headspace volatile organic compounds;
the actual percent reduction would be higher.
3-19
-------
TABLE 3-7 VOLATILE ANALYSIS OF LIQUID AND SLUDGE PROCESS SAMPLES (COMPUCHEM DATA)
Extrpction of 4 g of sampl. (nominal w.ight) with 10 nt. methanol, GC/US..
bollution of s.mples 1:1000 with methanol, GC/MS.
Cpurge and trap of 175 L sample, GC/MS.
dXyI.ne, obscured by high background of other chemicals, *ylenes may be present.
Surrogate measurement not possible due to lsrg. sampl, dilution required the analysis of the prepared sample.
C D
LLJWArun#
Sample number
Sample type
LUWA-186 LUWA—188
feed, feed,
mg/kg mg/kg
5 5
LUWA-68 LUWA-149
bottoms, org cond,
mg/kg mg/kg
7
WWA-69
bottoms,
mg/kg
7
LUWA-156
org cond,
mg/kg
8
LUWA-78
bottoms,
mg/kg
8
LIJWA-164
org cond,
mg/kg
10 10
LUWA-98 LIJWA-172
bottoms org cond,
mg/kg mg/kg
10
LUWA—168
sq cond,
mg/kg
Ssmpl. prep/analytical
a
a
a
b
e
b
a
b
a
b
c
Compounds
Benz.ne
2—H.x.none
Toluens
Ethylbenzene
Styren.
m—Xyl.n.
o,p-Xylene
230
2,800
180
180
280
280
0.2
1
0.175
0.39
d
d
0.97
5.8
0.7
0.84
1.3
1.4
4,500
1,800
85,000
6,600
3,700
7,000
12,000
BOL
6.1
2.1
2.6
3.8
4.4
6,000
120,000
7,800
5,000
9,700
14,600
0.64
4.8
0.58
1.6
0.93
0.9
1,300
25,000
2,000
BDL
2,200
3,800
0.58
2.7
0.39
1.2
0.71
0.72
1,900
34,000
24,000
1,300
2,800
4,900
1.4
4.8
0.18
0.16
0.21
0.2
Surrogat. recovery
D4—1,2-D lchloroet.hane
Broniofluorob.nzene
08—Toluen.
68
102
86
100
95
86
86
79
81
e
e
e
99
78
98
e
.
.
77
69
88
.
e
.
77
89
88
e
•
.
97
97
104
-------
Table 3—8 lists the semivolatile analysis for the same set of samples.
The agreement between the two feed samples is much better for this analysis
and reflects the separate preparation of samples for analysis In the two pro-
cedures. The aqueous condensate again showed relatively low concentrations of
semivolatiles relative to the organic condensates, although several additional
components were identified tn the aqueous condensate that were not identified
elsewhere. A metals analysis of the feeds, one bottoms sample, and one organ-
ic condensate sample was also performed. These results are listed in Table
3-9. This analysis was performed primarily for feed characterization and to
demonstrate where metals entering with the feed appeared. Only small amounts
of toxic metals were detected in the organic condensate, except for arsenic,
which was present at 0.128 ppm.
Tables 3—10, 3—11, 3—12, and 3—13 compile the volatile and semivolatile
results for runs 5, 7, 8, and 10. Percent removal from the feed sludge, per-
cent reduction in concentration between the feed and bottoms, and mass balance
closures are calculated for each component identified. The volatile removal
from the feed sludge was 99.5+ percent for all volatiles In runs 5, 8 and 10.
Runs 8 and 10 showed 99.9+ percent removal for all volatiles measured. The
removal of volatiles for run 7 was slightly lower (98.6+ percent), which is
attributable to the relatively low temperature of operation (150 °C) and high
feed rate (154 lb/h) for this run. The concentrations of heavier semivolatile
compounds were increased in the bottoms for runs 8 and 9. This would be
caused principally by the removal of the volatile material in the feed,
increasing the apparent concentration of these compounds.
The observed mass balance closures are somewhat disconcerting in that
there is a substantial bias toward more material flowing into the process than
Is leaving (runs 5, 8, and 10). The volatile components show better closure
than the semivolatile components. The closures for run 7 are much better,
however, where closure Is within about 25 percent. There are two reasonable
explanations for this. First, the semivolatiles may be thermally changed or
cross—linked as they pass through the TEE into larger, unanalyzable molecules,
showing loss of the components while also preventing closure of the mass
balance. Second, the bottoms may not have extracted as well as the feed sam-
ples, reducing the amount of the components extracted from the bottoms and
producing a low analysis for the bottoms samples. The bottoms from run 7
(which showed the best mass balance closure) resembled the feed substantially.
Very little material was evaporated from the feed during this run, and the
temperature was relatively low (150 C). Its components would have undergone
the least change as a result of processing, and it would be nearly as extract-
able as the original feed. The almost identical percent removal from feed and
mass balance closures for benzo(a)anthracene, chrysene, benzo(b)flouranthene,
benzo(k)flourantherie, and benzo(a)pyrene in runs 5, 7, and 10 show that these
materials are not actually removed from the feed; rather, they are bound to
the bottoms samples to reduce their apparent concentrations in the analyzed
samples.
The flow rate of gases from the condenser vent were also measured during
the testing of the TEE. Table 3-14 lists the measurements of vent gas flow
and concentrations for each of the runs. The vapor hydrocarbon concentrations
above the bottoms samples as they were removed from the TFE are also listed in
3—21
-------
TABLE 3-8 SEMIVOLATILE ANALYSIS OF LIQUID AND SLUDGE PROCESS SAMPLES (COMPUCHEM DATA)
LUWArun$
Sample number LUWA-185 LUWA-188
Sample type feed, feed,
mg/kg mg/kg
6 6
LL!WA-68 LIJWA-149
bottoms, org cond,
mg/kg mg/kg
1
LLJWA-59
bottoms,
mg/kg
7
LUWA-166
org cond,
mg/kg
8
LUWA-78
bottoms 1
mg/kg
8 10
LLJWA-164 LUWA-96
org cond, bottoms,
mg/kg mg/kg
10
LUWA-172
org cond,
mg/kg
10
LUWA-168
sq cond,
mg/kg
Sample prep/analytical
a
a
a
b
a
b
a
b
a
b
c
Compound.
Phenol
benzyl alcohol
2-Uethylphenol
4-Methy lphenol
2 ,4—Dimethy lphenol
Bis(2—.thylhexyl) phthalat.
Naphthalene
2-Methylnaphthsl .ne
Acenaphthyiene
Acenaphthene
Dibenzofursn
Fluorene
n—Nitrosodiphenylamine
Phenanthrene
Anthracene
Pyrene
Benzo(a)anthracene
Chrysene
Di-N—Octyl phthalate
Cenzo(b)fiuoranthene (f)
Benzo(k)fluoranthene (f)
Cenzo(a)pyrene
850
790
CDL
47
31
77
180
17
38
20
28
11
1 1
CDL
880
CDL
21
CDL
43
94
270
23
45
18
38
36
16
16
14
180
320
8.2
15
19
40
120
17
14
9.6
17
8.9
8.9
7.7
71000
5,000
CDL
110
76
126
120
56
CDL
CDL
CDL
CDL
CDL
CDL
CDL
520
880
CDL
CDL
28
64
200
18
24
22
31
13
13
12
5,000
3,000
CDL
CDL
CDL
83
40
CDL
CDL
CDL
CDL
CDL
CDL
CDL
48
120
1 1
CDL
20
82
220
28
64
36
70
44
44
28
5,000
8,000
180
240
240
600
000
70
9 5
CDL
CDL
CDL
CDL
CDL
24
99
14
CDL
30
89
430
430
80
38
88
49
49
31
6,400
6,000
190
300
280
680
760
54
80
CDL
CDL
CDL
CDL
CDL
0.21
ooga
065
0.41
0.17
0.022
1.4
0.74
CDL
CDL
CDL
0.1
0.098
CDL
CDL
CDL
CDL
CDL
CDL
CDL
Surrogate recovery
2-Fiuorophenol
06-Phenol
06-Ni trobenzene
2-Fl uorobenzene
2,4,6—Tribromophenol
D14—Terphenyl
D10-Pyrene
4
4
d
e
d
e
d
e
.
d
e
59
47
87
84
88
93
92
CDL = Celow detectiom limit; thi. limit varies with the sample analyzed, due to differences in sample preparation and dilution
prior to analysis; limits can be found with the results reported in Appendix F.
530 g (nominal) samples extracted with methylene chloride, final extract Volume 8-10 ci, CC/MS.
g oil diluted to 26 i tt with methylene chloride, CC/MS.
c 100 ml. of sample extracted with methylene chloride at pH )11 and pH <2, extract concentrated to 1.0 ci, CC/MS.
dsampie was diluted for analysis, surrogatds below detection limits.
eSurrogates were not added to diluted oil samples.
Co—eiuting isomers.
t)
l ’J
N )
-------
TABLE 3-9. METALS ANALYSIS OF LIQUID AND SLUDGE
PROCESS SAMPLES (COMPUCHEM DATA)
LUWA run #
10
10
Sample number
Method
LUWA—185
LUWA—188
LUWA-96
LUWA-172
Sample type
used
feed,
ppm
feed,
ppm
bottoms,
ppm
org
cond,
ppm
Aluminum ICP 1,470 2,060 1,590 0.034
Arsenic ICP 1.7 1.1 1.2 0.128
Barium Furnace 21 29 33
Calcium ICP 1,360 1,610 2,540 0.173
Chromium ICP 27 28 55
Cobalt ICP 2.9 2 3.4
Copper ICP 48 62 95
Iron ICP 1,370 1,750 2,210 0.065
Lead ICP 58 72 105 0.005
Magnesium ICP 317 352 816
Manganese ICP 17 19 30
Nickel ICP 9.8 14 16
Sodium ICP 349 399 1,200
Vanadium ICP 3.1 3.1 3.5
Zinc ICP 92 111 179 0.013
The following metals were below the detection limits for all
samples:
Antimony ICP
Beryllium ICP
Cadmium ICP
Mercury CV
Potassium AA
Selenium Furnace
Silver ICP
Thallium Furnace
ICP = inductively coupled plasma
Furnace = furnace atomic absorption
AA flame atomic absorption.
3—23
-------
TABLE 3-10. PERFORMANCE OF THIN-FILM EVAPORATOR, VOLATILE A SEMIVOLATILE
BDL = Below detection limits.
8 Feed concentrations used are a composite of the two semivolatilo and volatile analyses.
bCalculated as (1 — bottoms rate x concentration/feed rate x concentration) x 100.
CCalculathd as (feed concentration — bottoms concentration)/feed concentration x 100.
d((Feed_bottems_organic condensat.e)/feed) x 100%.
econdensate sample not analyzed for this run.
Co-eluUng isomers.
COMPOU DS, RIJI #5
1
r )
Run 5
Feed 8
Bottoms
Aqueous
condensate
Organic
condensate
% Removal
from feed ’
%
Reduction in
concentration
from feedc
Mass balance
closured
(% difference)
Feed rates (lb/h)
71.6
56.2
14.63
1.74
Concentrations
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
Benzane
230
0.97
e
4,600
99.87
99.68
52
2-Hexanone
1,800
Toluene
2,800
6.8
85,000
99.84
99.79
28
Ethylbenzeno
180
0.7
8,600
99.70
99.81
12
Styrene
180
0.84
3,700
99.80
99.48
43
m-Xylen.
280
1.3
7,000
99.84
99.54
39
o,p-Xylene
280
1.4
12,000
99.81
99.60
-5
Naphtha lena
766
160
7,000
83.88
79.08
82
2-Methylnaphthalene
790
320
5,000
88.77
59.49
53
Acenaphthyleno
21
8.2
BDL
89.90
60.95
70
Acenaphthene
41
16
110
16.40
68.09
70
Dibenzofuran
37
19
76
80.41
48.66
55
Fluoren.
86.6
40
125
63.93
63.22
80
n—Nitrosod lphenylamin.
120
Phenanthrene
226
120
65
58.88
46.87
58
Anthracene
20
17
BDL
34.47
15.00
34
Pyrene
40.5
14
BOL
73.35
86.43
73
Benzo(a)anthracene
19
9.6
BDL
81.05
49.47
61
Chrysen.
32
17
BDL
69.04
48.88
59
Di—n—octyl phthalat.
17.5
8enzo(b)fluoranthene
13
8.9
BDL
59.09
48.92
59
Benzo(k)fluoranthene’
13
6.9
BDL
59.08
48.92
59
Benzo(a)pyren.
14
7.7
BDL
57.80
45.00
58
-------
TABLE 3-11. PERFORW.i CE OF THIN-FILM EVAPORATOR, VOLATILE
ODL = Below detection limits.
5 Feed concentrations used are a composite of the two semivolatile and volatile analyses.
bCalcuIat as (1 — bottoms rate x concentration/feed rate x concentration) x 100.
Ccalculated as (feed concentration — bottoms concentration)/feed concentration x 100.
d((Feed_bottoms_organic condensate)/feed) x 100%.
°Condensate sample not analyzed for this run.
fuses reported detection limit of 0.62 mg/kg for benzene.
9 Co—eluting Isomers.
N.)
U,
Run #7
Feed 5
Bottoms
Aqueous
condensate
Organic
condensate
% Removal
from feedb
%
Reduction in
concentration
from feedC
Mass balance
closured
(% difference)
Feed Rates (lb/h)
153.7
139
11.93
2.67
Concentrations
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
Benzene
230
BOL
e
8,000
99 .lOf
99 . 73 f
65
2-Hoxanone
Toluene
2,800
6.1
120,000
99.80
99.78
25
Ethylbenzene
180
2.1
7,800
98.94
98.83
24
Styreno
160
2.5
5,000
98.59
98.44
44
m- .Xylene
280
3.8
9,700
88.77
98.64
39
o,p-Xylene
280
4.4
14,600
98.58
98.43
8
Napht.halene
785
520
5,000
38.63
32.03
27
2—Methylnaphthalene
790
880
3,000
24.46
16.48
18
Acenaphthylene
21
BOL
8DI
Acenaphthan.
47
80 1..
BOL
Dlbenzofuran
37
26
GDL
38.46
29.73
38
Fluorene
85.5
54
83
42.88
38.84
42
Phenanthrene
226
200
40
19.81
11.11
19
Ant.hracene
20
18
BOL
18.81
10.00
19
Pyrene
40.5
24
BOL
48.41
40.74
48
Benzo(a)anthracon.
19
22
BDL
—4.72
—15.79
—5
Chrysene
32
31
BDL
12.39
3.13
12
Oi—n—octyl pht.halat.
17.5
Benzo(b)fluoranthene 9
13
13
BOL
9.58
0.00
10
Bonzo(k)fluoranth.ne9
13
13
BDL
9.58
0.00
10
Benzo(a)pyrene
14
12
BDL
22.48
14.29
22
-------
BOL = Below detection limits.
•Feed concentrations used are a composite of
bCalculated as (1 — bottoms rate x concentration/feed rate * concentration) x 100.
Ccalculated as (feed concentration — bottoms concentration)/feed concentration x 100.
d((Feed_bettoms_organic condensate)/feed) x 100%.
°Condensate sample not. analyzed f or this run.
Isomers.
)
TABLE
3-12. PERFORMANCE OF
TIIIN—FILI4 EVAPORATOR, VOLATILE AJ’D SEMIVOLATILE
COMPOU1)S, RUl #8
Aqueous
Organic
Removal
% Reduction in
Mass balance
closured
Run #8
Feeda
Bottoms
condensate
condensate
from feedL
from feedC
(% difference)
Feed rates (lb/h)
88.6
8.5
63.22
6.78
Concentrations
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
Benzene
230
0.84
e
1,300
99.97
99.72
44
2—Hexanone
Toluene
2,800
4.8
25,000
99.98
99.84
12
Ethylbenzene
180
0.58
2,000
99.96
99.88
—10
Styrene
160
1.8
80L
99.88
99.00
100
m-Xy lane
280
0.93
2,200
99.96
99.87
22
o,p-Xylene
280
0.9
3,800
89.96
99.68
34
Naphthalene
785
46
5,0.00
99.26
93.99
35
2-Uethylnaphthalene
790
120
6,000
98.12
84.81
23
Acenaphthylene
21
11
160
93.60
47.62
18
Acenaphtheno
47
BOL
240
49
Dibenzofurar,
37
20
240
93.29
45.95
Fluorene
85.5
62
50.0
91.00
27.49
Phenanthrene
225
220
800
87.87
2.22
Anthracene
20
28
70
82.63
-40.00
53
Pyreno
40.5
54
95
83.45
-33.33
80
Benzo(a)anthr.cone
19
38
BDL
78.49
-89.47
78
Chrysene
32
70
BDL
72.88
—118.75
73
Di—n—octyl phthalate
17.5
Benzo(b)fluoranthene
13
44
BDL
8enzo(k)fluoranthene
13
44
BOL
58.00
68.00
—238.48
—238.46
58
58
Benzo(a)pyrene
14
28
BDL
75.18
—100.00
76
the two semivolatile and volatile analyses.
-------
TABLE 3-13 PERFORMANCE OF THIN-FILM EVAPORATOR, VOLATILE AND SEMIVOLATILE COMPOUNDS, RUN #10
R.,n •iO F..d
bottc ,na
Aq . .oo.
condonsat.
Or inic
cond .na.t..
3 R o al
from foedb
% Rod , ,ct on n
concontratlon
from f 6
lii• balanc.
cleaurod
(3 diff.r.nce)
Feed rat.. (lb/h) 143 4
18 6
1 )1 93
12.97
Conc.ntr.t on.
(m /k 9 )
(r,glkg)
( .g/k 9 )
(ne/kg)
Oenz.n. 230
0 68
1 4
1,900
99 97
99 79
2—Ho asnon.
Tolu.n. 2,806
2 7
4 8
34,000
99 69
99 90
-9
Ethylb.nz.n. 180
0 39
0 16
24,
99.97
99.78
-1097
Styron. 160
1 2
0 18
1,300
99 90
99 26
m—X 1 1.n. 260
0 71
0 21
2,800
99 97
99.76
o ,p-Xyl.n. 200
0 72
0 2
4,900
99.97
99.74
Ph.nol
0 21
B.nzil alcohol
0 098
2-I .thylph.nol
0 66
4- sth 1 lph.nel
0 41
2,4- Oi omUi 3 lph.nol
0 17
Bis(2- .thylhszyl) phthsl.t.
0.022
Naphthsl.n. 766
24
1.4
6,400
99 59
90 80
2-t4sth lnsphth sl.n. 790
99
0 74
6.000
98 37
87.47
30
Acen.phthyl.n. 21
14
801.
190
91 35
33.33
Acsn.phth.n. 47
801
801.
399
Dibanzotur.n 37
30
801.
260
89 49
18 92
43
Fluor.n. 96 6
69
0.1
680
66 69
-4 09
Phsnanthr.n. 226
439
0 096
760
76.21
-91 11
Anthrsc.ns 20
430
801.
64
-178 87
-2060 60
Pyr.n. 40 6
80
801
80
74 38
-97.53
-203
B .nxo(a).nthracen. 19
38
804.
801
74 08
-100 00
81
Chrys.n. 32
86
804.
901.
66 14
-168 76
74
66
Dl—n-octyl phthalst 17 6
bonzo(b)fluor snthon.’ 13
49
901
801.
61 11
276 92
Benzo(k)fliiorsnth.n. 13
49
801
80 1 .
61 11
-278 92
B .ngo(.)pyron. 14
31
801
801.
7) 20
-121.43
61
73
904. o Oslo ,, d .t.ction limits.
F.ed concentrations used or. • coaipo.it. of th. tso smolvoIstIl. and vol. 1 1 1. analyse..
ii (1 — bott.omi rat, a concentration/feed rat., a conc.ntr.tion) a 100.
Cc.lcul.t.d is (feed concsntration - bottom. cnnc.nt . .Lion)/fe.d concentration a 100
condon..t.—.queoo. condon. .t. ./fesd) a 100%.
°Co-,lutsn 9 ,.om.r.
C )
N)
-4
-------
TABLE 3-14. MEASUREMENTS OF VENT GAS FLOW RATE AND BACHARACH TLV
SNIFFER MEASUREMENTS
Run
Date No. Time Location Measurementa Value
9/09/86 1 09:47 S4, Cond vent Total HC >10.000 ppm Hexane
9/09/86 1 Condenser vent Flow rate <10 mL/min
9/09/86 2 10:05 S4, Cond vent Total HC >10,000 ppm Hexane
9/09/86 2 Condenser vent Flow rate <10 mL/min
9/09/86 3 10:40 S4, Cond vent Total HC >10,000 ppm Hexane
9/09/86 3 Condenser vent Flow rate <10 mL/min
9/09/86 5 13:19 S4, Cond vent Total HC >10,000 ppm Ilexane
9/09/86 5 Condenser vent Flow rate 250 mL/min
9/09/86 5 Feed tank headspace Total HC >10,000 ppm Hexane
9/09/86 6 13:56 S4, Cond vent Total HC >10,000 ppm Hexane
9/09/86 6 Condenser vent Flow rate 250 mL/min
9/09/86 7 14:33 S4, Cond vent Total HC >10,000 ppm Hexane
9/09/86 7 Condenser vent Flow rate 200 mL/mln
9/10/86 8 09:15 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 8 Condenser vent Flow rate 15 mL/min
9/10/86 9 11:33 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 9 Condenser vent Flow rate 16 mL/min
9/10/86 10 12:06 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 10 Condenser vent Flow rate <10 mL/min
9/10/86 11 Vacuum run, condenser flow could not be measured
9/10/86 12 16:09 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 12 Condenser vent Flow rate <10 mL/min
9/10/86 13 16:45 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 13 Condenser vent Flow rate 30 mL/min
9/10/86 14 17:35 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 14 Condenser vent Flow rate 30 mL/min
9/10/86 15 18:00 54, Cond vent Total HC >10,000 ppm Hexane
9/10/86 15 Condenser vent Flow rate 24 mL/min
9/10/86 15 18:09 Vapors above bottoms Total HC 110 ppm Hexane
when removed from TFE
(conti nued)
3—28
-------
TABLE 3—14 (continued)
Run
Date No. Time Location Measurement Value
9/10/86 16 18:28 S4, Cond vent Total HC >10,000 ppm Hexane
9/10/86 16 Condenser vent Flow rate <10 mL/min
9/10/86 16 Vapors above bottoms Total HC 200 ppm Hexane
when removed from TEE
9/10/86 17 19:46 Vacuum run, condenser flow could not be measured
9/10/86 17 Vapors above bottoms Total HC 170 ppm Hexane
when removed from TFE
9/10/86 18 20:44 Vacuum run, condenser flow could not be measured
9/10/86 18 Vapors above bottoms Total HC 200 ppm Hexane
when removed from TFE
9/11/86 19 09:40 S4, Cond vent Total HC >10,000 ppm Hexane
9/11/86 19 Condenser vent Flow rate 9 L/mln
9/11/86 19 Vapors above bottoms Total HC 250 ppm Hexane
when removed from TEE
9/11/86 20 10:50 S4, Cond vent Total HC >10,000 ppm Hexane
9/11/86 20 Condenser vent Flow rate 10.3 L/rnin
9/11/86 20 Vapors above bottoms Total HC 200 ppm Hexane
when removed from TFE
9/11/86 22 13:15 S4, Cond vent Total HC >10,000 ppm Hexane
9/11/86 22 Condenser vent Flow rate 2.4 L/mln
9/11/86 22 Vapors above bottoms Total HC 190 ppm Hexane
when removed from TEE
9/11/86 21 13:35 S4, Cond vent Total HC >10,000 ppm Hexane
9/11/86 21 Condenser vent Flow rate 0.75 L/min
9/11/86 21 Vapors above bottoms Total HC 700 ppm Hexane
when removed from TFE
aRC = Hydrocarbon as measured with Bacharach TLV Sniffer.
3-29
-------
this table. All of the condenser vent gas concentrations exceeded the range
of the Bacharach TLV meter used for this measurement. The flow rates from the
condenser were highly dependent upon the waste being processed In the TFE.
The emulsion tank sludge showed only minimal flows from the condenser vent,
with a maximum of 250 mL/niin being recorded during runs 5 and 6. The second
waste sample, however, showed much higher rates of vapor flow from the conden-
ser. Flows of 9 and 10 L/mln were measured during runs 19 and 20. Because
any IFE used for waste processing would probably handle a wide variety of
wastes from the plant environment, and many of these wastes would be expected
to have substantial amounts of light hydrocarbons that would not condense at
process cooling water temperatures, either a glycol—cooled condenser or a two-
stage condenser (first—stage cooling water, second—stage chilled glycol)
should be employed. As an alternative, the condenser vent gas could be routed
to either an incineration system or flare to destroy any uncondensed hydrocar-
bons.
Tables 3-15 and 3-16 show the results from measurements of vent gas con-
centrations. These samples were taken in evacuated canisters and analyzed by
GC and GC/MS. Table 3-15 has the results of GC analysis of these samples,
where the measured peak areas were grouped according to the number of carbon
atoms in each compound (for compounds contaIning 1 through 6 carbon atoms).
GC/MS analyses of one gas sample canister and the headspace of two feed
samples are listed In Table 3—16. The feed headspace analysis was used in
identifying components found in the onsite GC analysis of sample headspace
(Table 3-5).
The results of an oil, water, and solids analysis are listed in Table
3—17. The residue analysis for solids was a slightly modified Method 224G
(Standard Methods for the Analysis of Water and Wastewater). Exactly 25 g of
each sample were placed into weighing dishes, dried at 105 .C for 1 hour,
cooled In a desiccator, weighed, dried at 300 •C for 1 hour, cooled In a
desiccator, weighed, and then ashed at 550 C for 1 hour. The weights after
each drying were recorded, and the residue after drying at each temperature
recorded. This analysis was performed principally to determine the solids of
the samples, but the weight loss at 105 °C corresponds to the loss of water
and volatiles, heavier oils are lost between 105 and 300 C, and the final
residue after ashing represents the inorganic material of the samples.
The oil was analyzed by extraction of the samples with freon, followed by
spectrophotometric measurement of the oil (Method 413.2). Water content of
samples was determined by Carl-Fischer titration. These results do not appear
very good in that the feeds and bottoms show very high water contents and very
high oil contents. It would be impossible to have samples with both 880 to
950 gIL of oil and water contents between 59 and 63 percent. The very high oil
contents measured Indicate that the freon extraction technique was Inappropri-
ate for the analysis of sample oil content. Several valid points can be made
from these data, however. There was very little oil in the aqueous condensate
and very little water in the oil, indicating very clean splits between the oil
and aqueous condensates. The solids and heavy oil contents of the bottoms
from runs 8 and 10 were increased substantially from their feed concentra-
tions.
3-30
-------
TABLE 3-15. VENT GAS CONCENTRATIONS
Run Sample Cl Range C2 Range C3 Range C4 Range C5 Range C6 Range
No. No. (ppm) (ppm) (ppm) (ppm) (ppm) (ppm)
5 LUWA-70 84 3,100 6,900
7 LIJWA-62 340 130 3,800 3,800
7 LUWA-61 470 520 7,700 7,700
8 LUWA=84 210 650 640 6,500 4,200
8 L1J\ !A-85 28 43 100 1,100
10 LUWA-99 92 130 200 4,000 4,200
3—31
-------
( ‘4
NJ
TABLE 3-16. GC/MS ANALYSIS OF GAS SAMPLES, VENT GAS. AND HEADSPACE (lEA)
Compound
Vent gas
run #10
LLJWA-98 ,
JAg/L
Vent gas
run #10
LIJWA-98,
/.4g/L
Average
vent gas
run #10,
/4g/L
% Difference Feed headspace Feed headapace
vent, gas drum #3 drum #4
run #10 LUWA—44, LUWA-81,
/Sg/L
2—Methyl propane
But.n.
2,800
2,800
2,700
7
2,200
1,800
3,800
2-Butene
890
1,400
Cyclopentane
870
1,000
2—Methyl butane
8,000
4,800
5,400
22
4,500
8,300
Pentane
8,800
6,800
7,800
26
6,300
8,300
2-Methy I -1—pentene
1,200
Methyl cyclopentano
2,300
1,700
2,000
30
1,800
1,800
3—Methyl pentane
2,500
1,800
2,150
33
2—Methyl pentane
8,300
4,400
5,350
36
4,300
5,000
Benzene
3,000
2,000
2,600
40
1,800
2,000
Hexan.
3,900
2,500
3,200
44
2,300
2,800
Methyl cyclohexane
1,600
1,000
1,250
40
1,000
3—Methyl hexana
1,400
Toluene
10,000
7,200
8,800
33
7,800
8,000
Ethyl benzene
978
m-Xylene
108
-------
TABLE 3-17 OIL, WATER, AND SOLIDS ANALYSIS OF PROCESS STREAMS FOR RUNS 5, 7,8, AND 10 a
c..d F.ed F..d Run 16 Run •6 Run 16 Run I ’ Run Il Run I? Run 18 Run 110 Run hO Run •IO
drum •3 drum 14 drum 14 bottom. sq cond org cond bottoms sq cond org cond bottoms bottoms sq cond org cond
LU*A-45 LIIWA-83 LIJWA-830 LuWA-89 LUWA-146 LtJWA-160 LUWA-60 LLPWA-162 LUWA-167 IUWA-79 IUwA-97 LUwA-169 LUWA-173
Solid.
R .sldu.
106
C
34 8%
38.8%
34 4%
16.7%
42 8%
88 6%
89
R,sldu.
300
C
2 8%
3.4%
6.7%
2.1%
3.3%
26 9%
18.9%
0.07%
46 1
ResIdu.
660
C
0 1%
(2 1%)
1.8%
(2.8%)
1.8%
(2.3%)
0 9%
(2.84%)
1.6%
(3.04%)
7 3%
(12 44%)
4.6%
(13.14%)
0.04%
0 8
0.0
Oil (g/L)
888
[ 18 8%)
860
(21.96%)
840
(24.7%)
960
(20.11%)
<.02
(0.1%)
730
(23.68%]
(.02
(0.1%)
960
(76 6%)
1000
(81 66%)
810
Water (..eight
%)
71 00%
(81 3%]
69.00%
(74.8%)
69.00%
(73 0%)
83 00%
(77.06%)
0 04% 47.00%
[ 73.28%)
0
02% 8.70%
(20 44%]
1.70%
(6 18%]
0.40
Uumb.rs in block psr.nth...s Sr. tb. vslu.. r.port.d by Ch.vron r.s.srch (a.. App.ndis I)
-------
PROCESS RESIDUALS
Bottoms Sludge
The bottoms sludge from the TFE was very similar to the feed for the low-
temperature runs. The only difference between the feed and bottoms for these
runs was the removal of most of the volatile organics and some of the water
from the original feed. Any disposal methods employed for the original feed
should be successful with the bottoms from the low—temperature runs. The feed
from the high-temperature runs was substantially more viscous and had a much
higher solids content than the original feed. Most of the water, volatile
organics, and oils had been removed by the processing of the feed sludge.
This material could probably be incinerated successfully or disposed of by
other methods currently used for the raw feed sludge. This material may be
somewhat less biodegradable since most of the organics remaining in the bot-
toms are relatively heavy oils.
The bottoms from the high—temperature vacuum run were even more viscous
than the other bottoms and had much more of the oils removed from the bottoms
when compared visually to other samples. In this form, the volume of waste
for disposal was substantially reduced, but it may present some additional
disposal problems because of its tarry and viscous nature.
Organic Condensate
The organic condensate is a potentially useful effluent stream from the
treatment process. The organic condensate from the low—temperature runs was
an organic oil with a very high concentration of volatiles. It was very
clean, had very little solids, and had only a slight color. This organic
condensate could easily be recycled to numerous raw product streams within a
refinery and would be an economic credit to the process operation. The organ-
ic condensates from the high-temperature runs were very similar to product
from lower temperature runs, but contained more of the heavier boiling oils.
It had more of a caramel color, a heavier burned organic odor, but could still
be recycled as a raw product to the refinery operations. The organic conden-
sate Is the most valuable recovered material from the process and its reuse
within the refinery will reduce the overall cost of the waste treatment.
Organic condensates produced during the vacuum runs were In the form of an
organic/water emulsion. The emulsion from the high—temperature vacuum run
(run 11) never completely separated, even when left standing for a substantial
period of time. The emulsions from the low-temperature vacuum runs (runs 17
and 18) separated more substantially, but might require larger separations
tanks In Installed systems. A two—stage system where the water and volatile
organics are removed in the first TFE (at atmospheric pressure) and the heav-
ier organics are removed in a second TFE (under vacuum) would remove the emul-
sion problems while producing two different organic fractions for recycling to
refinery operations.
3-34
-------
Aqueous Condensate
There are several methods available for the treatment of the aqueous con-
densate. When only a relatively small amount of the total water in the feed
waste Is recovered as condensate (as in runs 5, 6, and 7), the aqueous conden-
sate stream could be recycled to the TFE feed stream. This would increase the
Inlet water percentage to the evaporator, and all water would eventually exit
with process bottoms for normal disposal of the bottoms sludge. This would be
appropriate for the removal of volatiles only from the waste sludges, but
would not work when either dewaterlng of the waste or recovery of lower boil-
ing oils from the feed is desired. Another rather easy method of handling the
aqueous condensate would be to send it to an API separator, where it would be
combined with other oil—contaminated water streams for treatment.
COST OF THIN—FILM EVAPORATION
Results from the pilot-scale TEE test were used to select two possible
modes of operation upon which a cost estimation was made.
In the first mode, the evaporator is operated under relatively mild condi-
tions, with a low heating temperature (150 C) and high feed rates. Almost
all of the volatiles and some of the water are removed under these conditions,
but the bottoms produced are only slightly changed from the initial feed.
These conditions correspond to run #7 of the pilot-scale tests and remove the
volatiles from the waste prior to ultimate disposal without requiring the
evaporation of all the water in the sludge. The second possible application
uses a higher drum temperature (310 °C) to evaporate essentially all of the
volatiles and water from the waste and to recover much of the higher boiling
oils. This second mode of operation corresponds to run #10 of the pilot-scale
tests. These conditions substantially reduce the amount of waste sludge for
ultimate disposal and recover much more of the hydrocarbon oils in the ori-
ginal waste. The additional recovered oils are an economic credit to the
process cost, but additional energy is expended in the process and heated oil
is required instead of steam to heat the process.
A wide range of TFE sizes is also possible. The Installed unit size will
be determined by the amount of sludge to be processed at any given site. Two
different sized units are examined here, one with 53.8 ft 2 of heat transfer
area and the other with 258.2 ft 2 of heated area. The cost of these two sizes
of TFEs was obtained from Luwa Corporation along with an estimate of the other
costs associated with TFE operation.
An estimate of the cost of refinery waste treatment using a IFE was made.
The estimate includes the land disposal costs for the TEE bottoms, water
treatment costs for the aqueous condensate, and a credit for the recovered
organic condensate. Table 3—18 presents the results of the cost estimate for
refinery waste treatment based on the different modes of TFE operation. The
cost of TFE treatment is compared with the cost of disposing the same wastes
by landfarmlng ($100/ton).
3—35
-------
TABLE 3—18. TEE COST ESTIMATION
TEE modes of operation
A
B
C
0
TEE size (ft 2 )
53.8
53.8
258.2
258.2
Operating temperature (°C)
150
310
150
310
Feed rate (lb/hr)
2,690
2,690
12,910
12,910
Product flow rates (% of feed)
011
Water
Bottoms
1.7
7.8
90.5
9.0
78.1
12.9
1.7
7.8
90.5
9.0
78.1
12.9
Treatment cost ($/ton)a
116
37
89
25
Difference from landfarming (%)
+16.1
—63.1
—11.4
—74.9
acost includes utilities, labor, maintenance, taxes, Insurance, disposal of
bottoms, and capital recovery. The cost of disposal of bottoms is the major
component of total cost.
3—36
-------
SECTION 4
FIELD TEST RESULTS: STEAM STRIPPING
RESULTS FOR PLANT I
Process Description for Plant I
Plant I produces mono-carbon chlorinated solvents such as methylene
chloride, chloroform, and carbon tetrachloride. A steam stripper is used to
recover solvents and to treat the plant’s wastewater. The major contaminants
that are removed and monitored by the plant include methylene chloride, carbon
tetrachioride, and chloroform with National Pollutant Discharge Elimination
System (NPDES) discharge limits of 50, 55, and 75 parts per billion (ppb),
respectively. Plant analyses showed that the concentrations in the effluent
were generally on the order of 50 to 75 percent of the limits given above.
Concentrations in the feed stream to the steam stripper are quite variable and
range from hundreds of parts per million to saturation of the water phase with
organics.
The wastewater at this plant consists of equipment wash water and rainfall
collected from diked areas around the plant; consequently, the flow rate and
composition of the wastewater is cyclical and dependent on the amount of rain.
Plant personnel indicated that the steam stripper operated roughly 75 percent
of the time with accumulation in storage when the stripper is not operating.
Once the stripper is started, it operates in an essentially continuous mode
until the wastewater in storage has been steam stripped.
A flow schematic of the treatment system is given In Figure 4—1. The
wastewater enters one of the two decanters (each approximately 75,000 L or
20,000 gal) where it is processed as a batch. Sodium hydroxide solution
(caustic) is added to the decanter to adjust the pH to the target of 7, and
flocculants are added to aid in solids removal. The mixture is recirculated
and mixed in the decanter and allowed to settle. The wastewater (upper layer)
Is sent to the stripper feed (or storage) tank (approximately 473,000 L or
125,000 gal). The organic layer (on the bottom) is removed periodically from
the decanter and sent to a surge or collection tank, and solids are removed
periodically with a vacuum truck for disposal. The cycle time of a batch of
wastewater in the decanter is about 1 day. Approximately 15,000 L (4,000 gal)
of sludge is generated each month and Is disposed of offsite, usually by
Incineration.
The steam stripper is started after a sufficient quantity of water has
accumulated in the storage tank. The stripper feed passes through a heat
exchanger for preheating by the effluent from the stripper. The stripper
column Is packed with 2.5—cm (1—inch) saddles and processes about 41.6 L/rnin
(11 gal/mm). The stripper effluent, after cooling by the heat exchanger,
4-1
-------
S8 (F T)
Vent
Figure 4-1. Simplified schematic of sampling points.
(S = sampling point, F = flowrate, T = temperature, 0 = quantity)
S12 (F,T)
Si
Water
In
l\)
Sb (T)
Condensate
S5 (F,T)
S3 (0)
Organics
Si (F,T)
(T)
Sludge
Collection
Tank
Discharge to River
-------
enters one of two open—topped holding tanks (about 19,000 L or 5,000 gal)
where it is analyzed for comparison with the discharge limits. If the analy-
sis Is satisfactory, the water Is pumped to a surge tank where the pH is
adjusted for final discharge to the river under the NPDES permit. If the
analysis Is not satisfactory, the water is treated again, either in the steam
stripper or carbon adsorber (described below).
The overhead vapors from the stripper pass through a condenser cooled with
cooling tower water. The condensate enters a decanter that separates the
heavier organic layer from water. The entire water layer Is returned to the
steam stripper, and the organic layer is drained periodically by the operator
to a small collection tank for recycle back to the process. The col1ection
tank is open-topped and has a layer of water and sludge floating on top of the
organic layer.
The condenser is vented through the decanter to a vent condenser (cooled
with cooling tower water), as shown in Figure 4—1. The vent condenser
receives vapors from the initial water/organics/solids decanters and the steam
stripper condenser/decanter. The initial decanters and storage tank are fixed
roof tanks and have conservation vents that open as necessary to prevent pres-
sure buildup.
The plant also has two liquid phase carbon adsorbers operated in series
with about 3,800 L (1,000 gal) of carbon in each (see the dashed lines in
Figure 4—1). This system is used rarely and was installed as a backup unit in
the event the stripper is down or if the capacity of the stripper is exceeded.
When the carbon adsorber is used, the wastewater Is pumped from the storage
tanks through the adsorbers and into the holding tanks prior to analysis.
The original process schematic was claimed to be confidential information
and was replaced by the more general block diagram in Figure 4-1. Similarly,
the basic design information on the process was provided to EPA and is not
given here because it was claimed to be confidential business information.
VO Removal From Water
A primary focus of the sampling effort was the assessment of VO removal
from water for the steam stripper. The concentrations of Individual VO con-
stituents entering the stripper with the feed and leaving with the stripper
bottoms are summarized for both test days in Tables 4—1 and 4—2. Methylene
chloride and chloroform are the major constituents in the feed. Compounds
detected at lower levels in the feed included chioromethane, carbon tetra—
chloride, trichloroethylene, and 1,1,2-trichloroethane. The concentrations
for both test days are averaged In Table 4-3 and show that an average VO con-
centration In the feed of 5,858 ppm (0.6 percent) is reduced to less than
0.037 ppm in the stripper bottoms. The only compounds detected in the strip-
per bottoms were methylene chloride (7 of 11 runs), chloroform (4 of 11 runs),
and carbon tetrachioride (1 of 11 runs) with detection limits of 0.005 ppm.
The only other compound detected in the feed and bottoms was 1,1-
dichloroethane during Run 2-1, with a feed concentration of 4.7 ppm and a
4-3
-------
TABLE 4-1 STRIPPER FEED (IN) AND BOl IOMS (OUT) CONCENTRATIONS FOR FIRST TEST DAY (ppm)
Compound
Run
1—1
Run
1-2
Run 1-3
Run 1-4
Run
In
1-5
Out
In
Out
In
Out
In
Out
In
Out
Chleromethane
31.1
<0.006
37.2
<0.005
47.8
<0.005
37.8
<0.005
36.7
<0.005
hlet .hylene chloride
6,788
0.0116
4,893
0.017
4,077
0.013
4,760
0.009
4,951
<0.005
CMoroform
2,829
0.009
928
0.009
1,250
0.008
864
0.008
868
(0.005
Carbon tetrachloride
85.2
(0.006
51.2
0 008
67.2
(0.005
40.0
<0.005
48.8
<0 005
Trichloroethylen.
10.7
<0.005
3.71
(0.005
--
(0.006
--
(0.006
<0.005
1,1,2-Trich loroethane
8.0
<0.005
4.98
(0.005
8.17
<0.006
4.35
(0.006
5.0
<0.006
-------
TABLE 4-2 STRIPPER FEED (IN) AND BOl IOMS (OUT) CONCENTRATIONS FOR SECOND TEST DAY (ppm)
Compound
Run 2-1
Run
2-2
Run 2—3
Run 2-4
Run
2-6
Run 2-6
In
Out.
In
Out
In
Out.
In
Out
In
Out.
In
Out
Chlorom.t.han,
26
3
(0 005
32.3
<0.005
29
0
<0 006
26.0
(0.005
33.1
(0.005
21.5
<0 605
I othyl.ni chlorid.
3,539
0 017
5,256
<0 005
3,411
<0.005
3,600
(0.006
4,746
0.007
3,419
0 023
Chloroform
1,316
(0 005
963
<0 006
1,200
<0 006
1,144
<0.005
1,353
(0.006
1,207
<0 005
C,rbon t.tr.chlorld.
66
3
(0 005
46.7
0 005
80
6
(0.005
63 4
(0 006
62 2
<0 005
57.0
<0 006
Tr.chloro.thyl .n .
6
46
(0995
39
<
006
•-
(0 006
6.2
(0 006
3.7
<0 006
1,1 .2-Trlchloro.thsns
--
(0.005
--
(0.005
--
(0 006
--
(0 006
3.9
<0.065
2.8
(0.005
U)
-------
TABLE 4-3. SUMMARY OF FEED AND Sot IUMS CONCENTRATIONS (ppm)
Compound
Average in
Range
Average out Range
Chloromethane
32.6
21.5—47.6
<0.005
—-
Methylene chloride
4,490
3,419-6,788
0.011
<0.005-0.125
Chloroform
1,270
864-2,829
0.006
<0.005-0.009
Carbon tetrachioride
54.8
40-67
0.005
<0.005-0.008
Trichloroethylene
5.6
ND-10.7
<0.005
--
1,1,2—Trichioroethane
5.3
ND-8.2
<0.005
——
Total VO
5,858
<0.037
4-6
-------
bottoms concentration of 0.011 ppm. No 1,1-dichloroethane was detected in any
of the other 10 samples of the feed or bottoms.
The stripper feed and bottoms also were evaluated for extractable organic
compounds. The only extractable compound found was hexachloroethane at a
level of 0.1 ppm In the feed and not detected (<0.01 ppm) In the bottoms.
Because of the expense of analysis for extractable organics relative to the
minor information obtained (i.e., low concentrations), no additional samples
were analyzed for extractable organic content.
Removal efficiencies from water are given in Tables 4—4 and 4—5 for the
two test days. The removal of the two major compounds, methylene chloride and
chloroform, averaged over 99.999 percent. The removal of methylene chloride
ranged from 99.996 to >99.9999 percent, and the removal of chloroform ranged
from 99.998 to >99.9996 percent. The number of nines (9’s) in the removal
efficiency for the other compounds was determined by the feed concentration
and detection limit values for the bottoms. The removal efficiency of these
other compounds was over 99.8 percent. The overall VO removal efficiency,
based on an average feed concentration of 5,858 ppm and an average bottoms
concentration of less than 0.037 ppm, was approximately 99.999 percent.
Air Emissions
The flow rates and concentrations of VO in the vapor phase were determined
at the primary condenser (S8), at the secondary vent condenser (S9), at the
storage (or feed) tank vent (Sb), and the solids decanter vent (S12). Head-
space samples were taken from the open-top tank used as a surge or pumping
tank for the organic condensate (Sib).
The results of the vapor-phase sampling and analysis are given In
Table 4-6. The major component of the vapors from the two condensers (SB and
S9) is methylene chloride at about 40 percent (by volume or total number of
moles), followed by chloroform (roughly 4 percent), chloromethane (0.5 per-
cent), and carbon tetrachioride (0.2 percent). Methylene chloride was also
the dominant compound found in the storage tank vapor (Sb) and In the vapor
from the solids decanter (S12). The lowest vapor-phase concentrations were
found over the open-top collection tank (Sib), which had a layer of water and
sludge floating on top of the organic layer and was exposed to the atmosphere.
The first sample taken from the storage tank vent (Sb) at 11:10 yielded
concentration results that were roughly a factor of 10 lower than the results
from the other five samples. This set of results was evaluated by following
the procedures outlined in Grubbs (Reference 1) to test for outliers. The
statistical test suggested that the results for the sample taken at 11:10 are
outliers at the 99-percent significance level. Similarly, the carbon tetra-
chloride result for the sample from the secondary condenser vent (S9) at 12:54
(0.57 percent) was found to be an outlier at the 99—percent significance
level.
The vapor-phase concentrations from the two condensers are compared in
Table 4-7. The two sets of average concentrations are very similar and are
4-7
-------
TABLE 4-4. REMOVAL EFFICIENCY FROM WATER FOR FIRST TEST DAY (PERCENT)
Compound
Run 1—1
Run 1—2
Run 1—3
Run 1—4
Run 1—6
Average
Chloromet.hane
>99.98
>99.98
>99.988
>99.98
>99.98
>99.98
Hathylene chloride
99.9996
99.9996
99.9998
99.9998
>99.9998
>99.9997
Chloroform
99.9998
99.998
99.9993
99.998
>99.9993
99.999
Carbon tetrachloride
>99.991
99.98
>99.991
>99.98
>99.98
>99.98
Trichloroethylono
>99.96
>99.8
>99.92
1,1 ,2—Trichloroethane
>99.93
>99.8
>99.9
>99.8
>99.8
>99.9
-------
TABLE 4-5. REMOVAL EFFICIENCY FROM WATER FOR SECOND TEST DAY (PERCENT)
Compound
Run 2-1
Run 2-2
Run 2-3
Run 2-4
Run 2—5
Run 2-8
Average
Chloromet.hono
>99.99
>99.98
>99.98
>99.98
>99.98
>99.98
>99.98
Methylone chloride
99.9995
>99.9999
>99.9998
>99.9998
99.9998
99.9992
>99.9997
Chloroform
>99.9998
>99.9994
>99.9998
>99.9995
>99.9998
>99.9998
>99.9998
Carbon tetrachloride
>99.991
>99.98
>99.991
>99.98
>99.98
>99.991
>99.990
Trichloroethylene
>99.90
>99.8
—-
—-
>99.91
>99.8
>99.8
1 9 1,2—Trichloroethane
—-
--
——
—-
>99.8
>99.8
>99.8
0
-------
TABLE 4-6 RESULTS OF VAPOR ANALYSES
Concent.ratlon
Date Time Location Chloromethane Methylene
(volume
percent.
at. 1 stm and 25
C)
chloride
Chloroform
Carbon tet.r.chlorid.
9—24-88 10:54 Primary condanser 0.62 37.8 3.87 0.18
9—24-88 12:32 vent (50) 0.63 39.2 3.99 0.19
9—24-88 12:34 0.62 38.7 3.88 0.17
9—24-88 12:37 063 38.8 3.91 0.18
9-24—86 14:45 0.59 40.1 4.39 0.20
9—25-88 00:50 0.48 40.0 4.42 0.24
9—26—88 11:10 0.60 38.4 4.38 0.27
9—25—88 12:09 0.48 40.4 4.47 0.26
9-24-88 10:40 Secondary condenser 0.69 40.2 4.60 0.23
9-24-88 12:54 vent ($9) 0.69 38.2 4.41 0.5?
9-24-08 12:56 0.58 38.5 4.13 0.32
9-24—88 12:58 0.45 31.8 3.21 0.21
9—24—88 14:59 0.59 40.0 4.28 0.22
9-25—86 09:08 0.49 39.1 4.43 0.23
9-25—88 10:40 0.48 39.9 4.47 0.24
9—26—88 11:49 0.47 41.2 4.71 0.28
9—24—88 11:10 Storag. tank 0.019 1.20 0.31 0.087
9—24—86 13:16 vent (510) 0.19 9.83 2.47 0.71
9-24-86 16:33 0.19 9.89 2.53 0.76
9-25-86 09:43 0.20 12.0 3.12 0.79
9—25—86 11:21 0.19 11.9 3.12 0.79
9-26-88 12:20 0.19 11.8 3.08 0.78
9—24—86 11:05 Organics collect.lon 0.0073 2.44 0.47 0.020
9—24-88 15:27 tank (SIl) 0.0029 1.17 0.23 0.044
9—24—86 10:47 SolIds decanter 1.04 27.7 7.11 2.63
9-24-86 16:10 vent ($12) 0.10 25.9 7.64 3.80
9-26-86 09:34 0.61 32.2 8.43 1.54
-------
TABLE 4-7. COMPARISON OF VAPOR CONCENTRATIONS AT PRIMARY (S8)
AND SECONDARY (S9) CONDENSER VENTS (VOLUME PERCENT)
Compound
Concentration
at S8
Concentration
at S9
RPD
Chloromethane
0.568
0.530
6.9
Methylene chloride
39.2
38.6
1.5
Chloroform
4.16
4.28
2.8
Carbon tetrachioride
0.210
0.286
31
Total VO
44.1
43.7
1.0
= relative percent difference.
4 -il
-------
not significantly different for the major components, considering the preci-
sion of sampling and analysis. The relative percent difference is only 1.5
percent for the major component (methylene chloride) and 2.8 percent for
chloroform. The difference between total VO concentrations was only
1 percent.
The vapor flow rate at S8 was measured by diverting the vapor flow to the
atmosphere at the condenser through a dry test meter. The measured flow rates
are listed In Table 4—8 and reveal an average of about 57 L/min. These flow
rates are representative of a primary condenser vented directly to the atmos-
phere. However, the vapors from the primary condenser travel through piping
to the secondary vent condenser. Consequently, the steam stripper is operated
at a backpressure of about 1.5 pounds per square inch gauge (psig) at the
primary condenser, which Increases condensation and decreases the flow rate
during normal operation. Vapor flow rates measured at the secondary condenser
(S9) are given In Table 4—9. The flow rate at this point ranged from 11 to
13.6 L/min and was measured with a wet test meter. The difference In measured
flow rates at S8 and S9 can be accounted for partially by the change in back-
pressure and In the flow route. However, the difference In flow rates also
may have been caused by undetected leaks In the vent system used for the
secondary condenser. Figure 4—1 shows that the two solids decanters also are
vented to the secondary condenser. Any leaks around these decanter tops or
vents or leaks In the overhead piping may have resulted in lower measured
flows at the secondary condenser (S9). Both test meters had been carefully
calibrated with a standard wet test meter. Consequently, the meter
calibration should not have contributed to any differences between the two
measured flow rates.
The mass flow rates into and from the primary condenser are given in
Table 4—10. The flow rate of organic compounds into the primary condenser
were calculated from the average mass rate entering with the stripper feed
water minus the average rate leaving with the stripper bottoms, which was
essentially negligible in comparison to the feed rate. The mass rate of
compounds leaving with the vapor were calculated from the measured flow rates
and concentrations. The condenser efficiency was on the order of 89 to 94
percent for Individual compounds with an overall efficiency of 91 percent for
total VO. Theoretical calculations indicated that the dew point of the vapors
leaving the primary condenser was 21 C. (The dew point is the temperature at
which drops of liquid begin to form from the vapor.) The dew point
corresponds approximately to the cooling water temperature and ambient
temperature during the 2-day test. The removal efficiency predicted for the
condenser was estimated from vapor—liquid equilibrium data, with the results
listed in Table 4—11. The predicted efficiencies agree well with the measured
efficiencies for the two major compounds, methylene chloride and chloroform.
The agreement was poorer for carbon tetrachioride and chiorornethane. Chloro-
methane Is a very volatile compound and is subject to loss during sampling,
transport, and analysis. If the actual feed concentrations of chioromethane
were higher than the reported values because of low recovery, then the meas-
ured efficiency would be lower and closer to the predicted efficiency.
The concentrations measured at S8 and S9 and dew point calculations indi-
cate that the secondary condenser was not removing VO from the vapor vented
from the primary condenser. Table 4-7 showed no significant difference
4-12
-------
TABLE 4-8. VAPOR FLOWRATE FROM PRIMARY CONDENSER (S8)
Date Time 8 Flowrate (L/mln)b
9—24-86 9:31 55.8
12:03 55.8
16:03 59.5
Average 57.0
9—25-86 10:00 54.9
11:25 56.6
12:26 58.1
Average 56.5
aMeasurements were taken over a 6- to 12-mm
bPe ’b0d.
Corrected to standard conditions.
4-13
-------
TABLE 4-9. FLOWRATE MEASUREMENTS FROM
SECONDARY CONDENSER VENT (S9)
Date Timea Flowrate (L/min)b
9-24—86 9:07 11.0
11:44 13.5
15:53 13.6
Average 12.7
9—25—86 9:27 11.3
10:55 11.1
12:04 11.2
Average 11.2
aMeasurements were taken over an 8- to 10-mm
bPeriod.
Corrected to standard conditions.
4-14
-------
TABLE 4-10. MASS FLOWRATES INTO AND FROM THE PRIMARY
CONDENSER (S8) AND CONDENSER EFFICIENCY
Compound
Flowrates
(g/h)
Efficiency
(percent)
Vapor
in 8
Liquid
outb
Vapor out
Water
1.16 x
i0 5
1.16 x
i0
13
99.99
Chioromethane
75.7
67.1
8.6
88.6
Methylene chloride
Chloroform
1.05 x
2.94 x
i0 4
i0 3
9.42 x
2.78 x
iü
1,050
160
90.0
94.4
Carbon tetrachioride
136
122
14
89.6
Total VO
1.37 x
i0 4
1.24 x
i0 4
1,230
90.9
8 From mass balance around stripper.
bBy difference between inlet and outlet vapor flows.
4-15
-------
TABLE 4-11. COMPARISON OF MEASURED AND PREDICTED
EFFICIENCIES FOR PRIMARY CONDENSER (S8)
Compound
Efficiency
(percent)
Predicted
Measured
Chiorometharie
49.4
88.6
Methylene chloride
91.0
90.0
Chloroform
95.8
94.4
Carbon tetrachioride
97.5
89.6
4- 1
-------
between the concentrations measured at S8 and S9. If material had condensed
at S9, the vapor samples from S9 would have had a higher concentration of
noncondensibles and reduced concentrations of the less volatile compounds.
The calculated dew point for both locations was approximately 21 oct which was
roughly equal to the cooling water temperature. The lowest cooling water
temperature recorded was 17 °C; consequently, the maximum driving force for
heat transfer was 4 C (21-17). Dew point calculations indicate that reducing
vapor temperatures to 17 °C does not Increase condensation significantly.
During the test, the secondary condenser did not provide any significant
Improvement in overall condenser efficiency and recovery of VO. However, the
cooling water temperature during the winter months will be lower than during
this test. The efficiency of both the primary and secondary vent condensers
should improve during the winter months with lower cooling water temperatures.
The plant currently is considering the replacement of the vent condenser that
uses cooling tower water with one that uses refrigerated glycol. If refriger-
ated glycol is used to cool the vapors to 2 •C, the efficiency of the second-
ary condenser will improve significantly. Predicted efficiencies for cooling
to 2 •C reveal levels of 67 percent for inethylene chloride and 83 percent for
chloroform, the two major compounds. The overall removal of total VO by the
secondary condenser would improve from around 0 to 68 percent.
Emissions from the storage tank vent (Sb) and the solids decanter (S12)
were estimated from the measured concentrations of VO in the vapor space. The
conservation vent on the storage tank remained closed during the test and
probably opens only when material is pumped into the tank. Attempts to meas-
ure the vapor flow rate at S12 were not completely successful because the con-
servation vents on the decanter opened as the decanter was being filled.
Sampling modifications had been made to the vent lines, which increased the
normal backpressure on the tanks. Consequently, a portion of the vent flow
escaped through the conservation vent and was not measured. However, the
vapor flow rate from these tanks should equal the volumetric rate at which
liquid is displacing the vapor. The emission estimates for these tanks are
provided in Tables 4—12 and 4-13. The estimates are based on the measured
vapor-phase concentrations and the quantity of water treated for operation
during 50 weeks of the year. If the system is operated only 75 percent of the
time, the estimates in Tables 4-12 and 4—13 would be commensurately lower.
The estimated emissions include only working losses, which are the primary
emissions from such process storage tanks. Breathing losses may be estimated
from API (American Petroleum Institute) storage tank equations; however, these
equations require assumptions on temperature changes and are not included.
Emission estimates are not included for the small collection tank. The rela-
tively low vapor—phase concentrations at Sli and low working losses from the
low rate of condensate generation suggest that those emissions would be low.
The theoretical estimate of emissions from this small tank are difficult to
derive because there is no information on the rate of mass transfer through
the upper layer of sludge and water. Numerous assumptions also would be
required on the average driving force concentration for mass transfer, typical
windspeed, etc.
A summary of the major air emission sources is given in Table 4-14. The
major source is the initial solids decanters followed by the storage tank and
4-17
-------
TABLE 4-12. ESTIMATES OF EMISSIONS FROM STORAGE TANK (Sb)
Emissions
Compound g/ha Mg/yrà gIL treatedb
Chloromethane 9.0 0.08 0.004
Methylene chloride 890 7.48 0.356
Chloroform 325 2.73 0.130
Carbon tetrachioride 112 0.94 0.045
Total VO 1 336 11.2 0.535
aBased on operation of 50 weeks per year.
bGrams per liter of water treated.
4-18
-------
TABLE 4-la ESTIMATES OF EMISSIONS FROM SOLIDS DECANTERS (S12)
Emi ssions
Compound c/ha Mg/yrd gIL treatedb
Chloroinethane 43.4 0.36 0.017
Methylene chloride 2,709 22.8 1.08
Chloroform 940 7.90 0.375
Carbon tetrachioride 449 3.78 0.180
Total VO 4,142 34.8 1.658
aBased on operation of 50 weeks per year.
bGrams per liter of water treated.
4- 19
-------
Q
TABLE 4-14 SUMMARY OF VAPOR EMISSIONS
Emission. (g il treated)
Compound Steam stripper Solids decanters Storage tank Total
Chioromethane 0.0034 0.0174 0.0036 0.024
Met.hylene chlorIde 0.420 1.084 0.356 1.86
Chloroform 0.0855 0.376 0.130 0.572
Carbon tetrachloride 0.0058 0.180 0.0448 0.230
Total VO 0.494 1.68 0.635 2.89
-------
the overhead system for the steam stripper. The results in Table 4-14 are
expressed in terms of grams of emissions per liter of water treated, which is
Independent of the time of operation. These results suggest that 2.7 gIL of
VO in the wastewater eventually may be transferred to the air. Based on dew
point calculations, these emissions can be reduced to about 1.1 gIL if the
vent condenser is cooled with refrigerated glycol at about 2 ‘C. The modified
condenser would reduce emissions from the solids decanter, storage tank, and
the steam stripper’s overhead system.
RESULTS FOR PLANT H
Process Description for Plant H
Plant H produces 1,2-dichioroethane (ethylene dichioride or EDC) and vinyl
chloride monomer. The following process description is taken from Reigel’s
Handbook of Industrial Chemicals.’
EDC is produced from the chlorination of ethylene, and about 87 percent of
the nationwide production of EDC Is used In the manufacture of vinyl chloride
monomer. The vinyl chloride Is used In the manufacture of polyvinyl chloride
(PVC) polymer.
EDC is produced at this plant by the oxychlorination reaction shown below:
C 2 H 4 + 2HC1 + 1/2 02 • C 2 H 4 C1 2 + H 2 O
The oxychlorinatlon reaction is carried out In a fluid bed of copper chloride-
Impregnated catalyst. Purified EDC from this process Is cracked In a furnace
at temperatures of about 400 ‘C and elevated pressures. The hot gases are
quenched and distilled to remove HC1 and then the vinyl chloride. HC1 is
returned to the EDC reactor, and unconverted EDC is returned to the EDC puri-
fication process.
Wastewaters from the EDC/VCM production operation and from other parts of
the plant, including storm water runoff, are treated in the steam stripper
shown schematically in Figure 4—2. Miscellaneous plant wastewater is collect-
ed In an equalization tank with a retention time of roughly 24 hours. The
separate, continuous wastewater stream from the process Is added to the feed
line from the equalization tank, and the two combined streams represent the
feed to the stripper. The wastewater from the equalization tank is passed
through a heat exchanger to provide preheating of the stripper feed. Heat to
the heat exchanger Is provided by the stripper effluent. The feed rate to the
stripper is in the range of 760 to 950 L/min (200 to 250 gal 1mm).
The feed stream enters the stripper at the top of the column and steam at
50 psig is injected countercurrently at the bottom of the stripping column.
Trays (instead of packing) are used Inside the column to promote liquid/vapor
contact. Overhead vapors from the column pass through a condenser cooled with
cooling tower water. Condensate is collected In a decanter/receiver vessel.
Vapors not condensed in the first condenser pass through a second condenser
cooled with refrigerated glycol. Condensate from this condenser is also col-
lected in the decanter/receiver vessel. Noncondensibles that pass through
4-21
-------
Figure 4-2. Schematic of steam stripper and sampling locations.
(S = sampling point, T = temperature, P = pressure, F = fiowrate)
Vapors to
Incinerator
1iscellaneous
Plant
Was tewater
To Wastewater
Treatment
S2 (T,F,pH)
Overhead
Vapors
Process
Water
(T)
Si
(T,F,pH)
S5 (T)
Condenser Condenser
Heat
N)
Exchanger
s (T,F)
Stripper
Decanter
Vent
Steam
(P,P)
Not
Used
S3 (T,F,pII)
Recycled
Effluent (T)
Condensate to
Process
-------
both condensers are collected and vented to an incinerator. The column is
maintained at about 4 psig; consequently, the vapor streams described above
are maintained at a positive pressure.
The receiving vessel for condensate was previously operated as a decanter.
However, both the organic layer and aqueous condensate are currently recycled
to the process. This current method of operation does not use the vessel as
the decanter. When the vessel is filled to an upper limit switch, the conden-
sate pump comes on, lowers the level to a lower limit switch, and shuts off.
The pump cycles on and off depending upon the rate at which condensate is
collected. The effluent from the steam stripper first passes through a heat
exchanger to obtain some cooling and to preheat the stripper feed stream.
Additional cooling of the effluent is provided in a second heat exchanger (not
shown in Figure 4-2) prior to sending it to the wastewater treatment system.
Details on the design of the steam stripper were classified as confiden-
tial business information and are not Included in this report.
Removal of VO From Water
Samples of the stripper feed and bottoms (water leaving the bottom of the
steam stripper) were taken for analysis of volatile organics and extractable
organics. The analysis of the pretest survey samples indicated that the over-
whelming majority of organic compounds in the wastewater were volatile organ-
ics. The only extractable organic found was hexachlorobenzene with a concen-
tration of 0.8 ppm entering the stripper and 0.12 ppm leaving with the bot-
toms. During the test, samples were taken for analysis of extractable organ-
ics but only one sample was analyzed Initially. This sample confirmed the
presurvey results that hexachlorobenzene was the only extractable organic
present and revealed a level of 0.044 ppm entering the stripper and less than
0.01 ppm leaving with the bottoms. Because of the high expense of analysis
for extractable organics relative to the minimal amount of Information obtain-
ed (i.e., no significant levels of extractable organics), the other extract-
able organic samples were not analyzed.
The results for volatile organics analysis are presented in Tables 4-15
and 4—16 and show that 14 constituents were identified and quantified. The
results are numbered as Run 1 (first day) and Run 2 (second day) followed by
Numbers 1 through 5 to identify the sequence in which they were taken during
the day. The results in Tables 4—15 and 4—16 show that 1,2—dichloroethane and
chloroform are the major components. The results for 1,2-dichloroethane show
a 4 to 5 order—of—magnitude reduction In concentration after the wastewater is
steam stripped. The chloroform results are quite variable and show very high
reductions for most of the first day of the test and smaller reductions for
the second day of the test. The other 12 constituents were present at levels
of roughly 1 to 20 ppm In the feed, and all were consistently reduced to below
detection limits (0.01 ppm) in the stripper bottoms.
The total VO concentration in the feed over the test averaged approxi-
mately 6,000 ppm (0.6 percent) with a range of 4,400 to 8,200 ppm. The total
VO concentration in the stripper bottoms was essentially determined by the
amount of chloroform remaining in the wastewater. The overall average VO
4-23
-------
TABLE 4-15. STRIPPER FEED (IN) AND BOHOMS (OUT) CONCENTRATIONS FOR
FIRST TEST DAY (ppm)
1—1 1-2 1-3 1—4 1-5
Compound I i i Out In Out In Out I i i Out In Out
1 .2-Dich1oIo lhdne 7.100 06 7.600 25 6,200 14 4,600 .20 5,100 11
Chloroform 370 29 440 13 220 20 240 60 310 63
Benzene 32 < 01 .36 < 01 26 < 01 < 20 < 01 < 50 < 01
Carbon tetrachioride 6.1 < 01 < 20 < 01 2 9 < 01 1.8 <.01 1 5 <.01
Ch lorobeiizene 64 < 01 56 < 01 50 <.01 25 < 01 <.50 <.01
Chioroethane 17 < 01 26 < 01 8 8 < 01 5.3 <.01 6 4 < 01
1,1—Dichloroethane 18 < 01 20 <.01 9.8 <.01 7 8 <.01 8 6 < 0]
l,1—D lchloroethene 11 <.01 16 < 01 2.5 <.01 2.1 <.01 2.4 <.01
1,2—Dichloroethene 18 <.01 20 < 01 7 1 <.01 5 2 < 01 6 2 <.01
Methylene chloride 1.5 < 01 1 6 < 01 1 0 < 01 1 3 <.01 80 < 01
Tetrachloroethene 4.8 <.01 4.4 <.01 1 3 <.01 75 <.01 70 < 01
1,1 .2—Trichloroethane 15 < 01 14 < 01 6 2 < 01 6 0 <.01 3.9 < 01
Trichloroethene 12 < 01 13 < 01 5 1 < 01 2 9 <.01 2.8 <.01
Vinyl chloride 11 < 01 24 <.01 6 9 < 01 7 0 < 01 6.5 < 01
-------
TABLE 4-16 STRIPPER FEED (IN) AND BOTTOMS (OUT) CONCENTRATIONS FOR
SECOND TEST DAY (ppm)
2-1 2—2 2-3 2—4 2-5
Compound In Out In Out In Out In Out In Out
1 .2-Dich loroethane 4.600 .07 5,900 .034 5.500 .017 4.200 042 5.500 .043
Chloroform 180 41 290 11 240 16 160 2 0 260 36
Benz ne < 20 < 01 22 < 01 .22 <.01 <.20 <.01 <.20 <.01
Carbon tetrachioride .82 < 01 1 1 < 01 1.1 <.01 < 20 <.01 1 0 <.01
Ch lorobenzene 36 <.01 30 < 01 30 < 01 < 20 <.01 <.20 < 01
r’ . Chioroethane 7.0 < 01 7 6 < 01 6 7 < 01 4 6 < 01 6 6 < 01
1,1-Dichloroethane 13 <.01 9 3 < 01 8 8 < 01 5 5 <.01 7 3 < 01
1,1-Dichioroethene 3 0 < 01 2.9 < 01 2 5 <.01 1 4 < 01 2 7 < 01
1.2-Dichloroethene 7 2 < 01 7.6 < 01 6 8 <.01 4 7 < 01 6 5 <.01
Methylene chloride .98 < 01 1 1 <01 1 1 <.01 .86 <.01 1.8 < 01
Tetrachloroethene 74 < 01 44 < 01 .12 < 01 20 <.01 40 < 01
1,1,2-Trichioroethane 5 5 < 01 7 2 < 01 5 4 < 01 6 0 < 01 5 7 <.01
Trichloroetherie 2 8 < 01 2 9 < 01 2 6 < 01 1.4 01 2 2 < 01
Vinyl chloride 5 2 < 01 6 2 < 01 5 7 < 01 4 3 < 01 6 7 <.01
-------
concentration leaving the stripper was 9.7 ppm with a range of 0.34 to 36 ppm.
However, the VO concentration in the bottoms for 6 of the 10 runs averaged 0.8
ppm and showed that the stripper was capable of a 4 order-of—magnItude reduc-
tion in total VO concentration. The swings In chloroform concentration are
consistent with the plant’s data, which showed occasionally high levels of
ch 10 ro form.
The removal efficiencies from water are given in Tables 4-17 and 4-18.
The removal of the major component (1,2-dichioroethane) consistently exceeded
99.99 percent and was generally on the order of 99.999 percent. The removal
of all other constituents except chloroform was greater than 95 percent and in
many cases exceeded 99.9 percent. Only 1,2-dichioroethane and chloroform were
detected in the bottoms; consequently, the removal of the other 12 components
is based on the detection limits of these compounds In the bottoms stream.
The removal efficiency for total VO was essentially determined by the amount
of 1,2—dichioroethane and chloroform In the influent and the amount of
chloroform in the effluent. An average VO concentration of 6,000 ppm in the
Influent was reduced to an average of 9.7 ppm in the effluent and represents a
removal efficiency of approximately 99.8 percent for total VO.
The removal of chloroform averaged 99.6 percent for 6 of the 10 runs and
averaged only 92.4 percent for the other 4 runs. The process data were
examined closely, and no positive correlation could be made between chloroform
removal and the steam rate, feed temperature, condensate rate, feed concentra-
tion of chloroform, pH variation, or the feed rate. However, the removal
efficiency for chloroform does appear to be related to fouling of the trays in
the column. For example, Table 4-19 shows that the system was backflushed
between the samples taken for runs 1—1 and 1—2. After backflushing, the
removal of chloroform markedly improved for the balance of the first test day.
Chloroform removal deteriorated again during the second test day. At about
13:00, the condensate showed the presence of solids and indicated that the
steam stripper’s overhead system had been flooded with liquid. This observa-
tion was confirmed by a sharp increase in the reported flow rate of conden-
sate. Flooding can be caused by excessive liquid holdup due to tray fouling,
and the flooding could also provide some unintentional backflushing of the
trays. Note that after the flooding was controlled, the sample taken for Run
2-4 at 13:30 shows a temporary drop in chloroform concentration. During the
second test day, the column pressure drop Increased slightly and plant person-
nel stated that they intended to shut the column down for cleaning.
The observations given above suggest that the chloroform removal efficien-
cy Is probably affected by tray fouling. However, column fouling can be
experienced by any steam stripping operation and periodic cleaning is re-
quired. Subsequent conversations with a representative of EPA’s Office of
Water indicated that chloral (C 2 HC1 3 O, trichloroacetaldehyde) is formed in the
production process and can decompose in the column to form chloroform.
Removal efficiency for chloroform is supposedly increased under these condi-
tions if the pH Is maintained consistently above u. 3
The headspace measurements in Table 4-20 show the effect of high and lower
removal efficiencies of chloroform on potential air emissions from the
stripper bottoms. The feed to the stripper for the two samples have VO
4-26
-------
TABLE 4.17 REMOVAL EFFICIENCIES FROM WATER FOR FIRST TEST DAY (PERCENT)
Compound 1-I 1-2 1-3 1-4 1—5 Average
1,2-Dichloroethane 99 9991 99.998 99.998 99 995 99.998 99.998
Chloroform 91.7 99.98 99.8 99.7 99.8 98.2
Beazene >96 7 >98 1 >96 5 >97 . 1
Carbon tetrdChloride >99.8 >99.7 >99 4 >99.3 >99.6
Chlorobenzene >98 4 >98 8 >98 2 >95.7 >97 8
Chloroethane >99 94 >99 98 >99.90 >99.8 >99.8 >99.9
f
1,1-Dichioroethane >99.94 >99.97 >99 91 >99 86 >99.88 >99.9
1.1—Dich loroethene >99 91 >99 96 >99 6 >99.5 >99.6 >99 7
1,2-Dich loroethene >99 94 >99.97 >99 87 >99.8 >99.8 >99.9
Methylene chloride >99.3 >99 96 >99.1 >99 2 >98 6 >99 2
Tetrachioroethene >99.8 >99 85 >99 3 >98 5 >98 5 >99.2
1,1,2—Trich loroetlian >99 93 >99 95 >99.85 >99 8 >99.7 >99 8
Trichioroethene >99.91 >99.95 >99 8 >99 6 >99.6 >99 8
Vinyl chloride >99.91 >99 97 >99 86 >99 85 >99.84 >99 9
-------
TABLE 4-18 REMOVAL EFFICIENCIES FROM WATER FOR SECOND TEST DAY (PERCENT)
Compound 2-1 2-2 2-3 2-4 2-5 Average
1,2-Dichioroethane 99.998 99.9994 99.9997 99.9990 99.9993 99.9991
Chloroform 99.8 96 3 93 2 98.5 88.4 95.2
Benzene >95 5 >95 5 >95.5
Carbon tetrachloride >98 8 >99.1 >99 1 >99 2 >99.1
Ch lorobenzene >97 1 >96.6 >96.7 >96 8
Chioroethane >99.85 >99 86 >99.85 >99 7 >99 87 >99 8
11-Dichloroethane >99.92 >99.89 >99 89 >99.78 >99.88 >99.87
1,1-Dichloroethene >99.7 >99.7 >99.6 >99.1 >99.7 >99.6
1,2-Dichloroethene >99 85 >99 86 >99.86 >99 7 >99.87 >99.8
Methylene chloride >98 9 >99.1 >99.1 >98.6 >99.5 >99.0
Tetrachioroethene >98.6 >97.7 >97.7 >93.9 >97.9 >97.2
11,2—Trichloroethdne >99.80 >99.86 >99.82 >99.80 >99.85 >99.8
Trichloroethene >99 6 >99 7 >99.6 >99 1 >99.6 >99.5
Vinyl chlorIde >99 8 >99.8 >99.8 >99 7 >99.9 >99 8
-------
TABLE 4-19 RESULTS FOR CHLOROFORM
Time
Run
Number
Conc
Chloroform
entrcition
(ppm)
Removal
Efficiency
(%)
Comments
In
Out
7,20
1-1
370
29
91 7
The system was backflushed
9: 00.
around
9 20
1-2
440
0.13
99 98
11:20
1-3
220
0 20
99 8
Operation normal per plant.
14,00
1-4
240
0.60
99 7
Operation normal per plant
15:30
1-5
310
0.63
99.8
Operation normal per plant.
8.00
2- 1
180
0 41
99 8
a
930
2-2
290
11
96.3
a
11.30
2-3
240
16
932
a
13:30
2-4
160
2 0
98.5
The overhead system flooded
13:00
at
15 30
2-5
260
36
88.4
a
aD ii img t1u- second test day, the pressure drop across the column appenred to iii.rease slightly throughout
the day and suggests column fouling Problems The plant planned to shut the steam stripper down upon
completion of our test to “tiydroblast” (high pressure water c1eaiiin ) the column dud trays
-------
TABLE 4-20. HEADSPACE RESULTS FOR FEED AND But IUMS (mgIL at 25°C)
Constituent
Headspace
Concentration of
Feed
Headspace
Concentration of
Bottoms
Day
1 a
Day
2 b
Day
1 C
1 d
1,2—Dichioroethane 120 34 <0.05 <0.05
Chloroform 12 12 2.5 <0 05
Benzene <0.05 <0.05 <0.05 <0.05
Carbon tetrachioride 5.3 .26 <0.05 <0.05
Ch lorobenzene .072 <0.05 <0.05 <0.05
Chioroethane 3 9 .88 <0 05 <0.05
1.1-Dichioroethane 3 3 .99 <0.05 <0.05
l,1-DichLoroethene 3.5 .46 <0 05 <0.05
1,2—Dichloroethene 3 0 .72 <0 05 <0.05
Methylene chloride 08 .11 <0 05 <0 05
Tetrachioroethene 1.1 .06 <0 05 <0 05
1,1,2-Trichloroethane .77 18 <0 05 <0 05
Trichioroethene 3 2 .27 0.05 <0.05
Vinyl chloride 3.4 1.3 <0.05 <0.05
Totals 160 51 <3 2 <0.7
acorresponding in time to Run 1— I
bcorresponding in time to Run 2-3.
Ccorrespondjng in time to Run 1-1 when the chloroform concentration in the
bottoms was 29 ppm.
dCorresponding in time to Run 1-2 when the chloroform concentration in th
bottoms was .13 ppm.
4-30
-------
concentrations in the vapor that ranged from 51 to 160 mg/L. When the
chloroform removal efficiency was 99.98 percent (Run 1-2), none of the
compounds were detected in the vapor phase of the stripper bottoms at a
detection limit of 0.05 mg/L. For Run 1-1, the chloroform concentration in
the bottoms was 29 ppm (removal efficiency of 91.7 percent) and the only
compound detected In the vapor phase was chloroform at 2.5 mgIL. Both samples
of bottoms show a significant reduction in headspace concentration; however,
when the stripper was performing well, none of the volatile compounds were
detected in the headspace of the stripper bottoms.
Condenser Efficiency
Condenser efficiency was evaluated from the organic loading (organics
entering the condenser with the vapor) and the quantity of organics leaving
through the condenser vent. The difference between the mass rates of organics
entering with the feed and the mass rates of organics leaving the stripper
with the bottoms represents the organic loading on the condenser. These mass
rates are sumarized in Table 4-21 for all constituents that were detected in
the vented vapor. The 1,2-dichioroethane was by far the major organic consti-
tuent entering the condenser.
The mass rate of organics leaving the condenser vent were determined from
the measurement of the vent flow rate and concentration. The mass rates
leaving the condenser vent are summarized in Table 4-22. The condenser effi-
ciency for the major component (1,2-dichloroethane) was consistently above 99
percent. However, as the vapor phase concentration decreases and the vola-
tility of individual constituents increases, the condenser efficiency drops.
Solubility of the vapor constituents in the condensate may also affect conden-
ser efficiency. The overall mass rates out of the condenser vent average
about 20 Mg/yr of VO for this system. In this installation, these vented VO
are destroyed in an incinerator. However, the results suggest that similar
steam strippers with the condenser vented to the atmosphere could have air
emissions of 12 to 51 Mg/yr. These rates represent emissions from the
secondary condenser cooled with glycol at about 2 °C. The emission rates
would be expected to be higher for condensers cooled only with cooling tower
water at ambient temperatures (e.g., 29 °C).
The overall condenser efficiency for total VO is high because the removal
is dominated by the high loading of a single constituent (1,2-dichioroethane).
An average VO loading of 68 g/s Is reduced to an average vent rate of 0.63 g/s
and represents a VO control efficiency of 99.1 percent.
Process Costs and Limitations
The company supplied cost information for the steam stripping operation to
permit an evaluation of cost effectiveness. The basic equipment Includes the
feed storage tank and surge tank, heat exchanger, the column and trays, two
condensers in series (one for cooling tower water followed by a refrigerated
glycol condenser), decanter, 8 pumps (4 @ 25 Hp, 2 @ 40 Hp, 2 @ 7.5 Hp), in-
strumentation (see Table 4-23), pipIng, and insulation. The installed capital
cost in 1986 dollars provided by the company is $950,000. The major annual
operating cost components are listed in Table 4-24 and include utilities
4-31
-------
TABLE 4-21. ORGANIC LOADING ON THE CONDENSERa
(gb)
Compound
Run
1-1V
Run
l-2V
Run
1 -3V
Run
2- IV
Run
2-2V
Run
2-3V
Vinyl chloride
.18
0.10
.092
.075
.050
.079
Chioroethane
.20
0.10
.091
.10
054
.078
1,1-Dichloroethene
12
0.034
.034
.043
016
.032
1,1—Dichioroethane
.15
.13
.12
19
0G4
.086
1,2-Dichioroethene
.15
091
088
.10
.055
.032
Chloroform
3.3
3 3
4 4
2.6
1.9
2.7
1,2-Dichioroethane
57
79
72
66
49
65
Totals
61
83
77
69
51
68
aDetermined by the difference in mass rates entering the stripper with the
feed and leaving with the effluent. Averaged over time intervals that
correspond approximately to the time that the vapor samples were taken
4-32
-------
TABLE 4-22 CONDENSER VENT RATES (g/s) AND CONDENSER REMOVAL EFF!CIENCY (percent)a
l-3V
Ccxrc i i nd
laj 2-1V
ti 2-2V
kin 2-3V
Average
-— R iye
Rate
Efficieicy Rate Eff iciii cy Rate Efficiaicj
Rate
Etficiacy
Rate Effici cy
Vinyl chloride 018 15 12 0 065 0 012 89 084 6 (0—15)
Chicroathane 032 65 068 32 035 35 036 54 043 47 (32-65)
1,I—O ichloroetts ie 016 53 058 0 020 0 030 60 031 15 (0-53)
1,1—O ich loroathane 0073 94 021 89 Ofl 83 011 87 013 88 (83—94)
1,2—D chloroetIe e < 0054 ‘94 0 17 83 008 85 0088 13 0098 84 (73—94)
Chloroform 065 98 5 17 93 10 95 10 96 11 96 (93-99;
1,2—Oichloroethane 17 998 54 992 40 992 26 996 34 995 (99 2—99 8)
Totals (gls) 31 99 064 52 63
( Jyr) 12 31 20 16 20
ased c ii tt propane tracer results
-------
TABLE 4-2a PROCESS INSTRUMENTATION
Steam flow rate
Bottoms level indicator
Column differential pressure
Decanter level indicator
Overhead pressure
Feed pH
Storage tank pH
Feed flow rate
Storage tank level
Surge tank pH and pressure
Column temperatures
Heat exchanger temperatures
4-. 34
-------
TABLE 4-24. COST ESTIMATE FOR THE STEAM STRIPPER
Total installed capital cost (1986 dollars)
$950,000
Annual operating cost
$250,000/yr
Components ot annual cost (stated by company to be in 1986 dollars)
Utilities (steam, water, electricity)
Operating labor
Maintenance labor
Laboratory support
Recovery credit for organicsa
$/yr
250,000
155,000
405,000
203
223
3.38
0.81
0.89
$225 .000/yr
1.100 h/yr
14,000 h/yr
$ 85,000/yr
$400,000/yr
bCapLtal recovery factor = 0.163 (10 years at 10 percent).
CBased on treatment of 1.2 x 108 gal/yr.
Total annualized cost estimate
Annual operating cost
Capital recoveryb
Total annualized cost
Cost effectiveness (S/ton VO removed)
(S/Mg VO removed)
Cost effectivenessC ($/1,000 gal treated)
(S/ton treated)
(S/Mg treated)
a 4 million lbs/yr at $0 1/lb.
4-35
-------
(primarily steam), operating and maintenance labor, laboratory support for
analyses, and a recovery credit for organics (primarily 1,2—dichioroethane)
that Is recycled to the production process. The annual operating cost was
estimated by the facility as $250,000 per year or about $2.08 per 1,000 gal-
lons of wastewater treated. A total annualized cost of $405,000 was derived
from the Information supplied by the company based on a 10—year lifetime and
an Interest rate of 10 percent.
Historical data supplied by the company indicated that about 4 million
pounds (1,820 Mg) of organics were recovered from the wastewater annually.
The cost-effectiveness for removal from water is about $220 per Mg of organics
removed. This steam stripper Is operated with solids in the feed with levels
over 1 percent for dissolved solids and about 0.1 percent for suspended
solids. The major operational problems that are experienced Include fouling
and plugging of the column’s trays and the heat exchanger. The annual down-
time Is about 350 to 450 hours per year or about 4 to 5 percent of the total
utilization time. The major limitation in the steam stripper’s operation is
the restriction of flow and heat exchange because of fouling from the accumu-
lation of solids. The pH control on the system is also difficult to maintain.
The pH of the stripper feed ranged from 4.9 to 8.7 and the pH of the effluent
ranged from 5.6 to 9.3. It is not known how the variations in pH affect the
stripper’s performance. These limitations and other operational problems are
probably reflected in the estimate of maintenance labor required annually
(14,000 hours).
4-36
-------
SECTION 5
FIELD TEST RESULTS: STEAM STRIPPING/CARBON ADSORPTION
SITE AND PROCESS DESCRIPTION FOR PLANT G
Plant G is an explosives manufacturing plant; the wastewater streams that
are produced are predominantly water soluble. The two major waste streams are
redwater and whitewater. The waste streams pass through decanters where the
oils are separated from the aqueous phase. A surface impoundment (lagoon) is
used as a large storage vessel to provide a stable flow to the stream strip-
ping unit. The steam stripper removes organic compounds and water from the
waste stream. The organics separate and are transferred to an organic slop-
sump. The water that separates from the steam stripper condensate is recycled
to the wastewater stream.
Figure 5-1 presents a plant layout showing the process flows. Whitewater
Is discharged from process operations to the whitewater settling tanks where
it is transferred to the organic condensate tank. During the tests of the
steam stripping column, the discharge from the whitewater organic settling
tank was diverted away from the organic settling tank. The aqueous stream
(top layer) of the organic condensate tank is transferred to the steam strip-
ping column feed tank. A 7.6—cm (3-inch) overflow from this feed tank is
discharged into a lagoon. Waste from the redwater tank is discharged directly
into the lagoon.
An 46—cm (18—inch) subsurface pipe transfers the contents of the lagoon
Into a decanter tank. Other feeds into the decanter tank include a pipe from
an organic sump that Is supplied with organic bottoms from the feed tank, line
water, and pad and walkway runoff. The organic layer from the decanter tank
Is transferred directly to the organic slop sump. The aqueous overflow from
the decanter tank is transferred to the feed tank. When the steam stripping
column is run on complete recycle, the discharge from the steam stripping
column is directed into the decanter tank, bypassing the carbon adsorption
(CA) units. Complete recycle Is used for start—up and Is not used during
normal operation. During normal operation, the effluent from the steam strip-
per passes through a carbon adsorption unit before discharge.
The steam stripping column treats wastes pumped from the feed tank. The
aqueous condensate from the steam stripping vapors separates from the organic
condensate in the organic condensate tank and is recycled to the feed tank.
Figure 5—2 presents the flow diagram of the continuous steam stripping
unit and carbon adsorber at Plant G. Table 5—1 lists the process streams,
their average temperatures, pressures, flow rates, concentrations, and
enthalpies. The aqueous waste from the feed tank (F) passes through a heat
exchanger, Is heated from 27 °C to 83 C, then fed to the steam stripping
5 —1
-------
Rainwater
Pad, Walkway
Closed
Whitewater
Settling
Tank
(I ,’
Decanter
Tank
L
18” Subsurface
Lagoon
Figure 5-1. Plant layout (not to scale).
-------
Feed, F
D tifl ate, D
Feed Vent,V
Tank
Condensate ® densate, C
Fl Steam Tank To
Stripping Feed
Heat Packed Tank
Exchanger Column
___ 19.2m J
(63 ft)
Bi
+
Liquid, Gas,
L G
Organic, 0
Carbon To
Adsorber Organic
0.46 m Sump
— (l.5ft)-’
dia
B2 Bottomif..........._. Steam, S
Figure 5-2. Flow diagram of continuous steam stripping and carbon absorption unit.
5-3
-------
TABLE 5-1 PROCESS STREAM CHARACTERIZATION
Temperature Pressure Flow rate
Process stream (°C) (kPag) (kg/h)
Concentration (mg/kg)
Enthalpy
Nit.robenzane
2-Nitrotoluene 4-Nitrot.oluene
(kJ/kg)
Fl Feed before heat. ex 27 NM 29,900°
605
78
51
111.7
F Feed after heat ex. 83 NM 299900 a
505
78
51
348.3
B Bottoms before heat ex. 112 b 103 314 90 C
41
2.4
44
4680 b
Bi Bottoms after heat ox 58 NM 3 l, 49 O
41
2.4
4.4
243.3
B2 Bottoms after carbon NM NM 31 , 490 C
adsorpt.i on
<0.8
<0.8
<0.8
—-
D Distillate 100 0 443 b
31 , 170 d
5 , 094 d
3 , 129 d
2,676.8
C Aqueous condensate NM 427.60
1,900
87
45
—-
O Organic condensate NM NM 15.40
787,000
193,000
97,000
V Condensate tank vent NM NM ø. 34
09
9
9
--
S Steam 128.8 283 2,033
0
0
0
2,709.2
NM = Not measured.
aAssumes liquid density = 1.0 kg/I.
bFrom energy balance around column (B, F, D, S).
CFrom mass and energy balance around heat exchanger and column (Fl,
dFrom component mass balance around column (B, F, D, S).
0 From organic component mass balance around condensate tank (Dl, C,
t Very low, assumed zero for calculations.
9<0.015 mg/L, gas phase.
81,
0,
S,
V).
0).
-------
packed column. Heat to the heat exchanger is supplied by the treated aqueous
waste (B) from the bottom of the packed column. Steam is fed into the bottom
of the packed column CS) and flows upward through the packed column, counter-
current to the liquid flow. The vapors (D) leaving the top of the column are
condensed and flow into a condensate tank. The aqueous fraction that sepa-
rates In the condensate tank (C) is transferred back to the feed tank while
the organic bottoms (0) are periodically transferred to the organic slop sump.
The steam stripping packed column Is 19.2 m (63 ft) high. The column’s
internal diameter is 0.46 m (1.5 ft) and Is packed with 3.17 m 3 (112 ft 3 ) of
2.5 cm (1—inch) diameter stainless steel rings. The packing is wetted by the
liquid and provides additional surface area for gas-liquid contact.
The effluent from the steam stripping column (B) was first passed through
the heat exchanger and then through one of two carbon adsorption units. The
carbon served as a polishing step removing residual organics from the waste
streams. The carbon was used at a replacement rate of 1,362 to 2,043 kg/day
(3,000 to 4,500 lb/day). The carbon was returned to the supplier for offsite
regeneration.
The carbon bed operated in a pulsed—feed mode. Fresh carbon was placed on
top of the bed an average of 1.5 times a day, with approximately 680 kg (1,500
Tb) of carbon per charge. The carbon beds were 9.2 m (30 ft) high with 1.7 m
(5.5 ft) outage. There were 18.2 to 20.4 Mg (40,000 to 50,000 ib) of carbon
in the carbon bed at a typical operating carbon depth of 7.5 m (24.5 ft).
One of the problems reported with the steam stripping column was heat
exchanger fouling on the feed preheater. The downtime on the units was
reported as less than 1 percent of running time by Plant G personnel. The
operating manpower requirements for the continuous steam stripping unit and
associated lagoons and tanks were reported as 8 h of manager time, 1/2 h of
laboratory time, and 24 h of operator time per day.
The instrument department at Plant G regularly calibrates the flow meas-
uring devices. The steam flow rate was controlled at 127 C (260 •F) on the
bottom. The pressure of the steam Is supplied at 25 to 40 psig. The plant
can raise the header pressure if necessary to maintain the bottom temperature
in the inlet steam. The bottom of the column is maintained at 15 psig. The
packed column was operated without any recycle from the condensate directly
into the packed column, but there was an indirect recycle of the condensate
aqueous phase through the feed tank, which included mixing with other feeds,
and then being transferred to the top of the packed column through the heat
exchanger.
RESULTS FOR PLANT G
Process Stream Composition
The process streams analyzed were liquid (aqueous and organic) and gas-
eous. The components measured in the liquid streams were nitrobenzene,
2-nitrotoluene, and 4—nitrotoluene. The only components identified In the
vent gas samples were 1,5—hexadlyne (HC:::CCH 2 CH 2 C:::CH) and toluene. The
5-5
-------
1,5-hexadlyne was identified using GC/MS. The individual analyses for each
stream were averaged, and these averages were used in the process calcula-
tions. Table 5-2 lists the average concentrations of each sampled process
stream.
Ideally, the flow rates and concentrations of each stream would be meas-
ured, and mass and energy balances would be used to check the validity of the
data. In this system, flow rates were measured on only three streams, temper-
atures on four streams, and pressure on one stream. Collection of samples and
process data was limited by the availability of flowmeters, thermocouples,
pressure transducers, and sampling ports installed In the system. Mass and
energy balances were solved for data on streams that were not completely char-
acterized by direct means. Table 5—3 lIsts the temperatures, pressures, flow
rates, concentrations, and enthalpies of the process streams for the steam
stripper-carbon adsorber. Table 5—4 shows the results of Individual component
mass balances around the steam stripper and condensate tank. The measured
organics In the liquids are considered semivolatiles because each boils above
200 •C and Is only slightly soluble in water. The toluene and 1,5-hexadiyne
are both volatiles. Characteristics of the three compounds are listed In
Table 5—5.
Removal Efficiencies of the Steam Stripper—Carbon Adsorber
The process consisted of two distinct operations: the steam stripping of
the feed waste and the carbon adsorption that follows. Removal efficiencies
can be calculated for each of the processes separately and for the combined
process. Table 5-6 shows the removal efficiency of the steam stripper, carbon
adsorber, the total process for each organic (nltrobenzene, 2—nitrotoluene,
and 4-nitrotoluene), and the total organics. The carbon adsorber effluent
samples had organic concentrations below the analytical detection limit, so
the calculated removal efficiencies involving the carbon adsorber are based on
the reported detection limits. Hence, they represent minimum removal effi-
ciencies and would be higher If the true concentrations of the adsorber efflu-
ent were actually measured. Steam stripping followed by carbon adsorption
effectively treats the feed waste stream. Steam stripping removed 92 percent
of the organics fed to the process, and the carbon adsorber further reduced
the concentrations below the analytical detection limit (<0.8 mg/kg, with
>95 percent removal in the carbon adsorber).
Process Limitations
The feed to be treated should have relatively little dissolved solids
because they will (1) coat the heat exchanger surfaces, (2) clog the column
packing, and (3) clog the carbon bed. The feed stream had a total solids
content of 23 g/L, and the plant reported heat exchanger fouling as a problem
associated with the process (but with downtime of only 1 percent operating
time).
Steam requirements could be reduced by using a larger, more efficient heat
exchanger between the feed (Fl) and bottoms (B) stream. Each 1 C increase In
the feed temperature decreases the steam requirements by 53 kg/h or 2.6 per-
cent of the steam consumption. The capital (purchase cost) and operating
5-6
-------
TABLE 5-2 MEASUR
ED CONCENTRAT
ION OF ORGANICS IN PROCESS STREAMS
Process
stream
Sampling
point
Average concentration (jig/mL)
•
Nitrobenzene
•
2-Nutrotoluene 4-Nutrotoluene
•
1,5-Hexadsyne
Toluene
F,F1
Feed
1-VOC
500
78 51
--
—-
B,61
Bottoms
2-VOC
40
2.4 4.4
--
—-
82
Carbon effluent
3-VOC
<0.8
<0 8 <0 8
--
--
0
Organic condensate
4-VOC
787,000
193,000 97,000
C
Aqueous condensate
6-VOC
1,900
87 46
--
--
V
Condensate tank
vent
6—VOC
< 0015 a
< .øib < O. 015 a
1 • 4 a
04 a
5 Gas
concentration.
-------
TABLE 5-3 PROCESS STREAM CHARACTEAIZATION
Process stream
Temperature
(°C)
Pressure
(kP ag)
Flow rate
(kg/h)
Concentration (mg/kg)°
Enthalpy
Nitrobenzeno
2-Nit.rotolueno 4-Nit.rotolueno
(kJ/kg)
Fl
Feed before heat ox
27
29 , 900 a
505
78
61
111.7
F
Feed after heat ex.
83
NM
29,900 °
505
78
51
348.3
B
Bottoms before heat ox.
112 b
103
31 , 490 C
41
2.4
4.4
468 • 0 b
Bi
Bottoms after heat ox
58
31490 C
41
2.4
4.4
243.3
B2
Bottoms after carbon
adsorpt on
NM
NM
3 l , 4 g O
(0 8
<0.8
<0.8
--
D
Distillate
100
0
443 b
31 , 170 d
5 , 094 d
3 , 129 d
2,675.8
C
Aqueous condensate
NM
NM
427.6°
1,900
87
45
—-
0
Organic condensate
NM
NM
l6.4
787,000
193,000
97,000
V
Condensate tank vent
NM
NM
O.34
09
09
9
--
S
Steam
128.8
263
2,033
0
0
0
2,709.2
NM = Not measured.
aAssumes liquid density = 1 0 kg/L.
bFrom energy balance around column (B, F, 0, 5).
cFrom mass and energy balance around heat exchanger and column (Fl, Bi, S, D).
dFrom component mass balance around column (B, F, 0, S).
0 From organic component mass balance around condensate tank (Dl, C, 0, V).
every low, assumed zero for calculations.
<ø.øi& mg/L, gas phase
-------
Table 5—4. COMPONENT MASS BALANCE AROUND STEAM STRIPPER AND CONDENSATE
TANK (STREAMS F, B, C, 0, 5)
Flow
rate (kg/h)
In
(S+F)
Out
(B+C+O) % Difference
Nitrobenzene
15.10
14.22
+6.0
2-Nitrotoluene
2.33
3.09
-28.0
4—Nitrotoluene
1.53
1.65
+7.6
Total organlcsa
18.96
18.96
00 b
Water
31,914.04
31
,914.04
0 • 0 b
Total
31,933
31
,933
0 • 0 b
aNitrobenzene + 2-nitrotoluene + 4—nitrotoluene. -
bRedundant calculation; these components are mass balances used in earlier
calculations.
5—9
-------
TABLE 5-5. CHARACTERISTICS OF MEASURED ORGANICSa
Molecular Melting Boiling
weight point (C) point (°C)
Density
(g/mL)
Nitrobenzene 123.11 5.7 210.8
1.204
2—Nitrotoluene 137.14 —9.6 221.7
1.163
4-N ltrotoluene 137.14 54.5 238.3
1 • 104 b
Toluene 92.15 —95 110.6
0.867
1,5 Kexadlyne 78.12 -6 86
0.8049
aFrom CRC Handbook of Chemistry and Physics, 58th editIon.
b 70 . c.
TABLE 5-6. REMOVAL EFFICIENCIES (%)
Nitrobenzene 2-Nitrotoluene 4-Nitrotoluene
lotaib
Steam stripper 91.5 96.7 90.9
92.1
Carbon adsorbera >98.1 67.1 >82.0
>95.0
Steam stripper anda >99.8 >98.9 >98.4
carbon adsorber
>99.6
aEffluent concentrations from the carbon adsorber were below quantitation
limits. Indicated efficiencies calculated were based on a <0.8-mg/kg
detection limit.
bTotal removal efficiency of nitrobenzene, 2-nitrotoluene , and 4—nitrotoluene.
5-10
-------
costs (fouling problems, pressure drop) of the heat exchanger increase with
the heat exchanger size, so some optimum (least total cost) sized heat
exchanger exists.
The carbon adsorption process produces 1,362 to 2,043 kg (3,000 to 4,500
ib) of spent carbon per day. This carbon contained approximately 31 kg of
nltrobenzene, 1.8 kg of 2—nitrotoluene, 3.3 kg of 4-nitrotoluene, and adsorbed
solids from the feed stream. Clearly, the regeneration or disposal of the
spent carbon Is a major process limitation.
PROCESS RESIDUALS
Air Emissions
Air emissions from the steam stripper process were minimal. The conden-
sate tank vent flow rate was too small to be measured reliably with the pitot
tube used, and the organic concentrations (measured both onsite and with grab
samples) varied widely. The condensate tank vent is a necessary part of the
apparatus in that It (1) provides a tank vent when adding or removing liquid
from the tank and (2) vents noncondensable components from the condenser. If
the vent were not present, noncondensable gases (e.g., air entrained in the
feed) would accumulate in the condenser and impair its performance. The con-
denser vent flow rate, in the absence of liquid level changes in the con-
densate tank, depends on the amount of noncondensable material entering with
the feed.
The vent gas flow was estimated by assuming that the feed wastewater to
the stripper was saturated with air and that the air was removed from the
water and exited through the condensate tank vent. This flow (0.34 kg/h, or
9.4 L/min) was very small compared to the other system flows, and It corre-
sponds to the maximum possible flow rate through the vent (in the absence of
other noncondensable components).
The observed pltot tube pressure drop and measured organic concentrations
are compiled In Table 5-7. The maximum vent gas emissions (Table 5—8) were
estimated by using the maximum concentrations measured In the vent and the
estimated vent flow rate. The actual vent emissions would be considerably
lower than this estimate because the average vent gas organic concentration
was observed to be much lower than the observed maximums. The average emis-
sions would probably be about one—tenth of the maximum emissions If the
observed concentration variation is representative.
The two compounds identified in the vent gas were not detected in any of
the liquid samples. The feed concentrations required to produce the maximum
observed vent concentration were 0.03 jsg/mL for toluene and 0.18 pg/mL for
1,5-hexadlyne. Both concentrations were well below the detection limit for
feed organlcs (0.8 j g/mL); therefore, they would not be detected In the feed
wastewater.
5—11
-------
TABLE 5-7. VENT GAS MEASUREMENTS, SAMPLE LOCATION 6 VOCC
Time
Sample
number
b
P pitot.
tube
(inches H 2 0)
Bacharach TLC
(ppm hexane)
Gas concentrations (
g/L)
Nitrobenzane
2—Nitrotoluene
4—Nit.rotolueno
l,6-Hexadiyne
Toluene
12:25
FCC—6—VOC—1
0.02
200-300
(16
(16
<15
120
77
12:52
FCC-8-VOC-2A
0.01
2,000
<15
(15
(lb
51500
1,400
12:58
FCC—8—VOC-2B
0.01
NM
-------
TABLE 5-8. ESTIMATION OF MAXIMUM CONDENSATE TANK VENT EMISSIONSa
Maximum measured Estimated maximum emission rate
concentration
Compound (mg/L) (g/h) (kg/yr )
Tolueneb 1.4 0.8 6.7
1,5_Hexadlyneb 5.5 3.1 26.0
Total hydrocarbonsC 7.0 4.0 33.6
aAssumes feed liquid saturated with air, all air exiting through vent.
bAnalyzed by GC.
COrjgjnally measured as ppm hexane with Bacharach TLC.
dEifty_week , 24-h/day year.
5-13
-------
Liquid and Solid Residuals
The treated feed liquid had levels of nitrobenzene, 2-nitrotoluene, and
4-nitrotoluene below the analytical detection limit (<0.8 mg/kg) and could
probably be discharged as wastewater after its acidity Is neutralized.
The major solid residual from the process is approximately 1,360 to
2,040 kg of spent carbon/day from the carbon adsorber. The plant reported
that the carbon is regenerated at another location by the supplier. This
carbon would contain approximately 31 kg of nitrobenzene, 1.8 kg of 2-nitro-
toluene, 3.3 kg of 4—nitrotoluene, and deposited solids from the feed stream.
Suspended solids within the feed stream may also settle as a sludge In the
lagoon.
PROCESS COST
Plant operators supplied adequate operating cost data, but did not supply
any capital cost information. Several attempts to collect capital cost Infor-
mation from the plant failed. Consequently, the capital cost of the process
was estimated and used with the plant—supplied operating costs to determine
the total annualized cost of the process.
The manpower requirements for the process were 8 hours of manager time,
one—half hour of laboratory time, and 24 hours of operator time per day. The
plant supplied manpower costs. Other operating costs supplied by the plant
were steam cost ($5.25/bOO lb), carbon ($1,260,000/yr), and spare parts/in-
ventory costs ($72,000/yr). The steam cost includes the capital and operating
costs of steam generation. The erection and operation of a steam boiler spe-
cifically for the operation of the steam stripper would produce annualized
steam costs in the same range as this reported steam cost (approximately
$6.00/bOO ib). Because most operating plants have existing Installed steam
boilers, steam costs are treated as utilities In the operation of processes
requiring steam. The manpower costs include labor for removing and replacing
carbon In the carbon absorbers. The carbon cost includes the transportation
costs of spent and fresh carbon, and the costs of thermal regeneration of the
spent carbon. Spent carbon is returned to the supplier for thermal regenera-
tion, and this cost Is included with the carbon cost.
The total equipment cost of the process was $872,800 while total capital
investment was estimated at $1,885,260. Direct annual costs of operation were
$1,700,000. Steam cost was 5.5 percent of the direct annual costs of opera-
tion, while carbon costs were 75 percent of the direct annual cost.
The cost effectiveness of the process is $14.13/kg organic removed from
the water. This Is also $O.0089/kg of water treated. The total annualized
cost of the process was $2,240,000.
5-14
-------
SECTION 6
PROCESS LIMITATIONS AND COMPARISONS
PROCESS LIMITATIONS FOR THIN-FILM EVAPORATION
The major process limitations to the use of TFEs are that feed and bottoms
product must be pumpable and that the feed must not foam excessively during
processing. The feed and bottoms can be heated to make pumping easier, but
they must be pumpable through positive displacement pumps in order to intro-
duce the feed Into the evaporator and to withdraw product. If the feed foams
excessively during heating and vaporization of material, the body of the TEE
will fill with foam and foam/feed will be carried out of the evaporator and
into the entrainment separator and condenser. Foaming also reduces the heat
transfer to the material being processed. This Is a potential problem in
treating DAF wastes, as flocculating agents are frequently added to the water
passing through this waste treatment process and appear in the waste sludges.
The condenser and vent gas control systems for the process should be
designed specifically for the wastes being treated. Most of the organics in
the feed that do not condense at the condenser operating temperature will exit
through the condenser vent. As a wide variety of petroleum wastes would be
handled by the process, some of the prospective feeds would probably have
substantial amounts of uncondensable material.
One potential problem is the relatively rapid wear of feed and bottoms
pumps when the process is used with feeds containing suspended solids. This
is not actually a limitation of the process, but will affect the frequency of
pump replacement and the overall operation and maintenance costs of the proc-
ess. Feed streams containing suspended solids also require mixing in order to
prevent phase separation before the material is fed into the TFE.
PROCESS LIMITATIONS FOR STEAM STRIPPING
The steam stripping results for Plants H and I revealed that steam strip-
ping can generally remove over 99 percent of specific volatile compounds from
the wastewater. Total VO removal ranged from 99.8 to 99.999 percent. These
volatile compounds are those analyzed by purge—and-trap procedures. The
results for Plant G showed that steam stripping could also remove semivola-
tiles (extractable organics) from wastewater, but a lower removal efficiency
was seen (92 percent for total VO). Additional treatment by liquid—phase
carbon adsorption improved the efficiency to greater than 99.6 percent.
Consequently, steam stripping is expected to remove organic compounds with a
relatively wide range of volatility. However, if semivolatiles are present,
additional treatment may be needed to decrease the effluent concentration.
Liquid—phase carbon adsorption Is a demonstrated technique to accomplish this.
6—1
-------
At Plant H, purgeable volatile organics were reduced from 6,000 ppm to
<9.8 ppm. At Plant I, purgeable volatile organics were reduced from 6,000 ppm
to <0.037 ppm. Steam stripping of semivolatiles at Plant G reduced a feed
concentration of 634 ppm to 48 ppm. Additional treatment of these semivola-.
tiles by carbon adsorption decreased their concentration to below detection
limits (<0.8 ppm).
The solids content of the wastewater may be a limitation in the use of
steam stripping. At Plant H, the feed contained 1.4 gIL of solids that was
processed through the column. However, this steam stripper requires frequent
cleaning and experienced occasionally high levels of chloroform in the bot-
toms, apparently because of column fouling. At Plant I, solids were removed
prior to steam stripping to levels of 0.01 gIL. No significant fouling prob-
lems are experienced at Plant I, and the operation consistently obtained ppb
levels of organics In the effluent. Consequently, solids removal may be an
Important step in preparing aqueous wastes for steam stripping. However, the
process generates a low volume of sludge that may be a troublesome disposal
problem. At Plant I, this sludge was handled offsite and incinerated.
Another limitation of steam stripping aqueous wastes is the presence of a
separate phase of organics. Flooding of the column was observed at Plant H
when a slug of organic layer entered with the feed. A separate organic phase
can cause operational problems with the steam stripper and may yield higher
than normal bottoms concentrations. The operation at Plant I included decant-
ing any separate organic phase prior to steam stripping.
These tests indicate that steam stripping may be used for aqueous hazard-
ous wastes that contain strippable organic compounds. The presence of solids
or a separate organic layer is the major limitation. However, preliminary
treatment for solids or organic phase removal can overcome these limitations
and produce a wastewater stream for steam stripping. These preliminary treat-
ment steps add to the total treatment costs and, as discussed earlier in the
report, create new sources of air emissions.
COMPARISONS
Plants H and I offer some useful comparisons for wastewater containing
fairly volatile compounds. The operation at Plant H treated about 850 L/min
compared to 42 L/min at Plant I. Solids were removed at Plant I before steam
stripping, and at Plant H, the wastewater received no preliminary treatment.
At Plant H, the stripper bottoms were treated in the wastewater treatment
plant for removal of solids and biodegradation of organics. At Plant I, no
additional treatment (other than occasional pH adjustment) is needed prior to
discharge.
The steam usage at these plants and others is compared in Table 6-1. The
values of 0.036 to 0.1 kg steam/kg water are well within the recommended
design values. A comparison of performance is given In Table 6-2. The steam
stripper at Plant I obtained very low effluent levels (average of 33 ppb)
compared to the other data.
6—2
-------
Plant I
Plant H
Plant K
Ehrenfeld and Bass
(design)
Shukia (design)
Nathan
TABLE 6—1. COMPARISON OF STEAM USAGE RATES
Steam usage
(kg/kg water) Comments
Total VO reduced from 6,000 ppm
to 33 ppb
1,2-Dichioroethane reduced from
4,600 to 0.16 ppm
1,2-Dichioroethane reduced from
15,000 to 0.037 ppm
Design range for steam strippers
Design range for steam strippers.
Theoretical design to reduce
1,2-dichloroethane from 8,700 ppm
to 9 ppm
4
5
6
7
Source
Reference
0.1
0.036
0.10
0.07-0.24
0.1—0.3
0.31
6-3
-------
TABLE 6-2. COMPARISON OF PERFORMANCE
1. Plant I continuous steam
Compound
stripper
Concentration
(ppm)
Percent
removed
In
Out
Chioromethane
33
<0.005
>99.98
Methylene chloride
4,490
0.020
>99.999
Chloroform
Carbon tetrachloride
1,270
55
0.005
<0.005
99.999
>99.99
Trichioroethylene
1,1,2—Trichloroethane
5.6
5.3
<0.005
<0.005
>99.9
>99.9
Total VO
5,860
0.033
99.999
2. Olin Chemical continuous
steam stripper 8
Compound
Concentration
(ppm)
Percent
removed
In
Out
Methylene chloride
3,600
0.22
99.99
Chloroform
52
5.3
89
Carbon tetrachiorlde
1.5
0.09
95
Total VO
3,600
5.6
98
3. Plant H continuous steam
stripper
Compound
Concentration
(ppm)
Percent
removed
In
Out
1,2—Dlchloroethane
5,630
0.097
99.998
Chloroform
270
9.6
96
Other Va
50
<0.01
>99
Total VO
5,950
9.7
>99.8
(conti nued)
6-4
-------
TABLE 6-2. (continued)
4. Plant B fractional batch distillation (5 percent organics in water) 8
Concentration (ppm ) Percent
Compound Initial Final removed
Methyl ethyl ketone 30,000 <10 99.98
2,2—Dimethyl oxirane 6,300 <10 99.9
Methanol 3,500 <10 99.9
Methylene chloride 3,100 <10 99.8
Isopropanol 1,900 <10 99.7
Carbon tetrachioride 1,700 <10 99.7
1,1,1—Trichioroethane 710 530 29
Total VO 49,000 575 99
5. Data from Reference 9 for continuous steam strippers
Concentration (ppm ) Percent
Stripper Compound In Out removed
1
Methylene
chloride
1,430
<0.015
>99.99
2
Methylene
chloride
4.7
<0.002
>99.95
3
Methylene chloride
Chloroform
1,2—Dichioroethane
34
4,509
9.0
<0.01
<0.01
<0.01
>99.97
>99.99
>99.99
6-5
-------
The cost and performance data from Plant H are compared in Table 6-3 with
data obtained from a similar stripper during a 1-day plant visit. 4 One
difference between the two types Is that the stripper examined in this test
used trays for vapor/liquid contact whereas the Plant K operation uses a pack-
ed column. The basic feed constituents are similar; however, Plant K has a
higher concentration of 1,2-dichloroethane in the feed. The difference in
annual operating cost is probably attributable primarily to the higher rate of
steam usage at Plant K where approximately 75 percent of the annual operating
cost Is for steam. Both of these steam strippers achieve similar effluent (or
bottoms) concentrations of VO in the range of 1 to 2 ppm. Although the steam
usage for Plant K appears to be higher, the steam usage for both plants in
terms of VO removed is very similar (6.2 and 6.7 kg steam/kg VO removed). The
small difference in cost-effectiveness Is probably not significant and can be
attributed to the higher feed concentrations observed at Plant K.
A comparison of condenser efficiencies is given In Table 6-4. A higher
efficiency was observed at Plant H (99 percent), which used refrigerated
glycol at 2 C, than was observed at Plant I (91 percent), which had a con-
denser cooled with cooling tower water at 21 C. The highest efficiency was
observed for ethylene dichloride at Plant H because of the high vapor-phase
concentration and its relatively low vapor pressure. The lowest efficiency
was observed for vinyl chloride, which has a very high vapor pressure and is
difficult to condense.
6-6
-------
TABLE 6-3. COST COMPARISON
Item Plant H Plant K 3
Capital cost ($) 950,000 700 000
Operating cost ($/yr) 250,000 450,000
Total annualized cost ($/yr)a 405,000 564,000
Average feed rate (L/rnin) 818 680
Steam rate (kg/h @ 50 psig) 1,790 4,090
Feed constituents (ppm )
1,2-Dichioroethane 5,600
Chloroform 270
Other VO 59
Bottoms constituents (ppm )
1,2-Dichioroethane 0.16 • 037 b
Chloroform 08 C 13 b
Other vo <0.01 14 b
Steam usage
kg/kg water 0.036 0.10
kg/kg Va removed 6.2 6.7
Cost-effectiveness
$/Mg Va removed 220 d 120 e
aBased on a 10-year lifetime at 10% (Capital recovery factor = 0.163).
bBased on a single sample analysis from presurvey trip.
CBased on 6 of 10 runs.
dBased on company data of 1,820 Mg/yr recovered.
eEstimated from single analysis and 329 days/year operation.
6-7
-------
TABLE 6—4. SUMMARY OF AVERAGE CONDENSER VENT RATES AND EFFICIENCIES
Plant H
Condenser Condenser
loading Vent rate efficiency
Constituent (g/s) (g/s) (percent)
Vinyl chloride 0.089 0.084 6
Chloroethane 0.081 0.043 47
1,1-Dichioroethene 0.036 0.031 15
1,1—Dlchioroethane 0.11 0.013 88
1,2-Dichloroethene 0.006 0.001 84
Chloroform 2.9 0.11 96
Ethylene dichioride 63 0.34 99.5
Total 66 0.62 99
Plant I
Condenser Condenser
loading Vent rate efficiency
Constituent (g/s) (g/s) (percent)
Chloromethane 0.021 0.0024 89
Methylene chloride 2.9 0.29 90
Chloroform 0.81 0.045 94
Carbon tetrachloricle 0.038 0.0039 90
Total 3.8 0.34 91
6-8
-------
SECTION 7
REFERENCES
1. Grubbs, F. E. Procedures for Detecting Outlying Observations in Samples.
Technometrics . Volume 11 (1). February 1969. 21 p.
2. Kent, J. A. (ed.). Riegel’s Handbook of Industrial Chemistry. Van
Nostrand Reinhold Co., New York. 1983. p. 930-934.
3. Telephone conversation with Hugh Wise. U. S. Environmental Protection
Agency, Office of Water. Washington, D.C. November 1986.
4. Branscome, M. and K. Leese. Trip Report: Visit to Plant K. June 26,
1986. 7 pages.
5. Ehrenfeld, J., and J. Bass. Handbook for Evaluating Remedial Action
Technology Plans. EPA 600/2-83-076. August 1983.
6. Shukia, H. M., et al. Process Design Manual for Stripping of Organics.
EPA 600/2—84—139. August 1984.
7. Nathan, M. F. Choosing a Process for Chloride Removal. Chemical
Engineering. January 1978. p. 93—100.
8. Vancll, M. A. Pretreatment Test Report Summaries and Analysis. Prepared
for Industrial Wastewater Project File. U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. November 7, 1986.
9. Metcalf and Eddy, Inc. Briefing: Technologies Applicable to Hazardous
Waste. Prepared for U.S. Environmental Protection Agency, Cincinnati,
OH. May 1985. 2.15, 2.16.
7—1
-------
|