&EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
RTF, NC 27711
EMB Report 87-IWW-2
NOVEMBER 1988
Air
INDUSTRIAL WASTEWATER
STEAM STRIPPER
PERFORMANCE
EMISSION TEST REPORT
RHONE-POULENC AG COMPANY
INSTITUTE, WEST VIRGINIA
SUMMARY REPORT
-------
FINAL REPORT
METHOD DEVELOPMENT AND TESTING
FROM INDUSTRIAL WASTEWATER FACILITIES
RHONE-POULENC AGRICULTURAL CHEMICALS
INSTITUTE, WEST VIRGINIA
ESED PROJECT NO. 84/11
CONTRACT NO. 68-02-4337
WORK ASSIGNMENT NO. 10
TRC PROJECT NO. 4683-E81
Prepared by:
James E. Canora, Principal Scientist
TRC ENVIRONMENTAL CONSULTANTS, INC.
800 CONNECTICUT BOULEVARD
EAST HARTFORD, CT 06108
Prepared for:
U.S. ENVIRONMENTAL PROTECTION AGENCY
EMISSION STANDARDS AND ENGINEERING DIVISION
EMISSION MEASUREMENT BRANCH
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
December 1988
-------
TABLE OF CONTENTS
SECTION
1.0 INTRODUCTION
1.1 Program Objectives
1.2 Site Description
1.3 Measurement Program
2.0 SUMMARY OF RESULTS AND CONCLUSIONS
2.1 Steam Stripper Performance
2.2 Method Development
3.0 PROCESS DESCRIPTION AND OPERATION
4.0 SAMPLING LOCATIONS
5.0 SAMPLING AND ANALYTICAL METHODS ,
5.1 Sampling Equipment/Procedures ,
5.2 Analytical Methods
5.2.1 Purgeable Organic Compounds (EPA Method 624)
5.2.2 Extractable Organic Compounds (EPA Method 625)
5.2.3 Total Organic Carbon (TOC) and Purgeable
Organic Carbon (POC) (EPA Method 415.2) . .
6.0 DETAILED RESULTS
6.1 Steam Stripper Performance Evaluation
6.2 Total Organic Carbon and Purgeable Organic
Carbon Analyses
7.0 QUALITY ASSURANCE AND QUALITY CONTROL
7.1 Data Validation
7.1.1 Accuracy
7.1.2 Precision
7.2 Calibration Procedures and Frequency
7.3 Sample Custody
7.3.1 Chain-of-Custody Form
7.3.2 Chain-of-Custody Tape
7.3.3 Shipping of Samples
7.4 Deviations from the Test Plan
APPENDICES
A TEST PLAN
B DESCRIPTIONS OF ANALYTICAL METHODS
C DATA TABLES
D ANALYTICAL DATA SHEETS
E CHAIN-OF-CUSTODY FORMS
PAGE
1
1
2
2
4
4
5
10
14
14
14
16
16
19
21
21
24
26
26
26
28
31
31
31
31
33
33
-11-
-------
LIST OF FIGURES
FIGURE PAGE
1 STEAM STRIPPER SYSTEM AT RHONE-POULENC AGRICULTURAL
COMPANY 7
2 STEAM STRIPPER SAMPLING LOCATIONS AT RHONE-POULENC
AGRICULTURAL COMPANY 12
3 SCHEMATIC OF HEAT EXCHANGER SAMPLING SYSTEM 15
4 CHAIN-OF-CUSTODY FORM 32
LIST OF TABLES
TABLE PAGE
4-1 STEAM STRIPPER SAMPLING MATRIX 13
5-1 PURGEABLE ORGANIC COMPOUNDS 17
5-2 EXTRACTABLE ORGANIC COMPOUNDS 18
6.1-1 RPAC STEAM STRIPPER BULK STREAM FLOWRATES 22
6.1-2 RPAC STEAM STRIPPER AVERAGE COMPONENT STREAM
CHARACTERIZATION 23
6.2-1 COMPARISON OF TOC AND TOTAL PURGEABLE EXTRACTABLE
CONCENTRATIONS 25
7-1 WATER MATRIX SPIKE/MATRIX SPIKE DUPLICATE RECOVERY .... 30
-111-
-------
1.0 INTRODUCTION
The Environmental Protection Agency is currently developing background
information on volatile organic (VO) emissions from industrial wastewater and
control techniques for reducing these emissions. The Emission Measurement
Branch (EMB) of EPA's Office of Air Quality Planning and Standards (OAQPS) is
evaluating sampling and analytical methods for measuring the VO emission
potential of a waste to support this program. In addition, EMB is evaluating
steam strippers as a control technology. The information may ultimately aid
in the development of an air emission standard or a Control Technique
Guideline (CTG) document.
Field sampling and analytical programs are to be conducted at several
organic chemical, plastic and synthetic fiber (OCPSF) facilities which use
steam strippers for wastewater treatment. Sampling was conducted at
Rhone-Poulenc Agricultural Chemicals (RPAC) in Institute, West Virginia,
during the week of September 14, 1987. TRC Environmental Consultants, Inc.
(TRC) performed sampling and analytical tasks, and Radian Corporation
monitored the process operation. Subsequently, TRC provided Radian with
analytical results and Radian applied these data to an evaluation of the
process.
1.1 Program Objectives
There were three primary objectives for the testing performed at RPAC:
1.0 Investigation of the cost effectiveness of controlling volatile
organic (VO) emissions from wastewater streams at OCPSF
facilities with steam strippers.
2.0 Evaluation of steam stripper efficiency on a compound-specific
basis using EPA methods for the analysis of wastewater.
3.0 Evaluation of EMB test methods (e.g., air purge and trap, steam
distillation, heated headspace) by comparison with approved EPA
methods. As of this date, this objective has not been met for
-1-
-------
the RPAC program. Samples were collected for method evaluation,
but were not analyzed because desired EMB conditions have not
been established.
1.2 Site Description
Rhone-Poulenc Agricultural Chemicals (RPAC) is an agricultural chemical
manufacturing facility. The process wastewater is treated by steam stripping
to remove organics prior to the wastewater treatment plant. The steam
stripper is in operation 24 hours per day.
1.3 Measurement Program
The measurement or sampling program was conducted by TRC over a three day
period from September 15-17, 1987. The objectives of the study at RPAC
required simultaneous sampling of the stripper feed, tails, and overhead waste
streams. During that time, the steam stripper feed, tails, and overhead
streams were sampled and analyzed for volatile and semi-volatile organic
compounds using modified EPA Methods 624 and 625. Additional samples were
taken for total organic carbon (TOG), purgeable organic carbon (POC), and EPA
method development analysis.
Concurrent with the wastewater sampling. Radian monitored the following
process parameters:
Base liquid level
Column pressure differential
Overhead flow
Feed flow
Tails flow
Jet decanter liquid level
Condensate pot liquid level
Steam stripper water level
Steam stripper temperature
Steam stripper head pressure
-2-
-------
Critical to the stripper evaluation were the liquid flow measurements of
the stripper feed, overhead and tails. The feed and overhead were equipped
with calibrated orifices, and the flows could be monitored in the control
room. The stripper tails was not equipped for monitoring and required a
periodic flow measurement implemented by diverting the flow into a 55-gallon
drum and timing the fill rate.
-3-
-------
2.0 SUMMARY OF RESULTS AND CONCLUSIONS
2.1 Steam Stripper Performance
Steam stripper performance was evaluated on a compound-specific basis as
well as an overall basis. The efficiency evaluation was made using flow and
concentration data to perform material balances. The average removal
efficiencies for the compounds detected are listed below:
Compound VOC Removal (%)
Benzene 99.62
Toluene 98.94
Ethylbenzene 99.44
Isophorone 96.02
Naphthalene 99.66
5-Ethyl-l,2-Methylpyridine 84.54
1,2,3,4-Tetrahydronaphthalene 98.33
Acetophenone 97.91
2-Methyl-l,3-Cyclopentanedione 98.29
Total VOC 92.25
The overall efficiency was calculated based on the flow weighted
concentrations of each compound in the inlet and outlet streams. The sampling
and analytical methods employed in the compound-specific stripper efficiency
determination were conducted in accordance with EPA Methods 624 and 625. The
analytical procedures were judged to be valid according to the quality
assurance measures as described in Section 7.1. Although no standard protocol
exists for sampling wastewater streams, sampling was conducted in such a way
as to minimize evaporative VO losses and assure representativeness. All
samples were collected from pump outlets where the waste streams should have
been sufficiently mixed, minimizing the potential for stratification and
non-representative sampling. Heat exchangers were employed to cool samples
under 20°C prior to contact with air. Flowrates were also monitored with
calibrated orifices at the feed and overhead and with the timed fill of a
-------
55-gallon drum from the tails stream. These flow measurements are assumed to
be accurate although orifice calibration data has not been provided.
2.2 Method Development
Samples were collected for method development purposes, but analyses have
not been performed. The decision to hold this portion of the program was
based on incomplete results from laboratory evaluations presently being
conducted on synthetic wastes.
-5-
-------
3.0 PROCESS DESCRIPTION AND OPERATION
The primary source of wastewater during production is the jet collection
system. A steam jet system is used to pull vacuum on the process refining
stills and the reactor recovery stills. This steam picks up organics during
operation of the vacuum jet system. The condensed steam containing organics
is collected in various jet collection pots and pumped to a central jet
collection pot which is shown in Figure 1. Other wastewater streams feeding
the central jet collection pot include the scrubber decant pot, periodic
reactor washes (once per month), and wastewater overflow from the jet
collection pot. Process wastewater from the central jet collection pot is
pumped to the decanter tank. Additionally, the reactor recovery wash-up
header which is used twice a year feeds into the decanter tank.
The decanter tank, which is 15,000 gallons in capacity, is used to recover
organics picked up by the jet collection system. The recovered organics
overflow to a collection tank and are recycled to the process. The water
layer is drawn off the side of the decanter tank and pumped to the steam
stripper. The temperature of the water normally ranges between 55 - 60°C.
Any sludge that enters the decanter settles to the bottom of the tank and is
removed as necessary, using a bottom suction line.
The steam stripper system is shown in Figure 1. The purpose of the steam
stripper is to remove volatile organic compounds (VOC's) from the wastewater
generated by the process. At least one of the VOC's removed by the stripper
is not biodegradable in the biotreatment basin which is located at the
facility wastewater treatment plant. The stripper column is 60 inches in
diameter and contains 14 Glitsch valve trays. The trays are spaced one foot
apart. A water-cooled condenser is installed in the top head of the column to
condense the overheads stream. Nitrogen is used to maintain the column
pressure at 5 pounds per square inch (psig). A control valve vents pressure
-6-
-------
From
WLSS
Condensale
Pol
Jel Collection Pols
A.
Reactor
Washes
^T^ "T""
"^te^f
^^^
Scrubber
Decani
Pot
^»^_ ^^^
Recycle
lo
Process
Organics
H,0
ca
7
/
^ S "j ^
Process
Waslewaler
Decanter
o "'" ^T.
-{XI »
Steam
Water Layer
Steam Stripper
(WLSS)
N, -IX] i-tlj-/
WLSS
Condensale
Pol
To Jel
Collection Pot
-Plant Sewer
To Waslewaler
Treatment Plant
Figure 1. Steam stripper system at Rhone-Poulenc Agricultural Company.
-------
from the column when the column pressure exceeds 5 psig. The column is
equipped with a condensate pot for collection of the condensed overhead.
Feed from the decanter tank passes through a heat exchanger and is heated
to ~110°C by the stripper tails stream. On startup, additional heating of the
feed is supplied by a preheater, heated with 75 psig steam. The feed rate to
the stripper is controlled by the level in the decanter tank. The feed rate
normally operates within a range of 8,000 to 14,000 Ibs/hr. The feed stream
leaving the decanter tank is a single phase wastewater stream; an organic
layer is present only when there is a process upset. The stripper cannot
process a stream with an organic layer.
Design of the column provides for use of 75 psig steam. However, waste
steam is supplied to the sparger at 25 psig under normal operation. The steam
flowrate is controlled by the column pressure differential which ranges from
30-40 inches. Additional heating in the column may be provided by steam coils
submerged below the liquid level at the bottom of the column. These coils are
only used during cold weather. The overheads stream flowrate normally
operates at about 1,000-2,000 Ibs/hr and is controlled to maintain a constant
level in the condensate collection pot. The condensed overheads are pumped
from the condensate pot to the jet collection pot which overflows to the
central jet collection pot. As previously discussed, the water and organics
in the central jet collection pot are pumped to the decanter tank for recovery
of the organics.
The column tails stream is pumped through an exchanger to heat the feed
stream. The flowrate of the tails stream is controlled to maintain constant
liquid level in the bottom of the column. The temperature of the tails stream
leaving the column ranges from 95-100°C. The stream is cooled in the heat
exchanger before being discharged to the process sewer at 75-80°C. The
-8-
-------
tails stream is combined with other wastewater in the process sewer and
collected in a sump which is located on-site at the process unit. Wastewater
from the sump is pumped to a decanter before the process wastewater flows to
the facility sewer system and on to the wastewater treatment plant.
-9-
-------
4.0 SAMPLING LOCATIONS
The RPAC stripper wastewater treatment system is monitored by plant
personnel for flow, temperature, pressure, and tank liquid level in the
control room. The following parameters were measured and recorded from the
control room by Radian personnel throughout the sampling program:
Base liquid level
Column pressure differential
Overhead flow
Feed flow
Jet decanter liquid level
Condensate pot liquid level
Steam stripper water level
Steam stripper temperature
Steam stripper head pressure
In order to accurately determine the stripper performance, the feed and tails
stream flowrates were monitored. RPAC measured both the feed and overhead
flowrates using calibrated orifices and pressure transducers. These devices
were recalibrated by RPAC prior to the beginning of the sampling program. The
feed and overhead lines are equipped with 0.920 and 0.413 inch orifices,
respectively. Each line was also outfitted with pressure taps on the pump
inlets and outlets to provide additional flow measurements based on pump
output versus pressure curves. TRC measured and recorded pressure data at
hourly intervals but the accuracy of the method was inadequate and results
were discarded. The tail or stripper bottom line was not equipped with a
calibrated orifice. Flow was measured at hourly intervals by loading a
55-gallon drum with the tails stream, recording time on a stopwatch, and
volumetrically measuring the stream collected.
Sampling taps were located at the feed, tails, and overhead pump outlets.
Sampling locations for required pressure measurements and liquid sample
-10-
-------
collection are indicated in Figure 2. Wastewater samples were collected at
hourly intervals six times daily according to the matrix presented in
Table 4-1.
In addition to feed, tails and overhead flowrates, steam flowrate at the
stripper was estimated based on a mass balance for water in the stripper. The
steam flowrate is normally not monitored at RPAC.
-11-
-------
Jet Collection Pols
A.
From
WLSS
Condcnsale
I Pol
K-*
to
I
Reactor
Washes
Sr*11
~^^^i
r-^"
"^^
^^
Scrubber
Decant
Pol
^^•^ *~^^
Recycle
to
Process
Organtcs
H,O
Process
Wastowater
Decanter
Plant Sewer
FEED
SAMPLES) H,?^£
—{3d »
Water Layer
Steam Stripper
(WLSS)
To Jet
Collection Pot
To Wastewater
Treatment Plant
Figure 2. Steam stripper sampling locations at Rhone-Poulenc Agricultural Company.
-------
TABLE 4-1
STEAM STRIPPER SAMPLING MATRIX
RHONE-POULENC AGRICULTURAL CHEMICALS
September 15-17, 1987
Date
9/15
9/16
9/17
Total
Feed
Method
Time Development
10:30
11:30 X
12:30
13:30*
14:30
15:40
14:00
15:00*
15:30
08:50
10:00
11:00
12:00
13:00
14:00
Number of Samples
(excluding duplicates) 2
and Tails
EPA
624/625
X
X
X
X
X
X
X
X
X
18
TOC/POC
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
30
Overheads
EPA
624/625 TOC/POC
X
X X
X X
X X
X X
X X
6 5
* Collected replicate samples for RPAC internal analyses
-13-
-------
5.0 SAMPLING AND ANALYTICAL METHODS
All sampling and analysis procedures used in the field study are fully
described in the Test Plan submitted to the EPA on September 11, 1987 which is
in Appendix A. A brief overview is included in this section along with a
diagram of the heat exchanger used for sampling. All sample containers and
sampling equipment were precleaned and decontaminated as described in the Test
Plan. Sample identification and preservation procedures for collected samples
were also adhered to during the field program. Any deviations or
modifications to the procedures and methods described in the Test Plan are
summarized in Section 7.0, Quality Assurance and Quality Control.
5.1 Sampling Equipment/Procedures
Samples taken at RPAC were collected from taps which were located at the
pipeline walls. The tap was reduced to a 1/4 inch tube to minimize aeration
of the sample and subsequent VO loss. The sampling valves were opened slowly
to prevent accidental exposure to wastes. The waste stream was cooled using a
heat exchanger as depicted in Figure 3. The heat exchanger consisted of a
6-foot by 0.25 inch stainless steel coil submerged in an ice bath which cooled
the liquid waste sample prior to containerization.
The system was purged prior to sampling. The flowrate through the system
was lowered during sampling to help reduce the sample temperature prior to air
contact. The wastewater temperature was monitored with a thermocouple and
maintained below 20°C.
5.2 Analytical Methods
This section provides a brief review of the analytical methods utilized in
the project. The detailed description of these methods is included in
Appendix B.
-14-
-------
PROCESS LINE
VALVE
P" REDUCER (I" TUBE FITTING)
i" TEFLON OR STAINLESS STEEL COIL
ICE BATH
SAMPLE CONTAINER
Figure
Heat Exchanger Sampling System
-15-
-------
5.2.1 Purgeable Organic Compounds (EPA Method 624)
This method was designed to determine volatile organic materials in
wastewater. It was designed to be used to meet the monitoring requirements of
the National Pollutants Discharge Elimination System (NPDES). The method was
designed to determine 31 volatile organic priority pollutants (see
Table 5-1). However, other volatile organic compounds could be determined
with a similar degree of accuracy and precision. The method involves a purge
and trap gas chromatographic/mass spectrophotometer (GC/MS).
An inert gas (Helium) was bubbled at a rate of 40 ml/min through a 5 ml
aqueous sample contained in a specially designed purging chamber at 30°C. The
purgeable organic compounds were transferred from the aqueous phase to the
vapor phase. The vapor phase was passed through a three-material sorbent trap
(SP-1000, silica gel, and Tenax GC) where the purgeables were trapped. After
purging was completed, the trap was heated to 80°C and backflushed with
nitrogen gas to desorb the purgeable onto a gas chromatographic column packed
with 60/80 mesh Carbopack B/1% SP-1000. The sample passed through the column,
which was temperature programmed to gradually increase at the rate of 8°C per
minute from 30°C to 220 °C, and held at 220°C for 15 minutes. The components
of the sample were separated and detected by the mass spectrometer.
Components (either one of the 31 priority pollutants or other than the 31
listed compounds) were identified by the mass spectra and quantified through
the use of the nearest internal standards. The system used was Hewlett
Packard Model 5985.
5.2.2 Extractable Organic Compounds (EPA Method 625)
This is a gas chromatographic/mass spectrometer (GC/MS) method applicable
to the determination of the 43 base neutral extractables and 11 acid
extractable components of the priority pollutants (see Table 5-2). A
-16-
-------
TABLE 5-1
PURGEABLE ORGANIC COMPOUNDS
Chloromethane
Brornomethane
Vinyl Chloride
Chloroethane
Methylene Chloride
1,1-Dichloroethene
1,1-Dichloroethane
Trans-l,2-Dichloroethene
Chloroform
1,2-Dichloroethane
Trichlorofluoromethane
1,1,1-Trichloroethane
Bromodichloromethane
Carbon tetrachloride
1,2-Dichloropropane
Trans-1,3-Dichloropropene
Trichloroethene
Dibromochloromethane
Benzene
1,1,2-Trichloroethane
cis-1,3-Dichloropropene
Brornoform
2-Chloroethyl vinyl ether
Tetrachloroethene
1,1,2,2-Tetrachloroethane
Toluene
Chlorobenzene
Ethylbenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
-17-
-------
TABLE 5-2
EXTRACTABLE ORGANIC COMPOUNDS
Base Neutral Compounds;
Acenaphthene
Ac enaphthy1ene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(g,h,i)perylene
bis (2-Chloroethoxy) methane
bis (2-Chloroethyl) ether
bis (2-Chloroisopropyl) ether
bis (2-Ethylhexyl) phthalate
4-Brornophenyl phenyl ether
Butyl benzyl phthalate
2-Chloronaphthalene
4-Chlorophenyl phenyl ether
Chrysene
Dibenzo (a,h) anthracene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
3,3-Dichlorobenzidine
Diethyl phthalate
Dimethyl phthalate
Di-N-butyl phthalate
2,4-Dinitrotoluene
2,6-Dinitrotoluene
Di-N-octyl phthalate
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadi ene
Hexachlorocyclopentadiene
Hexachloroethane
Indeno (1,2,3) pyrene
Isophorone
Naphthalene
Nitrobenzene
N-Nitroso-di-N-propylamine
N-Nitrosodiphenylamine
Phenanthrene
Pyrene
1,2,4-Trichlorobenzene
Acid Compounds;
4-Chloro-3-Methylphenol
2-Chlorophenol
2,4-Dichlorophenol
2,4-Dimethylphenol
4,6-Dinitro-2-methylphenol
2,4-Dinitrophenol
2-Nitrophenol
4-Nitrophenol
Pentachlorophenol
Phenol
2,4-6-Trichlorophenol
-18-
-------
one-liter sample was serially extracted with methylene chloride at a pH
greater than 11, and again at a pH less than 2 using a separatory funnel. The
methylene chloride extract was dried to a volume of 1 ml, and analyzed by gas
chromatographic/mass spectrometer. Both the listed compounds and unlisted
components were identified through mass spectra comparison and quantified
through internal standard response factors. The system used was Hewlett
Packard Model 5990 with Mass Spectrometric Detector. The column was fused
silica coated with DB-5 (Supelco, Bellefonte Park 0, PA) and operated between
50°C and 300°C with an increasing rate of 10°C per minute.
5.2.3 Total Organic Carbon (TOG) and Purgeable Organic Carbon (POC) (EPA
Method 415.2)
The method converts organic carbon to carbon dioxide and then methane.
The methane was measured by a flame ionization analyzer (O.I. Corp. Model
700). The instrument was designed for a two-step operation to distinguish
between purgeable and non-purgeable organic carbon.
A sample was combined with 1 ml of acidified persulfate reagent and placed
in a sparger. The sample was purged with helium which transfers inorganic CC>2
and purgeable organics to a C02 scrubber. The C02 was removed with at least
99.9% efficiency with a 2.5-minute purge. The purgeable organics proceeded
through a reduction system where the gas stream was joined by hydrogen and
passed over a nickel catalyst which converted the purgeable organic carbon to
methane. The methane was measured by a flame ionization detector. The
detector signal was integrated and displayed as the concentration of purgeable
organic carbon.
The sample was then transferred to a quartz ultraviolet reaction coil
where the nonpurgeable organics were subjected to intense ultraviolet
illumination in the presence of the acidified persulfate reagent. The
-19-
-------
nonpurgeables were converted to C02 and transferred to a second sparger where
a helium purge transferred the C(>2 to the reduction system and into the
detector. The signal was integrated, added to the purgeable organic carbon
value, and displayed as the concentration of total organic carbon.
-20-
-------
6.0 DETAILED RESULTS
6.1 Steam Stripper Performance Evaluation
Flowrate data were gathered during testing for the streams entering and
leaving the steam stripper. Using these data and the concentration data
discussed in Section 3.0 [of the Radian report], material balances were
performed to evaluate the removal performance obtained by the steam stripper
for individual pollutants and overall volatile organics (VO).
Flowrates for the feed and overheads streams were monitored during testing
using measurements provided by calibrated orifices. In addition, the flowrate
of the tails was periodically measured by diverting this stream to a 55-gallon
drum and recording the increase in volume per unit time. The volumetric
flowrates of the feed, tails, and overhead streams measured during testing
were converted to mass flowrates and are presented in Table 6.1-1.
Using concentration data obtained from periodic sampling and the flowrates
measured at the time of each sample, mass loadings were computed for each
pollutant in the feed, tails, and overhead streams. These results, along with
calculated percent removals, are presented in Appendix C (Tables C-l through
C-6). It should be noted that tails flowrates were not measured at each time
a sample was taken. Therefore, tails flowrates were estimated for these
sample times based on the feed flowrate and the average measured ratio of
tails to feed flow of 1.7.
The computed loadings for the feed, tails, and overheads streams were
averaged for the three days of testing. The average concentration of each
pollutant was also computed for the three days of testing. The average
loadings and concentrations are presented in Table 6.1-2. Also shown in Table
6.1-2 are the average percent removals for each pollutant. These removals
were estimated based on the average loadings in the feed and tails streams.
-21-
-------
TABLE 6.1-1
RPAC STEAM STRIPPER BULK STREAM FLOWRATES
Date
Time
Feed
Rate
(kg/hr)
Overheads
Rate
(kg/hr)
Tails
Rate
(kg/hr)
9/15/87a
9/16/87b
9/17/87b
11:15
12:15
14:15
15:15
16:15
14:00
14:30
15:00
15:30
16:00
9:
9:
30
10:30
11:00
11:30
12:30
13:30
14:30
2,077
2,168
2,043
2,043
2,168
1,771
1,771
1,771
1,771
1,771
1,445
1,445
1,351
1,615
1,623
1,652
1,670
1,668
1,373
1,426
1,435
1,267
1,283
1,298
1,289
1,283
3,448
3,599
3,391
3,391
3,599
3,004
2,921
2,394
2,442
2,512
a Tails rate values for this day were calculated from an average tails-to-
feed (T/F) ratio of 1.66. This was the average ratio obtained for T/F from
measurements taken on 9/16/87 and 9/17/87. Reliable measurements for the
tails rate were not obtained on 9/15/87. Tails rates for 9/15/87 listed here
are the calculated values.
b On this day, the tails rate was determined by measuring the liquid level
increase in the drum per unit time. The tails rate was calculated from the
following expression: Tails rate = (ird2p/4) x (level increase per unit time)
where d = inside diameter of the drum = 22.5 inches. Feed and overhead rates
were not recorded at these times.
-22-
-------
LO
I
TABLE 6.1-2
RPAC STEAM STRIPPER AVERAGE COMPONENT STREAM CHARACTERIZATION
Feed
Tails
Overheads
Flow Weighted Flow Weighted Flow Weighted VOC
Flowa Concentration" Flow Concentration*- Flow Concentration*' Removal6
Component
Benzene
Toluene
Ethyl Benzene
Isophorone
Naphthalene
5-Ethyl-l ,2-Methylpyridine
1 ,2,3,4-Tetrahydronaphthalene
Acetophenone
2-Methyl-l ,3-Cyclopentanedione
Total VO
Water
Total
(kg/hr)
9.26E-03
1.96E-03
2.66E-03
1.91E-03
2.20E-02
1.76E-01
1.40E-01
1.81E-02
2.21E-02
3.95E-01
1,828.61
1,831
(ppmw)
5.06E+00
1.07E+00
1.46E+00
1.04E+00
1.20E+01
9.64E+01
7.67E+01
9.91E+00
1.21E+01
2.29E+01
(kg/hr)
3.45E-05
2.07E-05
1.49E-05
7.58E-05
7.58E-05
2.73E-02
2.35E-03
3.79E-04
3.79E-04
3.06E-02
2,972.13
2,973
(ppmw)
1.19E-02
6.98E-03
5.00E-03
2.55E-02
2.55E-02
9.18E+00
7.89E-01
1.27E-01
1.27E-01
7.90E-02
(kg/hr)
7.45E-03
1.27E-03
1.33E-03
2.36E-03
1.39E-02
1.90E-01
9.16E-02
2.90E-02
1.77E-02
3.55E-01
1,436.94
1,437
(ppmw)
5.18E-00
8.85E-01
9.25E-01
1.64E-00
9.65E+00
1.32E+02
6.38E+01
2.02E+01
1.23E+01
3.12E+01
(V.)
99.62
98.94
99.44
96.02
99.66
84.54
98.33
97.91
98.29
92.25
Feed-to-
Closure Steam
Error Ratio
(%) (F/S)
19.14
34.09
49.57
-27.93
36.62
-23.38
33.04
-61.71
18.42
0.70
a The flow is calculated by averaging the individual compound flows presented in Tables C-5 through C-10 in Appendix C.
calculated by averaging the six total flows presented in Tables C-5 through C-10 in Appendix C.
b The feed pollutant flow weighted concentration = feed pollutant flow/total flow x 106,
i.e.. for benzene, 9.26 x 10~3 kg/hr / 1,831 kg/hr x 106 = 5.06 ppmw.
c The tails pollutant flow weighted concentration = tails pollutant flow/total flow x 106,
The total flow is also
i .e.,
for benzene, 3.54 x 10~5 kg/hr / 2.973 kg/hr x 106 = 1.19 x 10~2
ppmw.
The overheads pollutant flow weighted concentration = overheads pollutant flow/total flow x 10°,
for benzene, 7.45 x 10~3 kg/hr / 1,437 kg/hr x 106 = 5.18 ppmw.
.e.
e VOC removal = (feed pollutant flow - tails pollutant flow) / feed pollutant flow x 100,
i.e., for benzene, 9.26 x 10~3 kg/hr - 3.45 x 10~5 kg/hr / 9.26 x 10~3 kg/hr x 100 = 99.62%.
f Closure error (1.) = (feed pollutant flow - (tails pollutant flow + overheads pollutant flow)) / feed pollutant flow x 100,
for benzene, 9.26 x 10~3 kg/hr - (3.45 x 10~5 kg/hr + 7.45 x 10~3 kg/hr) / 9.26 x 10~3 kg/hr x 100 - 19.14%.
i.e.
-------
The loadings in the overhead streams were not used due to the difficulty in
obtaining a representative sample.
This overall VO removal was determined by summing the average pollutant
loadings in the feed and tails streams and comparing the difference. The
average VO removal during the test was 92.25 percent.
6.2 Total Organic Carbon and Purqeable Organic Carbon Analyses
Theoretically, the results from the TOC (total organic carbon) analyses
and the sum of the extractable and purgeable analyses (Methods 624 and 625)
should be equal if identical units are used for reporting the analyses and all
the organic material can be detected using Methods 624 and 625. Table 6.2-1
shows that TOC values were often lower than the sum of the 624 and 625
values. This difference can be attributable to the relative uncertainty of
results from methods 624 and 625. The TOC results may be lower than normal
due to incomplete combustion of the larger molecular weight compounds.
Purgeable organic carbon (POC) analyses were not conclusive as all samples
analyzed showed levels below the method detection limit. In comparison with
Method 624 analyses, the POC method should have shown concentrations slightly
above the limits of detection.
-24-
-------
TABLE 6.2-1
COMPARISON OP TOG AND TOTAL PURGEABLE EXTRACTABLE CONCENTRATIONS
Feed Concentrations
(ug of carbon/1)
Tails Concentrations
(ug of carbon/1)
Overhead Concentrations
{ug of carbon/1)
9/15/87
TOC
Total Purgeable
and Extractable8
ho
i
TOC
Time: 11:30
6,475
20,979
Time: 14:00
3,506
13:30
7,202
18,324
15:00
4,334
15:40 11:30
6,110 21
18,975 <71
9/16/87
15:30 14:00
4,615 226
15:30 15:40
19 20
<32 <31
15:00 15:30
215 38
13:30
9,964
26,099
14:00 15:00
6,684 5,614
Total Purgeable
and Extractable*
23,020 21,053 20,296
<82
<82
<82
9,665
9,033
9/17/87
Time: 10:0012:00 14:00
10:00 12:00 14:00
10:00
12:00
TOC
Total Purgeable
and Extractable8
298 3,007 3,047
44.3 46.0 48.1
14,360 15,250 14,910
<48
<82
<82
5,358
11,471
6,128
9,026
Individual concentrations of compounds determined by EPA Method 624 and 625 are shown in Tables C-l through C-3.
a Total Purgeable and Extractable = E (Concentration in ug of Compound)(Weight of Carbon)(# of Carbons)/(MW of
Compound)
Tentatively identified compounds were not included in the total concentrations. Concentrations determined by TOC
and POC are shown in Table C-4 in Appendix C.
-------
7.0 QUALITY ASSURANCE AND QUALITY CONTROL
The overall QA goal was to determine steam stripper efficiency within +5
percent. As stated in the test plan, because of the orders of magnitude
difference between stripper feed and tails pollutant concentrations,
measurement uncertainties have little impact on overall accuracy. Standard
deviation(s) of the mean (z) was calculated for each pollutant removal
efficiency based on six measurements (only five measurements for the Method
625 compounds) and is reported in the table below. The mean VOC removal rates
were calculated on an arithmetic basis rather than the flow weighted basis
used in Section 2.1.
Compound X S
benzene 99.68 0.25
toluene 98.96 0.37
ethyl benzene 99.41 0.17
isophorone 94.70 3.82
naphthalene 99.63 0.23
s-ethyl-l,2-methylpyridine 89.12 14.20
1,2,3,4 tetrahydronaphthalene 98.73 2.10
acetophenone 96.48 2.90
2-methyl-l,3-cyclopentanedione 97.41 1.51
All standard deviations were less than 5% with the exception of 5-ethyl-l,2-
methylpyridine. Based on standard deviation, the primary QA goal was met on
each compound except the pyridine. Analytical QA procedures were required and
results are outlined in the following subsections.
7.1 Data Validation
Data were validated by reviewing quality control sample analytical results
and comparing them with pre-determined criteria. The following discussions
present the validation results and the criteria.
7.1.1 Accuracy
To determine the accuracy of any analytical method, an assumption must be
made that appropriate sampling techniques have been employed to provide for a
-26-
-------
representative sample. Assuming a representative sample has been collected,
the analytical accuracy is defined as the degree of agreement between an
accepted reference value and the true value.
A laboratory sample spiking program was conducted on the samples to
determine the accuracy of measurement methods for volatile and semi-volatile
organic compounds. Four samples were spiked. Only two samples were spiked at
correct concentration ranges. The other two spiked samples had to be diluted
before the final analysis; therefore, the spike concentrations were too low to
be properly quantified. The analytical results of the two correctly spiked
samples are listed in Table 7-1. As indicated in this table, the matrix spike
percent recovery (%R) was calculated as follows:
_
%R = -=r- x 100
where:
X = Analytical result from the spiked sample
T = Analytical result from the unspiked sample
K = Known value of the spike
%R = Percent recovery = {% Accuracy)
The calculated matrix spike percent recovery (%R) is listed below:
%R
#1 »2
1,1-Dichloroethene 70% 82%
Trichloroethene 92% 86%
Chlorobenzene 94% 96%
Toluene 96% 98%
Benzene 94% 88%
Average 89% 90%
These matrix spike percent recovery values are within the range of 60% to
145%, i.e., the acceptable range determined in the test plan for volatile
organic analysis {Method 624).
The matrix spike percent recovery values for the semi-volatile organic
compounds are:
-27-
-------
%R
#1 t2
1,2,4-Trichlorobenzene 80% 79%
Acenaphthene 81% 84%
2,4-Dinitrotoluene 38% 65%
Pyrene 115% 112%
N-Nitroso-Di-n-Propylamine 91% 89%
1,4-Dichlorobenzene 63% 62%
Pentachlorophenol 67% 78%
Phenol 85% 77%
2-Chlorophenol 91% 89%
4-Chloro-3-Methylphenol 76% 78%
4-Nitrophenol 76% 82%
Average 79% 81%
These matrix spike percent recovery values are within the acceptable range
for the semi-volatile organic compounds analysis noted in Table 6 of
Appendix A of Method 625. In conclusion, results from both analytical
measurement methods are accurate.
There were no matrix spikes performed for the total organic carbon (TOC)
or purgeable organic carbon (POC) analyses. No accuracy information can be
derived from this project for the total organic carbon (TOC) and purgeable
organic carbon (POC).
7.1.2 Precision
Precision is the measure of the mutual agreement between individual
measurements of the same property. Precision is best expressed in terms of
Standard Deviation or Relative Percent Deviation (RPD) and is inferred through
the use of duplicate samples. RPD for each component was calculated using the
following equation:
where:
RPD = Relative Percent Deviation
A = Replicate Value 1
B = Replicate Value 2
2 = Number of Duplicates
-28-
-------
The relative percent deviation (RPD) calculated for volatile and
semi-volatile organic compounds is also listed in Table 7-1. Among the 16 RPD
calculated, two were determined visually to be outliers (1,1,-Dichloroethene,
and 2,4-Dinitrotoluene). The mean and standard deviation of the RPD were
calculated and are shown below:
Volatile Organic Compound (624):
With outlier 6.80 5.72
Without outlier 4.50 2.89
Semi-Volatile Organic Compound (625):
With outlier 9.27 14.81
Without outlier 5.00 4.55
The precision of total organic compound (TOC) and purgeable organic
compound (POC) measurement methods was determined by analyzing duplicate
results. For total organic compounds three sets of duplicate results are
available:
#1 #2 RPD
RPAC ST-1 19.0 mg/L 18.0 mg/L 2.1%
RPAC ST-12 38.0 rog/L 38.0 mg/L 0%
RPAC SO-16 6,128.0 mg/L 5,697.0 mg/L 3.6%
Since only limited RPD data were available, no further statistical analysis
was performed.
For purgeable organic compounds, there are also three sets of duplicate
data available:
#1 #2
RPAC ST-1 <0.1 <0.1
RPAC ST-12 <1.0 <1.0
RPAC SO-16 <1.0 <1.0
-29-
-------
TABLE 7-1
WATER MATRIX SPIKE/MATRIX SPIKE DUPLICATE RECOVERY
I
OJ
o
Cone.
Allowable
QC Limits*
Spike Added Sample
Sample No.
RPACST-4
Compound
1 , 1-Dichloroethene
Trichloroethene
Chlorobenzene
Toluene
Benzene
(pg/L)
50
50
50
50
50
Result
ND
ND
ND
ND
ND
Cone.
MS
35
46
47
48
47
%
Recovery
70
92
94
96
94
Cone.
MSD
41
43
48
49
44
%
Recovery
82
86
96
98
88
Cone.
RPD
16*
7
2
2
7
RPD
14
14
13
13
11
Recovery
61-145
71-120
75-130
76-125
76-127
* Asterisked values are outside QC limits:
RPD: VOAs 1 out of 5 ; outside QC limits
RECOVERY: VOAs 0 out of 10 ; outside QC limits
-------
No meaningful data analysis can be performed because all results are lower
than the detectable limits.
7-2 Calibration Procedures and Frequency
All analytical instruments, including the GC/MS, TOG and POC systems, were
calibrated in accordance with the procedures listed in the test plan, and
following the guidelines of the referenced EPA methodology.
7.3 Sample Custody
The purpose of using chain-of-custody procedures at RPAC was to ensure the
integrity of the sample from the time of collection through data reporting.
Custody records trace each sample from collection through all transfers of
custody from shipment to arrival at the analytical laboratory. Internal
laboratory records then document the custody of each sample throughout the
analysis and its final disposition. Use of chain-of-custody procedures are
described in the sections that follow.
7.3.1 Chain-of-Custody Form
A chain-of-custody form (Figure 4} accompanied each group of samples
collected for laboratory analysis. One person was responsible for maintaining
the samples at any given time. Upon shipment of the samples to the analytical
laboratory, both the person relinquishing the sample and the person receiving
the samples for the shipper signed and dated the chain-of-custody form.
Responsibility for the samples then passed on to the receiver. Copies of
chain-of-custody forms are in Appendix E.
7.3.2 Chain-of-Custody Tape
Chain-of-custody tape was used to detect unauthorized tampering of samples
prior to and during shipment. The tape contained the following information:
-31-
-------
Environmental
Consultants, Inc.
CHAIN OF CUSTODY FORM
N2
PROJECT NO.:
SAMPLERS (SIGNATURE):
PROJECT NAME:
SAMPLE NO.*
DATE
TIME
*SOURCE Teat Pit So»l...(TP)
CODES: Ground Water. ..(GW)
RELINQUISHED BY (SIGNATURE).
RELINQUISHED BY (SIGNATURE).
RELINQUISHED BY (SIGNATURF1:
DATE/TIME:
DATE/TIME:
DATE/TIME:
COMP
GRAB
NO.
OF
CON-
TAINERS
/
// f/
/
ay «/ REMARKS
Boring Hole Soil. ..IBM) Surface So!l...(SS) Waate...(W)
Surface Water. ..(SW) Sediment ...( SD) Waate Water...
RECEIVED BY (SIGNATURE).
RECEIVED BY (SIGNATURE):
RECEIVED FOR LABORATORY BY
(SIGNATURE):
RELINQUISHED BY (SIGNATURE):
RELINQUISHED BY (SIGNATURE):
DATE/TIME:
1
Distribution Original accompanies shipment, copy to coordinator field files
REMARKS:
DATE/TIME:
1
DATE/TIME:
(WW)
RECEIVED BY (SIGNATURE):
RECEIVED BY (SIGNATURE!
NJ
I
Figure 4 - Sample Chain of Custody Form
-------
• Project Name/Number
• Date
• Signature
• Printed Name
The person who had possession of the samples filled out the tape and placed it
on the sample cooler at times the samples were not in view, and during
shipment to the analytical laboratory. The tape was attached in such a way
that it was necessary to tear it in order to open the cooler.
7.3.3 Shipping of Samples
The samples were delivered to the laboratory for analysis within 24 to 48
hours after sampling. They were accompanied by the chain-of-custody record
and delivered to the person in the laboratory authorized to receive samples.
All material was transported as required in 49 CFR Subpart B Section 172,
which lists and classifies those materials which the Department of
Transportation has designated as hazardous for purposes of transportation. It
also details the requirements for shipping, marking, labeling and
transportation of these materials.
Upon receipt of the samples, the laboratory inspected the chain-of-custody
seal for signs of tampering, reconciled the information on the
chain-of-custody form with the sample label, and inspected the sample
container for leakage.
7.4 Deviations from the Test Plan
Some deviations in the test plan procedures were implemented in the
program. These are listed as follows:
-33-
-------
• Due to process delays on September 14 there were fewer numbers of
TOC/POC samples collected.
• Graduated cylinder measurements of phase separation in the
stripper overheads was not conducted.
• Temperatures of wastewater samples at the heat exchanger outlets
were less than 20°C, not 4°C as required by the test plan.
• Documentation of flow orifice calibrations were not obtained.
• Did not preserve EPA 624 samples with I^SO^ to a pH of < 2 as
stated in the test plan.
These deviations did not have significant effects on program goals. The
higher temperatures encountered on the sampling heat exchanger outlets may
have contributed to some increase in VO losses.
-34-
------- |