Technology Evaluation Report:
BIOLOGICAL TREATMENT OF WOOD PRESERVING
SITE GROUNDWATER BY BIOTROL, INC.
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
NOTICE
The information in this document" has been funded by the U.S.
Environmental Protection Agency under the auspices of the Superfund
Innovative Technology Evaluation (SITE) Program under Contract
Nos. 68-03-3485 and 68-CO-0048 to Science Applications Inter-
national Corporation. It has been subjected to the Agency's peer
and administrative review, and it has been approved for publication.
as an EPA document. Mention of trade names or commercial products
does not constitute an endorsement or recommendation for use.
-------
FOREWORD
The Superfund Innovative Technology Evaluation (SITE) program was
authorized in the 1986 Superfund amendments. The program is a joint
effort between EPA's Office of Research and Development and Office
of Solid Waste and Emergency Response. The purpose of the'program
is to assist the development of hazardous waste treatment
technologies necessary to implement new cleanup standards which
require greater reliance on permanent remedies. This is
accomplished through technology demonstrations designed to provide
engineering and cost data on selected technologies.
This project consisted of a demonstration of BioTrol, Inc.'s fixed-
film, amended biological treatment process and a careful and
extensive analysis of the effectiveness of the system. The study
was carried out at the MacGillis and Gibbs Company site in New
Brighton, Minnesota, where wood preserving operations have been
carried out over several decades using first creosote, later
pentachlorophenol, and most recently, chromated copper arsenate. In
1984 the ^site was added to the National Priorities List as one
where soil and groundwater were contaminated with hazardous
chemicals. Documentation for the project consists of two reports.
The field activities and analytical results of the study are
summeirized in this Technology Evaluation Report. The companion
Applications Analysis Report analyzes the broader applicability and
economics of the biological treatment process for the elimination
of pentachlorophenol and polynuclear aromatic hydrocarbons from
groundwater.
For further information, please contact the Superfund Technology
Demonstration Division at the Risk Reduction Engineering
Laboratory.
E. Timothy Oppelt, -Director
Risk Reduction Engineering Laboratory
iii
-------
ABSTRACT
The BioTrol, Inc. Aqueous.. Treatment S_ystem (BATS), a fixed-film,
aerobic biological treatment process for contaminated groundwaters
and other wastewaters, was evaluated at three different throughput
rates, each maintained for 2 weeks to provide steady-state
conditions for both operation and sampling.
This report presents a detailed description of the process and the
system and provides detailed results of the six weeks of monitoring
at the MacGillis and Gibbs Company wood preserving site in New
Brighton, MN. Technological effectiveness of the process is
assessed on the basis of an extensive analytical program coupled
with a quality assurance program. The economics of the process were
also assessed.
From the results of the pilot scale demonstration study it is
concluded that (a) the fixed film aerobic process is effective at
degrading pentachlorophenol, achieving more than 96% removal; (b)
effluent concentrations of pentachlorophenol well below 1 ppm are
attainable by controlling throughput rate; (c) removal of PGP is
largely by mineralization to carbon dioxide, water, and salt; (d)
acute toxicity of the groundwater to minnows and water fleas is
eliminated, and (e) operating cost is about $3.45/1000 gal at 5 gpm
and would decrease to $2.43/1000 gal at 30 gpm.
IV
-------
CONTENTS
Page
Foreword " iii
Abstract iv
Figures vi
Tables vii
Abbreviations and Symbols ix
Conversion Factors xi
Acknowledgements xii
1. Introduction 1
2. Process Description 4
Introduction 4
Process Description 6
Site Characteristics 11
3. Field Operations 13
Predemonstration Well drilling 13
Field operations 16
4. Sampling and Analysis Program 22
5. Demonstration Results 26
Introduction 26
Results 26
System Parameters 27
Pentachlorophenol 34
Polynuclear aromatic hydrocarbons 43
Dioxins/furans 45
Heavy metals 46
Volatile organics 47
Biomonitoring 47
6. Economics 48
Introduction 48
Conclusions . . . . 48
Operating Costs 48
Hypothetical Case Studies 51
7. Conclusions 53
8. Bibliography 54
Appendix
Quality Assurance Review 57
-------
FIGURES
Figure Page
1 PVC Media - as used and schematic 5
2 BioTrol's Mobile Treatment System Trailer 8
3 Schematic of Bioreactor 9
4 Carbon Adsorber 10
5 Site layout - groundwater wells shown 12
6 Schematic of wellhead . 17
7 Operating and sampling schematic 24
-------
TABLES
Table , Page
1 PGP Concentrations in Well #1 Borehole 14
2 Analysis of Well #1 Groundwater after Development 15
3 Analytical Results on Groundwater during Pump Test 15
4 Field Measurements 19
5 Sampling Schedule, methods 23
6 Nitrite/nitrate Analytical Results 28
7 Ammonia Analytical Results 29
8 Phosphate Analytical Results 30
9. Temperature Summary Across BioTrol System 31
10 Average TSS and Oil Across BioTrol System 31
11 Oil and Grease Analytical Results 32
12 TSS Analytical Results 33
13 VSS/TSS Analytical Results 35
14 Average PGP Removal by BioTrol System 34
15 Pentachlorophenol Analytical Results 37
16 Mass Removal of Pentachlorophenol 38
17 Soluble PGP/Total PGP Analytical Results 39
18 Comparison of Chloride and TOG Changes 40
19 Chloride Analytical Results 41
20 TOG Analytical Results 42
21 Potential Chloride Contributions from Partially
Chlorinated Phenols 43
vii
-------
TABLES (cont'd)
Table Page
22 Sludge Analysis Results 44
23 PAHs in Air Emissions from Bioreactor 45
24 Dioxins/Furans Found in System 45
25 Dioxins/Furans Found in Sludge 46
26 Metals Found in System 46
27 Acute Biotoxicity of Groundwater
and Treated Effluent 47
28 Operating Costs 49
29 Estimated Costs of Treating Various Wastewater
volumes . 52
30 Treatment Costs over Five Years 52
viii
-------
ABBREVIATIONS AND SYMBOLS
BATS BioTrol Aqueous Treatment System
BGS below ground surface
BOD biochemical oxygen demand (mg oxygen/liter)
BTEX benzene, toluene, ethyl benzene, and xylenes
CERCLA Comprehensive Environmental Response, Compensation, and
Liability Act of 1980
cfm cubic feet per minute
COD chemical oxygen demand (mg oxygen/liter)
EMSL Environmental Monitoring Systems Laboratory
GC/MS gas chromatograph/mass spectrometer
gpd gallons per day
gpm gallons per minute
HPLC high pressure liquid chromatography
HSWA Hazardous and Solid Waste Amendments to RCRA - 1984
kwh kilowatt-hour
LC(50) Lethal concentration to 50% of a test species
population
Mgd million gallons per day
mg/L milligrams per liter .
ng/kg nanograms per kilogram
ng/L nanograms per liter
NPL National Priorities List
O/G oil and grease
ix
-------
ORD Office of Research and Development
OSHA Occupational Safety and Health Act/Administration
OSWER Office of Solid Waste and Emergency Response
PAHs polynuclear aromatic hydrocarbons
PCP pentachlorophenol
PEL Permissible Exposure Limit
POTW publicly owned treatment works
ppb parts per billion
ppm parts per million
psi pounds per square inch
PVC polyvinyl chloride
QA/QC quality assurance/quality control
RCRA Resource Conservation and Recovery Act of 1976
RI/FS Remedial Investigation/Feasibility Study
RREL Risk Reduction Engineering Laboratory
SARA Superfund Amendments and Reauthorization Act of 1986
SITE Superfund Innovative Technology Evaluation
TCPs tetrachlorophenols
TOG total organic carbon (mg carbon/liter)
TSS total suspended solids (mg solids/liter)
-------
CONVERSION FACTORS
English (US)
x
Factor
Area:
Flow Rate:
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
ft*
in2
gal/min
gal/min
Mgal/d
Mgal/d
Mgal/d
ft
in
Yd
Ib
Ib
ft3
ft3
gal
.gal
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
9.2903 X 10"2
6.4516
6.3090 x 10'5
6.3090 x 10'2
43.8126
3.7854 X 103
4.3813 X 10'2
0.304S
2.54
0.9144
4.5359 X 102
0.4536
28.3168
2.8317 X 10"2
3.7854
3.. 7854 X 10'3
Length:
Mass:
Volume:
ft = foot, ft2 = square foot, ft3 = cubic foot
in = inch, in2 = square inch
yd = yard
Ib = pound
gal = gallon
gal/min (or gpm) = gallons per minute
Mgal/d (or MGD) = million gallons per day
m = meter, m2 = square meter, m3 = cubic meter
cm = centimeter, cm2 = square centimeter
L = liter
g = gram
kg = kilogram
xr/s == cubic meters per second
L/s = liters/sec
m3/d == cubic meters per day
= Metric
m2 '
= cm2
m3/s
L/s
L/s
ra3/cL
m3/s
= m
= cm
= m
g
kg
xi
-------
ACKNOWLEDGEMENTS
This report was directed and coordinated by Mary K. Stinson, EPA
SITE Project Manager in the Risk Reduction Engineering Laboratory -
Cincinnati, Ohio.
This report was prepared for EPA's Superfund Innovative Technology
Evaluation (SITE) Program by Herbert S. Skovronek and William Hahn
of Science Applications International Corporation for the U.S.
Environmental Protection Agency under Contract Nos. 68-03-3485 and
68-CO-0048.
The cooperation 'and participation of Thomas J. Chresand and
supporting staff of BioTrol, Inc. throughout the course of the
project and in review of this report are gratefully acknowledged,
as is the assistance of A.J. Bamby of MacGillis and Gibbs.
Mark Lahtinen of the Minnesota Pollution Control Agency (MPCA) and
Rhonda McBride, Linda Kern, and Darryl Owens, the Remedial Project
Managers of USEPA's Region V provided invaluable assistance and
guidance in initiating the project and in interpreting and
responding to regulatory needs of the project.
Ronald Lewis and Gordon Evans of USEPA's Risk Reduction Engineering
Laboratory and Linda D. Fiedler of the Technology Innovation Office
of OSWEK provided invaluable reviews of the draft report.
Finally, the project could not have been carried out without the
tireless efforts of the many SAIC and Radian field and laboratory
personnel.
xii
-------
SECTION 1
INTRODUCTION
In response to the Superfund Amendments and Reauthorization Act of
1986 (SARA), EPA's Office of Solid Waste and Emergency Response
(OSWER) and the Office of Research and Development (ORD)
established the Superfund Innovative Technology Evaluation (SITE)
Program. The goal of this program is to promote and accelerate
the development of innovative technologies for consideration in the
clean up of Superfund sites across the country.
The SITE Program seeks to meet new federal and state cleanup
standards by providing permanent remedies to waste problems, rather
than temporary measures. Such methods include destruction,
stabilization, and treatment processes that will assure - to the
maximum extent possible - that problems will not resurface in the
future.
The SITE Program is composed of two major elements: the
Demonstration Program and the Emerging Technologies Program. The
focus of the Demonstration Program is to provide reliable
engineering and cost data based on field tests of selected
technologies. The Emerging Technologies Program fosters the
investigation and development of technologies now at the laboratory
scale.- A third component of the SITE Program, the Measurement and
Monitoring Technologies Program, assists in the development of
innovative technologies to better characterize Superfund sites.
In this case, EPA was able to carry out two demonstration projects
and the assessment of a measurement and monitoring methodology at
one facility, thereby making maximum use of personnel and
resources. Documented in this report are the sampling and analyses
results used to evaluate the treatment of contaminated groundwater
by a fixed film bioreactor. The other two research efforts, soil
washing and an Immunoassay analytical technique for
pentachlorophenol, are briefly noted and are documented in separate
reports.
PROJECT OBJECTIVES
The major objectives of this SITE Demonstration Project were to:
-------
1. Determine the capability of the BioTrol technology to remove
pentachlorophenol (PGP) from groundwater at a wood treatment
facility, where a range of other contaminants may be present;
2. Determine the effect of throughput rate (flow) on removal
efficiency;
3. Determine the capability of the BioTrol process to remove
polynuclear aromatic hydrocarbons from a wastewater
contaminated by creosote; and
4. Determine the cost of the treatment and the factors that
impact on the cost of the process.
DESCRIPTION OF PROJECT
Biological treatment has a long history of use for a wide variety
of municipal and industrial wastewaters. Only recently has it
become apparent that even organics thought to be highly resistant
to biodegradation, such as the chlorinated phenols used to impart
rot resistance to wood, could be degraded biologically, under
carefully selected conditions.
On the basis of fundamental research over the last several years,
BioTrol, Inc. developed an innovative process to carry out such a
biological degradation process compactly, efficiently, and
economically. The process is carried but aerobically in stages on
an inert support matrix. Local soil bacteria already able to resist
and even degrade pentachlorophenol are supplemented or "amended"
with an inoculum of a pentachlorophenol-specific bacterium to give
maximum biodegradation in the first stage, where the concentration
of the pentachlorophenol is highest. The staged arrangement of the
technology then allows other constituents to be degraded in the
latter stages of the reactor, after the "toxic" pentachlorophenol,
has been reduced.
For the demonstration project, three different throughput rates
were examined over the course of about six weeks to evaluate the --
effect of flow rate on the efficiency of treatment, particularly in
terms of pentachlorophenol removal. An extensive sampling ^and
analysis program defined by an EPA Category II Quality Assurance
Project Plan (QAPP) was an integral part of the project.
The site selected for the demonstration was a wood preserving
facility in New Brighton, .Minnesota. Earlier screening of the site
as part of a Remedial Investigation/Feasibility Study indicated
significant contamination of both soil and groundwater with
pentachlorophenol and polynuclear aromatic hydrocarbons as the
result of many years of operation with creosote, pentachlorophenol.
Currently, chromated copper arsenate (CCA) is used and waste
-------
management is consistent with current regulatory requirements. The
site had been added to EPA's National Priorities List in 1984.
This report and the companion Applications Analysis Report
document, analyze, and interpret the results of the SITE
demonstration project.
PROJECT ORGANIZATION
Through a Cooperative Agreement between EPA and BioTrol, Inc., the
developer was responsible for operating its equipment while EPA and
its contractor, Science Applications International Corp. (SAIC),
located and installed a suitable well to supply the groundwater for
the project, prepared the demonstration plan, designed the sampling
plan, conducted sampling and analysis, evaluated the data, and
prepeired the final reports. In addition, EPA's contractor was
responsible for pre-demonstration sampling of the groundwater, for
managing public information meetings, and for site decontamination.
-------
SECTION 2
PROCESS DESCRIPTION
INTRODUCTION
Biological processes have been widely used for many years in the
treatment of industrial and municipal wastewaters, with aerobic
treatment being the most widely used technology. As industrial
products have been developed to provide resistance to degradation
by the environment, it has often been assumed that these chemicals
also would be resistant to conventional biodegradation. It has now
been recognized that such is not the case and, using proper
procedures and suitable biological populations, efficient
biodegradation of many organic chemicals, including chlorinated
aromatics such as pentachlorophenol, can be achieved.
To provide the most efficient and cost effective treatment of
wastewaters containing such contaminants, BioTrol, Inc. has
developed a proprietary process based on amended fixed-film aerobic
treatment. Using a multi-stage reactor, an initial biogrowth is
developed on an inert support matrix such as corrugated polyvinyl
chloride sheets (Figure 1) using indigenous bacteria from a source
such as soil at the site. This bacterial population, having been
derived from the local soil, is resistant to the toxicity of the
local contaminants and may even have a population distribution
which favors the destruction of chemicals from the site. After this
bacterial source has been allowed to acclimate on -the matrix in
the presence of nutrients, an inoculum of a Flavobacterium specific
to the target chemical, pentachlorophenol, is added. Further
acclimation then is allowed to occur in a total recycle mode before
the system is ready for once-through treatment of the groundwater.
The design of the BioTrol system allows the development of the
largest concentration or population of bacteria capable of
degrading pentachlorophenol in the first chamber, where the
concentration is highest. As the wastewater flows through the
reactor and the pentachlorophenol concentration decreases, other
bacteria more suitable to degradation of other contaminants, but
perhaps more sensitive to deactivation by pentachlorophenol, have
the opportunity to grow and consume those contaminants.
The BATS fixed-film process has a further advantage in that very
little sludge is generated and, consequently, sludge management is
not a significant factor in operation.
-------
BLOCKS
CROSS-STACKED
FIGURE 1. CORRUGATED POLYVINYL CHLORIDE MEDIA
'5
-------
BioTrol Soil Washing Process
In addition to the BioTrol Aqueous Treatment System (BATS),
BioTrol, Inc. also has developed a soil washing process that
separates coarser, relatively uncontaminated soil from more heavily
contciminated fine material such as clay. PCP contamination of the
fines; then can be treated by other methods such as biological
degrsidation in a slurry bioreactor. The relatively clean washed
soil and even the slurry reactor-treated fines may then be returned
to the site. The BATS which is the subject of this report also is
employed in the soil washing sequence to treat the washwater used
to separate the sand and fine materials. The Soil Washing process
is the subject of a separate SITE Demonstration Program at the
MacGillis and Gibbs facility and is reported separately.
Field Immunoassay for Pentachlorophenol
While this project was being developed, EPA's Environmental
Monitoring Systems Laboratory (EMSL) in Las Vegas was searching for
a facility where a new immunoassay method for field monitoring
pentaichlorophenol in wastewaters could be studied under real-world
conditions. The groundwater study at the MacGillis and Gibbs
facility offered an ideal environment to evaluate this technique in
parallel with the extensive analyses being done as part of the
demonstration. Consequently, arrangements were made to have EPA's
contractor carry out the field test as part of their assignment.
The immunoassay test, developed by Westinghouse, is based on the
inhibition of bacterial enzyme activity when a target chemical, in
this case pentachlorophenol, is present. The inhibition is
observed by a color change in a reagent matrix and is readily
quantified using standards. Only very small samples of material are
needed and the results are generated in about an hour. Field
experience was needed to learn how well results correlated with the
standard GC/MS analysis and larger, laboratory-scale immunoassays;.
how sensitive the test was with real-world wastewater matrices, and
how conveniently and reliably it could be carried out in the field
by relatively inexperienced personnel.
Reference is included to make the reader aware of this effort and
the potential availability of the method. A report documenting .the
procedure and results of this study will be available separately.
PROCESS DESCRIPTION
The BioTrol equipment used in this demonstration consists of a
single mobile trailer (20 ft. in length) fitted with all the
necessary tanks, pumps, etc. (Figure 2). A level area (ideally a
concrete apron) about 50 x 50 ft is needed to support the trailer
and auxiliary facilities. The capacity of the mobile system is
about. 5 gpm, depending on the concentration of the pollutants to be
-------
degraded; and its hydraulic capacity is about 10 gpm. Contaminated
water is brought to the 100 gallon mixing or conditioning tank
where pH is automatically adjusted to slightly above 7.0 by
metering in a caustic solution. A solution of nitrogen and
phosphorus, nutrients (urea plus trisodium phosphate) is also
metered into the conditioning tank at a pre-determined rate. For
this; study a mixture of 2.5 Ibs of urea and 5.0 Ibs of trisodium
phosphate in 50 gal of water was metered in at a rate of 2.5 ml/gal
of wastewater. The mixed water passes through a heat exchanger and
then a heater which is available to elevate the temperature to
about 70°F (21°C) , which is considered an optimal temperature for
biological treatment.
The influent is then introduced into the base of the first of the
3-celled biological reactor (Figure 3) and moves from cell to cell
by means of underflow weirs-. In the pilot unit each cell,
containing the PVC matrix, has approximately 160 gallon capacity,
giving a total system capacity of about 500 gallons. Thus, at 5
gpm the retention time is approximately 1.7 hours; at 3 gpm the
retention time would be 2.7 hours and at 1 gpm the retention time
would be 8.3 hours. Air is continuously injected to the base of
each chamber and distributed by a network of sparger tubes to
maintain sufficient dissolved oxygen (DO) to support aerobic
conditions for biodegradation (about 5 ppm). The combination of the
flow, the air sparging, and the design of the plastic media are
such that upward and lateral distribution of the water and the air
occur in each chamber. After moving through the three chambers, the
effluent exits from the third chamber via an overflow weir.
While the process is claimed to be relatively insensitive to
suspended solids and dispersed oil and the mobile trailer is not
equipped to remove these contaminants, oil/water separation or
solids removal could be added externally if necessary, depending on
the quality of the wastewater being treated. Such pretreatment was
deemed unnecessary with the MacGillis and Gibbs groundwater.
Similarly, while BioTrol's experience has been that post-treatment
such as suspended solids removal or carbon polishing are not
usucilly necessary,- EPA made the decision that a small bag filter
and a carbon adsorber would be added to the effluent line to
assure that the discharge to the Minneapolis Metropolitan POTW
would be of high quality. The bag filter was installed in the
effluent line leaving the bioreactor and was followed by a 50 cubic
foot carbon adsorber (Figure 4). EPA also chose to install a
smaller carbon adsorber on the air exhaust line exiting from the
bioreactor to assure that no hazardous volatile chemicals were
emitted. During the course of the demonstration, analyses were
carried out before and after both carbon adsorbers to assess the
need for such protective devices.
-------
HI
fe
LU
I
CO
I"
2:
UJ
LLJ
a:
1--
O!)
s
c!
<
LLI
5i
o
q
O
DC:
m
c\i
LLI
cc:
ID
8
-------
i
!•*
11
7*r A --I...
/^>.-.-.-.-.-----
o
JJ
o
CO
•
CO
3
•H
fa
i
-------
7'-6"
L
CD
3'-9"
-------
SITE CHARACTERISTICS
In the demonstration program the mobile pilot scale system requiired
only a level base (ideally a concrete pad) ; potable water and power
for pumps, the heater, lighting, etc. were supplied by the facility
operator. If necessary at other sites, potable water could be
trucked in and power could be provided by an on-site generator.
The MacGillis and Gibbs Company facility has been used for wood
preserving for several decades. Originally creosote was used in the
treatment; subsequently, in the 1950's, pentachlorophenol in oil
was substituted. Operations were carried out in open troughs,
presenting many opportunities for spills and leaks. Treated lumber
was stored on the site where further contamination could occur as
the result of rain wash-off. In addition, for a period in the 50's,
waste pentachlorophenol solution was used for weed control on the
site. As shown in Figure 5, the southwestern section of the
MacGillis and Gibbs property where disposal has frequently taken
place also collects water and forms a pond. There has been
suspicion that some of the contamination in this area may be due to
run-off from the adjacent Bell Pole property. In the 1970's,
MacGillis and Gibbs replaced pentachlorophenol with the chromated
copper arsenate process and substituted closed reactors for the
open troughs, thus minimizing the opportunities for inadvertent
spills and leaks.
The facility originally came to the attention of the EPA as part of
an investigation of the neighboring Bell Pole facility where wood
treatment was also being carried out. Ultimately, as the result of
an RI/FS,it was concluded that the soil and groundwater at both
sites was contaminated with pentachlorophenol and lesser
concentrations of polynuclear aromatic hydrocarbons. Both sites
were placed on EPA's National Priorities List in 1984.
MacGillis and Gibbs has a discharge permit for discharge of its
sanitary wastewater to the Minneapolis/St. Paul Treatment Works. An
agreement was reached for the POTW to accept treated effluent from
the bioreactor on the condition that it contained <2 ppm of PCP.
11
-------
CCA
AREA
n
i PCP
11.
PROCESS
AREA
B!:LL
POLE •
FIGURE 5. MACGILL1S & GIBBS SITE
12
-------
SECTION 3
FIELD OPERATIONS
PREDEMONSTRATION WELL DRILLING
Before the demonstration project could be initiated or a decision
reached as to whether this facility was appropriate for evaluation
of the process, it was necessary to establish whether a suitable
grouridwater was available. Criteria established by the developer
were (a) up to 100 ppm of pentachlorophenol; (b) flowrate of at
least 5 gpm, and (c) some concentration of polynuclear aromatic
hydrocarbons.
Reports by Twin Cities Testing Corp., a Remedial Investigation (RI)
contractor, documented soil borings at the site which strongly
suggested that suitable groundwater was present in the vicinity of
the disposal area and pond in the southwestern portion of the site.
Several of these borings had been found to contain very high levels
of PCP and PAHs, but some had also indicated high concentrations of
oil. Based on these reports, a 4-inch well (#1) was drilled by a
contractor in what appeared to be an attractive area, as shown in
Figure 5. The well was drilled to a depth of 45 ft. and screened
from approximately 30 ft to 45 ft. Samples were taken at 5 foot
intervals as the well was drilled and rapidly analyzed (24-hour
turnaround) using BioTrol's High Pressure Liquid Chromatography
(HPLC) method so that the operation could be terminated if the well
proved to be unsuitable in PCP analysis.
The results of the initial analyses (Table 1) showed considerably
less PCP than anticipated; while not quantified, PAH concentrations
also appeared to be quite low.
13
-------
TABLE 1. PGP CONCENTRATIONS IN WELL #1 BOREHOLE
depth
(ft)
10
15
20
25
30
35
40
PCP cone.
(ppm)
12.1
19.6
18.
13.4
29.2
31.4
45.7
Based on these initial results, and to minimize the risks
associated with relying on a single well for the demonstration
project, a second well was then drilled approximately 65 feet north
of Well #1, in an area where pentachlorophenol contamination also
was expected to be high. Starting at about 10 feet, considerable
contamination with an oil-like material was observed in this well
and groundwater was not detected in the borehole even down to 20
feet. A static water level of 11 feet below ground surface (BGS)
was observed after operations had been left for the weekend and a
floating liquid layer was also clearly evident. Drilling was
continued to a depth of 46 feet where clay-like material was
encountered. Evidence of oil was observed at. several different
depths.
Two groundwater samples were analyzed from Well #2, believed to be
representative of. 15 ft BGS and 45 ft BGS. Pentachlorophenol
concentrations at these depths were 97.8 and 5.7 ppm, respectively;
only traces of PAHs were detected in the 45 ft sample.
Each well was then developed using a surge block and submersible.
pump, continuing the pumping until the water was turbidity-free and
field measurements of pH, temperature, and conductivity had
stabilized to within 10%. A sample of Well #1 was then taken using
a Teflon bailer and submitted for extensive analyses. The results <
of those analyses are shown in Table 2.
14
-------
TABU! 2. ANALYSES OF WELL #1 GROUNDWATER AFTER DEVELOPMENT
Constituent Concentration Units
Volatile Organics
methylene chloride
toluene
Semivolatile Organics
pent achl or opheno 1
2 -methyl phenol
4 -methyl phenol
2, 4 -dimethyl phenol
naphthalene
2 -methyl naphthalene
Metals
leeid
zinc
Total Phenol ics
TOG
Oil/Grease
Chloride
Dioxins
5*, 5*
6, 5*
15, 13
20, 23
20, 24
32, 33
86, 90
50, 52
8
280
0.859, 0.706
51
14, 10
23, 30#
All
-------
Slug tests were also carried out on both wells using a Hermit data
logger to provide additional flow rate data. Analysis of the data
indicated flow rates of 28 gpm and 3 gpm for Well #1 and Well #2,
respectively.
On the basis of the foregoing data, a decision was made to proceed
with the demonstration project using the groundwater from Well #1
as the feed water to the BioTrol system. The well was cased and
protected as shown in Figure 6.
FIELD OPERATIONS
Once the well had been selected and constructed, a 1 inch PVC line
running some 800 feet to the pilot BATS unit was installed. The
line was buried where it traversed railroad tracks and a road to
avoid accidental rupture and interruption of operations. A-
submersible pump was used to withdraw the groundwater.
In anticipation of start-up, BioTrol personnel acclimated the
bioreactor to the groundwater contaminants, first with closed loop
recycle of the water from the well development and pump test, which
had been stored in tanks adjacent to the reactor trailer.
Acclimation consisted of introducing air, nutrients, arid a slurry
of soil from the site to supply indigenous bacteria. After about
one week of closed loop operation with this water source, the
flavobacterium specific to pentachlorophenol was introduced to the
conditioning tank and acclimation was continued. Delays in start-up
of the demonstration phase of the program made it necessary for
BioTrol to continue operation for several weeks, ultimately using
the groundwater directly from the well. In effect, this placed the
system in steady state operation even before the demonstration was
initiated.
A separate trailer was brought on-site for contractor personnel to,
use in preparing samples and related documentation (and in which to
carry out the immunoassay tests for EPA-EMSL/Las Vegas). Composite
samplers were installed adjacent.to the reactor trailer to collect
22-hour samples from the influent, each cell of the bioreactor, and ,_
the effluent. Procedures were reviewed to assure that all sampling
could be carried out as planned. The bag filter and the carbon
adsorbers for the effluent and for the combined exhaust from the
cover over the bioreactor were connected.
On July 24, 1989 the project was officially initiated at an initial
flow rate of 1 gpm. Nutrient solution was continuously added to the
conditioning tank and pH was adjusted to about 7.3 by the addition
of 50% caustic solution. Some problems were initially encountered
due to foaming in the bioreactor, but this soon subsided. Both
composite and grab samples were collected as scheduled over the
course of two weeks of operation at this flow
16 .
-------
Vented Stainless
Steel Cap
Steel Casing with
Cap & Lock
3"x5' Steel
Guard Post
Land Surface
4" I.D. Schedule 40 Type
304 Stainless Steel Riser
Cement Bentonite Grout
8" Borehole
Bottom of
Borehole 5®
Bentonite Pellet Seal
Sand Pack
4" I.D. Schedule 40
Type 304 Stainless
Steel Screen
(0.020" Slot)
Flush Plug
NOT TO SCALE
FIGURE 6. WELL CONSTRUCTION
17
-------
rate. System parameters (temperature, pH, dissolved oxygen, and
flow rate) were measured to assure that the system was operating
within planned parameters. The summary of the system data for the
entire project is presented in Table 4.
After two weeks of operation at 1 gpm, the flow rate was increased
to 3 gpm on August 6, 1989 by changing the discharge pump rate. One
24-hour period was allowed for equilibration and return to steady
state before sampling was restarted. Operation was again continued
for two weeks, to August 19, 1989.
On August 20, 1989, flow was again increased, to 5 gpm, and sample
collection resumed after one day of equilibration.
On September 1, 1989, the demonstration phase was completed.
Because of the long delay in receiving analytical results for key
parameters, it was not feasible to review data as a means of
deciding whether any segment of the project needed to be repeated.
Because the demonstration project took place during the summer
months, it was unnecessary to use the heater in the system.
[Originally, plans for the project had called for "spiking" of the
groundwater with a much higher concentration of PCP to determine
the impact of such shock loading. However, all parties agreed that
(a) such shock loading was not reasonable to expect from a
groundwater, and (b) dilution of such a spiked feedwater with
effluent or potable water would be a viable means of preventing
shock loading. In addition, plans called for the unit to be used in
the immediate future for treatment of wastewaters from the soil
washing study. If successful, the shock loading experiment would
have deactivated the biological growth and necessitated re-
acclimation. Since this was not desirable, the spiking segment of
the study was not undertaken.]
Normally, the system would have been decommissioned after
completion of the demonstration project; however, plans called for
the unit to be used for treating process waters from the soil
washing demonstration. It was left in a "stand-by" mode, operating
at a low flow rate on groundwater, until the second project was
operational. When that project was completed, the BATS was drained,
electrical and water- hook-ups disconnected, and the PVC matrix
removed and discarded. Wastes that were collected during the
project included sludge, analytical wastes such as cleaning
solvents, and contaminated protective gear. Liquids and solids were
stored, segregated, in 55 gal drums for off-site disposal as
hazardous wastes. The carbon from both adsorbers was removed to be
thermally regenerated.
18
-------
TABLE 4. SYSTEM FIELD MEASUREMENTS
DAY
OF
RUN
1
2
3
4
5
6
7-
3
9
10
11
12
1
2
3
4
5
6
7
8
9
10,
11
12
1
2
3
4
5
6
7
8
9
1O
11
12
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
ST1-C-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
Tl-C-06
Tl-C-07
Tl-C-08
Tl-C-09
Tl-C-10
Tl-C-11
Tl-C-12
PROCESS
FLOW
( MGD )
0.75
0.95
1.25
NA
NA
0.97
0.92
0 . 96
0.88
0.87
1.1
1.24
2.74
2.98
2.71
3.15
3.04
2.88
2.97
3.04
3.04
2.81
2.99
3. .24
NA
NA
4.96
5.24
5.17
5.21
5.05 .
5.04
5.21
4.84
4.97 .
5.06
pH ( s.u. )
""'•" SAMPLING POINT
#
INF
FIRST
GRAB
7.4
7.6
7.6
7.6
7.6
6.9
7.9
7.4
7.7
7.3
7.8
7.3
8.3
7.0
7.8
7.4
8.0
7.4
7.7
8.0
8.4
8.3
7.7
7.3
7.1
7.5
7.6
7.3
8.0
7.4
7.5
7.0
7.1
7.3
7.3
7.3
02
LUENT
SECOND
GRAB
7.9
7.6
7.6
7.6
6.9
7.9
7.7
7.6
7.9
7.5
7.4.
7.5
7.1
7.9
8.2
7.3
7.9
7.6
8.0
8.7
7.1
8.7
7.6
8.7
NA
7.2
7.5
"8.0
7.5
7.1
7.0
6.9
6.8
7.7
7.4
7.1
^
MID P
FIRST
GRAB
8.4
8.3
8.0
8.3
8.4
8.0
8.2
8.0
8.0
8.2
8.0
7.8
7.8
7.5
7.9
7.8
7.7
7.8
7.8
8.0
7.8
7.7
7.7
7.7
7.5
7.6
7.7
7.6
7.7
7.8
7.8
7.2
7.0
7.8
7.7
7.7
03
OINT 1
SECOND
GRAB
8.5
8.5
8.3
8.4
8.3
8.1
8.1
8.0
8.2
8.2
8.2
7.8
8.5
7.7
7.8
7.7
7.7
7.8
7.9
7.7
7.6
7.8
7.8
7.9
NA
7.8
7.6
7.9
7.7
7.2
7.4
7.3
7.2
7.7
7.7
7.7
#
MID F
FIRST
GRAB
8.4
8.3
8.3
8.3
8.3
8.0
8.4
8.0
8.1
8.3
8.2
8.5
8.0
7.7
7.8
7.7
7.8
7.9
7.9
8.0
7.9
7.8
7.8
7.7
7.5
7.9
7.7
7.6
7.7
7.8
7.5
7.1
7.0
7.7
7.7
7.8
04
OINT 2
SECOND
GRAB
8.3
8.4
8.3
8.4
8.2
8.1
8.2
8.2
8.3
8.4
8.4
8.1
7.8
7.9
7.8
7.9
8.0
7.9
7.9
7.8
7.6
7.8
7.8
7.9
NA
7.8
7.7
7.8
7.7
7.2
7.4
7.3
7.2
7.7
7.7
7.7
1
EFF
FIRST
GRAB
8.3
8.2
8.4
8.3
8.4
8.0
8.1
8.1
8.3
8.4
8.3
8.3
8.1
7.9
7.8
7.9
7.9
7.9
7.9
8.1
7.9
7.8
7.9
7.9
7.6
7.7
7.7
7.7
7.7
7.7
8.0
7.4
7.2
7.3
7.7
7.8
05
LUENT
SECOND
GRAB
8.4
8.4
8.2
8.4
8.3
8.2
8.3
8.4
8.4
8.4
8.4
8.2
8.1
7.9
8.0
8.0
7.9
8.0
7.9
7.8
7.6
7.9
7.8
8.0
NA
7.7
7.8
8.0
7.8
7.3
7.4
7.5
7.2
7.7
7.8
7.8
19.
-------
TABLE 4. SYSTEM OPERATING PARAMETERS (cont'd)
Dissolved Oxygen
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
a
9
10
11
12 .
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
ST1-C-01
ST1-C-02
ST1-C-03
STl-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
PROCESS
FLOW
( MGD )
0.75
0.95
1.25
NA
NA
0.97
0.92
0.96
0.88
0.87
1.1
1.24
2.74
2.98
2.71
3.15
3.04
2.88
2.97
3.04
3.04
2.81
2.99
3.24
NA
NA
4.96
5.24
5.17
5.21
5.05
5.04
5.21
4.84
4.97
5.06
Dissolved Oxygen ( mg/1 )
SAMPLING POINT
#02
INFLUENT
FIRST
GRAB
9.7
3.8
3.7
NA
6.56
5.8
6.12
6.7
3.6
5". 4
4.73
4.15
7.4
3.8
4.24
5.3
3.62
5.17
6.16
5.23
6.11
5.32
4.09
4.85
6.46
5.54
5.8
4.35
5.92
3.94
5.84
5.05
5.23
5.49
6.03
6.41
SECOND
GRAB
3.8
4.3
NA
NA
NA
NA
4.82
7.7
5.64
4.12
5.04
5.48
3.26
4.24
3.82
4.65
4.23
4.77
5.45
6.8
5.26
5.32
5.81
6.4
NA
5.85
5.12
6.4
5.96
6.31
6.03
5.16
4.95
5.33
5.77
5.33
#03
MID POINT 1
FIRST
GRAB
9.6
3.5
3.6
5.2
6.7
6.48
6.85
8.7
5.9
6.2
5.15
5.58
6.85
6.37
5.87
6.56
6o22
6.84
5.94
6.07
3.22
4.86
5.6
5.27
5.95
5.8
5.68
5.65
5.54
4.53
5.6
5.9
5.31
5.75
5.97
6.56
SECOND
GRAB
3.5
4.4
NA
NA
NA
NA
5.93
6.5
6.10.
6.88
5.76
6.14
6.83
6.45
6.08
6.48
5.78
5.92
5.81
5.06
4.96
5.51
5.52
4.37
NA
5.92
5.17
5.35
4.9
5.76
6.21
5.81
6.0
5.77
6.17
6.57
#04
MID POINT 2
FIRST
GRAB
9.6
3.5
3.3
5.86
6.64
6.62
7.17
8.7
5.6
6.03
5.75
5.75
7.63
5.5
5.92
6.54
6.11
6.49
6.11
6.56
5.93
5.22
6.04
5.5
6.1
5.57
5.81
5.3
5.49
4.44
6.29
5.52
5.07
5.1
5.47
6.16
SECOND
GRAB
3.6
4.4
NA
NA
NA
NA
8.16
7.4
6.52
5.92
5.99
6.63
6.03
6.58
6.58
6.52
6.42
6.43
5.85
6.23
5.75
5.24
5.31
5.07
NA
6.1
5.31
. 5.2
4.62
5.74
5.39
5.47
5.38
5.12
5.04
5.78
#05
EFFLUENT
FIRST
GRAB
9.8
3.2
3.8
5.28
7.3
0.42
6.79
8.1
7.1
6.5
5.96
5.42
7.26
4.75
4.85
6.35
5.72
6.6
5.81
6.02
5.67
2.0
4.88
5.81
5.83
5.92
5.06
5.65
5.92
3.61
5.88
5.96
5.64
5.89
6.03
6.2
SECOND
GRAB
3.5
4.7
NA
NA
NA
NA
6.17
7.7
6.37
5.86
5.89
6.21
6.8
6.1
5.34
4.95
4.74
6.33
6.23
6.06
5.26
5.51
5.62
5.24
NA
5.62
5.79
6.41
5.8
6.19
6.11.
6.32
6..1
6.13
6.31
5.58
20
-------
TABLE" 4. SYSTEM FIELD MEASUREMENTS (cont'd)
Temperature
DAY
OF
'RUN
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
8
9
1Q
11
12
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07.
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
ST1-C-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
PROCESS
FLOW
( MGD )
0.75
0.95
1.25
NA
NA
0.97
0.92
0.96
0.88
0.87
1.1
1.24
2.74
2.98
2.71
3.15
3.04
2.88
2.97
3.04
3.04
2.81
2.99
3.24
NA
NA
4.96
5.24
5.17
5.21
5.05
5.04
5.21
4.84
4:97
5.06
Temperature ( oF )
-------
SECTION 4
SAMPLING AND ANALYSIS PROGRAM
The Demonstration Test Plan for this project, dated July 1989, is
part of the Appendices to this report. The Plan provides a detailed
discussion of the overall sampling and analytical procedures and
methods discussed in this section. The Plan also references the
procedures used in the Predemonstration well drilling effort and
the HPLC method developed by BioTrol for PCP analysis.
In essence, the sampling and analysis plans developed in
anticipation of the study were adhered to during the investigation.
Composite samples of the influent, effluent, and midpoints of the
reactor were decreased from 24-hour samples to 22-hour samples to
allow equipment to be prepared for the next cycle within the same
24-hour period. Because of access problems, it was necessary to use
a stainless steel beaker to withdraw certain grab samples from the
reactor cells.
The one area where seripus sampling problems were encountered was
the sludge in the bag filter. The nature and amount of the sludge
was such that the original plan could not be adhered to. To obtain
any sample it became necessary to remove the bag and manually
separate solids from the layers of the fabric. Even with that
technique, only a small amount of sludge could be collected. These
difficulties should be considered when reviewing the sludge
analysis data.
Air monitoring was carried out in accordance with the EPA Modified
Method 5 (MM5), although it was necessary to provide additional air
by fan to obtain flow against the resistance" of the carbon canister
and to insure isokinetic sampling. The simultaneous 3-point
sampling could not be achieved because of the narrow chimney
exiting the bioreactor cover. Instead, 3 samples were taken; on
consecutive days. Organic vapors were adsorbed on XAD resin traps
for subsequent analysis by standard methods. The procedure is
discussed in more depth in the Demonstration Test Plan and the
Quality Assurance Project Plan.
The sampling schedule, sample type, and analytical methods for all
tests are summarized in Table 5 as originally developed for the. the
Quality Assurance Project Plan and the Demonstration Test Plan.
Sampling points are as shown on the schematic of the BATS in Figure
7.
22
-------
TABLE 5. ANALYTICAL SAMPLING SCHEDULE AND METHODS
medium parameter frequency method
WATER PCP
Sol PAHs/PCP
Volatiles
Semi-volatiles
Dioxins/furans
Metals
Be, Cd, Cr, Cu,
Ni, Ag, Zn
Sb
As
Hg
Pb
Se
Th
Chloride
TOC
Total Phenolics
Oil/grease
Alkalinity
Residue
Ammonia
Nitrate/nitrite
Phosphate
Biotoxicity
SLUDGE*
PCP/PAHs
Semi-volatiles
- Dioxins/furans
Metals
Be, Cd, Cr, Cu,
Pb, Ni, Ag, Zn
Sb
As
Hg
Se
Th
Total Phenolics
Residue
daily
3/wk
1/wk
1/wk
1/wk
1/wk
daily
daily
1/wk
3/wk
1/wk
3/wk
3/wk
3/wk
3/wk
1/wk
1/wk
1/wk
1/wk
1/wk
1/wk
1/wk
846 3510/8270
ii it
846 8240
846 3510/8270
846 8280
846 3005/6010
it ii
846 3005/7041
846 7060
846 7470
846 3020/7421
846 7740
846 3020/7841
846 9252
846 9060
846 9065
846 9070
SM 403
SM 209
EPA 350.1
EPA 353.1
SM 424
846 3550/8270
846 3550/8270
846 8280
846 3050/6010
it ii
846 3050/7041
846 3050/7060
846 7471
846 3050/7740
846 3050/7841.
846 9065
SM 209
AIR
PAHS/PCP
l/2wk
846 0010(collect)
3510/3540/8270
* As noted in report, difficulties were encountered in obtaining
sufficient sample for analyses.
23
-------
W
-p
c
•H
&
I
0}
•p
•H
i
•
r^
0)
CP
•H
fa
24
-------
As noted earlier, significant differences were observed between
results of the BioTrol HPLC method and the EPA Standard Method
(8270) for PGP during the predemonstration well drilling. While the
cause of the difference remains unclear, it may be attributable to
the fact that the samples for GC/MS were collected while the
grouridwater was standing in the well. In any case, recognition of
the differences led to a small study which confirmed that the two
methods were comparable, at least down to 1 ppm of PCP.
The HPLC data became much more important during the demonstration
when it was found that GC/MS results for PCP in the composite
influent samples were considerably lower than the the values (also
GC/MS) obtained for grab samples of the well water. Process
monitoring samples taken by BioTrol just ahead of the bioreactor
(Sampling point "B" in Figure 7) -— and also ahead of the influent
sampling point (sampling point #2 in Figure 7) — were analyzed
using the HPLC method and consistently gave higher values in
agreement with the well water. These data suggested that some other
factor was contributing to the inconsistency of the influent
character (sampling point #2). Ultimately, it was concluded that
backmixing under the weir separating each bioreactor cell from the
preceding cell was diluting the samples taken at sample points 2,
3 and 4, leading to lower PCP results on samples taken at those
points. While decreases across the bioreactor could still be
measured using either the HPLC results or the well water data,
these findings precluded any consideration of changes from one
chamber to the next for PCP or other pollutants.
Acute biomonitoring with two different species was used as the
ultimate test of the benefit of the biotreatment. The procedure
consisted of determining the LD(50) for sand fleas and fresh water
minnows when various amounts of the groundwater, influent, and
effluent was added to the test water.
25
-------
SECTION 5
DEMONSTRATION RESULTS
INTRODUCTION
The goal of this demonstration project was to study the
effectiveness of the BATS at removing relatively high
concentrations of PCP and PAHs from wood treating wastewaters.
Upon consideration of the BioTrol process and the groundwater
available for testing, it was determined that the key variable that
could be tested was flow rate. At a constant concentration, this
allowed the capacity of the system in terms of pentachlorophenol
concentration and removal to be studied. To avoid shock loading and
the need for re-acclimation, it was further determined that this
could best be accomplished by incrementally increasing the flow and
allowing the system to stabilize at each flow so that analytical
results of "steady state" operation could be obtained. Based on the
capacity of the system (pumps, tanks, etc.) three experimental
stages, each of two-week duration, at 1 gpm, 3 gpm and 5 gpm, were
selected.
Using sampling and analytical methods as described in Section 5 and
in the Appendices to this report, an extensive data base was
collected. These data allowed the system to be evaluated in terms
of (ci)- removal of PCP and other organics; (b) the ultimate fate of
the PCP; (c) the effect of other contaminants such as oil/grease
and metals on this and, potentially other applications of the BATS;/
and (d) the need to be concerned about air emissions and disposal
of Ceirbon, from either a health or an environmental perspective.
J
RESULTS
Extensive analyses using standard methods for sampling and analyses
were carried out over the course of the study. The following
sections discuss the individual parameters and present essentially
all of the analytical results. As noted earlier, it appears that a
backmixing or reverse flow phenomenon was occurring throughout the
study, with partially treated water passing back under the
underflow weir separating each cell of the bioreactor. .This was
detected in the PCP analyses of influent to the system, where two
methods and two different sampling points were fortuitously used.
The data supporting this backmixing hypothesis are presented in the
section on pentachlorophenol. From that data, it must be assumed
26
-------
that analyses at the intermediate sampling points in the bioreactor
(points #3 and #4 in Figure 7) are similarly affected while the
data for the groundwater and the effluent are unaffected.
Consequently, the most reliable assessments of changes in
parameters are considered to be those based on differences between
the groundwater and the effluent, at least at the lower flow rates
where, the backmixing seemed to have had its greatest impact.
System Parameters
Nitrogen, as nitrate, nitrite, and ammonia; and phosphorus, as
phosphate, were monitored throughout the course of the
investigation. No unusual effects were observed; the presence of
residual nutrients in the effluents suggests that reaction
(biodegradation) was not inhibited by insufficient nutrients. The
results are presented in tabular form for the three flow rates in
Table 6 (nitrate/nitrite), Table 7 (ammonia), and Table 8
(phosphate).
Similarly, dissolved oxygen was measured in each chamber of the
reactor twice daily over the course of the study and found to
remain between 5 and 6 ppm. There was a slight increase in the
dissolved oxygen as the water passed through the system and was
aerated by the air sparger system. The incoming groundwater
contained <1.0 ppm dissolved oxygen. This data was presented
earlier in Table 4.
Some variation in temperature was observed throughout the project,
as summarized in Table 9; complete data were presented earlier in
Table 4. First, the temperature of the influent to the bioreactor
was lower in the 3 gpm (14-17°C) and the 5 gpm (14-21°C) experiments
than in the 1 gpm study (22-25°C) . This probably was partially due
to a decrease in the temperature of the groundwater over the course
of the program (11-16°C) , and possibly partially due to the heat
exchanger returning insufficient heat to the incoming groundwater,
from the effluent to maintain a constant temperature for all three
runs. Since the study was carried out in the summer, the heater
which is a part of- the BATS was not brought into operation. As
would be expected, there was a slight temperature increase as the /
wastewater passed through the system, i.e., the effluent
temperature was slightly higher than the influent temperature in
all three study segments.
27
-------
TABLE 6. NITRATE/NITRITE ANALYTICAL
N02/N03 (mg/L)
RESULTS
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-0
ST1-A-0
ST1-A-0
ST1-A-04
ST1-A-0
ST1-A-0
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
SD(n-1)
1
•2
3
4
5
6
7
8
9
10
11
12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89.
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1rB-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
ST1-C-01
ST-1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
<0.100
0.120
0.11
0.01
0.054
0.094
0.07
0.03
0.160
<0.020
0.09
0.10
#02
INFLUENT
0.370
0.091
0.680
0.095
0.620
0.230
0.35
0.26
0.120
0.320
0.120
NA
0.280
0.025
0.17
0.12
1.200
0.200
0.340
0.400
0.240
0.200
0.43
0.39
#03
MID#1
#04
MID #2
#05
EFFLUENT
8.100
7.700
9.900
1.800
5.600
8.500
6.93
1.19
3.400
3.600
2.700
4.000
1.700
2.500
2.98
0.35
0.025
0.670
0.370
0.620
0.500
0.44
0.10
* - Not included in Calculation of Averages or Standard Deviations
28
-------
TABLE 7. AMMONIA ANALYTICAL RESULTS
NIL (mg/L)
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
1
2
3
4
5
6
7
.8
9
10
11
- 12
SD(n-1)
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
| ST1-B-01
ST1-B-02
ST1-B-03
| ST1-B-04
ST1-B-05
ST1-B-06
| ST1-B-07
ST1-B-08
| ST1-B-09
ST1-B-10
(ST1-B-11
! ST1-B-12
AVERAGE
1
2
3
4
5
6
7
8
9
10
11
12
SD(n-1)
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
| ST1-C-01
ST1-C-02
! ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-ia
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
0.06
<0.04
0.05
0.02
0.67
0.41
0.54
• 0.18
0.40
0.44
0.42
0.03
#02
INFLUENT
12.00
3.10
1.40
3.60
4.70
2.80
4.60
3.78
0.45
f
1.20
1.50
NA
2.10
2.20
1.49
0.71
0.24
0.64
0.56
0.83
0.80
0.68
0.63
0.21
#03
MID#1
#04
MID #2
#05
EFFLUENT
14.00
3.00
0.54
2.70
2.10
0.83
3.86
2.10
0.37
0.36
0.38
1.07
2.30
1.30
0.96
0.32
1.30
. 1.40
1.80
2.40
4.00
3.10
2.33
0.44
* - Not included in Calculation of Averages or Standard Deviations
29
-------
TABLE 8. PHOSPHATE ANALYTICAL RESULTS
PO, (mg/L)
DAY
OF
RUN
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
SD(n-1)
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1.0
2.0
3.0
4.0
.'5.0
15.0
7.0
3.0
9.0
10.0
11.0
12.0
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
ST1-C-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
2.4
0.4
1.4
1.4
0.3
0.2
0.2
0.1
1.4
0.2
0.8
0.8
#02
INFLUENT
2.4
2.4
0.4
1.5
2.2
3.3
2.0
1.0
0.4
,
2.0
1.0
1.2
1.0
1.1
0.6
1.5
1.3
1.2
0.2
1.2
1.4
1.1
0.5
#03
MID#1
#04
MID #2
#05
EFFLUENT
2.3
3.2
2.7
1.3
3.0
0.1
. 2.1
0.5
1.4
2.0
1.0
3.4
0.8
0.6
1.5
0.4
1.5
0.8
0.5
0.6
2.8
0.6
1.1
0.4
#14
3UPUCATE
„
- Not included in Calculation of Averages or Standard Deviations
30 .
-------
TABLE 9. TEMPERATURE ACROSS BIOTROL SYSTEM
Flow
gpm
1
3
5
gdwater
#1
21
11
13
Temperature ( °C ,
influent midpt
#2 #3
23.4
15.7
14.6
24.6
20.4
20.8
avg)
midpt
#4
24.5
20.4
21.0
effluent
#5
24.8
20.9
20.9
While the developer's specifications called for a pH of about 7.3
in the tempering (conditioning) tank, the pH measured in the three
chambers of the bioreactor usually was somewhat higher,
particularly at the two lower flows and, at least at the 1 gpm
flow, appeared to increase across the system. The pH of the
groundwater drawn from the well was consistently in .the range of
6.7-6.9 standard units (Table 4).
Initial oil and grease (O/G) concentrations of approximately 50 ppm
were reduced by passage through the system to <10 ppm (Table 10).
While the design of the bioreactor cells is such that formation of
an oil layer at the underflow weirs might have been expected, a
significant amount of oil was observed only occasionally during the
1 gpm portion of the study. At these influent oil levels, there is
no reason to expect the oil to adversely affect the bioreaction;
considerably higher levels are tolerated well in conventional
activated sludge systems. The complete data for oil and grease are
presented in Table 11.
TABLE 10. AVERAGE TSS AND OIL ACROSS THE BIOTROL SYSTEM
Flow
gpm
1
3
5
2.
13
1.
groundwater
TSS O/G
mg/L
5+.07
+12.7
5+0.7
54.5+ 2.1
61.0+ 1.4
47.5±10.6
29
24
15
influent
TSS O/G
mg/L
.6+9.4
.2+17.6
. 7+8 . 9
57
37
50
.5+10
.8+14
.8+10
.7
.9
.5
effluent
TSS O/G
mg/L
53.6+ 6.6
26.3+11.1
22.5+ 9.5
6.
6.
8.
0+0 . 4
0+1.3
0+2 . 4
Total suspended solids (TSS) levels in the incoming well water were
quite low (<5 ppm) and increased somewhat (to 54 ppm at 1 gpm, 26
ppm at 3 gpm, and 18 ppm at 5 gpm) over the course of the stxidy
(summarized in Table 10, full data in Table 12) . At these incoming
"levels, suspended solids would not be expected to interfere with
the reaction. In the effluent the suspended solids probably
contains biomass sloughed from the PVC matrix. Even suspended
solids levels of 54 ppm, as observed in the 1 gpm portion of the
study, do not represent a significant mechanism for removal of
31
-------
TABLE 11. OIL AND GREASE ANALYTICAL RESULTS
(mg/L)
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/28/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-Ar-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-?A-11
ST1-A-12
AVERAGE
SD(n-1)
1
2
3
- 4
5
- 6
7
8
'• 9
10
11
12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
. 4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
ST1-C-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
56
53
54.5
2.1
60
62
61.0
1.4
55
40
47.5
10.6
#02
INFLUENT
8 *
47
53
58
371 *
72
57.5
10.7
57
.
15
31
33
50
41
37.8
14.9
58
59
37
58
42
NA
50.8
10.5
#03
MID#1
<5
<5
5
<5
48 "
20 *
5.0
0.0
7
14
9
9
13
12
10.7
2.7
12
14
32 *
10
18
15
13.8
3.0
#04
MID #2
5 *
5
5
5
20 *
5
5.0
0.0
<5
6
<5
5
<5
8
5.7
1.2
<5
19
12
11
15
10
12.0
4.7
#05
EFFLUENT
<5
8
7
<5
21 *
<5
6.0
0.4
<5
6
<5
8
<5
7
6.0
1.3
<5
11
9
.10
7
6
8.0
2,4
#14
DUPLICATE
19(EFF) *
57 (INF)
5{EFF) .
40 (INF)
•- Not included in Calculation of Average or Standard Deviation
32
-------
TABLE 12. TSS ANALYTICAL RESULTS
(mg/L)
DAY SAMPLE SAMPLE
OF DATE NUMBER
HUN
1
2
3
4
5
6
7
8
9
10
11
12
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
ST1-C-01
ST1-6-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06.
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
3
2
2.5 ~
0.7
22
4
13.00
12.73
2
<1
1.50
0.71
#02
INFLUENT
29
38
39
4
26
16
29.6
9.4
10
54
23
NA
22
12
24.20
17.64
27
8
6
,13
26
14
15.67
8.91
#03
MID#1
26
48
53
4 *
46
50
44.6
2.7
10
84
22
20
18
8
27.00
7.78
34
12
11
8
50
12
21.17
4.67
#04
MID #2
35
47
52
4
54
69
51.4
3.6
10
72
16
14
94
6
35.33
12.05
31
12
10
8
36
9
17.67
4.01
#05
EFFLUENT
23
67
64
4
54
60
53.6
6.6
8
SB
12
67
11
4
26.33
11.09
32
14
8
6
66
9
22.50
9.52
#14
DUPLICATE
52 (EFF)
14 (INF)
6 (EFF)
11 (INF)
- Not Included In Calculation of Averages or Standard Deviations
33
-------
organic pollutants. For example, at a 51 ppm increase in suspended
solids and a PGP concentration of 34 ppm, this would account for
only 0.00025 Ibs of PCP over the course of the 12 days of operation
at 1 gpm. With total PCP removal during that period of 5.97 Ibs,
the amount in the sludge is negligible. The volatile suspended
solids (VSS) represent approximately 45% of the total suspended
solids, possibly increasing slightly (to "50%) at the highest flow
rate (Table 13).
Pentachlorophenol Removal
With the receipt of the PCP analytical data generated by SW-846,
Method 8270, it became clear that the results were not as
anticipated. Influent PCP values (point #2 in Figure 7) were
significantly lower than the concentrations leaving the well (point
#1) , as summarized in Table 14. Grab samples taken by BioTrol from
a "T111 just before the influent chamber (point B on Figure 7) and
analyzed by BioTrol's HPLC method agreed well with the well samples
(point #1). In addition, the groundwater (point #1) and influent
(point #2) PCP values approached agreement as the flow rate was
increased each two weeks. Within limits discussed later, free
chloride and Total Organic Carbon (TOC) data at the various
sampling points also supported these observations.
TABLE 14. AVERAGE PENTACHLOROPHENOL REMOVAL BY BIOTROL SYSTEM
Flow PCP
(gpm) (#1)
gdwtr
1
-3
5
42.0+7.1*
34.5+7.8*
27.5+0.7*
PCP
(#2)
infl.
6.9+3.4
19.0+5.8
24.2+6.8
PCP
(#5) [
effl.
0.13+.25
0.34+.15
0.99+.49
Removal
_________/ 9-N 1
\
gdw/eff
99.8
98.5
96.4
i^i j
infl/effl
98.1
98.2
95.9
* The gradual decrease in groundwater concentration may
be a result of well drawdown over the 6 weeks of
operation. -
At first it was thought that the difference might have been due. to
foaming during the extractions for the analytical procedure (SW-r846
Methods 3510/8270). While that problem did persist throughout the
study and did affect the recovery achievable for the samples, it
became clear that it was not the cause of the differences between
the well or BioTrol's samples and the influent chamber samples.
Other possible explanations for the anomalous results were
considered, including absorption of PCP on the walls of the
bioresactor, separation of a PCP-in-oil layer, and backmixing under
the underflow weirs separating the chambers. Calculations indicated
that the amount of an absorbed film of PCP or a PCP-in-oil layer
34
-------
Table 13. VSS/TSS ANALYTICAL RESULTS
(mg/L)
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE WT
SD(n-1)
,1
2
3
4
5
6
7
8
9
10
11
- 12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
. 8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE V/T
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/89
8/29/89
8/30/89
8/31/89
9/01/89
ST1-C-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
ST1-C-09
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE V/T
SD(n-1)
SAMPLING POINT
#01
WELL
2/3
1/2
0.58
0.12
-
4/22
3/4
0.47
0.40
<1/2
<1
0.75
0.35
#02
INFLUENT
8/29
16/38
16/39
2/4
10/26
7/16
0.40
0.07
4/10
27/54
11/23
«
8/22
6/12
0.45
0.06
16/34
6/8 *
3/6
4/13
*
6/14
0.49
0.16
#03
MID#1
11/26
17/48
20/53
2/4
18/46
21/50
O41
0.02
4/8
38/84
10/22
12/20
8/18
5/8
0.51
0.02
14/31
6/12 *
6/11
4/8
22/50
6/12
0.49
0.01
#04
MID #2
16/35
20/47
19/52
1/4
21/54
26/69
0.38
0.03
4/10
32/72
8/16
8/14
51/94
3/6
0.49
0.02
16/32
6/12 *
6/10
4/8
18/36
6/9
0.54
0.03
#05
EFFLUENT
10/23
29/67
25/64
1/4
20/54
22/60
" 0.37
0.04
2/8
27/56
6/12
33/67
5/11
4/4
0.53
0.13
1/2
8/14 *
4/8
3/6
32/66
5/9
0.52
0.02
- Not Included in Calculation of Average
35
-------
needed to explain the differences were simply not present.
Consequently, backmixing appeared to be the best explanation for
the observed differences. And, as flow through the system was
increased each two weeks, the impact of the backmixing decreased,
explaining the gradual improvement in agreement between well water
data, BioTrol process monitoring data by the HPLC method, and
influent data.
On the assumption that the "real" concentration of PGP in the
inflxient was about 45 ppm in the 1 gpm study but was measured at
about 7 ppm, it was estimated that approximately 5 gallons was
"leaking" back through the system for every gallon introduced in
the first two weeks. At higher flows, the effect of backmixing
would be expected to be less.
Because the influent sampling problem was only discovered when
samples had been analyzed, an additional QA study was carried out
to compare BioTrol' s HPLC method for determining PCP concentrations
with EPA's Standard Method (SW-846, Method 8270). This
investigation, reported in the Quality Assurance Review Appendix to
this report, confirms that the. two methods are comparable when
extraction efficiency in Method 3370/8270 is taken into
consideration.
On the basis of this additional information, it was concluded that
concentrations of PCP in the well water were better representations
of the concentrations entering the first cell of the bioreactor
than the influent samples. Because of the backmixing phenomenon,
the analytical data from the intermediate sampling points in the
bioreactor (points #3 and #4) were used only to confirm that
gradiial reduction in PCP (and other parameters) was taking place in
the expected direction. PCP analytical results for all sampling
points are presented in Table 15. Removal efficiencies have been
based on well water and effluent data (both by Method 8270) instead
of the influent samples.
It remains clear that the BioTrol Aqueous Treatment System does
effectively remove PCP from the groundwater as it moves through the
system. Effluent concentrations of PCP after treatment (but before
carbon polishing) averaged 0.13 ppm at the 1 gpm flow rate and
increased to approximately 0.99 ppm at the 5 gpm flow rate. Thus,
depending on the level of PCP required in the effluent, effluent
quality can be adjusted by adjusting the flow rate. Using the well
water data as the influent concentrations and the observed effluent
concentrations for PCP, the system achieves an average of 96.4% PCP
removal at the 5 gpm flow rate and this increases to 99+% at the
•lower flow rates.
36
-------
TABLE 15. PENTACHLOROPHENOL ANALYTICAL RESULTS
PCP (ug/L)
DAY
OF
HUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8'
9
10
11
1?.
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
1?
8/21/89 ST1-C-01
8/22/89 ST1-C-02
8/23/89 ST1-C-03
8/24/89 ST1-C-04
8/25/89 ST1-C-05
8/26/89 ST1-C-06
8/27/89 ST1-C-07
8/28/89 ST1-C-08
8/29/89 ST1-C-09
8/30/89 ST1-C-10
8/31/89 ST1-C-11
9/01/89 ST1-C-12
AVERAGE
SD(n-1)
SAMPLING . POINT
#01
WELL
47000
37000
42000.0
7071.1
40000
29000
34500.0
7778.2
27000
28000
27500.0
707.1
#02
INFLUENT
680
7000
6900
3000
7200
13000
7600
6600
12000
8900
4200
6100
6931.7
3434.0
8800
24000 .
NA
21000
19000
21000
NA
24000
22000
10000
25000
15000
18980.0
5812.2
15000
21000
NA
15000
19000
23000
35000
23000
28000
28000
25000
34000
24181.8
6720.4
#03
MID#1
11
1200
1500
400
400
1300
17
200
85
150
530
690
540.3
524.4
2600
4700
4800
NA
2900
3900
5500
6800
10000
4300
3700
7400
5145.5
728.1
4900
6700
NA
10000
8100
6000
5300
2600
4100
3700
6900
12000
6390.9
926.9
#04
MID #2
50
230
340
140
100
410
160
11
8
<50
<50
17
130.5
133.3
210
1000
160
1200
490
570
900
1400
1500
740
370
190
727.5
191.8
3500
3200
NA
4000
1200
1600
870
3200
8600
3900
5500
10000
4142.7
1112.3
#05
EFFLUENT
16
880
140
140
88
<5000
<50
5
5
8
<50
50
130.2
253.5
82
810
190
NA
320
430
200
1000
<50
390
230
78
343.6
147.9
3100
1200
NA
910
890
1100
<50
140
300
330
140
2700
987.3
487.4
% removal
02-05/02
97.6
87.4
98.0
95.3
98.3
99.3
99.9
100.0
99.9
98.3
99.2
98.1
3.7
99.1
96.15
,
98.3
98.0
95.13
99.13
96.1
99.1'
99.5
98.2
0.9
79.3
94.3
— ,
93.9
95.3
95.2
99.9
99.4 .
98.9
98.8
99.4
92.1
95.9
4.0
- Not included in Calculation of Averages or Standard Deviations
37
-------
The data are presented in terms of pentachlorophenol mass removal
in Table 16. Based on the calculated mass of pentachlorophenol
introduced to the system over each two week experimental period,
and assuming that all pentachlorophenol is lost by biological
degradation, mass removals of >95% are consistently achievable.
TABLE 16. MASS REMOVAL OF PENTACHLOROPHENOL ''.
Week
1
2
3
4
5
6
Flow
(gpm)
0.98
1.0
2.92
3.02
5.14
5.03
Total PCP
In (Ibs)*
(#D
3.32
2.65
8.39
6.29
9.99
10.13
Total PCP
Out (Ibs)
(#5)
.178
.002
.077
.075
.533
.221
Removal
(%)
94.6
99.9
99.1
98.8
94.7
97.8
* Based on well water (#1) analyses
Comparison of analytical results for total PCP in as-is samples
, and soluble PCP in samples that had been filtered to remove any
suspended solids confirmed that very little if any of the PCP in
.the as-is samples was absorbed on the filterable solids (Table
17) . There were, however, measurable concentrations of PCP in
the sludge samples (see Table 21) . Considering the small amount
of sludge exiting the bioreactor, this is not a significant
contributor to the removal of PCP from the system, as discussed
earlier. Similarly, the low concentration of oil in the effluent
(<10 ppm) strongly argues against loss of PCP by extraction into
that phase.
Analyses of the air exhausted from the bioreactor also confirmed
that no detectable quantities of pentachlorophenol were lost by
this route.
Mineralization. Chloride and TOC monitoring of the influent and
effluent produced results that are consistent with an assumption
that the bulk of the pentachlorophenol is mineralized, but
indicate that other contributors to chloride and TOC are present.
The expected reaction would be:
OH
ci /_\ Cl
!( )| + 5.5 02 + OH" ---- - --- > 5 Cl" + 6 C02 + H2O
Cl V-7 Cl
Cl
38
-------
TABLE 17. SOLUBLE PCP/TOTAL PCP ANALYTICAL DATA
PCP (ug/L)
DAY
OF
HUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1^A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE PCP(tOt)
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE PCP(tot)
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89 ST1-C-01
8/22/89 ST1-C-02
8/23/89 ST1-C-03
8/24/89 ST1-C-04
8/25/89 ST1-C-05
8/26/89 ST1-C-06
8/27/89 ST1-C-07
8/28/89 ST1-C-08
8/29/89 ST1-C-09
8/30/89 ST1-C-10
8/31/89 ST1-C-11
9/01/89 ST1-C-12
AVERAGE PCP(tOt)
SD(n-1)
SAMPLING POINT
#01
WELL
39000/47000
41000/37000
42000.0
7071.1
42000/40000
29000/29000
34500.0
7778.2
29000/21000
29000/28000
27500.0
707.1
#02
INFLUENT
690/680
— 17000
5200/6900
—13000
— /7200
15000/13000
5200/7600
— 16600
11000/12000
— /8900
— /4200
8800/6100
6931.7
3434.0
15000/8800
.— /24000
NA
— /21000
—/19000
21000/21000
NA
— /24000
22000/22000
— /10000
— /25000
22000/15000
18980.0
5812.2
3000/15000
—/21000
NA
— /15000
— /19000
27000/23000
— 135000
21000/23000
— /28000
28000/28000
— /25000
2000/34000
24181.8
6720.4
#03
MID #1
<50/11
— /1200
810/1500
— 7400
— /400
1800/1300
13/17
— 1200
96/85
— /150
— /530
460/690
540.3
524.4
3200/2600
— /4700
490/4800
NA
— /2900
3600/3900
4200/5500
— /6800
8900/10000
— /4300
— /3700
6100/7400
5145.5
728.1
4400/4900
— /6700
NA
— /10000
— /8100
7400/6000
— 15300
5000/2600
— /4100
3100/3700
— /6900
10000/12000
6390.9
926.9
#04
MID #2
6/50
— /230
350/340
— /140
— /100
710/410
66/160
— /11
4/8
— /<50
— /<50
<50/17
130.5
133.3
240/210
— /1000
540/160
— /1200
—7490
640/570
830/900
— /1400
1400/1500
— /740
— 1370
190/190
727.5
191.8
3200/3500
— /3200
NA '
— /4000
~/1200
2200/1600
— /870
3100/3200
•— /8600
3600/3900
— /5500
100/10000
4142.7
1112.3
#05
EFFLUENT
11/16
— /880
120/140
~/140
— /88
~/<5000
11/<50
— /5
4/5
— /8
— /<50
9/50
130.2
253.5
77/82
~/810
580/190
NA
~ 1320
350/430
330/200
— /1000
1600/<50
— 7390
—7230
100/78
343.6
147.9
3000/3100
--/1200
NA
~/910
— /89CI
810/1100
— 7
-------
or perhaps
OH
Cl A-A Cl
ci
+ 5.75
-> 5 Cl' +6 CO,
+ H20
ci
Table 18 summarizes the observed and expected changes in chloride
and TOC calculated on the basis of the observed decrease in
pentachlorophenol, again using the PCP values for the groundwater
rather than the influent. Chloride and TOC values were selected for
the same days. The actual analytical data are reported in Table 19
(chloride) and Table 20 (TOC).
TABLE 18. COMPARISON OF CHLORIDE & TOC CHANGES WITH PCP REMOVAL
1
1
1.
I
1
1
1
Flow PCP
(gpn)
1
3
5
change
-41.9
-34.1
. INCREASE IN j
Cl(fd) Cl(calc) !
DECREASE IN
TOC(fd) TOC(calc)
• i
i
(mg/L) j
+40.2
+37.2
-26.5 +27.2
+27.9
+22.7
+17.6 (
-25.5
-31.5
-21.0
-11.3
- 9.2
- 7.0
1
1
fd = found (effluent - groundwater)
calc = calculated from change in PCP, as 5C1/PCP & 6C/PCP
The concentrations of mono-, di-, and trichlorophenols (by SW846
Method 8270) all were below the detection limits in the
semivolatile scans of the groundwater, influent, and effluent but
the detection limits were often quite high since pentachlorophenol
was the primary "target" of the analyses. It is reported that
manufactured pentachlorophenol may have contained as much as 20% by
weight of tetrachlorophenols (TCPs). If, for example, the original
PCP concentration in the groundwater, 42 ppm, were accompanied by
20% by weight as- tetrachlorophenols (which were not target
analytes) , this would mean that 8.4 ppm of TCPs were present. If*.
these tetrachlorophenol isomers degrade as efficiently as PCP, the
8.4 ppm would contribute 5.1 ppm of chloride. And if one also
assumed that all of the less chlorinated isomers (mono-, di-, and
trichlorophenols) in the semivolatile scan were present at just
below their detection limits in the original sample, they could
contribute an additional 6.4 ppm of chloride. Table 21 presents
calculated chloride yields on the basis of these assumptions for
the groundwater sample during the 1 gpm study. [Subsequent review
of two archived scans of groundwater indicated TCP contents .of
about 4% and 10% of the PCP measured.] Other chlorinated phenol
species that are not reported as part of the semi-volatile scan
also could be present.
40
-------
TABLE 19. CHLORIDE ANALYTICAL RESULTS
Cl (mg/L)
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
1
2
3
4
5
6
7
8
9
10
11
-12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
SD(n-1)
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B-11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89 !
' 8/26/89 !
8/27/89 i
8/28789 5
8/29/89 5
8/30/89 S
8/31/89 J
9/01/89 £
ST1-C-01
ST1^C-02
ST1-C-03
ST1-C-Q4
3T1-C-05
3T1-C-06
3T1-C-07
3T1-C-08
3T1-C-09
3T1-C-10
3T1-C-11
5T1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
14
19
16.50
3.54
14
24
19.00
7.07
38
12
25.00
18.38
#02
INFLUENT
58
48
56
48
53
46
.48
48
44 '
53 '
53
48
50.82
3.95
jt
44
29 .
44
34
40
46
NA
34
34
38
34
34
37.36
5.45
41
27
29
27
26
28
29
29
18
16 '
28
21
27.55
5.70
#03
MID#1
53
58
53
63
63
58
53
53
53 '
58
58
58
57.09
1.31
48
53
9.7 *
50
53
53
60
50
46
51
. 50
51
51.36
1.24
53
46
51
45
43
52
51
51
37
48
46
44
47.25
1.67
#04
MID #2
58
68
63
82
58
58
58
58
9.7 '
63
63
53
62.00
3.19
58
58
106 *
53
55
55
54
54
57
54
55
56
55.36
0.69
55
51
53
- 50
54
54
55
52
52
46.
46
47
51.25
1.40
#05
EFFLUENT
58
53
58
34
63
63
58
53
100
58
63
63
56.73
4.44
52
58
58
63
56
56
55
55
55
56
56
55
56.25
1.44
54
52
54
53
55
55
55
55
54
27
42
46
50.17
4.50
#14
DUPLICATE
48 (INF1)
53(EFIF)
44 (INF)
54 (EFF)
(
54 (EFF)
6 (EFF)
4 (INF)
* - Not Included in Calculation of Average or Standard Deviation
.41
-------
TABLE 20. TOG ANALYTICAL RESULTS
TOC (mg/L)
DAY
OF
RUN
1
2
3
4
5
6
7
8
9
10
11
12
SAMPLE
DATE
7/24/89
7/25/89
7/26/89
7/27/89
7/28/89
7/29/89
7/31/89
8/01/89
8/02/89
8/03/89
8/04/89
8/05/89
SAMPLE
NUMBER
ST1-A-01
ST1-A-02
ST1-A-03
ST1-A-04
ST1-A-05
ST1-A-06
ST1-A-07
ST1-A-08
ST1-A-09
ST1-A-10
ST1-A-11
ST1-A-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/07/89
8/08/89
8/09/89
8/10/89
8/11/89
8/12/89
8/14/89
8/15/89
8/16/89
8/17/89
8/18/89
8/19/89
ST1-B-01
ST1-B-02
ST1-B-03
ST1-B-04
ST1-B-05
ST1-B-06
ST1-B-07
ST1-B-08
ST1-B-09
ST1-B-10
ST1-B^11
ST1-B-12
AVERAGE
SD(n-1)
1
2
3
4
5
6
7
8
9
10
11
12
8/21/89
8/22/89
8/23/89
8/24/89
8/25/89
8/26/89
8/27/89
8/28/69
8/29/89
8/30/89
8/31/89
9/01/89
ST1-G-01
ST1-C-02
ST1-C-03
ST1-C-04
ST1-C-05
ST1-C-06
ST1-C-07
ST1-C-08
&T1-C-O9
ST1-C-10
ST1-C-11
ST1-C-12
AVERAGE
SD(n-1)
SAMPLING POINT
#01
WELL
..',
79
79
79.0
0.0
82
82
82.0
0.0
80
82
81.0
1.4
#02
INFLUENT
63
65
62
64
63
62
60
57
120
65
58
59
61.6
2.8
69
72 -
70
69
68
68
NA
75
74
70
73
75
71.2
2.7
72
85
84
82
83
82
79
84
84
81
84
79
81.6
3.6
#03
MID#1
32
55
55
52
51
47
48
47
48
60
51
46
49.5
2.5
55
58
60
57
55
54
57
59
61
58
58
58
57.5
0.7
66
66
69
66
64
72
65
66
68
68
69
67
67.2
0.8
#04
MID #2
52
54
56
53
50
51
49
48
53
48
49
46
50.5
1.2
51
51
53
54
51
49
49
53
54
51
50
48
51.2
0.8
64
63
64
65
59
70
58
61
68
80
67
71
65.8
2.5
#05
EFFLUENT
56
57
66
60
61
2450
48
46
49
49
46
46
53.5
3.8
51
52
54
51
46
50
53
50
.54
50
47
48
50.5
1.4
63
61
60
56
53
68
.55
56
61
68
59
60
60.0
2.5
#14
DUPLICATE
69 (INF)
50 (EFF)
70 (INF)
49 (EFF)
64 (EFF)
60 (EFF)
82 (INF)
- Not Included in Calculation of Averages or Standard Deviations
42
-------
TABLE 21. POTENTIAL CHLORIDE CONTRIBUTIONS FROM
PARTIALLY CHLORINATED PHENOLS
substance detection limit chloride yield
PPM PPM (calc)
2-MONOCHLORO 2 0.6
2,4-DICHLORO 2 0.9
2, 4,6-TRICHLORO 2 1.1
2,4, 5-TRICHLORO 10 5.4
TETRACHLORO NA, .2PCP* 5.1
POSSIBLE NON-PCP TOTAL .......... 13 . 1
PENTACHLORO 42 found 27.9
GRAND TOTAL ............. . ........ 41.0
*Estimated only, on the basis that PCP
may contain as much as 20% tetra isomers.
Removal of other pollutants such as residual oil or biomass may
explain why the decreases in TOC levels are higher than calculated
for PCP removal.
Polvnuclear Aromatic Hydrocarbon Removal
Concentrations of the various polynuclear aromatic hydrocarbons
(PAHs) in the incoming well water were lower than anticipated and
.below the detection limits in both the well water and the influent.
The high detection limits for PAHs in the semivolatile organics
GC/MS scans, often in the range of 2 ppm when analyzing groundwater
or influent for pentachlorophenol in the 10-50 ppm range, precluded
measurements for PAHs at ppb level, but do confirm that significant
concentrations of the various PAHs were not present in the
groundwater. Two values for total PAHs obtained during the
predemonstration well drilling effort, 145 and 295 ppb, would be,
consistent with the absence at the indicated detection limits
during the experimental portion of the project.
Specific PAHs were below the detection limits in the effluents as <
well, even with the detection limits now in the range of 10-100
ppb. One effluent sample exhibited a single PAH component, reported
as 0.4 ppb of fluoranthene. The lack of numerical values for PAHs
in the well water or the influent makes it impossible to assess the
removal of these chemicals by the BATS at the MacGillis and Gibbs
facility.
As noted earlier, It was only possible to carry out limited
analyses, and only on a few sludge samples, for pentachlorophenol
or the PAHs. These analyses, indicated low but detectable
quantities of PCP and selected PAH compounds of interest, as noted
in Table 22. In every case the amount of sludge produced or
collected was so small that daily sampling of the sludge was not
practical. While not conclusive, the limited data does suggest that
43
-------
accumulation of PGP or PAHs in the sludge is not a significant
contributing mechanism for the removal of these species.
TABLE 22. SLUDGE ANALYSIS RESULTS
SAMPLE POLLUTANT CONG
(mg/kg dry)
B-09-07S PGP 34
pyrene 15
chrysene 5.3
Benzo(b)fluoranthene 10
benzo(a)pyrene 6.1
C-05-07S PGP 170
Anthracene 92
Benzo(k)fluoranthene 74
C-10-07S PGP 2.7
Anthracene 3.0
Pyrene 2.5
C-10-15S* PGP 18
phenanthrene 1.9
* This sample is "active" sludge found
adhering to the walls of bioreactor.
Interestingly, monitoring of the exhaust air stream, in a 4 inch
chimney in the lid over the bioreactor before and after the carbon
adsorber, indicated that some stripping of polynuclear aromatic
hydrocarbons does occur, probably due to the air bubbled through
system (Table 23). It was necessary to introduce additional air to
the stack during the monitoring to overcome the resistance of the
carbon. Small amounts (=< 1 ppb) of phenol, 2,4-dimethyl phenol,
and higher molecular weight" PAHs such as fluorene and dibenzofuran,
were found occasionally in the pre-carbon samples but not in the
after-carbon samples. Only naphthalene was found to pass through
the carbon at a detectable level (under 2 ppb), possibly by an air
stripping "regeneration" phenomenon.
44
-------
TABLE 23. PAHS IN AIR EMISSIONS FROM BIOREACTOR
Test Liquid naphthalene 2-me naphthalene acenaphthene
# flow at b.c.* a.c.* b.c. a.c. b.c. a.c.
*
#
UA,
1
2
3
4
5
b.c.
a.c.
nd
tuts uj. Ufciia
1
1
1
3
5
= before
= after
= not de
u L
6.5
3.8
4.6
4.6
34.6
carbon
carbon
tected
U
0.6
1.6
1.7
0.6
1.1
adsorber
adsorber
JP"
4
3
3
6
47
) — '
.7
.0
.7
.7
.9
nd f
nd
nd
nd
nd
0.
0.
nd
0.
2.
5
3
7
8
]
nd
nd
nd
nd
nd
Polvchlorinated Dibenzo-p-Dioxins/Dibenzofurans
Historically, there has been concern about the possibility of
chlorinated dibenzodioxins in manufactured chlorinated products
such as pentachlorophenol. Consequently there was some concern
about the possibility of these species being present in the
grounclwater. Selected samples were scanned for the various dioxins
and furans using high, resolution GC coupled with low resolution MS.
A number of the chlorodioxin and furan species were found to be
present in the influent and in the effluent, but only at
nanogram/liter levels except for the octachloro- isomer (OCDD). The
2,3,7,8-TCDD isomer was reported at above detection limit in only
one sample (62 ng/L in an effluent sample) . Review of the data
(Table 24) also discloses an increase in concentration for all the
isomers as the wastewater moves through the reactor. The increases
may be due to accumulation of the dioxins on the biomass, which
then is sloughed off into the effluent.
TABLE 24. DIOXINS/FURANS FOUND IN. SYSTEM
Week
HpCDD
HpCDF
Influent/Effluent Concentrations
(ng/L)
HxCDD HxCDF OCDD OCDF
TCDD 2378-
1
2
3
4
5
6
60/180
32/180
<4.4/4.3
—
25/62
^
<10/20
<2.8/30
-
-
2,1/7.0
*~
-* •
<2.2/8.8
-
-
-
••
—
<1. 5/4.1
. -
—
-
—
340/1100
170/910
28/42
<8.6/28 .
140/390
—
<17/23
<7.3/40 '
-
-
<4.8/12
—
„_ ~ _
> _ .
- ~ -
- - '
-
<3.2/62
* An '-' indicates the isomer was absent or below detection limit in
both influent and effluent.
Some dioxin/furan isomers were found in the sludge in the ng/Kg
range, based on wet weights (Table 25). Recognizing the small
45
-------
amount of sludge generated, it is not considered a problem except
that the sludge may require disposal as a dioxin-contaminated
material. This would add somewhat to the cost of the process.
TABLE 25. DIOXINS/FURANS FOUND IN SLUDGE
Week
Sludge Concentrations
(ng/gm)
1
2
3
4
5
6
HpCDD HpCDF
260 46
no sample
no sample
25 3.9
<.20 <.14
23 4
HxCDD
13
1.4
<. 17
1.5
HXCDF
13
1.7
<. 12
1.7
OCDD
1900
190
0.98
140
OCDF
41
3.0
<.22
3.1
TCDD
<.054
<.088
<. 16
<.08
2378-
<.054
<.088
<. 16
<.08
Heavy Metals
While, low concentrations of arsenic and various heavy metals were
found to be present in the groundwater, these appear to pass
through the system with little change in concentrations (Table 26)
- with the exception of the first effluent,, sample, which shows
anomalous results. The remaining data in the table are groundwater
and effluent data obtained on neighboring, but not necessarily the
same, day. At the levels encountered, metals are not expected to
interfere with the bioreaction. The analytical data show no
evidence for accumulation of metals on the biomass and subsequent
sloughing into the effluent.
TABLE 26. METALS FOUND IN SYSTEM
Week
Concentration of Metals (groundwater/effluent)
(ug/L)
As Cu Ni cr Pb Zn
1
2
3
4
5
6
DLA
6.4/220
4.1/5.6
5.4/5.3
— /6.0
6 . 5/7 . 7
5.9/5.7
1.5
25/4400
19/37
20/23
— /19
~/23
ND/30
12
60/390
54/67
81/73
-v/60
67/71
91/87
21
ND/450
ND/8 . 0
ND/ND
— /ND
ND/23
7 . 0/ND
6.7
7.7/580
11/ND
ND/11
— /8.1
3.9/6.1
6.9/5.9
2.9
32/20,000
20/20
ND/8.0
— /13
23/30
. 20/17
5.7
* An '—' indicates the analysis was not carried out; an ND
indicates the concentration was below detection limit.
A DL = Detection limit
46
-------
Volcitile Organics
Analyses for volatile organics indicated that few of these
materials were present in the groundwater, even at low
concentrations. Levels were further reduced by passage through the
treatment system, probably by stripping. Interestingly, volatile
organics were not detected in the exhaust air collected from the
reactor during the Modified Method 5 testing.
Biomonitoring
Considering the nature of the contaminants, it was suspected that
the groundwater could be toxic to aquatic species. With that in
mind, biomonitoring with two different species, minnows and water
fleas, was carried out on the groundwater, the influent, and the
effluent. The results confirmed that the groundwater was toxic to
these species and that treatment removed the cause of the toxicity.
The results presented as LC50 values in Table 27 reflect the
percentage of groundwater, influent, or effluent that could be
tolerated in the water used for the test before 50% of the species
succumbed. When 1% or even less of either groundwater or influent
water is introduced to the test water, 50% or more of the test
species die during the test period. After treatment in the BATS,
the effluent has essentially no adverse effect on either species,
even when 100% effluent is the test water.
TABLE 27. ACUTE BIOTOXICITY OF GROUNDWATER AND TREATED EFFLUENT
Week Flow LCeg-Daphnia # LC5Q-Minnow *
(gpm) grdwtr infl. effl. grdwtr infl. effl,
( % wastewater/test water)
1
2
3
4
5
6
1
1
3
3
5
5
__
—
—
—
1.0
-• 0.27
0.35
0.84
0.26
0.54
0.61
0.66
100
100
100
100
100
100
__
—
—
0.22
0.22
0.3
1.07
0.43
0.3
0.20
0.20
100
100
100
35
100
100
# 48 hour static test at 20°C, daphnia magna
* 96 hour static test at 20°C, pimephales promelas
47
-------
SECTION 6
ECONOMICS
INTRODUCTION
The purpose of this economic analysis is to provide realistic
costs and a knowledge of the basis for their determination, so that
it is possible to estimate the cost for systems for other sites.
The analysis is based on cost factors developed over the six weeks
of operations at the demonstration site as well as other
information provided by BioTrol, Inc. for operation at other sites.
The key conclusions are based on the two weeks of operation of the
mobile unit operating at 5 gpm with an influent containing
approximately 30 ppm of pentachlorophenol and achieving 96% removal
of that, pollutant.
CONCLUSIONS
Operating costs total approximately $3.45/1000 gallons when
treating approximately 7200 gallons/day at 5 gpm in a 5 gpm BATS.
The equivalent residence time is 1.8 hours.
Key components of the operating cost for the mobile unit were found
to be heat and labor. Secondary items included: nutrients,
electricity (pumps), and caustic. Related items such as
permitting, sinking wells, residuals disposal, etc. are not-
included.
OPERATING COSTS
Cost data were provided by BioTrol, Inc. for operation of -the
mobile unit at a flow of 5 gpm, assuming an incoming groundwater
with about 45 ppm of pentachlorophenol and a removal efficiency of
about 95%. Operating costs were also provided by BioTrol for a 30
gpm unit on the basis of this demonstration and other data from the
developer. Table 28 summarizes the essential operating cost delta
for systems of both sizes.
48
-------
TABLE 28. OPERATING COST OF TREATMENT
Cost item at 5 gpm at 30 gpm
($/1000 gal)
nutrients
electricity
heat
labor*
caustic
0.042
0.216
1.46
1.49
0.24
0.017
0.216
1.46
0.50
0.24
TOTAL 3.45 2.43
* includes operation, maintenance, routine
monitoring at $15/hr
The estimates do NOT include operating costs for oil/water
separation, suspended solids removal, exhaust air treatment, or
effluent polishing since these items are site and waste specific.
While carbon was used at the MacGillis and Gibbs site as part of
the demonstration program, the exhaust from the reactor and the
effluent from the reactor would have met the OSHA Permissible
Exposure Limit (PEL) for naphthalene (10 ppm) and POTW discharge
requirements (2 ppm PCP) without these controls.
In addition, the cost data do not reflect items such as well
drilling, obtaining permits, disposing of solid residues or other
wastes, site security, insurance, or decommissioning the site after
treatment. While these Items are part of EPA's list of 12 cost
categories, these costs are expected to be comparable for any above
ground system that was used to remediate a site. In addition, these
costs - would be very dependent on the site location and
characteristics.
A more detailed discussion of each of the cost elements included in
Table 28 is provided in the following paragraphs.
Nutrients; Urea and trisodium phosphate are added at the same t
dosage (0.16 Ib urea and 0.31 Ib TSP/lb PCP) regardless of system
size. For the pilot scale system the cost reflects purchase of.the
ingredients at a local fertilizer supplier; for the larger system
some economy of scale has been factored in on the assumption that
the materials would be purchased in bulk. (Cost for storage
facilities has not been included in the capital costs.)
Electricity; Groundwater (or other feedwater) is delivered to the
system by a pump .external to the system; the cost to operate that
pump has not been included. Electricity, at $0.06/kwh, is used to
pump effluent from the bioreactor and to power the air sparger
blower motor. The developer's calculations indicate that variations
49
-------
between the requirements for a 5 gpm and a 30 gpm system appear to
be nominal.
Heat: The amount of heat required at a particular site would be
dependent to some extent on the incoming water temperature,the
ambient temperature and resulting heat loss, as well as the
exothermicity of the reactions for a particular wastewater. With
the heat exchanger in use, the actual temperature difference
between influent and effluent is only about 5°F and is essentially
independent of the temperature of the water source except during
startup. At the MacGillis and Gibbs site during July to September
the average groundwater temperature was 55°F (13.25°C) because of
ambient heating and the heater was not used. For the cost
calculations, a 5°F difference in temperature is assumed.
Caustic; The cost for these estimates is based on the caustic used
in the demonstration program at a price of $2.60/gal of 50%
solution. For the MacGillis and Gibbs demonstration project, total
caustic use was about 15 gal over the six weeks. Depending on the
pH and alkalinity of the incoming water to be treated, more or less
caustic may be required at another site.
Labor; The vendor's experience is that operation of the 5 gpm
system requires only about 5 hours/week for operation, on-site
maintenance, replenishment of nutrient and caustic supply tanks,
and sampling for off-site monitoring. With the larger system, it is
estimated that labor requirements could increase to 10 hours/week.
Increasing the size of the system presents few economies of scsile
except in labor, which is very significant, and, to a much smaller
extent, the quantity of nutrients that will be needed during the
course of a cleanup.
Start-up of the system requires approximately two weeks and costs
would be about the same as two weeks of operation. Clearly,,
acclimation time and costs contribute a more significant portion of
total operating cost where a system is being used for a short
duration cleanup. •
Capital Costs
BioTrol, Inc. provided three bases to the capital cost of the
equipment:
a. A lease rate for a 5 gpm mobile unit of $2400/month, which
would be suitable for a short term cleanup;
b, A purchase price of $30,000 for a 5 gpm skid mounted
installation, such as might be needed for long term
treatment of a relatively low flow stream, leachate
from a pond, etc.; and
50
-------
c. A purchase price of $80,000 for a 30 gpm skid mounted
installation, as might be needed for a larger aquifer.
These capital costs also do not include the equipment or
installation costs for any pre- or post-treatment equipment nor the
cost of a concrete pad, building enclosure for the equipment, or
other site preparation. For the following comparisons, the capital
costs for both the 5 gpm and the 30 gpm units have been amortized
over 10 years and assigned a $0.00 salvage value at that time.
HYPOTHETICAL CASE STUDIES
Using the above operating and capital cost figures provided by the
vendor, various case studies can be formulated and conclusions
drawn as to the time and actual cost that would be involved in
remedliations of specific sizes. For example, if a site is expected
to yield 1,000,000 gallons of groundwater contaminated with 45 ppm
of pentachlorophenol, and all other factors are comparable to those
in the demonstration program, the most cost-effective scenario can
be predicted, as shown in Table 29. At 5 gpm, the demonstration
project approximates to the 0.1 MG example shown in the table.
While the 30 gpm unit would be the most cost-effective on these
bases, short term use of this unit is probably unrealistic. Even
for short term uses, lease of the 5 gpm mobile unit is more
realistic than purchase of the 5 gpm unit unless other uses can be
readily foreseen for the remainder of a purchased unit's life.
Mathematically, costs can be evaluated with the following general
equation using the appropriate operating and capital costs for the
system size selected:
Total Cost = (operating cost) x (total throughput)
1000 gal 1000 1000
+
( capital cost ) x (total throughput)
-amortize period throughput/day
Another way of examining the costs is to assume that a Superfund
site is a large reservoir of contaminated water. Costs can then be
estimated on the assumption that treatment (to a comparable removal
level) will be carried out at the site for an extended time, for
example, 5 years. For this analysis it would be assumed that the
equipment is 50% amortized by the use and has 50% salvage value at
the end of the five years. The results of such an analysis are
presented in Table 30. Of course, this analysis results in the
treatment of different volumes of wastewater over the course of the
cleanup.
51
-------
TABLE 29. ESTIMATED COST OF TREATING VARIOUS WASTEWATER VOLUMES*
5
System
0.1 MG 1 MG 10 MG ;
gpm skid mounted unit:
days to treat # 28 153 1.403
opereiting cost
$3.45/1000 gal $ 690 3,795 34,845
capital cost of unit @ $8.33/day 233 1,294 11^687
5
30
Total cost, $
Cost, $/lOOO
gpm mobile unit:
days to treat #
operaiting cost
lease cost ($2,
Total cost, $
Cost, $/1000
923 5,069 46,532
gallons
9.23 5.07 4.
65
DEMONSTRATION
2
$3.45/1000 gal $ 69
400/mo) 2,40
3,09
gallons j 3
8 153 1,403
0 3,795 34,845
0 12,000 112,800
0 15,795 147,645
0.90 15.80 14.
76
gpm skid mounted unit :
days to treat #
opereiting cost
capital cost of
Total cost, $
Cost, $/lOOO
17 37 243 ,'
$2.43/1000 gal 1,713 3,900 25,770
unit @ $22.22/day 378 822 5,422
2,091 4,722 31,192
gallons 20.91 4.72 3.
12
* At an assumed 45 ppm PCP concentration
# Acclimation time of about 2 weeks is included in the "days to
treat" and in capital and operating costs at the same operating cost
TABLE 30. TREATMENT COSTS OVER 5 YEARS (1800 days)
Cost item
MG water treated
operating cost
equipment cost
total cost
Cost/1000 gal
5 gpm
skid mounted
12.96
$ 44,712
$ 15,000
$ 59,712
$ 4.61
5 gpm
mobile
12.96
$44,712
$144,000
$188,712
$14.56
30 gpm
skid mounted
77.76
$188,957
$ 40,000
$228,957
$ 2.94
An equation that can be used to estimate total cost on this basis would
have the following form:
Cost = [BATS rate x total time x operating cost] + depreciation
52
-------
SECTION 7
CONCLUSIONS
A mobile (trailer-mounted) BATS with 5 gpm capacity tested at the MacGillis
and Gibbs Company site under the Super fund SITE program demonstrated the
ability of the system to remove pentachlorophenol from groundwater. At the
flow selected as optimum for the system, 5 gpm, removal of 96+% was
achieved and an effluent with about 1 ppm of pentachlorophenol was
attainable. At lower flow rates, 1 gpm and 3 gpm, pentachlorophenol removal
increased to 99+% and final concentrations down to 0.1 ppm were achievable.
Polynuclear aromatic hydrocarbons probably were also removed, either by
biodegradation or by adsorption on/in the biomass but low concentrations in
the groundwater source make it impossible to reach a firm conclusion.
Secondary pollutants such as oil, suspended solids, and even heavy metals
do not appear to interfere with the reaction, at least at the
concentrations present in this wastewater. Decreases in the Total Organic
Carbon (TOC) beyond that attributable to pentachlorophenol suggest that the
system removes other organic species as well.
Biomonitoring demonstrated that acute toxicity present in the raw
groundwater was essentially totally removed. Coupled with the measured
removal of specific chemical species, this suggests that any form of
discharge -or reuse would be safe for this wastewater.
Operating cost for the BATS ranges from $3.45/1000 gal when using a 5.gpm
system to $2.43/1000 gal when using a larger (30 gpm) unit.
53
-------
SECTION 8
BIBLIOGRAPHY
1. Bishop, J., GAINING ACCEPTANCE, Hazmat World, June, 1989, p. 37
ffd.
2* Brown, E.J., et al, PENTACHLOROPHENOL DEGRADATION: A PURE
BACTERIAL CULTURE AND. AN EPILITHIC MICROBIAL CONSORTIUM, Applied
and Environmental Microbiology, July, 1986, p. 92-97.
3. Bourguin, A.W., BIOREMEDIATION OF HAZARDOUS WASTE, HMC, Sept/Oct
1989, p. 16 ffd.
4. Dworkin, D. and R.M. Shapot, INNOVATIVE REMEDIAL TECHNOLOGIES AT
A CREOSOTE-BASED WOOD TREATMENT PLANT, presented at HAZMAT
'88, Nov. 8-11, 1988.
5. EPA, BIOLOGICAL' TREATMENT OF CHLOROPHENOLIC WASTES, Final
Report, Project No. 12130 EGK, June 1971.
6. EPA, METHODS FOR CHEMICAL ANALYSIS OF WATER AND WASTES, U.S.
Environmental Protection Agency, Environmental Monitoring and
Support Laboratory/ Cincinnati, OH., EPA-600/4-79-020. Revised
March 1983.
*•
7. EPA, SUPERFUND INNOVATIVE TECHNOLOGY EVALUATION (SITE) STRATEGY
AND PROGRAM PLAN, EPA/540/G-86/001, December 1986.
8. EPA, TEST METHODS FOR EVALUATING SOLID WASTE, SW-846, U.S.
Environmental Protection Agency, US Government Printing Office,
Washington, DC, Third Edition, November 1986.
9. Finlayson, G. , MICROBIAL CLEANUP OF TOXIC WASTES MAY PROVIDE
ALTERNATIVE SOLUTION, Occupational Health and Safety, Jan 1990, p.
36,38,40,57.
10. Kennedy, M.S., J. Grammas and W.B. Arbuckle, PARACHLOROPHENOL
DEGRADATION USING BIOAUGMENTATION, Research J. Water Pollution
Control Federation, 62 #3. 227-233 May/June 1990.
11. Lee, L.S., et al, INFLUENCE OF SOLVENT AND SORBENT
CHARACTERISTICS ON DISTRIBUTION OF PENTACHLOROPHENOL IN OCTANOL-
WATER AND SOIL-WATER SYSTEMS, Environmental Science and Technology,
24., |5, 654-661 (1990).
54
-------
12. Lee, K.M. and H.D. Stensel, AERATION AND SUBSTRATE UTILIZATION
IN A SPARGED PACKED-BED BIOFILM REACTOR, J. Water Pollution Control
Federation, 58 #11. 1066-1072 (1986).
13. Loper, J.C., THE FUTURE OF BIOREMEDIATION: PROSPECTS FOR
GENETICALLY ENGINEERED MICROORGANISMS, HMC, Sept/Oct 1989, p. 24
ffd.
14. Mathewson, J.R. and R.B. Jones, COMMERCIAL MICROORGANISMS,
Hazmat World, June, 1989 p. 48-51.
15. Mueller, J.G., P.J. Chapman and P.H. Pritchard, CREOSOTE-
CONTAMINATED SITES-THEIR POTENTIAL FOR BIOREMEDIATION,
Environmental Science and Technology, 23. #10. 1197-1201 (1989).
16. Nicholas, R.B. and D.E. Giamporcaro, NATURE'S PRESCRIPTION,
Hazmat World, June 1989, p. 30 ffd.
17. PEI Associates, Inc., REMEDIAL ALTERNATIVES REPORT FOR THE
MACGILLIS AND GIBBS "CO. HAZARDOUS WASTE SITE IN NEW BRIGHTON,
MINNESOTA, for Twin City Testing and Engineering Laboratory Inc.,
State PMN Contract No. 13726, Subcontract No. 1-35086, PN 3686,
April 1987.
18. Piotrowski, M.R., BIOREMEDIATION: TESTING THE WATERS, Civil
Engineering, Aug., 1989, p. 51-53.
19. Rusten, . B. , WASTEWATER TREATMENT WITH AERATED SUBMERGED
BIOLOGICAL FILTERS, J. Water Pollution Control Federation, 56 #5.
424-431 (1984).
20. Saber, D.L. and R.L. Crawford, ISOLATION AND CHARACTERIZATION
OF FILAVOBACTERIUM STRAINS THAT DEGRADE PENTACHLOROPHENOL, Applied
and Environmental Microbiology, Dec. 1985, p. 1512-1518.
21. Skladany, G.J. and K.M. Sullivan, "DECAY THEORY" BIOLOGICAL
TREATMENT FOR LOW-LEVEL ORGANIC CONTAMINATED GROUNDWATER AND
INDUSTRIAL WASTE, Superfund '87 Conference (HMCRI), Nov. 17, 1987,
Washington, D.C. /
22. Steiert, J.G. and R.L. Crawford, MICROBIAL DEGRADATION-. OF
CHLORINATED PHENOLS, Trends in Biotechnology, 3. #12. 1985, p. 300-
305.
23. Stinson, M.K., W. Hahn and H.S. Skovronek., SITE DEMONSTRATION
OF BIOLOGICAL TREATMENT OF GROUNDWATER BY BIOTROL, INC. AT A WOOD
PRESERVING SITE IN NEW BRIGHTON, MN, Presented at 16th Annual EPA
Research Symposium, Cincinnati, OH, April 3-5, 1990.
55
-------
24. Topp, E., R.L. Crawford, and R.S. Hanson, INFLUENCE OF READILY
METABOLIZABLE CARBON ON PENTACHLOROPHENOL METABOLISM BY A
PENTACHLOROPHENOL-DEGRADING FLAVOBACTERIUM SPECIES, Applied and
Environmental Microbiology, Oct 1988, p. 2452-2459.
25. Torpy, M.F., H.F. Stroo, and G. Brubaker, BIOLOGICAL TREATMENT
OF HAZARDOUS WASTE, Pollution Engineering, May 1989, p. 80 ffd.
26. Twin City Testing Corp., REMEDIAL INVESTIGATION REPORT:
MACGILLIS AND GIBBS COMPANY SITE, NEW BRIGHTON, MINNESOTA, Document
#120-86-414 for Minnesota Pollution Control Agency, June 25, 1986.
27. Valine, S.B., T.J. Chresand and D.D. Chilcote, SOIL WASHING
SYSTEM FOR USE AT WOOD PRESERVING SITES, Proceedings of the 1989
Air and Waste Management Association/USEPA International Symposium,
p. 257-268, Feb. 20-23, 1989, Air and Waste Management Association,
1989, Pittsburgh, PA.
28. Wilson, S., THE IMPORTANCE OF BIOREMEDIATION, Pollution
Equipment News, Dec., 1989, p. 67 ffd.
56
-------
APPENDIX A
QUALITY ASSURANCE/QUALITY CONTROL EVALUATION
A-l
'(57)
-------
APPENDIX A
QUALITY ASSURANCE/QUALITY CONTROL EVALUATION
BioTrol Aqueous Biological Treatment of Wood Preserving Site Groundwater
TABLE OF CONTENTS
Section Page
List of Tables iii
1.0 INTRODUCTION 1- 1
1.1 CRITICAL ANALYSES . . 1- 1
1.2 DATA QUALITY OBJECTIVES 1- 6
1.2.1 Data Quality Indicators 1- 6
2 .0 ANALYTICAL QUALITY CONTROL ' 2- 1
2.1 COMPLETENESS 2- 1
2.2 PCP/PAH (TOTAL, DISSOLVED) AND SEMIVOLATILE
ORGANICS ANALYSIS : 2- 2
2.2.1 Holding Time 2-4
2.2.2 Method Blank Results .. . 2-4
2.2 .'3 Surrogate Recovery 2- 4
2.2.4 Precision and Accuracy 2- 7
2.3 CHLORIDE 2-9
2..4 TOTAL ORGANIC CARBON 2- 9
2.5 POLYCHLORINATED DIBENZO-p-DIOXINS/DIBENZOFURANS 2-12
2.6 METALS 2-12
2.7 VOLATILE ORGANICS 2-14
2.8 OIL AND GREASE 2-14
2.9 TOTAL PHENOLICS 2-17
2.10 RESIDUE 2-17
2.11 NUTRIENTS 2-17
2.12 FIELD DUPLICATES 2-17
2.12.1 Pentachlorophenol 2-18
2.12.2 Other Analyses 2-21
3.0 BIOMONITORING AND EMISSION MONITORING 3- 1
3.1 BIOMONITORING ' 3- 1
3 .2 EMISSION MONITORING 3- 1
(58)
-------
QUALITY ASSURANCE/QUALITY CONTROL EVALUATION
BioTrol Aqueous Biological Treatment of Wood Preserving Site Groundwater
TABLE OF CONTENTS (CONTINUED)
Section Page
3.2.1 Sample Collection , 3 - 1
3.2.2 Sample Analysis 3- 2
4.0 ADDITIONAL STUDIES '. 4- 1
4.1 PENTACHLOROPHENOL SAMPLING LOCATION BIAS 4- 1
4.2 EXTENDED PENTACHLOROPHENOL CALIBRATION STUDY :. 4- (5
5.0 NON-CONFORMANCES 5- 1
6.0 DEVIATIONS FROM THE QAPjP 6- 1
6.1 FIELD PROCEDURES 6- 1
6.2 SAMPLING SCHEDULE 6- 2
6.3 ANALYTICAL METHODS 6- 3
7 .0 AUDIT FINDINGS/RESULTS 7- 1
7.1 FIELD AUDIT, WATER SAMPLING 7- 1
7.2 FIELD AUDIT, EMISSIONS SAMPLING 7- 2
7 .3 LABORATORY AUDIT ... '. 7- 3
8.0 CONCLUSIONS AND LIMITATIONS OF DATA 8- 1
ii
(59)
-------
QUALITY ASSURANCE/QUALITY CONTROL EVALUATION
BioTrol Aqueous Biological Treatment of Wood Preserving Site Ground-water
LIST OF TABLES
Table No.
A.I
A. 2
Title
Sampling Summary: BioTrol Aqueous Treatment SITE
Demonstration Project '
Completeness Data: BioTrol Aqueous Treatment SITE
Demonstration
Page
1-2
2- 2
A.3 Method Blank Results: PGP/PAH and Semivolatile
Organic Analyses 2- 5
A.4 Surrogate Recovery: PCP/PAH (total, dissolved) and
Semivolatile Organic Analyses 2- 6
A.5 QC Summary: Precision and Accuracy - Semivolatile
Organic Compounds 2- 8
A.6 Method Blank Results: General Chemistry, Metals,
Volatile Organics and Dioxin/Furan Analysis 2-10
A.7 QC Summary: Precision and Accuracy - General
Chemistry Analysis 2-11
A.8 Surrogate Recovery: Dioxin/Furan Analysis 2-13
A.9 QC Summary: Precision and Accuracy - Metals 2-15
A.10 Surrogate Recoveries: Volatile Organic Analysis 2-16
A.11 QC Summary: Precision and Accuracy - Volatile Organics ... 2-16
A.12 Field Duplicate Analysis Summary 2-19
A. 13 Pentachlorophenol Field Duplicate Analytical Re.sults 2-20
A.14 Surrogate Recovery: MM5 Sample Analysis 3- 3
A. 15 PCP Method Comparison 4- 4
A. 1.6 Method Precision Data for GC/MS and HPLC
PCP Analaysis 4- 5
. iii
(60)
-------
APPENDIX
QUALITY ASSURANCE/QUALITY CONTROL EVALUATION
BIoTrol Aqueous Biological Treatment of Wood Preserving Site Groundwater
1.0 INTRODUCTION
As a part of the USEPA Superfund Innovative Technology Evaluation (SITE)
program, the BioTrol Aqueous Treatment System (BATS) technology was demonstrated
at the MacGillis and Gibbs Superfund site in New Brighton, Minnesota. This
appendix-.to the final report discusses the results of the comprehensive quality
assurance program associated with this demonstration. Included in this
discussion are: (1) data quality objectives (as defined in the Quality Assurance
Project Plan), (2) analytical quality control, (3) additional studies undertaken
during the project (including a method comparison study initiated in an effort.
to resolve an apparent discrepancy in pentachlorophenol influent concentrations
as a function of sample location and system flow rate), (4) non-conformances
encountered during the demonstration, (5) deviations from the QAPjP in field or
laboratory protocols, (6) findings of audits conducted during the demonstration,
and (7) conclusions and limitations of the data.
1.1 CRITICAL ANALYSES
The primary technical objective of this SITE demonstration was to determine
the extent of removal of pentachlorophenol (PCP) by the BioTrol bioreactor
treatment system. The project was also designed to obtain data on polynuclear
aromatic hydrocarbons (PAHs) removal. Additionally, a secondary objective was
to determine the fate and extent of removal of other semivolatile organic
compounds, dioxins/furans, volatile organics, metals, oil and grease and total
phenolics. Chloride ion production was determined along with total organic
carbon as supporting measures of pentachlorophenol degradation/destruction. The
toxicity reduction between well water/influent and effluent wastestreams was
determined by conducting acute toxicity tests. Losses from the bioreactor via
sloughed biomass and exhausted air were monitored by sludge analysis and
emissions sampling and analysis, respectively. Residue and nutrient analyses
were also performed in the laboratory as system condition measurements. Table
A.I summarizes the analyses performed, matrices sampled, location sampled and
frequency of sampling. Each of these analyses will be summarized in sections 2.0
1-1
(61)
-------
•M
•s-
SITE Dewmstration Pr
*J
5
g
1
_
1
1
iS
^
9
i
A
g
i
S
1
§
L
u.
4
1
8
•5
..
4)
i
a
5-g
11
Is
41
P
- n
St. CO
a u
*C «2 *C
u a u
Q Of ^»
09
C
1 „
u c t.
2 32
o —* u
SH- a
H- Ot
O * O
15 g 15 2
< 3 <8.
O 09
i- 3:
O < 7
— * a. 09 35 w
.C **N Z ^ 31
U — * < — » <
to o CL o a.
-W C >* 0) "X,
CO! O. OJ Q.
41 JT U •»- O
tx ^L & rx **••••
41 4>
44 I?4*1
8. -g 5 1
U U Ik U
fi» &. &*
«l 0> V
«4 *J 4-»
5 53
g s s
a a a
N M M
„- ^ ^
w 1 o>
a c sj c a
- J - J -
S £ S £ S
• i
_3 ^
O 09 O OS
a. — • a.— •
•a a a
S to >>w S--M
"52 c S --JE •-«
W -~. X X OW 0 w
01 Q. O O •— C — C
•v>u *•• M— C> 4) 41 4) 41
4^ -M 4-» 4-* **
€8 OT O O j8
1-2
(62)
g g
S 5
¥ - fe -
a w a *• a
o u _o o _u
u .c u - f t-
u u u u u
Ji J* J< Ji — •
01 Q) 01 0) •—
1 1 s i s
s g
33 L.
— • — • O
4> (• 01 t* (O
§a c fi o
_ ^ 2 _ S 1
C 4) C 4) "^ O
•^ 3 *•* 3 4 ^L
S 8
"§ *S'
S 6 •- 2 i S
4J OJ 4-* (0
0) 41 .
?
01
3
c.
0>
ID
_
S
Q
11
^5
0 c
-^ Q
a 2
«•» c.
O ID
J2
u
(3
C.
01
4J
0
3
-------
il
»••*
oC
g
IP'
19
1.
i
g!
xj (7!
*** iE
£ i
i r
w H"
• S!|
** i>
.-3 |4"
if jr
•"' J
g
•I
>»
Sampling Su
w
1°
"*•
u.
V
I
g
f~*
Parameter
'"B
n
(
SS
1
g g
1 1
i_ U
"a 4^ 4^
o u S o
— ca at to
•IB CO CO CD
c- fi •>. fi
«J u en «j
1 "cp S "«
£ 3 £ 3
.3 2
4-» O
01 C. 41 i.
01 o> t- O
U 4-» ffl 4-»
^ « S .2 co S
"° c — "° c —
51 1 38. 1
CO CO
o o
7 -o "? "p
5 5 .. 5 •• S
•2 .2 S"S § "8
8~" "o -Q S -Q S
to — a. •— a.
•"• O •"» O 41 3 Q> 3
ot- OP— "oeiA eecA
4) ^^ 4}
•n 4J — H
0>~- u, —
&. CO ** CO
II 2 1 2
iZ CJ C3 U (3
u c, c. u
49 41 Of 41
S S 5' S
g g
1 I
c- ~
41 V
U 4^
V 0 V CO O
co a co CD ca
^•5 £ S Sf
1 is ' I- s s-
Z 4) •t3 4> 'D
2 S 2
St. CO C. CO
41 a> tu ai
*- 4-* t, -M t.
1 S |« 5 .2«
— S _ •** c
5 1 il I 5S.
•^ « T3
CO CO _. ^
4J 4J 41 4^
.. « ^ .. -S j3 4) 1^3 !it3 "003
•-"a- '-fflQ. vco » B. 6 a. 6 o o £
«
Q. ^ J3 ^ _ 5
g CO CO CO CO .
5 t- -.(._,
u o u a a
C- _. C. C- C.
41 Ot 41 41 01
1 . ^ ^ 4rf |^
S 5 S S 3
1-3
(63)
§
a
N
*U
41
U
L,
CD
U
1
^
41
««
3
i
_
4) «
Dissolved
Oxygen (fi
measuremen
CO
(3
1 '
g
+•*
1
I
fe
U
•— CO
5S.
i
2
3
41 M
Temperatur
(field mea
ment)
1
1
S
1
L.
I
CD
5
1
2
. .
I
^^
01 C
Temperatur
(field
measuremei
_S
2
<9
C.
41
I
-------
4V
Q
Q.
g
4V
a
L.
4V
O
I
a
I
i
2
D-
O
00
^
B
55
g
I
4rf
S
8°.
i!
Y
u.
«•
s
1
§•
«
b
I
1
P-n
- 8
.5
•|
S g g
** ** 4^
a a a
M N N
4> C 4^ 4^
S o u o E u
S c — a o a
4J L. 4J t. 4J I_
>• JZ t- J= >* JC
H» O U U CO U
*U J^ .^ \x 4) w
V CI O ' O 2 CJ
3 5 S Ot Kl 3»
4-* +4
* 1 1
4J «*- «*-
O »*- «4-
m u u t_ «> i_
« «> ^ 0) v» «|
o m c o c a
S2 * J * J *
51 S . I S £ • I
4V 4J 4_»
CO CO CO
g* (0 CD 0) g 8"
O 13 O C3 CJ CJ
L. t_ t. U t_ L.
.S
4V
N
^ t-
9 a 4V
u u o
5 5 g
3 ' «» —
111
3 10 3
4^
S C.
-• 0
H- 4-*
H- O
oat.
>» v a
4^ L. 4J
1 £« • S
I f S =
— < S. 3
s'
v e>
« m
a ca
>- e> cj
4-* t- L.
— a o
H 1 1
o a — • -^
CO H- O "5
V
1?
§- •§ -s
O L. t.
U U U
«- «- t-
tt O Q>
4J *J 4J
3 5 •. 5
c a.
u o
4^ 4>
t. (_
O U
1
_C J=
•*- v
« o a. >•
*^ M ^J ^^
_£ w »
<*C S 3
O O 4J
a u —
4^ X
to c v
<-J
O t- O
o o u
I- 4V 4V L,
o — •— o
•—XX -^
ca v v m
4»
t -*
t— TTT 4V
o < c *-*
-^ o. -v to ee
^ "N. o u nr
SS § w a.
VC O> L. CJ
a. a.%*r o a.
& ^1
— ™
g £11
1_ •— C
•
I 1
4V 4J
X X
S S
t_ c_
o o
•M 4V
u u
CD CO
V V
L. U
F m m
CO 4V
U X C
§""* "o *u a ^
4v o — * a
O »— — 4V
.e 4v i_ o o
a. w a. a. z
g g
VS ^s
O CT O D)
QJ.*- (g >^>
- g r g
U. (J U. U
f^ 0 T?
«i o> a>
CA O (0 O
O — ' O — '
.^
-------
$
'e
a.
g
I
1
a
S
M
i
i
§
I
i
s
m
d*
a
i
B
|
MO.
ll
2 H-
•v
1
u.
«„
•5
£
1
'
2
£
• J
t*
„
"a
u
£
u
2r
1
**
'i
c-
o
44
Bioreac
1
OS
tion/
fugation
o —
u- 1.
u. O
«IT»
fl
3 ?
13
•~ ia
a v»
"3
u
44
C.
U
i
I
4-*
'i
o
44
Bioreac
ox ins
a
tion/
fugation
Zt
IS
«i p
fl
§£
— to
CB >x
«
"m
u
'u
u
u
1
_^
i
•«
c.
o
44
m
1
a.
2
I
'5.
S
^
a
09 V
' "^
€ **
o u
••>
-------
and 3.0, including any potential bias, outlier data and precision and accuracy
data.
1.2 DATA QUALITY OBJECTIVES
Quality assurance objectives were quantitatively defined in the QAPjP for
precision, accuracy and completeness for the critical measurements for the
project (excluding biotoxicity). A primary concern was the establishment of a
comprehensive quality assurance program with defined objectives to ensure that
the data produced during the demonstration were of known and acceptable quality,
thus allowing an evaluation of the treatment technology and achievement of the
project's technical objectives (the primary objective being the determination of
the extent of removal of POP). Overall, analytical measurements met data quality
objectives with the following exceptions.
o High levels of PGP in the samples required that many semivolatile
extracts be diluted prior to analysis by GC/MS. As a result, a
significant amount of surrogate recovery data was unobtainable,
since the necessary dilution factors brought the surrogate compound
concentrations below the detection limit. However, the available
data suggest that overall approximately 90% of the recoveries were
within control limits.
o The high levels of PGP native to the samples also affected the
calculation of pentachlorophenol spike recoveries. For half of the
matrix spike duplicate pairs analyzed, the sample PGP concentration
was 6-20 times higher than the spike concentration and therefore
recovery could not be determined. The remaining available data met
precision and accuracy objectives overall.
o One set of emissions samples (from the 3 gpm flow rate) had signi-
ficantly low surrogate recoveries for the semivolatile organic
analysis of the MM5 sampling train, indicating that these results
are biased low. However, the sampling episodes from the 1 gpm arid
5 gpm flow rates, which had normal surrogate recoveries, indicate
that losses to the exhaust air were minimal, and that no PCP was
present in these emissions.
o Completeness objectives for sludge (sloughed biomass) collection/
analysis were not achieved for reasons detailed in section 6.1.
1.2.1 Data Quality Indicators:
Quality control analyses associated with the data quality indicators of
precision and accuracy include matrix replicates, matrix spikes and matrix spike
1-6
(67)
-------
duplicates. Precision, the ability of the measurement system to generate
reproducible data, was calculated as the relative percent difference (RPD)
between the results of duplicate samples or matrix spike/matrix spike duplicate
(MS/MSD) pairs using the following equation:
% RPD - (Rx - R,) x 100
(Rj. + R2)/2
where Rx and R2 are the results of the sample and duplicate, or the MS and MSD,
analyses. Accuracy, defined as the nearness of the analytical result to the
"true" value, is assessed by the analysis of matrix spikes (or MS/MSD pairs) and
reported as percent recovery according to the following equation:
% Recovery - Ci - C0 x 100
Ct
where CL — the measured concentration in the spiked sample, C0 — the measured
concentration in the unspiked sample and Ct - the known concentration of analyte
added to the sample. The data quality indicator of completeness is measured as
thes comparison of valid data obtained during the demonstration with the amount
of data expected:
% Completeness - Valid data points/Expected data points x 100
Completeness is discussed in section 2.1, and precision and accuracy are
discussed for each of the analyses in sections 2.2-2.11.
1-7
(68)
-------
2.0 ANALYTICAL QUALITY CONTROL
The following sections review the measure of completeness achieved for the
project, followed by discussions of the analytical QC results for each of the
parameters measured.
2.1 COMPLETENESS
QA objectives for completeness for all critical measurements were given in
the QAPjP, generally as 90-95%, and. results are summarized in Table A.2. The
data presented in this table include the number of samples and duplicates
proposed for collection in the QAPjP, the number actually collected during the
6-week demonstration and the number of samples analyzed. (Sample analyses which
may have exceeded QC criteria for surrogate recoveries are included in the totals
given for PGP/PAH, semivolatiles and emissions analysis).
Almost all analyses met the QA objective for completeness, with the
exception of PCP/PAHs-(total and dissolved) and sludge analyses. The treatment
technology was evaluated using average PGP concentrations for well water,
influent and effluent waste streams. The loss of eight individual data points
does not seriously impact data quality or the achievement of the project's
technical objective, and as noted in Table A.2, the completeness objective only
fell one-two percent short of the QAPjP objective. Sludge collection/analysis
was hindered by the type of sample collected (see section 6.1). Sludge data was
used qualitatively to estimate the potential PGP loss due to ad/absorption; thus
data quality was not greatly impacted.
2.2 PGP/PAH (TOTAL, DISSOLVED) AND SEMIVOLATILE ORGANICS ANALYSIS
Aqueous samples were analyzed for PCP/PAHs (on samples as collected and
after filtering), and occasionally for full-scan semivolatile organics, by SW846
method 3510/8270. Solid/sludge samples were analyzed for PCP/PAHs by SW846
method 3550/8270, and were phase-separated by centrifugation prior to extraction
of the solids. Method blank data, surrogate recovery results, and precision and
accuracy data (MS/MSD analysis) are summarized below. In addition to these
quality control analyses, method specified protocols for calibration (tuning,
multi-point standard curves, continuing calibration standards and internal
standards) were followed and met QC criteria.
2-1
(69)
-------
U Q.IF*
ill
i s
8 -M
I »
£ 3
•O •* <- •-•• ••(
o-5
*
in m
-------
TABLE A.2 (Continued)
Notes;
NC = Not a critical measurement; completeness QA objective given for guidance
N/A = Hot applicable
1. Does not include field, trip or equipment wash blanks
2. Aqueous matrix includes: well water, influent, effluent, reactor midpoints and carbon effluent
samples.
3. One influent sample lost in field; one influent, effluent and midpoint sample lost in lab (due to
breakage).
4. One influent sample lost in field, one influent sample lost in lab (both due to breakage).
5. One influent sample lost in field (breakage).
6. An extra duplicate was collected for this measurement parameter. Completeness is calculated based only
on the # of samples without the extra duplicate.
7. QAPJP originally called for 2 samples/week. An additional well water sample was added the last 2
weeks.
8. Sludge collection severely hampered by type of filter used, generating wastewater-type sludge samples.
The limited volume available prevented samples being collected during weeks 2 and 3 and made a field
duplicate impractical (as qualified in the QAPP). Completeness calculated without including proposed
duplicate.
9. Uncertainty about the opportunity to collect an active biomass sample was clearly documented in the
QAPP. (Section 5.4, Table 5-5, Table 5-8, and 5.4.3.2-page 29) A single sample was obtained the last
weak of sampling. A completeness objective/calculation is not appropriate given the nature of the
sampling activity.
10. 'The QAPP proposed an initial sampling episode at the start of the project, with subsequent bioreactor
and carbon effluent being sampled once per test period, for a total of 8 samples. Three consecutive
samples were to be taken with the first sampling episode, thus 1 sample and 2 "duplicates"; since both
effluents were sampled there are actually 4 "duplicates". The initial sampling episode prior to the
start of the demonstration was not performed due to schedule constraints; therefore only 6 samples were
collected. The QAPP was not fully revised to reflect this altered schedule.
11. This total does include samples which had surrogate recoveries outside of control limits due to
suspected matrix effect (see section 2.2).
12. An extra effluent sample was analyzed for dioxins/furans.
13. Eighteen samples were specified for collection in the QAPP; three grab samples were to be collected and
analyzed in the laboratory as a composite (excluding the raw well water sample). Following the lab
audit, the decision was made to analyze each individual grab separately.- thus the actual number of
samples analyzed were 38 and the duplicate was actually 3 grab analyses.
2-3
(71)
-------
2.2.1 Holding Time
The method-specif led holding time for aqueous samples (7 days from sample
collection to extraction and 40 days from extraction to analysis) was exceeded
for the influent, midpoint 1, midpoint 2, effluent and equipment wash samples
from day 3 of the 5 gpm flow test (STI-C-03-02/03/04/05/13) , and were met for all
other water samples. The results for samples exceeding hold time were not used
in the evaluation of the bioreactor. The method-specified holding time for solid
matrices (14 days from sample collection to extraction) was exceeded for two of
the sludge samples by one and five days; all samples met the analysis hold time
of 40 days. The elapsed time was considered to have minimal impact on data
quality and these data were used in evaluating the extent of PGP removal via
sloughed biomass. In addition, sludge data was used only as a quantitative
estimate as previously stated.
2.2.2 Method Blank Results
Method blank results are summarized in Table A.3. The only consistent lab
contaminant detected in aqueous method blanks was bis(2-ethylhexyl)phthalate (at
an average concentration of 22 ug/1.) Phthalates were only a target compound for
the 19 samples (and associated field blanks) analyzed for full-scan semivolatile
..organics. At the detected blank levels, there was no impact on sample data
quality.
2.2.3 Surrogate Recovery
Seven surrogate compounds were spiked into each aqueous sample prior to the
liquid-liquid extraction procedure. The percent recovery ranges and comparison
to control limits are give in Table A.4. As discussed previously, a significant
number of surrogate recovery data points were unobtainable when dilutions
required, due to high sample PCP concentrations, resulted in surrogate
concentrations below the detection limit.
Of the analyses for which surrogate recoveries were generated, an overall
90% of the recoveries were within specified control limits. However, approxi-
mately 18% of the samples had 2 or more surrogate recoveries (more than one each
of the acid and base/neutral compounds) outside control limits. Significantly,
two-thirds of these samples were collected from the second bioreactor midpoint
or the effluent, mostly from the first two weeks of the demonstration (1 gpm flow
'2-4
(72)
-------
TABLE A.3
METHOD BLANK RESULTS: PGP/PAH AND SEMIVOLATILE ORGANIC ANALYSES
Number
Compound of Blanks
AQUEOUS MATRIX BLANKS
Diethylphthalate 22
Di-n-butylphthalate 22
Bis(2-ethylhexyl)phthalate 22
Di-n-octylphthalate 22
Phenol 22
2-Chlorophenol 22
4-Chloro-3-methylphenol 22
All other compounds 22
SOLID MATRIX BLANKS
All compounds 3
* — Also found in 2 additional blanks
No . Above Concentration Detection
Detection Range Limit
Limit (ug/L) (ug/L)
1
1
12*
1
1
1
1
0
0
at levels
34
16
13-140
17
14
16
11
- -
-.
just below the
10
10
10
10
10
10
10
10 - 50
330 - 1600
mg/kg
detection limit.
2-5
(73)
-------
-r
tu
1
g
V
1
u
S
S
a
4-1
o
1
1 •
M
1 s
1 g
1 a
U M
"3 <
0
1
Ss
0.
t
i
J
£
1
^
fO
»• 01 3
O 0 5
L. >.T>
|||
O
CO
+J
t-S'l
f'n->
22
z 5 *•
CA
'i
5 IE """"
li|
Sot
£
11
U .J
COM
4) > 0)
u o in
£ S e
a! ac
(M
O CO
II
~
"o ^
U 01
Ji-s
•s a-
I i
2 en
_j
§
|
CJ
4)
1
c_
2
2fc~fcfc5~| »
*"
o co tn o o N- *oi *-
inin^OKi^^otnj co
»-
•>* ^O *- « O M O
*- *- -r «* o (M in
SJSRSSISS
*~ «Q r*t ^ tXJ »A in
O OO OO 0 O.
tMCMIMOJ- u ra
^ "S. *? * *~ ° *^i
*^ U- b. O) U. *** 4^ H-j
CO •
01 4-»
S. g
1 2
0 «
I i
0 t
a *
*•* p-
2 ."
.2 -5
j*^ £
-S 3
w **.e
f J 5
g§l
111
* wl
"S u'~
511
0 ** •
S 8.
0 g ^
- f-a i
J3 |^ t! !
1. ^a.. •
c . — IS S
llili I
if|2|| »
0 •£ =5 I S
(0 (O
§ t» 5 "3
S§ 01 9
•o«3S;>.
2- s« §
i-lSlJ
** *• ** >.
^s1|i
H— «» 3 S
slEi's
Ills!
GO o a a
— *> CD g
'•211^7
to CO «9 •> C
f iH5!
Illll
^
-------
rate). Factors which appear to have contributed to this matrix effect include
the1, suspended solids in the effluent which were highest during the first part of
the demonstration (average TSS=54 mg/1 at '1 gpm, 26 mg/1 at 3 gpm and 18 mg/1 at
5 gpm), and throughout the demonstration the samples collected at the second
midpoint of the bioreactor and from the effluent would have been the most
affected by biological activity (including any degradation by-products) thus
causing a residual microbe population which may have interfered with the
extraction or analysis.
In addition, the surrogate compounds d5-phenol and 2-fluorophenol were
primarily affected and biased low; tribromophenol recoveries were reported both
above and below the control limits and more frequently within limits. Similarly
PGP concentrations did not appear to be biased low for samples with low surrogate
recoveries - a review of influent samples from the 1 gpm flow rate indicated that
..the average PGP concentration was higher for samples (six out of twelve) with
surrogate recoveries below the control limits than for samples with surrogate
recoveries within limits.
It appears reasonable, therefore, to conclude that the low surrogate
recoveries observed for d5-phenol and 2-fluorophenol are attributable, at least
in part, to matrix effects, particularly for the bioreactor midpoint 2 and
effluent samples from the first two weeks. It is also significant that extrac-
tion recovery/efficiency was generally within control limits for tribromophenol
and for spiked pentachlorophenol - which for most samples was the only target
acid extractable compound. Therefore, given the low percentage of samples-with
surrogates below specified limits there is considered to be minimal adverse
impact on data quality basing conclusions on average PCP concentrations as a
result of the low-recovery for the .two surrogate compounds. In addition, the
comparison study of methods SW3510/8270 and the developer's HPLC method (which
is not dependent on extraction efficiency) shows PCP recovery was not adversely
affected in any given matrix (see section 4.1).4
2.2.4 Precision and Accuracy
A summary of aqueous sample MS/MSD analysis results is presented in Table
A.5. The full, complement of spiking compounds was included for each MS/MSD
analyzed although in many cases only PCP and PAHs were the target compounds. Due
2-7
(75)
-------
§
s.
1! a
a 5
(A 4-* •*•
SSJ
O U -J
>.
28
a!£ w
oooooruoo«-o «
*2
II
8S
£«
«s
U
-§,
s
•o
a
$
11 '5?
o a. t- *•
(MocooocotMt-ocoov- 2" c o
••f^rrufncM^fnmfrimft oo So.
«•= ^ 2 ?>
«_ - . g«-
g. . ° o
-Z"S .2s
°« Is
W •— 5 «B •
OtiAOth»^ro* Ji • 7 TJ
r\j«-«-ru*-xT(\jrotnrONr £« _ a o
^: is ?
O» Q "S W °
*2 fet J?
00 — a _
S So SS 2
, S §i || 1
:--! - || 1-: I
g" "S£ •»
•2 b «- -
S3! S 3 a
z ^ H> 4**
•I I «! 2
s ;i II *
.^•olu; -Jg .S«
li 2 il »
N | «| g
ug I S| 5
sssaasasssssi s g.-= s g§ .
sS S IS I
* . « s I.
• i i i i i i i t i i S ** oi o M0 *^
sifcs^asssa^s s-gi 1 |I g
ill 5 |1 8
-o«««^a—afe HI j M I
ii*Tiii*t"iitT u'o 2L -^ a a . S« §
||g f Sc -I
M e U o U 8 <-
o a. 9--c o t- i .c _ _ _
."S g f 5 ** ' 2 ** " ""*
>C £1 «t i-iui »4»X —1
.p. <'^cM&a. <
§? S
WO H-
H» T3 —J
°lt
(8 5.U f
^J f
o o v u
4^ O
-------
to the elevated background level of PGP in some of "the samples spiked, penta-
chlorophenol recoveries were not considered meaningful and are not reported if
the sample concentration was greater than 5 times the spike concentration.
Precision measurements for all PGP MS/MSD pairs reported were within
control limits (RPD - 0-50%); RPD values were not reported for the MS/MSD pairs
which had elevated PGP sample concentrations. However, if precision is calcu-
lated for these analyses using the observed spiked sample concentration (as
opposed to using the calculated % recoveries to determine the RPD), then
additional precision data is obtained which is particularly relevant for high PCP
concentrations. This measure of precision generates an RPD range of 0.8-17.2%
with all values within control limits.
Matrix spikes could not be performed on sludge/solid samples, due to
limited volume. .
2.3 CHLORIDE
During the initial stage of the analysis program, aqueous samples were
analyzed for chloride by SW846 Method 9252. After the laboratory audit (see
section 7.3), the method was changed to Standard Method 4Q7C for the remainder
of the program. All samples were analyzed within the method specified holding
time of 28 days. Method blanks (see Table A.6) were analyzed with each sample
batch, and chloride was less than the detection limit for all blanks. Precision
and accuracy were determined by the analysis of matrix duplicates and matrix
spikes/matrix spike duplicates (see Table A.7). Initial and continuing
calibration standard" analyses met applicable QC criteria.
2.4 TOTAL. ORGANIC CARBON (TOG)
Samples were analyzed for TOG as collected and after filtering (soluble
total organic carbon, STOC) by SW846 Method 9060. All samples were analyzed
within the 28 day method specified hold time. One method blank had a TOG valtae
of 1.4, all others were not detected at
-------
TABLE A.6
METHOD BLANK RESULTS: GENERAL CHEMISTRY, METALS, VOLATILE ORGANIC,
AND DIOXIN/FDRAN ANALYSIS
Parameter-Matrix (units)
Chloride-Aqueous (mg/L)
Total Organic Carbon-Aqueous
(mg/L)
DIoxins/furans-Aqueous (ng/L)
Number of
Blanks
28
30
Number Above
Detec. Limit
0
Concentration Detection
Ranee
1.4
Limit
OCDD
Others
Dioxins/furans- Solid (ng/g)
Metals -Aqueous (ug/L)
Chromium
Nickel
Lead
Metals - Sludge (ug/L) <2)
Volatile Organics- aqueous (ug/L)
Methylene chloride
Others
Oil and Grease - aqueous (mg)
Residue - aqueous (mg/L)
Nutrients -Aqueous (mg/L)
Nitrate -nitrite
Ammonia
Total phosphate
Total phenolics- aqueous- (mg/L)
6
6
3
6
,6
4
2
11
11
14
17
11
12
15
7
1
0
0
1
1
3
0
2<3)
0
9
0
0
0
0
0 '
21
--
--
9.0
31
1.4-3
--
16-29
. -
0.2-3
--
--
--
--
<4.5<"
<0,24
<7..1
<31
.9 <1.,2
variable
t
<5
<5-<10
.2 0.0
<1
<0.02
<0.02
<0.02
<0.005
(1) MDLs varied from 4.5-11 ng/L
(2) Due to the aqueous nature
of the
(3) Four additional blanks contained
sludge, an aqueous
blank is
methylene chloride at levels
applicable
just below the
detection limit
2-10
(78)
-------
1
(0
£
1
<
L.
CO
I
Q)
5
2
S '
S
(^ i
J X
^ o
- 1
|2 S
^
1
g
(A
O.
,,
u"
g
c/>
u
•S-
— o «
« c. *J
III
*u~1
c
'Jc "o w
•*• w «t-
•* J5
.
! _
x 22
II
CJ _J
<0
oc a
oe
«
4k t*
— • >.
4-» a
•§-—
— o w
CO U *»
III
»
!e 'o m
111'
Is
si-
0 •-
w t M
S > S
IIs
*s
-^ (0
2 >•
fc
«•«
U
9
a.
CM O '*O
in o M
CM CM
tn tn o
(M CM Kl
0 0 O
CM
! ? ?
000
^«
r^ o C
(M CM 1*1
CM O O
s a •*
o o tn
CM CM CM
1 1 1
S § 1C
*• **
S i 1
in ^O o
fc "s sc
CM CM *»
S
~s
3
£ 1
i 2
CD U
•S o -52
'C - a
o a
u t- o
o o rvj o o
**• >o vr o
- S3J3S S
~ i i ^ i
in fo-o o
0 0 O 0 0
-J- ^ ^o
^
Z CM CM O O
Z 2S2 t>
A
O O O 0
(M CM CM CU
< till
«v a o o o
z a coca co
o> in
^2in CM
1 i «~ O
5 ^-^2 S
Z *O N- CO &
* so -O O
z ^ *• *~ o
4^4^ tO
II I
« ? i
41 C •— ** "S.
15 5 c « — • ••
T5 •— Q t- O - «* t/»
(A *-• 5 ^~ O M . 1—
2-11
(79)
|
4)
.C
'3
I
X
1
Q
tL
i
1
•g
to
^
y
f«
u
u
CM
<5
^
g
=
i
s.
"3
a*
^
"K.
to
X
c.
M
1
4*
a
u
1
X
*u
1
"S?
1^
Hs
?T
3
1?
°- 3
Is
§•=
«l >
'S.o
S 2
Limited si
batches;
^>
1
VI
CM
oc
1
iH
»
s
V
•-
n
*
~
**»
%
O
in
Q
(M
S
DC
tn
u
-M
g
t_
g
U
^
i
A
CO
_g
<0
i
**•
i
4^
2
g
1
to
5
X
-5
CO
a
£
Control I
CM
,.
-------
Precision and accuracy for TOG was evaluated -by the analysis of matrix
duplicates and matrix spikes and measurements met QC objectives. In addition,
QC check standards were analyzed and had recoveries ranging from 96.8-103%.
Calibration standard analyses met QC criteria.
2.5 POLYCHLORINATED DIBENZO-P-DIOXINS/DIBENZOFURANS
Aqueous and sludge samples were analyzed for dioxins/furans by SW846 method
8280. Sludge samples were phase separated prior to extraction by decanting the
water and only the solid fraction was analyzed. The procedure involves a matrix
specific extraction and cleanup, followed by GC/MS analysis. Sample extracts,
blanks and standards were spiked prior to analysis with 13C12-1,2,3,4-TCDD as a
quantification standard, then analyzed by selective ion monitoring and
identified/quantified based on isotope dilution techniques. All samples were
analyzed within the method-specified holding time. of 30 days from sample
collection to extraction and 45 days for complete analysis.
All samples and,blanks were spiked prior to extraction with labelled
(13C12-) PCDDs and PCDFs. Surrogate recoveries are summarized in Table A. 8. Most
compound recoveries were within control limits. In one instance, a sludge sample
.(STI-A-6-7) had high recovery of the labelled OCDD and OCDF surrogates (144% and
141%, respectively); the sample results reported for OCDD (1900 ng/g) and OCDF
(41 ng/g) may consequently be biased somewhat high. For most of the other
samples having surrogate recoveries outside of control limits, the related
congener was not detected. Two sample results were reported as NC (not
calculated) for PeCDF due to the associated surrogate being non-detected.
2.6 METALS
Aqueous samples and four sludge samples were analyzed for priority
pollutant metals by SW846 methodologies. All samples were analyzed within the
method-specified holding times of 28 days for mercury and 6 months for the other
metals. Sludge samples were analyzed "as is" without phase separation, and
reported as ug/g, wet weight.
Method blanks were digested/analyzed with each batch of aqueous and sludge
samples. Three of the four lead analysis method blanks contained lead, at
concentrations ranging from 1.4-3.9 ug/L. In addition, four equipment wash
2-12
(80)
-------
23
CO
<
1
S iSH
| * s
1
15
••
o . o o o o
OJ CM fU CM CM
t i i i i
00 0 0 0
CM S
i i
O O
o o o
OJ CM *NJ
000
** "^ *«*
•5
I
V
C.
C9
K S
I
1
s
CD
O
"3.
i
u
£
a u.
8 S
u. o> u
8
co
h.
Kl
CM
i
CM
O
en
8
co
N.
m
CVJ
i
CM
O
en
£
CO
i»_
Kl
.«\J •
i
CM
CJ
en
a.
cb
N.
K1
N
i
U
en
«_
-O
IO
IM
i
CM
en"
t-i
CO
•
rvi
CM
U
en
•
-------
blanks and one field blank also had lead detected at concentrations ranging from
4.6-5.3 ug/1. Lead concentrations in the samples ranged from non-detected to 11
ug/L; one effluent sample had a concentration of 580 ug/L (all elements detected
in this sample had much higher concentrations) . Based on blank results, the low
levels of lead detected in the samples may be biased due to the background lead
level observed.
Precision and accuracy for metals analysis was determined by the analysis
of matrix spike duplicates (MS/MSD) and these results are summarized on Table
A. 9. In addition, laboratory control samples were analyzed and generally met QC
criteria.
2.7 VOLATILE ORGANICS
Aqueous samples were analyzed for volatile organics by SW846 Method 8240.
All samples were analyzed within the method-specific holding time of 14 days for
a preserved aqueous sample (pH < 2- with hydrochloric acid).
Method blanks were analyzed along with the samples each day. There were
eleven method blanks associated with these samples; the only detected contaminant
.was methylene chloride in 6 blanks at concentrations ranging from 3J-29 ug/L.
(The "J" refers to an estimated concentration reported below the MDL of 5 ug/L.)
Most trip, field and equipment wash blanks (11 out of 15) were also contaminated
with methylene chloride at concentrations ranging from 3J-28 ug/L. Sample
concentrations ranged from non-detected to 48 ug/L. One sample which required
a 50-fold dilution due to foaming during purging had a concentration of 1200
ug/L, which is equivalent to 24 ug/L prior to accounting for the dilution factor.
Thus, all methylene chloride values should be considered biased high.
Surrogate recovery results, summarized in Table A. 10 were all within
control limits. Precision and accuracy were measured by the analysis of MS/MSD
samples; the results are presented in Table A.11.
2.8 OIL AND GREASE
Samples were analyzed for oil and grease by SW846 Method 9070. All samples
were analyzed within the method-specifled holding time of 28 days.
2-14
(82)
-------
*-w
SIS
»
e
.115
I»I5
i
'i -
i 2S
s'«
u 3
8
2 §
et
JAJ •- O M
oooinoorj--inino-j-fn
fMCXJCMCMOJCMCMOJfMCMfUCMCM ^
OOOOOOOOOOOOO O
in
inisiirti/itntni/iominmintn
CMIMC\t(MOJCMC\JC>JON(Moeago
m
ooooooooooooo
-
ooooooooooooo
CM(M(\inirg(\IIMCMIMf>JCMIM(M
OQOOOOOOOOOOO
in v- v o ^ *- o •^
rJo^o^o^ooxio:*-**
ooooooooooooo
(M(M(MCMr>J(MeMCMfM(MJ
tninintntninininmintninin
(Mrunj(U(MrucM(>jNruiM *— —•
t •*» ^ 3 •*• c.
x if
t. .^ M- u »~
3 ^ c o —•
nis;
2-15
(83)
-------
000
c£
55
.« 5-
'(A t* O
l
« 2
I
S § g
a
g g
s s
5 Q. ^r
s
<••
•)
as
S "3 «»
b *• *•
*»
if
O CO CO
in tn in tn in
o *- o o o I —
w— *— »o «o tn
^O N- h» N. F»
H 2 s 5 ~
3 • 8! iS S fc
o o o o o I o
€ c5 g
? 1 I
2-16
(84)
-------
Precision and accuracy were to be determined by the analysis of matrix
replicates and matrix spikes; due to limited sample volume this was not always
possible. In addition, QC check standards were analyzed with each sample batch
and all recoveries were acceptable, ranging from 80-101%.
2.9 TOTAL PHENOLICS
Aqueous samples and sludge samples were analyzed for total phenolics by
.SW846 Method 9066. Sludge samples were analyzed as received, basically as
aqueous samples and reported as mg/L.'" All samples were analyzed within the
method recommended hold time of 28' days. Method blanks were free from any
contamination. All precision and accuracy measurements were within control
limits..
2.10 RESIDUE
Aqueous samples were analyzed for total suspended solids (TSS) and total
volatile suspended solids (TVSS) . Sludge samples were analyzed for total solids.
All samples were analyzed within the 7 day method-specified hold time. All
associated method blanks had non-detected (<1 mg/L) results. Matrix replicate
analyses were used to assess method precision and results were within control
'limits.
2.11 NUTRIENTS
Samples were analyzed for ammonia, nitrate-nitrite, and total phosphate as
system condition measurements. Samples were analyzed within the 28 day holding
times for each of these analyses excluding the following samples which exceeded
the holding time for phosphate: days 1 and 3 of the 1 gpm flow test, influent/
effluent/equipment washes (ST1-A-1-2/5/13, ST1-A-3-2/5/13). All method blanks
for these analyses were free from any laboratory contamination. Method precision
was evaluated by the analysis of matrix duplicates and accuracy was determined
by the analysis of matrix.spike analyses; results are summarized in Table A.7.
2.1.2 FIELD DUPLICATES
Field duplicates were collected and analyzed at a frequency of 5%. As
discussed in the QAPjP, field duplicates inherently measure the precision of both
the collection procedure and the analytical methodology. When the results of the
2-17
(85).
-------
field duplicate analyses meet the analytical precision objectives, then both the
sampling and analysis can be considered to be within control. If, however, the
measurements exceed the precision objectives, then it is difficult to ascertain
whether the collection procedures resulted in non-representative samples or if
the analysis was in error. A total of eight field duplicates, analyzed for a
variety of parameters, were collected during the demonstration. Table A. 13
summarizes the results of the field duplicate analyses.
2.12.1 Pentachlorophenol
Since the measurement of PGP was the most critical of the measurements, the
individual sample results for the field duplicate pairs are given in Table A.13
for total and dissolved PGP.
Of the eleven duplicate pair analyses, seven RPD values were less than the
analytical control limit (RPD 0-50%). This indicates both sampling and
analytical precision were "in control. In the cases where the analysis for total
and/or dissolved PGP exceeded the control limit, other parameters analyzed for
the sample pairs generally met the control limits. For sample pair ST1-C-6-05/14
the RPD for total PGP was 85.7%, but dissolved PGP and other analyzed parameters
had RPD values within the applicable analytical control limits. Similarly,
sample ST1-C-09-05 and the duplicate had dissimilar PGP concentrations, but
results for TOG and chloride agreed. For sample pair ST1-B-01-02/14 both total
and dissolved PGP had RPD values outside of control limits, although the
duplicate sample dissolved PGP concentration is suspect due to analysis at an
inappropriate dilution. In addition, other parameters analyzed for this field
duplicate pair were * generally within the applicable control limits. This
suggests that sample collection techniques resulted in representative samples;
if differences in PGP concentrations were a result of non-representative sample
composition this would be reflected in other analyses. The PGP analyses which
exceeded the control limit are more likely a result of differences in extraction
efficiency or analytical measurements which were affected by large dilution
factors. PGP extraction is known to be highly variable by SW846 Method 3510 and
because of this variability several field duplicates showed precision to be above
specified objectives. However, a majority of the field duplicate pairs, as well
as MS/MSD analyses, indicate that overall variability was in control for the
measurement system.
2-18
(86)
-------
Table A.12
Field Duplicate Analysis Summary
Analysis Parameter
Pentachlorophenol
Dissolved Pentachlorophenol
Chloride
Total Organic Carbon (TOG)
TOG -dissolved
Total Phenolics
Oil and Grease
Volatile Organic Compounds :
Methylene Chloride
Metals: Arsenic
Copper
Lead
Nickel
Zinc
Dioxin/furans :
HpCDD
HpCDF
OCDD
OCDF
Alkalinity
Residue - TSS
TVSS
Nutrients - Ammonia
Nitrate-nitrite
Phosphate
Total #
of Analyses
7
4
7
7-
4
1 '
3
3
1
1
1
1
1
1
1
1
1
1
4
4
2
2
2
RPD Ranges
0-200
0-195
0-61 -
1.4-9.1
0-9.2
33
0-33
0-200
5.3
12
15
20
0
45
35
44
20
1.2
3.8-40'
0-40
0-6.1
4.1-12
0-23
Analytical RPD
Control T.fm-tt
0-50
0-50
0-25
0-25
0-25
0-25
0-30
NAd)
0-20
0-20
0-20
0-20
0-20
0-30
0-30
0-30
0-30
0-20
0-20
0-20
0-20
0-20
0-20
There are no established control limits for this compound.
2-19
(87)
-------
TABLE A.13
Pentachlorophenol Field Duplicate
Analytical Results
Sample/Duplicate ID(1)
STI-A-5-02/14
STI-A-9-05/14
STI-B-1-02/14
STI-B-9-02/14
STI-B-11-05/14
STI-C-6-05/14
STI-C-9-05/14
PGP Concentration (ug/L)
Sample Dupl. %
Cone. Cone. RPD
7200
<50
8800
24000
230
1100
300
11000
<50
17000
24000
160
440
<50
41.7
0
63.6
0
35.9
85.7
200
NA
<50
15000
22000
NA
810
NA
NA
<50
200*
22000
NA
810
NA
Dissolved PCP (ug/L)
Sample Dupl. %
Cone. Cone. RPD
0
195
0
(1) The last number in the sample ID relates to influent (-02), effluent (-05) and
the field duplicate (-14).
* i- Extract was not re-analyzed at an appropriate dilution
2-20
(88)
-------
2.12.2 Other Analyses
One chloride field duplicate RPD value (61%) exceeded the analytical RPD
control limit; all other chloride RPD values were well within limits, again
indicating that any problem was limited to individual sample analysis. The one
sample pair collected as the field duplicate for total phenolics had an RPD of
33%. One oil and grease duplicate analysis also had an RPD of 33%. These
estimates of precision are considered to be indicative of an overall system
(sampling and analysis) which is in control, even though the analytical RPD
values were slightly exceeded. Methylene chloride was the only detected compound
in the volatile organic analysis of the field duplicate collected (as 3 grabs
over 24 hours) for this parameter. Due to the background levels encountered in
the laboratory for this compound, the high RPD can be attributable to lab
contamination of the duplicate. One effluent sample collected in duplicate for
dioxins/furans had RPD values which exceeded the analytical control limit for
three out of four detected congeners. The duplicate pair analyses for TOG, STOC
and metals all had results Which met the precision objectives for the analysis.
2-21
(89)
-------
3.0 BIOMONITORING AND EMISSIONS MONITORING
3.1 BIOMONITORING
Acute toxicity was determined on influent, effluent and well water samples.
The fathead minnow tests were performed at five concentrations, each in
replicate, using a total of 20 organisms. Controls were performed, in replicate,
with each test. Survival for the controls was well within the 80% survival
requirements; in most cases no mortality was observed (except a few tests
exhibiting 90% survival). The water flea bioassay tests also were performed at
five concentrations, using a total of 20 organisms divided between two or four
vessels. All control samples exhibited 100% survival.
Reference toxicant tests were performed using cadmium chloride and sodium
pentachlorophenate. (This latter compound could only be obtained as a technical
grade reagent since it is no longer available from EPA.) The replicate 48 hour
LC50 values for cadmium chloride for the water flea were 0.08 mg/L and 0.09 mg/L.
The reported EPA ranges were 0.01-0.09 mg/L. The replicate 96 hour LC50 values
for cadmium chloride for the minnow was 0.48 mg/L and 0.60 mg/L; the reported EPA
ranges were 0.10-0.41 mg/L. The replicate 96 hour LC50 values for sodium
pentachlorophenate were determined for the minnow to be 0.09 mg/L and 0.12 mg/L
(it was not run with the water flea). The values referenced in the literature
indicate LC50 values ranging from 0.198-0.60 mg/L.
3.2 EMISSIONS MONITORING
3.2.1 Sample Collection
Gaseous samples were collected from the exhaust in the lid of the
bioreactor at locations before (upstream) and after (downstream) the carbon
adsorption system using EPA Modified Method 5 sampling procedures (SW846 Method
0010). Pre- and post-sampling leak tests were conducted per the Method to assure
that the sampling systems were within tolerance of the leak specifications in the
method.
EPA guidelines for acceptable results based on percent isokinetic sampling
are 100% ± 10%. Isokinetic sampling percentages for the upstream and downstream
samples collected during Run 1 were 80.6% and 99.1% respectively. The low
'3-1
(90)
-------
measured percent of isoklnetic for the upstream sample was probably a result of
turning the bioreactor fan on and off during the sampling run. During the first
sampling run, flow measurements were made with the fan on and off to determine
impacts the fan had on flowrates. It was decided that the fan was necessary for
isokinetic sampling; therefore, the booster fan was left on for the remaining
runs. The percent of isokinetic sampling maintained for all of the remaining
runs (Periods 2 and 3) was within the acceptance criteria of 90 - 110%.
3.2.2 Sample Analysis
Each emissions sample was comprised of the particulate filter, XAD sorbent
trap, impinger condensate and the probe/nozzle solvent washings. The filter and
trap were combined, spiked with phenol-d5, 2-fluorobiphenyl and anthracene-dlO
as surrogate standards, and extracted by SW846 Method 3540 (soxhlet extraction).
The condensate was spiked with nitrobenzene-d5 and .2,4,6-tribromophenol, the
washings fraction was spiked with 2-fluorophenol and terphenyl-d!4 (see section
6.3); both were extracted (individually) by SW846 Method 3510 (liquid-liquid
separatory funnel). The three extracts were combined prior to analysis for PGP,
PAHs and phenols by SW846 Method 8270. Reagent blanks were prepared using XAD
and an aqueous blank and were extracted with each batch of samples.
Four reagent blanks were analyzed; one contained 11 ug of naphthalene.
Four field blanks were collected and three had naphthalene detected at 15-100 ug.
Sample concentrations of naphthalene ranged from 92-780 ug for inlet samples and
11-40 ug at the outlet; therefore the outlet levels of naphthalene may be
suspect.
Table A. 14 summarizes surrogate recoveries achieved for the analysis of MM5
samples. All surrogate recoveries which were outside control limits were for the
samples collected during the 3 gpm flow rate period. All of these field samples
(inlet, outlet, and field blanks), along with the associated reagent blank, had
low surrogate recoveries for almost all compounds (average recoveries were 16%
for the trap/filter fraction, 14% for the condensate fraction and 8% for the
washings fraction) . All results for these samples should therefore be considered
significantly biased low.
3-2
(91)
-------
a
i
21
u £•
• u
SS
.is
—. 4)
(0 CO
t- w 01
.S S .5
O* O*
in tn in in CM in N.
^- *- ^— •-c>j •- ^~
KICM in CM ° CM °
c
SO
•••
5- *-
i
9 £
oc
fc
0000*0
o o ;g o eo o
»• «~ eo *™ h* »—
•rrot rM-4-
5 "5
** *^
- 8
4> a
„
5
il
1^2 ll II
?lf 5'C ?f
rvia.«zi-CMi-
= ig
S« M
— u
3-3
(92)
-------
Three samples were collected during the 1 gpm flow rate period as a means
of evaluating sampling precision. Triplicate analyses were not possible,
therefore these three consecutive samples were the best alternative.
Naphthalene, phenol and 2-methylphenol were measured in all three inlet samples;
the relative standard deviations (RSD) were 27.9%, 23.4%, and 21.7%,
respectively. Naphthalene was detected in all three outlet samples, with an
RSD-46%; the elevated RSD is probably a function of the concentration bias
discussed previously (based on contaminated reagent and field blank analyses).
Acenaphthene was detected in two of the three inlet samples, but at levels barely
above the MDL for the compound. Overall, the results for these analyses indicate
that the sampling and analyses protocols were in-control and generated
representative concentrations.
3-4
(93)
-------
4.0 ADDITIONAL STUDIES
4.1 PENTACHLOROPHENOL SAMPLING LOCATION BIAS
As initial pentachlorophenol results were reported for samples collected
during the first two weeks of the demonstration, it became apparent that there
was a significant difference between the PCP concentration in the raw well water,
and the concentration in the sample collected as the influent. The well sample
was collected as a grab sample from a tap in the line prior to the groundwater
entering the conditioning tank. The influent samples were collected as 22 hour
composites taken from the bioreactor pre-chamber above the underflow weir where
the water entered the first cell of the bioreactor. As additional data became
available, the difference in PCP concentration between the well water and the
influent decreased as the flow rate increased, as shown below:
Average Concentration (ug/L)
Flow Rate Well Water Influent
1 gpm 42000 6900
3 gpm 34500 18300
5 gpm 27500 24200
Data available from grab samples collected and analyzed by the developer (using
an HPLC method) , and data from samples split with Lockheed as part of the field
immunoassay project provided additional data. The following observations were
made.
o The developer collected samples from a tap in the line after the
conditioning tank, but prior to the water entering the bioreactor.
Results indicated a fairly constant sample concentration averaging
42000 ug/L.
o Samples analyzed by GC/MS (SW846 Method 8270) by EMSL-LV as part of the
field immunoassay project were split samples of the composites
collected from the bioreactor. Results were comparable to the influent
concentrations obtained by the SITE contract laboratory.
o Other analyses - chloride and residue - showed a reverse trend in
-concentration: the well water concentrations were lower than the
influent sample, suggesting that some change was occurring in- the
bioreactor prechamber.
4-1
(94)
-------
It.was postulated that backmixing within the bioreactor, from one chamber
to the preceding chamber, could be occurring. Thus, the PGP concentration of the
"influent" would decrease as wastewater from midchamber in the bioreactor, having
a lower PGP concentration after undergoing biodegradation, mixed with and diluted
the incoming groundwater. Likewise, chloride and residue concentrations would
be higher within the cells of the bioreactor due to degradation by-product and
sloughed biomass, respectively, and would therefore increase the "influent"
concentrations after mixing with the incoming groundwater.
To determine whether the variation between the influent sample
collected/analyzed by the developer and the composite influent sample collected
from the bioreactor pre-chamber could be due to differences in methodology versus
sample location and to document how the methods compared, as well as to determine
if the developer's data could be used in support of the other demonstration data,
the following steps were taken:
1. Compared the developer's sample results (HPLC) with the SITE contract
lab sample results (SW8270) for high flow rate influent and effluent
samples where possible. Also determine what, if any, QC/validation
data the.developer had available for the HPLC method.
2. Have the SITE contract lab re-analyze several of their composite
influent samples (low and high flow rate) by SW8270 to evaluate hold
time effect.
3. Have the SITE contract lab re-analyze several developer influent
samples (low and high flow rate) by SW8270 to compare methodology.
4. Have the developer re-analyze several of their influent grab samples
by HPLC to evaluate hold time effect.
5. Have the developer re-analyze several of the composite influent
samples by HPLC to compare methodology.
6. Have each lab analyze the standard used for calibration by the other.
Eight samples (four each from the low and high flow tests) were chosen for re-
analysis by both labs; two were to be performed in duplicate. By re-analyzing
the samples which were also being sent to the other lab, data would be comparable
from within the same analysis time frame, and in addition, information on the
effects of holding time would be generated. Results of this study are discussed
below.
4-2
(95)
-------
Holding Time: The re-analysis of samples by both laboratories
provided inconclusive results. Twenty reanalyses, total, were
performed with a majority of the re-analyzed values lower than the
original sample concentrations; however, the percent decrease ranged
from 4.7-100%. (A few re-analysis sample concentrations actually
increased.) The results indicated that the effect of compound
degradation over time was unpredictable.
Method Comparison: Regardless of the effect of hold time on the
samples, since both laboratories were re-analyzing samples at the
same time, the results could be used to compare the GC/MS and HPLC
methods used. Table A. 15 presents the re-analysis values of
composite samples (collected from the bioreactor pre-chamber) re-
analyzed by both laboratories. Six of the eight analysis pairs had
RPD values less than 50 percent (the SW8270 precision criteria for.
MS/MSD pairs for pentachlorophenol).
Each laboratory also analyzed the other's analytical standard. The
SITE contract laboratory analyzed the 38.4 ppm standard used by the
developer at 25.0 and 26.0 ppm (average 66.4% recovery). The
developer analyzed the 2.0 ppm standard used by the SITE contract
lab at 1.8 and 2.1 ppm (average 98% recovery).
The average of the 8270 values compared to the HPLC data were
approximately 35% lower. The GC/MS method used the 3510 extraction
procedure, which is known to exhibit fair to moderate extraction
efficiency for PGP. The SITE contract lab MS/MSD recoveries, while
highly variable (65-204%), also suggest PCP recovery efficiency was
biased low. Therefore it appears the two different method are
comparable, that extraction efficiency causes the GC/MS analysis to
be biased low, and that HPLC methodology may be more accurate for
PCP concentrations above 1 PPM.
Precision: Several samples were analyzed in duplicate by both
laboratories. The precision estimates are presented in Table A.16.
In general, the HPLC duplicate analyses performed by the developer's
lab showed excellent agreement (maximum RPD - 12). The GC/MS
analyses performed by the SITE contract lab were also within QC
limits, meeting the method criteria for precision for pentachlo-
rophenol, RPD s50%, in 7 out of 8 anlayses. The larger RPD for the
SITE contract lab result was primarily a function of variable
extraction efficiency.
The results of these analyses indicated that the lower-than-expected influent
concentrations for the composite samples collected from the bioreactor prechamber
during the demonstration, specifically during the 1 gpm and 3 gpm flow rates,
were probably due to the suspected backmixing as opposed to method bias.' (See
reported concentrations in the Technical Evaluation Report) In general, the
extraction/analysis method used for the analysis of the well water demonstration
4-3
(96)
-------
TABLE A.15
FCF Method Comparison
Concentration. ug/L
SW846-3570/8270
Sample Date Re-Analysis HPLC Re-Analysis RPD
7/26/89 ND(4J) ND (<500)
7/27/89, dupl. 3400, 5600 5700, 5700 24*
8/01/89, dupl. 2900, -- 3700, 3300 19*
8/03/89, dupl. 6-100, 5500 8400, -- 37*
8/24/89 18000 37900 71
8/26/89, dupl. 33000, 35000 37300, 37500 9.5*
.8/28/89, dupl. 28000, 15000 33400, 33600 44*
8/29/89 9700 35800 115
* - RPD value calculated using the average concentration^ ) of the duplicate pair
4-4
(97)
-------
o o <
ut m 2
co r-
W W
O •< O
r* a p-
s
•g
JS
S I
I I
i s
t-t r* oo
g §
< o
Z 0
s >
M V
ig
X u
•^ •»
i;
f sr
M O
3 a a
?u
I r*
« M iH
Q H <0
--* W . •*
no. rt
o C
* -* •<
:»
!!
:a
CO CO CO CO CO
4-5
(98).
-------
samples, SW846 Method 3510/8270, did appear to achieve somewhat lower absolxite
recovery (60-70%) when compared to the HPLC method.
Well water PCP concentrations analyzed by GC/MS were similar to BioTrols
influent PCP concentrations analyzed by HPLC when a correction for a 65%
extraction efficiency was performed.
The QA data from the study indicated that the developer's results could
be used to support the demonstration data, and that the well water PCP
concentrations were more representative of the actual "influent" concentration.
4.2 EXTENDED PENTACHLOROPHENOL CALIBRATION STUDY
Many sample extracts for PCP, particularly the well water and the influent,
required analyses at multiple dilutions to bring the PCP concentration within the
calibration range of the analysis. Time constraints resulted in a number of
samples with reported PCP concentrations which exceeded the calibration range;
the highest concentration standard analyzed was 160 ng/uL which, for a one liter
sample extracted and concentrated to one milliliter, is equivalent to a sample
concentration of 160 ug/L. After reviewing the sample and calibration data, it
.was felt that the actual linear range of PCP analysis extended beyond 160 ng/uL.
Therefore, three studies were performed to determine the actual linear range of
PCP. Standards were analyzed at eight concentrations for the tests: 20, 50, 80,
120, 160, 200, 250 and 300 ng/uL. The first two tests used all eight points, the
third excluded the 20 ng/uL standard. (The routine laboratory calibration range
for PCP is 50-160 ng/uL.) The average response factors and relative standard
deviations were calculated for all eight standards (20-300 ng/uL) and a second
time excluding the lowest concentration standard (since PCP has a method
detection limit of 50 ug/L). The results are summarized below:
20-300 ng/uL 50-300 ng/uL
Test RF %RSD RF %RSD
1 • 0.1183 41.8 0.1316 26.4
2 0.1168 29.9 0.1257 20.8
3 NA NA 0.1310 13.4
where the response factor (RF) - area PGP/area internal standard x concentration
internal standard/concentration PCP. The method requirement for linearity for
4-6
(99)
-------
a compound designated as a calibration check compound (CCC), which PGP is not,
is an RSD <;30.0%. If this criteria is applied to the PCP results for the seven-
point calculations, then the % RSD all meet the linearity requirement.
Therefore, it was determined that pentachlorophenol was linear within 50-300
ng/uL, and almost all sample data did indeed fall within this range, and
therefore was determined to be useable for calculating process efficiency.
4-7
(100)
-------
5.0 NON-CONFORHANCES
Most analytical QC. results, including blank analyses, precision and
accuracy data, surrogate recovery results and QC check sample results, were
discussed in detail in previous sections (2.2 through 2.11) of this review.
Therefore, nonconformances regarding QC analysis .results which did not meet QC
criteria have already been discussed. However, a few additional non-conformances
are presented below.
o Dioxin/furan extraction was to be performed on a phase-separated solid
sample of the sludge. Percent solids were to have been determined in
this phase-separated sample to allow the calculation of dry-weight
concentrations. Percent solids were not performed and dipxin/furan
results for the sludge were reported as wet-weight concentrations.
Corrective Action: None possible.
o Trip blanks were often received (at the field and/or at the lab) with
air bubbles.
Corrective Action: Document on Chain-of-Custody. (This problem was
not encountered with the samples.)
5-1
(101)
-------
6.0 DEVIATIONS FROM THE QAPjP
In some instances the procedures and methods referenced in the QAPjP for
use during the BioTrol demonstration project were modified or changed in response
to physical constraints of the system, field requirements/limitations, laboratory
needs and/or EPA concerns. These deviations are listed below; each includes in
parentheses the QAPjP section reference and page number which applies.'
6.1 FIELD PROCEDURES
Samples for biotoxicity testing were to be collected as composites. Due
to the limited hold time, and large volume requirements, these samples were
actually collected as grab samples (5.4, pg. 32).
Total organic carbon (TOG) samples were to be preserved with hydrochloric
acid to a pH of less than 2. At the laboratory's request, the preservative used
was changed to sulfuric acid (5.4, pg. 16) in order to accommodate their TOG
analyzer.
Samples to be analyzed for dissolved parameters (TOG and PCP/PAHs) were to
be filtered in the field prior to preservation; time constraints in the field
required that samples for dissolved PCP/PAHs be filtered in the laboratory.
(Dissolved TOG samples were filtered in the field, preserved and then shipped to
the lab.) (5.4, pg. 28).
Oil and grease samples of the bioreactor midpoints could not be obtained
by direct immersion of the sample bottle; the Teflon suction line of the auto-
sampler was used to collect a sample (no equipment wash blank was collected)
(5.4, pg. 27).
Samples for total phenolics and volatile organic compounds analysis were
to be taken as three grabs over 24 hours and composited in the laboratory prior
to analysis. Instead, each individual grab sample was analyzed separately for
volatile organics, and samples for total phenolics were collected as a field
composite. (A measured volume was collected, preserved, poured into a sample
bottle and refrigerated. During the day, two more volumes were collected,
6-1
(102)
-------
preserved and added to the bottle which was then shipped as a composite of the
three grab aliquots.) (5.4, pg. 27).
The sample numbering system was modified to include two additional sample
identifiers. The last digit(s) of the sample number specifies the • discreet
ssimple location (e.g., influent, sludge). Field replicates were assigned a
sample point number "-14" (e.g., STI-A-9-1-14). An additional sludge sample was
collected from the walls of the bioreactor (not proposed in the QAPjP) and was
assigned a sample number "-15". [5.4, page 35]
Sludge samples were to be collected at the bioreactor filter location on
a weekly basis. The type of filter utilized (described as a felt sock bag
fitter) hindered the collection of a true sludge sample. Volume was limited, and
the sludge was collected as a composite by combining samples obtained each time
the filter was changed. [5.4, pg. 29]
Emissions monitoring was originally planned to take place during system
start-up, and once during each of the three test periods. At the time the QAPjP
was approved, it was no longer feasible to sample during the start-up period and
most references in the QAPjP to this proposed sampling were omitted. However,
Table 5-5, an overall sampling summary for the project, [5.4, pg. 8] still
(incorrectly) specified a total of four samples .(plus samples for precision) at
each sample point. Only three were collected (plus samples for precision), one
for each' flow rate test.
6.2 SAMPLING SCHEDULE
Seven more trip blanks (TB) than originally proposed were collected. The
raw well water was grab sampled once/week for all parameters including volatile
organics on a different day than planned, necessitating one additional TB per
week. An additional TB was also required when the field blank was collected and
sent for analysis.
As a result of the field audit, an equipment wash blank for dissolved TOG
was added to the sampling plan on a weekly basis, beginning with week 4.
6-2
(103)
-------
As a result of concerns expressed during the EPA audit of the procedure
used by laboratory for biotoxicity testing, the raw well water was added for
biotoxicity testing on a weekly basis (in addition to the influent/effluent
samples proposed).
An additional sludge sample was obtained the last week of sampling by
collecting scrapings from the side walls of the bioreactor. This sample, STI-C-
10-15, was analyzed for PCP/PAHs.
6.3 ANALYTICAL METHODS
Total phenolics were cited in the QAPjP as analysis method SW846 9065; the
laboratory actually used SW9066 which is an automated application of the same
procedure. (5.5, pg. 11).
During the EPA audit of the laboratory performing chloride it was agreed
to change the procedure referenced for chloride analysis, SW9252, to the Standard
Methods procedure 407C. (5.5, pg. 10).
Metals determinations require the analysis of a laboratory control sample
(LCS), often a known solution provided by EPA with statistically determined 95%
confidence limits. The lab performing metals used an NBS standard for the LCS,
and has established an acceptance criteria for the analysis of ±15% of the true
value. (5.7, pg. 4).
Sludge samples were expected to be a very wet solid sample. A wastewater
type sample was actually obtained. The sample was treated "as-is" for total
phenolics and metals as proposed in the QAPjP, and results were calculated on a.
wet-weight basis. Samples for dioxin and PCP/PAH analysis were separated into
phases and the solid phase was extracted. Total solids were used to calculate
dry-weight results for PCP/PAH; dioxins were reported as wet weight. (5.4, pg.
29).
Samples for semivolatile analysis (full scan or PCP/PAHs only) which
exhibited low surrogate recoveries of d5-phenol and 2-fluorophenol were re-
analyzed to confirm results. However, due to time constraints, this re-analysis
6-3
(104)
-------
was discontinued after it was determined that the re-analysis was confirming the
original results. (5.7, pg.2).
6-4
(105)
-------
7.0 AUDIT FINDINGS/RESULTS
A series of audits were conducted in the field and of the laboratories
involved with this BioTrol SITE demonstration project. The findings and results
of the audits and corrective actions taken in response to any concerns noted
during the audits are summarized below.
7.1 FIELD AUDIT, WATER SAMPLING
An audit of the water sampling operations was conducted by an EPA
contractor on July 26-27, 1989. The audit rating was satisfactory. A few issues
were identified and prompt corrective action was taken to resolve these issues.
Prior to the audit, it had been noted during the first few days of sampling
that the dissolved oxygen meter was not functioning properlyl'and steps were taken
to obtain an additional D.O. meter. Dissolved oxygen was a system condition
measurement, not a critical measurement. One day's worth of measurements were
lost (and the two prior day's measurements were suspect). While waiting, for the
new meter to arrive on site, dissolved oxygen measurements were taken using the
Winkler titration method. After receiving a new D.O. meter, problems were still
encountered-with calibration, and the decision was made .to .continue using the
Winkler method for dissolved oxygen determinations.
A second item of concern noted by the auditor was the manner in which
samples were labelled for shipment. A question arose as to whether or not
samples which were preserved with acid to pH <2 required shipment as corrosive
material. After shipments made according to DOT regulations were unsuccessful
and, after evaluating the regulations, the following procedure was implemented.
The samples were not considered corrosive and were shipped via regular overnight
carrier. All samples were preserved to a pH <2 without over-acidifying
(generally only a few drops of acid per several hundred milliliters of sample).
Bottles were sorbent wrapped, bubble wrapped, packed with ice, placed in a
plastic bag and shipped in a cooler.
A third issue noted on the audit corrective action form involved field
filtration of samples for dissolved TOG. The field filtration kit erroneously
7-1
(106)
-------
included an organic-based filter paper and consequently inorganic-based filter
paper was promptly obtained and substituted.
7.2 FIELD AUDIT, EMISSIONS SAMPLING
An audit of the emissions monitoring MM5 sampling protocols was conducted
by an EPA contractor on August 2-3, 1989. The audit rating was satisfactory.
One issue was raised concerning the representativeness of the MM5 samples.
Previously, a memorandum from Guy Simes dated July 20, 1989 raised the issue of
isokinetic sampling. As a result, the sampling team obtained a 4" PVC pipe to
attach to the BioTrol reactor exhaust for MM5 sampling. This was done to
increase duct size as required by the method. In order to measure the velocity
of both the reactor exhaust and carbon filter exhaust, an exhaust fan attached
to the reactor lid needed to be turned on. Once this decision was made, an issue
of representativeness was raised. Without the fan running it was evident that
no vacuum was pulled on the reactor bed, nor were there leaks in the system. It
was also obvious that since the fan increased the flow in the reactor outlet, the
extra air flowing through the system resulted from system leaks. Leaks into the
exhaust flow would produce a non-representative sample. Without the exhaust fan
operating, sample flow in the exhaust outlet would be almost impossible to
measure and the exhaust for the carbon bed filter had no measurable flow. This
presented a problem for obtaining isokinetic samples in the reactor exhaust and
for obtaining a sample at all in the carbon filter exhaust. All parties involved
were in agreement that isokineticity was non-critical because there were no
particles in the reactor exhaust. Not being able to obtain a measurement of the
carbon bed filter emission, however, would have presented a problem if it was
decided to turn off the exhaust fan. It was determined that representativeness
of the sample was not a problem providing no significant PGP or PAH concentration
existed in the dilution air. In order to do a mass balance of the system all
that was needed was a total PGP/PAH concentration regardless of whether or not
it was diluted. After close examination of the calculations used to determine
PGP/PAH concentration in air emissions, all involved parties were satisfied with
the decision to keep the exhaust fan operating. It was therefore decided that
sampling should take place only while the exhaust fan was operating. Additional
concerns about pulling a vacuum on the reactor or if operating the fan upset
reactor operations proved to be negligible.
7-2
(107)
-------
7.3 LABORATORY AUDITS
Audits of the three laboratories involved with the analysis of the BioTrol
demonstration samples were conducted by EPA contractors. Two audit ratings were
satisfactory and one was rated conditional.
On August 4, 1989 an audit was conducted of the laboratory performing
PGP/PAH analysis, full scan semivolatile organics, volatile organics arid metals
analysis. The audit rating was satisfactory. During the audit, the volatile
compositing procedure described in the QAPjP was changed. Rather than
compositing the three grabs collected over 24 hours into a single sample prior
to analyses, each of the grabs was to be analyzed separately as a discrete
sample.
The August 7, 1989 audit covered the analyses of total organic carbon,
dissolved organic carbon, oil and grease,, residue and alkalinity. A conditional
rating was assigned. The major concern noted involved the procedure being used
for distillation in the oil and grease method, a modification of the reference
method SW9070. The laboratory responded to the concern by providing
documentation of the distillation temperature and by performing duplicate
.analyses comparing the modified procedure and the referenced method. Based on
these determinations, which indicated that data quality was not compromised by
the variation, a change in audit rating was requested. To date, a final audit
report and rating have not been received.
A final audit was conducted August 14-15, 1989 covering the analysis of
dioxins/furans, chloride, phenolics, ammonia, phosphate, nitrate-nitrite,
biotoxicity, and the MM5 sampling trains for PCP/PAHs. The audit rating was
satisfactory. Minor concerns were noted regarding the biotoxicity method, and
the shipment of MM5 samples.
•7-3
(108)
-------
8.0 CONCLUSIONS AND LIMITATIONS:OF DATA
As noted throughout this Appendix, most of the data generated during the
BioTrol SITE demonstration can be used in its current form with the exceptions
noted below. Process efficiency was most representative when determined using
well water contaminant concentrations, as opposed to bipreactor influent sample
concentrations. The additional study performed on PGP influent concentrations
supports the use of these well water sample concentrations which were similar to
data obtained by the developer in an independent collection/analysis program.
Data which have limitations include the following; however, these
qualifications to the data did not preclude the ability to adequately evaluate
the treatment technology.
o Emissions samples collected during the 3 gpm flow rate period were
significantly biased low. However, sampling during the 1 gpm and 5
gpm flow rates indicated that losses to the exhaust air were minimal
and that PGP was not present in these emissions, therefore mass
balance calculations were not affected.
o Sludge samples were not the wet solids expected, but rather an almost
aqueous wastewater type sample. Analyses performed on the samples
were either reported on an as-is, wet weight basis (which could easily
be converted to dry weight) or on a phase-separated sample. Samples
were phase-separated for PGP/PAH by centrifugation and reported on a
dry-weight basis, and by decanting for dioxins/furans. Without
percent moisture on the solids obtained after decanting, overall mass
balances can only be estimated for dioxins.
8-1
(109)
------- |