.MAL
REPORT
600285109
Point-of-Use Reduction of Volatile
Halogenated Organics
   NATIONAL SANITATION FOUNDATION
   ASSESSMENT SERVICES

-------
                                                 PROPERTY OF THE
                                                OFFICE OF SUPERFUND
POINT-OF-DSE REDUCTION OF VOLATILE HALOGENATED ORGANICS
                   IN DRINKING WATER
                           by
      Gordon Bellen, Marc Anderson, Randy Cottier
            National  Sanitation Foundation
                      PO Box 1468
                  Ann Arbor, MI  48106
                Contract No. R809248010
                    Project Officer

                    Steven Hathaway
            Drinking Water Research Division
         Water Engineering Research Laboratory
                 Cincinnati, OH  45268
         WATER ENGINEERING RESEARCH LABORATORY
           OFFICE OF RESEARCH AND DEVELOPMENT
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                CINCINNATI,  OHIO  45268
           U.S. Environmental Protection Agency
           Region 5, Library  (PL-12J)
           77 West Jackson Boulevard, 12th Floor
           Chicago,  !L  60604-3590

-------
                                  DISCLAIMER
The information in this document has been funded wholly or in part by the
United States Environmental Protection Agency under Contract No. R809248010 to
the National Sanitation Foundation.  It has been subject to the Agency's peer
and administrative review, and it has been approved for publication as an EPA
document.  Mention of trade names or commerical products does not constitute
endorsement or recommendation for use.

-------
                                   FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water systems.  Under a mandate of
national environmental laws, the agency strives to formulate and implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life.  The Clean Water Act,
the Safe Drinking Water Act, and the Toxics Substances Control Act are three
of the major congressional laws that provide the framework for restoring and
maintaining the integrity of our Nation's water, for preserving and enhancing
the water we drink, and for protecting the environment from toxic substances.
These laws direct the EPA to perform research to define our environmental
problems, measure the impacts, and search for solutions.

The Water Engineering Research Laboratory is that component of EPA's Research
and Development program concerned with preventing, treating, and managing
municipal wastewater discharges; establishing practices to control and remove
contaminants from drinking water and to prevent its deterioration during
storage and distribution; and assessing the nature and controllability of
releases of toxic substances to the air, water, and land from manufacturing
processes and subsequent product uses.  This publication is one of the
products of that research and provides a vital communication link between the
researcher and the user community.

Contamination of drinking water by organic chemicals has become a national
problem of increasing concern.  Small communities faced with the problem of
removing these contaminants have limited resources and often cannot afford
traditional treatment approaches.  One approach which has the potential to be
cost effective is treatment at the home, or point-of-use treatment.  This
report discusses the results of a study of point-of-use treatment of
contaminated groundwater in two small communities.

                                       Francis T. Mayo, Director
                                       Water Engineering Research Laboratory
                                       ii

-------
                                   ABSTRACT
A study was conducted to determine whether point-of-use carbon treatment is
cost effective for the control of volatile halogenated organic chemicals in
small water systems and to study water quality district management techniques
for point-of-use treatment.  Point-of-use treatment efficacy and costs were
monitored in two small communities for the control of eight volatile organic
chemicals (VOCs) and four trihalomethanes.  The groundwaters in this study
were contaminated with one or more of the following VOCs: chloroform,
1,2-dichloroethane, 1,1,1-trichloroethane, 1,1-dichloroethylene,
1,2-dichloropropane, carbon tetrachloride, trichloroethylene, and
tetrachloroethylene.  The point-of-use treatment method used was granular
activated carbon in a line bypass configuration.

Two communities were studied: Silverdale, Pennsylvania, and Rockaway Township,
New Jersey.  Silverdale has four groundwater wells connected to a community
distribution system.  Rockaway Township is a collection of homes, served by
private wells.  Five models of point-of-use devices were selected and
installed in approximately 50 homes in Silverdale.  Rockaway Township had
already selected and installed devices, and was used as a case study.

The devices in Silverdale removed the VOCs from the groundwater for the entire
testing period (14 months).  Devices in use in Rockaway Township were still
removing all VOCs after 24 months.

Breakthrough (detection, in the effluent, of the VOC present in the influent
in consecutive samples from the same location) did not occur for any of the
devices tested during the 14 months of sampling in Silverdale,  However, trace
concentrations of VOCs were detected intermittently in postdevice samples for
all but one of the models.

Water samples collected in Silverdale had higher microbial densities in post
device samples than predevice samples, as measured by standard plate count
(SPC).  Postdevice microbial densities were affected by the sampling technique
used.  Samples of water collected after two minutes of flushing had SPC densi-
ties comparable to samples of water from  the distribution system.  If one
liter of water was flushed from the line  before sampling, postdevice samples
had mean densities one order of magnitude higher than predevice samples.  Un-
flushed postdevice samples had mean densities two orders of magnitude higher.

Mean densities of predevice samples collected in Rockaway  Township were more
than 10 times higher than predevice samples collected in Silverdale.  However,
Rockaway postdevice microbial densities were not significantly higher than
predevice densities.  Mean postdevice densities in Silverdale were more than
10 times higher than those in Rockaway.


                                       iii

-------
No evidence existed of coliform colinization on any of the devices in either
community.  Positive coliform results were intermittently found for postdevice
water samples in Silverdale, but follow-up samples were negative for coliform.
Fecal coliforms were never found in postdevice samples.  A one-time sampling
for total Pseudomonas and Pseudomonas aeruginosa in Silverdale was negative.

Postdevice samples from silver impregnated carbon did not have lower SPC
densities for devices tested in Silverdale.  In fact, mean densities were
slightly higher from silver impregnated devices.  No comparisons could be made
in Rockaway.

The average monthly cost per customer for point-of-use treatment in Silverdale
was $5.98, assuming a 20 year capital amortization at 10 percent.  The
estimated cost in Rockaway Township was $5.06 per customer per month.  In both
communities, the homeowner is responsible for maintenance.  If a water quality
district were established to monitor and maintain the devices, the estimated
administrative costs (including labor and supplies) would be $1.23 per
customer per month for both communities.

This report was submitted in fulfillment of Contract No. R8092A8010 by the
National Sanitation Foundation under the sponsorship of the U. S.
Environmental Protection Agency.  This report covers the period September 1982
to March 1985.
                                       iv

-------
                               TABLE OF CONTENTS

                                                                  PAGE

FOREWORD	   ii
ABSTRACT	ill
CONTENTS  	    v
FIGURES	vii
TABLES	viii
ACKNOWLEDGMENTS 	    x
ENGLISH TO METRIC CONVERSIONS 	   xi
    1.  INTRODUCTION  	    1
        BACKGROUND  	    1
        OBJECTIVES  	    2
    2.  SUMMARY AND CONCLUSIONS 	    3
    3.  PROCEDURES	    5
        STUDY SITE SELECTIONS	    5
        SITE DESCRIPTIONS 	    5
           Silverdale, PA   	    5
              Equipment Selection and Description 	    7
           Rockaway Township, NJ	 .  .    8
        SAMPLING AND ANALYTICAL PROCEDURES  	   12
           Volatile Organic Chemicals 	   12
           Microbiological  	   14
           General Water Quality Analysis 	   14
           Cost Data Collection	   14
    4.  RESULTS   	   15
        VOLATILE ORGANIC CHEMICAL REDUCTION 	   15
           Predicting Breakthrough  	   15
           Silverdale, PA	   17
           Rockaway Township, NJ  	   23
        MICROBIOLOGICAL GROWTH IN POINT-OF-USE DEVICES  	   25
           Silverdale, PA	   25
           Rockaway Township, NJ  	   31
        COSTS	   33
           Silverdale, PA	   33
              Equipment Purchase and Installation 	   33
              Operation and Maintenance 	   33
           Rockaway Township, NJ	   36
              Equipment Purchase and Installation 	   36
              Operation and Maintenance 	   36
           Monitoring Costs 	   37
           Water Quality District Administration Costs  	   37
    5.  CURRENT MANAGEMENT STATUS 	   39
        SILVERDALE,  PA	   39
        ROCKAWAY TOWNSHIP, NJ 	   39
                                       v

-------
REFERENCES	   40

APPENDIX A.  EFFECTS OF HEADSPACE AND HOLDING TIME
             ON VOC RECOVERY	   42
APPENDIX B.  VOC AND MICROBIOLOGICAL RESULTS FOR
             POINT-OF-USE EFFLUENT SAMPLES -
             SILVERDALE, PA	   52
APPENDIX C.  VOC AND MICROBIOLOGICAL RESULTS FOR
             POINT-OF-USE EFFLUENT SAMPLES -
             ROCKAWAY TOWNSHIP, NJ 	   71
APPENDIX D.  MICROBIOLOGICAL RESULTS USING R2A
             AND SMA MEDIA	   73
                                       VI

-------
                                    FIGURES
                                                                PAGE

1.   Mean VOC concentrations in influent samples
    presented by month	18

2.   Geometric means of Silverdale pre and
    postdevice standard plate counts by month   	 28

3.   Median postdevice SPC densities versus grams of carbon  ... 30

4.   Comparison of standard plate count agar, unflushed
    postdevice samples, March-April, 1984 	 32
                                       vii

-------
                                     TABLES
                                                                PAGE

 1.  OPTIONS AVAILABLE  FOR SOLVING GROUNDWATER
    CONTAMINATION  PROBLEMS  	    1

 2.  WATER QUALITY  MEASURED IN SAMPLES  FROM FOUR
    WELLS AND  THE  DISTRIBUTION SYSTEM  IN
    SILVERDALE,  PA	    6

 3.  DESCRIPTIONS OF  POINT-OF-USE  DEVICES USED
    IN SILVERDALE, PA	    9

 4.  GROUNDWATER  QUALITY MEASURED  IN SAMPLES FROM
    FOUR PRIVATE WELLS IN THE LAKE TELEMARK SUBDIVISION 	   11

 5.  PRECISION  AND  ACCURACY OF VOC ANALYSES USING
    LIQUID/LIQUID  EXTRACTION  	   13

 6.  RELATIVE CARBON  AFFINITY FOR  VOCs  AND VOC
    SOLUBILITY IN  WATER	16

 7.  SUMMARY OF INFLUENT VOC RESULTS,  SILVERDALE,
    PA (MARCH  1983 - APRIL 1984)   .	17

 8.  POSTDEVICE VOC CONCENTRATIONS BY MODEL AND
    VOLUME TREATED,  SILVERDALE, PA
    (MARCH 1983  -  NOVEMBER 1983)	21

 9.  POSTDEVICE VOC CONCENTRATIONS BY MODEL AND VOLUME
    TREATED, SILVERDALE, PA (MARCH 1983 - APRIL 1984) 	   22

10.  VOCs IDENTIFIED  IN TWELVE PRIVATE WELLS IN
    ROCKAWAY TOWNSHIP, NJ 	   23

11.  RESULTS OF ROUND ROBIN ANALYSIS OF PAIRED VOC
    SAMPLES FROM ROCKAWAY TOWNSHIP, NJ  	   24

12.  VOC CONCENTRATIONS AT FOUR POINT-OF-USE TREATMENT
    SITES IN ROCKAWAY  TOWNSHIP, NJ	24

13.  BACTERIOLOGICAL  SAMPLING METHOD COMPARISON,
    RESULTS OF PAIRED  SAMPLES, SILVERDALE, PA
    POSTDEVICE STANDARD PLATE COUNTS  	   26


                                       viil

-------
14.  STANDARD PLATE COUNT SUMMARY FOR UNFLUSHED
    SAMPLES, SILVERDALE, PA (MARCH 1983 - APRIL 1984) 	  26

15.  POSTDEVICE STANDARD PLATE COUNT RESULTS FOR
    FLUSHED AND UNFLUSHED SAMPLES BY MODEL TYPE 	  27

16.  MEDIAN POSTDEVICE STANDARD PLATE COUNTS BY
    MODEL AND CARBON WEIGHT (Unflushed Samples Only)  	  29

17.  POSTDEVICE STANDARD PLATE COUNTS FOR MODELS WITH
    AND WITHOUT SILVER IMPREGNATED CARBON
    (Unflushed Samples Only)  	  29

18.  ROCKAWAY TOWNSHIP, NJ SPG DATA COMPARING SAMPLING
    TECHNIQUES (Paired Samples Only)  	  33

19.  ROCKAWAY TOWNSHIP, NJ SPC DATA COMPARING
    SAMPLING TECHNIQUES 	  33

20.  CAPITAL COSTS FOR POINT-OF-USE DEVICES USED IN
    SILVERDALE, PA	34

21.  ESTIMATED AVERAGE MONTHLY CUSTOMER COSTS,
    SILVERDALE, PA	36

22.  ESTIMATED AVERAGE MONTHLY CUSTOMER COSTS,
    ROCKAWAY TOWNSHIP, NJ	  37
                                      ix

-------
                                ACKNOWLEDGMENTS

An advisory committee of water treatment professionals and USEPA personnel
participated in project planning and review of this report.  The authors wish
to recognize the helpful guidance and comments of the following committee
members:
    Raymond Barg, New Jersey State Department of Environmental Protection
    Frank Bell, USEPA Office of Water Supply
    Richard Christie, Rockaway Township Department of Health and Welfare
    Ira Markwood, Illinois EPA
    Glenn Maurer, Pennsylvania Bureau of Community and Environmental Control
    Juan Menendez, Dade County Department of Public Health
    Robert McCall, American Water Works Association
    Douglas Oberhamer, Water Quality Association (WQA)
    P. Regunathan, WQA Point-of-Use Committee
    Bernie Sarnowski, USEPA Region III
    William Seidel, Seidel Water Company
    Thomas Sorg, USEPA Office of Research and Development
    Lee Thomas, Bucks County Health Department
    Lee Wikstrom, WQA Point-of-Use Committee
Additional assistance and information were provided by other individuals
interested in this project.  Particularly valuable assistance was provided  by:
    Steven Levinson, Rockaway Township Department of Health and Welfare
    Esther Seidel, Seidel Water Company
    Gwendolyn Ball, National Sanitation Foundation
    Lorri White, National Sanitation Foundation
    The townspeople of Silverdale, Pennsylvania
                                       x

-------
                      ENGLISH TO METRIC CONVERSION TABLE
     TO CONVERT
        FROM

Inches
Feet
Miles
Square Miles
Gallons (U.S. liquid)
Gallons (U.S. liquid)
Pounds
Pounds/Square Inch
Dollars/Gallon
Acres
                              MULTIPLY
        TO                       BY

Centimeters                      2.54
Meters                           0.3048
Kilometers                       1.609
Square Kilometers                2.590
Liters                           3.785
Cubic Meters                     0.0038
Grams                          453.6
Kilograms/Square Centimeter      0.070
Dollars/Liter                    0.264
Square Kilometers                0.004
                                       xi

-------
                                  SECTION 1.

                                 INTRODUCTION
BACKGROUND

The presence of chemical contaminants in groundwater has become a potential
health threat of national significance.  Sources of synthetic organic chemi-
cals are diverse and can be found in rural communities as well as large indus-
trial communities.  A private well may even be inadvertently contaminated by
its owner through the use of cleaning solvents and other man-made chemicals.

In the Advance Notice of Proposed Rulemaking (ANPRM) for volatile organic
chemicals (VOCs), published in the Federal Register on March 4, 1982, 14
compounds were identified as those most commonly detected in drinking water
(1).  The 14 VOCs are: trichloroethylene, benzene, tetrachloroethylene chloro-
benzene, carbon tetrachloride, p-dichlorobenzene, 1,1,1-trichloroethane, tri-
chlorobenzene, I,2-dichloroethane, 1,1-dichloroethylene, vinyl chloride, cis-
1,2-dichloroethylene, dichloromethane, and trans-1,2-dichloroethylene.

In a random sampling of 500 groundwater systems in the United States,
approximately 21 percent of the systems contained at least one of the 14 VOCs
listed (1).  Once contaminated, groundwater supplies may remain so for years
or decades (2).

When faced with a groundwater contamination problem, communities and/or indi-
viduals must either find new sources of drinking water, or treat the
contaminated water.  Within those two general approaches, several options may
be available.  Examples are presented in Table 1.


   TABLE 1. OPTIONS AVAILABLE FOR SOLVING GROUNDWATER CONTAMINATION PROBLEMS

Options Involving New Sources             Options Involving Treatment

      New well                   Central Treatment:
      Purchased water              Aeration
      Bottled water                Blending
                                   Carbon adsorption
                                   Macroreticular resins

                                 Point-of-Use Treatment:
                                   Carbon adsorption
                                   Reverse osmosis (with appropriate membranes)
                                   Distillation

-------
For large communities, all of these solutions may be feasible; but small
communities may not be able to afford new wells or central treatment.  The
best economic choices for a small community may be bottled water or
point-of-use treatment.  Point-of-use treatment may include whole house
treatment, faucet mounted devices, or under the sink line bypass treatment.

Recommended maximum contaminant levels (RMCLs) were published in the Federal
Register for nine of the 14 VOCs listed earlier: p-dichlorobenzene,
1,1,1-trichloroethane, benzene, 1,1-dichloroethylene, vinyl chloride,
1,2-dichloroethane, carbon( tetrachloride, tetrachloroethylene, and
trichloroethylene.  RMCLs are nonenforceable goals for public water systems.
They are established based on health effects data only.  Before maximum
contaminant levels can be established, economics must be determined, including
best available technology.  The National Primary Drinking Water Regulations
state that appropriately designed point-of-use devices that have been shown to
be effective for VOC control can also be considered for some small water
systems, if they are cost effective and properly managed (3).

In a previous study, various point-of-use carbon treatment devices were tested
in the laboratory with spiked municipal water.  Some were found to be
effective for VOC control (4).
OBJECTIVES

The purpose of this study was to determine whether point-of-use carbon
treatment is cost effective for the control of VOCs and to study water quality
district management techniques.  Point-of-use treatment efficacy and costs
were monitored in two small communities for the control of six VOCs and four
trihaloraethanes (THMs).

In Silverdale, Pennsylvania, the water contamination problem had been
identified, but no course of action had been determined.  This community
provided the opportunity to observe and study point-of-use technology from
equipment selection through an extended period of operation.  Several
point-of-use treatment device models were selected to provide a representative
sampling of commercially available products.

In Rockaway Township, New Jersey, the local health department had already
initiated a water quality district program to evaluate the feasibility of
point-of-use treatment for controlling VOC contamination of private wells.
Rockaway Township provided an opportunity to observe a point-of-use strategy
already in operation.

-------
                                 SECTION 2.

                           SUMMARY AND CONCLUSIONS
1.  Data from this study indicate that point-of-use granular activated
    carbon (GAG) treatment devices effectively reduced concentrations of
    trichloroethylene, tetrachloroethylene, carbon tetrachloride,
    1,1,1-trichloroethane, 1,1-dichloroethylene, 1,1-dichloroethane, and
    chloroform at influent concentrations studied.  These results confirm
    bench and field results from a previous study (4).

2.  Breakthrough (defined as consistent detectable concentrations of effluent
    VOCs) was not observed for point-of-use devices that had been in use for
    up to 24 months.  However, trace concentrations of VOCs were detected
    intermittently in postdevice samples for all models in place through the
    entire study.

3.  Monitoring (which may include sampling, meter reading, and/or mechanisms
    for homeowner feedback) should be conducted periodically for water
    quality districts.  For most cases of VOC reduction, it will be more cost
    effective to replace cartridges prematurely rather than to pay for
    frequent sampling and analysis.  For this method to be used for control
    of water quality, a relatively consistent raw water quality is required.

4.  As with central treatment, maintenance must be provided after installa-
    tion.  Homeowners must be aware of how to request maintenance and moni-
    toring.  Some homeowners did not report operational problems immediately.

5.  Microorganisms measured by the standard plate count method were present
    in higher numbers in postdevice water than in predevice water.

6.  No evidence of coliform bacteria colonization was found in any of the
    point-of-use devices.

7.  Variation of sampling techniques for microbiological testing of water
    passing through point-of-use devices will significantly affect the
    results.  If microbiological monitoring of point-of-use devices is
    planned, sampling methods should be given careful consideration.  The
    methods chosen should be consistently followed.

8.  Flushing 1 L of water through the point-of-use device before collecting a
    sample for microbiological analysis reduced the mean standard plate
    counts by one order of magnitude over unflushed samples, to a level below
    500 organisms per mL.

-------
 9.   Flushing water through the point-of-use device for two minutes reduced
     the mean standard plate count by two orders of magnitude over unflushed
     samples, to a level of less than 100 organisms per mL.

10.   The mean standard plate count in treated water from silver impregnated
     GAG devices was not less than the mean standard plate count from
     non-silver impregnated devices.

11.   Carbon,  replacement cartridges,  and point-of-use devices should always be
     stored or packaged in such a way as to minimize possible contamination by
     solvents.  A quality control system should be established by manufactur-
     ers and users to help assure contamination has not occurred.  Immersing
     cartridges for 1-2 days in "organic free" water and then analyzing for
     organic contaminants was effective for detecting methylene chloride
     contamination in this study.

-------
                                  SECTION 3.

                                  PROCEDURES
STUDY SITE SELECTIONS

Silverdale, Pennsylvania, and Rockaway Township, New Jersey, were selected as
study sites for the project.  Selection of these sites was based on previous
sampling results and interest in both communities in participating in the
study.

The primary site for studying point-of-use removal of organic contaminants was
Silverdale, Pennsylvania.  New point-of-use devices were installed and
monitored for a period of 14 months beginning on February 28, 1983.  Data
collected from this site were used to study the feasibility, costs,
performance, and management of point-of-use technology.

Before this-project began, the Rockaway Township Department of Health and
Welfare had been conducting a study to verify, under actual use conditions,
previous U. S. Environmental Protection Agency (EPA) studies concerning GAG
point-of-use treatment devices for organics reduction in drinking water.  For
this project Rockaway Township was used primarily as a case study, although a
limited number of samples were collected to verify performance of the devices
used.  Point-of-use devices at this site had been in use for a year when the
project began.

SITE DESCRIPTIONS

Silverdale, Pennsylvania

The Village of Silverdale is located 30 miles north of Philadelphia and lies
within the Pleasant Springs Creek Drainage Basin in central Bucks County.  The
Village has an estimated population of 550, consisting of approximately 200
residences and 15 commercial establishments.  Individual septic tanks were
used for wastewater treatment within the community until 1981.

Sixty-three homes (about 150 people) in the Village get their water from the
W. C. Seidel Water Company.  The water company's wells, pumps, and
distribution system are privately owned by William and Esther Seidel.  The
Seidels also serve as part time water system operators.  The water source
consists of four eight-inch diameter wells with depths of 275, 180, 200, and
150 feet (a table of English to metric conversions is on page xi).  All wells
are located within a 100 foot radius.  Well No. 4 (150 feet) is used as the
primary water source.  There are two galvanized steel pressure tanks with
capacities of 1000 and 3000 gallons.  Pressure in the tanks ranges from 60 to

-------
80 psl.  Pressure at the farthest point in the distribution system has never
been measured at less than 40 psi.  Flow through the system is estimated to be
11,000 gallons per day.  The only treatment provided is gas chlorination for
disinfection.

The primary source of groundwater recharge in the area is local precipitation.
Water from the wells is moderately mineralized (334-455 mg/L total dissolved
solids), with low levels of iron (0.30-0.43 mg/L) and manganese (0.018-0.036
mg/L).  Total organic carbon (TOG) levels in the distribution system average
0.59 mg/L.  A summary of water quality in the distribution system and wells is
presented in Table 2.
           TABLE 2. WATER QUALITY MEASURED IN SAMPLES FROM FOUR HELLS
                 AND THE DISTRIBUTION SYSTEM IN SILVERDALE, PA.
                                            Mean Concentration
          Analyte

Alkalinity (mg/L as CaC03)
pH (units)
Color (units)
Odor (threshold no.)
Total Dissolved Solids (mg/L)
Total Organic Carbon (mg/L)
Turbidity (NTU)
Arsenic (mg/L)
Barium (mg/L)
Cadmium (mg/L)
Chloride (mg/L)
Copper (mg/L)
Fluoride (mg/L)
Iron (mg/L)
Lead (mg/L)
Manganese (mg/L)
Mercury (mg/L)
Nitrate-N (mg/L)
Selenium (mg/L)
Silver (mg/L)
Sulfate (mg/L)
MBAS (mg/L)
Well
133
7.
<5
1
384
2.
1.
0.
0.
0.
30
0.
0.
0.
0.
0.
0.
0.
<0.
53
<0.
1

66



1
0
012
232
0003

002
30
002
033
0002
7
002
0002

05
Well
122
7.
<5
1
455
2.
1.
0.
0.
0.
76
0.
0.
0.
0.
0.
4.
0.
<0.
42
<0.
2

54



6
0
005
282
0002

008
31
003
030
0002
1
003
0002

05
Well
108
7.
<5
1
455
5.
4.
0.
0.
0.
86
0.
0.
0.
0.
0.
3.
0.
<0.
36
<0.
3

56



0
2
012
263
0002

010
43
007
027
0003
0
003
0002

05
        Distribution
Well 4     System1
130
7.63
<5
1
334
<0. 1
0.6
0.008
0.362
0.0002
27
0.006
0.31
0.003
0.036
0.0004
1.2
0.003
<0.0002
31
<0.05
131
7.43
<5
1
361
0.59
0.5
0.008
0.309
0.0004
26
0.014
0.30
0.002
0.018
0.0002
1.2
0.003
<0.0002
32
<0.05
  Composite of well water  collected  at  a  homeowner's  faucet which  was  not
  connected to the point-of-use  device.   The  contribution  of  each  well to  the
  composite was  not known.
 In  October  1979,  the  Bucks  County  Department  of  Health  began  routine  monitor-
 ing of  the  W.  C.  Seidel  Company's  water  for VOCs.   Initial  results  showed
 trichloroethylene (TCE)  at  10 yg/L and tetrachloroethylene  (PCE)  at 2 yg/L.
 However,  within  a year,  concentrations had  reached  65 yg/L  for TCE  and 12 yg/L

-------
for PCE.  The source(s) of the contaminants could not be identified by the
Health Department.  Septic tank cleaning compounds which had been used in the
community until 1981 were suspected.  Many of the septic tanks in the
community were close to the wells.  The length of time the VOCs had been
present in the distribution system is unknown.

Silverdale was chosen as a study site because central treatment would be
extremely expensive for such a small community and point-of-use treatment
seemed feasible.  Also, the Seidel Water Company and its customers were
willing to participate.

Equipaent Selection and Description

Selection of point-of-use devices was based on criteria developed by the
Project Advisory Committee.  The Committee was composed of both regulatory
officials and representatives of the point-of-use industry (see Acknowledge-
ments). Criteria for selection of devices for this study were as follows:

    - Greater than 95 percent reduction of halogenated organics demon-
      strated in the Gulf South Research Institute (GSRI) Phase III study,
      or equivalent (4).  Manufacturers of units not involved in the GSRI
      study were required to provide validation of testing by an
      independent laboratory to establish comparable performance.  The
      independent laboratory had to have EPA certification for organics
      analyses.

    - Manufacturers were required to certify that their products will meet
      the National Sanitation Foundation (NSF) Standard 53, Section 3,
      concerning structural integrity, corrosion resistance, nontoxicity,
      etc.

    - Point-of-use devices were required to have a rated capacity
      exceeding 700 gallons (estimated one year service life).

    - Point-of-use devices were required to be installable and
      maintainable by a single contractor.  A replaceable treatment
      cartridge was a desirable feature.

    - Point-of-use devices were required to provide on-demand water with
      no maintenance required by homeowner.

    - Manufacturers were required to assure price and availability (up to
      15 units within 15 days at a competitive price).

    - The community would have the right to select from the list of
      qualified devices.

    - No fewer than eight (preferably 10) of each model selected were to
      be used in the study.  That was the minimum number the committee
      determined would provide sufficient data for each model.
      (Forty-seven homeowners participated; consequently the maximum
      number of models which could be tested was five.)

-------
Information concerning the study objectives and equipment  selection  criteria
was sent to known point-of-use manufacturers by the Water  Quality Association
and NSF.

Nine manufacturers provided the required product information.  Product
information was reviewed by NSF.  The information for products meeting  the
selection criteria was summarized and coded.  Trade names  and manufacturers'
names were not identified.  This information was given  to  a  committee composed
of six citizens of the community.  After careful review, the committee  listed
the coded products in order of preference.  The top five were selected  for
installation.  Selection of a total of five models was  based on the  number  of
homeowners expressing a desire to participate in the study (47) and  the
criteria for using a minimum of eight to ten of each model.

For this study, only line bypass models were selected.  Descriptive  summaries
of the models used in Silverdale appear in Table 3.

Rockaway Township, Hew Jersey

Rockaway Township is located 18 miles northwest of Newark, New Jersey,  in the
county of Morris.  It is 45 square miles in area with an estimated population
of 20,000.  The study was conducted in the 320 acre Lake Telemark subdivision
located in the northern part of the township.  The subdivision lies  within  the
Hibernia Brook Drainage Basin and is primarily residential,  consisting  of
approximately 310 private homes and a small commercial  district.  All homes
within the subdivision are served by individual wells.  Wastewater is treated
onsite with septic tanks.

The geology of the area is characterized by hills of crystalline bedrock
composed primarily of granite and gneiss of Precambrian origin.  The bedrock,
which is fractured and jointed, displays frequent outcroppings and may  be
covered by up to 15 feet of soil and glacial till (6).

The story of groundwater contamination in Rockaway Township began with  the
city of Rockaway.  In November 1979, high levels of TCE (50-220 yg/L) were
detected in two of three wells used to provide water for the 10,000  residents
living in the township.  This area receives centrally treated water.  The two
wells were removed from service.  The third and largest well remained in use.
In October 1980, two ether compounds, di-isopropyl ether and methyl  tert-butyl
ether, were found at levels sufficient to cause taste and  odor problems
(70-100 yg/L and 25-40 yg/L, respectively).  A water emergency was declared
and residents were advised to avoid drinking the water.  Nearby unaffected
supply systems were identified and residents were advised  to obtain  drinking
water from these sources.

Township officials considered the available treatment options and chose GAG
adsorption as the best approach.  A GAG adsorption system  was purchased for
$200,000 (including piping, site work, and a building housing the contactors).
The carbon was expected to last six to eight months; however, it was necessary
to replace the carbon once every four to six weeks (5).

Subsequent investigations indicated that the length of  time between  replace-
ment of GAG and breakthrough could be increased significantly by aerating the

                                    8

-------
water prior to treatment with the carbon.  By the end of 1981, a packed column
aerator was installed at a cost of $375,000.  The combined systems have proven
effective, but have increased the costs of potable water from $0.27/1000
gallons before treatment for organics removal to $0.48/1000 gallons after (5).
The source of the ether contamination was identified by the township.  As a
result, most of the installation and operating costs for the system have been
borne by the company responsible for contaminating the water.
       TABLE 3. DESCRIPTIONS OF POIHT-OF-DSE DEVICES USED IN SILVERDALE
Device
      Description
           Approximate
Estimated   Volume of
        1          2
Capacity     Carbon
(gallons)   (gallons)
             Estimated
Carbon Wt.2 Contact Time3
 (grams)	(seconds)
       Two PVC cartridges in     1000
       series containing
       carbon.  Carbon is
       silver impregnated.

       Single stainless steel    2000
       cartridge containing
       carbon.
                                       0.18
                                       0.65
                             417
                            1708
                  14
                  78
   E
Microfilter and two       2000
acetal copolymer
cartridges containing
carbon.  Carbon is
silver impregnated.

Single stainless steel    1000
cartridge containing
carbon, compressed into
a solid "filter".
                                              0.46
                0.20
                            1150
                  19
     300
1
       Two cartridges; one
       with GAG, one with PAG
       and filterable materials.
                          2000M
                0.44
     765
62
 Estimates for Silverdale water quality based on total weight of carbon and
 previous performance data (GSRI).

 Data from GSRI Phase II Report.

 Contact times based on manufacturers' specified maximum flow rate.
  Manufacturer only claims 1000 gallon capacity in product literature.

-------
Six to seven months after installation of the central GAG system, organic
contaminants were found in water samples from a cluster of private wells
located one mile north of the Lake Teleraark subdivision.  Because citizens
were already sensitized to water contamination, the Rockaway Township Health
Department became deluged with requests to sample private wells.  The health
department was legally limited in its involvement in monitoring private wells.
It was the homeowner's responsibility to provide proof of contamination before
health officials could take action.  As a result, homeowners were required to
pay for water analysis for the initial sample collected from their wells.
A low price of $50 per gas chromatography/mass spectrometry (GC/MS) analysis
for any township resident was obtained through competitive bidding.  The
health department assisted in obtaining the low bid and reviewed and monitored
analytical and sampling procedures.  Cost to the health department for review
and monitoring services was estimated to be $10.00 per sample, including
labor, fringes, equipment, and automobile expenses.

Groundwater in the Lake Telemark area originates almost entirely from local
precipitation, is relatively unconfined, and moves approximately parallel to
the slope of the land (7).  Recharging of the aquifer is difficult because of
the fractured nature of the underlying bedrock.  The groundwater has moderate
levels of dissolved solids (312 mg/L), low pH (5.76), low alkalinity (25.6
mg/L as CaC03), and low concentrations of total organic carbon (0.8 mg/L).
(Concentrations reported are averages from four wells.)   Wells in the Lake
Telemark subdivision range from 50 to 200 feet in depth and have an average
yield of six gallons per minute (gpm).  A summary of the analysis of water
samples from four Lake Telemark wells is presented in Table 4.

Samples from 50 wells in the subdivision were analyzed for organic
contaminants; 14 had detectable concentrations of VOCs.  The VOCs found
included:  trichloroethylene, 1,1,1-trichloroethane, 1,2-dichloropropane,
1,1-dichloroethylene, tetrachloroethylene, and 1,1-dichloroethane.  Ten wells
had VOC concentrations high enough that the township health department
recommended the temporary use of bottled water.

A consulting engineering firm was hired by the township in May 1981 to
determine the size of the contaminant plume, direction of plume flow,
probable contaminant sources, and to estimate the length of time the aquifer
would be contaminated (7).  Two contaminant plumes were found; both from an
undetermined source or sources.  It was estimated that the VOCs would remain
in the aquifer for 12 years.  The report also suggested that, given the slow
rate of movement of contaminants in the aquifer, remedial actions be
implemented to provide some type of relief to the affected homeowners (7).

A citizens' ad hoc committee was formed to investigate options available for
providing potable water to the affected residents.  Point-of-use treatment was
selected by the committee as the most economical treatment option.  The costs
of providing homes in this subdivision with centrally treated water was
estimated to be $4,000,000, compared to the estimated costs of $60,000 for
installing point-of-use devices in all dwellings.
                                     10

-------
      TABLE 4.  GRODHDHATER QUALITY MEASURED IN SAMPLES FROM FOUR PRIVATE
                     WELLS IN THE LAKE TELEMARK SUBDIVISION
          Analyte

Alkalinity (mg/L as CaCOs)
pH (units)
Color (units)
Odor (threshold no.)
Total Dissolved Solids (mg/L)
Total Organic Carbon (mg/L)
Turbidity (NTU)
Arsenic (mg/L)
Barium (mg/L)
Cadmium (mg/L)
Chloride (mg/L)
Chromium (mg/L)
Copper (mg/L)
Fluoride (mg/L)
Lead (mg/L)
Manganese (mg/L)
Mercury (mg/L)
Nitrate-N (mg/L)
Selenium (mg/L)
Silver (mg/L)
Sulfate (mg/L)
MBAS (mg/L)
Well 1

 25.0
  5.97
   Sample Source
Well 2      Well 3
  0
304
  0.55
  0.13
 <0.002
  0.021
 <0.0001
 66
 <0.0002
  0.27
 <0.10
 <0.0001
  0.018
 <0.0002
  2.84
 <0.001
 <0.0002
 30.0
  0.029
  21.3
   5.81
18.8
 5.54
Well 4
  37.5
   5.72
0
228
0.80
0.25
<0.002
0.020
<0.0001
38
<0. 0002
1.72
<0.10
<0.0001
0.010
<0.0002
4.16
<0.001
<0.0002
38.8
0.027
0
422
0.40
0.49
<0.002
N/A
N/A
140
N/A
0.10
<0.10
N/A
N/A
N/A
5.74
N/A
<0.0002
27.0
0.026
0
294
1.55
0.26
<0.002
N/A
N/A
41
N/A
S/B
<0.10
N/A
N/A
N/A
6.67
N/A
<0.0002
31.6
0.036
N/A - Not analyzed based on low concentrations found at first two sites.
S/B - Sample bottle broken in transit to lab.
The committee was then asked to oversee a demonstration project using
point-of-use treatment, and began the process of selecting a device which
would provide maximum health protection with minimum maintenance.  Several
types of devices (distillation units, whole-house treatment systems, and line
bypass devices) were investigated.  Manufacturers' representatives were
interviewed and allowed to provide equipment demonstrations.

One point-of-use model was selected for use in 12 homes for demonstrating
efficacy. The model used in Rockaway was also selected for the study in
Silverdale (Model F).

The township entered into a contract with the manufacturer which included a
price reduction and did not require payment until the devices had proven their
effectiveness.   A provision was also made that homeowners not selected for the
demonstration project could purchase devices for the same reduced price (25
homeowners have done so).  The manufacturer also agreed to arrange and pay for
installation and maintenance, including cartridge replacement, for a period of
one year, and to pay 50 percent of the analytical costs incurred to
demonstrate efficacy.  Monies for sampling costs and the remaining 50 percent
of analytical costs were provided by the Rockaway Township Health Department.
                                    11

-------
Participating homeowners were required to sign documents holding the health
department harmless for any negative effects resulting from use of the
devices.  Installation was completed in October 1981.

At the start of this EPA/NSF study, the devices had been in operation for
approximately one year, and were demonstrating greater than 99 percent
efficiency in removing VOCs.  Average total volume treated per device at this
time was estimated to be approximately 800 gallons.

SAMPLING AND ANALYTICAL PROCEDURES

Qualified sample collectors for each study site were selected by NSF staff.
Sampling personnel were provided with written sampling and preservation
instructions, and with required sampling containers for the analyses to be
performed.  Sampling techniques were reviewed onsite to ensure proper
sampling.  Sample types collected were grab samples of predevice and
postdevice water in both communities.  Grab samples were also collected
directly from the supply wells in Silver-dale.

Volatile Organic Chemicals

Sampling, sample preservation, and analytical procedures for VOCs were in
accordance with Federal Register, 44, No. 231, Part II, November 29, 1979, and
Federal Register, 44, No. 233, Part III, December 3, 1979, as amended.
Specific analytical methods used were: liquid/liquid extraction (EPA Method
501.2), purge and trap (EPA Method 601), and gas chromatography/mass
spectrometry (EPA Method 624).  Use of these methods allowed qualitative and
quantitative determination of a broad spectrum of VOCs.

Samplers were Instructed to let water run for 1 to 2 minutes before collecting
a VOC sample.  For predevice samples, this assured samples were representative
of water in the distribution system.  Postdevice samples collected from
flowing taps assured that the water sampled represented the minimum
water/carbon contact time.

All VOC samples were collected in duplicate.  A field blank accompanied every
sample batch.  VOC samples were shipped in freezer packs to NSF for analysis.

Although sampling procedures were in accordance with prescribed techniques
(11), 17 percent (134 samples) of the 788 samples collected had a small
headspace when they reached NSF.  The headspace volumes were estimated to be
zero to  1.0 mL.  Samples were collected in duplicate, and one of a pair was
usually free of headspace.

A study of the effect of headspace and holding time was conducted to determine
if sample headspace and headspace combined with holding time would cause
significant loss of VOC.

A description of the study and results is presented in Appendix A.  The
results  indicated that significnt loss of VOC did not occur for up to 0.5 mLs
of headspace.  Data from samples with apparent headspace greater than 0.5 mLs
were not used in this report.
                                     12

-------
EPA Method 501.2 for liquid/liquid extraction is used primarily for
determination of trihalomethanes in finished drinking water, drinking water
during the treatment process, and raw source water.  The technique is not
restricted to trihalomethanes, however, as it efficiently extracts a broad
range of non-polar organics, and also certain polar organics, with varying
efficiencies.  The technique involves extraction of the sample using pentane,
and injection of a portion of the extracted pentane into a gas chromatograph
equipped with a linearized electron capture detector.  A 30 mL aliquot of
sample was extracted instead of the normal 10 mL volume.  This modification
improved detection, precision, and accuracy.

EPA Method 601 uses a variation of the Bellar purge and trap technique for
measurement of 29 purgeable halocarbons, including the trihalomethanes,
trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane,
1,2-dichloroethylene, 1,1-dichloroethane, and carbon tetrachloride.  The
method is used to measure specific compounds which are expected to be present
in the water to be analyzed.  The technique requires injection of a sample
into a specially designed purging chamber through which an inert gas is
bubbled.  Halocarbons contained in the generated vapor are trapped in a
sorbent tube, which is then heated and backflushed to desorb the compounds
into a gas chromatograph.

Gas chromatography/mass spectrometry (GC/MS) (EPA Method 624) is applicable to
the determination of the same VOCs which can be detected by the purge and trap
method.  For this study, GC/MS was used as a screening mechanism to confirm
the presence of the compounds detected by the liquid/liquid extraction and
purge and trap methods.  This technique is essentially the same as the purge
and trap technique with the addition of a mass spectrometer, which provides
both qualitative and quantitative information.

The Quality Assurance plan was administered in accordance with the NSF Quality
Assurance Manual, NSF, Ann Arbor, Michigan, 1983, and the NSF Chemistry
Laboratory Quality Control Manual, NSF, Ann Arbor, Michigan, 1983.  Routine
analytical quality control included: duplicate analyses of all samples;
analysis of field blanks with each batch of samples; spikes, controls and
standards with each analytical series; and periodic anlaysis of EPA QC check
and performance standards.  A detailed quality control data package was
submitted to the EPA.  A summary of precision and accuracy results for VOCs
using liquid/liquid extraction is presented in Table 5.


TABLE 5. PRECISION AND ACCURACY OF VOC ANALYSES USING LIQUID/LIQUID EXTRACTION

       Analyte             Percent Relative Precision    Percent Accuracy

1,1,1-Trichloroethane                 8±5                   93 ^ 6
Tetrachloroethylene                   8 + 5                   98+6
Trichloroethylene                     2 + 1                   99 _+ 5
Chloroform                            5 + 5                   93+4
Bromoform                             4 +_ 4                   95 +_ 6
Bromodichloromethane                  9 +_ 6                   95 +_ 8
Chlorodibromomethane                 13+13                  99+5
                                    13

-------
Laboratories in New Jersey and Pennsylvania were selected to perform analyses
for which holding times did not allow sample transport to NSF.  Subcontract
laboratories had current EPA certifications for the analyses they performed.

Microbiological

Samples analyzed for standard plate count and total coliform content were
collected using several different sampling techniques.  Techniques included
the standard method of collecting water samples for microbiological analysis,
i.e., from disinfected taps which were flushed for two to three minutes.  The
two minute flush provides water which is representative of water in the
distribution system.  This technique was used for samples collected from
August 1983 through April 1984.  Samples were also collected from unflushed,
undisinfected taps (March 1983-February 1984), and from undisinfected taps
after a one liter flush (March 1984-April 1984).  The latter techniques were
efforts to collect samples representative of water in the pipe between the
carbon bed and the end of the tap, and of water in the carbon bed itself.
These procedures were used for both predevice and postdevice samples.

In Silverdale, all samples were collected in the morning.  Homeowners were
notified in advance of the dates for which their homes would be sampled.  They
were asked not to use the device on the morning of sampling until after the
samples were collected.  Samples for microbiological analysis were always
collected before VOC samples.

Preparation of sample containers, sample preservation and transportation, and
analysis of samples were performed as outlined in Microbiological Methods for
Monitoring the Environment, Water and Wastes, USEPA-600/8-78-017, December
1978.  The following sampling modification was made:

    . Chambers neutralizing solution (5 percent sodium thiosulfate, 7.3
      percent sodium thioglycolate) was added to samples collected from
      devices containing silver impregnated carbon (1 mL neutralizer to
      100 mL sample).

General Water Quality Analysis

Sampling, preservation and analytical procedures used for all other analytes
were in accordance with Methods for Chemical Analysis of Water and Wastes,
USEPA-600/4-79-020, March 1979.

Cost Data Collection

Reported costs for point-of-use GAC treatment in Silverdale and Rockaway were
based on either direct experiences during the study, community records, or  if
specific costs were unavailable, on the assumptions presented in the text.
                                     14

-------
                                  SECTION 4.

                                    RESULTS
VOLATILE ORGANIC CHEMICAL REDUCTION

Predicting Breakthrough

All devices used in this study used activated carbon as the adsorption media.
The carbon adsorption process is particularly amenable to the removal of
organics from water.  Certain specific organic contaminants may be more
effectively or more cheaply removed by other processes, but activated carbon
is the best process for reduction of a broad spectrum of organics (8).

When designing a carbon adsorption system for removal of organic chemicals
from water for a specific treatment goal, factors which affect the efficiency
of the adsorption process can be optimized.  Manufacturers of point-of-use
devices do not have the same opportunity to design for a specific water
treatment goal.  Point-of-use devices are designed with many important
variations; quantity of carbon, type of carbon, internal flow patterns, number
of cartridges, etc.  All of these variables can affect performance.
Point-of-use treatment devices are generally designed to provide effective
treatment for a variety of water qualities.  Looking at any one factor and
trying to predict performance for a specific water treatment problem or device
may be misleading.

Once a device is installed, however, the adsorption principles which affect
the removal of organics will not change.  Consequently, consistent performance
may be anticipated.  A knowledge of these adsorption principles, coupled with
performance information for a specific device, may be used to predict
breakthrough behavior and establish an effective monitoring plan.

The following discussion briefly focuses on the basic principles of carbon
adsorption.  An example is presented of how those principles might be applied
to predict the order of breakthrough for a mixture of VOCs using one type of
carbon.

Carbon is a good adsorbent because it provides a large surface area per unit
volume.  Although there are areas of localized charge on the carbon surface,
the adsorption mechanism is "physical" as opposed to electrostatic for
exchange sorption.  The energy for physical adsorption is predominantly
provided for by van der Waals forces.  The relatively weak adsorption energy
allows molecules to adsorb and desorb, moving along the carbon surface.  One
molecule may also be displaced by a molecule having stronger carbon affinity.
                                    15

-------
Adsorption of an organic molecule from water onto  carbon  is  the  result  of  two
properties of the solvent-water-carbon system.   The  first property  is  the
tendency of a molecule to leave solution  (hydrophobicity).   The  tendency to
leave solution can be indicated by  the solubility  of  the  molecule.   That is,
the less soluble the molecule, the  less likely  it  will  stay  in solution.
Larger molecules may exhibit both hydrophilic and  hydrophobic properties,  with
functional groups and/or branches of one  molecule  exhibiting contrasting
solubility properties (9).

The second property is the affinity of the molecule  for the  carbon  surface.
This affinity is the result of a number of properties of  the molecule and  the
carbon surface.  Molecular size and structure will influence the energy of
adsorption between the molecule and the surface  and  the rate of diffusion  of
the molecule within the internal pore structure.   The surface characteristics
of the carbon will also influence the energy of  adsorption.  Molecular
affinity for carbon can be demonstrated empirically with  isotherms.  Isotherms
represent the quantity of solute adsorbed onto  the carbon surface as a
function of solute concentration in water.

The properties affecting adsorption can be assessed collectively using
empirical solubility and isotherm data.   Isotherm  and solubility data for  the
VOCs of interest in this study are presented in  Table 6 (10).  Column one
shows the concentration of granular activated carbon  (mg/L)  required for a 100
yg/L reduction in the VOCs listed.  This magnitude of reduction fits the range
of VOC concentrations found in this study.  The  solubilities of the VOCs are
presented in column two.
    TABLE 6.  RELATIVE CARBON AFFINITY FOR VOCs AND VOC SOLUBILITY IN WATER

                                    Required Carbon                Solubility
         VOC                          Dose (mg/L)1                (mg/L @ 25°C)

Methylene chloride                       1,100                      20,000
Chloroform                                 210                       8,200
1,2-Dichloroethane                         190                       8,300
1,1-Dichloroethane                         120                       5,000
1,1,1-Trichloroethane                       90                       4,400
1,1-Dichloroethylene                        70                       5,000
1,2-Dichloropropane                         68                       2,700
Carbon tetrachloride                        61                         795
Trichloroethylene                           15                       1,000
Tetrachloroethylene                        7.1                         200
 Computed from Freundlich isotherm data for Filtrasorb 300 as the quantity of
 carbon in a column adsorber required for a 100 yg/L reduction of contaminant
 at neutral pH (10).
The VOCs are listed in increasing order of adsorbability, with methylene
chloride least adsorbable and tetrachloroethylene most adsorbable.  That order
would also be the same order for which breakthrough would likely occur.

                                    16

-------
Therefore, if methylene chloride and tetrachloroethylene were co-contaminants,
methylene chloride would be found in device effluent before tetrachloroethy-
lene.

The correlations between solubility and isotherm data are good for these
compounds.  One might surmise from both sets of data that breakthrough would
occur in the order presented.  The use of these data as a predictive tool
could theoretically reduce analytical costs for a point-of-use monitoring
program.  Analysis for only the least adsorbable chemical in the device
effluent would reduce analytical costs.

Silverdale, Pennsylvania

Table 7 presents a summary of VOCs found in Silverdale water samples during
the study.  Mean results of VOC analyses of well water and predevice samples
are presented.

Well water samples were collected prior to chlorination.  Predevice samples
were collected at kitchen cold water taps.  Predevice sample results reflect
the quality of the influent water after some residence time in the
distribution system.

TCE and PCE were the contaminants most consistently measured at relatively
high concentrations in predevice samples during the study (approximately 80
and 20 yg/L, respectively).  Chloroform and carbon tetrachloride were also
consistently found in predevice samples, but at concentrations generally less
than 10 yg/L.


                   TABLE 7. SUMMARY OF INFLUENT VOC RESULTS,
              SILVERDALE, PENNSYLVANIA (MARCH 1983 - APRIL 1984)

                           Mean Concentrations (yg/L)1
      Compound          Well 1    Well 2    Well 3    Well 4    Predevice

Trichloroethylene        36.2      12.0      22.0      83.8       80.4
Tetrachloroethylene       7.5       3.5       5.0      19.7       20.6
1,1,1-Trichloroethane     1.1      <1.0      <1.0       1.4        1.1
1,2-Dichloroethane       <1.0      <1.0      <1.0      <1.0       <1.0
Carbon Tetrachloride      1.4      <1.0       1.1       7.0        8.0
Chloroform                1.6       1.0       4.3       5.5        6.7
Bromodichloromethane      1.0      <1.0       3.4       1.2        1.5
Dibromochloromethane      1.4      <1.0       2.6      <1.0        1.4
Bromoform                 1.2      <1.0      <1.0      <1.0       <1.0
 Samples with <1.0 yg/L of a VOC were assigned a value of 0.9 yg/L for
 calculation of the mean.
                                    17

-------
140 -
130 -
120-
110 -
100 -
90 -
80 -
yg/L 70 -
60 -
50-
40 -
30 -
20 -
10 -















I
I
I








I
I
I
I
*•*



I
I
i
I *





i
I.
i :*







I
I
I
I *




i
I
I
l-.i




i

:}







I
| i
I i








i











f
!

	
*Nonde
TCE
_PCE ,
_ TTHM's
Other VOCs
itectable








I
I.

*






I
I
I !






,
I
':!







i
id
            M   A   M   J    J
O   N   D   J
                                                           M   A
Figure 1. Mean VOC concentrations in influent samples collected by month.
                                   18

-------
Figure 1 shows the mean influent VOC concentrations in distribution system
(predevice) samples.  The results are presented by month for TCE, PCE, TTHMs,
and the other VOCs found (1,1,1-trichloroethane and carbon tetrachloride).
The concentrations of VOCs found varied from sample to sample.  For example,
the range of results for TCE was 133 yg/L with a relative standard deviation
of +_ 31 percent.  Predevice samples were collected at different locations in
the distribution system during the study, but there was no apparent pattern in
VOC concentrations which could be attributed to sampling location.  There also
does not appear to be a seasonal pattern to VOC concentrations.

Figure 1 does seem to show an increase in influent TCE concentration during
the study.  In fact, the mean TCE concentration from March 1983 through
September 1983 was 69.4 yg/L.  From October 1983 through April 1984 the mean
TCE concentration was 80.9 yg/L, an average increase of 31 percent.  The mean
PCE concentrations for the same two time intervals were 19.5 yg/L and  21.7
yg/L, an 11 percent increase.  The VOC concentrations in Silverdale ground-
water are not decreasing.

The influent VOC concentrations closely paralleled the concentrations  found in
Well No. 4.  Well No. 4 was used as the primary water source for Silverdale.

The four wells are close together and of similar depths (see site descrip-
tions).  The local health department indicated that VOC concentrations in the
well water appear to increase with use rate (13).  Consequently, regardless of
which well is selected as the primary source, influent VOC concentrations will
not vary significantly.

Postdevice samples were collected from each home a minimum of one time every
two months during the 14 month sampling period.  Each location was sampled at
least seven times during the study.  Additional samples were collected when
VOCs were detected to verify the result.  Therefore, some locations were
sampled more than seven times.

For this study, breakthrough was defined as detection of the same VOC  in con-
secutive postdevice samples from the same location at a concentration  above
the routine detection limit of 1.0 yg/L.  Breakthrough did not occur for any
of the devices tested during the 14 months of sampling for TCE and PCE.
However, trace concentrations of VOCs were detected intermittently in
postdevice samples for all models which were in place during the entire study.
The VOC data are presented in Appendix B.

Overall, the reduction of VOCs achieved by the five models of point-of-use
devices was good.  Table 8 presents the mean and maximum VOC concentrations
for postdevice samples for the time interval March 1983 through November 1983.
Data are presented only through November of 1983, because Model A was  removed
from service in November.  Table 9 presents mean and maximum VOC concentra-
tions for the entire sampling interval, March 1983 through April 1984, for the
four models in service during the entire sampling interval.

In addition to the mean and maximum VOC concentrations, the number of  samples
in which VOCs were detected versus the number of samples analyzed is also
presented.  Of 332 postdevice samples collected during the study, 41 had at


                                    19

-------
least 1 VOC present.  The most frequently measured VOC was TTHMs (chloroform),
which was detected 32 times.  The number of times detected for other VOCs was:
carbon tetrachloride, 16; TCE, 15; 1,1,1-trichloroethane, 13; and PCE, 7.

TCE and PCE were detected at concentrations exceeding the routine detection
limit of 1 yg/L in two consecutive samples at two locations but, these
occurrences were not considered to be caused by breakthrough.  One location
used Model C.  A potential for physical bypass for this model was discovered
because of cartridges which did not match manufacturers size specifications.
All Model C cartridges were replaced with properly sized cartridges, and
additional TCE and PCE breakthrough did not occur.  The other location used
Model E.  A third sample at this location had non-detectable concentrations of
TCE and PCE, but the cartridge was replaced to assure that breakthrough and/or
bypass would not occur.

Model D had two occurrences of TTHMs (specifically chloroform) detected in
consecutive samples above the routine detection limit of 1.0 yg/L.  These were
the last two samples collected for the study at both locations.  The consecu-
tive concentrations measured were 1.6 and 5.0 yg/L, and 1.4 and 1.1 yg/L, con-
centrations well below the EPA maximum contaminant level of 100 yg/L. Although
the mean concentration of chloroform reported in Table 7 is 12 times less than
the TCE concentration, chloroform may be breaking thorugh first.  This would
support the predicted breakthrough order discussed earlier.  This is also
supported by the fact that chloroform was always detected in samples which had
TCE and PCE, but not all samples which had chloroform had TCE and PCE.

Although the overall VOC reduction performance for Model E was good, there was
a higher frequency of detection of trace concentrations of VOCs in Model E
effluent than the other models.  However, there was no pattern in the
detection of VOCs which could be related to volume of water treated.  It was
discovered late in the study that the carbon in some of the cartridges for
Model E had become contaminated with methylene chloride after they were
manufactured.  The presence of an organic contaminant on the surface of the
carbon would affect the adsorption properties and performance of the carbon.
This may be why the Model E effluent results were less consistent than for
other models, and why VOCs were detected more frequently in Model E effluent.

The right hand column in Tables 8 and 9 presents a summary of the volume of
water treated at each location at the time the last VOC sample was collected.
Cartridges for Model C were replaced during the fifth month, so the mean and
maximum flows are lower than the other models.  Model A was replaced by
Model E during the ninth month of sampling, so the final mean volume for Model
E is lower than for Models D and F.  Based on the mean volumes treated for
Models D and F, the average daily flow through the point-of-use devices was
one gallon.
                                    20

-------
      TABLE 8. POSTDEVICE VOC CONCENTRATIONS BT MODEL AND VOLUME TREATED,
             SILVERDALE, PENNSYLVANIA (MARCH 1983 - NOVEMBER  1983)
Model A
 Mean
 Maximum
 No. detected/total'
                      Mean1 and Maximum Concentrations (yg/L),
                      and Number of Samples with Detectable Concentrations
TCE
PCE
 1,1,1-Tri-
chloroethane
                                                    TTHMs
 Carbon
 Tetra-
chloride
0/33    0/33
          0/18
               0/19
   0/15
Volume
Treated
 (gal)2
                                   216
                                  <400
Model C
Mean
Maximum
No. detected/total
Model D
Mean
Maximum
No/ detected/total
Model E
Mean
Maximum
No. detected/total
Model F
Mean
Maximum
No. detected/total

2.0"
19.0
2/33

<1.0
<1.0
0/36

25
22.2
5/36

<1.0
<1.0
0/35

<1.0
<1.0
2/33

<1.0
<1.0
0/36

1.1
4.5
3/36

<1.0
<1.0
0/35

<1.0
<1.0
0/20

<1.0
<1.0
0/20

<1.0
<1.0
0/21

<1.0
<1.0
0/24

<1.0
2.0
3/22

<1.0
<1.0
0/23

<1.0
1.4
1/20

<1.0
<1.0
0/23

<1.0
<1.0
0

<1.0
<1.0
0/17

1.2
4.7
3/14

<1.0
<1.0
0/16

102
227


344
831


343
638


250
603

  Samples with <1.0 yg/L of a VOC were assigned a value of 0.9 yg/L for
  calculation of the mean.

  Volume of water treated at the time the last sample was collected for each
  model.

  The number of times a VOC was detected versus the number of samples
  analyzed.

  31 of 33 samples <1.0yg/l.  Mechanical problem with one cartridge produced
  2 high samples.
                                    21

-------
      TABLE 9. POSTDEVICE VOC CONCENTRATIONS BY MODEL AND  VOLUME  TREATED
              SILVERDALE, PENNSYLVANIA (MARCH  1983 - APRIL 1984^
                               Mean  and Maximum Concentrations  (yg/L)



Model C4
Mean
Maximum
No. detected/total5
Model D
Mean
Maximum
No. detected/total
Model E6
Mean
Maximum
No. detected/total
Model F
Mean
Maximum
No. detected/total


TCE

<1.0
1.0
3/59

<1.0
<1.0
2/68

1.9
24.3
9/104

<1.0
2.9
1/68

1,1,1-Tri-
PCE chloroethane

<1.0 <1.0
<1.0 1.2
2/59 2/46

<1.0 <1.0
<1.0 1.2
0/68 5/56

1.1 <1.0
7.4 1.6
5/104 5/89

<1.0 <1.0
<1.0 1.6
0/68 1/56
Carbon
Tetra-
TTHMs chloride

1.1 <1.0
4.2 <1.0
8/46 2/40

1.0 <1.0
5.0 <1.0
7/56 5/49

1.2 <1.0
19.8 4.7
11/89 7/83

1.0 <1.0
6.6 <1.0
6/56 2/49
Volume
Treated
(gal)3

235
394


516
1,126


287
730


328
721

  Model A was replaced in November and is not included in this table.

2
  Samples with <1.0 yg/L were assigned a value of 0.9 yg/L for calculation of
  the mean.

3
  Volume of  water treated at the time the last sample was collected for each
  model.

4
  Carbon cartridges for Model C were replaced in August because of mechanical
  problems.   Data for Model C are for August 1983 - April 1984.


5 The number of times a VOC was detected versus the number of samples
  analyzed.


  Model E cartridges were replaced in May 1984.
                                    22

-------
Rockaway Township, New Jersey

Twelve point-of-use treatment devices were installed on private wells  in  the
Lake Telemark area of Rockaway Township, New Jersey, on October 22  and 23,
1981.  Samples of the well water were anlayzed for VOC concentrations  a week
before installation.  The results of those analyses are summarized  in
Table 10.
    TABLE 10. VOCs IDENTIFIED IN TWELVE PRIVATE WELLS  IN ROCKAWAY TOWNSHIP1
                             Range of Concentrations         Number of
          VOC                     Found (yg/L)	      Wells With VOCs

1,1,1-Trichloroethane             1.0   -  240.0                 8
1,1-Dichloroethylene              6.7   -   20.7                 4
Tetrachloroethylene               1.0-12.3                 7
1,1-Dichloroethane               <0.4   -10.1                 6
Trichloroethylene                 0.7   -  240.2                 4
trans-1,2-Dichloroethylene        0.8-5.1                 2
Chloroform                        1.7-2.1                 2
Trichlorofluoromethane          <25.0                            1
 Results of samples collected October 14, 1981; courtesy of Rockaway Township
 Health Department.


The predominant organic contaminants were 1,1,1-trichloroethane,  1,1-dichloro-
ethylene, 1,1-dichloroethane, trichloroethylene, and tetrachloroethylene.
These contaminants and their respective concentrations varied considerably
from site to site.  The individual wells used for water supply probably tapped
different aquifiers.  It is also believed that there is more than one source
for these contaminants in the area.

After the twelve devices were installed, the local helath department
supervised the sampling and analysis of water treated by the devices from
October 1981 until October 1982.  The health department reported  100 percent
reduction of all VOCs through the first twelve months.

In September 1982, samples were split between the commercial laboratory which
had been providing analyses for Rockaway Township, the National Sanitation
Foundation, and'a referee laboratory selected by NSF.  The results of the
round robin sampling are presented in Table 11.  These are results of samples
from four wells.   Only results of VOCs reported at more than trace concentra-
tions are shown.   The agreement between the three laboratories was good.

The Rockaway Township Health Department continued limited sampling from
October 1982 through October 1983.  Four of the twelve locations were
scheduled for sampling.   Table 12 presents results for those four sites.
Sites 1, 2, and 3 were sampled in October 1982 and July and October of 1983.
                                    23

-------
              TABLE 11. RESULTS OF ROUND ROBIN ANALYSIS OF PAIRED
                   VOC SAMPLES FROM ROCKAWAT TOWNSHIP (yg/L)
                  (Paired Nondetectable Results Not Reported)
VOC

1,1,1-Trichloroethane
Trichloroethylene
1,1-Dichloroethane
1,1-Dichloroethylene
      NSF1

     220
      58

      70
     160

       8.8
       3.6

       7.2
       2.0
               Laboratory
                 RTS2

                 162
                  41
                  87
                 204
                   8.6
                   2.7

                   3.7
                   0.6
  SRL;

  230
   70

   60
  170

   10
1
 National Sanitation Foundation
?
"Rockaway Township subcontract laboratory

 Subcontract referee laboratory
       TABLE 12. VOC CONCENTRATIONS AT FOUR POINT-OF-DSE TREATMENT  SITES
                          IN ROCKAWAT TOWNSHIP (yg/L)
                          October 1982 - October  1983
Site 1
  Predevice Mean
  Predevice Range
  Postdevice Mean
  Postdevice Range

Site 2
  Predevice Mean
  Predevice Range
  Postdevice Mean
  Postdevice Range

Site 3
  Predevice Mean
  Predevice Range
  Postdevice Mean
  Postdevice Range

Site 4
  Predevice Mean
  Predevice Range
  Postdevice Mean
  Postdevice Range
  1,1,1-Tri-
chloroethane

   155
   82-230
29
10 - 57
      - 12
                                            Trichloro-
                                             ethylene
                1.3
                   - 2
                94
                52
                2
                - 161

                - 4
             155
             150 - 160
                           Tetrachloro-
                             ethylene
                                - 3
- 2
- 1
                                     24

-------
Four additional sites were also sampled in October 1983, the 24th month of
operation.  There were no VOCs measured in postdevice samples at 8 sites after
24 months of operation in Rockaway Township.  VOC data for Rockaway Township
are in Appendix C.  Flowmeters were not installed on devices in Rockaway
Township.

The average use rate in Rockaway was estimated at 2.2 gpd, based on several
readings taken from a total flow indicator on the device.  After two years of
service, average total flow was approximately 1650 gallons.  Efficacy data
demonstrate that the devices used were effective in removing VOCs at the con-
centrations encountered after two years of service.  Only one of the 21 post-
device samples collected yielded detectable levels of any VOC (4 yg/L TCE and
2 yg/L PCE).  One device was sampled after reaching its estimated treatment
capacity of 2000 gallons and still produced water with non-detectable VOCs.

As with most groundwater supplies, the Rockaway predevice TOG concentrations
were low.  This low total organic loading could improve the capacity of
carbon.

MICROBIOLOGICAL GROWTH IN POIHT-OF-OSE DEVICES

Silverdale, Pennsylvania

Bacteriological quality of predevice and postdevice water was monitored at all
installations in Silverdale using Standard Plate Counts (SPG) and coliform
isolation.  No attempt was made to measure or identify specific microbial
species in water from the carbon beds.

Three different sampling methods were used during the study to collect water
samples for microbiological testing.  One method was the standard technique of
faucet disinfection and flushing the line for at least two minutes prior to
collecting the sample.  The purpose of this method is to sample water in the
central distribution system and which has not been stagnant in the plumbing,
and possibly contaminated by growth on faucets or other plumbing appurten-
ances.  The second method was to collect an unflushed sample from an undisin-
fected faucet, or the first water out of the faucet.  This technique simulated
water which would be consumed if the faucet were not allowed to run before a
glass of water was drawn.  The third technique was to flush one liter of water
from the line prior to collecting the sample from an undisinfected tap.  This
method was used to discard the water in the plumbing between the carbon bed
and the faucet.  Partial flushing would more closely reflect the bacterial
density in the first flush of water through the carbon bed.  It would also
represent water from a faucet allowed to run a few seconds.

Efforts were made in Silverdale to collect the first flow of water through a
device on its sampling day.  Homeowners were notified of the days on which
samples would be collected at their home.  They were asked not to use their
device until after the samples were collected.  Sample collection usually was
started by 6:00 a.m. and completed by noon.  SPG data by sampling location are
presented in Appendix B.
                                    25

-------
Table 13 shows the geometric mean, median, and range of paired samples
collected using the three sampling techniques.  A one liter flush reduces the
mean postdevice standard plate count and brings the bacterial density down to
a level less than 500 organisms per mL, a level often cited as a possible
standard for SPG. Continued flushing reduces the level another order of
magnitude, but does not completely eliminate the microorganisms.
       TABLE 13. BACTERIOLOGICAL SAMPLING METHOD COMPARISON, RESULTS OF
       PAIRED SAMPLES, SILVERDALE POSTDEVICE STANDARD PLATE COUNTS (f/nL)
Sampling                Number
 Method                 Samples
Unflushed,
Undisinfected Tap         143
Flushed,
Disinfected Tap           143
Partial Flush (1 L),
Undisinfected Tap          33
Flushed,
Disinfected Tap            33
                     Geometric
                        Mean


                        1630

                          24


                         220

                          13
                    Median


                     2300


                       16


                      330

                        6
One liter flush samples were collected for the final two months in place of
unflushed samples.  The total number of one liter flush samples was not large;
therefore, most of the discussion of microbial growth will involve unflushed
samples.

Table 14 summarizes microbial results recorded for all samples collected from
unflushed faucets in Silverdale, the primary sampling technique used for the
study.  Mean SPG results from postdevice samples (1380 organisms/mL) were
significantly higher than those from predevice samples (4 organisms/mL),
indicating microbial colonization of the carbon bed.  This is an established
effect of treating water with GAG and has been documented in a number of
studies (14, 15, 16).  Background organisms were present in nontreated
(predevice) faucet water in spite of the maintenance of a mean chlorine
residual of 0.5 mg/L in the distribution system.
         TABLE 14. STANDARD PLATE COUNT SUMMARY FOR UNFLUSHED SAMPLES,
                     SILVERDALE (MARCH 1983 - APRIL 1984)
Location
Predevice
Postdevice
Number
Samples

  187
  297
Geometric
   Mean

      4
   1380
Standard Plate Count
       (#/mL)

       Median        Range

           6         1-1300
        3700        1->5800
                                    26

-------
Monthly means for unflushed plate counts in predevice and postdevice water are
shown in Figure 2.  During the last two months (March and April, 1984),
samples were collected after a partial flush.  All remaining samples were from
unflushed faucets.  Postdevice plate counts increased during the warmer months
(May-October), then returned to previous levels as temperatures decreased.

Mean and median values and ranges of postdevice standard plate counts for each
model used in Silverdale are presented in Table 15.  Data are presented for
flushed and unflushed samples for the time period February through November,
1983.  After November, Model A was replaced, and direct comparisons are not
possible.  Geometric means of SPG densities in unflushed samples were all
reasonably close for the five models; however, median values showed a larger
spread, 1600 to 3250 organisms per mL.

The flushed sample results are quite different.  There is no consistent
pattern to the differences which can be easily related to model design or
quantity of carbon used.  Models C and D do use more carbon than A and E, but
not significantly more than Model F.
               TABLE 15. POSTDEVICE STANDARD PLATE COUNT RESULTS
                FOR FLUSHED AND UNFLUSHED SAMPLES BT MODEL TYPE
    Model

    A Unflushed
      Flushed

    C Unflushed
      Flushed

    D Unflushed
      Flushed

    E Unflushed
      Flushed

    F Unflushed
      Flushed
      Standard Plate Count (#/mL)
Geometric
  Mean         Median         Range
  1672
    16

  1528
    51

  2252
   556

  1515
    12

  2186
     6
2900
   3

1600
  50

2700
 560

3250
   9

2700
   9
35->5800
1-2300

39-5800
3-1300

180-5800
130-4200

1-5800
1-280

6-5800
1-110
 For the period February through November  1983.  Not all samples were paired.


Table 16 shows median SPC densities for each model, as well as grams of carbon
and whether or not the carbon is silver impregnated.  Under the conditions
encountered in this study, silver impregnation does not appear to affect
raicrobial density.  Combined mean and median SPC densities for silver and
non-silver impregnated carbon are presented in Table 17.  There is  not a
significant difference in the microbial densities; however, the SPC densities
in postdevice samples from silver impregnated carbon devices  are higher.
                                    27

-------
STANDARD PLATE COUNT (if/ml)





Figure
counts

a- s»
^ •

S i?
0 Q
rt 1
0- ft
• ft
a
p**
rt

M
I
i
o
H,

M
H

ft*
It
0.
P.
A
TJ
8
1
•5
5
«
rt
B.
3
8
to
ft
1
p>
3,
•a

ft
n





CO
1
»-» 3 -<
ca

n
o > —
H1
CD
o
£ K -1
CD
a.
Ml
H ,_. „
0 ^ ^
8
S CH-
C
O
ex.
0 > -
H-
0
Ml
(D
O to-
rt
n>
CL
rt o -
(0
(U
ft *-
n
H

93 0.
O
9
n
H1 c_i •
H-
rt
(D
^
Ml
M
(^
CO
f 3 •<
•*

> -I

M 10 O O
o o o o
O O O 0
o o
1 1 1 1
1 •



X •



1 •



^

* •



x •




x •


, 9

t •



H*
m
1 • >
§
. t
o
^ • «g
(D
^»
1 • 5"
IB
A
3
o
































K •
TJ >T»
H 0
SB
< n
H- 
-------
All of these observations relate to the samples collected  in  Silverdale.
Microbial populations may vary significantly with changes  in  chemical  quality
of the water supply.
  TABLE 16. MEDIAN POSTDEVICE STANDARD PLATE COUNTS BT MODEL AND CARBON HEIGHT
                             (Unflushed Samples Only)
    Model

      A
      C
      D
      E
      F
 Weight of
Carbon (gms)

     417
    1700
    1150
     300
     765
  Silver
Impregnated

    Yes
     No
    Yes
     No
     No
Median Postdevice
   SPG WmL)

        2900
        1600
        2700
        3250
        2700
    TABLE 17. POSTDEVICE STANDARD PLATE COUNTS FOR MODELS WITH AND WITHOUT
                           SILVER IMPREGNATED CARBON
                           (Unflushed Samples Only)
                                  Standard Plate Count  (#/mL)
                        Geometric Mean           Median
With Silver
Without Silver
          1960
          1740
           2800
           2500
        35->5800
        1-5800
Median SPG densities for unflushed samples versus grams of carbon are
presented in Figure 3.  The data correlate well, with a slope of -1.008 and a
correlation coefficient of 0.933.

The correlation of median SPC densities with grams of carbon is an empirical
observation.  All five models have different design features which could
affect the postdevice SPC densities.  Although efforts were made to control
sampling by collecting the first flow of water through a device on its sampl-
ing day, it could not always be verified.  To adequately study any correlation
between SPC densities and carbon quantity would require an experiment designed
for the purpose.  If, in fact, postdevice bacterial densities could be reduced
by using more carbon, then that could be a good design parameter for point-of-
use treatment devices.  However, further study is needed.

Standard Plate Counts are typically performed using Tryptone Glucose Yeast
Agar, most often referred to as Standard Methods Agar (SMA), and a 48 hour
incubation period at 35°C.  It has been reported that some bacteria frequently
found in potable water will not grow on this media (17).  As a result, a media
used to enumerate bacteria in natural oligotrophic water (R2A) was adapted for
use in potable water.  The populations enumerated by R2A media with SMA media
were compared by parallel testing with the two media.  Unflushed postdevice
samples were collected once a day, five days a week for a 30 day period.
Samples were collected from four devices (two silver and 2 non-silver), split,
and plated on SMA and R2A media.  Plates were incubated at 35°C and were
counted at 48 and 72 hours.  A summary of the results appears in Appendix D.
                                    29

-------
                     4000
                                                 Slope = -1,008

                                                 R = 0.933
OJ
o
                  to

                  CO
                  •H
                  a
                  
-------
A comparison of R2A and SMA is shown in Figure 4.  The R2A and SMA media
yielded comparable results indicating that SPG organisms sampled during the
study were predominantly oligotrophic.  This would be expected for groundwater
like Silverdale's, with low nutrient and substrate concentrations.  Positive
coliform results were obtained from 27 of 297 postdevice samples collected
from unflushed, undisinfected taps, and from 4 of 176 postdevice samples
collected from flushed, disinfected taps.  Resamples collected at the four
sites yielding positive results were collected from flushed, disinfected taps.
Resamples tested negative for coliform organisms.  The occurrence of coliforms
could not be correlated with standard plate count densities.  The data collec-
ted in Silverdale do not indicate that coliform bacteria colonize GAG devices.
It is possible that coliform contamination of the water is occurring after
passing through the carbon.

Several investigators have identified organisms in carbon beds and in effluent
water (19, 20).  The most frequently reported genus is Pseudomonas.  Efforts
were made to isolate this genus in January 1983 when eight postdevice samples,
two from each model, were collected from unflushed, undisinfected taps.
Samples were analyzed for total Pseudomonas and Pseudomonas aeruginosa. No
organisms of this genus were detected.

Rockaway Township, New Jersey

Standard Plate Counts and coliform analysis were performed on predevice and
postdevice water samples collected at ten sites in Rockaway.  Table 18
presents a summary of SPC data collected from unflushed, undisinfected taps.
Mean plate counts were higher in postdevice samples than in predevice samples.
Although the difference is significant (tested at 95 percent confidence level
using student "t" distribution), it is not as pronounced as that found in
Silverdale.  The mean postdevice SPC was an order of magnitude lower than that
reported for Silverdale.  This is somewhat of a surprise, as chlorination is
not practiced on private wells in Rockaway, and the predevice SPC mean is two
orders of magnitude greater than that observed in Silverdale.  Several
possible explanations exist for this, including microbial floras composed of
organisms sufficiently different in their ability to colonize the carbon or
compete for nutrients, or inhibition or injury to the organisms present caused
by exposure to chemical constituents found in the water or in material
concentrated by the carbon.  It is also interesting that the point-of-use
devices used in Rockaway contained silver impregnated carbon.  Silver may be
more effective on the organisms native to Rockaway, but comparable data for
non-silvered models in Rockaway are not available.

Fourteen postdevice samples were collected from both unflushed, undisinfected
taps and from flushed, disinfected taps and analyzed by SPC.  Data are summa-
rized in Table 19.  Postdevice samples collected from flushed, disinfected taps
had microbial densities one order of magnitude lower than unflushed samples.

No coliform organisms were detected in any postdevice samples collected in
Rockaway.
                                    31

-------
  3000
              • R2 Agar



              X "Standard Methods" Agar
o
  2000
m
rH
IX
•a


5 1000

CO
1
                              X  *
X


i
                                                                         1      1
                                                                                     x
                                                                                     1
                              8     10     12
                              14



                            Day
                                                     16     18    20    22     24     26     28     30
                Figure  4.  Comparison  of standard plate count agar, unflushed postdevice

                samples, March-April,  1984.

-------
             TABLE  18. ROCKAWAY STANDARD PLATE COUNT DATA SUMMARY
                              (Paired Samples Only)
 Location

Predevice
Postdevice
 Number
Samples

   23
   23
       Standard Plate Count
              (#/mL)
Geometric
   Mean          Median
   130
   210
155
160
40-54,000
40-32,000
           TABLE 19.  ROCKAWAY SPG DATA COMPARING SAMPLING TECHNIQUES
                             (Paired Samples Only)
      Technique

Unflushed/Undisinfected
Flushed/Disinfected
        Number
       Samples

          14
          14
       Standard Plate Count
              (#/mL)
     Geometric
        Mean         Median
         250
          35
    350
   <10-3,2000
   <10-3600
COSTS

Silverdale, Pennsylvania

Equipment Purchase and Installation

Table 20 shows a detailed cost breakdown for the purchase and installation of
the five models installed in Silverdale.  Costs for system components reflect
manufacturers' discounted retail prices for purchases in lots of ten or more
for January 1983.  Data are presented in terms of a basic GAG line bypass unit
with sink mounted taps.  The system includes the housing and cartridge(s),
tap, water meter1, and all connective fittings.  The Seidel Water Company was
subcontracted at $20.00/hour to perform installation and maintenance.

Installation of 47 devices was completed in an 18 day period.  An additional
two devices were requested at a later date and installed within six days.
Installation of the devices, taps, and water meters took from 1.5 to 3.25
hours per household, depending on the model, type of sink, and installation
problems encountered.  The 3.25 hour installation time was probably longer
than most similar devices because of problems with improperly sized parts for
one model.   Average times for installation of the various models, neglecting
problems caused by improperly sized parts, ranged from 1.5 to 2.0 hours, with
an overall  average time for all models of 1.7 hours.
1
 Water meters are not required by the manufacturer, but are important for
 monitoring device life and performance.
                                    33

-------
Item Cost
Device and Tap1
Water Meter
Connective Fittings
Total Equipment Costs
Installation ($20/hr)
Total Capital Costs
A
$ 227
38
5
270
34
304
$ 170
38
22
230
40
270
$ 225
38
5
268
32
300
$ 232
38
5
275
30
305
F_
$ 192
38
5
235
31
266
      TABLE 20. CAPITAL COSTS FOR POIHT-OF-USE DEVICES USED IN SILVERDALE

                                     Model	
                                                                Average Cost

                                                                   $ 209
                                                                      38

                                                                       8

                                                                     256

                                                                      33

                                                                     289


 Device and tap costs include point-of-use device, sink mounted tap and
 shipping costs, based on a purchase of 10 or more units.
The average capital cost of a GAC installation in Silverdale was $289.
Differences in total capital costs for the five models are not very large,
with costs ranging from $266 to $305.  Installation costs were fairly
consistent.  If citizens paid for capital cost and installation over a 20 year
service life, amortization of the average capital cost at ten percent for 20
years would cost $2.83 per customer, per month.

Operation and Maintenance

Average repair costs incurred over one year of service for the devices used in
Silverdale ranged from $1.00 to $35.00 per installation.  The average annual
repair cost for the 49 devices was $17.14 per unit or $1.43 per device per
month.  Repair costs are based only on labor costs of $20.00 per hour.  All
materials needed for repairs were provided by the manufacturers free of
charge.

Repairs performed on devices in Silverdale were more frequent than expected.

The overall maintenance frequency was biased to the high side by the number of
different models used.  All models used in Silverdale required some attention
by maintenance personnel.  Problems encountered included leaking, intermittent
flow through the device, and contaminated carbon.

Three of the five models used in the study had some leaking problems.  For one
model, glued joints of the PVC housing leaked in six of ten installations.
Leaks developed over an eight month period.  The manufacturer reported that
leaking was the result of production quality control problems and replaced
several of the devices in the first month of the study.  Subsequent leaking in
one of the replacements and in several of the original devices eventually led
to the replacement of this model.  The replacement model was one in use in
Silverdale at the time, and was selected by the Seidel Water Company based on
low maintenance requirements.  Thus, at the conclusion of the study the
community had only four different models.
                                    34

-------
A second model had intermittent leaks at tubing-cartridge connections.  Leaks
developed on at least one occasion in all devices of this model installed in
Silverdale.  Connective fittings were replaced by the manufacturer's
representatives, but this was only a temporary solution; several devices
continued to develop leaks.  Leaking caused damages in two homes of $250 and
$300.  Damages were covered by the manufacturer's liability insurance, but
reimbursement took several months.  The cause of the leaks was eventually
determined.  The manufacturer replaced all units in service.  No further
leakage occurred.  GAG cartridges from the devices which were replaced were
installed in the new device to provide cartridge continuity for the study.

A single unit of a third model developed a small leak immediately after
installation.  A return visit by maintenance personnel revealed the leak had
stopped without intervention.  No further maintenance was required.

Several homeowners reported that the devices in their homes would not always
provide water on demand.  Investigation revealed that all these devices were
of one GAG model, and that the intermittent flow was common to all the devices
of that model installed in Silverdale. Discussions with the manufacturer
resulted in identifying the problem as slightly oversized carbon cartridges.
The oversized dimensions were attrituted to supplier error and led to the
replacement of the cartridges.  This corrected the problem, and no further
reports of intermittent flow were received.

Operational problems observed with another model used in Silverdale were
rather surprising.  Samples collected from some installations of this device
contained methylene chloride at levels up to 316 yg/L.  Investigation into VOC
sample collection, handling, and storage showed no source of this
contamination.  Repeated sampling, however, continued to reveal methylene
chloride at high concentrations in postdevice water, while predevice water
collected at the same locations showed no detectable levels.  Subsequent
investigation revealed that a methylene chloride spill had occurred in a
storage area at the manufacturer's warehouse, and methylene chloride was
adsorbed by nearby GAG cartridges.  The cartridges in use were replaced with
new cartridges in May 1984.

Before the replacement cartridges were used five were selected at random for a
brief immersion test to determine whether or not they were contaminated.  The
cartridges were unpacked, and placed in organic free water (Milli-Q) contained
in a jar which was sealed with no headspace.  Exposure times of up to 144
hours showed methylene chloride at a minimal background level.  When contam-
inated cartridges were tested this way, high concentrations of methylene
chloride were fo'und.  The remainder of the new GAG cartridges were then used
to replace those in use.  Replacements were installed in May 1984, and did not
affect VOC removal data.

Anticipated routine cartridge replacement frequencies in Silverdale were esti-
mated using each manufacturer's rated capacity to predict replacement costs.
An average measured water use rate of 1.0 gpd was used.  Replacement frequency
estimates ranged from 2 to 5 years.  Yearly replacement was assumed for par-
ticulate prefliters.  Replacement costs for each model were multiplied by the
number in service and divided by the estimated years of service.  The sum of
the cartridge replacement costs per year was divided by 49 (total devices in

                                    35

-------
Silverdale).  This gave an average replacement cost of $20.61 per year per
device, or $1.72 per month per device.  Model specific costs ranged from $0.48
to $3.11 per month.  GAG cartridge replacement costs are based on the
manufacturer's discounted retail prices for single purchases for January 1983.

Average monthly customer costs are summarized in Table 21.  The average total
monthly customer cost for a pointr-of-use device in Silverdale was $5.98.
      TABLE 21. ESTIMATED AVERAGE MONTHLY CUSTOMER COSTS - SILVERDALE, PA

Capital Cost (GAG equipment, connective
  fittings, and installation labor)1                          $2.83

Operation Cost (Including maintenance
  labor)                                                       1.43

Replacement Cartridge Cost                                     1.72

Total customer cost                                           $5.98
 Amortized at 10% for 20 years
Rockaway Township

Equipment Purchase and Installation

Equipment for point-of-use GAG treatment in Rockaway Township did not include
a water meter.  The model was equipped with a shutoff flow meter.  An average
installation cost of $30.00 (1.5 hours labor) was assumed based on installa-
tion costs for the same model in Silverdale.  Total capital costs for Rockaway
Township, including GAG equipment, connective fittings, and installation
labor, were $255 per installation.  The capital cost of $225 for equipment in
Rockaway was negotiated by the community.

Operation and Maintenance

In Rockaway, GAG cartridge life was assumed to be 2.4 years based on the
manufacturer's rated treatment capacity of 2000 gallons and an estimated use
rate of 2.3 gallons per day.  The use rate was estimated by flow shutoff
readings taken at the time of postdevice sample collection.  Total estimated
replacement costs were $1.77 per month per device based on these assumptions.

No maintenance was reported during the two year demonstration period.

Average monthly costs including cartridge replacement and maintenance labor
are presented in Table 22.  Average monthly customer costs are estimated to be
$4.23.
                                    36

-------
  TABLE 22. ESTIMATED AVERAGE MONTHLY CUSTOMER COSTS - ROCKAWAY TOWNSHIP, NJ

Capital Cost (GAG equipment, connective fittings,
  and installation labor)1

Cartridge Replacement Cost2

Total Customer Cost
 Amortized at 10% for 20 years
2
 No maintenance costs were reported during the two year demonstration period.
Monitoring Costs

Costs for sampling and microbiological analyses were $20 per sample in
Silverdale.  The same laboratory provided VOC analyses on request at $50 per
sample.  These costs will vary depending on laboratory capability, proximity,
and whether sampling is subcontracted or provided by the community. The
Silverdale costs are typical of fees charged by commercial laboratories.

An additional variable which affects monitoring costs is the sampling
frequency chosen by the community.  Unlike relatively inexpensive inorganic
analyses, VOC analyses generally will cost more than replacement cartridges.
Consequently, it could be more cost effective to replace cartridges before
they become exhausted than to try to fully use cartridge capacity by closely
monitoring for breakthrough.

It is recommended that communities selecting point-of-use treatment conduct a
pilot study by operating a device on the community water supply at continuous
flow until breakthrough occurs.  This pilot study will establish the device's
capacity for the quality of water, and could be completed in less than three
days for most devices.  Raw water quality should be monitored during normal
operation to assure that the quality has not changed and that the pilot study
results are still valid.

Water Quality District Administration Costs

The costs presented thus far would be costs incurred by homeowners, regardless
of whether they particpated in a water quality district program or not.  For
private homeowners, the only costs that would differ would be capital
amortization, and discounts provided by quantity purchases.  Private
homeowners could decrease one time capital and installation costs through
normal private financing routes.

If a community chooses to establish a water quality district, routine
administrative costs will be incurred.  These administrative costs could
include record keeping, billing, and parts inventory control.
                                    37

-------
Thunderbird Farms, Arizona has had a water quality district (for fluoride
removal) established for over four years (21).  Their records indicate that
200 hours per month are required to maintain 1500 records for a total of 643
customers (some customers have more than one record).  That amounts to 0.133
hours per month per record.  Assuming a district would operate on a quarterly
billing basis, 0.40 hours per quarter per record is required for maintenance.
Assuming $8.00 per hour (including fringe) for record keeping, the cost per
customer is $3.20 per quarter.

Telephone, postage, and miscellaneous supplies for the 643 customers are
$1,275 per year, or $0.495 per quarter, or $0.17 per customer per month.

Therefore, total administrative costs are $3.70 per quarter, or $1.23 per
month.  Labor rates may vary.  Costs could be reduced through voluntary labor
and/or more active homeowner participation.
                                    38

-------
                                  SECTIOH 5.

                           CURRENT MANAGEMENT STATUS
SILVERDALE, PENNSYLVANIA

All point-of-use devices installed in Silverdale were still in service as of
August 1984 (18 months).  Participating homeowners were notified of their
responsibility for service and for the associated costs of any maintenance and
cartridge replacements.  Homeowners were provided with conservative estimates
of projected cartridge life and the meter readings at which the current carbon
cartridges should be replaced.  Addresses and phone numbers of both local
distributors and manufacturers' headquarters, along with quoted prices for
replacement cartridges, were also provided.

By replacing cartridges at a conservatively estimated volume treated,
expensive sampling and monitoring will not be necessary, provided the
concentration of contaminants in the raw water does not change significantly.
Monitoring requirements may change if the VOCs become regulated under the Safe
Drinking Water Act.

ROCKAWAY TOWNSHIP, NEW JERSEY

The field verification study of point-of-use efficacy conducted in Rockaway
Township is completed.  Under the formal agreement between the township and
the manufacturer, participants in the study now have the option to either
purchase the devices at a reduced rate or have them removed at the
manufacturer's expense.  Should the homeowners choose to keep them, they will
have responsibility for maintenance and carbon cartridge replacement.  It is
assumed that those individuals wishing to keep the devices will make
arrangements for service with the manufacturer.  The township cannot require a
private well owner to install a point-of-use device.

The Rockaway Township Department of Health and Welfare is in the process of
developing recommendations for using GAG point-of-use treatment based on the
results of the verification study.  These are to include conditions for use
(not all VOCs are effectively removed by GAG), lists of products considered to
qualify for use (models used in the GSRI study or from manufacturers providing
the results of an independent evaluation), and monitoring considerations.
Health officials prefer models with automatic shut-off valves but will not
require that automatic shutoffs be part of a GAG installation.  Monitoring
recommendations will most likely include VOC and bacteriological sampling and
analysis on an annual basis.  Health officials are also considering picking up
the cost of annual VOC sampling of those devices treating water from the most
contaminated wells (22).
                                    39

-------
                                   REFERENCES
 1.   "Advanced  Notice  of  Proposed  Rule  Making for Phase I VOCs" USEPA,
     47 FR  9350, March 14,  1982.

 2.   Lehr,  J.H., Gass, I.E.,  Pettyjohn,  W.A.  and J.  DeMarre,  Domestic Water
     Treatment, McGraw-Hill,  Inc.,  1980.

 3.   "National  Primary Drinking Water Regulations; Volatile Synthetic Organic
     Chemicals; Proposed  Rulemaking" USEPA,  49 FR 24330, June 12,  1984.

 4.   Perry,  D.L.,  Smith,  J.K.  and  S.C.  Lynch, Study of Home Drinking Water
     Treatment  Units  Containing Activated Carbon for Organics Reduction,
     Interim Phase 3  Report,  USEPA, Washington, DC,  1982.

 5.   McKinnon,  Ronald  J., and John E. Dyksen, Removing Organics From
     Groundwater Through  Aeration  Plus  GAG.   Jour. AWWA 76:5:42 (May 1984).

 6.   Gill,  H.E., and  Vecchioli,  1965, Availability of Ground Water in Morris
     County, New Jersey,  U.S.  Geological Survey Special Report No. 25.

 7,   Geraghty and  Miller, Inc.,  Investigation of Contamination By Volatile
     Organic Compounds in the Lake Telemark  Subdivision of Rockaway Township,
     New  Jersey.   Syosset,  New York (1981).

 8.   Interim Treatment Guide  for  Controlling Organic Contaminants in Drinking
     Water  Using Granular Activated Carbon.   J.M. Symons, Ed. USEPA
     Cincinnati, OH  1978.

 9.   Weber,  W.J.,  Jr., Physicochemical  Processes for Water Quality Control,
     John Wiley &  Sons, Inc., 1972.

10.   Dobbs,  R.A. and  J.M. Cohen,  Carbon Adsorption Isotherms for Toxic
     Organics,  USEPA-600/8-80-023, Cincinnati, OH  1980.

11.   Bennett, D.L., Sample Collector's  Handbook, Illinois EPA, Springfield,
     IL.  March,  1982.

12.   Remington, R.D.  and  M.A. Schork,  Statistics With Applications to  the
     Biological and Health Sciences,  Prentice-Hall, Inc., Englewood Cliffs,
     NJ,  1970.

13.   Thomas, Lee,  Bucks County Health  Department, Doylestown, PA, personal
     communication, 1983.
                                    40

-------
14.  Ford, D.B., The Use of Granular Carbon Filtration For Taste and  Odor
     Control, In: Papers and Proceedings of a Water Research Association
     Conference at the University of Reading, United Kingdom, Paper  12
     (February 1974).

15.  Schalekamp, M., Use of Activated Carbon in the Treatment of Lake Water,
     In Translations of Reports on Special Problems of Water Technology, Vol.
     9-Adsorption, H. Sontheimer, Ed. EPA 600/9-76-030.

16.  Weber, Walter J.; Pirbazari, Massoud, and Melson, Gail L., Biological
     Growth on Activated Carbon: An Investigation by Scanning Electron
     Microscopy.  Envir. Sci. and Tech., 12:7:817 (July 1978).

17.  Reasoner, D.J. and Geldreich, E.E., A New Medium for Enumeration and
     Subculture of Bacteria From Potable Water, Presented at the 79th Annual
     Meeting of the American Society For Microbiology, May 4-8, 1979, Los
     Angeles, CA, Paper No. 47.

18.  Klotz, M.; Werner, P.; Schweisfurth, R., Investigations Concerning the
     Microbiology of Activated Carbon Filters, In Translations of Reports on
     Special Problems of Water Technology, Vol. 9 Adsorption, H. Sontheimer,
     Ed. EPA 600/9-76-030.

19.  Werner, P.; Klotz, M.; and Schweisfurth R., Investigations Concerning the
     Microbiology of GAG Filtration For Drinking Water Treatment.  Presented
     at the USEPA/NATO CCMS Symposium on Practical Application of Adsorption
     Techniques in Drinking Water, Reston, VA. (May 1979).

20.  Latoszek, A.; and A. Benedek, Some Aspects of Microbiology of Activated
     Carbon Colummns Treating Domestic Wastewater, Water Research Group
     Report, McMaster University. (1974).

21.  Bellen, Gordon E., and Marc Anderson. Defluoridation of Drinking Water in
     Small Communities:  Final Report.  USEPA, Cincinnati, (1985).

22.  Levinson, Steven, Rockaway Township Department of Health and Welfare,
     Rockaway, NJ, Personal Communication, 1984.
                                    41

-------
                                  APPENDIX A

             EFFECTS OF HEADSPACE AND HOLDING TIME ON VOC RECOVERY


Effect of Saaple Headspace and Holding Time on VOC Analytical Results

Most of the samples collected for VOC analysis were shipped to the NSF
Laboratory in Ann Arbor, Michigan for analysis.  Although sampling and
shipping procedures were in accordance with prescribed techniques, 17 percent
(134 samples) of the 788 samples collected had a small but perceptible
headspace when they arrived at the lab.  The volume of the headspace ranged
from zero to 1.0 mL.

Samples were collected in duplicate, so at least one of the pair usually was
free of headspace.  However, a study of the impact of headspace and holding
time was conducted to determine what effect sample headspace, and headspace
combined with holding time, might have on analytical results.

Tetrachloroethylene (PCE), trichloroethylene (TCE), and 1,1,1-trichloroethane
were selected as study chemicals because they were the primary chemicals of
interest in both Rockaway and Silverdale.  Spike concentrations of
approximately 20 yg/L, 50 yg/L, and 50 yg/L, respectively, were added to
Milli-Q water (pH 5.3) individually and as a mixture of all three.

Forty mL glass vials were filled to overflowing with the spiked water.  A
teflon septum was placed on the vial.  The septum was removed and a volume of
water was removed from the vial.  The volumes removed were 1.0 mL, 0.5 mL,
0.25 mL, and 0.0 mL.  The septa were replaced, and the vials capped and
sealed.  Vials were stored at 4°C for 1 hour, 1 day, 3 days, 7 days, 10 days,
and 14 days.  Duplicate vials for each headspace volume and time period were
analyzed.  The study was repeated one time, making a total of 384 analyses.

Mean percent recoveries were calculated based on the original sample
concentrations listed in Table A-l.  Figure A-l shows mean percent recoveries
versus headspace volume for each chemical from vials spiked with one chemical.
Figure A-2 shows percent recovery versus headspace for vials containing all
three chemicals (multiple spike).  Figures A-3 and A-4 show single and
multiple spike recoveries versus holding time.  The dotted lines indicate
100 +_ 10 percent recovery.

It is apparent that as headspace volume increases, there was some decrease in
percent recovery.  With the exception of the single spike for PCE, recoveries
remained within the +_ 10 percent range for headspaces of up to 0.75 mL.  In
all cases, percent recovery remained within a range of 100 +_ 10 percent for up
to 0.5 mLs of headspace.


                                    42

-------
£
a
&
s
o
      120
      100.
       80
                                                             TCE

                                                             1,1,1-Trichloroethane

                                                             PCE
       60
                     I
                                 I
                                                         I
                    0.0
                                                                    1.0
                     0.25        0.50        0.75

                          Headspace Volume (mL)

Figure A-l.  Percent recovery versus headspace volume for single spike  samples.
      120-
     100
      80
      60
                                                                       TCE
                                                                       PCE
                                                                       1,1,1-Trichloroethane
                                _L
                                              _L
                    0.0
                     0.25        0.50        0.75

                         Headspace Volume  (mL)
                                                                    1,0
           Figure A-2. Percent recovery versus headspace volume for multiple spike
           samples.
                                          43

-------
                          Percent  Recovery
                                                                Percent Recovery
 00


~?~
1-1
re
o
o

">      I—
Hi 03

2  ^
>-t  to

B  '"'
T3

h-1
(I)
I
•o
h-1
ro
ED
                                                 o
                                                ff-
                                               H T<
                                               O O ET-
                                               H W£~



                                                    O  I
                                                    fl> H
                                                    rt M
T)
H-
e
it
n>

^
OJ
                                       'tl
                                       re
                                       H
                                       o
                                       it
                                       a
                                       n

                                       H
                                       n
                                       o
                                       8
                                       «<


                                       3
                                       H
                                       CD

                                       £rt

                                       gr
                                       H n>
                                       m« °°

                                       S^
                                         tn
                                       oo **~*

                                       5-
                                       TO
                                                                         g-
                                       1
                                       M
                                       n
                                       D>
 o

T"
                                                                                                            00

                                                                                                            O
                                                                                         O
                                                                                         o
                                                                 ro
                                                                 o

-------
    TABLE A-l.  CONCEBITRATIONS OF THE THREE VOCs IN MULTIPLE SPIKE SAMPLES
                           USED FOR HEADSPACE STUDY
               Repetition I
                 1,1,1-Trichloroethane               46.8 yg/L
                 TCE                                 32.8 yg/L
                 PCE                                 24.4 yg/L
                 Mixture                             104.0 yg/L

               Repetition II
                 1,1,1-Trichloroethane               41.8 y g/L
                 TCE                                 38,8 yg/L
                 PCE                                 32.5 yg/L
                 Mixture                             113.1 yg/L
There was no loss of recovery for holding times up to 14 days.  There even
appeared to be some improvement in percent recovery versus time; however, the
increase is not significant.

A two factor analysis of variance (ANOVA) of the headspace study data appears
in Tables A-2 through A-7.  ANOVA was used to analyze the variation in percent
recovery attributable to headspace volume, holding time, and the interaction
between headspace volume and holding time.  A random model was used.
Therefore, even though the headspace volumes and holding times were fixed for
this study, the data can be used to infer the impact of randomly occurring
headspace/holding time interactions within their respective ranges.

If the computed F value (F ) is greater than the tabular F value (Ft>,
there is a significant effect.  F values in Tables A-2 through A-7 show no
significant effect caused by holding time.  Results for the single spike PCE
test (Table A-4) contradict the other results, but incomplete data for a
portion of the single PCE spike test biased the ANOVA.  It is clear from
Figure A-3 that there is little change in PCE recovery from a single spike as
a function of time.  Consequently, the ANOVA data in Table A-3 will not be
used in this discussion.
                                    45

-------
TABLE A-2.  ANALYSIS OF VARIANCE RESULTS FOR 1,1,1-TRICHLOROETHANE
                  RECOVERED FROM A 50 Ug/L SPIKE


          Large Sum   Degrees of  Large Mean       F 1           F
                                                                          2
Source
Headspace
Time
Headspace-Time
Interaction
Within
Cells3
Total
of Squares
53,784
171,575
202,205
305,338
732,903
F = computed F value.
2
F = F value from table.
3
Cells are subgroups of data
holding times.
TABLE
Source
Headspace
Time
Headspace-Time
Interaction
Within
Cells3
Total
Freedom
3
5
15
24
46
corresponding
Square (Random Model) (0.
17,928 1.33 3.
34,315 2.55 2.
13,480 1.06 2.
12,722
15,933
to specific headspace volumes
95)
29
90
11
-
and
A-3. ANAYLSIS OF VARIANCE RESULTS FOR TRICHLOROETHYLENE
RECOVERED FROM A 50 yg/L SPIKE
Large Sum
of Squares
56,243
8,729
11,611
444,328
523,474
Degrees of Large Mean F l F
Freedom Square (Random Model) (0.
3
5
15
24
47
18,748 24.22 3.
1,746 2.26 2.
774 0.04 2.
18,597
11,138
2
95)
29
90
11
-
-
 F  = computed F value.


2p  = F value from table.

3
 Cells are subgroups of data corresponding to specific headspace volumes and
 holding times.
                                    46

-------
        TABLE A-4.   ANALYSIS OF VARIANCE RESULTS FOR TETRACHLOROETHYLENE

                         RECOVERED FROM A 20 yg/L SPIKE
Large Sum Degrees of Large Mean Fc 1 F
Source of Squares Freedom Square (Random Model) (0.
Headspace 483,459
Time 713,441
Headspace-Time 420,819
Interaction
Within -1,309,749
Cells3
Total 307,970
F = computed F value.
c
2
F = F value from table.
3
Cells are subgroups of data
holding times.
3 161,153 5.74 3.
5 142,688 5.09 2.
15 28,055 0.51 2.
24 -54,573
36 7,897
corresponding to specific headspace volumes
2
95)
29
90
11
-
and
TABLE A-5. ANALYSIS OF VARIANCE RESULTS FOR 1 , 1 , 1-TRICHLOROETHANE
RECOVERED FROM 20/50/50 yg/L MIXTURE OF
TETRACHLOROETHYLENE, 1 , 1 , 1-TRICHLOROETHANE , AND TRICHLOROETHYLENE
Large Sum
Source of Squares
Headspace 178,404
Time 63,476
Headspace-Time 75,396
Interaction
Within 51,840
Cells
Total 369,116
Degrees of Large Mean ₯cl F
Freedom Square (Random Model) (0.
3 59,468 11.83 3.
5 12,695 2.53 2.
15 5,026 2.33 2.
24 2,160
47 7,853
2
95)
29
90
11
-
-
 F  = computed F value.



2p  = F value from table.
  t

3
 Cells are subgroups of data corresponding to specific headspace volumes and

 holding times.
                                    47

-------
         TABLE A-6.   ANALYSIS OF VARIANCE RESULTS FOR TRICHLOROETHYLENE
                   RECOVERED FROM A 20/50/50 yg/L MIXTURE OF
       TETRACHLOROETHYLENE, 1,1,1-TRICHLOROETHANE, AND TRICHLOROETHYLENE

                Large Sum   Degrees of  Large Mean       Frl            F 2
Source of Squares
Headspace 89,683
Time 18,435
Headspace-Time 32,513
Interaction
Within 73,080
Cells3
Total 213,711
F = computed F value.
2
F = F value from table.
3
Cells are subgroups of data
holding times.
Freedom Square (Random Model) (0.
3 29,894 13.79 3.
5 3,687 1.70 2.
15 2,168 0.71 2.
24 3,045
47 4,547
corresponding to specific headspace volumes
95)
01
62
11
-
and
TABLE A-7. ANALYSIS OF VARIANCE RESULTS FOR TETRACHLOROETHYLENE
RECOVERED FROM A 20/50/50 yg/L MIXTURE OF
TETRACHLOROETHYLENE, 1,1,1-TRICHLOROETHANE, AND TRICHLOROETHYLENE
Large Sum
Source of Squares
Headspace 197,331
Time 24,413
Headspace-Time 309,687
Interaction
Within 249,791
Cells3
Total 781,222
Degrees of Large Mean FCI F
Freedom Square (Random Model) (0.
3 65,777 3.19 3.
5 4,883 0.24 2.
15 20,646 1.98 2.
24 10,408
46 16,983
2
29
90
11
-
-
 F  = computed F value.
2
 F  = F value from table.
3
 Cells are subgroups of  data corresponding to specific headspace volumes and
 holding times.

                                    48

-------
Three of the five remaining tables show a significant headspace effect, but
none show a holding time effect.  There is not apparent headspace/holding time
interaction.  The fact that there is no trend of significant headspace/time
interaction means that the effect of headspace on recovery is independent of
holding time.

Figure A-l and A-2 indicate that the headspace effect may be more pronounced
for the multiple spike.  The multiple spike best reflects the composition of
raw water samples for this study.  Consequently, it was determined that the
headspace effect was only significant in raw water samples with headspace
volumes greater than 0.5 mLs.  Data from raw water samples with greater than
0.5 mLs headspace were not used in this report.
                                    49

-------
>
o
o
0)
OS
      120 -
      100-
       80
                                                                       TCE

                                                                       1,1,1-Trichloroethane

                                                                       PCE
       60
                     I
                                 I
                                                                      I
                    0.0         0.25         0.50         0.75          1.0
                                    Headspace Volume  (mL)
           Figure A-l.  Percent  recovery versus headspace volume for  single spike samples.
o
3
&
 Hi
PH
      12C-
      100
       80
       60
                                                                       TCE
                                                                       PCE
                                                                       1,1,1-Trichloroethane
                                                                      I
                    0.0
                                                                    1.0
                                0.25        0.50        0.75
                                    Headspace Volume (mL)

           Figure A-2.  Percent recovery versus headspace volume  for  multiple  spike
           samples.
                                           50

-------
    120
   100
S
o
M
01
CL,
    80
     60
  1,1,1-Tri-

  chloroethane

   TCE
   PCE
                                                                       I
                 2         4         6         8        10        12

                                         Time  (Days)

      Figure A-3.  Percent recovery versus time for single spike samples.
14
    120 -
     100-
      80-
01
u
u
      60-
  1,1,1-Tri-
   hloroethane

  PCE

  TCE
                 2          5        6         8        10        12^       14

                                         Time (Bays)

      Figure A-4. Percent  recovery versus  time  for multiple  spike  samples.
                                         51

-------
                              APPENDIX B.

                    VOC AND MICROBIOLOGICAL RESULTS
                                  FOR
              POINT-OF-DSE EFFLUENT SAMPLES - SILVERDALE
                         SILVERDALE VOC RESULTS

                          Concentration (yg/L)

                                Model A
LOCATION
/Flow
8/21
26
104
209
13/8
11
42
82
19/28
42
76
117
26/34
126
233
274
35/28
71
304
408
440
46/
38
86
109
52
66
91
93
55
36
310
64
36
104
67165
422
DATE
3/14/83
3/21/83
6/13/83
11/7/83
3/14/83
3/21/83
6/13/83
9/12/83
3/14/83
4/18/83
7/11/83
10/10/83
3/7/83
5/16/83
8/15/83
10/17/83
3/14/83
4/18/83
8/15/83
10/17/83
11/7/83

4/18/83
8/15/83
10/17/83

6/13/83
8/15/83
10/17/83

3/21/83
10/10/83

5/16/83
9/12/83
5/16/83
9/12/83
TRI-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND

ND
ND
ND

ND
ND

ND
ND
ND
ND
TETRA-
CHLORO-
ETHYLENE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND

ND
ND
ND

ND
ND

ND
ND
ND
ND
1,1,1-TRI-
CHLORO-
ETHANE
-'
-
ND
ND
-
-
ND
ND
-
-
-
ND
-
-
ND
ND
-
-
ND
ND
ND

-
ND
ND

ND
ND
ND

-
ND

-
ND
-
ND
1 ,2-DI-
CHLORO-
ETHANE
-
-
-
ND
-
-
-
ND
-
-
-
ND
-
-
-
ND
-
-
-
ND
ND

-
-
ND

-
-
ND

-
ND

-
ND
-
ND
CARBON
TETRA-
CHLORIDE
-
-
-
ND
-
-
-
ND
-
-
-
ND
-
-
ND
ND
-
-
ND
ND
ND

-
ND
ND

-
ND
ND

-
ND

-
ND
-
ND
CHLORO-
FORM
-
-
ND
ND
-
-
ND
1
-
-
""NO
ND
-
-
ND
ND
-
-
ND
ND
ND

-
ND
ND

ND
ND
ND

-
ND

-
ND
-
ND
BROMODI-
CHLORO-
METHANE
-
-
ND
ND
-
-
ND
ND
-
-
ND
ND
-
-
ND
ND
-
-
ND
ND
ND

-
ND
ND

ND
ND
ND

-
ND

-
ND
-
ND
CHLORO-
DIBROMO-
METHANE
-
-
ND
ND
-
-
ND
ND
-
-
NT) "•
ND
-
-
TO
ND
-
-
Nb "
Nb '
ND 	 •'

-
ND
ND

ND
ND
ND

-
ND

-
ND
-
ND
Bromoform
-
-
NTT"
"ND
-
-
ND
ND
-
-
m '
m
-
-
W>"
NT>""
-
-
NT)
NT)
NT)

-
MB '
"W

NT) '"
ND
Nfl

-
ND

-
ND
-
Nb
Methylene
Chloride



























_


_
_

_
_
_

1ND - not detected

  -  not  analyzed  for
52

-------
                        SILVERDALE  VOC RESULTS




                          Concentration (yg/L)




                                Model  C
LOCATION
/Flow
11/41
41
123
403
89
139
128
192
16/11
32
89
363
71
108
149
20/46
100
226
10'
36
54
130
28/43
121
20'
158
185
303
348
383
32/18
64
13S
187
237
313
336
42/136
99s
227
256
367
DATE
3/14/83
3/21/83
6/13/83
9/12/83
11/14/83
1/16/84
3/12/84
4/2/84
3/14/83
4/18/83
7/11/83
10/10/83
12/12/83
2/6/84
4/9/84
3/14/83
4/18/83
7/11/83
9/12/83
1/16/84
2/6/84
4/9/84
3/14/83
5/16/83
8/15/83
10/17/83
11/7/83
1/16/84
2/13/84
3/5/84
3/14/83
5/16/83
8/15/83
12/5/83
1/9/84
3/12/84
4/2/84
6/13/83
9/12/83
11/7/83
12/5/83
,2/13/84 .,
TRI-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
ND
ND
ND
ND
ND
18
19
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1.0
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
TETRA-
CHLORO-
ETHYLENE
ND
ND
ND
ND
ND
ND
ND
ND
ND
3
3
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
., ,ND
1,1,1-TRl-
CHLORO-
ETHANE
_2
-
ND
ND
ND
ND
0.6
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
_
-
ND
ND
ND
ND
1.2
ND

-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1,2-DI-
CHLORO-
ETHANE
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
-
-
-
-
-
ND
ND
-
_
-
-
_
ND
ND
ND
_
_
_
-
_
ND
ND
-
_
_
_
ND
ND
_
CARBON
TETRA-
CHLORIDE
-
-
-
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
0.5

-
ND
ND
ND
ND
ND
^
_
ND
ND
ND
CHLORO-
FORM
-
-
ND
2.
ND
ND
2.0
ND
-
-
.1
ND
ND
ND
ND
-
-
ND
1
ND
ND
ND
-
-
ND
ND
ND
ND
4.2
0.6

-
ND
ND
ND
0.6
ND
ND
1
ND
ND
KD..
BROMODI-
CHLORO-
METHANE
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND

-
ND
ND
ND
ND ..
ND
ND
ND
ND
ND
ND
CHLORO-
DIBROMO-
METHANE
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND

-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Bromoform j
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND

-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Methylene
Chloride
-
-
-
-
-
-
ND
ND
-
-
-
-
-
-
0.7
-
-
-
-
-
-
ND
-
-
-
-
_
-
-
ND

-
-
-
-
4.0
ND
_
_
-
-
-
ND - not detected
- not analyzed for




- new cartridge
                                53

-------
                         SILVERDALE  VOC RESULTS




                          Concentration (yg/L)




                            Model C  (cont.)
LOCATION
/Flow
42/394
44/41
81
46
100
106
49/51
146
1013
163
217
279
60 us
293
55
62
111

























DATE
3/5/84
4/18/fi,1
8/15/83
3,1/7/83
2/13/84
3/5/84
3/21/83
6/13/83
10/10/83
12/12/83
2/6/84
4/9/84
5/16/83
9/12/83
11/14/83
12/5/83
4/12/84

























TRI-
CHLORO-
ETHYLENE
NT)1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

























TETRA-
CHLORO-
ETHYLENE
ND
ND, ,.
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

























1,1,1-TRI-
CHLORO-
ETHANE
_ND

ND
ND
ND
ND

ND
ND
ND
ND
ND
-
ND
ND
ND
ND

























1,2-DI-
CHLORO-
ETHANE
5


ND

_
-
-
ND
ND
-
-
-
ND
ND
ND
-

























CARBON
TETRA-
CHLORIDE
ND

ND
ND
ND
ND

-
ND
ND
ND
ND
-
ND
ND
ND
0.7

























CHLORO-
FORM
ND

ND
ND
ND
ND

ND
ND
ND
ND
ND
-
1
ND
ND
0.5

























BROMODI-
CHLORO-
METHANE
ND

ND
ND
ND
ND

ND
ND
ND
ND
ND
-
ND
ND
ND
ND

























CHLORO-
DIBROMO-
METHANE
ND

ND
ND
ND
ND

ND
ND
ND
ND
ND
-
ND
ND
ND
Wb

























Bromoform
ND |

ND
ND
ND
ND

'WI-
ND
ND
ND
w
-
Ntl
NT)
'W
ND'

























Methylene
Chloride
ND




ND

-
-
-
-
'Tirr
-
-
-
-
TO1"

























ND - not detected
 ~ not analyzed for



 - new cartridge
                                54

-------
                          SILVERDALE VOC RESULTS


                           Concentration (yg/L)


                                  Model D
LOCATION
/Flow
5/62
181
398
445
485
503
551
15/31
37
64
109
125
138
142
201
18/21
31
82
135
172
199
201
237
25/63
113
197
255
291
311
335
359
27/49
112
213
298
321
401
449
460
36/51
211
225
DATE
3/21/83
6J13/83
10/10/83
12/12/83
1/16/84
2/6/84
4/9/84
3/14/83
4/18/83
7/11/83
10/10/83
12/12/83
1/16/84
2/6/84
4/9/84
3/14/83
4/18/83
7/11/83
10/10/83
12/12/83
1/16/84
2/6/84
4/9/84
3/14/83
5/16/83
8/15/83
11/2/83
12/12/83
1/9/84
2/13/84
3/5/84
3/14/83
5/16/83
8/15/83
10717/83
11/7/83
1/16/84
2/13/84
3/5/84
6/13/fn
10/17/8^
n/7/fn
TRI-
CHLORO-
ETHYLENE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.6
ND
ND
ND
ND
ND
ND
ND
n 6
ND
ND
NT)
NT)
TETRA-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT)
NT>
1.1,1-TRI-
CHLORO-
ETHANE
_2
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND


ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
0.8
1.0
_
_
ND
ND
ND
ND
n fi
NT)
NT)
NT)
ND
1,2-DI-
CHLORO-
ETHANE
-
_
ND
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
-
-
_
ND
ND
ND
_
_
_
_
_
ND
ND
KD
_

—
KT)
vn
CARBON
TETRA-
CHLORIDE
— M
_
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
0.8
_
_
ND
ND
ND
NT)
ND
r> fi
_
NT)
NT)
CHLORO-
FORM
„ —
ND •
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
1.6
5.0
-
-
ND
ND
ND
ND
1.3
0.8
ND
ND
ND
BROMODI-
CHLORO-
METHANE
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
.
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
KD
KD
KD
CHLORO-
DIBROMO-
METHANE
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
Bromoform
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
Methylene
Chloride
-
-
-
-
-
-
ND
-
-
-
-
-
-
-
ND
-
-
-
-
-
-
-
ND
-
-
-
-
-
-
-
ND
-
-
-
-
1
-
_
1.3
-
-
-
 ND - not detected ~-
2
 - not analyzed for
                                   55

-------
                         SILVERDALE VOC RESULTS




                          Concentration (yg/L)




                            Model  D (cont.)
LOCATION
/Flow
36/239
243
246
TNA

398
435
495
501
50/31
64
160
254
328
59/349
660
831
895
1123
1126
66/171
463
607
655
835
878
















DATE
2/5/83
1/16/84
2/13/84
4718/83
7/11/83
10/17/83
TI7l47#y
2V13/8"4
3/5/84
4/18/83
6/13/83
10/10/83
2/6/84
4/9/84
6/13/83
9/12/83
11/14/83
12/5/83
3/12/84
4/2/84
5/16/83
9/12/83
11/14/83
12/5/83
3/12/84
4/2/84
















TRI-
CHLORO-
ETHYLENE
vm
Km
m
NTi
m
m
m
NT)
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
r ND
ND
ND
ND
















TETRA-
CHLORO-
ETHYLENE
NT)1
NT1
wn
m
m
m
m
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
















1,1,1-TRI-
CHLORO-
ETHANE
m
m
1 7

m
NT)
m
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
0.7
ND
















1,2-DI-
CHLORO-
ETHANE
m
m

„
„
Km
m

-
-
-
ND
-
-
-
ND
ND
ND

_
_
ND
ND
ND

-
















CARBON
TETRA-
CHLORIDE
ND
ND
ND

-
ND
ND
0.7
0.6
-
-
ND
ND
ND
-
ND
ND
ND
ND
ND

ND
ND
ND
0.5
ND
















CHLORO-
FORM
ND
ND '
1.0

ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
1.4
1.1
















BROMODI-
CHLORO-
HETHANE
ND
ND
ND

ND
ND
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
















CHLORO-
DIBROMO-
METHANE
ND
ND
ND

ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
















Bromoform
ND
ND
ND

ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
















Methylene
Chloride
_ 2
_
_
_
_
_
_
.
ND
_
-
_
_
ND
-
_
_
_
0.6
ND
_
_
_
_
ND
ND
















ND - not detected
 - not analyzed for
                     NA - not available
                                 56

-------
                          SILVKRDALE VOC  RESULTS
                           Concentration  (pg/L)
                                  Model E
LOCATION
/Flow
6/38
46
156
398
487
546
577
662
8/113
152
176*
8$
12/31 |
35
131
408
490
608
656
13/8
28
52
61
17 /8
68
182
295
353
400
425
482
19 /O
27
87
21/69
150
235
278
373
413
451
466
DATE
3/14/83
3/21/83
6/13/83
10/10/83
12/12/83
1/16/84
2/6/84
4/9/84
12/12/83
1/16/84
2/6/84
4/9/84
3/14/83
3721/83
6/13/83
11/14/84
1/16/84
3/12/84
4/2/84
11/14/83
1/16/84
3/20/84
4/2/84
3/14/83
4/18/83
7/11/83
10/10/83
12/12/83
1/16/84
2/6/84
479/84
11/14/83
1/16/84
4J9/84
3/14/83
5/16/83
7/11/83
8715/83
12/5/83
1/16/84
3/12/84
4/2/84
TRI-
CHLORO-
ETHYLENE
ND
ND
. ND
ND
ND
ND
ND
ND
21.5
24.3
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
3
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT)
NT)
ND
NB
TTO
TETRA-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
ND
ND
ND
ND
7.4
3.5
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
TTO
ND
NT1
NT)
1,1,1-TRI-
CHLORO-
ETHANE
_2
-
ND
ND
ND
NP
TO
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
_
ND
NT)
Nn
NT)
m
Nn
1,2-DI-
CHLORO-
ETHANE
-
-
_
ND
ND
ND

-
ND
ND
-
-
-
-
-
ND
ND
-
-
ND
ND
-
-
_
-
-
ND
ND
ND
_
_
.ND
ND
_
_
_
_
^
Nn
_
_
_
CARBON
TETRA-
CHLORIDE
-
_
-
. ND
ND
ND
ND
ND
ND
ND
ND
ND
_
-
-
ND
ND
ND
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
1.1
CHLORO-
FORM


ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
16.7
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
BROMODI-
CHLORO-
METHANE
-
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
3.1
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
CHLORO-
DIBROMO-
HETHANE
_
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
Broaofora I
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
mi
ND
Hethylene
Chloride
-
-
-
-
-
-
-
o1:?11
_
-
-
'JoT
-
-
-
-
-
0.6
ND
_
-
ND
133
-
-
-
-
-
-
-
1.6
-
-
30.1
-
-
-
-
_
_
4.8
4.7
ND - not detected
 - not analyzed for
                                  57

-------
                          SILVERDALE VOC  RESULTS
                           Concentration  (yg/L)
                             Model E (cont.)
LOCATION
/Flow
26/8
25
27
.«..
30/NAr
105
240
342
403
456
522
554
35/98
143
168
41/63
428
549
638
411-
117
168
46/9
51
48/82
156
200
247
282
332
52/1
26
33
54/63
365
597
648
707
730
SS/2
infi
177
DATE
12/5/83
1/16/84
2/13/84
3/5/84
3/14/83
5/16/83
8/15/83
10/17/83
12/5/83
1/16/84
3/12/84
4/2/84
1/16/84
2/13/84
3/5/84
4/18/83
8/15/83
10/17/83
11/7/83
1/9/84
2/13/84
3/5/84
11/14/83
2/13/84
4/18/83
7/11/83
10/10/83
12/12/83
2/6/84
4/9/84
11/14/83
1/9/84
2/13/84
3/21/83
7/11/83
11/7/83
12/12/83
2/1 3/84
3/1 7/84
11 /14/R3
7/A/Ri
4/Q/R4
TRI-
CHLORO-
ETHYLENE
wn1
m
n..fi
mi
ND
m
12
, ND .
TO
ND
ND
ND
ND
6.1
m
3
NT)
ND
19. n
m
un
NT)
ND
NT)
ND
ND
??.2
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT]
KTl
NT!
wn
NTI
TETRA-
CHLORO-
ETHYLENE
ND
ND
ND
ND
ND
ND
4
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
4.5
ND
ND
ND
ND
ND
ND
ND
2.7
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1,1,1-TRI-
CHLORO-
ETHANE
ND
ND
0.8
ND
-
-
ND
ND
ND
ND
ND
ND
ND
1.6
ND
-
ND
ND
ND
ND
ND
ND
ND
1.4
-
ND
HD
ND
ND
ND
ND
ND
0.7
-
ND
ND
ND
ND
ND
ND
ND
ND
1,2-DI-
CHLORO-
ETHANE
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
ND
-
-
-
-
ND
ND
ND
-
-
ND
-
-
-
ND
ND
-
-
ND
ND
-
-
-
ND
ND
-
_
ND
-
-
CARBON
TETRA-
CHLORIDE
ND
ND
ND
0.5
-
-
1
ND
ND
ND
ND
ND
ND
ND
ND
-
ND
ND
4.7
ND
ND
0.5
ND
ND
-
-
1.5
ND
ND
ND
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
CHLORO-
FORM
ND
ND •
0.7
1.2
-
_
ND
ND
ND
ND
0.6
ND
ND
2.3
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
-
ND
1.4
ND
ND
ND
ND
ND
0.7
-
ND
ND
ND
2.8
ND
ND
ND
ND
BROMODI-
CHLORO-
METHANE
ND
ND
ND
ND
_
_
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
CHLORO-
DIBROMO-
METHANE
ND
ND
ND
ND
_
_
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
Bronoforn 1
ND
ND
ND
ND

_
ND
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
ND
ND
ND
_
ND
NP
ND
ND
ND
ND
ND
ND
-
ND
ND
ND
ND
ND
ND
ND
ND
Mcthylene
Chloride
1 Z


12.3

_
_

_
_
10,0
0.6
_
_
259
„
_
_
_
_
_
IP*-

_
_
_
_
_
_
324
_
_
_
_
_
_
_
_
ND
_
-
324
ND - not detected
 - not analyzed for
                            NA - not available
                             - new cartridge
                                   58

-------
                        SILVERDALE VOC RESULTS




                         Concentration (yg/L)




                           Model  E (cont.)
LOCATION
/Flow
56/43
68
94
130
136
64/1
9
63
74
67/11
38
204
242
68 /14
66
207
227
236
304
326






















DATE
6/13/83
11/14/83
12/5/83
3/12)84
4/2/84
11/14/83
12/5/83
3/12/84
4/2/84
11/14/83
12/5/83
3/12/84
4/2/84
3/7/83
5/16/83
9/12/83
11/14/83
12/5/83
3/12/84
4/2/84






















TRI-
CHLORO-
ETHYLENE
ND1
ND
im
ND
ND
ND
NT)
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND






















TETRA-
CHLORO-
ETHYLENE
NT)
ND
ND
ND
ND
ND
ND
ND
ND
ND ,
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT)






















1,1,1-TRI-
CHLORO-
ETHANE
ND
ND
ND
ND
WD
ND
ND
ND
ND
ND
ND
0.7
ND
_
_
ND
ND
ND
ND
ND






















1,2-DI-
CHLORO-
ETHANE
-Z
ND
ND
_
_
ND
ND
_
_
ND
ND
_
_
_
_.
ND
ND
WO
_
_






















CARBON
TETRA-
CHLORIDE
-
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.5
ND
-
-
ND
ND
ND
ND
ND






















CHLORO-
FORM
ND
ND
ND
ND
ND
ND
ND
ND
1.6
ND
ND
0.6
0.9
-
-
ND
ND
ND
ND
ND






















BROMODI-
CHLORO-
METBANE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND






















CHLORO-
DIBROMO-
METHANE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND






















Bronoforn
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
-






















Methylene
Chloride
-
-
-
16.2
53.0
-
-
9.6
548.0
-
-
30.5
L66.0
-
-
-
-
-
9.8
6.4






















ND - not detected



 - not analyzed for
                               59

-------
                       SILVEKDALE  VOC RESULTS




                         Concentration (yg/L)




                               Model F
LOCATION
/Flow
3/31
44
81
86
103
107
108
121
4/24
36
181
306
375
426
456
533
14/20
38
109
214
271
314
420
NA5
22/17
48
112
151
170
196
239
242
34/05
25
54
87
107
110
143
156
165
43/120
DATE
5/16/83
6/13/83
8/15/83
10/10/83
12/12/83
1/16/84
2/6784
4/9/84
3/14/83
3/21/83
6/13/83
10/10/83
12/12/83
1/16/84
2/6/54
4/2/84
3/14/83
3/21/83
6/13/83
9/12/83
11/14/83
1/9/84
3/12/84
4/2J84
3/14/83
4/18/83
8/15/83
10/17/83
12/5/83
1/16/84
3/12/84
4/2/84
3/7/83
4/18/83
7/11/83
9/12/83
10/17/83
11/7/83
1/16/84
2/13/84
3/5/84
4/18/83
TRI-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
2.9
ND
ND
TETRA-
CHLORO-
ETHYLENE
ND
ND
"ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND _
ND
ND
ND
ND
N)
ND
ND
ND
ND
ND
1,1,1-TRI-
CHLORO-
ETHANE
_z
ND
r ND
ND
ND
ND
ND
ND
—
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
1.6
ND
-
1,2-DI-
CHLORO-
ETHANE
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
ND
-
-
-
-
-
ND
ND
ND
ND
_
_
-
CARBON
TETRA-
CHLORIDE
-
-
ND
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
ND
-
-
-
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
0.6
ND
-
-
-
ND
ND
ND
ND
ND
ND
-
CHLORO-
FORM
-
ND "
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
2
ND
ND
ND
ND
-
-
ND
ND
ND
ND
0.6
6.6
-
-
ND
ND
ND
ND
ND
1.3
ND
-
BROMODI-
CHLORO-
METHANE
-
ND
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
-
CHLORO-
DIBROMO-
HETHANE
-
ND
' ND
ND
Mb ' .
' " ' ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
-
Bronoform
-
TO"'
ND""
ND
TO
TO
TO'
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
-
-
ND
ND
ND
ND
ND
ND
ND
-
Methylene I
Chloride
-
-
-
-
-
-
-
o.y
_
-
-
-
-
-
-
to
-
-
-
-
-
-
ND
ND
-
-
-
-
_
_
ND
ND
-
-
-
-
_
_
_
_
ND
-
ND - not detected




 - not analyzed for
NA - not available
                                60

-------
                           SILVERDALE VOC RESULTS




                            Concentration (yg/L)




                              Model F (cont.)
LOCATION
/Flow
43/316
568
603
721
51/130
130
137
152
SS/H
17
23
31
37
49
61/125
186
193
193
290
330
f,S/NA3
7Q4
369
387
518
545
















DATE
7/11/83
10/17/83
11/7/83
2/13/84
7/11/83
10/10/83
11/14/83
4/9/84
6/l-)/83
8/15/83
11/7/83
12/12/83
2/13/84
3/5/84
5/16/83
10/10/83
11/14/83
12/5/83
3/12/84
4/7/84
5/16/83
9/12/83
11/14/83
12/5/83
3/12/84
4/2/84
















TRI-
CHLORO-
ETHYLENE
ND1
ND
ND
ND
,ND , ,
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT)
vo
ND
ND
ND
ND
















TETRA-
CHLORO-
ETHYLENE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NT)
ND
ND
ND
ND
ND
















1,1,1-TRI-
CHLORO-
ETHANE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
_
ND
VI)
ND
VD
ND
















1,2-DI-
CHLORO-
ETHAKE
_ '1
ND
ND
_
_
ND
ND

-
-
ND
ND
_
_
_
ND
ND
ND
_
„
_
VD
ND
ND
_
_
















CARBON
TETRA-
CHLORIDE
_
ND
ND
ND
_
ND
ND
ND
-
ND
ND
ND
ND
0.6

ND
VD
ND
ND
NT)
_
VD
ND
ND
ND
ND
















CHLORO-
FORM
ND
ND
ND
ND
ND
ND
ND
ND
ND
VD
ND
ND
1.4
ND

ND
ND
ND
ND
ND
_
NT)
ND
ND
NT)
0.7
















BROMODI-
CHLORO-
METHANE
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
_
ND
ND
ND
VD
ND
















CHLORO-
DIBROMO-
METHANE
ND
ND
ND
ND
ND
ND
_. ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
_
ND
ND
ND
ND
ND
















Bromofora j
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND'
ND

ND
ND
ND
ND
,ND
_
m
ND
ND
ND
ND
















Mettiylene
Chloride
_
_
„
_
_
_
_
1.4
_
_
-
_
_
ND
_
_
_
_
ND
vp
_
_
_
,.
0.8
ND_
















ND - not detectable
 - not analyzed for
                            NA - not available
                                   61

-------
                  SILVERDALE MICROBIOLOGICAL DATA




                Standard Plate Count  Results (#/ml)




                              Model A
Location
7 Flow
8/21
26
104
209
13/8
11
42
82
19/28
42
76
117
26/34
126
233
35/28
71
304

46 /O
38
86
52/10
66
91
55 /10
36
NA.1
310
64 /O
36
104
67 /NAJ
165
207










Date
3/14/83
3/21/83
6/13/83
11/2/83
3/14/83
3/21/83
6/13/83
9/12/83
3/14/83
4/18/83
7/11/83
10/10/83
3/7/83
5/16/83
8/15/83
3/14/83
4/18/83
8/15/83
11/7/83
2/28/83
4/18/83
8/15/83
3/4/83
6/13/83
8/15/83
2/28/83
3/14/84
7/n/jn
in/in/s?
7/7S/8T
•i/lfi/Rl
Q/1 •> IH^
2/28/RT
5/16783
9/12/83










Unflushed/Un-
Dlsinfected Tap
1300
1200
5800
35
3700
5800
5800
3100
480
590
3100
5800
600
1800
470
2700
88
1000
110
4100
3200
5700
4200
2900
2800
>5800
5800
3400
130
1900
4800
5800
5100
2100
360










Flushed/Dis-
infected Tap



1



1



190


730


2
I


2300


2



81


410


3










1 L Flushed/
Disinfected Tap













































NA - not available
                              62

-------
                SILVERDALE MICROBIOLOGICAL DATA




              Standard Plate  Count Results (#/ml)




                            Model  C
Location
/Flow
11/41
41
123
40
139
192
16/11
32
89
361
94
108
20/46
100
226
in1
14
36
54
28/43
121
201
185
303
383
32/18
64
13 '
111
237
336
«/30
136
99
256
325
394
44/n

43
8J
46
89
106
49 /20
Date
3/14/83
3/21/83
6/8/83
9/12/83
1/16/84
4/2/84
3/14/83
4/18/83
7/11/83
10/10/83
1/16/84
2/6/84
3/14/83
4/18/83
7/11/83
9/12/83
10/10/83
1/16/83
2/6/83
3/14/83
5/16/83
8/15/83
11/7/83
1/16/83
3/5/84
3/14/83
5/16/83
8/15/83
12/5/83
1/9/84
4/2/84
3/7/83
6/13/83
9/12/83
12/5/83
1/23/84
3/5/84
2/28/83

4/18/83
8/15/83
11/7/83
1/23/84
3/5/84
2/28/83
Unflushed/Un-
Disinfected Tap
190
39
240
1300
2

1600
1200
340
1600
290
490
5800
1600
980
3200
3100
1100
4100
540
3900
2900
3600
830

4900
3200
560
350
630

5800
1200
1100
250
31
3
1200

1900
5800
2800
2100

2200
Flushed/Dis-
infected Tap



11
8
2



54
1
1



160
23
8
42


490
5
2
1


6
11
16
4


49
2
1
3



1300
230
47
32

1 L Flushed/
Disinfected Tap





1600


















11





3ftn





i






?on

new cartridge
                             63

-------
                  SILVERDALE MICROBIOLOGICAL DATA




                Standard Plate Count Results (#/ml)




                          Model C (cont.)
Location
/Flow
49/S1
lift
im a
195
217
60/10
48
29'
55
84
111


































Date
3/21/83
6/13/83
10/10/83
1/23/84
2/6/84
2/28/83
5/16/83
9/12/83
12/5/83
1/23/84
4/2/84


































Unflushed/Ui*-
Disinfected Tap
600
900
1400
780
110
2700
5800
5800
1300
83



































Flushed/Dis-
infected Tap


3
1
3


210
120
24
28


































1 L Flushed/
Disinfected Tap










170


































new cartridge
                               64

-------
                 SILVERDALE MICROBIOLOGICAL DATA




               Standard Plate Count Results (#/ml)




                             Model D
Location
/Flow
5/44
62
1ST
3Q8
485
503
15/31
37
64
. „. 109
138
142
18/21
31
82
135
199
•201
25/63
113
197
255
311
359
27/49
112
213
321
401
460
36/51
152
225
239
243
168
45/30
NA1
255
43S
AQA
501
^0 /10
1"
	 *-*ft —
Date
3/J/83
3/21/83
6/13/83
10/16/83
1/16/84
2/6/84
3/14/83
4/18/83
7/11/83
10/10/83
1/16/84
2/6/84
3/14/83
4/18/83
7/11/83
10/10/83
1/16/84
2/6/84
3/14/83
5/16/83
8/15/83
11/7/83
1/9/84
3/5/84
3/14/83
5/16/83
8/15/83
11/7/83
1/16/84
3/5/84
6/13/83
8/15/83
11/7/83
12/5/83
1/16/84
3/5784
2/28/83
4/18/83
7/11/83
11/7/81
1 / 2 3 / 8 f\
3/5/B'i
7/?R/R1
/, /l^/gl
6/13/83
Unfluahed/Un-
Dlslnfected Tap
1300
2700
4200
5100
3700
4200
4900
2700
1400
5800
5800
5800
5300
5800
2700
4900
5800
2600
600
1300
5100
2100
5800

4300
2700
5800
2900
3200

1400
5500
5800
5800
5800

1400
660
720
54 n
5800

7nn
180
1000
Flushed/Dis-
infected Tap



310
120
160



140
31Q
530



130
2900
1100


1200
620
540
480


4200
730
660
120

560
2300
900
3200
420



540
J?H . -
690



1 L Flushed/
Disinfected Tap























1300





3500





1500





5800



NA - not available
                              65

-------
  SILVERDALE MICROBIOLOGICAL DATA




Standard Plate Count Results (#/ml)




          Model D (cont.)
Location
/Flow
130
231
254
59 /60
349
660
875
988
1176
66 /60
171
463
655
752
878






























Date
10/10/83
1/23/84
2/6/84
2/28/83
6/13/83
9/12/83
12/5/83
1/23/84
4/2/84
2/28/83
5/16/83
9/12/83
12/5/83
1/23/84
4/2/84






























Dnflushed/Un-
Dislnfected Tap
5800
5800
1500
1200
1000
3400
1100
140

3900
1900
5800
5800
1400































Flushed/Dis-
infected Tap
480
130
110


1QO
510
6
24


600
550
180
1500






























1 t Flushed/
Disinfected Tap








110





3900






























              66

-------
                SILVERDALE MICROBIOLOGICAL  DATA




              Standard Plate Count Results  (#/ml)




                            Model E
Location
/Flow
6/138
46
156
398
546
577
8/152
176
12/131
35
131
291
490
656
•3/28
61
17/8
68
182
295
400
425
1^36
49
21/69
150
	 235
278
373
413
	 466
26/8
25
49
30 /NA1
105
	 240
	 422
45^

35 /QR
	 T£g
41 /2Q
	 f,^
	 42B_
Date
3/14/83
3/21/83
6/13/83
10/10/83
1/16/84
2/6/84
1/16/84
2/6/84
3/14/83
3/21/83
6/13/83
9/12/83
1/16/84
4/2/84
1/16/84
4/2/84
3/14/83
4/18/83
7/11/83
10/10/83
1/16/84
2/6/84
1/16/84
2/6/84
3/14/83
5/16/83
7/11/83
8/15/83
12/5/83
1/16/84
4/2/84
12/5/83
1/16/84
3/5/84
3/16/83
5/16/83
8/15/83
12/5/83
1/16/84
4/2/84
1/16/84
3/5/84
2/28/83
4/18/43
8/15/83
Unflushed/Un-
Disinfected Tap
1
180
1200
5800
2100
1000
1200
540
5200
5800
3600
2900
480

4100

1700
600
5800
1300
600
1300
130
2200
1900
5800
5100
5800
4800
4900

1300
5800

5800
5800
3700
4100
5100

3900

2300
660
5600
Flushed/Dis-
infected Tap



16
8
4
6
3



4
4
3
2
1



4
1
3
1
5



9
16
6
490
580
8
6


28(1
2
5
1
3
1


120
1 L Flushed/
Disinfected Tap













9

7














2100


480





Aonn

"n



not available
                             67

-------
                SILVERDALE MICROBIOLOGICAL DATA




              Standard Plate Count Results (#/ml)




                        Model E  (cont.)
Location
/Flow
• 41/638
4 11
168
46 /36
64
48 /10
82
156
200
271
282
52 /I
26
46
54 /30
63
365
597
691
730
55/89
108
56/10
43
68
94
118
136
64/9
30
74
67/38
118
242
68/14
66
207
236
263
326





Date
11/7/83
1/9/84
3/5/84
1/23/84
3/5/84
2/28/83
4/18/83
7/11/83
10/10/83
1/23/84
2/6/84
11/14/84
1/9/84
3/5/84
2/28/83
3/21/83
7/11/83
11/7/83
] /23/S4
3/5/84
T/23/84
2/6/84
2/28/83
J>/13/83
9/12/83
12/5/83
1/23/84
4/2/84
12/5/83
1/23/84
4/2/84
12/5/83
1/23/84
4/2/84
3/7/83
5/13J83
9/12/83
12/5/83
1/23784
4/2/84





Unflushed/Un-
Dis infected Tap
1500
4100

810

300
2700
1100
5600
3200
1200
580Q
1900

110
5800
1700
230
i ?nn

280
900
1
5800
4000
2900
2200

5200
5800

720
480

5800
3600
1300
5800
5800






Flushed/Dis-
infected Tap
6
9
81
59
4



12
6
1
1300
52
180



l
l
1
4
1


81
8
38
26
3

2
23
1
1


2
15
4
2





1 L Flushed/
Disinfected Tap


9QO

310








?pn





16







1400


380


1





i7n





new cartridge
                             68

-------
                 SILVERDALE MICROBIOLOGICAL DATA




               Standard  Plate Count Results (#/ml)




                              Model F
Location
/Flow
3/31
44
81
86
107
108
4 /24
36
181
306
426
456
14 /20
38
109
214
324
'HA1
22 in
48
112
170
196
242
34 /05
25
54
87
110
143
165
43 /30
120
316
603
709
721
51 /10
32
130
132
146
151
53 /O
11
Date
5/16/83
6/13/83
8/15/83
10/10/83
1/16/84
2/6/84
3/14/83
3/21/83
6/13/83
10/10/83
1/16/84
2/6/84
3/14/83
3/21/83
6/13/83
9/12/83
1/9/84
4/2/84
3/14/83
4/18/83
8/15/83
12/5/83
1/16/84
4/2/84
3/7/83
4/18/83
7/11/83
9/12/83
11/7/83
1/16/84
3/5/84
2/28/83
4/18/83
7/11/83
11/7/83
1/23/84
3/5/84
2/28/83
3/21/83
2/11/83
. 10/10/83
1/23/84
2/6/84
2/28/83
6/13/83
Unflushed/Un-
Disinfected Tap
5800
5800
5800
4400
4200
5800
310
2100
1100
260
900
900
3400
1600
5800
2700
2300

3200
3200
900
380
1300

5800
5400
4900
1600
610
4500

1900
2200
3400
3900
630

720
3200
920
960
3300
1700
710
3900
Flushed/Dls-
Infected Tap


49
12
36
24



1
1
6



110
61
41


54
6
3
11



62
30
180
20



5
1
27



2
3
2


1 L Flushed/
Disinfected Tap

















420





190






780





290








NA - not available
                             69

-------
                  SILVERDALE MICROBIOLOGICAL DATA




                Standard Plate Count  Results (#/ml)




                          Model F  (cont.)
Location
/Flow
53/17
23
23
49
61/20
125
186
19?
221
33(J
65/lQ
NAT
294
387
41?
545





























Date
8/15/83
11/7/83
1/23/84
3/5/84
2/28/83
5/16/83
10/10/83
12/5/83
1/23/84
4/2/84
2/28/83
3/16/83
9/12/83
12/5/83
1/23/84
4/2/84





























Unflu*hed/Un-
Dlslnfected Tap
2600
1100
150

3300
5800
2700
1200
340

2300
5800
1100
900
2300






























Flushed/Dis-
infected Tap
1
2
6
5


11
19
5
1


3
270
4
1





























1 L Flushed/
Disinfected Tap



180





6





11





























NA - not available
                              70

-------
                       APPENDIX C
VOC AHD MICROBIOLOGICAL RESULTS FOR POIHT-OF-DSE SAMPLES
                        Rockaway

                      VOC RESULTS
                        Pre/Post
                  Concentration (ug/L)
LOCATION
1



2



3



it


5

6

7

8

9



DATE

10/82
7/83
10/83

10/82
7/83
10/83

10/82
7/83
10/83

10/82
10/83

10/83

10/83

10/83

10/83

10/83


TRI-
CHLORO-
ETHYLENE

ND/ND
1/ND
ND/ND

ND/ND
2/ND
ND/ND

70/ND
52/4
161/ND

160/ND
150/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


TETRA-
CHLORO-
ETHYLENE

ND/ND
3/ND
ND/ND

ND/ND
2/ND
ND/ND

ND/ND
1/2
ND/ND

ND/ND
ND/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


1,1,1-TRI-
• CHLORO-
ETHANE

230/ND
154 /ND
82/ND

57/ND
10/ND
20/ND

ND/ND
ND/ND
0.8/ND

ND/ND
ND/ND

12/ND

ND/ND

-/ND

ND/ND

ND/ND


1,2-DI-
CHLORO-
ETHANE

—
—
—

—
—
—

—
—
—

—
—

—

—

—

—

—


CARBON
TETRA-
CHLORIDE

—
—
ND/ND

—
—
ND/ND

—
—
ND/ND

—
ND/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


CHLORO-
FORM

ND/ND
1/ND
ND/ND

ND/ND
2/ND
ND/ND

ND/ND
ND/ND
ND/ND

ND/ND
ND/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


BROMODI-
CHLORO-
HETHANE

—
ND/ND
ND/ND

—
ND/ND
ND/ND

—
ND/ND
ND/ND

—
ND/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


CHLORO-
DIBROMO-
METHANE

—
ND/ND
ND/ND

—
ND/ND
ND/ND

—
17ND
ND/ND

—
ND/ND

ND/ND

ND/ND

-/ND

ND/ND

ND/ND


Bromoforn

—
ND/NE
SO/NT

—
ND/NI
ND/NC

—
ND/NE
ND/Nt

—
ND/NI

ND/NI

ND/NI

-/ND

ND/NI

ND/NI


Methylene
Chloride

—
—
—

—
—
—

—
—
—

—
—

—

—

—

—

—


                        71

-------
               ROCKAWAT TOWNSHIP MICROBIOLOGICAL RESULTS
                            Pre/Post (f/nL)
                     Unflushed,  Undisinfected
Location
1
2
3
4
5
6
7
8
9
Date
7/25/83
10/26/83
7/25/83
9/8/83
10/26/83
12/1/83
7/12/83
9/8/83
10/26/83
12/1/83
7/12/83
10/26/83
9/8/83
10/26/83
9/8/83
10/26/83
12/1/83
10/26/83
12/1/83
10/26/83
10/26/83
12/1/83
SPC Coliform
<10/>10 
-------
                                   APPENDIX D
                MICROBIOLOGICAL RESULTS USING R2A AND SMA MEDIA
A specially developed plate count medium (R2A) was used to enumerate organisms
in unflushed postdevice water samples collected in Silverdale, Pennsylvania.
Samples were collected daily for one month.  Organisms were enumerated using
the R2A and standard methods agar (SMA).

The R2A media composition is presented in Table D-l.  The significant
difference in this media versus SMA is the inclusion of sodium pyruvate, which
enhances the recovery of stressed cells.  For the oligotroplic environment of
groundwater, the use of this media produces a greater potential for enumera-
tion of total heterotrophic organisms.

Table D-2 presents results of bacterial enumeration using both media.  Paired
samples of effluent from two of the same model non-silvered carbon devices and
silvered carbon devices were enumerated.

The differences between media are not as pronounced as the differences between
silver and non-silver impregnated carbon.  The difference between 48 and 72
hour incubation indicates the presence predominately slow growing organisms in
the effluent from the carbon beds.
                                    73

-------
                     TABLE  D-l.  COMPOSITION OF R2A MEDIUM

               INGREDIENT                 CONCENTRATION, yg/L

         Yeast Extract                            0.5
         Proteose Peptone No.  3                   0.5
         Gas ami no Acids                           0.5
         Glucose                                  0.5
         Soluble Starch                           0.5
         Sodium Pyruvate                          0. 3
         K2HP04                                   0.3
         MgSO!».7H20                               0.05
         Agar                                    15.0
Final pH 7.2, adjust with  K2HPOi+  or KfoPOit before adding agar.  Add  agar,  heat
medium to boiling  to dissolve agar and autoclave for 15 minutes at  121 °C,  15
psi.
             Table D~2.  GAG Postfilter Mean Standard Plate Counts
                       Standard Methods Agar vs R2A Medina

              Sample                  Standard Methods       R2A
               Type	Agar	Medium

              48 hr Incubation
                   Silver                   3556              4572
                   Non Silver               130                155
                   Combined                 585                720

              72 hr Incubation
                   Silver                   6145              7118
                   Non Silver               210                267
                   Combined                 990              1190
                          U.S. Environmental Protection Agency
                          Region 5, Library (PL-12J)
                          77 West Jackson Bouievand, 12th Floor
                          Chicago, IL  60604-3590
                                     74

-------