v»EPA

             United States
             Environmental Protection
             Agency
              Municipal Environmental Research
              Laboratory
              Cincinnati OH 45268
EPA-600/2-8O-132
August 1980
             Research and Development
Water Quality
Effects  Related to
Blending Waters in
Distribution Systems

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional  grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2,  Environmental Protection Technology
      3.  Ecological Research
      4.  Environmental Monitoring
      5.  Socioeconomic Environmental Studies
      6,  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has  been assigned  to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                         EPA-600/2-80-132
                                         August 1980
       WATER QUALITY EFFECTS RELATED TO
   BLENDING WATERS  IN DISTRIBUTION SYSTEMS
                       by

      Warren K.  Schimpff and Harold E. Pearson
 The Metropolitan Water District of Southern California
            Los Angeles, California 90054
                 Grant No. R804709
                  Project Officer

                 Marvin C. Gardels
          Drinking Water Research Division
     Municipal Environmental Research Laboratory
               Cincinnati,  Ohio  45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
      OFFICE OF RESEARCH AND DEVELOPMENT
    U.S. ENVIRONMENTAL  PROTECTION .AGENCY
             CINCINNATI,  OHIO  45268

-------
                              DISCLAIMER
          This  report has  been  reviewed  by  the  Municipal  Environmental
Research  Laboratory, U.S.  Environmental Protection Agency,  and approved
for publication.  Approval  does  not  signify that the contents necessarily
reflect the views and policies of the U.S. Environmental  Protection Agency,
nor does  mention of trade names or commercial  products  constitute endorse-
ment or recommendation for use.
                                    11

-------
                                FOREWORD
          The  Environmental  Protection  Agency  was  created because of
increasing public  and government concern about  the dangers of pollution to
the health and welfare of the American people.   Noxious  air, foul water,
and  spoiled  land  are tragic  testimony to the  deterioration of our natural
environment.   The complexity of that environment and the interplay between
its components  require a concentrated and integrated attack on the problem.

          Research and development  is that necessary first step in problem
solution  and it involves  defining the  problem, measuring  its impact,  and
searching  for  solutions.   The Municipal Environmental Research  Laboratory
develops  new  and improved technology  and  systems for  the prevention,
treatment, and management  of  wastewater and solid and  haaardous  waste
pollutant   discharges  from   municipal  and  community  sources,   for  the
preservation  and  treatment  of  public  drinking water  supplies,  and to
minimize  the  adverse  economic,  social,  health,  and  aesthetic effects of
pollution.  This publication is one of the products of that  research; a most
vital communications link between the researcher and the user community.

          This  study  evaluates  the  effects of blending two or more waters
of different quality and relates  their composition to the corrosive effects of
the  water  in  distribution  systems.    The  EPA's  mobile  water quality
monitoring laboratory  was used  to amass field data on parameters related to
corrosivity and stability of waters  representing those available in  Southern
California.  The study should provide a data base to support the states  and
EPA  in  their responsibilities under the  Safe Drinking Water  Act to determine
the  need  for  water  quality  control programs to minimize  health  effects
associated with  the   presence  of  contaminants  that are   the products of
corrosion in water distribution systems.

                                     Francis T. Mayo, Director
                                     Municipal Environmental Research
                                     Laboratory

-------
                                ABSTRACT
          This study was conducted to evaluate the effects of blending two
or more waters  of different  quality and to relate  their composition to the
corrosive effects  and calcium carbonate, deposition tendency of the water on
distribution  systems.  The  EPA mobile water  quality monitoring  laboratory
was  deployed at  30 selected sites within the service area of The Metropolitan
Water  District  of  Southern California  where  imported waters  from  the
Colorado  River and California  aqueducts are used  as  delivered or  blended
with local groundwaters.   Eighteen  computer controlled parametric  systems
on board the  laboratory  analyzed and recorded field  data to assess water
quality factors associated with  corrosion and stability.   The waters studied
could be  classified as having moderate  to high  hardness, alkalinity and total
dissolved solids content.

          The data were  analyzed for significant interrelationships relative
to   pH,  calcium  hardness,  alkalinity,   dissolved  minerals,  polarization
corrosion rates on day  7, calcium carbonate  deposition  test (CCDT)  results,
and  calculated values for the Langelier saturation index and Ryzner  stability
index.

          For waters  of similar chemical  composition  the CCDT  results were
more indicative  of  the benefits  to be  derived from  pH  control  or zinc
phosphate films   for  mitigating corrosion  than  the polarization corrosion
rates.

          Cost comparisons  for corrosion  control by use of caustic soda to
adjust  pH and  zinc  phosphate to  promote  protective film deposition  were
made.   An experimental program of intermittent  application  of zinc phosphate
was  proposed to optimize the  costs  and benefits of this treatment.

          The  continuous   monitoring   systems  have  the  capability  of
responding  to those differences in water  quality which can then be used to
make qualitative comparisons of the relative stability  and/or corrosivity of
waters blended in distribution systems.

          This report was  submitted in fulfillment of Grant No. R804709  by
The   Metropolitan  Water  District  of  Southern  California,   Los  Angeles,
California  under  the  sponsorship  of the   U.S.  Environmental  Protection
Agency.  This report covers  the period from October 1976 to October  1978.

-------
                            CONTENTS
Foreword	iii
Abstract 	 ...... 	   iv
Figures  	 .  	  ........  vii
Tables	ix
Acknowledgment ....  	    x

  1.   Introduction 	    1
        General	    1
        Objectives and  Study Plan	    4

  2.   Conclusions  	  .....    6

  3.   Recommendations	    9

  4.   The EPA Mobile Water Quality Monitoring Laboratory .....   11
        General	11
        Improvements to the Mobile Laboratory  ...  	   15
          Dual Power Hook-Up Capability  	   15
          Air-Conditioning and Temperature Control    	   15
          Modifications to Existing Monitoring Systems ......   17
            Schneider Robot Monitor  	  .  	   17
            Residual Chlorine Analyzers  . 	   17
            Calcium Carbonate Deposition Test  	   17
            Corrosion Rate Meter	18
            24-Bottler Sampler  	   18
          Monitoring Equipment Added by MWD  	  .....   20
            H-F Turbidimeter	20
            Turbine Flowmeter   	   20
        Mobile Laboratory Computer System  	   21
          General	21
          Worker Tasks  	   21
          The DAQ Worker Task	22
          Computer Cooling Problems  	   27

  5.   Data Handling and Presentation	28
        Paper Tape Data	28
        Recorder Chart Data	  .   29
        Data Processing	29
                                 V

-------
                     CONTENTS (continued)
6.  Experimental Field Monitoring  	  34
      Monitoring Sites .  ,	34
        Site Operating Conditions  	  34
      Parameters Monitored 	  41
      Sampling for Manual and Special Analysis 	  41
      Monitoring Period  	 	  43
      Cleaning Corrosion Electrodes   	 .  	  43
      Cleaning CCDT Electrodes 	  43
      Special Attention for the CCDT	44
      Instrument Calibration 	  44
      Instrument Problems	  44

7.  Results and Discussion	47
      General  .....  	  47
      Interrelations Between Water Quality Parameters  	  59
        Effect of pH on Corrosion Rates  .	59
        Effect of Calcium Content on Corrosion Rates 	  61
        Effect of Dissolved Minerals on Corrosion Rates  ....  61
      Calcium Carbonate Deposition Test (CCDT) 	  65
        Effect of pH Adjustment on CCDT	67
        Effect of Alkalinity and Hardness on CCDT  , ~v	67
        Effect of Zinc Phosphate on CCDT and Corrosion Rates .  .  67
      Corrosion Coupon Tests 	  75
      Interrelations between CCDT and Other Quality Paremeters  .  76
      Response of Monitoring Systems to Quality Changes  ....  81
      Cost of Corrosion Control Treatment  	  84

References	86
Appendix	88
                              VI

-------
                                   FIGURES
Number                                                                 Page

  1   MWD distribution system and service area  	  2

  2   Aqueducts serving Southern California ..... 	  3

  3   Photographs of EPA Mobile Water Quality Monitoring
        Laboratory—a) Exterior; b) Interior  	 12, 13

  4   Electrical service disconnect box 	 .... 16

  5   Schematic diagram of corrosion monitor flow cells
        with constant-head flow regulation	 19

  6   Schematic diagram of CCDT automation relay control panel  ,  ... 26

  7   Computer-plotted water quality data at Locations 25 — Graph A  . 32

  8   Computer-plotted water quality data at Locations 25 — Graph B  . 33

  9   MWD system delivery pattern showing monitoring locations—
        Pre-drought period (before March 1977)  	 35.

 10   MWD system delivery pattern showing monitoring locations—
        Drought period (March 1977-February 1978) 	 36

 11   MWD system delivery pattern showing monitoring locations—
        Post-drought period (after March 1978)  	 37

 12   Corrosion rates for mild steel, zinc, and copper on
        day 7 vs. Langelier saturation index  ....... 	 60

 13   Corrosion rates for mild steel, zinc, and copper on
        day 7 vs. Ryzner stability index	62

 14   Corrosion rates for mild steel, zinc, and copper on
        day 7 vs. calcium content	63

 15   Corrosion rates for mild steel, zinc, and copper on
        day 7 vs. electrical conductivity	64

 16   Effect of blending Colorado River water and
        State Project water on CCDT	66

                                     vii

-------
                             FIGURES (continued)


Number                                                                 Page

 17   Effect of pH adjustment of blended waters on CCDT	68

 18   Effect on CCDT due to in-line diurnal changes in
        source of water at Location 29	69

 19   Effect of adding zinc phosphate for corrosion
        control on CCDT	71

 20   Photographs of corrosion electrodes at (a) Location 13
        and (b) Location 28	74

 21   CCDT results vs. calcium and alkalinity levels  	 77

 22   CCDT results vs. corrosion rates for mild steel
        and zinc on day 7	78

 23   CCDT results vs. Langelier and Ryzner indices	79

 24   Data showing diurnal variations in water quality at
        Location 29 — Graph A	82

 25   Data showing diurnal variations in water quality at
        Location 29 — Graph B	83

 A-l  Flowchart of DAQ (MWD modification No. 3)	91

 A-2  Flowchart of DAQ Subroutine A	92

 A-3  Flowchart of DATA (MWD modification)  	92

 A-4  Flowchart of PUN	93
                                    Vlll

-------
                                   TABLES


Number                                                                 Page

  1   Sensors in the EPA Water Quality Monitoring Laboratory .....   14

  2   Mobile Laboratory Field Monitoring Locations 	   38

  3   Parameters Monitored for MWD Mobile Laboratory Project 	   42

  4   Chemical Analysis of MWD Source Waters—
        Averages from January 1977-June 1978 .	48

  5   Typical Chemical Analysis of MWD Filtered Waters—
        Monthly Composite of Daily Samples 	   49

  6   Typical Chemical Analysis of Groundwaters Used
        During Study 	   50

  7   Summary of Water Quality Data for Field Monitoring Stations  .  .   51

  8   Corrosion Rates for Mild Steel, Zinc, and Copper from
        Field Monitoring Stations  ..... 	  .  .   52

  9   Comparison of Long Term Corrosion Tests—Mild Steel  	   55

 10   Comparison of Long Term Corrosion Tests—Zinc	56

 11   Comparison of Long Term Corrosion Tests—Copper	57

 12   Trace Metal Levels at Field Monitoring Stations  	   58

 13   Comparisons of Corrosion Rates for Waters With and
        Without Zinc Added for Corrosion Control 	   70

 14   Zinc Levels for Locations Where Zinc Corrosion
        Inhibitor was Added	72

 15   Iron in Long Beach Distribution Water	   75
                                     IX

-------
                           ACKNOWLEDGMENTS
          Sincere  appreciation is  expressed  to  Robert C.  Thurnau,  who
served as  project officer during  the first year of the project.  Immediately
after  delivery of the mobile laboratory,  Mr. Thurnau spent eight days with
Metropolitan's personnel at La Verne,  California,  checking1 the instrumenta-
tion and  providing instruction on use  and maintenance  of  the monitoring
systems.   He gave  continuing support to requests by project personnel to
make  changes  in  the  on-board  computer  software  and  reprogramming to
increase the data collecting capability of the mobile laboratory.

          A  special note of thanks  goes to William B. Everett, formerly of
the  National Sanitation  Foundation,  who  helped  develop  the  original
computer   software  for  the  on-board  computer.   Mr.  Everett's  personal
interest in this  project  caused him to stop by La Verne,  California,  at his
own expense, and  spend a  day helping the MWD operators learn more about
the computer and its operating software.   The few  hours that Mr. Everett
spent in  California  were extremely  valuable  in that it was  from this  start
that we began acquiring the knowledge to make changes in the software and
reprogram  the computer.

          Thanks also to Gregory Kok,  Ph.D., of  the chemistry department
at Harvey  Mudd College in  Claremont, California, for the  temporary loan of
several  electronic  instruments that helped  to keep the mobile laboratory
operating  when  monitoring  problems  occurred that  would have otherwise
resulted in lost time and data.

          The cooperation of the  local water utilities that graciously allowed
the MWD  project staff to use their facilities  in order to  set up the mobile
laboratory  and  monitor  the  water in  their  respective  service  areas is
gratefully  acknowledged.  This project  could  not  have taken place without
the generous support of these water suppliers.  We  are also  grateful to the
many  people at the  local utilities who gave  special  help  in  preparing the
monitoring locations and who gave immediate assistance whenever necessary.
The water utilities  and  the  people  directly involved with  helping the mobile
laboratory  staff were:

     City of Alhambra, Department of Public Works
     L.  E.  Moeller,  Director of Public Works
     Orville Cheek, Water Distribution Foreman

     City of Anaheim, Utility Department
     Larry Sears, Water Planning Manager
     Burton H. Moore, Ph.D.,  Water Quality Supervisor

-------
                     ACKNOWLEDGMENTS (continued)
     City of Burbank,  Public Services Department
     Martindale Kile, Jr., Water Superintendent
     C. M. Leatherwood, Water Production Supervisor

     Foothill  Municipal Water District
     Ronald C. Palmer, General Manager
     Jack Esterly, Field Superintendent

     City of Long Beach, Water Department
     Clyde N. Moore, General Manager
     Harry G. Offner,  Laboratory Director

     City of Orange, Water Department
     Frank Page, Water Department Superintendent
     Steven P. Smith,  Junior Civil Engineer

     City of Pasadena, Water and Power Department
     Karl A.  Johnson,  General Manager
     Henry B. Steinbiser, Purification  Supervisor

     Walnut Valley Water District
     Edmund M. Biederman, General Manager
     with the cooperation of
       Libbey Glass, Division of Owens-Illinois
       Donald A. Heinz,  Engineering Services Supervisor

          Finally,  special  thanks  go  to the various MWD  sections without
whose  special help this project  and report would have been very difficult to
complete:  to Donna Squire  in  the Centralized  Control Section  for writing
the  programs and for  transferring the raw paper  tape  data to magnetic
tape;  to Earl  Cowden  in  the Data  Processing Section  for  writing  and
modifying the programs  that prepare the computer listing and graphs of the
data;  to Bill Webb  and  the  Graphic  Services  Section;  and  to  Arteemas
Greene  and  the Word Processing  Section for the preparation  of this  final
report.

          The MWD staff for this  project consisted of:  Harold E. Pearson,
Ph.D.,  project manager; Warren K. Schimpff,  Ph.D., principal investigator;
and Ignacio C. Valdivieso,  engineering technician.

-------
                                SECTION  1

                              INTRODUCTION
GENERAL

          Regional  water  systems  designed  for  the efficient  collection,
treatment,  and  distribution  of water  in  large  urban  areas  have  been
remarkably  successful in meeting  the  needs of the large  population served.
Inherent in most  regional plans is the right of water purveyors within the
regional boundaries to  retain  and utilize  their  historical  water sources.
Moreover,  due  to  the  higher  cost  of imported water, the  local supply  is
often used  exclusively  part  of the year  and  supplemental  water from the
regional system  is purchased as  needed  during  periods  of peak demand.
Thus, in an  area with  a  regional  supplemental water supply  there  is the
chance  that a given  distribution  system  may  be  exposed to two waters  of
different  quality and  origin and to various blends of the two waters.  With
daily, weekly, monthly,  or seasonal changes in water quality,  there  is the
possibility that these changes  could have an adverse effect on the distri-
bution  system  itself  or  on  the   quality  of   the  water  delivered  to the
consumer.

          The Metropolitan Water   District  of  Southern  California  (MWD)
operates one  such regional water  distribution system.  MWD is  a public and
municipal corporation  of the  State  of California which provides  supplemental
water,  as  a  wholesaler, through  its  27 member agencies (cities and water
districts) to  nearly 11  million  people  in a 4900-square-mile  service area on
the coastal  plain of Southern California,  as shown  in Figure  1.  MWD imports
water to  Southern  California  from  two  distant sources;  the Colorado River
via the Colorado  River  Aqueduct,  and the Sacramento-San Joaquin Delta  in
Northern  California via the California Aqueduct, see Figure 2. Approximately
one-half of  this  supplemental  water--700,000  acre-feet—is  being imported
annually from the  Colorado River.   Initial deliveries of Colorado  River water
(CRW)  began in 1941.   In  addition,  Metropolitan has contracted  to ultimately
receive more  than  2 million acre-feet  annually of  Northern  California water
through the State Water Project (SWP).   First deliveries of northern water
began in 1972.

          There  are 128 incorporated  cities within Metropolitan's boundaries
and   the  imported  water  comprises from  0 to  100  percent of the  water
delivered to the consumers in various cities.   MWD supplies three different
waters, either treated or raw, to its member agencies; Colorado  River water
(CRW), State  project water (SPW),  and  a  blend of Colorado  River and State
project waters  (CRW/SPW).   Most  of  the communities within Metropolitan's
service area  have  some local  groundwater supply  of  their own  that is used

-------
Figure 1.  Metropolitan Water District's distribution system and service area.

-------
   SAN
I RAN CISCO
                                                         SAN DIEGO
                     Figure 2.  Aqueducts serving Southern California

-------
intermittently  or  blended  more  or  less  continuously  with  this  imported
water.  Each local supply  may vary significantly from a quality  standpoint.
Thus,  within  the confines of Metropolitan's  service area there is  a  good
opportunity to study a wide range of different water blends.

          Mixing  waters  of  different  origin  and   quality in  distribution
systems may  cause the quality of water to  deteriorate before  it is  delivered
at the  consumer's tap.  To date,  there  have been few attempts to establish
functional relationships in  the  quality  of  a water supply before and  after
blending.  Monitoring  various parameters of water quality should provide a
basis for assessment  of  problems associated with corrosion and stability due
to blending.   Such  studies will help  in  developing  a  plan for  corrosion
control in  addition  to providing  a basis  for  establishing consumer  costs
related to changes in water quality.

          Attempts  have  been made  to establish  a  functional relationship
between water quality and  household cost  by interviewing consumers.   The
parameters  measured  in  these  surveys  have generally been related to  total
dissolved solids  (TDS) and/or  hardness of waters.  The  most  comprehensive
studies were  done by  Black and Veatch  (1), Metcalf and Eddy (2), and the
Orange County Water District (3).

          While these  surveys  have done  much  to  point out  the long-term
costs of poorer  quality water, they have done  little  to  provide  answers to
more immediate  problems.  Among these  are the  effects of periodic shifting
from one  source  of water to another or of blending different waters within a
distribution  system rather  than  at a central point.   Some of these  factors
were discussed in a paper by Pearson and Singer (4) for the  major regional
water importation  system operated  by MWD.

          Detailed investigation  of  water  quality parameters is needed to
delineate  factors  related  to corrosivity and  stability of blended water within
such regional systems in order to develop adequate treatment methods either
for the local water supplier or for the regional system to employ.

Objectives and Study  Plan

          The  principal objective of this project was the evaluation of the
effects  of blending  two or  more waters  of different  origin  and inorganic
chemical  content  in  public  water  systems.   A  second 'objective  was  to
quantify  corrosion rates  and  calcium  carbonate  deposition  tendency  in
dynamic  water  systems  before  and  after  blending.   It was  hoped  that
control measures could  be developed  that  would minimize  deterioration of
water  distribution system  facilities and household plumbing.   Methods for
mitigating  unfavorable effects due to unstable conditions caused by  mixing
different waters could then  be recommended  to water utility managers.

          In October 1976, MWD  was awarded a shared cost research grant
from the  Environmental Protection Agency  (EPA)  to  use  their mobile water
quality  monitoring  laboratory   designed  and  assembled by   the  National
Sanitation  Foundation  (5)  to study the  effects  on  water  quality  due to
blending  waters  of  different  origin.    As  indicated  above,   the  area  of

-------
Southern  California  served by MWD  is  an  especially  good  area  to  study
changes in  water quality due  to blending because of the variety of waters
and blends of waters used in the various Southern California communities.

          MWD has  three  400  mgd and one  150 mgd filtration plants  which
provide water for domestic  purposes in Los Angeles,  Orange,  San Diego and
Ventura Counties.  The effluent from  each of MWD's plants is stabilized by
final pH adjustment,  but the  treated  waters  are later commingled with local
groundwaters which may vary  in mineralization  from  200 to 750 mg/L of TDS
and  some  contain  significant  amounts of free  carbon  dioxide.  Monitoring
locations  for  this   project  were  selected  to  obtain  data  for making
comparisons  of the corrosivity  and stability of  MWD's  two waters, Colorado
River  Water  (CRW),  and  State project  water  (SPW),  blends  of  the two
(CRW/SPW), and blends of these waters with local water.

          Several communities  in  the Southern California  area  add zinc
phosphate  corrosion inhibitors  to  their water supplies.  Monitoring of water
quality in these systems above  and below the point of adding zinc phosphate
was  included in the study plan to provide  insight into  the  value of such
treatment.

          An important  corollary  objective  of this investigation was  to
accumulate a data base related  to corrosivity and stability parameters to be
made available to the  states  and EPA for support  of their responsibilities
under  the   federal  Safe   Drinking Water Act (SDWA).   As  amended  in
November  1977,  the  SDWA  authorizes EPA  to  prescribe  special monitoring
requirements  for unregulated  contaminants.  Such  monitoring may provide
criteria for  water quality control to minimize health effects associated with
the presence of contaminants,  including metals (e.g.  lead  and  cadmium) that
are products of corrosion in the distribution system.

          When   the  EPA  published proposed  amendments  to the  National
Interim Primary  Drinking Water Regulations  on  July 19,  1979  (6),  comments
were solicited on the  applicability and limitations of several corrosivity and
stability  indices to  assess  whether a particular water  is  corrosive.  The
monitoring of water  quality parameters  on-board the mobile laboratory was
designed  to procure data  for  evaluation  of  such  indices for the types of
water sources within the study area.

-------
                                SECTION 2

                               CONCLUSIONS
     1.   The  EPA  mobile  water  quality  monitoring  laboratory  was  an
effective tool for  concurrently collecting- data on a broad spectrum  of water
quality parameters.  Most certainly  the  quantity  of  data collected could not
have  been  possible without  the  continuous  monitoring  systems  and  the
on-board  minicomputer.  Such a  capability  is  very helpful in defining the
characteristics of  the distribution system with  respect to both water quality
and diurnal flow patterns.

     2.   During  the course of this project, which  was the  first extensive
investigation  employing the  EPA  mobile  laboratory,  a number of improve-
ments  and/or  modifications were  made  on the monitoring  systems.   These
changes were  made in order to  either  overcome problems  experienced in
actual  use  of  the laboratory  or to  save time both in servicing the instru-
ments and in analyzing  the data. The modifications were:

          (a)  Adding  zinc  and copper sensors  to  the  corrosion rate
     monitoring system.  A constant head flow control was  also  added
     to maintain a uniform flow in the corrosion cells.

          (b)  Automating   the   calcium   carbonate  deposition   test
     (CCDT) to allow for continuous  collection of data.

          (c)  Adding  a turbine  flow  meter which  was  then used  to
     regulate the  total water flow into the laboratory.

          (d)  Reprogramming  the  computer's  data  acquisition worker
     task  to  record  additional   parameters  by   the  computer.   The
     program  modification  also  allowed   for  greater   flexibility   in
     changing the time  sequence for  data collection.

     3.   The  waters studied at 30 field locations on this  project could be
classified as  having moderate to high  hardness, alkalinity  and total dis-
solved  solids content.   In general,  the  pH  was  near the saturation value
and  free  carbon  dioxide  was low  to moderate.   Attempts  were  made to
establish  interrelations  between   the   several  water  quality   parameters
measured,  calculated saturation and corrosion indices,  and  the  results of
intantaneous polarization corrosion  rate measurements on day 7.  For the
waters investigated the following relations were observed:

-------
          (a)   Corrosion  rates   for  mild  steel  and  zinc   decreased
     moderately as the Langelier saturation index shifted from negative
     to positive.

          (b)   A  slight trend toward higher  corrosion  rates for mild
     steel and zinc was noted as the Ryzner  index increased from 6.9
     to 8.8.

          (c)   There  was a  slight trend  toward lower corrosion rates
     for zinc as the calcium content increased.

          (d)   No  observable trend  toward higher corrosion rates for
     mild  steel and zinc were noted  as  the conductivity  of  the  waters
     increased.    A slight trend  toward   higher  corrosion  rates  for
     copper was noted at higher  conductivities.

          (e)   The more  complex problem  of  galvanic  corrosion could
     not  be studied by the polarization corrosion  rate  meter on board
     the mobile laboratory.

     4.    The  calcium carbonate  deposition test proved  to  be  one  of  the
most  sensitive  tests  performed by  the  mobile  laboratory.   It responded
immediately  to changes in water quality;  however,  the significance of this
response  was  not always evident.  When  the two MWD water sources were
blended in  varying ratios,  there  was  a  progressive delay  in the onset of
film  formation  as the  calcium hardness in  the blend decreased.  Once  the
film  began  to form the CCDT slopes were not markedly  different.  Diurnal
changes in  water quality due to in-line mixing of groundwater with the MWD
water which caused marked increases in alkalinity  and/or calcium hardness
were readily detected by the CCDT meter.

     5.    The  CCDT meter appears to  provide evidence  of  the  formation of
porous  or  discontinuous  films  deposited  on the rotating gold  sensor.   In
some  cases  the  microcurrent  flow did not drop to zero on long  exposure to
the water.  In  another,  the  current flow increased  after a  water of lower
alkalinity  reached the  laboratory.  We believe this  indicates  partial dissolu-
tion   of  the  original  film  to  satisfy  new  equilibrium  conditions.   Such
information  is of  value  for assessing the impact of water quality changes  on
the stability of pipeline deposits.

     6.    Interrelations between the  CCDT  data  and  other water quality
data were explored.   For the waters  not treated with zinc phosphate in this
investigation,  the following relations were observed:

          (a)   Because  of the  scatter  of points,  there  was no firm
     evidence of  an inverse  relation  between  the polarization corrosion
     rate for mild steel and CCDT values..

          (b)  A  moderate   inverse  trend was  noted  between  zinc
     corrosion rates and CCDT values.

-------
          (c)  A strong positive trend toward higher CCDT values was
     observed  as  the calcium  content  and  alkalinity  of  the  waters
     increased.

          (d)  There  was  only  slight  evidence  of  a  direct  relation
     between the  CCDT values and the  Langelier saturation index.  We
     believe  this is because  all of the waters studied contain moderate
     to high levels  of bicarbonates and calcium hardness.

          (e)  Lower  CCDT  values were associated with higher values
     for the  Ryzner index.  This should be  expected since the higher
     Ryzner  index  indicates  an  increase  in  the  calcium  carbonate
     dissolving  tendency,  the opposite  of  the CCDT representing  a
     deposition rate.

     7.   The  addition of  zinc orthophosphate  as a corrosion  inhibitor  in
three  of  the  water systems  monitored during  this project caused the most
rapid  film formation rate to  be recorded by  the CCDT meter.   There was,
however,  no consistent evidence of an  improvement in the day 7 corrosion
rates measured of the polarization method in the mobile laboratory. Corrosion
coupon tests  performed by others were not  always supportive of the need
for adding this chemical to these waters. However, such application appears
to be  most beneficial to bridge a transition period when shifting from  well
water  to  imported  surface water.   Because of  the  ambivalence concerning
the  beneficial effects of  this treatment  and  its  cost  effectiveness,  recom-
mendations  for  studying  the  possibility  of  intermittent  feeding  of  the
orthophosphate have been made.

     8.   In  one   case,  the  diurnal  shifting  from  one  water  source  to
another  resulted  in higher  corrosion rates   during the  period when  the
surface water was  blended with  the well water  than  when either water was
delivered alone.  The continuous  monitoring systems on  the mobile laboratory
provide a means for diagnosing problems  which may evolve from this type  of
operation.  The  data would represent  qualitatively a  directional tendency
toward more  or less  calcium  carbonate  film  formation  and higher OF lower
corrosion rates, but the data would not  quantify these differences. Possibly
by  the  use  of weight-loss  coupon test methods  in  conjunction  with the
laboratory tests,  quantification  of the magnitude  of  these effects could be
achieved.

     9.   On  the   basis of this  study,  the  blending  of imported surface
waters and their mixing with local groundwaters in  the southern California
communities served  by The Metropolitan Water District of Southern California
has not generated serious  water quality problems in the  distributions systems
of the  area.

-------
                                SECTION 3

                            RECOMMENDATIONS
     1.   The   EPA   mobile  laboratory  should  be  deployed  in  several
communities  across  the country  where  each is serving  water representing
one of the  many  types of  water being used in public  water supplies.  The
laboratory could  then provide a  data base for evaluating  those  parameters
needed for determining  the  stability or  corrosiveness  of waters.   Since  no
single  parameter seems to be reliable for assessing  these  quality characteris-
tics,  a  combination  of well-defined parameters  might  be  selected on  the
basis of concurrent measurements taken on-board the mobile laboratory.

     2.   For systems where water quality  changes due to  blending, mixing
or storage  result in water quality deterioration, the mobile laboratory is  an
excellent  diagnostic tool.   It can  be used  to identify changes  and to  study
the efficacy of  modifying operations,  including treatment,  to  improve water
quality.

     3.   When  the  mobile  laboratory is to be used for investigating corrosion
rates,  a protocol should be  adopted to place a set of mild steel, zinc, and
copper electrodes for  advance exposure  to  the  water  to be tested some
three or four weeks before the laboratory  will arrive at this  location.  The
water  being  tested  should  flow  continuously through  the  cells  at pre-
determined  velocities  during  this  conditioning  period.   A second  set  of
freshly cleaned  electrodes should be installed in  a flow cell  on-board  the
laboratory,  and  corrosion  rate measurements made on each  set will provide
comparative  data for  immediate  and  long-term  passivated  electrodes,  the
latter  approaching  equilibrium corrosion  rate measurements.   This  would
require modifying the corrosion  rate meter to handle more than  the four
stations presently available in the instrument.

     4.   Further study  is  needed  to  determine factors related  to  the
reported  effectiveness  of zinc  phosphate  as  a  corrosion  inhibitor'.  The
possibility  of  intermittant  feeding  of  the  chemical to  a  previously zinc
phosphate passivated  system should be  investigated as  a  means of making
the process more cost effective.

     5.   The  mobile  laboratory with its 18 integrated, computer-controlled,
parametric systems  should  be used as a research tool rather than for routine
surveillance activities.  This  recommendation is based on  the rather rigorous
siting  requirements  to  provide sufficient water, power,  and water  disposal
facilities for proper laboratory operation.   The time required  for setting  up
and  recalibration of  the  instruments at each  new location mitigate against
brief stops.

-------
     6.   Depending on the objectives  of the research project,  some of the
more  sophisticated  monitoring  systems  may  not  be  required  as  in  the
current project.  We believe that to operate the laboratory at its designed
monitoring capacity, two chemists familiar with all the chemical instrumenta-
tion  on-board are needed and  that  one or both chemist operators should be
familiar with electronics and the operation of minicomputers.
                                     10

-------
                                SECTION 4

      THE EPA MOBILE WATER QUALITY MONITORING LABORATORY


GENERAL

          The EPA mobile water quality monitoring laboratory was designed,
assembled, and initially  tested  by the National  Sanitation Foundation (NSF)
in Ann  Arbor under  a contract from the EPA,  "Water Quality Monitoring in
Distribution  Systems."   (5)  Exterior   and  interior  photographs  of  the
laboratory are  shown in  Figures 3a and 3b,  respectively.   It  was  the
purpose  of  the  EPA funded  project  to  develop  and  test  a variety  of
automated analytical techniques which could be  used to detect and measure
changes  in  quality which may  occur in  potable water during transmission
through distribution systems.

          The NSF had developed for the EPA a prototype mobile laboratory
for monitoring drinking water  quality with 18 integrated, computer-controlled,
parametric systems installed in  the mobile laboratory.   The 18 sensor systems
as developed by  the NSF and  in some cases  modified by  the  EPA Research
Center in Cincinnati,  Ohio, are listed in Table  1.  The  calcium system was
added  by the EPA  (7).   The  monitoring systems as  well as  the  complete
mobile laboratory  are described in detail in the  NSF project report (5).

          MWD  received  the  EPA  mobile laboratory  in October 1976.   The
unit  was  shipped  to  MWD from EPA's  Water  Supply Research Center  in
Cincinnati, Ohio,  lashed  onto a lowboy trailer.  On arrival at MWD's Central
Laboratory  in  La Verne,  California, a number  of nuts  and  bolts had
separated and several soldered connections in the monitoring equipment had
broken  as a result of road vibrations  during shipping.   Considerable time
was spent locating the problems  and making the necessary repairs.

          After  observing the  problems  caused during transportation  to
California, it was decided to  install some additional cushioning on several of
the mounted instruments.  While  working  with the  laboratory for several
weeks  during the initial training  and learning period,  it  became obvious
that  if  the  laboratory were going  to  be  moved often,  some  modification
needed  to be made that  would shorten  the  setup and takedown  times.  A
number  of  modifications   were  made  to instruments which  were  normally
shipped  and stored in boxes  and  had not been permanently  mounted.  As
much   as  feasible,   these  instruments   were  permanently  mounted   with
cushioning to available counter and wall space.
                                    11

-------
Figure 3a. Exterior photographs of the EPA mobile water quality monitoring laboratory.
                                        12

-------
Figure 3b. Interior photographs of the EPA mobile water quality monitoring laboratory.
                                     13

-------
        TABLE 1.   SENSORS IN THE E.P.A, WATER QUALITY
             MONITORING LABORATORY WHEN RECEIVED
Parameter
Sensor Type
Unmodified
Commercially
Available
System
Temperature
Conductivity
pH
Chloride
Dissolved
Oxygen
Thermistor
A-C Conductivity Cell
Glass Electrode
Solid State Ion-Selective
Electrode
Voltammetric Electrode
X
X
X
X
X
Free Residual
Chlorine

Total Residual
Chlorine

Turbidity

Corrosion Rate


Free Fluoride


Total Fluoride


Alkalinity


Hardness


Nitrate


Calcium


Cadmium


Lead

Copper

Calcium
Carbonate
Deposition  Test
Galvanic Cell
Galvanic Cell
Nephelometer

Polarization Admittance
Technique

Solid State Ion-Selective
Electrode

Solid State Ion-Selective
Electrode

Potentiometric Combination-
pH Electrode

Liquid Junction Ion-Selective
Electrode

Liquid Junction Ion-Selective
Electrode

Liquid Junction Ion-Selective
Electrode

Differential Anodic Stripping
Voltammetry (DASV)

DASV

DASV

Potentiostatic Rotating
Ring Disc Electrode
     x


     x

     X
                             14

-------
          During  the  initial  training and learning period considerable  time
was  spent attempting to get the  ion-selective  electrode monitoring systems
to operate correctly (Ca   ,  Hardness,  NO~  free F~, total  F~,  alkalinity).
There  were  continual problems  in obtaining good data because  of troubles
with the sampling and recording systems.   It became  obvious that the effort
expended in relation  to  the reliability of  the  information obtained was not
justified.   Too much time  was needed  for  operating  these systems and the
resultant  data was not completely reliable.  Thus,  it was  decided not  to
operate these  systems  and to perform manual  titrations for  calcium, hard-
ness, and alkalinity using  samples taken by the 24-bottle sampler.  Nitrate
and  the  fluoride species  were not monitored because  these species were not
considered to be  of prime  importance for the purposes of the study.  Using
the  data  acquisition   regime for  the  ion-selective  electrode  systems   as
controlled  by  the  computer,  one  analysis  value was  obtained  every  22
minutes.  Using  the 24-bottler sampler, a  sample  can be taken every hour.
This sampling  regime  gives fewer data points;  however,  it was felt that the
effort put into obtaining the  data and the reliability of the  data  were  in a
more nearly optimal ratio.

          The  anodic  stripping voltammetry  (ASV)  system  for  trace metal
analysis  was not used  during this study for two reasons.  Firstly,  the ASV
system was not functional when the mobile  laboratory was received by MWD.
The  ASV unit  in the  laboratory was  a newly designed instrument that had
not  been  thoroughly  tested or used.   Secondly, most  of  the  distribution
system  piping in  the  areas  where  the  water  was monitored was  made  of
materials  that would not add significant amounts of the ASV  detectable trace
metals:  lead, cadmium,  and  copper.    Samples  for trace  metals  were taken
periodically   at   each   location  and   analyzed  by  atomic   absorption
spectroscopy.

IMPROVEMENTS TO THE MOBILE LABORATORY

Dual Power Hook-Up Capability

          The  mobile laboratory was initially designed to be  operated using
a 240/220-volt  20-amp electrical  service.  The electricity  was  supplied to the
mobile  laboratory by  a cable attached to  an appropriate  electrical  source.
In selecting field  monitoring  sites for  the mobile laboratory,  it  was deter-
mined that  a number of the  sites chosen  did not have the  needed 240-volt
electrical service,  but  rather a 480-volt  supply.  A  10-kVA  480-to-240-volt
transformer  was  instaEed  on board the  laboratory.   A  special disconnect
box, Figure 4,  was also installed which was set up so that either a 480-volt
or a 240-volt supply could be used as the electrical service for the mobile
laboratory,  depending on which  way the lever was thrown.   Each of  the two
voltage  supplies  has a different type  of  cable  connector  attached to pigtail
cables   from  the  disconnect  box.  This  was  done  to  eliminate  possible
confusion regarding which voltage is connected to which input cable.

Air Conditioning and Temperature  Control

          The  mobile laboratory was initially designed with  one  12,000 BTU
roof mounted air-conditioner. During  days when  the mobile  laboratory was

                                     15

-------
     BREAKER BOX
      SQUARE "D"
       CAT. 82252
H
cor
480 VOLTS
1 PHASE
UOMIviUN —
OROUnlD —
GROUND

240 VOLTS
1 PHASE "
cot
CRO
A
M
LJBBELL
JNECTOR
rt_
^
O—
— •
~

o_


-v-u_
\s






TRANSFORMER
G.E.-9T2181009
10kVA60Hz





O
O
--0
k
P-




	 <





I











C
^
C







BLK.
f-
! c
5 C
BLK.


RED
I—

•)
RED




J 	
I


BREAKEF
METER BOX
\Sj LAB

F
gNECTOR
JSE-HINDS
PJ3485
OD. M54
                                                                        120
                                                                      V.A.C.
Figure 4, Electrical Service Disconnect Box.

-------
sitting  in  a  sunny  location,  the temperature inside the  laboratory would
commonly reach 85°-95° F,  even with the air-conditioner on maximum cool.
Even  though the mobile laboratory is  relatively small, the dark blue exterior
absorbs  great  amounts of heat from  the sun.  Added to this is a consider-
able amount  of heat  from the  instruments  and computer.  Not only was this
extremely  uncomfortable  for  the operators,  it was a  potentially  serious
hazard  to  the  computer.  On hot days the temperature on top of the central
processing unit would climb  to  115°-120°  F.   The  computer manufacturer
recommends  72° F  as the optimum operating temperature for the computer.
To  remedy overheating  situations,  a  second  roof-mounted  air-conditioner
was installed (Duotherm 13,500 BTU).

          The  large  amount of  window space in  the  front of  the  mobile
laboratory  admits  heat  from  the sun  which is  further intensified by  the
greenhouse effect.   Cardboard inserts  painted white were made and used to
cover  each window thereby eliminating this source  of heat.  These  covers
were  either  set in place  or held over  the  windows with tape and  could  be
easily   installed  and removed when  changing  locations.   In  addition  to
reducing heat  due  to sun,  these window coverings added to the security of
the laboratory.  It was  almost  impossible  to  see into the laboratory when
they  were  in place, and  the amount and kinds  of equipment contained inside
were not revealed.

Modifications to Existing  Monitoring Systems

Schneider Robot Monitor —
          Cup  inserts were made that  could be filled with water  and placed
into the monitor's flow  cells.   This way the electrodes could be kept in
water  during  transit,  eliminating problems that  might occur as a result of
the electrode sensors drying out.

Residual Chlorine Analyzers--
          The   pump  used to  supply  standard  chlorine solutions  when
calibrating the residual  chlorine  analyzers  was permanently mounted  to  the
outside  base   of  the   Schneider  Robot  Monitor   (SRM).   In  addition,
polyethylene tubing  was  installed permanently from the pump  to  the input
line  to  the chlorine  analyzers.  A shutoff valve  at this  point allowed for a
choice of either the standard solution or the  tap water to be analyzed.

Calcium  Carbonate Deposition Test —
          The   calcium   carbonate  deposition   test  (CCDT)   rotator   and
potentiostat  were permanently  mounted to  available  bench  and shelf space
respectively,  on the left side (as facing forward) of the laboratory.  These
units  had  not been previously mounted,  which  would  have necessitated
boxing  the instruments  for each move.  The  CCDT system when received
was  a completely manual  operation, that is, the test had to be started  and
stopped manually.  The  data,  which  was recorded on a  strip chart,  had to
be  manually   transposed to  tabular  form  so that data cards  could  be
punched.  The data from  the cards were  then  merged  with the  data from
the on-board computer so  that all the  laboratory analysis  data  could be in
computer storage.  A modification was  made in the  data  acquisition program
of the  on-board computer that allowed for the  computer to  record the CCDT


                                     17

-------
data.  This modification  was started with Location  6.  Near the end  of  the
project  the  complete  CCDT  analysis  cycle  was  automated  with  computer
control.  A more complete description of the computer modifications and data
handling will follow in subsequent sections of this report.

Corrosion Rate Meter --
          The  mobile laboratory came  equipped with a Petrolite four-station
corrosion  rate  meter (CRM); however, no provision had been made for flow
cells  in  which to  place  the electrodes which  measure the  corrosion rate.
Thus, before any field monitoring' was started flow  cells, shown in Figure 5,
had  to be  constructed and  proper cables,  electrodes,  and electrode holders
had  to  be  purchased  from  Petrolite.  A rack to accomodate  three flow  cell
chambers  was  made; one cell held the original mild steel electrodes supplied
with  the  laboratory  while  the  other  two  cells  were  added  to  house
sets of  zinc and copper  electrodes.  A special rack for securely holding  the
CRM  was  constructed and mounted in unused space under the bench on  the
right side  of the laboratory.   The unit is  easily removed  from  the rack  for
necessary access and servicing.

          Initially,  the flow  to the  electrode cells was  set  by adjusting  the
small globe valve  on  the water  source line, and  maintenance  of flow  was
attempted  using visual  observation  of the  output overflow from the flow
cells.  After  some  time  it was observed  that  this  type of  flow adjustment
was  quite   inaccurate  and  that  the flow to the cells  varied daily  due to
changes  in the source line  flow  and pressure and  poor flow control by  the
globe valve.   A flow  meter was  installed in the feed line to the  corrosion
cells  and the  total  flow set at 0.5 L/min for  the three cells.  Observing  the
flow meter  over a  period of several  days  confirmed the fact that the flow
changed periodically.  Several constant-flow restriction valves  were  tried,
but with no success.

          It also  appeared  that  the corrosion  rate  measurements were flow
dependent  and needed a better and  more accurate flow regulation system.
A constant-head flow regulation  device  was  designed, which had an  adjust-
able head  level that could be used  to adjust the flow (Figure 5).  This  was
installed and initially put in service at sampling Location 13.  The  constant-
head device for maintaining a constant flow proved to be very satisfactory.

24-Bottle Sampler--
          The   24-bottle  sampler  (Sigmamotor)  supplied   with  the  mobile
laboratory  was designed to  take a sample  only  upon  a  signal from some
external source.  Originally  this  external source was  the  computer.   There
was  a  separate worker  task  program  written  for the computer that   was
supposed to activate the sampler every hour  so  that it would take a sample.
However,  there was no  documentation information  available on exactly  how
to operate  the computer so that it  would take the sample.  Periodically  the
computer  would activate  the sampler,  but the time interval between samples
was  not always consistent  and sometimes there would be  no samples taken
for  a whole day.   The  sampler could not be used in this manner  and an
alternative  had to  be  found.  The sampler  was  taken  off  computer control
and  connected to  a 24-hour timer  with adjustable event control (Paragon
Model 25001-OS Program Time  Switch).  The new  timer can  be set to take


                                     18

-------
 OVERFLOW
DRAIN
                                    CONSTANT HEAD
                                  TANK-ADJUSTABLE
                                       HEIGHT
                                   2"PLEXIGLAS
                                     TUBING
ELECTRICAL
CONNECTOR
   TYP.
                               WALL BRACKET
                          ELECTRICAL
                           CABLE TO
                          CORROSION
                          METER, TYP.
   INFLUENT
                 FLOWMETER
                                 1TYGON TUBING
                                     TYP.
                                 2%  PLEXIGLAS
                                    TUBING
                                                        MILD
                                                       STEEL
                                                     CHANNEL 1
                           Figure 5, Schematic diagram of corrosion flow cells with constant-head flow regulation.
                                                                                                     DRAIN

-------
samples  every  hour or  in any desired  sequence with  5-minute minimum
intervals.  It has proved to  be very  reliable.  This type of control for the
sampler  is  very  versatile,   while  the  computer had  a  fixed  sampling
frequency.   Over- normal weekends, samples were taken every 3 hours and
for 3-day  weekends samples were taken every  4  hours.  In either case, 24
samples  are  taken over  the  weekend.  At locations where the distribution
water  is  a  blend and  the blend  varies  during high-use  periods  in the
morning  and  evening,  samples  can be taken  every 15 or  30 minutes,  and
during other hours samples can be taken every several hours.

Monitoring Equipment Added By MWD

HF Turbidimeter--
          The  Hach CR   low  range  turbidimeter,  Model  1720,  that  was
supplied with  the mobile laboratory had several  problems.  Firstly, it was
observed that the calibration  procedure,  using the standard reflectance rod
supplied with the unit, was not an  adequate means of calibration.  The unit
is calibrated on  the 0-5  scale  while the measurements  are made  on the 0-1
scale.   When samples  were taken into the main laboratory,  the  results did
not  agree,  even after  calibration.   It was found that the  best way of
calibrating the  Hach unit was  to take a sample into the main laboratory and
take  a  turbidity reading,  and  then  adjust  the mobile  laboratory unit
accordingly.

          This means  of  calibrating worked  well  when  the mobile laboratory
was  stationed  near the  central laboratory,  but  was  impractical when field
monitoring.  It  was therefore  decided to install a second turbidimeter,  a HF
Model DRT 200, into the mobile laboratory.  This unit  has  the flexibility of
being used for continuous monitoring  or  for discrete samples, and is easily
standardized with a standard reference solution vial.

          Since the HF unit had the  capability of  continuous monitoring,  it
was  decided  to record  turbidity  from  both  the Hach  and the HF  units.
Starting with Location 7,  the  HF turbidity  was  recorded on the computer.
Recorded data  from  the  two  turbidimeters allows  for   a  comparison of the
performances of the two units.

          The HF unit has another advantage over the Hach—the scale and
output  are  linear and  thus  easily converted  to" turbidity.  On the other
hand,  the scale and output from the Hach unit are non-linear and turbidity
must be interpolated  from a table  (turbidity units  (TU) vs.  millivolts)
stored in the computer.

Turbine Flowmeter--
          During the  early stages  of  the project  it was difficult to  maintain
a  constant water flow to the mobile laboratory. The flow would vary as the
line  pressure  in  the  distribution system  varied.  A flow  regulating valve
that was supposed to maintain a constant flow with changes in line pressure
was  tried, but  it did not keep  the flow constant.  In September  1977  it was
decided  that  a  flowmeter  should be installed that could be used to monitor
the  total  flow  into  the  laboratory.  In  this   way,   there  would  be  a
continuous  record of  the flow, which  could be used to  ascertain if any

                                     20

-------
parameters were flow-dependent.  A  turbine-type  flowmeter  (CE in-Val-Co
Model W3/0750, 1-10 gal/min), with analog readout and output (CE in-Val-Co
Model  531-1C  frequency to  analog1  converter) was  installed in  the  main
influent  line  to  the mobile laboratory downstream of the totalizing water
meter.  The  output from the  flowmeter  has been  logged by  the  computer
since Location 13.

MOBILE LABORATORY COMPUTER SYSTEM

General

          The mobile laboratory  is equipped with  an  on-board minicomputer
system designed to log  the  data output from the various  monitoring systems,
and  control  the  operations of some of the  monitors.   The computer system
includes  a Texas  Instruments (TI)  digital minicomputer,  Model 960-A;  a TI
Silent 700 teleprinter;  a Remex  combination high-speed  paper tape reader/
perforator;  and  a Computer  Products programmable  wide-range analog-to-
digital (A/D) converter, Model RTP 7480.

          The computer contains 24K  of  semi-conductor  core  memory with a
battery backup  that will maintain the core memory  for several hours during
a power  failure or when the mobile laboratory  is moved  from one location to
another  and  the  power must be  shut  off.   This eliminates the  need to
completely reprogram the computer  each  time the unit is moved.  During all
of the  MWD  field  work,   memory  was  never lost  while  moving  between
locations, even  with power off for periods  of  3 to 4 hours.   The  computer
has two 8-channel multiplexer boards mounted in the A/D converter for data
input and a  16-channel input/output digital  switching  card  for  computer
control operation.

Worker Tasks

          The original  computer  system as  set  up  and programmed by NSF
is described  in  greater detail in that project report  (5).  The original NSF
system, both hardware and software, had been modified by the EPA before
MWD  received  the grant for this study.  Within the  computer, all program
operations are under the control of the TI supervisor program PAM (Process
Automation  Monitor).   Worker  task  programs,  written  to  carry  out  the
functions  of  the computer,  are  installed  in  the computer and operate under
the control of PAM.  Ten  worker tasks had been written for  the  operation
of the mobile laboratory computer by NSF and/or  EPA.   The  seven worker
tasks used during this study are  as follows:

     1.    NSFC is a  program  containing  utility  routines, such  as a
          decimal core dump routine.

     2-    INIT  initializes output storage areas,  checks  punch (PUN)
          and  starts  the   data  acquisition  (DAQ).   It is  used   for
          start-up when the system is  initially loaded.
                                    21

-------
     3.   DAQ  is  the  basic, data acquisition program  which reads the
         various  analog input  channels  and  stores the digital values
         in a  data buffer.  DAQ contains the gains to be  used by the
         A/D converter when reading each analog input channel.

     4.   DATA obtains data from the DAQ  buffer  area and stores  it
         in an output storage area.

     5.   PUN  outputs  data  from the  output storage area of DATA and
         clears  that  area  for  fresh data.   DATA  unsuspends  PUN
         wherever more data is put  into the output storage area and
         suspends PUN when  the data  output is complete.  Data can
         be output to either paper tape or the printer.

     6.   CQNTAB has  calibration  data  for  a  number of  monitoring
         systems.  The free and total residual chlorine were  the only
         parts of this program used for this study.   This calibration
         data  is  used by the  CONMN  worker  task.   The  calibration
         data  for the  chlorine  system is the slope  and intercept from
         a calibration curve.

     7-   CONMN  is a conversion program  which takes the latest data
         from  output buffer,  calculates values for the parameters  in
         engineering  units,  and generates  a report that  is output  to
         the printer.

The  DAQ Worker  Task

         The  DAQ  (data acquisition) worker  task is  the program which
controls  the  gathering  of data from the  various monitoring systems.  It
controls what  data is collected  and the time sequence between  data samples.
The  DAQ  task  was originally written by  NSF and later modified by EPA.  It
was  further  modified  by  the MWD  project staff  in order to increase the
efficiency of  data collection  and to  improve the monitoring capability of the
computer.

         The  computer monitoring system  was used as  is for the first five
monitoring  sites.   During this  time  a  number of inherent problems  were
observed in the  system.   The original DAQ program was  set up to read 16
channels of input and output  23  different pieces  of  information, that is,
several  input channels  were used to gather more than one piece of informa-
tion.  The  basic  sampling  cycle was over 21 minutes and samples  of  various
parameters were  made at the end of each wait period in  a  sequence of 4, 6,
4,  7 minutes.   During  these   21  minutes,  41  readings  were  taken  and
punched onto  the paper tape.   However,  9 of the 41 punched readings were
for  parameters that  were no  longer used, but  the computer still  gave  a
print out  indicating  no  data  for  them.   Under  this sampling  regime  a
thousand-foot  roll of paper  tape for  the output would last a maximum  of 34
hours.   Considering  that  operators  were  at the  mobile laboratory only
during normal  working hours, this necessitated  servicing the computer once
a  day in  order  to change  the  paper tape.   Such  daily attention required
considerable overtime on weekends, however,  it was not practical since the


                                    22

-------
field  sampling  for the  project was to last for a year.  There were  two
alternatives,  first was  not to sample over  weekends, or second  to change
the sampling cycle and regime.  Since it is desirable to have uninterrupted
data in order to observe possible weekend  quality  changes the only choice
was to change  the computer sampling procedure.  This meant that the DAQ
worker task would have to be modified.

          Since  the MWD  project staff had only a basic amount of computer
training  and no  specific  experience on this TI  system,  it was  difficult to
reprogram  the  computer  after it  was decided  that  changes would improve
the utility of the mobile laboratory.   The  computer system came from the
EPA without  a  compiler or  assembler and  thus  all programming changes  had
to be  made at  the machine code level.   The  first  modifications  that  were
made  altered the sampling sequence and eliminated extraneous data from the
paper  tape  output.   With  the altered DAQ  worker task, readings for only
the parameters of interest  are taken and  punched  on the paper tape  every
10 minutes.  With  8 parameters being monitored, this  extended the length of
time that a roll of paper tape ran  to 68 hours and eliminated  the need for
daily  servicing of the computer.  The mobile laboratory could  then operate
unattended  over  a  normal 2-day  weekend.  The  procedure  to  make  the
necessary  changes (patches  in  the original DAQ  program) was relatively
easy, although somewhat troublesome and  time  consuming.  The  patches  had
to  be manually  entered into  memory  via the  teleprinter whenever  the
computer needed to be reprogrammed.

          After  spending considerable  time  studying the  computer and its
programming,  a  technique for generating  new  worker  task  programming
tapes  was developed and implemented.   Any  new  changes in worker task
programs need only to be entered into core  memory once, after which a new
programming tape is made.

          As the  project progressed, additional  changes were made in the
DAQ  program.   One  such  change  simplified  the  operation  to alter  the
sampling  cycle  time.   A  single number,  entered  into a  specific memory
location  via the teleprinter  is presently the  only operation needed to change
the  sampling  cycle  time.   After  this  DAQ program  change was made  the
standard  sampling  cycle was  10 minutes,  however,  it became convenient to
change to a  12 minute  cycle  over standard  2-day weekends  and appro-
priately  lengthen  the  cycle  to  14 or  more  minutes  for a  3- or  4-day
weekend.

          A second change  in the DAQ worker  task was made that increased
the utility of the  computer by  enabling it to  record  more data.   A number
of input channels  were made  available  because  the  ASV  and ion-selective
electrode monitoring  systems  were  not being  used.   Some of  these  spare
channels  were  put to  use recording data  from  other  monitoring systems.
Since not all input channels were read in the same sampling sequence,  major
revisions in DAQ  as  well as some minor revision  in the DATA program were
needed in  order to make this change.  The  following  systems were added to
the computer for data recording:
                                    23

-------
     1)   CCDT on channel 1, starting with sample Location 6

     2)   HF turbidity on channel 2, starting with Location 7

     3)   Corrosion rates on  channel 11, starting with Location 11

     4)   Flow  to the lab on channel 12, starting with Location 12.

         The  two  monitoring systems, CCDT and corrosion rate,  were not
being recorded by the computer when MWD received the mobile laboratory.
Raw data from  these two monitors  were recorded on  strip charts from which
the final data  had to  be manually  calculated.   This  data was then recorded
on  keypunch  transmittal forms  from which  data  cards were  subsequently
punched.   This  punched-card  data could then  be computer-analyzed  and
stored as with  the paper tape data.  The other two monitoring  systems,  HF
turbidity and  flow, were added to  the  monitoring  capability of the  mobile
laboratory by MWD  (discussed earlier in greater detail).

         For  three of  the  monitoring systems,  CCDT,  HF turbidity,  and
flow, the connections for computer recording were easy to make since no
special alterations in the instruments were necessary.  All three instruments
had continuous analog outputs which were put directly to the inputs  of the
computer's  analog-to-digital converter.

         Even  though continuously  recording the CCDT  by computer posed
no  special  problem, measurements were  still  not  continuous, because when
the test  finished, the instrument had to  be manually stopped and restarted.
This was to  allow cleaning of the gold disc electrode with dilute HC1.  Near
the  end  of  the project,  the CCDT was  completely automated.  The DAQ
worker task  was reprogrammed to control the acid washing of the electrode
by  controlling  the operation of several solenoid valves and a pump  for the
acid.  (See Appendix  A for program listing and flowchart for DAQ MWD3.)
This  CCDT  automation  modification was  put on  line at  Location 27 about
halfway through the monitoring period.

          The  primary items needed  for automating the CCDT were:  a small
DC  pump  (Micropump Corp.,  Paragon Division) for pumping  the electrode
cleaning  acid to  and from the cell; four solenoid valves (Skinner Valve Co.)
for switching flows to the cell; and a 24VDC  power supply for operating the
DC pump.   Because the  pump operates on DC power,  the direction  of  acid
flow can be changed by reversing  the  polarity of  the  DC  source voltage.
Since  250  ml  of  dilute  HC1  are  required  for  each  wash, changing the
direction of  flow was  a  necessity  for reducing the amount of HC1  consumed
during the cleaning operation.  With the  reversible pump,  the same acid can
be  used  repeatedly,  by pumping from a  reservoir to  the electrode cell and
back to  the  reservoir, with a great savings in chemical consumption.  The
solenoid  valves are used to start  and stop the test water flow  to the CCDT
cell,  drain  the cell, start  and stop the acid feed,  and to bypass the  acid
reservoir.   The bypass  valve is  necessary so  that the line and  pump can be
backwashed  with tap water  in order to eliminate residuals of the dilute  HC1
that  might  damage the  pump.   The  acid feed  solenoid valve  cannot be
backwashed, however, it is teflon lined and is not attacked by dilute HC1.


                                    24

-------
          The solenoid  valve  coils are activated by 110AC voltage and could
not be  operated  directly through the computer digital output (DO) card.
The  DO card  can only  be  used to switch DC  voltages up to 24V.   An
external relay  board  was  assembled with  11  miniature  24VDC relays  (10
SPST and 1 DPST) that could  be activated from the computer DO card.
These  relays are  then used  to  switch the  110VAC to  activate  the  solenoid
valves,  to  switch  the 24VDC  to start  and stop the acid  pump,  and to
change  the polarity  of the  DC  voltage  to the  pump (the DPST  relay).
Figure 6  shows a  schematic  diagram of  the  relay control  board  and  the
CCDT  automation setup.  It was  found that the coils of the  relays operated
by  the  computer DO  card  had  to be protected with  a diode.  Otherwise,
electronic noise from the relay coil operation would feed back  through  the
DO card to the computer and cause interrupts  that would halt the computer
so that it needed to be completely reprogrammed.

          Recording the  data from the  corrosion  rate  meter  (CRM)  by  the
computer  was  also no  serious  problem,  since  the analog output  from  the
CRM  was no  different  than  that from  the other instruments.  The  CRM
measures eight  different corrosion rates,  i.e., anodic and cathodic  corrosion
rates at four different  stations.   In normal operation  the CRM makes  each
of the  eight measurements  at a  given frequency  as  set  on  a timer on  the
CRM.   The main  problem arose  in trying  to  synchronize  the CRM timing
cycle with the computer data sampling cycle.   This  situation was resolved
by having the  computer take preference over  the internal clock for cycling
the CRM.

          The  CRM was  not  designed for  this type  of  operation  and  the
instrument had  to  be modified to accept external cycling  control.  It  was
observed  that the  CRM instrument cycles to the next  station whenever the
AC power  is turned off by  the main on/off switch.   The modification  was
made  by  installing a 24VDC relay  into  the  CRM  chasis  which  could be
controlled by the  computer and  used to  turn off  the main AC power to the
CRM,  thus  cycling  the  instrument.    One  relay  from  the  relay board
described earlier, Figure 6, was used for  the CRM cycling.

          After  the instrument  was properly modified, the DAQ worker task
had  to  be reprogrammed to  control  the  instrument  cycling  as  well as take
the readings.   The program  is  now written  so  that after all  the instruments
have  been  sampled the  CRM relay  is activated  for a short period which
turns off the AC  power, thus  advancing the  CRM to  the next  station.  It
should  be noted  that  the Petrolite Instruments Corp.   (manufacturer of the
CRM) approved of this modification  for  computer control of  the  unit  and
also  verified that  it would  not  damage the instrument or alter the measure-
ments.  The change in no way altered the instrument from  operating in the
standard  mode  without the computer.  When  the  CRM is not connected to
the computer,  or  if  the computer fails,  the internal  clock will control the
cycling operation.  In  other  words,  the  modification  is merely  a  computer
override.
                                    25

-------
                                                                    TERMINAL BOARD
|                    DRAIN           I
C.C.D.T. L*T~
INTERRUPT1^ *
•*
1* „ . . 	

87
—»-,nnn~*- —
Kc- 11
I 4 '"* 5
1 - /»»>
i
t <
a
>
h
_
	 1
^ Kc- 10


~^

O i
t 3 _
J^
L ^~
^ —






r
4 5
Kc- 1



















i
4 • *
+ T/C —
24 volt
regulated
power supply
AC AC
1 1


AC A
X. | Y
1
I AC j AC AC| AC

A
B
' ?
-)-
—
+
—
•f


4-
—
+
y
_
+
—
•f
—
+
—
+
—
-f
T
—
DI-8— •-
24 VDC
10x.
^^ DO — B
10*^^
^S po A
>DO-8
.
^\ QO 7
>OO-6
.
4N 00-5
4/
3
3
2
I
1
1
Connections
To Computer
1
                     Figure 6. Schematic diagram of CCDT automation relay control panel.
                                               26

-------
Computer Cooling Problems

          During  the hot  summer months,   until  the  installation  of the
second air-conditioner (discussed earlier),  the mobile laboratory would warm
up to extreme temperatures.  At the same time the computer would get very
hot,  in  excess  of the limits recommended by  the  manufacturer.   There was
no circulation within  the computer  cabinet and thus  no way to  vent the heat
generated by the various computer hardware components; primarily the CPU
and  the  paper tape  reader/perforator.  The  second air-conditioner helped
considerably  by decreasing  the  ambient air temperature around the computer
cabinet.   However, it was  decided that  this  was  not sufficient,  so  a small
circulating fan  was  installed inside the top  of  the computer  cabinet.  In
addition,  a portion of the cabinet top was cut out and replaced with  a filter
and  screen.  This further allowed  for  increased circulation  by  providing for
a vent  at the  top of the  cabinet  through which hot air could escape.   It
should be noted  that  even  though  the  computer  CPU  got extremely hot
(120° F) at times  during the day,  no problems were experienced that were
due  to  overheating.   The  computer   should  be  commended as it operated
extremely well with less  than  optimum operating conditions (large  changes in
temperature  and  bouncing  around in  a mobile laboratory)  and no problems
were experienced that could be related to these adverse conditions.
                                    27

-------
                               SECTION 5

                  DATA HANDLING AND  PRESENTATION
PAPER TAPE DATA

          The main source of raw data from the mobile laboratory monitoring'
systems  is  the paper  tape output  from  the on-board  computer.   As  was
stated in an earlier section of  this report, a considerable  volume  of paper
tape  data was  generated by the computer every day.   At first,  approxi-
mately 700  feet  and later,  after the  program changes,  approximately 350
feet of paper  tape was punched every 24 hours.  The  paper tape reader/
punch unit  does  not have a takeup reel for the punched  tape and  thus  this
tape would  merely pile up  on the floor in  front of the computer.   Since the
reader/punch unit has  only a small  wheel  for winding up tape, a  consider-
able amount  of time would  have to  be spent in rolling up the punched tape
by  hand.  The MWD  project staff designed,  had constructed,  and installed
on  the side  of the computer cabinet a paper tape winder that would handle
a whole  roll of paper  tape. The  winder  saves handling  time and  can be
used  to  wind up the tape  during the  day  and thus  eliminate the clutter of
tape on the floor.

          The  preparation  of tabulated listings of the raw data was to be
done  with  the IBM-370-138 computer in  MWD's  Data  Processing Center.
However, this  computer  is  not  able  to use paper tape  input.  This meant
that before  the mobile laboratory data on paper  tape could be analyzed it
had to  be  first transposed to magnetic  tape,  an  input  media which the
IBM-370  will accept.

          The  transposing of the paper tape data to magnetic tape  was done
by  MWD's  Centralized  Control  Section which has  a  Lockheed MAC-16 mini-
computer system  with both a paper  tape reader-punch and a magnetic tape
unit.   Since there  was  no existing conversion  program,  the Centralized
Control staff had to write the necessary program  to handle  the transposition
of the data between the two media.  The program was written such that the
data  is  stored on the magnetic  tape  in  large  blocks which minimizes the
amount of  tape needed to store the data from each location.  It was  written
to  be flexible  enough  to  handle various  data situations that might  arise,
such  as  being able  to eliminate the data  from  certain   parameters  on the
paper tape  that are invalid due to instrument  malfunctioning or some other
reason.

          A  major revision  to the transfer program was  necessary  when the
corrosion data was  added  to  the records  stored on paper tape.    All the
corrosion data was  punched on the paper tape  as parameter number 11,

                                    28

-------
however,  this data represented 8 different corrosion rates, i.e., anodic and
cathodic  corrosion  rates on the 4 CRM channels, mild steel, zinc, copper,
and  a  10-mil-per-year instrument  test  standard.   Each  of  these  eight
corrosion  rates had to be assigned new parameter numbers and in order to
carry out this parameter number reassigning properly,  the MAC-16 computer
had to be told for each paper tape  roll what number  to assign to the first
parameter  11 on  each roll.   After the first number is  known,  then each of
the other  seven  numbers  come  in  the same repeated  sequence.   Also,  the
recorded  value from each  CRM channel had to be multiplied by  an appro-
priate  sensitivity factor depending on how  the  sensitivities  were  set on  the
CRM for that channel.  Thus sensitivities  for each of the four CRM channels
(anodic and cathodic corrosion for a  given metal are on the same sensitivity)
also  had  to be  entered to  the MAC-16 before  the proper transfer of data
could be accomplished.

          Since   there  was  a  significant  possibility   for  error  in  either
recording  or  entering  all  the information needed for  the data  transfer
(beginning  channel . number and sensitivities)  a check  was added to  the
program  that would catch  errors.  Every fourth corrosion value  was either
the anodic or cathodic 10-mil-per-year  test standard.   Thus, an easy check
was  to make sure  that the  standard  points were  there and that they were
within  certain limits.  This  data check  was quite helpful in catching errors
that  would have otherwise caused  considerable  amount  of extra time  to
locate and correct.

RECORDER CHART DATA

          Throughout the  duration  of  the project  there  were  always a
number of parameters that  were not recorded by the computer and thus  not
on the paper tape.  At first there were more parameters  recorded manually
than recorded  by  the  computer, but  as the capabilities  of the computer
were  improved   and  expanded,  the  volume  of  manual  data decreased.
However,   there  have  always been  titration data and this could  not  be
recorded by the  computer.   As was stated earlier in the report,  the manually
recorded  data had  to be  listed on  data  transmittal forms,  from which  the
data  were punched onto computer cards.  The IBM-370 can  accept computer
card data as well as the magnetic tape data  generated from the  paper tapes.
The  card  data  from  each  monitoring location had to  be merged with  the
magnetic  tape file of the paper tape data, thus producing  one final magnetic
tape  file for that location.   Computer listings,  graphing, arid other analysis
could  be   done   on  all  of  the  data together  using  this file.  The  Data
Processing  Center  personnel  wrote  the programs necess'ary to  merge  the
card data  with the magnetic  tape data.

DATA PROCESSING

          The EPA had  developed a computer program (written in  Fortran)
that  would take  the paper tape data from  the mobile laboratory minicomputer
and  make both   a  printed   listing and a graphical plot  of the parameters
versus time. MWD's Data  Processing Center staff used this program as  the
basis for  the general data processing of the field data  for MWD's  monitoring
project, however,  it had to be greatly modified and expanded.  Some of  the

                                    29

-------
parameters for  which the EPA  program was  written  were not monitored in
the  MWD  study,  and  other  parameters  were  added  to  the  monitoring
capability  of  the mobile  laboratory.  The program  had  to  be modified to
compensate for these changes.

          The  program takes the number which  was recorded  on  the paper
tape  for  each  parameter  and  converts   it  to  the  actual  value  in  the
appropriate engineering units  and uses these  final values for the listings
and  plots.  The numbers on  the paper  tape are in  analog-to-digital units
(ADU), which  are the values from the analog-to-digital (A/D) converter and
are proportional to the  instrument signal voltage depending  on the gain of
the A/D converter when it  reads the instrument. The conversion program
must  include the A/D converter gains for each data channel plus a means of
converting ADU's to actual engineering units.

          For   converting   some  parameters,   simply  multiplying  by  a
conversion factor is  all  that is needed.   Other parameters require a slope
and  intercept  calculation from  a calibration  curve.   The signal  from the
Hach  turbidimeter  is  non-linear,  and thus turbidity  had to  be  calculated
from  a  table  stored  in  the computer.   The table  prepared by  the NSF
programmers was used at first, however it was not  completely satisfactory.
The  table  was a step rather than a  continuous function and was not set up
for turbidities  greater  than 0.65 turbidity units (TU).   The MWD  project
staff  and Data  Processing staff together made a  new turbidity table  for the
full range 0-1.0 TU  and calculated a best-fit equation for  a set of turbidity
values.  This  equation  was then used  to  convert the ADU  value from the
Hach  instrument to  turbidity units.

          The  data processing  program had  to be first modified to accept
magnetic  tape  input.  Later on  it was also modified  to merge  the card data
with  the magnetic  tape data, producing  the final magnetic  tape record of all
the data.  This final tape  was then used for the  converting, listing,  and
plotting.   The  program was also modified to allow for the omission of data
segments that have  been determined to be  bad and would  have otherwise
biased the results.

          The  plot program was  also modified to accept  the  changed data
situation.   Plotting  of  the data  versus  time  was  done  on  a  Houston
Instruments  Complot  DP-7  plotter  which   has  three pens for multicolored
plots.  All plots are made on graph paper rolls  with a 10 x 12 lines to the
major  grids,  which is well suited for  plotting  by  the  day with  one-hour
divisions.  The plotting format  was  changed  to  graph several  parameters in
different colors  on the  same sheet  and  to  add other pertinent information,
such  as location name, year monitored, and, on the time axis, actual calendar
dates.  It takes two different  graphs (A  and B) to include all the various
parameters  monitored.   Reduced-scale illustrations of each computer-plotted
graph for a 5-day period at Location  25 are shown in Figures 7 and 8.
                                     30

-------
          The data for parameters that are  taken on the 10-minute computer
data  sampling  sequence  are plotted  as  a continuous  curve.   The  pen is
lifted and  the curve stopped any time there is a pause  in data greater than
20 minutes.  Data taken  at  a less frequent rate is plotted on  a curve with
tick marks indicating actual data points.
                                     31

-------
                                  BURSfTK
    2,0
1 F-  1-3
                2.0


                2.0
                d.f
          1-8501C
                       |j
                            2.0
                       PS
                 20    =)
                         g
                         p
1.3     Eg  6.0
                            7.Q
                                              Figure 7.  Computer-plotted water quality data at Location 25 — Graph A.

-------
00
U)
                LOCATION *  25

                      300
                 51  15°
                 a!
BURBfiNK HELU NO.  U BUND (
                                                                                             1S76
                    •  ca
                      IOC
                £-85010
                                  3CO
                                   !0
                                   10
                                  200
                                                   H	H
                                                       i—r
                                                                                                                       ^^
                                       fPK.
                       234


            Figure 8.  Computer-plotted water quality data at Location 25 - Graph B.

-------
                                SECTION  6

                    EXPERIMENTAL FIELD MONITORING
MONITORING SITES

          The  mobile  laboratory  was deployed over  a period of 18 months,
January  1977  to  June  1978,  for field monitoring the  effects of blending
waters  of different quality  and origin.   Field  studies were  made  at 30
different  locations, representing many of  the  different  blends of waters
available  in  MWD's service area.   Six additional test  runs were  made at
MWD's Weymouth  filtration plant.  The list of  monitoring  locations and the
type  of  water monitored  is given in Table 2,  with comments pertaining to
the location or  the parameters monitored.

          The  severe  drought  conditions  in  California  in  1977  imposed
constraints on the selection of monitoring sites  because State project water
was  unavailable at  most locations where it was  normally  used.  Moreover,
none  was available for blending  with  Colorado  River water at the Weymouth
and  Diemer filtration plants from March 1977 to February 1978. Figures 9,
1'Q,,  and  11,  respectively,  show the  system  delivery  patterns during the
pre-drought  period prior to March  1977, the drought period, and  the post-
drought  period after March  1978.  The general location of sites monitored
during each period in relation to water distribution patterns is  shown in the
above figures; the  circled numbers  correspond  to  the locations listed in
Table 2.

Site Operating Conditions

          Potential  monitoring locations were reviewed  and selected  to meet
the operating  conditions  required for the mobile  laboratory.  Both MWD and
the  cooperating  water   agencies  had  to  provide sites with  the  following
utilities and conditions:

     1)   A  continuous  flow  of  the  water to  be monitored,  6 to  8
          gallons  per minute  (gpm), preferably 8 gpm  for  10  days  or
          longer.   The  6-gpm minimum flow is  needed to have suffi-
          cient velocity past the dissolved oxygen electrode.

     2)   Power   supply   of 20   amp at  240 VAC single phase or  an
          equivalent amperage  at  480 VAC  single  phase  to  operate
          instruments and air-conditioning equipment.
                                     34

-------
u>
Ul
                           JENSEN
                         FILTRATION
                           PLANT
                                                                          WEYMOUTH
                                                                          FILTRATION
                                                                            PLANT
  DIEMER
FILTRATION
  PLANT
        ...»   FILTERED -SPW/CRW BLEND

               FILTERED SPW

        ——   FILTERED CRW
                              SKINNER
                             FILTRATION
                               PLANT
                       Figure 9.  MWD system delivery pattern showing monitoring locations—Pre-drought period (before March 1977).

-------
CTi
                       JENSEN
                     FILTRATION
                       PLANT
                                                                    WEYMOUTH
                                                                    FILTRATION
                                                                      PLANT
  DIEMER
FILTRATION
  PLANT
                                                                                                  SKINNER
                                                                                                 FILTRATION
                                                                                                   PLANT
             FILTERED-SPW/CRW BLEND

             FILTERED -SPW

             FILTERED -CRW

-------
U)
                             JENSEN
                           FILTRATION
                             PLANT
                                                                         WEYMOUTH
                                                                         FILTRATION
                                                                           PLANT
                                                                         DIEMER
                                                                       FILTRATION
                                                                          PLANT
                                                                                                      SKINNER
                                                                                                     FILTRATION
                                                                                                       PLANT
FILTERED - SPW/CRW BLEND
FILTERED-SPW
FILTERED - CRW
                     Figure 11. MWD system delivery pattern showing monitoring locations—Post-drought period (after March 1978).

-------
                           TABLE 2.  MOBILE LABORATORY FIELD MONITORING LOCATIONS
00
Location
No.
1


2
3



4
5

6
7
8
9
10




11


12
13
14

15
16

17
18
19

a)
b)



a)
b)









a)
b)


a)
b)










Location Name
Pasadena


Long Beach Harbor
Long Beach-Will Johnson
Reservoir


Long Beach Harbor
Santiago Creek Pressure
Control Structure
Orange Fire Sta. No. 4
Orange Fire Sta. No. 3
Orange Well No. 14
Anaheim Fire Sta. No. 8
Anaheim Fire station
No. 8 -OH


Locations lla through
Anaheim Fire Sta. No. 7


Anaheim Wells Nos. 27 & 28
Jensen Filtration Plant
Sepulveda Canyon Pressure
Control Structure
Long Beach Service Yard
Long Beach Filtration
Plant
Long Beach Service Yard
Long Beach Alamitos Res .
Alhambra Fire Sta. No. 4
Dates
1/24/77 -


2/7/77 -

2/11/77 -


2/23/77 -

4/4/77 -
4/18/77 -
5/9/77 -
5/19/77 -
6/6/77 -

6/15/77 -


19 had the ground
9/16/77 -


9/28/77 -
10/11/77 -

10/21/77 -
11/7/77 -

11/16/77 -
11/30/77 -
12/12/77 -
1/5/78 -
2/7/77


2/11/77

2/23/77


3/7/77

4/18/77
4/20/77
5/19/77
5/28/77
6/15/77

6/27/77


loop in the
9/27/77


10/7/77
10/21/77

11/2/77
11/16/77

11/30/77
12/12/77
12/22/77
1/16/78
Water Analyzed

CRW/SPW/GW( low )
CRW/SPW/GW( high )
SPW + Zn3(PO4)2


SPW
CRW/SPW
CRW/SPW + zn3(P04)2

CRW
CRW
CRW/GW
GW
Anaheim Treated CRW (no pH


MWD Treated CRW
MWD Treated CRW/GW
corrosion meter circuit

MWD Treated CRW/GW (low)
MWD Treated CRW/GW (high)
GW
SPW

CRW/SPW
CRW/GW + zn3(PO4)2

GW
CRW/GW
CRW/GW
CRW/GW + Zn3(P04)_
Comment
No.










1
2,3,4
5
6
ad j ustment )



7,8



8,9

10,11





12
12

                                                                                               (Continued)

-------
                                             TABLE 2   (Continued)
to

Location
No.
20
21
22

23
24
25
a)
b)
26
27

28
29
a)
b)
30
31

32

33
Location Name
Walnut-Libby Glass Co.
Foothill MWD
Sepulveda Canyon Pressure
Control Structure
Burbank Service Yard
Burbank Well No. 11
Burbank Well No. 11 Blend


Foothill MWD II
Long Beach-Will Johnson
Reservoir
Long Beach Harbor
Anaheim Fire Sta. No. 7


La Verne Test 1
La Verne Test 2
Repeat of Test 1
La Verne Test 3
Continuation of Test 2 Varying
La Verne Test 4
Dates
1/30/78 - 2/8/78
2/8/78 - 2/22/78

2/22/78 - 3/7/78
3/7/78 - 3/17/78
3/21/78 - 3/30/78
3/30/78 - 4/11/78


4/11/78 - 4/21/78

5/4/78 - 5/16/78
5/16/78 - 5/30/78
5/30/78 - 6/8/78


6/8/78 - 6/22/78
6/22/78 - 7/1/78

7/7/78 - 7/27/78
Flow
7/27/78 - 8/15/78
Water Analyzed
CRW -f Zn3(P04)2
CRW

SPW
CRW/SPW
GW

CRW/SPW/GW ( low )
CRW/SPW/GW ( hi gh )
CRW/SPW

CRW/SPW
CRW/SPW + Zn_(P04)2

CRW/SPW/GW ( low )
CRW/SPW/GW ( hi gh )
CRW/SPW (no pH adjustment)
CRW/SPW (no pH adjustment)

CRW/SPW (no pH adjustment)

CRW/SPW (no pH adjustment)
Comment
No.
13
14

14
14
6,14


8,14
14

14,15
14


8,14
16
16

16

16
Continuation of Test 2 On & Off Ground Loop
34

35
La Verne Test 5
Repeat of Test 1
La Verne Test 6
8/17/78 - 9/7/78

9/7/78 - 9/25/78
Repeat of Test 1, but with Ground Loop Continuously
36
Skinner Filtration Plant
10/2/78 - 10/26/78
CRW/SPW (no pH adjustment)

CRW/SPW (no pH adjustment)

CRW/SPW
16

17

18

-------
                   Comments Related to Table 2


 1.    Start water sampler for hourly samples for  hardness, calcium, and
      alkalinity titrations.

 2.    Start new computer time sequence -10 min. cycle time.

 3.    Start recording CCDT on Computer CH-1.

 4.    Corrosion meter problem, suspended monitoring on 4/20/77.

 5.    Start recording HF turbidity on Computer CH-2.

 6.    No residual chlorines; no Cl? added to well water.

 7.    Malfunction of A/D converter on Computer 6/21/77.

 8.    Marked  diurnal  fluctuations  in quality due to  in-line blending of
      the 2 waters.

 9.    Start recording corrosion on Computer CH-11.

10.    Start constant head control of  flow to corrosion cells.

11.    Start recording flow on Computer CH-12.

12.    Locations  Nos.  17  and  18  are  the  same  water as No. 15  but
      without  zinc; that is, the zinc feed was stopped.  Monitoring  was
      not started  until the  zinc  level  at  the  monitoring site  had
      decreased to near background levels.  However, the water had to
      flow through 5 to 10  miles of  previously  pacified delivery mains to
      get to  Location 17.   Since  there was  a possibility of long-term
      residual  effects due  to  zinc  being in the delivery  mains, it  was
      decided  to  monitor  at  Location  18  which  is  only   about  a mile
      downstream from where the zinc is added.

13.    Corrosion ground  loop  discovered  and  corrected  by using  an
      external power  supply for the  corrosion meter relay.

14.    Chloride monitor  breakdown;  chloride  values  taken  from daily
      grab samples.

15.    CCDT   automation   started,   halfway  through   monitoring   at
      Location  27.

16.    Recorded only pH, EC, DO, Temp, corrosion rates.

17.    Added CCDT to parameters listed in 16.

18.    All parameters monitored except Chloride.
                                 40

-------
     3)   Drain,  essentially at ground  level, to dispose of wastewater
          from the laboratory by gravity flow from 1-1/2" PVC drain
          pipe.

     4)   Fenced parking area to secure laboratory from vandalism.

          In addition  to  meeting the above conditions,  sites were selected
where  the cooperating  water utility could modify water distribution patterns
to vary the sources of water,  or in  a  few cases to temporarily discontinue
feeding of  corrosion inhibitor  chemicals  to  observe subsequent effects  on
the parameters being monitored.

          Prior  to  moving  to  a  new location,  the potential site was visited
and  surveyed to make sure the  necessary utilities  were available  and that
the other conditions were met.  In this  way any necessary  alterations in the
site  or necessary  preparative  work could be done in advance of bringing in
the  mobile  laboratory.  Packing  up  the  lab at one location,  moving, and
setting up at the new  location generally required one day.  Packing up and
setting up  each  took  about  an  hour when  the necessary  site preparations
had  been  made.   It must  be remembered that the time for changing locations
had  to be kept  to  a  minimum, because  the battery back-up which holds the
computer  memory  is  only  good  for approximately 4 hours.  The  electrical
power  was,  therefore,  the last utility  disconnected  and  the  first one
connected when changing  locations.

          The mobile laboratory  was  leveled and stabilized at  each site  by
the use of four screw jacks placed under the frame  of  the  vehicle.  If much
leveling was necessary, the laboratory  was driven up on blocks after  which
fine  adjustments were made  with the jacks.

PARAMETERS MONITORED

          Table  3 is  a list  of the parameters  monitored  during this study.
The  table lists  the parameter, the frequency  with  which  it was measured,
and  the technique  used to make  the  measurement.  In addition,  daily  grab
samples were  taken back to the  MWD central laboratory for analysis and a
check  on  values  obtained by  the  mobile  laboratory  instruments.   Daily
comparisons  were  made  on  pH,  conductivity,  turbidity,  and  dissolved
chloride.   These daily  chloride values became  of  greater value later  in the
project when the chloride  monitor was malfunctioning.

SAMPLING FOR MANUAL AND SPECIAL ANALYSIS

          Calcium,  hardness,  and  alkalinity   were  monitored by manual
titrametric  techniques.   The  samples for these parameters were  taken  by
the  24-bottle  sampler  (Sigmamotor).    The  sampling  frequency  varied,
however  in general samples were taken  hourly during the week and every 3
hours  over the weekend.   The  sampler  collected about  200 ml of  water  in
250-ml polyethylene bottles.  Analyses  were  made  daily  on the  previous
day's samples.
                                    41

-------
                                       TABLE 3.  PARAMETERS MONITORED FOR MWD MOBILE LABORATORY PROJECT
                       Parameter Name
to
                       Frequency of
                         Analysis
                                                             Computer
                                                             Channel
                                                             Number
         Methods of Analysis
  Manufacturer if
Commercial Instrument
1)  Calcium Carbonate
      Deposition Test    Continuous*
2)  H-F Turbidity            "
3)  Chloride                 "

4)  pH
5)  Conductivity             "
6)  Dissolved Oxygen         "
7)  Temperature              "
8)  Free Residual Chlorine   "
9)  Total Residual Chlorine  "
10) Hach Turbidity           "
11) Corrosion Rates          "
     Mild Steel
     Zinc
     Copper
                     12) Total Flow to Lab
                     13) Hardness

                     14) Calcium
                     15) Alkalinity
                     16) Cadmium
                     17) Lead
                     18) Copper
                     19) Zinc
                     20) Iron
                     21) Manganese
                        every 1-4 hrs.
                        from 24-bottle
                           sampler
                        2 or more/loc.
       Potentiostatic Rotating
1      Ring Disc Electrode
2      Nephelometer
3      Solid-state Ion-
       Selective Electrode
4      Glass Electrode
5      A-C conductivity cell
6      Voltammetric Electrode
7      Thermister
8      Galvanic Cell Analyzer
9      Galvanic Cell Analyzer
10     Nephelometer
11     Polarization Admittance
       Technique

12     Turbine Flowmeter
       EDTA titration

       EDTA titration
       H2S(04 titration
         /
       Graphite Furnace AA
_            it
       Flame AA
Pine Instrument Co.
H-F Instruments Ltd.

Schneider Inst. Co.
Schneider Inst. Co.
Schneider Inst. Co.
Schneider Inst. Co.
Schneider Inst. Co.
Capital Controls Co.
Capital Controls Co.
Hach Chemical Co.
Petrolite Insts. Co.


C-E In-Val Co.
Perkin-Elmer
                     *Cohtinuous,  i.e.,  every TO to 12 mins.

-------
          The  daily  grab  samples, which were used  to  check  the  mobile
laboratory instrument readings  against  the  central laboratory values, were
taken  in  2-liter polyethylene  bottles.  Before  the  sample was taken,  the
bottle  was rinsed  several  times  with  the water to  be sampled.   The grab
samples for  trace  metal  analyses  were taken in 125-ml polyethylene bottles
with polyethylene  caps.   Both  bottles and  caps were prewashed with  1:1
nitric  acid and rinsed  thoroughly with  deionized  water  (Millipore  Corp,
Super  Q).  Each bottle  had 1 ml  of clean 1:1 nitric  acid  added so that  the
sample, when taken, would be preserved until the time of analysis.

MONITORING PERIOD

          The mobile laboratory was set  up at each location for a period of
about  10  days.   It was  thought that the  10  days was  a necessary minimum
period in  which to obtain  a good background  of water quality data.  This
time period would  allow for the observation of diurnal variations as well as
weekend  variations.  This  period  was arbitrarily chosen;  however,  it was
based  on  the fact that it takes some time to  obtain meaningful corrosion rate
measurements.   Corrosion of a clean metal surface  is greatly accelerated  on
the first  contact with water giving a  very high initial corrosion  rate.  The
corrosion  rate   drops in  an exponential  manner and begins  to  stabilize  in
about  a  week,  that  is,  the daily rate of change is  at a  minimum.   It was
hoped  that the  10  days would be sufficient to obtain corrosion  rates that
would be  comparable between the various  monitoring locations.

CLEANING CORROSION ELECTRODES

          Acid  cleaning  was  used to remove  corrosion products from  the
corrosion  electrodes  prior  to  the  start  of monitoring at each location.   At
first 20 percent hydrochloric acid (1:5 by volume) was used to clean  all  the
electrodes (mild steel,   zinc,  and  copper),  however,  it was observed that
the copper electrodes were  not completely cleaned of  oxidation products  by
the dilute HC1.  Subsequent to this observation,  dilute nitric  acid (1:1  by
volume)   was used "to  clean  the copper electrodes.   In  all  cases,  the
electrodes were dipped in the  cleaning acid and held there until  the  visible
gross  deposits  and corrosion products  were removed.  Brushing was also
used to help clean the  electrodes.  Immediately after  the  acid washing,  the
electrodes were rinsed thoroughly in  tap water to remove  any traces of  the
cleaning acid.   Extreme care was  taken  during  cleaning to never touch  the
electrodes or, in some other  way, get oil or grease on them.

CLEANING CCDT ELECTRODES

          Until  the" CCDT  was completely automated,  the CCDT electrode
had to be manually. cleaned prior to the start of each CCDT run.  Cleaning
was accomplished  by  removing  the  electrode  from  the  rotator  and then
putting a  drop  of  hydrochloric  acid (1:1 by volume) on the gold electrode.
Any deposited  CCDT film was  instantly  solubilized  by  the acid.  Most films
were  carbonaceous and there would  be  a little  fizzing after the acid was
applied.   The electrode was then  rinsed with tap water.   This acid-washing
procedure was  repeated once before the electrode  was  remounted  in  the
rotator.

                                    43

-------
SPECIAL ATTENTION FOR THE CCDT

          In  the early  stages  of  the project after the computer was  used to
record the CCDT data,  every effort  was made to manually start the test
just  before a reading was  to  be  taken by the  computer,  that  is, near  the
end  of a  10  minute  sampling cycle.  In this  manner an initial reading would
be taken  just  after  the  test  was started.   However,  this  was  not always
easily accomplished.  It was important to obtain this initial  reading  because
some waters  (those with zinc phosphate added)  had very steep  CCDT slopes
and  the  whole  test  was  nearly  complete  in 10 minutes.   If the test was
started  in the  middle of  a sampling  cycle, the test would  be almost  finished
by  the  time the computer took the  first  reading.   Midway  through  the
project, a small  program change  was written for  the DAQ worker task that
allowed  an external switch, connected to one of  the digital input channels to
the  computer,  to start and stop  the test.  When  the  switch was thrown in
the  appropriate direction the  CCDT  was either  started, or stopped,  several
seconds  before the reading  was taken  by  the computer.  Thus an initial
reading was assured just seconds after the test was started.

INSTRUMENT CALIBRATION

          The   monitoring   instruments   in   the   mobile  laboratory  were
periodically  checked  for  accuracy,  and  calibrated when  necessary.   In
general,  most  of the instruments needed  little attention  as to calibration.
The  residual  chlorine  analyzers  needed  to  be  recalibrated  only  once,
however,  monthly calibration checks were made using a Wallace and  Tiernan
amperometric titrator.   The  HF  turbidimeter was checked  daily against  a
standard  and the meter was  adjusted  when  needed.  The Hach turbidimeter
was  then  adjusted to  read the same as  the  HF  unit.  The dissolved oxygen
(DO) system was checked  periodically against a Winkler DO  titration.  When
the  sensitivity  of the  electrode  decreased  significantly,  the  electrode was
rebuilt  according to the  manufacturer's  directions.  After  rebuilding,  the
DO system was  calibrated against Winkler titrations.

          Daily grab  samples were  collected in the mobile laboratory and
taken back to  the  central lab where  the  pH,  conductivity, turbidity, and
chloride   were  measured  and  the values  recorded.   Values for the  same
parameters were recorded  from  the  mobile laboratory  instruments when  the
sample was taken, so that comparisons could be made  which would  indicate
the  need for  calibration.   The  various   monitors  were  adjusted  when
necessary.

INSTRUMENT PROBLEMS

          During the field monitoring phase of  the project,  a  number of
instrumental problems  and failures  were experienced,   some of which were
responsible  for  periods   in  which  no  monitoring  could be  carried  out.
However,   in most cases  these pauses  in the field monitoring program were
put  to  good use in calibrating instruments  and  in  making the  previously
described improvements in the monitoring equipment.
                                     44

-------
          The first problem  causing a time  loss was due to a failure in the
corrosion  rate  meter.   The  instrument had to be sent back to  the  factory
for  repairs.   The  problem  was   a   defective power  supply  which  was
replaced.  This caused a 3-week delay  in monitoring.

          The most serious time loss problem was a failure in the computer's
analog-to-digital  converter (ADC).  Here again the  piece  of equipment had
to be  sent  back  to   the  manufacturer  (Computer  Products)  for  repair.
However,   this   was  handled  through  the  equipment  supplier,   Texas
Instruments, who had  originally  supplied the complete computer  package.
The  extra handling caused  by not dealing  directly with the manufacturer
caused additional delays in completing the repair.  As a result,  a  total loss
of nearly  three months was  experienced due  to  this particular instrument
failure.   Some  of  the  delay was   caused by  the fact that  there were no
manuals or other documentation  on  the ADC.  Subsequently,  several manuals
were purchased to provide  this information.

          A major  problem was experienced with  the  corrosion  rate meter
(CRM) that  caused no  monitoring  time  loss,  however, it did introduce a
question   as to  the validity  of  the  corrosion  data collected  while the
undetected  problem  existed  (from   Locations 11 through  19).  The problem
was  inadvertently introduced when the computer  was set up  to record the
corrosion  data  and to  cycle  the CRM.  As was stated earlier in  the  report,
a 24VDC  relay that would  cause the CRM to cycle  was installed in the CRM.
An  existing 30VDC power  supply in  the  CRM  was  used as  the  supply
voltage to operate  the  relay coil  and  a  channel  on  the  computer's  digital
out-card was used  as  the  switch would complete the circuit and  activate the
relay.   The manufacturer  approved this  alteration  in the CRM and stated
that  it would not overburden the  power  supply.  However,  when the CRM
was  connected  to the  computer a  very complicated  and  circuitous  ground
loop,  that  affected  corrosion  rates,  was  introduced  into the  corrosion
measuring circuit.  It  was found  that this  ground loop included the digital
out-board  in the  computer,  the  power  supply  in  the  CRM,  the  CRM
measuring circuit,  and  the corrosion  electrodes in the flow cells.  When one
part of this loop was broken, the effect was removed,  as  was the case when
the  CRM  was  measuring  on  the  meter prover internal corrosion  standard
instead of  on  the  electrodes in  the  flow cell.  The  initial testing of the
computer  cycling  control  was  done  using   only  the  channel  that  was
connected to the meter  prover  and thus  no effect was noticed because the
ground loop was broken.  Since no anomalies  were  noted during the setup
and  testing it  was  assumed  that there  would be no  erroneous readout when
the  CRM  was set up  to measure  on  the four channels.   Thus no  special
attempt was made  to look for  possible problems.   The only way  to detect
the problem was to  observe  the meter readout  while one leg of the loop was
broken and this  was   exactly  how  the problem  was  discovered.   Some
additional changes  were being made in computer controlling other operations
and  the CRM  control wires were disconnected  from  the digital out board of
the computer.  During  this operation  the CRM meter was observed to jump.
The  full  extent of the problem  was determined  after further  inspection.
After it was identified,  the ground loop effect  was immediately eliminated by
changing  the power supply for the cycling relay  from the one in  the CRM
to a  power supply isolated from the  CRM.


                                     45

-------
          The  ground loop had different effects  on the corrosion rates for
the  three electrode materials  and was  not  the same  in the  anodic and
cathodic  modes.  Several  tests  have  been  carried  out  in  an  attempt  to
quantify  the   effect of  the ground  loop and  to  determine how best  to
interpret and/or  correct  the  results  obtained during  the period when the
ground  loop  was connected and affecting the corrosion  rates.   This was
done  by  measuring the  corrosion  rates  with the corrected  circuit for the
normal  monitoring period;  then,  the corrosion rates were  measured with the
ground  loop  reconnected  for  one  day  near  the  end  of each  monitoring
period.   Having  the  corrosion rates measured under both  conditions was
helpful in interpreting the  data  collected while the  ground loop  was in the
circuit.

          Other minor problems were experienced with the following:

     1)   Recorder  on  the  Petrolite  Corrosion  Rate Meter--sent  to
          manufacturer for repair

     2)   Residual chlorine analyzers—chemical feed  "star wheels"  had
          to be replaced a number of times due to  defective  parts

     3)   Conductivity module in the Schneider Robot Monitor—replaced
          operational amplifier.

          During a  considerable portion of  the monitoring  program,  the
chloride module of  the  Schneider Robot  Monitor  did not function properly.
It was  difficult to calibrate and when calibrated it would  almost immediately
begin  to slowly drift out of  calibration.  Late in the monitoring program,
after numerous attempts by the  MWD staff to adjust and  repair the unit, it
was  sent to  the  manufacturer for  repair.   Considerable  chloride  data was
lost  during the repair period; however, the  daily samples that were  taken
and titrated with silver nitrate provided sufficient chloride  data.

          It  should  be  pointed  out  that  in none  of the above  described
problems could the  failure be  attributed  to the fact that the  instrument was
installed in the mobile  laboratory  and  thus  was subjected to some  rather
rough treatment during transit  between locations.
                                     46

-------
                                SECTION  7

                        RESULTS AND DISCUSSION
GENERAL

          A  tremendous  amount  of data on  water quality parameters  that
could be measured by  the monitoring  systems on  the EPA mobile laboratory
was  collected during  this  study.  Without the  capabilities of the on-board
computer-recorder  and  the back-up data  processing  capabilities  at  MWD's
Data  Processing Center,  meaningful  data  reduction for  evaluation  of  the
results would have been a herculean task, if not impossible.

          An understanding of the  chemical composition  of  the source  waters
imported by MWD and the  ensuing quality after  blending and  filtration is
needed  to  set  the stage  for  discussion of the  results of the  distribution
system  monitoring  program.   Table 4 shows  the  average  chemical analysis
for the major  constituents in  the  Colorado River and  State  Project  source
waters, as  determined at MWD's central laboratory,  during this investiga-
tion.  Typical  chemical analyses  of the imported surface waters after  blend-
ing and  treatment at  the  filtration plants are  presented in Table 5.  Except
for  transient  differences  in  blending  ratios,  the  Diemer  filtration  plant
effluent is similar to the Weymouth plant filtered water.

          Typical  chemical analyses  of  the  groundwaters tested  during this
study  are  shown  in  Table 6.   These  analyses  were obtained  from  the
cooperating  water utilities where field monitoring was performed.

          The range in quality characteristics for all these waters  was:

                                  Units          Minimum        Maximum
     Electrical Conductivity      jumhos/cm           370      -      1195
     Total Dissolved  Solids         mg/L            240      -       760
     Total Hardness                 mg/L             61      -       325
     Total Alkalinity               mg/L             86      -       214
     Sulfate                       mg/L             17      -       290
     Chloride                      mg/L             16      -       106
     pH                                              7.4-8.4
     Carbon  Dioxide                 mg/L              1.6   -        11
     Dissolved  Oxygen               mg/L              2.5   -        11.6

The  total dissolved  solids content and alkalinity  of these waters  varied from
intermediate  to  high  and  they would be  classed as moderately hard to very
hard  waters.   These characteristics of the  waters  studied  should be kept in
                                     47

-------
              TABLE 4.   CHEMICAL ANALYSIS OF MWD SOURCE WATERS
                        AVERAGES FROM JANUARY  1977 TO JUNE 1978

Source of Water
Constituent
SILICA
CALCIUM
MAGNESIUM
SODIUM
POTASSIUM
CARBONATE
BICARBONATE
SULFATE
CHLORIDE
NITRATE
FLUORIDE
BORON
TOTAL DISSOLVED SOLIDS*
TOTAL HARDNESS AS CaC03
TOTAL ALKALINITY AS CaC03
FREE CARBON DIOXIDE (Calc.)
HYDROGEN ION CONCENTRATION
ELECTRICAL CONDUCTIVITY
TURBIDITY
TEMPERATURE
Symbols
and
Units
Si02 mg/L
Ca mg/L
Mg mg/L
Na mg/L
K mg/L
C03 mg/L
HC03 mg/L
S04 mg/L
Cl mg/L
N03 mg/L
F mg/L
B mg/L
mg/L
mg/L
mg/L
C02 mg/L
pH
Mmho/cm @25°C
TU
°C
Colorado
River
Lake
Mathews
9.9
80
30.5
104
4.6
0
153
290
92
0.2
0.31
0.13
688
325
125
1.4
8.27
1060
1.6
17
State
Project
Lake
Silverwood
11.5
27
16.0
73
3.7
1.4
103
56
106
0.6
0.1
0.18
347
133
86
1.0
. 8.27
618
3.2
14
Castaic
Lake
13.8
42
16.0
45
2.6
0.3
120
90
53
0.5
0.22
0.19
324
171
99
2.7
7.87
544
2.5
13
*  TDS determined by summation by method in Ref.  7,  p.  146-147.

mind  during  the ensuing discussion.   Moreover,  the concentration values
for some of the  constituents will  be useful in computing corrosion  indices
(10).

          Table 7 presents the mean values  for the water  quality parameters
related  to corrosion  and stability monitored by the mobile laboratory.   The
waters tested are divided  into groups  based on the major source or type of
water.   The  replicate  tests  within each  group  are listed  chronologically
together  with  site  location.  Within each  group  the locations where  zinc
phosphate was  added to  stabilize  the water  and  mitigate  corrosivity are
                                     48

-------
TABLE 5.  TYPICAL CHEMICAL ANALYSIS OF MWD FILTERED WATERS
             MONTHLY COMPOSITE OF DAILY SAMPLES
April 1977
CONSTITUENT
SILICA
CALCIUM
MAGNESIUM
SODIUM
POTASSIUM
CARBONATE
£ BICARBONATE
SULFATE
CHLORIDE
NITRATE
FLUORIDE
BORON
TOTAL DISSOLVED SOLIDS
TOTAL HARDNESS AS CaCO
tOTAL ALKALINITY AS CaCO
FREE CARBON DIOXIDE (Calc.)
HYDROGEN ION CONCENTRATION
ELECTRICAL CONDUCTIVITY
TURBIDITY
TEMPERATURE
PERCENT STATE PROJECT WATER
SYMBOLS
AND
UNITS
Si02 mg/L
Ca mg/L
Mg mg/L
Na mg/L
K mg/L
C03 mg/L
HC03 mg/L
S04 mg/L
Cl mg/L
N03 mg/L
F mg/L
B mg/L
mg/L
mg/L
mg/L
C02 mg/L
pH
Mmho/cm @25°C
TU
°C

Colorado
River Water
Weymouth
Filtration
Plant
10.2
83
30.5
108
4.7
0
154
299
92
0.4
0.34
0.12
705
333
126
2
8.15
1090
0.2
15.5
0
State
Project Water
Jensen
Filtration
Plant
13.2
36
14.5
43
2.2
1
110
76
52
0.2
0.16
0.22
293
150
92
0.7
8.4
500
0.3
14.5
100
Colorado
River Water
Skinner
Filtration
Plant
9.0
77
28.5
100
4.5
0
151
274
88
0.1
0.32
0.10
657
310
124
2
8.15
1030
0.2
16.5
0
April 1978
State
Project Water
Jensen
Filtration
Plant
14.5
45
16.5
52
3.3
0
118
107
60
0.2
0.32
0.18
358
180
97
2
8.03
600
0.2
12
100

CRW/SPW
Blend
Weymouth
Filtration
Plant
12.3
56
22
79
3.9
0
120
189
79
0.2
0.24
0.10
502
230
98
1
8.15
820
0.2
15.5
35-40

-------
                    TABLE 6.  TYPICAL CHEMICAL ANALYSIS OF GROUND WATERS USED DURING STUDY


CONr- :iTUENT



SILICA
CALCIUM
MAGNESIUM
SODIUM
POTASSIUM
CARBONATE
BICARBONATE
SULFATE
CHLORIDE
NITRATE
FLUORIDE
BORON
TOTAL DISSOLVED SOLIDS
TOTAL HARDNESS AS CaCO
TOTAL ALKALINITY AS CaCO
FREE CARBON DIOXIDE (Calc.)
HYDROGEN ION CONCENTRATION
ELECTRICAL CONDUCTIVITY
TURBIDITY
TEMPERATURE
SYMBOLS
AND
UNITS


SiO- mg/L
Ca mg/L
Mg mg/L
Na mg/L
K mg/L
C03 mg/L
HC03 mg/L
S04 mg/L
Cl mg/L
N03 mg/L
F mg/L
B mg/L
mg/L
mg/L
mg/L
C02 mg/L
pH
^mho/cm @25°C
TU
°C
Pasadena
Villa
Wel-
9-6-7


21.5
26
6.8
52
--
0
120
56
16
18.3
2.4
--
259
94
98
2.0
8.0
410
--
22
Orange
Well#14
12-13-76



20
73
13.9
42
2.8
0
206
69
52
4.2
0.4
0.19
380
240
169
8.6
7.6
650
--
--
Anaheim Wells
5-31-77
#27



12
83.2
26
96
4.4
0.5*
197
183
99
15.7
0.7
0.24
619
310
161
8.2
7.57
975
0.02
14.2
#28



11.0
84
23.8
90
4.4
0.3*
194
185
98
15.1
0.8
0.22
610
302
160
11
7.47
975
0.02
15.2
Long Beach
Well Water
Blend
Filtered
10-3-77

21.4
21.7
1.7
62
1.35
0
156
17.3
35.2
0.27
0.5
--
240
61.2
128
2.7
8.02
413
0.86
--
Burbank
Well #11
3-20-78



26.5
66
16.5
33
3.5
0
261
53
18
9.6
0.47
__
357
233
214
6.9
7.8
570
0.14
--
Alhambra
Wells
San Gabriel
Basin
Average
Values
29
37
13.3
33
1.5
--
172
22.1
28.1
23.1
0.68
--
274
148
141
5.7
7.7
--
--
--

Carbonate values calculated from equilibrium relationships-from nomograph in Ref. 9, p. 296.

-------
             TABLE 7.   SUMMARY OF WATER QUALITY DATA FOR FIELD MONITORING STATIONS
cn

Water
Type
SPW
SPW
SPW
SPW
CRW
CRW
CRW
CRW
CRW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
GW
GW
GW
GW
CRW/GW
CRW/GW
CRW/GW
CRW/GW
CRW/GW
CRW/GW
CRW/GW
CRW/GW
CRW/SPW/GW
CRW/SPW/GW
CRW/SPW/GW
CRW/SPW/GW
CRW/SPW/GW
LOC.
No.
2
3a
13
22
5
9
lOa
20
21
la
3b
4
14
23
26
27
28
35
36
8
12
16
24
7
lOb
lla
lib
15
17
18
19
Ib
25a
25b
29a
29b
Temp
°C
15
13
17
13
15
21
21
15
14
13
13
16
20
15
16
18
20
25
25
21
17
27
19
19
22
25
22
24
23
24
17
18
17
19
22
20
D.O.
mg/L
10.7
10.7
6.3
10.0
10.9
7.8
6.3
10.4
10.2
10.7
11.3
11.6
9.2
9.8
9.5
9.5
9.0
6.2
7.9
7.5
4.4
2.5
7.6
7.3
6.6
5.8
7.6
3.5
3.5
4.0
10.4
9.0
9.7
7.2
9.0
8.3
EC
micro-
mho/cm
@25°C
490
490
540
604
1117
1130
1110
1060
1090
760
840
795
750
950
790
565
530
713
880
690
980
370
510
900
1195
980
980
554
620
570
950
600
903
571
675
830
mg/L
35
38
44
44
82
82
82
80
82
52
56
55
58
67
57
39
40
45
61
75
86
20
69
82
117
80
117
38
42
38
78
62
66
70
53
84
T.H.
mg/L as
CaC03
146
144
174
172
332
337
330
326
325
218
230
222
236
266
228
176
159
190
244
256
329
61
237
280
406
312
395
137
150
144
326
240
262
240
202
280
Alk.
mg/L as
CaC03 pH
89
94
111
96
124
124
127
120
120
117
102
98
117
109
96
77
78
85
114
163
192
132
214
170
215
122
210
128
126
129
120
150
110
215
91
146
7.7
8.3
7.9
7.9
8.1
7.8
8.1
8.2
8.2
8.0
7.7
7.5
8.2
8.1
8.1
7.8
7.4
7.5
7.9
7.3
7.6
7.4
7.4
7.7
7.3
7.4
7.7
7.4
7.5
7.4
7.9
7.5
8.1
7.5
7.9
7.6
Free CO, pHs
mg/L ( nomo-
(Calc) graph)
3.6
1.1
2.3
2.7
2.0
3.6
1.9
1.5
1.6
2.4
2.5
6.2
1.4
1.8
1.6
2.5
6.0
4.8
2.5
16.0
9.5
9.4
17.0
6.5
19.0
8.5
7.5
9.4
7.4
9.4
3.1
10.0
1.7
14.0
2.1
7.2
8.2
8.2
8.05
8.18
7.9
7.79
7.78
7.92
7.95
8.08
8.1
8.08
7.9
8.0
8.06
8.22
8.12
8.03
7.88
7.6
7.6
8.05
7.52
7.63
7.35
7.72
7.30
7.90
7.93
7.90
7.85
7.75
7.97
7.52
7.97
7.70
Langelier
Index
-0.5
0.1
-0.15
-0.28
0.2
0.01
0.32
0.28
0.25
-0.08
-0.4
-0.58
0.3
0.1
0.04
-0.42
-0.72
-0.52
0.02
-0.3
0.0
-0.65
-0.12
0.07
-0.05
-0.32
0.40
-0.5
-0.43
-0.5
0.05
-0.25
0.13
-0.02
-0.07
-0.10
Ryzner
Index
8.7
8.1
8.2
8.46
7.7
7.78
7.46
7.64
7.7
8.16
8.5
8.66
7.6
7.9
8.02
8.64
8.84
8.56
7.86
7.9
7.6
8.7
7.64
7.56
7.40
8.04
6.9
8.4
8.36
8.4
7.8
8.0
7.84
7.54
8.04
7.8
CCDT
Slope
iifi/min
46.0
0.22
0.24
0.42
2.16
3.0
3.17
57.0
2.0
2.29
2.5
58.0
0.9
1.39
—
1.2
54.0
lis"1
1.5
0.33
—
—
2.58
—
2.8
1.5
27.0
—
—
61.0
5.0
—
17.8
1.4
3.8
Comments
Zn,(PO.) -Added
•j *x £,



No pH Adjustment

Zn, (PO.), Added
j % £


2n, (POA), Added
O fr <&



2n, (PO. )0 Added
j 4* Z.
-No pH Adjustment








Zn,.(PO.)0 Added
j ft £.

Zn3(P04>2 Added






-------
TABLE 8.  CORROSION RATES FOR MILD STEEL, ZINC, AND COPPER
 FROM FIELD MONITORING STATIONS (7TH DAY CORROSION RATES)
Location Water Mild Steel
No . Type Anodic
mils/yr
4
5
7
8
9
lOa
on lOb
to
20

21
22
23
24
25a
25b
26
27
28
CRW/SPW
CRW
CRW/GW
GW
Anaheim
Treat. CRW
CRW/GW
CRW/GW
CRW

CRW
SPW
CRW/SPW
GW
CRW/SPW/GW
CRW/SPW/GW
CRW/SPW
CRW/SPW
CRW/SPW
8.4
8.0
4.8
9.1
7.0
9.2
8.9
5.7

6.6
7.1
4.0
4.6
6.0
4.9
6.5
7.4
9.0
Mild Steel Zinc
Cathodic Anodic
mils/yr mils/yr
7.2
8.0
4.0
8.8
6.0
8.3
7.8
5.0

6.2
6.6
3.7
4.6
5.3
4.6
6.0
6.9
8.1
3.7
2.5
2.1
3.4
4.4
2.8
2.0
0.6

2.9
3.8
2.9
2.6
6.6
2.2
2.0
4.9
5.3
Zinc
Cathodic
mils/yr
3.0
2.2
2.2
3.4
4.3
2.8
1.7
0.6

2.3
3.1
2.4
1.8
5.0
1.8
1.5
4.3
4.4
Copper
Anodic
mils/yr
2.0
2.5
0.5
0.5
1.5
1.3
1.7
1.1

1.8
0.8
1.4
0.8
0.4
0.8
1.1
1.3
0.9
Copper
Cathodic
mils/yr
1.2
2.0
0.5
0.5
1.2
1.0
1.3
0.6

1.3
0.6
1.1
0.8
0.3
0.5
0.8
1.0
0.7
CCDT
M/min
Comments
58 Zn3(P04)2
Added
2.16
2.58
1.5
3.0 No pH Ad j .
3.17
57~" Zn (PO )
Added
2
0.42
1.39

17.8
0.8
54 Zn3(P04)2
Added

                                                                     (Continued)

-------
TABLE 8. (Continued)

Location Water Mild Steel Mild Steel Zinc Zinc
No. Type Anodic Cathodic Anodic Cathodic
mils/yr mils/yr mils/yr mils/yr
29a
29b
30
31
34
36
CRW/SPW/GW
CRW/SPW/GW
CRW/SPW
CRW/SPW
CRW/SPW
CRW/SPW
8.0
8.4
7.8
7.4
10.4
8.7
Locations lla through
ui lla
OJ
lib
12
13
14
15
16
17
18
19

CRW/GW
CRW/GW
GW
SPW
CRW/SPW
CRW/GW
GW
CRW/GW
CRW/GW
CRW/GW

9.1
7.7
7.6
6.1
5.0
7.0
7.0
7.5
8.3
6.8

7.3
7.7
6.9
7.0
9.4
7.7
19 had
10.0
9.0
7.3
6.4
4.9
7.4
6.9
7.6
8.3
8.3

3.6
2.1
1.9
2.7
4.1
2.2
the ground
4.9
1.7
5.9
3.4
2.8
2.8
4.2
4.7
5.5
4.2

2.9
1.8
1.5
2.1
3.2
1.6
loop in the
3.7
1.6
2.3
1.8
2.3
1.2
3.7
2.7
3.6
1.1

Copper Copper CCDT
Anodic Cathodic ptA/min
mils/yr mils/yr Comments
0.9 0.8 1.4
1.2 1.0 3.8
IT HO —__.__ Mo TV W 2X ("I n
1.1 0.8 	 No pH Adj.
1.3 0.9 	 NopHAdj.
0.6 0.6 1.3
corrosion meter circuit
2.8
1.5
0.33
0.24
0.9
27 Zn3(P04)2
Added


61 Zn3(P04)2
Added

-------
noted  in  the  far-right column marked  "comments".   The time  intervals
between   repeated  tests  on  a  particular  water  can  be  ascertained  by
reference to location numbers in  Table 2.

          In addition  to  summarizing  the monitoring  data,  Table 7 shows
calculated  values  for free  carbon dioxide, pH at calcium  carbonate satura-
tion (pH ),  Langelier index, Ryzner  index  and  the slope for the calcium
carbonate  deposition  test  (CCDT)  in microamperes/minute.    Correlations
between  these  indices  on  the tendency for protective film formation will be
discussed later.

          Table 8 indicates  the  instantaneous  corrosion rates for mild steel,
zinc, and  copper as measured by  the  current required to polarize the  test
electrode ten millivolts anodic or cathodic to  the reference electrode.   The
readout is  calibrated in mils per  year of surface corrosion.

          Except  as noted,  the corrosion rate measurements reported  here
represent  the average rate for  the seventh day  of exposure to each water.
As   discussed  in   Section 6,  these  data do  not  represent   the  minimum
corrosion  rate that  may be  expected,  particularly for mild steel and zinc,
but it was the time selected  for comparison of waters when the daily rate of
change was relatively low.

          Tables  9,  10,   and  11  indicate  the  gradual  decrease in  the
instantaneous  corrosion rate  after  longer exposure  periods  and provide
support  for the  rationale  of selecting  the  seventh  day   corrosion rate for
comparison of the different waters.  It appears  that the  corrosion rate for
copper,  Table 11,  approaches  a steady  state more quickly than for  mild
steel and zinc.   Moreover,  the  copper  corrosion  rate  was  not affected
adversely  by changes in the sample flow velocity  past the  electrode surface.
In  contrast, the  mild  steel and zinc electrodes  indicated  that  the  measured
corrosion  rates fell sharply when flows  decreased  to  less  than 300 ml/min
and increased markedly immediately after adjusting  the flow upward to the
normal  rate,  see La  Verne Test 5 in  Tables  9 and  10.  This  response  was
most noticeable when  low  velocity flows  occurred during the  first week of
exposure  after  cleaning  the electrodes.   As explained  in  Section 4,  the
problem' of variable flow  rates  to the  corrosion rate cell was  overcome by
installing a constant head device on the feed Line.

          Beyond the  effects of low sample flow on  the mild steel and  zinc
corrosion  rates,  the accidental ground loop  introduced  into   the  corrosion
rate meter circuit,  which  was  discussed in  Section 6, produced anomalous
readouts from the copper  electrode in both the anodic and cathodic modes at
Locations 11 to 19.  These  data are omitted from Table 8.  The ground  loop
circuit  also  affected  the corrosion rate measurements at Locations 11  to 19
for  mild steel and zinc.   However,  the  bias on readings was less marked
and the  data are presented for comparison  within the group.

           The  trace  metal  content of  the water  was determined on  grab
samples  taken  at each  field  monitoring location,  and the average values are
presented in Table  12.  Since  a continuous  flow condition  was required to
operate  the  corrosion rate meter  and  to obtain data  on  other parameters,

                                     54

-------
TABLE 9.  COMPARISON OF LONG TERM CORROSION TESTS
  MILD STEEL ANODIC CORROSION RATE (MILS/YEAR)
Time
Days
0
0.5
1.0
1.5
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
25
26
27
28
30
32
35
38
41
48
53
* 	 *
n 	 n
La Verne La Verne La Verne
Test 1 Test 2,3,4 Test 5
No pH Adjustment
18.0 18.2
14.7 14.5
13.5 13.1
11.8 12.5
11.2 11.7
10.0 10.0
9.0 8.7
8.0 7.8
7.7 7.5
7.7 7.5
7.7 7.5
7.7 	
7.7 	
7.7 	
7.4 	
7.3 	
	
3~67§~3
6.4
6.4
4_-___4
6.0
6.1
S-g-j-5
5.8
6-5-5-6
5.0
4.7
7~472~7
4.0
5~37T5
3.5
3.5
3.5
3.5
3.6
Flow Adjustment
Change in flow rtte
18.0
14.7
11.7
9.7
8.8
8.0
7.5
*__ 	 *
10.8
10.5
10.3
9.8
9.7
9.2
8.7
8.6
7.8
	
7.4
7.6
7.4
7.2
6.7











9
to nxlO ml/min
Skinner Foothill
Filt. Plant MWD
pH Adjustment
15.7
13.7
11.7
	
	
10.0
9.3
9.2
8.9
8.6
8.5
8.3
8.1
7.9
7.7
7.4
7.3
	
6.9
6.6
6.5
6.3
6.2
6.0
5.8









( standar
15.8
11.0
10.4
9.4
8.5
7.6
6.8
6.5
6.4
6.5























d flow is 500 ml/min)
                           55

-------
TABLE 10.  COMPARISON OF LONG TERM CORROSION TESTS
      ZINC ANODIC CORROSION RATE (MILS/YEAR)

Time
Days
0
0.5
1.0
1.5
2
3
4
5
6
7
8
9
30
11
12
13
14
15
16
17
18
19
20
21
22
25
26
27
28
30
32
35
38
41
48
53
La Verne La Verne
Test 1 Test 2,3,4
No pH Adjustment
15.0
13.5
12.5
11.6
8.7
3.8
1.7
1.4
1.7
2.6
1.4
2.3
2.7
3.2
2.8
3.2

















19.0
14.5
12.0
10.3
9.0
4.7
2.8
1.8
2.4
2.7
2.3
	
	
	
	
	
	
3~3"72~3
2.7
2.6
4_____4
2.0
2.2
S-j-g-5
2.0
6~lT7~6
1.6
1.4
7-i:r7
1.2
S-j-5-5
1.3
1.2
0.8
0.6
0.5
La Verne
Test 5
20.3
11.7
9.2
7.0
5.2
2.8
1.6
* 	 *
5.4
4.3
4.0
3.7
3.6
3.3
3.0
2.8
3.0
	
2.7
2.7
2.7
2.7
2.9











Skinner Foothill
Filt. Plant MWD
pH Adjustment
21.5 13.
12.5 5.
6.0 3.
	 3.
	 2.
2.3 2.
2.1 2.
2.2 2.
2.2 2.
2.2 1.
2.2
2.0
2.0
1.8
1.6
1.7
1.7
	
1.7
1.8
1.8
1.7
1.6
1.7
1.6









6
8
8
0
8
3
5
3
2
8
























* 	 *
n 	 n
Flow
Chanc
Adjustment
je in flow rate to nxlO ml/m
in (standard flow is
500 ml/min)
                       56

-------
          TABLE 11.  COMPARISON OF LONG TERM CORROSION TESTS
               COPPER ANODIC CORROSION RATES (MILS/YEAR)
Time
Days
0
0.5
1.0
1.5
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
25
26
27
28
30
32
35
38
41
48
53
La Verne
Test 1
No
3.1
1.8
1.5
1.3
1.3
1.1
1.1
1.1
1.1
1.1
1.2
1.2
1.1
1.2
1.1
1.1
















La Verne
Test 2,3,4
pH Adjustment
3.7
1.8
1.6
1.4
1.2
1.1
1.1
1.2
1.1
1.1
1..1
	
	
	
	
	
3-075"3
1.0
0.9
4~o7
-------
TABLE 12.   TRACE METAL LEVELS  AT  FIELD MONITORING STATIONS

Location
No.
1
2
3a
3b
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Water
Type
CRW/SPW/GW
SPW+Zn
SPW
CRW/SPW
CRW/ SPW+Zn
CRW
CRW
CRW/GW
GW
Anaheim CRW
Met CRW
CRW/GW
GW
SPW
SPW/ CRW
CRW/GW+Zn
GW
CRW/GW
CRW/GW
CRW/GW+Zn
CRW+Zn
CRW
SPW
CRW/SPW
GW
CRW/SPW/GW
CRW/SPW
CRW/SPW
CRW/SPW+Zn
CRW/SPW/GW
ND = None Detected at
detection limits :
Cd
-
-
ND
0.0001
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0001
ND
ND
ND
ND
ND
ND
ND
ND
0.0003
0.0002
ND
0.0001
Average
Pb
-
-
ND
ND
ND
ND
ND
0.0001
ND
ND
ND
0.0007
0.0004
0.0003
ND
ND
ND
ND
0.0004
ND
0.0002
0.0002
0.0001
0.0004
0.0005
0.0003
ND
ND
0.0002
0.0001
Values in
Cu
0.003
0.007
0.007
0.032
0.027
0.005
0.013
0.006
0.004
0.085
0.100
0.100
0.007
0.006
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.02
0.02
ND
0.01
mg/L
Zn
0.01
1.10
0.01
0.02
1.14
ND
0.01
0.01
0.02
0.01
ND
0.03
0.06
0.02
0.02
0.66
0.03
0.12
0.03
1.46
0.79
0.06
0.01
ND
ND
0.33
0.03
ND
0.95
ND
0.01
Fe
0.03
-
0.01
0.02
0.02
0.02
0.02
0.01
0.01
ND
0.02
0.02
0.01
0.01
0.02
0.02
0.05
0.03
0.03
0.02
0.04
0.02
ND
ND
ND
0.02
0.05
0.04
0.06
0.01
Mn
-
-
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
0.01
Number
of Samples
7
4
5
4
7
1
3
2
2
2
2
2
2
2
2
2
2
2
2
2
3
2
2
2
2
2
2
2
2
2

                           58

-------
metal uptake after long residence times  or  "no  flow" conditions could not be
monitored  on board the  mobile  laboratory.   At  the  outset, several grab
samples  were taken at  each field  location  for  analysis  at MWD's central
laboratory at  La Verne;   but when  the  analytical  results  indicated  that
metal  concentrations  were  below  the  detection  limit  in  most  cases the
frequency of sampling was reduced.

          The  aforementioned tables  summarize  the massive amount of  data
collected  into single mean values.  For  each parameter the  value  reported
may  represent  an  average  of  100  to  3,000  sets of  data for the  tests
performed aboard the mobile laboratory.   In a dynamic water system, signif-
icant variations  in water quality  may  occur  due to blending  and in-line
mixing  of water  sources.  The  capability of  the laboratory for detecting
such  changes  will  be  brought  out  in  greater  detail  as  the  results are
discussed below.

INTERRELATIONS BETWEEN WATER QUALITY PARAMETERS

          Larson  (10),  Ryder  (11)  and  others  have  pointed  out the
strengths  and  weakness of  the several  accelerated tests,  water quality
parameters  and indices used  for comparing the stability and  corrosivity  of
waters.  In this water  quality investigation > an attempt was made to develop
and  evaluate  some  of  the correlations and interrelationships suggested by
earlier  workers.    First,  we  will look  for  correlations  based on  the  data
obtained  at all  the locations  monitored  and  then  compare data illustrating
specific   water  quality  conditions,   such  as  use   of   corrosion  control
chemicals.

Effect  of  pH on Corrosion Rates

          Experience has taught water  works  personnel the  value  of pH
control for corrosion protection,  and the principles of the saturation index
as derived by  Langelier (12) has been  useful  as  a guideline for pH control
for many years.   Larson and Buswell (13) pointed out some limitations on
the interpretation  of  Langelier's index  as a measure of  calcium carbonate
deposition capacity for corrosion  control.  Using the field water quality  data
obtained  in  this  study,  the Langelier  saturation  index  was  computed,
Table 7.   Then,  the  anodic linear  polarization measurement  of  corrosion
rates on  the seventh day was • plotted against this saturation index for the
waters tested in Figure 12.

          As  would  be  expected  when the  saturation  index  shifts  from
negative to positive, the linear polarization corrosion rate for mild steel and
zinc  decreased moderately.   However,  the corrosion rate measurements' for
copper  did not show any  adverse effect of lower pH and negative saturation
values for the waters tested.  The range of pH values was from 7.3 to  8.3,
and the Langelier index varied from -0.72 to +0.32.
                                     59

-------
•»     6
r-



1     4


o

 -
Z     0
o

i
oc
oc
o
o
o
5
o

1     12


      10
                                                            WATERS WITH Zn3(PO4)2 ADDED
                                             COPPER
                                         •    I   ••••>•• .*.     ••
                                               ZINC
                                                   OiB«® 9
        MILD STEEL
                                                           (©)         ©

                                                                9      fi
                         I
I
I
I
            -0.8         0.6         0.4       -0.2          0          0.2        0.4



                                LANGELIER SATURATION INDEX



            Figure 12.  Corrosion rates for mild steel, zinc, and copper on day 7 vs. Langelier saturation index.
                                            60

-------
          Another  "stability index"  was  devised  by  Ryzner  (14).  It is
equal to 2 pH -pH,  an impirical value which places greater emphasis on the
concentration of calcium and alkalinity  as related to scale-forming character-
istics of  the  water.   Figure 13 shows anodic linear polarisation corrosion
rates on the  seventh  day plotted against the Ryzner stability index.

          Waters with  a stability index of  <7.5 are  scale-forming according
to Ryzner,  and  waters with indices  >7.5 are  increasingly scale  dissolving.
Recently,  Ryder  (11) indicated  that waters in the  Seattle  area  were
increasingly  corrosive with values above 8.0.

          It   should  be  noted  that  there was a  slight  trend  for  higher
corrosion  rates for mild steel and  zinc as the Ryzner index (RI) increased,
but  there  was  no  sharp  demarkation  between   scale-forming  and  scale-
dissolving  characteristics  for these waters with RI values  ranging from 6.9
to 8.8.   As in Figure 12,  the pH effect as measured  by the Ryzner stability
index showed no measurable difference in  the corrosion rate  of copper in
these tests.   The  Ryzner index will  be discussed again in relation to the
calcium  carbonate deposition test results.

Effect of Calcium Content  on Corrosion Rates

          The  seventh  day  anodic  linear  polarization  measurement  of
corrosion  rates  was  plotted  against the  calcium  content  of these  waters,
ranging from 39 to 84 mg/L, in Figure 14.  There is  a slight  trend toward
lower corrosion  rates as  the calcium  content increases.   The  scatter of
points,  however, suggests  that other factors, such as, pH and alkalinity
probably exert a more profound effect on  corrosion rates than calcium as an
independent  parameter. Calcium levels may provide insight into the capacity
of a water to deposit  films  which  will be discussed in relation  to the CCDT
measurements.

Effect of Dissolved Minerals on Corrosion Rates

          In  this  investigation,  electrical conductivity was the  parameter
used to measure the  variations in  mineral content of the waters examined in
the  field.    Conversion of  conductivity measurements  to the more  familiar
term,  total  dissolved  solids  (TDS)  may  be  approximated  by  multiplying
conductivity in micromhos/cm by 0.65  for the waters studied.  More  precise
estimates  for this  conversion factor can be obtained for each blend of water
tested by  using conductivity  and  TDS values  obtained by  analysis of the
waters  before blending.  These data  are  shown  in Tables 5  and 6.   The
anodic  polarization  corrosion rates  for  mild steel,  zinc,   and  copper are
plotted  against conductivity  in Figure 15 for the waters studied.

          Numerous studies  (1,2,3,10)  have suggested  that a strong positive
correlation may  be expected between  total dissolved solids or conductivity
and  corrosion  rates.  For  the waters studied with conductivities  ranging
from  370 to  1195 micromhos/cm and TDS from 240 to 760 mg/L,  there  was no
observable trend toward  higher corrosion rates for mild steel and zinc as
the  conductivity increased.   There  was an indication  of slightly  higher
corrosion rates for copper  at higher conductivities.  Whether this is  related

                                     61

-------
                                                        (§) WATERS WITH
                                                                                   ADDED
       4 _
       2 —
                                              COPPER
                                               ZINC
£
o
LU
t-
                                                     »
                                                t
                                                 *
V)
O
oc
DC
O
u
O
5
O
z
      10
        6.5
                      7.0
                                            MILD STEEL
7.5            8.0           8.5


 RYZNER STABILITY INDEX
                                                                               9.0
9.5
           Figure 13. Corrosion rates for mild steel, zinc, and copper on day 7 vs. Ryzner stability index.
                                             62

-------
-5     6
      12
      10
                                                             WATERS WITH Zn3(PO4>2 ADDED
                                               COPPER
                                                 ZINC
                                              MILD STEEL
                    20
                                40
                                                                   100
                                                                               120
                              60          80



                              CALCIUM  (mg/L)



Figure 14. Corrosion rates for mild steel, zinc, and copper on day 7 vs. calcium content.
                                            63

-------
                                             COPPER
                                                •                           Q

                                               ^  &          •         ft    (
                                                                    I
                                                                             I
                                               ZINC
5  6
I  *
o
LU  2
oc

2

2  o

o
oc
QC
O 12

O
Q
 «    c

     t>
"»        .**.•••*•     *
                                          J	I
                  MILD STEEL
    2 —
                                                O  O
         400
                                                                                      J.
                            600                800                1000               1200



                      ELECTRICAL CONDUCTIVITY (micromho/cm at 25° C)



           Figure 15. Corrosion rates for mild steel, zinc, and copper on day 7 vs. electrical conductivity.
                                             64

-------
to higher mineralization, per  se,  or due to a  specific effect, such  as, free
carbon dioxide,  pH or alkalinity will be explored later.

          It   should  be  pointed out  that  the  instantaneous  polarization
corrosion rate determined by  the mobile  laboratory instrumentation does not
measure   the  effect  of  mineralization  on  galvanic  corrosion.    This   is
associated with  the  contact  of two  different  metals or  alloys  in water.
Galvanic corrosion is most troublesome in household and institutional facilities
where two or more  dissimilar  metals may be joined together in the plumbing
system consisting of pipes, valves, and fittings.

          Galvanic  corrosion is generally increased  when  the difference  in
potential  between two metals  results in current flow  and also by increased
mineralization or conductivity of the  water.   However,  it is one of the
important corrosion mechanisms which  could not  be  evaluated by the  tests
performed on board the mobile  laboratory.

          Despite  these  limitations,  the  corrosion  rate   instrument  was
sensitive  enough to respond to diurnal changes in water quality  where the
variations  in blending of  source waters occurred at a  single  station.  In
effect, this  resulted in paired sets of data where other variables remained
constant.  Even though the differences in corrosion  rates  were qualitative,
the  response  to changes  in  water  quality  gave  insight concerning the
tendency of  one water  to  be  more  or  less aggressive than another  to mild
steel,  zinc,  or  copper sensors.  These observations  will be discussed after
presenting  the  data  on  the  other parameters  monitored on  the mobile
laboratory.

CALCIUM CARBONATE DEPOSITION  TEST (CCDT)

          For   the   waters  analyzed  during   this   project,   the   calcium
carbonate  deposition  test was the most  sensitive to  differences  in quality
due  to blending of source waters or diurnal changes in water  characteris-
tics.   The rate of  CaCQ-  film formation is determined from the slope of the
linear portion  of the  current-time curve generated  by a rotating gold disc
electrode.  The slope is then stated in microamperes per minute  (/iA/min).

          The  effect  on CCDT  of blending Colorado  River  water and  State
project water is shown  in  Figure 16.   In this  family  of curves where lower
calcium hardness is the most significant change in water quality characteris-
tics,  the CCDT  slope progressively decreases..  This  indicates that  the time
required  to  form a protective film  under  the  test conditions for the  State
project water with a  calcium content of 44 mg/L was  four  times  as  long as
for Colorado River water  containing 85  mg/L  of  calcium.   The significance
of  these  differences  in   slope  will   be  discussed later  after   further
comparisons  are made.

          Alkalinity  and  pH  are  also  basic water   quality  characteristics
which affect the tendency for  deposition  or  dissolution  of  CaCCL  films.
Recognizing  the importance  of maintaining pH near Langelier's  saturation
pH  or higher, many utilities add caustic  soda or lime  as  a final treatment
step  before  the  water enters  the distribution system. Metropolitan  adopted

                                     65

-------
           700
                                                                            Curve
                                                                            No.

                                                                              1
                                                                              2
                                                                              3
                                                                              4
                                                                              5
Location   % Blend
  No.    CRW/SPW
    5
   23
   36
   27
   22
100/0
70/30
60/40
25/75
0/100
                                                                                                        PH
8.1
8.2
73
7J8
73
               Ca    Alk.  EC
              nr»0/L   mg/L
                                                                                                                                CCDT Slope
85
67
60
45
44
125   1120
110    960
115    890
 87    612
 96    600
IS
1.4
1.3
1.2
0.4
           600
CTi
           500
      ,~   400
           300
           200
           100
                                                                 8           10

                                                                     TIME  (hours)
                                                                                          12
                                                                                                       14
                               16
                                                                                                                                18
                                                                                                                                             20
                                     Figure 16. Effect of blending Colorado River water and State Project water on CCDT.

-------
the practice of pH  adjustment at each of its filtration plants from the outset,
first using milk of  lime and later using liquid caustic soda because of better
control of dosages and increments in post-treatment turbidity .

Effect of pH Adjustment on  CCDT

          During  the period  when  long-term  corrosion tests  reported  in
Tables 9 to 11 were being performed,  the Weymouth plant filtered water was
tested  prior to pH  adjustment.   The slope of the CCDT  curve  for this  water
containing 45  percent CRW at  Location 35  is shown  in   Figure 17.  For
comparison, the curves  for two pH  adjusted  blends  of  water  from  the dis-
tribution  system at Locations 27  and 36 are repeated from  Figure 16.  The
quality  of  these waters  containing  25  percent and 60 percent  CRW,  respec-
tively bracketed the quality of  filtered water  tested before pH adjustment.
The CCDT  curve  indicates that the film  formation  rate for  filtered  water
was only  one-half  the rate, 0.6 ^A/min. compared to 1.2 and  1.3, observed
for  the  pH  adjusted waters of  similar  quality.  This  difference in  CCDT
resulted from  a pH  adjustment of  only 0.3  to  0.4  and  changed  the film
forming characteristic from  "slight"  to "moderate" by the suggested criteria
of McClelland and Mancy (5) .

Effect of Alkalinity and Hardness on CCDT

          After the  CCDT  unit was set up to cycle  automatically and data
could be  collected  for 24 hours  each day, some interesting  data showing the
effect  of  water quality  variations on calcium  carbonate  deposition were
observed.   At Location 29, groundwater was  pumped into  the distribution
system  daily  after  midnight.   This  intermittent pumping  caused  in-line
mixing of higher amounts of groundwater with MWD  water  for 9 to  12  hours
each day.  Figure  18 shows CCDT data for the variations  in water quality
conditions  when calcium hardness and total alkalinity changed significantly:
Curve  1 indicates a CCDT  value of 3.8 MA/min. when the greater percentage
of  ground water reached  the mobile laboratory; Curve  2 for MWD water has
a  slope  of  1.4;  and Curve 3  represents   a  run that included a transition
period where the  quality  changed after about  6 hours causing a  distinct
slope break to 6.7
          From  the  above it appears that the CCDT is sensitive to  changes
in  the quality  of  water being  monitored and  responds immediately to  the
change.   Insufficient  data  were obtained after  full automation  of the  CCDT
system to determine the extent of a carryover effect  when the electrode is
partially  covered  by  a  CaCOg  film from  one. water before exposure  to
another.   At the  very  least,  a  qualitative comparison  of  waters can  be
observed.  Moreover, when the residence time  of  each water in the distri-
bution is long enough to complete a CCDT curve, such as, Curves  1  and 2
in  Figure 18,   good  agreement on  film  formation rates  was obtained  on
successive days.

Effect of Zinc Phosphate on CCDT and Corrosion Rates

          Three water  suppliers  in MWD's  service area  were  using zinc
phosphate for corrosion control when  field monitoring, was  being  done  on


                                     67

-------
                                                                             Curve
                                                                              No.

                                                                              1
                                                                              2
                                                                              3
Location
  No.

   36
   27
   3S
  % Blend
 CRW/SPW

   60/40
   25/75
   45/55
                                                                                                                       PH
7.9  Adjusted
7.8  Adjusted
7.5  Not Adjusted
CCDT Slope
  jUA/mln

    1.3
    1.2
    0.6
CTi
00
           700
           600
           500
      —   400
      Q
      O
      0    300
           200
           100
                                                                 8            10

                                                                      TIME (hours)
 12
14
             16
                                                                                                                                18
                                                                                                                                            20
                                              Figure 17. Effect of pH adjustment of blended waters on CCDT.

-------
                               Curve
                                No.

                                1
                                2
                                3
  Date and
   Time

6/1/78® 0803
6/1/78 @ 1216
5/31/78 O2015
 Water Type
 Ca   Alk.
mg/L  mg/L
                                                                             EC
High%GW    130   195   960
Low % GW     53    91   690
Period when water changed
 from low GW   53    90   700
 to high GW   130   200   960
CCDT Slope
  /M/min

    3.8
    1.4

    1.3
    6.7
600
                             WATER QUALITY CHANGED
                                      TIME  (hours)

   Figure 18.  Effect on CCDT due to in-line diurnal changes in source of water at Location 29.
                                          69

-------
this  project.   At  Long  Beach  the zinc  orthophosphate  was prepared by
mixing  zinc  sulfate and  orthophosphoric  acid  on site;  whereas,  a  propri-
etary  formulation  sold under  the trade  name  Virchem  932 was  used  at
Alhambra  (Location 19)  and  Walnut  Valley Water  District  (Location 20).
Paired comparisons of  corrosion rates and CCDT  were made on these waters
before  and after adding  the  zinc  phosphate.  A summary of these results
are presented in Table 13  and Figure 19  shows  the  marked effect of  the
zinc compound on the CCDT.

    TABLE 13.   COMPARISONS OF  CORROSION RATES FOR WATERS WITH AND WITHOUT
      	 ZINC ADDED  FOR  CORROSION CONTROL (7TH DAY CORROSION RATES)
Location
 Number
 Water
 Type
 Zn    Zn Cone.      Anodic Corrosion Rate
Added    mg/L   pH
                                        Mild
                                        Steel
                                      Zinc   Copper
                                     CCDT
                                    juA/min
  3b      CRW/SPW    No

  4       CRW/SPW    Yes
                   O.J02    7.7   7.5*   4.5*    1.0*      2.5

                   1.14    7.5   8.4    3.7     2.0      58
  27

  28
CRW/SPW

CRW/SPW
 No

 Yes
0.01

0.95
7.6

7.4
7.4

9.0
4.9

5.3
1.3

0.9
 0.8

54
  21      CRW        No      0.06    8.2   6.6    2.9     1.8       2

  20      CRW        Yes     0.79    8.2   5.7    0.6     1.1      57

  19      CRW/GW     Yes     1.46    7.9   6.8*   4.2*    -        61
17
18
15
CRW/GW
CRW/GW
CRW/GW
No
No
Yes
0.12
0.03
0.66
7.5
7.4
7.4
7.5*
8.3#
7.0*
4.7* -
5.5* -
2.8* -
9
6
27

*  8 days from water quality change; 12 days from clean electrodes
#  With ground loop in corrosion measuring circuit

          The addition of  zinc phosphate caused a thin zinc-containing film
to form  very rapidly on  the CCDT electrode  which blocked the microcurrent
flow in the instrument.   The film formation rate ranged from 27 to 61 nA/min.
irrespective of the film-forming characteristics of the water prior to  adding
the  corrosion control chemical.   Because of the short time, less than  20
minutes  to  complete  a single CCDT run,  the  amount of material deposited  on
the  electrode  was very  small.  Its exact nature could not be  determined,
but  a  qualitative  test  indicated  the presence of  precipitated  zinc.  The
amount  of  corrosion inhibitor added was   not  the  same  at  all  sampling
                                     70

-------
                                                                         Curve
                                                                         No.
 Location
   No.

2, 19, or 26
500
    % Blend
   CRW/SPW

Typical curve with
Zn3(PO4)2 added
     100/0
     25/76
     0/100
CCDT Slope
  /IA/min

   54

    3.3
    1.2
    0.3
400
300
200
100 - •
                                                               6
                                                                                   8
                 12
                                     14
                                                              TIME  (hours)

                                  Figure 19. Effect of adding zinc phosphate for corrosion control on CCDT.

-------
locations, Table  14.   Those water suppliers  that ha,d  used MWD  water for
several years were  feeding lesser dosages,  equivalent  to  1 mg/L of zinc or
less.

          At Location 19 where  MWD water was  recently introduced into the
system, a zinc level of about 1.5 mg/L was being maintained.  Lower dosages
caused  an  upsurge  in consumer complaints  about  "red  water,"  and  the
chemical was  being  fed  to  ease  the  transition  from  groundwater   to  a
CRW/GW blend.

          Despite the  excellent  film forming potential  shown by the CCDT
results,  the corrosion rates measured for mild  steel exposed  to  the  zinc-
treated waters  were  not  significantly different than  those for  the   same
quality  of  water without the  corrosion  control chemical.   Obviously  the
results  of  the polarization  corrosion  test were not  in  agreement with  the
"consumer  acceptance" parameter, i.e., red water complaints,  for  the water
distributed adjacent to Location 19.

          The  corrosion  rates  for  zinc  metal  were  more variable.  Paired
comparisons  of zinc corrosion rates in water  before adding  the  corrosion
control chemical  and after treatment,  that is,  at Locations 36 and  4, 27  and
28, respectively, showed no  difference in corrosion rate.  Likewise,  when
data from Locations 21 and 20,  also 18 and 15,  respectively, are compared,
it   appears  that  the  corrosion rates  were   markedly  reduced by   the
treatment.
               TABLE 14.   ZINC  LEVELS FOR LOCATIONS WHERE ZINC
                          CORROSION  INHIBITOR WAS ADDED

Location
No.
2



Date Water
2/8/77 SPW
2/9/77
2/10/77
2/11/77
Zinc
mg/L
1.04
1.09
1.27
0.99
Location
No.
15


19*
Date
11/7/77
11/9/77

1/5/78 '
Water
CRW/GW


CRW/GW
Zinc
mg/L
0.70
0.63

1.43
          2/23/77

          2/24

          2/25

          2/26

          2/27

          2/28

          3/1
CRW/SPW
1.18

1.12

1.40

1.08

1.07

1.14

1.01
                     20
                     28
1/10/78


1/30/78

2/1/78

2/8/78
CRW
1.50


0.81

0.80

0.76
5/16/78   CRW/SPW   0.90

5/18/78             1.00
                                     72

-------
          With the limited data  available,  the corrosion rates for copper did
not  appear  to  be  appreciably  effected  by   the  zinc  corrosion  control
chemical.

          Photographs   were  taken   of   the   corrosion  rate  sensors  on
completion of the  tests at  all locations.  Figure 20 shows sets  of  the mild
steel,  zinc  and  copper electrodes:   (a) at Location 13 where no zinc was
added and (b) at Location 28 where zinc phosphate  was added for corrosion
control.   In general  no  specific conclusion concerning  differences in  the
corrosivity of the  various waters can be made  by observing the appearance
of deposits formed on the electrodes in  the photographs.

          Special arrangements were required to obtain  the  data at the Long
Beach Locations  15-18 because the corrosion control  chemicals were added to
the groundwater  upstream of the point of blending CWR/GW.  Zinc phosphate
had  been fed to  the blend  of MWD  water and  very soft  groundwater which
was  then distributed  for several years  in  a large  area of Long Beach. A
direct comparison of water quality before and after  adding the zinc corrosion
inhibitor  was desired.   To accomplish this, a plan was  implemented  with the
cooperation  of the Long  Beach  Water Department  staff  to  discontinue  the
application of zinc phosphate for six weeks to perform  the" mobile laboratory
tests and to  monitor  14 established points  in  their distribution system for
iron content.

          During this temporary cessation of the zinc  phosphate treatment,
the  laboratory   staff  of  the Long  Beach  Water Department  collected and
analyzed  six  sets of  samples   taken  weekly to determine  the iron levels.
These data  are compared in Table 15 with the average iron content of  seven
sets  of   samples  taken  immediately  before  the above  test  period.   The
difference in iron levels for the six-week period after stopping the corrosion
control treatment was not significant.

          It  seems probable that the protective deposits laid down during
long-continued application  of  zinc  phosphate may  have  a  carryover  effect
for several  days or weeks after terminating  the treatment.  For example,  a
zinc concentration  of  0.12 mg/L was noted at  Location 17 some three weeks
after ceasing to  feed  the zinc  compound.   At that  same time the  zinc  back-
ground levels in  the water supplies  were 0.03 mg/L  and 0.01 mg/L for Long
Beach  groundwater   and   Metropolitan's  CRW,  respectively.   Later  at
Location 18,  there  was  no  evidence  of leaching  or  sloughing of  these
deposits.  Nonetheless,  some residual  benefits  of the fili^s deposited earlier
may have persisted for  the entire  six weeks  when no zinc phosphate was
being  fed.  While the  data definitely suggest that continuous feeding of zinc
phosphate corrosion control  chemicals may not be required, a longer cessation
of feeding the inhibitor is needed to prove this conclusively.

          In the summer of  1978,  six  months after the mobile  laboratory
tests  at  Location 20,   the feeding  of this  corrosion   control chemical was
discontinued in the service area of  the  Walnut  Valley Water District  because
of  budgetary  constraints.   Even  though  they  have  not  conducted an
extensive  water  quality -monitoring   program,  their  staff  reports   that
                                     73

-------
       (a)    Location 13.  10/21/77
             Jensen Filtration Plant
             No Zn3(P04>2 added.
        (b)   Location 28 5/26/78
             Long Beach Harbor
Figure 20. Photographs of corrosion electrodes.
                   74

-------
             TABLE 15.  IRON IN LONG BEACH DISTRIBUTION WATER

Long Beach
Sampling
Points
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Avg. of 6 samples
when Zn feed
was off
mg/L iron
0.04
0.03
0.02
0.03
0.02
0.04
0.05
0.02
0.02
0.03
0.02
0.04
0.06
0.08
Avg. of 7 samples
when Zn feed
was on
rag/L iron
0.02
0.03
0.02
0.02
0.02
0.03
0.03
0.02
0.02
0.03
0.02
0.03
0.05
0.07
Difference in
iron when Zn
feed was off
mg/L iron
0.02
0
0
0.01
0
0.01
0.02
0
0
0
0
0.01
0.01
0.01

consumer  acceptance of the water  has remained  good with  no appreciable
rise in corrosion complaints.

CORROSION COUPON TESTS

          Mild  steel coupons placed in the water received at Walnut Valley
Water  District  in  April  1977  (CRW)  before  and  after treatment  with  the
corrosion  control chemical indicated a  16 percent reduction in corrosion rate
after  adding  the  zinc  compound.  Coupons  exposed in  August 1978  to  a
blend  of  CRW/SPW  showed no difference in corrosion rates  after treatment
with  zinc  phosphate.   Thus,  the  coupon tests  appeared to confirm  the
consumer observations.

          Similar tests performed in June  and July 1976 by the Long Beach
Water  Department  staff indicated the  great influence of flow  rates through
the corrosion cell on the results:
                                    75

-------
                                           Corrosion Rate
                         No  zinc         With zinc         Percent
     Flow Rate           or  phosphate    and phosphate      Reduction

     8.0-8.6 f.p.s.       2.9 mils/yr.     1.2 mils/yr.        59
        O.Q5 f.p.s.       12.0 mils/yr.    11.0 mils/yr.          8.3

Duration  of  these   tests  was  four  to  five  weeks.   These  results  are
presented  to show  the  complexity  of the  problem  when  one attempts to
evaluate the  benefits  of a  corrosion  control  procedure.   Obviously,  flow
rates must  be taken into account on coupon tests just as  was  found to be
true  for the  polarization   corrosion  rates  measured on-board the  mobile
laboratory.

          Corrosion  rates were determined at Alhambra  on coupons  of  SAE
1018  mild  steel exposed for  65  days  to various blends of Colorado River and
State project water  during the period from November  1977 to December 1978.
The  tests  were performed  by  the  producer  of  Virchem 932  inhibitor.
Corrosion  rate   reductions  of  90-94 percent  were reported for  the water
treated with the corrosion inhibitor.  Unfortunately, the flow rates past the
coupons exposed to untreated water were not the same as for the  treated
water.   Consequently,  the magnitude  of   the  beneficial  effect  is  more
qualitative  than quantitative,  but it   agrees  with  consumer acceptance
reported  in  this  distribution system where surface water from MWD  is being
blended with the local groundwater.

INTERRELATIONS BETWEEN  CCDT
     AND OTHER WATER QUALITY PARAMETERS

          The   effects  of   calcium hardness,  alkalinity,  pH  and  zinc
phosphate on  the CCDT  measurements  have  indicated some  definite  corre-
lations  for  several  of the  waters  discussed  in  the  above  sections.  The
usefulness  of CCDT as a general parameter  to characterize the stability or
corrosivity  of  a water will depend on its  relation to other  water  quality
parameters.   To  look  for such interrelations for  all the waters studied, the
CCDT  data  were  plotted   against  the  calcium and  alkalinity levels in
Figure  21,  against  the polarization corrosion rates for mild  steel and zinc in
Figure  22,   and  against  the  Langelier  index  and  the Ryzner  index in
Figure  23.   All  of  these   figures   illustrate  the  marked effect  of  zinc
orthophosphate  on  accelerating  film  formation   as  measured  by  CCDT.
Consequently,  those  few points must be considered as a separate group in
the  ensuing discussion.

          A strong  trend  toward higher CCDT  values  was associated  with
increases in calcium content and  alkalinity of the southern  California waters
studied (Figure 21).   This  was not  unexpected since the  theory  of calcium
carbonate film formation is based on  calcium hardness,  alkalinity and pH.

          McClelland  and Mancy  (5,15)  measured  the scaling characteristics
of  tap  water from  a  number  of  Michigan cities  by the CCDT  method.  On
the  basis  of  their  data,  they  suggested  guidelines  for  interpreting the
scaling  potential of  waters by  this test.   With CCDT values >3,  waters

                                     76

-------
   100
    50
                                            WATERS WITH
                                                                                                                   ADDED
    20
^  10
 c
ui   5.0
a.

O
-i
CO
Q
U
O
    2.0
    1.0
    0.5
    0.2
                                    _L
I
                                       V
                                                                                 .I  *l
I
        20
                   40
                               60         80         100        120    60        100        140        180



                                CALCIUM (mg/l)                                      ALKALINITY (mg/LasCaCO3)




                                           Figure 21.  CCDT results vs. calcium and alkalinity levels.
                                                                   220

-------
                                                                                       () WATERS WITH
oo
                 100
                  50
                  20
                  10
•»  5.0
m
a.
O

V)

Q
O
0  2.0
                 1.0
                 0.5
                 0.2
                                           MILD STEEL
                                            T
                                                              ,
                                                                                 ZINC
                                                        810                2

                                                     CORROSION RATE ON DAY 7 (mils/yr)
                                  Figure 22.  CCDT results vs. corrosion rates for mild steel and zinc on day 7.

-------
                                                                                        WATERS WITH
    100
     50
                                                                                                                    ADDED
a.

o
_l
in
a
o
o
     20
    10.0
5.0
    2.0
     1.0
    0.5
    0.2
                                             I    •    I
                                                                                I  i  i  i
                                                                                              .  I •  f I  I  1  I   I  I  I
               -0.6      -0.4       -0.2        0        0.2        0.4
                          LANGELIER SATURATION INDEX
                                                                        7.0          7.5          8.0         8.5
                                                                                  RYZNER STABILITY INDEX
                                           Figure 23. CCDT results vs. Langelier and Ryzner indices.

-------
known to  form films that reduce carrying  capacity were  indicated.  Waters
with  CCDT values  between 0.2 and 0.75 can be  expected to  produce thin
hard  films of  calcium  carbonate,  which  will  effectively  protect against
corrosion.   Waters  showing CCDT  values >0.75 but <3.0 may  cause  films
to be deposited which will, in time, affect carrying capacity.

          Excluding those  waters  treated with  zinc phosphate,  only  three
waters were found to have CCDT values  >3.0 in this study.  Five locations
had measured  CCDT values between  0.2  and 0.75, and none was  less than
0.2.

          We  believe  that  the data  obtained  from  this project and the
previous  study by  McClelland and  Mancy  have shown  that  the  CCDT is a
very  useful parameter  for evaluating film  formation potential of drinking
water supplies.   Nonetheless,  additional  data on  many types  of water are
needed  to develop  more  meaningful  guidelines for interpretation  of the
results.

          For  the waters  not  treated  with  the zinc compound,  there was no
evidence of an inverse relation between the corrosion  rate for mild steel and
CCDT values (Figure 22).  There  was, however,  a moderate inverse trend
between  corrosion rates  for zinc and  the CCDT  measurements as also  shown
in Figure 22.

          When the  CCDT  data were plotted against the Langelier  index,
Figure  23,  the  points  were  quite  scattered  and there  was  only  slight
evidence  of  a  relationship between these parameters for the  types  of waters
studied in this area.

           In  general,   CCDT  results plotted  against  the  Ryzner  index
showed  lower  CCDT   values  correlated  -with  high  Ryzner  indices  in
Figure 23.   This  should  be expected  since  higher Ryzner  indices indicate
an increase in the calcium  carbonate  dissolving  tendency  and  low  CCDT
values  indicate  slight to  moderate  rates  of calcium carbonate  deposition.
The  impirical Ryzner index was devised to assign a greater  influence  to the
calcium  content  of a  water  than  the Langelier index.  These  data  would
indicate  that Ryzners's modification to the Langelier index has made it more
sensitive  to the  calcium carbonate deposition tendency  as measured  by the
CCDT method.

           It  is   interesting  to note  that  the  great  impact  of  the  zinc
phosphate treatment on CCDT did not cause a  cluster  of points  with  lower
corrosion  rates  for mild  steel and  zinc.   This  failure  of the  monitored
corrosion  rate for mild  steel to  be  reduced  was also  pointed  out  when
discussing paired  data  for  the  same  quality  of water  before and  after
adding the  corrosion control  chemical.  Other investigators  have suggested
the need  for  coupon tests and observations of metallic ion picJc-up in  the
distribution  system as  a  means  of overcoming   the  inadequacies  of the
various instantaneous or accelerated corrosion parameters.
                                     80

-------
RESPONSE OF MONITORING  SYSTEMS TO QUALITY CHANGES

          The  mobile  laboratory  monitoring  systems  afford  an  excellent
means  of  detecting quality changes in  a  distribution system.  They have the
capability  of  responding  to those differences in water  quality which can
then be used to make qualitative comparisons of the relative stability and/or
corrosiveness of waters blended in a distribution system.

          This capability of sensing changes  can  be seen  in typical portions
of the  graphs plotted from data recorded  by the laboratory  computer at
Locations  25 (Figures 7 and 8) and 29 (Figures 24 and 25).  The well water
pumped into the  distribution system had a much higher alkalinity  than the
blend  of  CRW/SPW  being  delivered  to Location  25,  but  there  was  little
difference in the  hardness of  the  two waters.  At  Location 29,  both the
hardness  and  alkalinity  of the  well  water  was much higher  than in the
CRW/SPW blend.

          Whenever a high percentage  of groundwater arrived at Location 25
lower readings were observed on  polarization corrosion rates for  mild  steel
and  zinc  but slightly  higher rates were  recorded for copper.  The latter
response  is  probably due  to  the higher  free  carb'on  dioxide   content,
14 mg/L,  and lower  pH of the well water.   The  corrosion rates during the
periods when a low  percentage  of groundwater was  present are designated
in Table  8  as Location 25a; the  rates for the high percentage of  ground-
water  are shown as  Location 25b.  During the 9 hours each day when the
blend  contained  the  minimal amount of well water,  about  11 percent, the
corrosion  rates for  mild  steel  and zinc  were higher than for either the
CRW/SPW  delivered  by  MWD  (Table  8, Location 23)  or  the  groundwater
(Location  24)  when fed continuously to  the corrosion rate cells.

          Because   it  is   probable   that   the  electrode   surfaces  were
conditioned by each cyclic change in  water quality,  one  must  question the
significance of the numerical values for the  corrosion rates  under  each set
of conditions.  Nonetheless,  the  values do  point up  directional tendencies
which  indicate that one water may be more  or less aggressive than another
to a particular metal in  the  distribution  system.  The  data  also provide
some support to observations that constant changing  of water  quality in a
water  system  may  often be  less desirable than using one water  continuously
for long periods of  time irrespective of  the qualities of each.

          The strip  chart  in   Figure  25  for  Location  29  is  presented
principally  to show the effect of  changing both  hardness  and  alkalinity on
the  CCDT  slope  and on corrosion rate  curves.   The  automatic cut-off and
start-up  of the CCDT system enabled  the collection of data on these diurnal
changes that was not available at the onset of this study.

          When the well water  with higher alkalinity  and  hardness  than the
CRW/SPW blend  reached the laboratory at  Location 29, the  corrosion rate
for zinc  decreased  significantly though not as much as at Location  25 where
the well water alkalinity was  nearly twice that of MWD's blended water.
                                     81

-------
CO

NJ
          3 5
                I
                ti

                I
                8
                si-
                     20
                g
                >f
                o „
                S2  1.3
               L-B5Q13
V

i
                           in
                                           Figure 24. Data showing diurnal variations in water quality at Location 29 - Graph A.

-------
                LOCATION «

                      JDS
CO
u>
                             F'£  2K
                t-esoic
                 2                       3


Figure 25.  Data showing diurnal variations in water quality at Location 29 - Graph B.

-------
          Attention  is  called  to  Figure 8  where an  interesting sequence  of
film  formation  and partial dissolution was  recorded by the CCDT instrument
at Location 25.  The high alkalinity well water always reached the laboratory
during  the normal  work  day when the operators were there to  clean the
rotating gold  disc  and start another  test run  except on  weekends.   Film
formation occurred  at  a rapid rate, 17.8 jjA/min, until the  current flow was
diminished  to  about 50-60 nA when the  groundwater  was  flowing into the
sampling  system.   Then,  as  the  CWR/SPW  entered  the test cell  at night,
there was a decrease in the continuity of the original film  deposited on the
rotating  disc,  sufficient  to allow a current  flow of  about  110  jiA.  This
increase  in current indicates either a  partial dissolution  of the  film or a
change in its porosity.

          This  cyclic pattern for single 24-hour periods  was repeated each
day  after cleaning  the rotating disc.   In  the  absence of cleaning on week-
ends,  the  current  flow  dropped to  15-20 MA on  April 1 and 2 during the
hours  when well  water entered the test cell and increased to about 110 MA
again  at  night when  the  surface  water arrived.  Further study of the
characteristics  of the  film formed by the  well water  which  seems to dissolve
partially  on contact with  a water of lower alkalinity might  yield interesting
information on protective film formation.

           Changes  in  fluoride  concentration  by a  planned interruption  in
operation  of  chemical  feeders  were used  by  McClelland  and Mancy (5)  to
determine the  residence time of water in portions of  Chicago's  distribution
system.  In the  present  study,  changes in  concentration  levels  of one  or
more parameters,  such as,  alkalinity,  hardness,  chlorides or conductance,
could  have been used to trace the movement  of local well  water or MWD
water through several of the distribution systems.

COSTS OF CORROSION CONTROL TREATMENT

           The  most widespread  practice  employed to  improve water quality
by  mitigating corrosion  is  pH  adjustment with caustic soda or Lime.   MWD
uses caustic  soda  to offset the  pH  lowering  effect of   alum  added as  a
coagulant during treatment and chlorine added for disinfection.   The cost of
pH  adjustment varies from an  average  of  $0.55  per acre foot  of  water
treated to a maximum of  $1.52.   The  dosage of caustic  soda is varied to
maintain  a positive Langelier  Index of 0.1 to  0.2  as the  water leaves the
filtration plants.

           Blending of MWD's  treated  water  with local well waters having
moderate  amounts  of free  carbon  dioxide  and  lower pH may  cause  the
blended  water to have a negative saturation  index.   Several years  ago the
 application of  pH  adjustment chemicals in  a  few systems  was found  to  be
 efficacious  in  stabilizing   the   blended  water and  minimizing  consumer
 complaints concerning water quality.

           More recently three  utilities have  been using zinc phosphate to
 stabilize  pipeline  deposits  and  to minimize "red  water"  occurrences.  The
 annual cost of the treatment chemicals at Long Beach, Walnut Valley  Water
 District, and  Alhambra is about  $100,000, $47,000 and $20,000,  respectively.

                                     84

-------
Unit costs were $1.57, $3.91 and  $6.56 per  acre-foot of water treated in
each of  these  distribution  systems, respectively.  Because these costs are
considerably  higher  than  the  cost  of pH adjustment and  must be added to
the overall cost,  each utility is  concerned about more economical means of
achieving adequate corrosion control in their system.

          As pointed  out  earlier, the addition of zinc phosphate certainly
enhances  film formation as  measured by  CCDT, but  the comparative polari-
zation  corrosion rates  were not  conclusive in  support of a benefit  there-
from.  In view of the  persistence of zinc in  the  Long  Beach  system for
some  three weeks after  cessation  of zinc phosphate  feed,  a program of
intermittant feeding of the  corrosion inhibitor is.suggested to reduce costs.

          To further define such a feeding program:  the present chemical
application should  be cut  off completely  for a period of 4 to 6 weeks;  and,
then, the  chemical should be fed continuously for 4 to 6 weeks.  This cycle
should be repeated until monitoring of iron and zinc content at representa-
tive sampling  points  indicate that longer  or shorter  on-  and  off-periods
result  in  optimum stabilization of  water quality at  minimum cost.  The possi-
bility that the  most  desirable off-period may be  longer than  the on-period
should not be overlooked.
                                     85

-------
                         REFERENCES
1.    Black and Veatch,  Consulting Engineers.   Economic Effects  of
     Mineral  Content  in  Municipal  Water  Supplies.   Research and
     Development  Progress  Report  No.  260,  Office  of  Saline  Water.
     U.S. Government Printing Office, Washington,  B.C., 1967.

2.    Metcalf  and  Eddy,  Engineers.   The  Economic  Value of  Water
     Quality.   Research  and  Development  Progress  Report  No. 779,
     U.S. Government Printing Office, GPO:  -Washington, D.C., 1972.

3.    Orange  County Water  District.   Water  Quality  and Consumer
     Costs, Santa  Ana, California, 1972.

4.    Pearson, Harold E.  and P. R.  Singer.  Water Quality Considera-
     tions  in Water Distribution  from Two  Sources.   Journal  of the
     American Water Works Association,  66:600-605,  1974.

5.    McClelland, N.  I. and K. H.  Mancy.  Water  Qaulity  Monitoring in
     Distribution Systems. U.S. Environmental Protection Agency
     Technical   Report   No. EPA-600/2-77-074.    National  Technical
     Information Service,  Springfield, Virginia, 1977.

6.    Environmental Protection Agency.  Proposed  Amendments  to the
     National   Interim   Primary   Drinking  Water.  Federal  Register
     44(140):42246-42260.  July 19, 1979.

7.    Thurnau,   R.C.,   Improvements  in   the  Continuous  Analysis  of
     Calcium, Total Hardness, and Nitrate by  Ion-Selective Electrode.
     American   Water   Works   Association   Technology   Conference
     Proceedings,  San Diego,  California, December, 1976.

8.    Brown,  Eugene; M.  W.  Skougstad;  and M. J. Fishman.  Methods
     for  CoEection and Analysis of Water  Samples for Dissolved Minerals
     and  Gases:   Techniques of Water-Resources  Investigations of the
     U.S.  Geological Survey;  Book 5, Chapter Al, Wasington,   D.C.,
     Superintendent of Documents, U.S.  Government  Printing  Office,
     1970, 160 pp.

9.    Standard . Methods  for the Examination of Water  and Wastewater.
     14th Edition.  American  Public Health Association, American Water
     Works   Association,   and  Water  Pollution   Control   Federation,
     Washington, D.C. 1975.  pp.  1193
                               86

-------
10.    Larson,  T. E.   Corrosion  by Domestic  Waters.   Bulletin No. 59.
      Illinois State Water Survey, Urbana, Illinois, 1975, 48 pp.

11.    Ryder,  Robert A.,  Methods of  Evaluating  Corrosion.   American
      Water Works   Association,  Water  Quality  Technology  Conference
      Proceedings, Louisville, Kentucky, December,  1978.

12.    Langelier, M.  A.  The analytical  Control of Anti-Corrosion Water
      Treatment.   Journal  of the American  Water  Works  Association,
      28:1500, 1936.

13.    Larson-,  T. E.  and A.  M.  Buswell.  Calcium  Carbonate  Saturation
      Index and Alkalinity  Interpretations.   Journal of the  American
      Water Works Association, 34:1664, 1942.

14.    Ryzner,   J.  W.  A  New Index  for  Determining  the  Amount of
      Calcium  Carbonate  Scale  Formed  by  a Water.   Journal  of the
      American Water Works Association,  36:472, 1944.

 15.   McClelland, N.  I. and  K. H.  Mancy.  CCDT Bests Ryzner Index
      as  Pipe   CaCOs Film  Predictor.   Water and  Sewage Works, 126,
      No. 6, pp. 77-81, June 1979..
                                87

-------
                       APPENDIX A

                   DAQ PROGRAM LISTING
              DATA ACQUISITION WORKER TASK
        (As Modified by MWD--Modifications Only)
 LOG
   OBJECT
1530  0006
1531  0000
1532  OOOA
1533  0000
            SIX
            CNTR
            TIME
            TIME1
                                  SOURCE
16F1
16F3
16F5
16F7
16F9
16FB
16FD
16FF
1701
1703
1705
1707
15AF
1430
15AF
3406
3407
3408
3409
340A
340B
340C
340F
7082
9012
9031
9033
0000
0000
0000
0000
0800
0000
0000
0000
1749
INIT
1711
1713
1715
1717
1719
171B
171D
171F
1721
1723
1725
1727
1729
172B
172D
1C14
7082
7082
7082
1430
7082
4401
OC1F
7082
4881
4400
5080
4880
4483
7980
8140
171D
1719
171D
9031
1725
1531
1723
1749
1531
1532
0001
1533
31AB
007F
WAIT



CNTUE

COUNT


STORE
HOLD

HOLD1


MOV
MOV
MOV
SETB
SETB
SETB
SETB
SETB
SETB
SETB
SETB
B
CMI
< B
> B
= B
MOV
B
L
ARB
B
ST
L
SA
ST
LA
SXBS
ZERO,FG
SIX, CNTR


ZERO, TIME 1
D06,0
007,0
008,0
009,0
DOA,1
DOB,0
DOC,0
DOF,0
CCDT
ADI, 320
COUNT
CNTUE
COUNT
SIX, CNTR
HOLD
1,CNTR
-1, STORE,
CCDT
1,CNTR
0,TIME
0,1
0,TIME1
3,X'3000'
*127
















1





+@DEL2

                            88

-------
172F
1731
1733
1735
1737
1739
173B
173D
173F
1741
1743
1745
1747
1749
174B
174D
174F
1751
1753
1755
1757
1759
175B
175D
175F
1761
1763
1765
1767
1769
176B
176D
176F
1771
1773
1775
1777
1779
177B
177D
177F
1781
1783
1785
1787
1789
OCOF
3008
340A
7082
340A
4483
7980
7882
340F
4483
7980
340F
7082
3 40 A
7482
340C
7482
3409
7482
3408
4483
7980
3408
7482
3407
7482
340B
4483
7980
340B
4483
7980
340C
4483
7980
340B
4483
7980
340B
7482
3407
7482
3409
7482
3406
7082
1729
0878
0800
1739
0000
31A9
007F
178F
0800
31A9
007F
0000
184C
0800
1830
0800
1830
0800
1830
0800
31 AC
007F
0000
1830
0800
1830
0800
31 AC
007F
0000
31A9
007F
0000
31A9
007F
0800
31 AC
007F
0000
1830
0000
1830
0000
1830
0800
17CA

TEST
CLOSE

OPEN
READ







CCDT
































                                       ARB   -1,HOLD1,0
                                       BBNE  DI8,1,OPEN
                                       SETB  DOA,1
                                       B     READ
                                       SETB  DOA,0
                                       LA    3,X'3000'+@DELO
                                       SXBS  *127
                                       SSB   SRTA
                                       SETB  DOF,1
                                       LA    3,X'3000'+@DELO
                                       SXBS  *127
                                       SETB  DOF,0
                                       B     FINI
                                       SETB  DOA,1
                                       BL    2,HOLD2
                                       SETB  DOC,1
                                       BL    2,HOLD2
                                       SETB  D09,1
                                       BL    2,HOLD2
                                       SETB  D08,1
                                       LA    3,X'3000'+@DEL3
                                       SXBS  *127
                                       SETB  D08,0
                                       BL    2,HOLD2
                                       SETB  D07,1
                                       BL    2,HOLD2
                                       SETB  DOB,1
                                       LA    3,X'3000'+@DEL3
                                       SXBS  *127
                                       SETB  DOB,0
                                       LA    3,X'3000'+@DELO
                                       SXBS  *127
                                       SETB  DOC,0
                                       LA    3,X'3000'+@DEL3
                                       SXBS  *127
                                       SETB  DOB,1
                                       LA    3,X'3000'+@DEL3
                                       SXBS  *127
                                       SETB  DOB,0
                                       BL    2,HOLD2
                                       SETB  D07,0
                                       BL    2,HOLD2
                                       SETB  D09,0
                                       BL    2.HOLD2
                                       SETB  D06,1
                                       B     CONTUE
(178F - 17C9  SRT A)
                           89

-------
17CA
17CC
17CE
17DO
17D2
17D4
17D6
17D8
17DA
17DC
17DF
17EO
17E2
17E4
17E6
17E8
17EA
17EC
17EF
17FO
17F2
17F4
17F6
17F8
4483
7980
340B
4483
7980
340B
7482
340C
4483
7980
340B
4483
7980
340B
7482
340C
7482
3406
4400
1C32
7082
7007
5080
7082
31A9
007F
0800
31AE
007F
0000
1830
0800
31A9
007F
0800
31AB
007F
0000
1830
0000
1830
0000
1532
8009
1731
0000
0009
1729
CONTUE
1830  4480 0000  HOLD2
1832  OC02 1832
1834  72A2 0002
LA
SXBS
SETS
LA
SXBS
SETB
BL
SETB
LA
SXBS
SETB
LA
SXBS
SETB
BL
SETB
BL
SETB
L
CMI
< B
> HOP
= SA
B
LA
ARB
B
3,X'3000
*127
DOB,1
3, X1 3000
*127
DOB,0
2,HOLD2
DOC,1
3, X* 3000
*127
DOB,1
3,X'3000
*127
DOB,0
2,HOLD2
DOC,0
2,HOLD2
D06,0
0,TIME
TIME, 9
TEST

0,9
HOLD1
0,0
2,$,0
2,2
1 +@DELO


1 +@DEL5




1 +@DELO


1 +@DEL2















184E  8400 0852
                                  BFNE   (0,0),1,WAIT
                            90

-------
                       (     Start      j
Initialize Set Values
FG
CNTR
TIME 1
DO-6
DO-7
DO-8
= 0
= 6
= 0
= 0
= 0
= 0
DO-9
DO-A
DO-B
DO-C
DO-F

= 0
= 1
= 0
= 0
= 0

c
     WAIT
Begin  time  cycle
  [   Set CNTR = 6    [
B) b,
1 W
NU<

Wait 1 min "TIME" times |
                                    Set DO—A on
               NO
      Set DO-A off
       Wait  3.4  sec
           CallSRT A
         (sample instruments)
       Set DO-F on
       Wait 3.4  sec
       Set DO-F  off
       Go  to Wait
        Figure A-1.   Flowchart of DAQ (MWD modification No. 3)

                                               91
                                                                           Set  DO—A  on
                                                                           Set  DO—C  on
                                                                           Set  DO—9 on
                                                                           Set  DO—8 on
                                                                           Wait  45  sec
                                                                           Set DO-8 off
                                                                           Set DO-7 on
                                                                           Set DO—8 on
                                                                           Wait 45  sec
                                                                                J_
                                                                           Set  DO-B off
                                                                     Wait  3.4  sec
                                                                           Set  DO-C off
                                                                           Wait 45  sec
                                                                                _L
                                                                          Set DO—B on
                                                                           Wait  45  sec
                                                                          Set DO-B off
                                                                          Set DO-7 off
                                                                          Set DO-9 off
                                                                          Set DO—6 on
                                                                                _L
                                                                           Set DO—B on
                                                                           Wait  4  min
                                                                          Set DO-B off
                                                                          Set DO-C on
                                                                               _L
                                                                     Wait  3.4  sec
                                                                [    Wait  3.4 sec   |
Set DO-B on

Wait


1 min

Set DO-B off
Set DO-C off
Set DO— 6 off

-------
     SAMPLE AND STORE
 CHANNELS AD-1 THRU AD-12
 BID DATA PROCESSING TASK
           DATA
          RETURN
                                                                               NO
                                                                 GET FLAGWORD, DATE/TIME AND
                                                                  TWELVE READINGS FROM DAQ
                                                                                              STORE DATE/TIME
                                                                                               AND READING IN
                                                                                              OUTPUT STORAGE
                     PARAMETER^  YES
                       READ?
SET POINTER
 FOR NEXT
PARAMETER
Figure A-2. Flowchart of DAQ Subroutine A.
      Figure A-3. Flowchart of DATA (MWD modification).

-------
f    START   J
 in:
   WRITE
   HEADER
 SET PARAMETER
   COUNTER
 GET TIME AND
 VALUE, WRITE
   RECORD
  DECREMENT
   LENGTH
                      UNSUSPENDED
GET LENGTH
^
^
NEXT
PARAMETER
                PARAMETERS
                  DONE?
 WRITE
HEADER
SUSPEND
 SELF
              Figure A-4. Flowchart of PUN.
                    93

-------
                                       TECHNICAL REPORT DATA
                               {Please read Instructions on the reverse before completing)
  REPORT NO.
     EPA-600/2-80-132
                                                                   3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
     WATER QUALITY EFFECTS RELATED TO BLENDING
     WATERS IN DISTRIBUTION SYSTEMS
                                                                   5. REPORT DATE
                                                                      August 1930 (Issuing Date)
              6. PERFORMING ORGANIZATION CODE
 . AUTHOR(S)

     Warren K, Schimpff and Harold E. Pearson
                                                                   8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

     The Metropolitan Water District of Southern California
     Los Angeles, California 90054
               10. PROGRAM ELEMENT NO.
                  61C1C, SOS 1, Task 16
               11. CONTRACT/GRANT NO.

                  R804709
12. SPONSORING AGENCY NAME AND ADDRESS
     Municipal Environmental Research Laboratory — Cin., OH
     Office of Research and Development
     U.S. Environmental Protection Agency
     Cincinnati, Ohio 45268
               13. TYPE OF REPORT AND PERIOD COVERED

                  Final, Oct. 1976 -Oct. 1978
               14. SPONSORING AGENCY CODE
                  EPA/600/14
15. SUPPLEMENTARY NOTES
     Project Officer: Marvin C. Gardels (513/684-7236)
16. ABSTRACT
            This study was  conducted to evaluate the effects of blending two or more waters of
       different quality and to relate their composition to the corrosive effects and calcium carbonate
       deposition tendency  of the water on distribution systems.  The  EPA  mobile water quality
       monitoring laboratory was deployed at 30 selected sites in Southern California where imported
       waters from the Colorado River and California aqueducts are used as delivered or blended with
       local  groundwaters. Eighteen computer-controlled parametric systems on board the laboratory
       analyzed and  recorded field data to assess water quality factors associated with corrosion and
       stability. The  waters studied could  be classified as having moderate to high hardness, alkalinity
       and total dissolved solids content.
            The  data were analyzed for significant interrelationships  relative to  pH, calcium hardness,
       alkalinity, dissolved minerals, corrosion rates, calcium carbonate deposition test (CCDT) results,
       and calculated values for the Langelier saturation and Ryzner stability indices.

            For  waters of similar chemical  composition the CCDT results were  more indicative of
       the benefits  to be derived from pH control or zinc phosphate films for mitigating corrosion
       than the polarization corrosion rates.

            Cost comparisons  for corrosion control by  use of caustic  soda to adjust pH and zinc
       phosphate as a corrosion  inhibitor were made. An experimental approach for reducing costs of
       the latter was proposed.
17.
                                    KEY WORDS AND DOCUMENT ANALYSIS
                    DESCRIPTORS
                                                    b.lDENTIFIERS/OPEN ENDED TERMS
                              c.  COSATI Field/Group
     Water quality
     Chemical Analysis
     Corrosion Prevention
     Data Acquisition
     Distribution Systems
     Potable water
     Mobile Laboratories
     pH Control
13 B
 18. DISTRIBUTION STATEMENT


      RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)

     UNCLASSIFIED
                                                                                   21. NO. OF PAGES
 106
                                                    20. SECURITY CLASS (Thispage)

                                                         UNCLASSIFIED
                                                                                   22. PRICE
 EPA Form 2220-1 (Rev. 4-77)
                                                  94
                                                                   U.S. GOVERNMENT PRINTING OFFICE:  1980--657-165/0062

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati 01-1*45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA-335
Official  Business
Penalty for Private Use, $300
                                                                                                           Special Fourth-Class Rate
                                                                                                                     Book


                                                               Please make all necessary changes on (he above label.
                                                               detach or copy, and return to the address in the upper
                                                               left-hand corner

                                                               If you do not wish to receive these reports CHECK HERE c.
                                                               detach, or copy this cover, and return to the address in the
                                                               upper left-hand corner
                                                                          EPA-600/2-80-132


-------