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
------- |