A SURVEY
TANTS ON
9
AND ECONOMIC ASSESSMENT OF THE EFFECTS OF AIR POLLU-
ELECTRICAL COMPONENTS. VOLUME I - SECTIONS I THROUGH
ITT Electro-Physics Laboratories, Incorporated
Columbia, Maryland
August 1971
.'to foster, serve and promote the
nation's economic development
and technological advancement.'
NATIONAL TECHNICAL INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
-------
ELSCYRO-fWSJCS LABORATORIES INS.
PROJECT REPORT 179
A SURVEY AND ECONOMIC ASSESSMENT
OF THE EFFECTS OF
AIR POLLUTANTS ON ELECTRICAL COMPONENTS
Volume I
Sections 1 through 9
PREPARED FOR
DIVISION OF ECONOMIC EFFECTS RESEARCH
AIR POLLUTION CONTROL OFFICE
ENVIRONMENTAL PROTECTION AGENCY
ATTENTION: MR. JAMES B. UPHAM
FINAL REPORT ON
CONTRACT CPA 70-72
August 1971
-------
TABLE OF CONTENTS
VOLUME I
Section
1.
2.
3.
SUMMARY
INTRODUCTION
THE ELECTRONIC COMPONENTS INDUSTRY
2. 1 Characterization of the Industry
2. 2 Major Electronic Component Manufacturers
INFORMATION SOURCES
3.1 Literature Search
3. 2 Telephone Interviews and Field Visits
3. 3 Special Survey of ITT Corporate Divisions
TECHNICAL ASSESSMENT OF THE EFFECTS OF
AIR POLLUTION
4. 1 General Effects of Air Pollutants on Electronic
Components
4.1. I Theoretical and Experimental Determinations
of Materials Damage Effects
4.1. 2 Observed Damage Effects to Electronic
Components
4. 2 Case Histories
4.2.1 Navigational Aids Facility - FAA
4.2. Z Weather Radar - U. S. Weather Bureau
4.2. 3 Telephone Switching Equipment - N. Y. Bell
Telephone Co.
4.2.4 Precision D-C Motors - Western Gear Company
4.2. 5 Integrated Circuits
4.2. 6 Vacuum Components - ITT Jennings
4. 2. 7 Instrument Cleaning - M. P. Odell Company
4.3 Statistical Observations
4.3.1 Methods of Analysis
4. 3. 2 Navigational-, Aids
4. 3. 3 Weather Radar
4.3.4 Conclusions
ECONOMIC ASSESSMENT OF THE EFFECTS OF
AIR POLLUTANTS
5.1 The Cost of Missing Information
5.2 Basis of Economic Analysis
5. 3 Cost Assessment as Determined by the Survey
5. 3.1 Semiconductors
5.3.2 Integrated Circuits
5. 3. 3 Television Picture Tubes
5.3.4 Capacitors
Page
1-1
2-1
2-1
2-3
3-1
3-1
3-1
3-2
4-1
4-2
4-2
4-8
4-10
4-10
4-11
4-11
4-12
4-12
4-14
4-15
4-17
4-17
4-18
4-19
4-24
5-1
5-1
5-2
5-4
5-5
5-5
5-6
5-6
ii
Section
6.
7.
8.
9.
TABLE OF CONTENTS (Cont'd)
5. 3. 5 Resistors
5. 3. 6 Power and Special Purpose Tubes
5. 3. 7 Connectors
5. 3. 8 Transformers
5. 3. 9 Relays
5. 3. 10 Receiving Tubes
5.3.11 Crystals
5.3. 12 User Costs
5. 3. 13 Maintenance Costs
5. 3. 14 Summary of Costs
RECOMMENDATIONS
ANNOTATED BIBLIOGRAPHY
GENERAL BIBLIOGRAPHY
EXHIBITS
9.1
Interviews with Manufacturers of Electronic
Components
9. 2 Interviews with Equipment Manufacturers
Using Electronic Components
9. 3 Interviews with Users of Electronic Equipment
9. 4 Interviews with Government Agencies, Professional
Associations, University Groups and Others
VOLUME II
APPENDIX A. INTERVIEWS
APPENDIX B. FIELD VISITS
Page
5-6
5-7
5-7
5-8
5-8
5-8
5-9
5-9
5-11
5-12
6-1
7-1
8-1
9-1
9-2
9-9
9-10
9-12
A-l
B-l
iii
-------
LIST OF TABLES
Table
2-1
2-2
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
5-1
Component Categories
Representative Manufacturers
Metals Used in Electronic Components
Primary Air Pollutants
Air Pollutant Damage Mechanisms
Effects of Air Pollutants on Electronic Components
Electronic Equipment Failures at Seven VOR Stations
Standard Correlation Values
Electronic Equipment Failures at 31 VOR Stations
Weather. Radar Component Failures
Summary o'f Costs, Effects of Air Pollution on
Electronic Components
Page
2-2
2-5
4-3
4-5
4-6
4-7
4-20
4-20
4-21
4-23
5-13
SUMMARY
An assessment was made of the economic impact of air pollu-
tion on electronic components. The study was performed for the National
Air Pollution Control Administration, now the Air Pollution Control
Office of the Environmental Protection Agency, during the latter part of
1970 and extending into the spring of 1971.
Electronic components were divided into eleven different
categories as defined by the Department of Commerce. The manner and
amount of damage from air pollutants that might have been expected in
each of these component categories was assessed by surveying the litera-
ture describing pollutant material damage mechanisms. The expected
effects were compared with the actual experience of major manufacturing
companies that in total probably accounted for most of the sales in each
category. Where the literature survey indicated that gaseous sulfur
compounds, notably sulfur dioxide, should be expected to account for
most of the damage to electronic components, interviews with manu-
facturers revealed that particulate matter, acting in at least five different
ways, actually was responsible for most of the electronic component
and equipment malfunctions currently being experienced.
An attempt to relate levels of particulate concentration with
equipment failure rates by means of a statistical analysis was not
conclusive.
The economic effects of air pollution were represented by
three cost categories. These were applied to each component category
and then summed to arrive at a total annual cost of about $15. 5 million.
-------
1. INTRODUCTION
Air pollution is costly in its effects, its control, and its
abatement. It damages materials, degrades performance or, in extreme
cases, leads to catastrophic failure; and these are only the physical
effects. Biological effects, especially those perceptible to humans, can
be equally devastating. Materials and components are costly to replace,
repair or clean; protecting these things from damaging pollutants or
abating the source of pollutants, if this is even possible, may be more
costly.
This aspect of air pollution has been surmised for as long as
air pollution has been a disturbing factor in our economy, but only re-
cently have studies been undertaken to make quantitative assessments of
the magnitude of its economic impact. These studies have been conducted
by die Materials Branch of the Economic Effects Research Division, Air
Pollution Control Office, a major component of the Environmental Pro-
tection Agency, and are designed to answer the question, "How much does
air pollution cost?" The data from this study, "An Assessment of the
Economic Impact of Air Pollution on Electrical and Electronic Components,
will add to that already accumulated in several other studies.
The primary objective of the study was to answer the question,
"How much does air pollution cost the electrical and electronic industry?"
The electrical and electronic industry is, of course, much too large to
be encompassed by a single study such as this, but, fortunately, the field
can be delineated rather easily. The method for doing so, described more
fully in Section 2, consisted essentially of obtaining an inclusive categori-
zation of electronic components from a U. S. Department of Commerce
publication, and then referring to several other publications to select a
list of representative companies in each category.
Thus, when "electrical and electronic industries" are referred
to in the report, the reader will recognize that the term is meant to be
all-inclusive, but that specific data usually will be an extrapolation of
results obtained from a sampling procedure.
1-1
The current state of knowledge of the mechanisms by which air
pollution damages materials and components was assessed by surveying
all available literature on the subject. The primary purpose of the survey
was to provide a basis for. meaningful discussions with electrical manu-
facturers at a later date, and these in turn led to the perception of damage
mechanisms not discovered in the literature.
This is the framework within which the survey of electrical and
electronic manufacturers is presented. Initially, it was believed that
between 50 and 60 contacts would yield a statistically significant sample.
When early results revealed a disappointing paucity of data, the sample
was expanded to over 100. In contacting manufacturers, an effort was
made to interview the Manager of Manufacturing, the Manager of
Engineering, the Manager of Quality Assurance, or some other highly
placed individual who was fully cognizant of the company's manufacturing
methods and of the steps necessary to assure a high-quality product.
Further discussion would elicit information necessary to determine which
of these steps were necessary to reduce the effects of air pollution. Ex-
cept in a few isolated cases, however, it proved to be difficult to assign
an unambiguous cost to these measures, and in no case was it possible to
determine from interviews with manufacturers a quantitative relation
between rejection rates and air pollution.
Several users of electronic equipment, on the other hand, main-
tained carefully detailed records of failure rates, and by comparing these
data with air quality data obtained in the immediate locale, statistical
relationships were calculated.
It is gratifying to note here the cooperative responses obtained
from almost all of the persons interviewed. Almost without exception,
each individual became interested in the project when its purpose was
explained to him, and supplied the requested information without reservation.
As was to be expected, however, the amount of information varied
widely from one manufacturer to another. In by far the majority of cases,
no added manufacturing costs were uncovered which could be related
l-Z
-------
solely to air pollution. In many cases, preventive measures such as clean
rooms, filters and the like would have been installed even if the plant had
been located in the cleanest spot on earth. In some cases, the manufac-
turer first became aware, as a result of our interview with him, that air
pollution might be a problem. In a few cases, we encountered companies
which had had unusual experiences or had maintained comprehensive
records or were taking unusual control or abatement measures. These
companies were visited and, when possible, case histories were prepared
describing the situation.
An additional major objective of the survey was to obtain data
needed to determine total annual cost. This was done by extrapolating
preventive costs, maintenance costs, and residual costs, all estimated
from interview data, to annual nationwide or industry-wide costs on a
case-by-case basis and then combining them into a total.
The number obtained in this way is subject to two uncertainties.
First, gaps in the data were identified during the study. Second,
the need to contain the study within prescribed time and cost limitations
means that some areas of possible significance have been omitted. On
this basis, we estimate that the total annual cost of air pollution to
electronic components is $15. 5 million and that the figure may be un-
certain by + 30%.
2. THE ELECTRONIC COMPONENTS INDUSTRY
There are almost 150 recognizable products distinguishable as electronic
components and approximately 1450 companies in the United States that manu-
facture them. To assess the economic impact of air pollution on these manu-
facturers, it is necessary to classify all of the components into manageable
categories and then to select groups of manufacturers who will represent each
category. The object, of course, is to select the classification system so that
the conclusions resulting from this study will be statistically valid representations
of the entire industry.
2. 1 CHARACTERIZATION OF THE INDUSTRY
In a broad sense, the electronic industry is engaged in applications
of that branch of science and technology which deals with techniques and devices
to direct and control the conduction of electricity through a gas, vacuum, liquid or
solid state material.
More specifically, for purposes of this report, we shall define the
Components part of the Electronic industry to include companies engaged in the
manufacture of the types of devices listed in Table 2-1. This is the categorization
developed and utilized by the Business and Defense Services Administration (BDSA)
of the United States Department of Commerce, and was selected because
(a) the number of categories is easily manageable; (b) each category is
relatively homogeneous, that is, the components contained in each category are
sufficiently similar to permit straightforward economic analysis; and ic* because
statistical and economic data were available in these categories.
The magnitude of each component category is also shown in Table 2-1.
Semiconductor devices and integrated circuits represent over one quarter of the
total electronic components industry and, as will be seen, they are also among
the components most susceptible to the effects of air pollution.
An alternate classification scheme would have been the Standard Indus-
trial Classification (SIC) established by the Bureau of the Budget in 1967. This
1-3
2-1
-------
Table 2-1. Electronic Component Categories
Component Category
Semiconductor devices
Integrated circuits
Television picture tubes
Capacitors
Resistors
Power and special purpose tubes
Connectors
Transformers
Relays
Receiving tubes
Crystals
1969 .
($ Millionf
936
751
613
472
434
380
372
315
295
238
57
4,863
Percentage
of Market
19.2
15.4
12.6
9.7
8.9 .
7.8
7.7
6.5
6._1
4.9
1.2
100
*From Shipments of Selected Electronic Components. 1960-1969, U.S.
Department ot Commerce, October 1970. Other sources may quote
different figures.
is a logical, well-structured, all-inclusive system which has the further ad-
vantage of widespread use, both within the government and in industry. The
SIC is intended to cover the entire field of economic activity, but in so doing.
Electronic Components and Accessories (SIC 367) are represented by only five1
major categories, the last of which contains all components not elsewhere
classified and includes such diverse products as crystals, resistors, recording
heads, and pulse forming networks. Assessment of the economic impact of air
pollution on such a varietal collection would seem to be difficult, if not impossible.
The classification system adopted permits simple summation of the
economic effect of air pollution on each category into a value representative
of the Electronic Components industry as a whole.
2. 2 MAJOR ELECTRONIC COMPONENT MANUFACTURERS
Two criteria were established in selecting manufacturers to represent
each component category:
• The companies selected must manufacture a determinable and
statistically significant fraction of the total components in each
category.
• The total number of manufacturers must be small enough to allow
meaningful contacts with all of them.
Ideally, the first criterion would have been satisfied by selecting the
five or six companies with the largest sales in each category. Practically, it
turned out that sales figures, except those published in stockholders reports, are
secrets guarded with militant zeal. Thus, an alternate approach was developed
which provided a qualitative but not quantitative solution to the dilemma.
In June, 1969, McGraw-Hill polled a sampling of subscribers to
Electronics magazine and asked them to indicate the names of manufacturers
whose products they preferred. A listing of 192 products was provided to each
subscriber who was asked to write in the names of the three manufactuers he
preferred for each product and to indicate his first choice. The results were
tabulated and published later in the year.
2-2
2-3
-------
In one sense, the results do nothing more than represent a popularity
poll, but in a broader sense they would seem to be valid for this study. The
readers and subscribers to Electronics, among whom are the authors of this
report, as a group probably are a reasonably representative cross section of
the electronic community. Thus, their opinions could be expected to be
reflected in their buying habits and this in turn should be reflected in the sales
of the companies from which they buy.
Another factor which was considered before deciding to utilize the
poll was the size of the sample. Over 9, 500 questionnaires were sent out; over
1, 500 returns were received, thereby assuring good representation.
In consideration of these two factors, the composition and magnitude
of the sample polled, we have assumed that the ranking of manufacturers based
on preference would be the same as a ranking based on sales.
The Product Preference Poll (PPP) categories are not identical to
the Department of Commence (DOC) categories adopted for this study, but by
simply combing appropriate PPP categories, the DOC categories were duplicated
almost exactly. Proceeding in this way, we produce Table 2-2 which lists the
major manufacturers of each type of electronic component.
What is needed in addition to this qualitative listing is a quantitative
evaluation of how large a fraction of the total sales for each category is repre-
sented by the companies listed. Since individual company sales are unavailable,
all we can say is that the companies listed probably account for most of the
sales.
There are 48 companies appearing on the list, several appearing
in more than one category, and each was contacted by telephone. During the
interviews, other branches of the company being interviewed or other
companies were sometimes suggested as sources of information. Additionally,
it soon became apparent that users of electronic components and equipment
were valuable sources of information. To complete the picture, several
interviews were conducted with government agencies, professional associations,
and university groups. All together, over one hundred interviews were conducted
to obtain the data utilized in the following sections.
2-4
Table 2-2. Representative Manufacturers
Category
Semiconductor devices
Integrated circuits (active)
Integrated circuits (passive)
TV picture tubes
Capacitors
Resistors
Connectors
Manufacturer
Motorola
Texas Instruments
General Electric
RCA
Fairchild
Fairchild
Motorola
Texas Instruments
Signetics
National Semiconductor
Vector Electronics
Sprague
Centralab
Triad
Photo Circuits
RCA
Sylvania
National Electronics
Tungsol
Admiral
Zenith
Sprague
Cornell DubUier
Mallory
General Electric
Aero vox
Erie
Sangamo
International Resistance Corp.
Allen Bradley
Ohmite
Dale
Beckman
Am phenol
Amp, Inc.
ITT Cannon
Cinch Electronics
Elco, Inc.
2-5
-------
Table 2-2. Representative Manufacturers (cont'd)
3. INFORMATION SOURCES
Transformers and inductors
Relays
Receiving tubes
Crystals
Manufacturer
RCA
Varian Associates
General Electric
Raytheon
Amperex
United Transformer
Stancor Div. of Essex International
Triad
J. W. Miller Co.
Thordarson-Meissner
Potter & Brumfield
C. P. Clare Company
General Electric
Sigma Instruments
Guardian Electric
RCA
General Electric
Sylvania
Amperex
Raytheon
CTS Knight Inc.
International Crystal
Bliley Inc.
Texas Crystals
Peterson Radio
3. 1
LITERATURE SEARCH
Almost everyone at one time or another has encountered air pollution.
It is also true that documented evidence indicates the existence of air pollution
as early as the 16th century. Thus, history and experience have combined to
produce an abundance of literature on the subject which is at once gratifying in
extent and staggering in magnitude. The job of perusing it, fortunately, has
been simplified by several well-indexed abstract bulletins, those published
by APCA and the Defense Documentation Center being of especial value.
In addition, information was obtained from the following sources:
Clearinghouse for Federal Scientific and Technical Information
Institute of Electronic and Electrical Engineers
Institute of Environmental Sciences
National Bureau of Standards
NASA Electronics Research Center Library
National Assocation of Corrosion Engineers
American Association for Contamination Control
American Society for Testing Materials
Pertinent references are included in the General Bibliography, Section
8, and an annotated bibliography of the most significant documents is contained
in Section 7.
3.2
TELEPHONE INTERVIEWS AND FIELD VISITS
The background and theoretical information obtained from the
literature was supplemented by conditions currently being experienced by electronic
component manufacturers and users selected as outlined in Section 2. The com-
panies and organizations actually contacted are listed in Exhibits in Section 9.
Some of the names on the planned list are omitted from the tables, either
because a sufficiently cognizant individual could not be contacted or because company
policy forbade dissemination of the requested information. On the other hand.
several companies were contacted that were not on the planned list. These
were added often at the suggestion of persona being interviewed.
2-6
3-1
-------
The 105 telephone interviews uncovered 13 companies and organi-
zations with sufficiently interesting stories to warrant personal visits. These
are indicated in the tables by an asterisk. Case histories of some of these
facilities are included in subsection 4. 2.
3.3
SPECIAL SURVEY OF ITT CORPORATE DIVISIONS
Within the International Telephone and Telegraph System are several
companies engaged in the manufacture of six of the major components listed
in Section 2. A special survey of these companies was conducted in advance
of the planned survey, both to develop the interview procedure and to uncover
any previously unanticipated problems. Several useful results were obtained:
e It was discovered that a telephone conversation of a few minutes
could elicit a great deal more information than a written questionnaire.
• The unique story of ITT Jennings problems with the State of
California was uncovered.
• The survey forewarned of the paucity of correlative data.
• The importance of long-term effects was suggested.
4. TECHNICAL, ASSESSMENT OF THE EFFECTS OF
AIR POLLUTION
One of the earliest laboratories providing an experimental deter-
mination of the effects of air pollution was the city of London. It was an
inadvertent and unwilling laboratory, to be sure, but the effluent from every
chimney in town, combined with fog and dampness, produced conditions which
could be duplicated elsewhere only with the greatest difficulty and not at all
on the same scale. The most noteworthy effects, of course, were those
afflicting people; other effects were overshadowed by the hazards to health.
Those few who ventured to theorize on the relationship between air pollution
and the corrosion of metals (e.g., Davy in 1800 and deLa Rive in 1856) had
no impact.
At about the turn of the century, however, the situation began to
change and by 1924 W. H. Vernon, U. R. Evans and others were establishing
causal and quantitative relationships between tarnishing, moisture and hydrogen
sulfide.
Today, air pollution is a viable science and like other sciences,
it is the subject of an extensive literature.
What was once thought to be simply explainable, the interactions
between air pollutants and materials, are now recognized to be exceedingly
complex. The change has arisen, at least in part, from the recognition of
greater numbers of variables so that controlled laboratory experiments are
difficult to perform and harder to interpret in terms of the real world. This
should not be construed as an admonishment against further experimentation;
on the contrary, it emphasizes the need for additional well-conceived experiments.
The analysis and experiments which have been performed so far
lead to rational theories of pollution damage mechanisms, but for electronic
components only two pollutants appear to be of major significance. These
are participate matter and gaseous sulfur compounds and the damage caused by
the former exceeds the latter by a wide margin.
3-2
4-1
-------
In the following sections, the general effects of air pollution on
electronic components with emphasis on theoretical and observed damage
mechanisms will be examined, and this will be followed by a discussion of
what was actually observed in the survey.
4. 1
GENERAL EFFECTS OF AIR POLLUTANTS ON ELECTRONIC
COMPONENTS
Some air pollutants may attack the materials from which electronic
components are made while others may affect the component as a whole. The
literature abounds with studies of the interactions between various gaseous
pollutants and a whole host of materials, but with only a few exceptions, studies
of gross overall effects are lacking. A complete technical and economic
picture must, of course, include both microscopic and macroscopic effects.
4.1.1 Theoretical and Experimental Determinations of Materials Damage
Effects
Corrosion of metals may occur by reaction with gases, such as
high-temperature-oxidation, or with liquids. In the latter cases, metals
normally go into solutions as ions, and the basic mechanism, therefore, is
electrochemical. The theory is summarized by Muffley' ' and by Larrabee' ',
and Evans'"' devotes an entire chapter to it.
Briefly stated, local dissimilarities on the surface of a metal will
cause local anodes Tend cathodes to be formed in the presence of moisture.
Galvanic action then proceeds in well-known ways. Sparling' ' generalizes on
this theory by stating that corrosion is a frequent cause of electronic equipment
malfunction and that galvanic corrosion is the factor most often blamed.
Sparling is referring to the electromotive potential which results when certain
dissimilar metals are in close contact in the presence of an electrolyte, and he
gives tables of "couples " which should be avoided.
More specifically, many studies, "both in the laboratory and in the
field, have led to qualitative and quantitative understanding of how various
atmospheric contaminants interact with a wide variety of materials. Many of
the materials tested are used in electronic components. Table 4-1 lists the
major metallic materials used in the eleven categories of components listed
in Section 2, Table 2-1. Only metals are included because the damage effects
described in the literature are aimed mostly at these materials.
Table 4-1. Metals Used in Electronic Components
1. Semiconductors
Ag, Ni, Cu, Al, Pb, Sn, Au
2. Integrated Circuits
Same as semiconductors
3. TV Picture Tubes
Ag, Cu, Pb, Sn
4. Capacitors
Cu, Al, Pb, Sn, Brass, Steel, Cd
5. Resistors
Ag, Cu, Pb, Sn
6. Connectors (including Printed Circuits)
Ag, Cu, Al, Brass, Au, Pb, Sn, Cd
7. Power Tubes
Ag, Cu, Pb, Sn, Au, Pd, Pt
8. Transformers and Inductors
Cu, Fe, Al, Pb, Sn, Brass, Steel, Cd
9. Relays
Ag, Cu, Fe, Al, Pb, Sn, Brass, Steel, Cd, Pd
10. Receiving Tubes
Ag, Cu, Pb, Sn, Au
11. Crystals
Ag, Cu, Al, Pb, Sn
4-2
4-3
-------
The primary air pollutants are identified by Stern (ref. 7, p. 12)
and are listed here in Table 4-2. Radioactive compounds have been specifi-
cally excluded from this study. The remainder on the list, except carbon
compounds, were found to have some kind of effect on the materials from
which electronic components are made.
The possible damage effects likewise have been identified and
listed by Stern (ref. 7, p. 618). These mechanisms are included here as
Table 4-3.
The data in Tables 4-1, 4-2, and 4-3 are combined to form Table
4-4 which shows the kind of interaction to be expected between all of the
currently classified air pollutants (except radioactive compounds) and the
electronic components being considered in this study. The entries in Table
4-4 consist of one or more letters, each representing one of the damage
mechanisms listed in Table 4-3, followed by one or more numbers in paren-
thesis. The latter are references keyed to the General Bibliography, Section 8.
A few general observations are immediately obvious.
First, gaseous sulfur compounds will affect every category of
electronic component, either by chemical attack from H,S or by the sulfur acid
formed from SO, and moisture. Judging-from the amount of effort spent
studying sulfur compounds, one would conclude that these gases are by far the
most serious attackers. This conclusion is confirmed, in fact, by Tice' ', who
states that ". . . from the standpoint of corrosion, SO-, content is probably the
most critical air pollutant. "
A ^econd general observation is that very little seems to have been
done to study quantitatively.the first two damage mechanisms - (a) abrasion
and (b) deposition and removal. The first of these, abrasion, might not be
too important because electronic components in the field are almost always
protected from high velocity particles. The second, however, would seem to
merit some study, because, as will be shown in the next section, the cleaning
of electronic components is rapidly becoming a recognized method of main-
tenance. If, as suggested by Stern, cleaning of deposited material is accom-
panied by the removal of some base material, then there are definite limits to
the number of times a component or instrument should be cleaned.
4-4
Table 4-2. Primary Air Pollutants
(From Stern, A. C., Air Pollution, Volume 1, p. 12)
1. Fine Solids (less than 100 microns)
2. Coarse Solids (greater than 100 microns)
3. Sulfur Compounds (e. g. , 803, HjS)
4. Organic Compounds (e.g., aromatic and aliphatic hydrocarbons)
5. Nitrogen Compounds (e.g., NO, NOz, NH,)
6. Carbon Compounds (e. g., CO, CO2>
7. Halogen Compounds (e.g., HF, HC1)
8. Radioactive Compounds
4-5
-------
i-*
b.
c.
d.
Table 4-3. Air Pollutant Damage Mechanisms
(From Stern, A. C., Air Pollution, Volume 1, p. 618)
Abrasion
Deposition and removal - effect which occurs after repeated cleaning
to remove deposited material
Direct Chemical Attacks - Example is tarnishing of silver by H,S
Indirect Chemical Attacks - SO, may be converted to H^SO.
which deteriorates materials
Electrochemical Corrosion - Stern includes only ferrous metals in
this category; we include also metal couples capable of forming
galvanic cells
z r
»
B p
I—
-------
A third observation is that the carbon compounds, CO and COj,
cause no damage to materials. This is not surprising.
A fourth observation, not quite so obvious, is that each of the
five damage mechanisms is a function of, or at least is affected by, several
other factors, among which are temperature, humidity, sunlight and electric
field. Thus, a true representation of what we have shown in two dimensions
really requires many dimensions. This is further evidence of the complexity
of the whole problem.
The major result of this survey of the literature is that the primary
effect of air pollution on electronic components would be expected to come
from the gaseous sulfur compounds, primarily SO, and H,S. Our survey of
manufacturers, however, revealed quite another conclusion.
4.1. 2 Observed Damage Effects to Electronic Components
A series of telephone interviews and plant visits to manufacturers
of electronic components failed to reveal the widespread effect of SO2 and
HgS expected from the literature survey. Instead, most manufacturers
disclaimed aay problems with any forms of air pollution whatsoever, and
for those who had experienced difficulty, the greatest offender by far was
airborne particulates.
In general, particulates cause damage to electronic components
in five different ways, none of which is included in Stern's list of damage
mechanisms (Table 4-3). These are:
1. Interference with the photoresist process for integrated
o'rcuit fabrication.
2. Spot formation on the screens of TV and cathode ray tubes.
3. Release in vacuum of adsorbed gas.
4. Provision of microscopic leakage paths in semiconductors
and integrated circuits.
5. Provision of gross current leakage paths in all electronic
equipment.
It is to be noted that the first four of these mechanisms cause
damage during manufacture to semiconductors, integrated circuits, TV and
cathode ray tubes, and vacuum components such as vacuum capacitors,
relays and tubes. The last named mechanism, gross current leakage, occurs
in the field on finished products.
A sixth damage mechanism, also attributable to particulates, is
not included in the list because it was noted for only a specific piece of
equipment and is, therefore, a special case. Traffic controllers have rotating
parts which lose their lubrication, apparently due to the capillary action of
accumulated particulate matter. As noted in section 5, this effect results in
a significant cost.
Most manufacturers of components subject to damage by particulates
provide some degree of protection during the manufacturing process. This
can vary from nothing more than the air filters in the air conditioning system
to elaborate clean rooms with laminar flow. The amount of protection depends
upon the component and the environment in which it is manufactured.
A note of caution is necessary in interpreting these results. Manu-
facturers generally do not identify the source of the particulate matter.
Experience has taught them, however, that it is either airborne or human-
borne and many believe that the latter may be the more prolific source.
Regardless of the source, one company, the Odell Company, has
learned that accumulations of dirt in electronic equipment frequently leads to
failure or malfunction and has devised a system to clean equipment. Experience
has shown that this often eliminates the need for repairs or replacement of
parts. The company and its system are described more fully in the section
on Case Histories.
It is apparent from this survey, therefore, that particulate matter,
operating in one of the five damage mechanism modes listed above, is
responsible for most of the detrimental effects on electronic components. It
seems plausible that many of the preventive measures first used on military
equipment and later on space hardware are becoming commonplace. The
document in reference 6, for example, is in essence, a design manual for the
4-8
4-9
-------
prevention of corrosion in electronic components used by the U. S. Army.
Since, as we have seen, almost all of the work in the field has been directed
toward the study of corrosion, and since the preventive measures described
in reference 6 likewise are aimed at reducing or preventing corrosion, it
seems logical to assume that these measures have been successful and that
what we are observing today are the effects which were masked by the
heretofore more prevalent corrosive effects.
4. 2 CASE HISTORIES
Telephone interviews with manufacturers and users of electronic
components and equipment resulted in uncovering several cases directly
pertinent to this study or of unusual general interest. These cases are
presented here because they aid in-understanding both the technical and
economic problems encountered.
4. 2. 1 Navigational Aids Facility - FAA
The Federal Aviation Administration (FAA) operates approximately
900 Very High Frequency Omnidirectional Ranging (VOR) stations in the United
States to provide navigational assistance to commercial, military, and private
aircraft. These stations radiate a coded radio signal by means of which a
pilot may determine his position. The VOR station located near Charleston,
West Virginia is maintained by personnel who report to Mr. John McGivern.
Mr. McGivern relates that in 1965 and early 1966, about one year after the
equipment was installed, evidence of corrosion started to appear. The
biggest problem was sulfiding of silver-plated contacts with consequent arcing
and pitting to the extent that replacements were required. Fortunately,
aircraft traffic delays resulting from equipment downtime were minimal
because of the redundancy built into the VOR system.
The two main sources of corrosive agents in the Charleston River
Valley were the many chemical plants that emitted a variety of substances,
including gaseous sulfur compounds, and a nearby municipal burning dump
that produced all combustion products.
The solution arrived at by the FAA was to install an air-conditioning
system with special activated-charcoal filters to eliminate pollutants. This was
done late in 19.66 at an approximate cost of $6,500. Corrosion appears to
be under control at this time.
4.2.2
Weather Radar - U. S. Weather Bureau
The U. S. Weather Bureau operates approximately 40 type WSR-57
weather radar stations in the United States for the purpose of accumulating
data concerning storm systems in the area. One of these radars has its
antenna, transmitter, and receiver mounted on a chimney of a power plant
at the University of Chicago. This radar is maintained by personnel who
report to the area Engineering Manager, Mr. Harold Anderson, in Kansas
City, Missouri. Mr. Anderson related that this radar equipment was exposed
to high concentrations of SO. and participates from the stack for several
years. The equipment became very dirty and its appearance deteriorated
greatly. The structural tower supports corroded sufficiently to cause the
weather bureau to call in a structural consultant to check the integrity of the
tower. The tower'was deemed structurally sound and the equipment still
> f^mctioned normally. A check of maintenance records revealed that electronic
component failures for this equipment were no greater than for identical
equipments in locations like Wichita, Kansas, and Missoula, Montana, which
have low pollution levels (these sites are not near local sources of pollution).
It would appear from this case that the weather radar equipment is
relatively immune to attack by air pollution. This is not all surprising because
most of the components in this radar are hermetically sealed against humidity
and this also affords protection against attack by air pollutants.
4.2.3 Telephone Switching Equipment - N. Y. Bell Telephone Co.
The New York Bell Telephone Company operates a number of
central telephone switching centers in New York City. These switching
centers contain complex electronic switching networks, many of which are
computer-controlled. In 1965, one of these centers, which was approximately
five blocks away from a power-generating station, was exhibiting erratic
computer behavior. IBM engineers were called in to evaluate the situation.
4-10
4-il
-------
IBM claimed, according to Mr. E. P. Hodges, New York Bell Telephone
Buildings Engineer, that the computer was suffering from outside particulate
contamination and recommended that all intake air for the installation be
filtered.
A high-efficiency (approximately 85%) deep-bed air filter was
developed for this use and placed in series with a coarse prefilter for removal
of large particles, bugs, etc. The installation of this filter eliminated the
computer malfunctions and decreased maintenance to other switching equipment.
Over the next few yeara this type of filter system was installed in all New
York central switching buildings. The company estimates that the initial
cost was approximately 1% of the cost of the air-conditioning system.
4. 2.4 Precision D-C Motors - Western Gear Company
The Western Gear Co. manufactures a line of very high reliability
precision d-c motors for Sandia Laboratories. These motors must tolerate
several years of storage at temperatures of 225°F; then they must start
immediately and run at 15, 000 rpm. These motors are built in an ultra-clean
facility that uses banks of-High-Efficiency Particle Air (HEPA) filters in a
recirculated-air system. In addition, this type filter cleans outside air used
for down-draft cleaning stations.
The filters are changed when the particulate buildup in the filter
causes a pressure drop which exceeds a predetermined value. According
to Mr. C. W. Yost, one of Western Gear's product managers, the first set
of down-draft filters was replaced after 1-1/2 years of operation. The second
set became clogged with dirt in approximately 4 months. The dirt in these
filters was analyzed and found to have a high iron content. The source of
this dirt was traced to a construction site where an excavation was being
made for a school building. The company estimated that the cost of pre-
mature replacement of these filters was $3, 000.
4. 2. 5 Integrated Circuits
The manufacture of semiconductor devices and integrated circuits
relies heavily on photographic processes to delineate the areas on the surface
4-12
of a silicon chip into which carefully controlled contaminants (dopants) are
diffused. One of the key steps in this process is the application of a photo-
resist material to the wafer surface. This photo-resist material is a cross-
linked organic polymer that is relatively insoluble in the washing solution
until exposed to ultraviolet radiation. After certain areas of the wafer
surface are exposed to a controlled ultraviolet source through a precision
mask, the photo-resist polymer is broken down by the photochemical action
(breaking of cross-link bonds). Both the exposure time and wavelength of
the ultraviolet radiation are critical in preventing additional cross-linking
from occurring and producing changes in the characteristics of the photo-
resist material.
In 1963 and 1964, engineers at the Fairchild Company discovered
that during certain times of the year and at certain times of the day, an
insoluble scum or film would remain on the silicon wafer after the washing
operation •even though all precautions were taken. This film would prevent
diffusion of dopants and interfere with adhesion of deposited metal conductors.
In some cases, it appeared that the areas that were intended to be unexposed
actually had been partially exposed. These effects led to unacceptably poor
quality in the final devices.
It was suspected that these effects were related to photochemical
oxidants present at certain times in the atmosphere. Qualitative correlations
were established between "smoggy days" and poor yield from the photo-resist
operations. Some experiments were conducted by performing resist operations
utilizing a dry nitrogen atmosphere into which small amounts of ozone were
injected. When pure nitrogen was used, no film formed and no problem
existed; when ozone was injected, film formed with the attendant problems.
It was deduced that the photochemical oxidants were creating
additional cross-linking in the exposed photo-resist material, producing an
insoluble polymer. In the case of the unexposed material, it was assumed
that the oxidants were creating the same effect as exposure to ultraviolet
radiation, i.e., breaking cross-link bonds.
From this point on, solution to the problem took different forms
in different branches of the semiconductor industry. Some companies chose
to use different types of photo-resist materials which were immune to attack by
4-13
-------
oxidants, while some resorted to air cleaners and use of dry nitrogen
atmospheres.
The modified types of photo-resist material proved to be more
difficult to process .and handle, since they had a high solid content and required
more filtering. These materials also required more closely controlled
ultraviolet wavelength and higher intensities.
The Signetics Company chose, to utilize carbon filters in conjunction
with water screens or scrubbers to remove these damaging oxidants from
plant intake air, and the use of dry nitrogen atmospheres in critical assembly
areas. The estimated maintenance costs 'Of the air-cleaning system were
$30, 000 (7. 5 million cubic feet per month). Approximately 30% of the nitrogen
is used in the photo-resist operation.
4.2.6 Vacuum Components - ITT Jennings
The Jennings Radio Manufacturing Corporation, a subsidiary of
the International Telephone and Telegraph Corporation (ITT), operates a facility
which manufactures high-voltage vacuum capacitors, switches and relays at
San Jose, California. These components depend on a hard vacuum to operate
properly, and are adversely affected by contamination of all types. They are
particularly affected by particulates that become lodged within the vacuum
envelope and which, subsequent to vacuum pumping, outgas and produce a
conductive atmosphere within the device.
In the fall of 1967, Jennings learned that the State of California
was planning to condemn a portion of its property for construction of a free-
way ramp approximately 400 feet from their main manufacturing building.
Jennings became concerned about the possibility of contamination from con-
struction and use of the freeway, and instituted a study program to determine
whether the effects would be sufficiently deleterious to warrant a counter-suit
against the condemnation.
Jennings was not able to delay the freeway, but the company engaged
a consultant who measured particulate levels at strategic locations within the
plant, both before and after construction was started. Significant increases
4-14
were noted at all monitor stations. Thus, Jennings was faced with increased
rejection rates for its sensitive components or an immediate expenditure to
control the added dust and dirt.
Choosing the latter course, the company expended approximately
$50, 000 for upgrading air handling systems, providing entrance protection,
sealing windows and doors, installing additional air handling equipment in
some locations, and providing more efficient filters. In addition, the company
estimates that its annual costs for equipment maintenance and added janitorial
services amount to $20, 000.
4, 2. 7 Instrument Cleaning - M. P. Ode 11 Company
The M. P. Ode 11 Company started business in 1947 as a manu-
facturer's representative. The company established a service department
with facilities for calibrating both a-c and d-c instruments, and quickly noted
that a large number of instruments being returned from the Akron, Ohio,
area were inoperative simply because of a black deposit on silver switch
contacts. The black deposit, although never identified formally, was attributed
to the presence of sulfur gas arising from the numerous rubber and tire
factories in the Akron area. This was the company's first hint that air
pollution could cause malfunctioning in electronic instruments.
Later, in the 1960's, Odell began receiving inoperative or unstable
instruments in which no defective parts could be found. The company cites
one early experience of an FM tuner which was drifting so badly that it was
useless. Cleaning was the only thing necessary to restore it to its original
factory stability.
Mr. Odell then conducted further tests and experiments which
convinced him that a major cause of instrument instability or malfunction
was accumulations of dirt and grime, which caused current leakage paths.
Additional experimentation led to the adoption of a solution of detergent in
water as the ideal cleaning agent. A chemist was engaged to develop a
detergent formula which, in addition to being a good cleaning agent, would also
be non-toxic, non-dermatologic, non-explosive and biodegradable. The presently
used solvent was the result.
4-15
-------
A cabinet equipped with a turntable, lights, a spray gun for
applying the solvent and rinse water, and an air hose, plus a temperature-
controlled drying cabinet, completes the system.
In the four years since starting manufacture of its cleaning systems,
the company has encountered many examples of the deleterious effects of
any contaminants which can settle or condense on electronic components in
such a •way as to cause leakage paths. It has been observed that the most
common offenders appear to be tobacco smoke and dust, although special
- conditions have been encountered under unique circumstances. The Jones
and LaughUn Steel Company, for example, operates oscillographs in an
atmosphere apparently laden with iron oxide and cutting-oil vapor. This
particular combination rapidly degrades the performance of the instruments,
but fortunately, the dirt yields quite readily to the cleaning process.
The Magnavox Company utilizes approximately 30, 000 instruments
of various kinds in its daily operations, and while the operating environment
is not unusually dirty, a regular routine for cleaning all instruments several
times a year has been established. Other companies are apparently establishing
similar procedures.
The total annual cost of cleaning instruments may be estimated as
follows:
We know from data supplied by the M. P. Odell Company that
approximately one million instruments now go through their cleaning systems
each year. Since many instruments are cleaned routinely several times per
year, we can estimate that the one million cleanings represent, conservatively,
250, 000 instruments.
We use values of 50 cents per cleaning for the cleaning solution and
one-half hour per cleaning for labor. The latter figure includes time for dis-
assembly and reassembly, as well as the actual cleaning. Both figures were
supplied by Odell. Assuming a value of $5 per hour for labor, we arrive at
the total cost for one million cleanups of $3 million.
There is ho way of determining accurately how much of this cost
s attributable to air pollution, but from information supplied by Odell about
he environments in which typical instruments are located and the type of
ontaminants encountered, we would estimate that at least two-thirds of the
osts are related to air pollution.
:. 3 STATISTICAL OBSERVATIONS
In the preceding two sections, the damaging effects of air pollutants
>n the materials composing electronic components was examined from theo-
•etical and experimental considerations as discussed in the literature and
hese results were compared with field experiences. To complete the picture,
t would be desirable to correlate equipment failure rates (or some similar
ndex of malfunction) with pollutant levels. As may be surmised, data of this
lature do not seem to exist. People who keep records on failure rates of
ieir eq^pment do not have air quality measuring equipment and vice versa.
Lacking specific data of the kind required, but having available
iailure rate data from both the Federal Aviation Authority and the Weather
Bureau, along with air quality data from EPA, an attempt was made to
establish a correlation.
The air monitoring stations from which data were used were always
located in the same city as either the VOR or Weather Bureau stations, but
not necessarily as near as would be desired. Thus, this was in the nature of
an experiment more than anything else, and as experiments are sometimes
inclined to do, this one produced negative results.
The following paragraphs describe the method of analysis and the
results obtained.
4.3.1 Methods of Analysis
Since failure of electronic components and equipment involve many
environmental and other variables, and since the problem is in part one of
statistical failure analysis, multivariable analysis is an appropriate research
method for examining this problem. Computer methods are .essential, even
considering the amount of data involved with this effort. The computer
4-17
4-16
-------
employed was the ITT-EPL IBM 1130. In particular, multiple regresion
analysis patterned after the methods of Ezekiel and Fox'9' and Sprent'10'
was used.
The basic data comprising this study were mean annual concentrations
of the three pollutants (suspended particulates, sulfur dioxide, and nitrogen
dioxide), taken from 1966 NAPCA air quality data'11', and failures of electronic
equipment taken from FAA and U. S. Weather Bureau data.
4. 3. 2 Navigational Aids
The FAA operates approximately 900 VOR systems in the U. S.
for use by aircraft pilots in establishing their position. These systems
basically consist of signal generation and coding circuitry and transmitter
modules. Data regarding failure of these VOR stations are gathered and
compiled by the System Maintenance Service Branch of the FAA. The data
are coded so that failures due to electronic equipment are identified for each
station. Data for fiscal years 1966 and 1967 were used and covered all VOR
electronic equipment failures from July 1965 through June 1967. These data
were compared to air quality data for 1966.
The first sample utilized data from the seven cities identified in
Table 4-5.
The regression analysis was carried out utilizing the three
pollutant concentrations as the independent variables Xj, X2, X3 in the
regression equation below. X,j, the dependent variable, was VOR failures.
X4 = A0 + BjXl + B2X2 + 83X3
The regression Coefficients Bi and correlation coefficients Ri4
are given below:
Variable
Mean Value Std. Dev. Correlation Regression
Ri4 Bi
Xj Particulate level 132. 7
X2S02
X3 NO2
X4 VOR failures
Intercept Ao -6.16
Multiple correlation 0. 78
coefficient
132.7
117.5
196.8
5.28
27.60
80.76
109.7
2.19
0.661
-0.132
-0.058
0.084
-0.018
0.012
In order for a correlation between two variables to be considered
significant (i.e. , reason exists to doubt the independence of the two variables),
it must exceed standard values for a given sample size. These values from
ref. (9) for a number of sample sizes N and confidence levels r are given in
Table 4-6.
It can be seen that correlation coefficient between particulates
and VOR failures of 0. 661 does not exceed the 95% value of 0. 669 from
Table 4-6, but is sufficiently close to indicate a possible trend. There is .
no apparent correlation in this sample with either of the gaseous pollutants.
A second, larger sample of 31 cities was chose'n as indicated in Table 4-7.
Applying the same analysis to these data yielded the following
results:
- Variable
Xl Particulates .
X2S02
X3 NO2
X4 VOR Failures
Intercept Ao
Mutiple correlation
Mean Value
128.4
70.00
140. 1
4.58
0.963
0. 307
Std. Dev.
33.07
74.50
73.61
3.22
Correlation
Ri4
0. 080
0.031
0.244
Regression
Bi
0. 015
-0.007
0.015
4-18
coefficient
Again, by comparing these correlation coefficients with those of
Table 4-7, it is found that none is significant at the 95% confidence level.
One may conclude from the results of these rather limited
samplings of navigational aids failures that there is no significant correlation
with the air pollutants considered.
The author wishes at this point to acknowledge the cooperation
and assistance provided by the System Maintenance Services Division of
the FAA, and iii particular by Mr. Ed Kaneko.
4. 3. 3 Weather Radar
The U. S. Weather Bureau operates approximately 40 weather radars
in the continental United States. These S-band radars have both transmitter and
4-19
-------
Table 4-5. Electronic Equipment Failures at
Seven VOR Stations
Annual Pollutant Concentration
Location
Denver, Col.
Cincinnati, Ohio
Chicago, 111.
St. Louis, Mo.
Philadelphia, Pa.
Washington, D. C
Pittsburgh, Pa.
•
*Arithmetic mean of 26 randomly
collected over a one-year period.
Table
N_
5
- 7
10
20
30
50
100
1,000
4-6. Standard
y.95
.805
.669
.549
.378
. 306
.235
.165
.052
Failures Susp. Part.
5 126
8 154
5 124
2 143
6 154
2 77
9 151
selected 24 -hour -tesfefester
Correlation Values
y.99
.934
.833
. 715
.516
.423
. 328
.232
.062
SO2 NO2
18 68
44 234
221 333
132 89
225 288
90 271
93" 95
samples
y. 995
-.959
.875
.765
.561
.463
.361
.256
.081
4-20
Table 4-7.
Electronic Equipment Failures at
31 VOR Stations
Location
Washington, D. C.
Pittsburgh, Pa.
Baltimore, Md.
Nashville, Tenn.
Buffalo, N. Y.
Chattanooga, Tenn.
Cleveland, Ohio
Charleston, W. Va.
Cincinatti, Ohio
Dayton, Ohio
Dubuque, Iowa
Denver, Colorado
Des Moines, Iowa
£1 Paso, Texas
Hartford, Conn.
Indianapolis, Ind.
New York City
Louisville, Ky.
Kansas City, Mo.
Milwaukee, Wise.
Marlton & Mville, N. J.
Minneapolis, Minn.
Oklahoma City, Okla.
•Chicago, Illinois
Portland, Oregon
Providence, R. I.
Philadelphia, Pa.
Seattle, Wash.
Salt Lake City, Utah
St. Louis, Mo.
Youngstown, Ohio
Annual
Failures
2
9
3
8
1
4
14
5
8
4
9
5
0
7
7
4
1
2
4
0
5
4
7
9
1
3
6
7
2
4
2
Pollutant Concentration \ue/i
Susp. Fart. SO,
77
151
144
115
139
140
116
226
154
143
132
126
123
193
90
154
134
129
116
150
98
87
107
124
75
121
154
76
93
143
152
90
93
107
29
25
34
78
29
44
49
16
18
13
63
62
54
346
38
12
28
42
44
10
221
22
125
225
35
20
132
66
ND2
271
95
75
65
169
66
191
68
234
78
189
68
82
73
161
103
228
175
136
142
186
112
201
333
62
99
288
93
109
89
105
4-21
-------
receiver mounted with the antenna on a tower exposed to the outside environment.
The balance of the components making up the weather radar (data processing,
display, and communications) are located inside in an air-conditioned environ-
ment. Maintenance and component failure data for these radars are filed on
computer tape for ready retrieval. Component failures are identified as to
which portion of the radar they came from. If there are any significant effects
of air pollution on the electronic components used in this radar, one would
expect higher failure ratios for those components in non-air conditioned areas.
Component failure data for calendar year 1966 and air quality data for the
same locations are presented in Table 4-8.
Multiple regression techniques were applied to total component
failure data:
Correlation Regression
Variable
Xj SO2
X2 NO2
X3 Particulates
X4 Component
failures
Intercept, Ao
Mean Value
87.16
169.00
126.33
48. 66
160.33
Std. Oev.
103. 00
88.25
26.19
23.45
Ri4
0. 186
-0. 056
-0. 641
Bi
0.118
-0.155
-0.757
Multiple correlation 0.819
Note that none of the Ri4 are significant at the 95% level when
compared to Table 4-6 except for the seeming negative correlation between
suspended participate matter and component failure. At this time there is
no readily available physical interpretation for this result. Possibilities
include such things as a higher level of preventive maintenance in areas
where particulate levels are high. This possibility could not be confirmed
by discussion with Weather Bureau personnel.
4-22
Table 4-8. Weather Radar Component Failures
Annual
Component Failure
Non
Pollutant Concentration vug/m^/
!!?2.S<.*
-------
•-- The regression analysis was repeated for the same location using
only component failures in non-air-conditioned areas. The results are:
Variable
X1S02
Xz N02
Xj -Particulates
X4 Component
failures
Intercept Ao
Multiple correlation
Mean Value
87. 16
169.00
1Z6.33
12.41
31.79
0.563
Std. Dev.
103.00
84.25
26. 19
5.64
Correlation
Ri4
0. 030
-0.073
-0.470
Regression
Bi
0. 013
-0.025
-0. 129
Again, none of Ri are significant at the 95% level except for the
negative correlation with suspended particiilai.ee,. While this value decreased,
it ia still not explained.
_ Ti"> »eeming immunity of weather radar components to air
pollution suggested by this analysis is borne out as indicated by further
discussion with Weather Bureau personnel. These discussions (paragraph
4.2.2) revealed no higher failure rates in Chicago (high SC-2 level) than in
rural areas.
The author wishes to acknowledge the cooperation and assistance
provided by Mr. Russell Hovey of the U. S. Weather Bureau.
4. 3. 4 Conclusions
Although the results of the data analysis for both the VOR and
Weather Bureau Stations lead to the conclusion that no valid correlation exists
between equipment failures and pollutant levels, two factors mitigate the
validity of this conclusion.
First, of course, is the size of the sample. Using the VOR case
as an example, it may be seen by reference to Table 4-6 that a sample-size
approaching 1000 would be necessary to establish 95% confidence in the
magnitude of the correlation coefficients calculated. The negative coefficients
obtained from the Weather Bureau calculations are quite another matter and
4-24
do not lead to the same quantitative conclusion, although qualitatively the
same argument can be made. In any event, the small sample in both cases
casts some doubt on the validity of any correlation or lack of it.
Second, air monitoring stations from which data were used were
not necessarily located in proximity to either VOR or Weather Bureau stations,
even though the city was a common factor in all cases. Air monitoring
stations frequently are located in downtown areas, weather bureau stations
at local airports and VOR beacons several miles from the town. Thuo, there
is no assurance that the air monitored by the air station is the same quality
air experienced by either the Weather Bureau equipment or the VOR beacons.
While no valid conclusions can be drawn from these calculations,
they do suggest an approach to obtaining valid data. Simply locate air moni-
toring equipment close by the electronic equipment in question and measure
air quality and failure rates for periods of several months to a year or more.
This could be done quite economically by utilizing mobile laboratories.
4-25
-------
5. ECONOMIC ASSESSMENT OF THE EFFECTS OF AIR POLLUTANTS
5. 1 THE COST OF MISSING INFORMATION
The ubiquitous effects of air pollution ultimately manifest themselves
in our pocketbooks and yet we have only vague notions about what the effects
really cost us.
The main reason for this state of affairs seems to be that the
costs attributable to air pollution are not separable from other costs.
Manufacturers of electronic components known to be sensitive to airborne
contaminants generally filter the air entering manufacturing areas; and for
some especially sensitive components, operations are conducted in clean rooms.
These measures provide protection against particulate matter, but whether they
are borne by some other mechanism, such as humans, is not known in most
cases, nor do most manufacturers believe the distinction is important because
they would take the same precautions under any circumstances. As one
manager put it, he would build his plant the same way even if it were to be
located in the cleanest spot on earth.
On the other hand, the Manager' of Manufacturing Engineering of a
large semiconductor plant stated that he had expended about $1 million (not
at his present plant) to get class 100 air, only to discover that the amount of
particulate matter in the building was not reduced appreciably, and moreover,
that it varied with the particulate concentration in the outside air. He
concluded, therefore, that dust and dirt were being brought into the building
by his own personnel. Consequently, air filters were installed at each sensitive
work station, positive pressure ventilation was provided, and rugs were placed
at all building entrances.
Here was a case where contamination, initially thought to be air
pollution, provoked a million dollar investment which turned out to be needless.
Thus, it seems that manufacturers have no other choice than to
take the maximum practicable precautions against air pollution, simply be-
cause they cannot separate the effects of airborne contaminate from others.
5-1
Whenever these effects are separable, however, we see that it may be
possible to achieve better results at lower cost. Another factor which would
seem to contribute to lower costs, but for which data are lacking, is the
effect of contaminant concentration on component damage. Nowhere in this
study did we encounter anyone who could correlate contaminant concentration with
rejection rates or failure rates or maintenance costs, and yet it seems
intuitively obvious that to provide more protection than necessary is to incur
an unnecessary expense.
Thus, these two bits of missing informational) separability of air-
borne contaminants from others and (2) relationship between concentration and
damage effects, may be costly. In addition, the absence of them has a profound
effect on the economic analysis that follows.
5. 2 BASIS OF ECONOMIC ANALYSIS
A study of the costs of air pollution has recently been published by
Ridker. We have adopted the general approach to the problem discussed
in his Introduction, but have arrived at a somewhat different expression of
costs. Where Ridker seems to express the point of view of the economist, we
have tried to look at the picture through the eyes of a manager who has
responsibility for the quality of his product. His purpose, and the purpose of
the sections which follow, is to determine how much air pollution actually costs
him, so that he can take the most effective steps toward minimizing the costs.
With this approach in mind, we identify three major cost categories (instead
of Ridker's two). Thus:
CT - Loss Cost - if nothing is done to control or abate air
pollution effects, components are degraded and may ultimately fail. C_ represents
the cost to the manufacturer of components lost through air pollution damage and
may include secondary losses as well; that is, cost of equipment which fails
because of air pollution-damaged components.
C = Prevention Costs - costs of installing filters, scrubbers, clean
rooms, etc., necessary to provide uncontaminated air to the manufacturing
area. Costs of design changes and of special materials necessary to prevent
or abate the effects of air pollution should be included in this category, but as a
_ practical matter are not because they cannot be identified uniquely.
5-2
-------
C. = Maintenance Costs - costs to remove the effects of air pollution
once they have contaminated a component may be incurred in the factory or
in the field.
The manager is presented with the problem, in principle at least,
of selecting which of the three alternative costs, or combinations of them,
to incur. He may decide, for example, to do nothing and simply accept the
costs of the failed components. This he might do if these costs were
demonstratively lower than the cost of installing control equipment, but in this
case, he woulu doubtless rely upon the Air Pollution Control Office of the
Environmental Protection Agency to take some positive steps, or he might
be able to consider some litigation if he can identify the source of the
offending pollution.
Obviously, he also has many other strategies available to him.
What he needs is a system for expressing all of these costs which will allow
him to select a strategy for minimizing his costs. The following formulation
was developed with this idea in mind and, at the same time, it was intended
to serve as a vehicle upon which to base an estimate of total costs.
Let E=E. + E,+---E = set of electronic components (1)
P = P. + P, +---P = set of pollutants (2)
1 £, m
Then
D = £ P. \..E. = number of components damaged by all (3)
pollutants where \ .. - probability that i pollutant will damage j component.
The probability, X • *s some function of the concentration of air pollution,
but as we have seen in the section on damage mechanisms, no rigorous
relationship yet exists. The best we can do at the present time is to say
that as the concentration of air pollutants increases, the probability increases
that components will be damaged.
Returning now to costs, we see that as preventive measures are
installed, then losses ought to be reduced by some amount. Similarly,
cleaning or otherwise removing the effects of air pollution also ought to
reduce losses. This can be represented as
Cff, — CT-J + C. . + , CT - f (Co) - f (£.»«) I \^l
5-3
or the total cost of the effects of air pollution (C_,) is the sum of the cost
to prevent it (Cp) plus the cost to remove the effects (C..) plus the loss
cost (Cj) reduced by the amount of the loss saved by preventive and
maintenance measures (f(C ) and f(C,.). The loss cost may be further
represented by using Equation (3)
CT = cD = c
•L
Pi
-.,
ij j
(5)
where c is the unit cost of the component in the set E.
At one extreme, if no preventive or maintenance measures are
taken, then
(6)
At the other extreme, if preventive and/or maintenance measures
are taken, then CL may equal 0, but Cp and CM are large, and CT is also
large. The equation, therefore, has a minimum somewhere between the two
extremes, but a quantitative determination of it depends upon quantitative
knowledge of the functions f(Cp) and f(C..). On the other hand, Equations (3)
and (4) depict an approach to minimizing air pollution costs and, at the same
time, they provide a basis for a discussion of the actual costs that were
uncovered during this survey.
5. 3 COST ASSESSMENT AS DETERMINED BY THE SURVEY
In the preceding section, the total cost was represented by three terms,
one of which, the loss cost C, , varied inversely with the preventive and
maintenance costs. While this approach shows the interrelation of all
costs, another form seems more suitable for measurements. We define
a residual cost Cj.as cost of those components still lost, even after preventive
and maintenance measures have been taken
The total cost now becomes
(7)
(8)
5-4
-------
Each of the three quantities on the right can be measured, at least
in principle. Since C_ varies inversely with both C_ and C,., however,
any measurements of C_. must be interpreted with caution.
K
5.3.1 Semiconductors
Without exception, semiconductors are manufactured in controlled
atmospheres, provided either by clean rooms or filters, but in no case were
these installed exclusively to control air pollution. Most manufacturers seemed
-more concerned about human-borne dust and dirt--hence, the clean rooms--and
one manufacturer supplies filtered air at each work station, again primarily
as a safeguard against human-borne contamination. Preventive costs, there-
fore, are low.
These devices do not lend themselves to cleaning or otherwise
removing the damage effects, and one manufacturer stated that no devices
returned from the field had failed due to air pollution. Maintenance and - -
residual costs likewise are low.
5. 3. 2 Integrated Circuits
Some years ago, a major manufacturer of integrated circuits
discovered that airborne photo-chemical oxidants could corrode the photo-
resist material used in the manufacturing process (paragraph 4.2.5). Subse-
quently, other types of photo-resist material were developed which were less
sensitive to oxidants, but which were much more .difficult to apply. Thus, the
choice lies between using the old material and scrubbing the incoming air, or
using the newer material and accepting higher manufacturing costs. The one
company for which cost data are available has adopted-the first approach and
estimates that it expends approximately $30,000 per year to provide scrubbed
and filtered air. In addition, the company expends approximately $130, 000 per
year for high purity dry nitrogen, which it supplies to critical assembly and
fabrication areas. We estimate that 1/3 of the nitrogen cost, or about $43, 000,
actually goes for abating the effects of air pollution, while the remainder must
be attributed to the abatement of contamination from other sources. In 1969,
this company's sales were about $40 million. * Thus 3%000,.t.43'°-°x 100 = 0. 18%
r ' . 40 million
of the company's gross sales were expended for air pollution abatement.
Extending this figure to the remainder of the $750 million integrated circuit
market we arrive at $1.36 million as the cost of abating air pollution for this
segment of the market.
*Source: Electronic News
5.5
No maintenance costs are incurred, since the integrated circuits
do not lend themselves to this sort of treatment. Of the approximately 20%
which are rejected after manufacture, at least half fail because of some
form of contamination, but there the company doubted that any of this was
due to air pollution. We have attributed no residual cost to air pollution.
5. 3. 3 Television Picture Tubes
The effect of particulate matter upon the screens of television
picture tubes invariably is catastrophic and all manufacturers take special
precautions to eliminate all dust and dirt from the screening areas. The
source of the contamination, however, turns out to be mainly in-plant and
human-borne, so the precautions take the form of protective clothing, work
stations with a supply of clean air and other similar measures, none of which
can be assigned primarily to the abatement or control of air pollution.
In use, television picture tubes sometimes develop leakage paths
due to accumulations'of airborne oily dirt. This substance, which degrades
the operation of the tube, and the entire television set for that matter,
can be removed by cleaning. These maintenance costs are included in
paragraph 5. 3. 13.
There are, however, some tubes which fail in the field, and of
these, some failures almost certainly are attributable to air pollution.
Thus, there probably is a residual cost associated with this category.
but we have no data and no means of estimating it.
5. 3. 4 Capacitors
Without exception, capacitor manufacturers interviewed took
no special measures to control air pollution. Almost all factories are
air-conditioned and this seemed to provide sufficient filtering for all
purposes. We could find no preventive, maintenance or residual costs.
5. 3. 5 Resistors
The comments noted in the previous section can apply to resistor
manufacturers also. Most plants are air-conditioned but no other precautions
S-6
-------
..are found necessary. We could find no preventive, maintenance or
residual costs.
5. 3. 6 Power and Special Purpose Tubes
Included in this category are a wide variety of electron tubes, each
of which seems to have a special problem connected with it. Oxide-coated
cathodes, for example, are harmed by particulates and high humidity. Similarly,
particulates are extremely damaging to x-ray image intensifiers, Plumbicons
and others. Tubes with silver-plated parts are susceptible to tarnishing from
H,S. No company, however, could associate any of these damaging
contaminants with their present location and experience.
On the other hand, one company, formerly located in Newark, New
Jersey, found it necessary to install an'electrostatic precipitator to obtain
sufficiently high quality air for its manufacturing process. Since moving to
a more suburban location, the problem has disappeared, and we are led to
suspect that there may still be others who may be experiencing similar difficulties.
Since we did not discover any such companies, however, we conclude
that if they do exist, they must represent only a small fraction of manufacturers
in this category and so we attribute no preventive costs to air pollution.
Maintenance costs for cleaning are included in paragraph 5. 3. 13,
and we believe residual costs to be vanishingly small.
5. 3. 7 Connectors
The contaminants most damaging to connectors, and this category
Includes printed circuit boards, are gaseous sulfur compounds, either as SO, or
H,S which react with copper and silver to form non-conducting films. To prevent
these damaging effects, exposed metallic parts are gold-plated. One company
expends $800, 000 per year for gold and estimates that the connector industry
as a whole may expend $10 million per year. Of this, we estimate that 75%
may be attributed to air pollution, with the remainder being used to protect
against finger marks, salt spray and other corrosive agents not associated
with air pollution. The preventive cost for the industry is thus about
$7. 5 million.
In a study on the economic effects of air pollution on electrical
contacts, the Stanford Research Institute assigned a value of $20 million
per year for precious-metal plating. This includes not only connector contacts,
but also relay contacts and all others. We assume, therefore, that our
value is contained in the $20 million figure and have not duplicated the value
in our summary.
Connectors easily collect airborne dust which forms extraneous
leakage paths. This is removable, however, and the concomitant costs
are included in paragraph 5. 3. 13. We have identified no residual costs.
5. 3. 8 Transformers
Transformers are not at all affected by air pollution during
manufacture. Bushings and insulators can develop leakage paths due to the
deposition of particulates on them,but this is easily removed by cleaning.
Maintenance costs are included in paragraph 5. 3. 13. No residual costs
have been identified.
5. 3. 9 Relays
Experience in the manufacture of relays varies from one company
which stated that the only problems were internal, and indeed, did not
even provide air conditioning, to another which stated that all relay
manufacturers face a constant battle with the effects of contamination on *
contacts. The latter manufacturer and others gold-plate relay contacts. The
cost of this operation is included in the SRI study' ' referenced in paragraph
5. 3. 7 and is not duplicated herein. Maintenance costs are probably low,
but relay contacts can become corroded and non-conductive if exposed to
contaminants like SO, and H,S. Replacement of defective relays should be
included as a residual cost but we have no data to ascertain its magnitude.
5. 3. 10 Receiving Tubes
In general, air pollution does not add significant costs to the
manufacture of receiving tubes. One manufacturer in the San Francisco
Bay area believes that H,S from the Bay may be causing some sulfiding of
5-7
5-8
-------
silver contacts. Another in Connecticut has also experienced some isolated
problems with the formation of silver sulfide. The number of tubes involved,
however, and the costs, while not known, are probably insignificant. Main-
tenance and residual costs also should be vanishingly small for receiving
tubes.
5.3.11 Crystals
The crystal manufacturers interviewed adopt no special precautions
to abate or control air pollution. All costs related to air pollution are
insignificant.
5. 3. 12 User Costs
Quite apart from air pollution costs incurred by the manufacturers
of electronic components are the costs which the user must absorb. Two
examples in this category are traffic signal controllers and telephone switching
centers. Both types of equipment are degraded by particulate matter. In the
fir.* --»se, no preventive measures are possible since the equipment is located
out of doors, hence, cleaning is resorted to as a remedy. In the second case,
high efficiency air filters have been installed.
The Washington, D. C. highway department has estimated that the
1085 traffic light control systems in the District of Columbia each require
2 hours of maintenance per year. At least 50% of this maintenance is for
cleaning of accumulated particulate matter which tends to interfere with
proper operation of lubricated electro-mechanical components such as power
contactors and timing motors. At a cost of $5. 00 per hour for maintenance,
this represents an estimated annual cost of $5, 400 to the D. C. Highway Department
for cleaning traffic light controllers, a unit cost of about $5.00 per controller.
The annual mean suspended particulate level in Washington, D. C. ,
77 jigm/m , is less than the national average of 10Z n gm/m for all U. S.
urban areas. We assume, therefore, that other areas, on the average, will
find it necessary to maintain their controllers with at least the same frequency
as Washington. A simple extrapolation to the 244, 500 traffic signals in the
U. S., most of which are in urban areas, yields a total annual maintenance cost
of about $1. 2 million.
The New York Bell System has been forced to install high-efficiency
air filters in all of its New York City switching centers, and these, of course,
must be maintained. New York Bell estimates that switching centers in the
borough of Manhattan alone use 15, 000 tons of air-conditioning, and that initial
filter cost represents about 1% of the cost of this air conditioning. Cost of
commercial air conditioning equipment is estimated at approximately $500 per
ton. This results in an initial cost for air filtering of $5. 00 per ton or
$75,000 for Manhattan, an annual cost of $7, 500, since the amortization period
is 10 years. New York Bell estimates that filter maintenance in a building
requiring 2,000 tons of air conditioning takes two men 15 days every 6 months,
or 60 man days per year. Assuming $5. 00 per hour, the filter maintenance
labor per ton of air conditioning is
60 X 8 X 5
2000
= $1.20 ton
or $18,000 per year in Manhattan. Filter replacement materials cost
$10. 00 for 4 square feet of filter element. Total filter replacement costs
can be calculated by assuming 400 cubic feet per minute of air per ton of
air conditioning, and an output velocity of 1, 000 linear feet per minute.
This leads to a duct cross section of 0. 4 square feet per ton of air
conditioning. At $2. 50 per square foot, the cost for filter materials is
$1. 00 per ton. of air conditioning. This figure leads to a total filter replace-
ment cost of $30, 000 per year in Manhattan.
In summary, the costs to the New York Bell System to prevent
the damaging effects of air pollution to approximately 1.47 million telephones
in Manhattan are estimated to be:
Equipment Amortization
Yearly Maintenance Labor Cost
Yearly Filter Material Cost
Annual Cost
$ 7,500
18,000
30. OOP
$55, 500, or $0. 0378/telephone
Manhattan is like no other city, of course, but it seems safe to assume that
other large cities (population in excess of 100, 000) and with particulate levels
in excess of 100 figm/m3 will eventually find it necessary to install similar
5-9
5-10
-------
equipment. There are 50 cities in the United States fulfilling both requirements,
and they include almost 28 million telephones. Extrapolating the Manhattan
data to these areas yields a total annual cost to the country of slightly
over $1 million. These are classed as preventive costs, even though part is
for maintenance of the filters.
5. 3. 13 Maintenance Costs
Maintenance costs as we use them here are the costs of cleaning,
repairing, replacing, and in any way restoring a piece of otherwise defective
electronic equipment to enable it to perform its initial function.
When maintenance costs represent a significant fraction of the
economic effect of air pollution on a category of components, we have not
listed them for each category, but instead, have lumped them all together
as one item. We do this because maintenance, as we have defined it, is
performed mostly on equipment rather than components and costs, therefore,
are not separable by component category.
Maintenance costs result from a wide variety of operations, but for
simplicity, we will divide them into two major groups: cleaning and everything
else. Cleaning was selected separately because electronic equipment operators
are becoming increasingly aware of the damaging effects of dirt and may have
adopted formal cleaning programs. One company, the M. P. Odell Company
of Cleveland, Ohio, has discovered that electronic instruments frequently can
be reconditioned without replacing parts simply by cleaning the entire instrument
to remove conduction paths created by dust, dirt, tobacco smoke or other
air-borne contaminants. The company has developed a solvent which is used
with a cleaning unit and a dryer to perform the cleaning operation. The units
and the solvent are sold to other companies; M. P. Odell does not do the cleaning
itself.
The time required on the average to clean an instrument such as a
vacuum tube voltmeter is about one-half hour. This estimate was supplied by
Odell and corroborated by a user of the system, and it includes time pocessary
to remove the instrument from its case and to replace it after drying. The
cost for the solvent is about $0. SO per cleaning. Odell estimates that his
systems are currently used for approximately one million instrument cleanings
per year. If labor costs are $5 per hour, it may be seen that the present yearly
cost of cleaning instruments with the Odell system is about $3 million. Of this
we estimate that two-thirds, or $2 million, is attributable to air pollution.
Extending this analysis to the instruments that are not cleaned
leads to an even greater cost. Sales of testing and measuring equipment in
the United States has averaged about $500 million per year for the last 5 years.
Assuming an average life of 5 years and an average acquisition cost of $1,000
per instrument, we estimate that there are about 2. 5 million instruments of
this type in use at the present time. Maintenance costs, in our experience,
average about 2% of acquisition costs, or about $50 million per year for these
2. 5 million instruments. .
The most common maintenance procedures involve nothing more
than changing tubes, a function not at all related to air pollution. The most
expensive maintenance, however, probably is related to air pollution because
it must include relating malfunctions to leakage paths, discovering the leakage
paths and then removing them. Odell1 s experience has convinced him that
accumulation of dirt and grime are."major" causes of instrument instability or
malfunction. The fraction of maintenance costs attributable to air pollution,
therefore, is certainly greater than 10% and probably less than 50%. If we
conservatively estimate 20%, we arrive at an annual cost of $10 million for
maintenance. Adding this to the preceding $2 million yields an annual total
cost of $12 million for the maintenance of all electronic instruments.
5. 3. 14 Summary of Costs
The economic impact of air pollution on electronic components as
determined by the survey of over 100 manufacturers and users of these com-
ponents is significant for its low magnitude. All of the data discussed in the
previous sections are summarized in Table 5-1, which shows that the total
annual cost is about $15, 5 million.
There are two major uncertainties in this figure. First, some of
the data simply are not available. In this category, the most obvious
5-11
5-1Z
-------
Table 5-1. Summary of Costs, Effects of Air Pollution on Electronic
Components
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Semiconductors
1C
TV Picture Tubes
Capacitors
Resistors
Power 8t Special
Purpose Tubes
Connectors
Transformers
Relays
Receiving Tubes
Crystals
Maintenance
Users
Preventive
Costs
($000)
0
1,360
0
0
0
0
(Included in
SRI Study)
0
(Included in
SRI Study)
0
0
1,000
Maintenance
Costs
($000)
0
0
(Included in
Line 12)
0
0
(Included in
Line 12)
"
"
0
0
0
12,000
1,200
Residual
Costs
($000)
0
0
Not available
0
0
Not available
0
0
Not available
0
0
2,360
13,200
5-13
pieces of missing data are those which would tell us the cost of components
that fail or otherwise become inoperative due to the damaging effects of air
pollution. The three components most likely to be vulnerable to this fate
would seem to be TV picture tubes, cathode ray tubes, (included with power
and special purpose tubes) and relays. Other components are protected by
encapsulation, potting, protective coatings and the like, to the extent that
they would suffer a similar fate only in a small degree.
Second, while the procedure for selecting companies to be interviewed
assured that all major manufacturers were included, still limitations of time and
money meant that some related users were left out. To estimate the magnitude
of these omissions, the New York City Police Department, a major user of
communications equipment, was contacted. No new significant data were
developed, and we conclude, therefore, that other electronics equipment
users likewise would account for tangible but diminishingly small
increments of cost.
The combined effect of these two factors results in an uncertainty
we estimate to be no less than +20% and certainly no greater than +50%. A
value of +30% seems probable.
5-14
-------
6. RECOMMENDATIONS
One area of potentially great economic impact indicated by the
present study is the cost of long-term effects of air pollution on semiconductors
and integrated circuits. At present, costly steps are being undertaken to
protect these components from effects of contamination during both manu-
facturing and use. These include the use of filtration of incoming air at
manufacturing facilities, the use of clean rooms for fabrication and assembly,
and the use of expensive materials for encapsulation of the components. In
part, this is done to protect against contamination induced by handling and in
part to protect against airborne contaminants. Little has been reported as
to the potential long-term effects of the airborne contaminants on the function
or reliability of the components. The protective measures are being taken
on the assumption ih-t there may be such long-term effects. This is sup-
ported by discussion with ITT Technical Staff personnel. In the event that
it could be demonstrated that the long-term (10 to 20 years) effects are not
significant, less costly materials could probably be employed with a conse-
quent reduction in the cost of manufacture. To develop data to determine if
the long-term effects do require protective measures, a program of accel-
erated life testing would be required. It is recommended that such a program
be considered. In general, such a program would include controlled experi-
ments to determine the effects of major air pollutants on the operation and
life of a variety of semiconductor and integrated circuits using no protective
measures except for cleaning and also using the present and planned encap-
sulation techniques. The output of the study would be a determination of the
contribution of these measures to the reliability of the components.
Although a survey of major manufacturers (paragraph 4. 1.2)
exposed particulate matter as the major source of damage to electronic
components, the statistical survey (subsection 4.3) illustrated a deplorable
deficiency in our knowledge of the quantitative relationship between
particulate concentration and equipment failure. Capabilities exist,
however, which would enable measurement to be made from which data
necessary for significant statistical analyses could be made. We refer to
mobile monitoring laboratories equipped to measure a wide variety of
pollutants and which can be moved easily to the exact site where it is
desired to make measurements. The impact of such a study would be a
more quantitative understanding of the economics of particulate control.
6-1
6-2
-------
7. ANNOTATED BIBLIOGRAPHY
Influence of Atmospheric Contaminants on Corrosion--Literature Report,
"Muffley, Harry C. , Department of the Army Project Number
1-A-0-24401-A-109, AMC Code No. 5026.11.803, Report No. 63-2041,
June 13, 1963.
Corrosion preventive coatings are currently designed to
prevent moisture from coming in contact with the metal surfaces.
Corrosion due to moisture alone is very slow; however, the inter-
action of moisture and atmospheric contaminants accelerates the rate
of corrosion. The fact that these two, moisture and atmospheric con-
taminants, vary independently, makes it difficult to predict the corrosion
behavior of a given area.
Literature in this field of endeavor was reviewed with the
following questions in mind:
1. What contaminants contribute to atmospheric
corrosion?
2. What is the mechanism of atmospheric corrosion?
3. What techniques are available for determining
atmospheric contamination?
"Mechanisms by Which Ferrous Metals Corrode in the Atmosphere, "
Larrabee, C. P. , Corrosion. Volume 15, October 1959, p 526t-529t.
The electrochemical theory of corrosion, by which the
corrosion of steels in solutions is readily explained, also is used to
explain the start of atmospheric corrosion on ferrous surfaces. Effect
of contaminants both in the atmosphere and on the steel surface is
described. Experiments leading to this conclusion are illustrated.
The part played by various alloying elements in retarding
atmospheric corrosion, once a rust film is formed, is discussed. The
relative effects of heavy and light deposits of sea salt on ferrous specimens
are shown.
7-1
Corrosion Prevention/Deterioration Control in Electronic Components and
Assemblies, Sparling, R. H. , U. S. Army Missile Command, Redstone
Arsenal, Alabama, December 1965 (143 pages).
This engineering report provides information to designers and
manufacturers concerning the effect of corrosion, the subsequent degrada-
tion of electronic products and the techniques for the prevention of corrosion.
It covers both metallic and non-metallic materials and includes discussions
of metal couples which will result in galvanic corrosion. No separate
deliberate categorization is made by the author for corrosion due to natural
elements and corrosion due to atmospheric pollution, but most of the
information is clearly presented so that the causes are readily separable.
The deterioration of both metals and non-metals is discussed.
"Effects of Air Pollution on the Atmospheric Corrosion Behavior of Some
Metals and Alloys, " Tice, E. A. , Journal of the Air Pollution Control'
Association, Volume 12, Number 12, December 196Z. '
When exposed to atmospheric pollutants, metals corrode in a
manner which depends upon the constituents of the pollutants. Designers,
therefore, are faced with the problem of choosing a particular alloy or
metal which will show the best resistivity to pollutants known to exist at
the location of a proposed structure.
Data are presented showing the relative corrosion rates for
various metals for areas of Europe and the United States.
It is concluded that SC"2 is the most critical air pollutant from
the standpoint of corrosion and that it is aggravated by soot.
Inquiry Into the Economic Effects of Air Pollution on Electrical Contacts,
Robbins, Robert C. , Stanford Research Institute, April 1970 (58 pages).
Prevention of contact failure in low voltage contacts is the
largest problem area in which pollution effects on electrical contacts are
found. Protection is afforded by the use of air conditioning and purification
equipment or by precious metal plating. Air conditioning costs attributable
to air pollution are estimated to be $25 million per year, plating costs are
$20 million, other costs (loss, research, high voltage flashover) are
$19 million, for a total of $64 million.
7-2
-------
The Corrosion and Oxidation of Metals, Evans, U. R. Arnold, London,
1960, Chapter XIII, "Atmospheric Corrosion. "
Atmospheric corrosion can be divided into three types: dry,
damp, and wet. The first category includes the development of oxide films
and the tarnishing of certain metals when exposed to sulfur compounds.
Damp corrosion becomes serious when humidity exceeds a critical value--
about 70%. The fogging of nickel, bronze disease, production of rust-spots, and
filiform attacks are examples. The effect of air pollution on outdoor
corrosion is discussed.
"The Effect of Air Pollution on Electric Contacts, " Antler, Morton,
Gilbert, Jay, Journal of the Air Pollution Control Association, Volume 13,
No. 9, September 1963. pp 4o5-415:!
Air pollution is a major course of the degradation of electric
contacts. Laboratory accelerated aging tests involving polluted environ-
ments are developed, based on field experience. Cause-effect relationships
are established which aid in materials selection and the development of
connector designs.
"Beware of Electronic Dirt, " Manko, H. H. , Electronic Design 24,
November 22, 1969.
Erratic behavior and certain frequently encountered-equipment
malfunctions can be drastically reduced through a thorough understanding
of the role of electronic dirt. Of the two types of dirt, particulate matter
and foreign films, the latter is the most harmful. Final cleaning is not
always feasible, however, and may be costly. Solvents are discussed and
three case histories illustrate the dramatic results achieved by cleaning.
"Design of Telephone Exchanges for Corrosive Atmospheres, " Elliott, J.F.F..
Tin and Its Uses. Vol. 81, 1969, pp 8-12.
Tests were conducted with copper and its alloys to determine
the effects of hydrogen sulfide corrosion. The tests were prompted by the
deterioration of telephone equipment in an area of New Zealand remarkable
for great geothermal activity. The concentrations of hydrogen sulfide
ranged from 0.004 to 0. 18ppm.
Sulfides of copper and silver are porous and do not provide a
barrier against further attack. Consequently, the sulfide layer gradually
increases in thickness, and having low adhesion and weak mechanical
strength, eventually flakes off. In this way progressive corrosion of the
metal takes place and prolonged exposure causes serious loss of the metal
and possible failure of the part.
Possible solutions involving air purifying equipment, application
of special finishes, and use of alternate materials are discussed.
"Corrosion of'Metals by Aqueous Solutions of the Atmospheric Pollutant
Sulfurous Acid, " McLeod, W. , and Rogers, R. R. , Electro-Chemical
Technology, Volume 6, No. 7-8, July/August 1968, pp 231-235. "
It has been found here that the corrosion rate of a metal in an
acid such as H^SO^, HNOs, H£SO4, or HC1, having a normality between
N/l and N/10,000, is related to the concentration of the acid in accordance
with the equation Corrosion rate = a x (Acid Normality) where a and b are
constant for each combination of acid and metal and where the termperature
is 25°C. Having determined the values of a and b for a considerable number
of these acid-metal combinations, it was possible 1—to compare the
corrosion rates of the various metals in sulfurous acid with those of the same
metals in the well-known nitric, sulfuric, and hydrochloric acids, and 2 —
to determine the corrosion rates of the metals in sulfurous acid of
different normalities.
"The Deterioration of Materials in Polluted Atmospheres, " Yocum, John E. ,
Journal of the Air Pollution Control Association, Vol. 8, No. 3, November 1958,
pp 203-208.
The deterioration of materials resulting from atmospheric pollution
is a serious economic loss. It is related to a number of complex atmospheric
and surface reactions. Further study is required of the physical factors and
the chemistry of the specific agents which lead to the ultimate damage to
materials.
"Creeping Silver Sulfide, " Egan, T. F. and Mendizza, A. , Journal of the
Electrochemical Society, Vol. 107, No. 4, April 1960.
When a gold-plated silver plate was exposed to sulfur vapor,
silver sulfide formed at pores in the gold plating, but in addition, spread
radially outward over the gold surface.
The same experiment using rhodium and silver revealed needles
of silver sulfide at the bimetallic juncture, but no creepage of silver sulfide.
7-3
7-4
-------
"Mechanism by which Non-Ferrous Metals Corrode in the Atmosphere, "
Aziz, P. M. and Godard, H. P., Corrosion. Vol. 15, October 1959,
pp 529t - 533.U
Literature dealing with the mechanism by which non-ferrous
metals corrode in the atmosphere is reviewed, particular emphasis being
given to the important series of papers published by W. H. J. Vernon and
his co-workers in the period extending from 1923 to the present. The
influence of the common atmospheric constituents is discussed and related
to the chemical composition and properties of the corrosion-produced film
formed. The-influence of these on the atmospheric corrosion rate is also
discussed in relation to the nature of the corrosive atmosphere. The atmos-
pheric corrosion of nickel, copper, zinc, aluminum and magnesium is
discussed.
The Tarnishing Behavior of Silver-Palladium Contact Surfaces, and its
Relationship to Accelerated Testing, Crossland. W. A., and Knight, E. ,
Standard Telecommunications Laboratories, Harlow, Essex, England.
Silver and silver -palladium alloys were exposed to various
sulfur-containing atmospheres and the following results were noted. The
rate of sulfidation of silver increases with illumination. Quantitative rates
of sulfidation are given for silver and silver-palladium exposed to various
concentrations of HzS and to flowers of sulfur. A given thickness of Ag2S
gives rise to a larger increase in contact resistance on AgPd than on Ag.
"Combating the Effects of Smog on Wire-Spring Relays, " Hermance, H.' W. ,
Bell Laboratories Record, Volume 44, February 1966, pp 48-52.
-where tire-dust deposited on the wire. No cracking was observed on wires
free of dust deposits nor was there any cracking when the humidity was kept
below 50%. Further investigation correlated the cracking with moisture
absorbing dusts containing high nitrate concentrations.
7-5
"The Nature of Corrosion of Zinc by Sulfurous Acid at Ordinary Temperatures, "
McLeod, W. , Rogers, H. R. , Corrosion-NACE, Volume 25, February 1969.
pp 74-76.
Corrosion rate data are presented for zinc in (1) a combination of
sulfur dioxide, water vapor and air, and (2) aqueous solutions of sulfurous
acid in the absence of air, at ordinary temperatures. It is shown that the
rate depends to a considerable extent on the form in which the sulfurous
acid occurs. Information as to the nature of the corrosion products is
presented.
"Sulfur Dioxide Porosity Tests for Coatings of Gold and the Platinum Metals
on Substrates of Copper and its Alloys, Nickel, Silver, " Clarke, M. , and
Leeds, J. M. , Transactions of the Institute of Metal Finishing, Volume 46,
1968.
This report describes tests devised to detect porosities in platings
of gold, platinum, alloys of both, nickel and silver. The test, which involves
a 24-hour exposure to sulfur dioxide, is intended to replace currently'used
electrographic tests and 48-hour sulfur dioxide-hydrogen sulfide tests.
Results show that such a test could complement the latter two tests and, in
certain cases,replace either.
The significance of these tests is the implication that SO^ permeates
noble metal coatings and attacks the substrate.
"Evaporated Metal Films as Indicators of Atmospheric Pollution" (Part I),
Lodge, J. P. and Havlik, B. R., International Journal of Air and Water
Pollution, Vol. 3, No. 4, pp 249-252.
Glass slides covered with films of aluminum, iron, lead, copper,
and silver were prepared and exposed to the atmosphere in the Cincinnati
area. Corrosion products were identified by X-ray and microchemical means.
Results showed that aluminum was the most dependable indicator
of general atmospheric corrosiveness, although silver appeared specifically
sensitive to a small group of pollutants such as halogens, ozone, and sulfur
compounds. Lead corroded so rapidly that it was difficult to interpret. Iron
was primarily responsive to humidity.
7-6
-------
"Evaporated Metal Films as Indicators of Atmospheric Pollution" (Part II),
Lodge, .P,-, and Frank, E. R. , International Journal of Air and Water
Pollution, Vol. 6, pp 215-221.
An attempt was made to utilize the electrical resistance of the
metallic films as an index of the rate of attack of the corroding agents. A
simple, inexpensive technique was evolved for the continuous recording of
the rate of corrosion by pollutants in the atmosphere. The recording system
and the resulting graphs are presented and analyzed. It was concluded that
there is a direct relationship between gross atmospheric features associated
with the accumulation of substances in the air that are corrosive to metals
and the rate of resistance change in metallic films.
Systems Analysis of the Effects of Air Pollution on Materials, Salmon, R. L. ,
MRl Report, Contract CPA-2Z-69-113, MRI Project No. 3323-D, 15 June 1969-
14 January 1970.
While it has been known for some time that materials are damaged
by the action-of p'r pollutants, information on the exact mechanisms and
rates of attack is available only for a small number of specific material-
pollutant combinations, notably zinc-SC>2, cotton-SC>2, and rubber-O^.
In spite of the paucity of information, it is possible to estimate
the economic losses caused by air pollution. In order to maintain all
materials as clean as they-would be in the absence of air pollution would
require the expenditure of an estimated $100 billion per year. The direct
economic losses caused by deterioration are estimated to be as high as
$9. 5 billion per year.
Soiling of paint, zinc, and Hal glass (all used primarily as
outdoor-exposed surfaces of buildings) is the cause of the major portion of
economic loss.
In order to make further progress-in the assessment of air
pollution damage to materials, research is needed to obtain precise data
on both physical-chemical and economic aspects of significant material-
pollutant interactions. There is also a need for evaluation of the procedures
used to .obtain such data.
"The Role of Copper (II) in the Oxidation of Ferrous Hydroxide Colloid with
Special Reference to the Corrosion of Iron in an SO2-Containing Environment, "
Inouye, Katsuya, Journal of Colloid and Interface Science, Volume 27, No. 2,
June 1968, pp 171-179 (Extract).
The presence of sulfur dioxide influences the rate of corrosion of
iron through a change of SOz to SC>4--ions, further accelerating the oxidation
7-7
reaction. Evidence of the existence of FeSO4 in the corrosion product,
usually in the boundary between the metal surface and the oxide layer,
is reported. In the presence of copper, however, the oxidation is inhibited.
"Dirt and Calibration, " Odell, M. P., Measurements and Data,
January/February, 1969. " ~
Malfunctions of electronic equipment are frequently caused by
contamination such as dirt and oily deposits that create electrical leakage
paths around input terminals and components. These deposits can be
successfully removed by the use of proper solvents and if repair time and
costs are to be kept low, cleansing ought to precede the replacement of
parts. Examples are given.
"Freeways vs. Contamination Control in California, " Contamination
Control. April 1970. —
This article is of interest mainly because it involves the first
legal case in history wherein a suit was filed against a state to obtain a.
settlement to offset the cost of combating contamination due to the intended
construction and operation of a freeway passing near the plant. The company
involved is ITT Jennings and the settlement request involved a figure of
approximately $1,000,000. The article includes a chart which shows
environmental conditions before the construction of the freeway, during the
construction of the freeway, after the completion of the freeway, and the
normal future which takes into account a normal community growth over a
five-year period, from 1971 to 1976'. The case against the California
State Highway Department was denied a hearing in February 1969, but the
case has been appealed to California higher courts.
"The Electronics Industry, " Dunlap, Lloyd \V. , Chemical and Engineering
News, November 30, 1970.
A number of chemical problems in the manufacture of electronic
components, several of which either originate directly, or are indirectly
caused by pollution of the atmosphere are cited.
Semiconductor manufacture requires process water with
resistivities of 15 megohms. To obtain water with such high resistivity,
distilled water, which normally has resistivities of about 1 ohm. must be
deionized a number of times.
7-8
-------
Similarly, the manufacture of MOS'integrated circuit cards
requires procedures that avoid any contaminatro«-of the surface layer of
silicon dioxide by sodium, potassium, lithium, calcium, aluminum, mag-
nesium, barium, and strontium ions. Such purity control inevitably involves
extraordinary measures (which are described) to prevent atmospheric
pollution of the materials through all stages of manufacture.
"Environment Simulation for Studying the Effects of Air Pollutants on
Computers, " Steading, T, Walter, APCA Journal, Vol. 15, No. 3,
March 1965.
This paper describes a study directed towards the laboratory
simulation of parameters of the office environment related to airborne
gaseous and particulate pollutants and initial operating experience of a
particulate test chamber, instrumentation, and control techniques. Also,
discussed is the design of a simple gaseous test chamber being constructed.
Gas dilution procedures, control and general operating approach are described
"Measurement and Simulation of the Computer Environment, " Steading, T. W.
and Pauls en, J. N., Proc. of the Institute of Environmental Science, Annual
Meeting, 1967.
This paper describes three field investigations made by IBM
which resulted in better definitions of earth environmental parameters of
man-operated commercial and industrial computers. The environments
of interest were floor vibration, airborne dust, and conducted electrical
switching transient noise. The techniques used in defining these environ-
ments as well as experimental methods used in simulating them are
described.
General summaries derived from the collected data are presented.
"The Effects of Air Pollution of Electrical Contact Materials: A Field Study, "
Chiarenzelli, Robert V. , and Joba, Edwin L. , Journal of the Air Pollution
Control Association, Vol. 16, No. 3, March 19~6TT
A long-term field and laboratory program designed to determine
and understand the effects of air pollutants on the performance of electric
contact materials has reached the one-year mark. An extensive variety of
metals has been exposed at six field environments, for periods up to one year
(August 1963 to August 1964). These environments were selected to provide
7-9
-------
a wide range of air pollutants in typical data processing or process.control
situations. The program undertakes to determine material degradation as
a function of time and environment. The important air pollutants at the
field sites are measured regularly, and materials are returned periodically
to the laboratory for evaluation using techniques developed specifically for
this program. The results of the program to date are presented, and pre-
liminary, correlations are drawn.
"The Relation of Thin Films to Corrosion, " Rhodin, R. H. , Corrosion,
Vol. 12, No. 9, 465t-475t (1956) September.
Recent studies of thin oxide films isolated from stainless steels
and other ferrous alloys have provided new information on surface properties.
Correlations between the unique compositional and structural properties
of thin films and alloy corrosion behavior in various corrosive environments
have been established.
Evaluation of the resistance to pitting corrosion of silicon-modified •
Type 316L stainless steel is described as a typical example of an application
of a film study to corrosion research. In this case, improved corrosion
behavior was correlated with a mutual film-enrichment of silicon and
molybdenum and a film-depletion of iron. Experimental alloys of improved
purity were prepared-and evaluated. Effects of alloy purity, carbon content,
and alloying additions were found to influence strongly surface behavior and
corrosion resistance.
The mechanism of resistance to pitting corrosion and the protective
nature of the associated oxide films are interpreted in terms of the chemistry
of hydrated heteropoly acids containing oxides of silicon and molybdenum.
The following group of nine papers, all delivered at the IEEE
Winter Power Meeting in New York, January 31 - February 5, 1971, are on
the general topic of high voltage insulator flashover. a subject discussed in
the SRI Report (Ref. 13 ), but published too late to be included in that report.
They are included in this bibliography to augment the SRI Report and because
they are part of the literature on air pollution damage effects.
"A Survey of Insulation Contamination in the United States and Canada, "
Parti, Transactions of the IEEE, Paper No. 71 TP 133-PWR. February 19, 1971.
Because of the growing problem caused by insulator contamination
due to both industrial pollution and natural deposits, a Working Group of the
Lightning and Insulator Subcommittee was formed to study the problem. The
first objective was to determine the extent of the problem in the U. S. and
7-10
-------
Canada, and to determine the various kinds of contaminants and the most
severe weather conditions that cause transmission line outages. A survey
was conducted by sending questionnaires to 90 utilities. A total of 309 case
histories were reported. This report summarizes the answers to questions
regarding contaminants and insulator types. It also sheds some light on the
preventive measures most frequently undertaken by the utilities and the
degree of success of each action.
of this reduction, lightning impulse flashover characferistics of disc
insulators.are to have great significance for insulation design of UHV lines.
This reduction in flashover voltage of insulator strings varied the flashover
paths of insulator assemblies equipped with arcing horns; the lightning
impulse flashovers tended to take the paths creeping along the insulators under
polluted conditions, in spite of the fact that the flashovers take the paths
between the arcing horns and stay away from the insulators under clean
conditions. With deep-rib type disc insulator strings, the reduction due
to pollution was almost nil for positive polarity.
"Switching Surge Performance of Contaminated Insulators, " Turner, F. J. ,
Macchiaroli. Transactions IEEE, Paper No. 71 TP 141-PWR, December 1, 1970.
The "wet contaminant" test method, originally developed for use
with alternating voltage, is extended for use with switching surges. In
order to evaluate insulators under conditions where dry bands would form,
it was found necessary to first energize the contaminated insulators with
alternating voltage before applying the surges. Included in the investigation
are the effect of front time and tail time on the critical flashover voltages,
and a comparison of the switching surge performance of five insulator types -
under various levels of contamination.
"A Study of the Performance of Contaminated Insulators Under Various Test
Conditions, " Kawai, M. , Transactions IEEE, Paper No. 71 TP 132-PWR.
December 15, 1970.
Two contamination test methods have been examined for the
difference in assumptions used as the basis of each method. Dynamic
surface impedance of insulators in a slow-wetting condition is remarkably
influenced by the voltage across the insulator unit, which causes nonlinearity
of insulation strength with insulator length. On the other hand, the surface
condition can be represented as resistance in wet contaminant tests. This
difference in surface condition naturally leads to disagreement in the results.
It has been concluded to be of limited utility to discuss the order of merit
for several insulators with different test methods, since any contamination
method cannot simulate all aspects of this problem.
"The Effects of Pollution and Surface Discharges on the Impulse Strength
of Line Insulation, " Lushnicoff, N. L. and Parnell, T. M. , Transactions
IEEE, Paper No. 71 TP 143-PWR, December 1, 1970.
The effect of surface discharges on the ability of polluted
insulators to withstand lightning and switching surges has long been a
matter of conjecture. This paper describes combined power frequency
and impulse voltage tests on 6.6 kV and 132 kV insulator strings. The
tests have shown that an appreciable lowering of both lightning and switch-
ing surge strength may occur in the presence of pollution layers. The
worst lowering of impulse strength (about 50%) has been found in cases
where drying of the pollution has reduced the leakage current almost to
zero by the production of extensive dry bands on the insulator surface.
"Effects of Dew and Fog on Insulator Breakdown Strength, " Berger, K.
and Chowdhuri, P., Conference Paper IEEE, Paper No. 71 CP 140-PWR,
December 7, 1970.
This paper describes the sparkover characteristics of insulators
subjected during 20-80 minutes to a 50-Hz voltage under simulated condi-
tions of dew and fog in the laboratory. Results obtained with different
materials are discussed. A test procedure is proposed for the purpose
of reducing the duration of tests. It consists of the superposition of a surge
voltage of slow rising front on the rated 50-Hz voltage applied to the
insulator.
"Lightning Impulse Flashover Characteristics of Long Disc Insulator Strings
Under Polluted Conditions, " Okada, T. , Koga, S. , and Kimoto, I.,
Conference Paper IEEE, Paper No. 71 CP 144-PWR, December 1, 1970.
Lightning impulse flashover voltage of long disc insulator strings
showed great reduction under polluted conditions: maximum 25% reduction
for 25-unit string and maximum 35% reduction for 45-onit string. Because
7-11
7-12
-------
"A Possible Solution to the Insulator Contamination Problem, " Moran, J. H.
and Powell, D. G. , Conference Paper IEEE, Paper No. 71 CO 41-PWR,
January 31 - February 5, 1971.
A new semi-conducting glaze is introduced as an excellent
solution to the growing problems posed by contamination of electrical
insulation. The method of use is given, as are the results of both
laboratory and field tests on station posts and suspension insulators.
Discussions of the electrical characteristics and operation of the glaze
and the economics of its use are included.
"Analysis of Dielectric Recovery with Reference to Dry-Zone Arcs in
Polluted Insulators, " Rizk, F.A. M. , Conference Paper IEEE,
Paper No. 71 CP 134-PWR, January 31 - February 5, 1971.
the conditions of reignition and recovery of dry
c u,.,^...
complicated nature of the problem the agreement between theory and
previous field and laboratory experiences is satisfactory.
"A Criterion for A. C. Flashover of Polluted Insulators, " Rizk, F.A.M.,
Conference Paper IEEE, Paper No. 71 CP 135-PWR, December 29, 1970.
On the basis of the analysis of dielectric recovery of dry-zone
arcs presented in a companion paper, the critical a.c. voltage per unit
length of creepage path necessary to maintain the arc on a polluted insulator
surface is calculated. Calculations cover the range of pollution resistance
per unit creepage length of 1 - 6 kohm/cm. From this a new criterion for
a.c. pollution flashover of insulators with uniform pollution resistance along
the leakage path is deduced. It is shown that the present theory agrees
better with the experimental findings than does the previous d. c. theory.
Moreover the present theory agrees satisfactorily with the predictions of
dimensional analysis.
8. GENERAL BIBLIOGRAPHY
1. Shipments of Selected Electronic Components for Calendar Years
1960 - 1969, Publication BD 70-35, U. 5. Department of Commerce,
August 1970.
2. U. S. Industrial Outlook, 1970, U. S. Department of Commerce,
December 1969.
3.
"Product Preference Poll, A Study of Purchasing Attitudes Among
Electronics Subscribers - 1969," McGraw-Hill, New York, 1969.
4. Influence of Atmospheric Contaminants on Corrosion--Literature
Report, Muffley, Harry C. , Department of the Army Project Number
1-A-0-24401-A-109, AMC Code No. 5026.11.803, R'eport No. 63-2041,
June 13, 1963.
5. "Mechanisms by which Ferrous Metals Corrode in the Atmosphere , "
Larrabee, C. P. , Corrosion, Volume 15-, October 1959.
6. Corrosion Prevention/Deterioration Control in Electronic Components
and Assemblies, Sparling, R. H, , U. S. Army Missile Command,
Redstone Arsenal, Alabama, December 1965.
7. Air Pollution, Stern, A. C. , ed. , 3 volumes. .Academic Press,
New York - London, 2nd ed. , 1968.
8. "Effects of Air Pollution on the Atmospheric Corrosion Behavior of
Some Metals and Alloys, " Tice, E. A. , Journal of the Air Pollution
Control Association, Volume 12, Number 12, December 19o2.
9. Methods of Correlation and Regression Analysis, Ezekiel, M. , and
Fox, K. A. , Wiley and Sons, New York, 1967.
10. Models in Regression. Sprent, P., Methuen & Co. Ltd., London, 1969.
11. Air Quality Data from National Air Sampling Networks and Contributing
U. S. Department of Health, Education and
rvice, Consumer Protection and Environment
Slate and Local Networks, .
Welfare, Public Health Service, onsumer rotecon
Health Service, 1966 Edition, Durham, N. C. , 1968.
12. Economic Costs of Air Pollution. Ridker, R. G. , Frederick A. PraegtT,
13. Inquiry Into the Economic Effects of Air Pollution on Electrical Contacts,
Robbins, Robert C. , Stanford Research Institute, Contract PH-22-b8-35,
SRI Project PSU-7345, April 1970.
7-13
8-1
-------
14. The Corrosion and Oxidation of Metals, Evans, U. R., Arnold, London,
I960.
15. "The Effect of Air Pollution on Electric Contacts, " Antler. Morton,
Gilbert, Jay, Journal of the Air Pollution Control Association,
Volume 13, No. 9, September 1963, pp 405-415.
16. "Beware of Electronic Dirt, " Manko, H. H. , Electronic Design 24,
November 22, 1969.
17. "The Deterioration of Materials in Polluted Atmospheres, " Yocum,
John E. , Journal of the Air Pollution Control Association, Vol. 8,
No. 3, November 1958, pp 203-208.
18. "Design of Telephone Exchanges for Corrosive Atmospheres, "
Elliott, J. F. E. , Tin and Its Uses. Vol. 81, 1969, pp 8-12.
19. "Corrosion of Metals by Aqueous Solutions of the Atmospheric
Pollutant Sulfurous Acid, " McLeod, W., and Rogers, R. R. ,
Electro-Chemical Technology. Volume 6, No. 7-8, July/August 1968,
pp 231-235.
20. "Creeping Silver Sulfide, " Egan, T. F., and Mendizza, A. , Journal
of the Flertrochemical Society, Vol. 107, No. 4, April I960.
21. "Mechanisms by which Non-Ferrous Metals Corrode in the Atmosphere,'
Aziz, P. M. , and Godard, H. P., Corrosion. Vol. 15, October 1959.
22. The Tarnishing Behavior of Silver-Palladium Contact Surfaces, and
its Relationship to Accelerated Testing, Crossland, W. A., and
Knight, E. , Standard Telecommunication Laboratories, Harlow,
Essex, England.
23. "Combating the Effects of Smog on Wire-Spring Relays, " Hermance,
H. W. , Bell Laboratories Record, Volume 44, February 1966, pp 48-52.
24. "The Nature of Corrosion of Zinc by Sulfurous Acid at Ordinary
Temperatures," McLeod, W., Rogers, R. R. , Corrosion-MACE,
Volume 25, February 1969, pp 74-76.
25. Atmospheric Pollution. Juda, J. , and Budzinski, K. , U. S. Department
of Commerce, Office of Technical Services, JPRS 18455, OTS 63-21460,
31 March 1963.
26. "Sulfur Dioxide Porosity Tests for Coatings of Gold and the Platinum
Metals on Substrates of Copper and its Alloys, Nickel, Silver, "
Clarke, M. , and Leeds, J. M. , Transactions of the Institute of Metal
Finishing, Volume 46, 1968.
27. "Evaporated Metal Films as Indicators of Atmospheric Pollution"
(Part I), Lodge, J. P., and Havlik, B. R., International Journal of
Air and Water Pollution, Vol. 3, No. 4, pp 249-252.
28. "Evaporated Metal Films as Indicators of Atmospheric Pollution"
(Part II), Lodge, J. P., and Frank, E. R., International Journal of
Air and Water Pollution, Vol. 6, pp 215-221.
29. Systems Analysis of the Effects of Air Pollution on Materials, Salmon,
R. L. , MRI Report, Contract CPA-22-69-113, MRI Project No. 3323-D,
15 June 1969"- 14 January 1970.
30. "Atmospheric Effects Can Be Correlated, " Canadian Chemical
Processing, Vol. 51, No. 8, August 1967, p4~T£
31. Air Quality Criteria for Particulate Matter, U. S. Department of Health,
Education and Welfare, NAPCA Publication AP-49, January 1969.
32. Air Quality Criteria for Sulfur Oxides, U. S. Department of Health,
Education and Welfare, NAPCA Publication AP-50, January 1969.
33. Air Quality Criteria for'Photochemical Oxidants, U. S. Department of
Health, Education and Welfare, NAPCA Publication AP-63, March 1970.
34. Air Quality Criteria for Nitrogen Oxides. U. S. Department of Health,
Education and Welfare, NAPCA Publication AP-84, January 1971.
35. "The Role of Copper (II) in the Oxidation of Ferrous Hydroxide Colloid
with Special Reference to the Corrosion of Iron in an SO2 - Containing
Environment, " Inouye, Katsuya, Journal of Colloid and Interface Science,
Volume 27, No. 2, June 1968^ pp 171-179 (Extract).
36.
"Dirt and Calibration, " Odell, N. P. , Measurements and Data,
January/February 1969.
37. "Freeways vs. Contamination Control in California," Contamination
Control, April 1970.
38. "The Electronics Industry, " Dunlap, Lloyd \V. , Chemical and Engineer-
ing News, November 30, 1970.
39. "Environment Simulation for Studying the Effects of Air Pollutants on
Computers, " Steading, T. Walter, APCA Journal. Vol. =>, No. 3,
March 1965.
40. "Measurement and Simulation of the Computer Environment, " Steading,
T. W. , and Paulsen, J. N. , Proc. of the Institute of Environmental
Science, Annual Meeting, 1967.
8-2
8-3
-------
41. "The Effects of Air Pollution on Electrical Contact Materials: A Field
Study, " Chiarenzelli, Robert V. , and Joba, Edwin L. , Journal of the
Air Pollution Control Association, Vol. 16, No. 3, March 1966.
42. "A Survey of Insulation Contamination in the United States and Canada, "
Part I, Transactions of the IEEE. Paper No. 71 TP 133-PWR,
February 19, 1971.
43. "The Relation of Thin Films to Corrosion, " Rhodin, R. H. , Corrosion,
Vol. 12, No. 9, 465t-475t (1956) September.
44. "Switching Surge Performance of Contaminated Insulators, "
Turner, F. .J. , Macchiaroli, Transactions IEEE, Paper No. 71 TP 141-PWR,
December 1, 1970.
45. "A Study of the Performance of Contaminated Insulators Under Various
Test Conditions, " Kawai, M., Transactions IEEE. Paper No. 71
TP 132-PWR, December 15, 19751
46. "Lightning Impulse Flashover Characteristics of Long Disc Insulator
Strings Under Polluted Conditions, " Okada, T. , Koga, S., and Kimoto, I.,
Conference Paper IEEE, Paper No. 71 CP 144-PWH, December 1, 1970.
47. "The Effects of Pollution and Surface Discharges on the Impulse Strength
of Time Insulation, " Lushnicoff, N. L. and Parnell, T. M., Transactions
IEEE. Paper No. 71 TP 143-PWR. December 1, 1970.
48. "Effects of Dew and Fog on Insulator Breakdown Strength, " Berger, K.
and Chowdhuri, P., Conference Paper IEEE, Paper No. 71 CP 140-PWR,
December 7, 1970.
49. "A Possible Solution to the Insulator Contamination Problem, " Moran, J. H.
and Powell, D. G. , Conference--paper IEEE, Paper No. 71CP41-PWR,
January 31 - November 23, 1970.
50. "Analysis of Dielectric Recovery with Reference to Dry-Zone Arcs in
Polluted Insulators, " Rizk, F. A. M. , Conference Paper IEEE, Paper
No. 71 CP 134-PWR, December 14, 1970.
51. "A Criterion for A. C. Flashover of Polluted Insulators, " Rizk, F.A. M. ,
Conference Paper IEEE, Paper No. 71 CP 135-PWR, December 29, 1970.
52. "Contamination Generation of Internal Combustion Engines, " Fourth
Annual Technical Meeting and Exhibit, Hotel Fontainebleau, Miami
Beach, Florida, May 25-28, 1965.
53. "Investigation to Determine the Possible Need for a Regulation on
Organic Compound Emissions from Stationary Sources in the San
Francisco Bay Area, " Wohlers, H. C, and Feldstein, M. , JAPCA,
Vol. 15 (3), May 1965, pp 226-229.
54. "Detection and Estimation of Microgram Quantities of Carbon "
tgan, T. F. , Microchemical Journal. Vol. 13, No. 4, December
55. Automobile Traffic Control Systei
1968.
Rogers, L. M. , 1969.
8-4
-------
9. EXHIBITS
EXHIBIT 9. 1
Interviews with Manufacturers of Electronic Components
9-1
Company and Location
Admiral Corporation
Chicago, Illinois
Aerovox Corporation
Clean, .N. Y.
Allen Bradley
Milwaukee, Wisconsin
Amperex Electronic
Corp.
Slatersville, R. I.
Amperex Electronics, Inc.
Hickesville, N. Y.
Amphenol Connector Div.
Broadview, 111.
Amphenol RF Division
Danbury, Connecticut
Bliley Electric Corp.
Erie, Pennsylvania
Centralab Div. of
Globe-Union, Inc.
Milwaukee, Wisconsin
Person Interviewed
Vincent Kamler
Director-Engr.
Manager, Mech.
Engineering
Harold Zoble
R. Polakowski
Al Smith
Design Engineers
Gullio Masullo
Chemist
Michael J. Pawelko
Engrg. & Applications
Ned Spangler
Vice President,
Manufacturing
Eugene Makl
Vice President,
Manufacturing
George Wright
Mgr. Marketing
Gerald Stoughton
Mgr. -Corporate
Facilities
Product
Color TV Tubes
Ceramic Capacitors
. 050" sq. x .088
long to I. 5" square
Resistors and
Potentiometers
X-ray image
intensifiers
Plumicon tubes
Tx Tubes 5 W to 1 5 KW
Microwave tubes for
radar and cooking
ranges
X-ray tubes
Radiation counters
Connectors
RF Connectors 3/8"
dia. x 1/2" long
to 3" dia. x 6'*long
Coaxial switches
Microwave components
Crystals
Ceramic capacitors
Package circuits
Hybrid circuits
Potentiometers
Switches
Semiconductors,
electroluminescent
devices - Calif.
9-2
-------
Company and Location
Cinch Manufacturing
Elk Grove, 111.
Person Interviewed
R. Scott Modjeska
Concerned with
air and water
pollution control
Cornell Dubilier Electronics Robert Stone
New Bedford, Mass. Mgr. -Marketing
Cornell Dubilier Electronics Robert M. Butler
Fuquay-Verina, N. C. Mgr. -Marketing
Cornell Dubilier Electronics John Hand
Mgiv. -Engineering
(Electrolytic
Capacitors)
William Colman
Mgr. -Engineering
(Paper/film
capacitors)
C. P. Clare & Co.
Chicago, Illinois
CTS Knight. Inc.
Sandwich, Illinois
Dale Electronics, Inc.
Columbus, Nebraska
W. O. Thompson
Vice President Mfgr.
R. T. Bensen
Quality Control Engr.
Elsie Ita
Secretary to
Regional
Manager
Product
Connectors
Tube sockets
Terminal blocks
Oil-filled,
hermetically sealed
capacitors-1 /2 in. ^
to 2000 in. 3
Relays-Enclosed and
open
Antenna Rotors
Vibrators-DC & AC
Power Supplies
Electrolytic Capacitors
(Al & Ta foils)
Paper and Film
Capacitors
(non-metallic cases)
1/8 x 5/16 to 1" x 3"
Relays milliwatts to
100 amps.
P. C. boards
Lighted pushbuttons
Keyboards
Stepping switches
Timers
Crystals
Crystal controlled
oscillators
Filters
Components
Bare wire wound
resistors
Carbon film resistors
Metal film resistors
Bobbin wire wound -
coated wire
9-3
Company and Location
Electronics Center
(Thordarson &
Meissner)
Mt. Carmel, Illinois
Erie Technical Products,
Inc.
Erie, Pennsylvania
Fairchild Semiconductors
Mt. View Calif.
Fairchild Semiconductor
Products
Mountain View, Calif.
General Electric Semi-
conductors
Auburn, N. Y.
General Electric
Hudson Falls-, N. Y.
General Electric Co.
Tube Products Dept.
Owensboro, Ky.
Guardian Electric Mfg. Co.
Chicago, Illinois
International Crystal
Mfg. Co. , Inc.
ITT Cannon
Santa Ana, Calif.
Person Interviewed
Wayne Morris
Admin. Asst.
W. T. Crotty
Vice President,
Sales
Dr. J. Early
Director R & D
Dr. O. J. Trapt.
Mgr., Quality
Assurance
Finis Gentry
Mgr. -Engineering
Ed Izzo
Mgr. -Mfg. Engineering
Donald J. Meyers
Mgr. -Manufacturing
James F. Madole
Mgr. -Engrg.
Donald Schuette
Works Manager
Paul Freiland
Vice President,
Production
Ralph Alexander
Staff Chemist
Product
Transformers
(Ultra miniature
to 7 KVA)
Fixed &; variable
ceramic capacitors
Chip size up to 2"
Mica Capacitors
Carbon Resistors
Semiconductors
Semiconductors and
I.C. 's
Integrated Circuits
Semiconductors
Capacitors
(includes plastic
and metal foils)
Receiving Tubes
Microwave Tubes
Thyratrons
Industrial CRT's
LRT building
Relays
Solenoids
Switches
Crystals
Crystal Oscillators
Connectors
9-4
-------
Company and Location
ITT Headquarters
New York, N. Y.
• ITT Jennings
San Jose, Calif.
J. Wi Miller Co.
Campton, Calif.
Macklett Labs.
Stamford, Conn.
Macklett Labs.
(Div. of Raytheon]
Stamford, Conn. .
Mallory Capacitor Corp.
Indianapolis, Ind.
Motorola Semiconductor
Products, Inc.
Phoenix, Arizona
National Electronics
Geneva, 111.
'National Semiconductor
Santa Clara, Calif.
National Semiconductor
Towson Office, Md.
Ohmite Corp.
Skokie, Illinois
Person Interviewed
Ray Berry
Asst. Tech. Director
for Switching
Components
M. C. Budde
Mgr. -Mfg.
F. T. Scott
Or. Doolittle
Engr. Mgr. - RfeD
Carmen Ramie
Mgr. -Marketing
Harry Nieders
Mgr. -Marketing
Product
Electrical Components
Vacuum capacitors
and switches
RF Coils and
transformers
Microwave and
Power Tubes
Image tubes
Power tubes
Ta Electrolytic
Capacitors
Al Electrolytic
Capacitors
AC Oil Capacitors
Joe Flood Semiconductors and
Mgr. -Quality Assurance Integrated circuits
Robert Streaker
Vice President, Div.
Manager SEMCO Div.
Ed Thompson
Manager, Quality
Assurance
Tom Reynolds
Dist. Sales Mgr.
L. Berkelhamer
Vice President,
Engineering
Semiconductors
Integrated circuits
Semiconductors and
Integrated Circuits
Resistors
Switches
Variable transformers
Relays
*Field visit in addition to telephone interview
9-5 - '
Company and Location
Person Interviewed
Product
Philadelphia Division of
TRW (IRC)
Philadelphia, Pa.
Photocircuits, Inc.
Glen Cove, N. Y.
Potter & Brumfield
Princeton, Indiana
Raytheon - Industrial
and Cathode Ray Tube
Division
Raytheon - Microwave
Tube Division
RCA - Electrio-Optical
and Power Tube Div.
Lancaster, Pa.
RCA Electrio-Optical
and Power Tube Div.
Lancaster, Pa.
RCA (Image Tube Division)
Lancaster, Pa.
RCA Microwave Tube Div.
Harrison, N. J.
RCA Microwave Tube
Division
Harrison, N. J.
Sangamo Electric Co.
Capacitor Division
Fred Rocchi
Operations Manager
George Messner
Director - Research
Wayne Harper
Sales Supervisor
Warren Davis
Chief Engineer
Ed Downing
Manager .
Quality Assurance
Torn Lewis
Application
Engineer Photo-
multiplier Tubes
Dr. Morris Slater
Engineering Specialist
Frank Keith
Mgr. -Marketing
Carbon Resistors
1/4W to 1W
Printed circuit boards
DC motor with
printed armature
Relays
Cathode Ray Tubes
Microwave Tubes
Photomultiplier
Tubes
Electric Optical Div.
Power tubes
Image tubes
Photomultiplier tubes
John Kucera Microwave tubes
Mgr. -Quality Assurance
Joe Smack
Environ. Engineer
Emil Blase
Director-Engineer ing
Microwave tubes
Mica capacitors
Electrolytic
capacitors
Paper, Oil
capacitors
Power factor
capacitors
size range-picofarad
to farad
9-6
-------
Company and Location
Person Interviewed
Product
Sigma Instrument Inc.
Braintree, Mass.
* Signetics, Inc.
Sunnyvale, Calif.
Sprague Electric Co.
North Adams, Mass.
Sprague Electric Co.
North Adams, Mass.
Stancor Corp.
Chicago, 111.
Sylvania
Seneca Falls, N. Y.
Texas Instruments
Dallas, Texas
Triad/Utrad
Venice, Calif.
TRW-IRC
Philadelphia, Pa.
Tungsol
Livingston, N. J.
D. A. Belcher
Sales Service Mgr.
K. G. Lemmons
Manager, Qual.
Assurance
Dr. Dan Norton
Engrg. Specialist
William Allison
Technical Advisor
R. J. Reigel
Robert Peters
Manager-Marketing
Dr. Gordon Peattie
Mgr. -Quality
Assurance Dept.
Bob Larson
Market Manager
Hank Jajaroski
Quality Control Mgr.
Dale Fisher
Mgr. - Marketing
J. J. McGarth
Plant Engineer
United Transformer Corp. Burt Yudin
New York, N. Y. Engrg. Manager
*Field visit in addition to telepnone interview
Relays
Reed
General Purpose
Sensitive
Digital stepping motors
Outdoor lighting controls
Active Integrated
Circuits
Ceramic Capacitors
Passive integrated
circuits
Film capacitors
Paper & oil capacitors
Electrolytic capacitors
Transformers
Audio to 3 KVA
CRT's
Color TV Tubes
Semiconductors
and active
integrated circuits
Transformers
Carbon Resistors
1 /4 to 1 W
Wire Resistors
1 to 2 W
Hydrogen Thyratrons
Silicon Diodes
Electronic Power
Packs
Transformers 1/10 oz
to 520 Ibs. mv to
10 kv.
open with varnish dip,
hermetically sealed
canned with bakelite
terminals
9-7
Company and Location
Varian Associates
Palo Alto, Calif.
Varian Associates
Eimac Division •
San Carlos, Calif.
Varian Associates
Eimac Division
San Carlos, Calif.
Vector Electronic Co.
Sylmar, California
Western~Electric Co.
Allentown, Pa.
Western Electric Co.
Allentown, Pa.
Person Interviewed
H. Gardner
Staff Engineer
R. Albertson
Chief-Mfg. Engineer
Bill McAulay
Manager-Appls. Engrg.
C. D. Grey
Mgr. -Administration
Phillip Morrison
Jack Coughlin
Product
Microwave Tubes
Power Tubes
Instruments
Microwave tubes
Power tubes
Microwave tubes
Power tubes
Printed circuit
boards & hardware
Integrated Circuits
for Telephone
System
Integrated Circuits
for use in Telephone
System
-------
EXHIBIT 9.2
EXHIBIT 9. 3
Interviews with
Components
Company and Location
Beckman
Fullerton, California
IBM Federal System
Division
Raleigh, N. C.
McGraw Edison Company
Milwaukee, Wisconsin
* Royco Instruments
Menlo Park, Calif.
Varian Associates
Union, New Jersey
* Western Gear Co.
Pasadena, Calif.
Equipment Manufacturers Using Electronic
Person Interviewed
William Newton
Director-Mfrg.
Dr. Don Preiss
H. A. Van Dusen, Jr.
Senior Engineer
Dr. A. Lieberman
Bill.Sylvernal
Mechanical Engineer
C. W. Yost
Program Manager
Product
Analytical instruments
Precision components
Potentiometers
Microcircuits
Computers
Street Lighting
Equipment
Particle
Detecting
Instruments
Power tubes &
instruments
Precision
DC Motors
Interviews witl
Company and Location
ASTM (American Society
for Testing Materials)
Philadelphia, Pa.
AT&T
New York, N. Y.
AT&T Long Lines Division
Washington, D. C.
AT&T
New York, N. Y.
J*«U Telephone Laboratories
Holmdel, New Jersey
* Bell Telephone Laboratories
Holmdel, New Jersey
i Users of Electronic Equipment
Person Interviewed
Jim Dwyer
E. P. Hodges
Bldg. Engineer - NYC
Bob Scatchard
Dist. Bldg. Engineer
Dan Lyon
Mgr. Bldg. Engrg.
C. A. Russell
Supervisor: Electrical
Contact Contamin-
ation Studies
T. F. Egan
Product
Maintenance of
Telephone Co.
Building-NYC
Telephone Relay
Station
Maintenance of
AT&T Buildings
Materials Research
in Electrical
Contact
Contamination
Materials Research
on Electrical
Contacts
Field visit in addition to telephone interview
IBM Burlington Components
Division
Essex, Vermont
IBM Engr. Eval. Lab.
San Jose, Calif.
* M. P. Odell Co.
West Lake, Ohio
H. Frankel
T. W. Steading
M. P. Odell
President
Potomac Electric Power Co. P. Benziger
Asst. Vice President
Generation
Washington, D. C.
Potomac Electric Power Co.
Washington, D. C.
Sandia Laboratories
Albequerque, New Mexico
Stewart Bouchet
Transmission Engineer
Leo Klameris
Sandia Laboratories
Albequerque, New Mexico
9-9
J. D. Sivinski
Willis Whitfield
Contamination Control
Specialists
*Field visit in addition to telepnone interview
Computers and
component parts
Computers
Equipment cleaning
systems
Power generation
Materials and
contamination
research
0-10
-------
EXHIBIT 9. 4
Interviews with Government Agencies, Professional Associations,
University Groups and Others
Company and Location
Western Electric Co.
Cockeysville, Md.
Western Union
Washington, D. C.
** Merck and Co. , Inc.
Rahway, N. J.
Person Interviewed
Len Cochran
Envir. Engineer
Art Carroll
Area Maintenance
Supervisor
R. R. Johnson
Project Manager
** Visited but not formally interviewed by telephone
Product
Telephone
Equipment
Maintenance
Maintenance of
Telephone and
Telegraph
Equipment
Pharmaceutical
Chemicals
Organisation and Tdocation
AACC - American
Association for
Contamination Control
Bureau of Ships
Equipment Maintenance
Group
Washington, D. C.
District of Columbia
Highway Department
Traffic Signal Branch
Drexel Institute
Philadelphia, Pa.
Environmental
Engineering, Inc.
Gainesville, Florida
FAA
Falls Church, Va.
Person Interviewed
W. T. Maloney
Ralph Thresher
A. J. Breckinridge
Branch Manager
Dr. Henry Wohlers
Prof, of Environmental
Engrg. & Science
Dr. A. Hendrickson
John McGivern
Area Facilities Chief
* NASA-Goddard Research
Center
Greenbelt, Maryland
* NASA-Goddard Research
Center
Greenbelt, Maryland
Calvin Hirt
Reliability Engineer
Tom Sciacca
Matls. Res. Engineer
Product
Shipboard
Electronic
System
Maintenance of
Traffic Signals
Consulting
services
Maintenance of
Navigation Aid
Equipment for
Washington area FAA
FAA, System Maintenance
Services
* Federal Aeronautics
Administration
Washington, D. C.
* Federal Aeronautics Admin-
istration
Washington, D. C.
Walt Quitter
Director, System
Maintenance Services
Mr. Williams
Runway lighting
Maintenance
Engineer
Frank Opeka
Branch Manager
Maintenance Engineer
Maintenance of
FAA Equipment
Maintenance of
airport runway
lighting
Engineering of F.a
Maintenance of
Equipment
Tracking Station for
manned space flights
Materials research
for space craft
*Field visit in addition to telephone interview
9-11
9-12
-------
Organization and Location
NASA-Goddard Research
Center
Greenbelt, Maryland
National Bureau of Standards
Gaithersburg, Maryland
Naval Test Engineering
Center
Hyattsville, Maryland
Pima County, Arizona
Health Dept.
Tucson, Arizona
San Francisco Bay Area,
APCD
U. S. Air Force
Radar Equip. Maintenance
Group
Sacramento, Calif.
*U. S. Weather Bureau - ESSA
Silver Spring, Md.
U.S. Weather Bureau
Kansas City. Mo.
Person Interviewed
Dr. Frankel
Post Director-
Materials Research
Dr. Charles Marsden
Mr. Smollen
Test Engineer
Dr. Roy McColdin
Dir. -Environ.
Health
Dr. Milton Feldstein
Director
Donald Hinshaw
Russell Hovey
Engrg. Supervisor
Harold Anderson
Regional Engrg. Chief
Product
Materials research
for manned space .
craft
Navy Communications
System
Maintenance of AF
Radar Equipment
Maintenance of
Weather Bureau
Equipment
Maintenance of
Weather Bureau
Equipment
* Field visit in addition to telephone interview
9-13
BIBLIOGRAPHIC DATA
SHEET
II. Report No.
AFTD-0797
[me and Subtitle
A 'Survey and Economic Assessment of the Effect of
Air Pollutants on Electrical Components
Volume I - Sections 1 through 9
7. Authot(s)
9. Performing Organization Name and Address
ITT Electro-Physics Laboratories Inc.
9140 Old Annapolis Road
Columbia, Maryland 21043
12. Sponsoring Organization Name and Address
Division of Economic Effects Research
Air Pollutuion Control Office
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. Recipient's Accession No.
5. Report Date
August 1971
8. Performing Organization Rept.
No.
10. Project/Task/Work Unit No.
11. Contract/Grant No.
CPA 70-72
13. Type of Report & Period
Covered
Final
15. Supplementary Notes DISCLAIMER - This report was furnished to the Uttlce ot Air
Programs by ITT Electro-Physics Laboratories Inc., 9140 Old Annapolis
Road, Columbia, Maryland 21043 in fulfillment of Contract CPA 70-72
16. Abstracts
An assessment was made of the economic impact of air pollution on electronic components
lo begin the study, electroniccomponents were divided into eleven different categories
as defined by the Department of Commerce. The manner and amount of damage from air
pollutants that might have been expected in each of these component categories was
assessed by surveying the literature describing pollutant material damage mechanisms.
The expected effects were compared with the actual experience of major manufacturing,
companies which appeared to account for most of the sales in each category. Where the
literature survey Indicated that gaseous sulfur compounds, notably sulfur dioxide.
should be expected to account for most'of the, damage to electronic components, Inter-
views with manufacturers revealed that particulate matter, actually was responsible
for most of the electronic component ana equipjaetiK malfunctions currently being ex-
perienced. The economic effects, which were represented by three cost categories,
were applied to each component category and then summed to arrive at a total cost of
about 15.5 million.
17. Key Tords and Document Analysis. 17o. Descriptors
Air pollution
Economic analysis
Corrosion
Deformation
Electronic devices
17b. Identifiers Open-Eoded Terms
Air pollution effects (materials)
Particulates
17c. COSATI Field/Group 13B, 9A
Electric equipment
Electric devices
Particles
Sulfur compounds
Sulfur dioxide
18. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Securiry Class (This
Page
UNCLASSIFIED
21. No. of Pages
89
Tla-39 t1O-7OI
UICOMM-OC 4011»-
"' (L
------- |